Rights / License: Research Collection In Copyright - Non ... · Tersar, K., Volkmer, B., and Suter,...

Research Collection

Doctoral Thesis

Function, regulation and involvement of Mtmr2 and Mtmr13/Sbf2in the hereditary human diseases CMT4B1 and CMT4B2

Author(s): Tersar, Kristian

Publication Date: 2008

Permanent Link: https://doi.org/10.3929/ethz-a-005767040

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.

ETH Library

https://doi.org/10.3929/ethz-a-005767040http://rightsstatements.org/page/InC-NC/1.0/https://www.research-collection.ethz.chhttps://www.research-collection.ethz.ch/terms-of-use

DISS. ETH.Nr. 17844

Function, Regulation and Involvement of Mtmr2 and MTMR13/Sbf2 in the Hereditary Human Diseases

CMT4B1 and CMT4B2

A dissertation submitted to

ETH ZURICH

for the degree of

Doctor of Science

presented by

Kristian Tersar

Dipl. Biol. Technische Universität Darmstadt, Germany

Born November 25, 1973 in Speyer, Germany

Citizen of Slovenia

accepted on the recommendation of

Prof. Dr. Ueli Suter, examiner

Prof. Dr. Sabine Werner, co-examiner

Dr. Philipp Berger, co-examiner

May 2008

1 THESIS OUTLINE 5

2 SUMMARY 6

3 ZUSAMMENFASSUNG 8

4 INTRODUCTION 10

4.1 The Nervous System 10

4.2 The peripheral Nervous System 11 4.2.1 Cells of the PNS 11 4.2.2 The Schwann Cell 11 4.2.3 Structure of the myelinated Axon 13 4.2.4 Myelin composition – Myelin Proteins 16

4.3 Charcot-Marie-Tooth diseases 17 4.3.1 CMT4B1 and CMT4B2 – clinical characteristics and pathology 19 4.3.2 Animal models of CMT4B1 19

4.4 The Myotubularin family 20 4.4.1 Myotubularin Related Protein 2 and Myotubularin Related Protein 13/SBF2 22 4.4.2 Myotubularin Related Protein 2 and intermediate filaments 25

4.5 Aim of the Study 27

5 MTMR13/SBF2-DEFICIENT MICE 28

5.1 Animal model for CMT4B2 28

5.2 Gene-trap disruption of Mtmr13/Sbf2 28

5.3 Expression analysis of Mtmr2, Mtmr13/Sbf2 and Dlg1/Sap97 in sciatic nerves of mutant animals 30

5.4 Behavioral analysis 33

5.5 Electrophysiology of peripheral nerves of Mtmr13/Sbf2-deficient mice 34

5.6 Progressive myelin abnormalities in peripheral nerves of Mtmr13/Sbf2-deficient mice 35

5.7 Motor and sensory nerves are affected in Mtmr13/Sbf2-deficient mice 39

5.8 Complex structures of misfolded myelin in Mtmr13/Sbf2-deficient mice 40

I

6 INTERACTION PARTNERS OF MTMR2 44

6.1 Principle of the Screening procedure and mass spectrometry-based identification of isolated proteins 44

6.2 Screening tool – Immunoprecipitation via monoclonal anti-Mtmr2 antibody 46

6.3 Identification of novel Mtmr2 interaction partners 47 6.3.1 In-gel digestion - Electrophoresis based screening 47 6.3.2 In-solution tryptic digestion – screening of total immunopreciptates 48

6.4 Expression of intermediate filament mRNA in cultured Schwann cells 51

6.5 Interaction of Mtmr2 and intermediate filaments 52 6.5.1 Immunoprecipitation with anti-intermediate filament antibodies 52 6.5.2 Coimmunoprecipitation with an anti-Mtmr2 antibody 54 6.5.3 Colocalisation of Mtmr2 and intermediate filaments 55

7 DISCUSSION 59

7.1 MTMR13/SBF2 deficient mice – an animal model for CMT4B2 59

7.2 Mtmr2 interacts with intermediate filaments 63

7.3 Conclusions and future directions 68

8 EXPERIMENTAL PROCEDURES 69

8.1 Generation of Mtmr13/Sbf2-/- mice 69

8.2 Genotyping PCR 69

8.3 Protein expression analysis on mutant and wt sciatic nerves 70

8.4 Rotarod test 70

8.5 Neurophysiology 70

8.6 Electron microscopy 71

8.7 Morphometric analysis and quantification of focally folded myelin 71

8.8 Cell Culture 72

8.9 Monoclonal anti-Mtmr2 Antibodies 72

II

8.10 Covalently coupling of antibodies to Protein A Sepharose 73

8.11 Immunoprecipitation for identification of new binding partners of MTMR2 74

8.12 Tryptic in-gel digestion of protein bands (IG) 75

8.13 Automated desalting of tryptic protein digests for MALDI measurements 75

8.14 Identification of of proteins from 1-D gels – MS/MS Acquisition and Mass spectrometry 76

8.15 Identification of of proteins in solution – Tryptic digest in solution (IS) 77

8.16 MS/MS Aquisition via LCQ Deca after tryptic in-gel digestion 77

8.17 Protein identification 78

8.18 Co-Immunoprecipitaion 78

8.19 RT PCR 79

8.20 Primary Rat Schwann Cell (RSC) Culture 82

8.21 Immunstaining of differentiated primary RSC 82

8.22 VSV-tagged intermediate filament Constructs 83

8.23 Specificity Test of mouse monoclonal anti-intermediate filament antibodies 85

9 LIST OF MATERIALS 88

9.1 General Solutions and Fixatives 88

9.2 Blocking solutions 88

9.3 Slides, Plastic ware, Mouting and Embedding Media 88

9.4 Semi Thin Staining solution 89

9.5 Antibody coupling solutions 89

9.6 Immunoprecipitation 89

9.7 Western Blotting 89

9.8 Tail lysis Solutions 90

9.9 Polymerase Chain Reaction 91

III

9.10 Agarose Gel Elecrtophoresis 91

9.11 Primary Antibodies 91

9.12 Secondary Antibodies 92

9.13 Mass spectroscopy 92

9.14 Instruments and Software 92

9.15 Mouse Lines 93

10 REFERENCES 94

11 LIST OF FIGURES 103

12 ACKNOWLEDGEMENTS 105

13 CURRICULUM VITAE FEHLER! TEXTMARKE NICHT DEFINIERT.

IV

Thesis outline

1 THESIS OUTLINE

This thesis is based on the following publications:

Tersar, K., Boentert, M., Berger, P., Bonneick, S., Wessig, C., Toyka, K.V., Young,

P., and Suter, U. (2007). Mtmr13/Sbf2-deficient mice: an animal model for CMT4B2.

Hum Mol Genet 16, 2991-3001.

Berger, P., Berger, I., Schaffitzel, C., Tersar, K., Volkmer, B., and Suter, U. (2006).

Multi-level regulation of myotubularin-related protein-2 phosphatase activity by

myotubularin-related protein-13/set-binding factor-2. Hum Mol Genet 15, 569-579.

All Experiments described in chapters 5 and 6 have been carried out by the author,

with the exception of results presented in chapter 5.5 (performed in collaboration with

Dr. Carsten Wessig and Prof. Dr. Klaus V. Toyka from the Department of Neurology,

University of Wurzburg, Wurzburg, Germany) and in chapters 5.6 – 5.8 (performed in

collaboration with Dr. Matthias Boentert and Dr. Peter Young from the Department of

Neurology and Interdisciplinary Center of Clinical Research, University of Munster,

Munster, Germany).

The thesis begins with a general introduction to the nervous system, Schwann cell

biology, the myotubularin family proteins MTMR2 and MTMR13/SBF2 and the related

CMT4B disease. The results of the first publication are described in chapter 5.

Chapter 6 describes the results of the screening for new Mtmr2 interaction partner

and contains an experiment from the second publication. Afterwards the results are

discussed in detail in chapter 7 and an outlook on future directions is provided.

Subsequent to the discussion the experimental procedures and materials used for

the experiments are described in chapters 8 and 9. Chapters 10 and 11 give a

detailed reference list and the list of figures.

5

Summary

2 SUMMARY

Charcot-Marie-Tooth (CMT) diseases comprise a large group of genetically

heterogeneous hereditary motor and sensory neuropathies (HMSN). With a

prevalence of 1:2500 they range among the most common inherited neurological

disorders affecting the peripheral nervous system (PNS). Mutations in the

Myotubularin-Related Protein-2 (MTMR2) or MTMR13/Set-Binding Factor-2 (SBF2)

genes are associated with the autosomal recessive disease subtypes CMT4B1 or

CMT4B2. Both CMT4B subtypes share similar pathological and clinical

characteristics, including a demyelinating neuropathy associated with reduced nerve

conduction velocity (NCV), focally folded myelin, and an onset of the disease in early

childhood. The proteins MTMR2 and MTMR13/SBF2 belong to the Myotubularin

protein family, which contains 8 active and 6 inactive members. MTM1, an active

member, is the founder of the Myotubularin family and responsible for the X-linked

Myotubular Myopathy. The active Myotubularin family members, including MTMR2,

are phosphoinositide D3-phosphatases. The inactive family members, including

MTMR13/SBF2, contain inactivating substitutions in their phosphatase domain. Two

animal models have been generated to demonstrate that CMT4B1 is caused by the

loss of MTMR2. To better understand the disease mechanism in CMT4B2 the

generation and analysis of a mouse model mimicking the human disease is of major

importance. For a better understanding of the cellular functions of MTMR2 it is

important to identify protein complexes this protein is involved in.

The first goal of of my thesis project was to generate and analyze a Mtmr13/Sbf2-

deficient mouse line. Therefore I used embryonic stem cell line XH212 from

BayGenomics carrying a gene trap plasmid between exon 14 and 15. Only mice

homozygous for the gene trapped Mtmr13/Sbf2 gene displayed a phenotype, as was

expected in comparison to the human disease and the disease models for CMT4B1.

These animals reproduced myelin outfoldings and infoldings in motor and sensory

peripheral nerves as the pathological hallmarks of CMT4B2, concomitant with

decreased motor performance. The number and complexity of myelin misfoldings

increased with age, and was associated with axonal degeneration and decreased

compound motor action potential amplitude. Prolonged F-wave latency indicated a

6

Summary

mild NCV impairment. Loss of Mtmr13/Sbf2 did not affect the levels of its binding

partner Mtmr2 or the Mtmr2-binding Dlg1/Sap97 in peripheral nerves. With the

Mtmr13/Sbf2-deficient mouse line I generated a suitable animal model for the human

disease CMT4B2, and provided further evidence that MTMR13/SBF2 is the disease-

causing gene in CMT4B2.

An important issue in understanding the characteristics of CMT4B is to know the

protein interactors of MTMR2 in Schwann cells, as such interactors might integrate

and control the functions of MTMR2 and explain the involvement of MTMR2 in the

CMT4B disease. I used mass spectrometry (MS)-based proteomics to screen for

protein complexes MTMR2 is involved in. In an MS-based screen I could confirm the

myotubularin family members Mtmr5/Sbf1, Mtmr13/Sbf2, and Mtmr12/3-PAP as

binding partners of Mtmr2. Besides the already described binding partners of

MTMR2, I could confirm the interaction of MTMR2 with proteins of the intermediate

filament family. Using co-immunoprecipitation and western blot analysis I showed

that Vimentin, Desmin, Peripherin, Glial fibrillary acidic protein (GFAP) and the

already described binding partner Neurofilament light chain protein (NF-L) interact

with MTMR2 in lysates of mouse Schwann cell line MSC80 cells.

7

Zusammenfassung

3 ZUSAMMENFASSUNG

Charcot-Marie-Tooth (CMT) Krankheiten beschreiben eine grosse Gruppe

vererbbarer motorischer und sensorischer Neuropathien mit einem heterogenen

genetischen Hintergrund. Mit einer Prävalenz von 1:2500 gehören sie zu einer der

meist verbreiteten neurologischen Krankheiten die das periphere Nervensystem

(PNS) betreffen. Mutationen im Myotubularin Related Protein-2 (MTMR2) oder

MTMR13/Set-Binding Factor-2 (SBF2) Gen führen zu den autosomal rezessiven

Krankheitssubtypen CMT4B1 oder CMT4B2. Beide CMT4B Krankheitssubtypen

haben sehr ähnliche pathologische und klinische Ausprägungen mit einem

Krankheitsbeginn im Kindesalter, gekennzeichnet durch eine demyelinisierende

Neuropathie, welche mit einer verminderten Nervreizleitgeschwindigkeit (NCV) und

örtlich auftretenden Myelinfaltungen einher geht. Das MTMR2 Protein und das

MTMR13/SBF2 Protein gehören beide zur Myotubularin Familie, einer sogenannten

dual-spezifischen Phosphatase-Familie, die 14 aktive und inaktive Proteine

beinhaltet. MTM1, ein aktives Mitglied, ist das Gründerprotein der Myotubularin

Familie und verantwortlich für die sogenannte X-chromosomal vererbte myotubuläre

Myopathie. Die aktiven Mitglieder der Myotubularin Familie, zu denen auch MTMR2

gehört, sind Phosphoinositid-D3-Phosphatasen. Die inaktiven Familienmitglieder, zu

denen MTMR13/SBF2 gehört, haben eine inaktivierende Substitution in ihrer

Phosphatase Domäne. Es wurden zwei Tiermodelle hergestellt um zu zeigen, dass

die CMT4B1 Krankheit durch eine Null-Mutation im MTMR2 Protein hervorgerufen

wird. Um den Mechanismus der CMT4B2 Krankheit besser zu verstehen, ist es von

grösster Bedeutung, ein Maus Modell herzustellen, welche die menschliche

Krankheit imitiert. Um die zellulären Funktionen von MTMR2 besser zu verstehen, ist

es ebenso wichtig die Proteinkomplexe zu identifizieren, an welchen es beteiligt ist.

Der erste Teil meiner Doktorarbeit war es, eine Maus Linie mit einem fehlerhaften

Mtmr13/Sbf2 Gen herzustellen. Dafür habe ich die embryonale Stammzelllinie XH212

von Baygenomics, welche ein sogenanntes „Gen-Trap“ Plasmid zwischen den Exons

14 und 15 trägt, verwendet. Nur Mäuse welche homozygot für das „Gen-Trap

insertierte“ Mtmr13/Sbf2 Gene waren, zeigten eine phänotypische Ausprägung, wie

es auch von der menschlichen Krankheit und den beiden CMT4B1 Mausmodellen

8

Zusammenfassung

her bekannt ist. Diese Tiere bildeten Myelin Ein- und Ausfaltungen in den

motorischen und sensorischen Nerven aus, dem Kennzeichen der CMT4B2

Krankheit, begleitet durch einen Rückgang in der motorischen Leistungsfähigkeit. Die

Anzahl und Vielschichtigkeit der Myelin Fehlfaltungen stieg mit zunehmendem Alter

der Tiere, einhergehend mit axonaler Degeneration und einen verminderten

Amplitude des Muskel- und Nervensummenpotentials. Eine längere F-Wellen Latenz

weist auf eine leichte Beeinträchtigung der Nervreizleitgeschwindigkeit (NCV) hin.

Der Verlust des Mtmr13/Sbf2 Proteins beeinträchtigte jedoch nicht den Pegel des

Bindepartners Mtmr2, sowie den Pegel des an Mtmr2 bindenden Proteins

Dlg1/Sap97 in Nerven des peripheren Nervensystems. Mit der Mtmr13/Sbf2

defizienten Maus Linie habe ich ein entsprechendes Mausmodell für die menschliche

Krankheit CMT4B2 hergestellt und gleichzeitig bewiesen, dass das MTMR13/SBF2

Gen, das Gen ist, welches für die CMT4B2 Krankheit verantwortlich ist.

Um die Eigenheiten der CMT4B Krankheiten zu verstehen, ist es wichtig die

Proteinbindepartner von MTMR2 in der Schwann Zellen zu kennen. Dies ist wichtig,

da solche Bindepartnerproteine die Funktion von Mtmr2 integrieren und kontrollieren

und die Beteiligung von MTMR2 an der CMT4B Krankheit erklären können. Daher

habe ich eine massenspektrometrische Raster-Methode für Proteine verwendet, um

nach Proteinkomplexen zu suchen an denen MTMR2 beteiligt ist. Mit der

massenspektrometrische Raster-Methode konnte ich alle bekannten Bindepartner

von MTMR2 die zur Myotubularin Familie gehören, nachweisen. Neben den schon

beschrieben Bindepartnern von MTMR2, konnte ich die Bindung von MTMR2 an

Proteine der Intermediärfilament-Familie nachweisen. Durch Zuhilfenahme der Co-

Präzipitationsmethode und anschliessender Western Blot Analyse, zeigte ich, dass

die Proteine Vimentin, Desmin, Peripherin, Glial fibrillary acidic Protein (GFAP) und

der unlängst beschriebene Bindepartner Neurofilament light chain Protein (NF-L) in

Lysaten aus den Maus Schwann Zellen MSC80, mit MTMR2 interagieren.

9

http://dict.leo.org/ende?lp=ende&p=wlqAU.&search=Vielschichtigkeit

Introduction

4 INTRODUCTION

4.1 THE NERVOUS SYSTEM

The mammalian nervous system is divided into the central nervous system (CNS)

and the peripheral nervous system (PNS). The CNS consists of the brain and the

spinal cord, while the PNS is made up of nerves that connect theCNS with peripheral

structures. The nerves of the PNS innervate skeletal, cardiac, and smooth muscle, as

well as the glandular epithelium. Sensory fibers of the PNS connect the CNS with the

surrounding environment of the organism, and with its internal state. The nervous

system is composed of two major cell types, neurons and glial cells. As the basic

structural and functional units of the nervous system, neurons are specialized to

receive information, transmit electrical impulses, and influence other neurons and

effector tissues (Haines et al., 2002). Glial cells provide neurons with structural

support and maintain the appropriate environment that is essential for neuronal

function.

The fast transmission of electrical impulses in the mammalian nervous system is

achieved by saltatory conduction of action potentials along the axon towards their

target cell. Axons are enwrapped by glial cells, which built up a myelin segment

called an internode. The nodes of Ranvier are myelin-free segments located between

adjacent internodes. Thereby axons are discontinuously insulated. Only at the nodes

of Ranvier are the axons electrically charged relative to the fluids surrounding them.

Thus the electrical impulses (action potentials) jump from node to node, achieving a

conduction up to 100m/sec, which some 10 times faster than in an unmyelinated

axon (Brown, 2002).

10

Introduction

4.2 THE PERIPHERAL NERVOUS SYSTEM

4.2.1 CELLS OF THE PNS

The glial cell types in PNS are the Schwann cells and the satellite glia. These may be

compared with the major glial cell types of the CNS, the astrocytes and

oligodendrocytes. The astrocytes are responsible for nutrition of the neurons and are

involved in the control of the blood brain barrier (Kandell et al., 2000). The

oligodendrocytes enwrap with their numerous processes different axon and built up

the myelin sheath in the CNS.

The PNS contains myelinating as well as non-myelinating Schwann cells

(Kuhlenbaumer et al., 2005). The Schwann cells insulate axons (with a diameter

more than 1µm) by building up a myelin sheath. In contrast to the oligodendrocytes of

the CNS, Schwann cells do not myelinate multiple axons, but rather establish a one-

to-one relationship at the promyelinating state. Non-myelinating Schwann cells show

similarities to astrocytes and are likely to have metabolic and mechanical support

functions of small caliber axons (smaller than 1µm in diameter) (Jessen, 2004).

Besides Schwann cells and axons, fibroblasts are found, (Berger et al., 2002b) and

also macrophages (Maurer et al., 2002). Macrophages play a role as antigen-

presenting cells and as effector cells that phagocytose damaged myelin (Craggs et

al., 1984).

4.2.2 THE SCHWANN CELL

The Schwann cells of the peripheral nerves originate mainly from the neural crest

cells, except for a fraction derived from boundary cap cells at the margin of the neural

tube. Neural crest cells (NCC) emigrate after the closure of the neural tube from the

most dorsal part of the neural tube. These highly migratory and multipotent stem cells

migrate along two major streams, in a lateral stream and a ventral stream (Bhatheja

and Field, 2006; Jessen and Mirsky, 2002). The lateral stream gives rise to

melanocytes in skin and the ventral stream gives rise to sensory neurons in the

dorsal root ganglia, to autonomic neurons, and to glia cells. The NCC give also rise to

smooth muscle cells and fibroblasts (Jessen and Mirsky, 2005; Lobsiger et al., 2002).

11

Introduction

Before a Schwann cell (either myelinating or non-myelinating) is formed two, other

cell types appear in this lineage, namely the Schwann cell precursor cells (SPCs) and

the immature Schwann cells (Fig. 4.1). Numerous molecules have been identified

regulating Schwann cell development. The Sox 10 protein is essential for the

generation of the peripheral glia from NCC, and is expressed by all neural crest cells

(Britsch et al., 2001). Neuregulin1 (NRG1) derived from axons is important for SPC

survival and promotes the SPC to Schwann cell transition (Garratt et al., 2000). To

maintain the survival of immature Schwann cells, NRG1, Laminin, and ETS

transcription factors have to be expressed (Fig. 4.1). Myelination is promoted by

NRG1, BRN2, KROX20 and OCT-6. The transition from SPCs to Schwann cells

come along with major changes in the cytoarchitecture of the peripheral nerves. The

immature Schwann cells exit the cell cycle and start to form myelinating and non-

myelinating Schwann cells (Fig. 4.1). The fate of the immature Schwann cells

(myelinating or non-myelinating) is determined by the diameter of the axon which

they contact.

Figure 4. 1 The mouse Schwann cell lineage. The developmental transitions of the main cell types in Schwann cell development is illustrated from embryonic day 9 (E9) to postnatal day 15 (P15). The dashed lines indicate reversible (postnatal) transitions. The cell fates are illustrated from the neural crest cell (NCC) to the myelinating and non-myelinating Schwann cell. Bold letters indicate the proteins that are important for the transition steps or maintain a cell fate. The proteins expressed by myelin forming and non-myelinforming Schwann cells are depicted in bold/italic letters (adapted and modified from Jessen and Mirsky, 2005).

12

Introduction

If a Schwann cell contacts a large caliber axon (with a diameter more than 1µm) then

myelination occurs. When a Schwann cell ensheaths small diameter axons (with a

diameter less than 1µm), it become a non-myelinating cell (Jessen and Mirsky,

2005). Myelination involves a combination of down-regulation and up-regulation of

proteins. Marker proteins expressed in myelinating Schwann cells are the myelin

proteins P0, MAG, MBP or PMP22. Non-myelinating Schwann cells show expression

of p75, L1, O4 and GFAP (Fig. 4.1, bold italic) (Jessen and Mirsky, 2002).

4.2.3 STRUCTURE OF THE MYELINATED AXON

The precise arrangement of Schwann cells along axons is an important event before

myelination starts. Initially Schwann cells select individual axons from a nerve bundle

and establish a 1:1 relationship through a process termed "radial sorting" (Simons

and Trotter, 2007). Basal lamina is produced by the Schwann cells before they start

to wrap around an axon. Adhesion molecules (Necl1 and 4) establish axon-glia

contact, and the intracellular asymmetrically distributed Par-3 protein together with

the basal lamina establish a radial axis (Fig. 4.2. A)

Figure 4. 2 Glia-axon recognition in the PNS (A) Necl4 and Necl1 establish glial-axon contact. The asymmetric distribution of the intracellular protein Par-3 and the basal lamina on the outer side establish radial axis. (B) MAG and Necl4 establish the longitudinal polarity and form the internode. Gliomedin, NF-155 and TAG-1 accumulate to establish the presumptive node of Ranvier. (adapted and modified from Simons and Trotter, 2007)

13

Introduction

The longitudinal axis is defined by accumulation of MAG and Necl4 proteins at the

axon-glial junction, which establishes the region forming the future internode.

Accumulation of the proteins Gliomedin, NF-155 and TAG-1 occurs at or around the

future node of Ranvier (Fig. 4.2 B) This compartmentation of the myelin membrane

reaches its maximum extent in the fully myelinated axon (Simons and Trotter, 2007).

The myelin sheath is formed by extension of the Schwann cell plasma membrane in

a spiral growing around the axon. Thereby the Schwann cell nucleus is located

ouside the myelin sheath. Only a small amount of cytoplasma persists at the outer

myelin compartment, which is called the abaxonal domain. The adaxonal domain is

the innermost wrap of the Schwann cell containing cytoplasm adjacent to the axon.

Adaxonal and abaxonal domains are linked via cytoplasmic channels in the Schmidt-

Lanterman incisures (Fig. 4.2, A). They enable transport of small molecular weight

substances between the inner and the outer cytoplasmic domains of the Schwann

cell. The adaxonal domain mediates the contact to the axon. The abaxonal domain

expresses extracellular matrix receptors (Bhat, 2003; Previtali et al., 2001).

Figure 4. 3 Schematic presentation of myelin sheath compartments. (A) Schematic depiction of the longitudinal organization of the PNS myelin. Node, Paranode, Juxtaparanode and Internode are depicted. The coloured axon represents the compartmentation borders and shows important proteins expressed by the axon. Neurofascin 155 indicates the paranodal region. Microvilli, the basal lamina and the incisures are depicted. (B) The myelinating Schwann cell in an “unrolled” presentation. The compartments of the Schwann cell are depicted. Also the regions of compact and non-compact myelin. (adapted and modified from Scherer and Arroyo, 2002)

14

Introduction

The compartment of the myelin sheath can be divided into compact and non-compact

myelin. The internodal compartment consists mostly of compact myelin. The length of

an intermodal compartment is (depending on the axon diameter) about 0.5mm, with

a diameter of 2.5 – 2.8 µm (Salzer, 2003). The compact myelin sheath is formed by

fusion of adjacent Schwann cell membranes. Thereby the major dense line is formed

by cytoplasmic membrane leaflets, while extracellular leaflets form the interperiod

lines (Scherer and Arroyo, 2002). The compact myelin inhibits ion exchange during

nerve conduction by having low capacitance and high resistance. The Schmidt-

Lanterman incisures (Fig. 4.2, A and B) radially traverse the compact myelin. A

cytoplasmic channel extends as well over the entire length of the internode,

containing the outer mesaxon (comprising the membranes that connect the outer,

abaxonal Schwann cell membrane and compact myelin). A comparable channel is

present at the inner surface containing the inner mesaxon. The major dense line

opens towards the node and the paranodal loops contact the axon and form the non-

compact myelin (Fig. 4.3 and 4.4.). The axo-glial junctions are mediated in the

paranodal region via the proteins Caspr and Contactin on the axonal side, and by

neurofascin (NF) 155 on the Schwann cell side. In the juxtaparanodal region

(adjacent to the paranodal region) a protein called Caspr2 is enriched, as are the

potassium channels Kv1.1 and Kv1.2 (Rasband et al., 1998; Schafer and Rasband,

2006) (Fig. 4.4). The node of Ranvier is located between two paranodal junctions and

its length depends on the axon diameter (1-5µm). Voltage gated Na+ and K+

channels are accumulated at the node of Ranvier and mediate the transmembrane

currents to enable rapid saltatory conduction. The axonal side at the node of Ranvier

contains the voltage gated channels Nav1.6 (main representative) as well as Nav1.2,

Nav1.8, Nav 1.9, and the cytoskeletal and scaffolding proteins AnkyrinG βIV and

Spectrin that cluster the sodium channels. Other important proteins are the CAMs

neurofascin-186 and neuri-glia related NrCAM (Susuki and Rasband, 2008).

Schwann cell microvilli contact the node via gliomedin, which binds to axonal NF-186

(Schafer and Rasband, 2006).

15

Introduction

Figure 4. 4 Node, Paranode and Juxtaparanod. Composition and structure of the nodal region and the adjacent paranodal and the juxtaparanodal regions. The regions can be defined morphologically and via their protein compositions (adapted and modified from Scherer and Arroyo, 2002).

4.2.4 MYELIN COMPOSITION – MYELIN PROTEINS

The dry mass of of the PNS myelin is characterized by high lipid (70 – 85%) and low

protein (15 – 30%). Though there are no “myelin-specific” lipids, myelin is enriched

for cerebrosides and cholesterol (Garbay et al., 2000). Lipids play a crucial role in

assisting nerve conduction and provide an inert insulation (Menon et al., 2003). The

PNS myelin contains a high portion of glycoproteins (~60% of total protein) and is

enriched with basic proteins (~20% of total protein). Glycoproteins are Protein zero

(P0) and Peripheral myelin protein 22 (PMP22). The major basic protein is the

Myelin-basic protein (MBP) (Berger et al., 2002b; Suter and Scherer, 2003) (Fig. 4.5).

16

Introduction

Figure 4. 5 Proteins of the compact and non-compact myelin Compact myelin contains P0, MBP and PMP22. Non-compact myelin contains MAG, DM20, E-cadherin Cx32, and claudin (of unknown subtype) (adapted and modified from Scherer and Arroyo, 2002).

P0 is the major protein in the PNS myelin, constituting about 50 to 60% of the total

myelin protein, and interacts directly with PMP22. PMP22 is also part of the compact

myelin. PMP22 has major impact on myelination and maintenance of the myelin

sheath. The function of PMP22 is highly dosage dependent (Suter and Scherer,

2003). MBP is a minor compact myelin component. The loss of MBP is a reliable

marker for demyelinaton (Martini and Schachner, 1997). Myelin-associated

Glycoprotein (MAG) is a protein of the non-compact myelin. MAG is also located in

the paranodal loops and is important for axonal growth and regeneration (Hu and

Strittmatter, 2004). Connexin 32 (Cx32), also a protein of the non-compact myelin, is

located in the gap junctions of Schwann cells. These gap junctions mediate radial

diffusion across incisures (Meier et al., 2004). E-cadherin is localized in the

paranodes, the Schmidt Lanterman incisures, and the outer mesaxon. There it forms

adherens junctions with α- and β-catenin (Young et al., 2002).

4.3 CHARCOT-MARIE-TOOTH DISEASES

J.M. Charcot and P. Marie and, independently, H.H Tooth described in the late 19th

century a hereditary peripheral neuropathy. What is now called Charcot-Marie-Tooth

(CMT) disease or Hereditary Motor and Sensory Neuropathies (HMSN), comprise a

genetically heterogeneous group of inherited disorders affecting myelinated axons in

the peripheral nervous system (Berger et al., 2006b; Dyck et al., 1993; Niemann et

17

Introduction

al., 2006) with a prevalence of approximately 1:2500 (Skre, 1974). The disease is

characterized by progressive distally accentuated muscle weakness and atrophy.

Based on clinical, electrophysiological and histological data, CMT has been

subdivided into demyelinating (CMT1 CMT3 and CMT4) and axonal (CMT2) forms

(Berger et al., 2002b). The demyelinating subtypes CMT4 belong to the autosomal

recessive forms of CMT. CMT4 forms account for about 5% of CMT neuropathies in

western countries and often apperar as isolated cases because of small numbers of

siblings (Niemann et al., 2006; Suter and Scherer, 2003). Demyelinating

neuropathies are diagnosed by reduced nerve conduction velocity (NCV). Axonal

loss and muscle atrophy are also observed, most likely as secondary effects due to

the tight interaction and communication between myelinating Schwann cells, axons,

and muscle cells (Suter and Scherer, 2003). Axonal forms of CMT are characterized

by a reduction of the compound muscle action potential (CMAP) amplitude due to a

loss of myelinated axons (Zuchner and Vance, 2006). Dissection of the cellular

functions of the gene products altered in CMT (Fig. 4.6) as well as the generation of

detailed pathophysiological models are of crucial importance to understand the

underlying common as well as distinct disease mechanisms which may affect

Schwann cells, axons, or both.

Figure 4. 6 Schematic overview highlighting proteins that are mutated in CMT. The figure depicts the locations of the wild-type proteins encoded by the genes that are mutated in CMT. Proteins have been assigned to Schwann cells and/or neurons, respectively, when expression and the observed form of CMT overlap (adapted and modified from Niemann et al., 2006)

18

Introduction

4.3.1 CMT4B1 AND CMT4B2 – CLINICAL CHARACTERISTICS AND

PATHOLOGY

The gene responsible for the severe autosomal recessive CMT type 4B1 has been

identified as the Myotubularin Related Protein-2 gene (MTMR2; (Bolino et al., 2000)).

The disease onset is in early infancy, and the symptoms are those typical for a

demyelinating neuropathy. Patients exhibit reduced nerve conduction velocity, and

the histological analysis revealed typical focally folded myelin sheaths and

demyelination (Previtali et al., 2007). Muscle atrophy and weakness proceed towards

the proximal muscles and a wheelchair is needed from late childhood on. Intelectual

functions are not affected. The nerve conduction velocities are markedly reduced and

range from 9-20m/sec. The distal latency is prolonged and the amplitude is markedly

reduced (0.7 – 1mV) (Houlden et al., 2001). The autosomal recessive form CMT type

4B2 also shows focally folded myelin sheaths, which is due to mutations in the

MTMR13/Set-binding-Factor-2 (SBF2) gene. The phenotype in human CMT4B2

patients is usually less severe than in CMT4B1. The disease onset is around the age

of 8, involving motor and sensory defects. The neurophysiology is similar to the 4B1

subtype, although a NCV of 22m/sec suggests a milder phenotype. In some families

the neuropathy segregates with early onset glaucoma (Azzedine et al., 2003; Hirano

et al., 2004; Senderek et al., 2003).

The histological hallmark of both diseases is focally folded myelin from the outer

myelin sheaths (myelin outfoldings). The protrusions contain axon as well as

Schwann cell cytoplasm (Quattrone et al., 1996). Inward protrusions are present

extending towards the axon and are called “myelin infoldings”. Demyelination and

remyelination events can be inferred from numerous Schwann cell processes or

basal lamina structures encircling some fibers ("onion bulbs") (Quattrone et al.,

1996).

4.3.2 ANIMAL MODELS OF CMT4B1

Two MTMR2 "knockouts" have been generated as animal models of CMTB41 (Bolino

et al., 2004; Bonneick et al., 2005). In addition one conditional Mtmr2-null mouse

model was generated using the Cre/loxP system (Bolis et al., 2005). The conditional

19

Introduction

null mice were viable and showed no significant functional impairment. Behavioral

and electrophysiological tests suggest a neuromuscular defect. Motor and sensory

nerves are affected, showing typically myelin outfoldings, mainly at the paranodal

loops, starting at 3-4 weeks after birth. 12 month-old Mtmr2-null animals show myelin

outfoldings also at Schmidt-Lanterman incisures. These mice also show a

spermatogenesis defect, consistent with one CMT4B1 family (Bolino et al., 2004).

The mouse model produced by Bonneick and colleagues (Bonneick et al., 2005) also

mimics a mutation found in one familial CMT4B1 case. The animal model has been

produced by inserting an E276X mutation in exon 9. Mice homozygous for this

mutation show myelin outfoldings similar to those observed in the Mtmr2-null mouse

model produced by Bolino and colleagues. No electrophysiological and behavioral

alterations were observed and the testis appeared normal. Axonal loss was observed

in later stages (15 months) in distal nerves (Bonneick et al., 2005). The mouse model

based on the Cre/loxP system was used to generate two conditional Mtmr2-null lines

with specific ablation in Schwann cells or motor neurons. Only the Schwann-cell

specific ablation displayed the phenotype also achieved with the Mtmr2-null mouse

model (Bolis et al., 2005). It is possible to conclude that Mtmr2 in Schwann cells is

sufficient and necessary to provoke a condition resembling CMT4B1 with myelin

outfoldings, although Mtmr2 still might have a function in motor neurons (Previtali et

al., 2007).

4.4 THE MYOTUBULARIN FAMILY

MTMR2 and MTMR13/SBF2 both belong to the myotubularin family, which in turn

belongs to the tyrosine/dual-specific phosphatase superfamily (PTP/DSP). The

myotubularin family consists of 14 members in humans (Fig. 4.7, A). The

myotubularin family comprises both catalytically active and inactive phosphatases.

Inactive myotubularin phosphatases have divergent residues in the catalytically

active Cys-X5-Arg motif in the catalytic pocket (containing substitutions in the

Cysteine and Arginine residues) (Laporte et al., 2003). Proteins of the MTM family

share the same protein domain core. The structural hallmarks of myotubularins are a

PH-GRAM (pleckstrin homology glucosyltransferases, Rab-like GTPase activators

and myotubularins) domain, a large Phosphatase domain (PTP/DSP) and a coiled

coil domain. The Phosphatase domain is a large structural unit. It contains N-

20

Introduction

terminally a RID (Rac-induced recruitment domain) and C-terminally a SID (SET-

interacting domain) motif (Fig. 4.7, B) (Begley et al., 2003; Robinson and Dixon,

2006).

Figure 4. 7 The human Myotubularin family Panel (A) shows the phylogenetic tree of the MTM protein family. (B) Domains within the family members are indicated from N- to C-terminus. DENN (purple), PH-G (red), Phosphatase (dark blue) or inactive Phosphatase (blue), coiled-coil (green), FYVE (grey), PH (orange in hMTMR13 and hMTMR5), and PDZ binding motif (yellow in hMTMR1 and hMTMR2). MTMs with active phosphatase domains are shown in bold (adapted and modified from Clague and Lorenzo, 2005).

The phosphatase domain of MTM1 was found to catalyze the removal of 3-phoshate

from PtdIns3P at the D3 position of the inositol ring (Taylor et al., 2000).

Subsequently it was shown that also the MTMR proteins 1, 2, 3, 4, 6, and 7 hydrolyze

the 3-phoshate from PtdIns3P (Berger et al., 2002; Kim et al., 2002; Laporte et al.,

2002; Schaletzky et al., 2003). MTM1, MTMR1, MTMR2 and MTMR6 have been

reported to hydrolyze the 3-phosphate from PtdIns(3,5)P2 (Begley et al., 2003;

Berger et al., 2002; Schaletzky et al., 2003). Three members of the MTM family are

known to be involved in human diseases. Myotubularin, the founding member of the

family, was originally identified as the disease-causing gene in X-linked myotubular

myopathy (Laporte et al., 2003). Several heteromeric interactions of MTM family

members have been described. It seems to be common that an active Myotubularin

family member interacts with an inactive family member. It has been shown that

MTM1 and MTMR2 interact with MTMR12/3-PAP (Nandurkar et al., 2003), MTMR6

and MTMR7 with MTMR9 (Mochizuki and Majerus, 2003). MTMR2 also interacts with

MTMR5/Sbf1 and MTMR13/Sbf2 (Berger et al., 2006; Kim et al., 2003; Robinson and

Dixon, 2005). MTMR2 has also been shown to dimerize through its coiled-coil

domain (Berger et al., 2003).

21

Introduction

4.4.1 MYOTUBULARIN RELATED PROTEIN 2 AND MYOTUBULARIN

RELATED PROTEIN 13/SBF2

Mutations in the MTMR2 gene and MTMR13/SBF2 gene lead to CMT4B1 and

CMT4B2 respectively. For a detailed list of mutations in MTMR2 and MTMR13/SBF2

gene see: http://www.molgen.ua.ac.be/CMTMutations/Home/Default.cfm.

These two CMT forms are clinically indistinguishable, suggesting that these two

proteins have related functions. MTMR2 and MTMR13/SBF2 proteins have been

shown to exist as a complex (Berger et al., 2006; Robinson and Dixon, 2005). The

domain structure of both proteins is shown in Figure 4.8.

Figure 4. 8 Schematic presentation of the MTMR2 and MTMR13/SBF2 domain structure. DENN (white, in MTMR13/SBF2), PH-G (yellow), Phosphatase (green) or inactive Phosphatase (green – crossed through), coiled-coil (orange), RID (grey, in MTMR2), SID (blue) PH (off-white, in MTMR13/SBF2) and PDZ (red, in MTMR2). The indicated RID and SID sub-domains belong structurally to the Phosphatase binding motif. Some selected mutations are depicted that lead to CMT4B1 and CMT4B1.

The pleckstrin homology-GRAM (PH-G) domain, present in MTMR2 and

MTMR13/SBF2, mediates the membrane attachment by binding to

phosphoinositides. The PH-G domain of MTMR2 binds to PI(4)P, PI(5)P, PI(3,5)P2,

and PI(3,4,5)P3. The PH-G domain of MTMR13/SBF2 can bind to the same

phosphoinositides as MTMR2 (Berger et al., 2006; Berger et al., 2003).

MTMR13/SBF2 contains a C-terminal PH domain. A consensus motif of this PH

domain is required for binding to PI(3,4,5)P3 (Robinson and Dixon, 2005). The

function of the DENN domain of MTMR13/SBF2 is not understood. Several DENN

domain-containing proteins have been shown to regulate or associate with Rab

family GTPases, suggesting involvement in membrane trafficking. The coiled-coil

dimerization module is involved in membrane association of MTMR2 (Berger et al.,

2003). The coiled-coil domains do not directly mediate the interaction of MTMR2 and

22

Introduction

MTMR13/SBF2, and therefore further domains must be involved (Berger et al.,

2006). Also present in both proteins is a SET-interacting domain (SID) that was

initially described as mediating the interaction of MTMR5 with the SET-domain of

ALL1, human orthologue of trithorax. This suggests a possible nuclear localization of

MTMR2 and MTMR13/SBF2, which has already been reported in some studies (Cui

et al., 1998). MTMR2 contains an additional PDZ (PSD-95/DLG1/ZO-1) binding site

at the C-terminus (Robinson and Dixon, 2005). The RID sub-domain is the putative

binding site for the class IV intermediate filament neurofilament light chain (NF-L)

protein. The interaction of MTMR2 and NF-L was shown in Schwann cells as well as

neurons (Previtali et al., 2003a). The phosphatase activity of MTMR2 shows a

substrate specificity towards PI3P and PI(3,5)P2, and dephosphorylates the inositol

ring at the D3 position (Berger et al., 2002). Berger et al., also demonstrated with an

elegant expression system that MTMR2 and the MTMR2//MTMR13/SBF2 complex

use PI3P and PI(3,5)P2 as specific substrates (Fig. 4.9). The activity towards other

phosphoinositides was at background level. The activity of the

MTMR2//MTMR13/SBF2 complex is more than 25 fold increased towards PI(3,5)P2

compared to the activity of MTMR2 alone. The activity of the

MTMR2//MTMR13/SBF2 complex towards PI3P increases 10 fold when compared to

the activity of MTMR2 alone.

Figure 4. 9 Phosphatase activity of MTMR2 and MTMR2//MTMR13/SBF2 complex. In this experiment only PI3P and PI(3,5)P2 were dephosphorylated by the MTMR2 and the MTMR2//MTMR13/SBF2 complex. The complex formation of MTMR13/SBF2 and MTMR2 strongly increases the phospahtase activity towards PI3P and PI(3,5)P2 (adapted and modified from Berger et al., 2006).

23

Introduction

The MTMR2//MTMR13/SBF2 complex has an approximately two times higher activity

PI(3,5)P2 than towards PI3P (Berger et al., 2006). The MTMR2 substrate PI3P is

highly enriched on early endosomes in mammalian cells. If PI3P is depleted the

trafficking of a number of proteins through the early endosomes is delayed. There

PI3P plays an important role in endosome function and recruits effector proteins to

the endosomal membranes (Gruenberg and Stenmark, 2004). The enzyme

producing the other MTMR2 substrate PI(3,5)P2, the PIKfyve kinase, is located on

the late endosomes (Sbrissa et al., 1999). Changes in the level of PI(3,5)P2,

provoked through over-expression of the kinase or dominant-negative mutants of the

kinase, produce in yeast and in mammalian cells swollen endosomes and

vacuolization. This implicates a function of PI(3,5)P2 in membrane homeostasis

(Robinson and Dixon, 2006). When MTMR2 and MTMR13/SBF2 were expressed in

COS-7 cells a broad but incomplete overlap of MTMR2 and MTMR13/SBF2 was

observed in the cytoplasm. PI(3,5)P2 levels increase in COS-7 cells during hypo-

osmotic stress. When hypo-osmotic conditions were applied to COS-7 cells

expressing MTMR2 and MTMR13/SBF2, MTMR13/SBF2 was bound to the

membranes of the vesicles (formed by the hypo-osmotic conditions) and MTMR2

remained in the cytoplasm (Berger et al., 2006). This experiment shows the nature of

intracellular vesicle compartments to which MTMR2 and MTMR13/SBF2 bind.

Changes in the phosphoinositide levels might subsequently de-localize MTMR2 and

MTMR13/SBF2 also in Schwann cells. The events that might lead to these changes

need to be assessed. The interaction of MTMR2 and MTMR13/SBF2 is most likely

the molecular basis for the identical phenotypes when the gene encoding either

protein is mutated, showing that these proteins act together in an important pathway.

Loss of MTMR2 and MTMR13/SBF2 function lead to a disruption of the

MTMR2//MTMR13/SBF2 complex, leads to a lack of phosphatase activity, and might

be responsible for the human diseases CMT4B1 and CMT4B2. Thereby

MTMR13/SBF2 functions as a regulator fo the phosphatase activity of Mtmr2 (Berger

et al., 2006).

24

Introduction

4.4.2 MYOTUBULARIN RELATED PROTEIN 2 AND INTERMEDIATE

FILAMENTS

Mutations in neurofilament light chain gene NF-L are associated with dominantly

inherited axonal CMT type 2E and dominant demylinating CMT type 1F forms. The

identification of Neurofilament light chain protein (NF-L) interacting with MTMR2

draws additional attention to the intermediate filament family (Perez-Olle et al., 2005;

Previtali et al., 2003a). The mutations in NF-L appear as an axonal or an intermediate

form showing features of axonopahty and demyelination, with rare excessive myelin

which resemble myelin outfoldings (Zhu et al., 1997). NF-L is mainly expressed in

neurons NF-L mRNA is also upregulated after injury in Schwann cells and is present

during development. Whether the MTMR2/NF-L interaction also contributes to the

CMT4B1 disease isnot yet clear (Previtali et al., 2007). IFs are important cytoskeletal

polymers and the proteins are encoded by a large family of differentially expressed

genes. They are important for intracellular organization, provide structural support in

the cytoplasm and nucleus, and account for a large number of genetic human

diseases. In this summary I will focus on the IFs of class III and class IV type (Kim

and Coulombe, 2007) (Table 3.1).

IF name type Size

(kDa)

Main tissue distribution

Vimentin III 55 Mensencymal, Fibroblasts, endothelium,

Schwann cells

Desmin III 53 Muscle

GFAP III 52 Asctrocytes, Schwann cells

Peripherin III 54 Peripheral neurons

NF-L IV 68 Neurons, Schwann cells

Lamin A/C V 62-68 Ubiquitous expression in differentiated cells Table 4. 1. Class III and Class IV IF proteins Subset of intermediate filamentsimplicated in CMT diseases and Schwann cell biology (adapted and modified from Kim and Coulombe, 2007).

General features of intermediate filament proteins are the presence a central rod

domain. The rod domain of the individual molecules in subdivided into the coil

segments 1A, 1B, 2A, and 2B. L1, L12 and L2 are linker segments between the coil

25

Introduction

domains (Fig. 4.10). These domains are flanked by a N-terminal head domain and a

C-terminal tail domain.

Figure 4. 10 Domain signature of cytoplasmic class III and calss IV IFs. Head and Tail domain in green. The α-helical rod domain is the major determinant for self assembly. The heptate repeat-containing segments within the rod domain are indicated with 1A, 1B, 2A, and 2B, the flexible linker regions with L1, L12 and L2. Highly conserved rod domain boundaries are indicated in orange.

The rod domain boundaries consist of highly conserved amino acid regions (15-20

amino acides). These amino acids are frequently mutated in human disease and are

important for the polymerization of the IFs. The linker regions provide flexibility to the

stiff coiled-coil structure (Herrmann et al., 2007; Kim and Coulombe, 2007). IFs have

important roles in tissue integrity and cell-shape determination in mammalian cells.

IFs also coordinate mechanical forces and have diverse functions in embryonic

development, growth and maturation of specific tissues. GFAP and Vimentin are both

upregulated in the Schwann cell during nerve regeneration. They interact physically

in two signaling pathways involved in proliferation and regeneration. GFAP regulates

mitotic signals after nerve damage via αVβ8 integrin. Vimentin binds to α5β1 and

regulates thereby proliferation and differentiation later in regeneration (Triolo et al.,

2006). Recent publication provides a new link between disease-linked MTMR2 and

MTM1 mutations and NF-L assembly. The co-expressed disease mutant proteins of

MTMR2 and MTM1 produce missassembly of NF-L and aggregation of NF-L occurs.

The over expression of the wild type proteins have no effect on the assembly of NF-L

(Goryunov et al., 2008). Individual mutations can affect the biophysical properties of

the desmin filaments and afterwards interfere with cellular functions. Cellular

response is changed upon physiological alterations in the affected cell-type. These

cellular events lead to tissue-wide pathogenic changes like skeletal muscle atrophy

(Myopathy) or heart failure (Herrmann et al., 2007). An interesting link is also the

involvement of intermediate filament associated proteins (IFAPs) in the formation of

26

Introduction

axonal membrane domains at nodes and paranodes, as Vimentin is a known binding

partner of Ankyrin and Spectrin (Green et al., 2005). Ankyrin B and Spectrin (αII and

βII) mediate the correct positioning of the paranodal loops. Ankyrin G and the βV-

Spectrins are involved in positioning of the sodium channels at the node of Ranvier

(Schafer and Rasband, 2006).

4.5 AIM OF THE STUDY

The major aim of the first part of the study was to prove that MTMR13/SBF2 is the

disease-causing gene in CMT4B2 and provide a suitable animal model using gene-

trap disruption of MTMR13/SBF2. The mutant mice were used to analyze the protein

expression in sciatic nerves. The new mouse model for CMT4B2 also helped to

analyze in detail the behavioral, electrophysiological and histological changes that

occur upon a mutation in MTMR13/SBF2. The MTMR13/SBF2 deficient animals will

help to unravel the disease mechanism of CMT4B and to elucidate the critical

functions of protein complexes that are involved in phosphoinositide-controlled

processes in the peripheral nerves. In combination with the MTMR2 mutant animals

the MTMR13/SBF2 mutant mice will give rise to new insights into the mechanism of

hereditary neuropathies.

In the second part of the project, a mass-spectrometry based screen was performed

to identify additional binding partners of MTMR2 using co-immunoprecipitation from

the mouse Schwann cell line MSC80.

Specific mouse monoclonal antibodies were used for the precipitation of

endogenously expressed MTMR2. New identified interaction partners will be tested

for coexpression and colocalization to elucidate their influence on the complex and a

possible role in disease mechanisms of CMT4B.

27

Mtmr13/Sbf2-deficient mice

5 MTMR13/SBF2-DEFICIENT MICE

5.1 ANIMAL MODEL FOR CMT4B2

Mutations in the Myotubularin-Related Protein-2 (MTMR2) or MTMR13/Set-Binding

Factor-2 (SBF2) genes are associated with the autosomal recessive disease

subtypes CMT4B1 or CMT4B2 (Azzedine et al., 2003; Senderek et al., 2003). Both

forms of CMT share similar features including a demyelinating neuropathy associated

with reduced nerve conduction velocity (NCV) and focally folded myelin. To unravel

the disease mechanism in CMT4B the known Mtmr2-deficient animals were of major

value (Bolino et al., 2004; Bonneick et al., 2005). To prove that MTMR13/SBF2 is the

disease causing gene in CMT4B2 and to provide a suitable animal model, we have

generated Mtmt13/Sbf2 deficient mice. The CMT4B2 animal model should help to

elucidate the critical functions of protein complexes that are involved in

phosphoinositide –controlled processes in peripheral nerves. The possibility to

generate also Mtmt2//Mtmt13/Sbf2 double deficient mice will give us the possibility to

dissect the role of this pair of myotubularins in health and disease.

5.2 GENE-TRAP DISRUPTION OF MTMR13/SBF2

We have used mouse embryonic stem cells carrying a gene trap insertion in the

Mtmr13/Sbf2 locus (XH212; Baygenomics Gene Trap Resource) for the generation of

an Mtmr13/Sbf2-deficient mutant mouse line using established procedures (Bonneick

et al., 2005). The insertion site of the gene trap cassette was mapped 1267 bp

downstream of exon 14 of Mtmr13/Sbf2 (Fig. 5.1 A). Based on this information,

primers I, II and III were designed to discriminate between different alleles and for

genotyping (Fig. 5.1 B). Western blot analysis of sciatic nerve lysates of

Mtmr13/Sbf2-deficient and wt littermates revealed that the Mtmr13/Sbf2 protein was

absent (Fig. 5.1 C). Mice with a disrupted Mtmr13/Sbf2 allele are viable and were

born according to Mendelian expectations. No obvious signs of impaired

spermatogenesis were observed, in contrast to some Mtmr2 mutants (Bolino et al.,

2004). Having this Mtmr13/Sbf2 allele at hand, we also generated

28


Mtmr2//Mtmr13/Sbf2-double deficient mice by appropriate cross-breeding with Mtmr2-

deficient mutant animals (Bonneick et al., 2005).

Figure 5. 1 Gene trap disruption of Mtmr13/Sbf2. A, Ideogram of the Mtmr13/Sbf2 protein structure (first row) and the Mtmr13/Sbf2 gene (second row). The gene trap vector and the locus of the gene-trap integration into intron 14 of Mtmr13/Sbf2 is schematically depicted in rows three and four, respectively. Arrows marked with I and II represent the forward primers for the wt and trapped Mtmr13/Sbf2 alleles, respectively, and III the reverse primer for the genotyping PCR (SA, splice acceptor; beta-geo, beta-galactosidase and neomycin-resistance fusion gene; pA, polyadenylation site). B, Genotyping PCR for homozygous (-/-), or heterozygous (+/-) Mtmr13/Sbf2 mutant mice, or wt (+/+). C, Western blot analysis of sciatic nerve lysates of Mtmr13/Sbf2-deficient (-/-) and wt (+/+) control mice. A rabbit polyclonal antibody was used to detect the 210 kDa Mtmr13/Sbf2 protein. Purified CBP-tagged Mtmr13/Sbf2 protein from a baculovirus expression system served as positive control (Berger et al., 2006). Bands below 210 kDa represent degradation products of Mtmr13/Sbf2.

These Mtmr2//Mtmr13/Sbf2-double deficient mice were also viable and born

according to Mendelian ratios. Upon visual inspection, the behavioral phenotype of

both Mtmr13/Sbf2-deficient and Mtmr2//Mtmr13/Sbf2-double deficient mice appeared

normal compared to control littermates. Starting at the age of two months, however,

both mutant lines showed an unusual but very mild hind limb clamping upon tail

suspension (data not shown). Double-heterozygous Mtmr2//Mtmr13/Sbf2 mutant

29


animals appeared indistinguishable from their wt littermates up to fifteen months of

age (latest time point examined).

5.3 EXPRESSION ANALYSIS OF MTMR2, MTMR13/SBF2 AND DLG1/SAP97

IN SCIATIC NERVES OF MUTANT ANIMALS

In a first step, we analyzed whether alterations in Mtmr2 or Mtmr13/Sbf2 expression

alter the protein levels of their respective binding partners in the sciatic nerve of

mutant animals. Western blot analysis of sciatic nerve lysates from twelve month-old

animals revealed that Mtmr2 levels were unchanged in Mtmr13/Sbf2-deficient mice

(Fig. 5.2 A). Similarly, Mtmr13/Sbf2 levels remained unaltered in Mtmr2-deficient

mice (Fig. 5.2 B). Bolino et al. (Bolino et al., 2004) and Bolis et al. (Bolis et al., 2005)

have reported an interaction of Mtmr2 with Sap97. They also detected reduced

expression of Sap97 in the sciatic nerves of their strain of Mtmr2-deficient mice.

Here, we confirmed these findings in our strain of Mtmr2-mutant mice (Bonneick et

al., 2005). We continued to test whether loss of Mtmr13/Sbf2 would also reduce the

levels of Sap97 by reasoning that loss of the Mtmr2 interaction partner Mtmr13/Sbf2

might affect indirectly the interaction between Mtmr2 and Sap97 within a putative

larger complex. However, the levels of Sap97 were not significantly different

compared to wt in Mtmr13/Sbf2-deficient mice (Fig. 5.2 C). Consistent with these

findings, we found a comparable reduction of Sap97 in Mtmr2//Mtmr13/Sbf2-double

deficient mice as in Mtmr2-single mutants (Fig. 5.2 D). We conclude that the

interaction of Mtmr13/Sbf2 with Mtmr2 and the interaction between Mtmr2 and Sap97

are unlikely to be intimately connected.

30


Figure 5. 2 Western blot analysis of the relative Mtmr13/Sbf2, Mtmr2 and Dlg1/Sap97 levels in sciatic nerve lysates from twelve month-old wt, Mtmr2-deficient (Mtmr2-/-), Mtmr13/Sbf2-deficient (Sbf2-/-), and Mtmr2//Mtmr13/Sbf2-double deficient (MTMR2-/- Sbf2-/-) mice. Each pool contains the sciatic nerves from two or three animals. Protein levels of Mtmr13/Sbf2, Mtmr2 and Sap97 were quantified by normalizing the relative protein levels to beta-actin, illustrated in a bar chart. A, The relative protein level of Mtmr2 does not differ significantly between wt and Mtmr13/Sbf2-deficient (Sbf2-/-) sciatic nerves. B, The Mtmr13/Sbf2 expression level shows no difference in Mtmr2-deficient (Mtmr2-/-) compared to wt control lysates of sciatic nerves. The Sap97 protein levels in Mtmr2-deficient (Mtmr2-/-) lysates are significantly reduced (p


This conclusion was also supported by co-immunoprecipitation experiments revealing

no apparent differences in the interaction of Mtmr2 with Sap97 between wt and

Mtmr13/Sbf2-deficient sciatic nerves (Fig. 5.3).

Figure 5. 3 Co-Immunoprecipitation of Sap97 with mouse monoclonal anti-Mtmr2 antibody Each lysate pool contains sciatic nerves from three animals. The immunoprecipitation was performed with mouse monoclonal anti-Mtmr2 antibody covalently coupled to Protein A sepahrose. Western blot was performed with mouse monoclonal anti-Sap97 antibody and polyclonal rabbit anti-Mtmr2 antibody. Mtmr2 and Sap97 were present in the IPs from wild-type and Mtmr13/Sbf2 deficient sciatic nerves. Mtmr2 was not precipitated from Mtmr2 deficient sciatic nerves and Sap97 was not co-precipitated.

32


5.4 BEHAVIORAL ANALYSIS

Visual examination of both Mtmr13/Sbf2-deficient and Mtmr2//Mtmr13/Sbf2-double

deficient mice revealed no obvious signs of tremor or major functional disability,

similar to what we had observed in the Mtmr2-deficient model of CMT4B1 (Bonneick

et al., 2005). Therefore, we performed a Rotarod test to assess whether a behavioral

difference related to motor function was detectable using this assay.

Figure 5. 4 Rotarod analysis of wt, Mtmr13/Sbf2-deficient (Sbf2-/-) and Mtmr2//Mtmr13/Sbf2-double deficient (Mtmr2-/- Sbf2-/-) mice.

33


Mice were tested four times per day on four consecutive days, and the time spent on the rotating rod was plotted versus the trial number. For statistical analysis Students t-test was used. Error bars show the Standard Error of the Mean (S.E.M.). A, Analysis of four month-old Mtmr13/Sbf2-deficient (Sbf2-/-) and wt control mice (n=6). No significant difference between the two groups was detected. B, Analysis of five twelve month-old Mtmr13/Sbf2-deficient (Sbf2-/-) mice and six wt control animals. Trials 2, 5, 6, 7, and 9-15 were significantly different for p


Figure 5. 5 Motor nerve conduction studies of sciatic nerves of four months-old (A-C) and twelve months-old animals (D-H). Error bars indicate the Standard Error of the Mean (S.E.M.). A, At four-months of age, Mtmr13/Sbf2-deficient (Sbf2-/-) mice showed mild but not significant NCV slowing compared to wt mice (n=6). B, In comparison to wt animals, Mtmr13/Sbf2-deficient (Sbf2-/-) mice showed no significant difference in CMAP amplitudes at this age. C, F-wave latency was slightly but significantly increased in Mtmr13/Sbf2-deficient (Sbf2-/-) mice compared to wt (*p


Figure 5. 6 Histological analysis of cross sections of sciatic nerves at four months of age. Wt (A, C) and Mtmr13/Sbf2-deficient (Sbf2-/-) (B, D) mice were compared. Mtmr13/Sbf2-deficient (Sbf2-/-) mice show numerous nerve fibres with redundant myelin loops scattered across the nerve section. Affected nerve fibres exhibit different morphologies including myelin sheath outfoldings and infoldings (white arrows in D), and likely degradation of the axon-Schwann cell unit (white arrowhead in D). Scale bars for A,B: 100 µm; for C,D: 25 µm.

Abnormalities included both infoldings and outfoldings of the entire myelin sheath,

which particularly affected large caliber fibres but also smaller, thinly myelinated

fibres. Non-myelinated fibres appeared normal. At higher magnification, sciatic nerve

cross sections of mutant mice were littered with various degrees of dysmyelination

ranging from focal budding of the myelin sheath to multiple or combined infoldings

and outfoldings. Abnormal myelin structures were first but rather sporadically

observed in the sciatic nerves of Mtmr13/Sbf2-deficient mice already at the age of

three weeks (Fig. 5.7 A, B).

36


Figure 5. 7 Time course of morphologic changes indicates a progressive neuropathy in Mtmr13/Sbf2-deficient (Sbf2-/-) mice. A qualitative and quantitative analysis of sciatic nerve cross section was performed at the age of three weeks, four months, and fifteen months. Three animals were analyzed per age and genotype, and representative images are shown. At 3 weeks (A, B), mutant animals already show focally folded myelin (white arrows in B), ranging from mere “budding” (oblique white arrow in B) to a major outfolding of the myelin sheath (horizontal white arrow in B). At four months (C, D) and fifteen months (E, F), mutant mice exhibit numerous nerve fibers with redundant myelin loops (oblique white arrows in D and F). Progression of the neuropathy is reflected by an increasing number of affected fibers, a higher morphological complexity of myelin abnormalities, and signs of additional axonal degradation. Focal folding of the myelin sheath may be also observed, albeit very rarely, in wt littermates, possibly reflecting age-dependent dysmyelination. Scale bars: 25 µm.

Irregular myelin folds were easily detectable although of low complexity. Quantitative

analysis of these pathological structures revealed a significant increase in numbers

compared to wt animals (Fig. 5.8 A). Next, since CMT is usually associated with a

37


clinically progressive time course, we followed the qualitative and quantitative

progression of the pathology over time. Thus, we examined sciatic nerves at the age

of four months and fifteen months (Fig. 5.7 C-F; Fig. 5.8 A).

Figure 5. 8 Quantification of myelin misfoldings, whole fiber morphometry and relation between axon diameter and myelin sheath thickness. A, Quantification of myelinated fibers at three weeks, four months-, and fifteenth months of age. For each stage, the sciatic nerves were isolated from three mice per genotype. Two semithin sections per nerve, taken at an interval of 1.5 mm, were used for quantification. After the total number of fibers had been determined, the proportion of myelinated fibers exhibiting misfolded myelin was calculated. B, Relation between axon diameter and myelin sheath thickness of three week- and four month-old mice. For each age, the sciatic nerves of three animals per genotype were analyzed. Myelin thickness was determined and plotted as a function of the corresponding axon diameter. Each line represents the trend line for one mouse. Slope and intercept did not differ significantly between wt and Mtmr13/Sbf2-deficient (Sbf2-/-) mice at either age. For statistical analysis a Student`s t-test was used (*, p


signs of axonal damage were recognized although we did not observe an obvious

major loss of myelinated axons. Next we reasoned, considering the suggested

molecular function of myotubularins, that the misfoldings of myelin observed in

Mtmr13/Sbf2-deficient mice might be due to altered vesicular trafficking and myelin-

membrane overgrowth. This could potentially lead to generally altered myelin sheath

thickness as we have observed in myelin mutants with multi-folded myelin extensions

in the central nervous system (Thurnherr et al., 2006). However, using computer-

aided morphometry, we did not detect significant alterations in myelin thickness and

axon diameter in Mtmr13/Sbf2-deficient sciatic nerves (Fig. 5.8 B,C), consistent with

our identical previous findings in Mtrm2-deficient mice (Bonneick et al., 2005). In

agreement with these data, we did not observe Schwann cell onion bulb formation as

the classical indicator of demyelination and remyelination.

5.7 MOTOR AND SENSORY NERVES ARE AFFECTED IN MTMR13/SBF2-

DEFICIENT MICE

CMT4B2 is classified as a motor and sensory neuropathy. Thus, we analyzed

whether these mixed symptoms were reflected in pathological aberrations in both

motor and sensory nerves. We chose to examine the ventral roots containing

exclusively axons derived from motor neurons and dorsal roots for sensory axons. In

both locations, the pathological hallmarks of myelin misfoldings were barely

detectable at the age of four months (Fig. 5.9 A, B), in contrast to the more distally

(with respect to the neuronal cell bodies) located sciatic nerve (Fig. 5.9 D) which

contains both motor and sensory axons. These findings indicate that the pathology is

more severe in distal compared to proximal parts of PNS nerves. At the age of fifteen

months, myelin misfoldings were prominently visible in ventral and dorsal roots

suggesting that both motor and sensory nerves become affected in a progressive

manner over time (Fig. 5.9 C, D).

39


Figure 5. 9 Involvement of proximal motor and sensory nerves in Mtmr13/Sbf2-deficient mice. Histological analysis of ventral and dorsal roots from mutant mice at four and fifteen months of age. On cross sections from four months-old mutants (A, B), focally folded myelin is only scarcely present. At fifteen months, both infoldings and outfoldings of the myelin sheath are present in dorsal and ventral roots (arrows in C, D), reflecting a progressive involvement of proximal nerves in the neuropathy. Scale bar for (A–D): 25 µm

5.8 COMPLEX STRUCTURES OF MISFOLDED MYELIN IN MTMR13/SBF2-

DEFICIENT MICE

In order to gain more detailed insights into the fine structure of aberrant Schwann

cell-axon units in our mutant mice, we performed ultrastructural analysis using

electron microscopy. Figure 5.10 shows a collection of pictures to provide a sampling

of the different aberrant structures that we have observed. The myelin misfoldings

invariably originated from compacted myelin and showed an identical number of

myelin lamellae in both myelin misfoldings and the myelin sheath they originated from

(Fig. 5.10 D, quantitative data not shown). Within myelin misfoldings, we observed

normal compaction and periodicity of the myelin sheath. In general, the impression of

pathological alterations was dominated by myelin outfoldings with strongly variable

complexity (Fig. 5.10 A-D).

40


Figure 5. 10 Electron microscopic analysis of focally folded myelin and axonal degeneration in sciatic nerve of Mtmr13/Sbf2-deficient mice. The figure shows a collection of different morphologies at age P21 (A-D), four months (E-F, M, O ,P), and 15 months (G-L, N). Outfolding of the myelin sheath is the most frequent type of dysmyelination (A-C). Single or multiple redundant myelin loops are visible adjacent to the original myelinated fibre. The Schwann cell membrane surrounds both the outfoldings and the Schwann cell-axon unit they arise from. Simple outfoldings and the original myelin sheath share the same periodicity and number of lamellae (D). Infoldings of the myelin sheath may severely affect the axonal shape by leading to constriction of the axonal cytoplasm (arrow in E) or by forming extensions giving the cross-sectioned fibre a target-like appearance (F,G). The double circles in F and G likely reflect the retrograde inversion of infolded myelin loops. Note that the interspaces between the inner and outer infolding and the original myelin sheath show the structure of axonal cytoplasm (insert in F). Some fibres exhibit

41


both infoldings and outfoldings (G). Apart from abnormalities of the myelin sheath, some fibres show disintegration or even degradation of the entire Schwann cell-axon unit. We observed widening of the periaxonal space (white arrow in H), vacuolar disruption of the inner myelin layers (I), compression and lateralization of the axon by massive infoldings (J), and various stages of axonal degeneration (K, L). The white arrow in L points to residual myelin. Longitudinal sections (M-P) revealed preferential location of myelin abnormalities in the nodal and paranodal segments (white arrow in M). Note that the aberrant myelin loops ensheath axonal processes and lead to massive disruption of the normal architecture of the node, which here is forced off the cutting plane (white arrow in N). Infoldings and outfoldings also occur in the internodal segment of the myelinated fibre (white arrowhead in O) or near the Schmidt-Lantermann incisures (arrows and white arrowhead in P). Scale bars for D: 1 µm, L-P: 5 µm.

The most common formation consisted of one or multiple outfoldings of different size

adjacent to a myelinated large caliber axon, and multiple outfoldings showed a

tendency to form groups (Fig. 5.10 C). Aberrant myelin loops were always

ensheathed by the plasma membrane of the related Schwann cell (Fig. 5.10 A).

Myelin infoldings were also prominent. They usually presented as finger-like

inversions of the myelin sheath (Fig. 5.10 E) or circular inclusions within the

myelinated nerve fiber (Fig. 5.10 F). Entraining Schwann cell cytoplasm on their outer

surface, they protrude far into the axon and displace the axoplasm (Fig. 5.10 E). The

formation of double circles (Fig. 5.10 F,G) is most likely due to the retrograde

inversion of a single infolding since, at higher magnifications, we observed axonal

material in the gap between the inner and outer infolding. Alternatively, the nesting of

two distinct infoldings may have led to the double-circle appearance.

On electrophysiological examination, older Mtmr13/Sbf2-deficient mice showed a

reduction of the CMAP indicating axonal loss or damage. Thus, we also carefully

looked for axonal pathology. Degeneration of whole Schwann cell-axon units (Fig.

5.10 K,L) was not observed in young mutants, but was sporadically present at four

months and rather frequent at fifteenth months. Lateral dislocation of the axon by

myelin infoldings and vacuolar alteration of the axoplasm (Fig. 5.10 I,J) was often

observed. We occasionally found also myelinated nerve fibers not affected by myelin

misfoldings but with a conspicuous widening of the periaxonal space (Fig. 5.10 H).

The amazing complexity of myelin misfoldings and the consequences for the affected

Schwann cell-axon units, however, can be best appreciated on longitudinal sections

(Fig. 5.10 M-P). Complex myelin formations are preferentially, although not

exclusively, located in the nodal and paranodal regions. At the same time misfoldings

contain axoplasm-like structures suggesting the entrainment of axonal parts during

their formation (Fig. 5.10 M, P). Occasionally, misfoldings of the myelin sheath had a

larger diameter than the original nerve fibre, or lead to spatial disarrangement of the

42


normal fibre anatomy like affecting the node of Ranvier (Fig. 5.10 N). Relating these

structural changes back to the molecular and cellular functions (and misfunctions in

disease) of MTMR13/SBF2 will be a major challenge for the future. On a pure

morphological level, the pathological observations in Mtmr13/Sbf2-deficient mice are

very reminiscent of our observations in Mtmr2-deficient animals consistent with the

biochemical finding suggesting a crucial role for a high-molecular complex containing

both MTMR13/SBF2 and MTMR2 in the biology of myelinated peripheral nerves,

possibly in the regulation of membrane trafficking.

43

Interaction partners of Mtmr2

6 INTERACTION PARTNERS OF MTMR2

6.1 PRINCIPLE OF THE SCREENING PROCEDURE AND MASS

SPECTROMETRY-BASED IDENTIFICATION OF ISOLATED PROTEINS

To date several interactions between Mtmr2 and other myotubularin family members

have been characterized (Berger et al., 2006; Lorenzo et al., 2006). Furthermore

Mtmr2 has been shown to interact with NF-L (Previtali et al., 2003a) and Dlg1/Sap97

(Bolino et al., 2004). Proteins usually do not act alone, but fulfill their cellular roles in

complex interactions with other proteins (Pandey and Mann, 2000). In recognition of

this we wanted to identify additional binding partners to better understand the role of

MTMR2 in the pathogenesis of the peripheral neuropathy CMT4B. Mass-

spectrometry based proteomics is becoming an indispensable tool for molecular and

cellular biology.

Figure 6. 1 Screening scheme for the detection of new binding partners of Mtmr2 Mouse monoclonal anti-Mtmr2 antibody was covalently coupled to Protein A Sepharose. This antibody matrix was used to isolate protein complexes from cell or tissue extracts. Afterwards detection of precipitated proteins was performed by Western blotting, MALDI TOF/TOF MS or LC-MS analysis.

44


Therefore we used a screening scheme (Figure. 6.1) based on immunoprecipitation

followed by MALDI TOF/TOF mass spectrometry (MS) (Gstaiger et al., 2003; Yart et

al., 2005) or liquid-chromatography integrated ESI (Electrospray ionization)-MS

analysis (Aebersold and Mann, 2003).

The MS-based screening procedure consists of five stages as described elsewhere

(Aebersold and Mann, 2003). In the first stage protein complexes were isolated from

lysate of a mouse Schwann cell line (MSC80) by affinity selection using a specific

mouse monoclonal anti-Mtmr2 antibody covalently coupled to Protein A Sepharose

beads (see Experimental Procedures 8.11) to isolate the Mtmr2 interaction partners

(Figure 6.1). Afterwards the isolated protein complex was either separated on a one-

dimensional gel by electrophoresis and proteins were excised prior to digestion, or

the isolated protein complex was used for digestion in solution (IS). In the first

method the excised proteins are enzymatically digested by trypsin (in-gel digestion,

IG). For the second method the proteins were enzymatically digested by trypsin in

solution. The difference between these two procedures is that separating the isolated

protein complex on a one-dimensional gel by electrophoresis defines a sub

proteome, e.g. on a 10% SDS-PAGE only proteins between 250 kDA and 37 kDA

can be separated. The digestion of the isolated protein complex in solution has no

cut-off of proteins. In the third step the peptides obtained by one-dimensional gel

electrophoresis and in-gel digestion were processed and spotted on a MALDI plate

(see Experimental Procedures 8.14) prior to MALDI TOF/TOF MS (Granvogl et al.,

2007). The peptide mixture obtained from the isolated protein complex by in-solution

digestion was separated by high-preasure liquid chromatography (HPLC) in fine

capillaries and eluted into an ion source where they were nebulized in small, highly

charged droplets (Aebersold and Mann, 2003). In the fourth stage the mass spectra

of the peptides are recorded. In the fifth stage a computer prioritized list of these

peptides is generated and a series of tandem mass spectra (MS/MS) is produced. A

given peptide ion is thereby isolated and fragmented by energetic collision with gas

and an MS/MS spectrum is recorded. The last step to identify the proteins from the

in-gel (IG) or from the in-solution (IS) digestion experiment was performing a

database search with GPS (Global Proteome Server) Explorer Software version 3.5

(Applied Biosystems) and Mascot version 2. 1. 0 (Matrix Science, London, UK) being

utilized as the search engine.

45


6.2 SCREENING TOOL – IMMUNOPRECIPITATION VIA MONOCLONAL ANTI-

MTMR2 ANTIBODY

The isolation of interaction partners of Mtmr2 was done using immunoprecipitation. A

highly specific affinity selection was obtained by using mouse monoclonal anti-Mtmr2

antibodies (Berger et al., 2006) covalently coupled to Protein A Sepharose with

dimethylpimelimidate (Yart et al., 2005).

Figure 6. 2 Immunoprecipitation with a monoclonal anti-Mtmr2 antibody (Berger et al., 2006) Immunoprecipitation was performed from mouse sciatic nerve and various Schwann cell lines, followed by Western blotting. In the upper panel polyclonal rabbit anti-Mtmr2 antibody was used to detect Mtmr2. In the lower panel polyclonal rabbit anti-Sbf2 was used to detect co-precipitated Sbf2. As controls purified Mtmr2 and Sbf2 were used. Sciatic nerve extracts show besides the full length Mtmr2 also major degradation products.

In the experiment shown in the upper panel of Figure 6.2 endogenously expressed

Mtmr2 was precipitated from lysates from Schwann cell lines and sciatic nerve using

the mouse monoclonal anti-Mtmr2 antibody 4H10 (see Experimental Procedures

8.10). Endogenously expressed Mtmr13/Sbf2 was co-precipitated from sciatic nerve

and Schwann cell lysates (Fig. 6.2 lower panel). In this experiment we show for the

first time the endogenous interaction of Mtmr2 and Mtmr13/Sbf2 and the suitability of

the mouse monoclonal anti-Mtmr2 antibody (Berger et al., 2006). MSC80 cells

express antigens of myelin-forming Schwann cells such as S-100 and Laminin

(Boutry et al., 1992). On the basis of this experiment MSC80 cells were used to

screen for new binding partners of Mtmr2.

46


6.3 IDENTIFICATION OF NOVEL MTMR2 INTERACTION PARTNERS

6.3.1 IN-GEL DIGESTION - ELECTROPHORESIS BASED SCREENING

To find new protein complexes that contain endogenous Mtmr2, a one-step

immunoprecipitation with mouse monoclonal anti-Mtmr2 antibody covalently coupled

to Protein A sepharose was carried out. After extensive washings the bound proteins

were eluted from the beads by boiling the samples with 1x SDS-sample buffer,

electrophoresed on a 10% SDS-PAGE and the separated proteins were stained with

colloidal blue (Fig. 6.3).

In this ideal situation the endogenous protein Mtmr2 serves as the bait for the affinity

purification with the monoclonal mouse anti-Mtmr2 antibody. Thereby a number of

proteins should be enriched in the immunoprecipitation with anti-Mtmr2 antibody. In

parallel a control immunoprecipitation with monoclonal mouse anti-Myc antibody was

performed. The anti-Myc antibody belongs to the same IgG subclass as the anti-

Mtmr2 antibody, namely the IgG1 subclass. In the control immunoprecipitation the

enriched proteins, especially from ca. 120 to 150 kDa, appeared as background with

a very low Protein Score. The relevant bands were excised from the gel (Fig. 6.3, IP:

mouse anti-Mtmr2) and processed for tryptic digestion within the gel material. The

peptides obtained by tryptic digestion were eluted and subjected to MALDI TOF/TOF

mass spectrometry analysis. This analysis provided several proteins that precipitated

together with the bait Mtmr2. Among them were known interaction partners of Mtmr2

belonging to the myotubularin family, Mtmr5/Sbf1 (Kim et al., 2003) and Mtmr13/Sbf2

(Berger et al., 2006; Robinson and Dixon, 2005). Mtmr12/3-PAP, also a member of

the myotubularin family, was to date not shown to interact with Mtmr2. All proteins

belonging to the Myotubularin family appeared on the SDS-PAGE according to their

molecular mass described in the Swiss-Prot database, http://expasy.org/. Another set

of proteins that was of particular interest due to their Protein Score were Plectin,

Spectrin, F-Actin cross linking protein, Peripherin, Vimentin Desmin and beta-Actin.

The Protein Score specifies a confidence interval for those proteins that are

considered to be significant. The more MS/MS spectra are recorded from one protein

on the basis of the MS spectrum, the higher the Protein Score and significance

becomes for identified proteins. Besides their high Protein Score these proteins also

appeared on the SDS-PAGE according to their molecular mass described in the

Swiss-Prot database, http://expasy.org/.

47

http://expasy.org/


Figure 6. 3 Separation of immunoprecipitated proteins by 1D SDS-PAGE and identification by MALDI TOF/TOF MS. The proteins were precipitated from MSC80 lysate with monoclonal mouse anti-Mtmr2 antibody and as negative control with monoclonal mouse anti-Myc antibody covalently coupled to Protein A Sepharose (both antibodies belong to the IgG1 subclass). The immunoprecipitates were separated on a 10% SDS-PAGE, stained with GelCodeBlue (Pierce) and the protein bands were excised with a 1D gel spotting pencil. After tryptic in-gel digestion of the proteins each band was analyzed with a MALDI TOF/TOF MS system. The collected MS and MS/MS data were used to identify the proteins that correspond to the excised bands. The control immunoprecipitation with a mouse monoclonal anti-Myc antibody revealed none of the proteins that were precipitated with the anti-Mtmr2 antibody. Polypeptides that yielded unambiguous mass spectrometry spectra are indicated. The major protein bands in both IP lanes at about 55 kDa represent IgG heavy chains of anti-Mtmr2 and Anti-Myc antibody. In the control IP lane with anti-Myc also the light chain of the antibody appears at about 38 kDa.

6.3.2 IN-SOLUTION TRYPTIC DIGESTION – SCREENING OF TOTAL

IMMUNOPRECIPTATES

In comparison to the electrophoresis based screening procedure (see Results 6.3.1)

two things are changed in the screening procedure of the total immunoprecipitates.

48


First, the proteins that were affinity purified with mouse monoclonal anti-Mtmr2

antibody covalently coupled to Protein A Sepharose and co-precipitated with Mtmr2

were eluted with Glycine pH 2.5, immediately neutralized and prepared for tryptic in-

solution digestion (see Experimental procedures 8.16). Second, prior to mass

spectrometry the peptides were separated with an integrated HPLC-System. This

procedure allows a more rapid and generic analysis of the precipitated proteins

(Gingras et al., 2007), since the whole immunoprecipitates are analyzed and no cut-

off occurs due to SDS-PAGE limited size exclusion. Also the integrated liquid-

chromatography ESI-MS systems are preferred to analyze complex samples

because all components are analyzed in one experiment. A limiting factor is that the

false-positive error rates can be generally large, but can be reliably estimated at the

level of the whole data set (Rinner et al., 2007) and are also backed up by the

electrophoresis-based screening method (see Results 6.3.1). In the screening of the

total immunoprecipitates all proteins were found that were also identified in the

electrophoresis based screening (Fig. 6.3). In addition five new proteins were

identified, namely Ankyrin G, Lamin A and Voltage-gated potassium channel subunit

beta-1 (see Table 5.1).

Of particular interest were the proteins Vimentin, Desmin and Peripherin, because

they appeared in both screening experiments, feature a significant Protein Score,

and display high protein-sequence coverage by their identified peptides. These three

proteins belong to the type III intermediate filament family; they provide cel

Rights / License: Research Collection In Copyright - Non ... · Tersar, K., Volkmer, B., and Suter,...

Documents

Transcript of Rights / License: Research Collection In Copyright - Non ... · Tersar, K., Volkmer, B., and Suter,...