Investigation of protein induction in vascular-targeted strategies.

COLE, Laura Margaret.

Available from Sheffield Hallam University Research Archive (SHURA) at:

http://shura.shu.ac.uk/19491/

This document is the author deposited version. You are advised to consult the publisher's version if you wish to cite from it.

Published version

COLE, Laura Margaret. (2013). Investigation of protein induction in vascular-targeted strategies. Doctoral, Sheffield Hallam University (United Kingdom)..

Copyright and re-use policy

See http://shura.shu.ac.uk/information.html

Sheffield Hallam University Research Archivehttp://shura.shu.ac.uk

i-cammy ana ipionTiation ServicesAdsetts Centre, City Campus

Sheffield S1 1VVD

1 0 2 0 1 9 7 9 8 6

Sheffield Hailam UniversityLearning and Information Services

Aaseas Centre City Campus Sheffield S1 1WB

REFERENCE

ProQuest Number: 10694372

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a com ple te manuscript and there are missing pages, these will be noted. Also, if material had to be removed,

a note will indicate the deletion.

uestProQuest 10694372

Published by ProQuest LLC(2017). Copyright of the Dissertation is held by the Author.

All rights reserved.This work is protected against unauthorized copying under Title 17, United States C ode

Microform Edition © ProQuest LLC.

ProQuest LLC.789 East Eisenhower Parkway

P.O. Box 1346 Ann Arbor, Ml 48106- 1346

Investigation of Protein Induction in Vascular-targetedStrategies

Laura Margaret Cole

A thesis submitted in partial fulfilment of the requirements of Sheffield Hailam University

for the degree of Doctor of Philosophy

In collaboration with the University of Sheffield

February 2013

ACKNOWLEDGEMENTS

First and foremost, I would like to express my gratitude and appreciation of my principal supervisor Professor Malcolm R Clench, not only fo r his support and encouragement throughout my studies but for giving me many opportunities to extend my knowledge of

analytical Mass Spectrometry through conferences and instilling faith in me during the

presentation of my work. I am now a 'convert to Mass Spectrometry'!

I would also like to thank Dr Vikki A Carolan for her expertise and advice received during the

writing of my thesis and comments on the structure of the many abstracts and posters I have

emailed her over the last four years.

Special thanks and recognition is owed to the collaboration with the University of Sheffield. The

guidance by Professor Gillian M Tozer has been invaluable. I would also like to express my

gratitude to Professor Martyn N Paley and Professor Nicola J Brown for their input and support

during the many Imaging update meetings.

I would like to say thank you to Dr Chris Sutton at the Institute of Cancer Therapeutics in

Bradford for his patience and training during my time there.

For help with training and assistance on a variety of new techniques I am truly grateful to Dr Marie Claude Djidja, Emmanuelle Claude, Dr Adam Dowle, Dr Samira Kazan, Dr Florian Wulfert,

Kevin Blake, Rachel Daniels, Jenny Globe and M att Fisher.

I gratefully acknowledge Dr Simona Francese and Dr David Smith who not only have been

helpful minds of information during my PhD but great conference travel cohorts too.

I would like to declare a special thank you to colleagues past and present for their friendship

and support during my PhD who include Dr Paul J Trim, Leesa Ferguson, Dr Kate Phillips, Dr Louise Vickers, Eva llles-Toth, Dr Philippa J Hart, Bryn Flinders, Robert Bradshaw, Chris Mitchell

and Afnan Batubara, some of which are within the Mass Spectrometry group here at the BMRC.

A huge acknowledgement is extended to the 'coffee club ladies'; Dr Claire M Bradford, Dr Helenne Nyamwaro and Dr Rachel E Doherty who not only introduced me to cherry scones but also shared much laughter and discussion!

A very special thank you to my family and parents Tony and Laura who as always, support me

in everything that I have ever set out to achieve.

Last, but certainly not the least, much love and endless thanks to my husband Wayne and

daughter Alicia for support and encouragement words just cannot describe.

Abstract

The aims of the study reported in this thesis were to develop and utilise mass spectrometry

imaging techniques (MALDI-MSI), in combination with conventional proteomic methodologies,

to investigate protein induction in vascular-targeted strategies.

Proteins thought to be involved in tumourigenesis and drug treatment resistance were

observed along with the responses from proteins identified via the techniques used, in this

global analysis study. MALDI-MSI, LC-ESI-MS/MS, LC-MALDI-MS/MS with iTRAQ labelling and

immunohistochemistry intended to provide cross validation of the effects post administration

of vascular disrupting agent CA-4-P. Two mouse fibrosarcoma models (expressing VEGF120/ VEGF188 isoforms only) following treatment with the tubulin-binding tumour vascular disrupting agent, combretastatin A-4-phosphate (CA-4-P) have been studied.

The gross haemorrhagic pharmacological response elicited by CA-4-P was visible by MALDI-MSI throughout the fibrosarcoma 120 time course. The latter encouraged the prospect that other proteins could potentially be observed induced via a dose response relationship. The

haemoglobin time course using the resistant 188 tumour model gave quite different results to

those previously seen in the MALDI-MSI of the fibrosarcoma 120 data set. The first indication

of the 'switch back to tissue viability' concept was revealed.

The experimental work using LC-ESI-MS/MS revealed many proteins connected with necrosis, apoptosis, cell structural reorganisation, polymerisation, tumour survival and stress induced

molecular chaperones. The inverse correlation of structural proteins, haemoglobin and heat shock molecular chaperones gave the required validation and identification to relate these

responses to those seen in MALDI-MSI. The relationship pathways generated by using STRING9.0 proteomic network software gave an invaluable insight into the activity of the active

tumour milieu and provided a means of linking the identified proteins to their functional partners. Protein-protein interactions could be observed to help interpretation of the MALDI-

MSI, LC-ESI-MS/MS and iTRAQ LC-ESI-MS/MS response graphs.

Overall, the dose relationships observed in the iTRAQ data by the proteins involved in

haemorrhaging, structural remodelling, were in good agreement with the other techniques

employed here.

It could be said that MALDI-MSI could potentially forge a place in the workflow of clinical diagnostics. Targeted approaches for the observation of disease biomarkers could be visualised

using MALDI-MSI and serve as a complimentary technique to standard clinical imaging. A

novel method reported here using a multi-peptide recombinant standard could prove an

important diagnostic tool for the analysis of patient biopsies and tissue micro-arrays. The

exciting prospect is the diversity of a multi-peptide recombinant standard, an artificial construct that can be engineered to include any prospective biomarkers for both research and

diagnostic screening applications.

Table of ContentsTable of Contents.......................................................................................................................................1

List of Figures.............................................................................................................................................. 6

Chapter 1 ................................................................................................................................................ 6

Chapter 2 ................................................................................................................. 7

Chapter 3 ................................................................................................................................................ 8

Chapter 4 .............................................................................................................................................. 10

Chapter 5 .............................................................................................................................................. 11

Abbreviations........................................................................................................................................... 13

Chapter 1 ...................................................................................................................................................17

General Introduction...............................................................................................................................17

1.1 Cancer............................................................................................................................................. 18

1.1.1. Statistical information, incidence/ prevalence............................................................... 18

1.1.2 Tumour types........................................................................................................................ 22

1.1.3 Overview of current anti-cancer therapies.......................................................................23

1.1.4 Vascular targeted strategies................................................................................................ 30

1.1.5 Vascular disrupting agents.................................................................................................. 33

1.1.6 Biomarker Discovery............................................................................................................ 39

1.1.7 Considering Vascular disrupting agents for anti-cancer therapeutics..........................41

1.2 The Role of Mass Spectrometry in Biomarker Discovery....................................................... 41

1.2.1. Overview of mass spectrometry as an analytical tool.................................................... 41

1.2.2. Basic principles of mass spectrometry............................................................................. 42

1.2.3. Ionisation techniques and Mass analysers.......................................................................46

1.3 Mass spectrometry instrumentation........................................................................................ 49

1.3.1 LC-MS.................................................................................................................................... 49

1.3.2. Electrospray ionisation (ESI)............................................................................................... 51

1.3.3. Matrix assisited laser desorption ionisation (MALDI)................................ 52

1.3.4. Quadrupole Time of flight mass spectrometry (QqToF)................................................54

1.3.5. Synapt HDMS - travelling wave ion mobility separation............................................... 56

1.3.6. MALDI-mass spectrometry imaging.................................................................................. 59

1.3.7. Quantitative Proteomics............................ 60

1.3.8. Aims of the study.................................................................................................................. 63

References............................................................................................... 66

2.1 Introduction.................................................................................................................................75

2.2 Materials and Samples................................................................................................................ 76

2.2.1 Chemicals and Materials..................................................................................................... 76

2.2.2 Tissue samples.......................................................................................................................77

2.2.3. Experimental groups...........................................................................................................77

2.2.4 Tissue preparation................................................................................................................ 77

2.2.5 In situ tissue digestion and trypsin deposition................................................................. 77

2.3Methods and instrumentation.................................................................................................... 78

2.3.1 Matrix deposition................................................................................................................. 78

2.3.2 Instrumentation....................................................................................................................78

2.3.3 Haematoxylin and eosin staining....................................................................................... 79

2.3.4 Data pre-processing using SpecAlign and Marker View software................................ 79

2.3.5 Statistical analysis..................................................................................................................79

2.3.6 Data pre-processing usingWaters MassLynx™ Software and MATLAB®...................... 79

2.4 Results and Discussion.............................................................................................................. 80

2.4.1 Initial MALDI profiling and imaging of in situ tissue tryptic digests in mouse

fibrosarcoma 120 models.............................................................................................................. 80

2.4.2 Haematoxylin and Eosin staining of fibrosarcoma 120 tissue........................................83

2.4.3 Initial imaging of proteins according to theoretical digests using MALDI-MSI ......86

2.4.4 Principle component analysis-discriminant analysis (PCA-DA).................................. 88

2.4.5. Employment of Ion Mobility MALDI-Mass Spectrometry to study//? situ tryptic

digestion in fibrosarcoma 120 models........................................................................................ 91

2.4.6 Rho GTPases: a target to help understand resistance to CA-4-P?................................ 99

2.4.7.Statistical analysis of mouse fibrosarcoma 120 tumour tissue using MATLAB® 107

2.4.8 Region of interest (ROI) study for the analysis of tumour rim and necrotic regions

within fibrosarcoma 120 tissue................................................................................................... 112

2.4.9 Concluding Remarks........................................................................................................... 118

References.......................................................................................................................................... 120

3.1 Introduction................................................................................................................................. 124

3.2 Materials and Samples...............................................................................................................126

3.2.1 Chemicals and Materials....................................................................................................126

3.2.2 Tissue samples..................................................................................................................... 126

3.2.3. Experimental groups......................................................................................................... 126

3.2.4 Tissue preparation...............................................................................................................126

3.2.5 In situ tissue digestion and trypsin deposition............................................................... 127

3.3 Methods and instrumentation..................................................................................................127

2

3.3.1 Matrix deposition............................................................................................................... 127

3.3.2 Instrumentation..................................................................................................................127

3.3.3 Haematoxylin and eosin staining.....................................................................................128

3.3.4 Statistical analysis............................................................................................................... 129

3.3.5 Data pre-processing using Waters MassLynx™ Software and MATLAB®...................129

3.3.6Protein network analysis.................................................................................................... 129

3.3.7 Immunohistochemical staining........................................................................................ 129

3.3.8 Protein precipitation and digestion................................................................................ 130

3.3.9 Protein estimation - BCA assay....................................................................................... 131

3.4 Results and Discussion...........................................................................................................132

3.4.1 MALDI-MSI of fibrosarcoma 188 CA-4-P treated tumour tissue................................ 132

3.4.2 Statistical analysis of mouse fibrosarcoma 188 tumour tissue using MATLAB® 150

3.4.3 Label free LC-ESI -MS/MS for the identification of proteins in a CA-4-P treated

fibrosarcoma 188 mouse model............................................................................................... 159

3.4.4 Protein relationship mapping of fibrosarcoma 188 LC-ESI-MS/MS data using String9.0................................................................................................................................................... 182

3.4.5 Concluding remarks........................................................................................................... 186

References......................................................................................................................................... 188

4.1 Introduction................................................................................................................................ 194

4.2 Materials and Samples..............................................................................................................195

4.2.1 Chemicals and Materials...................................................................................................195

4.2.2 Tissue samples....................................................................................................................195

4.2.3. Experimental groups..................................................................................................... 196

4.2.4 Tissue preparation............................................................................................................. 196

4.2.5 Tissue homogenisation and precipitation of protein................................................... 196

4.2.6 Protein Digestion................................................................................................................197

4.2.7 Preparation of samples for column loading.................................................................. 197

4.2.8 C18 bond-elute preparation.............................................................................................197

4.2.9 Sample re-suspension.......................................................................................................198

4.2.10 iTRAQ labelling.................................................................................................................198

4.2.11 SCX..................................................................................................................................... 199

4.3Methods and instrumentation................................................................................................. 199

4.3.1 Sample fractionation and matrix deposition................................................................. 199

4.3.2 Instrumentation................................................................................................................ 199

4.3.3 Data pre-processing..........................................................................................................200

4.3.4 Statistical analysis.............................................................................................................. 201

3

4.3.5 Protein network analysis................................................................................................... 201

4.4 Results and discussion - Fibrosarcoma 120............................................................................ 202

4.4.1. HPLC UV profiles of fibrosarcoma 120 tumour fractions............................................ 202

4.4.2 Frequency of unique peptides identified........................................................................203

4.4.3. Analysis using excel...........................................................................................................203

4.4.4. Analysis with Scaffold Q + .................................................................................................207

4.4.5 Protein relationship mapping of fibrosarcoma 120 data using STRING 9.0.............. 218

4.5 Results and discussion - Fibrosarcoma 188............................................................................ 224

4.5.1. HPLC UV profiles of fibrosarcoma 188 tumour fractions............................................ 224

4.5.2 Frequency of unique peptides identified........................................................................225

4.5.3. Analysis using excel........................................................................................................... 225

4.5.4. Analysis with Scaffold Q + ................................................................................................. 231

4.5.5 Protein relationship mapping of Fibrosarcoma 188 data using STRING 9 .0 ............. 242

4.5.6 Protein dose response relationship analysis using String 9 .0 ......................................247

4.5.7 Relation of MALDI-MSI to iTRAQ proteomic response.................................................256

4.5.8 Concluding Remarks........................................................................................................... 259

References..........................................................................................................................................260

5.1 Introduction................................................................................................................................. 266

5.2 Materials and Methods............................................................................................................. 267

5.2.1 Peptide sequences chosen for a recombinant protein standard based on MALDI-MSI and conventional proteomic analysis of fibrosarcoma 120 and 188 tissue.........................267

5.2.2 Bacterial culture..................................................................................................................268

5.2.3 Plasmid preparation........................................................................................................... 269

5.2.4 E. coli- transformation........................................................................................................269

5.2.5 Protein Purification............................................................................................................ 269

5.4.6 Digestion...............................................................................................................................270

5.4.7 MALDI-MSI - Chemicals and Materials............................................................................ 271

5.4.8 Methods and instrumentation..........................................................................................272

5.3 Results and Discussion.............................................................................................................. 273

5.3.1 MALDI-IMS-MS Peptide mass fingerprint of the digested recombinant DNA construct ............................................................................................................................. 273

5.3.2 MALDI-IMS-MSI of fibrosarcoma 188 tumour tissue, Control and post CA-4-P

treatm ent.......................................................................................................................................275

5.3.3 Recombinant MALDI-IMS-MSI visualised through HDI 1.1 software..........................280

5.3.4 Cross validation of Plectin and HSP-90 response in fibrosarcoma 188 tissue

employing a multimodal proteomic strategy............................................................................292

4

5.4 Concluding Remarks................................................................................................................295

References..........................................................................................................................................297

Appendices.............................................................................................................................................304

Appendix 1: Publications................................................................................................................. 305

Appendix 2: Oral presentations..................................................................................................... 306

Appendix 3: Poster presentations..................................................................................................307

5

List of FiguresChapter 1Figure 1.1: Percentage of all deaths due to cancer in the different regions of the world (WHO)

Regions of the World............................................................................................................................... 19Figure 1. 2: Cancer incidence worldwide map.....................................................................................20Figure 1.3 : The 20 Most Common Causes of Cancer Death, UK..................................................... 21

Figure 1.4 : Diagrammatical representation of the possible routes of cell division..................... 22Figure 1. 5 : Proposal of adding CSC specific predictive tests to therapy regimes for potentialincrease in tumour control probability.................................................................................................26Figure 1. 6 : Images to illustrate the substandard tumour vasculature.......................................... 33Figure 1. 7 : Chemical structures of vascular disrupting agents....................................................... 34Figure 1. 8 : Proposed mechanism of action of Combretastatin A-4-3-0-phosphate...................35Figure 1. 9 : Fluorescent staining of Human umbilical vein endothelial cells, Control and postCA4P administration................................................................................................................................ 36Figure 1 .10 : CA-4-P and its effect on hypoxia induction and vascular shutdown........................37Figure 1 .1 1 : Exploitation of tumour vasculature by antiangiogenic treatment and vasculardisrupting agents......................................................................................................................................37Figure 1.12 : Therapeutic targeting of the hallmarks of cancer...................................................... 40Figure 1 .13:The cells of the tumour microenvironment...................................................................40Figure 1 .14 : Basic principles of a mass spectrometer...................................................................... 43Figure 1.15 : Basic ionisation techniques............................................................................................ 46Figure 1.16 : Mass analysers used in mass spectrometers...............................................................47Figure 1.17 : Protein identification methods......................................................................................48Figure 1.18 : Schematic representation of a liquid chromatography mass spectrometer.........50Figure 1.19 : Schematic of Electrospray ionisation............................................................................ 51Figure 1. 20 : Ionisation using MALDI....................................................................................................52Figure 1. 21: Diagrammatical representation of a tandem QqTOF mass spectrometer........... 55Figure 1. 22 : Schematic representation of a ToF modulator............................................................56Figure 1. 23 : SYNAPT™ G2 HDMS system............................................................................................57Figure 1. 24 : Stacked ring ion guide.....................................................................................................57Figure 1. 25 : Travelling wave ion guide- schematic representation................................................58Figure 1. 26 : MALDI-MSI workflow diagram...................................................................................... 60Figure 1. 27 :The concept of area peptide measurement coupled with a specific time

domain,visualisedthrough Progenesis LC- MS software....................................................................61

Figure 1. 28 : Simple workflow using Isobaric Tags for Relative and Absolute Quantitation(iTRAQ).......................................................................................................................................................62Figure 1. 29 : VEGF isoform expression and exon structure..............................................................64

6

Chapter 2Figure 2.1: Representative peptide mass fingerprint (PMF) arising from an on-tissue digest of aVEGF120 tumour 24 h after dosing with 100 mg/kg i.p, CA-4-P...................................................... 80Figure 2. 2: Peptide mass fingerprints obtained from the fibrosarcom 120 tumour tissue

sections treated with saline or 100 mg/kg i.p..................................................................................... 81

Figure 2. 3: MALDI-MS images for the distribution of m/z 1274throughout a fibrosarcoma 120timecourse................................................................................................................................................ 82Figure 2 .4 : Haematoxylin and eosin stained sections...................................................................... 84Figure 2. 5: Higher magnification images of the H&E staining of the tumour rim in the VEGF120tumour sections....................................................................................................................................... 85Figure 2. 6: MALDI-MSI of peptides from tumour 5:1 (24hrs CA-4-P).............................................86Figure 2. 7: PCA-DA of fibrosarcoma 120 tumour in situ tryptic digests.........................................88Figure 2. 8: Representative peptide mass fingerprint (PMF) arising from an on-tissue digest of aVEGF120 tumour 6 h after dosing with 100 mg/kg i.p, CA-4-P.........................................................92Figure 2. 9: IMS-MS/MS data obtained from the peak observed in Figure 2.8 at m/z 1819....... 93Figure 2 .10 a: A Drift scope plot of a tumour section 24 h post CA-4-P showing the ion mobilityseparation as a function of drift time................................................................................................... 94

Figure 2.11: The distribution of two known peptides in tumour 5_6, i.e. a tumour 24 h aftertreatment with CA-4-P........................................................................................................................... 97Figure 2.12: Overlaid MALDI MSI images showing differing spatial distribution and co­registration of peptides...........................................................................................................................98Figure 2.13: Major Functions of Rho GTPases in vitro.................................................................... 100Figure 2.14: Peptide mass fingerprints obtained from the fibrosarcom 120 tumour tissue30min post CA-4-P(100 mg/kg i.p.) with peak of interst at m/z 844...................... 101Figure 2.15: IMS-MS/MS data obtained from the peak observed in Figure 2.14 at m/z 844.5. 102Figure 2.16: Mascot score histogram detailing the results in the identification of N-chimaerin. 102

Figure 2.17: Zoomed in Driftscope plot showing ion of interest, arrow indicating the peakoriginally selected.................................................................................................................................. 103

Figure 2.18: PMF's throughout a fibrosarcoma 120 time course with putative peptide of interest N-chimaerin visible in early time points .Control (saline), 0.5 h post CA-4-P, 6 h postCA-4-P and 24 h postCA-4-P................................................................................................................. 104Figure 2.19: MALDI-MSI throughout a fibrosarcoma 120 time course of the putative peptideRho GTPase activating protein 2 m/z 844 (183-189, RLTSLVR).......................................................105Figure 2. 20 : PCA of fibrosarcoma 120 tumour tissue in situ tryptic digests............................... 108Figure 2. 21: Partial least squares discriminant analysis (PLSDA) regression vector plot comparing MALDI "on-tissue" digest data from samples of fibrosarcoma 120 Control/Saline

and 30 min post combretastatin-4-phosphate (CA-4-P) (a vascular disrupting agent), treatmenttumour samples......................................................................................................................................110Figure 2. 22: HDI software displaying MALDI-MSI from a fibrosarcoma tumour section 24h postCA-4-P.......................................................................................................................................................113Figure 2. 23: PLSDA regression vector plot showing fibrosarcoma 120 Control necrotic regions

versus fibrosarcoma 120 necrotic regions post 24h CA-4-P............................................................114

7

Figure 2. 24: PLSDA regression vector plot showing fibrosarcoma 120 Control rim regionsversus spectra from rim regions post 24h CA-4-P............................................................................ 115Figure 2.25: VIP scores for the Control fibrosarcoma 120 tumour rim regions...........................116Figure 2.26: VIP scores for the 24h post CA-4-P fibrosarcoma 120 tumour rim regions........... 117

Chapter 3Figure 3.1: Diagrammatical cross section of the region present at the end of growing longbones showing the areas involved in epiphysis and metaphysis................................................... 124Figure 3. 2: Screen shot of protein delocalisation in MALDI-MSI, visualised using BioMapimaging software................................................................................................................................... 133

Figure 3. 3: Screen shot of an on tissue tryptic digest using MALDI-MSI, visualised using BioMap

imaging software................................................................................................................................... 135

Figure 3.4: PMF's of a fibrosarcoma 188 time course study. Spectra shown are Control, 0.5h,6h, 24h and 72h post CA-4-P treatment.............................................................................................137Figure 3. 5: Zoomed in region of fibrosarcoma 188 PMF's. The blue arrows indicate the gross

differences visible in the 24h and 72h time points, m/z 1416 being Hb and an unknown speciesat m/z 1577............................................................................................................................................. 138Figure 3.6: Comparison control tissue peptide mass fingerprints from fibrosarcoma 120 and188 on tissue digests............................................................................................................................. 139Figure 3. 7: MALDI-MSI multi image of a fibrosarcoma 188 time course..................................... 141Figure 3. 8: Fibrosarcoma 188 MALDI-MSI multi image showing spatial distribution of HSP-90.MALDI images showing spatial distribution of HSP-90 alpha at m/z 1168................................... 143Figure 3. 9: Immunohistochemical staining of a fibrosarcoma 188 time course.........................144

Figure 3.10: Immunohistochemical staining of HSP-90 in Fibrosarcoma 188 tissue..................145Figure 3.11: SYNAPT G2 MALDI imaging plate indicating tissue image regions..........................147Figure 3 .1 2 :40pm MALDI-MSI of 24h post CA-4-P fibrosarcoma 188 tumour tissue 148Figure 3 .1 3 :40pm MALDI-MSI of m/z 944.5 in fibrosarcoma 188 tumour tissue..................... 148Figure 3 .1 4 :40pm MALDI-MSI of a on tissue trytic digest and corresponding H and Ehistological section................................................................................................................................ 149Figure 3.15: 30 pm MALDI-MSI of a Control fibrosarcoma 188 tumour on tissue tryptic digest.................................................................................................................................................................. 149

Figure 3.16: PCA score plot of fibrosarcoma 188 tumour tissue in situ tryptic digests.............150Figure 3.17: PCA loadings plot fibrosarcoma 188 tumour tissue in situ tryptic digests............ 151Figure 3.18: Partial least squares discriminant analysis (PLSDA) regression vector plot comparing MALDI "on-tissue" digest data from samples of fibrosarcoma 188 Control and 0.5hpost combretastatin-4-phosphate (CA-4-P), treatment tumour samples....................................152Figure 3.19: Partial least squares discriminant analysis (PLSDA) regression vector plot for thecomparison of Control and 6h post CA-4-P fibrosarcoma 188 on tissue digests.........................153Figure 3. 20: Partial least squares discriminant analysis (PLSDA) regression vector plot for thecomparison of Control and 24h post CA-4-P fibrosarcoma 188 on tissue digests...................... 154Figure 3. 21: Partial least squares discriminant analysis (PLSDA) regression vector plot for thecomparison of Control and 72h post CA-4-P fibrosarcoma 188 on tissue digests.......................155Figure 3. 22: VIP scores for the 0.5h post CA-4-P fibrosarcoma 188 on tissue digest................ 156

8

Figure 3. 23: VIP scores for the 6h post CA-4-P fibrosarcoma 188 on tissue digest....................156Figure 3. 24: VIP scores for the 24h post CA-4-P fibrosarcoma 188 on tissue digest................. 157Figure 3. 25: VIP scores for the 72h post CA-4-P fibrosarcoma 188 on tissue digest................. 157Figure 3. 26: Scaffold proteomic tool software used for LC-ESI-MS/MS analysis....................... 160Figure 3. 27: Graph A - Fibrosarcoma 188 treatment time course results, exported fromScaffold 3.................................................................................................................................................161Figure 3. 28: Graph B - Fibrosarcoma 188 treatment time course results, exported fromScaffold 3.................................................................................................................................................162Figure 3. 29: Graph C - Fibrosarcoma 188 treatment time course results, exported fromScaffold 3....................................................... 163Figure 3. 30: Graph D - Fibrosarcoma 188 treatment time course results, exported fromScaffold 3.................................................................................................................................................164

Figure 3. 31: Graph E - Fibrosarcoma 188 treatment time course results, exported fromScaffold 3.................................................................................................................................................165Figure 3. 32: Graph F - Fibrosarcoma 188 treatment time course results, exported fromScaffold 3.................................................................................................................................................166Figure 3. 33: Graph G - Fibrosarcoma 188 treatment time course results, exported fromScaffold 3.................................................................................................................................................167

Figure 3. 34: Fibrosarcoma 188 time course results post CA-4-P treatment using LC-ESI-MS/MS. 168Figure 3. 35: MS/MS spectrum and normalised intensity graph of Alpha-2-macroglobulin infibrosarcoma 188 LC-ESI-MS/MS results............................................................................................. 170

Figure 3. 36: Immunohistochemical staining of macrophage cell surface marker F4/80 usingfibrosarcoma 188 tissue lOx................................................................................................................ 172Figure 3. 37: MS/MS spectrum and normalised intensity graph of Plectin in fibrosarcoma 188

LC-ESI-MS/MS results. Insert displays the normalised spectral counts for Plectin throughout thetime course post CA-4-P........................................................................................................................174Figure 3. 38: Fibrosarcoma 188 MALDI-MSI multi image showing spatial distribution of Plectin.MALDI images showing spatial distribution of Plectin at m/z 977.................................................174Figure 3. 39: Immunohistochemical staining of Plectin in Fibrosarcoma 188 tissue post CA-4-Ptreatment................................................................................................................................................ 175Figure 3.40: Optical scans of anti-Plectin and HSP-90 Immunohistochemical staining infibrosarcoma 188 serial sections, 72h post CA-4-P treatm ent...................................................... 177Figure 3.41: Example MS/MS spectrum and normalised intensity graph of IQGAP1 infibrosarcoma 188 LC-ESI-MS/MS results............................................................................................ 178Figure 3.42: Example MS/MS spectrum and normalised intensity graph of Tenascin C infibrosarcoma 188 LC-ESI-MS/MS results............................................................................................ 179Figure 3.43: Example MS/MS spectrum and normalised intensity graph of Alpha-enolase infibrosarcoma 188 LC-ESI-MS/MS results............................................................................................ 180Figure 3.44: Example MS/MS spectrum and normalised intensity graph of Transforming

growth factor-beta-induced protein ig-h3 (Tgfbi) in fibrosarcoma 188 LC-ESI-MS/MS results. 181

Figure 3.45: LC-ESI-MS/MS proteins of interest from fibrosarcoma 188, visualised throughproteomic pathway software STRING 9.0.......................................................................................... 182

Figure 3.46: Proteomic network from LC-ESI-MS/MS fibrosarcoma 188 data............................ 184

9

Chapter 4Figure 4 .1: The principles of multiplex tagging chemistry............................................................ 194Figure 4.2: UV profiles from the fibrosarcoma 120 fractions after SCX...................................... 202Figure 4. 3: Frequency of detection of unique peptides by LC-MALDI-MS/MS......................... 203Figure 4 .4 : iTRAQ results from fibrosarcoma 120 fixed filtered search......................................204Figure 4. 5: iTRAQ ratio response relative to the control (113), normalised to an average of 1.

................................................................................ 205

Figure 4. 6: Scaffold proteomic software used for iTRAQ analysis................................................ 208Figure 4 .7: Scaffold proteomic software with the employment of 'Q+' function for iTRAQ ratioanalysis....................................................................................................................................................209Figure 4. 8: iTRAQ fold change ratio using fibrosarcoma 120 tumour digest samples.............. 210Figure 4. 9: Vimentin peptides identified, response bar graph and related MS/MS spectrumfrom iTRAQ LC-MALDI-MS/MS Fibrosarcoma 120 tissue................................................................ 212Figure 4.10: Haemoglobin subunit beta-1 peptides identified, response bar graph and relatedMS/MS spectrum from iTRAQ LC-MALDI-MS/MS Fibrosarcoma 120 tissue................................ 213Figure 4.11: Peroxiredoxin-1 (Macrophage 23kDa stress protein) peptides identified, response

bar graph and related MS/MS spectrum from iTRAQ LC-MALDI-MS/MS Fibrosarcoma 120tissue...................................................................................................... 214Figure 4.12: Tubulin beta-5 chain peptides identified, response bar graph and related MS/MSspectrum from iTRAQ LC-MALDI-MS/MS Fibrosarcoma 120 tissue.............................................. 215Figure 4 .13: Annexin A5 peptides identified, response bar graph and related MS/MS spectrumfrom iTRAQ LC-MALDI-MS/MS Fibrosarcoma 120 tissue................................................................ 216Figure 4.14: Keratin -1 (type II skeletal 1) peptides identified, response bar graph and relatedMS/MS spectrum from iTRAQ LC-MALDI-MS/MS Fibrosarcoma 120 tissue................................ 217Figure 4.15: Fixed modification search showing proteins from iTRAQ using Fibrosarcoma 120tumours, visualised by STRING 9.0 (Confidence view).....................................................................219Figure 4.16: Fixed modification search of iTRAQ fibrosarcoma 120 proteomic data, compliedusing STRING 9.0, evidence view.........................................................................................................220Figure 4.17: Zoomed in region of a fixed modification search, iTRAQ fibrosarcoma 120 samples. 221

Figure 4.18: Zoomed in region of a fixed modification search, iTRAQ fibrosarcoma 120 samples. 222

Figure 4.19: UV profiles from the fibrosarcoma 188 fractions after SCX..................................... 224Figure 4. 20: Frequency of detection of unique peptides by LC-MALDI-MS/MS........................225Figure 4. 21: iTRAQ results from fibrosarcoma 188 fixed filtered search.................................... 226Figure 4. 22: a iTRAQ ratio response relative to the control (113), normalised to an average of 1...........................................................................................................................................................................227

Figure 4. 23: Scaffold proteomic software used for iTRAQ analysis.............................................232Figure 4. 24: Scaffold proteomic software with the employment of 'Q+' function for iTRAQ

ratio analysis........................................................................................................................................... 233Figure 4. 25: iTRAQ fold change ratio using fibrosarcoma 188 tumour digest samples............ 234Figure 4. 26: Vimentin peptides identified, response bar graph and related MS/MS spectrumfrom iTRAQ LC-MALDI-MS/MS Fibrosarcoma 188 tissue.................................................................237Figure 4. 27: Haemoglobin subunit beta-1 peptides identified, response bar graph and related

MS/MS spectrum from iTRAQ LC-MALDI-MS/MS Fibrosarcoma 188 tissue .................... 238

10

Figure 4. 28: Pyruvate kinase isozymes M 1/M 2 peptides identified, response bar graph andrelated MS/MS spectrum from iTRAQ LC-MALDI-MS/MS Fibrosarcoma 188 tissue.................. 239Figure 4. 29: Tenascin peptides identified, response bar graph and related MS/MS spectrumfrom iTRAQ LC-MALDI-MS/MS Fibrosarcoma 188 tissue................................................................ 240Figure 4. 30: Fibronectin peptides identified, response bar graph and related MS/MS spectrumfrom iTRAQ LC-MALDI-MS/MS Fibrosarcoma 188 tissue................................................................ 241

Figure 4. 31: iTRAQ Fixed modification search using Fibrosarcoma 188 tumours, visualised bySTRING 9.0 (Confidence view)............................................................................................................. 243Figure 4. 32: iTRAQ Fixed modification search using Fibrosarcoma 188 tumours, visualised bySTRING 9.0 (Evidence view)..................................................................................................................244Figure 4. 33: Zoomed in region of a fixed modification search, iTRAQ fibrosarcoma 188 samples,visualised through STRING 9.0.............................................................................................................245Figure 4. 34: iTRAQ fibrosarcoma 188 zoomed in region of a fixed modification search,visualised through STRING 9.0.............................................................................................................246Figure 4. 35: iTRAQ fibrosarcoma 188 zoomed in region of a fixed modification search,visualised through STRING 9.0.............................................................................................................247Figure 4. 36: iTRAQ proteomic response results using a Fibrosarcoma 120 treatmenttimecourse.............................................................................................................................................. 248Figure 4. 37: Fibrosarcoma 120 proteomic response using String 9.0.......................................... 250Figure 4. 38: iTRAQ combinations of fibrosarcoma 120 increased and decreased proteins.... 251

Figure 4. 39: iTRAQ proteomic response results using a Fibrosarcoma 188 treatment timecourse...................................................................................................................................................... 252Figure 4.40: iTRAQ proteomic response results using a 24h post CA-4-P Fibrosarcoma 188treatment................................................................................................................................................ 253Figure 4.41: iTRAQ proteomic response results using a 0.5h post CA-4-P Fibrosarcoma 188treatment................................................................................................................................................ 254Figure 4.42: iTRAQ combinations of fibrosarcoma 188 increased and decreased proteins. ...255

Figure 4.43: Bar graph showing protein response relative to the control (113) normalised to anaverage of 1, using fibrosarcoma 188 tissue..................................................................................... 256Figure 4.44: MALDI-MSI multiple Fibrosarcoma 188 sample on tissue digest of Plectin at m/z977............................................................................................................................................................257Figure 4.45: MALDI-MSI multiple Fibrosarcoma 188 sample on tissue digest of Hb subunit alpha at m/z 1416.................................................................................................................................. 258

Chapter 5Figure 5.1: pETBIue-1 vector and sequence used for the synthesis of the recombinant proteinstandard.................................................................................................................................................. 268Figure 5. 2: The final sequence of the recombinant standard........................................................ 268Figure 5. 3: PMF of a novel recombinant standard for MALDI-IMS-MSI validation.................... 274Figure 5.4: MALDI-MSI of on tissue tryptic digests and digested recombinant standard in 72h

treated tissue..........................................................................................................................................275Figure 5. 5: MALDI-MS images for the distribution peptides in fibrosarcoma 188 72h tissue,employing a DNA artificial construct for image validation..............................................................276Figure 5. 6: MALDI-MS images displaying peptides distribution in fibrosarcoma 188 6h treated

tissue, employing a DNA artificial construct for image validation. The spatial distribution of peptides; Actin (m/z 1198) with arrow indicating position of the spotted recombinant standard,

11

Histone HB (m/z 1032)showing the absence of the standard and CHCA peak at 1066 to showinverse image correlation.....................................................................................................................277Figure 5. 7: MALDI-MS images for the distribution peptides in fibrosarcoma 188 Control tissue,

employing a DNA artificial construct for image validation............................................................. 278Figure 5. 8: HDI visualisation of Actin at m/z 1198.7. Arrow indicates the ion of interestseparated by drift time and yellow square denotes the ROI for exportation..............................281Figure 5. 9: HDI visualisation of m/z 1198.6 to display the differentiation in tissue and standardimage due to separation in drift time compared to m/z 1198.7....................................................283Figure 5.10: HDI visualisation of HSP-90a at m/z 1168.5. Arrows indicates the ion of interestseparated by drift time, and yellow ROI square is shown..............................................................284Figure 5.11: HSP-90 a at m/z 1168.5 after exportation to Mass Lynx software from theselected ROI in HDI 1.1......................................................................................................................... 285Figure 5.12: HDI visualisation of Histone H3 at m/z 1032.7. The arrow indicates the ion ofinterest separated by drift time and no signal apparent in the standard.....................................286Figure 5.13: HDI visualisation of Actin at m/z 1198.7. The arrow indicates the ion of interestseparated by drift time......................................................................................................................... 287Figure 5.14: HDI visualisation of Histone 2A at m/z 944.5. The arrow indicates the ion ofinterest separated by drift time.......................................................................................................... 288Figure 5.15: HDI software with image correlation filter for Plectin m/z 1032 displaying fellowHistones...................................................................................................................................................290Figure 5.16: HDI software with image correlation filter for Plectin m/z 1350 showing nodiscrimination between other peaks............................................................. 291Figure 5.17: MALDI-MSI of fibrosarcoma 188 on tissue tryptic digests for the spatialdistribution of Plectin at m/z 1385......................................................................................................292Figure 5.18: MALDI-MSI and conventional proteomic techniques to asses a dose relationship inPlectin...................................................................................... 293Figure 5.19: MALDI-MSI and conventional proteomic techniques to asses a dose relationship inHSP-90..................................................................................................................................................... 294Figure 5. 20: Issues and factors to consider when performing on tissue digestion for MALDI imaging experiments............................................................................................................................. 296

List of Tables

Table 1: Cancer incidence throughout the regions of the w orld.................................................... 18Table 2: Current anti-cancer treatments..............................................................................................29Table 3: Categorisation of vascular targeted therapies.....................................................................30

Table 4 : Example of drugs within each category................................................................................31

Table 5 : Inhibiting the process of angiogenesis.................................................................................32Table 6 : Mass spectrometers for proteomics.................................................................................... 45Table 7 : Common matrices used in UV- MALDI..................................................................................54Table 8: Identities and Mascot scores from the IMS-MS/MS analyses of peptide signals...........91Table 9: Twelve peptides from target proteins chosen for positive identification..................... 267

12

AbbreviationsACN: acetonitrile

Aldoartl: Fructose-bisphosphate aldolase A

AL022784: alpha-enolase 1 (Enol)

ANI: aniline

APCI: atmospheric pressure chemical ionisation

APPI: atmospheric pressure photoionisation

BCA: bicinchoninic acid assay

Calr: Calreticulin

CHCA: a-cyano-4-hydroxycinnamic acid

CHCI3: chloroform

CID: collision-induced dissociation

CA-4-P: combretastatin A-4 -3-0 phosphate

CSC: cancer stem cell

dH20: deionised water

DMXAA: 5,6-Di-methylxanthenone-4-acetic acid

Eeflal: Elongation factor 1-alpha 1

ESI: electrospray ionisation

ESI-LC-MS: electrospray ionisation-liquid chromatograpgy-mass spectrometry

ESI-LC-MS/MS: electrospray ionisation-liquid chromatograpgy-tandem mass spectrometry

EtOH: ethanol

FFPE: formalin fixed paraffin embedded

GAPDH: Glyceraldehyde-3-phosphate dehydrogenase

13

Gpil: Glucose-6-phosphate isomerase

GRP-78: glucose- regulated protein 78 (Hspa5)

HE: haematoxylin and eosin

HIF-1 a: hypoxia- inducible factor 1 alpha

HPLC: high performance liquid chromatography

HSP: heat shock protein

IHC: immunohistochemistry

IMS: ion mobility separation

IQGAP1: Ras GTPase-activating-like protein

iTRAQ: isobaric tag for relative absolute quantitation

KTR1: Keratin-l(type II skeletal-1)

LC: liquid chromatography

LC-MALDI: liquid chromatography-matrix assisted laser desorption ionisation

LC-MALDI-MS/MS: liquid chromatography-matrix assisted laser desorption ionisation-tandem

mass spectrometry

LC-MS/MS: liquid chromatography -mass spectrometry tandem mass spectrometry

Lgalsl: Galectin-1

MALDI: matrix assisted laser desorption ionisation

MALDI-MS: matrix assisted laser desorption ionisation-mass spectrometry

MALDI-IMS-MS: matrix assisted laser desorption ionisation-ion mobility separation-mass

spectrometry

MALDI-IMS-MSI: matrix assisted laser desorption ionisation-ion mobility separation-mass

spectrometry imaging

MALDI-MSI: matrix assisted laser desorption ionisation- mass spectrometry imaging

MALDi-ToF: matrix assisted laser desorption ionisation-time of flight

Mcm5: DNA replication licensing factor MCM5

MS: mass spectrometry

m/z: mass- to -charge ratio

MeOH: methanol

MudPIT: multidimensional protein identification technology

N2 laser: nitrogen laser

Nd: YAG: neodymium-doped yttrium aluminium garnet

Nd:YV04: yttrium ortho-vanadate

NH4HC03: ammonium bicarbonate

NHS: N-hydroxysuccinimide

OcGIc: Octyl-a/p-glucoside

PCA: principle component analysis

PCA-DA: principle component analysis-discriminant analysis

PLSDA: partial least squares discriminant analysis

PMF: peptide mass fingerprint

PRDX1: Peroxiredoxin-1

Q: quadrupole

QqTOF: quadrupole time of flight

ROI: region of interest

S100a6: Calcylin

SMA: smooth muscle actin

TAA: tumour associated antigen

TFA: trifluoroacetic acid

Tgfbi: Transforming growth factor-beta-induced protein ig-h3

15

Tnc: Tenascin C

TOF: time of flight

TOF MS: time of flight mass spectrometry

Tpil: Triosephosphate isomerase

TSP: thermospray

TWIGS: travelling wave ion guides

VDA: vascular disrupting agent

VEGF: vascular endothelial growth factor

VEGFR: vascular endothelial growth factor receptor

VIP: variable importance in projection

VTS: vascular targeted strategy

VTT: vascular targeted therapy

16

Chapter 1General Introduction

17

1.1 Cancer

1.1.1. Statistical information, incidence/ prevalenceAt the time that this account was written the latest official current world population estimate

was ~7,094,391,160. This approximation was made based on population figures on the

27thJanuary 2013 (Worldometers 2013). Demographic predications state that by 2028 the

number of our planet's inhabitants will exceed 8 billion (Cancer Research UK 2007).

The current population is described as 'aging', the relevance of this from a cancer perspective

is that the disease has been known to predominantly affect the older generations (Cancer

Research UK 2007). Worldwide life expectancy is increasing, people are living longer and the

average lifetime expectancy at birth is said to be 65yrs possibly escalating to 76yrs by the year

2050. Throughout the world there are still ~12.7 million people who are annually diagnosed

with cancer with estimations likely to increase to ~26 million by 2030 (World Cancer Research

Fund 2010). The UK has the 22nd highest overall cancer rate in the world.

Africa 715,571 6

Asia 6,092,359 48

Europe 3,208,882 25

Latin America & 906,008 7Caribbean

North America 1,603,870 13

Oceania 135,864 1

Developed regions 5,555,281 44

Developingregions 7,107,273 56

World 12,662,554 100

Table 1 : Cancer incidence throughout the regions o f the world (Adapted from Cancer Research UK 2011).

It is evident from Table 1 that Asia has the largest proportion of new cases diagnosed but this

is understood to be a reflection of the magnitude of population size within that particular

continent.

18

On a global scale the mortality rate is such that in 2008 there were an estimated 7.6 million

deaths as a result of cancer, the bar graph below shows the proportion of deaths due to cancer

in the regions of the world (World Health organisation 2012).

25

Western Europe Amencas World South-East Eastern Africa Pacific Asia Mediterranean

Area

Figure 1 .1 : Percentage of all deaths due to cancer in the different regions o f the world (WHO) Regions of the World, 2008 Estimates (Cancer Research UK 2007).

The following map (Figure 1.2) entitled 'Cancer Incidence Worldwide' (Cancer Research UK

2011) represents the most frequently diagnosed cancers within the areas shown in 2008. The

cancer types which account for the largest proportion of cases diagnosed are found to be; lung,

breast, bowel, stomach and prostate.

19

2 \ ^

I 1 I I

sr.fitrf?

5

Clj y S :: g - -S r . r- f-

fewi isJsflg*

I

O T r

AmsJSS*iy 4) oj) g 8 3 - e

ifS C c ¥ £

qj t ) !E

h JI Jill

i?«ss'

zi

orsj

«N

s3CS»ill

It is stated that 47% of all cancer related death are from lung, bowel, breast and prostate

cancers (Cancer Research UK 2011). Figure 1.3 displays the 20 most common causes of death

from cancer in the UK.

Male ■ Female

Lung Bowel* Breast

Prostate Pancreas

Oesophagus

Stomach Bladder

Leukaemia Non-Hodgkin Lymphoma

Ovary Kidney

Brain and Central Nervous SystemLiver

Other Digestive Organs Myeloma

Mesothelioma Malignant Melanoma

Oral Uterus

Other Sites**

0 5.000 10,000

Number of Deaths

15,000 20,000

Figure 1.3 : The 20 Most Common Causes of Cancer Death, UK, 2010 (Cancer Research UK 2011).

21

1.1.2 Tumour typesThere are over 200 different kinds of cancer which can be classified by cell of origin (Cancer

Research UK 2009). Tumours are generally classified as benign or malignant to illustrate that

not all are cancerous (Medline Plus 2010). Malignant is the term used to describe an abnormal

neoplasm which is cancerous and benign is a non-cancerous growth. The cells which constitute

the cancerous tumour types have metastatic capabilities and are able to reach other regions of

the body to form secondary tumours whilst benign tumour types do not metastasize.

Figure 1 .4 : Diagrammatical representation o f the possible routes o f cell division (National Cancer Institute 2009).

The cancer cell has the ability to elicit morphological changes as depicted in Figure 1.4., to

ultimately disregard the usual signals for cell suicide or apoptosis.

It is possible to divide solid tumours into 3 broad categories these being; carcinomas which

originate from altered epithelial cells which function as linings of a body surface region i.e. skin,

gut, cervix, mouth, lung, sarcomas occur in connective tissue i.e. bone, muscle, cartilage and

lymphomas/ leukaemia which describes cancers that arise from lymphoid organs (lymph nodes,

spleen) and cells from bone marrow (http://www.chemotherapy.com/ 2011).

Loss of Normal Growth Control

Normal cell division

CeB SujckV or ApoptoiH

Cancer cell division

Flrct Second mutation mutation

TTWd Fourth Of mutation later mutation

Uncontrolled growth

22

1.1.3 Overview of current anti-cancer therapiesPre-treatment clinical assessment of patient tumours is generally based on tumour size,

through histological staining, tumour grade and is regularly coupled with the employment of

various imaging techniques (Koch et al 2010). The tumour grading is a system that is used as a

reference in order to classify cancer cells. Analysis of the abnormalities exhibited by cancer

cells are visualised by microscopy, here the aggressiveness of the tumour can be determined.

Although factors used for tumour grading vary in cancer type the cellular architecture and

growth patterns are studied (National Cancer Institute 2009). The outcome of these

evaluations will then determine the appropriate treatment route.

An example of a grading system commonly used by Doctors is the TNM system. This grading

system is an accepted method by the International Union against Cancer (UICC) and the

American Joint Committee on Cancer (AJCC) (National Cancer Institute 2009). In brief the

rationale behind the TNM method is that the T refers to the extent of the tumour, the 'N'

amount of colonisation in the lymph nodes and 'M ' the presence of metastasis. A number is

assigned to the T /N /M indicative of size and metastatic status. The TNM system is further

annotated, details are as follows:

Primary Tumours (T); TX - Primary tumour cannot be evaluated, TO - no evidence of primary

tumour, Tis - Carcinoma in situ and T l, T2, T3, T4 refer to the size/ severity of the primary

tumour.

Regional Lymph Nodes (N); NX-regional lymph nodes cannot be evaluated, NO - no regional

lymph node involvement and N l, N2, N3 depicts the involvement of regional lymph nodes and

to what extent.

Distant Metastasis (M) - MX distant metastasis cannot be evaluated, M0 - no distant

metastasis and M l distant metastasis is present.

Further categorisations of the above TNM combinations are then graded as follows;

Stage 0 - Carcinoma in situ, Stage I, Stage II, and Stage III Higher are indicative of the extent

ofcancer: Larger tumour size and/or spread of the cancer beyond the organ in which it first

developed to nearby lymph nodes and/or organs adjacent to the location of the primary

tumour.Stage IV - the cancer has metastasised to another organ(s).

The above information regarding the TNM system was taken from (National Cancer Institute

2009, http://www.cancer.gov/cancertopics/factsheet/detection/staging).

23

Currently the main course of action for the treatment of tumours is by surgical removal,

chemotherapy, radiotherapy, systemic adjuvant treatments and in recent times some agents

have been used to achieve a more targeted approach (Koch et al 2010).

A brief account follows which considers the history of how both chemotherapy and

radiotherapy have been implemented as anticancer therapies within the clinical setting.

Secondly, a compilation anticancer therapies in table format aims to summarise the range of

treatments available with an overview of how they are administered, anti-cancer mechanisms

and possible side effects.

It is documented that work by German scientist Paul Ehrlich lead to the adoption of the widely

used term 'chemotherapy' (DeVita and Chu 2008). During turn of the 20th century, the focus

of his work was the discovery and advancement of drugs to combat infectious disease.

Amongst Ehrlich's achievement was the use of animal models to enable assessment of

potential anticancer agents.

With reference to a cancer chemotherapy timeline by DeVita and Chu 2008, it took

approximately 40 years to develop a suitable experiment model (DeVita and Chu 2008).

Although this period included the revolutionary success of transplantable tumour rodent

models by New Yorker George Clowes a surgeon/ researcher from the Roswell Park Memorial

Institute.

With regard to early anticancer therapeutics, 1908 saw the use of arsenicals which had been

originally used to eradicate Syphilis (DeVita and Chu 2008). 'Fowlers solution1 (arsenicals) as it

was known, was used to treat leukaemias and apparently a common practice well into the

1930's (Papac 2001).

Optimism arising from these early triumphs in cancer therapy was dampened following work

investigating the use of nitrogen mustard (1943) to eradicate lymphomas. (DeVita and Chu

2008). Reports stated that patients suffering from lymphomas experienced only short-lived

remission and thus began a decline of confidence in the concept of 'a drug to cure cancer'.

It was not until the opening of the Cancer Chemotherapy National Service Centre (CCNSC) in

1955 that a possible future for cancer drug development was confirmed (DeVita and Chu 2008).

This change of heart was said to be encouraged by the positive response to the drug

methotrexate in children with acute leukaemia. Since the 1960s, the development of anti­

cancer drugs has been aided by many advances. Experimental models which employ

xenografts and novel cell culture methods have been used. Adjuvent therapy (chemotherapy,

radiation therapy, hormone therapy, targeted therapy, or biological therapy) given in addition

24

to the primary treatment was developed and investments into molecular biology have been

made. From the beginning of the 21st century, there seems to have been a different approach

in anticancer therapeutics emerging (DeVita and Chu 2008). Completion of sequencing of the

genome appears to have led to a more targeted approach to cancer treatment. Through

advancements in chemotherapy a major decrease in mortality is seen by 2007, the latter

considering knowledge of prevention and improvements in diagnostics.

Radiotherapy/ Radiation Oncology are terms which are commonly associated with cancer

treatment with an estimated 50% of patients receiving this form of anticancer therapy for both

palliative and therapeutic means (Delaney et al 2005). In combination with other forms of

management, radiation therapy features in approximately 40% of those patients who are

cured of their disease (Shariq et al 2009).

After the German physicist Wilhelm Roentgen discovered x-rays in 1895, the concept of using

x-rays as a therapeutic aid was not long to follow, assisted greatly through work on radioactive

elements (radium, polonium) by Marie Curie and husband Pierre (Connell and Heilman 2009).

There have been numerous clinical, technological and biological advances in radiotherapy

since Roentgen's breakthrough over 115 years ago. Examples of key clinical advances are Henri

Coutard using fractionated external beam radiotherapy (XRT) in 1928 to cure head/ neck

malignancies, the use of the proton beam (1961) and the high precision accuracy of

stereotactic body radiotherapy (1995) (Connell and Heilman 2009).

A large proportion of patients, however, still exhibit metastasising cancers even after

treatments such as combination radiotherapy. (Koch et al 2010). One explanation could be the

inadequate targeting of cancer stem cells (CSCs), i.e. CSCs could be sheltered by hypoxic

tumour environments enabling them 'radio-resistant'. Therefore increased doses of radiation

could be concentrated to these specific resistant areas. The following proposal (Figure 1.5) by

Koch et al (2010) suggested how CSC-predictive test regime and radiotherapy could be

implemented.

25

Current therapy

Entity I Histology Tumour volume Tumour stage Metabolic imaging (PET)

CancerFuture therapypatient

(Clinical) predictors(Clinical) predictors

Entity I Histology Tumour volume Tumour stage

• Metabolic imaging (PET)

Additional CSC-specific predictive tests

CSC densityCSC distributionCSC sensitivity to therapyImaging of CSC niches I enriching micromilieuSignature (DNA, RNA, protein etc.) of CSCs

Estimation of response to therapy

Decision for empirical class solution

r.......

±Estimation of response to therapy

CSC-based design of individual therapy within class solution

Frequently incomplete eradication of CSCs Improved eradication of CSCs

Frequent tumour recurrences

El 80

--------Higher cure rates ____________

& 0

0

@

| £ BOo I’ GO

| | 40

K & 20i O

T e r*

m Tumour eradicated

O Non-CSC

0 C S C

Figure 1. 5 : Proposal of adding CSC specific predictive tests to therapy regimes fo r potential increase in tumour control probability (Koch et a 12010).

Table 2 gives a list of current anti-cancer therapies, administration routes, mechanism and

possible side effects related to the particular therapy.

26

Anti-cancer Therapy Administration Route Anti-cancer Action Examples of Possible

Side Effects

Biological therapy Orally, in jected,

in travenous,

in tra d e rm a l/

in tram uscu la r

Targets th e im m une

system, re ta rd a tio n o f

cancer cells, e.g. IL-2,

In te rfe ro n alpha,

Rashes, nausea, flu like

sym ptom s, hypotension

Chemotherapy Orally, in jected, Targets cells o f high Hair loss, rashes,

in travenous, in tra - p ro life ra tive na tu re d ia rrhoea , nausea/

a rte ria lly , in su ffla tio n a lte ring DNA synthesis/ vom iting , d igestion

fun c tion , in te rfe rence issues, fa tigue

w ith cell cycle e.g.

a lky la ting

agents(C isplatin),

an tim e tab o lite s

(m e tho trexa te ),

a n titu m o r an tib io tics

(B leom ycin),

anthracyclines

(Idarubic in), p lan t

alkalo ids (V inblastine),

taxanes (Paclitaxel),

m onoclona l an tibod ies

(Bevacizumab),

Radiation Therapy Patient exposure to E xp lo ita tion o f rap id Hair loss, rashes.

source o f rad ia tion , d iv id ing cell types, d ia rrhoea , nausea/

localised beam exposure o f rad ia tion to vom iting , d igestion

tu m o u r regions issues, fa tigue ,

ne u trop en ia

Targeted Therapies O rally (small m olecu le Target high Nausea/ vom iting ,

inh ib ito rs ), p ro life ra tive cells, rashes, d ia rrhoea , skin

In travenously Inh ib ito rs o f d isco lo ra tion ,

(m onoclona l angiogenesis, ty ros ine dyspepsia, flu c tu a tio n s

an tibodies), kinase rece p to r in b lood pressure

in h ib ito rs , to im pede

tu m o u r en la rgem ent

and survival

27

Proton Beam Therapy Patient exposure to

source o f rad ia tion ,

h igh ly specific localised

beam

Based on s tab ility o f

p ro tons enabling 3D

m an ipu la tion o f beam.

Highly specific to

tu m o u r region w ith

m in im al exposure to

norm al tissue

Hair loss, rashes,

d ia rrhoea , nausea/

vom iting , d igestion

issues, fa tigue ,

neu tropen ia

Surgical Oncology Perform ed by Surgeon

p re /p o s t chem othe rapy

or rad io the rapy if

m alignan t

Surgical rem oval o f

tu m o u r tissue plus

som e hea lthy tissue

fro m periphery,

transp la n ta tio n o f bone

m arrow

Risk o f in fe c tion , ha ir

loss, by 'p rep a ra tive

reg im en ' (bone m arro w

transp la n t

surgery)d iarrhoea ,

nausea/ vom iting ,

fa tigue

Gene Therapy A fte r transduc tion cells

are re in troduced in to

pa tien t by in travenous

in fusion, insu ffla tion ,

d irec tly in to tu m o u r if

accessible

Cancer cells th a t have

taken up transduced

cells are said to have

increased suscep tib ility

to chem othe rapy and

o th e r anticancer

tre a tm e n ts

Long te rm side effects

unknow n, f lu -like

sym ptom s,

hypo tens ion , vom iting ,

nausea

Hormone Therapy O rally, in jec tion ,

surgical in te rve n tio n

In te rfe rence w ith

ho rm one cell surface

receptors to block

activa tion o f a

Flu-like sym ptom s,

w e igh t gain, th rom b os is

in rare cases, a lopecia,

m u s c le / jo in t pain

pa rticu la r horm one

th o u g h t to encourage

tu m o u r g ro w th , e.g.

Tam oxifen - a n ti­

oestrogen, Fulvestrant

- oestrogen recep to r

an tagon ist

Vaccine Therapies Presently tria lled on Aim to induce tu m o u r Long te rm side e ffec ts

pa tien ts w ith advanced im m unogen ic ity unknow n, flu - like

illness, no t ye t am ongst sym ptom s, vo m itin g ,

the m ainstay o f nausea

28

anticancer trea tm e n ts ,

the search fo r novel

tu m o u r associated

antigens is underw ay

Complementary / Various rou tes

Alternative Medicine ad m in is tra tion

o f Num erous categories Advice requ ired fo r the

some w h ich include; p a tien t to consider

sp iritua l healers, poss ib ility o f cross

d ie ta ry in te rven tion , rea c tiv ity w ith

herbal m edicines conven tion means o f

psychological m ethods, tre a tm e n t

cu ltu ra l techniques

based on tra d itio n

Table 2: Current anti-cancer treatments (Oncolink 2010., National Cancer Institute 2009).

29

1.1.4 Vascular targeted strategiesTargeting the tumour vasculature is a promising anticancer strategy and drugs of this nature

are in various stages of clinical trials. Here, the types of vascular -targeted therapies (VTTs) will

be discussed looking at pitfalls, limitations and future aspirations regarding this kind of

approach to anticancer medicine. Firstly a brief overview will describe the targets and

mechanisms of VTTs focusing then on the vascular disrupting agent Combretastatin A-4-3-0-

phosphate (CA-4-P/ ZybrestatTM/ Fosbretabulin).

The importance of angiogenesis in the development and rejuvenation of tumour tissue was

initially suggested by Folkman in 1971 (Folkman 1971). In vivo experiments using rodent

models carried out by Denekamp, showed that blocking and interference of the tumour blood

supply resulted in deterioration of tumour tissue (Denekamp 1990). Denekamp then proposed

the tumour endothelium as a therapy target by exploiting its high proliferative fragile nature to

ultimately interfere and/or destroy the tumour vasculature.

Vascular targeted agents can be divided in to two main categories (Table 3) i.e. those which

inhibit formation of new blood vessels (antiangiogenic) and the vascular disrupting agents

which take advantage of the substandard micro-architecture of the tumour vasculature

(Siemann et al 2005). Examples of VTTs and targets are shown in Table 4.

Vascular-targeted therapy

Antiangiogenic agents Vascular disrupting agents

Stop new vessel fo rm a tio n (neovascularisa tion) D isrupt a rch itec tu re o f tu m o u r b lood vessels

Continual tre a tm e n t Acute tre a tm e n t

Early stage - asym ptom atic metastases Established disease, e ffec tive against large

tu m o u r masses, cause cen tra l necrosis

Table 3: Categorisation of vascular targeted therapies (Siemann 2005).

30

Vascular-targeted strategies

Antiangiogenic Vascular disrupting

Target

VEGF/VEGFR 1-3

inhibition

Tie-2/angiopoietin

Av03 integrin

Endogenous inhibitors

VE-cadherin

Unknown

Example

Bevacizumab, VEGF-

Trap, ZD6474

A422885.66

V itaxin , c ileng itide

Ang iosta tin

E4G10

Thalidom ide

Target

Tubulin

Cytokine induction

Ab Toxin

Av|33 integrin

Unknown

Example

C-A-4-P, AVE8062A

DMXAA

VEGF-gelonin

Gene the rapy , pep tide

ta rge ting

PDT

Table 4 : Example o f drugs within each category (Siemann 2005).

The design hypothesis of antiangiogenic agents is to hamper the cascade of biochemical events

that result in new vessel development (Siemann 2005). Amidst the complex array of signals

and receptors which govern the latter process, vascular endothelial growth factor (VEGF) is

thought to be an important target for impeding antiangiogenesis. An example of anti

angiogenic therapy is Bevacizumab/ Avastin, this uses monoclonal antibodies to target VEGF or

its' receptors, as does ZD6474 a tyrosine kinase inhibitor. (Wedge et al 2002). One treatment

success story of Bevacizumab was when it was used in fluorouracil based combination

chemotherapy to treat colorectal cancer. (Hurwitz et al 2004).

31

Table 5 illustrates the stages of angiogenesis and goals of the 'antiangiogenic 'approach.

(Siemann et al 2005).

The steps towards

Angiogenesis

Antiangiogenisisapproach

Endothialc«llproliferation

Angiogenic switch

Upre gulation of factors and docking of

endothelialcell receptors which promote

angiogenesis

Matrix metalloproteinase activation and

degradation of basement membrane

* \

Endothelial cell migration and tube formation

t ......

Inhibition of proangioge me factors

k I II III III II I IMI

Inhibition of stimulus amplification of

e ndoge nous suppre ssors of angiogenesis

Receptor antagonists signal transduction

inhibitors

Ihibition of matrix breakdown

f ............

Inhibition of tube formation

Table 5 : Inhibiting the process o f angiogenesis (Siemann et al 2005).

32

1.1.5 Vascular disrupting agentsVascular disrupting agents (VDAs) exploit the characteristic tumour vasculature (Figure 1.6)

which is notoriously sinusoidal and leaky, and is hence susceptible to vascular disruption.

(Tozer et al 2008). Shortly after administration of VDAs major shutdown occurs of the tumour

vascular network.

Healthy vasculature, m ature blood

vessels, smooth muscle and pericyte

support

Tum our vasculature, im m ature blood

vessels, lack o f support by smooth

muscle and pericytes

Figure 1. 6 : Images to illustrate the substandard tumour vasculature (Oxigene 2010).

Flavanoids (e.g. DMXAA) and tubulin-binding agents (e.g. CA-4-P) are VDAs, both are designed

to disrupt tumour blood vessels but work in different ways as the chemical structures suggest

in Figure 1.7.

33

CA-4-P (Fosbretabulin)

HoCO

O— P— ONa

ONaOCH

H

DMXAA (Vadimezan)

O

CH, c h 2c o o h

CA-l-P (Oxi4503)

H3CO.

h 3CO

OCH

ONa 0 \ | ONa

I, 0

o/ /

'O — P — ONa\

OCH3 0NaAVE8062 (Ombrabulin)

H3CO.

H3C 0 ‘

OCHNHF.HCI

OCH

Figure 1. 7 : Chemical structures o f vascular disrupting agents (Kanthou and Tozer 2007).

The mechanism of 5,6-dimethylxanthnone-4-acetic acid (DMXAA/ ASA404) is still not yet fully

understood but it is thought to operate via the induction of TNF-a. (Thorpe et al 2003).

Through the action of DMXAA it is suggested that TNF-a is retained within the tumour

microenvironment thus stimulating additional amounts of TNF-a amongst other cytokines.

The tumour blood vessels are weakened with the subsequent changes in morphology leading

to haemorrhagic necrosis (Kanthou and Tozer 2007). In Phase I clinical trials DMXAA showed

the drug had a well tolerated anti-tumour action, Phase II involved combination with34

conventional chemotherapy agents with positive outcomes when used in non-small cell lung

cancer patients. DMXAA then moved into Phase III clinical trials which unfortunately did not

produce the same effects as in its early trial success (Buchanan 2012).

Microtubule depolymerising agents have a destructive effect on the architectural integrity of

3D capillary composition (Bayless and Davies 2004). The following diagrammatical

representation (Figure 1.8) from a paper by Kanthou and Tozer (2007) details the proposed

mechanism of action of a microtubule depolymerising agent (e.g. Combretastatin A-4-3-0-

phosphate).

H IV '

A< m \ IV

Untreated endothelial cells

• intact micronjbulss• non-contractile cells• tight VE-cadherin junctions• low permeability

• low RhoA-GTPase activity• low Rho kinase activity

Microtubules

— Actin stress fibers

%% Cell surface actin blebs

VE-cadherin junctions

VDA-treated endothelial cells

• disrupted microtubules• contractile cells• actin stress fibers• strong fecal adhesions• disrupted VE-cadherin junctions• high permeability

• high RhoA-GTPase activity• high Rho kinase activity• high SAPK-2'p3d activity• high ERK activity?

VDA-treated endothelial cells

• disrupted microtubules• conlractile-blebbrng cells• actin-lined surface blebs• (Disassembled focal adhesions• lost VE-cadherm junctions• high permeability

• necrotic celts

• high RhoA-GTPase activity• high Rho kinase activity• high SAPK-2/p38 activity• low ERK activity?

Dug Dacmvry Today S‘rat*yes

Figure 1. 8 : Proposed mechanism of action o f Combretastatin A-4-3-0-phosphate (Kanthou and Tozer 2007).

It is suggested that increased permeability post CA-4-P administration is a major factor in the

disruption of endothelial organisation; this is evident from Figure 1.8 (Kanthou and Tozer

2007). Interference in vascular integrity due to abnormality in the cytoskeleton, subsequently

leads to stress fibre formation and endothelial cells as a result become more rounded

('blebbing'). Haemorrhagic necrosis follows exacerbating the already hypoxic regions within

the tumour microenvironment.

35

The remodelling of the actin cytoskeleton and formation of actin stress fibres shown in Figure

1.8 is clearly demonstrated in the Figure 1.9 adapted from Kanthou and Tozer 2002 and has

also been shown by Siemann (2010).

(b)

(d )

Figure 1. 9 : Fluorescent staining o f Human umbilical vein endothelial cells, Control and post CA4P administration (Kanthou and Tozer 2002). Figure 1.9 shows dual staining for filamentous actin (Texas Red Phalloidin (red)) and beta tubulin (green): (a) control actin, (b) control tubulin, (c) CA4P (1 uM for 30 min) actin, (d) CA4P (1 uM for 30 min) tubulin.

In vivo studies using CA-4-P have clearly shown how the effects of the drug can induce vascular

shutdown resulting in increased hypoxia. (Tozer et al 2008). Studies on mouse models (Figure

1.10) looking at hypoxic responses in mammary carcinomas after 50mg/kg CA-4-P, indicated

increased pimonidazole staining compared to non treated. Similarly, dorsal skin flap

experiments in mice (Figure 1.10) displayed the de-vascularising properties of CA-4-P

HOmg/kg.

(a)

(c)

36

»*«n

Figure 1 .1 0 : CA-4-P and its effect on hypoxia induction and vascular shutdown (Tozer et al 2008). Panel A shows pimonidazole staining o f the control untreated tumour at the top with the bottom being 1 hour after CA-4-P, panel B is the untreated image of a rat bearing P22 sarcoma directly below is the tumour 1 hour 40 min post CA-4-P administration in vivo using a window chamber method for analysis o f treatments effects.

Abnormal tumor vascular network with poor perfusion, high interstitial fluid pressure and areas of hypoxia

Tumor-VDAs

Blind ends

Abnormal bulges

Direct endothelial cell effects Vascular disruption

Remaining rin of viable cells that can be targeted with chemotherapeutic agents and radiation

Necrotic central core

Decreased nascent Vessel number

Anti-VEGF effectsInitial vascular normalization 1Anti-Angiogenic

Agents

I p a l i i[ ’.vSi I VS f j Improved blcod flow'

loading to improved delivery of

chemotherapeuticagents

Figure 1 .1 1 : Exploitation of tumour vasculature by antiangiogenic treatment and vascular disrupting agents (Siemann 2010).

To summarise the different actions of both VDAs and antiangiogenic treatments a diagram

(Figure 1.11) by Siemann (2010) highlights the preclinical effects of each type.

With all new drugs/ therapies in development there are numerouscomplications to overcome

and consider i.e. the therapeutic window with the balance between safe dosage and the onset37

of adverse effects. The phrase 'tumour evasion' frequents many article titles and appears to be

a key issue in the advancement of this type of anti-tumour medicine.

Zhi-Chao and Jie (2008) present valid points within their review on tumour evasion in a section

entitled "Hypoxia: Two sides of the coin". They state that disruption by vascular targeted

treatment to increase hypoxia severing the lifeline of the tumour, can indeed have undesirable

consequences. It is suggested that there is a requirement to unravel some 'cause-effect'

connection. Hypoxia-inducible factors (HIF) have been known to correlate with the up-

regulation of VEGF, involved in angiogenesis regulation. So could VTTs amplify the angiogenic

process after the creation of an increased hypoxic tumour environment? Hypothetically

speaking this could even help the metastatic process ensuing tumour invasion. (Abdollahi and

Folkman 2010).

Evasion of the hosts' immune system is one feat that the tumour is capable of and there is a

school of thought that claims there could be a similar evasion mechanism in place against VTTs.

The heterogeneous nature of the tumour mass could well contain a cell population deemed

oblivious to vascular targeted strategies. (Zhi-Chao and Jie 2008). The latter is supported by

the hypothesis of four possible vascular phenotypes in existence; normal pre-existing, tumour

phenotypic vessels, neo-vasculature which is normal and abnormal neo-vasculature. This

could explain why some cells remain unaffected including the well documented viable tumour

rim. (Hinnen and Eskens 2007).

The viable tumour rim being a speculative resistance factor is also thought to achieve evasion

through its positioning, adjacent to the non-diseased tissue. (Tozer et al 2008). Consequently,

seizing oxygen and essential nutrients as it resides in this prime location.

The explanations why resistance issues exist against VTTs appears to be complex and multi­

factorial. That said progress is being made through various combination therapies. Xenograft

experimental models displayed increased quantities of VEGF and basic fibroblast growth factor

following doses of CA-4-P, with angiogenic circulating progenitor cells in murine studies. (Tozer

et al 2008) Promising responses have been reported by using VDAs and anti-angiogenic VEGFR

tyrosine kinase inhibitor treatments in combination therapy.

38

1.1.6 Biomarker DiscoveryWhen considering biomarker discovery as it relates to oncology many complexities exist. There

are countless biochemical pathways with numerous prospective targets, with more layers of

complexity added by the multifaceted characteristics of the cancer cell. A cell unchallenged by

host immunogenicity, that has mastered tissue invasion and anti-cancer drug evasion.

Nevertheless a good place to start is given bythe 'hallmarks' of cancer created by Hanahan and

Weinberg (2000).

Whilst the 'hallmarks' of cancer reveal the attributes and capabilities of this lethal cell type, a

further proposal by Lord and Ashworth (2010) relates the possible drug targets to each

characteristic of the cancer cell to provide inspiration for biomarker exploration. (Lord and

Ashworth 2010). A diagram in the article by Lord and Ashworth (2010) supplies a wealth of

information on the latter subject however, one consideration to be made is the possible

upshot from molecular interactions and when choosing one trait to target such as 'sustained

angiogenesis' or 'proteotoxic stress'.

This concern is echoed by Lord and Ashworth (2010) as they discuss how in order to decipher

cancer mechanistics, medicine should bear in mind how drugs disturb whole biochemical

networks as opposed to solitary pathways. A concept which has been termed, "drugging the

undruggable".

Over a decade after their original paper, Hanahan and Weinburg (2011) proposed an updated

version of 'The Hallmarks of Cancer' entitled 'Hallmarks of Cancer: The Next Generation'. The

idea of two additional hallmarks of cancer were put forward these being the "reprogramming

of energy metabolism"and "evading immune destruction".

Figure 1.12 displays the original hallmarks with the addition of 'deregulating cellular

energetics' this referring to reprogramming of cellular metabolism in favour of proliferation

and 'avoiding immune destruction' which describes the ability to fend off advances from

macrophages, natural killer cells and T/ B lymphocytes (Hanahan and Weinburg 2011). The

article proposes two enabling characteristics, genome instability and mutation and tumour-

promoting inflammation. Both of which are said to be key orchestrators in the original and

emerging characteristics.

Hanahan and Weinburg also included therapeutic suggestions to hinder a particular cancer

hallmark as shown in Figure 1.8.

39

Aerobic glycolysis inhibitors

( Proapoptotic BH3 mimetics

PARPinhibitors

C ) CEGFRinhibitors

Cyclin-dependent kinase inhibitors

Sustainingproliferative

signaling

Evadinggrowth

suppressorsImmune activating anti-CTLA4 mAb

Avoidingimmune

destruction

C Enablingreplicativeimmortality

TelomeraseInhibitors

Tumor promoting

inflammation

Inducingangiogenesis

Selective anti inflammatory drugs

Activating invasion & metastasis

) cInhibitors of VEGF signaling

Inhibitors of HGF/c-Met

Deregulating cellular

energetics.

Resistinqcell

death

Genome instability &

mutation

V *

Figure 1 .1 2 : Therapeutic targeting o f the hallmarks o f cancer (Hanahan and Weinberg 2011).

Cancer-Associated Fibroblast , (CAF)~

Endothelial Cell (EC) 1

Cancer Stem Cell (CSC)

r - Cancer Cell (CC)

v w n r immunePericyte (PC)' inflammatory Cells

' :V (ICs)

Local & Bone marrow- derived Stromal Stem & Progenitor Cells

Invasive Cancer Cell

sat'

Core of Primary Tumor microenvironment

Invasive Tumor microenvironment

Metastatic Tumor microenvironment

Figure 1 .13: The cells o f the tumour microenvironment. (Hanahan and Weinberg 2011).

Advances in the knowledge of the individual cancer cells that make up the solid tum our have

enabled appreciation of their influential drive on the tumour microenvironment. Figure 1.13

40

depicts the varied cell types and sub types. Each cell has a specialised role and collectively

supports the plasticity required for tumourigenesis (Hanahan and Weinberg 2011).

Interestingly it is known that the inflammatory immune cells present within this cellular

assemblage, possess both pro-oncogenic and anti-tumour activity. The latter is thought to aid

future tumour seeding during the metastatic process.

1.1.7 Considering Vascular disrupting agents for anti-cancer therapeuticsIt is known that one vessel is primarily responsible for supplying the tumour cell population

with oxygen and nutrients for growth, development and excretion of waste metabolites.

(Thorpe et al 2003). Disruption of this vital sustenance by vascular disrupting agents causes

rapid shutdown of blood flow to and from the tumour.

Unlike some conventional cancer therapies which focus on cell killing VDAs create damage via

morphological changes to produce drastic effects on the tumour microenvironment. Drug

deliverance is simple due to endothelial positioning near the bloodstream.

In addition to the above, another reason to support the use of vascular targeted therapy is that

the diploid target cells are thought not to be as drug resistant due to low risk of mutations. The

drug has good clinical measurability being blood flow, in most cases short-term exposure of a

VDA creates sufficient hypoxic conditions. To conclude, intermittent dosage with regular

treatments may be adequate to provide a synergistic effect, not true in the case of

antiangiogenesis agents. (Thorpe et al 2003).

1.2 The Role of Mass Spectrometry in Biomarker Discovery

1.2.1. Overview of mass spectrometry as an analytical toolMass spectrometry can be applied to numerous analytical subject areas from the food industry

to forensic applications to complement various diagnostic and prognostic techniques (Griffiths

2008). Techniques using mass spectrometry have supported the advancement of proteomics,

metabolomics, lipidomics and atomic physics to forge a place within the clinical setting.

Proteomics is a platform for potential biomarker discovery and mass spectrometry is widely

used as a discovery tool (Anderson 2005). Advances in mass spectrometry instrumentation

have resulted in high mass accuracy and vast improvements in sensitivity and specificity (Al-

Shahib 2012). The forward momentum of mass spectrometry has led to the analysis of

diseased and non-diseased tissues supported by bioinformatics.

41

Recent experimental work by Rower et al (2011) showed development in the early diagnosis

of invasive ductal breast carcinoma through protein signatures and structural analysis. Such

progress in the identification of disease biomarkers could not only benefit patients directly but

support future drug design and development.

1.2.2. Basic principles of mass spectrometryThe mass spectrometer uses an ion source (liquid-phase/ solid-state) to transfer the analyte

into gas phase ions which then travel to the mass analyser to be sorted according to mass-to-

charge ratio (m/z), finally a detector accounts for the relative abundance of each ion at a

particular m/z (Hoffmann and Stroobant 2007). Instrumental calibration and use of analytical

standards is also vital to produce accurate mass measurements whilst using the many different

mass spectrometric techniques. External calibration is performed before sample analysis

where a known selected compound is used for spectral comparison against a reference list

where assignment of exact masses can then be made. Internal calibration describes a

calibration method where a reference compound is introduced into the mass spectrometer

during sample analysis. This technique compensates for any possible changes in conditions in

the analytical instrumentation (Hoffmann and Stroobant 2007).

Figure 1.14 is a representation of the basic principles of the mass spectrometer.

With regard to proteomics there are a variety of mass spectrometers available some of which

are described in more detail in later.

The following table (Table 6) adapted from Han et al (2008) details commonly used

instruments, applications and performance indicated by mass analyser type.

42

o<3-t->

■§G

a"Gdj+oo(Uccou

vjdj+Jdj

"Ocdj

£OCG50«0c

CDCO> >

03C03COCO03

\ooI—3OCO

A

t<D .£2

5 5ra ctf) ro

<U83Ovoc

.ocG

dJ,<oCQ.

CdjcoO.58u5?O

-O<u

o +j8 . $ w> 3** S 'O GE 8 o £:

H»» O dJ

.9- QU VO•5 5Q- G y cvo GO o>

cOQ

^ I T"1 -2*-1 "odj dJ C v j3 S .Oi s U. "G

ro

Inst

rum

ent

Mass

re

solu

tion

Mass

Se

nsiti

vity

m

/z ra

nge

Scan

rat

e Dy

nam

ic

rate

MS/

MS

Ion

sour

ce

Mai

n

O <u ■

3 CL

cl •*= Ol “

2u! .c S=i +-1 c| -c<U; 00■r.1 --j-

co'+->toJ = £<U 4-> CO OJQ. S

Mu-SI

£ E2 rev> re

•15

o

§ §° ?

o o o in (n

o .. o

o o

ooCM

i (Ui o

i§IN

OJ: ro— .2 o ore ai<§ s

o —.■K a> £S “S 5 ; «° ° £1 3E E a ic f

)

15

oo00

c r

6

c C M-

c O _g!

(U +Jo oj

E lOaio oo<u <Uo o

n . i -o a . o i a . “

i Q: _I, <

]5j on

*3'

oooo '

OHul.O

from

in-g

eldi

gest

ion

from

(1) T3 2

C C MW

D" CuO d) (D

X-SI u

(Uto*to

i c| 5

sz.SPQ. 4-J _>• S O sz

E c ro j -ao c tou 03 i o

<uCO0) C j E

“O i— oE 4-»Q. •MX

'554-»o 5

at+->oo <v k_ I- o s_M— CL E Q. cl ! I— Q.

•ao

Sio

a6

to 5 CD ^ *- r- r—Q. E Q_

00LU

15 oi o ^!0* 6 i o ot in rsl

O; EISi i1 u-

! <N

i o■! o i o I oifC

c sz 5 op o szT5 ^

Eod)

§•?I - CL

<E -S .§ 5 c E '-S. | - ^ o S' -

o o-i. o^ 6 o oLD rM

<3•2«QOQ)

■§C3Cp

o$ 5k k! o!?• QJ■c -J* •2» 8

■*! *<U O■Kl </)Q +-

J5C go Q.t> is

^ .QC3 C)5 -O o €£ § O O “Q CQJ O

^ I■5 c8 I

•2 & c 2o

2J cr?

6 ° O 4-

a «N g c o *c?

S C'O p■Jr* -C Qj o

I I5- 25 oE 2O Q)a c2 2 a $

so *ea gQ) .O

| I■fc £

a u 00

§ 2 I f i3 <£> ^ <u .t;23 i £

LD

1.2.3. Ionisation techniques and Mass analysersIn physical chemistry there are four main ways to produce gas phase ions from a molecule

interest which are; spray ionisation, desorption ionisation, chemical ionisation and electron

ionisation (Cooks 2010). Figure 1.15 shows these ionisation techniques schematically.

(A) (B)SampSe

Capilary

SprayingGas

Energy Beam

(C) (D) © v/

o r m —

M — * M H ++

H hvM|

XQ '

Figure 1.15 : Basic ionisation techniques. (Pol et al 2010). (A) spray ionisation, (B) desorption ionisation, (C) chemical ionisation and (D) electron ionisation.

Examples of spray ionisation are electrospray and nanospray, these techniques require the

sample to be in solution and regularly coupled to an LC system (Pol et al 2010). Secondary ion

mass spectrometry (SIMS), desorption electrospray ionisation (DESI) and matrix assisted laser

desorption ionisation (MALDI) are all methods using desorption ionisation from a solid surface

(Pol et al 2010). The desorption methods previously mentioned are used for image acquisition.

Chemical ionisation and electron ionisation use techniques involving gas-phase ion-molecule

reactions at atmospheric pressure, atmospheric pressure chemical ionisation (APCI) and

electron impact are such techniques. APCI can be applied to polar and relatively non-polar

molecular analysis with electron ionisation is useful for organic compounds.

Figure 1.16 displays current mass analysers at use in the analytical laboratory.

46

(A)Entrance

Slit QuadrupoleRods

| j

(B)

0„

ExitSlit

/3

— j ILinear

Detector

ReflectorDetector

\ Reflector

(Ion Mirror)

(D)

(E)

CentralElectrode

Electrode

EntranceEndcap Endcap

Electrode

(C) Magnet

Detector DetectorElectrode Electrode

TrappingPlates

Plates

SupercondutiveMagnet

Excitation

Figure 1 . 1 6 : Mass analysers used in mass spectrometers, (a) Quadrupole, (b) Time-of-flight (TOF), (c) Magnetic sector, (d) Ion trap, (e) Orbitrap, (f) Ion cyclotron resonance. (Pol et al 2010,. Trim et al 2012).

47

1.2.4. Proteomics using mass spectrom etry

Currently mass spectrometers allow interpretation of complex protein mixtures that contain

post translational modifications (PTMs) through fragmentation techniques (Hoffmann and

Stroobant 2007). Tandem mass spectrometry (MS/MS) describes the selection of a precursor

ion which then undergoes collision induced dissociation (CID) by an inert gas to generate a

product ion spectrum, b and y ions are generated due to shattering of the peptide C-N

backbone (Hoffmann and Stroobant 2007). To illustrate the stages of protein identification in

more detail the flowchart shown in Figure 1.17 adapted from Han et al (2008) segregates the

various methods into either bottom-up or top-down proteomic methods (Han et al 2008).

Proteins

(Shotgun approach) digestion

Protein separation Protein fractionation separation

Digestion ^Onlinemultidimensional separation (nD) LC- MS

Online (ID) LC-MS

(PMF) Target precursor

ion selection

Peptides

Peptide precursor mass measurement

Offline static MSPeptide mass measurement

Online LC-MS

Peptide identification

Protein ID from MS

Single protein or less complex mixture Peptide mixture

Data dependant MS/MS

Single protein or less complex mixture

Bottomup

Topdown

Data dependant MS/MS

Protein ID and characterisation from peptide MS/MS

Protein ID and characterisation from intact protein MSr

Protein precursor mass measurement

Figure 1 . 1 7 : Protein identification methods (Han et al 2008).

An article by Djidja et al (2009) regarding the profiling and imaging of glucose-regulated

protein 78 kDa (Grp78) contains an clear example of peptide identification from the peptide

mass fingerprint (PMF) of a tryptic digest, a bottom-up method. Here in-situ tryptic digestion

was performed directly onto tissue sections, the resulting PMF allowed selection of precursor

48

ion for MS/MS. Peptide sequences/ MS/MS spectra are then entered into algorithmic

bioinformatics programmes i.e. MASCOT that can result in protein identification through

searches performed against a protein database such as UniProt KB/Swiss-Prot.

Interpretation of comparative proteomics data in itself presents a huge challenge and

Hodgkinson et al (2011) published an antibody microarray proteomic study based on the

theory proposed by Petrak et al (2008) that such experiments possibly result in 'repeatedly

identified differentially expressed proteins' (RIDEPs). Petrak et al (2008) devised a list of 15

'RIDEPs' that were commonly observable in human and rodent samples. Amongst the RIDEPs

were heat-shock protein 27, peroxiredoxins, annexins and enolase 1 with proteins such as

zyxin and smad4 found by Hodgkinson et al (2011). The latter could provide a good place to

start in any shotgun proteomics experiment, especially assessment of a treatment response

time course and how these proteins appear to be differentially expressed.

Frequently MS/MS alone is not sufficient enough to provide a good clear fragmentation

spectrum from the parent ion. Especially when analysing the high abundance of ions

generated from an on tissue digest. The novel concept of ion mobility-mass spectrometry

which describes the further separation of isobaric ions indeed facilitates the task of precursor

ion selection. (McLean et al 2005). Ion mobility separation (IMS) can be applied throughout MS

acquisition and then again during MS/MS enabling a 'cleaner' peptide mass fingerprint with

improved subsequent precursor fragmentation. IMS will be discussed further later on.

To conclude this section, the vast amount of data currently now attainable in proteomics

experiments is remarkable. Dealing with the computational capacity required to process and

analyse these data will remain a challenge. However this is imperative to contribute to and

advance proteomic analysis.

1.3 Mass spectrometry instrumentation

1.3.1 LC-MSThe diagram in Figure 1.18 shows how the analytical technique of high performance liquid

chromatography (HPLC) can be coupled to a mass spectrometer a combination generally

referred to as LC-MS. The nature of this method requires the analyte to be in solution for

injection onto a column packed with coated silica particles (stationary phase) (University of

Bristol 2005). Varying the concentration of the mobile phase passing over the column will then

determine what is eluted in to the mass spectrometer for analysis.49

The sample can either be collected in fractions (Figure 1.18) or sent to the mass spectrometer

via an interface: Electrospray ionisation (ESI), atmospheric pressure chemical ionisation (APCI),

thermospray (TSP) and atmospheric pressure photoionisation (APPI) have been or are

currently used (Hoffmann and Stroobant 2007).

Column selector valveAutosampler

Mass spectrometerColumns□

UV detector

Fraction collector

Flow splitter

HPLC pumps

Figure 1 . 1 8 : Schematic representation o f a liquid chromatography mass spectrometer. (The LC/MS Unit 2009)

LC-MS/MS is regularly used in clinical chemistry for its universal capacity for the detection and

quantification of small molecules (Jong et al 2011). The technique of LC-MS/MS has proven to

be invaluable in the diagnosis of pheochromocytoma, a tumour that can develop in the centre

of the adrenal gland (Lagerstedt et al 2004). Renowned for its specificity and sensitivity LC-MS

has the ability to measure plasma free metabolites metanephrine and normetanephrine

occurring in nmol/L, known markers for pheochromocytoma.

50

1.3.2. Electrospray ionisation (ESI)Electrospray ionisation mass spectrometry (ESI-MS) is a highly sensitive soft ionisation

technique capable of analysing low volume, non-volatile samples (Ho et al 2003).

During flow through the sample (in solution) is nebulised under atmospheric pressure via an

electrical potential which has been applied to the needle (Gates 2004). Overtime as the

multiply charged droplet travels towards the sample cone solvent evaporation occurs resulting

in shrinkage. This is due to the 'Raleigh effect' which describes an event where the surface

tension can no longer retain the charge (Figure 1.19). The droplet containing the analyte is

then ripped apart by 'Coulombic explosion'.

Cone<counterelec?ro<fo)

Spfay needle tip mullply charged droplet

sowentx evaporaton

analytemolecule

•Coulonnbic''^ explosion + +

fie “Rayleigh' limit is reached

muHply charged droplet

analyts

Power Supply

University o f0 B B R IS T O L

Figure 1 .1 9 : Schematic o f Electrospray ionisation (Gates P 2004).

The multiply charged droplet is expelled from the needle, overtime solvent evaporation occurs

resulting in shrinkage of the droplet leading to "Coulombic explosion". There are two models

of ion generation in ESI, the ion evaporation model (IEM) and the charged residue model

(CRM) (Wand and Cole 2000). The IEM was first introduced in 1976 by Iribarne and Thomson

and due to an increasing electrical field present on the surface of the droplet, solvent-ion

clusters are discharged (Wand and Cole 2000). In the case of the CRM, continuous droplet

fission through continuous solvent evaporation results in the formation of gas-phase ions.

51

1.3.3. Matrix assisited laser desorption ionisation (MALDI)Matrix assisted laser desorption ionisation (MALDI) is a soft ionisation technique which is

widely used for analysis of non-volatile compounds in many applications including proteomics,

lipidomics, synthetic polymers and the study of small molecules.

The matrix plays a key role in the successful ionisation of the analyte. After deposition of the

matrix solution onto the sample, formation of matrix co-crystals occurs after drying/

evaporation which form a protective coat from the ablation of the laser (Hoffmann and

laser

analyte/matrix spo: be3IT

i:

analyteions

cation

sampe plate

matnxtons

//

extractiongrid

w To mass spectrometer

University o fQ B BRISTOL

\focussing

lensStroobant 2007). Under vacuum the laser is pulsed onto the matrix embedded sample and

although not completely understood, it is thought that the energy is transferred from the

surface of excitation of the matrix crystals to the analyte. The transfer of energy from the laser

to the matrix then matrix to the analyte is a two step process resulting in the ionisation of both

enabling the sample molecules to remain structurally intact. It is reported that gas phase ion

proton transfer is the most accepted explanation for MALDI matrix ionisation (Breuher et

ol1999). There are also other proposals for this mechanism these being; excited state proton

transfer, gas phase photoionisation, ion-molecule reactions and desorption of preformed ions

(Hoffmann and Stroobant 2007). Figure 1.20 aims to depict ionisation principles of MALDI.

Figure 1. 2 0 : Ionisation using MALDI. (Gates 2004). The protonated cloud o f ions is shown above the analyte/matrix co-crystallisation.

The formation of Cations with the analyte takes place in the following way;

[M+X]+ (X= H, Na, K etc.) (Gates 2004).

52

Selection of a MALDI matrix requires tailoring to the specific analyte chosen and wave length

of the laser (Hoffmann and Stroobant 2007).

The matrix needs to exhibit stability under vacuum, be soluble in selected solvent, form

adequate co-crystals with the analyte and absorb the laser wavelength (Sigma Aldrich 2001).

Examples of common matrices used and the associated applications are listed in Table 7.

The MALDI source consists of a pulsed laser (varying pulse widths) to produce packets of ions

to be introduced into the time of flight (ToF) analyser (Hoffmann and Stroobant 2007).

Nitrogen lasers were the initial choice for MALDI and are relatively inexpensive compared to

otheralternatives, although reports claim that irregularity of the beam can lead to 'hot spots'

typically due to the heterogeneous nature of the N2 laser. (Trim et ol 2010). However the short

lifespan of the N2 laser is an issue when selecting this laser type for MALDI imaging. Therefore

these have been largely superseded by the use of ND:YAG solid state lasers. These solid state

lasers are a common choice for MALDI-MSI due to their high repetition rate and extended

lifespan (Trim et ol 2012). It is stated that up to 1 kHz having been employed using ND:YAG

compared to 20 Hz on a standard N2 laser and ~ lx l0 9 shots compared to 2x107-6x107 shots

respectively.

Trim et al (2010) proposed 2 other alternatives these being; Nd:YLF (yttrium lithium fluoride)

and Nd:YV04 (yttrium ortho-vanadate). The latter paper continues to investigate the Nd:YV04

laser and details experiments to progress the appearance of the beam using mechanical

oscillation to reduce laser 'speckle'. Results with employment of beam enhancement displayed

improved spectra with increased intensity.

Proteins/ peptides a - C yano-4-hydroxycinnam ic acid (CHCA), 2,5-

D ihydroxybenzo ic acid (DHB), 3 ,5 -D im e thoxy-4 -

hyd roxycinnam ic acid (sinapic) (SA_

Lipids D ith rano l (DIT)

Inorganic molecules T rans-2 -(3 -(4 -te rt-B u ty lphe ny l)-2 m ethy l-2 -

p ro pe ny lied ene )m a lo no n itr ile (DCTB)

Organic molecules 2,5 -D ihydroxybenzoic acid (DHB)

53

Carbohydrates 2,5 -D ihydroxybenzoic acid (DHB), a - Cyano-4-

hydroxycinnam ic acid (CHCA),

T rihyd roxyace tophenone (THAP)

Oligonucleotides T rihyd roxyace tophenone (THAP), 3-

H ydroxyp ico lin ic acid (HPA)

Synthetic polymers 2,5-D ihydroxybenzoic acid (DHB), D ith rano l (DIT),

T rans-3-indoleacrylic acid (IAA)

Table 7 : Common matrices used in UV- MALDI.(Hoffmann and Stroobant 2007).

1.3.4. Quadrupole Time of flight mass spectrometry (QqToF)The term QqToF refers to a hybrid mass spectrometer which the employs a quadrupole, (an

arrangement of four metal rods with applied potentials; +(U+Vcos(wt)) and -(U+Vcos(cot))

where fixed U and alternating Vcos(cot) potentials can be applied to selected ions for MS/MS.

The quadrupole arrangement used in a typical QqTOF arrangement is shown in Figure 1.21,

where Q refers to the mass resolving quad and q describes a hexapole collision cell which can

further dissociate precursor ions which are then separated again by another quadrupole or

time of flight mass analyser. (Chernushevich et ol 2001). The diagram (Figure 1.21) shows Q0

which functions as a "collisional dampening" r.f. quad to focus ions prior entry into the Q1

mass filter and Q2 collision cell. In some cases instruments can feature Q0 (Figure 1.21) and Q2

both as hexapoles. The schematic diagram above also illustrates the location of the ToF mass

analyser and other components within the Q-ToF instrumentation.

54

Electrospray QO

Ion„ M odu la to r

-CL /■Kid** TorrId ni rorrj M C P

D e tec to r

Accelerating Column

TurboPump

Shield (“Liner" l

Figure 1. 2 1 : Diagrammatical representation o f a tandem QqTOF mass spectrometer. (Chernushevich et a12001).

If only mass spectra in ToF MS mode is required the Q1 quad operates in r.f. mode to simply

provide a means of transmission for the ions generated which then continue to the ToF mass

analyser. (Chernushevich et al 2001). Q2 parameters are maintained (<10 electron volts (eV))

as such to not induce fragmentation.

In MS/MS mode the precursor ion of interest is selected via the mass filtering Q1 which then

reaches the Q2 and undergoes collision induced dissociation (CID) with an inert gas i.e.

nitrogen/ argon. Subsequent acceleration then occurs of ion fragments and corresponding

parent ions to the ion modulator (see Figure 1.21) in the formation of a continuous beam

focused by ion optics. The pulsation of ions via an electric field at discreet bursts admits ions

into the ToF mass analyser (Figure 1.22) in a way governed by the ion’s characteristic trajectory.

55

60

C i alinfi —I r—I * u M n

IIII

l“ l In f r a c t io n - * pulse

Detector

Collision cell Q2

c = a 11!!5. .5t T i B H

i)

T roF

Figure 1. 22 : Schematic representation of a ToF modulator (Chernushevich et al 2001).

1.3.5. Synapt HDMS - travelling wave ion mobility separation.Briefly mentioned earlier in section 1.2.4. (Proteomics using mass spectrometry), is the

revolutionary concept of ion mobility separation (IMS) coupled with mass spectrometry. This

novel concept allows visualisation into another dimension of spectral data.IMS has structural

elucidation capabilities and the capacity to provide a further separation method of isobaric

ions. (Pringle et al 2007).

The theory of ion mobility describes the separation of bundles of ions under a static electric

field with the use of a 'background' gas (Pringle et al 2007,. Clemmer and Jarrold 1997,.

Valentine et al 1997). A novel idea which uses molecular separation by mass, shape (cross

sectional area) and chiral spin. Work reported in this thesiswas performed usingSYNAPT™

G1/G2HDMS systems (Waters Corp., Milford, MA) (Figure 1.23). A series of 3 travelling wave

stacked ring ion guides (Figure 1.24) constitute the ion mobility segment of the Synapt series

instruments where packets of ions are allowed to accumulate from previous IMS then released

to the ion guide for further mobility separation.

56

Ionizationsource

A <

Time-of-f light mass analyzer

Tri Wave High-fieldpusher

Ion detection system

r - s . . T-wave 1 ion guide

..........

Ion mobility separationQuadrupole. Ion

mirror

Transfer

mHelium cell

sDual-stage reflectron

Figure 1. 23 : SYNAPT™ G2 HDMS system (Waters Corp., Milford, MA). (Waters 2012).The Tri Wave ion mobility cell is shown where selected ions of interest are separated dependant on size, cross sectional area and chiral spin. The mass analyser is set to dual reflectron mode to aid the focusing of high energy ions to achieve high mass resolution.

In

Ion*

O ut

Figure 1. 24 : Stacked ring ion guide (Pringle et al 2007).

57

Travelling wave ion guides (TWIGs) which carry the ions through are the result of a temporary

dc voltage applied to the RF of opposite electrodes in constant succession throughout the

instrument. (Pringle et al 2007). The motion of the ions is likened to the "surfing o f ions on the

wove” . (Giles et al 2004). Figure 1.25 illustrates the travelling wave ion guide, the driving

force for the movement of ions.

K nit* Hectrode Pan

l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l l

Time

■ f f f f

r e

1

n *

1 's

■

D C Pulsc ,ons

Figure 1. 2 5 : Travelling wave ion guide- schematic representation. (Pringle et al 2007).

The principle of ion mobility separation is based on the collision cross-section (Q) as the ions

traverse through a field of inert gas, this can be expressed in the following equation. The

values relate to gas-phase collision cross-section (Clemmer and Jarrold 1997):

Q = -------16 N

Where N is the background gas number density, ze is the ionic charge, p the reduced mass of

the ion and the drift gas molecules, kb Boltzmann's constant, T gas temperature, A^is the

mobility to 273.2 K and 760 Torr.

58

1.3.6. MALDI-mass spectrometry imagingMatrix assisted laser desorption ionisation - mass spectrometry imaging (MALDI-MSI) is an

advanced analytical tool that allows molecular profiling and imaging of many classes of

compounds including; proteins, peptides, lipids, drugs and many other molecules directly from

tissue sections. The use of this technique for the study of biological tissue was first described

by Caprioli et al. in 1997 and it has been improved and adapted for use in many other studies

(Caprioli et al 1997., Chaurand et al 2006., Chaurand et al 2003., Stoeckli et al 2007,. Burrell et

al 2007,. Djidja et al 2008,. Trim et al 2008,. Fournier et al 2008).

MALDI-MSI allows the acquisition of multiple single mass spectra across the tissue section at a

spatial resolution predefined by the operator (typically 20-200 pm) (Figure 1.27). These mass

spectra are then combined together in order to generate molecular maps or images which

represent the distribution and the relative abundance and/or intensity of a specific ion signal

detected within the tissue section. MALDI-MSI has been shown to be a powerful technique for

direct protein analysis within tissue sections and in tumour tissue samples, it has been used for

discrimination between tumour and non tumour regions with no requirement for predefined

targets (Schwartz et al 2004,. Chaurand et al 2001,. Lemaire et al 2007,. Schwamborn et al

2007,. Lemaire et al 2007). A recent and exciting development in the technique is the use of

"on-tissue" tryptic digestion in order to achieve direct identification of proteins within a tissue

section (Lemaire et al 2007,. Shimma et al 2006,. Groseclose et al 2007).

Such molecular profiling and imaging could be described as a bottom-up shotgun approach to

protein identification, performed directly on cryo-sectioned tissue samples, rather than

through protein extraction methods. The chance to visualise the position of peptides within an

image generated by MALDI-MSI could indeed advance our knowledge in cancer research

studies. Also being able to observe co-localisation of peptides and possibly relate them to

disease states can provide complementary information regarding cancerous tissues.

A 2009 article by Djidja e ta l mentioned earlier, described the profiling and imaging of glucose-

regulated protein 78 kDa (BiP/Grp78) in pancreatic tumour sections. As described within the

article, the MALDI images were the first of their kind, with the discussion emphasising the low

abundance of this particular heat shock protein. However its importance is not to be

underestimated in the aggressiveness, progression and drug resistant nature of tumours. The

paper then goes on to describe how the combination of ion mobility and MALDI helped to

selectively target specific proteins for imaging. This technique has been called Matrix-Assisted

Laser Desorption-lonisation-lon Mobility Separation-Mass Spectrometry Imaging (MALDI IMS-

59

MSI) and its advantages for imaging of specific proteins has also been described by others

(McDonnell et al 2010).

deposition of matrix solution samples local peptide/protein content

formation of pcptide'proteins doped matrix crystats

Tissue section placed onto a conductive glass slide for combined imaging mass spectrometry and histological analysis.

discrete droplets spray

uniform coalingdiscrete arraymatrix

deposition »

mass analysis measures local peptides'protoin conten*Datacube of position

correlated spectraUV laser

MS analysis of discrete array

MS anatys>s of array within uniform coating

MS analysis of all positions

* V

X

Figure 1. 2 6 : MALDI-MSI workflow diagram (McDonnell et al 2010).

In addition to the matrix deposition shown above in Figure 1.26 displaying a discreet

arrangement of matrix/ reagent droplets, other methods include manual or automated

spraying to achieve a finer more homogenous coating for co-crystallisation.

Examples of commercially available sample preparation for enzyme and/or matrix deposition

are; Chemical Inkjet Printer by Shimadzu a sprayer using vibration vaporisation, Suncollect by

SunChrom an automated spraying system, Portrait™ 630 Multispotter by Labcyte which works

using high precision acoustic ejection from a reservoir source. (McDonnell et al 2010).

1.3.7. Quantitiative ProteomicsA number of methods have been proposed for quantitative protoemics.which include both

label and label free quantitation.

1.3.7.1 Label free quantitationImmunoassays and immunoblots are methods which are widely used, however development

and optimisation of protein specific antibodies can be considered both time expensive and

time consuming (Tate et al 2012). Cross reactivity can also be one of the main pitfalls of

immune targeted approaches. It can be said that the latter includes some of the reasoning

behind a shift towards mass spectrometry.

60

One of the most basic label free methods of quantitation in proteomics is spectral counting.

Shotgun proteomics (as shown in Figure 1.17) describes an experiment where MS/MS results

are summed for each peptide and comparison of summed events for each peptide can then be

analysed throughout a time course of results (Tate et ol 2012). More detailed analysis of

spectral counting can involve i.e. counting of unique peptides before and after normalisation.

Measurement of the area generated from peptide signal intensity in the MSI scan is another

quantitative label free approach (Tate et ol 2012). Quantitative information is acquired via an

extracted ion chromatogram. This value from the ion of interest is taken at a specific region in

time. This information is then used to associate the LC elution peak with the selected mass of

interest. Processing of these data involves the alignment of MSI chromatographic peaks

throughout the samples/ time course samples analysed, the areas are then profiled.

Figure 1.27 shows an example using Progenesis, one of the proteomic software packages that

can be used to analyse area measurements as previously described.

Control Treated

Figure 1. 27 :The concept of area peptide measurement coupled with a specific time domain,visualisedthrough Progenesis LC~MS software. (http://www.nonlinear.com/products/progenesis/lc-ms/overview/)

61

1.3.7.2 Labelling fo r proteomic quantitationIsobaric Tags for Relative and Absolute Quantitation (iTRAQ) have been applied to numerous

mass spectrometry studies (Song et a/2008,. Ross et ol 2004,. Gygi et ol 1999). iTRAQ enables

samples to be labelled and then combined to perform 4 plex / 8 plex experiments to analyse

tryptic digests from cells/ tissue (Shilov et al 2007). Each isobaric tag is comprised of a reporter

group, balance group and a peptide reactive group. This proteomic bottom up approach

requires tandem MS to give relative absolute quantitation information using reporter groups

from 113 - 121. A control is included as a reference in the combined labelled samples and

quantitation ratios are then relative to the control. MS/MS spectra can then be submitted to

protein search software which can result in identification of the labelled peptides.

An example of an iTRAQ workflow is shown in Figure 1.28.

Tw\*trr**r* K TfM$r*fw8 TwMfrwtC Tf*»cnwit D

Protein extraction followed by enzymatic digestion

Combine labeled digests

Analysis by LC-MS/MS

iTRAQ label rr R A 0114 (TRAQ11S I ITRAQ 116 HRA0117

QUANTIFICATION FROM ITRAQ

REPORTER ONS

1PEPTIDE SEQUENCE IDENTIFICATION FROM

PEPTIDE BACKBONE FRAGMENT IONS __________________A__________________

m/z

Figure 1. 28 : Simple workflow using Isobaric Tags fo r Relative and Absolute Quantitation (iTRAQ). ( Broad Institute 2013 ).Relative quantitation can be measured by observation o f the isobaric MS/MS fragments in the region o f the spectra that correspond to a particular labelled specimen.

62

1.3.8. Aims of the studyThe aims of the study reported in this thesis are to develop and utilise mass spectrometry

imaging techniques (MALDI-MSI), in combination with conventional proteomic methodologies,

to investigate protein induction in vascular-targeted strategies.

Here MALDI-MSI intended to compliment conventional proteomics to provide information

relating the spatial distribution of protein/ peptides throughout time course experiments. Both

CA-4-P susceptible (fibrosarcoma 120) and resistant (fibrosarcoma 188) tumour models were

used to perform profiling and imaging experiments.

The tumour models used within this study originate from a mouse embryonic fibrosarcoma cell

line and had been genetically engineered to either express solely the vascular endothelial

growth factor (VEGF) 120 or 188 isoforms. Vasculature derived from the VEGF 120 growth

factor characteristically favours cell proliferation but produces an erratic disorganised network

of blood vessels with 'blind ends' lending itself to haemorrhagic disruption by CA-4-P. Tumour

blood capillaries in the fibrosarcoma 188 model are less substandard and as a consequence

mimic a more normal vascular arrangement (Tozer et al 2008). Tozer et al (2008) found that

VEGF 188 tumour vasculature had increased pericyte support hence adding to CA-4-P

resistance.

Nine different VEGF isoforms have currently been discovered in humans, with VEGF189,

VEGF165 and VEGF121 isoforms being the most documented and VEGF120, VEGF164 and

VEGF188 in mice ( Tozer et al 2008,. Harris et al 2012). VEGF mRNA is transcribed from eight

exons and splice variants result in the various isoforms mentioned previously. Each isoform is

known to have its own distinctive vascular modelling and angiogenic configuring capability.

Each VEGF isoform is thought to govern angiogenesis in relation to whether it is matrix bound

or not and yet little is known with regards to VEGF biological / mechanistic function (Harris et

al 2012). VEGF120 does not possess a heparin-binding site and is described as being 'readily

diffusible'. VEGF188 on the other hand is bound to proteoglycans, components of the

extracellular matrix or at the cell surface (Figure 1.29) (Tozer et al 2008).

In tumours VEGF is associated with tumour growth and metastasis, increased levels of VEGF is

said to correlate with poor patient survival.

63

VEGF-188

VEGF-164

DiffusibleVEGF-120

Exons

Matrix-bound

Heparin binding site

Neuropiiin-binding site

Dimerization and VEGF-R binding

sites

Bstd control 188 164 120 std control 188 164 120

500400300

200200

100 100Cell lines Tumor extracts

Figure 1.29: (A) The exon structure of the mouse VEGF isoforms showing the differences / similarities between exons and nature of each isoform whether diffusible or matrix bound. (B) The reverse transcription-PCR of cDNA showing the bands representative of each VEGF isoform. Traces of bands relating to other isoforms are present in the control but still align with the specific isoform bands (Tozer et al 2008).

PMFs were acquired from in-situ tissue tryptic digestion and proteins thought to be involved in

tumourigenesis and drug treatment resistance were observed and targets which could be

linked with hypoxic conditions of the tumour microenvironment.

Multivariate Statistical analysis aimed to observe grouping and classification within data time

courses post CA-4-P administration.

The techniques used; MALDI-MSI, ESI-LC-MS/MS, LC-MALDI-MS/MS with iTRAQ labelling

intended to provide cross validation of the effects post administration of vascular disrupting

agent CA-4-P.

MALDI imaging of proteins related to tumour biology and the stress response could forge a

place in the clinical setting providing alternative tool in the field of cancer research.

64

ReferencesAbdollahi A, Folkman J (2010) Evading tumour evasion: Current concepts and perspectives of

anti-angiogenic cancer therapy. Drug Resistance Updates. 13.16-28

Al-Shahib A, Misra R, Ahmod N, Fang M, Shah H, Gharbia S (2010) Coherent pipeline for

Biomarker discovery usingmass spectrometry and bioinformatics. BMC Bioinformatics.il. (437)

1-9

Anderson NL (2005) The Roles of Multiple Proteomic Platforms in a Pipeline fo r New

Diagnostics. Molecular and Cellular Proteomics.4.1441-1444

Bay less K and Davies GE (2004) Microtubule depolymerisation rapidly collapses capillary tube

networks in Vitro and angiogenic vessels in Vivo through small GTpase Rho. The Journal of

Biological Chemistry.279. (12) 11686-11695

Breuher K, Knochenmuss R, Zenobi R (1999) Proton transfer reactions of matrix assisted laser

desorption/ionisation matrix monomers and dimmers. Journal of American Society of Mass

Spectrometry.10.1111-1123

Broad insitute (2013) iTRAQ. Available at URL: http://www.broadinstitute.org/. Last accessed

29/07/13.

Buchanan CM, Shih JH, Astin JW, Rewcastle GW, Flanagan JU, Crosier PS, Shepherd PR (2012)

DMXAA (Vadimezan, ASA404) is a multi-kinase inhibitor targeting VEGFR2 in

particular. 122. (10) :449-57.

Burrell M M, Earnshaw G , M. Clench MR (2007)lmaging Matrix Assisted Laser Desorption

Ionization Mass Spectrometry: a technique to map plant metabolites within tissues at high

spatial resolution Journal of Experimental Botany 58. (4) 757-763.

Cancer Research UK (2007) Demographic facts for the world population.Cancer Stats

Worldwide.Available at URL: http://info.cancerresearchuk.org/. Last accessed 05/07/10

Cancer Research UK (2009) How many types of cancer are there? Available at URL: h ttp://

info.cancerresearchuk.org/. Last accessed 05/07/10

Caprioli RM, Farmer TB, Gile J (1997) Molecular Imaging of Biological Samples: Localization of

Peptides and Proteins Using MALDI-TOF MS.Analytical Chemistry.69. (23) 4751-4760.

66

Chaurand P, DaGue BB, Pearsall RS, Threadgill DW, Caprioli RM(2001) Profiling proteins from

azoxymethane-induced colon tumors at the molecular level by matrix-assisted laser

desorption/ionization mass spectrometry.Proteomics.l. (10) 1320-1326.

Chaurand P, Cornett DS, Caprioli RM (2006)Molecu/ar imaging of thin mammalian tissue

sections by mass spectrometry.Current Opinion in Biotechnology.17. (4) 431-436.

Chaurand P, Fouchecourt S, DaGue BB. Xu BJ, Reyzer ML, Orgebin-Crist M, CaprioliRM (2003)

Profiling and imaging proteins in the mouse epididymis by imaging mass spectrometry.

Proteomics 3.(11)2221-2239.

Chemotherapy (2011) Types of cancer.Available at URL: http://www.chemotherapy.com/. Last

accessed 21/ 11/12

Chernushevich IV, Loboda AV, Thomson BA (2001) An introduction to quadrupole-time-of-flight

mass spectrometry.Journal of Mass Spectrometry.36.849-865

Clemmer DE and Jarrold MF (1997) Ion mobility measurements and their applications to

clusters and biomolecules.Journal of Mass Spectrometry.32. 577-592

Connell PP, Heilman S (2009) Advances in radiotherapy and implications fo r the next century: A

historicalperspective.Cancer Research.69. (2) 383-392

Cooks RG (2010) Foreword: desorption ionization and spray ionization: connections and

progress. In Cole RB (ed) Electrospray and MALDI mass spectrometry: fundamentals,

instrumentation, practicalities, and biological applications. Wiley. Hoboken.

Delaney G, Jacob S, Featherstone C, Barton M (2005) The role of radiotherapy in cancer

treatment. Cancer.104. (6) 1129-1137

Denekamp J (1990) Vascular attack as a therapeutic strategy fo r cancer. Cancer Metastasis Rev.

9. (3) 267-282

DeVita VT and Chu E (2008) A history of cancer chemotherapy. Cancer Research.68 (21) 8643-

8653

Djidja M, Carolan V, Loadman PM, Clench MR (2008). Method development fo r protein

profiling in biological tissues by matrix-assisted laser desorption/ionisation mass spectrometry

imaging.Rapid communications in mass spectrometry, 22. (10) 1615-1618.

Djidja M-C, Claude E, Snel MF, Scriven P, Francese S, Carolan V, Clench MR (2009) Maldi-ion

mobility separation-mass spectrometry imaging of glucose-regulated protein 78 kDa (Grp78) in

67

human formalin fixed paraffin-embedded pancreatic adenocarcinoma tissue sections.iourna\ of

Proteome Research. 8. (10) 4876-4884

Folkman J (1971) Tumor Angiogenesis: Therapeutic Implications. The New England Journal of

Medicine. 285.1182-1186

Fournier I, Wisztorski M, Salzet M (2008) Tissue imaging using MALDI-MS: a new frontier of

histopathology proteom/cs.Expert Review of Proteomics.5. (3) 413-424.

Gates P (2004) Mass spectrometry resource: Electrospray ionisation (ESI). University of Bristol.

Available at URL: http:// http://www.chm.bris.ac.uk/. Last accessed 28/11/12

Giles K, Pringle SD, Worthington KR, Little D, Wildgoose JL, Bateman RH (2004) Applications of

travelling wave-based radio-frequency-only stacked ring ion guide.Rapid Communications in

Mass Spectrometry.18. 2401-2414

Griffiths J (2008) A brief history of mass spectrometry .Analytical Chemistry .80. 5678-5683

Groseclose MR, Andersson M, Hardesty WM,Caprioli RM (2007)ldentification of proteins

directly from tissue: in situ tryptic digestions coupled with imaging mass spectrometry. Journal

of Mass Spectrometry.42. (2) 254-262.

Gygi PS, Rist B, Gerber SA, Turecek F, Gelb MH, Aebersold R (1999) Quantitative analysis of

complex protein mixturesusing isotope-coded affinity tags. Nature Biotechnology.17. (10) 994-

999

Han X, Aslanian A, Yates JR (2008) Mass spectrometry for proteomics. Current Opinion in

Chemical Biology.12.483-490

Hanahan D and Weinberg RA (2000) The hallmarks of cancer. Cell.100. 57-70

Hanahan D and Weinberg RA (2011) Hallmarks of Cancer: The Next Generation. Cell.144. 646-

674

Harris S, Craze M, Newton J, Fisher M, Shima DT, Tozer GM, Kanthou C (2012) Do Anti-

Angiogenic VEGF (VEGFxxxb) Isoforms Exist? A Cautionary Tale.PLoS ONE 7. (5) 1-14

Hinnen P, Eskens FALM (2007) Vascular disrupting agents in clinical development. British

Journal of Cancer.96.1159-1165

68

Ho CS, Lam CWK, Chan MHM, Cheung RCK, Law LK, Lit LCW, Ng KF, Suen MW M, Tai HL

(2003) Electrospray Ionisation Mass Spectrometry: Principles and Clinical Applications. The

Clinical Biochemist Reviews.24.(l) 3-12

Hodgkinson VC, ELFadl D, Drew PJ, Lind MJ, Cawkwell L (2011) Repeatedly identified

differentially expressed proteins (RIDEPs) from antibody microarray proteomic analysis.

Journal of Proteomics. 74. 698-703

Hoffmann E de and Stroobant (2007) Mass spectrometry principles and applications.3rd Edition.

Chichester: John Wiley & Sons Ltd

Hurwitz H, Fehrenbacher L, Novotny W, Cartwright T, Hainsworth J, Heim W, Berlin J, Baron A,

Griffing S, Holmgren E, Ferrara N, Fyfe G, Rogers B, Ross R, F Kabbinavar (2004J Bevacizumab

plus irinotecan, fluorouracil and leucovorin fo r metastatic colorectal cancer. New England

Journal of Medicine.350.(23) 2335-2342

Jong W HA de, Vries E, Kema IP (2011) Current status and future developments of LC-MS/MS in

clinical chemistry for the quantification of biogenic amines. Clinical Biochemistry .44. (1) 95-103

Kanthou and Tozer (2007) Selective destruction of the tumour vasculature by targeting the

endothelial cytoskeieton.Drug discovery today: Therapeutic strategies. 4. (4) 237-243

Kanthou C, Tozer GM (2002) The tumor vascular targeting agent combretastatin A -4-

phosphate inducesreorganization of the actin cytoskeieton and early membraneblebbing in

human endothelial cells. BLOOD.99. (6) 2060-2069

Koch U, Krause M, Baumann M (2010) Cancer stem cells at the crossroads of current cancer

therapy failures - Radiation oncology perspective. Seminars In Cancer Biology. 20. (2) 116-124

Lagerstedt SA, O'Kane DJ, Singh RJ (2004) Measurement of plasma free metanephrine and

normetanephrine by liquid chromatography-tandem mass spectrometry fo r diagnosis of

pheochromocytoma.Clinical Chemistry.50. 603-611

Lemaire R, Desmons A, Tabet JC, Day R, Salzet M, Fournier I (2007) Direct analysis and

MALDI imaging of formalin-fixed, paraffin-embedded tissue sections. Journal of Proteome

Research 6. (4) 1295-1305.

Lemaire R, Menguellet SA, StauberJ, MarchaudonV, Lucot J, Collinet P, Farine M, Vinatier D,

Day R, Ducoroy P, Salzet M, Fournier I (2007) Specific MALDI Imaging and Profiling for

69

Biomarker Hunting and Validation: Fragment of the I IS Proteasome Activator Complex, Reg

Alpha Fragment, Is a New Potential Ovary Cancer Biomarker. Journal of Proteome Research. 6.

(11) 4127-4134.

Lord G and Ashworth A (2010) Biology-driven cancer drug development: back to the

future.BMC Biology.8. (38) 1-12

Medline Plus (2010) Tumour.Available at URL: http://www.nlm.nih.gov/medlineplus/ . Last

accessed 21/ 11/12

McLean JA, Ruotolo BT, Gilling KJ, Russell DH (2005) Ion mobility-mass spectrometry: a new

paradigm for proteomics. International Journal of mass Spectrometry. 240. 301-315

McDonnell LA, Corthals GL, Willems SM, Remoortere van A, Zeijl van RJM, Deelder AM (2010)

Peptide and protein imaging mass spectrometry in cancer research. Journal of Proteomics. 73.

(10) 1921-1944.

National Cancer Institute (2009) Understaing Cancer series - Loss of Normal Growth

Control.Available at URL: http://www.cancer.gov/. Last accessed on 21/11/12

Oncolink (2010) Cancer treatment information.Available at URL: http://www.oncolink.org/.

Last accessed 21/11/12

Oxigene (2010) Abnormal vasculature.Available at URL: URL http://www.oxigene.com/Last

accessed 21/09/10

Papac RJ (2001) Origins of cancer therapy. Yale Journal of Biology and Medicine. 74. (6): 391 -

398.

Petrk J, Ivanek R, Toman O, Cmejla R, Cmejlova J, Vyoral D, Zivny J, Vulpe CD (2008) Deja vu in

proteomics. A hit parade of repeatedly identified differentially expressed proteins. Proteomics.

8.1744-1749

Pringle SD, Giles K, Wildgoose JL, Williams JP, Slade SE, Thalassinos K, Bateman RH, Bowers MT,

Scrivens JH (2007) An investigation of the mobility separation of some peptide and protein ions

using a new hybrid quadrupole/travelling wave /MS/oo-ToF/nsfrumenf.International Journal of

Mass Spectrometry.261.1-12

Ross PL, Huang YN, Marchese JN, Williamson B, Parker K, Hattan S, Khainovski N, Pillai S,

Dey S, Daniels S , Purkayastha S, Juhasz P, Martin S, Bartlet-Jones M, He F, Jacobson A, Pappin

70

D J (2004) Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive

isobaric tagging reagents. Molecular and Cellular Proteomics.3. (12) 1154-1169

Rower C, Koy C, Hecker M, Reimer T, Gerber B, Thiesen H-J, Glocker MO (2011) Mass

Spectrometric Characterization of Protein Structure Details Refines the Proteome Signature for

Invasive Ductal Breast Carcinoma. Journal of American Society for Mass Spectrometry.22.440 -

456

Saharinen P, Eklund L, Pulkki K, Bono P, Alitalo K (2011) VEGF and angiopoietin signaling in

tumor angiogenesis and metastasis. Trends in Molecular Medicine.17. (7) 347-362

Schwamborn K, Krieg RC, Reska M, Jakse G, Knuechel R, Wellmann A (2007) Identifying

prostate carcinoma by MALDI-lmaging. International Journal of Molecular Medicine. 20. (2)

155-159.

Schwartz SA, Weil RJ, Johnson MD, Toms SA, Caprioli RM (2004) Protein profiling in brain

tumors using mass spectrometry: feasibility of a new technique for the analysis of protein

expression. Clinical Cancer Research: an official journal of the American Association for Cancer

Research.10. (3) 981-987.

Shariq J, Barton M, Noble D, Lemer C, Donaldson U (2009) An international review of patient

safety measures in radiotherapy practice.Radiotherapy and Oncology.92.15-21

Shilov IV, Seymour SL, Patel AA, Loboda A, Tang WH, Keating SP, Hunter CL, Nuwaysir LM,

Schaeffer DA (2007) The Paragon Algorithm: A next generation search engine that uses

sequence temperature values and feature probabilities to identify peptides from tandem mass

spectra.Molecular and Cellular Proteomics.6. (9) 1638-1655.

Shimma S, Furuta M, Ichimura K, Yoshida Y, Setou M, (2006) A Novel Approach to in situ

Proteome Analysis Using Chemical Inkjet Printing Technology and MALDI-QIT-TOF Tandem

Mass Spectrometer. Journal of Mass Spectrometry Society Japan. 54. (4) 133-140.

Siemann DW, Bibby MC, Dark GG, Dicker AP, Eskens F A.L.M, Horsman MR, Marme D, LoRusso

PM (2005) Differentiation and definition of vascular-targeted therapies. Clinical Cancer

Research.416. (11) 416-420

Siemann DW (2010) The unique characteristics of tumour vasculature and preclinical evidence

for its selective disruption by tumour-vascular disrupting agents. Cancer Treatment Reviews.37.

(1) 63-74

71

Sigma Aldrich (2001) MALDI mass-spectrometry.Available at URL: http://sigmaaldrich.com/

Last accessed 27/09/10

Song X, Bandow J, Sherman J, Baker JD, Brown PW, McDowell MT, Molloy MP (2008) iTRAQ

Experimental Design for Plasma Biomarker Discovery. Journal of Proteome Research.7. 2952-

2958

Stoeckli M, Staab D, Schweitzer A, Gardiner J, Seebach D (2007) Imaging of a 6-Peptide

Distribution in Whole-Body Mice Sections by MALDI MassSpectrometry. Journal ofAmerican

Society of Mass Spectrometry 18. (11) 1921-1924.

Tate S, Larsen B, Bonner R, Gingras A-C (2012) Label-free quantitative proteomics trends for

protein-protein interactions. Journal of Proteomics. ARTICLE IN PRESS

The LC/MS Unit (2009) The LC instrument details. Available at URL:

http://home.iprimus.com.au/ Last accessed 24/09/10

Thorpe PE, Chaplin DJ, Blakely DC (2003) The first international conference on vascular

targeting: Meeting overview. Cancer Research.63.1144-1147

Tozer GM, Akerman S, Cross NA, Barber PR, Bjo'rndahl MA, Greco O, Harris S, Hill SA,

Honess DJ, Ireson CR, Pettyjohn KL, Prise VE, Reyes-Aldasoro CC, Ruhrberg C, Shima DT,

Kanthou C (2008) Blood Vessel Maturation and Response to Vascular-Disrupting Therapy in

Single Vascular Endothelial Growth Factor-A Isoform-Producing Tumors. Cancer Research. 68.

(7) 2301-2311

Tozer GM, Kanthou C, Lewis G, Prise VE, Vojnovic B, Hill A (2008) Tumour vascular disrupting

agents: combating treatment resistance. The British Journal of Radiology.81. S12-S20

Trim P J, Atkinson S J, Princivalle AP, Marshall PS, West A and Clench MR (2008) Matrix-

assisted laser desorption/ionisation mass spectrometry imaging of lipids in rat brain tissue with

integrated unsupervised and supervised multivariant statistical analysis.Rapid Communications

in Mass Spectrometry.22. (10), 1503-1509.

Trim PJ, Djidja M-C, Atkinson SJ, Oakes K, Cole LM, Anderson DMG, Hart PJ, Francese S, Clench

MR (2010) Introduction of a 20kHz Nd:YV04 laser into a hybrid quadrupole time-of-flight mass

spectrometer for MALDI-MS imaging. Analytical and Bioanlytical Chemistry.397. (8) 3409-3419

Trim PJ, Djidja M-C, Muharib T, Cole LM, Bryn Flinders, Carolan VA, Francese S. Clench MR

(2012J Instrumentation and software for mass spectrometry imaging—Making the most of

what you've got. Journal of Proteomics.75. (16) 4931-4940

72

University of Bristol (2005) High performance liquid chromatography mass spectrometry. URL

http://www.bris.ac.uk/ Last accessed 24/09/10

Valentine SJ, Counterman AE, Clemmer DE (1997) Conformer-dependant proton-transfer

reactions ofubiquitin ions.Journal of the American Society for Mass Spectrometry.8.954-961

Wang G and RB Cole (2000) Charged residue versus ion evaporation fo r formation of alkali

metalhalide cluster ions in ESI. Analytica Chimica Acta.406. 53-65

Waters (2012) SYNAPT G2-S HDMS. Available at URL: http://www.waters.com/Last accessed

on 29/11/12

Wedge SR, Ogilvie DJ, Dukes M, Kendrew J, Chester R, Jackson JA, Boffey SJ, Valentine PJ,

Curwen JO, Musgrove HL, Graham GA, Hughes GD, Thomas AP, Stokes ESE, Curry B,

Richmond GHP, Wadsworth PF, Bigley AL, Hennequin LF (2002) ZD6474 inhibits vascular

endothelial growth factor signalling, angiogenesis, and tumour growth following oral

administration.Cancer Research.62.4645-55

Worldometers (2013) Worldometers - realtime world statistics. Available at URL:

http://www.worldometers.info/. Last accessed on 27/01/13

World cancer Research Fund (2010) World cancer statistics. Available at URL:

h ttp://www. wcrf-uk. org/. La st a ccesse d o n 21/11/12

World Health organisation (2012) Cancer . Available at URL: http://www.who.int/. Last

accessed on 21/ 11/12

Zhi-chao SI, Jie L (2008) What 'helps' tumours evade vascular targeting treatment?Chinese

Medical Journal.121. (9) 844-849

73

Chapter 2Protein induction in a CA-4-P

susceptible tumour model usingMALDI-MSI

74

2.1 IntroductionThe discovery of biomarkers that can be used to determine success of cancer therapy or

resistance to it, at an early stage, presents a very complex problem. There are countless

biochemical pathways to be considered and the multifaceted characteristics of the cancer cell

have also to be taken into account; a cell that can largely evade the host's immune responses,

has mastered tissue invasion and anti-cancer drug resistance.

One promising class of anti-cancer drug currently under developmentis the vascular disrupting

agents (VDAs) (Kanthou and Tozer 2007). VDAs cause rapid, selective and sustained shutdown

of tumour blood flow. This produces drastic effects on the tumour microenvironment. Certain

aspects of blood flow modification are due to the direct effects of the drug on the endothelial

cells that line the luminal surface of blood vessels.The article by Kanthou and Tozer describes a

proposed mechanism of action for CA-4-P (Kanthou and Tozer 2007). It is suggested that

increased permeability after the administration of CA-4-P is one major factor in the disruption

of endothelial organisation. Interference in vascular integrity due to abnormality in the

cytoskeleton, subsequently leads to stress fibre formation and endothelial cells become more

rounded ('blebbing'). Haemorrhagic necrosis follows exacerbating the already hypoxic regions

within the tumour microenvironment.

As mentioned in the introductory chapter, the remodelling and formation of actin stress fibres

is clearly demonstrated using immunofluorescence by Siemann (Siemann 2010). The images

contained in this article are good examples of the "rounding" of endothelial cells and

"condensation" of the microtubules post CA-4-P treatment.

Drug delivery of VDAs is relatively simple due to endothelial positioning near the bloodstream.

The diploid target cells are thought not to be directly drug-resistant due to low risk of

mutations (Thorpe et al 2003).

VDAs have relatively good clinical measurability, i.e. bloodflow and in most cases short-term

exposure to a VDA creates sufficient hypoxic conditions to induce necrosis. However, the use

of VDAs to increase hypoxia can also have undesirable consequences (Zhi-chao and Jie 2008),

(Tozer et al (2008) . Hypoxia-inducible factors (HIF), which are prevalent under hypoxic

conditions, cause up-regulation of pro-angiogenic factors such as VEGF. VDA smight amplify

the angiogenic process after the creation of an increased hypoxic tumour environment.

Hypothetically speaking this could even help the metastatic process ensuring tumour invasion

(Zhi-chao and Jie 2008).

75

In addition to acquired resistance to treatment, the heterogeneous tumour mass could well

contain a cell population oblivious to vascular targeted strategies, i.e. innate resistance

(Thorpe et al 2003). Support for this comes from the hypothesis that there are four possible

micro-vascular phenotypes in existence i.e. normal pre-existing, tumourphenotypic vessels,

normal neo-vasculature and abnormal (pathological) neo-vasculature. It may explain why

some tumour regions remain unaffected by VDA treatment, including the well documented

viable tumour rim (Abdollahi and Folkman 2010). The viable tumour rim is also thought to

achieve evasion through its positioning, adjacent to the non-diseased tissue, which is well

supplied with oxygen andessential nutrients (Kanthou and Tozer 2009). Resistance

mechanisms to VDAs are clearly complex and multifactorial but increased knowledge in this

area could provide strategiesfor future combination therapies.

The largest group of VDAs are tubulin binding, microtubule-depolymerising drugs such as

combretastatin A-4-3-0-phosphate (CA-4-P/Fosbretabulin/ZybrestatTM), which is currently in

late stage clinical trials. Combination of CA-4-P or a related compound, Oxi4503 (CA-l-P), with

anti-angiogenic therapysuch as VEGF receptor tyrosine kinase inhibitors has already shown

promise and novel concepts aimed at targeting circulating angiogenic endothelial progenitor

cells have also been proposed. This area has been recently reviewed by Siemann (Siemann and

Horsman 2009).

The experimental work reported here, describes a study, by MALDI MSI and MALDI IMS-MSI,

of the proteins induced in a mouse transplanted fibrosarcoma model (VEGF120 tumours), at a

number of time points, following treatment with CA-4-P. Imaging of peptide signals after in

situ tissue tryptic digestion was carried out using MALDI-MS and MALDI IMS-MSI to observe

their spatial distribution. MALDI-IMS-MS/MS of the peptide signals was performed to identify

proteins involved in the pharmacodynamic responses to treatment.

2.2 Materials and Samples

2.2.1 Chemicals and Materialsa-Cyano-4-hydroxycinnamic acid (CHCA), aniline (ANI), ethanol (EtOH), chloroform (CHCI3),

acetonitrile (ACN), octyl-a/b-glucoside (OcGIc), tri-fluoroacetic acid (TFA), ammonium

bicarbonate, haematoxylin, eosin, xylene and DPX mountant were from Sigma-Aldrich (Dorset,

76

UK). Modified sequence grade trypsin (20 pg lyophilised) was obtained from Promega

(Southampton, UK).

2.2.2 Tissue samplesSevere combined immunodeficiency (SCID) mice were injected sub-cutaneously in the flank

with a 50 pi tumour cell suspension containing 1 x 106 cells in serum-free medium.The cells

employed in this study were from the mouse fibrosarcoma cell line, VEGF120. This has been

engineered to express only the VEGF 120 isoform. Tumours were allowed to grow to

approximately 500 mm3, before CA-4-P treatment (a single dose of 100 mg/kg i.p). The dosage

for CA-4-P are consistent with animal studies performed at clinically relevant doses (Galbraith

2003., Prise 2002). Treatment was either Saline/ Control, Mice were killed and tumours excised

at various times after treatment. In the case of Oh CA-4-P, the tumour was excised

immediately (within 2mins) after dispatchment. All animal work carried out documented

herein was performed by Dr. J. E. Bluff, Tumour Microcirculation Group, University of Sheffield,

UK. Samples were provided as frozen excised tumours and stored at -80°C.

2.2.3. Experimental groupsControls (no treatment, saline i.p), n = 6 (labelled tu m o u rl_ l - tumour 1_6), C-A4-P (0 h after

treatment), n = 6, (labelled tumour2_l - tumour 2_6), C-A4-P (0.5 h after treatment), n =

6,(labelled tumour 3_1 - tumour 3_6), C-A4-P (6 h after treatment),n = 6, (labelled tumour 4_1

- tumour 4_6), C-A4-P (24 h after treatment)n = 6, (labelled tumour 5_1 - tumour 5_6).

2.2.4 Tissue preparationFrozen tissue sections were cut to give approximately 10pm sections, using a Leica CM3050

cryostat (Leica Microsystems, MiltonKeynes, UK) (Djidja et al 2009). The sections were then

freeze thaw mounted on poly-lysine glass slides. Mounted slides were either used

immediately or stored in an airtight tube at -80 °C for subsequent use.

2.2.5 In s itu tissue digestion and trypsin depositionThe triplicate tissue samples to undergo MALDI-MS/MSI were washed initially with 70% and

then 90% ethanol for 1 min then left to dry, subsequently slides were immersed in chloroform

for 10 s. Prior to matrix application,

in situ tissue digestion was performed with trypsin solution prepared (from lyophilised trypsin)

at 20 pg/ml by addition of 50 mM ammonium bicarbonate (NH4HC03) pH 8, containing 0.5%

octyl-a/b-glucoside (OcGIc) based on a protocol by Djidja et al (2009).77

Two automated systems were used for trypsin application. The "Suncollect" (SunChrom,

Friedrichsdorf, Germany) automatic pneumatic sprayer was used to spray trypsin in a series of

layers. The Portrait™ 630 Multispotter (Labcyte, Sunnyville CA) was used to apply trypsin in a

200-300 pm array of spots. The sections for MALDI-MS and MALDIMSI were incubated

overnight in a humidity chamber containing H20 50%: methanol 50% overnight at 37 °C and

5% C02.

2.3Methods and instrumentation

2.3.1 Matrix depositionThe matrix, a-cyano-4-hydroxycinnamic acid (CHCA) and aniline in acetonitri!e:water:TFA

(1:1:0.1 by volume), was applied using either the Suncollect (at 5 mg/ml) or the Portrait™ 630

Multispotter (at 10 mg/ml), as described above. Identical coordinate settings were used as

with the trypsin deposition, to ensure sample uniformity. Equimolar amounts of aniline were

added to the CHCA solution, i.e. 1 mlof 5 mg/ml CHCA solution contained 2.4 pi of aniline.

These matrix deposition parameters were based on methods from Djidja et ol (2009).

2.3.2 InstrumentationPeptide mass fingerprints and images were acquired by MALDI-MS/MSI using a Q-Star Pulsar i

hybrid quadrupole time-of-flight mass spectrometer (Applied Biosystems/MDS Sciex, Concorde,

Ontario, Canada) fitted with a variable repetition rate Nd:YV04 laser set at 5 kHz. Image

acquisition was performed using raster imaging mode at 150 Im spatial resolution, Biomap

3.7.5.5 software (http://www.maldi-msi.org/) was used for image generation. Instrument

calibration was performed using standards consisting of a mixture of 217 polyethylene glycol

(Sigma-Aldrich, Gillingham, UK) ranging between m/z 100 to 218 3000 Da prior to MALD1-IMS-

MSI analysis.To enable simple visual comparison between images all data were normalised to

m/z 877 (a peak arising from the aCHCA matrix) and intensity scales in the BioMap software

were all set to the same value. MALDI- IMS/MS, MALDI- IMS/MSI and MALDI- IMS-MS/MS

were performed using a HDMS SYNAPT™ G1/G2 system (Waters Corporation, Manchester, UK)

and Drift scope 2.1 software (Waters Corporation, UK). In order to achieve good quality

MS/MS spectra, they were acquired manually moving the laser position and adjusting the

collision energy to achieve good signal to noise for product ions across the full m/z range of

the spectrum. Collision energies were adjusted from 70 to 100 eV during acquisition and

acquisition times were generally of the order of 5-10 s per spectrum. MS/MS spectra were

uploaded to perform a Mascot (Matrix Science, London, UK) search which used the UniProt

database in order to generate a sequence match (see Table 1 for results).

78

2.3.3 Haematoxylin and eosin stainingSlides were immersed in haematoxylin for 1 min, rinsed in tap water until the water ran clear,

immersed in 1% eosin for 30 s and rinsed in tap water until the water ran clear. They were

then dehydrated as follows: 50% ethanol for 2 min, 70% ethanol for 2 min, 80% ethanol for 2

min, 95% ethanol for 2 min and 4 changes of xylene applied to each slide for 1 min at a time.

Finally, they were mounted with DPX mountant and left to dry in the fumehood overnight.

2.3.4 Data pre-processing using SpecAlign and Marker View softwareData lists were exported from Analyst QS software (Applied Biosystems/MDS Sciex, Concorde,

Ontario, Canada) as text files, then imported into SpecAlign (Oxford, UK) to undergo the

following data processing; baseline subtraction, smooth, denoise, normalise TIC, remove

negative, generate average spectrum, processing spectral alignment, PAFFT correlation

method max shift 20. Files were then exported as text files for import into Marker View

software 1.2 (Applied Biosystems/MDS Sciex, Concorde, Ontario, Canada).

2.3.5 Statistical analysisPCA-DA was performed using Marker View software 1.2 (Applied Biosystems/MDS Sciex,

Concorde, Ontario, Canada). Post SpecAlign text files were imported and data was transposed

into table format. A minimum intensity of 0.1 was selected, with a maximum number of peaks

of 20,000. Monoisotopic peaks selected by Marker View were used in the supervised PCA-DA.

PCA and PLSDA was performed using MATLAB®(Matrix Laboratory) (MathWorks, Inc., Natick,

MA 486USA).The PCA-DA/ PCA/ PLSDA data that follows is representative of the fibrosarcoma

120 data where 6 biological replicates per time point with 6 technical replicates for each,

(control - 24 hours). PLSDA Data for ROI study using HDI software (Waters Corporation,

Manchester, UK) was analysed using one tumour per time point with 9 technical replicates per

ROI. Variable Importance in Projection (VIP) scores were used from PLSDA regression

modelling to determine the importance of variables within the data time course.

2.3.6 Data pre-processing usingWaters MassLynx™ Software and MATLAB®MS results are imported into MATLAB® in .txt or .csv format either post "SpecAlign" or after

application of "automatic peak detection" if using the instrument data processing software

(Waters MassLynx™ Software). Normalisation (2-Norm) and mean centre were selected,

"contiguous blocks" was used for cross validation.

79

2.4 Results and Discussion

2.4.1 In itial MALDI profiling and imaging of in situ tissue tryptic digests in mouse fibrosarcoma 120 modelsInitial profiling and imaging experimental work was performed using an Applied Biosystems Q-

Star Pulsar i Quadrupole ToF Mass Spectrometer. Preliminary peptide mass fingerprints

(PMF's) (Figure 2.1) where acquired to provide insights into typical high abundant potential

peptides in order to observe any gross pharmacological responses post CA-4-P administration.

+TGF MS: 42 MSA scans from Sanple 3 (190210_5_6_centre) of 5_6.wiff a=3.57031177631956190e-004, t0=5.60672130395284330e+001, Smoothed, Baseline...

Max. 9540.2 counts,

9500

9000

8500

8000

7500

7000

6500

6000

5600

5000

4500

4000

3500

3000

2500

2000

1500

1000

500

1416.5415

1274.6311

1048.4783

850.3543

1013.4083

949.4176

1161.5557

11625584

1302.5300

13125544

1213.4991 1819.69341126.4847 13 S7.69771070.463c _788.4076 1577.59541551.580-1089.4631 1841.6850

1718.6566

1529.5902

1555.6108

1300 1400rrVz.arrti

1500 1600 1700 1800 1900 230C

Figure 2 .1 : Representative peptide mass fingerprint (PMF) arising from an on-tissue digest of a VEGF120 tumour 24 h after dosing with 100 mg/kg i.p, CA-4-P. Prior to matrix application in situ tissue digestion was performed with trypsin solution prepared (from lyophilised trypsin) at 20 pg/m l by addition o f 50 m M ammonium bicarbonate (NH4HC03) pH 8.12, containing 0.5% octyl-a/b-glucoside (OcGIc). The matrix employed was a-cyano-4- hydroxycinnamic acid (CHCA, 5 mg/ml) and aniline in acetonitrile:water:TFA (1:1:0.1 by volume).

80

68ZESZEI

<u SQ

ezsvtS lt> 6 8 n -

ZCZE SZEl

Z90E1

8 8 8 8

E80Z6ZSI*SZE*68H

J)OOOt'Sltl

Z9lZ>ZZl

90EZ86U-— ^

VE8EZE0V-

SOIZ **6 ------

tLdZVLlli l8 £ 8 6 l t -

8 8 8 8

0)€ S•5 cC* 7 ^ <1

? 3

& 0)§ §

Qj§ ■c

4-1

c0 q: cu ^

^ FC5 ^VO 4 - ~ VO

? s1 as -c

£ <N§ S+- "Cvo c

■2 Q.

Qj i »o(U O 3 +-VO VO,vo o+= a

1 5is0 cCa 71 3 8 +-1 IVO « •

2 -c■o in ^ dQj ^€ >2. c5 V | 2 7 |

*C3 U Q 0) +- Q.

.C vo .2"5 §. T3■9 -c «u ° o

.§ JS.-STk -5 C2" <U ’£^ .c #2 Ot ^ «N•S 2 .2^ S ■«vo +■*

1 £ I O =*•

Q. *0 ^£ ^ * a £ s- S .§

? •§ *<u c <u^ CU .£

<N

•S' 2- gu. (U u

The spectra shown in Figure 2.2 are from the following samples: (a) control/saline, (b) 0 h post

CA-4-P, (c) 0.5 h post CA-4-P, (d) 6 h post CA-4-P and (e) 24 h post CA-4-P.The increase in the

relative intensity of m/z 1819.3 along with other readily identifiable haemoglobin peptides (e.g.

m/z 1529.7HbAa 18-32 (IGGHGAEYGAEALER), m/z 1274.3 HbAb 52-41(LLVVPWTQR)) can be

seen. This increase in tissue haemoglobin is as would be expected considering the vascular

damaging propertiesof CA-4-P. Due to disruption of the 3D capillary architecture, endothelial

cell necrosis, apoptosis and leakage of blood cells into tumour tissues is inevitable. This is

clearly illustrated by Figure 2.3, where MALDI-MS images for the distribution of m/z 1274

(HbAb 52-41) at each time point are shown along with photographs of the corresponding

haematoxylin and eosin stained sections. It is debateable in the Oh treatment how much of the

drug reaches the tumour, however there is an apparent signal in Figure 2.3 corrresponding to

m/z 1274. Further drug distribution studies would clarify presence of CA-4-P and its

metabolites.

Figure 2. 3: MALDI-MS images fo r the distribution of m /z 1274 throughout a fibrosarcoma 120 timecourse. (HbAb 52-41) in (a) control (saline), (b) 0 h post CA-4-P (c) 0.5 h post CA-4-P, (d) 6 h post CA-4-P and (e) 24 h postCA-4-P, along with photographs of the corresponding haematoxylin and eosin stained sections.

24 hourS CA-4-P

/ V ' , • r•r. J*. vyjdespread

necrosis

82

2.4.2 Haematoxylin and Eosin staining of fibrosarcoma 120 tissueHaematoxylin and eosin (H&E) staining of each fibrosarcoma 120 serial section was performed

after imaging for identification of viable and necrotic tissue and to make observations of the

tumour rim.

Examples of viable and necrotic tissue are shown in Figure 2.4: (a) control saline-treated

tumour which is essentially viable, (b) control tumour with a small necrotic region, (c) 0 h CA-4-

P showing viable tissue, (d) 0 h CA-4-P showing viable and necrotic regions, (e) 0.5 h after CA-

4-P showing viable tissue, (f) 0.5 h after CA-4-P showing viable and increasing necrotic regions,

(g) 6 h after CA-4-P showing partially viable regions, (h) 6 h after CA-4-P with haemorrhage and

necrosis and (i) 24 h after CA-4-P treatment showing total necrosis.

Higher magnification images of the H&E staining of the tumour rim in the VEGF120 tumour

sections are shown in Figure 2.5: (a) control/ saline, (b) 0 h CA-4-P, (c) 0.5 h after CA-4-P, (d) 6

h after CA- 4-P and (e) 24 h after CA-4-P. The H&E tissue sections indeed reflect the

heterogeneous nature of the tumour especially considering the control with no CA-4-P (Figure

2.5 a), however the effect of the vascular disrupting agent is still apparent and the 0.5 h image

(Figure 2.5 c) suggests a good example of the viable tumour resistant rim.

In all images, viable tumour cells have densely packed nuclei which appear as regions of

"organised cells" which stain blue in H&E staining whereas regions of necrosis, due to cell lysis

tend to show as a disorganised region with less blue staining. In these data, haemorrhaging

also gives a pink colour to the tissue e.g. Figure 2.4h.

83

M i

V- «f] ». »Vv‘ f £

£ a

pca . cs

■ QjSt1S t

3p

0 S-c CS)© c

"o ' 1o

s?o

■cCOQ.

CJl ■QJ S tc. 1

.Cj’■C 2

8O

-CoCOo

QJ O .C -c

"o ixs

5 dCOO S

cn Qj'c 3

COi .COo ■G

-c -SiCO"o Co+o 5c csto cCj

Io

0?-cCO

3 Q.CO ■

.CO St■Co s t-Si 0

■Co.5 CO

OS aCS) -c

p ix)i do

-c "oTCO . VQj CO

.c C.O"5 B)JOqjV .

O .Cj

£co

’■Co

go QJ—v C

P. cCO 3c

.o ■Si■C pCJ 3QjCO ‘5

P cs)Qj pC §aCO

o■c

.CCOQ.<o ■

O StQj

PC

1

2cs -C

.c o—

"Ss P* ’s-—o qT■Coo 3

CO5Qj

•52PCS -Si15• • ps t S<N O)Qj

cV.3 I■5* o

-ci l CO

s£QSi■Co

Q.+2v.Qj

f-cStfM

gQjCPCcsCS)c5)cs

0 5QjQ•C

1a .Isi-I2 •Cvo

c/fc

.o5)£-Sip*5

cE0 Q. cs)

p10

-ccoQ.■S t12 -cVOE !co'c

,o3> ,co Sj <o^ ocj i ; +3 °1 £ c jQj o C £O) S .5 &co C O ‘5QJ £

•S ■§

00

■53^ H•“1 -

Figure 2. 5: Higher magnification images of the H&E staining o f the tumour rim in the VEGF120 tumour sections. Control (saline), (b) 0 h post CA-4-P, (c) 0.5 h post CA-4-P, (d)6 h post CA-4-P and (e) 24 h post CA-4-P. The 0.5 h image displays a good example o f the viable tumour resistant rim.

85

2.4.3 In itial imaging of proteins according to theoretical digests using M A L D I-

MS1Proteins which could be of relevance to tumour biology and the stress response were

tentatively identified according to theoretical digests (Li et al 2011). The following preliminary

MALDI images were observed to aid focus on potential future protein targets to consider in

this dose response relationship study.

CHCA matrix peak m/z 877

r■k

r * \

D jlaLC 101I09_5_1(D911101C .h3fxTO _l_m 3.W B *

(c) TNF-a m/z 954' ' 2 10

.vt * r '

(a) m/z 1887 Grp78

%*

(d) VEGF 120 m/z 1485

(b) m/z 1749 HIF-1a0 .9 6

(e) VEGF A m/z 1850

Figure 2. 6: MALDI-MSI o f peptides from tumour 5:1 (24hrs CA-4-P). CHCA matrix peak m /z 877 used for normalisation ,(a) m /z 1887 Grp78, (b) m /z 1749 HIF-la, (c) TNF-a, (d) m /z 1485.7 VEGF 120, (e) m /z 1850.8 VEGF A (188), according to theoretical digests and (f) tumour 5:1 histological section after H8cE staining with magnified insert showing necrotic tissue.

86

The ion signals shown in Figure 2.6 were generated according to predicted theoretical tryptic

digests. Although at this stage protein the identifications were not yet confirmed, it was

fascinating to notice the differing spatial distribution of the potential peptides thus reflecting

the heterogeneous nature of tumour tissue.

The possible detection and spatial distribution of proteins like glucose-regulated protein 78/

Binding immunoglobulin protein (Grp-78/ BiP) (Figure 2.6a) could reveal vital information of

how CA-4P can effect tumorigenesis thus inducing other pathological stimuli.

The image tentatively identified as belonging to VEGF A (188 isoform using UniProKB database

(mouse) with 2 missed cleavages) in Figure 2.6 e is quite contradictory to the mouse model

employed which had been genetically engineered to solely express the VEGF 120 isoform. The

latter could be due to immune cell infiltration.

It was found however, that through further interrogation of peptide sequence databases i.e.

Blast (http://blast.ncbi.nlm.nih.gov/) and reference to the article by Tischer et al (1991) that

the peptide sequence given regarding the VEGF 120 ion at m/z 1485 is incorrect. The latter

discovery came about whilst checking whether or not the sequence giving the ion signal

(Figure 2.6 e) at m/z 1850 really matched part of the VEGF 188 protein sequence to support

the theory of possible immune cell infiltration. It appeared that none of the VEGF 120 peptide

'RWQKIADELGNR' (2 missed cleavages) giving m/z 1485 provided by UniProKB theoretical

digest is detectable in the original VEGF nucleotide sequences (Tisher et al 1991). The

sequence in question when submitted to Blast database gave a result suggesting that the

peptide could be a zinc finger protein. The VEGF 188 peptide 'FKSWSVHCEPCSERR' (although 2

missed cleavages) giving m/z 1850, matched with the nucleotide sequences detailed in

UniProKB, Blast and original VEGF nucleotide sequences presented in Tischer et al (1991). Part

of the VEGF 188 peptide sequence coincided with nucleotide sequences present in the VEGF

exon 6, a region of the VEGF gene known exclusively to VEGF 188 (Figure 1.29), hence

supporting the presence of VEGF 188 in the MALDI-MSI (Figure 6 e).

87

2.4.4 Principle component analysis - discriminant analysis (PCA-DA)Multivariate analysis was chosen to statistically analyse the complex data sets which arise from

on tissue PMF's. PCA-DA was performed to try and correlate peptide inductionwith dose

response and to provide a method of further analysingthe PMF data.

Scenes lot Dl 141.5 veisus D2 (33 3 5S). Pareto (DA)

22-

20-

18-

16-

14-

12-

10-

8-

64-

2o 0o

<M 2O

-4-6 ■

•8-

-10-

•12-14-16-18•20.

•22.

-24-

?cmouf3!4 d M 2

(a )tumoui^5_l txt

fumour3 3clxt tumom3_bf t*l i

• 'a lwmout3_3d lx*»ijriour3_3e ixl

turnout 4_c.jxt - T t urrour3_3b.lxt

lymolw3_3d txt ■ turoour2_2_d.btl

tumoui3_6_b;lxt - * -£ ^li^mow3_4_e lit - t f lumoui3_&_c txt

tumour 2_5_b.lii( — H i tumour 3_5_b.(xt turnout 3 ftict

tusnour5_3_d tx*Hirnc«ji5 T " a i.W YumourS”5 c fy t " 'fC in o u fT ^ e tk f

(urr»ur5 4 f M

tumoui5 1 b Ixli "" T

MrtQu5_2_<S Ixl

lumout5_3_4.lxl

t •-tumourS 3 Mxi

,• • •lumour5 4 d tx t

lumoui5_5_blxl

tumoui5 5 a Ixt

i >}umout4_5J txt

tumour4_2..e.bct

tumori4_4_e Ixt ^ turnout 4j_4_b Ixl

tumour4i 4 dtxt

1x1

V-j turnout 4 4 f txl

i tumout4_3_f 1)4 lj3_b W

iumoui4_3_d Ixllixnour4_|3_b txt

Ujrncxji4_3_a ixlt

/tumour4 6 dtxt

-40 -35 -30 -25 -20 -15Dl Score

-10

; tumour 4_3_c Ixt-i - A ------------

tumourl turi»ur2 tumour 3 tumour 4 tumour5

Figure 2. 7: PCA-DA of fibrosarcoma 120 tumour in situ tryptic digests.(a) The scores plot showing groupings and variability between tumour time point spectra, Tumour 1-5 indicates Saline/Control - 24h post CA-4-P (b) the loadings plot displaying the separation and spatial distribution o f m /z values in relation to score plot positions, (c) illustrates the variability of two haemoglobin peptides within each tumour section. The red arrow on (b) indicates m /z 1274 (indicative o f haemoglobin b chain) which is associated with the tumour 4 set (6 h post CA-4-P).

88

Loadings for D1 (41.5 Z) versus D2 (333 Z). Pareto (DA)

(b)

1528.20

009 0.08

007

0.06

0.05-

004

0.03

0.02

001

0.00 ■■

-001

-002

-003

-004

-005

-006

-007

-008

-009

-010 •011

-012 -0.13

-014

-0.15

-016

943.26- •

132558

9 4 4 1 2 - • ^

1032 71 853.48 ,

^ 813 2 4 - *

•

945.98-

1514.94A

1031.21

1032.31

103332 A

91323 J — - 842.2?...®

1820.35 • «1 5 3 0 8 7 * j j .

1 8 2 2 .3 3 -= ^

: 800 35889.23 I

j l 808.39 r *

4-1325.38

j • £ - 1 3 2 7 .8 0

• | t * 82526

1m • -------1322.39(*ryf 983 36

1530.43i

1274 99

A 1302 79 1

130304"t103023 <

* 875 19 § • [ 849.50

1303.31

94495A

1530.27A

1531 30

1

-020 -015

m /z 1 5 2 9 - H b a

•0.1001 Loadng

J 873.32| . ,27880 i

1415.482*—. . / * 5 3 t.8 ,

“ V s • P i S - |1S20 214 « i i 1272.59 ^

1T ‘ L . ’4 “11273 40 • (

i , 1C0£. 943361416.86 -« l

1532.35 945.09

1032.87

274 41 A A 1275 39

_1_____0 05 000 0,05

(Default)(Isotope)(Monorsotopic)

806.35

968.37

831 54

1197.52

| 847 31

J 853 29 i

to

tt mur2_3_fkt

m /z 1274 - Hb ft

% m M r4j.ik t

kmur3 3b.kt kui»ur3_5_i.fct . \ * ! * 4-3j

— \ knourd 5 i.k t4 3 i.k t V , ” J

!L«J *******. . . . . . . . . r " 1i m«nr3_t_e.

jWfrWmTnmmHtu

im tu rk M lil knoi>4 6 fk t

kmour: 2 i.k timourb 4 i kt

4inMr1.4_d.kt kirnoui2.3_fkt kmom3_1_b.kt knour3.4_d.kt timour4_1_fkt kmMr4.5_b.kt knour5_2_d.kt kmourS_5_fkt

Sanplc (by hd«x)

m /z 9 0 1 .3 5 u n k n o w n

turrour1_5_c.txt

tu ra r2 2f.txt i2_2_atxt i h.m>ir25 htxt

turmr3_5_f.txt

ttunnur3 4_ctxt

1unoir3_.4_b.txt

% j | tunoui3_6_ctxt

i

K

I tunxr4 4 c.txt

La .««***—**•+turar4_4 e.txt

turoir4 5 a txt

fcjrrair1_4_dtxt tunxi2_3_f.txt Inmr3_1_bitxt tumDur3_4_dtxt turrair4_1_f.txt

Sar»|ie(ty index)

89

Figure 2.7 shows the results of PCA-DA of VEGF120 in situ tryptic digests: (a) the scores plot

showing groupings and variability between tumour time point spectra, (b) the loadings plot

displays the separation and spatial distribution of m/z values in relation to score plot positions,

(c) illustrates the variability of two haemoglobin peptides within each tumour section and an

unknown at m/z 901. The simple response present from the haemoglobin peptides was

regarded as a standard for future time course results to aid validation of dose response

relationships. The red arrow on Figure 2.7(b) indicates m/z 1274 (indicative of haemoglobin (3

chain) which is associated with the tumour 4 set (6 h after CA-4-P). In Figure 2.7(b) the green

data points are peaks identified as monoisotopic by the Marker View 1.2 software and blue

data points are peaks which have not been assigned as monoisotopic. These are termed

"default" by the software.

For the chemometric investigation of the proteomics data reported here, multivariate analysis

was deemed essential because of the large data sets thousands of signals can arise from just a

single pixel due to the acquisition of full-range mass spectra (McCombie et al 2005).

Comparisons of spectral signals are required between both biological and technical replicates

throughout a time course experiment and multivariate analysis allows a better visualisation of

the "vast data clouds". PCA-DA was chosen to provide preliminary insights into ion signals

generated by in situ tryptic digestion using MALDI-MSI with the hypothesis that the samples

should show separation of unique peptides due to known individual treatment time points.

The scores plot in figure 2.7a provides clear separation of tumour treatment groups 3, 4 and 5

which refer to 0.5 h, 6 h and 24 h post CA-4-P respectively. The earlier time points being

tumour 1 (control/ saline) and 2 tumour (0 h) are not visible but can been seen if the region

indicated is expanded. The peaks that appear on the loadings plot (Figure 2.7b) seem to be

representative of high abundant peptides i.e. histones (m/z 944, m/z 1032, m/z 1325) and

haemoglobin (m/z 1274, m/z 1416) which relate to the occurrence of haemorrhagic necrosis,

DNA damage and hypoxic conditions within the tissue.

90

2.4.5. Employment of Ion Mobility MALDI-Mass Spectrometry to stud yin situ tryptic digestion in fibrosarcoma 120 modelsIon mobility enabled MALDI profiling and imaging describes a further separation method of

isobaric species thus opening up another dimension to spectral analysis.

*<3 U iaB I ^ lrlI*

m m-m m

Haem oglob in P01942 2858.4 VADALASAAGHLDDLPGALSALSDLHAHK 24 >20subun it alpha +21.9819 AT D14 C a tion:M g [II] (DE)

[+21.9694], Ca tion:Na (DE) [+21.9819]2836.4 VADALASAAGHLDDLPGALSALSDLHAHK 87 >201819.8 TYFPHFDVSHGSAQVK 106 >361529.7 IGGHGAEYGAEALER 98 >181416.7 GGHGAEYGAEALER 87 >43

H is tone H4 P62806 1325.7 DNIQGITKPAIR 76 >14

Haem oglob in P02088 1302.6 VNSDEVGGEALGR 79 >38beta chain 1161.6 LVVYPWTQR 29 >20

1048.5 VVYPWTQR 32 >20

H is tone H2A Q8CGP5 1300.6 NDEELNKLLGR 33 >21Type 1-F

Ac tin , aortic P62737 1198.7 AVFPSIVGRPR 30 >13sm ooth 976.4 AGFAGDDAPR 64 >19muscle

H is tone H3 P68433 1032.6 YRPGTVALR 11 >10715.4 DIQLAR 19 >17

H is tone H2A Q8CGP5 944.5 AGLQFPVGR 24 >19Type 1-F

Rho GTPase- Q91V57 844.5 RLTSLVR 21 >18ac tiva tingp ro te in 2 / N-ch im aerin

Table 8: Identities and Mascot scores from the IMS-MS/MS analyses of peptide signals showing Mascot score o f mouse protein identifications and corresponding threshold score at 95% confidence level.

91

Q»17J- 5 ;3

a £cs

u3o

£ us £3

1c.o

C.cCS5

■C o*

§"33

oV.■3>

£ -Q -c1i

1s*10

CS -S' c3o X*§ PM 0■3 4-3 3V. 100) 3§

,c £o a "3■c >s QjVO ■fc XV. 73 O3 QJ aO£

•£ £Qj3 3Q.cs 9 •W«N cu! £ £<0 o Qj

s iS3 3 •

Si .CjO 3 <34i Q. uSPQj Si 0.cs a Qj

c 3Qj .o 553 ■C OCO 3 cj3£■ "o</> "SCo ,C

5?1

§c Q. 3 «3£ 1

■is4-.

"oT£

8 3OO 3

3) I 1.C ■3 Us,«/i Qj cs -QC £ cs3 ,c duT •& £ *-i§3^

QjQ.

"34)c</> 04> 3 CJ lC.cc £ pm" h-i*Q. C QJV.QJ ,o"43 00 4-»O3iC coQj a.CS -CDcoV)3

■3Qj

fvs'8 §

£ 3«/> 3: £0Qj;$4w

4>■C3

QJCjQ8-

.4-15? 'o74- .C

a C 3 QJQj C .Cc O •3"■C34-1c

o-G3u

*830£33Qj •3 £.3

"3CJ

ccoQjV.8QC

8

C4x

3

100«N

3£o

£3c

£us

Qj 4o 0V.3CSi l

oa

££3

05

CMCD

A representative peptide mass fingerprint obtained from a tumour from an animal sacrificed 6

h after administration of CA-4-P (hereafter referred to as 6 h CA-4-P) using the Synapt

instrument, is shown in Figure 2.8. These data illustrate the vast number of peptides that arise

from an in situ tissue tryptic digest, as previously reported (Schwamborn et a/2007.,Le maire et

al 2007., Shimma et al 2006., Groseclose 2007).

LC_1OG302_W 35 (1.061) S611.40.00). Srt (Mn. aOOOl Cm (3 JO) TOf MSMS 1819 35UD*too-

M S/M S Fragm en ta tion o f TYFPHFDVSHGSAQVK

b fragment ions = bluey fragment ions = red

o

1124> S’ id

726 3657 6131,78 9C63965

41213151231 5

111 itniik ill Ini bt903 1000 1100 ---

3 446 6233 15f3 E915I ” iI I300100 200 700 eoo 1200 1330 1403

’ 15

Figure 2. 9: IMS-MS/MS data obtained from the peak observed in Figure 2.8 at m /z 1819.3. This was identified by Mascot search as arising from Mouse HbAa 42-57 (TYFPHFDVSHGSAQVK).

Figure 2.9 shows IMS-MS/MS data obtained from the peak at m/z 1819.8. This was identified

by Mascot search as arising from Mouse HbAa 42-57 (TYFPHFDVSHGSAQVK) (Table 1). These

data are also a further indication of the advantages of IMS for the identification of "on-tissue"

generated tryptic peptides. The complexity of the obtained data sets is shown in Figure 2.10a,

where ions arising from doubly charged species and singly charged species are clearly

separated by the use of ion mobility and further those singly charged ions arising from

peptides, matrix adducts and lipids are also separated. Use of IMS separation prior to MS/MS

yielded good quality MS/MS spectra for peptides that produced a high and significant Mascot

score.

93

1003_1

uooo

oaooo

oo)

LC 100

302

02.ra

w: 1

ma“

3000

nvsVOQ>.C

T3C£

«£!<o«oOQ,

&C.o■Goc

c«oCSco

Q8 -VO

-505

1 ■§ 5 8cnc -S 1 &2si a

<oo ct - .2vo •*■§ .S-! CM OC S.2 i4-> o U r=q j c £ t3

O UC QjI fQ V0

O c +- O5 T3 a. qj Qj ^8- o 8 -5 OsQ -S z. <*^ -C3• • <u

P'0 i r

1 |* iQj -Cb -Q2» 3 .o> oU. t3

CD

-TIiODOo*s883e

Figure 2.10 c: Drift scope plot shows ion mobility separation of ion of interest haemoglobin subunit alpha m/z 1819.8 and C13 as shown on mass directly below.

The additional specificity afforded by ion mobility that allows the selection of individual ion

species can clearly be seen in Figure 2.10c . The white arrow indicates the ion of interest (m/z

1819.8) and the ion signal directly above depicts the isotope C13 which can be seen on the

mass spectum below the plot.

Figure 2.11: The distribution o f two known peptides in tumour 5_6, i.e. a tumour 24 h after treatment with CA-4-P (a) m /z 1416 from haemoglobin a chain depicting central necrotic haemorrhage and (b) m /z 1198 corresponding to actin.

The MALDI MSI images shown in Figure 2.11 show the distribution of two known peptides in

tumour 5_6, i.e. a tumour 24 h after treatment with CA-4-P: (a) m/z 1416 from haemoglobin a

chain depicting central necrotic haemorrhage and (b) m/z 1198 corresponding to actin. The

inverse nature of the ir distribution is striking.

97

Figure 2 .12: Overlaid MALDI MSI images showing differing spatial distribution and co­registration of peptides, (a) m /z 944 Histone 2A in red and m /z 1274 Hb in solid green, (b) m /z 1819 Hb in red and m /z 944 Histone 2A in solid green.

Figure 2.12 shows examples of overlaid images to show differing spatial distribution and co­

registration of peptides using MALDI MSI: (a) m/z 944 histone 2A in red and m/z 1274 Hb in

solid green, (b) m/z 1819 Hb in red (from rainbow) and m/z 944 histone 2A in solid green. The

Mascot scores for the IMS-MS/MS analyses of these peptides are given in Table 8.

98

2.4.6 Rho GTPases: a target to help understand resistance to CA-4-P?The Rho GTPasesare proteins known to be involved in cell permeability and actin/cytoskeleton

dynamics. (Modolo et al 2012., Aznar and Lacal 2001., Aznar et al 2004., Schmitz et al 2002).

Influence on extra cellular matrix (ECM) integrity, apoptosis, cell adhesion and proliferation,

makes them potential targets in the comprehension of anti-tumour drug resistance.

A review by Vega and Ridley (2008) discusses the role of Rho GTPases in cancer cell biology

and describes them as "molecular switches" which continuously revert from the active and

inactive forms which are GTP-bound and GDP-bound respectively. In the activated state they

can bind to many effectors which subsequently elicit a cascade of events leading to the

processes mentioned previously.

Interestingly with relevance to the work reported here, some Rho proteins promote

neovascularisation which could be as a result of association with angiogenic factors and drug

resistance effectors (Vega and Ridley 2008).

Over a decade ago work by Kozma et al (1996) suggested that N-chimaerin (Rho GTPase

activating protein 2) is implicated in morphological changes of the cytoskeleton when

associated with cancer related protein Racl and cytoskeleton signalling protein Cdc42Hs

(Kozma 1996). It is stated that n-Chimaerin induces the formation of the actin-based structures

lamellipodia (actin projections) and filopodia in migratory mammalian cells and have pro-

oncogenic properties. It has been suggested that Rho GTPases have strong influence not only

on tumourigenesis but the migratory functionality of the cancer cell (Figure 2.13), therefore

key players in the metastatic process (Karlsson et al 2009).

A potential combination therapy for colorectal cancer patients is administration of

bevacizumab and cetuximab to target vascular endothelial growth factor (VEGF) and epidermal

growth factor receptor (EGFR) (Leve and Morgado-Di'az 2012). It is known that in gastric

cancer VEGF interacts with Racl whereas EGFR acts as a stimulatory factor of Rac to encourage

the migratory process of colonic cells. Therefore the therapeutic likelihood of inhibiting cancer

progression is evident. In an article by Leve and Morgado-Di'az (2012) it is suggested that it

could be highly beneficial for the future of anti-cancer therapy to identify inhibitors of the Rho

GTPase-associated functions.

99

Tumor cell

Cell cycle

Epithelial cell polarity

Rho A, Racl

Cdc4^

Angiogenesispromoting

factors

MigrationMesenchymal: Racl Amoeboid: Rho A D i recti onal ity: Cdc42

Figure 2 .13: Major Functions o f Rho GTPases in vitro ( Karlsson et al 2009).

A putative peptide assigned to the Rho protein N-chimaerin as mentioned earlier (Rho GTPase-

activating protein 2) was identified in the follow ing MALDI-IMS-MS/MS experimental work

(Table 8). Mascot search results produced the peptide 183-189 (RLTSLVR) for the ion at m/z

844.5288.

The peak of interest is indicated in the PMF from fibrosarcoma 120 tissue 30min post CA-4-P

administration (Figure 2.14). The corresponding MALDI image is featured showing the spatial

distribution of the selected ion at m/z 844. The peptide appears to be located mostly around

the periphery of the tumour possibly indicative of its involvement in resistant to CA-4-P. The

IMS-MS/MS data obtained from the peak observed in Figure 2.14 at m/z 844.5 is shown in

Figure 2.15, the mascot score histogram for the said results is featured in Figure 2.16.

100

Qj£

O)e'■ff5 - a> 41O .«P :<j>V) •"roQ. cs

5 f| 2 a Q.

i f # ^ $ 2 ' |■- -44 ?Mti; v > r JjBft eQ* ^ ■ „ (* ' 5?- ?<? .^ ' v£P P»r •. . 0 • “

j f e v i l f : ' :.j>!;v*. c :-v , ?. . f . .r-W .% .0

W t iw jB P im c W *.

'+hO

=x<us <~4*c O1—4 38«5oooll_o 8 o ffi

* Ir**eft

© 2i ? nin —s •t1lit (r>Ul (N

*8,» 8 «g 5s|/) i Ol a s

101

IMS-M SMS 8 44.5288LC_100720_1826(0.736)Cm(1228)

175.1413 Rho GTPase activating prote in 2 183-189 (RLTSLVR)

8145748 84 4.5231

115.1004

129.1127 2662001405231-

490.43454492624461.; 716.40063782197 I.3297 68!■ | 5893490 616.4066 I . I L500.4341J I 6443606 668.4182 7475066

Figure 2 .15: IMS-MS/MS data obtained from the peak observed in Figure 2.14 at m /z 844.5. This was identified by Mascot search as arising from Mouse N-chimaerin (Rho GTPase activating protein 2) 183-189 (RLTSLVR), the insert features the matched MS/MS fragmentation pattern.

{Science} M a s c o t S e a r c h R e s u l t s

OsarEmailSearch title MS data file Database Taxonomy Timestamp Protein hits

Laura Cole ImooleSmy.shu.ac.ok

a l l . p k lSwissProt 2012_10 (538259 sequences; 191113170 residues)Mus. (16624 sequences)16 Nov 2012 at 11:25:09 GMTH2A1F_MOOSE Histcne H2A type 1-F OS=Hus musculus GN=Histlh2af FE=1 SV: CHIN MOOSE N-chimaerin 05=Mus muscuius GM=Chnl FE=1 5V=2

M ascot Score Histogram

Ions score is -10*Log(P); where P is the probability that the observed match is a random event Individual ions scores > 18 indicate identity or extensive homology’ (p<0.05).Protein scores are derived from ions scores as a non-probabilistic basis for ranking protein hits

Protein Score

Peptide Summary Report

Figure 2 .16 : Mascot score histogram detailing the results in the identification of N-chimaerin.

102

57

20.0 40.0 60.0 80.0 100.0 120.0 140.0 1600 180.0Drift Ttrte (Bins)

(O.M.202)(l4».:oO.M))(«M.IH_2>M.tO) (845:55043)

Figure 2 .17 : Zoomed in Driftscope plot showing ion of interest, arrow indicating the peak originally selected.

It could be said that there are many overlapping signals in the tryptic peptide MS/MS spectrum

shown in Figure 2.15 where both the arginine y l ion (m/z 175) and evidence of potential

hydoxy fatty acids (m/z 187) are visible. This is also evident in the Driftscope plot (Figure 2.17)

where the arrows point to the actual peak that was selected for MS/MS fragmentation. The

dominant peak to the left of the indicated peak in Figure 2.17 would be part of the MS/MS

spectrum, despite this the peptide was matched and identified (Figure 2.15) having 1 missed

cleavage from the fragments present in the data.

103

+o

LLO

CL

tnto

aQ.

SS

o>\_2L

oEOCOCO<oooo'poo'oooo>

CMo

ooo

o.II

<u+-•tooac

EoCO

CM p _=

a>c

"roCO

cou

o —1

«cind"qT.cs8c<szi.c5aCD

5

-Si.<os.ccCDo5•§

■C§•aCD

V5<3

4 o3Q.-C

PM CDQ5o Oo 4oV. COQ oCOo Q.c. -C-Q

«NC 134o c3 (3O-c a.cr>3O»o Cj-c4o 4oCOCO OU. Q.5 -Ca. to00 eC*-h ■PM 1

8 O3o>

4oCOOul Q.

104

The apparent absence/ low abundance of the ion at m/z 844 ( putative peptide, Rho GTPase

activating protein 2) in the later time points post CA-4-P administration, could be indicative of

how the drug effects the Rho protein's usual involvement in tumour progression and neo­

genesis of the tumour vasculature. The latter response is evident in the follow ing MALDI-MSI

(Figure 2.19) of the ion found at m/z 844.5 corresponding to N-chimaerin.

LC_100720_3_ 1 _M $01 J)02J301 _max j j i v (MEAN 844 530-84*St1OJ. 100 720_ I _5_MS0 t_3G2_M *. {MEAN £4a -yj-AM W0|

1312T0 ;MEAN $4,3 j

Figure 2 .19: MALDI-MSI throughout a fibrosarcoma 120 time course o f the putative peptide Rho GTPase activating protein 2 m /z 844 (183-189, RLTSLVR).(a) control (saline), (b) 0.5 h post CA-4-P, (c) 6 h post CA-4-P and (d)24 h postCA-4-P, the intensity scales are all set a t 80 to aid observation of response.

As mentioned previously in this chapter the spatial distribution of this putative protein appears

to be localised around the periphery of the tumour sections in the Control/ saline and 30min

105

post CA-4-P (Figure 2.19). The images mirror the spectra in Figure 2.18 with the peak of

interest decreasing in abundance as the time course progresses.

106

2.4.7.StatisticaI analysis of mouse fibrosarcoma 120 tumour tissue using MATLAB®Principle component analysis (PCA)is an unsupervised test to produce unbiased clustering of a

data set, once confident that there were some differences emerging due to none

discrimination of data then partial least squares discriminant analysis (PLSDA) was then

chosen to provide a classification model.

MATLAB® (Matrix Laboratory) (MathWorks, Inc., Natick, MA 486USA) is used in many areas as

a computational statistical tool for mass spectrometry multivariate data analysis, and also via

the Bioinfomatics Toolbox" provides tools for spectral alignment and pre-processing.

Spectral data from both profiling and imaging experiments can be imported into MATLAB® in

order to build predictive models. This allows the user a powerful means of understanding

complex data sets. In addition to script writing using the usual MATLAB® command format, use

of the Eigenvector PLS_Toolbox (http://www.eigenvector.com) enables access to a number of

advanced statistical tools for the analysis of mass spectrometry (MS) data (Eigenvector

Research Inc 2012). Recognition of trends within a data time course experiment can be

detected via decomposition mathematical modelling i.e. principle component analysis (PCA).

Classification studies of MS spectra were performed using PLSDA, where regression vector

plots can illustrate presence of ions unique to a particular sample. MS results are imported

into MATLAB® in .txt or .csv format either post "SpecAlign" or after application of "automatic

peak detection" if using the instrument data processingsoftware (Waters MassLynx™

Software). There are various pre-processing options available to aid linearization of variables

within MS data with normalisation (2-Norm) and mean centre being frequently selected. 2-

Norm refers to the sum of the squared value of all variables of the sample (Eigenvector

Research Inc. 2012). During building of a PCA/PLSDA model the choice of cross validation is

crucial to determine complexity of the data i.e. to consider data splits samples and/or latent

variables/principle components. An example for a time course experiment would be selection

of "contiguous blocks" which would look for assessing predictability between repeat blocks of

spectra to circumvent the "replicate sample trap" which could lead to overly optimistic fitting

of data.

In MATLAB® complex data clouds can be modelled with ease via the Eigenvector PLS_Toolbox

allowing a deeper perception of variability throughout time course experiments.

107

Score PlotSamples/Scores Plot oftest.txt

■A6hrs

sCO.

c24hrs Oand

0.5 hrsC/5

Control in red with| Oand 0.5 hrs

-0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6

Scores on PC 2 (6.64%)T tumourl* tumour2■ tumour3

tumour40 tumourS

— — 95% Confidence Level

Scores on PC 1 (7.97%)

1276 4193 Variables/Loadings Plot fortest.txt

Hb peaks

1275.1464

1530.2719

2 4 h rs

. *1276.6537

,*4416.8600

*15-584417 6404 •852.3960

• *1198.4343

♦947.3414

•942.3239 •948.3452 .

•811.4t)73811.2206

’ *808.3930

•1531.5824 %*1532.5170 •

•1530*0632 *. * • ^15,30-723* 49'3483 *1035.368© 033.4313

• 1 2 6 ^ 8 ^ , ^ 0 ^ 8 5 1 .5 1 3 ? *1?p1.377*5 ’ ‘ .1032.4232

• 00 *J l96.-3 4 :ii27*40l627<35&4153f.fl^V.-._^4^ •1625*3835

$ 0 4 9 0 3 ^ 2 5 _____ i3 4 3 474R1274.9925 , \ I I . • •

•1326.5821$0.3100 ,

944.5917

Control in red with Oand 0.5 hrs

-0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2 0.25PC 1 (7.97%)

Figure 2. 2 0 : PCA of fibrosarcoma 120 tumour tissue in situ tryptic digests. The score plot (a) shows groupings between the tumour time points, the loading plot (b) contains the peak information which spatial relate to the groupings seen in (a) with dominant Hb peaks.

Figure 2.20a although not aesthetically pleasing as the PCA-DA model featured earlier in Figure

2.7 shows similar groupings shared by the earlier time points w ith separate regions assigned

108

for the 6h and 24h post CA-4-P. The Hb peaks are clearly visible in the later time points of the

corresponding loadings plot (Figure 2.20b). This now, seemingly a characteristic of the CA-4-P

susceptible fibrosarcoma model.

The conclusions that can be drawn from these data are fairly limited as the very intense

haemoglobin and histone peaks appear to dominate the statistics. It could be argued that

biomarker discovery requires supervised discriminant analysis i.e. PLSDA, opposed to non

discriminant analysis. The rationale behind this idea is that non discriminant analysis could

mask expected trends, an example of this scenario would be a treatment time course study.

The treatment added throughout a time course experiment is a known factor that is expected

to influence the data generated. Pre-grouping the data before building the statistical model

could aid the separation of relevant peaks of interest.

PLSDA aimed to provide classification between tumour time points. The PLSDA data that follow

are representative of the fibrosarcoma 120 data set as described in 2.3.5 Statistical analysis

section.

An example of the clarity of the data possible from analysis using MATLAB® is shown in Figure

2.21. Here, the regression vector plots from PLSDA of "on-tissue" digests of fibrosarcoma 120

Control/Saline and 30 min post combretastatin-4-phosphate (CA-4-P) (a vascular disrupting

agent), treatment tumour samples show the increased abundance of peaks at m/z 944 and

1032. This allows classification between data obtained from the control/saline and 30 min post

treatment tumour samples based on the induction of Histone 2A and Histone H3. The latter

could possibly be a reflection of DNA damage necrosis and hypoxia as a result of CA-4-P

administration as discussed in a previous section.

109

Vai

iabl

es/L

oadi

ncp

Plot

fc

rtes

t.tx

t

JCIT)

CL

-J <D < Q. > O

OvJo

CNc? ■*frCDO O2A-DJ-DPSAB j

0 O50u

QjVJ3

C 33 •3VOs

33

-3 «33

"33 Qj

52-2

vo3Qj

"S. V.O5 .3C3VO QJ5 <5£ Qj0 a4oO 5

*3 3VO

VO k-Qj 33> O

53

\ ) +->3vo,vo

4-13QJ

■JO 513

JO4o3QjV.4o

Q-J N4oc

§ QJ3>

3 ) 3.3c Cn

33 4o3 .5 &0 V.04o ■3V.a ,3

33QJ

20O<03

QJVO2s

Qj 2k2k 3cO "vjVOQj

11 Qj

-3C 5C c3 ) Cj Q» VJV. QJ4o? 3 *3.Q -3a E—j »o 3

0 VO-C 4oQ- 3

QJs1

"3 .c 4oQ33

■C04-j £4-1 V)

33.3

2Si

4oVO0a

|C

-O5

.3cc o 3

•52 ° O -* XZ "i

O Qj 3 . .3

VOQjo c t ;3 ? .£S ' J «N^ ^ O 3 73 **i-Si C -C3— O C .O Qj 3t -S ^

-N 8 E«N 3 o

8 3 3 a o s.5 * I'M QjU. r-H 3 .

110

\feria

bes/

Load

ngs

Plot

for

test

.txt

CMo

I, x JOi xipoA 6ay

CUD0)cn

2l/D

to

3 O

O) -c£ Tf L E CM

< u 1Q O - 1

— I 0 < Q- > O

' toM’ -ic c °•5 qj

« 5-+-> 58 Si■t* 3Si -c~Q *s 2si4o -^ T to 4 0o toa «c

o5

M1 r-T «N f \ "3 Pvj

§ Ka. 3 + £ 3 • £4 o JOto * -o o a o.to V .

3 4 oO tJ - C Q>

Co' o<N to^ Jf1Qj3 C £ O

O Sivi o c >o +-

£ £ £

o' 8S ^u QJQj - Q

c - * •2 § 8 «■ QJ T 3

O SiQJ Q

5 * *Q c 53 § § . E.to Qj 4 o

V . c

3

3

SiQjtoSic

cQj

-C Q.K- -Q+ Jc toQj

3:o 4oc c

a 3E E .Ccu oto E

.to o*3 3 ~3to O QJSi E 533

34o 3 )

Cr 4oIS) C 4 o

4 o*/*C3

QJ

1cQjtoQjQj 3 2!

Qj 3 ."c V .

4 o Sir* ET QJ

toVJQ» 4■

O-C

-O 3 4o

QJ

*1<N Qj

4 o3

~3«N O Qj«c 4o

Si a 3*■» to u■3cn£

O■cQ. C

111

The nature of this CA-4-P susceptible model could mean suppression of vital low abundant

peptide information.

2.4.8 Region of interest (ROI) study for the analysis of tumour rim and necrotic regions within fibrosarcoma 120 tissue.To enable further examination of tryptic digest spectra from MALDI images acquired using the

fibrosarcoma 120 tumour set a ROI study was performed. ROI's were exported from the

tumour rim and necrotic regions using the H&E tissue viable and necrotic areas for image co­

registration. Straightforward selection of 'central necrosis' and Viable rim' regions was not

possible as the complex patterns of necrotic infiltration required a mixture of microscope and

whole scanned image analysis.

Peptide induction was observed via PLSDA and Variable Importance in Projection (VIP) scores

were used from PLSDA regression modelling to determine the importance of variables within

the data time course.

HDI software (Waters Corporation, Manchester, UK) was used to view the MALDI-MSI for the

following statistics. ROI's were selected according to areas either being necrotic or directly on

the tumour rim for export to WatersMassLynx™ Software. After using the automatic peak

detection processing option the text files were then imported intoMATLAB® where predictive

statistical models were built.

112

Proc

ess

Anal

ysis

Galle

ry

Con

vert

L *

X #

0

3CT>u :.cc£O

< 0

3+j

OQ.«C<NC

.o•Cu0 )v*v.3O

3

5 C§ - 2 i ; + - O O£ C 8 $

§ 5 iv«-, qj° ^ Qj5 o i o o< JAO

-s*

Q 9-

<31 ^ .C .£

I !

=) oQj <3i43 Cv> !-~5 ■?

c C J2 o

.2 p4-J Qj

&<32Qj4«j

IQjv.C

£

§voQa:<N<N

Qj

V)

<3iQjv .

Qj

QJ

QJ

r\j <3Q j .V J

3 <N .31 «Ni£ csj

(siunoo) Ajisudji'l

113

Var

iabl

es/L

oadi

ngs

Plot

for

Nec

rotic

_rim

_V12

0_20

000_

peak

s.tx

t

CL-cK-CL4 V.ist;0 3VJ•c COs t 3<N 5»4o ■3coOa

C33

c 5o OS tSi

§ :-ccl'•C

23 t3

VJ ECLC 3

*3©CN s

£3E

QE8

oQJE >

3 43co2

-O

3VJ"3.E

CO32 3CL CL

3 .10 -CC 4->o |3 )

.c8.VJ"43 0O 3 .EVJ CLCL SC ■3

2V.34o o

c cQ E

O 2OCN o

35

CL•E

o 4 -.

88

cCLCL

o £V.-O

4-1CL

^ -Q3i Cc oI 43

3 Eo VJ o■cvo iSCO

"43VJ•e CO CL_o

a3VJ

COCL

V.O £ , 4i4-1VJCL

3"5 . 3,vo 4-1

C ■3 COo.o

jovoCL

4-1oa

a.83

V .3)0)

E

2O

-EV. VJ

CL S tCN

Q CL3 .O -C

4->Q. COCo .E

CLV. c•N 3) .3CN 8 43

VJCL 3

e v-33 ) O

-o4-1COCL

i l 3 "3

9 9I a JOj-iopaA 6ay

114

N-c

him

aeri

n

§o -cco

QJco.C■oQjs o oN

s■CPM+->COOQ.coC

.o&Qj

5 ■a s+-CJQjaCOCO3 2 5COC O 6 8!

S -S3 a0 g*M Q.^ rt C gE 'SP o * £

1 2-Q to ^ r?c ^I ^5 5■S t

r>■I ■=Q. Qj V. Qa s£ Ic S .2 oCO 4-a IOl £ Qj O

s ?p53 ^Q- 55• • CO

r$ §.N tQj Qj 3 °-•5» Jiu. t :

LOrHrH

9I, a JO) JO pa a 6ay

The zoomed insert in Figure 2.24 shows a clear separation between control and 24h post CA-4-

P.

To examine this data further the follow ing VIP scores show the variables w ithin the

fibrosarcoma 120 ROI study that are deemed most important.

Rhoprote in

Variables/Loadings Plot forNecrotic_rim_V120_20000_peaks.txt150

1028.5925

1844.5084 VIP scores for Control tum our rim

H2A944.5335

100

1489.8930>-

coQ_> 763.4 )18 Increased

abundanceof Rho GTPase activating prote in 2

1325.7394 ,1529.74721820.11530.7177

861.2214

1.6887945.!.4035(42.50 )(11326.7450

75.69751198.69?

32.5862 12'>9.5886

,121(1.70

S0p6.426212O1.65;

782. 5089 1821.8392(1842.84001531.7265725;. 1347.766!1843.8267

1348.77! ,7390 t.1' ^55373^

1000 1400 1600200 400 600 800 ' 1200 Variable

2200'1800 2000

Figure 2.25: VIP scores for the Control fibrosarcoma 120 tumour rim regions. The Rho protein N-chimaerin is indicated at m/z 844, Histone 2A at m/z 944 and actin aortic smooth muscle at m/z 1198. These data is in good agreement with the increased ion signal for m/z 844.5 displayed in earlier results for this protein.

116

Variables/Loadings Plot for Necrotic_rim_V120_20000_peaks.txt

VIP scores for 24 hours post CA-4-P tumour rim

944.5335

H2A

Rhoprotein

>-

0oHb Decreased

abundanceof Rho GTPase activating protein 2

: 1530.7177

945.5380

881 5(j

\asm.Variable

Figure 2.26: VIP scores for the 24h post CA-4-P fibrosarcoma 120 tumour rim regions. The Rho protein N-chimaerin is indicated at m/z 844, Histone 2A at m/z 944 and actin aortic smooth muscle at m/z 1198 and a Hb peaks at m/z 1529. The above results reflect the decrease in m/z 844.5 evident in results shown previously.

The VIP score plots provide an interesting compliment to the data shown in the PLSDA

regression vector plots. Figure 2.25 illustrates the ion at m/z 844.5 as a key peptide in the rim

region, a response echoed in the peptide's spatial distribution in Figure 2.19a and 2.19b.

In contrast to this Figure 2.26 displays the contrary showing Hb and markers of necrosis as the

key players influencing variability, with marked decrease of importance for N-chimaerin.

117

2.4.9 Concluding RemarksThe potential for investigations into protein induction following drug treatment strategies is

phenomenal. Currently one major challenge is the management and analysis of the large

amounts of data generated. Considering each of the peptide mass fingerprints (Figures 2.2 and

2.8), on first inspection these seem to be very noisy, however this could be a signature of a

successful tryptic digestion. Therefore the many low-abundance peptide signals contained in

the apparent noise could house evidence of significant proteins such as heat shock proteins in

addition to the expected and readily visible abundant peptide signals (actin, histones etc.). Due

to the nature of these results one has to question whether or not each peak is worthy of

further scrutiny, or should the vast amount of peaks be simply treated as noise/interference.

The use of ion mobility separation and its associated data extraction software certainly

appears to ease this dilemma making many more signals visible for subsequent examination by

MS/MS. In this work the suspected pharmacological response of haemorrhaging to treatment

with CA-4-P was investigated (Figure 2.3). The MALDI-MS images generated from m/z 1274 in

Figure 2.3 show a clear increase in Hb with time after treatment with CA-4-P. The latter is then

confirmed in Figure 2.4 in the H&E sections. At 6 h after CA-4-P treatment, the tumour in

Figure 2.4 was partially necrotic and Hb was evident in both necrotic and viable tumour

regions. The tumour excised at 24 h was almost completely necrotic. The increase in

haemoglobin is also clearly visible in Figure 2.7(c), the time course plot extracted from the

PCA-DA data, where again the increase in m/z 1274 and m/z 1820 can be clearly observed.

Many other peptides were observed in this work, including some tentatively identified as

arising from VEGF isoforms other than VEGF120 Figure 2.6e. This mostly likely could be due to

infiltration of immune cells (McDonnell et al 2012) especially after further investigation into

the VEGF 188 peptide 'FKSWSVHCEPCSERR' which confirmed that it was present in the exon 6

region known exclusively to VEGF 188. This influx and subsequent induction of certain

chemokines is a studied response to tumour vascular targeted therapy. In the case of the VEGF

120 signal at m/z 1485, it was found to be vital to cross check protein database outputs and

refer back to the original publication detailing the sequencing of that particular gene to be sure

analysis of the correct peptide sequence information.

In PCA-DA of complex data sets, the scores plot (in this case Figure 2.7a) represents the

variance of the original variables, i.e. the obtained sample groups. The loadings plot (in this

case Figure 2.7b) describes the variable behaviour and differences between the observed

groups. The PCA-DA of tryptic digest spectra from fibrosarcoma 120 tumour tissue reported

here shows good grouping of the tumour time points (Figure 2.7a). Confirmation of the

118

increase in haemoglobin due to CA-4-P treatment is given by the association of haemoglobin

ions with the 24 h tumour in the loadings plot (Figure 2.7b) and this is shown further in Figure

2.7(c). Further analysis of this large data was required. The aim of this was to examine whether

changes in signals arising from proteins of lower abundance provided an insight into other

pharmacodynamic responses to the treatment.

After the discovery via Mascot search of protein N-chimaerin (Figure 2.16) (Rho GTPase

activating protein 2), spectra in Figure 2.18 and MALDI images in Figure 2.19 depicted the

proteins decreased expression over time and peripheral spatial distribution in the early time

points. The latter could be due to interference by CA-4-P. Driftscope plots (Figure 2.17) helped

to build evidence of its response throughout the time course of experiments. Statistical

analysis using MATLAB® confirmed the earlier findings from this chapter namely reflecting

haemorrhaging, necrosis and N-chimaerin responses.

MALDI MSI has been used to study the effect of treatment with CA-4-P, a vascular disrupting

agent in late stage clinical trials, on mouse VEGF120 tumours. A strategy incorporating "on-

tissue" tryptic digestion, MS/MS identification peptides and ion mobility separation to improve

specificity was employed. By taking tumour samples at a number of time points after

treatment, gross effects were clearly visible through changes in the expression of certain

peptides. These were identified as arising from haemoglobin and indicated the disruption of

the tumour vasculature. It was hoped that the use of PCA-DA, PCA and PLSDA would reveal

more subtle changes taking place in the tumour samples however at present these are masked

by the dominance of the changes in the haemoglobin signals and other proteins of high

abundance.

The MALDI-IMS-MS/MS reported here was performed manually by direct peptide/ protein

profiling. The numerous signals from the PMF's generated most likely contain many peaks

which have been post translationally modified thus hampering peptide identifications. The

MS/MS spectrum in Figure 2.15 is an example of this, clearly containing overlapping MS/MS

fragmentation patterns. Therefore the experimental work reported in the following chapters

which describes similar experiments conducted using a CA-4-P resistant mouse model explores

not only MALDI-MSI but the use of conventional proteomic techniques to compliment and aid

interpretation of the MALDI-MSI data.

119

ReferencesAbdollahi A and Folkman J (2010) Evading tumour evasion: Current concepts and perspectives of anti-angiogenic cancer therapy. Drug Resistance Updates. 13.16-28

Aznar S, Fernandez-Valeron P, Espina Carolina, Lacal Juan Carlos (2004) Rho GTPases: potential candidates for anticancer therapy.C ancer Letters.206.181-191

Aznar S and Lacal JC (2001) Searching new targets for anticancer drug design: the families of Ras and Rho GTPases and their effectors.Progress in Nucleic Acid Research and Molecular Biology.67.193-234.

Djidja M-C, Claude E, Snel MF, Scriven P, Francese S, Carolan V, Clench MR (2009) Maldi-ion

mobility separation-mass spectrometry imaging of glucose-regulated protein 78 kDa (Grp78) in

human formalin fixed paraffin-embedded pancreatic adenocarcinoma tissue sections. Journal

of Proteome Research.8. (10) 4876-4884

Eigenvector Research Inc (2012) About PLS Toolbox and Solo, available at URL:

http://wiki.eigenvector.com, Last accessed 18/11/12

Galbraith SM, Maxwell RJ, Lodge MA, Tozer GM, Wilson J, Taylor NJ, et al. (2003)

Combretastatin A4 phosphate has tumor antivascular activity in rat and man as demonstrated

by dynamic magnetic resonance imaging. Jornal of Clinical Oncology. 21. 2831-2842

Groseclose MR, Andersson M, Hardesty WM,Caprioli RM (2007)ldentification of proteins

directly from tissue: in situ tryptic digestions coupled with imaging mass spectrometry. Journal

of Mass Spectrometry.42. (2) 254-262.

Kanthou C and Tozer GM (2007) Selective destruction of the tumour vasculature by targeting the endothelial cytoskeleton.Drug discovery today: Therapeutic strategies. 4. (4) 237-243

Kanthou C and Tozer GM[2009)Microtubule depolymerizing vascular disrupting agents: Novel therapeutic agents for oncology and other pathologies. International Journal of Experimental Pathology 90 (3) 284-294

Karlsson R, Pederson ED, Wang Z, Brakebush C (2009) Rho GTPase function in tumorigenesis. Biochimica et Biophysica Acta - Reviews on cancer. 1796.(2), 91-98

Kozma R,Ahmed S, Best A, Lim L (1996) The GTPase-Activating Protein n-Chimaerin Cooperates with Racl and Cdc42Hs To Induce the Formation of Lamellipodia and FilopodiaVol. Molecular and Cellular Bio!ogyl6.(9) 5069-5080

Lemaire R, Desmons A, Tabet JC, Day R, Salzet M, Fournier I (2007) Direct analysis and MALDI imaging of formalin-fixed, paraffin-embedded tissue sections.Journal ofProteome Research6. (4) 1295-1305.

Lemaire R, Menguellet SA, StauberJ, MarchaudonV, Lucot J, Collinet P, Farine M, Vinatier D, Day R, Ducoroy P, Salzet M, Fournier I (2007) Specific MALDI Imaging and Profiling for

Biomarker Hunting and Validation: Fragment of the 11S Proteasome Activator Complex, Reg Alpha Fragment, Is a New Potential Ovary Cancer Biomarker. Journal of Proteome Research.6. (11) 4127-4134.

Leve F and Morgado-Di'az JA(2012) Rho GTPase Signalling in the Development of Colorectal Concer.Journal of Cellular Biochemistry Volume.113. (8)

Li X, Zhang K, Li Z (2011) Unfolded protein response in cancer: the Physician's perspective. Journal of Hematology & Oncology.4. (8) 1-10.

Lord Q and Ashworth A (2010) Biology-driven cancer drug development: back to the future. BMC Biology.8. (38) 1-12

McCombie G, Staab D, Stoeckli M, Knochenmuss R (2005) Spatial and spectral correlations in MALDI mass spectrometry images by clustering and multivariate analysis. Analytical Chemistry.77. (19) 6118-6124

McDonnell LA, Corthals GL, Willems SM, Remoortere van A, Zeijl van RJM, Deelder AM (2010) Peptide and protein imaging mass spectrometry in cancer research. Journal of Proteomics.73. (10) 1921-1944.

Modolo MF, Tillmann Biz M, Machado de Sousa S, Fachinelli RL, Crema VO (2012) Immunohistochemical expression of Rho GTPases inameloblastomas. Journal of Oral Pathology and Medicine.41. (5) 400-407

Prise VE, Honess DJ, Stratford MRL, Wilson J, Tozer GM (2002) The vascular response of tumor

and normal tissues in the rat to the vascular targeting agent, combretastatin A-4-phosphate, at

clinically relevant doses. International Journal of Oncology. 21. 717-26.

Schmitz AAP, Govek EE, Bottner B, Aelst LZ (2000) Rho GTPases: Signaling, Migration, and Invasion.Experimental Cell Research. 261. (1) 1-12

Schwamborn K, Krieg RC, Reska M, Jakse G, Knuechel R, Wellmann A (2007) Identifying prostate carcinoma by MALDI-lmaging.\nterr\at\ona\ Journal of Molecular Medicine.20. (2) 155-159.

Shimma S, Furuta M, Ichimura K, Yoshida Y, Setou M, (2006) A Novel Approach to in situ Proteome Analysis Using Chemical Inkjet Printing Technology and MALDI-QIT-TOF Tandem Mass Spectrometer. Journal of Mass Spectrometry Society Japan.54. (4) 133-140.

Siemann DW (2010) The unique characteristics of tumour vasculature and preclinical evidence for its selective disruption by tumour-vascular disrupting agents. Cancer Treatment Reviews.37. (1) 63-74

Siemann DW and Horsman MR (2009)Vascular targeted therapies in oncology. Cell and Tissue

Research.335.(l) 241-248

Thorpe PE, Chaplin DJ, Blakely DC (2003) The first international conference on vascular targeting: Meeting overview. Cancer Research.63.1144-1147

121

Tozer GM, Kanthou C, Lewis G, Prise VE, Vojnovic B, Hill A (2008) Tumour vascular disrupting agents: combating treatment resistance. The British Journal of Radiology.81. S12-S20

Trim PJ, Djidja MC, Muharib T, Cole LM, Flinders B, Carolan VA, Francese S, Clench MR (2012)

Instrumentation and software for mass spectrometry imaging—Making the most of what

you've got. Journal of Proteomics.75. (16) 4931-4940

Vega FM and Ridley AJ (2008) Rho GTPases in cancer cell biology.FESS Letters.582. 2093-2101

Zhi-chao SI, Jie L (2008) What 'helps' tumours evade vascular targeting treatment?Chinese Medical Journal.121. (9) 844-84

122

Chapter 3Protein induction post CA-4-P

administration in a resistant fibrosarcoma 188 mouse model

123

3.1 IntroductionIn this Chapter data from the analysis of a mouse fibrosarcoma model tumour genetically

engineered to solely express the insoluble matrix bound VEGF 188 isoform is described. In

contrast to the soluble VEGF 120 isoform type that were the focus of Chapter 2, which have

been found to be susceptible to VDA , the VEGF188 isoform type have been reported to be

resistant to treatment with VDA (Tozer et al 2008) . In bone development, VEGF 120 and

VEGF 188 both have d ifferent roles in vascularisation (Maers et al 2004). VEGF 188 is amongst

the matrix bound isoforms thought to be involved in vascularisation of the metaphyseal bone

regions, whereas the soluble VEGF components are concerned w ith the epiphyseal area. It is

stated that the various isoforms of VEGF all play unique roles in the proliferation and

chondrocyte survival in a hypoxic environment (Maers et al 2004).

Capitalfemoral

epiphysis

Epiphysis

Growth plate

Metaphysis ' MMG 2mi5

Figure 3.1: Diagrammatical cross section of the region present at the end of growing long bones showing the areas involved in epiphysis and metaphysis. (Ortho paediatrics group (2012) (http://www.orthopediatrics.com/docs/Guides/slipped_CFE.html)

The term fibrosarcoma describes a malignant fibrous mesenchymal neoplasm which is said to

be found in the femur, humerus, jaw bone and is sometimes present in organs/ soft tissues

(Encyclopaedia Britannica Inc 2013),. Angiero et al 2007).

In tumours VEGF is known to be a potent angiogenic stimulator and levels are found to be

elevated in both the tumour cells and stroma (Harris et al 2012, Saharinen et al 2011). The

tumour microenvironment is said to promote the recruitment of inflammatory cells due to

chemoattractants. These actions are said to result in such cells becoming a major component

124

of the actual tumour mass (Saharinenet al 2011). Pro- and/or antitumorigenic functions can

result as a consequence of immune cell infiltration which is largely tissue/ tumour dependant.

Tumour associated myeloid cells (TAMs) are known to secrete numerous factors which favour

angiogenesis some of which include; VEGFs, TNF-a, interleukin (IL)-lb, matrix

metalloproteinases, PDGFB, basic fibroblastgrowth factor.

In the data to follow using fibrosarcoma 188 tumours, a further time point of 72h post CA-4-P

was added based on the more resistant nature of this matrix bound VEGF isoform model

(Tozer et al 2008). The rationale for this decision was to try to observe any response which

could be indicative of drug resistance and a potential switch back to tumour viability.

The use of CA-4-P treatment could enhance knowledge of the proteins induced in tumours

expressing either fibrosarcoma 188/ 120 isoforms under hypoxic conditions. The following

data aims to provide proteomic comparisons throughout a treatment time course using

fibrosarcoma 188 models, and allow observational differences to me made between the

expression of proteins in the resistant fibrosarcoma 188 and susceptible 120 tumour models.

125

3.2 Materials and Samples

3.2.1 Chemicals and Materialsa-Cyano-4-hydroxycinnamic acid (CHCA), aniline (ANI), ethanol (EtOH), chloroform (CHCI3),

acetonitrile (ACN), octyl-a/b-glucoside (OcGIc), tri-fluoroacetic acid (TFA), ammonium

bicarbonate, haematoxylin, eosin, xylene and DPX mountant were from Sigma- Aldrich

(Dorset, UK). Modified sequence grade trypsin (20 pg lyophilised) was obtained from Promega

(Southampton, UK).

3.2.2 Tissue samplesMice were injected sub-cutaneously in the flank with a 50 pi tumour cell suspension containing

1 x 106 cells in serum-free medium. The cells employed in this study were from the mouse

fibrosarcoma cell line, VEGF188. This has been engineered to express only the VEGF188

isoform. Tumours were allowed to grow to approximately 500 mm3, before CA-4-P treatment

(a single dose of 100 mg/kg i.p). The dosage for CA-4-P are consistent with animal studies

performed at clinically relevant doses (Galbraith 2003., Prise 2002). Mice were sacrificedand

tumours excised at various times after treatment. All animal work carried out documented

herein was performed by Dr. J. E. Bluff, Tumour Microcirculation Group, University of Sheffield,

UK. Samples were provided as frozen excised tumours and stored at -80°C.

3.2.3. Experimental groupsControls (no treatment, saline i.p), n = 4 (labelled tumour 6_1 - tumour 6_4), C-A4-P (0.5 h

after treatment), n = 5, (labelled tumour 7_1 - tumour 7_5), C-A4-P (6 h after treatment), n =

5, (labelled tumour 8_1 - tumour 8_5), C-A4-P (24 h after treatment), n = 5, (labelled tumour

9_1 - tumour 9_5), C-A4-P (72 h after treatment) n = 4, (labelled tumour 10_1 - tumour 10_4).

3.2.4 Tissue preparationFrozen tissue sections were cut to give approximately 10pm sections, using a Leica CM3050

cryostat (Leica Microsystems, Milton Keynes, UK) (Djidja et al 2009). The sections were then

freeze thaw mounted on poly-lysine glass slides. Mounted slides were either used

immediatelyor stored in an airtight tube at -80 °C for subsequent use.

126

3.2.5 In situ tissue digestion and trypsin depositionThe tissue samples in triplicate to undergo MALDI-MS/MSI were washed initially with 70% and

then 90% ethanol for 1 min then left to dry, subsequently slides were immersed in chloroform

for 10 s. Prior to matrix application, in situ tissue digestion was performed with trypsin solution

prepared (from lyophilised trypsin) at 20 pg/ml by addition of 50 mM ammonium bicarbonate

(NH4HCO3) pH 8, containing 0.5%octyl-a/b-glucoside (OcGIc) based on a protocol by Djidja et al

(2009).

The "Suncollect" (SunChrom, Friedrichsdorf, Germany) automatic pneumatic sprayer was used

to spray trypsin in a series of layers. The sections for MALDI-MS and MALDI- MSI were

incubated overnight in a humidity chamber containing H20 50%: methanol 50% overnight at

37°C and 5% C02.

3.3 Methods and instrumentation

3.3.1 Matrix depositionThe matrix, a-cyano-4-hydroxycinnamic acid (CHCA) and aniline in acetonitrile:water:TFA

(r.l:0.1 by volume), was applied using either the Suncollect (at 5 mg/ml) or the Portrait™ 630

Multispotter (at 10 mg/ml), as described above. Identical coordinate settings were used as

with the trypsin deposition, to ensure sample uniformity. Equimolar amounts of aniline were

added to the CHCA solution, i.e. 1 mlof 5 mg/ml CHCA solution contained 2.4 pi of aniline.

These matrix deposition parameters were based on methods from Djidja et al (2009).

3.3.2 InstrumentationMALDI- IMS/MS, MALDI- IMS/MSI and MALDI- IMS-MS/MS were performed using a HDMS

SYNAPT™ G2 system (Waters Corporation, Manchester, UK) and Drift scope 2.1 software

(Waters Corporation, UK). Instrument calibration was performed using standards consisting of

a mixture of 217 polyethylene glycol (Sigma-Aldrich, Gillingham, UK) ranging between m/z 100

to 218 3000 Da prior to MALDI-IMS-MSI analysis. In order to achieve good quality MS/MS

spectra, they were acquired manually moving the laser position andadjusting the collision

energy to achieve good signal to noise for product ions across the full m/z range of the

spectrum.Collision energies were adjusted from 70 to 100 eV during acquisition and

acquisition times were generally of the order of 5-10 s per spectrum. MS/MS spectra were

uploaded to perform a Mascot (Matrix Science, London, UK) search which used the UniProt

database in order to generate a sequence match.Image acquisition was performed using raster

imaging mode at 30-100 pm spatial resolution, Biomap 3.7.5.5 software (http://www.maldi-

127

msi.org/) was used for image generation. To enable simple visual comparison between images

all data were normalised to m/z 877/ m/z 1066 (peaks arising from the aCHCA matrix).

The LC-ESI-MS/MS analysis of prepared tissue digests was performed by Adam Dowle at the

Proteomics Technology Facility, Department of Biology (Area 15), University of York. LC-ESI-

MS/MS analyses were performedwith a Bruker nanoESI source fitted with a steel needle using

Ion spray voltage of 1500V. MS/MS was acquired using the following Auto MSMS settings: MS:

0.5 s (acquisition of survey spectrum), MS/MS (CID with N2 as collision gas): ion acquisition

range: m/z 300-1,500,0.1 s acquisition for precursor intensities above 100,000 counts, for

signals of lower intensities down to 1,000 counts acquisition time increased linear to Is, the

collision energy and isolation width settings were automatically calculated using the

AutoMSMS fragmentation table:, 5 precursor ions, absolute threshold 1,000 counts, preferred

charge states: 2 - 4 , singly charged ions excluded. 1 MS/MS spectrum was acquired for each

precursor and former target ions were excluded for 30 s. The data obtained from the

University of York was in the MASCOT .data file format. The .dat files selected used the

following modification searches and parameters using the MASCOT Mus (Musculus) protein

database:

Fixed modifiation: Carbamidomethyl (C) and variable modifications: Acetyl (Protein N-term), Gln->pyro-Glu (N-term Q),Glu->pyro-Glu (N-term E),Oxidation (M).

Massvalues: Monoisotopic, protein Mass : Unrestricted, peptide Mass Tolerance : ± 10 ppm, fragment Mass Tolerance: ± 0.1 Da and max Missed Cleavages : 1

These data were then processed using Scaffold 3 proteomic software tool for visualisation and

analysis of the LC-ESI-MS/MS data (http://www.proteomesoftware.com/). The data files (.dat)

produced from MASCOT which corresponded to each digested analysed, were uploaded

individually. Analysis with X! Tandem was selected in order to improve protein identifications

with searching through an additional database.

3.3.3 Haematoxylin and eosin stainingSlides were immersed in haematoxylin for 1 min, rinsed in tap water until the water ran clear,

immersed in 1% eosin for 30 s and rinsed in tap water until the water ran clear. They were

then dehydrated as follows: 50% ethanol for 2 min, 70% ethanol for 2 min, 80% ethanol for 2

min, 95% ethanol for 2 min and 4 changes of xylene applied to each slide for 1 min at a time.

Finally, they were mounted with DPX mountant and left to dry in the fumehood overnight.

128

3.3.4 Statistical analysisPCA and PLSDA was performed using MATLAB® (Matrix Laboratory) (MathWorks, Inc., Natick,

MA 486 USA). The PCA/ PLSDA data that follows is representative of the fibrosarcoma 188 data

where either n= 4/5 biological replicates per time point with up to 6 technical replicates for

each, (control - 72 hours). PLSDA Data for ROI study using HDI software (Waters Corporation,

Manchester, UK) was analysed using one tumour per time point with 9 technical replicates per

ROI. Variable Importance in Projection (VIP) scores were used from PLSDA regression

modelling to determine the importance of variables within the data time course.

3.3.5 Data pre-processing using Waters MassLynx™ Software and MATLAB®MS results wereimported into MATLAB® in .txt or .csv format after application of "automatic

peak detection" using the instrument data processing software (Waters MassLynx™ Software).

Normalisation (2-Norm) and mean centre were selected, "contiguous blocks" was used for

cross validation.

3.3.6Protein network analysisAccession lists generated by results from the LC-ESI-MS/MS Mascot searches were imported

into the STRING 9.0 database (from Mus (Musculus) MASCOT protein database). Observations

of the relationships were made between the proteins indentified throughout a fibrosarcoma

188 time course. The sample data was used to build predictive proteomic pathways and study

the predicted functional partners of known protein-protein interactions (http://string-db.org/).

3.3.7 Immunohistochemical staining

3.3.7.1 Chemicals a nd M a te ria lsMethanol, Acetone, Hydrogen peroxide solution 30% wt., Xylene, Ethanol, Gill's Haematoxylin

and DPX mountant were all purchased from Sigma Aldrich UK. Phosphate buffer saline tablets

(Dulbecco 'A' Tablets) were from Oxoid Ltd. Normal goat serum, ImmEdge hydrophobic barrier

pen, 10X Casein solution, Avidin/Biotin horseradish peroxidise complex blocking kit, ABC

solution kit, Diaminobenzidine (DAB) substrate kit were from Vector Laboratories Ltd UK. HSP-

90 (abl3494,16F1) and Plectin (ab32528, E398P) antibodies were from Abeam, Cambridge UK.

Macrophage cell surface marker F4/80 (MCA497GA, CI:A3-1) antibody was from AbD Serotec.

129

3.3.7.2 Im m unoh is tochem ica l m ethodsMounted frozen tissue sections were allowed to equilibrate to R.T for 5 min. Slides were then

fixed in ice cold acetone for 20 min then rinsed in PBS. Endogenous peroxidases were blocked

using 30% H202 and Methanol. After rinsing in PBS, tissues were blocked for lhr using Goat

sera containing 10% Casein. After PBS washes, Primary Ab was added and left O.N at 4°C. To

aid optimisation of immunohistochemistry, Ab's were first diluted using dilutions ranging from

1:50 - 1:1600 depending on Ab and manufacturer's guidelines. Primary antibody dilutions

were; 1:400 for both Plectin and F4/80 and 1:2000 for HSP-90. Isotype IgG controls were used

for F4/80 and for Plectin and HSP-90 primary antibody was just omitted a negative control.

After overnight incubation and subsequent PBS washes, Secondary Ab was added diluted

(1:200) in 2% Goat sera and left for an incubation time of lh at R.T. After this period, ABC

solution was added after rinsing in PBS and left to incubate for 45min at R.T. After PBS

washing, DAB solution was applied and left to allow development of the staining. Slides were

rinsed in tap water prior to immersion in Gill's Haematoxylin for 2min. After slide dehydration

using 70% -100% EtOH tissue was immersed in 2 changes of Xylene for 5 min each. Slides were

mounted using DPX mountant.

3.3.8 Protein precipitation and digestion

3.3.8.1 Chemicals and m a te ria lsChloroform (CHCI3), methanol (MeOH), acetonitrile (ACN), hydrochloric acid (HCI), tri-

fluoroacetic acid (TFA), ammonium bicarbonate, ,DL-Dithiothreitol solution, lodoacetamide,

urea, potassium di-hydrogen phosphate (KH2P04), TRIzol were from Sigma-Aldrich (Dorset,

UK). Modified sequence grade trypsin (20 pglyophilised) was obtained from Promega

(Southampton, UK,).

3.3.8.2Tissue hom ogen isa tion and p re c ip ita t io n o f p ro te inThe fibrosarcoma 188 tumour tissue was homogenised in 800pl of TRIzol solution using a

micro-homogeniser (Kline and Wu 2009). Homogenised solution was centrifuged (1,500 rpm)

for 5 min to pellet out nuclei/ unbroken cells. Post-nuclear supernatant was then centrifuged

(14,000 rpm) for 30 min. Resulting supernatant was discarded and cellular membrane pellet

was retained for protein precipitation. 200pl of MeOH and 50pl of CHCI3 were then added to

each protein pellet sample. After vortexing, 150pl HPLC H20 was added with further vortexing.

After centrifugation (14,000 rpm) for 2 min, the bottom CHCI3 layer was removed. A further

50pl CHCI3 was added, removal of the bottom CHCI3 layer was repeated following

130

centrifugation (14,000 rpm) for 2 min. Subsequent removal of the H20 layer resulted in the

remaining protein precipitate layer. CHCI3 (50pl) and MeOH (150pl) were added directly to the

protein precipitate, after vortexing solution was centrifuged (14,000 rpm) for 2 min.

Supernatant was removed and protein pellet was then allowed to air dry for 2 min.

3.3.8.3 P ro te in D igestionlOOpI of 0.1% RapiGest in 50mM NH4HC03 buffer (pH 7.8) was added to the air dried protein

pellet. Each sample pellet/ solution was consecutively vortexed, incubated at -80°C (~ lhr) and

heated to 70°C (1 min) until solubilised. Once fully solubilised the sample was heated to 100°C

(2 min) and then left to reach room temperature. Each sample was reduced with DTT (final

concentration 5mM) for 30 min at 60°C the left to reach room temperature. Solutions were

then alkylated with iodoacetamide (final concentration 15mM) in the dark for 30 min at room

temperature. Sequence grade modified trypsin was added (20pg/ml) to 80pg of protein

following the BCA Protein Assay (see 3.3.9) used for protein determination. In-solution digests

were incubated over night at 37°C with shaking.

3.3 .8A P repa ra tion o f sam ples f o r co lum n lo ad ingFibrosarcoma 188; HCL (final concentration lOOmM) was added to the overnight digest,

solution was then incubated for a further 45 min at 37°C. After this step the sample was then

centrifuged (14,000 rpm, 4°C) for 10 min, the supernatant removed for analysis. This

remaining solution was lyophilised and stored at -80°C until further use.

3.3.9 Protein estimation - BCA assay

3.3.9.1 Chemicals and m a te ria lsBCA protein assay reagent (bicinchoninic acid), copper (II) sulfate pentahydrate 4% solution,

protein standard solution, 1.0 mg/mL bovine serum albumin (BSA) were from Sigma Aldrich

UK. RapiGest detergent solution was purchased from Waters (UK).

3.3.9.2BCA m ethodSolubilised protein pellets were thawed, vortexed and centrifuged ready for the BCA assay.

BSA standards were prepared within an analytical range of 0-4 mg/ml. BCA reagent was added

to BCA standards and tumour tissue solution samples and left to incubate at R/T for 30min.

BCA standards were measured in triplicate and Samples were measured in duplicate in a 96

well plate using a Wallac plate reader (spectrophotometer) at 570nm.

131

3.4 Results and Discussion

3.4.1 MALDI-MSI of fibrosarcoma 188 CA-4-P treated tumour tissueThe MALDI spectra from both fibrosarcoma 188 and 120 on tissue digests contain numerous

peaks, the challenge so far has been exploration and identification of the low abundant species

present.

One of the main aims at this point of the project was to develop a system of techniques that

could be used in collaboration with each other in order to co-validate proteomic responses

observed within a treatment time course.

Proteins of interest possibly induced as a result of CA-4-P treatment were firstly visualised by

MALDI-MSI as putative protein targets. These responses therefore could then be supported by

immunohistochemical staining. Similarly protein responses observed employing conventional

proteomic techniques i.e. Liquid Chromatographic methods coupled with ESI or MALDI, were

then used to relate back to MALDI-MSI.

132

3.4.1.1 Trypsin application fo r the avoidance of protein delocalisationOne of the main potential pitfalls of in situ tissue tryptic digestion is protein delocalisation. A

factor which when occurs eliminates the analysis of spatial protein distribution. To ensure

good tryptic digest MALDI-MSI, a homogeneous coating is essential during trypsin application.

Other issues to consider would be the tissue type, section size and to ensure a suitable flow

rate is employed when using a pneumatic sprayer for trypsin deposition. Too much trypsin

sprayed per cycle could lead to 'over wetting' of the tissue and subsequent protein

delocalisation. The follow ing images were as a result of initial trypsin application optimisation

involving flow rates, cycles and gas pressure in order to avoid any issues previously mentioned.

File E d* LaycsA W in d o w Oisptav Roi Analysts lock He lp

« j ± | * | l lL i|BL*J«JJxj

0 0 0 0 - ■ ...........

- • - .....

* “• -= •

Pot 110 22 0

Scan iJbiJ&<&i2j00um

3 Plot Tod

File Ccnfiq

1 0 0 0 1 2 0 0 1 4 0 0 1 6 0 0 1 8 0 0

Portion [1795 97. 374 400}. 012449

File Layout Window Display Roi Analysis Tools Help

• 000000 J00 " 00

- ■ -

.. >■:: s. ....

E mm E2mm | I 00000 2<xvr.

Pm 10$ 0 0

Scan 8.5bL.2202l2_100um

1 4 0 0 1 6 0 0 1 8 0 0m/z _____ . .

Figure 3. 2: Screen shot of protein delocalisation in MALDI-MSI, visualised using BioMap imaging software. Figure 3.2 displays a screen shot of MALDI-MSI of an on tissue tryptic

133

digest where trypsin applied using a pneumatic sprayer has caused delocalisation of proteins as a result of an incorrect spraying speed. The trypsin therefore; has not had sufficient drying time and over wetting of the tissue has occurred. The cross hairs which map the ion signals in the top MALDI-MSI in Figure 3.2, show lots of tryptic peptide signals outside of the tissue, indicative that the spatial distribution of the proteins have changed. The signals given from the cross hairs in the bottom image in Figure 2 show a substandard tryptic spectrum as a result.

The MALD-MSI featured in Figure 3.3 can be regarded as an exemplar of on tissue tryptic

digestion. The characteristic contrast of discreet CHCA matrix clusters seen in the left hand

MALDI-MSI, compared to the tryptic PMF evident from the cross hair mapping on the tissue

image shown to the right.

The most suitable parameters for spraying trypsin were found to be the employment of a

series of 5 layers at a flow rate of 2.5pl/ minute. To avoid protein delocalisation and ensure

sample spray uniformity the regions set for trypsin application allowed enough drying time

between each spray pass. A fine spray was found to be essential with a view to maintain

protein positioning within the tissue and the nitrogen gas pressure used was ~15 psi for

spraying trypsin.

For matrix application, the same principles were applied but the gas pressure was decreased to

~10 psi. The flow rate was adjusted to 4pl/ minute to take into account of solvent evaporation

and aid sample coverage and co-crystallisation.

134

m p m

t w m m

135

The spectra in Figure 3.4 are typical representations of fibrosarcoma 188 peptide mass

fingerprints (PMF) throughout the CA-4-P treatment time course studied.

The spectra within the time course of Figure 3.4 all appear to be quite similar upon first

inspection. The zoomed in regions of the fibrosarcoma 188 time course now shown in Figure

3.5 shows subtle differences of the appearance of Hb at m/z 1416 and appearance of an

unknown at m/z 1577. The high abundant Hb peaks (m/z 1274, m/z 1302, m/z 1529 and m/z

1819) which are evident in the spectra from Figures 2.1 and 2.2 in Chapter 2, do not dominate

the fibrosarcoma 188 spectra. This could be due to the CA-4-P resistant nature of this tumour

mouse model.

136

4.n S H co - 1 h- CO lO

? r< Su I■ , o■H tol/> T-VO OJQ . CO .

rvj

cn\_

tp CM- u) o in in

^ ,- coZ. £ 3° Z0) cm *2

s E1 S

EDO■**

O O oin r * !

. in Q in

co csi5u_Oh- - \

CLI

I<U+-<inOQ .

JC

PM

Si-_. ^ : CT) 7

l llO

in oo if) CM CO

O n)O’ -J CMco csi2ILoh-

Q- S -i

<U+ jCOOQ .

JCVO

<=>/'

i : CM2 <

mCM £P OD

..saA-

%

—1 ooCO

: 2

I 7

-= : o 10■ L

"D7

CL

■ rin

.o1Q .

j :o

co 4 : o) :

J L=. :

i : o

-j : o- J - i n

: o:-a rg

o

§

£30)v.•Si

Q .■■

S■SimOa

- c

R"3Co

«c

<N

■CVO

-c "VOd

coVjQjv.3c

o■52•S iO<L

*

5s"3■SilO0)23

8QjE

*-Coooo

3Eo83mO

3

*S*v</>u .§a.

rr)(Ls.33li i

137

8.3OJ OO

a;+Q CO —I CD CO 2 LLO ’-V

■"Sf J

o> .c 53 -o

tS <©CO*-!

1 1J2coaoj

"O

+->aj<U+->c3

Co

u

S -

ol_ -rm

■r- 01

3 - i -

CT) J -CO -4 .Is- d ■

c r-I a ° . T\ -

-cR■QCQ

o «clo

«N<u

•c

'm

sc/>S

I

Sis«3)CU

-c

°roQu■5.c£

§cso>.3

-QCUiS«ou.

0000

Q IS8 ^ 8 ^ -O

O 4oc oO co•p- 01

Gv. O) .£ &

cSi ^ £ o

?! lh ccvi °^ *o* 1 3 O3i -Ou: a:

13

8

CO T“5LLO

CO T-

CM CM

L- s ® 01fe S 3 3

O oo

<0 CO

1sqjc0■Li,</>1Qj

■Li30

.C

1§3QJ3

CUiS£if1QjCn"6Qj3JO•S2■Cco29oo'-i73C3OPM35oa8a-Q

5£.ccS-Qj3)C</>V)3EQJ

73•C

8-aQJ3»o.*0

(Timr—I

Oi: °°C >Ns >O .c .52 C L. Qj 3 3a c 5<3

VOooa3

oI

73C3

»o*N•S' ooU_ «-l

3.4 .I.2 . S pectra l com parison o f pep tide mass f in g e rp r in tsFigure 3.6 displays the spectra from examples of control tumour tissue from each tumour

model, fibrosarcoma 120 and 188. The peak tentatively relating to the Rho GTPase protein

discussed in Chapter 2 can be seen in the fibrosarcoma 120 spectra (indicated by an arrow),

but not at the same intensity in the spectra corresponding to fibrosarcoma 188. Similarly, the

same can be deduced for the ion at m/z 1325 relating to Histone H4 seen in the 188.

Although further work is required to determine the relevance of increased Histone H4

intensity seen here, there have been many studies on the analysis of Histone H4 acetylation/

deacetylationwith the premise of histone deacetylase inhibitors as potential anti-cancer

treatment (Advani et al 2010). It is stated that expression of Histone acelylation in cancers can

influence patient prognosis. Reports claim that increased levels of acetylated Histone H4 in

patients with esophageal squamous cell cancer can correlate with an encouraging clinical

outcome (Toh et al 2007).

Ultimately, histones are said to be key in gene regulation (Grunstein 1997). In the 188 Control

spectra seen here (Figure 3.6), the prominent Histone H4 could be a characteristic of this more

drug resistant fibrosarcoma 188 tissue.

140

3.4.1.3 MALDI-MSI and immunohistochemistry to analyse a fibrosarcom a 188 treatm ent courseInitial MALDI-MSI using the fibrosarcoma 188 tissue was carried out at a resolution of 100pm

employing a spot to spot laser pattern with the SYPNAT G2 HDMS system (Waters, UK). The

gross haemorrhagic pharmacological response that already has been shown in Chapter 2 using

the fibrosarcoma 120 tumours was considered a good place to start when validating responses

seen by other ions. Figure 3.7 shows a multi time point MALDI image of a fibrosarcoma 188

time course ranging from Control - 72 hours post CA-4-P. The MALDI images were taken from

a slide containing multiple samples.

Switch back to viable tissue.

Corresponding to Hb m/z 1819

24 hours CA-4-P

Control/saline

72 hours CA-4-P

0.5 hours CA-4-P

50711 _V188_combined_678910_ma* div (MEAN: 1819.29-1819.49;

6 hours CA- 4-P

Figure 3. 7: MALDI-MSI multi image of a fibrosarcoma 188 time course. The images show the spatial distribution of a peptide of haemoglobin subunit alpha at m/z 1819 throughout a treatment time course from Control-7 2 hours post CA-4-P administration.

The MALDI-MSI shown in Figure 3.7 of the matrix bound VEGF 188 model shows a different

response to that seen of a Fib peptide in Figure 2.3 (Chapter 2). The haemorrhagic response is

observed much later in the 24h tissue opposed to the 6h sample in the fibrosarcoma 120

MALDI-MSI. In contrast to to the evidence of complete necrosis and vascular disruption in the

last time point of the fibrosarcoma 120 (Figure 2.3 - 24h), here the 72h time point exhibits a

very different response. There appears to be evidence of a switch back to viable tissue in this

more resistant tumour model.

141

Heat shock protein 90 (HSP-90) is ubiquitous in all normally functioning cells and serves to

prevent the misfolding of proteins by helping to retain the correct structural format of the

protein (Neckers 2007). The characteristic tumour microenvironment renders cancer cells to

anoxia, alterations in pH and attack via pharmacological anticancer agents. HSP-90 has been

suggested as a potential target protein for anticancer therapy due to the dependence on

structural conformity by the cancer cell (Den and Lu 2012).

It is known that inhibition of HSP-90 results in the degradation of the HSP-90 stabilised protein

(client protein) via the proteasome.

Elevated levels of HSP-90 in breast cancer patients have been found to correlate with poor

patient survival due to the conserving effect of HSP-90 on human epidermal growth factor

receptor 2 (Jameel et al 1992). It is also stated that HSP-90 is highly expressed in solid tumours

and it is suggested that HSP-90 plays a key role in the evasion of apoptosis thus promoting

tumour cell survival (Den and Lu 2012).

There are over 20 known clinical trials taking place which currently involve the use of HSP-90

inhibitors (Den and Lu 2012). A review article by Den and Lu (2012) poses the question of

which tumour type is the most suitable for HSP-90 inhibition study and whether or not the

anticancer target should be the HSP-90 'client protein' i.e. epidermal growth factor receptor in

non-small cell lung cancer/ human epidermal growth factor receptor 2 in breast cancer.

The MALDI images in Figure 3.8 display the spatial distribution and relative abundance of an

ion at m/z 1168 corresponding to HSP-90. The optical scans shown in Figure 3.6 show the

immunohistochemical staining of HSP-90.

142

Corresponding toHsp 90 m/z 1168 % -

24 hours CA-4-P -<.*h A / C ' C " j*> ‘ . V - . * * . *

. , v • **~w y * * }

;£ % w • : • V ;

Control/saline

0.5 hours CA-4-P

Switch back to viable tissue.

72 hours CA-4-P

250711 V IS3 combined 678910 man diw (MEAN: 1153.29-1163.49) 6 hours CA-4-P-,

Figure 3. 8: Fibrosarcoma 188 MALDI-MSI multi image showing spatial distribution of HSP-90. MALDI images showing spatial distribution of HSP-90 alpha at m/z 1168.

The presence of HSP-90 is shown throughout the fibrosarcoma 188 time course. The MALDI

image corresponding to 24hr post CA-4-P is indicative of complete HSP-90 saturation. This

response is matched by HSP-90 immunohistochemical staining seen in Figure 3.8.

The possible 'switch bock to viable tissue' is also exhibited by the 72h time point in Figure 3.8.

This potential evidence of tumourigenesis can also be seen in the immunohistochemical

equivalent in Figure 3.9.

143

Control/Saline

•v

$

0.5 hours post CA-4-P

\

24 hours post CA-4- P - complete Hsp90 saturation

Figure 3. 9: Immunohistochemical staining of a fibrosarcoma 188 time course. Evidence of complete HSP-90 saturation can be observed in the 24h optical scan with the indication of tumour tissue rejuvenation in the 72h tumour time point.

72 hours post CA-4-P

Switch back to viable tissue.

The microscopic images in Figure 3.10 were captured in brightfield 10X to enable closer

inspection of the immunohistochemical staining seen here in Figure 3.9.

J£2 -V? - 1 : .. W . ^ 'a ^

’

'o'cu"Ccuo*ocu3c£cocu3V)•3oo00

§3</)s-o

.cC3IQ.

s■&cn.c.c5

sEcu-cVJo-coc3EEo*i

Si3cn

145

Figure 3.10; (a) displays the negative Control for HSP-90the time point used was 24h post CA-4-P, (b) HSP-90 staining using fibrosarcoma 188 Control tissue/ (c) HSP-90 staining ofO.Sh post CA-4-P, (d) 6h post CA-4-P,(e) HSP-90 staining of24h post CA-4-P showing HSP-90 tissue saturation status and (f) HSP-90 of the saturated necrotic central region and viable tissue regional indicated by the bracket.

The HSP-90 immunohistochemical staining in Figure 3.10 was captured at 10X magnification to

allow closer inspection of the HSP-90 optical scans previously shown. The apparent saturation

of HSP-90 in the 24h time point visible in the MALDI-MSI and in Figure 3.9 can be seen in

Figure 3.10(e). The lOx image of HSP-90 showing the regional switch back to viability can be

clearly observed in Figure 3.10(f), the viable tumour cells are highlighted by the bracket.

146

3.4.1.4 High resolution MALDI-MSI using fibrosarcom a 188 tum our tissueThe MALDI-MSI featured here employed high resolution imaging with a view to focus down on

the tumour rim regions.

Tissue was imaged using vertical sliced regions which encompassed both the tumour rim and

central area.

Figure 3.11: SYNAPT G2 MALDI imaging plate indicating tissue image regions. The regions selected for MALDI-MSI analysis have been superimposed over the histological tissue sections.

The 40pm resolution image show in Figure 3.12 shows the spatial distribution of m/z 844.5

which corresponded to the putative Rho GTPase activating 2 protein mentioned earlier in

Chapter 2. Although not a dominant peak within the fibrosarcoma 188 spectra, the spatial

distribution of this ion is shown to be localised around the periphery of the tumour tissue in

the follow ing MALDI-MSI of a 72hr treatment time point.

147

"2 Ins CA-4 Ps iZ •

m/z 844.5 Rho GTP- ase activating protein

290711 8 8 _ 4 0 ijm _ 1 0_ I j t »a ■ j i i v (M E AN 84 4 510 -8 44 65 0 )

Figure 3 .1 2 :40pm MALDI-MSI of24h post CA-4-P fibrosarcoma 188 tumour tissue. Peripheral localisation of n-Chimaerin at m/z 844.5.

"2 lu t C’A -I p

(C)

2 9 0 7 1 (MEAN 944 470 9*4f.l0'j SL 1/1

Figure 3 .1 3 :40pm MALDI-MSI of m/z 944.5 in fibrosarcoma 188 tumour tissue. The differing spatial distribution of an ion is shown here corresponding to Histone 2A.

An image from the same acquisition shown in Figure 3.12 is displayed in Figure 3.13. The ion at

m/z 944.5 (Histone 2A) is located in the necrotic region of the tissue, emphasising the

distinction in spatial distribution between m/z 844.5 and 944.5. The necrotic region is

indicated by the H and E in Figure 3.14.

148

40umHistone2AXm/z944.5

• i*7k v ? ’ * 7*

Vi r ■ *

290711 '■'183 40urn 10 1

Figure 3.14: 40pm MALDI-MSI of a on tissue try tic digest and corresponding H and E histological section. A necrotic region is seen here to correlate with the spatial distribution of Histone 2A.

k Jk - «

LC_6_1_080911_30um_max (MEAN:1 198.74-119* LC_6_1 _0S0911 _3Dum jn5^_diV;^EAN:344.430-844.54Ci) SL 1/1

Figure 3.15: 30 pm MALDI-MSI of a Control fibrosarcoma 188 tumour on tissue tryptic digest. The spatial distribution of Actin (a) and Rho GTpase activating protein 2 (b) is shown in the 30pm MALDI-MSI.

30pm images (Figure 3.15) were performed to investigate the tumour rim and although images

appear slightly 'grainy', distribution of known peptides could still be visualised. This 'grainy'

appearance of the MALDI-MSI here could be an artefact of 'over sampling' due to the laser

spot size measuring ~60-70pm teamed w ith the 30pm high resolution laser spot to spot

pattern.

149

3.4.2 Statistical analysis of mouse fibrosarcoma 188 tumour tissue using MATLAB®Initial statistical analysis on the fibrosarcoma 188 tryptic digest spectra was performed using

Principle component analysis (PCA) via MATLAB® through the Eigen Vector tool box, to

produce an unbiased clustering of spectra acquired from MALDI-MSI of the fibrosarcomal88

data. As discussed in Chapter 2, there is a requirement for an element of discrimination in a

statistics model for known time course data. This considered, a preliminary PCA test was done

to predict potential groupings between time points. This PCA consisted of 9 technical

replicates from MALDI images acquired in triplicate to include an example of each CA-4-P

treatment time point. The results are shown in Figures 3.16 and 3.17.

Sarfles®OTES FH cf Vt88_100rrpe_n_3_9_repeefc.txf

Sccres cn FC2(22627o)

♦ 6.

* M *9 /o CmkBxe ted

r

-1 -08 -Q6 -04 -02 0 02 04 Q6 0.8 1Sccres a iP C 1 (3135%)

Figure 3.16: PCA score plot of fibrosarcoma 188 tumour tissue in situ tryptic digests. The score plot in Figure 3.16 shows the potential groupings between the tumour time points in this fibrosarcoma 188 study. From this preliminary PCA test there does appear to be a separation between tumour treatment groups with an apparent overlap observed here in the Control (6:4) and 0.5h post CA-4-P (7:2)

150

VariaHes/Loadngs HotforV188_1example_n_3_9_nepeats.txt

•861.32^2

•878.^420

•862.5570 *1348,8272

•757.4354 V o/A ,„ . . . ,(a ) • 1326.9014

J J&k*7856*,* ’ g ^*|| t *844.4536

*1276.9177 * aacn

* . *1327.6876

_Vi$.jea(f,2j?^,, { * *i4fe8qp6', ' * 1 41851.1325 * *873.4352

•1199.to8o* . 1476.9374

(b)

(c)

/

-0.1 0 0.1 Loadngs on PC 1(31.37%)

Figure 3.17: PCA loadings plot fibrosarcoma 188 tumour tissue in situ tryptic digests, (a) corresponds to a peak relating to N-Chimaerin mentioned earlier, (b) Histone H4 and (c) Histone 2A.

The loadings plot contains the peak information which spatial relate to the groupings seen in

Figure 3.17. The loadings plot seen here does not appear to exhibit the Hb dominance that was

visible in the PCA loadings plot featured in Figure 2.20, Chapter 2.

The region in Figure 3.17 that spatially relates to the 72h post CA-4-P tissue contains many

unknown peaks between the mass range of 700-800 Da. Further confirmation is required to

elucidate peak identifications here as to whether the peaks are indicative of the viable tissue

switch discussed previously, or due to matrix/ lipid peak signals.

PLSDA was then carried out on a larger data set then the VIP scores for each time point were

then analysed. For the PLSDA analysis, 3 biological repeats per tumour time point were used

via MALDI-MSI acquisitions and 6 technical spectral repeats were taken from each biological

replicate.

PLSDA models were built to compare the treatment time points against the Control tumour

tissue.

151

Variables/Loading Rot for Table from MarkeAiewfo M M LABV188.txt

10326512

1033.6576

0.5 hours post CA-4-P1198.7821

0.5119976Z

1481.CM> 1850.13.7197 ! 1490.0318£ot3<D>O)®tr

n 823.91211275.11- 1530.79621048.1

1499.14726

-0.5909.2135

Control874.5519

873.5749

-1.5500 1000 1500 2000

Variable

Figure 3. 18: Partial least squares discriminant analysis (PLSDA) regression vector plot comparing MALDI "on-tissue" digest data from samples of fibrosarcoma 188 Control and 0.5h post combretastatin-4-phosphate (CA-4-P), treatment tumour samples.The increased abundance of Actin at m/z 1198.7 in the 0.5h post treatment sample can be clearly seen.

As previously mentioned, there are numerous peaks in the low mass range seen here in the

regression vector plot especially in the Control tissue. The peptide corresponding to Actin is

seen to be increased in the 0.5h treatment. A similar treatment response was reported by

Kanthou and Tozer (2002) (discussed in the Introduction) with fluorescent staining of Human

umbilical vein endothelial cells, Control and post CA4P administration. The results from this

study showed an increase of filamentous Actin overtime after treatment.

The follow ing regression vector plot generated using PLSDA is between Control and 6h post

CA-4-P treatment.

152

PI ot for Table from M afreruew fro MATLAB V18& txt342.5?%)

1032.6512

1491.0247756.!

1348.85111.02930.6 .4259

1246.75197 1527.97976 hours post CA-4-P1033.6576

0.4 1349.!971

1113.6 1570.10.2:

1594.88$2007.1926

'2056.•18259! '2257.2742Q: -0.2 1584.89091232.7853

'2131.9541

-0.4

1481.1817-0.6

ControlI5E0 6

909.: 135

900 1000 1500 2000

Figure 3.19: Partial least squares discriminant analysis (PLSDA) regression vector plot for the comparison of Control and 6h post CA-4-P fibrosarcoma 188 on tissue digests. The bracket designates areas which illustrate an abundance of shared peaks.

The results from Figure 3.19 display copious amounts of shared peaks, especially in the region

indicated by the bracket andit is quite d ifficult to distinguish trends between the two time

points.

The main response considered could be the increase in Histone H3 at m/z 1032, also present in

Figure 3.18.

Due to the substandard tumour vascular and high proliferative state, increased hypoxia is

typical of the tumour microenvironment. It is thought that the latter triggers hypoxia-inducible

factor 1 (HIF-1) which is considered a key transcription factor along with, nuclear factor kappa-

light-chain-enhancer of activated B cells (NF-kB), activator protein 1, tumour protein 53 (p53),

and c-Myc (Kenneth and Rocha 2008). The outcome via certain genes from the transcriptional

activation of these factors would be anti-apoptotic, pro-angiogenic, an increase in glycolysis

and support in the formation of metastasis (Zhou et al 2010).

The methylation of genes; Histone H3 lysine 4 (H3K4), H3K36, and H3K79 are said to be

transcriptionally active and a study by Zhou et al (2010) concluded that under hypoxic

conditions, H3K4me3 was found to be increased in human lung carcinoma A549 cells.

153

This link between the increase in Hypoxia and Histone H3 could be an explanation for the m/z

1032 peak observed in the regression vector plots of the fibrosarcoma 188 tumour tissue. The

increase in Histone H3 could therefore be linked to drug resistance.

Variables/Loadings Plot for Table frcm Markeruewfno MATLAB V188.txt

756.!

24 hours post CA-4-P385725.!

0.61032.6512

734|5783

1757.5405 |26.|395

7355728 804.2S

.3468 (784.5604

0.41417.75521944.5722

j 966.5237 1033.6576

i1054.6116(967.5228 j|

34.4833 988.5-456 1 _ „„„„E L X l II . 1055.6260 3 5 4 9 t l .999,3772; . fi11<

1861.0293CM>- ,1274.7575 1439.8243

i1198.7821

o(U>S’

1275.7490 >1348.8511

3> 128317i)78 1349.86041440.84391220.72!1.4651

'1143.6562 ,1500.7638'1240.7552 1325.8464714.ct 6 909.2135

1.5519,.5749

-0.2

52491.1725-0.4

1.5741

Controli.5775-0.6

100 200 300 400 500 600 700 800 900 1000

Figure 3. 20: Partial least squares discriminant analysis (PLSDA) regression vector plot for the comparison of Control and 24h post CA-4-P fibrosarcoma 188 on tissue digests.The blue arrows are indicative of Histone 2A (m/z 944), Histone H3 (m/z 1032), Actin (m/z 1198) and the red arrows correspond to Hb peaks; m/z 1274, m/z 1416 and m/z 1529.

Overtime in the 24h treatment (Figure 3.20) PLSDA the appearance of the Hb peaks can be

seen in this CA-4-P resistant tumour model. The Actin peak is still increased in relation to the

Control vector plot with the same true for Histone H3. Histone 2A is now also observed,

possibly reflecting the occurrence of cellular necrosis and DNA damage as a result of

Combretastatin administration.

154

Vaiables/Loadings Plot forTablefrom MarkerVewfro MATLAB V188.txt-725:5385

0.8 756.5454

3.62750.672 hours post CA-4-P1033.6576 1416.7760

0.4 1417.7552876.991?

105461161440.1

0.21954.1092

1792.00311516.90231181.7222

K -0.2 1335.7940574.!

1325.1844.6042

-0.4

824.5775

Control-0.65590

798.5131-0.8

7934361

500 1000 1500 2000Variable

Figure 3. 21: Partial least squares discriminant analysis (PLSDA) regression vector plot for the comparison of Control and 72h post CA-4-P fibrosarcoma 188 on tissue digests.Histone H3 (m/z 1032) and Haemoglobin subunit alpha (m/z 1416) are highlighted by arrows.

The Hb peaks are still observable in Figure 3.18 from the 72h treatment sample. The Histone

H3 ion could be regarded as a dominant factor again resulting in suppression of the other

peaks. One explanation for this could be the requirement for tumour rejuvenation/

proliferation as signified by the viable tissue regions seen in earlier MALDI-MSI and HSP-90

immunohistochemistry.

155

VIP scores were then observed which considered 0.5h, 6h, 24h and 72h time points based on

the PLSDA predictive models in Figures 3.18-3.21. Inspection of these data aimed to observe

the variables w ithin the fibrosarcoma 188 study that are considered most important to that

time point.

Variables/Loadings Plot forTablefrom Markervewfro MATLAB V188.txt120

100

VIP scores for 0.5h post CA-4-P

80

60

>>29 944.5722

40

20874.J 5V

0Variable

Figure 3. 22: VIP scores for the 0.5h post CA-4-P fibrosarcoma 188 on tissue digest.

Variables/Loadings Plot for Table from Markerview fro MATLAB V188.txt180

140 VIP scores for 6h post CA-4-P

120

11325.8464100

80

1032.6512

60

17)3'40

1199.762

1490.031820

0Variable

Figure 3. 23: VIP scores for the 6h post CA-4-P fibrosarcoma 188 on tissue digest.

156

Variables/Loadings Plot for Table from Markerview fro MATLAB V188.txt200

180

VIP scores for 24h post CA-4-P

160

140

120

100

80

60

40

l| 876.9919

20

0Variable

Figure 3. 24: VIP scores for the 24h post CA-4-P fibrosarcoma 188 on tissue digest.

Variables/Loadings Plot for Table from Markerview fro MATLAB V188.txt120

VIP scores for 72h post CA-4-P

100

80

60

40

20

0!2000

Variable

Figure 3. 25: VIP scores for the 72h post CA-4-P fibrosarcoma 188 on tissue digest.

The VIP scores featured in Figures 3.22-3.25 highlight Actin, Histone 2A, Histone H3 ,Histone

H4 and Hb as some of the important targets, albeit with differing intensities. The main trend

gained from these VIP score data gives the impression of transcriptionally active tumour

tissues. Throughout the treatment time course structural modifications are apparent via the

Actin and Hb peaks present teamed w ith increases in hypoxia related Histone H3 (m/z 1032).

157

Histone H4 (m/z 1325) is deemed to have a key position in the time points apart from in the

24h samples which could be indicative of CA-4-P interference before the regenerative tumour

action commences, seen in the 72h treatment.

158

3.4.3 Label free LC-ESI -MS/MS for the identification of proteins in a CA-4-P treated fibrosarcoma 188 mouse modelRecognition of proteins via label free MS/MS peptide identification describes the widely used

technique of Bottom Up shotgun proteomics. As mentioned earlier (see Introduction), this

methods uses enzymatic digestion of proteins prior to data-dependent mass spectrometric

analysis (Old et al 2005). MS/MS fragmentation spectra are matched against a proteomic

database in order to achieve a unique, rank 1 peptide identification.

The method reported here is based on multidimensional protein identification technology

(MudPIT) involving reversed-phase (RP) HPLC (Kline and Wu 2009,. McCormack et al 1997,.

Link et al 1999,. Wolters et al 2001).

Label free LC-ESI-MS/MS was employed to study protein response throughout the

fibrosarcoma 188 treatment time course. Control, 0.5h, 6h, 24h and 72h post CA-4-P

treatments were used for analysis.

Scaffold 3 proteomics software tool was used for the analysis of the following Label free LC-

ESI-MS/MS. Time course protein responses were plotted using normalised spectral counting

calculated and presented via Scaffold 3. Gene ontology pie charts were included in the

experimental output from Scaffold to aid understanding of biological function of the protein

indentified.

Figure 3.26 displays a screenshot from the fibrosarcoma 188 Label free LC-ESI-MS/MS data list.

After importing of data, 201 proteins were listed which included those which have a minimum

of 2 peptides assigned per protein, 99% minimum protein and 90% minimum peptides.

The data options for exportation to excel are; protein identification probability, percentage of

total spectra, number of assigned spectra, number of unique peptides, number of unique

spectra, percentage coverage, un-weighted spectrum count and quantitative value.

The Control/ CA-4-P protein response bar graphs to follow use the quantitative value based on

normalised spectrum counts.

Figures 3.27 - 3.33 show the entire Scaffold 3 protein list and corresponding normalised

spectrum counts that were generated. Figure 3.34 features a response bar graph of selected

proteins of interest.

159

A «]

% nh

liiL i M

in P

rote

in:

99.0

%

w Mi

n -

Pep

tides

: 2

▼ Mi

n P

eptid

e:

90%

O /

z 3|duueso!g

I 9|duu990|g

AjiAjpe ja jodsuej;/t}!A|pe Joqe|n0aj uoqe|sue.q

AjiAjpe jcqe|n6aj uopduosupj^ X A|pe 8|nD8|ouj lejnpnqs

A Ajpe jojoujAjjAipe jaDnpsuet} je|nDS|oui

uopounj -ie|ma|owX^A ipe Jo p |n 5s j auiAzua

AqiAqoe tau ted u o jp a |9

A ARDeDfiXp D • • •O M p m q

A^Arpe u jaio td podsue t} AteuixneA^Apoe ;uep|xopue

aujosoquauejqiuaiu aujse|d ,

pad a||aue6joauajqujauj a||aue6jo

snapnu uoupuoipojiuj

auejqujauj < a p u e G t o iP |rflaD e jq u ! <

uoiBi71?pp5ejqxa, aiuosopua

• • • •

• • • •

• • • •

• 4

• 4

• 4

ujn|roqaj nwuseidopuauop|ao|so D .

uise|do^Dsn p je d d e © jo^

ssaxud D ju jq i/q j 0

snjnujqs 0 } asuodsat , • •ssaoojd aA|pnpo jda j

uoqonpojdaj ssaoojd ipujsiupBjo je|n||a3q|nw (

ssoodjd ujaueBjo-qinui• • •

ssaaojd Di|oqepiu UOpOUJODO| uoqez!|eDO|

ssaDOjd uiaqsAs aunuiuji qpMOJ6

uoqez!|eao| jo juawqsyqejsa , ssaiojd |ejuaujdo|aAep

ssaocud je|n||aD&UjS|||03,

uope|n5aj pDiBojoiq (

• •

uoisaqpe |B3j6o|0!q

• •

i j I i l n n n n iq I

a S

ra m

2 2 2 2 2 2 2Awouoxej. ^

(9DueueA) ueueA QAi nuenf) o ^ ^ r? cd $CM CM - tf CM CM v—

if- £CO on co co

AyiBjquuv Suidnoj*) ura cud

}i|6!9M Je|rD9|0WS . .3 9 9 9 2 9 3 S« l S T O « ! r < O i O

• 4

1 i2 ImJ m£ 2 ^ *

£ <£ O

ic » co n n n r t1-1 n rt ^rs ro rs o fN th n0 0 0 0 0 0 0 0 0 _ 0 0 0 0 0 0 0 0 0 0S S E S 2 2M M M H M •

5jj c .ii tj « E s •£ - ! *

E 2 x ■■ * S * i £ 1 1 1

l | j | . s $ £ i s

D O 3 3 3 3U. Q U. u_ U_ u.h) r< hj t i * j w E -Hi E E E E

s & s

* * ju c 5 « curt re o 'c u 53 -* a .»j 8 £2 "5 MZ +J 3 £ 10 js>. w 34 I h II II II, = « .u. u . u . Li. 9 U U U ail E E E E

i l f

i l il/l "J “[ •'«= ■§. ! IV t U < 3§ 4 § -2T? ii i11 P ■a £►H ItJI* r—I -V W 10 <0 <0 <0 10 (O r I M 10 f

I s' 1 f 1 1 1 5 1 SB s' 11ipajjejS

i1siqis! a >] S H 0 S a S S S S 0 S H E

8 . I i is S I«D OJ £U Q $JJ « inQ. p . V

0000

QE0 u v»5s

■Q

1§■p iUj

0 )j :

£

-

>O'

_oacOOJ£o

~trim

£ <

.C’S+jov.acu

c

o■c

.Vi

•*1

1?1! m 1 •

• ! . sw> *0*♦ t i

* E j l % 43 V) * 55 s i ° y cl

e s s ? s. s

IN O CO O

a>

voUj

2 CL

u c

S 50 3X V3

c "o £

,v? t 3

£ §1 ? ® E

I Ia

=r at

"O to5 5f»Vi &

o C

4 2o ' Oi

0

o . 2O -Q

01 C

£ ^0 £ at o*-> Oio at

c5"o

0 it vo jO

to .5

<u 'i1 2.Oi o t i T3

160

■ a 1 a a

DOC

5o£wt/>C‘534->01.a

>4-Oco

axCD<u.c+4H-oc0toZ(0aEou0)>

aro>_e>

o o o orvl o oo torH H1jmaDuxupadspaajauJCN

'O ' /<?>&

§- % * ^

v 4

vvv

noT3

1utoE

£•30>tro&4-J

*5E0)23Ou0)

cQjS*4oCDv»+4000 0

oEoov_3«/>0V.-Qi l1

-Ca3k .

K•Nnri0jV .33)UL

161

■ a

<u+■>1/1oa•Mca/E

cosza

%%

^ <5>

% \ % x +\ x\ S S■fn ”

xX X \ Xxx %

X X X x. \ %XX.V */>

\ \ <6. o.

^ \ V\ %* \ %

*0 \ X X%> \ x %cSL c$

<3, ■'O'- A ,V o ■'-■

X %X X XSfi, 0. X*% x *■ . \ \X tj, %. \

X \yoX X

X -hA ^ X

*% \X, *■% c, yo.\ V^ A

\ \ \\ ^ w✓ sJ> C°x V /.

$ ^ \ %% \ \ X 0A, % / - r *

\ \ \ \ t \ X x X\ \ X \ x

X /% X X " v %

X X ^\ \ s % °0 o. \ ?<,\ ** V

<5>>0,

fe X 0'v<-*», A

* V % °v °* * \, ^ V ^A/ * k <

■o. <6- "A,

%O

o o o o o^ m (N rHsiunoo u in jp e d s pasjpa ixifvj

.. »TM .

Figu

re

3. 28

: Gr

aph

B - F

ibro

sarc

oma

188

trea

tmen

t tim

e co

urse

re

sults

, exp

orte

d fro

m Sc

affo

ld

3.

162

lunoo uirupads pasiieiujoN

163

a

Vv'% 'h.

O’

-Q

junoD lun jpads pss!|euuo|\j

164

a a 3 a a

axaivx+jM-0c0u»'CroaEou<U>

UJXaro

jJxaxTLqpBdspBBijtucN

A, 9,. On

4. \ V * «

% V 'a

'9. '0, \ ^/ \ % % V\ \ \ \ °

*0 V/ V 6-9s V % % °ft\ \ »% V(y \ ,y°f> %\ \ \ \ X \ 5 \ V \ \ v*‘X X \ °* X, v \ \%

•/<y %f t ' f t fts> " ft 9 /V V 1 V ^

\ \ > V°J " " • > . % % V .

4 \\ \ \v”0 'ft •'ft, %A O' ft "to ’Vs

\ \ \ \

^ \ 'ftVV% % X x xft % % 0

90 % X\ % X \V \V %v w \ % %%% %

f t , / f t *!•ft \xx\ x

%-y

Figu

re

3. 31

: Gr

aph

E - F

ibro

sarc

oma

188

trea

tmen

t tim

e co

urse

re

sults

, ex

porte

d fro

m Sc

affo

ld

3.

165

166

jLnoourupack paapujJCM

00•Q

1to5£■ocut01a'"3SP51 0)523OuQj.5

S£4«jcSi4-10000Q£oSio</>s-o

<3-cQ.0 <3000000Si301i l

167

c x -C Xo *0 X 01u O # N Sa a j i ■

Ofics0£I/)I/)c'5

ax<UOJr4-1H-0c0.2'CroaEouai>

' "..^

*rtr]uruiP0cfep0qpucfsj

Figu

re

3. 34

: Fi

bros

arco

ma

188

time

cour

se

resu

lts

post

CA-

4-P

trea

tmen

t us

ing

LC-E

SI-M

S/M

S.

The rationale for selection of the proteins included in Figure 3.34 was from proteins that either

had a high percentage of co-variance throughout the time course (as seen in Scaffold 3),

provided a known validating response or may be of relevance to the stress response and could

be characteristic of tumour biology.

The increases in Figure 3.34 from both Haemoglobin subunit alpha and Haemoglobin subunit

beta-2 exhibit the distinctive gross pharmacological response from administration of a vascular

disrupting agent.

Disruption in the architectural integrity of the vasculature is indicated by decreases in

structural Tubulin beta-5 chain and Tubulin alpha-lB chain compared to the Control tumour

tissue.

Actin, cytoplasmic l(Beta-actin) and Actin, alpha skeletal muscle precursor (Alpha-actin-1)

appear to show different trends throughout the LC- ESI-MS/MS data here in Figure 3.34.

It is known that the reorganisation and/or disruption of the cytoskeleton results in stress fibre

formation (Kanthou and Tozer 2007). Administration of CA-4-P elicits such a response on the

tumour cells thought to be induced by stress activated apoptosis.

Overtime the tumour appears to 'retaliate' favouring actin polymerisation after a brief

decrease of Beta-actin in the 6h time point with increased levels evident in the 24h and 72h

time points.

An overall decrease of alpha skeletal muscle precursor (Alpha-actin-1) towards 72h post CA-4-

P is visible from Figure 3.34. This decline could be due to disruption by CA-4-P, further shotgun

proteomics and immunohistochemical studies could ascertain the trend in the decreased seen

here ofAlpha-actin-1. An article by Hemingway et al (2012) reports of a study looking at the

expression of Alpha isoform of smooth muscle actin (SMA) in primary, benign and malignant

bone tumours. It was found that SMA was differentially expressed in the primary tumours and

although findings were inconclusive, interestingly SMA expressing pericytes (peri-vascular

cells) present in bone marrow/ were also mentioned as a source of SMA levels.

The concept of progenitor cells, infiltrating immune cells or indeed a full stromal cell

population could be regarded as a key factor to consider when analysing protein dose

response relationships.

169

v22100% 2,507.40 AMU, +3 H (Parent Error. 0.42 ppm)

•L —(—s ~ I— L —H C+57—|-A

E— l- P - f — K — l - L H - L -. —I—T —hA -f-V —(—D • L H -v H -s - t— Q—f

-A -f-A - l- P — K — i-A -

IJ —K — ' A + A “■v —HAH—t —|— l — Ha

75%-1.50

?lpha-2-macroglobulin*sc«>

c« 50%- > i «

y18 y2025%-

y21y17y14

y,130%

2500 500 750 25001250 1500 22501750 2000m/z

Figure 3. 35: MS/MS spectrum and normalised intensity graph of Alpha-2-macroglobulin in fibrosarcoma 188 LC-ESI-MS/MS results.lnsert displays the normalised spectral counts for Alpha-2-macroglobulin throughout the time course post CA-4-P.

Alpha-2-macroglobulin is a carrier protein and is known to have strong associations w ith

growth factors and cytokines i.e. basic fibroblast growth factor, Interleukin 1- (3 (IL-1(3) and

transforming growth factor beta (as seen in Figure 3.34) (Feige et al 1996,. Wellcome trust

Sanger Institute 2012). Alpha-2-macroglobulin is also known to be synthesised locally in tissues

by infiltrating macrophages.

The dose response relationship of this protein from the LC-ESI-MS/MS results can be observed

in Figure 3.35 via the Scaffold 3 viewer proteomics tool. Levels here appear to be undetected

by this technique in the early fibrosarcoma 188 treated time points but a sudden increase is

evident from the 6h treated sample, a surge in the 24h time point with the 72h representing a

drop in expression.

With aim to study this response further, immunohistochemical staining was performed on

F4/80, a macrophage cell surface marker to see whether or not this response could be studied

indirectly as a reflection of infiltrating immune cells in response to CA-4-P treatment.

170

m E B S m

400 \im

4Q0 pm

.^,-::v

TVr- • i, ■ •. . w ■■• : r ■>'■"

1x1) .

W $ 0 .

400 ym

171

Figure 3. 36: Immunohistochemical staining o f macrophage cell surface marker FA/80 using fibrosarcoma 188 tissue lOx. (a) negative FA/80 isotype control using 72h post CA-A-P fibrosarcoma tissue - central region, (b)negative FA/80 isotype control using 72h post CA-A-P fibrosarcoma tissue - rim region, (c) FA/80 staining of Control - central region, (d) FA/80 staining o f Control - rim region, (e) FA/80 staining o f 0.5h post CA-A-P - central region, (f) FA/80 staining of0.5h post CA-A-P - rim region, (g) FA/80 staining of6h post CA-A-P - central region, (h) FA/80 staining o f 6h post CA-A-P - rim region, (i) FA/80 staining o f 2Ah post CA-A-P - central region with bracketed region showing macrophage population within the necrotic tissue, (j) FA/80 staining of 2Ah post CA-A-P - rim region, (k) FA/80 staining o f 72h post CA-A- P - central region, the arrows indicate evidence o f macrophages still present and an example of a necrotic region is labelled with N (I) FA/80 staining of 72h post CA-A-P - rim region, the bracket indicates a macrophage population with the V labelling an example o f viable tissue.

There appears to be few macrophages present mainly around the periphery of the tumour in

the Control tissue (Figure 3.36 c and d) as indicated by the arrow with a meagre presence

visible w ithin the central region shown. Similar staining is evident in the 0.5h treatment (Figure

3.36 e and f) central and rim areas. The latter could explain the undetected presence of Alpha-

2-macroglobulin in the equivalent Scaffold 3 LC-ESI-MS/MS results.

An army of macrophages are now seen to be infiltrating the 6h post CA-4-P rim and central

tissue regions in Figure 3.36 (g) and (h) with populations still present in the 24h treatment

fibrosarcoma tissue.

172

v>v •■

400 pm

The isotype controls for the immunohistochemical staining of F4/80 are shown in Figure

3.36(a)-(b) using the same tissue from 72h images in 3.36 (k) and (I). There appears to be

macrophages still present in the necrotic and viable tissue as indicated.

The engulfing of cellular debris due to the phagocytic nature of the macrophage could explain

the decrease at 72h following the surge at 24h. The steep increase could be as a result of the

widespread necrosis observed at 24h post CA-4-P and influx of debris clearing immune cells

overtime between 6h/24h time points henceAlpha-2-macroglobulin accumulation in the

tissue. The decreased levels of Alpha-2-macroglobulin at 72h could be indicative of an efflux in

macrophage cell populations.

The tumour associated macrophages (TAM's), known to stimulate angiogenesis undoubtedly

play a major role in drug resistance (Welford et al 2011,. Ey et al 2006,. Bingle et al 2002). The

complex tumour microenvironment can result in the promotion of anti-tumour T cells yet

conversely studies have shown that biological events within the tumour can 're-programme'

TAM's to possess pro-oncogenic, assisting tumour growth and progression. The F4/80 staining

in the 72h could suggest a possible pro-tumourigenic action via the macrophage population

considering their presence in the viable tissue region.

The HSP-90 beta levels in the label free quantitation graph (Figure 3.34) show a steady

increase in response to CA-4-P administration. The HSP-90 alpha however is in close

agreement with the earlier MALDI-MSI in Figure 3.8 and HSP-90 immunohistochemical staining

in Figure 3.9 mirroring the 24h HSP-90 saturation prior to a slight decrease in the 72h due to

the switch back to viable tissue concept.

Plectin (Figure 3.34) was found to have a high value of percentage co-variance (170%)

throughout the sample time course so this effect was investigated inorder to relate the

response to the biological action oftumour tissue. The graph of Fibrosarcoma 188 time course

results post CA-4-P treatment using LC-ESI-MS/MS (Figure 3.34), indicates a marked

abundance of Plectin in the Control then levels dramatically decrease overtime.

Figure 3.37 displays an example of a Plectin MS/MS spectra visualised through Scaffold 3

software.

173

vio100% 1,529.74 4JWU, *2 H (Parent Error: 1.9 ppm)

80%

Plectin'•2 6 0% -

40%

20%

0%0 200 400 500 800 1000 1200 1400

mlz

Figure 3. 37: MS/MS spectrum and normalised intensity graph o f Plectin in fibrosarcoma 188 LC-ESI-MS/MS results. Insert displays the normalised spectral counts fo r Plectin throughout the time course post CA-4-P.

Plectin m/z977 Control/saline

24 hours CA-4-P

72 hours CA-4-P

0.5 hours CA-4-P

6 hours CA-4-R250711 V I38 combined 678810 max iMEAN 377.>20-377. ?• 7CD

Figure 3. 38: Fibrosarcoma 188 MALDI-MSI multi image showing spatial distribution of Plectin. MALDI images showing spatial distribution of Plectin at m /z 977.

The MALDI-MSI shown here in Figure 3.38 are in good agreement with the LC-ESI-MS/MS label

free quantitative results w ith an apparent abundance of Plectin in the Control with the

reduction in levels overtime post CA-4-P.

174

Immunohistochemical studies using anti-Plectin were performed to visualise this response in

histological tumour sections.

400 fjm

Figure 3. 39: Immunohistochemical staining of Plectin in Fibrosarcoma 188 tissue post CA-4-P treatment.Figure 3.39- (a) anti-Plectin staining of fibrosarcoma 188 Control tissue, (b) anti- Plectin staining o f fibrosarcoma 188 tissue 0.5h post CA-4-P, (c) anti-Plectin staining of fibrosarcoma 188 tissue 6h post CA-4-P, (d) anti-Plectin staining of fibrosarcoma 188 tissue 24h post CA-4-P, (e) anti-Plectin staining o f fibrosarcoma 188 tissue 72h post CA-4-P, the 'V' is indicative of the viable tissue stained here using anti-Plectin and the 'N' relates to the necrotic area which remained unstained and (f) the negative control using fibrosarcoma 188 tissue 72h post CA-4-P.

175

The anti-Plectin staining in Figure 3.39 (a) suggests that Plectin is highly expressed in the

Control tissue which correlates with both MALDI-MSI and LC-ESI-MS/MS data seen earlier

(Figure 3.38 and 3.37 respectively). In addition to this, there is also evidence of a reduction in

Plectin expression as the time course progresses as seen in Figures 3.39 (b)-(e).

From the anti-Plectin immunohistochemistry it is also apparent that Plectin is distributed in

the viable tissue regions.

Plectin is decribed as a 'cytolinker' and has a key role in the stabilisation of the cytoskeleton via

the formation of a mesh-like scaffold (Liu et al 2011,. Wiche 1998). An article by Katada et al

(2012) suggested that Plectin could be a 'novel prognostic marker' for head and neck

squamous cell carcinoma. The article concluded that Plectin had a strong influence on tumour

survival and metastatic processes and further work would elucidate how Plectin mediates

cancer migration and invasion. Plectin is amongst the proteins said to be increased in tumour

tissue and studies have found elevated levels of in colorectal cancer, prostate cancer and

pancreatic ductal adenocarcinoma (Winter and Wiche 2013,. Nagel et al 1995,. Lee et al 2004,.

Katada e ta l 2012).

The response reported here using CA-4-P treated fibrosarcoma 188 tissue suggested that

Combretastatin greatly interferes with Plectin expression. In addition to this, the decrease in

Plectin expression overtime could be an indication of apoptosis and/or necrosis within the

tumour tissue. It is known that initiator caspase 8 translocates to Plectin after apoptotic

induction leading to remodelling of actin/ microfilaments (Stegh et al 2000).

Remodelling and degradation of proteins occurs via the intrinsic and extrinsic apoptotic

pathways (Elmore 2007). Executioner caspases (i.e. caspases 3, 6, 7) supersede initiator

caspases (caspases 2, 8, 9 ,10) to then play a role in protein degradation.

Gelsolin, a known substrate for executioner caspase 3 will then go on (after itself being cleaved

by caspase 3) to cleave Actin to essentially destroy the architectural integrity of the

cytoskeleton (Elmore 2007).

Interestingly, from the immunohistochemistry reported here there is strong evidence of an

inverse correlation between the HSP-90 and Plectin. This enabled visualisation of regions

associated with tumour stress response versus those exhibiting tumour survival, growth and

progression. This inverse correlation is displayed in Figure 3.40.

176

Plectin HSP-90

Figure 3. 40: Optical scans o f anti-Plectin and HSP-90 Immunohistochemical staining in fibrosarcoma 188 serial sections, 72h post CA-4-P treatment.The necrotic (N) and viable (V) tissue regions are indicated.

A clear inverse correlation is displayed between the viable Plectin stained tissue and the

necrotic HSP-90 stained regions in Figure 3.40.

In Chapter 2 a Rho GTPase activating protein 2 was identifiable from the MALDI-MSI-MS/MS of

fibrosarcoma 120 tissue, here in the fibrosarcoma 188 samples Ras GTPase-activating-like

protein (IQGAP1) has been matched.

IQGAP1 is said to be a multimodal scaffolding protein and is implemented in actin dynamics,

tubulin multimerisation, cell motility and migration via numerous signalling pathways

(Malarkannanet ol 2012). There are that studies report that increased expression of IQGAP1

resulted in disruption of cell-cell junctions thus promoting a migratory, invasive phenotype

(Mataraza et al 2003). In Figure 3.34 IQGAP1 is undetected in the Control but visible from the

graph in the 0.5h from which point levels decrease in an 'L' trend through to the 72h

treatment, suggesting possible CA-4-P pharmacological action.

An example of an MS/MS spectrum corresponding to IQGAP1 and normalised spectrum time

course graph is featured in Figure 3.41.

177

1IQGAP1

_____ _______________ I .

Biological Sample

1 Control ■ 2 30 Minute Sam ple ■ 3 6 H our S am ple 4 24 H our S am ple ■ 5 72 H our Sam ple

100% J.95 AMU, +2 r y ^ e n t Error: 3.1 pprfe)

7 5%

£</>C*> y13+lcz 5 0 % ' >1■cc

y 1 6

0%0 250 500 750 1000 1500 17501250

mfz

Figure 3. 41: Example MS/MS spectrum and normalised intensity graph o f IQGAP1 in fibrosarcoma 188 LC-ESI-MS/MS results. Insert displays the normalised spectral counts for IQGAP1 throughout the time course post CA-4-P.

It can be said that the glycoprotein Tenascin (C), has similar functions to those of IQGAP1

having involvement in tumour survival and the outgrowth of circulating cancer cells (Oskarsson

et al 2011).

The levels seen in Figure 3.34 show a marked increase in the 0.5h compared to the Control

with decreases evident in the later time points to follow, a similar response to that observed

from IQGAP1.

An example MS/MS spectra with intensity graph is shown in Figure 3.42.

178

10

30 Tenascin C20

10Mr 11 1 1

^

Biological Sample

1 Control ■ 2 30 Minute S am ple * 3 6 Hour S am p le 4 24 Hour S am ple * 5 72 Hour S am ple

100% 1.050.50 AMU, *2 H (Parent Error: 5.7 ppm)

80% -

s eo%

b6

40%

321.M 5, '* f20%

0%0 200 400 600 1000800

m/z

Figure 3. 42: Example MS/MS spectrum and normalised intensity graph o f Tenascin C in fibrosarcoma 188 LC-ESI-MS/MS results. Insert displays the normalised spectral counts fo r Tenascin C throughout the time course post CA-4-P.

The action of CA-4-P does appear to have an inhibitory effect on Tenascin C however there is

literature to suggest that its fellow glycoprotein, Tenascin W could prove to be a more suitable

marker in solid tumours (Brellier et al 2012). This article reports that the expression of

Tenascin W is more specific in tumours compared to healthy tissues therefore facilitating the

grading/ scoring of diseased tissue. Future studies involving Mass spectrometric techniques

could help to ascertain this claim.

179

Alpha-enolase

Biological Sam ple

1 Control ■ ? 3(1 Minute S am nle * 3 fi Hnur S a m p l e A 74 Hnur S am nle n s 7? Hnnr Sarnnle I

b8100% I.032.06 AMU. +3 H (Parent Error 3.7 ppm)

-V — |-G H— D -

- D - » - G ± r V - ■g H a - f — s

8 0% -

(/>c 6 0%

c«>=s■5 4 0 % ' DC

1,923.97y1120%

0%250 5000 750 1000 20001250 1500 1750

m/z

Figure 3. 43: Example MS/MS spectrum and normalised intensity graph o f Alpha-enolase in fibrosarcoma 188 LC-ESI-MS/MS results. Insert displays the normalised spectral counts for Alpha-enolase throughout the time course post CA-4-P.

Alpha-enolase is thought to be implemented in many processes. In addition to its role as a

glycolytic enzyme it is known to have transcriptional capacity and molecular chaperone

capabilities (Pancholi 2001,. Kang et al 2008). The overall reaction to CA-4-P treatment by

Alpha-enolase is a steady increase from Control - 72h. Alpha-enolase is regarded as a tumour

associated antigen (TAA) and when elevated levels are present in tumour tissue activation of

immune responses occur in patients w ith cancer (Capello et al 2011). Increased expression of

Alpha-enolase is also indicative of aggressive tumour progression.

The response seen here by Alpha-enolase in Figures 3.34 and 3.43 could be a reflection of this

CA-4-P resistant tumour type which is now shown to have regenerative capability.

Transforming growth factor-beta-induced protein ig-h3 (Tgfbi) is featured in the follow ing

Figure (3.41).

180

Tgfbi

Biological Sam ple

1 C on tro l ■ 2 30 M inute S am p le ■ 3 6 H our S am p le 4 24 H our S a m p le ■ 5 72 H o u r S am p le

100% 1,406.70 AMU, +2 H (Parent E rror 2.5 ppm)

8 0%

“ 6 0% -

b3

4 0%

b4 a5

20%b5

0%0 200 400 600 800 1000 14001200

m/z

Figure 3. 44: Example MS/MS spectrum and normalised intensity graph o f Transforming growth factor-beta-induced protein ig-h3 (Tgfbi) in fibrosarcoma 188 LC-ESI-MS/MS results. Insert displays the normalised spectral counts fo r Tgfbi throughout the time course post CA- 4-P.

From the LC-ESI-MS/MS time course results in Figure 3.34/ 3.44, Tgfbi is shown to demonstrate

a 'L' shape trendline similar to that seen from Plectin (Figure 3.34) w ith high levels in the

Control compared to the samples post treatment.

The latter response is a trend similar to the results documented in an article by Li et al (2012)

entitled ‘The role o f TGFBI in mesothelioma and breast cancer: association with tumor

suppression'. It is stated that Tgfbi is a known tumour suppressor and expression of TGFBI is

reported to be greatly reduced in many tumour cell lines (Kim et al 2000,. Zhao et al 2004,.

Shao et al 2006). However a dual role is suggested for Tgfbi which proposes an anti-

tumourigenic function in early tumour growth and as a pro-oncogenic supporter in late stage

tumours.

It is hard to ascertain whether the response seen here is solely by Tgfbi in this resistant tumour

type or as a result of a synergistic effect from Tgfbi-CA-4-P interaction.

181

3.4.4 Protein relationship mapping of fibrosarcoma 188 LC-ESI-MS/MS data using String 9.0Visualisation of the complex protein pathways and networks that govern tumour biological

function could help the understanding of protein dose response relationships.

The follow ing protein-protein interactions visualised by String 9.0, are depicted by spatial

positioning and linear connections w ith thicker lines representing stronger associations

between proteins.

HspaS

Hsp90aalHba x

Plecl

¥

Tnc

Hsp90abl

Tubain

Hbb b2

Iqoapl

TubbS

AL0227BJ1

Figure 3. 45: LC-ESI-MS/MS proteins o f interest from fibrosarcoma 188, visualised through proteomic pathway software STRING 9.0.

The String 9.0 association network in Figure 3.45 consisted of the proteins chosen in the LC-

ESI-MS/MS proteins of interest graph (Figure 3.34).

182

The focal pathway association seen here (Figure 3.45) within this group of proteins is the

linkage between the heat shock proteins; HSP-90a, HSP-90P, GRP-78 (Hspa5) and Actin,

cytoplasmic 1 and Ras GTPase-activating-like protein (IQGAP1).

The stronger association within this pathway runs directly from HSP-90a through to IQGAP1

encompassing Actin, cytoplasmic 1. The latter is characteristic of an active tumour micro­

environment incorporating the morphological changes exhibited by architectural remodelling

of Actin due to CA-4-Padministration.

Although not directly linked Plectin and Tenascin display close positioning to IQGAP1 all of

which as previously mentioned, are known to be involved with promoting tumour survival,

cellular migration and metastatic invasion.

A strong link is visible between fellow tubulin proteins and interestingly AL022784 (Alpha-

enolase) is in close proximity to carrier protein Alpha-2-macroglobulin (A2m). Close positioning

of AL022784 to macrophage related protein A2m is indicative of allergic stress response which

remains true to the pharmacological intervention here. Haemoglobin subunit beta positioned

above (A2m and AL022784) supports the haemorrhagic theory.

Tumour suppressing adhesion protein Transforming growth factor-beta-induced protein ig-h3

(Tgfbi), is joined in this structural reorganisational milieu, as mentioned earlier the protein has

involvement in cell-collagen interactions.

The full pathway from the LC-ESI-MS/MS results recognised by String 9.0 is featured in Figure

3.46.

183

d) ® ®

The proteomic network association tool String 9.0 is featured in more detail in Chapter 4 in the

iTRAQ proteomic analysis section.

185

3.4.5 Concluding remarksThe data reported here in this Chapter appears to be suggestive of an active tumour

microenvironment possibly relating to this resistant nature of this fibrosarcoma 188 model.

The VIP scores (Figures 3.22-3.25) based on the PLSDA results gave the impression of

transcriptionally active tumour tissues. Structural modifications were apparent via the Actin

and Hb peaks present teamed with increases in hypoxia related Histone H3. As mentioned

earlier, Histone H4 (m/z 1325) is believed to have a chief position in all time points apart from

in the 24h samples. Further work would be required to confirm if this is due to CA-4-P

interference before the regenerative tumour action commences, seen in the 72h treatment.

Collectively the results from the LC-ESI-MS/MS comprise of proteins connected with necrosis,

cell structural reorganisation, polymerisation, tumour survival and stress induced molecular

chaperones. The inverse correlation of molecular chaperone HSP-90 and survival promoting

Plectin is evidence of this showing the distinct regional differences in the MALDI-MSI,

immunohistochemistry and LC-ESI-MS/MS results.

Overall the levels of the proteins involved in heat shock response i.e. GRP-78, HSP-90(3, HSP-70

are increased overtime, but interestingly HSP-90a displays the reduction in the 72h treatment,

possibly highlighted the switch back to tumour viability. Mentioned previously in the

introductory chapter regarding the concept of 'repeatedly identified differentially expressed

proteins' (RIDEPs). A few were also seen in the full list of proteins identified which included

HSP-70, enolase protein and peroxiredoxin.

Future studies could also consider the potential pharmacological interactions with the tumour

suppressor Tgfbi, to clarify whether a dual role is undertaken, dependant on dose response

relationships.

Combination anticancer therapy is a known route of choice in the treatment of patients. The

future of disease diagnostics could include MALDI imaging as a diagnostic tool in combination

with conventional imaging and proteomics. A novel and exciting approach was reported by

Balog et al (2010) which described how mass spectrometry could be incorporated directly into

the surgical theatres to identify tumour tissue boundaries and proximal metastases via Rapid

Evaporative Ionization Mass Spectrometry. The latter subsequently substituting the wait for

histological confirmation.

Label free LC-ESI-MS/MS shotgun proteomicsis a powerful technique with the potential to

identify thousands of proteins. Like all methodologies in science it can be said that

optimisation is key when performing shotgun proteomics. Further optimisation not only during

186

sample preparation but throughout data-dependent acquisition mode could prove invaluable

to advance proteome analysis and coverage (Kalli and Hess 2012).

187

ReferencesAdvani AS, Gibson SE, Douglas E, Jin T, Zhao X, Kalaycio M, Copelan E, Sobecks R, Sekeres

M, Sungren S, Hsi ED (2010) Histone H4 acetyiation by immunohistochemistry and prognosis

in newly diagnosed adult acute lymphoblastic leukemia (ALL) patients. BMC Cancer.10. (387) 2-

7

Angiero F, Rizzuti T, Crippa R, Stefani M (2007) Fibrosarcoma of the Jaws: Two Cases of Primary

Tumors with Intraosseous Growth. Anticancer Research.27. 2573-2582

Balog J, Szaniszlo T, Schaefer K-C, Denes J, Lopata A, Godorhazy L, Szalay D, Balogh L, Sasi-

Szabo L, Toth M, Takats Z (2010j Identification of Biological Tissues by Rapid Evaporative

Ionization Mass SpectrometryAnalytical Chemistry. 82.7343-7350

Bingle L, Brown NJ, Lewis CE (2002) The role of tumour associated macrophages in tumour

progression: implications for new anticancer therapies. The Journal of Pathology.196. (3) 254 -

265

Brellier F, Martina E, Degen M, Heuze-Vourc'h N, Petit A, Kryza T, Courty Y, Terracciano L,

Ruiz C, Chiquet-Ehrismann C(2012)Tenascin-Wis a better cancer biomarker than tenascin-Cfor

most human solid tumours. BMC Clinical Pathology.l2.(14) 1-10

Capello M, Ferri-Borgogno S, Cappello P, Novellia (2011) F-enolase: a promising therapeutic

and diagnostic tumortarget. FEBS Journal.1064-1074

Den RB and Lu B (2012) Heat shock protein 90 inhibition: rationale and clinical potential.

Therapeutic Advances in Medical Oncology.4. (4) 211-218

Elmore S (2007) Apoptosis: A Review of Programmed Cell Death. Toxicol Pathol. 35. (4): 4 9 5 -

516.

Encyclopaedia Britannica Inc (2013) "fibrosarcoma." Encyclopaedia Britannica. Encyclopaedia

Britannica Online Academic Edition.)Available at URL:

http://www.britannica.com/EBchecked/topic/205941/fibrosarcoma. Last accessed 04/01/13

Ey L,LiJF, Gnatovskiy L, Deng Y, Zhu L, Grzesik DA, Qian H, Xue XN, Pollard JW (2006)

Macrophages regulate the angiogenic switch in a mouse model of breast cancer.Cancer

Research.66 (23) 11238-11246

Feige JJ, Negoescu A, Keramidas M, Souchelnitskiy S, Chambaz EM (1996)Alpha 2-

macroglobulin: a binding protein for transforming growth factor-beta and various

cytokines.Hormone research.45. 227-32.

188

Grunstein M (1997) Histone acetylation in chromatin structure and transcription.NATURE. 389.

349-352

Harris S, Craze M, Newton J, Fisher M, Shima DT, Tozer GM, Kanthou C (2012) Do Anti-

Angiogenic VEGF (VEGFxxxb) Isoforms Exist? A Cautionary Tale.PLoS ONE 7. (5) 1-14

Hemingway F, Kashima TG, Mahendra G, Dhongre A, Hogendoorn PCW, Mertens F, Athanasou

NA (2012) Smooth muscle actin expression in primary bone tumours. Virchows Arch. 460.525-

534

Jameel A, Skilton RA, Campbell TA, Chander SK, Coombes RC, Luqmani YA (1992) Clinical and

biological significance of HSP89 alpha in human breast cancer. International Journal of Cancer

50 409-415.

Jin Kang HJ, Jung S-K, Kim SJ, Chung SJ (2008) Structure of human a-enolase (hENOl), a

multifunctional glycolytic enzyme. Acta Crystallographica Section D. D64. 651-657

Kanthou C, Tozer GM (2002) The tumor vascular targeting agent combretastatin A -4-

phosphate inducesreorganization of the actin cytoskeleton and early membraneblebbing in

human endothelial cells. BLOOD.99. (6) 2060-2069

Kanthou and Tozer (2007) Selective destruction of the tumour vasculature by targeting the

endothelial cytoskeleton.Drug discovery today: Therapeutic strategies. 4. (4) 237-243

Kalli A and Hess S (2012) Effect of mass spectrometric parameters on peptide and protein

identification rates fo r shotgun proteomic experiments on an LTQ-orbitrap mass analyzer.

Proteomics,12.21-31

Katada K, Tomonagab T, Satoh M, Matsushita K, Tonoikea Y, Koderad Y, Hanazawa T, Nomura

F, Okamotoa Y (2012) Plectin promotes migration and invasion of cancer cells and is a novel

prognostic marker for head and neck squamous cell carcinoma. Journal of Proteomics.1803-

1815

Kenneth NS, Rocha S (2008) Regulation of gene expression by hypoxia. Biochemical Journal.414.

19-29

Kim JE, Kim EH, Han EH, Park RW, Park IH, Jun SH, Kim JC, Young MF, Kim IS (2000) A TGF-beta-

inducible cell adhesion molecule, betaig-h3 isdownregulated in melorheostosis and involved in

osteogenesis. Journal of Cellular Biochemistry.77.169-178

189

Kline KG and Wu CC (2009) MudPIT Analysis: Application to human heart tissue. Methods in

Molecular Biology.528. 281-293

Li B; Wen G, Zhao Y, Tong J, Hei TK (2012)77ie role of TGFBI in mesothelioma and breast cancer:

association with tumor suppression. BMC Cancer.12. (239) 1-12

Lee KY, Liu YH, Ho CC, Pei RJ, Yeh KT, Cheng CC, Lai YS (2004) An early evaluation of malignant

tendency with plectin expression in human colorectal adenoma and adenocarcinoma. Journal

of Medicine. 35.141-149

Link, AJ, Eng J, Schieltz DM, Carmack E, Mize GJ, Morris DR, Garvik BM, Yates JR (1999) Direct

analysis of protein complexes using mass spectrometry.N ature Biotechnology.17. 676-682

Liu Y-H, Ho C-C, Cheng C-C, Chao W -T, Pei RJ, Hsu Y-H, Lai Y-S( 2011) Cytokeratin 18-mediated

disorganization of intermediate filaments is induced by degradation of plectin in human liver

cells. Biochemical and Biophysical Research Communications.407. 575-580

Maes C, Stockmans I, Moermans K, Van Looveren R, Smets N, Carmeliet P, Bouillon R,

Carmeliet G (2004) Soluble VEGF isoforms are essential fo r establishing epiphyseal

vascularization and regulating chondrocyte development and survival. The Journal of Clinical

lnvestigation.113. (2) 188-199

Malarkannan S, Awasth A, Rajasekaran K, KumarP, Schuldt KM, Bartoszek A, Manoharan N,

Goldner NK, Umhoefe CMr, Thaka MS (2012)IQGAP1: A Regulator of Intracellular Spacetime

Relativity.The Journal of lmmunology.188. (5) 2057-2063

Mataraza JM, Briggs MW, Zhigang L, Entwistle A, Ridley AJ, Sacks DB (2003)IQGAP1 Promotes

Cell Motility and Invasion. The Journal of Biological Chemistry.278. (42) 41237-41245

McCormack, AL, Schieltz, DM, Goode B, Yang S, Barnes G, Drubin D, Yates JR (1997) Direct

analysis and identification of proteins in mixtures by LC/MS/MS and database searching at the

low-femtomole level. Analytical Chemistry.69. 767-776

Nagle RB, Hao J, Knox JD, Dalkin BL, Clark V, Cress AE (1995) Expression of hemidesmosomal

and extracellular matrix proteins by normal and malignant human prostate tissue. American

Journal of Pathology. 146.1498-507

Neckers L (2007)Heat shock protein 90: The cancer chaperone. Journal of Biosciences.32. (3)

517-530

190

Old W M, Meyer-Arendt K, Aveline-Wolf L, Pierce KG, Mendoza A, Sevinsky JR, Resing KA, Ahn

NG (2005) Comparison of Label-free Methods for Quantifying Human Proteins by Shotgun

Proteomics. Molecular and Cellular Proteomics.4.1487-1502

Ortho paediatrics group (2012) A Patient's Guide to Slipped Capital Femoral Epiphysis.Available

at URL: http://www.orthopediatrics.com/docs/Guides/slipped_CFE.html. Last accessed

28/12/12

Oskarsson T, Acharyya S, Zhang X H-F, Vanharanta S, Tavazoie S, Morris PG, Downey RJ,

Manova-Todorova K, Brogi E, Massague J (2011) Breast cancer cells produce Tenascin C as a

metastaic niche component to colonize lungs.Nature Medicine.17. (7) 867-874

Pancholi V (2001^ Multifunctional -enolase: its role in diseases. CMLS Cellular and Molecular

Life Sciences.58. 902-920

Saharinen P, Eklund L, Pulkki K, Bono P, Alitalo K (2011) VEGF and angiopoietin signaling in

tumor angiogenesis and metastasis. Trends in Molecular Medicine.17. (7) 347-362

Shao G, Berenguer J, Borczuk AC, Powell CA, Hei TK, Zhao Y (2006) Epigenetic inactivation of

Betaig-h3 gene in human cancer cells. Cancer Research.66.4566-4573

Stegh AH, Herrmann H, Lampel S, Weisenberger D, Andra K, Seper M, Wiche G, Krammer PH,

Peter ME (20000000) Identification of the Cytolinker Plectin as a Major Early In Vivo Substrate

fo r Caspase 8 during CD95- and Tumor Necrosis Factor Receptor-Mediated Apoptosis.

Molecular and Cellular Biology. 20. (15). 5665-5679

Toh Y, Ohga T, Endo K, Adachi E, Kusumoto H, Haraguchi M, Okamura T, Nicolson GL (2007)

Expression of the metastasis-associated MTA1 protein andits relationship to deacetylation of

the histone H4 in esophageal squamous cell carcinomas.International Journal of Cancer.110.

362-367

Tozer GM, Kanthou C, Lewis G, Prise VE, Vojnovic B, Hill A (2008) Tumour vascular disrupting

agents: combating treatment resistance. The British Journal of Radiology.81. S12-S20

Welford AF, Biziato D, Coffelt SB, Nucera S, Fisher M, Pucci F, Di Serio C, Naldini L, De Palma M

Tozer GM, Lewis CE (2011) TIE2-expressing macrophages limit the therapeutic efficacy of the

vascular-disrupting agent combretastatin a4 phosphate in mice.The Journal of Clinical

lnvestigation.l21.(5) 1-5

191

Wellcome trust Sanger Institute (2012)Available at URL:

http://pfam.sanger.ac.uk/family/PF01835). Last accessed 22/01/13.

Wiche G (1998) Role of plectin in cytoskeleton organization and cfynom/csJournal of Cell

Science.111. 2477-2486.

Winter L and Wiche G(2013) The many faces of plectin and plectinopathies: pathology and

mechanisms. Acta Neuropathologica.125.77-93

Wolters DA, Washburn MP, Yates J R (2001) An automated multidimensional protein

identification technology fo r shotgun proteomics. Analytical Chemistry.73. 5683-5690

Zhou X, Sun H, Chen H, Zavadil J, Kluz T, Arita A, Costa M (2010) Hypoxia Induces

Trimethylated H3 Lysine 4 by Inhibition of JARID1A Demethylase. Cancer Research.70. (10)

4214-4786

Zhao Y, Shao G, Piao CQ, Berenguer J, Hei TK (2004) Down-regulation ofBetaig-h3 gene is

involved in the tumorigenesis in humanbronchial epithelial cells induced by heavy-ion radiation.

Radiation Research.162 655-659

192

Chapter 4LC-MALDI-MS/MS using iTRAQ labelling to investigate protein

induction in CA-4-P treated fibrosarcoma 120 and 188 mouse

models

193

4.1 IntroductionIt is known that potential disease biomarkers are of relative low abundance thus making data

validation a challenge (Ye et al 2010). The 'matrix suppression effect' is considered to

exacerbate analysis of low abundant species (Knochenmuss and Zenobi 2003). An article by

Knochenmuss and Zenobi (2003) states that in MALDI, matrix ions can be suppressed if

sufficient analyte is present. In on tissue tryptic digestion, this is seldom the scenario for ions

of interest. Suppression from high abundant Hb/ histones within MALDI-MS spectra is another

example of ion suppression. Such ion suppression by high abundant Hb and histone species

was evident in Figure2.1 in Chapter 2.

iTRAQ labelling coupled with LC-MALDI is a technique which aims to provide relative absolute

quantitation of fractionated samples to give protein response information using the reporter

ions generated (Chen et al 2012). Up to 8 samples can be combined for analysis and the

method is described as a robust labelling technique implemented by N-hydroxysuccinimide

(NHS) ester chemistry (Evans et al 2012, Noirel et al 2011). Sensitivity is improved as the mass

spectra displays summed peptide ion signals from the isobaric labelled sample combination.

The principles of iTRAQ multiplex chemistry are shown in Figure 4.1 (Ross et al 2004).

A Isobaric Tag Total mass = 145

A .Reporter Group mass

114-117 (Retains Charge)

Amine specific peptide * reactive group (NHS)

PEPTIDE

Balance Group Mass 31-28 (Neutral loss)

c11141—| 31 [—NHS

m/z114 (+1) 13C

m/z 115 (+2) 13C2

m/z 116 (+3) 13C2 «N

m/z 117 (+4) 13C3 15N

ISq 18Q

18Q

13C

(+3)

(+2)

(+1)

(+0)

+ P E P T ID E

115 - 30 —WHS + P E P T ID E MS/MS

116 - 29 -NHS + P E P T ID E

NHS + P E P T ID E

30 l-NH-P E P T ID E29 l-NH-PEPTIDE

-NH-PEPT DE

114

115

116

117

Reporter-Balance-Peptide INTACT - 4 samples Identical m/z

-Peptide fragments EQUAL -Reporter ions DIFFERENT

Figure 4 .1 : The principles of multiplex tagging chemistry. (Ross et al 2004). A, illustrates the 3 main components of the isobaric tag molecule; reporter, balancer and reactive group. B, shows the formation of an amide linkage created with a peptide amine, CID induces fragmentation here similarly to that of the peptide backbone. C, depicts the individually labelled peptides which when combined cannot be resolved by MS intact analysis, upon CID

194

the reporter ions emerge as separate masses retaining the b-/y- ion information. The intensity of the corresponding reporter ions determines the concentration of the peptides.

The technique of iTRAQ is a recognised method widely used throughout Oncology (Sutton et al

2010). Ruppen et al (2010) studyingKiSS-1 (a metastasis suppressor gene) in bladder cancer

reported the use of iTRAQ to help predict the involvement of proteins in biological

disease.iTRAQ analysis was performed on protein extracts from transfected (vector containing

KiSS-1 gene) bladder cancer cells (EJ138). iTRAQ comparisons were done of the transfected,

mock, and empty vector-exposed cells. A recent paper by McMahon et al (2012) employed

iTRAQ to studyglobal changes in protein expression in a regional analysis experiment using

multicell tumour spheroids. The article also used immunohistochemistry, Western blotting and

enzyme assays to complement the iTRAQ findings. The aim of the work reported in this

Chapter was to support/validate the results found using MALDI-MSI, LC-ESI-MS/MS and

immunohistochemistry described in Chapter 3.

4.2 Materials and Samples

4.2.1 Chemicals and Materialsa-Cyano-4-hydroxycinnamic acid (CHCA), aniline (ANI), ethanol(EtOH), chloroform (CHCI3),

methanol (MeOH), acetonitrile (ACN), hydrochloric acid (HCI), tri-fluoroacetic acid (TFA),

ammonium bicarbonate, BCA protein assay reagent (bicinchoninic acid), copper (II) sulfate

pentahydrate 4% solution, protein standard solution, 1.0 mg/mL bovine serum albumin

(BSA),DL-Dithiothreitol solution, lodoacetamide, urea, potassium di-hydrogen phosphate

(KH2P04), TRIzol were from Sigma-Aldrich (Dorset, UK). Modified sequence grade trypsin (20

pglyophilised) was obtained from Promega (Southampton, UK,).Modified sequence grade

trypsin (lmglyophilised) was obtained from Roche (Roche Products Limited Pharmaceuticals,

UK). iTRAQ reagents (AB Sciex, Warrington, UK). Isolute C18 RP LC column (Kinesis Ltd., St.

Neots, UK)./?op/Gest SF Surfactant was from Waters (UK). Peptide Calibration Standard II

(Bruker Daltonics, Bremen, Germany).

4.2.2 Tissue samplesMice were injected sub-cutaneously in the flank with a 50 pi tumour cell suspension containing

1 x 106 cells in serum-free medium.The cells employed in this study were from the mouse

195

fibrosarcoma cell lines, VEGF120 and VEGF188. Thesehave been engineered toexpress either

the VEGF120 or VEGF188 isoforms only. Tumours were allowed to growto approximately 500

mm3, before CA-4-P treatment (a single doseof 100 mg/kg i.p). The dosage for CA-4-P are

consistent with animal studies performed at clinically relevant doses (Galbraith 2003., Prise

2002). Mice were killed and tumours excised at varioustimes after treatment. All animal work

carried out documented herein was performed by Dr. J. E. Bluff, Tumour Microcirculation

Group, University of Sheffield, UK. Samples were provided as frozen excised tumours and

stored at -80°C.

4.2.3. Experimental groupsFibrosarcoma 120 tumour samples: Controls (no treatment, saline i.p), n = 6 (labelled

tumourlJL - tumour 1_6), C-A4-P (0 h after treatment), n = 6, (labelled tumour2_l - tumour

2_6), C-A4-P (0.5 h after treatment), n = 6,(labelled tumour 3_1 - tumour 3_6), C-A4-P (6 h

after treatment),n = 6, (labelled tumour 4_1 - tumour 4_6), C-A4-P (24 h after treatment)n = 6,

(labelled tumour 5_1 - tumour 5_6).

Fibrosarcoma 188 tumour samples: Controls (no treatment, saline i.p), n = 4 (labelled tumour

6_1 - tumour 6_4), C-A4-P (0.5 h after treatment), n = 5, (labelled tumour 7_1 - tumour 7_5),

C-A4-P (6 h after treatment), n = 5, (labelled tumour 8_1 - tumour 8_5), C-A4-P (24 h after

treatment), n = 5, (labelled tumour 9_1 - tumour 9_5), C-A4-P (72 h after treatment) n = 4,

(labelled tumour 10_1 - tumour 10_4).

4.2.4 Tissue preparationThe following preparation was performed on sections from fibrosarcoma 120 and 188 tissue

with one example used from each tumour treatment time point. Frozen tissue sectionswere

cut to give approximately 10 x lOpmsections per sample for digestion, using a Leica CM3050

cryostat (Leica Microsystems, MiltonKeynes, UK) (Djidja et al 2009). Sample sections were

stored in airtight tubes at -80°C until further use.

4.2.5 Tissue homogenisation and precipitation of proteinThe tissue as described in 4.2.4 was homogenised in 800pl of TRIzol solution using a micro-

homogeniser (Kline and Wu 2009). Homogenised solution was centrifuged (1,500 rpm) for 5

min to pellet out nuclei/ unbroken cells. Post-nuclear supernatant was then centrifuged

(14,000 rpm) for 30 min. Resulting supernatant was discarded and cellular membrane pellet

was retained for protein precipitation. 200pl of MeOH and 50pl of CHCI3were then added to

each protein pellet sample. After vortexing, 150pl HPLC H20 was added with further vortexing.

196

After centrifugation (14,000 rpm) for 2 min, the bottom CHCI3 layer was removed. A further

50|il CHCI3 was added, removal of the bottom CHCI3 layer was repeated following

centrifugation (14,000 rpm) for 2 min. Subsequent removal of the H20 layer resulted in the

remaining protein precipitate layer. CHCI3 (50pl) and MeOH (150pl) were added directly to the

protein precipitate, after vortexing solution was centrifuged (14,000 rpm) for 2 min.

Supernatant was removed and protein pellet was then allowed to air dry for 2 min.

4.2.6 Protein DigestionlOOpI of 0.1% RapiGest in 50mM NH4HCO3 buffer (pH 7.8) was added to the air dried protein

pellet. Each sample pellet/ solution was consecutively vortexed, incubated at -80°C (~ lhr) and

heated to 70°C (1 min) until solubilised. Once fully solubilised the sample was heated to 100°C

(2 min) and then left to reach room temperature. Each sample was reduced with DTT (final

concentration 5mM) for 30 min at 60°C the left to reach room temperature. Solutions were

then alkylated with iodoacetamide (final concentration 15mM) in the dark for 30 min at room

temperature. The fibrosarcoma 120 samples were then lyophilised and re-suspended in 5pl of

8M urea in 400mM NH4HCO3 buffer prior to tryptic digestion. Sequence grade modified trypsin

was added (20pg/ml or lmg/ml) to80pg (fibrosarcoma 188) / lOOpg of protein (fibrosarcoma

120), a BCA Protein Assay Reagent (bicinchoninic acid) was used for protein determination. In­

solution digests were incubated over night at 37°C with shaking.

4.2.7 Preparation of samples for column loadingFibrosarcoma 120; after overnight digestion the solution was centrifuged (1,500 rpm) for 1 min

and reaction stopped by placing samples on ice where 10pl of 10% TFA was added. Each

sample for analysis was de-salted using C18 Bond-elute columns and subsequently lyophilised

until required.

Fibrosarcoma 188; HCL (final concentration lOOmM) was added to the overnight digest,

solution was then incubated for a further 45 min at 37°C. After this step the sample was then

centrifuged (14,000 rpm, 4°C) for 10 min, the supernatant removed for analysis. This

remaining solution was lyophilised and stored at -80°C until further use.

4.2.8 C18 bond-elute preparationThe C18 columns were wetted with 1ml MeOH, equilibrated with (x2) 1ml solvent A (2% ACN,

0.05% TFA), after the addition of the sample two wash steps were performed with (x2) 1ml

solvent A (2% ACN, 0.05% TFA) and peptides were collected in a fresh tube after 1ml solvent B

(80% ACN, 0.05% TFA).

197

4.2.9 Sample re-suspensionEach lyophilised sample was re-suspended in 10pl of TEAB, 0.1% SDS, vortexed and centrifuged

(14,500 rpm) for 1 min at room temperature.

4.2.10 iTRAQ labellingThe iTRAQ reagents (AB Sciex, Warrington, UK) were removed from -20°C storage and allowed

to equilibrate to room temperature. After brief centrifugation(bench top), 40pl of isopropanol

was added to each iTRAQ vial, vortexing and brief centrifugation followed. Transfer of each

individual iTRAQ vial to each corresponding sample tube was as follows:

8-plex Fibrosarcoma 120; (a)Control - sample 1 - iTRAQ 113, (b)Control- sample 2 - iTRAQ

114, (c)0h post CA-4-P - sample 3 - iTRAQ 115, (d)0.5h post CA-4-P - sample 4 - iTRAQ 116,

(e) 0.5h post CA-4-P - sample 5 - iTRAQ 117, (f)6h post CA-4-P - sample 6 - iTRAQ 118, (g)6h

post CA-4-P - sample 7 - iTRAQ 119, (h)24h post CA-4-P - sample 8 - iTRAQ 121.

All iTRAQ ratios were calculated relative to the control labelled 113.

(a)Control - sample 1 - iTRAQ 113, (c)0h post CA-4-P - sample 3 - iTRAQ 115, (d)0.5h post CA-

4-P - sample 4 - iTRAQ 116,(f)6h post CA-4-P - sample 6 - iTRAQ 118,(h)24h post CA-4-P -

sample 8 - iTRAQ 121 combinations are shown here for comparison to fibrosarcoma 188.

5-plex Fibrosarcoma 188;(a)Control - sample 1 - iTRAQ 113, (b)0.5h post CA-4-P - sample 2 -

iTRAQ 114, (c)6h post CA-4-P - sample 3 - iTRAQ 115, (d) sample 4 - iTRAQ 116, (e)24h post

CA-4-P - sample 5 -iTRAQ 117.

All iTRAQ ratios were calculated relative to the control labelled 113.

To ensure maximum iTRAQ reagent transfer a further lOpI of isopropanol was added to each

iTRAQ reagent tube to then vortex and centrifuge. The contents were transferred to the

corresponding iTRAQ/ sample contents. Each sample with iTRAQ reagent was then thoroughly

vortexed and centrifuged (1 min) (NB - no turbidity observed). The pH was tested for each

sample and adjusted to pH 7.0 - 10.0 with 1M TEAB (maintain organic concentration of 60%).

50pl of HPLC H20 was added to each reaction, vortex followed by bench top centrifugation.

The corresponding samples were all combined into 1 reaction tube. A further 25pl of HPLC H20

was added to the combined reaction tube before lyophilisation (50°C aqueous and then stored

overnight at 4°C).

198

4.2.11 SCXThe SCX column was washed/wetted using 1ml HPLC grade H20 (assisted by a syringe)

followed by 1ml (x2) of loading buffer. The combined sample was re-suspended in 1ml loading

buffer (25% ACN: 75% HPLC H20) and pH checked to adjust to approximately pH 2.5-3 with

10% TFA. The sample was added to the SCX column by passive hydrostatic pressure and the

flow through collected (FT1). A further 1ml of was added directly to the SCX column and flow

through collected (FT2). For the elution of peptides varying concentrations of 500pl 60mM -

lOOOmM KCL were added to the column resulting in fractions El - E6. Fractions were de-salted

with C18 Isolute columns. Prior to de-salting 1.5ml HPLC solvent A was added (2% ACN, 0.05%

TFA) to the fractions. After de-salting the samples were lyophilised and stored at -20°C until

further use. As required, each eluate was re-suspended in 13pl of 10% ACN.

4.3Methods and instrumentation

4.3.1 Sample fractionation and matrix depositionThe Bruker PROTEINEER fc LC-MALDI Fraction Collector (Bruker Daltoniks, GmbH Bremen,

Germany) was used to deposit the LC separated peptide fractions in a combination of CHCA

matrix and peptide solution. Fractions were automatically spotted onto MALDI AnchorChip

targets. The matrix, a-cyano-4-hydroxycinnamic acid (CHCA) was used, in EtOH: Acetone (2:1),

lOOmM ammonium phosphate, 10% TFA.

4.3.2 InstrumentationLiquid chromatography was performed with LC Packings/Dionex Ultimate 3000 capillary HPLC

system (Dionex, Camberley, Surrey, UK). Samples were injected onto a C18, 300pm x 5mm,

5pm diameter, 100A PepMap pre-column (LC Packings, Sunnyvale, CA) on an LC Packings

UltiMate 3000 (McMahon et al 2012,. Montgomery et al 2012). The sample was washed with

mobile phase A for 3.5 minutes at a flow rate of 300 nl/min on the pre-column before being

switched onto a C18, 75pm x 15cm, 3pm diameter, 100A PepMap column (LC Packings)

equilibrated with mobile phase A at a flow rate of 300 nl/min. Mobile phase B (80%

acetonitrile, 0.05% TFA) was increased to 10% and then increased with a linear gradient to

either 30% - 40% over 105 minutes. The fractions after HPLC were coeluted with 1.2 pL of

saturated a-cyano-4-hydroxy cinnamic acid (CHCA) matrix solution. Peptide Calibration

Standard II (Bruker Daltonics, Bremen, Germany) was as listed; Angiotensin I, Angiotensin II,

Substrate P, Bombesin, ACTH clip 1-17, ACTH clip 18-39, Somatostatin 28, Bradykinin

fragment 1-7 and Renin Substrate Tetradecapeptide porcine. The calibrant covered a mass of

199

range 700-3200 Da. Peptide mass fingerprints and MALDI-MS/MS were acquired by using a

MALDITOF/ TOF UltraFlex II instrument (Bruker Daltonics, Bremen, Germany) with a 200 Hz

smartbeam laser (>250 pJ/pulse) in reflectron positive ion mode. WarpLC software (vl.2) was

used in automated mode which comprised of data acquisition (FlexControl version 3.0) and

peak detection (FlexAnalysis version 3.0, using the Snap peak detection algorithm), initially in

MS mode screening LC fractions between 700 and 3000 Da to generate a list of peptides.

External calibration was carried out for each during MS analysis. For Each peptide MS/MS

analysis was performed using LIFT mode (FlexControl 3.0). Mass lists (FlexAnalysis 3.0, using

TopHat baseline subtraction, Savitzky- Golay smoothing and Snap peak detection algorithms)

from each MS/MS spectrum were compiled into a batch file for exportation to perform protein

database searching.

4.3.3 Data pre-processingMascot software version 2.2 (Matrix Science, U.K.) was used for protein identification

searching, comparisons of the MS fragmentation data were made against the Swiss-Prot 2010

Mus (Musculus) protein database. The Mascot search parameters included used Mudpit

scoring with the following post translational modification parameters for both tumour types;

Fixed modifications : Carbamidomethyl (C) (fixed)

Fixed modifications : Carbamidomethyl (C) with a filter of protein score> 28 (fixed filtered)

Variable modifications : Carbamidomethyl (C), Acetyl (Protein N-term),Gln->pyro-Glu (N-term

Q),Glu->pyro-Glu (N-term E),Oxidation (M) (variable - fixed)

Variable modifications : Carbamidomethyl (C), Acetyl (Protein N-term),Gln->pyro-Glu (N-term

Q),Glu->pyro-Glu (N-term E),Oxidation (M) with a filter of protein score> 28 (variable filtered -

fixed)

The confidence threshold selected was 95% (p < 0.05) for MS/MS data mining. Proteinscape

software v2.1 (Bruker Daltonics, Bremen, Germany) was used to collate the total 7 LC-MALDI

runs in eacg case (fibrosarcoma 120/fibrosarcoma 188) into a single nonredundant protein list

(comprising at least one unique peptide and in the filtered search cases with a Mascot score of

>28) (McMahon et al 2012,. Montgomery et al 2012). The protein lists generated by

ProteinScape was manually filtered to flag proteins iTRAQ labelled/ un-labelled, peptides that

were not ranked first were discarded.

200

4.3.4 Statistical analysisExcel spreadsheets were generated via exportation of data lists generated by ProteinScape for

both fibrosarcoma 120 and fibrosarcoma 188 tumours analysed. The ratios were all quantified

relative to the control at time zero and normalised to the average of 1. Thus allowing the study

of quantitative information from the iTRAQ reporter ions throughout the time course of

samples, post CA-4-P administration. Scaffold 3Q+ was used for additional visualisation and

analysis of the iTRAQ data (http://www.proteomesoftware.com/). The data files (.dat)

produced from ProteinScape which corresponded to each fraction analysed, were uploaded

individually selecting 'iTRAQ 8-plex' and 'MudPit Experiment (combine samples)'. The .dat files

selected were using a fixed modification search. Analysis with X! Tandem was selected in order

to improve protein identifications with searching through an additional database.

4.3.5 Protein network analysisAccession lists generated by ProteinScape were imported into the STRING 9.0 database to

observe the relationships between the proteins indentified and study the predicted functional

partners of known protein interactions (http://string-db.org/).

201

4.4 Results and discussion - Fibrosarcoma 120

4.4.1. HPLC UV profiles of fibrosarcoma 120 tumour fractionsVEGF 12 D E I V E G F 123r :

VEG F12D E3 VEGF 12D,

VEGF 120 UV[ 280 . 0nm>

U V | 2 1 5 . 0 n mt

I ii toy iliw |M««»

,'EGF 12D F-

Figure 4.2: UV profiles from the fibrosarcoma 120 fractions after SCX. (a) shows the UV profile from the first eluant (E l) after 60mM KCI following the flow through (FT) fraction from strong cation exchange, (b) 120mM KCI (E2), (c) 180mM KCI (E3), (d) 300m M KCI (E4), (e) 500mM KCI (E5), (f) lOOOmM KCI (E6) and (g) (FT) after addition o f the combined iTRAQ labelled sample. The FT UV profile (Figure 4.2 (g)) does not show many peptide peaks which could be a sign o f good binding of the sample to the CSX column. It appears from the UV profiles that the most o f the peptides are present in fractions E3-E6. For these HPLC runs the gradient was reduced to 30% to maximise peptide elution from the column.

202

4.4.2 Frequency of unique peptides identified

Number of unique peptides identified andcorrespondingfrequency

120

1 00 —

Q.aao<u-QE3Z

60 —

40

20

-W -52 6 81 3 4 7 9 10 11 12

Frequency

Figure 4. 3: Frequency of detection of unique peptides by LC-MALDI-MS/MS. The graph shows the range o f unique peptides detected, for example 112 unique peptides were only seen once with only 1 peptide having 12 unique peptides assigned to it.

4.4.3. Analysis using excelSpreadsheets were prepared which featured exported lists generated from ProteinScape.

Figure 4.4 displays a screen shot of an excel spreadsheet used for protein analysis and ratio

comparison. This relevance of this data output was to observe the full protein list supplied by

MASCOT (using the 'mus' search database) prior to importing into Scaffold proteomic tool

software using a simple colour coded thresholds shown in Figure 4.4. The thresholds were

determined for each column by using the standard deviation (SD) value. To depict down

regulation values lower that the SD were coloured red, up regulation was determined by

values of 1 + SD and values that fell between the thresholds were coloured yellow.

203

nwcvi

>rp

c l . 2

IT

CO2CL

* EoCO

LL. —

I

LU 2 oCL

Q .2

<L TOo ^ ^ in (D t- cn

n id cn n K cm"1- s m o ir) no n in o T f cnCO «X) IX) CM ^CO CM CO CTi CM -rt­f'-'- CO O CO

CM CM C\J CO CM CM C\| kO O

CM N- ^ CM COID n O) r -CD ▼“* CO O ID CO IX) 00 CNJ O) r f

1 s ^ 5 1m cd m r-~ cop5cn cd cm **—CD CD

CD Ocn

o o

n m cm s cd to to __ S cn D N

CD O O CD O

CON-OlDTfCNJlDCDCO Q ^ C N r - 0 0 0 0 0 0 0 0 0

CD O O O O O

CMo CMIs- COr- COCMCMocn CO IX)CMCDCOCOCOy- coh- COCOCOanLX)cnCOCMOCOcoCOIX) oCMCM COCM OIX) CDCOO) cb coCMZ1 COCMCMIX)

O tX) IX) IX) CD CM — CM CM CM CO CM CN CM tX) O

D r r i D f O D C T C O C M l D r - r f T -C M C M c o i o K C N n o o i j n n c o c oO O C N J O C N O W C M O O K S CDLDCNCDTftDTj-COCOnCOTt T t O C M C O n r J - O O I C n Q C D O C O S 00 COCO^DinCOO)(7) O CD ▼- O O r - O r - O O O

LD in r r CM r f oj co n cd m h- mCD CO CO CD

o o o o o o T - o o o r o ^ - o o o

? Q < 55

T-cncncnncNJscocNinr-rfCDSr-inTfcnknnincDincD ST-rf^bcoco^coincvisT-ncvicoo) cd co cd n c\i i- cm n cn cm t- t- c n

o in cn m cd CM CM CM CO CM CM CM tX) O

C D C M O C N C D n m ' - C O T t i n C D D C D C N J O O C N C M C O C D C O C O i n f O n K N C v i m C M N C O C D C D C O t - r - r - O J C D C O C D r O ^ C D r f - n o c N j c D S S - r f c n T t o f o tn co

IX) CO CM

CM CM CM CO T— CO CD IX) IX) C7) IX) -rt CD *—

•X) X—

C O C O C M lD C O C \ IC O C M G )S nC O C D C T ) CD Ol CO CO CD IX) IX) h— IX) CD Tf -rf -rf-

ncncococNt-cooncno in c n o n c n L n in r-nnr-nNCO T-t-T-CM »-

O r o i D f M C D O C O C D C O C D C D N N CD O) CD CJ) COin cd i n Tf Tf in

(/) .= 2 O o cu 2 ro UL. (3 LUCO LU

2 => 2 0 J 2

O T -' o o

UJ

-id o o o o o

y ro To— a< Q)CD X X

UJ LU CO S

UJ 3 3

o2 0 m

□. a Sr o CO CO CD X X

E E E £ £ a> a> a> _ _

X X X X X

53 nj E c

E E Ia CL cn h h ^ >

LU UJ UJ LU CO CO LU CD CD X) =5 CO 3o o o o 2 2 2 0 2

s s s s ' sX X X X X

UJCO LU UJ 3 CO CO O UJ o 3 2 CO O O

I 3 ■»= To OCO l O cvj

£ 5 2 2

' - ' io

p

I s

3 =o < cl o x a

LU LU . . .co co ty2 => 3O O n2 5 2t o 1 - s ' I I I

o > CO * - CL _ l CO ^ <

UJCO UJCO co LU LU 3 D CO CO2 O 3 3 5 | 2 o O

£S c n m g

0 <u Q) oj a> Q) at a) ai (D Q) Q) at Q) ai a) a' a) n a) Q) a) a>E E 2 2 2 E 2 2 2 2 2 E 2 2 2 2 E 2 2 2 2 E E

Tf IX) CM CM

c0) c0) ca> c9 c0) ca) ca) cO) c<Dc C,a) C<Dc<Dc0) cai ca) C<Dc<Dca) ca) c<Da> Q) a> 0) Q) O) (V a) 0) CDa» O) 2 O) a) 2 <Da> 2 2 2c5c5O c56 O o o 5 o (5 (5 O o <5O 5 O 6 (5 c5Is- COCO r IX) f -CMIX) coCM CMCDCMK COCOCMCOCMCDCOClCMCMCMCM

nra°5S.

n n - j in id s

T 3<U

a ^ ^ v - § ° ! = £ CO OCO ^ CD <U

2 c "a So g j• u

■ o ?* I=S 5

o 5

. e £ ■ £ .o »

o ' ^ ,o >

- cco <U

. 3

O2k

.QJ

Q -

EO

Sc oco .co 0 > - 3

■Q OS 5O * -S ° Q . c

<uCU _ r-

" O O

S !

s- a

t 3CU

O « -l ■Oc o<U cu

c n

gcu o

Iccu _CO C5

" ? oQJ

<U "O 4o CU

- Q O

. 8 § ^ o o c »N cu

■ S ^

E C L

a c 5 ucu cu * - - c

a t ;

§ ?

cu3o » t

* a : 2 indi

cate

re

spon

se

rela

tive

to the

co

ntro

l ei

ther

in

crea

sed,

no

chan

ge

or de

crea

sed

resp

ectiv

ely.

iTRA

Q ra

tios

show

ing

resp

onse

re

lativ

e to

the

cont

rol,

norm

alis

ed

to th

e av

erag

e .

j= ^o o 3 1 1

-c '3-<£> fM

oo

*0✓ O '-*v *0

O' ^*0 '*&, 0,

* , ^ \% \ Vo> K \ °o% \ % \

A , ^ ®^ Ov ^

%<y

ctv *><■*

% >

*'A*

" a ^

X/ Q \

\ %

v v

0>o **0 4>Or \ y

J-9,

\ \^ %

%»0- %

A . /0A

V % 'fy0,

xf/o. "O,

A * \ \ ^•■%. \ \ \

^ % *• % ( <i, -6 . ❖

\ % s • * > ,&O'

*b.

(EXT) Iojjuod am o* aApejaa asuodsaj opej t)VHl!

J y

C‘a)4->Oi_CL

i ?Qj '*-■>c 28 I

.c 2?-C fMC A l* £ o o■5 8j2 t-.5 ocu b!s |a •=ft) O

rC ~Ch- o

O Q<u *oS' ^5 52r* cui*Q ft) C 3 : O Q.2 E^ o ■fcX O •52 *->

"or= Qj

\

C <A >r 3

C Irr) -E »*i Q. <1 ft)

aO ft) w -Cc n 8 gft) *c ■f: o*- uo ^-to 4-1ft) Hi• ■» •C 3-2 §cuC Oi <A c^ Ss a o5 S-| § s*H Q O -QE 2 o o8 -oC ft)</> «AO 3k °r9 c ^ c C oS gCO -3 C SQ0 3Q. -C «o o cu 8 *- o* 2I I1 st -Q .. cu“! ^ ^ 4; <u

•S' x u,

LOo(N

mod

ifica

tion

Car

bam

idom

ethy

l (C

).

The results featured in Figure 4.5 are from the complete excel spreadsheet in Figure 4.4.

The iTRAQ ratio response relative to the control (113) shown in Figure 4.5, displays various

outcomes throughout the tumour time course post CA-4-P. Due to the CA-4-P susceptible

nature of the fibrosarcoma 120 model the dominant factors apparent from the bar graph

(Figure 4.5) appear to be from the Hb peptides Haemoglobin subunit alpha, Haemoglobin

subunit beta-1 and Haemoglobin subunit beta-2.

Keratin-1 (KRT1) also produced an increased response compared to the control tumour tissue

and it is difficult to explain the increase seen here in the Oh time point. Further work is

therefore required to validate such a response so soon after treatment. Interestingly, an

article by Tang et al (2012) proposes Keratin 1 as acis-diamminedichloroplatinum-resistant

(cDDP) protein innasopharyngeal carcinoma cell lines. The data from this Nasopharyngeal

carcinoma study concluded that the expression of KRT1 was related to the issue of drug

resistant in this particular cancer and suggested KRT1 as a potential biomarker and key player

in the mechanism of multidrug resistance in Nasopharyngeal carcinoma (Tang et al 2012).

It is known that the type I and II keratins (polypeptides of the non-hair category) are co­

expressed during the differentiation of tumours, consist of either acidic or basic proteins and

are found mostly in epithelial cells (Karantza 2011). KTR1 is present in normal endothelia,

benign and malignant tumour tissue (Remotti et al 2001). It is also suggested that KRT1 has a

key role in the architectural integrity of the cytoskeleton in tumour cells (Reichelt et al 2004).

The increased response found in Figure 4.5 (0h-24h post CA-4-P) relative to the control

tumour tissue could be a evidence of the need to maintain epithelial cellular architecture as

KRT1 levels are still elevated in the later time points (Karantza 2011).

In contrast to the latter, type 3 mesenchymal intermediate filament Vimentin, is shown to be

greatly decreased in response to CA-4-P administration (Figure 4.5). Vimentin is expressed in

normal mesenchymal tissue and considered to have involvement in migratory, adhesive and

cell survival mechanisms (Lahat et al 2010).

Vimentin is said to play a role in mesenchymal transition in cancerous tissues with a study by

Handra-luca et al (2011) reporting a three-fold higher expression of vimentin in pancreatic

neoplasms. Increased expression of Vimentin is not seen in the time points following the

control, possibly suggesting an inhibitory action due to CA-4-P. However due to the results

here (Figure 4.5) being n=l then further repeats would clarify the response from Vimentin.

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is commonly used as a housekeeping

gene in quantitative experimental studies (protein/ RNA analysis) where expression levels are

used for normalisation purposes (Barber et al 2005). GAPDH levels therefore are renowned for

consistency throughout various tissues strengthening its use as an internal standard. Reports

by Said et al (2009) revealed that under hypoxic conditions GAPDH in hepatocellular

206

carcinoma, human lung adenocarcinoma epithelia and colon cancer cell lines (in vitro) remains

unregulated by hypoxia. The article concludes that a GAPDH-hypoxia study is not a suitable

strategy for future anti-cancer therapeutics in hepatocellular carcinoma, human lung or human

colon cancer. The GAPDH results in Figure 4.5 are in opposition to this hypothesis. A

considerable increase in the 0.5h post CA-4-P time point is evident whilst the Oh post CA-4-P

remains consistent with the control tissue. Unfortunately the protein remained undetected in

the later time points therefore making it impossible to deduce the complete response

throughout the time course. The other assumption could be that the absence of GAPDH in the

6h and 24h (post CA-4-P) samples is due to the effects of CA-4-P induced hypoxia.

With a view to validate the uncharacteristic alteration in GAPDH response in the later time

fibrosarcoma time points, the responses exhibited by Hb could be considered. GAPDH showed

a marked increase relative to the control as a result of treatment with the vascular disrupting

agent CA-4-P. Peptides relating to GAPDH were absent from the 6hand 24h samples. The

internal standard in the case of this iTRAQ experiment validating the GAPDH reaction post CA-

4-P could be the characteristic haemorrhagic response evident from Haemoglobin subunit

alpha, Haemoglobin subunit beta-1 and Haemoglobin subunit beta-2.

The protein lists reported here were generated using ProteinScape applying fixed modification

Carbamidomethyl (C), and filteredusing a Mascot score >28 which yielded 29 protein

identifications.

The protein lists generated using the alternative modification parameters; fixed, variable

filtered and variable gave the following proteomics outcomes:

fixed - 99 protein identifications, variable filtered - 31 protein identifications, variable - 352

protein identifications.

4.4.4. Analysis w ith Scaffold Q+A screen shot of a protein list generated in Scaffold proteomic software from the LC-MALDI-

MS/MS results is shown in Figure 4.6. The corresponding list generated after applying iTRAQ

quantitative 3Q+ using the parameters above in displayed in Figure 4.7.

207

f

I giduues 5)

^ ia iid 5 je u o d su e

£ JcqqnGaj uofiepueI jtyupe 9im9iouj p

uoipunj je|n38AtAflDe JO|nD0J 9UJAZU

AjAtpe a pue

Aqwq* qus>p|X0!iuguejquuaiu e

}jed 3||9ue6j9U0jqUI9UJ 9||9U9DJ

uoupuoqDoqiujsuejqujgw

9||9U9DJ0 J9|n||8DPqUuo|D9i J9|n||99eqx

uinimqgj D!uise|dopu9 uoqappoqA]

a 9 9 S S 9 S S 9 S S S S S * *

a a a a a a a a a a a a a a! ?Sa 3 f 8 KR! SS! I 3 a ! S

i'iiii in

1 !

w ® :'> s w v y > vs ss-s v

. 1 . 5

Hi1 M

Figu

re

4. 6:

Scaf

fold

pr

oteo

mic

so

ftwar

e us

ed

for

iTRAQ

an

alys

is.

An

exam

ple

is sh

own

of the

pr

otei

n lis

t fro

m the

LC

-MA

LDI-M

S/M

S fib

rosa

rcom

a 12

0 da

ta

with

in

form

atio

n re

gard

ing

biol

ogic

al p

roce

ss,

cellu

lar

com

part

men

t an

d m

olec

ular

fun

ctio

n us

ing

Scaf

fold

pr

oteo

mic

so

ftw

are.

File

Edit

Vie

w

*17

T5&

9 quent) %

° 5 5 ° 3 3 3 2 5 ° S 2 ° ° °

z u m s t s n i m * *

I iuen&2

uoi unj /tqjAtpe jo |nBaj

a a

i i i i i i i i i i i i i i i2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

liiiii jiiiiiiiiiii1 !i!il|lliliilil|ji Ii

*!i

.Sam

ples

' /«.

I Pro

teins

/'

> Sta

tistics

>Q- . 5 T *?J cn

c o■-£'■5

3o>

209

iTRA

Q fo

ld

chan

ge

ratio

s of

pr

otei

ns

iden

tifie

d fro

m

fibro

sarc

oma

120

tum

ou

rs

-c<\lI

C<383OVj<U5

O O VO CM

9 9 1 1

— -TJ

o6

om

o(N

OoID

o

3E

§<3vo8-o<u-c

3o-c3)38£<uvocoajo8VOC

8a.VO-8a£3vo

+->10<U3lTjv.3O£3+-0 fNI3£§3vo8-o01cvo3

Qj3>C3

•6

a§

o ij e j a S U a p p p j

« 4

+■* i VO,3> Q•I a

210

The fold change ratios which can be seen in Figure 4.7 are displayed in Figure 4.8.

The search parameters applied by Scaffold proteomic software using the .dat file from

fibrosarcoma fixed search gave information for 17 proteins (Figure 4.8). The subsequent data

shows examples of protein sequences found and spectra used for protein identification and

relative quantitation by Scaffold 3 Q+ using LC-MALDI-MS/MS. The following proteins detailed

are; Vimentin, Hb subunit beta, Peroxiredoxin-1 (Macrophage 23kDa stress protein), Tubulin

beta-5 chain, Annexin A5 and Keratin -1 (type II skeletal-1).

Vimentin (Figure 4.9), Hb subunit beta (Figure 4.10), Tubulin beta-5 chain (Figure 4.12) and

Keratin -1 (type II skeletal-1) (Figure 4.14) are portrayed similarly to the excel analysis in Figure

4.5 as previously discussed. The employment of an unfiltered fixed modification search

(proteins > 10 Mascot score) yielded Annexin A5 and the missing peptides for 6h and 24h post

CA-4-P corresponding to Peroxiredoxin-1 (Macrophage 23kDa stress protein). The Log2 of

normalised intensity in Figure 4.11 shows an initial decrease of Peroxiredoxin-1 with levels

recovering during the 6h and 24h time points. The latter matches the iTRAQ ratio initial

response of Peroxiredoxin-1 shown in Figure 4.5 evident in the Control and Oh time points.

Peroxiredoxin-1 is thought to be involved in hypoxia/ reoxygenation process and levels appear

elevated in a number of cancers (Park et al 2007). A recent study by Kalinina et al (2012)

looking at cisplatin resistance formation in cancer cells using peroxiredoxin genes included

PRDX1.It was found that elevated levels of PRDX1 was found in 3 cisplatin resistant cell lines.

Over expression of the PRDXlgene is said to act as a protective mechanism against H20 2

induced apoptosis (Kalinina et al 2012). Could this be an explanation for the regenerative

response seen here by Peroxiredoxin-1?

The biological role of Annexin A5 has yet to be fully determined however it is commonly used

as a marker of apoptosis (Rand et al 2012). This is due to the favourable binding affinity of

Annexin A5 to phosphatidylserine which is expressed on the cell surface of cells undergoing

programmed cell death. Expression levels of Annexin A5 are known to be increased in vascular

endothelium cells. Interestingly the results in Figure 4.13 show an increase in Annexin A5 as

the time course progresses fitting in with the increase in necrotic tissue postulated by CA-4-P

administration.

211

VIM

E_M

OUS

E (1

00%

) 53

,688

.7

Da

RecN

ame:

Ful

l=Vi

men

tin.

< X LU UJ 111 Q COO O < D w O to0 — J X J >- U.Q. > Z - I Q Z I-W Z Q i 4 < CL

co < qc _l Si. < _iCO u. in _i CiC UJ CL>- DC > LJ- CO _l - IJ Q LU < m < COCO Z > — > U- —

QC —I CL LU 5 Z QCCO UJ < LU UJ UJ COh- O O ui tu uj

W J Q J < 2 IDCL UJ z CO LU LU O LUCL > I - LU Q CL UJ _lJ p C - l > - l > J Q< u i a x B o j q

CO z Q CL 2 CL | X 0p- > UJ < UJ CL X

_ i q ; o _ i < _ i > o

COI— Q C O > C O I — CO > - Z C L - l « L U < t -i— x - J oc u i z — u j

C L I L I U O I - I U Z L— LU QC —I O O -------c o i - 2 « d x o >I— Z UJ QC —I —1 0 1--------I U Z > < < 0X I 1 J Q D Q 5 Q

> Q > > C L > H C LC O < _ J O _ I I U > J -Ct J o ! o < o z i uZ C O 0 Q C < I - - I >COU I LL I - _J _l h-C OQQCC O- ICOQX LL > CO 0 Q 0 O — q c c o j c - J C l > > - _ j

C O Q 0 l - ^ 0 | b _ l CO00COCOQCQCL-

h J O 0 _l X CO CL -J

0 > O 0 0 UJ

U- 0 Is > UJ

CL CO <

CL CO _ l

>- CL -J

LU > CL -J CL

< O > x |LU > LU CL <C>

1 U Q Z < I

QC — CO 2 h 0 0 UJ LU Q

J > 0 UJ >

2 I L J J

UJ O < 2 cl LU LU O Z - I

C O J - 0 0 C L 0 W

CO CL | —I — —1 — in C0>ZLO<UJJ C O < < J U J < D O Z

> C 0 O Q C O z O H J J Z j |

QC LU DC CL QC

t - l - H 2 0 Z 0 - ICO > QC < — Z2 > - > O i u « t J - j

CO CO UJ

QC QC - I

LOCMCM

CM

Otf>

oou>

o

5> W

oLOCM

_ l ooo^ — “CO >

LOJ2

LOr -

ooto

-Q

LOCM

55LOCM

OLO

A);sud)U|

i q q i q o i o q i q p i o q i o q i q q i q q i q q

A}|SU9)U| p9Z!|BUUJ0N j f i c q

or§

ofN|Q£Ouv .

22uZS'*

is— i5

aIi££5uQj8-

I1*oCD4«lO

2 T3 Co

- ca

2C3>v.0■Q a

s + ! 3S s • o ' lS PM4-1 I

1 s a sLO O Qj O£ -oQ. ^ CD Cu Q. -QC . 5

C ?CD S

go>, 4-1

OJ LO

s 8? • I sU. rM

212

HB

B1M

0USE

(1

00%

) 15

,840.2

Da

RecN

ame:

Ful

l=He

mog

lobi

n su

buni

t be

ta-1

; Al

tNam

e:

v o< _J

2 x

0 x

2 0 — _i< >

CO —

< >

CO —

Q 0

0 - J

U I

CO CC

O U- U . 2

>- LU QC Q.

0 D I- >£ X0 I

>-> o> o _i x_i _i tic LU

O CO

_ l _J

< CO

i l l < X

O l L >O i - i :> 0 xUI <O _J _ lto CO <

2 Q I -

t i t 0 2 0

5 -J <- i 0 >

■ ° > m z x > 1 1 - 0

< < U_

■ I— < * - < LU > O

< X < Q X < t - CD CL

J l t -

X < U-

> v; o 2 > X

m

oLOCM

CO.X I

ooo

to.in

oLO_1

in

-Q

ooin

osP0sinCM

vPo'O>Po'in s?oin

A)jsud)U| OAjjEioy

i q q i o q i o q i q q q qa j a j i o i f i ' t ' t n c o w wA}!SU9}U| P9ZHBUJJ0M z6 0 "|

io oq

c50

ctos-Qi l

1Q-jg

aIi£53v .4-1QJS'

Iyv

Qj

8!T3Co-ca2CD

o a? ^ 3i g £ oQ. to to ogi a

*C T3 SJ- 0) fM^ - i4 - Oc £ £ c 2 oto >* <U T35 -2 ao7 -qa 5 7 ao gV Ct0) lZ-q rz •t; o c +-3 £

-§ £10 Qjc v~ '- I

"2 *N o T-I3» ")E 3I IZ EO*1 (Ug to Q; .52V . 4o

• I sU» *-i

213

18

PRDX

1_M

0USE

(1

00%

) 22

,177

.5

DaRe

cNam

e: F

ull=

Pero

xire

doxi

n-1;

AltN

ame:

Ful

l=M

acro

phag

e 23

kDa

Q- j3

CO I CDLL_] CL

< > iX- O£—> CDLUQ <I- o CL

a . < OO —]>>H LU

LLcnCD1- i x XLLCL :X

Q D Q_l CO 1-CL—LL> -1 OLLCLLLLL — <LLz O

> 2 >> CL_l> CDcn:x —O CD—:x CDLU> O aLUiX >CO CO

_J CLcn

CO I- CD- >a — CL

$ _ILL < oa _i zCD x —D o t-Cl LL—

ZEX o

><i—<IKLi.

ZCL

<a .

co cn O _ J IX

> - O CO CD X < :x co OQiL — Q >> — LD

o — *O U - CO

^ ± o 2c? a + cois:

5 r 00.X>

X>

CO.XI

>“

co.XL

CN.XI

>5"

o k oLO

o L u)CM

OOO

oLOr->

©oLO

oLOCM

sPo'OO

T —

0sLOl*~

0sOlOA)|suo;u| 9Ane|8y

(5 J O

< LL CO

CD Q 0CO < LUuj cn a m m n »-co <o k> » *

Aj|SU9}U| p9Z||BUJJ0N z6o~i

aI1££2 4muCUaco50

■O<U

2"Occ-cQ.CL .Clv .C-QcucoCoa

alii

3L0 LO QJ o*- a. “c jc<U ^

5S *>c

<u73

CU

.cQj4m

2aLOLO<U

2c8|"cu

cr

I

2QJ

LO

1 CM CM 7QJ CO

o ^-c a

o h- o • -I s *-4 *0 7 .a c «- 5 oo CM73 *1 CU O

E2 § cu h

Q O ^ LO. . ors |

* 2 §. 1 ^ u. <

214

TBB5

_M0U

SE

(100

%)

49.6

70.6

Da

Re

cNam

e: F

ulls

Tubu

lin

beta

-5

chai

n.> > > _l 2 0 co

> > > o 2 CL>1 Cl 1- 1- z CL_io _l D v ; CLQo > CO > < - Z

f - CO > 0 Q o 2< a CO LLo zUI > CL5 > > CO

z _i CO 1- a < UJ> LUCL< a 1- <

> < > CO H LU

> 0 > > _l > H

w LU CO _ j UJ Z LLm 1- IL i CLz LU

CL>- 1- z > CL2D X z _J \— —LU_l 0 2 Q _l $ Q0 ¥ - 0 < UJ 2_ l < CL> CL> 0D 5 a 1- > LLUJ

COz CLCL0 >- 0D z >- 1- 0 COi-0 0 UJ1- CO CO>X < UJ_l 0 z>. 0 a X. CL * XI- CO—_l CO Z _l0 0 x. h- 1- 0 LLH- 0 COcl -1 > <CLLL- u. CL ZQ > _l o < _l CL_ LL_l _ LL. 2 CL0 Z L—a 0 O LLX □ 0 >- CL UJ2UJCL > 2 a <Cl CL0 < LL > l-COLLCOLULL UJLL——0 Z X X. 0 <> O 1- Q _J 2 UJUJUJ0 0 —CL COCOUJ

IL 0 O cl 2 - LU

IL CL0 > IL CLCL <■~y_0 _JL—a. 0 V UJ< COCOUJ> o: LLLU0 QLX Q 2 U- -J 0—> 1- 1- z > UJLL0 CO_l Z > < O QZ a o UJ< < - UJ0 2 LL> _i > < UJo i- 0 _l LL H- 1- UJ0 0 0 0 CL -1 coLU

0 LL_l X J >- z << LUo > □ CL0 t—0 _l Q CO< 0 —<- Q a _JZ X LLQX > CO(- -1 CL1- 0> _JUJ< a cl > >——< z o o < 0LU< UJp- a. O 2 0CLa: XLCl u|< ^ >-2 CLCLLU CL < _l UI

LOtMIX.

Tfo(O

CD

LO

lO

oLOCM

LL

O00

i?'oLO

'■SS'LO

A}!sueju| 9A!)B|0y

O iq o iq o iqiq o o o 10 oiq

V

Aj!su0}u| pazjiELUJON !6o~|

cu3vo,«o■C0 <N *-<QE§cvo2-Qo:

1voQ—iS

Cf§

Is3J3ocu&SO

1"3cu4-D2 ■Q C3

«ca2o>v.C5

-Q

<o2t?cu

m

.c'5■5LOIcsh->cu-Qc

43.3

CU is ? a s

vo Oa

— ■*=5 ^£3 *N•5 - i

:s -vo cu [3 ■C6

cou~3CU

Q CU ^ -Q

3a1o■Li

SO- >2 . 3 a?

<N*1■CO

LOt-HCM

CN *1"st H

ANXA

5 M

0USE

(9

9%)

35,7

54.2

D

aR

ecN

ame:

Ful

lsAn

nexi

n A5

; A

ltNam

e: F

ull=

Anch

orin

Ci

l; A

ltNam

e:

a§

CD -J < y 0Li. ui a > ycJ UJ LL > —i- a. j < 5* * - < _i COu. QC O _j >■UI CO < _iO < Q O CO< — _J UJ h------- LU _i <

UJ LU > z LLo v- a 0 Zqc _i < CO yO > a H QC< □ LU LLz LU — QC LUtO Q C Q yQC 1- l - — QCCO 0 Q 1- —1- < CL LU z

- l CD Q LU LL—I y QC — _lZ -J Z o Q-J < < LL —- I a 0 LUm y j CO COa _i _J — a:LU LU > 1— COQ >- > 2 >L- < - I > >

CD Q 2 y >-1 > QC Q QCcd - j a LL —y QC >- > _l

QC 1-< Cl CD QC Xy y co _J DQC 2 *— I Q-1 2 a CO 1-> < CD > 0

UJ > > CO << - > QC 0 OQ J Q L- y a< X Q 0 5 UJQC UJ UJ IL < 0CD U__ I — > 0q y z L- >- ou_ CD co — _J _JCD I - 0 IL L- _lQ__ 1 >- y LU _l

LL LU LU LU < -JQ CO UJ UI -J <i— yc LU Q >- y> -1 > (- < yl- a > 0 Q- >-0 a a $ — Qa > B yc CO 0i- _j — _i QC CO< Q < LU — i—2 QC w 0 co o

A)jsue;u| 9AHB|8y

/4!SU8ju| p9Zi|euj.iON 2Pen

5! 2,g> r-H U.

216

K2C1

_M0U

SE

(100

%)

65,6

07.1

Da

Rec

Nam

e: F

ull=

Ker

atin

, typ

e II

cyto

skel

etal

1;

AltN

ame:

Ful

l=67

kD

a

0 0 0 UJ Q CO — CO 0C0 0 CL LL > - LU 1- 0 >CO 0 Q LL Q - CO z 0 CO

0 LL > a X 0 CX 2 0 0

o CO UJ a * Q _ l _ j 0 0lL 0 > _ i - _1 CX LU >- CO

0 0 z z H D a 0 QC CO

0 0 _J 0 > _J — LL 0 0

0 LL CL 1- LL CO 2 a 0 0CO 0 0 <x UU oc QC QC 0 0

LL0 _J h- z z Z _ I CO 0

oc 0 _ l UJ z _ l _J 0 0

0 0 CO h- < a LU QC 0 CO

0 >- 0 Q z 5 CO < 0 0

0 CO z > H CO — _J QC 0

0 0 - 0 (X _ i LU Q CO QC

0 0 h- 0 X . > 2 UJ CO I

0 0 > z > CsC iC CO CO

CO LLLU - z l - O 0 0

qc 0 0 UJ LU 1- z 0 CO CO

QC 0 _ $ Q LU QC CO 2 CO

> 0 0 * LUCO > _ l CO 0CO 0 0 h- > — CO < 1- 0CO >- CL O X O o a I 0 CsCCO CO CL _ i H L— 0 UJ CO CO 1-

L- 0 o > qc 0 i — 1— 0 0co 0 > o >■ > x . UJ CO 0 QC

qc 0 Q. z LU 0 0 z > 0 CO

c t LL 0 a LU CO < CO >- >

a 0 > o > 5 I- X . > CO CO

u. 0 CO UJ _J LU _ z CO 0 L-

CO o 0 —j Q 2 O O > 0 co

> IL LLm 0 0 _ i < z 0 >_ i CO QC cx 2 >- LUo CL CO LL

0 0 0 > z J UJ 1- 0 CsC

< 0 0 CSC X. < > _ i o 0 >CO 0 0 Q _J COX < LU0 1-0 0 IL- LUILCOX. 0 >- co

CO < 0 llCO u. 0 UJ CO CO CO

CO > 0 co Q Q >- 0 2 0 0

LL CO 0 < 2 — LL QC QC 0 CO

z X CO LL QC o 1— 0 — CO CO

QC CO 0 CO 0 LU UJ UJ 0 XCO — IL Q 0 0 < < LU 0 >

0 CO 0 z a _1 LU o 0 0 0

0 QC 0 _ l CO < < z UJ 0 0

0 0 0 CO ^c o X - _ l CO 0

0 0 0 isC _ l < CO z _ l CO 0

QC 0 0 —CO ^c QC 0 0 CO

o _J _Jo Q < 0 0 X . 0 CO

_1 z 0 LU > 0 o _ > >- CO

CO 2 0 QC X _J —0 h- CO 0

QC —I 0 UJ QC LUCO CO < 0 0CO CO LLo QC > a — — 0 QCCO X . 0 CO _ l X . > 0 UJ 0 0

o CO CO X . - L- o iC2 CO 0

o 0 0 > CO 2 < X. o 0 0_JLL0 —>- o _J 0 CO

CO 0 LLa >- < > 0 < 0 CO

2 CO 0 —z CO UUa _J > 0

A)!SU0)U| 9A I)B |3y

U > < 0 I O I 0 ' t * < o r o M M ' r « -

Aiisuajui psziiBUJJON ?£|o_|

T~i Qj^ «o <u .523Olui

217

4.4.5 Protein relationship mapping of fibrosarcoma 120 data using STRING 9.0Observation of protein relationship maps facilitates the analysis of complex proteomic data

sets. Understanding known protein-protein interactions could provide an insight into potential

biomarkers. The proteomic pathway software STRING 9.0 proposed known protein pathways

and networks with an additional predicted functional partners feature (http://string-db.org/).

The following proteomic networks generated by STRING 9.0 will feature zoomed in sections of

the original networks shown in Figures 4.15 and 4.16 to ease interpretation. Evidence and

confidence views are included which depict evidence of association and strength of association

respectively. To relate to the fixed unfiltered search list imported into Scaffold 3 Q+ the same

list will be used for compilation of the following protein relationship maps.

218

219

m m

£uiJ^<u<ucoi 03t a»Cl Ac 0 JQU. 3ir:=rc = o £ U l X ) P rr a j u x v n £ t r c o n a « x o <u<uoox'*>a>T Z O U U i u Q h i

/ / / / / / /

sQjuCQj~C5sQj

O '

o<5

£QOCn.c3

T3.QJ

Q.5oQJ

■QU5oQj4osQ.

O*N

Cl5o

hi52

-Q

a§

j :oV .C5QJ«/>

Co4oCJu

iS■Qo

5■ oQj

•E

23Cn

220

Pgam 2

Ptpre

P rd x l LcthaN m e l

Apoa 1

Pkm2P4hb

H s p 9 0 b l A t p 5 a l

Pdia3 Atp5b‘t n o l

HspaS

H s p d lMdl>2

Figure 4 .17: Zoomed in region o f a fixed modification search, iTRAQ fibrosarcoma 120 samples.

There is a link suggested in Figure 4.17 between Peroxiredoxin-1 (Prdxl) and prolyl 4-

hydroxylase beta (P4hb). Considering Prdxl potential involvement in reoxygenantion and anti-

apoptosis, a link with P4hb fits. P4hb function includes the inhibition of aggregated malfolded

proteins (http://string-db.org/). P4hb protein then has strong associations w ith Hspa5 (78 kDa

glucose-regulated protein) and Hsp90bl. Hsp90bl was mentioned earlier in Chapter 3 along

with its relevance to tumour biology and the stress response. Calreticulin (Calr) is also included

to have strong associations w ith P4hb. Trace levels of Calr are found on normal lung cells but

highly expressed in lung cancer cells (Liu et al 2012). Alpha Enolase 1 (Enol, 8 peptides found,

collective Mascot score of 241) can be seen in the zoomed region in Figure 4.17, it also

features in iTRAQ ratio response graph (Figure 4.5) and the iTRAQ fold change graph (Figure

4.8). Both graphs display an early increase in the Oh and 0.5h post CA-4-P with levels

decreasing towards Control levels in the 6h and 24h samples. Enol is known to be involved in

tumourigenesis, proliferation and cancer cell metastasis due to its role as a plasminogen

receptor (Capello et al 2011). It could be suggested that its sudden decrease in the later time

points maybe as a result of increased necrosis by CA-4-P affecting the role of hypoxia tolerance

by Enol (http://string-db.org/).

221

D7Bw<)1348f

Mapkapk2

GmS138

AtpSalH>stlhlc

PrcJx2CcdclOS

Mdh?L p a l s l

A m a5

Aiixa?

Npm3S100a4

Aclb Ccabpl

ZwlO

H2a1/ RUSC2

Capnl

TnlihS

GeminS

Tuba8638031 Fam l79bH3f3a

Tubalc AcvrPaf’smdl

IllOrtJ(- ^

Figure 4.18: Zoomed in region of a fixed modification search, iTRAQ fibrosarcoma 120 samples.

Annexin A5 (Anxa5) where functional properties are mentioned earlier in this chapter, has

several associations. Galectin-1 (Lgalsl - 2 peptides identified w ith a collective Mascot score

of 54) is one such protein and is thought to have functional influence on cell proliferation,

apoptosis and differentiation, therefore enhancing tumour progression (Banh 2011). Figure 4.5

displays a contradictory trend to this where all the other time points relative to the control are

notably decreased, again suggesting an inhibitory effect by CA-4-P.

The same response exhibited by Galectin-1 is true for S100 calcium binding protein A6 (see

Figure 4.5) (Calcyclin -2 peptides identified w ith a collective Mascot score of 56). Functions of

Calcyclin include an indirect role in cell motility and cytoskeleton restructuring (Ning et ol

2012).

222

The Tubulin proteins (Tubalc, Tubb5, Tuba 8) are shown with associations feeding into

Vimentin with branching from Tubb5 to Cofilin 1, a protein which too intervenes in the

regulation of morphological cell changes and cytoskeletal architecture.

Actin cytoplasmic 1 (Actb) is positioned prominently linking the proteins relevant in structural

integrity, necrosis and heat shock molecular chaperones expressed due to malfolded un­

glycosylated proteins.

223

4.5 Results and discussion - Fibrosarcoma 188

4.5.1. HPLC UV profiles of fibrosarcoma 188 tumour fractionsUV| 280 . 0nmJ

UV) 215 . 0nm|V E G F 1 S S E l VEGF1SS E2

VE G F1SS E3 VEG F1SE E4

VEGF1ES E5 VE G F1E S E 6

VEGF1ES r

Figure 4. 19: UV profiles from the fibrosarcoma 188 fractions after SCX. (a) shows the UV profile from the first eluant (El) after 60mM KCI following the flow through (FT) fraction from strong cation exchange, (b) 120mM KCI (E2), (c) 180mM KCI (E3), (d) 300m M KCI (E4), (e) 500mM KCI (E5), (f) lOOOmM KCI (E6) and (g) (FT) after addition o f the combined iTRAQ labelled sample.

224

The UV profiles in Figure 4.19 appear to indicate a greater protein elution compared to those

in Figure 4.2 w ith a more even spread of UV signal, possibly indicating improved separation of

peptides.

4.5.2 Frequency of unique peptides identified

Num ber o f unique peptides identified and corresponding frequency

B no o f peptides

y _ t._1 2 3 4 5 6 7 8 9 10

F re q u e n c y

Figure 4. 20: Frequency of detection of unique peptides by LC-MALDI-MS/MS. The graph shows the range o f unique peptides detected. Here 49 unique peptides were only seen once whereonly 2 proteins had 10 unique peptides assigned to them.

4.5.3. Analysis using excelSpreadsheets were prepared employing the methods describedin section 4.43. Figure 4.21

displays a screen shot of an excel spreadsheet containing the fibrosarcoma 188 iTRAQ samples

used for protein analysis and ratio comparison.

225

ct .55

O cS

> o

CL ■**

Q:

TO

CD * - 0 CM T” CD O CD CD CD T“ ’T“ CM CD

CM t- CM cn cn CD CM 1^- CD CO CO N - CO m COcn m T_ CD ,r” 1^- CO cn CO CM CO CO CD

CO un CO CO - ' t CO cn CM cn T“ CO CO CO cn

►

r f n T -

s n cm m cm n CD ^

c 6 OT ^ ^CD i - S CD CO

n c n n c n i n o i n ^ N r - c i CO ( I I CO I D CD

^ C M i n - ^ c o c o Z t S ^

> 2CD 3

to Q jjj CO CO

n c M C D ^ i - s s n c v j o

0 3

03

73

6

1 3

67

27

6

1 3

11

84

6

0 8

36

38

2 COr- -

CDCO

0 3

22

21

3

1 1

27

07

9 S 3CD >i n r -

3 ur>0 O 0

31

29

75

1 0

06

93

CD

0 7

29

83

1.2

28

7

0 8

59

16

6 r - co *— r*-

3 3cn 0 CD CO CO ’T— 0

30

37

36

1 8

29

19

3

: 5

45

06

3

0 7

94

49

8

0 9

51

55

0 8

03

73

6

1 3

67

27

6

0.8

7

1.4

8

1.4

2

0 9

6 i nCD

0.8

9

1.2

2

9S

0

65 0

60 L

88 0

CO

CD 0.7

9

1.3

3

0.9

3

3.2

9

1.9

5 h -CO0 1

.98

0 5

9

0.8

6

1.0

3

0.8

7

1.4

8

< CO CDs ^ c d m co o o cn •<- o

N C D T f O O C O C O N .

C n C O C D ^ - C M r O G l T f T - C M C D N - O T a i C O C M L X l k n T r r f C M C D C M C O C M k O r f S ^ t O C O O f O CM t- O CO CM ^ CM CM CM ^ Csi CO ’r - C M ’* - ’* - ^ - t - CM

C M C M r f C M ^ - n n C M C M m r - r - r - 0 m

r o n o i n c D c o N c n c o n o ^ C M ^ i a i s c n c M c o c o c n n

C M r i - n c n n c o o c j )

r t n c o c o c o c o N r fC O O C D N S S ^ - C M l D T f S C MC M C M C D n o T f c n n i n o r r n

cm n t - cm

T - O O C n C f ) T - o c O C M S ^ - O C M f f ) S « > C D m i f i i D O T T - , _ T - ^ O T k n u ? O T c n s c n -rf

c o c o c d s n a ) in r t o m -=*

i - c M Q T f r o i f i ^ t n c M C T c o n o c o c D T t s c n n c M o s N N C D C D N S r - C O i x j r - C M r i ^ C O ^ S C O N t - T f m C M C ^ T r S m r - ^ c o n C M r - r - r - C M H CD l D 0 i r ) n S C O C Q T f T - i - i n

s. s. p z : o Z S . O <0 OCO

Q) Q) Q) W — COJ f _ _ 3 3 3 CM O O O O "=

* = = = 1 3 1i 5 5 5 3 e 3= c c c E

7 ^ 0 )

§ pW m X

- ^ 3 0

cn E E to

s <f> T O to o ,

g o5 ^ u ISto t : 3E 2

CO £

■= o u o_ ii w ii C Cfl o V) cn o > oO <N < <N 2? ' r- c\l n£ x X Icn a) cd a)

= CO 5 O

c 2S II o CO

o o o c c c c ' c . E w

- 8 E T3 o— s zTO ou ^1■O Sg o

! E ro

a' a> X X

3 5X Qlcn cn

a) ^oJ W (O w

X X X X

ygig3 S - 5 , S IO 2 3 Picn X X X

i ! o o» 5 ®

CLo

- | S .

X CO =3O T- ~ | 5 x 0 iQ LUS en 1

CL CL

o oa! cl i £

* 12 — "| o I |0 0 CL 5

5 <

U CO . | ° !

O 0) <co _a 3 ^E E ■3<f= =}

c i co II _.

CO _0 £ . 0) ■it;.O X '

" t1 Ei ?

CO I - (0 o <1 2 :

I IX> CD - CD CL :

< I- I

vi 3® o n '4=

— <J

js E•C (0? "S- o o

to CO

i ICO 3o 2a) II to COra O= I =<0 t ra

- 5 ® LJ

00 “ > . E to§ i

>£ -Eto CO

u o 5 , 5

■ n J J

i lCD I -

S-g.ifCD x =

S 5

-cr c IIJ ® -7= s>o

CO0 (0 3Q . 3 3 _ t ) CJ CO CO 2

1 • IO -n_Q o ^

£ E ^r« ! g

1 ^ 52 <

?■ E" c .a 1q .E Z> 3 >

5 o 2 9o ' x ' u j 13 ^ Scn 3 >

( l ) Q ) ( D Q ) ( D l D ( D ( D ( D Q ) Q ) ( D ( D ( D t t ) Q ) © ( D ( D 0 ( D 0 ( D ( DE E E 2 S E E E E E E E 2 E E 2 2 E 2 E E E 22

c cz c c c c cC<D

cCD

CZa>

C0)

c<v

02 CD CD p CD

c 5 O 6 0 c 5

COCDCM

CD3

CDCO

c c C C cz cz c c c Ca) <D a> a) a) a> a) CD ai 0)p p a> cd 0) p p p (V CD

3 3 O O 3 O 6 3 3 5CO CM c d r*-. CD CO T— CM CD m

*r_ CO CO ^ t CO T” CO

cm n m 0 sr t ' 'T

0 CD<0 Q

Siu

CD OJ O l T 3 Q

§io

cQ

2 ■QCD .

•S2 * Q

V.oCDC ncQ

■5oc

0

1o c

s JrCD ^S 22 53■2 o o i : =* c

o uCD

r n - c

r n

cooCD

o ^2 1/1 c$j o

a . - <t; if1 ^5 Si oQj CD ■*=

^ o Si0 o h q ' S * * 2 .c -C3a -o -g § Si 8

c o o0 ^ 0 CD Jl OIC| i j

■? c ' c 2 § . 2 cd <u .t:

^ V- ft ■Q CD ° 53 f : c .x .t; -

>c; cd "og c Si 00 — 0) "C t3O "535 ic C

8 8co . 1.o -cC CD

•s. S |1 1 * O ^ CD

2 .§ .2 » "3 o ^ 2 5 o— *«. «2 a £

£ t S>l-ec S

<N

CDL .3Cn

iH

c - 2 sO CD.i2 2»O <0 CL CD£ P- S C^O CD U v

22

6

£oa*<u1 1 S +_O Q-c .uV) jC,2 ^ •S oa e s 'Q£ 5 - i I

Si *' 3 S!

8</>

TS *=Qj Q £ *0 C O £ -CO •§

** & *■* S CL -c4*»*0c Io E o o<u o£ o ■+» +-S "o+•» CL)cu 52 > 3 '45 </i O QJ"5 -5

$ §■

r-CMCN

£ 05 59 e oo .§^ }?o 85 ^ o -t; o o.wig 2rQ cn ^ -5C'5 -2</>vj r- Q_ <U V) r-S! |0 P45 SJ2 S° I1 Is |*N £<N 45

Qj "O 2- 23 .5 .cn ^ u. o C

arba

mid

omet

hyl

(C).

oa

U3'J

^ CMcm r''-a a

CU00COi_01>rocu

JC

T 3<U

"roEoc

oucu

CU>+3

_ ro

cu

cu1/1coQ.i/I<u1—txoc

oJCV )</)O

’-Mnji—a<DC

4__j------ |

5=

'T?

. —rs

—-------1

iiuSBF

Hhl“ f c .

_ < * * i -

oo

\%

% VOc

% %% %

S^ °o

%

O y . ^

%

<Sr.■ % ,%V

%%

v%/

V

%%% \

% WCy

O '.X< r

y '

<*>

w % / \ * >

- ''f<y

% ,

vV O'

\ *%

A %\% %X

\% %V

< \

y><

<>' V *

( E l l ) 10 J J U 0 3 0 l | J 0 J 9 A ! J B | 0 J 0 S U O d S 0 J O J J E J t ) V 8 i !

cu+JoI —Q.

Figu

re

4.22

: b

Fibr

osar

com

a 18

8 iTR

AQ

sam

ples

sh

owin

g ra

tio

resp

onse

re

lativ

e to

the

cont

rol

(113

), no

rmal

ised

to

an av

erag

e of

1.

228

The proteins seen in Figures 4.22a and*4.22b are from the same excel spreadsheet separated

into two bar graphs to aid interpretation. These data has been generated via ProteinScape and

exported to excel for analysis employing the same parameters prior to examination as detailed

in the previous section entitled '4.3.4 Statistical analysis'.

The fixed filtered search performed on the fibrosarcoma 188 samples resulted in a greater

peptide yield in comparison to the fibrosarcoma 120 results. 45 proteins were identified,

opposed to 29 proteins in the fibrosarcoma 120 data set. The latter could be due to differing

factors during the tissue digestion, SCX or LC fractionation. Another factor to consider could be

the more CA-4-P resistant nature of the fibrosarcoma 188 tumour where different peptides

identified could be as a result of decreased suppression from high abundant proteins i.e.

Haemoglobin/ Histones.

The discussion that follows aims to highlight key responses from Figures 4.22a and 4.22b.

The haemoglobin peptides (Haemoglobin subunit alpha, Haemoglobin subunit beta-1,

Haemoglobin subunit beta-2 and Haemoglobin subunit epsilon-Y2) are consistently present in

Figure 4.22a to aid validation of the other protein responses however even with the fixed

filtered search parameters many different proteins are observable.

Heat shock protein HSP 90-beta is increased in the 6h-72h time points with what appears to be

a saturation point achieved in the 24h. This response by Heat shock protein HSP 90-beta was

discussed in Chapter 3 where evidence of the switch back to viable tissue was observed both in

the MALDI-MSI and immunohistochemistry.

A similar surge pattern of GAPDH levels (Figure 4.22a) seen previously in Figure 4.5 can be

observed. The sudden increase appears in the 0.5h fibrosarcoma 120 time point (Figure 4.5)

whereas in the fibrosarcoma 188 samples the surge is seen 6h post CA-4-P with decreased

levels apparent in the 24h and 72h samples. An explanation of the delayed response in the

Fibrosarcoma 188 data could be indicative of more resistant tumour model.

Tumour enhancing protein Galectin-1 (Figure 4.22a) is shown to be decreased relative to the

control but later on in the 72h sample a small increase is observable possibly indicative of

tumour regeneration.

Elongation factor 1-alpha 1 is a protein giving very different responses in the two tumour

models. In the fibrosarcoma 188 samples levels are elevated throughout the time points in

Figure 4.22b relative to the control. In contrast to this, levels in the fibrosarcoma 120 iTRAQ.

results are all decreased relative to the control. Elongation factor 1-alpha 1 is a known

229

regulator of the cytoskeleton (Shiina et al 1994,.Veremievaet al 2010). One explanation could

be that the elevated levels seen in the fibrosarcoma 188 samples maybe due to some form of

pharmacological conflict elicited by this CA-4-P resistant tumour model.

The Complement system exhibits regulatory influences on host immune, inflammatory

processes. Activation of this complex cascade of biological events is triggered via 3 main

activation routes; the alternative, classic, and lectin pathways. (Nunez-Cruz et al 2012)

Complement C3 is the key activator of the 3 main routes previously mentioned. Studies have

shown that increased Complement anaphylatoxin levels have been detected in the peritoneal

cavity fluid of ovarian cancer patients (Bjorge et al 2005). In vivo studies by Nunez-Cruz et al

(2012) used complement deficient mouse models crossed with epithelial ovarian cancer

susceptible mice. One of the key findings in this article was that tumour growth was either

absent or greatly retarded in the offspring, compared to the wild- type. Thus indicating the

importance of Complement C3 in tumour progression. Figure 4.22a shows that the iTRAQ ratio

responses from 0.5h, 6h and 24h are reduced after CA-4-P treatment with a slight increase

emerging from the 72h time point.

Plectin, not identified in the fibrosarcoma 120 data set is shown to have a greatly reduced

response relative to the control in Figure 4.22b. Plectin is has been shown to be present in high

levels in cancers such as head and neck squamous carcinoma and correlates with poor patient

survival (Katada et al 2012). The ITRAQ data presented here show good agreement with the

findings presented in Chapter 3 where MALDI-MSI and LC-ESI-MS/MS label free results

displayed increased Plectin levels in the Control fibrosarcoma 188 samples with a dramatic

decrease evident over the 0.5h-72h time course. Hence a decrease in Plectin expressions

protein associated with tumour aggressiveness, does therefore appear to be a response to CA-

4-P dosage.

The protein lists reported here were generated using ProteinScape applying fixed modification

Carbamidomethyl (C), and filteredusing a Mascot score >28 which yielded 45 protein

identifications. The protein lists generated using the alternative modification parameters; fixed,

variable filtered and variable gave the following proteomics outcomes:

fixed - 318 protein identifications, variable filtered - 51 protein identifications, variable -1 0 88

protein identifications.

230

4.5.4. Analysis with Scaffold Q+A screen shot of a protein list generated in Scaffold proteomic software from the LC-MALDI-

MS/MS results is shown in Figure 4.23. The corresponding list generated after applying iTRAQ

quantitative 3Q+ using the parameters detailed in section 4.4.4.

231

2<3■o29oo

£

O/

uar

PSI•9

I

I 8|duies t w a i ! j ££££££! o o o o o o oW i p e ja jjo d s u e j} « »

Wff® JoefiBaj uope|Suejq |AQ3e afuejoui jejrtpn

uoqrxjnj je|nD3|oui

Wff* oqAjeieofiupuq

suejquuaw ewse|d 5*dajaS6ib

snapnu(uoupuoqoojiw

auejqwaiua||aup6jo jp|n||8Dequ!» uojBaj jP|n[|3DPJixa

uoja |50j/j > uisefdoiXo

ssaoojd DjiuqqXqj

• •

• • •

N M II1 I ft f(tJ15 U y uI iO §

nri

sninuj s oi asuodsajssaoojd |pujs|up6jo jp|n||soq|nui*

ssaoojd uis!ue6jo-q|nuu ssaoojd D!|oqeiauj

uoqez!|PDO|LpMOjB

uoqpz{|P30| jo iu3uuqs||qpisa55330Jd |91U3UldO|9A3p <■

SMOOjd JPffSD »______ BmiPi iim

uoqe|n6sj po|Bo|0|q uojsaqpa |e3i6o|0|q

• •

• •• • *1

• • • •

I * I S4

4) • • • • <

• • 4

% |

t i l lE c =

! ! ! : : ! ! ! !raftjrerararararararaXUJOUOXCl £

8 s 2 to ^ 3

2 2 S 2 2 2 2 2 S M S S S S 5AlinBjquitf DUjdno.0 inajdJd

S _ IO IU «D3 3 2 3 9 S 9 2 53i|5i9M JS|rD9|ow 2 999

jgqujnN uoiss9ddv

S' s s IJ2 ? u £o O g .2U U4 u. XII II II 21 _H U i l l i U yt'l u u E E E E E E E E E I 1 E £ E E

iipajjeiS

is v > ms]S I3 B I& ' • E S S a S ElO A 01 O'

iaiqisiA >

"5f a j s

m

i i • i f E t f J i- ' | / o E

aO « w S °

I ? 2s 0. a. I£ | J ,

C£o2cto2-Q

OS

2is—j3

Qj2E4

.c5

2a. C qj .2-P* +j

E 4- £ 2 o 3 -c uto Qj

.52 OQj ^

a -o E cQ Q8C

.to

CQJ

I_ Cto 3 .

t 5 2 °5 ° o »*,a 55 =§ E 3j 5 ££ 3Q j O «o v. 3 Q.2 OO .y * 2* S i S t§ fO t jQJ 2

2 ^ a 2

5 c^ .2 Cfc* 4-0 oU E40 5«o »2 , <N c

1 i. 1 5U.

232

*Qj

- C

Si

fk& 4* ii ^

■s &

0 |Dnv O 'O 'lf lO O 'C O C O ^ N rt ° »-S T - J ^T- ^^oo r ^oo

aa< .

o cr* o o rv i*n£ juenb ° S 5 3

Z }uent) “?

I juenb' O O O O IO

ii“ i ? jS i ” o” S“ o“ t " i “ 9DueueA jo ju a p ! j ja o 3 3 6e j3A tf g o g o g o g o o g j

Buussnp OAijejijuenbAjjAjjoe jajjodsueq

XjiApae jojop&aj uoqe|SUB.qAjiaijdb a|m s|ouj |e jnjD n jjs

uoqounj je|n38|oui Ajwipe o!jA|ejeo

fiupuqauejq tuauu euuse|d

j j e d 8 ||a u e 6 josnapnu (

uoupuoqxqjui auejquiauj

3||3ueEuo je|n||a3ejju| * uoiBai je[n||aDPjjxa

U0J3|3>|S0}A3 •

uuse|dojjtosssDOJd D|Ujq jAqj •

sn|n u jq s o j s s u o d s s j • • • •sssDOJd laa is ju e B jo j 3 |n||3 o q |n u i» • • •

a s s s a o jd uu s iue b jo -q in w •sssDOJd 3{|oq3 js u j

£ uoqez||B30| • •q jM 0 j 6

2 uopBZ||B30| j o jusu u q s iiq e jss • •SSSDOJd |3 JU3UjdO|3 A3 p • •

BuillM ll» uoqe|n6aj |33{6o|0|quojssqpe |K>|Co|0|q

Aujouoxex m £ £ £____________S 3 3 3

A jinbiquiv 6 u |d n o jg u ia jo jd «

i i ia § 2

• •

E l l 2 2 2 2 2 £ £ £ i £2 2 2 - 2

^6iqm je|nD9|ow S 9 3 9 3 ^ 9 3 3 ^ S S $ S p 3 g S 8 p : ' t

l i i i i i = l §jsqujnN uoisssddv -r IS J , JS ri I = 3 l< t KS S S C 5 S i S i |< x z < £ C u S z >

i i i i s i i i i R 4 ktlllipILs I f l i l i l l l1£ £ t5 * I S o .o o > 3 J• ■ 6 < x r J ( S C u i x >Z 2 II II II II II II II II II M

S o S S

s i ! 5 25 8 u §* o c J,

ii n n ii3 3 3 3 3u . u . u . u . u.o3 (J oj 43 i3E E E E E

« od u ce e*ipajjejSi8|q!S|A *>

ID N OO O'

□ D

1>

I1

135

ii % 58

•£ §cg

<vc & V

a o .3£ *5 1

fef i<u •cb

~ d i

2 £

S t

C.o■C0

§1o> -< O oo

■R 00

a c

s §c rg .2 ^ 4mIo c

£"5cc.o‘■Csa§

*8,c.o4m0 c£*+pr

cQj

1.OQ.5QJQJ

IQJv .C£$is,U

EoQj4m0v .a■o1 u

«N

8i3cni i ch

ange

ra

tio

thro

ugho

ut t

he

time

cour

se

sam

ples

can

be

obse

rved

; Q

uant

1 (C

ontro

l) - Q

uant

5 (72

h po

st C

A-4-

P,

iTR

AQ

fo

ld

ch

an

ge

ra

tio

s

of

pro

tein

s

ide

nti

fie

d

from

fi

bro

sa

rco

ma

18

8 tu

mo

urs

-3R

§3

<3

23 O uQJ5

c -C -C0 “1 j: inu o in r3 1 3 3 3

% %

— «

1 i

.....1..... "J.......L_....•

% % r>X \ x 1 %

% \ \ y<t> "ju0/ $ X *\ %k \

■ " — -

oqsjcftapppy

38O83vo8

-3

Qj■3

3O-33>38€0)</ic0

1 2! V)

.3

8o.

-Si 3 . 8 3</>

Sf1QJ3)T3v.3O53-t-i

29oo35

§

83

3>.3553O

(U3)33

-3u■3

or§

uS ol* 40J O> -»W.§» su. a

234

The fold change ratios which can be seen in Figure 4.24 are displayed in Figure 4.25.

To analyse the data from fibrosarcoma 188 models the search parameters applied by Scaffold

proteomic software using the .dat file from fibrosarcoma fixed search gave information for 16

proteins (Figure 4.25). The subsequent data shows examples of protein sequences found and

spectra used for protein identification and relative quantitation by Scaffold 3 Q+ using LC-

MALDI-MS/MS. The following proteins detailed are; Vimentin, Hb subunit beta, Pyruvate

kinase isozymes M1/M2, Tenascin and Fibronectin precursor. The rationale for selection of

these proteins was to assess differences and similarities in the Scaffold 3 Q+ results in the

fibrosarcoma 120 model and observe proteins which did not appear in the VEGF 120 samples.

The Vimentin response observed by the resistant 188 model in Figure (4.26) bears some

similarity to the study mentioned by Handra-luca et al (2011) reporting elevated Vimentin in

pancreatic neoplasms. Although levels are still low in the 72h post CA-4-P sample a

considerable increase is evident from the 6h and 24h time points, not observed in the earlier

Vimentin VEGF 120 Log2 normalised intensity graph (Figure 4.9).

Hb subunit beta in Figure 4.26 does not exhibit a steady haemorrhagic response but levels still

remain increased toward the final 72h time point.

Results displayed in Figure 4.25 of Pyruvate kinase isozymes M1/M2 show greatly increased

fold change ratios when compared to the equivalent fibrosarcoma 120 graph in Figure 4.8. The

fibrosarcoma 188 6h value shows over a 3 fold increase along with a 2.5 fold increase in the

72h for Pyruvate kinase isozymes M1/M2, opposed to the 120 model which only shows a 1.8

fold in the 24h. Pyruvate Kinase isozyme M2 is found in tumour cell and proliferating tissues

alike and hypoxia-inducible factor 1 is said to play a role in its induction (Ye et al 2012).

Glycoprotein Tenascin, is seen in the fold change ratio graph (Figure 4.25). Tenascin appears to

exhibit a 'peak-trough-peak' response throughout the timecourse. Due to the non

identification of this protein in the fibrosarcoma 120 data set, it is difficult to validate or

comment on this reaction. Tenascin is reported to promote metastasis (Minn et al 2005,.

Oskarsson et al 2011). Studies by Oskarsson et al{2011) link Tenascin expression to the

aggressiveness of breast cancer cells and their ability to further produce lung metastases.

Considering the latter could the last 'peak' previously mentioned be a feature supportive of

the 'switch back to viability concept' overtime after treatment.

Fibronectin precursor is known to bind to Tenascin, interestingly when comparing the fold

change responses (Figure 4.25), both proteins appear to have an inverse fold change

235

relationship; Tenascin 'peak-trough-peak', Fibronectin 'trough-peak-trough'. Further study is

needed to confirm and comprehend this trend of protein-protein interaction.

236

VIM

E_M

OUS

E (1

00%)

53,

688.7

Da

Re

cNam

e: F

ull=

Vim

entin

.

< V LU LU LU a CO

0 D < □ v> o CO

0 - _ l I > u.CL >- Z _ l a z Hcn z Q V < < Q.co < CL * H I < _lCO U LU _J * UJ CL

> DC > Li. Vi _l _ l_j a LU < iL < COVi z > - > u -

CL _] a: LU § z CLVi LU < LU LU UJ CO

a iL a UJ UJ UJVi _l Q _i < 2 UJa. LU z Vi LU UJ 0 UlCL■ i - LU 0 CL UJ _l_1 * _ i > _1 2 _l Q< LU a * z O _J □Vi Z Q CL iL CL iL X

0 L— > LU < UJ CL X

_l cl a _ l < _ l > oVi H a: D > CO t— CO

> z CL _1 Vi UJ < 1-

H X _ i CL LU z - UJCL u LU < > 1- UJ z\- LU cl _l a 0 — -cn L— 2 Vi o o >\- z LU < CL _i _ l aj - - LU z > < < 0> < LU Q Q □ 2 Cl

> Q > > CL > CLVi < D _l LU > L—

CL _J Q O < O z LU

Z Vi CD DC < 1- _J >

Vi u _l IL h- _J _J (—Vi Q CL in _J CO Q iCCL > Vi a Q 0 0 -

CL Vi -1 Cl > > _JVi Q CD 1- X. 0 Ul _lVi O a V) CO CL CL 1-

\- _l CD LU > CL _ l CL0 _l iL < D > X JsCVi cl _J LU > UJ CL Vio > o Hi Q z < X0 0 LU oc — CO 2 Hu. CL _ l o O UJ Ul Q

2 > LU _ i > o Ul >CL Vi < 2 X iL X. _ l

CL Vi _J LU a < 2 a

> CL LU UJ 0 z _j

Vi _l _ O o CL 0 COVi cl X. _ l — _J - UlVi > Z X. o < UJ _ iVi < a LU < Q a z> Vi o CL 0 Z 0 ( -(Ji Vi LU _J _ l Z _ i UJ

a cl _ l CL LU OC CL CLt - H Ll_ 2 O z 0 _ l

co > a: - - < - z2 >- > D UJ < 1- _l

il

t

O Cr

A)!SU0)U| 8A!)B |3y

0000O5o2ss-O£50

i IQ—J5

OfI1££2 4»ucuamI"tsQj

Sit 3 Co «c Q.2 Cn v.0

-QQj01 CoO.

Si■o' rs«u

ccu*C3

QiII

s+J01OQ.

-C«N

Co

coO) cuT3Q.cuac

’•CcQj£

s

Qj■O_oas

CO H=- «N ^'St fsQJ 2

ApU0)U| p0ZI|BUJJON J6C'-|3Cnii

237

HB

B1

M0

USE

(100

%) 1

5.84

0.2

DaR

ecN

ame:

Ful

l=H

emog

lobi

n su

buni

t be

ta-1

; AltN

ame:

X CD< - l

X X CD x 2 CD— _J

< > CO — < > CO —

CO 2_J XQ CDCD - lLL _lCO QC

Q ILu_ z>- Ul1X CL

O ai— >

5 ICL §3>> □> o_l X_l _lDC LU

o CO_ l _l< COUl < XO LL >-

O 1- N£

> CD XUJ X <a _l _lCO CO <

z o 1-

1 -J <

* X >O z CD5 -1 <- I CD >

f t o >

<o X ■> U- o< < IL

R1— <t

* — <UJ > o< X <a X <i- CD CL_ i X 1-X < u.f t X if t2 > 'XL

r-_-Q

CO.n

"t.

n2 r

> JQ~

a .o<£

fte

oo

-T —

5*to oI f )

- r -spO'toCM

OolO

o k >oCM

ooo

oh r-

oolO

oIf)CM

A ) jS U 9 )U | 9 A j} B |9 y

Aj|SU^U| P9ZIIBUJJ0N ! 6o~|

350iis2-Qc

1Q—J

a1i£s3V.+-QJS-S'*

Co

*3Qj4»>.oQjV.■Cc3

-ca2CDV.3 0 .

4Qj ■CO ^

2 +-з . to«o O QJ Q.

tT «N q j rsS -s +-> oc £QJ c^ o<o £ QJ "3

15 .5C "‘T1* D Qj Qj -Qa o 7 a2 ^ ■Jr* ct Qj £-o fc: .t: o c *-3 £

■§ £ co Qj

1 *-O t-M2 ^ 3> cn

5 3

I Irs **CM Qj^ to Qj .CO V. +o ? 00 •S' 00и. t-H

238

«E>NO

n .2 Q «<o V)

2> S

7 >O EOr . o. IU il W 3 D IL

i j

MCL «* cl

2 a j > hi > m z lu -0 5 — <-i >LLJQQOOHJ> WCL Q > _ I Q > < IL0 >O O <<O « X >- OR CD 0 X <i— hi o: □ > o o .q x > < -i< - jO J 0) 0 > > < > -<0U-2<<CLCL

l - > > X - z n J l L > i < U J _ J O < l CL- > > o o > clcd> - cl> O wqqq:> oqczs: — < coco>-q: H > O Q U J O I - _ I 2

— X - J H J - X l -

a : - Z m-<CL<z < clvcoujq: < oi- h o >->qi-_ jclh.

- c d o < h . c l c l < 0

q. < _ l CL a QC CL 1- co< U- $ _ l CL Cl _J 0 0

</>(/)—! □ > X HI CL X

Q HI — > 0 i HI z cl- 1- Z < UJ - LLCL £

Q < Hi < x 2 _) 1- 0

B Hi Q 0 z co 0 > 1-

C C C O CL HI HI _J < -1

O > X _ l - _ l X - >

2 z hi z x 2 >- —x X 2 > m o - cl> HI — > 0 — !-< < >_i i- < x - < < cloh .h jzxxoh ic l0I- < O CO < CL XQ U O z > l l J > - i C<>-i-_ ixclclq: h2 U J - W0 i < < u .< X X 0 X 0 - > 0

<h-_l0HJ<_l0CL 101—Z 0 Q C II< _II<UJ_JZ0<X 0Cd(D>>Q5C/)0 0U-XHJX£LCLQ:>i-z x h q :0 > 0 o0 -I -j > > - < «0 2 — CLUI_JHI5HJX< L -< > H I2 — I <>HIQ>^Q 1 Z

-

1- Z0 2 i- 0 < < d - q:< O 0 ^ < j 0 - j

UiC/)COHJXHX<Clc/j x 0 x c l > < 0 >

X - X > - B w QCL2-0lLlUliJOliJ x h j _ i _ i c / ) <0 o <Q . ^ 0 ( / ) < I L ( / ) v § q m o

^ a> oo oo2 - 1 1 U- - _ l IL <

A}jSUd)U| 8AHB|9y

( O C O I f l O i f i n n N I N r rAl!SU0JU| P0ZIIBLUJON *6CT|

II I

Q- j

S

aIi

£E3v.■U.CjQj

*SO

SO

T3cu

T3C3

- C

Io»

acu

Qjcno.c

■5CU

§3L.X

Q.

<30<N

QJ

v. cn <3 O«Q a cu -cco fvjs S& °<u c o

u

iCL 'O It? cu ■*-1 ~QJ J3

^ 5 o

& a£ 5C

&3 .

§</><u

iO

o+J

Qj

.52 or 5CgQJ3cn.50■C0000*1o

Eo£3cnO3 __& -O

LU d

239

or

I

UJ X CL UJ CO . > - - _ l l - > Z C O > _ J O - C L - l L 2 t - >

0 > - Q ' a _ l _ J < - a x a < 0 O K X K0 «O. Q; C t OT > U. _ l _ l 2 l UC C U. I -> C

< o j q O q: > - j z i u c l u . i- ujcd- ( i i -_ l I I- O UJ 0 X I < _l UJ < H <( J f f l t J J i L U J C L > J O ( ! ) i - v Q m i u -i- a i - 1- x i— < > > q i - q : < > i - <

O t - < < / 3 O 0 U J < i - a > O J - a Q aq; iuu. v _i < < uj> q; < _ ( - 5 o i - c> - < k l - a x < a w _ l i - i - 0 ( / 3 C / 5 0 5co_i P) i- i- > q: < _ i co$ z c o _i q; q 5

u j x < 0 § < z £ < a _ i o : i - a 0 i - _h i l c l O ij-D.CDz i - > z - u. O ci; > <a ^ Z M U - Z H - i O h - U - K l U J a m S U Lat-wxc-M>i-a_j«aoa-wa: J - < 0 . J ^ h J D z ^ 4 H ^ O Q . C

c o _ J 2 > m i i J . < O O a - Q < > - < i -i < Q Q Q D : z Q i - < H O J D J O -K t - > - O 0 I $ > < > C O _ J M U J 0 Z- i- ci: u j h x < 0 0 u j o _i x > - h <u j q q : < j- u j h > q: > m > - 0 . j u j 2 -0 1 - W X I U < U J U J _ 1 Z > > W _ I > - C L _ U J > - Q C O - > < ( - Z Z - Q Z U J > _ Q Q J 0 > J Z > - < J I L > D i - > C L J i : f f l 0 ( i : J z i - > - i u z Q > > < i - K ^ > > O T _ i > O > - 0 0 ( - K U j i u - a

z i u u j - q i - ujujq; > - ( » > i i (i ; d1 02 ->-ax>-00 i-0uja_iuj_jacZ I l U < a W > _ l < > Q . Q . L L < 0 W _Q l K > K - > - < 0 0 Z < > U J _ l 0 U J J - U J <l L l L Q i Q 0 - Q . ^ l L Q ; 0 D I >1 -- < £ L a > l - _ I X < C f l U . - Q O -— d q ; q: ujuju. i- 0 x h - 4 i- q; < _ l U _ l _ l > - l - _ l > 0 a > O T I Z U J K | - H ^ a : z « 4 Q i u i - j < > < u j < a ; o . 0 K I - 0 Q < h - > - W Q _ I Q I - r o D : i l j B c C

a a >j i u M Q i n u u i - Q . w $ ao < a : i - _ i < 5 ^ a a i cQ!/)ji-50ffli-j>a<_i0$aco>owi-u.H - > 0 Z I — t- x - < u. $ a a _ i o : z a 0 i j j < a Z I - > H > I 0 0 0 Z 0j o : i - _ i 0 u j > . i - a u j x

a:>_i00 <aaiiji-2> l - Q I - 0 O J 0 Z Q - l - Z 0 t O - 4i u z - - * > 0 o : o £ > a t - w a o -> Z H f f l > U J Q 4 > t O 0 J Q > 1 ( 0 > « X H z < 2 § > c - > > - - i < 0 -J*aa: i - > Q ; Q a i M 0 0 0 i i J - Q ; - > i - 2 >U J - Q K > O < > H I - C 0 Z I - U J Q - Q :x i 0 O 0 0 u j - t - > > c o _ i > - Q : u j >- _ i a O T - ( - H - C O Q « a Q Z 0 - Y U .

x — QQ; > - _ i i - Q: i i j0 a - _ j 0 — to 0

- X 0 Y < W

Q . O U | V ( C Q□ a x $ lu

•ji 0 i— uj x aUJ (x —I —I Q —I

O► o

< t +oto

%-

--UJ

£-

A)!SU9^I| 9 A p |9 y

r%

(i)<i>inio*e*<np)N«Al!SU9JU| P9Z||EUJJ0N *601

CU3CO•52■D29ooc5ohioCO

ssCoi I5—J

I I- Jor

I

£

5+-CjCUQ.CO

CO

co

"CJCU+J.5"SV.

T3Cc-ca2OiQ-O<ucoCoaCOQjV."O'CU

iS

3

CO0a-cR12

C C Qj O

"T3 V* T3

s ^"o "5■5 43cu —a a•5 S G ocCO h-o •-C OQjK- W). . Qjcncm QJ

cuV.3cniE

240

ors

Aa o» £n c » 5 a 2N C ^ O

S f io Ll Z. JLHI 3 0) U.

f aiiIL tt

2 O w a o o - 2 0 n > J > - i - i L Q ; > W D . o 0 i i _ i Q ; C L « H I - l 0 L j Q ; x a - Q . - i O O l - Q . O > l - ( / ) > D . Q . O ) U J o 5 q ; i - i u < > - O x i - c o d i - i - z q . h > i — </) < - a § ^uj i -0002i i J$<>-a - i - i i : i -<-a )Q. ( / )§coQ. co n ( DQ a z z > i u i uu JC L- Q. i uz 5v a 3 t t ;< DH a j > - a : i : o o m K ' O x O ^ < z < O > i - D . ( D i i J 0 ) > > x [ i : j UJ(DQ>-ZZUJ£i : -$H>OOQ.>C003§UJtLQ:aQ_l > Z Z - O O z Z ( D Q ; i l J « a - _ J _ J O D « ^ D 2 l i J O 0 ) i - 0 o < 0 0 i - i Q : 0 0 i - Q : u _ o ' > O w > O i u t - > a : i -

0 O Z H : i - l - O I - > O T > l - 5 Q . 0 [ i : 0 > - l ^ - C L K i Q ; i -> - jq: > 0 0 - q o i - i- coq; i- x q x i i j > > clO coO(o

Z O W Q I - > l U Z > ^ Z Q . > a ' U I H > K O l > > <> L U I - | - > - I W > O a C 0 _ l > l U ( J l U < O . I - > 0 Z C 0 « ) - H J o i L < > i z O o ; i - Q - o o . z > < Q > u j i - > r > z 2 i - « > i - < < > > § > - i w z - J > - J i i i a . o 5 0 i L _ n u i i j 0 > - _ O i i j i < i- q; q; cl> h i u («ili - O z ><C i: > < - I i - 0 i i ; « O Q 2 - > - l ' - O Q l U > O > O < Z < _ l l U i Q V O O O O I —CLQO — C / ) H Q > t _ I H O L U > > Q I —i q ; ^ H Z < m O O Q l > - < l l J > < J Z > - » - l l U > H 0 l i : 0 .

O U J 0 W > l L l L > > X < / > O O > - > l - l U < L l J h - HC L-----I L I I J J < 0 ( O 0 - U . I - K 0 Q - I I I | 1 ( / M I C _ I 0 - I 1 J - Qo a . 0 « z > - i i . o x - i - i u > i - i | < 0 > z Q . i i ; Q . j > w i - Q ; _ i O 0 i : ^ O i u i u . i i J i - - Q ; D a O i - z > > z 0 > U J Q ; O J - l O O Q § m Q « Q . Z > - I O B c L _ l 0 < l - l - | - l u5 l -<ODQQ: C0a i L_JQ> - 0 J J ^ Q . Z - l - - l - < ll x d | L-0<_IOLUO:iuaO(/5HzLLCOlU — 0 > a O l i l Q Q Q : O l U Q O 0 Q . H > - a > > K Q ; 0 - l l L I - O > < O a 0 > u j j > > t o - O i u < 0 > O J - a i - m i - < Q ; 0 i - 2 O 2 z z > 0 z m > Q a « O £ L 0 ; i - o m J z c L -

c o x 2 ^ 0 o ) O a > - § Q ; o ; a O D < > - a > ( o z i - Q < JU J > § § X H I - - U . O l U . < _ I C L h ^ < > C n _ I H K I > Q .o ( O 0 u j z h h o i — 00 i - Q a j c o a a : 0a a a $ a z O O 0 0 m 0 < Q : a . > - m i - r o i - i . < i - z D i - < o ) i « a 0 > o ; O o o 5 ^ i O > O - i § Z - i $ > < i $ K i - > u i 0 0 a 0 m § § i u u j m o i - c o i - 5 » a i — i c t u j co h ui Q ; i i i x z ^ - i ^ 0 w ) z - Q i - > - J | 0 - a > _ i 0 _ i a Q ; « i u J 0 J i 2 Q ; Q a c D C o m ( O i - Q ; > > < > 0 0 u j t o 0 ; 0 O § H $ 0 Z 0 0 O - < - I Z > 0 Z - Q U J - Z < Z Z0 Q : i - O 0 D Q Z O « - J « - < v 0 > . | - Q ; i u t O _ > < >

> z i u > 2 O Q : 0 O a o > - i H > i O K i i j _ i c o j Q i - < o < 0o 0o : Q : > Q i -05w i u Q > x i i J 0mQHa j — I > 0 > 0 0 O - L U W W 0 a i - < a i - 0 l - l LU Q —

> 0 Z U J > 0 I U Q ; 0 Q ( / 3 X t O >_i o t w z > i- i- q; o - > i luj_j > x 0 i- - j O > - > q >< - ! - O > O O 2 ^ > 0 ( O l U l l i l L l L Q : Z > ( 5 l L D C L l L > -z o : 0 o : 0 c o o 5 5 O Q- cow,- O - l t - > - O z i 0 > - w 0 i u0 0 < Z » > L - l Q : > l - < - ( » J > - J O . - l - J Q > - h l -- i o : < Q Q a > - 0 0 a _ j i - 0 ( / ) Q i - 0 a i - a Q - a : a 0

> 0 x x i - i - z J Q : 0 0 - > - 0 i - > < - > - H i - 0 > m a H < Q ^ > o : Q ; Q 0 i - x i u > i u a 0 i - o ) i - i - j a _ i Q z a O i L i B u O z i O i — i - z > a a j q : > - i - o i u j j j u j - o $ i z z 0 0 0 i - > > o ) a m a j i u - a _ i z 0 z ! $ G X H 0> > x a a > - ( o - a > j > - a > a : > 0h-L-i- O ( - U i D > - l - J Q O > - » I U > - « ) Z 0 0 O O ) - l l l l - J Y J O O < 0 a a a $ i - i - z m x > - Q H < j a 0 a > i u O K z q — < a a 0 o O J H i > « ) — t— — q: — - § a o ; a j a Q z J i < 0 - Q : L U Q ; H _ i - > > - K i 0 o )0 - X V 0 > i 0 - i a 0 > a Q ; a < < a 0 Q i - > - a Q ; z

/tySU9)U| 3AI}E|9y

>i,\

e‘u

-tr,

Ajisusjui paziieuiicN Jf5o~|

0)3co■52■C0000*135oii8

8- ou :

so

so

a

§

£E3V.

CU

&so

so

■3cu

■3c3

■C

I3 )L .3

-Qcu coca -

§■ 8 s a

P§ 8 Qj <4o

5 §CO Cj

■3•S CUQ. Qj

& ■ §

•J§ 5CJ ^cu (Vg g

u . CO

O 2 * «■) QJ

IS

8 3 § , '

3

CO

U .

241

4.5.5 Protein relationship mapping of Fibrosarcoma 188 data using STRING 9.0The proteomic networks to follow will feature the same format as the protein relationship

analysis in section 4.4.5.

242

Figure 4. 31: iTRAQ Fixed modification search using Fibrosarcoma 188 tumours, visualised by STRING 9.0 (Confidence view).

244

Col4a2

A c tn l '

Anxal

Angp?S3Pled

Apoal

S100a6

Abca?"TV-

Anxa2Tubb5

Dnaja3

Hspa4

Faml35a

Prdx4

H istlh lc

Figure 4. 33: Zoomed in region of a fixed modification search, iTRAQ fibrosarcoma 188 samples, visualised through STRING 9.0.

Vimentin, Tenascin and Plectin are all grouped together in this zoomed region from Figure 4.33.

The roles of this trio all involve tumour progression through individual unique mechanisms as

discussed earlier in this chapter. A clear link can be seen with Tenascin and Fibronectin (fn l) to

support findings in the literature reported back in section 4.5.4.

Actin has associations either directly or indirectly with Vimentin, Tenascin and Plectin to

ascertain the ir roles in actin dynamics and general integrity and structural reorganisation.

Another strong association with Actin is the Zyxin Gene anadhesion plaque protein further

linking the latter to a Heat shock 90 containing group (not shown) via the Cingulin gene

(http://string-db.org/). Cingulin is thought to be involved w ith cell junction permeability.

245

p fTnfsfl2Pmpcb

Pkm2

Mdh2

AtpSal

AnxaB /

G m S lG B

Prdxl

Slc'tal

Prdx2

Kcna6

Figure 4. 34: iTRAQ fibrosarcoma 188 zoomed in region of a fixed modification search, visualised through STRING 9.0.

This section of the STRING fixed fibrosarcoma 188 search displays most of the proteins seen in

the 120 STRING results; Galectin, Annexin A5, Peroxiredoxin, Enolase 1. A protein in Figure

4.34 however does not feature in the fixed fibrosarcoma 120 STRING results, this being

Glucose-6-phosphate isomerase (Gpil). It has strong associations w ith hypoxia tolerance

regulator Enolase 1, interestingly Gpil is known to function as a 'tumour-secreted cytokine'

and features in angiogenesis encouraging endothelial cell motility (Ahmad et al 2011).

246

Csnkld H2afzLmna

Dctnl

ENSMUSCHsp90abl

HspOOaal

Cox4i2

MyhlOHsp90bl

Ubc2jl

Pmpcb

Figure 4. 35: iTRAQ fibrosarcoma 188 zoomed in region of a fixed modification search, visualised through STRING 9.0.

Many proteins that feature in Figure 4.35 are already familiar i.e. Heat shock proteins,

Calriticulin, Elongation factor 1 alpha. Perhaps disregarded is the important connection

between Histone 2A and Actin. Although widely acknowledged and easily detected in both

MALDI profiling and imaging, this connection could be regarded as a marker of the re-

organisational transition towards tissue necrosis during anti-cancer treatment. Numerous

associations can be clearly seen branching from these links downstream from Hsp 90 and

Elongation factor 1 alpha.

4.5.6 Protein dose response relationship analysis using String 9.0The follow ing data presented here analyses the protein dose response relationships and

associated pathways using the iTRAQ results from fixed medication searches, employing a filter

score of > 28. Fibrosarcoma 120 and 188 lists were prepared for import into proteomic

pathway software String 9.0 to observe proteins that were flagged either to exhibit a marked

increased expression or decreased expression. This proteomic information was taken from the

original iTRAQ spreadsheets featured in Figures 4.4 and 4.21. The normalised values were

highlighted in either green, yellow or red indicating response relative to the control e ither

increased, no change or decreased respectively.

247

4.5.6.1 iTRAQ F ib rosarcom a 120 dose re la tio n sh ip resu lts us ing S tr in g 9.0The protein pathways featured in Figure 4.36 show the proteins that were found to

demonstrate either increased or decreased responses in the particular treatment sample.

Aldaartl

Up - 24h post CA-4-P

Aldoartl

Gapdh

Up - 0.5h post CA-4-P

TubbS

Tuba8

Down - 24h post CA-4-P

U spell

H2afz

@ ~~ ^Down-0.5h post CA-4-P

0Pdia3

S100a6

H2afz

Figure 4. 36: iTRAQ proteomic response results using a Fibrosarcoma 120 treatment timecourse. Proteins are shown to either increase or decrease relative to the Control fibrosarcoma tissue, relationship links are shown between the proteins that have strong biological associations with each other.

In Figure 4.36 the increased proteins (Up - 24h post CA-4-P, Up - 0.5h post CA-4-P) displayed

do not show any associations w ith each other. However both 0.5h and 24h share increased

Keratin, type II cytoskeletal 1 (Krtl) and Fructose-bisphosphate aldolase A (Aldoartl).

Mentioned earlier in this Chapter, K rtl is thought key player in the mechanism of multidrug

resistance in some cancers (Tang et ol 2012). The unusual increase in GAPDH as previously

discussed is seen here in the 0.5h.

A ldoartl is a component of the glycolytic pathway and increased expression of this enzyme is

found in many cancers including cancer of the lung, kidney and cervix (Mhawech-Faucegliaet ol

2011,. Takashi et ol 1990,. Oijika et al 1991). Perhaps increased response of A ldoartl here is an

248

indicator of a tumour survival strategy against CA-4-P treatment. Elevated levels of this

enzyme could be indicative of a high glycolytic rate, a key factor in tumour survival and

metabolism (Saw 2006,. Beckner et al 1990).

The proteins that are decreased as a result in the 0.5h time point in Figure 4.36 include Tubulin

alpha (Tuba8) and Tubulin beta-5 chain (Tubb5), linked by a strong association. This decrease

could be as a result of disturbance due to the vascular disrupting agent. Although no direct link

is seen between Tubb5 and adjacent proteins Galectin-1 (Lgalsl) and Heat shock protein 60

(Hspdl), a close positioning is observed. The latter could echo the morphological transitions

underway by Tubulin hence induction of Hspdl and decrease in Lgalsl which mentioned

previously to be involved in cell proliferation, apoptosis and differentiation (Banh 2011).

The 24h decreased proteins in Figure 4.36 show a greater degree of protein-protein interaction

with triplet links between Lgalsl, Calcyclin (S100a6) and Vimentin (Vim) and a further linkage

between Elongation factor 1-alpha 1 (Eeflal) and Histone H2A.V (H2Afz).

S100a6, involved in restructuring of the cytoskeleton along with Vimentin and Lgalsl are

decreased in the 24h and appear to be retarded by the effects of CA-4-P. Low levels of Eeflal

in this CA-4-P susceptible tumour model reflect that at 24h the tumour tissue is almost totally

necrotic.

249

SlQ0a6

HspaS

U p - Oh postCA-4-P Down - Oh post CA-4-P

Figure 4. 37: Fibrosarcoma 120 proteomic response using String 9.0.iTRAQ increased and decrease protein response for tumour time point Oh post CA-4-P administration.

In the case of the fibrosarcoma 120 Oh time point there was only one marked early increase,

this being Keratin, type II cytoskeletal 1 (Krtl) (Figure 4.37). This remained elevated

throughout the later time point as shown in Figure 4.36. No associations were seen in the

corresponding decreased set of proteins in the Oh post CA-4-P. 78 kDa glucose-regulated

protein (Hspa5, Grp78) remains low in this early time point and would be expected to be

elevated here in the later time points. However the threshold levels determined in the

spreadsheet featured in Figure 4.4 did not highlight the values to be included in the increased

group of proteins. Upon close observation of the spreadsheet the values w ithin the 'no

change' (yellow) bracket for this protein did show a gradual increase throughout the time

points.

250

Aldoartl

Up com binationGapdh

H2afz

Pdia3

S100a6Eeflal

Down com b ination

Figure 4. 38: iTRAQ combinations of fibrosarcoma 120 increased and decreased proteins. The protein relationships in Figure 4.38 are a representation of combined responses throughout the time points previously discussed.

In summary of the combined responses seen in Figure 4.38, the increased protein combination

depicts the haemorrhagic pharmacological response, the potential multidrug strategy of Krtl,

an uncharacteristic influence on GAPDH and evidence of a need for tumour cell survival due to

the elevated levels of glycolytic enzyme Aldoartl. Amongst the proteins shown to decrease in

Figure 4.38 is a triangular triplicate linkage between Actin, Histone 2A and E efla l. This appears

to reflect the disrupting effects post CA-4-P treatment. The combination seen here

encompassing the transitional changes in cellular integrity and attempts by the drug to retard

growth and induce necrosis. The same effect is true for the other linear trio of proteins w ith

the decrease of Lgalsl, S100a6 and Vimentin. Links between Protein disulfide-isomerase A3

and Hspa5 imply evidence of the stress response and apoptotic processes respectively. Down

251

regulation of Protein disulfide-isomerase A3 was only seen in the 24h time point indicating

that extensive tissue necrosis was already present.

4.5.6.2 iTRAQ F ibrosarcom a 188 dose re la tio n sh ip resu lts using S trin g 9.0

Pfdn2

Rnfl60

Cct6a

Up - 72h post CA-4-P Down - 72h post CA-4-P

Figure 4. 39: iTRAQ proteomic response results using a Fibrosarcoma 188 treatment time course. Predicted functional partners are shown for increased Cct2 (indicated by the arrow) in the 72h time point and proteins that were decreased in the 72h post CA-4-P sample.

In Figure 4.39, the T-complex protein 1 subunit beta (Cct2) is shown with predicted functional

partners which include fellow chaperone proteins. Cct2 is known to have actin/ tubulin re-

organisational properties (http://string-db.org/). This increase could reflect a 'switch back to

viable tissue' strategy with an army of chaperones which favour cell proliferation and tumour

survival. RING finger protein 160 (Zinc finger protein 294, Rnfl60) and Triosephosphate

isomerase (Tp il) are shown to decrease in the 72h time point. No known links are shown

between the two proteins, Rnfl60 can be implicated in cell death and T p il involved in

glycolysis (Orosz et al 2006,.http://www.uniprot.org/).

252

Hist2h4

Hnmphl

HspyQabl

I I r f r l l r l c

Cct2

' ■

Colla2

Alb

Flna

Mem 5

Up - 24h post CA-4-P Down - 24h post CA-4-P

Figure 4. 40: iTRAQ proteomic response results using a 24h post CA-4-P Fibrosarcoma 188 treatment. The proteins shown here depict the responses seen that were either increased or decreased as a result of CA-4-P administration.

In Figure 4.40 there is a strong link evident connecting Actin and Hsp90 flanked by Hbb-y and

Heterogeneous nuclear ribonucleoprotein H (Hnrnphl).

The notable increase in 24h Hsp90 is seen in the immunohistochemical/ MALSI-MSI in Chapter

3, where abundant appears to reach saturation point. Elevated levels of Hsp90 could be

explained by the direct link here to Actin. The changes in Actin dynamics due to the drug could

indeed have a positive effect on Hsp90 expression, due to the need for protection against the

misfolding of Actin. Cct2 is also seen here to be increased as shown in the 72h (Figure 4.39)

response, again involved in Actin re-organisation. The Histone HI increased response could be

another indication of architectural remodelling of the tumour cells. The Haemoglobin subunit

epsilon-Y2 (Hbb-y) is again indicative of vascular disruption with Hnrnphl having mRNA

transcriptional influence on Hsp90.

P le d

Pdcd6ip

253

Rnfl60

Histlhl

VcpTpil

Cct2

Prdxl

_ McmSMem 10

TncHist2M

Up - 0.5h post CA-4-P Down - 0.5h post CA-4-P

Figure 4. 41: iTRAQ proteomic response results using a 0.5h post CA-4-P Fibrosarcoma 188 treatment. The proteins in the 'Up' 0.5h time point do not have any associations here, in the 'Down' group a strong associative link is present between McmlO and Mcm5 with a weaker relationship clear between Prdxl and Tpil.

The protein corresponding to Tenascin (Tnc) in Figure 4.41 is shown to be increased in the 0.5h

time point. This protein was discussed earlier in this Chapter, said to be involved in the

aggressiveness of breast cancer cells and the ir ability to further produce lung metastases

(Oskarsson et al2011). The other proteins in this group have no association here, but all seem

to be a sign of an active proliferating tissue, characteristic of this CA-4-P resistant tumour

model.

There are 2 linkages seen in Figure 4.41, a strong associative link is present between McmlO

and Mcm5 and a weaker relationship clear between Prdxl and Tpil.

The precise function of Protein MCM10 homolog (McmlO) is said to be unclear, it is thought to

play a role in DNA replication and the prevention of DNA damage, but is reported to perform

in different ways dependent on the particular organism (Ricke and Bielinsky 2004,.

http://string-db.org/). DNA replication licensing factor MCM5 (Mcm5) is known to have similar

functions to McmlO (Schultz et al 2009).

254

There is a link between Peroxiredoxin-1 (Prdxl) and glycolytic enzyme Triosephosphate

isomerase (Tpil). As mentioned previously in this Chapter, elevated expression of the

PRDXlgene is said to act as a protective mechanism against H20 2 induced apoptosis (Kalinina

et ol 2012). The level here is shown to be markedly decreased which could relate to the nature

of the resistant tumour type, where at this early treatment time point there is no requirement

as yet for protective mechanism against H20 2 induced apoptosis.

H istlhlc

Hsp90abl

Up combination

Hbb-yCct2

TncHnrnphl

Pdcd6ip

v *Prdxl

'

Rnfl60

Colla2

Tpil

t j

P led

Hist2Mf V I ■ ■

McmlO Mem 5

Hspaa

Down combination

Flna

Figure 4. 42: iTRAQ combinations of fibrosarcoma 188 increased and decreased proteins. The protein response relationships in Figure 4.42 are a representation of combined responses throughout the time points previously discussed in the iTRAQ fibrosarcoma 188 treatment time points.

255

To summarise the combination of groups seen in Figure 4.42, the 'Up' combination of proteins

similarly to those in the Fibrosarcoma 120 proteins includes increased Hb, however this is

where the similarities cease. Tnc is now seen, a protein observed in aggressive cancers and

along with Hsp90, Cct2 gives the notion of a more active, drug resistant class of tumour tissue.

4.5.7 Relation of MALDI-MSI to iTRAQ proteomic responseThe relation of MALDI-MSI to other complementary proteomic results aims to authenticate the

protein induction responses observed. Figure 4.40 demonstrates some of the proteins that

have featured in the previous chapters in the format of an iTRAQ ratio response graph as seen

in Figure 4.22a/b.

iTRAQ ratios showing response relative to the control, normalised to the average of 1.

■ Control

PLECTIN Elongation factor 1- HB alpha HB beta-1 HB beta-2 Hbepsilon-Y2alpha 1

Protein

Figure 4. 43: Bar graph showing protein response relative to the control (113) normalised to an average of 1, using fibrosarcoma 188 tissue. The responses shown here depict a decreased expression if values are below 1 and increased expression if values are above 1 on the bar graph.

Figures 4.44 and 4.45 show MALDI-MSI images which are in agreement to the iTRAQ responses

of Plectin and Hb subunit alpha seen in in Figure 4.34.

256

The time course images shown here in Figure 4.44 and Figure 4.45 match the response

exhibited in the iTRAQ results graph (Figure 4.22a/ 4.22b) and label free quantitative LC- ESI-

MS/MS results for plectin in Chapter 3.

P lectin m / z 9 7 7 Control/saline

.850.5C

24 hours CA-4-P

72 hours CA-4-P

0.5 hours CA-4-P

L515.97

.181.44

250711 V I88 combined 678910 max (MEAN:977.320-977.370)

6 hours CA-4-PSL 1/1

Figure 4. 44: MALDI-MSI multiple Fibrosarcoma 188 sample on tissue digest of Plectin at m/z 977.

257

Corresponding to Hb rrt/z 1819

Control/saline

72 hours CA-4-P

0.5 hours CA-4-P

11_V1$ 3 t0_n-,3«_* « |M & *M 13*9 * ) )6 hours CA-4-PSC I/i

24 hours CA-4-P

Figure 4. 45: MALDI-MSI multiple Fibrosarcoma 188 sample on tissue digest o f Hb subunit alpha at m /z 1416.

258

4.5.8 Concluding RemarksOne of the basic challenges with iTRAQ is not only successful labelling but accurate

interpretation and consistent peptide-protein identifications. Decisions regarding search

thresholds greatly influence peptide/ protein yield and are important to reduce the false

positives scenario. The alternative selection of search parameters to mine the iTRAQ data

herein aimed to assess differences in peptide/ protein yield. The choice was made to use

mainly fixed filtered search data to reduce data volume and ease interpretation/ validation for

the purpose of this project. It would however be beneficial to perform further repeats of the

iTRAQ experiments using the two tumour models to aid validation of proteins found using

variable modification parameters. Over 1000 proteins were identified with the employment of

variable database search parameters, validation of treatment response and elucidation of key

effector molecules could be facilitated in conjunction with iTRAQ data processing packages.

The complimentary/confirmatory technique employed here in this chapter provided a

powerful means of quantitatively assessing a treatment time course. A diverse approach to

biomarker discovery could indeed aid validation of the temporal stability of a chosen

experimental model. The more conventional gel based approaches i.e. difference gel

electrophoresis could be employed in future work along with cleavable isotope-coded affinity

tags involving the biotinylated labelling of cysteine residues (Wu et al 2006).

The experimental work described in the following chapter describes a novel method for

MALDI-MSI image validation, employing an artificial construct as a standard which could be

engineered to include any protein target of interest.

259

ReferencesAhmad A, Aboukamee A, Kong D, Wang Z, Seth S, Chen W, Sarka FH, Raz A (2011)1,2

Phosphoglucose Isomerase/Autocrine Motility Factor Mediates Epithelial-Mesenchymal

Transition Regulated by miR-200 in Breast Cancer Cells. Cancer Research.71. (9)

BanhA, Zhang J , Cao H, Bouley DM, Kwok S, Kong C, Giaccia AJ, Koong AC, Le QT

(2011)Tumor Galectin-1 Mediates Tumor Growth and Metastasis through Regulation of T-Cell

Apoptosis. American Association for Cancer Research.71. (13) 4423-4431

Barber RD, Harmer DW, Coleman RA, Clark BJ (2005) GAPDH as a housekeeping gene: analysis

of GAPDH mRNA expression in a panel of 72 human tissues. Physiological Genomics.21.(3) 389-

395

Beckner ME, Stracke ML, Liotta LA, Schiffmann E (1990) Glycolysis as primary energy source in

tumor chemotaxis.}ourna\ of the national cancer lnstitute.82.1836-1840.

Bjorge L, Hakulinen J, Vintermyr OK, Jarva H, Jensen TS, Iversen OE, Meri S (2005).Ascitic

complement system in ovarian cancer. Br J Cancer 92.895-905.

Capello M, Ferri-Borgogno S, Cappello P, Novellia (2011) F-enolase: a promising therapeutic

and diagnostic tumortarget. FEBS Journal.1064-1074

Chen Z, Wang Q, Lin L, Tang Q, Edwards JL, Li S, Liu S (2012) Comparative evaluation of two

isobaric labelling tags, DiARTand iTRAQ.Analytical Chemistry.84. 2908-2915

Evans C, Noirel J, Yen S, Ow SY, Salim M, Pereira-Medrano AG, Couto N, Pandhal J, Smith D,

Pham TK, Karunakaran E, Zou X, Biggs CA, Wright PC, (2012) An insight into iTRAQ: where do

we stand now?Analytical and Bioanalytical Chemistry.404.1011-1027

Handra-Luca A, Hong S-M, Walter K, Wolfgang C, Hruban R, Goggins M (2011) Tumour

epithelial vimentin expression and outcome of pancreatic ductal adenocarcinomas. British

Journal of Cancer.104.1296 -1 3 0 2

Karantza V (2011) Keratins in health and cancer: more than mere epithelial cell marker.

Oncogene.30. (2) 127-138

Kline KG, Wu CC (2009) MudPIT Analysis: Application to human heart tissue. Methods In

Molecular Biology. 528. 281-293.

Knochenmuss and Zenobi (2003) MALDI Ionization: The Role of In-Plume Processes.

Chemical.Reviews.103.441-452

260

Guy Lahat, Zhu QS, Huang K-L, Suizhao W, Bolshakov S, Liu J, Torres K, Langley RR, Lazar AJ,

Hung MC, Lev D (2010) Vimentin Is a Novel Anti-Cancer Therapeutic Target; Insights from In

Vitro and In Vivo Mice Xenograft Studies. PLoS ONE.5. (4) el0105

Kalinina EV, Berezov TT, Shtil' AA, N. N. Chernov, Glazunova VA, Novichkova MD,

Nurmuradov NK (2012) Expression of Peroxiredoxin 1, 2, 3, and 6 Genes in Cancer Cells during

Drug Resistance Formation.ONCOLOGY.153. (6)879-882

Katada K, Tomonagab T, Satoh M, Matsushita K, Tonoikea Y, Koderad Y, Hanazawa T, Nomura

F, Okamotoa Y (2012) Plectin promotes migration and invasion of cancer cells and is a novel

prognostic marker for head and neck squamous cell carcinoma. Journal of Proteomics.1803-

1815

Liu R, Gong J, Chen J, Li Q, Song C, Zhang J, Li Y, Liu Z, Dong Y, Chen L, Jin B (2012) Calreticulin

as a potential diagnostic biomarker fo r lung cancer. Cancer Immunology, lmmunothrerapy.61.

855-864

McMahon KM, Volpato M, Chi HY, Musiwaro P, Poterlowicz K, Peng Y, Scally AJ, Patterson LH,

Phillips RM, Sutton CW(2012) Characterization of Changes in the Proteome in Different Regions

of 3D Multicell Tumor Spheroids. Journal of Proteome Research.ll. 2863-2875

Mhawech-Fauceglia P, Wang D, KestersonJ, Beck A, de Mesy Bentley KL, ShroffS, Syriac S, r

Frederick P, Liu S, (2011)Kunle OdunsiAldolase mRNA expression in endometrial cancer and the

role of clotrimazole in endometrial cancer cell viability and morphology. Histopathology.59 (5)

1015-1018

Minn AJ, Gupta GP, Siegel PM, Bos PD, Shu W, Giri DD, Viale A, Olshen AB, Gerald WL,

Massague J (2005) Genes that mediate breast cancer metastasis to lung. NATURE

Medicine.436. (28)518-524

Montgomery H, Rustogi N, Hadjisavvas A, Tanaka K, Kyriacou K, Sutton CW (2012) Proteomic

profiling of breast tissue collagens and sitespecific characterisation of hydroxyproline residues

collagen alpha-l-(l). Journal of Proteome Research. ARTICLE IN PRESS

Ning X, Sun S, Zhang K, Liang J, Chuai Y, Yuan Li, Wang X (2012)S100A6 Protein Negatively

Regulates CacyBP/SIPMediated Inhibition of Gastric Cancer Cell Proliferationand Tumorigenesis.

PLoS ONE.7. (1)

261

Noirel J, Evans C, Salim M, Mukherjee J, Ow SY, Pandhal J, Pham TK, Biggs CA, Wright PC (2011)

Methods in quantitative proteomics: setting iTRAQ on the right track. Current Proteomics

8.(l):17-30

Nunez-Cruz S, Gimotty PA, Guerra MW, Connolly DC, Wu Y-Q, DeAngelis RA, Lambris JD,

Coukos G, Scholler N (2012) Genetic and Pharmacologic Inhibition of Complement Impairs

Endothelial Cell Function and Ablates Ovarian Cancer Neovascularization. Neoplasia. 14. (11)

994-1004

Ojika T, Imaizumi M, Abe T, Kato K (1991) Immunochemical and immunohistochemical studies

on three aldolase isozymes in human lung cancer. Cancer 67. 2153-2158.

Oskarsson T, Acharyya S, Zhang X H-F, Vanharanta S, Tavazoie S, Morris PG, Downey RJ,

Manova-Todorova K, Brogi E, Massague J (2011) Breast cancer cells produce Tenascin C as a

metastaic niche component to colonize lungs.Nature Medicine.17. (7) 867-874

Orosz F, Olah J, Ovadi J [2000)Triosephosphate isomerase deficiency: facts and doubts.WJBMB

Life.58. (12) 7 0 3 -7 1 5

Park SY, Yu X, Ip C, Mohler JL, Bogner PN, Park YM (2007) Peroxiredoxin 1 interacts with

Androgen receptor and enhances its transactivation.Cancer Research. 67. (19) 9294-9303

Rand JH, Wu XX, Lin EY, Griffel A, Gialanella P, McKitrick JC (2012) Annexin A5 Binds to

Lipopolysaccharide and Reduces Its Endotoxin Activity. 3. (2)1-6

Reichelt J, Furstenberger G, Magin TM (2004) Loss of Keratin 10 Leads to Mitogen-Activated

Protein Kinase (MAPK) Activation, Increased Keratinocyte Turnover, and Decreased Tumor

Formation in Mice. Journal of Investigative Dermatology.123. (5) 973-81

Remotti F, Fetsch JF, Miettinen M (2001) Keratin 1 expression in endothelia and mesenchymal

tumors: An immunohistochemical analysis of normal and neoplastic tissues. Human

Pathology.32. (8) 873-879

Ricke RM, Bielinsky A-K {2004)Mem 10 Regulates the Stability and Chromatin Association of

DNA Polymerase-a. Molecular Cell.16.173-185

Ross PL, Huang YN, Marchese JN, Williamson B, Parker K, Hattan S, Khainovski N, Pillai S, Dey

S, Daniels S, Purkayastha S, Juhasz P, Martin S, Bartlet-Jones M, He F, Jacobson A, Pappin DJ

(2004) Multiplexed Protein Quantitation in Saccharomyces cerevisiae Using Amine-reactive

Isobaric Tagging Reagents. Molecular & Cellular Proteomics.3.(12) 1154 -1 1 6 9

262

Ruppen I, Grau L, Orenes-Pinero E, Ashman K, Gil M, Algaba F, Bellmunt J, Sanchez-Carbayo

M (2010) Differential Protein Expression Profiling by iTRAQ-Two-dimensional LC-MS/MS of

Human Bladder Cancer EJ138 Cells Transfected with the Metastasis Suppressor KiSS-1 Gene.

Molecular and cellular Proteomics.9. (10) 2276-2291.

Said HM, Polat B, Hagemann C, Anacker J, Flentje M, Vordermark D (2009) Absence of GAPDH

regulation in tumor-cells of different origin under hypoxic conditions in - Wtro.BMC Research

Notes.2. (8) 1-9

Saw RJ (2006) Glucose metabolism and cancer. Current Opinion in cell Biology.18. 598-08

Shiina N, Gotoh Y, Kubomura N, Iwamatsu A, Nishida E (1994) Microtubule severing by

elongation 1 a. Science .266. 282-285

Shultz RW, Lee T-J, Allen GC, Thompson WF, Hanley-Bowdoin L (2009) Dynamic Localization of

the DNA Replication Proteins MCM5 and MCM7 in Plants.Plant Physiology.150. 658-669

Sutton CW, Rustogi N, Gurkan C, Scally A, Loizidou MA, Hadjisavvas A, Kyriacou K (2010)

Comparative analysis among transfected, mock, and empty vector-exposed cell. Journal of

Proteome Research.9. 3891-3902

Takashi M, Haimoto H, Koshikawa T, Kato K (1990) Expression ofaldolaseC isozyme in renal cell

carcinoma. American Journal of Clinical Pathology.93. 631-636.

Tang S, Huang W, Zhong M, Yin L, Jiang H, Hou S, Gan P, Yuan Y (2012) Identification Keratin 1

as a cDDP-resistant protein innasopharyngeal carcinoma cell lines. Journal of Proteomics.75.

2352-2360

Veremieva M, Khoruzhenko A, Zaicev S, Negrutskii B, El'skaya A (2010) Unbalanced expression

of the translation complex eEFl subunits in human cardioesophageal carcinoma. European

Journal of Clinical lnvestigation.41. (3) 269-276

Wu WW, Wang G, Baek SJ; Shen R-F (2006) Comparative Study of Three Proteomic

Quantitative Methods, DIGE, cICAT, and iTRAQ, Using 2D Gel- or LC-MALDITOF/TOF. Journal

of Proteome Research. 5. (3) 651-658

Ye H, Sun L, Xiaojun Huang X, Zhang P, Zhao X (2010) A proteomic approach fo r plasma

biomarker discoverywith 8-plex iTRAQ labeling and SCX-LC-MS/MS. Molecular and Cellular

Biochemistry.343. (1-2)91-99

263

Ye J, Mancuso A, Tong X, Ward PS, Fan J, Rabinowitz J, Thompson CB (2012) Pyruvate kinase

M2 promotes de novo serine synthesis to sustain mTORCl activity and cell proliferation.

PNAS.109. (18) 6904-6909

264

Chapter 5A novel multi-peptide recombinant

standard used in MALDI-IMS-MSI

265

5.1 IntroductionMany developments and innovations have been achieved in the field of mass spectrometry

over the last decade. Technological advances to enhance both mass resolution and mass

accuracy have been accomplished with various pioneering techniques.

Ion mobility has enabled observation of another dimension to mass spectrometry data with a

further separation method of isobaric ions. Nano scale chromatographic instrumentation now

provides a means to perform highly sensitive, high throughput proteomic analysis for both

targeted and undirected investigations (Angel e ta l 2012).

It can now be said that electrospray ionization (ESI) and matrix assisted laserdesorption

ionization (MALDI) are key techniques employable in all aspects of biomarker discovery and

proteomic investigative analysis.

Work by Atkinson et al (2007) displayed overlaid MALDI images using BioMAp which showed

the distribution of the bioreductive drug AQ4N and its active metabolite AQ4 in solid tumours.

Here in this paper, standards where included in the MALDI-MSI using AQ4N, AQ4 and ATP. This

enabled observation of the regional spatial distribution of the drug/ metabolite and whether

any co-localisation was visible.

An article by Djidja et al (2009) studying GRP-78 within pancreatic adenocarcinoma ffpe tissue

sections, reported the technique of using a digested recombinant standard for image

validation. MALDI-MSI with the employment of ion mobility produced BioMap images which

could be related to GRP-78 immunohistochemical staining.

Targeted proteomic MS methods aim to facilitate analyte identification and a novel method

using a multi-peptide recombinant standard for the MALDI-MSI of on tissue tryptic digests is

reported here.

266

5.2 Materials and Methods

5.2.1 Peptide sequences chosen for a recombinant protein standard based on MALDI-MSI and conventional proteomic analysis of fibrosarcoma 120 and 188 tissue.The pET vector detailed here was supplied by MWG-biotech (Germany). Unless otherwise,

stated the chemicals were purchased from Sigma Aldrich (UK).

The peptides listed in Table 9 were chosen based on the Mass spectrometric identifications

reported in Chapters 2-4 (with the exception of EGFR and Epiregulin).

The design of the 12 peptide construct was undertaken by Dr David Smith of Sheffield Hallam

University (UK). Organisation of the sequence was arranged to ensure thatthe charges were

evenly spaced across the synthetic protein in order to prevent intra cellular aggregation during

expression. The final protein had an expected mass of 16,667.26 Da.The DNA construct was

then sub cloned into the expression vector pET23a(+) and E-coli BL21 DE3 was chosen for

transformation.

Plasmid preparation, transformation and purification was performed by Dr David Smith of

Sheffield Hallam University (UK).

GRP-78 1887.9 VTH AVVTVPAYFN DAQREGFR 1564.7 MHLPSPTDSNFYRHSP- 90 beta 1513.7 GVVDSEDLELNISRPlectin 1385.7 AQAELEAQELQR

1350.5 DSQDAGGFGPEDRHaemoglobin beta chain 1302.6 VNSDEVGGEALGR

Tenascin 1219.6 ETFITGLD

Actin 1198.7 AVFPSIVGRPR

HSP-90 alpha 1168.5 LGIHEDSQNR

Vim entin 1093.5 FADLSEAANR

Epiregulin 983.4 CEVGYTGVR

Rho GTPase activating protein 2 844.5 RLTSLVR

Tag reporter A WLE HHH HHH

Table 9: Twelve peptides from target proteins chosen fo r positive identification.

267

MHl PSP TDS NFY RVN SDE VGG EAL GRA VFP SIV GRP RRL TSL VRE TFI TGL DAP RGV VDS EDL ELN ISR LGI HED SQN RFA DLS EAA NRA QAE LEA QEL QRD SQD AGG FGP EDR CEV GYT GVR VTH AW TVP AYF NDA QRAWLE HHH HHH

Figure 5 .1 : pETBIue-1 vector and sequence used fo r the synthesis o f the recombinant protein standard.

MHLPSPTDSNFYRVNSDEVGGEALGRAVFPSIVGRPRRLTSLVR ETFITGLDAPRGWDSEDLELNISRLGIHEDSQNRFADLSEAANR AQAELEAQELQRDSQDAGGFGPEDRCEVGYTGVRVTHAVVTVP AYFNDAQRA IWlLE iHHHHHHl

\ \Tryptophan His Tag

Figure 5. 2: The final sequence of the recombinant standard.

5.2.2 Bacterial cultureBacterium - BL21(DE3) HsdS gal (llts 857 ind l Sam7 nin5 lacUV5-T7genel)

Growth Media - Lurnia-Bertani (LB) media, Bacto-tryptone (lOg/L), Yeast extract (5g/L), NaCI

(10g/L).

The above materials were dissolved in 900 ml deionised water and made up to 1 L. LB media

was sterilised by autoclaving for 20 min at 15 Ib/sq inch. After cooling filter-sterilised

carbenicillin (final concentration of 250 mg/ml) was added as appropriate.

1.5 g of bacto-agar was added to 100 ml of LB media before autoclaving. Antibiotics were

added (carbenicillin, final concentration of 250 mg/ml), to the cooling agar and mixed

268

thoroughly. 15-20 ml of molten agar mixture was added per 100 mm diameter Petri dish and

allowed to set, before storage at 4 °C.

Cultures were inoculated from freshly transformed bacteria (Section 2.2.7) grown on agar

plates (Section 2.2.3) or directly from existing glycerol stocks (15 % v/v glycerol). Liquid

cultures of less than 100 ml were inoculated directly, whereas for larger scale cultures (1 L),

overnight starter cultures of 5 ml were used for inoculation. Cultures were incubated at 37 °C

in an orbital incubator with shaking at 200 rpm.

5.2.3 Plasmid preparationPlasmid DNA was isolated from fresh overnight liquid cultures of E. coli BL21(DE3) with

carbenicillin selection using a Qiagen plasmid midi kit as described in the manufacturer's

instructions. Isolated DNA was deemed of suitable purity if the observed ratio of the

absorbance at 260 nm (A260) to that at 280 nm (A280) w as3 1.8. An A260 of 1.0 was assumed

equivalent to 35 mg/ml of single stranded (ss) and 50 mg/ml of double stranded (ds) DNA.

5.2.4 E. coli- transformationTubes containing 200 ml aliquots of competent cells were thawed at 4 °C and 20 ng or 200 ng

of plasmid DNA were added to separate 14 ml round bottom Falcon tubes.

A negative control lacking DNA was also included. Cells / DNA mixtures were incubated on ice

for 30 min and then heat shocked at 42 °C for 90 sec with mixing every 10 min. Cells where

then put on ice for 2 min and 800 ml of liquid LB media was added to each tube. These cultures

were incubated at 37 °C for 45 min to allow the cells to express antibiotic resistance.

Transformed cells were spread onto LB / agar plates containing carbenecillin (250mg/ml) and

incubated at 37 °C for 12 h.

A 20 ml starter culture was grown overnight at 37 °C in liquid LB media (Section 2.2.2) with

antibiotic selection (carbenecillin, 250 mg/ml). This culture was used to inoculate 10 x 2 L

flasks each containing 1 L of LB liquid media and carbenecillin (100 mg/ml). Cultures were

grown at 37 °C with shaking until the OD600 reached 0.6, protein expression was induced with

1 mM (IPTG). The cells were then incubated at 37 °C overnight and harvested by continuous-

flow centrifugation at 15,000 rpm at 4 °C (Heraus Contifuge 17RS ). The cell paste was frozen

and stored at -80 °C until required.

5.2.5 Protein PurificationThe required DNA sequence was synthesised by MWG-biotech (Germany), and sub cloned into

the bacterial expression vector pET23a(+) (Novagen). The plasmid was then transformed into

269

the E-coli strain BL21 DE3 (Promega Corporation). All bacterial cultures were grown in Luria

Broth(LB) media with 100 jag/ml ampicillin for selection. 10 ml of a 100 ml overnight starter

culture was used to inoculate a 1 L of LB media. Cultures were grown at 37 °C with shaking

until the OD600 reached 0.5, protein expression was induced with 1 mMisopropyl-(3-D-

thiogalactoside (IPTG). The culture was incubated for a further 2.5 h before the bacterial cells

where harvested by centrifugation at 10,000g for 20 min using a Sorvall RC G+.

Bacterial pellets where resuspended in binding buffer (20mM Sodium Phosphate + 20mM

imidazole pH 7.4)3ml per lg of cells before being lysed by sonication (lOx 20sec pulses ata 35%

duty cycle) using a Vibra Cell VCX 750 (750 Watts). Phenylmethylsulphonyl fluoride (PMSF) (50

|ng/ml), DNase (20 jug/ml), RNase (20 |ig/ml) to the lysate which was then cleared of insoluble

material by centrifugation at 8,000 g for lh using an Eppendorf5804R centrifuge. The clear

lysate was filtered through a 0.2 jam syringe filter and load onto a 1 ml Histrap FF column (GE)

at a flow rate of 1 ml/min. The protein was eluted over 20 column volumes using a 0 to 100%

gradient into elution buffer (20mM Sodium Phosphate + 500mM imidazole pH 7.4). Fractions

containing the recombinant protein where then desalted into 50mM ammonium bicarbonate

using PD10 column (GE). Trypsin digestion was performed on a 50 |al sample of desalted

protein using 1 j l x I of trypsin (lm g/m l) added for lh plus a further 1 j j . 1 of trypsin overnight.

Samples were stored at -80C until further use.

5.4.6 DigestionTrypsin was added (lul) (20ug/ml) for lh plus a further lu l of trypsin for overnight incubation

37°C and 5% C02.

270

5.4.7 MALDI-MSI - Chemicals and Materialsa-Cyano-4-hydroxycinnamic acid (CHCA), aniline (ANI), ethanol (EtOH), chloroform (CHCI3),

acetonitrile (ACN), octyl-a/b-glucoside (OcGIc), tri-fluoroacetic acid (TFA), ammonium

bicarbonate, were from Sigma- Aldrich (Dorset, UK). Modified sequence grade trypsin (20 pg

lyophilised) was obtained from Promega (Southampton, UK).

Tissue samples

Mice were injected sub-cutaneously in the flank with a 50 pi tumour cell suspension containing

1 x 106 cells in serum-free medium. The cells employed in this study were from the mouse

fibrosarcoma cell line, VEGF188. This has been engineered to express only the VEGF188

isoform. Tumours were allowed to grow to approximately 500 mm3, before CA-4-P treatment

(a single dose of 100 mg/kg i.p). The dosage for CA-4-P are consistent with animal studies

performed at clinically relevant doses (Galbraith 2003., Prise 2002). Mice were killed and

tumours excised at various times after treatment. All animal work carried out documented

herein was performed by Dr. J. E. Bluff, Tumour Microcirculation Group, University of Sheffield,

UK. Samples were provided as frozen excised tumours and stored at -80°C.

Experimental sections for on tissue trytic digestion and MALDI-MSI

6_2 Control (no treatment, saline i.p), 8_1 (6 h after treatment with CA-4-P) and 10_1 (72 h

after treatment with CA-4-P).

Tissue preparation

Frozen tissue sections were cut to give approximately 10pm sections, using a Leica CM3050

cryostat (Leica Microsystems, Milton Keynes, UK) (Djidja et al 2009). The sections were then

freeze thaw mounted on poly-lysine glass slides. Mounted slides were either used

immediatelyor stored in an airtight tube at -80 °C for subsequent use.

In situ tissue digestion and trypsin deposition

The tissue samples were washed initially with 70% and then 90% ethanol for 1 min then left to

dry, subsequently slides were immersed in chloroform for 10 s. Prior to matrix application,in

situ tissue digestion was performed with trypsin solution prepared (from lyophilised trypsin) at

20 pg/ml by addition of 50 mM ammonium bicarbonate (NH4HC03) pH 8, containing 0.5%octyl-

a/b-glucoside (OcGIc) based on a protocol by Djidja et al (2009).

271

The "Suncollect" (SunChrom, Friedrichsdorf, Germany) automatic pneumatic sprayer was used

to spray trypsin in a series of layers. The sections for MALDI-MS and MALDI- MSI were

incubated overnight in a humidity chamber containing H20 50%: methanol 50% overnight at

37°C and 5% C02.

5.4.8 Methods and instrumentation Matrix deposition

The matrix, a-cyano-4-hydroxycinnamic acid (CHCA) and aniline in acetonitrile:water:TFA

(1:1:0.1 by volume), was applied using either the Suncollect (at 5 mg/ml). Identical coordinate

settings were used as with the trypsin deposition, to ensure sample uniformity. Equimolar

amounts of aniline were added to the CHCA solution, i.e. 1 ml of 5 mg/ml CHCA solution

contained 2.4 pi of aniline. These matrix deposition parameters were based on methods from

Djidja et al (2009).

Instrumentation

MALDI- IMS/MS and MALDI- IMS/MSI were performed using a HDMS SYNAPT™ G2 system

(Waters Corporation, Manchester, UK) and Mass Lynx software (Waters Corporation, UK).

Image acquisition was performed using raster imaging mode at 100 pm spatial resolution,

Biomap 3.7.5.5 software (http://www.maldi-msi.org/) and HDI 1.1 software was used for

image generation. Instrument calibration was performed using standards consisting of a

mixture of 217 polyethylene glycol (Sigma-Aldrich, Gillingham, UK) ranging between m/z 100

to 218 3000 Da prior to MALDI-IMS-MSI analysis. HDI 1.1 parameters were set to the following;

specificity type - IMS MS, MS, Number of most intense peaks -1 0 0 0 , resolution -1 0 ,0 0 0 , Low

energy intensity threshold - 50. The low intensity threshold was to allow low abundant species

to be potentially included. HDI images were chosen for analysis based on the highlighted

recombinant standard and m/z ratio listed in the software displaying the peak intensity and

corresponding drift time. HDI correlation filter tool was set to R1 minimum 0.5 and R

maximum 1.0 for peaks analysed at m/z 1350 and m/z 1032.

To enable simple visual comparison between images, BioMap data were normalised to m/z

877/ m/z 1066 (peaks arising from the aCHCA matrix).

272

5.3 Results and Discussion

5.3.1 MALDI-IMS-MS Peptide mass fingerprint of the digested recombinant DNA constructThe following MALDI-IMS peptide mass fingerprint is from a tryptic digest of the recombinant

synthetic DNA construct, used as a protein standard for MALDI-IMS image peptide validation.

The PMF in Figure 5.3 displayed a clean tryptic digest spectra with chosen peptides easily

observable showing excellent spectrum intensity.

273

O' <v 00

-J o CO <J3

■ r°> d ; T-

O

0C CL

CD

‘L "a1

- oL O- co

EO

oo

— <1. li.

—3 - O

o

^ n rs9 5-

o

Figu

re

5. 3:

PMF

of a

nove

l re

com

bina

nt s

tand

ard

for

MA

LDI-I

MS

-MS

I va

lidat

ion.

5.3.2 MALDI-IMS-MSI of fibrosarcoma 188 tumour tissue, Control and post CA-4-P treatm entThe follow ing MALDI-MSI was acquired at 100pm resolution using various fibrosarcoma 188

treatment time points to investigation this novel multi-peptide recombinant standard.

The images featured in Figures 5.4- 5.7, display an assortment of peptides from the digested

construct visualised by BioMap imaging software, therefore there is no ion mobility

component to ion selection here. Such images feature further on within this Chapter.

CHCA m /z 8 7 7 -

post 72h post CA- .

4-P ■

*(P_291D.C

170113 10 1 recomb max (MEAN :S77.015-877.055 ) SL 1/1

A c tin m /z 1198 -

post 72h post CA-

4-p f . i

1-4.33

170113 10 1 recomb max div (MEAH:1198.70-1198.731 SL 1/1

T e n a s c in C m /z H H is tone 2A m /z

1 2 1 9 - p o s t 72h

post CA -4-P9 4 4 - post 72h

post CA -4-P

170113_ 10_ 1 _recomb_ma_div (ME*N: 1213.57-1213.70) SL 1/1 170113_10_1 _recomb_ma* (MEAN:944.535-944.575) SL 1/1

Figure 5. 4: MALDI-MSI of on tissue tryptic digests and digested recombinant standard in 72h treated tissue. The spatial distribution o f peptides; Actin (m/z 1198) with arrow indicating position of the spotted recombinant standard, Tenascin (m /z 1219), Histon 2A (m /z 944) with bracket showing the absence o f the standard and CHCA peak at 877 to show inverse image correlation.

275

276

Figure 5.5 - The spatial distribution of peptides; Rho GTPase activating protein 2 (N- chimaerin m /z 844), GRP-78 (m /z 1887), HSP-90 a (m /z 1168), Histone H4 (m /z 1325) displaying absence o f the standard to validate recombinant signal and Plectin (m /z 1385).

CHCA m /z 1066 - post 6h post CA -4-P

190113_8_1_recomb_ma>: (JulEAN: 1066.21-1066.33) SL 1/1

A c tin m /z 1198 - post 6h post CA -4-P

190113_8_1 _recomb_maK (MEAN: 1198.78-1198.90) SL 1/1

H istone H3 m /z 1032 - post 6h post CA -4-P

Figure 5. 6: MALDI-MS images displaying peptides distribution in fibrosarcoma 188 6h treated tissue, employing a DNA artificial construct for image validation. The spatial distribution o f peptides; Actin (m/z 1198) with arrow indicating position o f the spotted recombinant standard, Histone H3 (m/z 1032)showing the absence of the standard and CHCA peak at 1066 to show inverse image correlation.

277

CHCA m/z 877- I Actin m/z 1198 - I Grp78 m/z 1887 -

Control . J Con trol/Saline _ C o n tro l/S a lin e

278

Figure 5.7 - The spatial distribution of peptides; Actin (m/z 1198) with arrow indicating position of the spotted recombinant standard, GRP-78 (m/z 1887) displaying apparent low intensity as seen in Chapter 3 Figure 3.31 also usingFibrosarcoma 188 time course results post CA-4-P treatment except via LC-ESI-MS/MS., Vimentin (m/z 1093), Histone 2A (m/z 944) showing the absence of the standard and CHCA peak at 877 to show inverse image correlation.

The MALDI-MSI in Figures 5.4 - 5.7 all display good validation of the multi-peptide standard

due to the absence in signal intensity of peptides; Histone 2A, Histone H3 and Histone H4

which are typically high abundant within on tissue tryptic digests. These peptides were not

chosen for the construct and therefore co-validate the MALDI images in conjunction with the

recombinant standard.

Interestingly, there is a signal for Rho GTPase activating protein 2 in the fibrosarcoma 188

tissue here (Figure 5.5), albeit low intensity. Some of the proteins identified by ESI-LC-MS/MS

and iTRAQ LC-MALDI are highlighted in the tumour tissue images which again are of low

intensity. This reiterates the apparent ion suppression discussed in previous Chapters, a factor

typical of high abundant species in MALDI digest spectra. It has to be considered what ions of

interest are present within the so called 'spectral grass' and in future experiments how this

data can be amplified for analysis.

279

5.3.3 Recombinant MALDI-IMS-MSI visualised through HDI 1.1 softwareHDI 1.1 imaging software is a useful tool for the processing and visualisation of MALDI-IMS-MSI.

Interactive spectra and drift time plots enable further confirmation of species of interest.

The HDI analysis that follows has the capacity for the user to preset threshold and peak picking

parameters to facilitate the analysis of low abundant species.

The peptides Actin, HSP-90 and Histone H3 are featured in the following HDI analysis using the

identical recombinant standard MALDI-MSI results in Figures 5.4 - 5.7.

The HDI images here are shown as superimposed images on top of the original histological

section used for on tissue tryptic digestion and CHCA deposition.

280

A n a ly te s Functions jl H □ O : © V X V IlC * lei @ I -

(sju

no

ol A

j|sua

)ii|

exportation.

CD

(3

Q, "

Q.

UQj&?O4sc<uu

■Ococ

- §

£

2QJ:C

*1 SI

- .CoQC■fcQJ+->o

</><b

5

£2c£

V)Xc, 1

« c

I -§ O

2 1^ -CI gI*t; c o o 9- P* o(U (X>. ■*»• q3 ^

v fc . ^

rs (X)oo Q-CT) o *H -C '-i o> o£ sc o•S 5 t> «- ^ o'

°9 g ^ S,QJ .O 3 O.•S' 2U . (U

(N00rsl

> \ T M • • - d n M lr / . I A.ft.5 | E \ k L W f * 1 v : 1

.

■■■■ r * * * V '

« J *l 5 m r t

/ ifflC5cz\fli i*

(s)uno3) Aijsusuil

283

(siunoo) Aiisuajui

ocr>iQ_iS)

CL

LLl

o

Figure 5.11: HSP-90 a at m/z 1168.5 after exportation to Mass Lynx software from the selected ROI in HDI 1.1.

285

(£v<Hl (siunoo) Aiisuaiui

0)■c

[N2

I■*oQIQjC0 +->

1

co£o•52~o3S53:IN

^ "58 ■§ 3 C CJ> 34»u. </>

286

(Cv

Ol) (sju

no

o) A

jisua

jui

Figure 5.13: HDI visualisation of Actin at m/z 1198.7. The arrow indicates the ion of interest separated by drift time.

287

A n a ly te s F u n c tio n s I: \ □ o ./ O V X < He * U @ |

(CvO

li (sum

oa

) Atisua

jui

288

The MALDI-IMS-MSI visualised by HDI 1.1 software in Figures 5.7 and 5.9 gives a good

conceptual illustration of separation by drift time. Furthermore how this unique technology

can benefit MALDI image analysis and validation due to the vast dissimilarity in the

corresponding images for m/z 1198.7 and m/z 1198.6.

Another useful tool of the HDI software is the image correlation filter as shown in Figures 5.15

and 5.16. After selection of a particular ion of interest, running the image correlation filter

option displays the peaks within the spectrum that have the same spatial distribution pattern

as the peak picked. Therefore peaks that co-exist are observable. The following HDI screen

images are by using this option and show examples of Plectin m/z 1350 and Histone H3 m/z

1032. The correlation peak lists generated by selecting m/z 1032 mostly contained other

histones i.e. Histone 2A m/z 944.5, which confirms the correlation between the spatial

distribution of other Histones. However this was unsuccessful for Plectin m/z 1350 as fellow

Plectin peaks were not selected. This was thought to be due to the strong peptide signals from

the spotted recombinant standard which where all listed in the peak list correlation results. In

order to discriminate between each peak when using a recombinant standard in future work,

the standard could be sprayed directly to the tissue when using the HDI correlation filter tool.

Teamed with the recombinant digested construct, the MALDI-IMS images reported here

therefore propose a novel for methodology for inclusion into the MALDI-MSI workflow.

The potential for this dynamic recombinant standard is very promising and could be

engineered to include any prospective biomarkers for both research and diagnostic screening

applications in the clinical workplace. This concept alone is in accordance with the forward

momentum of MS applications.

289

Ana

lyte

s Fu

nctio

ns

(siunoo) (tysueiui

3:£.o

Oi.cSxC

APM2

50)co

4m

£V.

QJ4-1

c.o4mo

£ou<uClc

£

8ic£

$</>suS

lo8i3Cl

£

29

0

" 8 (G v O I-J ( s u in o o ) A1|SU01U|

291

5.3.4 Cross validation of Plectin and HSP-90 response in fibrosarcoma 188 tissue employing a multimodal proteomic strategy.

The confirmation and validation of any scientific hypothesis is regarded a challenge, a range of

complementary techniques could indeed affirm marked responses seen in data generated.

The changes in the levels of Plectin throughout the treatment time course displayed a high

percentage of co-variance in the LC-ESI-MS/MS, with the same trend visible in the iTRAQ and

immunohistochemical results of the fibrosarcoma 188 data set.

The MALDI-IMS-MSI of plectin that follow include images from the Control fibrosarcoma 188

tissue and the 72h post CA-4-P time point with recombinant digested standard.

Plectin m /z 1385 -

C on tro l/S a line

• % « .I

'.V. . . - . • v• % * ♦. v

, A I *7 * v '■'< * ' t : • *>1 \» • .» .* ' V .

7. t « *-» K .* ' *

15.46

Plectin

m /z 1385 - post * * *m ___? 13

72h post CA-4-P f * .

* *

\*

Figure 5 .17: MALDI-MSI o f fibrosarcoma 188 on tissue tryptic digests fo r the spatial distribution o f Plectin at m /z 1385.

The spatial distribution of Plectin in Figure 5.17 is in good agreement with the anti-Plectin

immunohistochemical staining featured previously in Chapter 3.

Figures 5.18 and 5.19 use Plectin and HSP-90 to show some of the cross validation that the

techniques in this project aimed to achieve. The samples used for ESI-LC-MS/MS and iTRAQ

were performed on different biological sample time course sets but are still in good agreement.

292

A«U©JU|

If n)|Oiiun W

|| 01 tiuodui ottu q«U!

Figure 5.18: MALDI-MSI and conventional proteomic techniques to asses a dose relationship in Plectin.293

itMM

J

Aji$u®UJ| »Aijc|ay

Figure 5.19: MALDI-MSI and conventional proteomic techniques to asses a dose relationship in HSP-90.

5.4 Concluding RemarksPrediction of the future requirements to progress MS experimental analysis could focus on

further technological advancement of ion mobility for proteins. In scientific publications, much

of the subject matter has focused on pharmaceutical and metabolomic analysis (Thelen and

Miernyk 2012).

Future progression in the separating capacity of ion mobility could improve peptide resolution

thus peptide identification in MALDI-MS, during the acquisition of complex tryptic digest

samples. This could potentially result in a peptide yield comparable to high throughput

shotgun LC chromatographic systems.

If such advancements in ion mobility separation come into fruition, this technology along with

many other additional components could reveal numerous concealed biological molecules.

Optimisation in sample preparation, solution chemistry and enzymology could also result in

prevention of lost proteins to minimise the suppression and spectral dominance by

haemoglobin, structural proteins and such like (Lapthorn et al 2013).

The chart in Figure 5.20 contains points to consider when performing on tissue enzymatic

digestion for MALDI-MSI. The issues and factors included are mainly based on personal

experience throughout the project reported herein. Insights were also gained from an oral

presentation by Hamzah Neil Freeman during a worshop entitled 'On-Tissue Enzymatic

Digestion for MALDi MS Imaging' (H Freeman 2012).

The recombinant protein reported here could hold great potential for quantification

techniques; N15 labelling is one such method for relative quantification. Labelled and

unlabelled samples can be combined and analysed using LC-MS/MS. The only difference

between the two samples being their mass, this then will be used to quantify the peptides

within the sample. The latter is a novel method that could also be applied to MALDI-MSI where

the N15 labelled recombinant could be applied as an internal standard directly to tissue

section and used for peptide quantification analysis.

295

!s a washing step required

for buffer salts

Figure 5. 20: Issues and factors to consider when performing on tissue digestion for MALDI imaging experiments.

ReferencesAngel TE, Aryal UK, Hengel SM, Baker ES, Kelly RT, Robinson EW, Smith RD (2012) Mass

spectrometry-based proteomics: existing capabilities and future directions. The Royal Society of

Chemistry.41. 3912-3928

Atkinson SJ, Loadman PM, Sutton C, Patterson LH, Clench MR (2007) Examination of the

distribution of the bioreductive drug AQ4N and its active metabolite AQ4 in solid tumours by

imaging matrix-assisted laser desorption/ionisation mass spectrometry. Rapid Communications

in Mass Spectrometry. 21.1271-1276

Djidja M-C, Claude E, Snel MF, Scriven P, Francese S, Carolan V, Clench MR (2009) Maldi-ion

mobility separation-mass spectrometry imaging of glucose-regulated protein 78 kDa (Grp78) in

human formalin fixed paraffin-embedded pancreatic adenocarcinoma tissue sections.Journal of

Proteome Research. 8. (10) 4876-4884

Freeman HN (2012) Optimisation of a low-volume on-slide protocol for imaging collage-specific

peptide fragments. Workshop - On-tissue enzymatic digestion for MALDI MS Imaging.

Stevenage BioScience Catalyst, GlaxoSmithKline.

Lapthorn C, Pullen F, Chowdhrylon BZ (2013)Mobility Spectrometry-Mass Spectrometry (IMS-

MS) of Small Molecules: separating and assigning structures to ions. Mass Spectrometry

Reviews.32.4 3 -7 1

Thelen JJ and Miernyk JA (2012) The proteomic future: where mass spectrometry should be

taking us. Biochemical Journal.444.169-18

297

Chapter 6Conclusion and Future Work

298

The aims of the study reported in this thesis were to develop and utilise mass spectrometry

imaging techniques (MALDI-MSI), in combination with conventional proteomic methodologies,

to investigate protein induction in vascular-targeted strategies.

Proteins thought to be involved in tumourigenesis and drug treatment resistance and were

observed along with the responses from proteins identified via the techniques used, in this

global analysis study.

The collaboration between the techniques used; MALDI-MSI, ESI-LC-MS/MS, LC-MALDI-MS/MS

with iTRAQ labelling was intended to provide cross validation of the effects post

administration of vascular disrupting agent CA-4-P.

During the initial experimental work reported in Chapter 2 using MALDI-MSI, one of the major

challenges was the management and analysis of the large amounts of data generated. The

latter especially applicable to ion mobility MALDI-MSI.

The gross haemorrhagic pharmacological response elicited by CA-4-P was visible by MALDI-MSI

throughout the fibrosarcoma 120 time course. The latter encouraged the prospect that other

proteins could potentially be observed induced via a dose response relationship.

It was hoped that the use of PCA-DA, PCA and PLSDA would reveal more subtle changes taking

place in the tumour samples however at present these are masked by the dominance of the

changes in the haemoglobin signals and other proteins of high abundance. MALDI-MS of the

fibrosarcoma 120 tissue in Chapter 2 was one such example, most likely to be due to the CA-4-

P susceptible nature of this tumour model.

The manually performed MALDI-IMS-MS/MS in Chapter 2 contained numerous signals which

may had been post translationally modified in ways other than what was pre-selected for

searching, thus hampering peptide identifications. Many of the MS/MS spectra studied clearly

contained overlapping MS/MS fragmentation patterns.

The Fibrosarcoma 188, a CA-4-P resistant experimental was introduced in Chapter 3. The use

of immunohistochemistry and label free LC-ESI-MS/MS shotgun proteomics aimed to

compliment and aid interpretation of the MALDI-MSI data.

The experimental work carried out using LC-ESI-MS/MS and label free quantitation revealed

many proteins connected with necrosis, cell structural reorganisation, polymerisation, tumour

survival and stress induced molecular chaperones. The inverse correlation of structural

proteins, haemoglobin and heat shock molecular chaperones gave the required validation and

identification to relate these responses to those seen in MALDI-MSI.

299

In addition to this the striking differences between HSp-90 and Plectin showed the distinct

regional differences in the MALDI-MSI, immunohistochemistry and LC-ESI-MS/MS results.

The haemoglobin time course using the resistant 188 tumour model gave quite different

results to those previously seen in the MALDI-MSI of the fibrosarcoma 120 data set. The first

indication of the 'switch back to viability' concept was revealed. This regenerative attempt by

the tumour could also be correlated to the HSp-90 and Plectin responses recently discussed.

The relationship pathways generated by using STRING 9.0 proteomic network software gave

an invaluable insight into the activity of the active tumour milieu and provided a means of

linking the identified proteins to their functional partners. Protein-protein interactions could

be observed to help interpretation of the MALDI-MSI and LC-ESI-MS/MS response graphs.

One such pathway revealed a link incorporating Ras GTPase-activating-like protein (IQGAP1), a

protein thought to be involved with tumour progression and metastasis. The decreased

response from IQGAP1 over the time course suggests an inhibitory effect by CA-4P. Similar

responses to treatment were exhibited by Plectin and Transforming growth factor-beta-

induced protein ig-h3 (Tgfbi). Future anti-IQGAPl/ anti-Tgfbi immunohistochemistry of the

fibrosarcoma 188 tissue could determine whether the staining of these two proteins reveals

co-localisation with Plectin, potential implicated in the 'viable tissue switch'.

Another protein that would be expected to show an inverse correlation to Plectin/

IQGAPl/Tgfbi, is the protein Alpha-enolase. Increased expression of Alpha-enolase is said to be

indicative of aggressive tumour progression as claimed of Plectin. However in these data

Alpha-enolase showed an increase in abundance overtime whereas Plectin quite markedly

decreases.

Over a thousand proteins were present in the LC-ESI-MS/MS results but only a fraction were

selected by Scaffold 3 proteomic software for analysis. Conclusions were made that further

optimisation not only during sample preparation but throughout data-dependent acquisition

mode could prove invaluable to advance proteome analysis and coverage. This would result in

a greater number of unique peptides per protein.

An opportunity to analyse the fibrosarcoma 120 samples with LC-ESI-MS/MS could serve to

complete the comparison between the susceptible and resistant tumour models

iTRAQ. quantitation of both fibrosarcoma 120 and 188 samples was performed in Chapter 5.

Similarly to the LC-ESI-MS/MS results, numerous proteins were identified and various PTM

parameters were employed in order to reduce and affirm the relative responses observed.

300

Different biological tumour samples were used compared to those of the ESI study with an aim

to validate overall responses previously seen.

The main differences in the responses observed between the two tumour models employed

(120 and 188) here were that of Elongation factor 1-alpha 1 and Vimentin.

In the fibrosarcoma 188 samples Elongation factor 1-alpha 1 levels are elevated throughout

the time points relative to the control. In contrast to this, levels in the fibrosarcoma 120 iTRAQ

results are all decreased relative to the control. It was thought that elevated levels seen in the

fibrosarcoma 188 samples maybe due to some form of pharmacological conflict elicited by this

CA-4-P resistant tumour model.

Vimentin levels decrease overtime in the fibrosarcoma 120 iTRAQ response results with a

more erratic response leading to an increase seen by the 188 results.

Overall the dose relationships observed in the iTRAQ data by the proteins involved in

haemorrhaging, structural remodelling, tumour survival and the stress response were in good

agreement with the other techniques employed here. Further optimisation and/or more

experimental repeats could both ascertain the relative responses seen and achieve optimum

percentage of iTRAQ labelling.

Although a widely practised technique, it was concluded that MALDI on tissue digestion

imaging still has many potential courses for optimisation. The chart in Chapter 5 aimed to

summarise points and issues for consideration. The requirement for the best buffer, enzyme,

solvent and matrix could indeed reveal numerous concealed biological molecules, essential in

the understanding of cancer. Additionally, another consideration in the assesment of

proteomic dose response relationships using MALDI-MSI is the effect of post translational

modifications. The apparent disappearance of a protein within an image time course could be

due to such alteration i.e. phosphorylation of peptides due to having a serine/

threonine/tyrosine present in the sequence or having undergone methylation/acetylation.

MALDI-MSI has the potential to forge a place in the workflow of clinical diagnostics. Targeted

approaches for the observation of disease biomarkers could be visualised using MALDI-MSI

and serve as a complimentary technique to standard clinical imaging.

The novel method reported here using a multi-peptide recombinant standard could prove an

important diagnostic tool for the analysis of patient biopsies and tissue micro-arrays. The

exciting prospect is the diversity of a multi-peptide recombinant standard. As mentioned

301

earlier in Chapter 5, the artificial construct can be engineered to include any prospective

biomarkers for both research and diagnostic screening applications.

This initiative could indeed further the impetus of MS applications in cancer research.

302

Appendices

304

Appendix 1: PublicationsCole LM, Mahmoud K, El Enazi M, Mohamed S, , Tozer GM, Smith DP, Clench MR (2013)

Recombinant "IMS TAG" Proteins - A new method fo r validating bottom up matrix assisted

laser desorption - ion mobility separation - mass spectrometry imaging 9MALDI-IMS-MSI).

Rapid Communications in Mass Spectrometry. In Press

Cole LM, Djidja M-C, Bluff J, Claude E, Carolan VA, Paley M , Tozer GM, Clench MR

{2011)lnvestigation of protein induction in tumour vascular targeted strategies by MALDI MSI.

Methods.54. 442-453

Trim PJ, Djidja M-C, Muharib T, Cole LM, Bryn Flinders, Carolan VA, Francese S. Clench MR

[2012)lnstrumentation and software fo r mass spectrometry imaging—Making the most of

what you've got. Journal of Proteomics.75. (16) 4931^4940

Trim PJ, Djidja MC, Atkinson SJ, Oakes K, Cole LM, Anderson DM, Hart PJ, Francese S, Clench

MR[2010)lntroduction of a 20 kHz Nd:YV04 laser into a hybrid quadrupole time-of-flight mass

spectrometer for MALDI-MS imaging.397. (8) 3409-19

305

Appendix 2: Oral presentations• Cole LM, Bluff J, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR. MALDI-MSI

and Conventional Proteomic Techniques to Compare Protein Induction in

Combretastatin Resistant and Susceptible Tumours. Ourense Imaging Conference,

Spain, September 2012

• Cole LM, Bluff JE, Claude E, Wulfert F, Carolan VA, Paley M, Tozer GM, Clench MR,

Investigating Treatment Response of Combretastatin Resistant and Susceptible

Tumours using MALDI-MSI and Conventional Proteomic Techniques, ASMS Conference,

Vancouver, May 2012

• Cole LM, Bluff JE, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR. Proteomics

of Combretastatin Treated Fibrosarcomas Using lon-Mobility MALDI Mass

Spectrometry Imaging. East Midlands Proteomics Workshop, Loughborough University,

November 2011.

306

Appendix 3: Poster presentations• Cole LM, Bluff J, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR. Potential

Markers of Tumour Viability and Necrosis in CA-4-P Treated Fibrosarcomas Employing

a Multimodal Proteomic Approach. ASMS Minneapolis, June 2013.

• Cole LM, Bluff J, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR. MALDI-MSI

and Conventional Proteomic Techniques to Compare Protein Induction in

Combretastatin Resistant and Susceptible Tumours. NCRI Liverpool, November 2012

• Cole LM, Bluff JE, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR.

Investigations into Treatment Response in Tumour Vascular Targeted Strategies Using

MALDI-MS. British Mass Spectrometry Society, Cardiff, Wales, September 2011.

• Cole LM, Djidja M-C, Bluff JE, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR.

MALDI-MSI to Investigate Treatment Response in Tumour Vascular Targeted Therapy.

American Society for Mass Spectrometry, Denver, USA, June 2011.

• Cole LM, Djidja M-C, Bluff JE, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR.

MALDI-Mass Spectrometry Imaging to Investigate Treatment Response in Tumour

Vascular Targeted Therapy. Cancer Imaging Conference, City University, London, April

2011.

• Cole LM, Djidja M-C, Bluff JE, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR.

Peptide Analysis with In-situ Tissue Tryptic Digestion using MALDI-MS Techniques to

Investigate Protein Induction in Tumour Vascular Targeted Strategies. East Midlands

Proteomics Workshop, Nottingham November 2010.

• Cole LM, Djidja M-C, Bluff J, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR.

In-situ Tissue Tryptic Digestion to Facilitate Peptide Analysis using MALDI-MS

Techniques to Investigate Protein induction in Tumour Vascular Targeted Strategies.

BMSS Cardiff 2010

• Cole LM, Djidja M-C, Bluff J, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR.

MALDI-MS Techniques to Investigate Protein induction in Tumour Vascular-Targeted

Strategies. British Association for Cancer Research, Hallmarks of Cancer Conference,

Edinburgh, June 2010.

• Cole LM, Djidja M-C, Bluff J, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR.

In-situ Tissue Tryptic Digestion to Facilitate Peptide Analysis using MALDI-MS

307

Techniques to Investigate Protein induction in Tumour Vascular Targeted Strategies.

American Society for Mass Spectrometry, Salt Lake City, USA, May 2010.

• Cole LM, Djidja M-C, Bluff J, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR.

MALDI-MS Techniques to Investigate Protein induction in Tumour Vascular-Targeted

Strategies. Cancer Imaging Conference, Oxford University, April 2010.

• Cole LM, Djidja M-C, Bluff J, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR.

Comparative Study of Detergent Addition for In-situ Tissue Tryptic Digestion to

Facilitate Peptide Analysis Using MALDI-MS Techniques. East Midlands Proteomic

Workshop 2009

• Cole LM, Djidja M-C, Bluff J, Claude E, Carolan VA, Paley M, Tozer GM and Clench MR.

A Comparative Study of Detergent Addition for In-situ Tissue Tryptic Digestion to

Facilitate Peptide Analysis Using MALDI-MS Techniques. Bremen IMSC 2009

308