Characterization of the Vitreous Proteome in Diabetes without Diabetic

Diabetic retinopathy (DR) is the most common microvascular complication caused by diabetes mellitus and is a leading cause of vision loss among workin...
0 downloads 0 Views 682KB Size
Download PDF
Characterization of the Vitreous Proteome in Diabetes without Diabetic Retinopathy and Diabetes with Proliferative Diabetic Retinopathy Ben-Bo Gao,†,§ Xiaohong Chen,† Nigel Timothy,| Lloyd Paul Aiello,†,‡,| and Edward P. Feener*,†,§ Research Division, Beetham Eye Institute, Joslin Diabetes Center, One Joslin Place, Boston, Massachusetts 02215, and Departments of Medicine and Ophthalmology, Harvard Medical School, Boston, Massachusetts 02215 Received February 12, 2008

An understanding of the diabetes-induced alterations in vitreous protein composition in the absence and in the presence of proliferative diabetic retinopathy (PDR) may provide insights into factors and mechanisms responsible for this disease. We have performed a comprehensive proteomic analysis and comparison of vitreous samples from individuals with diabetes but without diabetic retinopathy (noDR) or with PDR and nondiabetic individuals (NDM). Using preparative one-dimensional SDS-PAGE and nano-LC/MS/MS of 17 independent vitreous samples, we identified 252 proteins from human vitreous. Fifty-six proteins were differentially abundant in noDR and PDR vitreous compared with NDM vitreous, including 32 proteins increased and 10 proteins decreased in PDR vitreous compared with NDM vitreous. Comparison of noDR and PDR groups revealed increased levels of angiotensinogen and decreased levels of calsyntenin-1, interphotoreceptor retinoid-binding protein, and neuroserpin in PDR vitreous. Biological pathway analysis revealed that vitreous contains 30 proteins associated with the kallikrein-kinin, coagulation, and complement systems. Five of them (complement C3, complement factor I, prothrombin, alpha-1-antitrypsin, and antithrombin III) were increased in PDR vitreous compared with NDM vitreous. Factor XII was detected in PDR vitreous but not observed in either NDM or noDR vitreous. PDR vitreous also had increased levels of peroxiredoxin-1 and decreased levels of extracellular superoxide dismutase, compared with noDR or NDM vitreous. These data provide an in depth analysis of the human vitreous proteome and reveal protein alterations that are associated with PDR. Keywords: proliferative diabetic retinopathy • human vitreous • LC-MS/MS • quantitative proteomics • kallikrein-kinin system • extracellular matrix

Introduction

sive medical therapies for the prevention of PDR remain an unmet clinical need.3

Diabetic retinopathy (DR) is the most common microvascular complication caused by diabetes mellitus and is a leading cause of vision loss among working-age adults in developed countries.1 The initial clinical stages of DR are characterized by the development of microaneurysms, retinal hemorrhages, hard exudates, and intraretinal microvascular abnormalities.2 Further retinopathy progression can lead to sight-threatening pathological retinal vascular leakage (macular edema) or retinal neovascularization, termed proliferative diabetic retinopathy (PDR). Although management of risk factors, including hyperglycemia, hyperlipidemia, and hypertension, have been shown to ameliorate diabetes-induced vision loss, effective noninva-

While the histological changes that occur during the pathogenesis of DR have been described in detail,2 less is known regarding the intraocular biochemical changes associated with the sight-threatening stages of this disease. Previous studies have demonstrated that the analysis of the vitreous fluid from people with DR can provide insight into mechanisms that potentially contribute to the pathogenesis of DR.4–6 The retinal interstitium and vitreous fluid are separated from the peripheral circulation by the blood-retinal barrier. Factors released from the retina can diffuse into the vitreous, and components within the vitreous can affect retinal function, suggesting that the vitreous may play an integral role in retinal physiology.

* To whom correspondence should be addressed. Edward P. Feener, Joslin Diabetes Center, One Joslin Place, Boston, MA 02215. Tel.: (617) 732-2599. Fax: (617) 732-2637; E-mail: [email protected]. † Research Division, Joslin Diabetes Center. § Department of Medicine, Harvard Medical School. | Department of Ophthalmology, Harvard Medical School. ‡ Beetham Eye Institute, Joslin Diabetes Center.

Previous mass spectrometry-based proteomic studies of human vitreous from our laboratory and others have begun to catalogue the vitreous proteome and have identified a limited number of changes in protein abundance in the vitreous obtained from people with PDR as compared with nondiabetic control subjects. Fractionation of vitreous fluid by 2-D SDS-

2516 Journal of Proteome Research 2008, 7, 2516–2525 Published on Web 04/24/2008

10.1021/pr800112g CCC: $40.75

 2008 American Chemical Society

research articles

Proteomic Analysis of Human Vitreous with PDR 7,8

PAGE has revealed from 600 to 1400 spots of protein staining, providing a visual estimate of vitreous proteome complexity. Analyses of proteins isolated on 2-D SDS-PAGE have resulted in the identification of between 38 and 51 proteins in the vitreous, including a limited number of proteins that differed in abundance based on protein staining of PDR vitreous compared with vitreous from nondiabetic individuals (NDM).7–12 Recently, using samples fractionated by 1-D SDS-PAGE and analysis by LC-MS/MS, we have identified 117 proteins in human vitreous.5 Investigation of some of these newly identified proteins in PDR vitreous revealed a novel mechanism of extracellular carbonic anhydrase-induced retinal vascular permeability.5 A recent study using pooled vitreous samples and protein identification based on at least one peptide match has further expanded the list of vitreous proteins.13 In this report, we have expanded the number of independent vitreous samples analyzed and have compiled and compared results generated from X!Tandem and SEQUEST algorithms. This analysis resulted in the identification of 252 proteins, which was based on at least 2 unique peptide matches and confirmation in independent vitreous samples. Using a labelfree quantitative approach, we have compared the abundance of these proteins among study groups. We have also characterized the vitreous proteome based on annotated protein cellular location, function, and biological process.

Materials and Methods Study Subjects and Sample Collection. Vitreous fluid was obtained from individuals undergoing pars plana vitrectomy at the Beth Israel Deaconess Medical Center, MA, Santa Barbara Cottage Hospital Eye Center, CA, and Asociacio´n Para Evitar la Ceguera en Me´xico Hospital, Mexico City, Mexico, in accordance with approved Human Discarded Specimen Research Protocols from institutional review boards. Undiluted samples were collected at the time of surgery, immediately placed on ice, spun at 15 000g for 1 min to remove insoluble material and stored at -80 °C. Proteomic Analysis. Proteomic analysis was performed on 50 µL of undiluted vitreous from noDR (diabetes with no apparent diabetic retinopathy), PDR, and NDM subjects. The absence of DR and presence of PDR were defined as Early Treatment Diabetic Retinopathy Study (ETDRS) grading scale2 level 10 or >60, respectively. Soluble proteins were separated by 12% SDS-PAGE, and the entire lane for each sample was divided into slices of 1-2 mm width. Tryptic digests of individual gel slices were analyzed by nano-LC/MS/MS using a linear ion trap mass spectrometer (LTQ, Thermo Electron). Assignment of MS/MS data was performed using SEQUEST (Bioworks 3.2, Thermo Electron) and X!Tandem (version 2006.09.15, http://www.thegpm.org) search against the International Protein Index (IPI) human sequence database (IPI_ HUMAN_v3.24, 66921 sequences, http://www.ebi.ac.uk) and a randomized version of the same IPI database generated by a Perl script, decoy.pl (Matrix Science, London, U.K.). The SEQUEST and X!Tandem search parameters were potential residue mass modification of +16.0 for oxidized methionine and +71.0 for acrylamide alkylated cysteine; peptide tolerance 2 Da, fragment ions tolerance 0.5 Da; digestion with trypsin; a maximum of one missed tryptic cleavage. Resultant matches were compiled into a MySQL database, and proteomics computational analyses were performed using MS Result Manager, which is PHP-MySQL based software developed in our laboratory (Figure 1A). First, peptide identifications were made based

Figure 1. Proteomic analysis process and the number of proteins identified. (A) Schematic of gel LC-MS/MS analysis and data processing. (B) Venn diagram of proteins identified using X!Tandem and SEQUEST algorithms. The number of proteins identified from 17 independent vitreous samples and percent of total number of proteins identified by each algorithm are shown.

on the criteria described previously5 for SEQUEST search results and -log(E) g 2.00 for X!Tandem search results. Second, protein identifications were assigned when the following criteria were met: the unique peptide match number was greater than or equal to two; peptides contributing to protein matches were derived from a single gel slice or from adjacent slices, and the protein was identified in at least two vitreous samples. In the case that shared peptides map to more than one protein identifier (ID), the ID with the larger number of peptides was retained. If multiple IDs had the same matched peptides, the ID with more Gene Ontology (GO) annotations14 was retained. Gene symbols have been reported in the proteome list because most of the duplicate IDs shared the same gene symbols. The false-positive rate (FPR) of protein identification was calculated by dividing the number of random sequences by the sum of “random” and “real” sequences and multiplying by 100.15 Spectral count (the total number of observations of spectral-peptides matches) for each protein was calculated by the summation of peptides matched using SEQUEST and X!Tandem algorithms. Bioinformatic Analysis. GO annotations were extracted from Gene Ontology Annotation Database (GOA, Human 46.0)16 and generic GO slim provided by European Bioinformatics Institute using MS Result Manager. Grouping by GO term and counting protein number within one group were performed using MS Result Manager. Computational methods (TargetP 1.1,17 MultiLoc,18 SubLoc19) were used for predicting protein subcellular location. Gene symbols corresponding to identified proteins Journal of Proteome Research • Vol. 7, No. 6, 2008 2517

research articles

Gao et al. a

Table 1. Demographics of Study Subjects group and level of retinopathy

n

age (years)

gender (F/M)

DM (type 1/type 2)

DME (y/n)

indications

NDM noDR PDR

6 4 7

73.8 ( 7.5 74.8 ( 6.0 45.3 ( 10.7*

1/5 1/3 1/6

n/a 0/4 5/2

n/a 0/4 5/2

MH(1), ERM(4), Glau Impl(1) MH(3), RD(1) RD(1), TRD(6)

a Non-diabetes mellitus (NDM), diabetes with no apparent diabetic retinopathy (noDR), proliferative diabetic retinopathy (PDR), diabetic macular edema (DME). Values represent means ( SD. Surgical indications are epiretinal membrane (ERM), glaucoma Implant (Glau Impl), macular hole (MH), retinal detachment (RD), traction retinal detachment (TRD). Values in parentheses indicate the number of individuals. Significant differences are indicated as * P < 0.001 compared with both NDM and noDR groups.

were analyzed through the use of the Database for Annotation, Visualization and Integrated Discovery (DAVID).20 Statistics. Linear regression and correlation analyses were performed by GraphPAD Prism (GraphPAD Software, San Diego, CA). Unpaired t tests were performed by in-house PHP script incorporated in MS Result Manager based on PHP statistics extension (http://www.php.net/). Results were expressed as mean ( SEM. Values of P < 0.05 were considered significantly different.

Results Identification of the Proteins in Human Vitreous. Vitreous samples were obtained from NDM (n ) 6), noDR (ETDRS < 10, n ) 4) and PDR (n ) 7) patients (Table 1). Five subjects in the PDR group had diabetic macular edema (DME). The indications for vitrectomy included macular hole, epiretinal membrane, and retinal detachment. Vitreous samples were fractionated by SDS-PAGE and analyzed by LC/MS/MS, and peptide matches were generated by using both SEQUEST and X!Tandem algorithms (Figure 1A). The number of protein matches generated from SEQUEST was 226 and from X!Tandem was 231. The FPRs were 14.8% and 0.4%, respectively, comparable to a recent report.21 The total number of proteins identified by both search algorithms was 192. In addition, 35 and 40 proteins were identified only by SEQUEST and X!Tandem, respectively (Figure 1B). We used 2 unique peptide matches generated from SEQUEST using HUPO low confidence score criteria along with single peptide match using X!Tandem to increase the number of proteins detected in the vitreous proteome beyond those based on at least 2 unique peptides using X!Tandem alone. Twenty-one of the 35 proteins initially identified by SEQUEST alone which contained a single peptide confirmed by X!Tandem were retained. The remaining 14 proteins, not confirmed by X!Tandem, were excluded. Of the 40 proteins initially identified by X!Tandem alone, 30 proteins were confirmed by a single peptide match from SEQUEST. A single protein (IPI00737886: similar to mucin 19) in the X!Tandem only group was deleted because it did not contain 2 unique peptides in the same gel slice. Since the FPR for X!Tandem with 2 peptide matches was 0.4%, we retained the remaining proteins. A total of 252 proteins [192 (both) + 21 (SEQUEST + 1 peptide in X!Tandem) + 39 (X!Tandem alone)] were thus identified from the collection of 17 vitreous samples. The detection of peptides using X!Tandem and SEQUEST for each protein is indicated in Supporting Information Table. Relative Abundance of Proteins in Human Vitreous. Previous reports indicate that spectral counting can be used as a semiquantitative measurement of protein abundance.22–24 However, the number of unique peptides generated from a given protein is limited by primary sequence, which severely limits the concentration range where this value can be used as a semiquantitative measure. In this study, we used spectral 2518

Journal of Proteome Research • Vol. 7, No. 6, 2008

Figure 2. Fractional distribution of the most abundant proteins in human vitreous. (A) Chart showing a summary of the relative amounts of highly abundant proteins in PDR vitreous. (B) Table showing the mean percent of number of total peptides for the 15 most abundant proteins identified in NDM, noDR, and PDR samples relative to the number of total peptides detected from respective samples.

counting to assess protein abundance. As expected, the most abundant protein in all vitreous samples was albumin, which represented 40-42% of the total spectral count. Additional high-abundance proteins in the vitreous from all three groups includes serotransferrin, transthryretin, clusterin, immunoglobulin, prostaglandin-H2 D-isomerase, alpha-1 antitrypsin, pigment epithelium-derived factor (PEDF), complement C3, and apolipoprotein A-1 (Figure 2). The inventory of vitreous proteins is listed alphabetically and the average number of total peptides detected in each group is shown in Supporting Information Table. Comparison of Vitreous Proteomes among NDM, noDR, and PDR Samples. The spectral count for each protein from each sample was used to compare protein abundance among groups. In our previous report, using both the number of

research articles

Proteomic Analysis of Human Vitreous with PDR

Table 2. Proteins Differently Changed in Human Vitreous Proteome in Patients with NDM (n ) 6), noDR (n ) 4) or PDR (n ) 7)a gene symbol

IPI

AGT APOC3 ALB SERPINF1 CFI APOA1 IGHG1 PRDX1 HRG GC APOA2 ACTA2 HSPG2

IPI00032220 IPI00387120 IPI00657670 IPI00745872 IPI00006114 IPI00291867 IPI00021841 IPI00448938 IPI00000874 IPI00022371 IPI00555812 IPI00021854 IPI00021439 IPI00024284

SPON1 PRSS3 LRG1 HBD ORM2 APLP2 HBA1 AHSG CA1 A1BG TTR SERPINA1 GSN RBP4 CHI3L1 CD14 NAV3 F2 SERPINA3 COL5A3 C3 B2M SERPINC1 IGKV2-40 IGL@ CFB AFM APOA4 APP C4A ATP6AP1 SERPING1 AMBP RBP3 PGK1 HPX SERPINI1 SOD3 IMPG2 CLSTN1 C1QC ENPP2

IPI00171473 IPI00220839 IPI00022417 IPI00473011 IPI00022429 IPI00031030 IPI00410714 IPI00022431 IPI00215983 IPI00022895 IPI00022432 IPI00553177 IPI00641047 IPI00022420 IPI00002147 IPI00029260 IPI00217051 IPI00019568 IPI00550991 IPI00018279 IPI00164623 IPI00004656 IPI00032179 IPI00387107 IPI00154742 IPI00019591 IPI00019943 IPI00304273 IPI00006608 IPI00418163 IPI00784119 IPI00291866 IPI00022426 IPI00022337 IPI00169383 IPI00022488 IPI00016150 IPI00027827 IPI00296866 IPI00413959 IPI00022394 IPI00156171

protein name

NDM

noDR

Angiotensinogen 5.00 ( 1.31 Ig kappa chain V-IV region Len 0.58 ( 0.58 Apolipoprotein C-III 1.17 ( 1.17 Serum albumin 1558.67 ( 349.38 Pigment epithelium-derived factor 81.92 ( 23.89 Complement factor I 3.50 ( 1.60 Apolipoprotein A-I 40.33 ( 11.09 IGHG1 protein 48.75 ( 3.95 Peroxiredoxin-1 0 Histidine-rich glycoprotein 3.50 ( 1.31 Vitamin D-binding protein 37.67 ( 5.77 Apolipoprotein A-II 3.00 ( 1.37 Actin cytoplasmic 1 2.33 ( 0.80 Basement membrane-specific heparan 0.67 ( 0.67 sulfate proteoglycan core protein Spondin-1 5.25 ( 1.48 Protease serine 3 16.08 ( 4.11 Leucine-rich alpha-2-glycoprotein 1.42 ( 0.92 Hemoglobin subunit delta 0 Alpha-1-acid glycoprotein 1 13.75 ( 2.97 Amyloid-like protein 2 6.00 ( 2.40 Hemoglobin subunit alpha 0 Alpha-2-HS-glycoprotein 11.58 ( 2.21 Carbonic anhydrase 1 0 Alpha-1B-glycoprotein 6.50 ( 1.28 Transthyretin 184.67 ( 54.85 Alpha-1-antitrypsin 103.00 ( 27.55 Gelsolin 14.33 ( 2.79 Plasma retinol-binding protein 2.25 ( 2.25 Chitinase-3-like protein 1 6.67 ( 3.58 Monocyte differentiation antigen CD14 0.33 ( 0.33 Steerin3 protein 3.67 ( 1.69 Prothrombin 1.83 ( 0.65 Alpha-1-antichymotrypsin 11.83 ( 1.79 Collagen alpha-3(V) chain 1.50 ( 0.72 Complement C3 60.17 ( 16.42 Beta-2-microglobulin 2.92 ( 2.22 Antithrombin III 12.83 ( 4.26 Ig kappa chain V-II region Cum 2.33 ( 0.76 IGL@ protein 25.83 ( 4.51 Complement factor B 6.50 ( 2.52 Afamin 3.08 ( 0.66 Apolipoprotein A-IV 9.92 ( 1.50 Amyloid beta A4 protein 1.17 ( 0.75 Complement component 4B preproprotein 27.17 ( 6.13 Vacuolar ATP synthase subunit S1 1.17 ( 0.83 Plasma protease C1 inhibitor 13.08 ( 1.81 AMBP protein 2.42 ( 1.24 Interphotoreceptor retinoid-binding protein 71.83 ( 10.68 Phosphoglycerate kinase 1 0 Hemopexin 26.58 ( 7.24 Neuroserpin 0.58 ( 0.58 Extracellular superoxide dismutase [Cu-Zn] 0.92 ( 0.92 Interphotoreceptor matrix proteoglycan 2 1.00 ( 1.00 Calsyntenin-1 4.58 ( 2.85 Complement C1q subcomponent subunit C 1.17 ( 1.17 Ectonucleotide pyrophosphatase/ 3.33 ( 2.16 phosphodiesterase 2

PDR

8.50 ( 1.85 13.86 ( 1.17***# 7.25 ( 2.93* 10.14 ( 2.22** 5.25 ( 5.25 16.64 ( 3.90** 3631.75 ( 1068.89 6921.43 ( 1386.68** 168.38 ( 72.10 258.64 ( 43.84** 7.88 ( 2.20 15.93 ( 3.17** 122.38 ( 47.15 413.43 ( 107.47** 174.88 ( 122.57 313.79 ( 78.77** 3.00 ( 2.38 7.43 ( 2.27* 8.13 ( 1.39* 13.36 ( 2.85* 54.50 ( 12.85 121.64 ( 25.64* 8.63 ( 5.32 16.64 ( 4.11* 13.75 ( 7.10 6.79 ( 1.25* 3.38 ( 1.95 7.64 ( 2.17* 4.25 ( 3.19 23.13 ( 13.94 4.13 ( 3.20 83.13 ( 77.56 47.75 ( 26.80 5.88 ( 2.13 122.88 ( 118.25 20.13 ( 4.49 13.75 ( 13.75 16.63 ( 3.77* 249.50 ( 45.27 236.25 ( 67.52 31.50 ( 9.44 17.63 ( 9.24 12.75 ( 6.22 0.88 ( 0.88 0 8.88 ( 4.07 42.25 ( 12.82* 0 193.75 ( 79.32 4.00 ( 4.00 22.25 ( 2.92 7.75 ( 2.06* 81.13 ( 22.22* 23.00 ( 8.01* 6.75 ( 1.69* 36.88 ( 12.53* 2.25 ( 1.31 65.13 ( 17.15* 2.00 ( 1.17 35.13 ( 9.38* 7.75 ( 1.49* 281.38 ( 119.09 2.75 ( 1.36* 97.25 ( 24.74* 2.00 ( 1.17 1.25 ( 0.75 1.13 ( 0.66 6.00 ( 1.74 5.50 ( 1.04* 9.13 ( 2.35

1.07 ( 0.52* 3.86 ( 1.93* 9.50 ( 2.55* 623.71 ( 208.08* 107.71 ( 31.44* 0*# 725.93 ( 254.90* 24.79 ( 4.29* 119.21 ( 42.75* 27.07 ( 7.35* 841.86 ( 240.05* 638.50 ( 198.24* 36.71 ( 8.00* 58.00 ( 20.71* 76.86 ( 26.52* 5.86 ( 2.12* 0* 7.79 ( 2.26* 161.93 ( 58.82* 0* 246.21 ( 75.04* 10.57 ( 2.55* 45.36 ( 13.00* 7.64 ( 2.27 167.07 ( 63.47 15.79 ( 3.60 5.29 ( 0.86 107.21 ( 49.95 0# 98.43 ( 42.29 0# 51.86 ( 24.16 42.14 ( 25.50 38.79 ( 20.42# 1.00 ( 0.72 51.43 ( 19.49 0# 0# 0# 2.64 ( 0.54# 2.50 ( 1.92 2.07 ( 0.97#

a The number of total peptides for each protein is listed as mean ( SEM. * P < 0.05, ** P < 0.01, *** P < 0.001 compared with NDM. # P < 0.05 compared with noDR. IPI ID, International Protein Index identifier.

unique peptides and immunoreactivity, we demonstrated increased levels of CA-I, angiotensinogen, and PEDF in vitreous from PDR subjects compared with vitreous from NDM subjects.5 These differences were confirmed by spectral count (Table 2). In the present study, we identified 56 proteins that

were differentially abundant in vitreous from noDR and PDR groups as compared with the NDM group (Table 2). Among them, the levels of 32 proteins were increased in PDR vitreous compared with NDM samples. Angiotensinogen was increased in PDR vitreous compared with noDR and NDM samples. Ig Journal of Proteome Research • Vol. 7, No. 6, 2008 2519

research articles

Figure 3. Comparison of proteins abundance in noDR or PDR vitreous relative to NDM vitreous. Ratio of the mean total peptides detected in noDR or PDR groups relative to the NDM group. The absence of protein detection in a group is indicated by >20-fold. Spectral count and significant differences (P < 0.05) are listed in Table 2.

kappa chain V-IV region Len, histidine-rich glycoprotein, alpha-1B-glycoprotein, and alpha-1-antichymotrypsin were increased in both PDR and noDR groups compared with the NDM group. Ten proteins were either not detected in PDR or decreased compared with NDM vitreous, including spondin1, protease serine 3, and steerin3. Twelve proteins were decreased in PDR compared with noDR vitreous, including neuroserpin and interphotoreceptor retinoid-binding protein, extracellular superoxide dismutase [Cu-Zn], interphotoreceptor matrix proteoglycan 2, and calsyntenin-1. We compared the fold difference in vitreous proteins in PDR and noDR samples compared with the NDM group. Forty-four proteins in the noDR group and 22 proteins in the PDR group were increased between 1- and 5-fold compared with the NDM group. Twenty-one proteins in the PDR group increased greater than 5-fold compared with the NDM group. Proteins showing the largest fold increases in the PDR group included plasma retinol-binding protein (RBP4, 25.8-fold), monocyte differentiation antigen (CD14, 17.8-fold), alpha-1-microglobulin (AMBP, 17.4-fold), alpha-1-antichymotrypsin (SERPINA3, 13.7-fold), and Chitinase-3-like protein 1 (CHI3L1, 11.5-fold). Eight proteins increased by more than 5-fold in the noDR group compared with the NDM group. Four and 13 proteins appeared less abundant in the noDR and PDR groups, respectively, as compared with the NDM group (Figure 3). Bioinformatic Analysis of Human Vitreous Proteome. To further characterize the vitreous proteome and the changes associated with diabetes and PDR, proteins in Supporting Information Table were analyzed by GO annotation and DAVID functional annotation. Protein subcellular localization was predicted by TargetP, MultiLoc, and SubLoc. Among the total 252 proteins, 233 (92%) were associated with at least one GO identifier. GO annotations were analyzed using the GO Slims list for “GOA and whole proteome analysis” provided by Gene Ontology. The most represented terms in the biological process category were “physiological process”, “response to stimulus”, “metabolism”, and “transport”. The protein numbers in these groups were increased in noDR and PDR groups compared with NDM group by 1.44-, 1.49-, 1.55-, and 1.47-fold, respectively. Analysis of molecular function terms distribution showed that the most represented terms were “binding”, “protein 2520

Journal of Proteome Research • Vol. 7, No. 6, 2008

Gao et al. binding”, “catalytic activity”, “enzyme regulator activity”, and “transporter activity” (Figure 4A). The proteins which have functions of binding, catalytic activity, hydrolase activity, and transporter activity increased in the PDR group compared with the NDM group. The analysis of cellular component and subcellular localization prediction shows that 38%, 44%, 66%, and 44% proteins are annotated or predicted as extracellular proteins or secretory proteins by GO term (Figure 4A), MultiLoc (Figure 4B), TargetP (Figure 4C), and SubLoc (Figure 4D), respectively. We also found more intracellular proteins in noDR and PDR samples compared with NDM samples by both GO annotation and subcellular prediction analysis. IPI identifiers were converted to gene symbols and a total of 234 genes were submitted to DAVID for KEGG-pathways analysis. This analysis revealed that the vitreous proteome contains multiple pathways with more than 10 protein matches, including the complement and coagulation cascades, extracellular matrix (ECM)-receptor interaction, local adhesion, and cell communication. The most striking match between the vitreous proteome and KEGG pathway was observed with the complement and coagulation cascades, which were represented by 30 proteins in the vitreous. Eight of these proteins were increased in the PDR group compared with the NDM group, including coagulation factor XII, complement C3, complement C9, prothrombin, antithrombin III, R1-antitrypsin, complement factor B, and complement factor I (Table 3). Protein Modification Analysis. We identified 7 proteins (cadherin-2, clusterin, vitronectin, antithrombin III, fibrinogen alpha chain, interphotoreceptor retinoid-binding protein, osteopontin) with predicted molecular weights >35 kDa based on amino acid sequence but with apparent molecular weight 0.05). We also observed a trend for increased levels of interalpha-trypsin inhibitor heavy chain, zinc-alpha-2-glycoprotein, apolipoprotein H (beta-2-glycoprotein 1), fibrinogen A, and complement C9 in PDR patients, which are consistent with previous reports. Quantitative proteomics approaches based on the number of peptide-spectral matches have been widely applied on large-scale and complex samples.5,22–24 Since the dynamic range of unique peptide count is limited according to the number of available potential peptides that can be generated by a given sequence, we used spectral count,26 as a relative quantitative method. Previous studies have shown that the total protein concentration in the vitreous is increased in diabetic retinopathy,27 which is likely due to loss of retinal endothelial barrier function. The most abundant protein in vitreous is albumin, which is 40% of total protein and increased 4.4-fold in PDR patients. Since changes in total vitreous protein content is likely an important characteristic of diabetic retinopathy, we have utilized volume of vitreous analyzed, rather that total protein concentration, to normalize across samples. Normalizing to total protein concentration, which is mainly affected by serum albumin and other abundant proteins, would alter the evaluation of low-abundance proteins and could lead to an overestimate of the number of down-regulated proteins in PDR. We found 56 proteins that were differentially detected 2522

Journal of Proteome Research • Vol. 7, No. 6, 2008

among the NDM, noDR and PDR groups, including 37 proteins that were either increased or decreased in the PDR group compared with the NDM group. From the results of relative quantitative analysis and comparative bioinformatic analysis, we found several protein categories that might be related to the pathogenesis of PDR. Intracellular Proteins. Intracellular proteins were increased in the vitreous from noDR and PDR patients according to both quantitative and bioinformatic analysis. Many of the proteins exclusively detected in noDR and PDR are annotated as intracellular proteins and abundant in erythrocytes,28 including carbonic anhydrase-I (CA-I), hemoglobin-R1, -δ, peroxiredoxin1, and catalase. The detection of CA-I and hemoglobin chains in the noDR group was primarily due to their appearance in a single sample, whereas all 7 PDR samples contained these proteins. These results suggest that intraocular hemorrhage in PDR significantly alters the vitreous proteome. In addition, erythrocyte lysis in PDR vitreous likely contributes to the appearance of CA-I and CA-II, which we have recently shown to increase retinal vascular permeability.5 These findings suggest that bleeding into the vitreous, erythrocyte lysis, and the release of intracellular erythrocyte components, including iron, heme, and carbonic anhydrase, may contribute to the pathogenesis of PDR.

Proteomic Analysis of Human Vitreous with PDR

research articles

Figure 5. Identification of protein fragments in the vitreous. (A) Schematic of peptide coverage for proteins with a lower than predicted molecular weight. The location of peptides identified is indicated. (B and C) The spectral count for cadherin-2 and vitronectin on SDSPAGE at different molecular weight from gel slices.

Proteins Response to Stimulus. Proteins annotated as “response to stimulus” displayed some of the largest fold increases in the PDR vitreous compared with NDM samples. Plasma retinol-binding protein (increased by 25.8-fold in PDR) is a plasma protein previously reported to be associated with insulin resistance and the metabolic syndrome.29 Monocyte differentiation antigen (increased by 17.8-fold) is a cell surface protein on monocytes that contributes to the innate immune response, leading to NF-kappa-B activation, cytokine secretion and the inflammatory response. Alpha-1-microglobulin (increased by 17.4-fold) is a secreted protease inhibitor for trypsin and plasmin and present in plasma, urine, and cerebrospinal fluid. Alpha-1-antichymotrypsin (increased by 13.7-fold) is a serpin family serine protease inhibitor. Like the related alpha1-antitrypsin, its concentration increases in the acute phase of inflammation or infection. GP-39 (YKL-40, increased by 11.5fold) is an inflammatory marker with relation to inflammation, cancer and with a reported role in extracellular remodeling and angiogenesis. GP-39 was recently shown correlated with insulin resistance and was elevated in patients with type II diabetes. GP-39 expression has been reported in human retinal pigment epithelium (RPE),30 neural retina and in the RPE-choroid complex, where it has been implicated in contributing to agerelated macular degeneration.31 Many of the vitreous proteins associated with “response to stimulus” have been associated with inflammation. These findings suggest points in which proinflammatory pathways may affect the pathogenesis of PDR. Oxidoreductase Activity. Oxidative stress has been implicated in the pathogenesis of diabetic retinopathy.32 GO analysis

identified 4 enzymes with oxidoreductase catalytic activity. Analysis of PDR vitreous revealed increased levels of peroxiredoxin-1 and decreased levels of extracellular superoxide dismutase [Cu-Zn] compared with noDR or NDM samples. In addition, a trend for increased glutathione peroxidase 3 and catalase levels was detected in PDR vitreous, suggesting that PDR may alter intravitreal peroxide and superoxide metabolism via multiple mechanisms. Further analyses of the relative abundance and activities of these enzymes will be required to determine their contributions in intraocular oxidative stress. Plasma Proteins. Plasma proteins such as transporter proteins, enzymes, protease inhibitors, complement and coagulation cascade increased moderately compared with intracellular and acute phase proteins. Most of the proteins increased in PDR and noDR are plasma proteins. Vitreous albumin concentration was previously shown to be increased in PDR (1.6 g/L) compared with controls (0.3 g/L).27 This increase is comparable with our proteomic data, which showed increases in noDR, and further increases (4.4-fold) in PDR compared with NDM vitreous. Therefore, it is likely that many of the proteins that appear increased in the vitreous of noDR and PDR groups compared with the NDM group could be due to increased permeability of retinal vessels and hemorrhage. Our data provides insight on the extent of plasma protein content in the vitreous fluid. Additional transport and cargo proteins that were increased in PDR vitreous compared with NDM vitreous included plasma retinal binding protein, apoliporoteins C-III, A-I, and A-II, transthyretin, vitamin D-binding protein, and afamin. We identified 30 proteins that are associated with the Journal of Proteome Research • Vol. 7, No. 6, 2008 2523

research articles kallikrein-kinin system, complement and coagulation cascade in the vitreous and levels of several components of these systems such as complement C3, complement factor I, prothrombin, alpha-1-antitrypsin and antithrombin III were found to be increased in PDR patients relative to control subjects. Activation of kallikrein-kinin system, complement and coagulation cascade can both compound and initiate thrombosis, leukostasis and inflammation, all of which are processes involved in vascular lesions of DR. One of the primary roles of the blood-retinal barrier is to separate the vitreous and interstitial fluid of neuroretina from factors in the plasma. The effects of plasma protein infiltration into the vitreous in PDR on retinal physiology are poorly understood. The identification of plasma proteins in the vitreous may suggest novel mechanism and therapeutic targets for the treatment of PDR. Extracellular Matrix and Adhesion Proteins. Extracellular matrix (ECM) and adhesion proteins are detected in vitreous. While the viscous composition of the vitreous fluid has been attributed to an abundance of proteoglycan, collagen glycoproteins, and hyaluronan,33 the specific components of this ECM has not yet been described using mass spectrometrybased proteomics. We detected collagen type I-R1 and -R2, type II-R1, type IV-R5, type V-R1 and -R3, type VII-R1, and type XVIII-R1. We also detected multiple proteoglycans, including interphotoreceptor matrix proteoglycan -1 and -2, chondroitin sulfate proteoglycan 2, biglycan, lumican, basement membranespecific heparan sulfate proteoglycan core protein (heparin sulfate proteoglycan 2, perlecan), and glycoproteins and matrix binding proteins, including fibronectin, fibulin 1, galectin-3binding protein and opticin. Heparan sulfate proteoglycan 2 (increased by 11.4-fold in PDR) possesses angiogenic properties and promotes tumor angiogenesis by inducing high affinity binding of FGF-2 to heparin sulfate deficient cells or to soluble FGF receptors.34 Cell adhesion proteins, including thrombospondin 4, spondin 1, and dystroglycan, were also detected in human vitreous. Since changes in vitreous gel structure may contribute to posterior vitreous detachment, retinal tears, retinal detachment, vitreous hemorrhage, intravitreal drug pharmacokinetics, and intraocular angiogenic responses, the further analysis of these components may provide insight into vitreoretinal disease associated with DR. In conclusion, we identified 252 proteins in human vitreous, including 37 proteins that were increased or decreased in the PDR group compared with the NDM group. Bioinformatic analysis revealed that the vitreous contains multiple components of the kallikrein-kinin system, and complement and coagulation cascade, indicating the presence of biological process pathways in the vitreous. Many of the proteins identified in the vitreous are annotated with catalytic activity or enzyme regulator activity, suggesting effects on extracellular metabolic and proteolytic pathways. Moreover, an abundance of intracellular proteins were detected in the vitreous from PDR subjects, suggesting cell lysis contributes to the vitreous proteome. Further understanding of the biological processes mediated by the vitreous proteome may provide new insights into novel mechanisms and therapeutic targets for proliferative diabetic retinopathy.

Acknowledgment. The authors thank Dr. Paul G. Arrigg and Dr. Robert L. Avery for providing vitreous samples and Susan Rook for technical assistance. This work was supported in part by the U.S. National Institutes of Health (grants DK 60165, DK 36836), the Juvenile Diabetes Research 2524

Journal of Proteome Research • Vol. 7, No. 6, 2008

Gao et al. Foundation (1-2005-1047), and the Massachusetts Lions Eye Research Fund. Dr. Ben-Bo Gao is a recipient of a Mary K. Iacocca Fellowship.

Supporting Information Available: Vitreous proteome in patients with NDM, noDR or PDR (Supporting Information Table). This Excel table contains the total list of IPI, protein name, gene symbol, sequence coverage, unique peptides, spectral count of NDM, noDR, and PDR group (Mean ( SEM), p-value, search engine, GO term (biological process, cellular component, molecular function), and protein subcellular localization prediction for the 252 proteins identified in this study. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Kempen, J. H.; O’Colmain, B. J.; Leske, M. C.; Haffner, S. M.; Klein, R.; Moss, S. E.; Taylor, H. R.; Hamman, R. F. The prevalence of diabetic retinopathy among adults in the United States. Arch. Ophthalmol. 2004, 122 (4), 552–63. (2) Early Treatment Retinopathy Study Research Group Grading diabetic retinopathy from stereoscopic color fundus photographss an extension of the modified Airlie House classification. ETDRS report number 10. Early Treatment Diabetic Retinopathy Study Research Group. Ophthalmology 1991, 98, (5 Suppl.), 786-806. (3) Mohamed, Q.; Gillies, M. C.; Wong, T. Y. Management of diabetic retinopathy: a systematic review. JAMA 2007, 298 (8), 902–16. (4) Aiello, L. P.; Avery, R. L.; Arrigg, P. G.; Keyt, B. A.; Jampel, H. D.; Shah, S. T.; Pasquale, L. R.; Thieme, H.; Iwamoto, M. A.; Park, J. E.; et al. Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N. Engl. J. Med. 1994, 331 (22), 1480–7. (5) Gao, B. B.; Clermont, A.; Rook, S.; Fonda, S. J.; Srinivasan, V. J.; Wojtkowski, M.; Fujimoto, J. G.; Avery, R. L.; Arrigg, P. G.; Bursell, S. E.; Aiello, L. P.; Feener, E. P. Extracellular carbonic anhydrase mediates hemorrhagic retinal and cerebral vascular permeability through prekallikrein activation. Nat. Med. 2007, 13 (2), 181–8. (6) Limb, G. A.; Chignell, A. H. Vitreous levels of intercellular adhesion molecule 1 (ICAM-1) as a risk indicator of proliferative vitreoretinopathy. Br. J. Ophthalmol. 1999, 83 (8), 953–6. (7) Yamane, K.; Minamoto, A.; Yamashita, H.; Takamura, H.; Miyamoto-Myoken, Y.; Yoshizato, K.; Nabetani, T.; Tsugita, A.; Mishima, H. K. Proteome analysis of human vitreous proteins. Mol. Cell. Proteomics 2003, 2 (11), 1177–87. (8) Garcia-Ramirez, M.; Canals, F.; Hernandez, C.; Colome, N.; Ferrer, C.; Carrasco, E.; Garcia-Arumi, J.; Simo, R. Proteomic analysis of human vitreous fluid by fluorescence-based difference gel electrophoresis (DIGE): a new strategy for identifying potential candidates in the pathogenesis of proliferative diabetic retinopathy. Diabetologia 2007, 50 (6), 1294–1303. (9) Koyama, R.; Nakanishi, T.; Ikeda, T.; Shimizu, A. Catalogue of soluble proteins in human vitreous humor by one-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electrospray ionization mass spectrometry including seven angiogenesis-regulating factors. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2003, 792 (1), 5–21. (10) Nakanishi, T.; Koyama, R.; Ikeda, T.; Shimizu, A. Catalogue of soluble proteins in the human vitreous humor: comparison between diabetic retinopathy and macular hole. J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 2002, 776 (1), 89–100. (11) Ouchi, M.; West, K.; Crabb, J. W.; Kinoshita, S.; Kamei, M. Proteomic analysis of vitreous from diabetic macular edema. Exp. Eye Res. 2005, 81 (2), 176–82. (12) Kim, S. J.; Kim, S.; Park, J.; Lee, H. K.; Park, K. S.; Yu, H. G.; Kim, Y. Differential expression of vitreous proteins in proliferative diabetic retinopathy. Curr. Eye Res. 2006, 31 (3), 231–40. (13) Kim, T.; Kim, S. J.; Kim, K.; Kang, U. B.; Lee, C.; Park, K. S.; Yu, H. G.; Kim, Y. Profiling of vitreous proteomes from proliferative diabetic retinopathy and nondiabetic patients. Proteomics 2007, 7 (22), 4203–15. (14) Harris, M. A.; Clark, J.; Ireland, A.; Lomax, J.; Ashburner, M.; Foulger, R.; Eilbeck, K.; Lewis, S.; Marshall, B.; Mungall, C.; Richter, J.; Rubin, G. M.; Blake, J. A.; Bult, C.; Dolan, M.; Drabkin, H.; Eppig, J. T.; Hill, D. P.; Ni, L.; Ringwald, M.; Balakrishnan, R.; Cherry, J. M.; Christie, K. R.; Costanzo, M. C.; Dwight, S. S.; Engel, S.; Fisk, D. G.; Hirschman, J. E.; Hong, E. L.; Nash, R. S.; Sethuraman, A.; Theesfeld, C. L.; Botstein, D.; Dolinski, K.; Feierbach, B.; Berardini,

research articles

Proteomic Analysis of Human Vitreous with PDR

(15) (16)

(17) (18)

(19) (20)

(21)

(22) (23)

(24)

T.; Mundodi, S.; Rhee, S. Y.; Apweiler, R.; Barrell, D.; Camon, E.; Dimmer, E.; Lee, V.; Chisholm, R.; Gaudet, P.; Kibbe, W.; Kishore, R.; Schwarz, E. M.; Sternberg, P.; Gwinn, M.; Hannick, L.; Wortman, J.; Berriman, M.; Wood, V.; de la Cruz, N.; Tonellato, P.; Jaiswal, P.; Seigfried, T.; White, R. The Gene Ontology (GO) database and informatics resource. Nucleic Acids Res. 2004, 32 (Database issue), D258–61. Elias, J. E.; Gygi, S. P. Target-decoy search strategy for increased confidence in large-scale protein identifications by mass spectrometry. Nat. Methods 2007, 4 (3), 207–14. Camon, E.; Barrell, D.; Lee, V.; Dimmer, E.; Apweiler, R. The Gene Ontology Annotation (GOA) Database-an integrated resource of GO annotations to the UniProt Knowledgebase. In Silico Biol. 2004, 4 (1), 5–6. Emanuelsson, O.; Nielsen, H.; Brunak, S.; von Heijne, G. Predicting subcellular localization of proteins based on their N-terminal amino acid sequence. J. Mol. Biol. 2000, 300 (4), 1005–16. Hoglund, A.; Donnes, P.; Blum, T.; Adolph, H. W.; Kohlbacher, O. MultiLoc: prediction of protein subcellular localization using N-terminal targeting sequences, sequence motifs and amino acid composition. Bioinformatics 2006, 22 (10), 1158–65. Hua, S.; Sun, Z. Support vector machine approach for protein subcellular localization prediction. Bioinformatics 2001, 17 (8), 721–8. Dennis, G., Jr.; Sherman, B. T.; Hosack, D. A.; Yang, J.; Gao, W.; Lane, H. C.; Lempicki, R. A. DAVID: Database for Annotation, Visualization, and Integrated Discovery. GenomeBiology 2003, 4 (5), P3. Balgley, B. M.; Laudeman, T.; Yang, L.; Song, T.; Lee, C. S. Comparative evaluation of tandem MS search algorithms using a target-decoy search strategy. Mol. Cell. Proteomics 2007, 6 (9), 1599–608. Pang, J. X.; Ginanni, N.; Dongre, A. R.; Hefta, S. A.; Opitek, G. J. Biomarker discovery in urine by proteomics. J. Proteome Res. 2002, 1 (2), 161–9. Blondeau, F.; Ritter, B.; Allaire, P. D.; Wasiak, S.; Girard, M.; Hussain, N. K.; Angers, A.; Legendre-Guillemin, V.; Roy, L.; Boismenu, D.; Kearney, R. E.; Bell, A. W.; Bergeron, J. J.; McPherson, P. S. Tandem MS analysis of brain clathrin-coated vesicles reveals their critical involvement in synaptic vesicle recycling. Proc. Natl. Acad. Sci. U.S.A. 2004, 101 (11), 3833–8. Gilchrist, A.; Au, C. E.; Hiding, J.; Bell, A. W.; Fernandez-Rodriguez, J.; Lesimple, S.; Nagaya, H.; Roy, L.; Gosline, S. J.; Hallett, M.; Paiement, J.; Kearney, R. E.; Nilsson, T.; Bergeron, J. J. Quantitative proteomics analysis of the secretory pathway. Cell 2006, 127 (6), 1265–81.

(25) Duh, E. J.; Yang, H. S.; Haller, J. A.; De Juan, E.; Humayun, M. S.; Gehlbach, P.; Melia, M.; Pieramici, D.; Harlan, J. B.; Campochiaro, P. A.; Zack, D. J. Vitreous levels of pigment epithelium-derived factor and vascular endothelial growth factor: implications for ocular angiogenesis. Am. J. Ophthalmol. 2004, 137 (4), 668–74. (26) Turck, C. W.; Falick, A. M.; Kowalak, J. A.; Lane, W. S.; Lilley, K. S.; Phinney, B. S.; Weintraub, S. T.; Witkowska, H. E.; Yates, N. A. The Association of Biomolecular Resource Facilities Proteomics Research Group 2006 study: relative protein quantitation. Mol. Cell. Proteomics 2007, 6 (8), 1291–8. (27) Fosmark, D. S.; Bragadottir, R.; Stene-Johansen, I.; Berg, J. P.; Berg, T. J.; Lund, T.; Sandvik, L.; Hanssen, K. F. Increased vitreous levels of hydroimidazolone in type 2 diabetes patients are associated with retinopathy: a case-control study. Acta Ophthalmol. Scand. 2007, 85 (6), 618–22. (28) Pasini, E. M.; Kirkegaard, M.; Mortensen, P.; Lutz, H. U.; Thomas, A. W.; Mann, M. In-depth analysis of the membrane and cytosolic proteome of red blood cells. Blood 2006, 108 (3), 791–801. (29) Graham, T. E.; Yang, Q.; Bluher, M.; Hammarstedt, A.; Ciaraldi, T. P.; Henry, R. R.; Wason, C. J.; Oberbach, A.; Jansson, P. A.; Smith, U.; Kahn, B. B. Retinol-binding protein 4 and insulin resistance in lean, obese, and diabetic subjects. N. Engl. J. Med. 2006, 354 (24), 2552–63. (30) Sharon, D.; Blackshaw, S.; Cepko, C. L.; Dryja, T. P. Profile of the genes expressed in the human peripheral retina, macula, and retinal pigment epithelium determined through serial analysis of gene expression (SAGE). Proc. Natl. Acad. Sci. U.S.A. 2002, 99 (1), 315–20. (31) Rakic, J. M.; Lambert, V.; Deprez, M.; Foidart, J. M.; Noel, A.; Munaut, C. Estrogens reduce the expression of YKL-40 in the retina: implications for eye and joint diseases. Invest. Ophthalmol. Visual Sci. 2003, 44 (4), 1740–6. (32) Kowluru, R. A.; Odenbach, S. Effect of long-term administration of alpha-lipoic acid on retinal capillary cell death and the development of retinopathy in diabetic rats. Diabetes 2004, 53 (12), 3233– 8. (33) Los, L. I.; van der Worp, R. J.; van Luyn, M. J.; Hooymans, J. M. Age-related liquefaction of the human vitreous body: LM and TEM evaluation of the role of proteoglycans and collagen. Invest. Ophthalmol. Visual Sci. 2003, 44 (7), 2828–33. (34) Aviezer, D.; Hecht, D.; Safran, M.; Eisinger, M.; David, G.; Yayon, A. Perlecan, basal lamina proteoglycan, promotes basic fibroblast growth factor-receptor binding, mitogenesis, and angiogenesis. Cell 1994, 79 (6), 1005–13.

PR800112G

Journal of Proteome Research • Vol. 7, No. 6, 2008 2525