Aquaphotomics: Water Spectral Pattern as a

Biomarker for Diagnosis

Roumiana TsenkovaKobe University, Japan

Bio Measurement Technology Lab rtsen@kobe-u.ac.jp

http://nirslab.org

Roumiana TsenkovaKobe University, Japanrtsen@kobe-u.ac.jp

Water and Light

Contents

1. Aquaphotomics concept

2. Near Infrared Spectroscopy

3. Water Spectral Patterns as a

biomarker of water

4. Water spectral pattern of urine as biomarker for estrus in Giant Panda

5. Water Spectral Patterns of soy bean plant as a biomarker of biotic and abiotic stress

1. Aquaphotomics concept

- From biosystems to water -

Water, Biosystem, NIRS

Water in tissues change from one conformation to another, when biological reaction change from steady state to no steady state.

� �

The spectral difference at water bands, i.e. 1440 nm (S1), 1398nm could provide valuable information for mammary gland inflammation diagnosis, but WHY???

Water spectral changes in bio systems under various perturbation

Physical

Chemical Biological

Bio monitoring

Accumulating real time spectral data base

Fractals???

�Aquaphotomics� is a new term introduced by Prof. Roumiana Tsenkova openes a new area in biological sciences and engineering. It describes a new way for exploring biological systems through a non- destructive monitoring of their interaction with VIS-NIR light. Multivariate spectral analysis reveals that changes with the water matrix under perturbation reflect, like a mirror, the rest of the molecules surrounded by water. As a result, characteristic water absorbance patterns are used to measure very small concentrations of solutes and for disease diagnosis.

Extended Water Mirror Approach Explore the universe through the behavior of water

Aquaphotomics: a new field of science dawn to biological sciences and engineering

VIS-NIR spectrum

Water

Water spectra analyzed by Multivariate Analysis

Changes in activated WAMACS

WAMACS = Water Matrix Coordinates, i.e. water absorbance bands in VIS-NIR range

Matter

Perturbation

WAMACSIdentify the rest of the molecules surrounded

by water

R. Tsenkova J. NIR infrared spectroscopy, 17, 303-314 (2009)

�Altered milk components trace signals influence the bonding structure of water measured by NIRS all at once and provide rapid and accurate mastitis

diagnosis�

��consecutive exposure of hydrogen bond network, in the body of mouse, to NIR light allowed successful diagnosis of fatal prion disease�

Aquaphotomics: B.M Jinendraa, K. Tamakib R.Tsenkovaa * Use of Near Infrared Spectroscopy for Nondestructive and Rapid Detection of Soybean Plants Infected with Soybean Mosaic Virus. Abstract published at Japanese society of Near Infrared Spectroscopy, Tsukuba, 2007.

Aquaphotomics: R. Tsenkova, S. Atanassova, S. Kawano, K. Toyoda, �Somatic cell count determination in cow milk by near infrared spectroscopy: A new diagnostic tool�, J. Animal Sci. 2001 79: 2550-2557.

Aquaphotomics: Tsenkova, R., Iordanova, I., Toyoda, K and Brown, D., : Prion protein fate governed by metal binding: BBRC. pp.1005-1012, 2004.

� Monitoring of Second Overtone of Water Absorbance Bands Reveals Hypersensitive Response from Virus Infected Plants�

Soybean Mosaic disease

Prion disease

Aquaphotomics: in bio monitoring and diagnosis

Mastitis disease

Virus diagnosis - HIV

�Regression vector coefficients reveals variations in water structures related to functionality of HIV�

Aqua-Photomics: Sakudo A., Tsenkova R., Onozuka D., Morita K., Li S, Warachit J., Iwabu I., Li G., Onodera T., Ikuta K., A novel diagnostic method for human immunodeficiency virus type-1 in plasma by near-infrared spectroscopy. Microbiol Immunol, 49 (7), 695-701, 2005.

Cd ?Cd ?Cd ?

�Mg/L concentration measurement of Cd, Mg, Mn, Zn in water solution possible because of the interaction with water �seen� by the NIR light at various water absorbance bands.

Prediction

R² SEP

0.96 0.98

Prediction

R² SEP

0.96 0.98

Wavelength range (nm)

400-2500

Wavelength range (nm)

400-2500

Prediction

R² SEP

0.96 0.98

Prediction

R² SEP

0.96 0.98

Wavelength range (nm)

400-2500

Wavelength range (nm)

400-2500

�Better prediction of polystyrene mesoscopic particle concentration is obtained by multivariate analysis (a) based on water absorbance bands than univariate method based on polystyrene absorbance (b) band at 1680nm�

Aquaphotomics: Sakudo, A., Tsenkova, R., Tei, K., Onozuka, T., Ikuta, K., Yoshimura, E., and Onodera, T.: Comparison of Vibration Mode of Metals in HNO3 by Partial Least Squares Regression , Bioscience, Biotechnology and Biochemistry, 2006.

Aquaphotomics: in micro particle analysis

Aquaphotomics: A NIRS Investigation into the Perturbation of Water Spectrum in an Aqueous

Suspension of Mesoscopic Scale Polystyrene Spheres. Asia Near Infrared conference, Tsukuba 2008

Heavy metal in water

Micro particles in water

0.020.640.001-0.0001%

0.0140.750.01-0.001%

0.3630.690.1-0.01%

R2(b)R2(a)Particle size

0.020.640.001-0.0001%

0.0140.750.01-0.001%

0.3630.690.1-0.01%

R2(b)R2(a)Particle size

Microbial organism detection (bacteria)

On going work

AQUAPHOTOMICS*Aqua - : water

Photo - : light

Omics - : all about,

complement of something

*Proposed by Roumiana Tsenkova in 2005

The PROTEOME is the entire complement of proteins expressed by a genome

The GENOME of an organism is its whole hereditary information and is encoded in the DNA

AQUAPHOTOME

is

the entire complements of water matrix absorbance

bands found under various perturbations.

Water Matrix Coordinates:

Water absorbance bands corresponding to the water

structural equilibria(coherent domains???)

WAMACS

Water Matrix Coordinates, WAMACSC2 C7 WAMACS

(nm)

WS1 V1+ free S0 H5O2 S1 V1, V2 (assignment)

V3 OH

温度変化Standard Deviation

PCA

PCA (2nd D)

ナノ粒子添加2nd D

PCA (MSC)

SD

PLS

スクロース・クエン酸添加

スクロース

PCA (2D)

PLS (MSC)

PLS (2D)

クエン酸

PCA (2D)

PLS (MSC)

PLS (2D)

アクアポリン発現

細胞＆細胞外液

PCA (2D)

PLS-DA　(MSC)

PLS-DA (2D)

細胞

PCA (2D)

SD

PLS （AQP-WT）

細胞外液

PCA (2D)

SD

PLS （AQP-WT）

S2

C1 C3 C4 C5 C6 C8 C9

V3 WS4

C10 C11 C12

S4S3

1344 1364 137213821398 1410 14381444 1464 1474 1492 1518

Temperature

Nano particles

SucroseCitric acid

Aquaporin

1330 1340 1350 1360 1370 1380 1390 1400 1410 1420 1430 1440 1450 1460 1470 1480 1490 1500 1510 1520

Water Illumination

Water Temperature

Water Time (Vapor)

HSA(Concentration High)

HSA(Concentration Low)

Globulin

Prion protein PrP

Prion protein PrP (Cu)

Prion protein PrP (Mn)

Prion protein Water (Mn)

Milk Protein (Healthy)

Milk Protein (Mastitis)

Milk Cell count (Mastitis)

Quarter milk(Healthy)

Quarter milk(Mastitis)

Milk Healthy

Milk Mastitis

Blood Healthy

Blood Mastitis

Urine Healthy

Urine Mastitis

Particle Polystyrene

Wheat Protein

Wheat PSI (Kernel texture)

Wheat MTI (Kernel texture)

Wheat Water (Kernel texture)

1506 15161370 1376 1398 14181336 1348

1380 1388 1421 14301432 1444

1448 14541462 1468

1472 14821482 1495

1360 1366C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 C11 C12

Water Absorbance Pattern, WAPAQUAGRAM

Aquagram was devised to visualize the WASP.

The aquagram displays normalized absorbance values at specific water

bands on the axes originating from the center of the graph.

Absorbance values at the WAMACs are placed on the respective radial axes.

Change in temperature

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

1344nm(V3)

1364nm(OH-H2O)

1372nm(V1+V3)

1382nm(O2-(H2O)4

1398nm(free OH)

1410nm(S0, free water )

1438nm(H5O2)

1444nm(S1)

1464nm(S2)

1474nm(S3)

1492nm(S4)

1518nm(V1, V2)

30℃

35℃

40℃

45℃

50℃

55℃

60℃

65℃

1300 1410 1490 1580 1680Wavelength (nm)

Abs

orba

nce

65℃60555045403530

1362nm, free water OH stretch (OH-H2O) 1374nm, 1371nm O2-.(H2O)4 1382nm, 1383nm interwater / DD stretch (OH-(H2O)4

Water Vapor Spectrum

Water and Water Vapor

-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5

3

3.5

1344nm(V3)

1364nm(water shell)

1372nm(V1+V3)

1382nm(water shell)

1398nm(free OH)

1410nm(S0, free water)

1438nm(H5O2)

1444nm(S1)

1464nm(S2)

1474nm(S3)

1492nm(S4)

1518nm(V1, V2)

moisture

water1300 1410 1490 1580 1680

Wavelength (nm)

Abs

orba

nce

1362

1382

1372

1452

water vapor water

MPA

http://www.gallery-sakura.com/search/higashiyama_kaii/gs1026.html

EXTENDED WATER MIRROREXTENDED WATER MIRROR

http://www.gallery-sakura.com/search/higashiyama_kaii/gs1026.html

EXTENDED WATER MIRROREXTENDED WATER MIRROR

EXTENDED WATER MIRROREXTENDED WATER MIRROR

with near infrared light, NIRwith near infrared light, NIR

As WATER is a molecular NETWORK, NIR lightNIR light

can penetrate, get partially absorbed and bring information for the

surrounding molecules making water a mirror on molecular level

Why Extended Water Mirror: Why Extended Water Mirror: ExampleExample

� Cd2+ has no spectral features in NIR�

� Prediction of Cd in water possible!!!

HOW?

� Cd2+alters water structure

� We detect Cd2+ indirectly, using the interaction of water and NIR light

24

-1 0 1 2 3 4 5 6 7 8 9 10-2

0

2

4

6

8

10

Concentration (mg L-1

)

Pre

dic

ted C

oncentr

ation (

mg L

-1)

30 oC

NaCl in Water

2. Near Infrared Spectroscopy

Overtone Vibrational Spectroscopy: a powerful tool

for non- invasive water observation:immense information about water

molecular conformations and respective function

for better understanding of the biological world

X-rays UV VISInfrared

NIR IR �Êwave

R

Abs

T

small

small

big

big

small

small

0.01 0.4 wavelengths�i����j2.5 500.6 1.0

R

T

Abs

NIR range

Relative intensities of C-H stretch bands for infrared and the overtone NIR regions

Band Wavelength region

Relative Intensity

Pathlength Optical Density

Fundamental band (n)

3380 - 3510 nm

100 0.01 mm 2

1st overtone (2n) 1690 - 1755 nm

1 1.0 mm 2

2nd over tone (3n) 1127 - 1170 nm

0.1 1.0 cm 2

3rd overtone (4n) 845 - 878 nm

0.01 10.0 cm 2

4th overtone (5n) 690 - 780 nm

0.005 10.0 cm 1

Why Near Infrared Spectroscopy?

� Long path lengths

� Non-destructive

� Fast

� Immense amount of information in one spectrum

Bio Fluid Analysis in Test Tubes

32

PORTABLE

INSTRUMENTS and

FIBER PROBES

FOR

BIO MONITORING AND

BIO DIAGNOSIS

Water spectra

NIR range

NIR and HNIR and H22OO

� Vibration frequency of hydrogen bonds in water are highly influenced by nearby molecules & environment

� Changes in these bonds may be used to gain information on the changing environments causing them (e.g. Presence of solutes, changing temperature)

37

Structures of the OH-·(H2O)3,4 clusters showing (A) OH-·(H2O)3, the completed first solvation shell, and (B) OH·(H2O)4, the onset of the second solvation shell.

Spectroscopic

Determination of the OH

Solvation Shell in the

OH·(H2O)n Clusters

OH·(H2O)2 1366 nm

OH·(H2O) 1480 nm..

.

William H. Robertson, Eric G. Diken, Erica A. Price, Joong-Won Shin, Mark A. Johnson* Science, Vol 299, Issue 5611, 1367-1372, 28 February 2003

EXAMPLE

Wavelengths, nm

Ａｂｓｏｒｂａｎｃｅ

WATER DYNAMIC SPECTRA2D and 3D Synchronous Correlation Maps

represent positive correlation

represent negative correlation

WATER DYNAMIC SPECTRAPower Spectra at 1530nm

3. Water Spectral Patterns as a biomarker of waters

Total 315 spectra

Methods for Aquagram Calculations

AA'

Calculating formula

A�: Value of AquagramA : Absorbance after MSC applied on 1st overtone region: Mean of all spectra : SD of all spectra: 12 Wavelengths(*)

<1342, 1364, 1372, 1382, 1398, 1410, 1438, 1444, 1464, 1474, 1492, 1518>(*): R. Tsenkova J. NIR infrared spectrosc. 17 303-314 (2009)

Aquagram of Various water types

-3

-2.5

-2

-1.5

-1

-0.5

0

0.5

1

1344nm(V3)

1364nm(water shell)

1372nm(V1+V3)

1382nm(water shell)

1398nm(free OH)

1410nm(S0)

1438nm(H5O2)

1444nm(S1)

1464nm(S2)

1474nm(S3)

1492nm(S4)

1518nm(V1, V2)

pure

16CONTROL 1

1/6NaCl1

19Duble Herixm3 1

5/6DoubleHerix 1

pen101

NW00133 1

NW00138 1

16CONTROL 1

Kongou1

19Duble Herixm3 1

pen51

Silver1

Austlia1

2/6D3-water1

1/6NaCl1

UPW1

Gold1

2/6CONTROL1

Copper1

5/6NaCl1

tap1

Kobe1

NW00145 1

Aquagram of Various Waters

-0.050

0.050.1

0.150.2

0.250.3

1344

1364

1372

1382

1398

14101438

1444

1464

1474

1492

pure

Kongou

-0.1-0.08-0.06-0.04-0.02

00.020.04

1344

1364

1372

1382

1398

14101438

1444

1464

1474

1492

pure

tap

-0.3-0.25

-0.2-0.15

-0.1-0.05

00.05

1344

1364

1372

1382

1398

14101438

1444

1464

1474

1492

pure

Gold

-0.1

-0.05

0

0.05

0.11344

1364

1372

1382

1398

14101438

1444

1464

1474

1492

pure

Kobe

Preprocessing:Smooth (21) + AutoscalingAquagram = Sample (avg) � Pure water (avg)

4. Water spectral pattern of urine as biomarker for estrus in Giant Panda

Fig. 2 Aquagrams of urine spectra from a female giant panda in(A: 2007, B: 2008, and C: 2009) and the Aquagram of median values (D) classified by hierarchical cluster analysis (HCA). Normalized absorbance values at the water matrix coordinates (WAMACs) (1344, 1364, 1372, 1386, 1410, 1424, 1444, 1456, 1464, 1474, 1494, and 1510 nm) were plotted on each axis. Red color is used to show high estrus aquagram.

A

DC

(B)

Fig. 1 Water matrix coordinates (WAMACs) related to the estrus in a female giant panda.

(A)Smoothed raw difference spectra of urine calculated as a difference between averaged spectra from the previous to the next day of E1G peak and the other averaged spectra

(B) Second derivative spectra of (A)(C) Second derivative spectra of 3 years� urine samples after smoothing. From this

plot, 12 absorbance bands were selected as WAMACs.

(A)

(C)

B

K. Kinoshita, K. et al. , Water spectral pattern of urine reveals estrus in Giant Panda, Nature Science Reports, accepted, 2012

5. Water Spectral Patterns of soy bean plant as a biomarker of biotic and abiotic stress

49

Inoculated plants set Control plants set

- Soybean, 20 plants were grown in a greenhouse.

- Only 10 plants were inoculated with SMV

- All the plants were kept under similar conditions.

Method

Mosaic Virus Diagnosis

B. Jinendra, K. Tamaki, S. Kuroki et al., Biochemical and Biophysical Research Communications 397 (4), 685 (2010).

50

Accumulated , 20 plants x 10 spectra per plants x 12 days = 2400

USED ELISA reference method data

- to confirm the infection,

- in the model building

Raw spectra

Actual number of disease plantsNumber of disease plant identified

Actual number of healthy plantsNumber of healthy plant identified

Disease diagnosisspecificity =

Disease diagnosis sensitivity =Actual number of disease plantsNumber of disease plant identified

Actual number of healthy plantsNumber of healthy plant identified

Disease diagnosisspecificity =

Disease diagnosis sensitivity =

SIMCA discrimination approach

Method

51

Factor1

Factor2

Factor3123

456789

1011121314151617181920212223242526272829

303132333435363738

39 40414243444546474849

5051525354555657585960616263646566676869

707172737475767778

7980818283848586878889909192939495969798

99101102103104105106107108109110111112113114115116117

118125126127128129130131132133134135136137

138139140141142143144145146147148149150151152153154155156157

158159160161162163164165166167168169

170171172173174175176177178179180181182183184185186187188189190191192193194195196

201203205208211213216217218219222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255

256258259260261263264266267268269270271272273274275346347348349350351352353354355

366367368369370371372373374375376377378379380381382383384385386447449450451452454455

552554582583

595596597598599600601602603

604605606607608609610611612613 614615616617618619620621622623624625626627628629630631632

633634635636637638639640641642643644645646647648649651652653654655656657658659660661662663664665

686687688689690691692693694695696697698699700701702703

704705706707708709710711712713714715716717718719720721722723724725726727728729730731732733

746747748749750751752753755758760762763764775776777778779780

781782783784

788790791792793794795796797799800801802804

805806807

808809810811812813814818819820821823824

825827828829830831832833834835840841842843844

845846847850854857

865866867868869870871872

873874

876877878879881883884885886887890893 895896897898899900901902903904

905906907908911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944

975976977978979981983984985986987988989990991992993994

10641065106610671068106910701071107210731082108310841085

10861087108810891090

113011311132113311341135113611371138

1238124212431245124712491250125112521253125412551256

12571258125912601261126212631264126512661277127812791280128112821283128412851286137613831384

1428142914301432143314341435

1536153715381539154015411542

154315441545154615471548154915501551

15521553155415551556155715581559156015911592159315941595

1716171717181719172017211722172317241725

1729173417521754

17851786178817891791179217931794

1796

180518061808180918101811181318141826

183618371838183918401841184218431844

1856185718581859186018611862

1865186618671868186918701871187218731874

19051906190719081909191019111912191319141955195619571958195919601961196219631964100

297298299300301302303304305

306307308309310311312313314315316317318319320321322323324325

326327328329330331332333334335336337338339340341342343344345

356357358359360361362363364365

387388389390391392393394395396406407408409410411412413414415416417418419420421422423424425

426427428429430431432433434435

436437438439440441442443444445

446448453

456457458459460461462463464465466

467468469470471472473474475

476477478479480481482483484485

486487488489490491492493494495

496497498499500501502503504

505506507508509510511512513514

515516517518519520521522523524

525526527528529530531532533534

535536537538539540541543544545546547548549550551553

555556557558559560561562563564565566567568569570571572573574

575576577578579580581584

585586587588589590591592593594666667668669670671672673674675676677678679680681682683684685888889891892894

90991099599699799899910001001100210031004

1015101610171018101910201021102210231024

1035103610371038103910401041104210431044

123912401241124412461248

1267126812691270127112721273127412751276

1287128812891290129112921293129412951296

1297129812991300130113021303130413051306

130713081309131013111312131313141315

1316

1317131813191320132113221323132413251326132713281329133013311332133313341335

13361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370

13711372137313741375

1377137813791380138113821385

1386138713881389139013911392139313941395

1396139713981399140014011402140314041405

1406140714081409141014111412141314141415

1416141714181419142014211422142314241425142614271431

1436143714381439144014411442144314441445

1446144714481449145014511452145314541455145614571458

1459146014611462146314641465

1466146714681469147014711472147314741475

147614771478147914811482148314841485

14861487148814891490149114921493149414951496149714981499150015011502150315041505

150615071508150915101511151215131514151515161517151815191520152115221523152415251526

152715281529153015311532153315341535

1775177617771778177917801781178217831784

1845184618471848184918501851185218531854

1855186318641885188618871888188918901891189218931894

1895189618971898189919001901190219031904

1936193719381939194019411942194319441965196619671968196919701971197219731974

19751976197719781979198019811982198319841995199619992000200220032004

2005200620072008200920102011201220132014

Factor1

Factor2

Factor3

11

12

13

14151617

18

2425

26

27282930

3132

37

38

3940

41424344

4551

5253

61 6267

71

72

8485

86

97

9899100

101

102110

111123136

137

3

4

5

6

7891019

20212223

33343536

46474849

5455

5657

5859

63646566

68

6970

7374

7576

77787980818283

87 8889

909192

93949596103104

105106107108109

112

113114

115

116117118119120

121122

124125

126

127128129130

131132133134135

138139140141142143

144145146147

Before averaging After averaging

PCA specificationsPreprocessing: Mean-centerTransforms: 2nd derivative Factors: 8

Healthy and infected plants were clearly differentiated after averaging

Principal component space

52

Standard deviation profiles of two plant types

946 nm = (S0) Non hydrogen bonded water976 nm = (S2) Bonded water (with two hydrogen bonds) (Abe et al., 1995)

Highly variable wavelengths with the progress of disease

Water index WI = R970/R900 (Penuelas and Filella,1998)

0.003

0.0035

0.004

0.0045

0.005

0.0055

0.006

730 780 830 880 930 980 1030

Wavelength nm

Sta

nd

ard

dev

iati

on

Healthy SD

Diseased SD

900 nm946 nm

976 nm

6 nm blue shift

964 nm

53

Specific water absorbance patterns (WAP) at 870 nm, 912 nm, 936 nm, 952 nm specific for SMV infection.

Fig. (A) Second derivative average spectra

AQUAPHOTOMICS

The blue shift at 970 nm virus influence on common leaf water stress.

(Penuelas and Filella,1998)

740 760 780 800 820 840 860 880 900 920 940 960 980 1000 1020

-0.025

-0.020

-0.015

-0.010

-0.005

0.000

0.005

0.010

2nd

deriv

ativ

e ab

sorb

ance

Wavelength , nm

870 912 936 952 970

6 nm blue shift for infected plants Infected

Non infected

(Jinendra et al., 2010)

Homework1. Development of new multivariate methods

for spectral data analysis and assignment

2. Understanding the 12 or more absorbance ranges in the water spectrum (coherent domains??? or molecular structures)

水の音

光の遊び

生きている

Sound of a stream,Sunlight dancing on waters,Life wakes up again.

R. Tsenkova, 2004