Memoria de investigación y docencia. AÑO 2014 · COMUNICACIONES AL CONGRESO ARVO Differential proteomics study in serum of patients with primary openangle glaucoma and pseudoexfoliation glaucoma M. Coca-Prados1,2, L. Álvarez1, M. García1, J. Escribano3, P. P. Rodríguez-Calvo1, L. Fernández-Vega1, and H. González-Iglesias1 YALE UNIVERSITY School of Medicine 1Fundación de Investigación Oftalmológica. Instituto Oftalmológico Fernández-Vega, Oviedo, Asturias, Spain. & Visual Science, Yale University School of Medicine, New Haven, CT, United States. 3Laboratorio de Genética Molecular Humana, Universidad de Castilla-La Mancha, Albacete, Castilla-La Mancha, Spain. 2357 - B0033 2Ophthalmology INTRODUCTION RESULTS Primary Open-Angle Glaucoma (POAG) and Pseudoexfoliation Glaucoma (PEXG) are among the most prevalent types of glaucoma in developed countries. Investigations conducted in the field of differential proteomics in search for molecular biomarkers for glaucoma diagnostic and management, in readily available biological fluids, have been limited [1]. In the present work, we carried out a comparative differential proteomic analysis of blood serum from patients with POAG, PEXG, and healthy controls. The aim was to identify proteins quantitatively altered, which are significantly discriminatory between glaucoma patients and healthy subjects [2]. STUDY DESIGN Study population Protein equalization 1.  Bhattacharya SK, Lee RK, Grus FH, Seventh ARVO/Pfizer Ophthalmics Research Institute Conference Working Group. Invest. Ophthalmol. Vis. Sci. 2013, 54(1):121-131. 2.  González-Iglesias H, Álvarez L, García M, Escribano J, Rodríguez-Calvo PP, Fernández-Vega L, Coca-Prados M, J. Proteomics. 2014, 98:65-78. Q14624 P01008 P02787 P02768 P04004 P01859 P02743 P01009 P06727 O14791 P0C0L4 O75636 P02647 P02766 2D-DIGE Verification on newly recruited subjects Protein name Complement factor H Fibulin-1 Complement C3 Inter-alpha-trypsin inhibitor heavy chain H4 Antithrombin-III Serotransferrin Serum albumin Vitronectin Ig gamma-2 chain C region Serum amyloid P-component Alpha-1 antitrypsin Apolipoprotein A-IV Apolipoprotein L1 Complement C4-A Ficolin-3 Apolipoprotein A-I Transthyretin Gene symbol No of spots (spot numbers) CFH 3 (30, 1127, 31, 36) FBLN1 1 (55) C3 5 (100, 1153, 113, 1128, 1145) ITIH4 SERPINC1 TF ALB VTN IGHG2 APCS SERPINA1 APOA4 APOL1 C4A FCN3 APOA1 TTR New cohort’s patient recruitment 5 (192, 593, 549, 617, 506) 10 (297, 281, 287, 507, 565, 745, 1907, 739, 423, 422) 4 (274, 373, 360, 346) 10 (462, 467, 446, 438, 436, 447, 2812, 433, 358, 291) 11 (564, 561, 651, 1086, 3550, 1109, 557, 536, 652, 661, 658) 1 (627) 4 (746, 1611, 1908, 1840) 3 (749, 734, 740)) 1 (959) 4 (1045, 964, 941, 960) 2 (1308, 1375) 3 (1359, 1372, 3556) 4 (1911, 1906, 1917, 1920) 1 (2768) Altered Proteins Identification (MALDI TOF/TOF and nLC-MS/MS) Multiple Reaction Monitoring-Mass Spectrometry (MRM-MS) n=53 2D-DIGE ELISA Quantification Mascot Score Histogram CFAH_HUMAN Mass: 143680 Score: 110 Expect: 5.3e-06 Matches: 11 Quantification of candidate biomarkers Statistical Analysis Functional Pathways Analysis Candidate markers Diagnostic test for glaucoma screening APOA4, TF, C3, APOL1, IGHG2, C4A C) Product ion M2+ 613.8 A) Selection of the peptide by HPLC-ESI-MS Selected peptide ALDFAVGEYNK ! Sequence is unique ! High sensitivity ! Soluble in water ! Unmodify groups 518.3 ! Labelling 13C2-Glycine B) Scan of the T3 Peptide A) Enzymatic digestion 13C T* 2 3 Previously quantified Reduction S-S, DTT 13C Blocked SH, IAA 0.1 mL serum Urea Buffer Cleaning sample by: 2-peptide Tripsin 37ºC, 3h B) Scan of target Peptide Normal Vision 1) HPLC-RP-UV 2) HPLC-SCX-UV Serum Tryptic peptides C) Multiple Reaction Monitoring (MRM) Glaucoma Figure 1. Study design. Step 1, Differential expression analysis; Step 2, Elisa screening; Step 3, Validation by MRM-MS. pI 4 MW 150- 7 A 30 36 55 60 176 455451 31 110 Ranking UniProt 14 113 100 90 226 274 288 235 276 281 284 287 346 564 571 431 283 192 297 373 397 372 528 566 291 358 536 422 423 533 447 351 658 651 740 749 594 296 360 514 661 433 606 627 739 593 436 699 652 506 544 565 438 446 745 617 734 462 467 507 952 1007967 959 964 666 609 559 645 941 636 1086 1127 851 830 953 746 1055 3544 997 938 958 1145 1045 1109 960 1153 968 1039 966 1184 1282 1128 1219 1214 1211 1379 1329 1212 1468 1375 1483 1308 450 8070- 605 4538- 557 561 1316 13191371 1397 1390 1461 1848 30- 143515371382 1840 1611 1906 1920 1911 1917 1907 3560 2708 153550 2768 1406 1359 1908 2167 1372 1846 3559 2760 P08603 P23142 P01024 Q14624 5 P01008 Fibulin-1 Complement C3 Inter-alpha-trypsin inhibitor heavy chain H4 Antithrombin-III 6 7 P02787 P02768 Serotransferrin Serum albumin 8 P04004 Vitronectin 19 20 2812 3552 21 22 23 24 kDa B 25 26 27 Figure 2. A.- 2D-DIGE analysis of serum proteome. Equalized serum samples (53 POAG, 45 PEXG, 51 Control) were analyzed by 2D-DIGE, resulting in 149 significantly altered spots, marked in the gel and subsequently identified. B.- Principal Component Analysis (PCA). Projection on the axes of the most significant spots for each group. ! Protein name Complement factor H 1 2 3 4 9 10 11 12 13 14 15 16 17 18 3556 Overall, this work identified a panel of candidates for glaucoma biomarkers that they are part of a network linked to regulating immune and inflammatory-related processes. This study provides a new basis to validate the identified proteins as biomarkers of glaucoma in a large-scale-multiplexed screening in serum. The data from this study offers new perspectives in the discovery of glaucoma biomarkers in serum of POAG and PEXG patients. REFERENCES 4 5 6 7 8 9 10 11 12 13 14 15 16 17 serum blood Compression of the dynamic range of proteins serum RESULTS Ranking UniProt 1 P08603 2 P23142 3 P01024 n>600 Sample collection/ Serum preparation Equalization of serum proteins with ProteoMinerTM, two-dimensional fluorescent difference gel electrophoresis (2D-DIGE), matrix assisted laser desorption/ionization time-of-flight/time of flight MALDI-TOF/TOF, nano-liquid-chromatography tandem mass spectrometry (nLC-MS-MS), functional analysis (Ingenuity® Systems, www.ingenuity.com), and enzyme-linked immunosorbent assay (ELISA). A total of 202 blood samples from subjects with POAG (n=73), PEXG (n=59), and controls (cataract, n=70) were analyzed following the proteomic workflow showed in Figure 1. In the first step, after 2D-DIGE analysis of n=149 equalized sera, we established the identity of 149 significantly altered spots resulting in the identification of 35 distinct proteins. The proteomic workflow implemented revealed that these proteins were altered in patients with glaucoma relative to healthy controls. The functional analysis identified the signaling network of these proteins correlated to an immunological and inflammatory pathway. In the second step, the top-17-ranked proteins of the 35 panel were analyzed by ELISA on n=53 non-equalized sera samples from newly recruited subjects (POAG, PEXG and control), and confirmed their alteration in individual protein levels. Stepwise discriminant analysis of the ELISA data resulted in a six-protein panel (APOA4, TF, C3, APOL1, IGHG2, C4A) that was used to generate multivariate predictive models. These models achieved a diagnostic efficiency of 81% (correct assignment), specificity of 83%, and sensitivity of 92%. Specifically, 100% efficacy was reached in the discrimination of the POAG group from control, and 86% efficacy in the discrimination of POAG from PEXG. In the ongoing third step, the clinical validation of candidate biomarkers is being carried out at large scale by means of Multiple Reaction Monitoring-based Mass Spectrometry (MRM-MS). Potential Candidate Markers n=149 Step 3: Validation (ongoing) Step 2: Elisa screening Step 1: Differential expression analysis MATERIALS & METHODS CONCLUSIONS P01859 Ig gamma-2 chain C region P02743 Serum amyloid P-component P01009 Alpha-1 antitrypsin P06727 Apolipoprotein A-IV O14791 Apolipoprotein L1 P0C0L4 Complement C4-A O75636 Ficolin-3 P02647 Apolipoprotein A-I P02766 Transthyretin P27169 Serum paraoxonase/ arylesterase 1 Q15485 Ficolin-2 Q03591 Complement factor H-related protein 1 P02748 Complement component C9 P22352 Glutathione peroxidase 3 P01860 Ig gamma-3 chain C region Q9BXR6 Complement factor H-related protein 5 Q6UX53 Methyltransferase-like protein 7B P02774 Vitamin D-binding protein P06681 Complement C2 Figure 3. Functional pathway analysis (IPA). The highest scoring network (Antigen presentation, humoral immune response, inflammatory response; score 52) has been obtained from the 35 proteins identified. Gene symbol No of spots (spot numbers) CFH FBLN1 C3 ITIH4 3 (30, 1127, 31, 36) 1 (55) 5 (100, 1153, 113, 1128, 1145) 5 (192, 593, 549, 617, 506) SERPINC1 10 (297, 281, 287, 507, 565, 745, 1907, 739, 423, 422) 4 (274, 373, 360, 346) 10 (462, 467, 446, 438, 436, 447, 2812, 433, 358, 291) VTN 11 (564, 561, 651, 1086, 3550, 1109, 557, 536, 652, 661, 658) IGHG2 1 (627) APCS 4 (746, 1611, 1908, 1840) SERPINA1 3 (749, 734, 740)) APOA4 1 (959) APOL1 4 (1045, 964, 941, 960) C4A 2 (1308, 1375) FCN3 3 (1359, 1372, 3556) APOA1 4 (1911, 1906, 1917, 1920) TTR 1 (2768) PON1 3 (952, 967, 968) TF ALB FCN2 CFHR1 2 (1219, 1184) 2 (958, 953) C9 GPX3 IGHG3 CFHR5 4 (528, 533, 283, 514) 1 (2167) 1 (636) 1 (397) METTL7B 1 (645) GC C2 1 (666) 1 (699) 1 (830) 28 P19823 P00734 Inter-alpha-trypsin inhibitor heavy chain H2 Prothrombin ITIH2 29 30 31 P02649 P02741 Apolipoprotein E C-reactive protein APOE CRP 32 P10909 Clusterin CLU 33 P01871 Ig mu chain C region IGHM 11 (451, 1211, 1319, 455, 450, 1212, 1316, 1282, 1371, 1390, 1397) 1 (351) 34 35 P01834 Ig kappa chain C region O00187 Mannan-binding lectin serine protease 2 IGKC MASP2 1 (1846) 1 (3560) F2 7 (966, 1039, 1382, 1406, 276, 288, 296) 5 (1379, 1435, 1468, 1483, 1537) 1 (1848) Table 1. Identification of the most-differentially-altered proteins. The identity of the 149 spots shown in Figure 2 was established, resulting in the identification of 35 distinct proteins differentially expressed between the three sample groups. SUPPORTED BY: CENIT-20091021 6 FACULTAD DE MEDICINA Gene name POAG group mg/dL APOA4 9.14 ± 3.01 C3 422.27 ± 62.40 TTR 152.25 ± 47.75 TF 445.86 ± 181.71 VTN 103.89 ± 40.52 FBLN1 26.61 ± 7.89 APOA1 234.04 ± 48.58 SERPINA1 289.65 ± 65.72 CFH 121.45 ±27.81 APOL1 8.47 ± 3.05 FCN3 114.74 ± 26.64 IGHG2 826.84 ± 265.30 ITIH4 543.34 ± 275.33 ALB 3695.73 ± 1066.51 SERPINC1 76.41 ± 10.80 C4A 0.15 ± 0.05 APCS 15.25 ± 9.00 PEXG group mg/dL 5.18 ± 1.55 311.28 ± 36.95 98.32 ± 40.59 289.62 ± 56.89 74.93 ± 23.50 21.26 ± 6.76 199.99 ± 28.36 226.98 ± 59.32 102.39 ± 27.62 7.87 ± 3.14 88.79 ± 19.15 800.59 ± 223.03 362.38 ± 189.81 4092.33 ± 954.84 80.39 ± 13.95 0.15 ± 0.05 19.43 ± 13.71 Kruskal-Wallis Test Control group FOLD CHANGE (Non-parametric ANOVA) mg/dL POAG vs CT PEXG vs CT POAG vs PEXG POAG vs CT PEXG vs CT POAG vs PEXG 3.44 ± 0.85 2.7 1.5 1.8 p < 0.001 p > 0.05 p < 0.01 276.40 ± 36.07 1.5 1.1 1.4 p < 0.001 p > 0.05 p < 0.001 82.81 ± 25.60 1.8 1.2 1.5 p < 0.001 p > 0.05 p < 0.01 263.44 ± 57.89 1.7 1.1 1.5 p < 0.001 p > 0.05 p < 0. 01 47.22 ± 12.06 2.2 1.6 1.4 p < 0.001 p < 0.01 p > 0.05 14.10 ± 3.88 1.9 1,5 1,3 p < 0.001 p < 0.05 p > 0.05 183.56 ± 37.04 1.3 1.1 1.2 p < 0.001 p > 0.05 p > 0.05 192.30 ± 58.78 1.5 1.2 1.3 p < 0.001 p > 0.05 p < 0.05 93.60 ± 33.54 1.3 1.1 1.2 p < 0.01 p > 0.05 p > 0.05 5.90 ± 2.28 1.4 1.3 1.1 p < 0.05 p > 0.05 p > 0.05 85.30 ± 21.66 1.3 1.0 1.3 p < 0.01 p > 0.05 p < 0.05 1144.26 ± 402.83 0.7 0.7 1.0 p < 0.05 p < 0.05 p > 0.05 372.13 ± 222.88 1.5 1.0 1.5 p > 0.05 p > 0.05 p > 0.05 3364.20 ± 792.07 1.1 1.2 0.9 p > 0.05 p > 0.05 p > 0.05 81.36 ± 13.06 0.9 1.0 1.0 p > 0.05 p > 0.05 p > 0.05 0.19 ± 0.03 0.8 0.8 1.0 p < 0.05 p > 0.05 p > 0.05 18.42 ± 14.19 0.8 1.1 0.8 p > 0.05 p > 0.05 p > 0.05 Table 2. ELISA Screening. The 17-top-ranked differentially-altered serum proteins were quantified by ELISA on sera samples (non-equalized) from a new population (20 POAG, 14 PEXG and 19 control). The concentrations (mg protein/dL serum, expressed as mean value ± SD) for each of the proteins estimated by ELISA in POAG, PEXG and control groups are shown, as are the fold-change and the Kruskal-Wallis Test (nonparametric ANOVA). Protein panel: C3, APOA4, Method 1 Naive Bayes 2 kNN 3 Classification Tree 4 Random Forest 5 AdaBookst.M1 C4A, TF, IGHG2 and APOL1 Evaluation Results CA Sens 0.7987 0.8214 0.6948 0.8214 0.7078 0.8036 0.7078 0.8571 0.8117 0.8214 Clinical Diagnosis CT CT 33 PEXG 5 POAG 0 Total 38 Error Rate CA (%) 0.19 0.81 PEXG 7 23 2 32 Sens. 0.92 Spec 0.9286 0.8673 0.8571 0.8367 0.9490 AUC 0.9583 0.9228 0.8154 0.9042 0.9210 POAG Total 0 40 7 35 33 35 40 110 Spec. POAG vs PEXG 0.83 0.86 Table 3. Machine learning models. Stepwise discriminant analysis was performed, resulting in a six protein panel: C3, APOA4, C4A, TF, IGHG2 and APOL1. Multivariate predictive models were generated, and the best classification efficiency was obtained training the data with Naive Bayes with area under the curve (AUC) of 96%. Table 4. Confusion matrix by means of Random Naive Bayes algorithm when comparing CT, PEXG and POAG groups. Each column corresponds to a correct class, and rows represent the predicted classes. Diagnostic efficiency reached 81% (correct assignment, CA), and a specificity (Spec.) of 83% and sensitivity (Sens.) of 92% were obtained.