RESEARCH ARTICLE

Aberrant protein expression in plasma and kidney tissue

during experimental obstructive nephropathy

Eleni Giannakis1, Chrishan S. Samuel1,2, Tim D. Hewitson3,4, Wee-Ming Boon5, Mary Macris1,Shane Reeve6, Josie Lawrence6, A. Ian Smith6, Geoffrey W. Tregear1,2 and John D. Wade1,7

1 Howard Florey Institute, The University of Melbourne, Victoria, Australia2 Department of Biochemistry and Molecular Biology, The University of Melbourne, Victoria, Australia3 Department of Nephrology, The Royal Melbourne Hospital, Victoria, Australia4 Department of Medicine, Royal Melbourne Hospital, University of Melbourne, Victoria, Australia5 Department of Physiology, Monash University, Victoria, Australia6 Department of Biochemistry, Monash University, Victoria, Australia7 School of Chemistry, University of Melbourne, Victoria, Australia

Received: February 16, 2009

Revised: May 26, 2009

Accepted: June 29, 2009

Kidney failure is a major health problem worldwide. Patients with end-stage renal disease

require intensive medical support by dialysis or kidney transplantation. Current methods for

diagnosis of kidney disease are either invasive or insensitive, and renal function may decline

by as much as 50% before it can be detected using current techniques. The goal of this study

was, therefore, to identify biomarkers of kidney disease (associated with renal fibrosis) that

can be used for the development of a non-invasive clinical test for early disease detection. We

utilized two protein-profiling technologies (SELDI-TOF MS and 2-D) to screen the plasma

and kidney proteome for aberrantly expressed proteins in an experimental mouse model of

unilateral uretric obstruction, which mimics the pathology of human renal disease. Several

differentially regulated proteins were detected at the plasma level of day-3-obstructed animals,

which included serum amyloid A1, fibrinogen a, haptoglobin precursor protein, haptoglobin

and major urinary proteins 11 and 8. Differentially expressed proteins detected at the tissue

level included ras-like activator protein 2, haptoglobin precursor protein, malate dehy-

drogenase, a enolase and murine urinary protein (all po0.05 versus controls). Immunohis-

tochemistry was used to confirm the up-regulation of fibrinogen. Interestingly, these proteins

are largely separated into four major classes: (i) acute-phase reactants (ii) cell-signaling

molecules (iii) molecules involved in cell growth and metabolism and (iv) urinary proteins.

These results provide new insights into the pathology of obstructive nephropathy and may

facilitate the development of specific assay(s) to detect and monitor renal fibrosis.

Keywords:

2-D / Protein profiling / Renal fibrosis / SELDI-TOF MS / Unilateral uretric obstruction

1 Introduction

Obstructive nephropathy encompasses a group of disorders

with impaired urine flow due to abnormalities of the urinary

tract. Clinical outcomes depend on many factors including

age, severity (complete or partial), totality (bilateral or

unilateral) and duration of obstruction [1]. In the adult

kidney, permanent unilateral obstruction leads to destruc-

tion of the renal parenchyma through interstitial inflam-

mation, tubular apoptosis and interstitial fibrosis [2];

ultimately leading to end-stage renal disease (ESRD), which

requires intensive medical support by dialysis or trans-

plantation.

Despite its significance, the mechanisms associated

with obstructive nephropathy remain poorly understood.

A National Institutes of Health working party, therefore,

Abbreviations: ESRD, end-stage renal disease; MUP, major

urinary protein; SAA1, serum amyloid A1; UUO, unilateral

ureteric obstruction

Correspondence: Professor John D. Wade, School of Chemistry,

University of Melbourne, Victoria 3010, Australia

E-mail: john.wade@florey.edu.au

Fax:161-3-9348-1707

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

Proteomics Clin. Appl. 2009, 3, 1211–1224 1211DOI 10.1002/prca.200900021

concluded that development of biomarkers in appropriate

clinical and experimental models was a major research

priority in obstructive nephropathy. These would enable

both the measurement of renal injury and the functional

impairment, and a means of elucidating the cellular and

molecular basis of the disorder [3, 4].

Although examination of biopsy tissue remains the best

current diagnostic tool for biomarker discovery, the proce-

dure is invasive, difficult to interpret and by its nature is

limited in frequency. Biochemical measures of renal func-

tion are therefore routinely used. The increase in serum

creatinine for instance is proportional to the loss of renal

output. However, such biochemical parameters are inher-

ently deficient, as they are both reactive and relatively

insensitive. The clinical reality is therefore that most

patients present with established renal damage, where their

renal function may have declined by as much as 50% before

it is detected.

It is therefore hypothesized that identifying biomarkers of

obstructive nephropathy will lead to the development of a

non-invasive clinical means for early detection of kidney

disease, which will allow timely therapeutic intervention

and prevention of established/irreversible renal damage.

This may subsequently lead to enhanced diagnosis, prognosis

and treatment monitoring of high-risk patients. The identi-

fication of biomarkers may also lead to the subsequent

discovery of potential drug targets to control the cascade of

destructive effects, thereby minimizing and ultimately

preventing the adverse outcomes associated with the disease

state. This is of particular importance since current treatment

strategies are only effective at impeding ESRD by a matter of

months.

The objective of this study was therefore to use protein-

profiling techniques to identify potential novel biomarkers

of renal disease in a murine model of unilateral ureteric

obstruction (UUO). Although complete unilateral obstruc-

tion is not a common form of renal disease in humans, it

closely mimics, in an accelerated manner, the different

stages of obstructive nephropathy [5, 6]. Furthermore, its

pathogenesis shares many features with other forms of

tubulointerstitial renal disease, where the final fibrotic

phase is very similar in all progressive renal diseases,

regardless of the underlying etiology. Thus, delineating the

factors that contribute to UUO will not only provide

potentially specific therapeutic targets, but will also contri-

bute to our understanding of renal disease in general.

In this study, we compared the protein profile of renal

tissue and plasma from normal mice and animals 3 days after

UUO. The experimental time point was chosen on the basis

of our previous studies demonstrating early rapidly progres-

sive inflammation and fibrogenesis in this model [7]. To

ensure extensive investigation of the proteome we coupled (i)

SELDI-TOF MS, which is useful for analyzing the lower-

molecular-weight proteome (o45 kDa) and (ii) complemen-

tary 2-D which, in contrast, is advantageous for assessing the

higher-molecular-weight proteome.

2 Materials and methods

2.1 Animals and experimental protocol

This study primarily focuses on protein profiling of plasma

and kidney tissue from the UUO model on the basis that

UUO is a widely used model of primary renal tubulointer-

stitial fibrosis, which occurs independently of species and

strain and demonstrates changes that mimic the pathology

of human progressive renal disease [5–8]. It should be noted

that urine was not analyzed in this study. UUO results in a

single non-functioning kidney (no urine output) and a

functioning contralateral kidney (normal or increased urine

output). We chose to examine tissue and plasma, and avoid

any confounding variables (e.g. urine flow and intact

glomerular filtration barrier) associated with the uninjured

contralateral kidney.

For studies involving the UUO model, 4- to 6-wk-old

male wild-type mice (on a C57B6J� 129SV background;

n 5 4 control and n 5 5 disease animals were supplied for

the SELDI MS studies) were generated from heterozygous

parents and housed in a controlled environment with a fixed

lighting schedule and free access to rodent lab chow (Bara-

stock Stockfeeds, Pakenham, Victoria, Australia) and water.

UUO surgery was performed under general anesthesia. The

left ureter of each mouse was ligated with 5.0 surgical silk

while the contralateral ureter remained intact. The incision

was sutured and mice were allowed to recover after temgesic

administration (Buprenorphine; Reckitt Benckiser, West

Ryde, NSW, Australia). The outlined experiments were

approved by the Howard Florey Institute’s Animal Ethics

Committee, which adheres to the Australian Code of Prac-

tice for the Care and Use of Laboratory Animals for Scien-

tific Purposes.

2.2 Tissue and plasma collection

Cardiac puncture was used to collect plasma while kidney

tissues were isolated from wild-type animals at day-0 (control;

n 5 4) and day-3 (n 5 5) post-UUO, as described previously

[7]. UUO affects different parts of the kidney in different

ways. In order to ensure that all anatomical compartments

were represented in our analysis, equivalent cross-sections of

the kidney were collected from each animal. All samples were

then stored at �801C until required.

2.3 Tissue extraction

Frozen tissue was pulverized into a fine powder using a

mortar and pestle and immediately transferred to 1.5 mL

tubes. Extraction buffer (10 mM Tris-HCl, 10 mM NaCl,

0.1% Triton X-100, pH 7.6, and protease inhibitor, complete

mini EDTA free, Roche Diagnostics, Mannheim, Germany)

was then added to the disrupted tissue (10 mL/mg tissue),

1212 E. Giannakis et al. Proteomics Clin. Appl. 2009, 3, 1211–1224

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

before samples were further homogenized by passing the

extract through a 19G needle ten times. The homogenates

were incubated on ice for 1 h with periodic mixing, followed

by centrifugation at 15 000� g at 41C for 10 min. The

supernatants (containing total cellular proteins) were

aliquoted into fresh tubes and stored at �801C until

required.

2.4 Sample handling and standardization

All tissue and plasma samples were collected, stored and

processed according to a strict standard operating procedure

to ensure that both (i) conformity of sample preparations was

achieved and (ii) differences observed between groups were a

true reflection of the phenotypes being examined rather than

a consequence of variations introduced during processing.

Furthermore, analysis was performed with replicate samples

to ensure reproducibility, as described previously [9]. Protein

standardization was one of the key steps taken to ensure

conformity of the tissue preparations. Once protein concen-

trations from the tissue homogenates were quantified (as

described in Section 2.5), the concentration of samples was

subsequently standardized using extraction buffer; hence,

both the protein concentration and the buffer constituents of

all samples were identical prior to performing a final dilution

in the binding buffer. This is a critical step as the protein

concentration affects both peak number and intensity and the

buffer constituents determine the binding stringency. The

resulting protein patterns produced were similar, demon-

strating conformity of sample preparations.

2.5 Total protein quantitation assay

A Bradford protein quantification assay was performed

according to the manufacturer’s instructions (Quanti-

chrome Protein Determination Assay, Bio Assay Systems,

Hayward, CA, USA) to determine the total protein concen-

tration of each sample.

2.6 Biomarker detection – plasma preparation for

SELDI-TOF MS analysis

Plasma samples were diluted in 8 M urea, 1% CHAPS, PBS

(1:1.5) and incubated on a shaking platform for 30 min on ice.

Samples were then diluted in the appropriate wash buffer

(1:30) in preparation for analysis with SELDI-TOF MS.

2.7 Biomarker detection – tissue preparation for

SELDI-TOF MS analysis

Total protein concentrations of each sample were adjusted

to the same concentration by using the extraction buffer, to

ensure they were identical with respect to protein concen-

tration and buffer constituents. The samples were subse-

quently diluted to a final concentration of 1 mg/mL in the

appropriate wash buffers in preparation for analysis with

SELDI-TOF MS.

2.8 SELDI-TOF MS analysis

Samples were analyzed in duplicate using the following

arrays and wash buffers: (i) Q10 (anion exchange) and

100 mM Tris-HCl, pH 9, (ii) H50 (hydrophobic) and 5%

ACN, 0.1% TFA and (iii) IMAC30 (immobilized metal affi-

nity capture) pre-coated with 100 mM ZnSO4 according to

the manufacturer’s instructions (Bio-Rad Laboratories,

Hercules, CA, USA) and PBS, pH 7.2.

Arrays were initially assembled into a bioprocessor (Bio-

Rad Laboratories) and pre-equilibrated into the appropriate

wash buffer. Each well was loaded with 50 mL of wash

buffer, centrifuged for 1 min at 1500� g to remove air

bubbles, followed by an incubation step for 5 min on a

shaking table. The buffer was decanted and the equilibration

step was repeated for a total of two washes. The samples

(50 mL) were applied to each well in duplicate and incubated

for 1 h. Samples were discarded, chips were washed with

the appropriate wash buffer three times for 5 min, followed

by two 1 min washes with 1 mM HEPES, pH 7.2. The

chips were removed from the bioprocessor, air dried and

l mL of 50% saturated sinapinic acid in 50% v/v ACN,

0.5% TFA was applied onto each spot twice, arrays were air-

dried between each application. Chips were analyzed by

SELDI-TOF MS using a PBSIIc instrument (Bio-Rad

Laboratories) at the low (0–20 kDa) and high (20–100 kDa)

mass settings. Resulting spectra were examined using

ProteinChips software (Bio-Rad Laboratories) and the

expression difference mapping tool which groups peaks of

the same molecular weight across all samples and statisti-

cally displays the differences in expression between sample

groups.

2.9 Biomarker detection – 2-D

Tissue and serum were prepared for 2-D using the 2-D gel

clean up kit (Amersham Biosciences, Piscataway, NJ, USA)

according to the manufacturer’s instructions. Briefly, 500mg

of total protein was precipitated, the pellet was washed to

remove non-protein contaminants and finally resuspended

in rehydration buffer (8 M urea, 2% CHAPS and 0.002%

bromophenol blue) to a final volume of 100 mL.

In total, 100 mg of protein in rehydration buffer was then

combined with 2.5 mg DTT, 2.7 mL IPG buffer, pH 3–10

(Amersham Biosciences) and rehydration buffer to a

final volume of 125 mL. Samples underwent isoelec-

trofocusing (EttanTM IPGphorII; GE Healthcare Bios-

ciences, Chalfont, UK) with a 7 cm immobiline gel strip, pH

Proteomics Clin. Appl. 2009, 3, 1211–1224 1213

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

3–10 (GE Healthcare Biosciences) under the following

conditions: rehydrate 50 V 14 h, focus 300 V 1 h, gradient

1000 V 0.5 h, gradient 5000 V 1.5 h, step 5000 V 0.5 h and

hold 50 V 12 h. IPG strips were subsequently washed for

15 min with 5 mL of equilibrating solution (0.5 M Tris-HCl,

235 mM urea, 4% SDS and 34.5% glycerol) containing

12.5 mg DTT, followed by a 15 min wash in equilibrating

buffer containing 0.225 g iodoacetamide. IPG strips were

assembled in the well of a 2-D gel (12% Tris-Tricine, Invi-

trogen, Carlsbad, CA, USA) and sealed with an agarose

solution containing 0.002% bromophenol blue, 100 mM

Tris base, 0.76 M glycine and 0.4% SDS. Protein standards

(10 mL, Bench Mark, Invitrogen) were loaded onto each gel.

Gels were run for 3 h at 40 mA (Hoefer miniVE vertical

electrophoresis system; Isogen Lifesciences, IJsselstein, The

Netherlands) and subsequently stained with 0.025%

Coomassie RT50 brilliant blue in 10% acetic acid overnight

followed by destaining in 10% acetic acid. Biomarker

analysis was conducted on samples run on three separate

gels, using PDQuest Advanced 8.0.1 (Bio-Rad Laboratories),

to ensure accuracy in the estimation of expression levels.

2.10 Protein purification and identification of

SELDI-TOF MS biomarkers

The protein markers identified by SELDI-TOF MS were

purified using RP-HPLC and/or resins matched to the

chemistry of the array surface, which was used to detect the

biomarker (Bio-Rad Laboratories). Resulting elution frac-

tions were screened on ProteinChip arrays to confirm the

fraction, which contained the peak of interest. In some

circumstances, it was necessary to complete several

column runs as the protein of interest was not present in

high concentrations. Elution fractions were subsequently

pooled, concentrated using a vacuum concentrator or

freeze drier and electrophoresed on a 1-D/2-D SDS-PAGE

gel as a final separation step. After excising the band

and eluting the protein from the gel, the purified

sample was analyzed on an NP20 array (Bio-Rad Labora-

tories) to confirm the band was having the correct molecular

mass. Once purified, biomarkers were digested with

trypsin and the masses of the proteolytic fragments

determined by MALDI-TOF MS (Applied Biosystems

Proteomics Analyser 4700) were submitted to the

MASCOT database for identification. Further confirmation

of the protein identification was achieved by performing

collision-induced dissociation using MALDI-TOF/TOF

MS (Applied Biosystems Proteomics Analyser 4700). Briefly,

MS spectra were acquired by reflectron in positive mode,

scanning a mass range of 800–4000 Da. Total shots perspectrum 5 1500 with a fixed laser intensity of 3400. The

spectra were assessed using a processing method with

internal calibrants (trypsin peaks), minimum signal-to-

noise 5 20 and a mass tolerance of m/z 1/�0.5. An inter-

pretation method was subsequently used to obtain MS/MS

information using a max precursor/spot 5 10 (strongest

first) as a precursor selection criteria, and weakest precursor

first as MS/MS selection criteria, also excluding

adducts such as sodium and potassium. The MS/MS

acquisition was in positive reflectron mode, total shot/

spectra 5 2500 and laser intensity of 3700. MS/MS proces-

sing was performed using a min. signal-to-noise 5 10 with a

default calibration. The acquisition, processing of both MS

and MS/MS spectra was done in batchmode using 4700

series explorer V3.0.

Sequencing of biomarkers was also achieved using

Edman degradation methodology as described previously

[10].

2.11 Protein purification and identification of 2-D gel

biomarkers

Protein markers identified by 2-D electrophoresis

were excised using a scalpel blade. Gel plugs were

washed with 40% methanol, 5% acetic acid for 15 min

followed by a 5 min wash in water to remove SDS.

Coomassie stain was removed from gel plugs by washing

with 50% ACN, 50 mM NH4HCO3 for 20 min. In some

circumstances, it was necessary to repeat this step to

ensure the stain was completely removed. Gel plugs were

then washed with 20 mM NH4HCO3 for 60 min and dehy-

drated by incubation in ACN for 10 min, followed by

centrifugation in a speed vacuum. Gel plugs were subse-

quently rehydrated in a trypsin solution (5 mL trypsin

500 ng/mL in 0.1% TFA) and 44 mL of 20 mM NH4HCO3,

1 mM CaCl2. Gel pieces were finally incubated at 371C

overnight and the identity of the peptide fragments was

determined by MALDI-TOF MS and MALDI-TOF/TOF MS

as described in Sections 2.8 and 2.10.

2.12 Immunohistochemistry

To further confirm that the identified biomarkers (by

SELDI-TOF MS and 2-D) could also be detected by a non-

proteomic-based method, immunohistochemistry was

performed to examine changes in fibrinogen expression in

4% formalin-fixed, paraffin-embedded sections from all

control and obstructed animals. Once paraffin-embedded

samples were de-waxed, they were quenched with

hydrogen peroxide, blocked with normal serum and incu-

bated with rabbit anti-fibrinogen (Dako, Glostrup,

Denmark) for 1 h. After washing, sections were incubated

with a species matched biotinylated anti-IgG (Vector,

Burlinghame, CA, USA). Sections were washed again and

incubated with avidin–biotin complex (ABC Elite; Vector).

The chromogenic substrate 3,30-diaminobenzidine (DAB;

Sigma-Aldrich) was used for detection. Finally, sections

were counterstained with haematoxylin and examined

microscopically.

1214 E. Giannakis et al. Proteomics Clin. Appl. 2009, 3, 1211–1224

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

2.13 Statistical analysis

The SELDI-TOF MS data (using the Expression Difference

Mapping tool in ProteinChip software) and 2-D gel

analysis (using both PD Quest Advanced, version 8.0.1 and

Excel software) were analyzed by the Mann–Whitney

(non-parametric) test. Results are expressed as the

mean7SEM with statistical significance accepted at the

level of po0.05.

3 Results

3.1 SELDI-TOF MS analysis of kidney tissue

Kidney tissue extracts from control and obstructed

animals were screened on several arrays including H50,

Q10 and IMAC30 arrays pre-coated with 100 mM ZnSO4.

A total of 212 peaks with a signal-to-noise value of 3

present in at least 20% of spectra were detected over the

Figure 1. Representative spec-

tra from control and UUO

kidney extracts generated by

SELDI-TOF MS analysis using

on H50, Q10 and IMAC-Zn

arrays.

Table 1. SELDI-TOF MS analysis of differentially expressed kidney tissue biomarkers on IMAC, Q10 and H50 ProteinChip arrays

IMAC arrays Q10 Arrays H50 Arrays

Up-regulated indisease

Down-regulated indisease

Up-regulated indisease

Down-regulated indisease

Up-regulated indisease

Down-regulated indisease

14 803 4121 3793 5147 780615 792 9206 6097 7130 7913

9903 6171 9897 15 58918 313 6227 973118 649 8186

11 05216 76329 964

All biomarkers listed are significantly (po0.05) different in 3-day-obstructed kidney tissue versus control samples; and were differentiallyaltered in all n 5 5 UUO samples analyzed.

Proteomics Clin. Appl. 2009, 3, 1211–1224 1215

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

three surfaces examined at both low- and high-molecular-

weight settings (H50: 58 peaks, Q10: 69 peaks, IMAC: 85

peaks).

The observed protein profiles were unique for each

surface examined (Fig. 1), demonstrating how specific

chromatographic chemistries can be utilized to enrich for

subsets of the proteome. The surface which revealed the

greatest peak count was the IMAC ZnSO4 arrays. However,

comparison of the protein profiles between control versusday-3-obstructed animals revealed that the Q10 arrays in fact

yielded the highest number of differentially expressed

proteins as determined by univariate t-test analysis with

p-valueso0.05. Q10 revealed 13 differentially expressed

proteins, eight of which were down-regulated, whereas the

IMAC analysis revealed two differentially expressed proteins

both of which were down-regulated and H50 arrays revealed

seven differentially expressed proteins, three of which were

down-regulated (Table 1).

Figure 2. Biomarkers with m/z of 18 338 and 18 666 Da up-regu-

lated in the obstructed kidney tissue were identified to be ras-like

activator protein 2. Representative spectra generated by SELDI-

TOF MS using Q10 array, virtual gel view and scatter plot analysis

of these peaks are shown in panels A, B and C, respectively.

Figure 3. Biomarkers with m/z of 11 052 Da down-regulated in the

obstructed kidney tissue were identified to be murine urinary

protein. Representative spectra generated by SELDI-TOF MS

using Q10 array, virtual gel view and scatter plot analysis of this

peak are shown in panels A, B and C, respectively.

1216 E. Giannakis et al. Proteomics Clin. Appl. 2009, 3, 1211–1224

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

One of the most significant changes detected at the tissue

level of the UUO model was that of a cluster of peaks with

m/z value of 18 338 (mean7SD; control: 0.5670.23 versusUUO: 4.2772.6) and 18 666 Da (mean7SD; control:

0.5170.16 versus UUO: 3.2971.19); which both showed

significant increases in average peak intensity in the

obstructed animals (both po0.02; Fig. 2). These peaks were

identified by MALDI-TOF/TOF MS analysis to represent

Ras-like activator protein 2, which is a small G protein

involved in signaling and cell communication.

A biomarker with m/z value of 11 052 Da (Fig. 3) was also

purified and subjected to MALDI-TOF/TOF MS and was

identified to be murine urinary protein, and which was

significantly (po0.05) down-regulated in obstructed animals.

Figure 4. Representative spec-

tra from control and UUO

plasma samples generated by

SELDI-TOF MS using H50, Q10

and IMAC-Zn arrays.

Table 2. SELDI-TOF MS analysis of differentially expressed plasma biomarkers on IMAC, Q10 and H50 ProteinChip arrays

IMAC arrays Q10 Arrays H50 Arrays

Up-regulated indisease

Down-regulated indisease

Up-regulated indisease

Down-regulated indisease

Up-regulated indisease

Down-regulated indisease

11 597 14 807 5881 13 618 580011 679 7493 13 742 581011 738 7919 18 652 11 599

11 597 8303 11 74011 747 8902 11 95411 953 13 953 26 04714 956 14 06515 16515 81015 994

All biomarkers listed are significantly (po0.05) different in the plasma of 3-day-obstructed animals versus control plasma samples; andwere differentially altered in all n 5 5 UUO samples analyzed.

Proteomics Clin. Appl. 2009, 3, 1211–1224 1217

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

3.2 SELDI-TOF MS analysis of plasma

Plasma samples collected from obstructed and control

animals were denatured in 8 M urea/1% CHAPS buffer and

screened on H50, Q10 and IMAC30 arrays pre-coated with

100 mM ZnSO4. In total, 210 peaks with a signal-to-noise

value of 3 present in at least 20% of spectra were detected

over the various surfaces examined at both low- and high-

molecular-weight settings (H50: 68 peaks; Q10: 80 peaks,

IMAC: 62 peaks).

The observed plasma profiles on each surface were

generally unique, however, considerable similarity was

observed between the H50 and Q10 profiles in the low-

molecular-weight range with a representative spectra being

shown in Fig. 4. The surface which revealed the greatest

peak count and the highest number of differentially

expressed proteins (n 5 17) was the Q10 arrays, the majority

of which (n 5 10) were up-regulated. The IMAC arrays

revealed four differentially expressed proteins, three of

which were up-regulated. H50 arrays revealed six differen-

tially regulated proteins, all of which were up-regulated

(Table 2).

Figure 5. Plasma biomarkers with m/z of 11 592 and 11 737 Da up-

regulated in UUO animals were identified to be SAA1. Repre-

sentative spectra generated by SELDI-TOF MS using Q10 array,

virtual gel view and scatter plot analysis of these peaks are

shown in panels A, B and C, respectively.

Figure 6. Plasma biomarker with m/z of 13 742 Da was identified

to be down-regulated in the disease group, the identity of this

protein is unknown. Representative spectra generated by SELDI-

TOF MS Q10 arrays, virtual gel view and scatter plot analysis of

this peak are shown in panels A, B and C, respectively.

1218 E. Giannakis et al. Proteomics Clin. Appl. 2009, 3, 1211–1224

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

One of the most significant changes detected in the

plasma, was the up-regulation of a cluster of peaks with m/zvalues of 11 592 and 11 737 Da which showed an approx-

imate four and sixfold increase in average peak intensity in

the disease group, respectively (both po0.02; Fig. 5). These

peaks were detected on all surfaces examined and on each

occasion showed significant increase in expression. During

the purification process, these peaks were difficult to sepa-

rate by chromatography methods suggesting that they may

potentially be related. Standard sequencing metholodologies

of tryptic digestion and MALDI-TOF/TOF MS analysis

revealed that this protein was serum amyloid A1 (SAA1).

A separate biomarker with m/z of 13 742 Da (Fig. 6) was

also purified and subjected to MALDI-TOF/TOF MS;

however, no significant hits were detected suggesting that

this protein may be novel. Several other peaks which were

differentially regulated remain to be identified.

3.3 2-D

Kidney extracts and plasma collected from control and

disease animals were separated by 2-D (pH 3–10, 12% Tris-

tricine). Up to 137 kidney proteins and 106 plasma proteins

(representing only the soluble proteins; as other fractions

caused substantial smearing on 2-D gels) were detected after

Coomassie blue staining. 2-D gel analysis software was

consequently used to match the corresponding spots among

gels and compare the spot intensity in pixel units. Of the

137 kidney proteins detected, 21 were identified to be

differentially regulated. The majority of these proteins were

identified to be down-regulated, which is consistent with the

SELDI analysis. Of the 106 plasma proteins, six were iden-

tified to be differentially expressed, all of which were up-

regulated. This also showed consistency with the SELDI

studies in which the majority of the plasma proteins were

identified to be up-regulated. Figures 7 and 8 show repre-

sentative 2-D kidney and plasma gels, respectively, from

control and disease animals with differentially regulated

spots labeled; whereas Tables 3 and 4 show the pixel unit

values and p-value calculations of these biomarkers,

respectively. Altered proteins were subsequently excised and

subjected to MALDI-TOF/TOF MS. Several differentially

regulated spots were, however, not subjected to further

assessment by MS analysis as the spot intensity was

considered to be too low for sequence analysis. Sequence

data were obtained from three of the differentially regulated

kidney proteins, which included haptoglobin precursor

protein, a enolase and malate dehydrogenase. The hapto-

globin precursor protein was significantly increased

(p 5 0.01 versus levels in control tissue), whereas a enolase

and malate dehydrogenase were markedly decreased (both

po0.05 versus levels in control tissue) in extracts obtained

from day-3-obstructed kidneys (Table 3). Many of the tissue

proteins subjected to MALDI-TOF/TOF MS were, however,

unidentified. These proteins could be post-translationally

modified, thus difficult to identify by classical proteomic

technologies. Additionally, of the several plasma proteins

that were subjected to sequencing, the identity of four was

confirmed to be haptoglobin, haptoglobin precursor, fibri-

nogen a and major urinary protein (MUP) (Table 4).

Interestingly, haptoglobin precursor protein was identified

to be up-regulated in both tissue extracts (Table 3) and

plasma (Table 4) obtained from day-3-obstructed animals.

3.4 Immunohistochemical confirmation of

biomarker detection

Although there is limited availability of suitable antibodies to

the identified biomarkers (characterized by SELDI-TOF MS

Figure 7. 2-D gel analysis of kidney extracts from UUO and

control groups. Statistically significant differentially regulated

spots are highlighted and numbered which correspond to data

summarized in Table 3.

Proteomics Clin. Appl. 2009, 3, 1211–1224 1219

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

and 2-D gel analysis) in the mouse a polyclonal antibody to

fibrinogen was used to confirm (by immunohistochemical

means) that fibrinogen expression was up-regulated in all

kidney tissues of mice subjected to UUO (Fig. 9). The

increased staining was particularly evident in glomerular

capillary loops and the interstitial space adjacent to peri-

tubular capillaries (Fig. 9); and confirmed the increase in

fibrinogen detected by the proteomic-based methods.

4 Discussion

The prognostic significance of the proteome in renal

diseases is frequently speculated upon. Studies of the urine

proteome in the obstructed kidney of newborns identified a

number of biomarkers that were able to discriminate clinical

prognosis [11]. The pathogenesis after UUO in the adult is,

however, quite different from the neonate, and equates

better with the generic responses found in tubulointerstitial

fibrosis. This has been highlighted by the recent proteomic

identification of a hitherto unrecognized tissue mediator of

fibroblast recruitment in a rat model of UUO [12]. Our work

therefore extends these studies by examining the adult

kidney, and by specifically examining both plasma and

tissue in an attempt to better understand the relationship

with known pathological changes.

Using a well-described model of experimental obstructive

nephropathy [5–7], we describe the proteomic assessment of

changes that occur at both the tissue and the plasma

level during progression. We combined the proteomic

Table 3. Differentially regulated proteins from 2-D gel analysis of kidney tissue

Control DiseaseSpot IDno.

Protein ID Accessionnumber

Average SEM Average SEM

p-Value Up/down indisease

t1 Unidentified NA 10 425.7 2024.4343 No signal NA 0.001 Downt2 Unidentified NA 5561.1 1226.6559 No signal NA 0.003 Downt3 Unidentified NA 9055.3333 579.91917 No signal NA 0.00002 Downt4 Unidentified NA 6082 2383.8405 No signal NA 0.02 Downt5 Unidentified NA 11 846.1 3235.8772 No signal NA 0.006 Downt6 Unidentified NA 7907.8333 699.47674 No signal NA 0.00009 Downt7 Unidentified NA 5853.9 876.21646 No signal NA 0.0006 Downt8 Unidentified NA 7785 450 No signal NA 0.0003 Downt9 a Enolase ENOA_MOUSE 8247.3 854.21145 No signal NA 0.0001 Downt10 Unidentified NA 9288.9667 1221.8753 No signal NA 0.0004 Downt11 Unidentified NA 6912.3667 1527.0629 No signal NA 0.003 Downt12 Unidentified NA 4911.0667 2032.9963 No signal NA 0.02 Downt13 Unidentified NA 8696.7 1609.6994 No signal NA 0.001 Downt14 Unidentified NA 16 143 6981.1709 No signal NA 0.03 Downt15 Malate dehydrogenase MDHC_MOUSE 8642.5667 4638.8855 No signal NA 0.05 Downt16 Unidentified NA 17 113 3989 No signal NA 0.003 Downt17 Unidentified NA 12 000 180 No signal NA 0.0000001 Downt18 Haptoglobin precursor HPT_MOUSE No signal NA 229 809.9 73 821.61 0.01 Up

The pixel unit values of each protein and p-value calculations of these biomarkers (comparing levels detected in 3-day obstructed versuscontrol samples) are summarized, as are the results from the MS identification studies. All detailed differentially regulated proteins weredetected in all n 5 5 UUO samples analyzed.

Table 4. Differentially regulated proteins from 2-D gel analysis of plasma

Control DiseaseSpot IDno.

Protein ID Accessionnumber

Average SEM Average SEM

p-Value Disease

p1 Fibrinogen a gi/33563252 3008 2223 33 508 11 576 0.02 Upp2 Haptoglobin gi/8850219 2988 2055 12 065 1255 0.005 Upp3 Unidentified NA 4381 4677 40 083 10 229 0.009 Upp4 MUPs 11 and 8 MUP8_MOUSE No signal NA 5581 2460 0.03 Upp5 Haptoglobin Precursor HPT_MUSCR No signal NA 7925 3483 0.03 Up

The pixel unit values of each protein and p-value calculations of these biomarkers (comparing levels detected in 3-day obstructed vscontrol samples) are summarized, as are the results from the MS identification studies. All detailed differentially regulated proteins weredetected in all n 5 5 UUO samples analyzed.

1220 E. Giannakis et al. Proteomics Clin. Appl. 2009, 3, 1211–1224

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

approaches of SELDI-TOF MS, MALDI-TOF/TOF MS and

2-D electrophoresis to search for and identify biomarkers in

animals 3-days after UUO when renal inflammation and

fibrogenesis are established [7]; and confirmed the differ-

ential regulation of one of these biomarkers by immuno-

histochemistry. Although the number of analyzed samples

is not large (five UUO versus four controls), the findings are

acceptable for this exploratory stage. We detected several

differentially regulated proteins at the plasma level of

obstructed animals which included SAA1, fibrinogen a,

haptoglobin precursor protein, haptoglobin and MUPs 11

and 8, whereas aberrantly expressed proteins detected at the

tissue level included ras-like activator protein 2, haptoglobin

precursor, malate dehydrogenase, a enolase and murine

urinary protein. Interestingly, these proteins are largely

separated into four major classes: (i) acute-phase reactants

(ii) cell-signaling molecules (iii) molecules involved in cell

growth and metabolism and (iv) urinary proteins.

SAA, haptoglobin and fibrinogen are all plasma proteins

that play a key role in the acute-phase response following

inflammation. The role of SAA in disease progression

after UUO may be related to its ability to induce matrix

metalloproteinases-1 (collagenase-I) and -3 (stromelysin-I)

[13, 14], which are responsible for the degradation of several

interstitial extracellular matrix proteins including collagens

I, II and III. Although this is the first reported study

demonstrating SAA as a marker for UUO, SAA has been

identified to be up-regulated in a number of related renal

pathologies where fibrosis is a final common pathway.

Examples include renal allograft rejection [15], renal abla-

tion [16] and glomerular basement membrane nephritis [17].

Interestingly, SAA has also been identified as a marker for

several other conditions including ovarian cancer [18],

coronary artery disease [19], chronic obstructive pulmonary

disease [20] and rheumatoid arthritis [21], suggesting that it

may be relatively non-specific as a univariate marker but

may provide the sensitivity and specificity required for

diagnosis when combined in a multimarker panel.

Haptoglobin plays a key role in the acute-phase response

by binding to and clearing toxic-free hemoglobin released

during hemolysis. It is a known marker for liver fibrosis,

believed to play a protective effect in response to organ

injury [22] and is part of the commercially available

FibroTest (Bioproductive, Houilles, France). Recently,

microarray analysis of kidney tissue from an ESRD model

revealed up-regulation of haptoglobin gene expression [23],

suggesting its potential importance in yielding a protective

response during renal disease, a result which we have

confirmed and extended by proteomic assessment in this

study.

Fibrinogen is involved in the coagulation cascade and is

converted to fibrin via thrombin activity to form a clot [24].

Elevated fibrinogen levels have been significantly associated

with cardiac/renal damage in hypertension [25] and severe

liver disease [26] potentially related to the thrombotic action of

the protein. Fibrin deposition has long been known to be an

important part of the pathogenesis of glomerulonephritis

[27]. Although less commonly associated with primary tubu-

lointerstitial disease [28], activation of the coagulation cascade

and deposition of fibrin have specifically been observed in

this model. Importantly, fibrinogen immunostaining was

also found to be up-regulated in kidney tissues.

Ras-like activator protein 2 is a GTPase involved in cell

signaling and communication. Although this is the first

reported study demonstrating up-regulation of ras-like acti-

vator protein 2 in UUO, ras-related molecules have been

shown to be involved in fibrotic disease, consistent with the

findings of this study. Activation of RhoGTPase is a key step

in epithelial mesenchymal transdifferentiation to form

myofibroblasts [29], Rap1b plays a critical role in various

signaling events of diabetic nephropathy [30], whereas

Rho-ROCK plays an important role in the development of

Figure 8. 2-D gel analysis of plasma from UUO and control

groups. Statistically significant differentially regulated spots are

highlighted and numbered which correspond to data summar-

ized in Table 4.

Proteomics Clin. Appl. 2009, 3, 1211–1224 1221

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

tubulointerstitial fibrosis involved in smooth muscle actin

expression, macrophage infiltration and interstitial fibrosis

[31].

Decreased lipid and protein synthesis are common

phenomena in renal failure [32], which may explain the

general decrease in kidney protein expression, which was

observed in this study. Recently, Schanstra et al. [23] utilized

microarray analysis and demonstrated strong down-regula-

tion of genes implicated in cell growth and metabolism in

an ESRD model. Consistent with these findings, we have

demonstrated down-regulation of several proteins that are

involved in growth and metabolism including a enolase and

malate dehydrogenase.

a Enolase is a glycolytic enzyme involved in growth

control. Its role in maintaining healthy tissue may also be

related to its ability to function during the fibrinolytic

pathway and activate collagenase and degrade fibrin

and several extracellular matrix proteins including

laminin and fibronectin. Although a enolase plays an

important role in autoimmune disease, cancer and

bacterial infections [33], this is the first reported study

demonstrating that the remodeling capabilities and cell

growth-associated functions of a enolase are switched off

after UUO.

Malate dehydogenase is involved in oxidative metabo-

lism. Consistent with our findings, it has been shown to be

inhibited in a model of partial bladder obstruction, which is

believed to interfere with metabolic energy production and

tissue contractile function and consequently may explain its

role in UUO [34]. Our study is, however, the first to

demonstrate malate dehydrogenase involvement in the

pathogenesis of obstructed nephropathy.

Schanstra et al., found four isoforms of MUP up-regu-

lated during ESRD, which are believed to be involved in

disruption of protein secretion and transport [23]. Our

results also demonstrated an up-regulation of MUPs 11 and

8. Other related studies have found MUP to be up-regulated

in a 2-wk remnant mouse model and in a model of poly-

cystic kidney disease [35, 36]. Further work will now be

required to determine the function of this up-regulated

MUP in addition to the impact it has, along with the other

significantly altered proteins detected, on the progression of

obstructive nephropathy.

Although we have detected several novel markers that

were differentially regulated in our experimental model

of renal pathology, a number of limitations related to

the proteomic analysis in this study should be noted.

First, the number of visualized spots that is expected on the

2-D gels may vary depending on the amount of protein

loaded per gel, whereas using larger gels may have

yielded better separation of these proteins. Second, the

approach taken was easily able to detect high-abundant

proteins, whereas the identification of low-abundant

proteins was limited. In this respect, depleting high-abun-

dant proteins or using fractionation techniques would

allow for improved detection of lowly expressed proteins.

Third, many of the proteins present on the 2-D gels, parti-

cularly the tissue gels, did not yield sufficient quantities for

MS analysis and remained unidentified. Loading larger

quantities of protein per gel would overcome such limita-

tions. Finally, we examined renal protein expression and

changes in the plasma proteome at a fixed time point,

when renal inflammation and fibrosis were previously

demonstrated in the model used [7]. Assessing the

presence of potential biomarkers at additional time points

would be of importance. Additionally, further work will be

undertaken using complementary approaches including

Western blottings.

Figure 9. Immunohistochemical staining for fibrinogen in kidney tissues of (A) an un-operated control mouse and (B) a mouse subjected

to UUO for 3 days. Aberrant fibrinogen staining is localized to glomerular capillary loops (arrows) and the interstitial space adjacent to

peri-tubular capillaries (arrow heads). Scale bar 5 100 mm.

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

1222 E. Giannakis et al. Proteomics Clin. Appl. 2009, 3, 1211–1224

5 Concluding remarks

In conclusion, we have reported a proteomic assessment of

the UUO model at the tissue and plasma levels. We have

identified several differentially expressed proteins, which

play diverse roles in regulating renal structure and function.

Although a number of these markers have altered expres-

sion in other disease states, when combined in a multi-

marker panel these biomarkers may facilitate the

development of a highly discriminatory assay for both

obstructive nephropathy and the progression of renal

disease in general. Such an approach offers vastly improved

sensitivity and specificity over a single marker detection

system and is likely to be used for future diagnostics/prog-

nostics. Finally, the biomarkers identified in this study may

provide directions for the development of potential ther-

apeutic targets against diseases associated with fibrosis.

This work was supported by an Ian Potter Grant to G.W.T.and J.D.W., a National Health & Medical Research Council(NHMRC) of Australia R. D. Wright/National Heart Foun-dation of Australia (NHFA) Career Development Fellowship toC.S.S. and NHMRC Fellowships to A.I.S., G.W.T. and J.D.W.

The authors have declared no conflict of interest.

6 References

[1] Provoost, A. P., Molenaar, J. C., Renal function during and

after a temporary complete unilateral ureter obstruction in

rats. Invest. Urol. 1981, 18, 242–246.

[2] Chevalier, R. L., Obstructive nephropathy: towards biomar-

ker discovery and gene therapy. Nat. Clin. Pract. Nephrol.

2006, 2, 157–168.

[3] Chevalier, R. L., Peters, C. A., Congenital urinary tract

obstruction: Proceedings of The State-of-The-Art Strategic

Planning Workshop. National Institutes of Health, Bethesda,

MD, USA, 11–12, March 2002. Pediatr. Nephrol. 2003, 18,

576–606.

[4] Lemley, K. V., An introduction to biomarkers: applications

to chronic kidney disease. Pediatr. Nephrol. 2007, 22,

1849–1859.

[5] Klahr, S., Morrissey, J., Obstructive nephropathy and renal

fibrosis. Am. J. Physiol. Renal. Physiol. 2002, 283,

F861–F875.

[6] Diamond, J. R., Ricardo, S. D., Klahr, S., Mechanisms of

interstitial fibrosis in obstructive nephropathy. Semin.

Nephrol. 1998, 18, 594–602.

[7] Hewitson, T. D., Mookerjee, I., Masterson, R., Zhao, C. et al.,

Endogenous relaxin is a naturally occurring modulator of

experimental renal tubulointerstitial fibrosis. Endocrinology

2007, 148, 660–669.

[8] Becker, G. J., Hewitson, T. D., The role of tubulointerstitial

injury in chronic renal failure. Curr. Opin. Nephrol. Hyper-

tens. 2000, 9, 133–138.

[9] Giannakis, E., Samuel, C. S., Boon, W. M., Macris, M. et al.,

SELDI-TOF mass spectrometry-based protein profiling of

kidney tissue. Methods Mol. Biol. 2009, 466, 237–249.

[10] Buzza, M. S., Zamurs, L., Sun, J., Bird, C. H. et al., Extra-

cellular matrix remodeling by human granzyme B via clea-

vage of vitronectin, fibronectin, and laminin. J. Biol. Chem.

2005, 280, 23549–23558.

[11] Decramer, S., Wittke, S., Mischak, H., Zurbig, P. et al.,

Predicting the clinical outcome of congenital unilateral

ureteropelvic junction obstruction in newborn by urinary

proteome analysis. Nat. Med. 2006, 12, 398–400.

[12] Kypreou, K. P., Kavvadas, P., Karamessinis, P., Peroulis, M.

et al., Altered expression of calreticulin during the devel-

opment of fibrosis. Proteomics 2008, 8, 2407–2419.

[13] O’Hara, R., Murphy, E. P., Whitehead, A. S., FitzGerald, O.

et al., Local expression of the serum amyloid A and formyl

peptide receptor-like 1 genes in synovial tissue is asso-

ciated with matrix metalloproteinase production in patients

with inflammatory arthritis. Arthritis Rheum. 2004, 50,

1788–1799.

[14] Mullan, R. H., Bresnihan, B., Golden-Mason, L., Markham, T.

et al., Acute-phase serum amyloid A stimulation of angio-

genesis, leukocyte recruitment, and matrix degradation in

rheumatoid arthritis through an NF-kappaB-dependent

signal transduction pathway. Arthritis Rheum. 2006, 54,

105–114.

[15] Hartmann, A., Eide, T. C., Fauchald, P., Bentdal, O. et al.,

Serum amyloid A protein is a clinically useful indicator of

acute renal allograft rejection. Nephrol. Dial. Transplant.

1997, 12, 161–166.

[16] Denk, G. U., Cai, S. Y., Chen, W. S., Lin, A. et al., A

comparison of gene expression in mouse liver and kidney

in obstructive cholestasis utilizing high-density oligonu-

cleotide microarray technology. World J. Gastroenterol.

2006, 12, 2536–2548.

[17] Kim, J. H., Ha, I. S., Hwang, C. I., Lee, Y. J. et al., Gene

expression profiling of anti-GBM glomerulonephritis

model: the role of NF-kappaB in immune complex kidney

disease. Kidney Int. 2004, 66, 1826–1837.

[18] Moshkovskii, S. A., Serebryakova, M. V., Kuteykin-Teplya-

kov, K. B., Tikhonova, O. V. et al., Ovarian cancer marker of

11.7 kDa detected by proteomics is a serum amyloid A1.

Proteomics 2005, 5, 3790–3797.

[19] Johnson, B. D., Kip, K. E., Marroquin, O. C., Ridker, P. M.

et al., Serum amyloid A as a predictor of coronary artery

disease and cardiovascular outcome in women: the

National Heart, Lung, and Blood Institute-Sponsored

Women’s Ischemia Syndrome Evaluation (WISE). Circula-

tion 2004, 109, 726–732.

[20] Bozinovski, S., Hutchinson, A., Thompson, M., Macgregor, L.

et al., Serum amyloid A is a biomarker of acute exacerba-

tions of COPD. Am. J. Respir. Crit. Care Med. 2007, 177,

269–278.

[21] O’Hara, R., Murphy, E. P., Whitehead, A. S., FitzGerald, O.

et al., Acute-phase serum amyloid A production by rheu-

matoid arthritis synovial tissue. Arthritis Res. 2000, 2,

142–144.

Proteomics Clin. Appl. 2009, 3, 1211–1224 1223

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com

[22] Tolosano, E., Fagoonee, S., Hirsch, E., Berger, F. G. et al.,

Enhanced splenomegaly and severe liver inflammation in

haptoglobin/hemopexin double-null mice after acute

hemolysis. Blood 2002, 100, 4201–4208.

[23] Schanstra, J. P., Bachvarova, M., Neau, E., Bascands, J. L.

et al., Gene expression profiling in the remnant kidney

model of wild type and kinin B1 and B2 receptor knockout

mice. Kidney Int. 2007, 72, 442–454.

[24] Hewitson, T. D., Martic, M., Kelynack, K. J., Pagel, C. N.

et al., Thrombin is a pro-fibrotic factor for rat renal fibro-

blasts in vitro. Nephron Exp. Nephrol. 2005, 101, e42–e49.

[25] Sechi, L. A., Zingaro, L., Catena, C., De Marchi, S., Increased

fibrinogen levels and hemostatic abnormalities in patients

with arteriolar nephrosclerosis: association with cardio-

vascular events. Thromb. Haemost. 2000, 84, 565–570.

[26] Arif, S., Khan, A. S., Khan, A. R., Changes in fibrinogen level in

liver cirrhosis. J. Ayub. Med. Coll. Abbottabad. 2002, 14, 19–21.

[27] Clarkson, A. R., Seymour, A. E., Jackson, B., Coagulation

and renal disease. Med. J. Aust. 1976, 1, 573–578.

[28] Wendt, T., Zhang, Y. M., Bierhaus, A., Kriegsmann, J . et al.,

Tissue factor expression in an animal model of hydrone-

phrosis. Nephrol. Dial. Transplant. 1995, 10, 1820–1828.

[29] Patel, S., Takagi, K. I., Suzuki, J., Imaizumi, A. et al.,

RhoGTPase activation is a key step in renal epithelial

mesenchymal transdifferentiation. J. Am. Soc. Nephrol.

2005, 16, 1977–1984.

[30] Lin, S., Chugh, S., Pan, X., Wallner, et al., Identification of

up-regulated Ras-like GTPase, Rap1b, by suppression

subtractive hybridization. Kidney Int. 2001, 60, 2129–2141.

[31] Nagatoya, K., Moriyama, T., Kawada, N., Takeji, M. et al., Y-

27632 prevents tubulointerstitial fibrosis in mouse kidneys

with unilateral ureteral obstruction. Kidney Int. 2002, 61,

1684–1695.

[32] Kalantar-Zadeh, K., Mehrotra, R., Fouque, D., Kopple, J. D.,

Metabolic acidosis and malnutrition-inflammation complex

syndrome in chronic renal failure. Semin. Dial. 2004, 17,

455–465.

[33] Liu, K.-J., Shih, N.-Y., The role of enolase in tissue invasion

and metastasis of pathogens and tumour cells. J. Canc.

Mol. 2007, 3, 45–48.

[34] Haugaard, N., Potter, L., Wein, A. J., Levin, R. M., Effect of

partial obstruction of the rabbit urinary bladder on malate

dehydrogenase and citrate synthase activity. J. Urol. 1992,

147, 1391–1393.

[35] Zhang, H., Wada, J., Kanwar, Y. S., Tsuchiyama, Y. et al.,

Screening for genes up-regulated in 5/6 nephrectomized

mouse kidney. Kidney Int. 1999, 56, 549–558.

[36] Valkova, N., Yunis, R., Mak, S. K., Kang, K. et al., Nek8

mutation causes overexpression of galectin-1, sorcin,

and vimentin and accumulation of the major urinary

protein in renal cysts of jck mice. Mol. Cell. Proteomics

2005, 4, 1009–1018.

1224 E. Giannakis et al. Proteomics Clin. Appl. 2009, 3, 1211–1224

& 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.clinical.proteomics-journal.com