Analysis of Disease-Associated Protein Expression Using Quantitative

Subscriber access provided by NEW YORK MED COLL

Article

Analysis of disease-associated protein expression using quantitative proteomics – Fibulin-5 is expressed in association with hepatic fibrosis Thilo Bracht, Vincent Schweinsberg, Martin Trippler, Michael Kohl, Maike Ahrens, Juliet Padden, Wael Naboulsi, Katalin Barkovits, Dominik A. Megger, Martin Eisenacher, Christoph H. Borchers, Jörg-Friedrich Schlaak, Andreas-Claudius Hoffmann, Frank Weber, Hideo Andreas Baba, Helmut Erich Meyer, and Barbara Sitek J. Proteome Res., Just Accepted Manuscript • Publication Date (Web): 25 Mar 2015 Downloaded from http://pubs.acs.org on March 27, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Proteome Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Analysis of disease-associated protein expression using quantitative proteomics – Fibulin-5 is expressed in association with hepatic fibrosis

Thilo Bracht§, Vincent Schweinsberg§, Martin Trippler$, Michael Kohl§, Maike Ahrens§, Juliet Padden§, Wael Naboulsi§, Katalin Barkovits§, Dominik A. Megger§, Martin Eisenacher§, Christoph H. Borchers&, Jörg F. Schlaak$, Andreas-Claudius Hoffmannǂ, Frank Weber∆, Hideo A. Baba≠, Helmut E. Meyer§,# and Barbara Sitek§

§ $ &

Medizinisches Proteom-Center, Ruhr-Universität Bochum, Germany

Department of Gastroenterology and Hepatology, University Hospital of Essen, Germany

University of Victoria - Genome British Columbia Proteomics Centre, Victoria BC, Canada

ǂDepartment of Medicine (Cancer Research), Molecular Oncology Risk-Profile Evaluation, University Hospital of Essen, Germany ∆

Department of General, Visceral and Transplantation Surgery, University Hospital of Essen, Germany ≠

Department of Pathology, University Hospital of Essen, Germany

#

Leibniz Institute for Analytical Sciences - ISAS, Dortmund, Germany

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract Hepatic fibrosis and cirrhosis are major health problems worldwide. Up to now, highly invasive biopsy remains the diagnostic gold standard despite many disadvantages. In order to develop non-invasive diagnostic assays for the assessment of liver fibrosis it is urgently necessary to identify molecules which are robustly expressed in association with the disease. We analyzed biopsied tissue samples from 95 patients with HBV/HCV-associated hepatic fibrosis using three different quantification methods. We performed a label-free proteomics discovery study to identify novel disease-associated proteins using a subset of the cohort (n = 27). Subsequently, gene expression data from all available clinical samples were analyzed (n= 77). Finally, we performed a targeted proteomics approach (MRM) to verify the diseaseassociated expression in samples independent from the discovery approach (n = 68). We identified Fibulin-5 (FBLN5) as a novel protein expressed in relation to hepatic fibrosis. Furthermore, we confirmed the altered expression of microfibril-associated glycoprotein 4 (MFAP4), lumican (LUM) and collagen alpha-1(XIV) chain (COL14A1) in association to hepatic fibrosis. To our knowledge, no tissue-based quantitative proteomics study for hepatic fibrosis has been performed using a cohort of comparable size. By this means, we add substantial evidence for the disease-related expression of the proteins examined in this study.

Keywords: Liver, Fibrosis, Hepatic Fibrosis, Cirrhosis, Fibulin-5, Label-free Proteomics, Multiple Reaction Monitoring, Selected Reaction Monitoring, MFAP4, Lumican


Page 2 of 40

Page 3 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Abstract Graphic:



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Introduction Fibrosis and cirrhosis of the liver are the common final path of a wide range of different liver diseases, which cause chronic liver injury and thereby provoke the formation of scar-like tissue inside the organ. This scarring may then progress towards fibrosis and ultimately cirrhosis of the liver. Liver cirrhosis is characterized by the destruction of the normal liver-vasculature, which leads to portal hypertension 1. The most frequent underlying diseases that cause liver fibrosis and cirrhosis are alcoholic and non-alcoholic fatty liver disease as well as hepatitis B and C. Fibrosis and cirrhosis of the liver are associated with high prevalence and mortality worldwide. It has been assumed that about 1% of the world population suffers from cirrhosis. More accurate data are hard to assess because an unknown proportion of affected but asymptomatic persons are left undiagnosed 1,2. Despite its alarming epidemiology, diagnosis and long-term monitoring of liver fibrosis and cirrhosis remain difficult, because early disease stages display no or only few and unspecific clinical symptoms. Invasive, percutaneous liver biopsy still represents the diagnostic gold standard. Histological examination of tissue samples allows the evaluation of the amount of fibrotic tissue in the liver and sequential biopsies permit monitoring of disease progression or regression. However, a considerable proportion of patients have to be hospitalized after liver biopsy due to complications. Up to 40% of patients suffer from massive pain during the procedure and death by bleeding occurs in 1 of 10000 patients 3. The invasive and risk-bearing nature of liver biopsy may increase the psychological threshold for the attending physician in deciding on the realization of this procedure in clinical practice. This could carry the hazard of delayed diagnosis, as liver biopsy, in many cases, might be performed only after major complications already have become apparent in the course of advanced disease. An early diagnosis of liver fibrosis is important because of its therapeutic consequences. In many cases, progression can be stopped if the noxious agent is eliminated and recent studies have shown that regression of fibrosis is possible 4. 4 ACS Paragon Plus Environment

Page 4 of 40

Page 5 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


A number of non-invasive tests for the assessment of liver fibrosis have been described in the last years 5. Some of these tests measure proteins which are directly related to the pathophysiological process of fibrosis, like Metalloproteinase inhibitor 1 (TIMP1; e.g. FibroSpect II 6, ELF panel 7). Other tests combine several indirect parameters in one algorithm to score fibrosis stages (e.g. FibroTest 8, Hepascore 9, APRI 10). However, most of these tests were found to be insufficiently validated in independent patient cohorts and none provides sufficient resolution to differentiate between intermediate fibrosis stages 11. The need to develop new and more effective non-invasive tests is therefore still obvious. Proteomics technologies facilitate the discovery of novel disease-associated proteins in an unbiased and non-hypothesis driven way and various proteomics studies have been performed to identify proteins expressed in relation to liver-fibrosis in the past (reviewed in 12). Numerous of such proteins (usually termed biomarker candidates) have been reported but most of these proteins have not been followed up and lack verification in independent cohorts. Many of the reported approaches facilitate gel-based approaches, which nowadays are usually outperformed by state-of–the-art LC-MS-based technologies regarding sample throughput and proteome coverage. In many cases serum or plasma samples were analyzed which exhibits some major drawbacks, since usually only a few high abundant proteins can be investigated in an unbiased manner. On the other hand, the discovery of a disease-associated protein using tissue samples and the subsequent transfer to a verification using blood samples have been performed successfully 13. In order to develop new biomarkers for liver fibrosis it is urgently necessary to have a profound knowledge of molecules which are robustly expressed in association with the disease. The aim of the presented study was to identify new proteins which are expressed in association with hepatic fibrosis and re-evaluate already described ones by a combination of label-free proteome analysis, gene expression analysis and targeted protein quantification using multiple reaction monitoring (MRM). We examined liver biopsies from 27 patients with 5 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

hepatitis C background suffering from different stages of liver fibrosis using an LC-MS– based label-free proteomics approach. We found 70 proteins to be differentially expressed between patients with low fibrosis stages (F0 – F2) and high fibrosis stages (F3 and F4, according to the Batts-Ludwig classification). Among 60 proteins which were found to have higher expression levels in high fibrosis stages we selected 7 proteins which were already described in the context of hepatic fibrosis or remodeling of extra cellular matrix (ECM). We analyzed gene expression data from 77 patients to evaluate the disease-associated expression of these proteins. Finally, we set up a multiple reaction monitoring assay and quantified 7 proteins in an independent cohort of 68 patients with HBV and HCV-related hepatic fibrosis. Altogether, we analyzed biopsied tissue samples from 95 patients suffering from hepatic fibrosis.


Page 6 of 40

Page 7 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Materials and Methods

Experimental Design A differential label-fee proteomics approach was performed using 27 biopsies from patients with HCV-associated hepatic fibrosis. For statistical analysis, the patients were grouped into a low and a high fibrosis group matched for age and BMI. The low fibrosis group contained 13 patients of fibrosis stages 0, 1 and 2 and the high fibrosis group contained 14 patients of fibrosis stages 3 and 4 (Supplementary Table 1). We selected 7 candidate proteins and examined the expression levels of the genes, coding these proteins in 77 samples (including 25 samples which were also analyzed in the label-free study, Table 1, Figure 1). Finally, we set up an MRM assay and performed relative quantification of these proteins in 68 samples independent from the discovery study (Supplementary Table 2).

Acquisition of Liver Specimens Samples were obtained via sonography-guided percutaneous liver biopsy at the Department of Gastroenterology, Essen University Hospital, Germany (permission from local ethics committee 11-4839-BO). All patients underwent invasive diagnosis for liver fibrosis and a part of the obtained biopsy was used for this study. The procedure and the use of a part of the biopsy for research purposes were explained to the patients, which allowed acquisition of informed consent. After sampling the biopsies were immediately divided for pathological assessment, gene expression analysis and proteome analysis. For proteome analysis the samples were snap-frozen and stored at -80°C. The part of the biopsy which was designated for gene expression analysis was stored at -80°C in RNAlater (Qiagen, Hilden, Germany). The biopsies were assessed pathologically according to Batts-Ludwig classification. For the verification experiments using gene expression analysis and MRM, patients with HBVassociated hepatic fibrosis were considered as well (Supplementary Table 1 and 2). 7 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Sample Preparation and Tryptic Digestion Biopsied specimens were lysed and homogenized in 40 µL of sample buffer (30 mM Tris, 7 M Urea, 2 M Thiourea, 0.1% SDS, pH 8.5) and the protein concentration was determined using the Bradford assay (Bio-Rad, Hercules, CA, U.S.A.). 15 µg of protein per sample were loaded on an 18% Tris-glycine polyacrylamide gel (Anamed Elektrophorese, Groß-Bieberau, Germany) and allowed to run slightly into the gels (15 min, 100 V). The protein bands were stained with Coomassie and cut from the gels. Digestion of proteins was performed overnight at 37 °C with trypsin (Serva Electrophoresis, Heidelberg, Germany). The peptides were extracted from the bands with 20 µL of 50% ACN in 0.1% TFA (1:1), vacuum dried and subsequently dissolved in 0.1% TFA. The peptide concentration was determined by amino acid analysis as previously described

14

. For MRM experiments, the samples were

processed identically with the only difference that 50% ACN in 0.1% FA was used for peptide extraction and finally the peptides were dissolved in 0.1% FA.

Label-free LC-MS/MS Analysis The LC-MS analysis was performed using an Ultimate 3000 RSLCnano system coupled online to an Orbitrap Elite mass spectrometer (both Thermo Scientific, Bremen, Germany). 300 ng of peptides were injected and pre-concentrated on a trap column (Acclaim PepMap 100, 300 µm × 5 mm, C18, 5 µm, 100 Å; flow rate 30 µL/min). Subsequently, the peptides were separated on the analytical column (Acclaim PepMap RSLC, 75 µm × 50 cm, nano Viper, C18, 2 µm, 100 Å) by a gradient from 5% to 40% solvent B over 98 min (solvent A: 0.1% FA, solvent B: 0.1% FA, 84% ACN; flow rate 400 nL/min; column oven temperature 60 °C). The instrument was operated in a data-dependent mode. Full scan mass spectra were acquired in the Orbitrap analyzer and the 20 most abundant precursors were


Page 8 of 40

Page 9 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


selected for MS/MS analysis. The tandem mass spectra were measured in the linear ion trap after peptide fragmentation by collision-induced dissociation.

Peptide Identification and Quantification The peptide identification was performed using Proteome Discoverer Software (ver. 1.4.0.288, Thermo Fisher Scientific, Rockford, IL, U.S.A.). The mass spectra were searched against UniProtKB/Swiss-Prot database (2013_05; 540.052 entries) restricted to homo sapiens using Mascot (ver. 2.3.0.2). The mass tolerance was set to 5 ppm for precursor ions and 0.4 Da for fragment ions. One tryptic miscleavage was considered as well as chemical modification of methionine (oxidation) and cysteine (propionamide). The percolator function implemented in Proteome Discoverer was used to estimate peptide confidence and only peptides which passed a false discovery rate < 1% were considered for analysis. The peptide quantification was performed using Progenesis LC-MS software (ver. 4.1.4832.42146, Nonlinear Dynamics Ltd., Newcastle upon Tyne, U.K.). All runs were aligned to a reference run automatically chosen by the software and a master list of features considering m/z values and retention times was generated. The peptide identifications were exported from Proteome Discoverer and imported in Progenesis LC-MS software where they were matched to the respective features. Proteins quantified with at least two peptides, an ANOVA p-value ≤ 0.05 and an absolute fold change ≥ 1.5 were considered to be significantly differentially expressed and used for further evaluation. The data presented in this manuscript is available via ProteomeXchange with the identifier PXD001474.

Gene Expression Analysis Total RNA was isolated from liver biopsies of all 95 HBV and HCV patients and analyzed as previously described

15

. Briefly, 77 RNA samples passed RNA quality controls

and were preprocessed with the 3’IVT Express Kit (Affymetrix, Santa Clara, CA, U.S.A.). 9 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Samples where then hybridized on Human Genome U219 16-Array Plates using the GeneTitan MC Instrument (Affymetrix) according to the manufacturer’s instructions. Robust multi-array average (RMA) normalization was carried out with the Expression Console Software (version 1.2.1, Affymetrix).

Selection and Synthesis of SIS Peptides Proteotypic tryptic peptides were selected considering the UniProtKB/Swiss-Prot data base restricted to homo sapiens as background proteome. Known post translational modifications or naturally varying amino acids were exclusion criteria as well as internal methionine, cysteine, lysine or arginine residues. All stable isotope labeled synthetic (SIS) peptides were labeled C-terminally with stable isotope-labeled amino acids (either [13C6, 15N2]-lysine or [13C6, 15N4]-arginine). One peptide was labeled with [13C9, 15N]phenylalanine since it was a C-terminal peptide and did not contain any K/R residues. SIS peptides were purchased from INTAVIS Bioanalytical Instruments (Cologne, Germany). Briefly, peptides were synthesized by means of standard solid phase peptide synthesis and purified using preparative RP-HPLC. Purity was controlled by analytical RP-HPLC and identity was controlled via MALDI-TOF MS. Peptide concentration was determined via quantitative amino acid analysis. Three peptides (representing MFAP4) were synthesized in Genome British Columbia Proteomics Center (Victoria B.C., Canada) as previously described 16

.

Triple Quadrupole LC-MS MRM experiments were conducted using an Agilent 6490 triple quadrupole MS coupled to an Agilent 1290 Infinity Binary HPLC standard-flow system (both Agilent Technologies, Santa Clara, CA, U.S.A.). A sample volume of 20 µL was injected and separated with a flow rate of 0.4 mL/min on an analytical C18 column (ZORBAX Eclipse 10 ACS Paragon Plus Environment

Page 10 of 40

Page 11 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Plus Rapid Resolution HD, 2.1x150 mm, 1.8 µm, column temperature 50 °C). Peptides were separated using a multi-step gradient from 97% solvent A (0.1% FA) and 3% solvent B (84% ACN, 0.1% FA) to 90% solvent B (time: % B): 0 min: 3% B; 2 min: 11% B; 15 min: 19% B; 20 min: 29% B; 22 min: 39% B; 25 min: 45% B; 27 min: 90% B; 29 min: 90% B; 30 min: 3% B. Peptides were ionized by a standard-flow ESI source (Agilent Jet Stream, for detailed parameters see Supplementary Material). The QQQ-MS was operated using the Mass Hunter workstation (ver. B.06.00 Service Pack 1).

Multiple Reaction Monitoring In order to setup the MRM assay, the optimal collision energies for each peptide were experimentally defined using Agilent automation implemented in Skyline (Skyline ver. 2.5.0.6157, MacCoss Lab, University of Washington, Seattle, U.S.A.). For each peptide, the three peptide/product ion pairs (transitions) resulting in the highest signal intensities were chosen and peptide retention times were identified (Supplementary Table 4). Calibration curves for SIS peptides were calculated to determine the limit of quantitation and the linearity of the assay (Supplementary Figure 1). The used approach considers two criteria: first, the coefficients of variation (CV) of the measured areas under the curves (AUC). Second, the average accuracy values (AAV) of each concentration level (for further details see supplementary material). For MRM experiments, the tissue samples were processed as described and subsequently SIS peptides were spiked into 1 µg of tryptic peptides resulting in final concentrations of SIS peptides as shown in Figure 3. MRM data were visualized and analyzed using Skyline. The most signal-intensive transition was used for quantification (quantifier), whereas the two additional transitions were used to ensure correct identity of the analyzed peptide (qualifiers). The ratios between the quantifiers of natural peptides and SIS peptides were used for relative quantification. The data presented in the manuscript has been deposited at PASSEL and is accessible via the identifier PASS00653. 11 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Statistical Analysis The label-free discovery study was analyzed by means of unpaired ANOVA implemented in the Progenesis LC-MS software. The false discovery rate (FDR)-correction implemented in the software was used (q-value). For the analysis of gene expression data, we only considered those features corresponding to the seven genes of interest. We conducted one-way ANOVA for each feature comparing the 4 groups of fibrosis stages. Afterwards we corrected the resulting p-values for multiple testing (controlling the FDR using the method of Benjamini and Hochberg

17

). For features with an adjusted p-value below 0.05 a Tukey's

‘Honest Significant Difference’ post-hoc test was conducted. The MRM data were tested for statistical significance using the Kruskal-Wallis test. The resulting p-values were adjusted to control the FDR. The peptides with an adjusted p-value smaller than 0.05 were further analyzed using FDR-adjusted Mann-Whitney-U test for pairwise comparisons between groups. Computations regarding adjustment and post-hoc tests have been conducted using R (Core Team (2014). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-project.org/.).


Page 12 of 40

Page 13 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Results

Quantitative Label-free Proteomics Analysis With the aim of identifying new proteins expressed in association with liver fibrosis we performed a label-free LC-MS/MS study. 3228 protein groups (according to protein grouping algorithm implemented in Proteome Discoverer) were identified and 19180 peptides representing 2381 proteins without conflicting peptides were quantified by Progenesis LC-MS software. Among these 1715 proteins were quantified with minimum 2 peptides (Figure 2). 70 proteins were found to be significantly differentially expressed between low and high fibrosis groups (Supplementary Table 3). 60 of these proteins showed higher expression levels in high fibrosis stages whereas 10 showed lower expression in this experimental group. Overall the number of significantly differentially expressed proteins was comparably low and none of the proteins passed the FDR-correction (q-value ≤ 0.05). Only 24 proteins showed a significant ANOVA p-value and an absolute fold change greater than 2 (23 of these with higher abundance in the high fibrosis group). Among these proteins we chose seven candidates, associated with either extracellular matrix organization or activation of hepatic stellate cells, for further verification of disease-associated expression: Lumican (LUM), fibulin 5 (FBLN5), cysteine and glycine-rich protein 2 (CSRP2), calponin 2 (CNN2), transgelin (TAGLN), collagen alpha-1(XIV) chain (COL14A1) and microfibril-associated glycoprotein 4 (MFAP4) (Table 2).

Gene Expression Analysis and Verification on Transcript Level In order to evaluate the expression of the selected proteins in association with hepatic fibrosis we used a second quantification strategy by analyzing gene expression data generated on an Affymetrix microarray platform. Thereby we investigated a level of evidence complimentary to the protein level (e.g. transcript level). We analyzed a total of 77 samples 13 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

representing the same patients that were analyzed in the proteomics experiments (18 of all 95 samples did not pass RNA quality controls, Table 1, Figure 1). Four different groups, representing fibrosis stages 0 (n =2) together with stage 1 (n = 23), stage 2 (n =22), stage 3 (n = 15) and stage 4 (n = 15), were analyzed. We found significant elevation and relation of gene expression levels with increasing fibrosis stages for LUM, FBLN5, TAGLN, COL14A1 and

MFAP4 (Figure 3). For LUM, FBLN5, TAGLN and COL14A1, we observed high levels of significance for differences between fibrosis stages 0/1 and stage 4, respectively (Supplementary Table 5). LUM and TAGLN also showed high significance between stage 2 and stage 4. For CSRP2 results were less informative since high variance in the individual groups was found. CNN2 transcript levels were at noise level and could not be analyzed.

Verification of disease-associated expression by MRM In order to verify the disease-associated expression of the selected candidate proteins we analyzed an independent cohort of 68 liver biopsies from patients with HCV and HBVrelated hepatic fibrosis (fibrosis stages 1-4). For relative quantification, the ratios between internal standard (SIS peptides) and the endogenous peptides were used and the acquired data was tested for statistical significance (Supplementary Table 6). CSRP2, CNN2 and TAGLN did not show a consistent relation of expression levels with increasing fibrosis stages (Supplementary Figure 2). TAGLN showed a slight increase of measured peptide levels from stage 1 to stage 4. However, although this was partially also statistically significant, the relation of TAGLN abundance to hepatic fibrosis does not seem to be robust. CSRP2 and CNN2 both showed significant increase of expression levels between stage 1 and 2 (Supplementary Figure 2). Although no or only slight additional increase of expression levels was observed from stage 2 to stage 4, a relevance of CSRP2 and CNN2 in early fibrosis might be considered. Mean expression levels of LUM, COL14A1 and MFAP4 were found to increase with increasing fibrosis stages (Figure 4). Additionally, FBLN5, which was not 14 ACS Paragon Plus Environment

Page 14 of 40

Page 15 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


described in the context of liver fibrosis before, was found with elevated expression levels corresponding to increasing fibrosis stages (Figure 4). Significant differences were observed for all four proteins between stages 1 and 3 and 1 and 4 as well as between stages 2 and 4. LUM, MFAP4 and FBLN5 also showed significantly differential expression levels between stages 1 and 2.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Discussion Hepatic fibrosis still represents one of the major health problems worldwide. In African and Asian countries chronic HBV and HCV infections are still a major burden, while non-alcoholic fatty liver diseases and especially NASH are becoming a major risk factor in developed western countries. As biopsy still remains the gold standard for diagnosis of liver fibrosis, the need for non-invasive biomarkers is obvious. A prerequisite for the development of such biomarkers is the knowledge of disease-associated molecules (e.g. proteins). Here, we used a label-free discovery approach to analyze a cohort of 27 clinical samples to identify new proteins associated with hepatic fibrosis. We were able to identify 70 proteins which were found to be significantly differentially expressed between fibrosis stages 0 - 2 and stages 3 and 4 according to the applied filter criteria. Out of these proteins we selected 7 proteins whose disease-associated expression was further verified in an independent cohort of 68 patients. We selected the proteins according to a known relation to hepatic fibrosis or ECM formation. During liver fibrosis, hepatic stellate cells (HSCs) are being activated and differentiate to myofibroblasts which are responsible for deposition of ECM and show several characteristics of smooth muscle cells (e.g. αSMA). Cysteine and glycine-rich protein 2 (CSRP2) has already been described as a marker for activation and differentiation of HSCs. In the liver CSRP2 is solely expressed in HSCs and its expression has been shown to be elevated during early HSC activation. Expression can be induced by TGF-β which plays a fundamental role in the induction of hepatic fibrogenesis 18. However, in vitro and in vivo the expression of CSRP2 was shown to decrease with differentiation of HSCs and therefore CSRP2 might serve as a marker for early fibrosis stages. Our data supports this hypothesis to a certain extent, since we observed a highly significant increase of CSRP2 levels between fibrosis stages 1 and 2 by MRM (Supplementary Figure 2, Supplementary Table 6).


Page 16 of 40

Page 17 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Transgelin (TAGLN; also known as SM22) and calponin-2 (CNN2; also known as h2calponin) are structurally closely related proteins associated with the actin cytoskeleton. TAGLN is expressed mainly in smooth muscle cells but also in fibroblasts and its expression can be induced by TGF-β

19

. TAGLN is a marker of smooth muscle differentiation and has

been demonstrated (together with CSRP2) to be expressed during activation and differentiation of HSCs in vitro 20. TAGLN has also been reported to be expressed in relation to hepatic fibrosis

13

. Therefore, we decided to re-inspect its expression in a larger cohort.

Although it was found to be up-regulated in high fibrosis stages in the discovery study and with a consistent up-regulation on transcript level we were not able to verify disease-related expression of TAGLN by MRM. Hence, the contradiction of transcript and protein levels underlines the necessity of a verification approach on protein level. Calponins are actin-regulatory proteins of which three isoforms are known (CNN1, CNN2 and CNN3). While CNN1 and CNN3 are mainly expressed in smooth muscle cells, CNN2 is present in various tissues and cell types, amongst them also smooth muscle cells and fibroblasts

21

. The expression of calponins in the liver is likely to reflect the activation and

differentiation of HSCs into myofibroblasts as it has been demonstrated for the expression of CNN1 in HSCs stimulated with TGF-β

22

potential biomarker for hepatic fibrosis

. As CNN1 has previously been described as a

13

, we included CNN2 as a candidate in our

verification study. Our data does only partially support a diagnostic capability of CNN2 in hepatic fibrosis. Its expression was found to be significantly up-regulated between fibrosis stages 1 and 2, but was not found to be further elevated with increasing fibrosis stages. Collagen is a major component of the extracellular matrix and it is well known that collagen is deposited during tissue remodeling associated with hepatic fibrosis

23

. Collagen

alpha-1(XIV) chain (COL14A1; also named undulin) is expressed in the healthy liver, mainly located in portal tracts and around central veins. Increased expression of COL14A1 in hepatic fibrosis was observed by immunostaining and in-situ hybridization provided evidence that it 17 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

is also expressed by myofibroblasts

24

. Our data confirms that COL14A1 is increasingly

abundant with advancing fibrosis stages both on transcript and protein level. The proteoglycan lumican (LUM) is another component of the ECM and functions in assembly of collagen fibrils. The relevance of LUM in hepatic fibrosis is well established, as already in the early nineties its expression has been found to be up-regulated in fibrotic rat livers

25

. LUM has also been shown to be expressed in the livers of patients suffering of

NAFLD and NASH

26

. LUM-deficient mice have been shown to be protected against

chemically induced hepatic fibrosis indicating a substantial function of LUM in matrix deposition and fibre formation during hepatic fibrosis 27. Furthermore, elevated expression of LUM was demonstrated for hepatitis B and C 27,28. Our data supports the importance of LUM in hepatic fibrosis and shows that it represents a robust indicator for fibrosis in the liver. Microfibril-associated glycoprotein 4 (MFAP4) is an extracellular matrix protein which is located in elastic fibers in the entire body

29

. The biological function of MFAP4 is

not fully clarified. However, it binds to collagen and elastin and has been described as a biomarker candidate for various conditions including COPD, cardiovascular diseases and liver fibrosis 13,29,30. In liver fibrosis MFAP4 serum levels could be quantified by ELISA and were shown to be elevated in patients with fibrosis, therefore marking MFAP4 as a promising candidate for non-invasive assessment of liver fibrosis 13. Our data comprehensively confirms the association of MFAP4 expression to liver fibrosis and therefore supports its diagnostic potential. FBLN5 (Fibulin-5; also known as DANCE) is an extracellular matrix protein which has been shown to play a fundamental role in elastic fiber assembly. Knockout of FBLN5 leads to a disorganized elastic fiber system in the whole body, which results in a phenotype resembling many features of the human ECM disorder cutis laxa 31. Mutations of the FBLN5 gene were shown to be qualified to cause this genetic disorder 32. FBLN5 is expressed during embryonic development, mainly in the vascular and pulmonary system. After being down18 ACS Paragon Plus Environment

Page 18 of 40

Page 19 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


regulated in adult tissues, its expression is reactivated after injury for example in blood vessels or during angiogenesis in the uterus

33

. FBLN5 is expressed by several cell types,

including fibroblasts and smooth muscle cells, thereby showing a likely relation of its expression during hepatic fibrosis to differentiation of HSCs. We found FBLN5 expression to be correlated with increased fibrosis stages. A significant differential expression between patients with low and high fibrosis stages was observed in the label-free approach as well as an increasing abundance with advancing fibrosis stages, both on transcript level and on protein level. Recently, FBLN5 together with MFAP4 was shown to be up-regulated in relation to tissue remodeling in COPD

34

. Together with our data, this suggests a general

function of FBLN5 and MFAP4 in disease-associated ECM remodeling. As the role of FBLN5 (concomitant with MFAP4) in COPD is discussed controversially 35 our results might be helpful in clarifying the function of these proteins in pathophysiological tissue remodeling. It is worth noting that all samples that were used in this study were biopsies, a clinical sample which is comparably easily available. However, although biopsy still represents the diagnostic gold standard for the assessment of liver fibrosis, the diagnostic quality of biopsied tissue is limited. A needle biopsy only displays a very small fraction of the liver (approximately 1/50000), leading to high sampling variability 36,37. Sampling error is assumed to be responsible for 10% of false negative diagnoses, especially because the fibrotic tissue is not distributed homogenously inside the organ 36. On top of this inter-observer variability may be considered, resulting in vague diagnoses especially for intermediate fibrosis stages (stages 1-3). Taking these facts into account, the disease-associated expression of the proteins presented in this study is characterized by a significant robustness, since they are qualified to discriminate several fibrosis stages irrespective of the heterogeneity of biopsied tissue samples. Clearly, the ultimate goal in biomarker discovery is the development of blood-based biomarkers allowing the minimally-invasive assessment of a disease state. The potential of the 19 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

presented disease-associated proteins to serve as such biomarkers needs to be addressed in further studies. The MRM assay as it is presented here did not reach the sensitivity required to quantify the candidates in serum samples from patients suffering fom liver fibrosis (except LUM which is a plasma protein). Further adjustment of the assay (e.g. 2D-LC, prefractionation of samples) might lead to higher sensitivity but also higher complexity of the assay thereby marking ELISA as the gold standard for quantification of low abundant proteins in blood samples in clinical routine. For MFAP4 the applicability as blood-based biomarker has already been demonstrated. As FBLN5 also is an extracellular protein and a similar biological function can be assumed, it seems to be advisable to follow up especially this novel disease-associated protein. As a plasma protein, LUM on the one hand is easily detectable in blood samples, on the other hand bigger amounts of LUM need to be set free to circulation to lead to significant changes in plasma concentrations. Possibly, LUM levels might be monitored intra-personally in high risk patients in order to detect exacerbations. For CSRP2 and CNN2 a potential usage for detection of early fibrosis stages might be investigated. However, if the amounts of these proteins which are produced in early fibrosis will be sufficient to be detectable in the blood stream needs to be tested. As none of the presented proteins is functioning in anti-viral response it seems likely that our findings might be transferable to non-HBV/HCV-related liver fibrosis as it has already been demonstrated for MFAP4 in alcoholic liver disease.


Page 20 of 40

Page 21 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Conclusions Here, we present a differential study comparing tissue samples of 95 patients suffering from hepatic fibrosis. We examined the samples by means of three individual quantitative methods, thereby giving evidence for disease-associated expression of the presented proteins both on transcript and protein level. To our knowledge, up to now no such study has been performed for hepatic fibrosis analyzing a cohort of comparable size. With Fibulin-5 (FBLN5) we discovered a new protein expressed in association to hepatic fibrosis. For lumican (LUM), collagen alpha-1(XIV) chain (COL14A1) and microfibril-associated glycoprotein 4 (MFAP4), which were already described in the context of liver fibrosis, we added substantial evidence for their disease-related expression. Since a robust association of protein expression with hepatic fibrosis is a prerequisite for the development of non-invasive biomarkers for this disease, we suggest a further follow up investigation of these proteins as biomarker candidates for the assessment of hepatic fibrosis.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1: Schematic representation of the analyzed sample cohorts. The sample cohorts used for the label-free discovery study and the MRM-based verification of disease-associated expression were independent of each other. The samples in which gene expression levels were analyzed (n = 77) overlapped with both cohorts analyzed in the proteomics approaches. In total 95 samples were analyzed.


Page 22 of 40

Page 23 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2: Volcano plot representing the results of the label-free discovery study. All displayed proteins were quantified with a minimum of 2 peptides. Dashed lines represent the applied thresholds (ANOVA p-value ≤ 0.05, Fold change ≥ 1.5). Proteins which fulfill the filter criteria are colored in red. Proteins which were selected for verification experiments are highlighted and labeled with the corresponding gene names.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3: Gene expression (transcript) levels of six genes expressed in association with hepatic fibrosis. Boxes represent first quartile, median and third quartile. Whiskers represent the standard deviation. The mean value is indicated by a filled circle. Individual data points are displayed as empty circles beside the boxes. Asterisks represent levels of significance, between fibrosis stages 0/1 and the marked stage (Tukey’s post-hoc test: * p < 0.05, ** p < 0.01, *** p < 0.001).


Page 24 of 40

Page 25 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4: Relative quantification of peptides representing proteins expressed in association with hepatic fibrosis by multiple reaction monitoring (MRM). Peptides representing MFAP4, FBLN5, COL14A1 or LUM were quantified. The concentrations of SIS peptides are indicated for each peptide. Boxes represent first quartile, median and third quartile. Whiskers represent the standard deviation. The mean value is indicated by a filled circle. Individual data points are displayed as empty circles beside the boxes. Asterisks represent levels of significance (Mann-Whitney U test: * p < 0.05, ** p < 0.01, *** p < 0.001).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 40

Table 1: Composition of the different sample cohorts analyzed using three individual quantification approaches.

1

Quantification Method

Label-free proteomics

Gene expression analysis1

Multiple reaction monitoring2

No. of analyzed samples

27

77

68

Age

54.1 ± 7.1

48.4 ± 10.9

46 ± 11.5

Sex

10 female; 17 male

34 female; 43 male

33 female; 35 male

Medical Condition

HCV

HBV/HCV

HBV/HCV

The samples in which the gene expression levels were analyzed overlapped with both cohorts

analyzed in the proteomics experiments: 25 samples which were analyzed in the label-free proteomics experiment and 52 samples which were analyzed using MRM (Figure 1). 2

The Samples which were analyzed by MRM were independent of the sample cohort used for the label-free discovery study.


Page 27 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 2: Proteins expressed in association with hepatic fibrosis which were selected for verification experiments. UniProt

1

ANOVA

Fold Change2

Accession

Gene

Protein

p-value

P51884

LUM

Lumican

0.026

2.30

Q9UBX5

FBLN5

Fibulin-5

0.009

4.53

Q16527

CSRP2

Cysteine and glycine-rich protein 2

0.001

6.23

Q99439

CNN2

Calponin 2

0.003

2.15

Q01995

TGLN

Transgelin

0.003

3.79

Q05707

COL14A1

Collagen alpha-1(XIV) chain

0.018

2.08

P55083

MFAP4

Microfibril-associated glycoprotein 4

0.006

7.21

Low fibrosis group (stages 0- 2, n = 13) and high fibrosis group (stages 3 and 4, n=14) were

compared. 2

1

All proteins were up-regulated in the high fibrosis group.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Supporting Information, this material is available free of charge via http://pubs.acs.org/. Detailed description of materials and methods. Supplementary Figure 1: Calibration curves of SIS peptides. Supplementary Figure 2: Relative quantification of TAGLN, CSRP2 and CNN2 by multiple reaction monitoring (MRM). Supplementary Table 1: Characteristics of clinical samples analyzed in the label-free discovery approach. Supplementary Table 2: Characteristics of clinical samples analyzed in the MRM verification study. Supplementary Table 3: Proteins which were found to be significantly differentially expressed between low and high fibrosis groups in the label-free discovery study. Supplementary Table 4: Peptides and fragment ions as well as corresponding m/z values which were used for quantification in the MRM approach. Supplementary Table 5: Statistical analysis of the gene expression data. Supplementary Table 6: Statistical analysis of the MRM verification study.

Corresponding author Dr. Thilo Bracht, Medizinisches Proteom-Center, Ruhr-Universität Bochum, 44801 Bochum, Germany, Tel. +49-(0)-234/32-29985, E-mail: [email protected]


Page 28 of 40

Page 29 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Acknowledgements The authors would like to thank Don Marvin Voss, Kristin Rosowski, Birgit Korte and Stephanie Tautges for their excellent technical assistance. This work was supported by the PROFILE project, which is co-funded by the European Union (European Regional Development Fund - Investing in your future) and the German federal state North RhineWestphalia (NRW), project number z0911bt004e. A part of this study was funded by P.U.R.E. (Protein research Unit Ruhr within Europe), a project of North Rhine-Westphalia, a federal state of Germany. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http://proteomecentral. proteomexchange.org) via the PRIDE partner repository with the data set identifier PXD001474 and DOI 10.6019/PXD001474. The authors thank the PRIDE Team as well as Julian Uszkoreit for their assistance during the data upload.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abbreviations in the order of appearance: LC-MS, liquid chromatography-mass spectrometry ECM, extra cellular matrix MRM, multiple reaction monitoring HBV, hepatitis B virus HCV, hepatitis C virus ACN, acetonitrile FA, formic acid TFA, trifluoroacetic acid MS/MS, tandem mass spectrometry SIS peptide, stable isotope-labeled synthetic peptide RP-HPLC, reversed phase-high performance liquid chromatography MALDI-TOF, matrix-assisted laser desorption/ionization-time of flight QQQ, triple quadrupole NASH, non-alcoholic steatohepatitis αSMA, alpha smooth muscle actin HSC, hepatic stellate cell TGF-β, transforming growth factor β NAFLD, non-alcoholic fatty liver disease COPD, chronic obstructive pulmonary disease


Page 30 of 40

Page 31 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References 1.

Schuppan, D.; Afdhal, N. H., Liver cirrhosis. Lancet 2008, 371, (9615), 838-

2.

Wiegand, J.; Berg, T., The etiology, diagnosis and prevention of liver cirrhosis:

51.

part 1 of a series on liver cirrhosis. Dtsch. Arztebl. Int. 2013, 110, (6), 85-91. 3.

Bravo, A. A.; Sheth, S. G.; Chopra, S., Liver biopsy. N. Engl. J. Med. 2001,

344, (7), 495-500. 4.

Marcellin, P.; Gane, E.; Buti, M.; Afdhal, N.; Sievert, W.; Jacobson, I. M.;

Washington, M. K.; Germanidis, G.; Flaherty, J. F.; Schall, R. A.; Bornstein, J. D.; Kitrinos, K. M.; Subramanian, G. M.; McHutchison, J. G.; Heathcote, E. J., Regression of cirrhosis during treatment with tenofovir disoproxil fumarate for chronic hepatitis B: a 5-year openlabel follow-up study. Lancet 2013, 381, (9865), 468-75. 5.

Adams, L. A., Biomarkers of liver fibrosis. J. Gastroenterol. Hepatol. 2011,

26, (5), 802-9. 6.

Patel, K.; Gordon, S. C.; Jacobson, I.; Hezode, C.; Oh, E.; Smith, K. M.;

Pawlotsky, J. M.; McHutchison, J. G., Evaluation of a panel of non-invasive serum markers to differentiate mild from moderate-to-advanced liver fibrosis in chronic hepatitis C patients. J. Hepatol. 2004, 41, (6), 935-42. 7.

Rosenberg, W. M.; Voelker, M.; Thiel, R.; Becka, M.; Burt, A.; Schuppan, D.;

Hubscher, S.; Roskams, T.; Pinzani, M.; Arthur, M. J.; European Liver Fibrosis, G., Serum markers detect the presence of liver fibrosis: a cohort study. Gastroenterology 2004, 127, (6), 1704-13. 8.

Imbert-Bismut, F.; Ratziu, V.; Pieroni, L.; Charlotte, F.; Benhamou, Y.;

Poynard, T.; Group, M., Biochemical markers of liver fibrosis in patients with hepatitis C virus infection: a prospective study. Lancet 2001, 357, (9262), 1069-75.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

9.

Adams, L. A.; Bulsara, M.; Rossi, E.; DeBoer, B.; Speers, D.; George, J.;

Kench, J.; Farrell, G.; McCaughan, G. W.; Jeffrey, G. P., Hepascore: an accurate validated predictor of liver fibrosis in chronic hepatitis C infection. Clin. Chem. 2005, 51, (10), 186773. 10.

Wai, C. T.; Greenson, J. K.; Fontana, R. J.; Kalbfleisch, J. D.; Marrero, J. A.;

Conjeevaram, H. S.; Lok, A. S., A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology 2003, 38, (2), 518-26. 11.

Tsochatzis, E. A.; Crossan, C.; Longworth, L.; Gurusamy, K.; Rodriguez-

Peralvarez, M.; Mantzoukis, K.; O'Brien, J.; Thalassinos, E.; Papastergiou, V.; Noel-Storr, A.; Davidson, B.; Burroughs, A. K., Cost-effectiveness of noninvasive liver fibrosis tests for treatment decisions in patients with chronic hepatitis C. Hepatology 2014, 60, (3), 832-43. 12.

Hannivoort, R. A.; Hernandez-Gea, V.; Friedman, S. L., Genomics and

proteomics in liver fibrosis and cirrhosis. Fibrogenesis Tissue Repair 2012, 5, (1), 1. 13.

Molleken, C.; Sitek, B.; Henkel, C.; Poschmann, G.; Sipos, B.; Wiese, S.;

Warscheid, B.; Broelsch, C.; Reiser, M.; Friedman, S. L.; Tornoe, I.; Schlosser, A.; Kloppel, G.; Schmiegel, W.; Meyer, H. E.; Holmskov, U.; Stuhler, K., Detection of novel biomarkers of liver cirrhosis by proteomic analysis. Hepatology 2009, 49, (4), 1257-66. 14.

Megger, D. A.; Bracht, T.; Kohl, M.; Ahrens, M.; Naboulsi, W.; Weber, F.;

Hoffmann, A. C.; Stephan, C.; Kuhlmann, K.; Eisenacher, M.; Schlaak, J. F.; Baba, H. A.; Meyer, H. E.; Sitek, B., Proteomic differences between hepatocellular carcinoma and nontumorous liver tissue investigated by a combined gel-based and label-free quantitative proteomics study. Mol. Cell. Proteomics 2013, 12, (7), 2006-20. 15.

Kohl, M.; Megger, D. A.; Trippler, M.; Meckel, H.; Ahrens, M.; Bracht, T.;

Weber, F.; Hoffmann, A. C.; Baba, H. A.; Sitek, B.; Schlaak, J. F.; Meyer, H. E.; Stephan, C.; Eisenacher, M., A practical data processing workflow for multi-OMICS projects. Biochim. Biophys. Acta 2014, 1844, (1 Pt A), 52-62. 32 ACS Paragon Plus Environment

Page 32 of 40

Page 33 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


16.

Percy, A. J.; Chambers, A. G.; Yang, J.; Borchers, C. H., Multiplexed MRM-

based quantitation of candidate cancer biomarker proteins in undepleted and non-enriched human plasma. Proteomics 2013, 13, (14), 2202-15. 17.

Benjamini, Y.; Hochberg, Y., Controlling the false discovery rate: a practical

and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B 1995, 57, (1), 289–300. 18.

Herrmann, J.; Borkham-Kamphorst, E.; Haas, U.; Van de Leur, E.; Fraga, M.

F.; Esteller, M.; Gressner, A. M.; Weiskirchen, R., The expression of CSRP2 encoding the LIM domain protein CRP2 is mediated by TGF-beta in smooth muscle and hepatic stellate cells. Biochem. Biophys. Res. Commun. 2006, 345, (4), 1526-35. 19.

Assinder, S. J.; Stanton, J. A.; Prasad, P. D., Transgelin: an actin-binding

protein and tumour suppressor. Int. J. Biochem. Cell. Biol. 2009, 41, (3), 482-6. 20.

Herrmann, J.; Arias, M.; Van De Leur, E.; Gressner, A. M.; Weiskirchen, R.,

CSRP2, TIMP-1, and SM22alpha promoter fragments direct hepatic stellate cell-specific transgene expression in vitro, but not in vivo. Liver Int. 2004, 24, (1), 69-79. 21.

Jiang, W. R.; Cady, G.; Hossain, M. M.; Huang, Q. Q.; Wang, X.; Jin, J. P.,

Mechanoregulation of h2-calponin gene expression and the role of Notch signaling. J. Biol. Chem. 2014, 289, (3), 1617-28. 22.

Ueki, N.; Ohkawa, T.; Yamamura, H.; Takahashi, K.; Tsutsui, T.; Kawai, Y.;

Yokoyama, Y.; Amuro, Y.; Hada, T.; Higashino, K., Induction of calponin-h1 by transforming growth factor-beta1 in cultured human ito cells, LI90. Biochim. Biophys. Acta 1998, 1403, (1), 28-36. 23.

Bataller, R.; Brenner, D. A., Liver fibrosis. J. Clin. Invest. 2005, 115, (2), 209-

18.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

24.

Milani, S.; Grappone, C.; Pellegrini, G.; Schuppan, D.; Herbst, H.; Calabro, A.;

Casini, A.; Pinzani, M.; Surrenti, C., Undulin RNA and protein expression in normal and fibrotic human liver. Hepatology 1994, 20, (4 Pt 1), 908-16. 25.

Krull, N. B.; Gressner, A. M., Differential expression of keratan sulphate

proteoglycans fibromodulin, lumican and aggrecan in normal and fibrotic rat liver. FEBS Lett. 1992, 312, (1), 47-52. 26.

Charlton, M.; Viker, K.; Krishnan, A.; Sanderson, S.; Veldt, B.; Kaalsbeek, A.

J.; Kendrick, M.; Thompson, G.; Que, F.; Swain, J.; Sarr, M., Differential expression of lumican and fatty acid binding protein-1: new insights into the histologic spectrum of nonalcoholic fatty liver disease. Hepatology 2009, 49, (4), 1375-84. 27.

Krishnan, A.; Li, X.; Kao, W. Y.; Viker, K.; Butters, K.; Masuoka, H.;

Knudsen, B.; Gores, G.; Charlton, M., Lumican, an extracellular matrix proteoglycan, is a novel requisite for hepatic fibrosis. Lab. Invest. 2012, 92, (12), 1712-25. 28.

Furuta, K.; Sato, S.; Yamauchi, T.; Ozawa, T.; Harada, M.; Kakumu, S.,

Intrahepatic gene expression profiles in chronic hepatitis B and autoimmune liver disease. J. Gastroenterol. 2008, 43, (11), 866-74. 29.

Wulf-Johansson, H.; Lock Johansson, S.; Schlosser, A.; Trommelholt Holm,

A.; Rasmussen, L. M.; Mickley, H.; Diederichsen, A. C.; Munkholm, H.; Poulsen, T. S.; Tornoe, I.; Nielsen, V.; Marcussen, N.; Vestbo, J.; Saekmose, S. G.; Holmskov, U.; Sorensen, G. L., Localization of microfibrillar-associated protein 4 (MFAP4) in human tissues: clinical evaluation of serum MFAP4 and its association with various cardiovascular conditions. PloS One 2013, 8, (12), e82243. 30.

Johansson, S. L.; Roberts, N. B.; Schlosser, A.; Andersen, C. B.; Carlsen, J.;

Wulf-Johansson, H.; Saekmose, S. G.; Titlestad, I. L.; Tornoe, I.; Miller, B.; Tal-Singer, R.; Holmskov, U.; Vestbo, J.; Sorensen, G. L., Microfibrillar-associated protein 4: a potential biomarker of chronic obstructive pulmonary disease. Respir. Med. 2014, 108, (9), 1336-44. 34 ACS Paragon Plus Environment

Page 34 of 40

Page 35 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


31.

Nakamura, T.; Lozano, P. R.; Ikeda, Y.; Iwanaga, Y.; Hinek, A.; Minamisawa,

S.; Cheng, C. F.; Kobuke, K.; Dalton, N.; Takada, Y.; Tashiro, K.; Ross Jr, J.; Honjo, T.; Chien, K. R., Fibulin-5/DANCE is essential for elastogenesis in vivo. Nature 2002, 415, (6868), 171-5. 32.

Hu, Q.; Loeys, B. L.; Coucke, P. J.; De Paepe, A.; Mecham, R. P.; Choi, J.;

Davis, E. C.; Urban, Z., Fibulin-5 mutations: mechanisms of impaired elastic fiber formation in recessive cutis laxa. Hum. Mol. Genet. 2006, 15, (23), 3379-86. 33.

Yanagisawa, H.; Schluterman, M. K.; Brekken, R. A., Fibulin-5, an integrin-

binding matricellular protein: its function in development and disease. J. Cell. Commun. Signal. 2009, 3, (3-4), 337-47. 34.

Brandsma, C. A.; van den Berge, M.; Postma, D. S.; Jonker, M. R.; Brouwer,

S.; Pare, P. D.; Sin, D. D.; Bosse, Y.; Laviolette, M.; Karjalainen, J.; Fehrmann, R. S.; Nickle, D. C.; Hao, K.; Spanjer, A. I.; Timens, W.; Franke, L., A large lung gene expression study identifying fibulin-5 as a novel player in tissue repair in COPD. Thorax 2014, 70, (1), 21-32. 35.

Bruselle G., Dysregulated fibulin-5 expression and elastogenesis in COPD

lungs: pyromaniac or fire fighter? Thorax 2015, 70(1),1-2. 36.

Bedossa, P.; Dargere, D.; Paradis, V., Sampling variability of liver fibrosis in

chronic hepatitis C. Hepatology 2003, 38, (6), 1449-57. 37.

Ratziu, V.; Charlotte, F.; Heurtier, A.; Gombert, S.; Giral, P.; Bruckert, E.;

Grimaldi, A.; Capron, F.; Poynard, T.; Group, L. S., Sampling variability of liver biopsy in nonalcoholic fatty liver disease. Gastroenterology 2005, 128, (7), 1898-906.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract Graphic 69x50mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 36 of 40

Page 37 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1: Schematic representation of the analyzed sample cohorts. The sample cohorts used for the label-free discovery study and the MRM-based verification of diseaseassociated expression were independent of each other. The samples in which gene expression levels were analyzed (n = 77) overlapped with both cohorts analyzed in the proteomics approaches. In total 95 samples were analyzed. 158x64mm (300 x 300 DPI)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2: Volcano plot representing the results of the label-free discovery study. All displayed proteins were quantified with a minimum of 2 peptides. Dashed lines represent the applied thresholds (ANOVA p-value ≤ 0.05, Fold change ≥ 1.5). Proteins which fulfill the filter criteria are colored in red. Proteins which were selected for verification experiments are highlighted and labeled with the corresponding gene names. 136x94mm (300 x 300 DPI)


Page 38 of 40

Page 39 of 40

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3: Gene expression (transcript) levels of six genes expressed in association with hepatic fibrosis. Boxes represent first quartile, median and third quartile. Whiskers represent the standard deviation. The mean value is indicated by a filled circle. Individual data points are displayed as empty circles beside the boxes. Asterisks represent levels of significance, between fibrosis stages 0/1 and the marked stage (Tukey’s post-hoc test: * p < 0.05, ** p < 0.01, *** p < 0.001). 126x80mm (300 x 300 DPI)



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 4: Relative quantification of peptides representing proteins expressed in association with hepatic fibrosis by multiple reaction monitoring (MRM). Peptides representing MFAP4, FBLN5, COL14A1 or LUM were quantified. The concentrations of SIS peptides are indicated for each peptide. Boxes represent first quartile, median and third quartile. Whiskers represent the standard deviation. The mean value is indicated by a filled circle. Individual data points are displayed as empty circles beside the boxes. Asterisks represent levels of significance (Mann-Whitney U test: * p < 0.05, ** p < 0.01, *** p < 0.001). 114x130mm (300 x 300 DPI)


Page 40 of 40

Analysis of Disease-Associated Protein Expression Using Quantitative

Recommend Documents