Article

Urinary Angiotensinogen and Risk of Severe AKI

Joseph L. Alge,* Nithin Karakala,*† Benjamin A. Neely,* Michael G. Janech,*† James A. Tumlin,‡ Lakhmir S. Chawla,§ | Andrew D. Shaw, ¶ and John M. Arthur,*† for the SAKInet Investigators

Summary Background Biomarkers of AKI that can predict which patients will develop severe renal disease at the time of diagnosis will facilitate timely intervention in populations at risk of adverse outcomes.

*Medical University Design, setting, participants, & measurements Liquid chromatography/tandem mass spectrometry was used to of South Carolina, identify 30 potential prognostic urinary biomarkers of severe AKI in a group of patients that developed AKI after Charleston, South cardiac surgery. Angiotensinogen had the best discriminative characteristics. Urinary angiotensinogen was Carolina; †Ralph H. subsequently measured by ELISA and its prognostic predictive power was verified in 97 patients who underwent Johnson Veterans cardiac surgery between August 1, 2008 and October 6, 2011. Affairs Medical Center, Charleston, South Carolina; Results The urine angiotensinogen/creatinine ratio (uAnCR) predicted worsening of AKI, Acute Kidney Injury ‡University of Network (AKIN) stage 3, need for renal replacement therapy, discharge .7 days from sample collection, and Tennessee College of Medicine, composite outcomes of AKIN stage 2 or 3, AKIN stage 3 or death, and renal replacement therapy or death. The Chattanooga, prognostic predictive power of uAnCR was improved when only patients classified as AKIN stage 1 at the time of Tennessee; §George urine sample collection (n=79) were used in the analysis, among whom it predicted development of stage 3 AKI or Washington death with an area under the curve of 0.81. Finally, category free net reclassification improvement showed that University, Washington, DC; the addition of uAnCR to a clinical model to predict worsening of AKI improved the predictive power. | Duke University, Durham, North Conclusions Elevated uAnCR is associated with adverse outcomes in patients with AKI. These data are the first to Carolina; and ¶ demonstrate the utility of angiotensinogen as a prognostic biomarker of AKI after cardiac surgery. Durham Veterans Clin J Am Soc Nephrol – Affairs Medical 8: 184 193, 2013. doi: 10.2215/CJN.06280612 Center, Durham, North Carolina

Introduction proteomics to identify prognostic urinary biomarkers Correspondence: One of the most important factors underlying the poor of AKI after cardiac surgery. We performed a verifi- Dr. John Arthur, outcomes seen in AKI is our current method of Division of cation of angiotensinogen in patients who developed Nephrology, diagnosis, which is based on either an increase in AKI after cardiac surgery. This is the first study to Department of serum creatinine (SCr) or decreased urine output (1,2). demonstrate the utility of angiotensinogen as a prog- Medicine, Medical SCr and urine output values at the time of diagnosis nostic biomarker of AKI. University of South are of limited prognostic value, making it difficult to Carolina, 96 Jonathan Lucas Street, PO Box discriminate between mild and severe AKI at the time 250623, Charleston, of diagnosis. The need for better biomarkers of AKI Materials and Methods SC 29425. Email: has been recognized as a crucial barrier to improve- Urine Samples [email protected] ment of the outcomes of AKI patients. Newer bio- Urine samples were obtained from 99 consecutively markers of AKI include kidney injury molecule 1, enrolled patients who had cardiac surgery at one of neutrophil gelatinase-associated lipocalin (NGAL), the SAKInet institutions between August 1, 2008 IL-18, and cystatin C (3–9). Many of these biomarkers and October 6, 2011. Informed consent was obtained initially appeared capable of early, accurate detec- in accordance with the institutional review board– tion of AKI, but subsequent verification studies have approved protocol at each institution. Samples were reported lower accuracy (10–17). In addition, the em- collected and stored using a standard operating pro- phasis on early detection has overshadowed investiga- cedure that included centrifugation, addition of pro- tion of their prognostic predictive power. Available tease inhibitors, and storage at 280°C. Urine samples data on the prognostic value of these biomarkers were collected as early as possible after AKIN serum suggest that they are limited predictors of adverse creatinine criteria were met (2). Inclusion criteria outcomes (18,19). The limitations of biomarkers un- were surgery of the heart or ascending aorta and de- derscore the need to discover new prognostic bio- velopment of AKI within 2 days of surgery. Partici- markers. pants with baseline SCr .3 mg/dl were excluded. Approximately 20% of patients who undergo car- Twelve samples were used in the proteomic studies, diac surgery develop AKI. The timing of the injury can 10 of which were also used in a validation set that be readily determined in these patients (20). We used included samples collected from the remaining 87

184 Copyright © 2013 by the American Society of Nephrology www.cjasn.org Vol 8 February, 2013 Clin J Am Soc Nephrol 8: 184–193, February, 2013 Urinary Angiotensinogen Predicts Worsening AKI, Alge et al. 185

participants in the study. Of the samples used in the val- combination of P values from the Wilcoxon rank-sum idation set, 79 were from patients classified as having test and mean fold-change between the experimental AKIN stage 1 at the time of sample collection. groups. The relationship between these two measures was visualized by “volcano plot.” In verification studies, ’ Proteomic Analyses count data were analyzed using the chi-squared or Fisher s A detailed description of the methods is available in the exact test as appropriate. Continuous variables were ana- t U Supplemental Material. HIV protein gp160 (200 ng; Bio- lyzed using the test or Mann Whitney test. ANOVA or – post hoc clone Inc) was spiked into each urine sample. Proteins Kruskal Wallis ANOVA on ranks test and the ’ were denatured, alkylated, and digested with trypsin. Dunn s test for pairwise comparison were used to evaluate Samples were prefractionated by solid phase extraction continuous variables when more than two groups were using a Strata-X SPE cartridge (Phenomenex). Sample frac- compared. Odds ratios (ORs) were used to test the associ- tions were reconstituted in mobile phase A (MPA) (98% ation of uAnCR with selected outcomes. Patients were stratified by uAnCR into quartiles. The effect of uAnCR H20, 0.1% formic acid; 2% acetonitrile). Five microliters of each fraction was injected onto an Acclaim PepMap100 on the risk of developing an outcome was tested by calcu- trap column, washed with 100% MPA, and separated on lating the OR of the upper and lower quartiles and esti- fi an Acclaim PepMap100 analytical column (75 mmID3 15 mating the 95% con dence interval of the OR. Receiver cm, C18, 3 mm, 100 Å; Thermo Scientific) using a 45-min- operator characteristic (ROC) curves were constructed to ute two-step gradient. Tandem mass spectrometry (MS/ determine the prognostic predictive power of uAnCR. MS) was performed using an AB SCIEX Triple TOF 5600 Univariate ROC curves were considered statistically sig- fi mass spectrometer. Acquired spectra were searched ni cant if the area under the curve (AUC) differed from z against the 2011_6 release of the Human UniProtKB/ 0.5, as determined by the test. Optimal cut-offs were de- Swiss-Prot database (20,127 entries) with addition of com- termined by selecting the data point that minimized the mon contaminants (112 entries) using the Mascot search geometric distance from 100% sensitivity and 100% speci- fi engine with trypsin as the specified enzyme. Monoisotopic city on the ROC curve (24). To visualize the relationship masses were used, and the error tolerances were 10 ppm between uAnCR and length of stay, patients were strati- fi – and 0.5 Da for peptides and MS/MS fragments, respec- ed into tertiles by uAnCR. Kaplan Meier curves with tively. Mascot search results were loaded into Scaffold censoring for death were plotted. The log-rank test was (Proteome Software Inc), which used the Peptide Prophet used to compare the curves and the Holm-Sidak test was post hoc and Protein Prophet algorithms to validate peptide and used for pairwise comparison. Category free net fi protein identifications (21,22). The Scaffold quantitative reclassi cation improvement (cfNRI) was used to deter- values of identified proteins were normalized to the inter- mine if addition of uAnCR to a multivariate logistic re- nal standard HIV protein, and the relative abundance of gression model for prediction of risk increased the ability each protein is reported in normalized spectral counts. of the model to predict worsening of AKI (25,26). The risk prediction model consisted of the Cleveland Clinic cardiac surgery risk score and percent change in serum creatinine Angiotensinogen ELISA at the time the urine sample was collected (27). Statistical The Human Total Angiotensinogen Assay Kit (Immuno- tests were performed in Matlab or SigmaPlot. Biologic Laboratories Co. Ltd.) was used according to the manufacturer’s protocol. Values for intra- and interassay variability were 2.4% and 9.9%, respectively. Results Discovery of Candidate Prognostic AKI Biomarkers Urine Creatinine Determination We used LC-MS/MS to compare the urinary proteomic Urine creatinine was measured using the Jaffe assay. profiles of 12 patients who developed AKI after cardiac surgery, 6 of whom later required RRT and 6 of whom did Outcomes not. There were no statistically significant differences fi The primary outcome was worsening of AKI, de ned as between the two groups with respect to the demographic progression to a higher AKIN stage after the time of sample characteristics, sample collection time, use of cardiopul- collection. Secondary outcomes were progression to AKIN monary bypass, bypass time, type of surgery, preoperative stage 3, the need for renal replacement therapy (RRT) SCr, and SCr at the time of sample collection (Supplemental within 10 days of sample collection, progression to AKIN Table 1). We identified 343 proteins, of which 59 were stage 2 or 3, progression to AKIN stage 3 or death, RRT or unique to $1 patients who required RRT and 5 were . death, and discharge 7 days from the time of sample unique to $1 patients who did not (Figure 1A). The rela- collection or in-hospital mortality. Outcomes were tested tive abundance of 30 proteins was statistically different using the entire cohort and in the subset of patients clas- between the two groups (Table 1 and Supplemental Table fi si ed as AKIN stage 1 at the time of sample collection. 2). The abundance of 26 proteins was increased in the urine of patients who required RRT and four were de- Statistical Analyses creased. We selected angiotensinogen as the most promis- Differential abundance of proteins quantified by LC- ing candidate marker based on the combination of its low MS/MS was tested using the Wilcoxon rank-sum test P value (P=0.002) and relatively large mean fold-change which has been shown to be a robust test for identifica- (9.67-fold) difference between groups (Figure 1B). Relative tion of candidate biomarkers in proteomic studies (23). abundances of angiotensinogen for the individual partici- Candidate biomarkers were selected based upon the pants (Figure 1C) demonstrate that urinary angiotensinogen 186 Clinical Journal of the American Society of Nephrology

discriminated with 100% accuracy between patients who required RRT and those who did not in this group. On the basis of these data, we attempted to verify the potential of urinary angiotensinogen as a biomarker of severe AKI after cardiac surgery.

Verification of the Prognostic Ability of Urinary Angiotensinogen We measured urinary angiotensinogen by ELISA and verified its association with outcomes in patients who had developed AKI after cardiac surgery (n=97). These patients were divided into three groups by maximum AKIN stage: stage 1 (n=59), stage 2 (n=19), and stage 3 (n=19). Seventy- nine patients were classified as AKIN stage 1 at the time of urine sample collection. Ten of these patients progressed to a maximum AKIN stage of 2, 10 progressed to AKIN stage 3 and 59 did not progress. There were no statistically significant differences among the groups with respect to sex, race, age, use of bypass, bypass time, preoperative SCr, and type of surgery (Table 2). In addition, SCr was not statistically different among the groups in the subset of patients who were classified as AKIN stage 1 at collection. Among all patients who had developed AKI of any stage at the time of urine sample collection, uAnCR (nanograms of angiotensinogen per milligrams of creatinine) was cor- related with both maximum SCr (r=0.49; P,0.001) and maximum percent change in SCr (r=0.29; P=0.01), and uAnCR increased with maximum AKIN stage achieved in both the whole cohort and the subset of patients classi- fied as AKIN stage 1 at collection (Supplemental Figure 1). Pairwise comparison revealed a statistically significant dif- ference in uAnCR between the patients who developed AKIN stage 3 and those who reached a maximum of stage 1 (Table 2 and Supplemental Figure 1). Patients with higher uAnCR had increased risk of adverse outcomes (Supplemental Table 3). Comparing pa- tients in the top quartile of uAnCR to those in the bottom quartile, the odds ratio for the primary outcome worsening of AKI was 5.0 (95% CI, 1.2–21.5) in the whole cohort and 4.6 (95% CI, 1.0–21.0) in the subset of patients who were classified as AKIN stage 1 at the time of sample collection. uAnCR discriminated between patients who experienced worsening of AKI after sample collection and those who did not in both the whole cohort (Figure 2A; AUC=0.70) Figure 1. | Proteins identified by LC-MS/MS in the urine of patients and in the subset classified as AKIN stage 1 at collection who developed AKI after cardiac surgery. (A) The Venn diagram (Figure 2B; AUC=0.71). At the optimal cut-off (33.27 ng/ shows the number of proteins identified in patients who later de- mg), the sensitivity and specificity of uAnCR were 70.8% veloped severe postoperative AKI requiring RRT versus those who did not. (B) The volcano plot shows the significance of the difference in and 66.2%, respectively, in the whole cohort. In patients fi mean abundance between the two groups for each identified protein. who were classi ed as AKIN stage 1 at collection, sensi- It facilitates the selection of candidate biomarkers that have a large tivity and specificity were 75.0% and 66.1%. ROC curves magnitude fold-change (positive fold changes indicate elevated for the other tested outcomes showed similar results (Sup- protein levels in the RRT group) and highly significant P values. The plemental Figures 2 and 3). Notably, the predictive power arrowhead indicates the data point for angiotensinogen. Proteins for most outcomes among those patients classified as , above the dashed line had a P value 0.05. (C) The scattergram shows AKIN stage 1 at collection appeared to be slightly aug- the angiotensinogen abundance in each patient in the discovery mented compared with the analysis of the entire cohort. analysis by group. Angiotensinogen was identified in the urine of two For example, uAnCR discriminated with high accuracy of six AKI patients in the No RRT group and was undetectable in the (AUC=0.81) between patients who later met the outcome other four patients. LC-MS/MS, liquid chromatography/tandem mass spectrometry; RRT, renal replacement therapy. of AKIN stage 3 or death and those who did not. In addition to the prediction of the renal and mortality outcomes, we noted a relationship between uAnCR and length of hospital stay. Among all AKI patients and in the subset of patients classified as AKIN stage 1 at the time of Clin J Am Soc Nephrol 8: 184–193, February, 2013 Urinary Angiotensinogen Predicts Worsening AKI, Alge et al. 187

Table 1. Candidate biomarkers of severe AKI requiring RRT

Uniprot Accession Mean Mean Mean Identified Proteins (349) P Number No RRT RRT Fold Change

Angiotensinogen P01019 1.69 16.33 9.67 0.002 Serum albumin P02768 653.89 2892.67 4.42 0.002 Apo A-IV P06727 2.36 21.45 9.09 0.01 C3 P01024 8.82 50.09 5.68 0.01 Vitamin D-binding protein P02774 4.51 54.58 12.11 0.01 C4-B P0C0L5 2.71 23.25 8.58 0.01 SOD [Cu-Zn] P00441 10.10 23.91 2.37 0.01 Epididymal secretory protein E1 P61916 3.17 10.39 3.28 0.01 Phosphatidylethanolamine-binding protein 1 P30086 0 5.90 N/C 0.02 Complement factor D P00746 0 5.81 N/C 0.02 Coactosin-like protein Q14019 0 3.70 N/C 0.02 Serotransferrin P02787 72.61 472.14 6.5 0.02 Profilin-1 P07737 5.87 24.04 4.1 0.02 Cystatin-B P04080 0.43 4.94 11.63 0.02 Fibrinogen a chain P02671 7.33 35.39 4.83 0.02 Brain acid soluble protein 1 P80723 0.87 5.05 5.82 0.02 Zinc-a-2-glycoprotein P25311 112.02 228.64 2.04 0.03 a-1-antitrypsin P01009 21.55 239.32 11.11 0.03 a-1-acid glycoprotein 1 P02763 44.94 112.82 2.51 0.03 Hemopexin P02790 7.94 30.11 3.79 0.03 Fibrinogen b chain P02675 3.65 14.29 3.92 0.03 Pigment epithelium-derived factor P36955 2.50 22.71 9.08 0.03 Fatty acid-binding protein, adipocyte P15090 6.25 9.99 1.6 0.03 a-1-acid glycoprotein 2 P19652 18.15 47.56 2.62 0.04 Metallothionein-2 P02795 1.43 19.25 13.47 0.04 Apo A-I P02647 2.39 20.36 8.53 0.04 Keratin, type II cytoskeletal 5 P13647 3.74 2.16 21.74 0.03 Secreted Ly-6/uPAR-related protein 1 P55000 4.49 3.18 21.41 0.03 Nonsecretory ribonuclease P10153 17.15 7.72 22.22 0.04 Keratin, type II cytoskeletal 1 P04264 35.85 18.85 21.9 0.05

Relative protein abundance was estimated using normalized spectral counts. RRT, renal replacement therapy. N/C, the value was not calculated because the denominator is zero.

collection, those patients with higher uAnCR concentra- added benefit of including uAnCR in the model. Addition tions had longer hospital stays (Figure 3, A and B). ROC of uAnCR in the new model improved the ability to curve analysis indicated that elevated uAnCR was predic- predict a patients risk of worsening of AKI in both the tive of longer length of stay defined as discharge .7 days entire cohort and the subset of patients who were classi- from the time of sample collection or death (Supplemental fied as AKIN stage 1 (cfNRI=45.7% and 42.8%, respec- Figures 2 and 3). Tables 3 and 4 summarize the perfor- tively). To visualize the improvement in prediction, we mance characteristics of uAnCR as a predictor of the tested constructed a risk assessment plot, as proposed by Picker- outcomes in patients who had AKI of any stage at the time ing and Endre (26). This plot compares the sensitivity and of sample collection and those who had not progressed 1-specificity of the reference and new models across the beyond AKIN stage 1 at the time of sample collection. spectrum of calculated risk for each model. Figure 4 shows We determined the ability of uAnCR to improve the that the addition of uAnCR into the model resulted in prediction of worsening AKI of a clinical risk model in both patients who met the outcome (events) having a greater the entire cohort and in the subset of patients classified as calculated risk, and patients who did not meet the out- AKIN stage 1 at the time of urine collection. The clinical come (nonevents) had a lower calculated risk. Therefore, model was a multivariate logistic regression model con- both sensitivity and specificity were improved by includ- sisting of the percent change in SCr from baseline at the ing uAnCR in the prediction model. time of sample collection and the patient’sClevelandClinic score, a perioperative risk score that predicts AKI severity after cardiac surgery (27,28). When uAnCR was added to Discussion the clinical model, we found that it predicted worsening of We identified candidate biomarkers for the prediction of AKI independently of the percent change in SCr and the the development of severe AKI, and the prognostic poten- Cleveland Clinic score (P=0.02). We used cfNRI, which tial of the most promising candidate, angiotensinogen, was compares each patient’s calculated risk for an outcome verified in a larger set of patients who developed AKI after using a reference model to a new model, to capture the cardiac surgery. We found that uAnCR was elevated in 8 lnclJunlo h mrcnSceyo Nephrology of Society American the of Journal Clinical 188

Table 2. Characteristics of cardiac surgery patients used to verify the potential of urinary angiotensinogen as a biomarker of postoperative AKI

AKI of Any Stage at Time of Sample Collection (n=97) AKIN Stage 1 at Time of Sample Collection (n=79) Maximum AKIN Stage Achieved AKIN Stage 1 AKIN Stage 2 AKIN Stage 3 P AKIN Stage 1 AKIN Stage 2 AKIN Stage 3 P (n=59) (n=19) (n=19) (n=59) (n=10) (n=10) uAnCR 22.6 34.1 58.8 0.01 22.6 35.3 77.0 0.01 (13.1–54.0) (11.1–50.4) (20.4–217.1) (13.1–54.0) (22.6–270.3) (30.9–329.4) Male 42 (71) 14 (74) 13 (68) 0.94 42 (71) 6 (60) 7 (70) 0.78 Caucasian 38 (64) 12 (63) 15 (79) 0.47 38 (64) 6 (60) 8 (80) 0.58 Age (yr) 65.8610.8 64.5610.0 68.5611.9 0.53 65.8610.8 68.2610.8 69.0614.7 0.58 Weight (kg) 88.2624.3 94.416 23.8 88.9627.3 0.63 88.2624.3 84.7621.7 88.1633.3 0.69 Sample collection time 27.9611.8 31.6614.5 36.0611.0 0.04 27.9611.8 26.2615.6 35.2612.2 0.23 (postoperative h) Operative variables CABG 36 (61) 10 (53) 11 (58) 0.81 36 (61) 4 (40) 5 (50) 0.41 Valve replacement 17 (29) 7 (37) 4 (21) 0.56 17 (29) 5 (50) 2 (20) 0.30 CABG + valve replacement 6 (10) 2 (11) 4 (21) 0.44 6 (10) 1 (10) 3 (30) 0.21 Bypass 45 (76) 16 (84) 14 (74) 0.71 45 (76) 8 (80) 7 (70) 0.88 Bypass time (min) 143.2672.5 145.4675.6 118.9667.3 0.31 143.2672.5 154.4674.7 146.4677.8 0.95 SCr (mg/dl) Preoperative SCr 1.260.3 1.260.4 1.260.5 0.19 1.260.3 1.260.4 1.460.5 0.74 SCr at collection 1.760.4 2.060.7 2.560.8 0.001 1.760.4 1.760.6 2.360.8 0.14 Max SCr 1.960.4 2.760.8 4.061.9 ,0.001 1.960.4 2.660.8 4.362.6 ,0.001 Outcomes Days to max SCr 2.0 (1.0–3.0) 2.0 (2.0–3.0) 4.0 (2.0–5.0) 0.001 3.0 (2.0–5.25) 4.5 (2.7–6.75) ,0.001 (postoperative) 2.0 (1.0–3.0) RRT 10 days 0 0 9 (47) ,0.001 0 0 8 (80) ,0.001 Death 0 2 (11) 6 (32) ,0.001 0 2 (20) 3 (30) ,0.001

Data are shown as median (interquartile range), n (%), or mean 6 SD. AKIN, Acute Kidney Injury Network; uAnCr, urine angiotensinogen/creatinine ratio; CABG, coronary artery bypass graft; SCr, serum creatinine; Max, maximum; RRT, renal replacement therapy. Clin J Am Soc Nephrol 8: 184–193, February, 2013 Urinary Angiotensinogen Predicts Worsening AKI, Alge et al. 189

Figure 2. | Angiotensinogen is associated with worsening of AKI. ROC curves were used to test the ability of uAnCR to predict worsening of AKI after sample collection among (A) patients who were any stage AKI at the time of collection (n=97) and (B) the subset who were classified as AKIN stage 1 at collection (n=79). ROC, receiver operator characteristic; uAnCr, urine angiotensinogen/creatinine ratio; AUC, area under the curve.

Figure 3. | Angiotensinogen is associated with increased length of stay in patients with postoperative AKI. Patients who developed AKI after cardiac surgery were stratified into tertiles by uAnCR. Kaplan–Meier survival curves show that (A) among patients with AKI of any severity at the time of collection and (B) among patients with AKIN stage 1 at the time of collection, those who have higher uAnCR have increased length of stay (defined as days to discharge from time of sample collection). *P,0.05 compared with the low uAnCR group. uAnCr, urine angiotensinogen/creatinine ratio.

patients who developed more severe AKI after sample could be used to predict adverse outcomes among patients collection. Elevated uAnCR was associated with worsening who have not yet developed severe AKI as measured by of AKI, independent of changes in SCr and Cleveland serum creatinine. Clinic score, and it was also associated with several Our data suggest that angiotensinogen could be used at secondary outcomes. The prognostic predictive power of the time of diagnosis with AKI to assess the risk of adverse uAnCR was improved when only patients who were outcomes. This risk assessment could lead to improved classified as AKIN stage 1 at the time of sample collection outcomes by identifying high-risk patients in need of were used in the analysis, indicating that angiotensinogen therapeutic intervention as was highlighted in the Kidney 9 lnclJunlo h mrcnSceyo Nephrology of Society American the of Journal Clinical 190

Table 3. Performance characteristics of uAnCR as a prognostic AKI biomarker among patients who were any stage AKI at the time of sample collection (n=97)

Outcome AUC Cut-Off (ng/mg) n (%)a Sensitivity (%) Specificity (%) LR+ LR2 PPV (%) NPV (%)

Worsening AKI 0.70 (0.57–0.82) Best .33.27 43 (44.3) 70.8 66.2 2.09 0.44 67.7 69.4 Max PPV .392.5 6 (6.2) 16.7 98.5 11.34 0.85 91.9 54.2 Max NPV .12.55 20 (20.6) 91.7 26.5 1.25 0.31 55.5 76.1 AKIN stage 3 0.71 (0.59–0.84) Best .34.33 40 (41.2) 68.4 65.4 1.98 0.48 66.4 67.4 Max PPV .572.0 4 (4.1) 15.8 98.7 12.34 0.85 92.5 54.0 Max NPV .14.06 25 (25.8) 94.7 30.8 1.37 0.17 57.8 85.4 RRTb 0.71 (0.54–0.88) Best .58.63 26 (26.8) 66.7 77.3 2.93 0.43 74.6 69.9 Max PPV .572.0 4 (4.1) 11.1 96.6 3.26 0.92 76.5 52.1 Max NPV .20.01 37 (38.1) 88.9 40.9 1.50 0.27 60.1 78.6 AKIN stage 2 or 3 AKI 0.64 (0.52–0.73) Best .34.33 40 (41.2) 57.9 69.5 1.90 0.61 65.5 62.3 Max PPV .392.5 6 (6.2) 13.2 98.3 7.79 0.88 88.6 53.1 Max NPV .6.777 7 (7.2) 97.4 10.2 1.08 0.26 52.0 79.5 AKIN 3 or deathc 0.75 (0.64–0.87) Best .37.36 35 (36.1) 66.7 72.4 2.41 0.46 70.7 68.5 Max PPV .392.5 6 (6.2) 23.8 98.7 18.04 0.77 94.7 56.4 Max NPV .14.06 25 (25.8) 95.2 31.6 1.39 0.15 58.2 86.9 RRT or death 0.71 (0.55–0.86) Best .58.63 26 (26.8) 61.5 78.6 2.87 0.49 74.2 67.1 Max PPV .466.6 5 (5.2) 23.1 97.6 9.70 0.79 90.7 55.9 Max NPV .16.28 31 (32) 92.3 35.7 1.44 0.22 58.9 82.3 Length of stayd 0.74 (0.64–0.84) Best .26.38 52 (54.6) 69.6 68.3 2.19 0.44 68.7 69.2 Max PPV .67.97 24 (24.7) 37.5 92.7 5.12 0.67 83.7 59.7 Max NPV ,13.78 24 (24.8) 89.3 43.9 1.59 0.24 61.4 80.3 uAnCr, urine angiotensinogen/creatinine ratio; AUC, area under the curve; LR1, positive likelihood ratio; LR2, negative likelihood ratio; PPV, positive predictive value; NPV, negative predictive value; Max, maximum; RRT, renal replacement therapy. aThe percentage and number of patients who were above the best and max PPV cut-offs or below the max NPV cut-off. bRRT initiated within 10 days of surgery. cDeath defined as 30 day in-hospital mortality. dLength of stay outcome defined as discharge .7 days from sample collection or death before postoperative day 7. lnJA o eho :1413 eray 2013 February, 184–193, 8: Nephrol Soc Am J Clin

Table 4. Performance characteristics of uAnCR as a prognostic AKI biomarker among patients who were classified as AKIN stage 1 at the time of sample collection (n=79)

Outcome AUC Cut-Off (ng/mg) n (%)a Sensitivity (%) Specificity (%) LR+ LR2 PPV (%) NPV (%)

Worsening AKI 0.71 (0.57–0.85) Best .33.27 35 (44.3) 75.0 66.1 2.21 0.38 68.9 72.6 Max PPV .392.5 5 (6.3) 20.0 98.3 11.83 0.81 92.2 55.1 Max NPV .19.95 29 (36.7) 85.0 44.1 1.52 0.34 60.3 74.6 AKIN stage 3 AKI 0.75 (0.58–0.92) Best .58.63 22 (27.8) 70.0 78.3 3.22 0.38 76.3 72.3 Max PPV .572.0 3 (3.8) 20.0 98.6 13.79 0.81 93.2 55.2 Max NPV .19.95 29 (36.7) 90.0 40.6 1.51 0.25 60.2 80.2 RRTb 0.68 (0.49–0.87) Best .34.33 32 (40.5) 75.0 63.4 2.05 0.39 67.2 71.7 Max PPV .572.0 3 (3.8) 12.5 97.2 4.43 0.90 81.6 52.6 Max NPV .19.95 29 (36.7) 87.5 39.4 1.44 0.32 59.1 75.9 AKIN 3 or deathc 0.81 (0.66–0.95) Best .58.63 22 (27.8) 75.0 80.6 3.87 0.31 79.4 76.3 Max PPV .392.5 5 (6.3) 33.3 98.5 22.37 0.68 95.7 59.6 Max NPV .19.95 29 (36.7) 91.7 41.8 1.57 0.20 61.2 83.4 RRT or death 0.76 (0.59–0.93) Best .58.63 22 (27.8) 70.0 78.3 3.22 0.38 76.3 72.3

Max PPV .466.6 4 (5.1) 30.0 98.6 20.69 0.71 95.4 58.5 191 al. et Alge AKI, Worsening Predicts Angiotensinogen Urinary Max NPV .19.95 29 (36.7) 90.0 40.6 1.51 0.25 60.2 80.2 Length of stayd 0.74 (0.63–0.85) Best .26.38 43 (54.4) 72.7 68.6 2.31 0.40 69.8 71.5 Max PPV .109.0 13 (16.5) 27.3 97.1 9.5 0.75 90.5 57.1 Max NPV .13.78 59 (74.5) 89.3 43.9 1.59 0.24 61.4 80.3 uAnCr, urine angiotensinogen/creatinine ratio; AKIN, Acute Kidney Injury Network; AUC, area under the curve; LR1, positive likelihood ratio; LR2, negative likelihood ratio; PPV, positive predictive value; NPV, negative predictive value; Max, maximum; RRT, renal replacement therapy. aThe percentage and number of patients who were above the best and max PPV cut-offs or below the max NPV cut-off. bRRT initiated within 10 days of surgery. cDeath defined as 30 day in-hospital mortality. dLength of stay outcome defined as discharge .7 days from sample collection or death before postoperative day 7. 192 Clinical Journal of the American Society of Nephrology

Figure 4. | Risk assessment plots showing the improved prediction of worsening AKI when uAnCR is included in the model. Results from both (A) the entire cohort and (B) the subset of patients who were classified as AKIN stage 1 at the time of sample collection are shown. Of 97 patients in the whole cohort (A), 39 patients met the outcome worsening of AKI after sample collection, whereas in the subset analysis, 20 patients met the outcome. Two multivariate logistic regression models were created to predict risk of worsening of AKI after sample collection. The first model (reference) used percent change in SCr from baseline and Cleveland Clinic score created by Thakar et al. The second model included these variables plus uAnCR. Each patient’s probability (i.e., risk) of meeting the outcome worsening of AKI was calculated with both models. The sensitivity (proportion of events with a calculated risk equal to or above the defined threshold) and 1-specificity (proportion of nonevents with a calculated risk below the defined threshold) was calculated across all possible unique thresholds using both models. uAnCr, urine angiotensinogen/creatinine ratio.

Disease Improving Global Outcomes (KDIGO) Clinical subsequent bioactive molecules in the RAS cascade. The iden- Practice Guideline for Acute Kidney Injury (29). The tified portions of angiotensinogen in our proteomic study guidelines suggest several interventions in patients with did not include the proximal domain of angiotensinogen stage 2 and 3 AKI that are not recommended for patients from which angiotensin I is cleaved (Supplemental Fig- with stage 1 AKI, including checking drug dosing, consid- ures 4 and 5). Likewise, the ELISA used to quantify ering RRT, and considering intensive care unit (ICU) ad- angiotensinogen recognizes an epitope distal to the angio- mission. Elevation in uAnCR could suggest a population tensin I domain, so it is insensitive to the detection of of patients with stage 1 AKI who are likely to worsen and cleavage of angiotensinogen by renin. Also unclear is could benefit from more intensive intervention. whether increases in angiotensinogen are systemic or Although we did not directly compare the prognostic intrarenal in nature. However, others have shown that predictive power of angiotensinogen to that of other intrarenal angiotensin II is increased after renal ischemia- biomarkers, our results are comparable with what has reperfusion injury in a rat model (36). In order to under- been reported in the literature for previously described AKI stand the relevance of urinary angiotensinogen in the biomarkers. For example, Hall et al. reported unadjusted pathobiology of AKI, it will be necessary to determine AUCs of 0.71, 0.64, and 0.63 for the prediction of the com- the status of the renin-angiotensinogen system during posite outcome of worsening of AKI or death for urine AKI. NGAL, kidney injury molecule-1, and IL-18, respectively (18). Koyner recently reported unadjusted AUCs of 0.58, Acknowledgments 0.63, and 0.74 for urine NGAL, urine IL-18, and plasma Additional members of the SAKInet consortium (www.sakinet. NGAL, respectively, for the outcome of worsening of AKI org) include Juan Carlos Q. Velez, Elizabeth G. Hill, Milos N. (19). Thus, uAnCR, alone or in combination with other Budisavljevic, Rick G. Schnellmann of the Medical University of South biomarkers could improve risk classification models in Carolina, and Frederick T. (Josh) Billings of Vanderbilt University. these patients. This study was supported by the National Institutes of Health Angiotensinogen is the principal substrate of the renin- (Grants R01 DK080234 and UL1 RR029882) and by a Merit Re- angiotensin system (RAS), a hormonal cascade that has view award from the Biomedical Laboratory Research and Devel- pleiotropic effects in the kidney, including the regulation of opment Program of the Department of Veterans Affairs. M.J. was hemodynamics, sodium reabsorption, aquaresis, cellular supported by the Nephcure Foundation as a young investigator. proliferation and apoptosis, fibrosis, and inflammation The contents do not necessarily represent the views of the Depart- (30). Our findings regarding the prognostic utility of an- ment of Veterans Affairs or the US Government. giotensinogen as a biomarker of AKI suggest that the RAS Disclosures could be mechanistically involved in AKI. This is suppor- None. ted by observational studies that have noted an associa- tion between pharmacologic inhibition of the RAS and risk of developing AKI, although there are conflicting re- References ports in the literature (31–34). The ACE II genotype has 1. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P; Acute been associated with increased risk of AKI in the ICU Dialysis Quality Initiative workgroup; : Acute renal failure - definition, outcome measures, animal models, fluid therapy and (35). It is unclear whether the elevated levels of urinary fl information technology needs: The Second International Con- angiotensinogen observed in severe AKI re ect cleav- sensus Conference of the Acute Dialysis Quality Initiative (ADQI) age of existing angiotensinogen into angiotensin I and Group. Crit Care 8: R204–R212, 2004 Clin J Am Soc Nephrol 8: 184–193, February, 2013 Urinary Angiotensinogen Predicts Worsening AKI, Alge et al. 193

2. Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock predict progression of acute kidney injury after cardiac surgery. J DG, Levin A; Acute Kidney Injury Network; : Acute Kidney Injury Am Soc Nephrol 23: 905–914, 2012 Network: Report of an initiative to improve outcomes in acute 20. Englberger L, Suri RM, Li Z, Casey ET, Daly RC, Dearani JA, Schaff kidney injury. Crit Care 11: R31, 2007 HV: Clinical accuracy of RIFLE and Acute Kidney Injury Network 3. Han WK, Bailly V, Abichandani R, Thadhani R, Bonventre JV: (AKIN) criteria for acute kidney injury in patients undergoing Kidney Injury Molecule-1 (KIM-1): A novel biomarker for human cardiac surgery. Crit Care 15: R16, 2011 renal proximal tubule injury. Kidney Int 62: 237–244, 2002 21. Keller A, Nesvizhskii AI, Kolker E, Aebersold R: Empirical sta- 4. Mishra J, Ma Q, Prada A, Mitsnefes M, Zahedi K, Yang J, Barasch J, tistical model to estimate the accuracy of peptide identifications Devarajan P: Identification of neutrophil gelatinase-associated made by MS/MS and database search. Anal Chem 74: 5383– lipocalin as a novel early urinary biomarker for ischemic renal 5392, 2002 injury. J Am Soc Nephrol 14: 2534–2543, 2003 22. Nesvizhskii AI, Keller A, Kolker E, Aebersold R: A statistical 5. Mishra J, Dent C, Tarabishi R, Mitsnefes MM, Ma Q, Kelly C, Ruff model for identifying proteins by tandem mass spectrometry. SM, Zahedi K, Shao M, Bean J, Mori K, Barasch J, Devarajan P: Anal Chem 75: 4646–4658, 2003 Neutrophil gelatinase-associated lipocalin (NGAL) as a bio- 23. Dakna M, Harris K, Kalousis A, Carpentier S, Kolch W, Schanstra marker for acute renal injury after cardiac surgery. Lancet 365: JP,Haubitz M, Vlahou A, Mischak H, Girolami M: Addressing the 1231–1238, 2005 challenge of defining valid proteomic biomarkers and classifiers. 6. Melnikov VY, Ecder T, Fantuzzi G, Siegmund B, Lucia MS, BMC Bioinformatics 11: 594, 2010 Dinarello CA, Schrier RW, Edelstein CL: Impaired IL-18 pro- 24. Pepe M: The Statistical Evaluation of Medical Tests for Classifi- cessing protects caspase-1-deficient mice from ischemic acute cation and Prediction, New York, Oxford University Press, 2004 renal failure. J Clin Invest 107: 1145–1152, 2001 25. Pencina MJ, D’Agostino RB Sr, D’Agostino RB Jr, Vasan RS: 7. Parikh CR, Jani A, Melnikov VY, Faubel S, Edelstein CL: Urinary Evaluating the added predictive ability of a new marker: From interleukin-18 is a marker of human acute tubular necrosis. Am J area under the ROC curve to reclassification and beyond. Stat Kidney Dis 43: 405–414, 2004 Med 27: 157–172, discussion 207–212, 2008 8. Herget-Rosenthal S, Marggraf G, Hu¨sing J, Go¨ring F, Pietruck F, 26. Pickering JW, Endre ZH: New metrics for assessing diagnostic Janssen O, Philipp T, Kribben A: Early detection of acute renal potential of candidate biomarkers. Clin J Am Soc Nephrol 7: failure by serum cystatin C. Kidney Int 66: 1115–1122, 2004 1355–1364, 2012 9. Portilla D, Dent C, Sugaya T, Nagothu KK, Kundi I, Moore P, Noiri 27. Thakar CV, Arrigain S, Worley S, Yared JP, Paganini EP: A clinical E, Devarajan P: Liver fatty acid-binding protein as a biomarker of score to predict acute renal failure after cardiac surgery. JAmSoc acute kidney injury after cardiac surgery. Kidney Int 73: 465–472, Nephrol 16: 162–168, 2005 2008 28. Englberger L, Suri RM, Li Z, Dearani JA, Park SJ, Sundt TM 3rd, 10. Liangos O, Tighiouart H, Perianayagam MC, Kolyada A, Han WK, Schaff HV: Validation of clinical scores predicting severe acute Wald R, Bonventre JV, Jaber BL: Comparative analysis of urinary kidney injury after cardiac surgery. Am J Kidney Dis 56: 623–631, biomarkers for early detection of acute kidney injury following 2010 cardiopulmonary bypass. Biomarkers 14: 423–431, 2009 29. Kidney Disease; Improving Global Outcomes (KDIGO) Acute 11. Koyner JL, Vaidya VS, Bennett MR, Ma Q, Worcester E, Akhter SA, Kidney Injury Work Group; : Clinical practice guideline for acute Raman J, Jeevanandam V, O’Connor MF, Devarajan P, Bonventre kidney injury. Kidney Int Suppl 2: 1–138, 2012 JV, Murray PT: Urinary biomarkers in the clinical prognosis and 30. Velez JC: The importance of the intrarenal renin-angiotensin early detection of acute kidney injury. Clin J Am Soc Nephrol 5: system. Nat Clin Pract Nephrol 5: 89–100, 2009 2154–2165, 2010 31. Arora P, Rajagopalam S, Ranjan R, Kolli H, Singh M, Venuto R, 12. Wagener G, Gubitosa G, Wang S, Borregaard N, Kim M, Lee HT: Lohr J: Preoperative use of angiotensin-converting enzyme in- Urinary neutrophil gelatinase-associated lipocalin and acute hibitors/angiotensin receptor blockers is associated with in- kidney injury after cardiac surgery. Am J Kidney Dis 52: 425–433, creased risk for acute kidney injury after cardiovascular surgery. 2008 Clin J Am Soc Nephrol 3: 1266–1273, 2008 13. Parikh CR, Mishra J, Thiessen-Philbrook H, Dursun B, Ma Q, 32. Benedetto U, Sciarretta S, Roscitano A, Fiorani B, Refice S, Kelly C, Dent C, Devarajan P, Edelstein CL: Urinary IL-18 is an Angeloni E, Sinatra R: Preoperative Angiotensin-converting en- early predictive biomarker of acute kidney injury after cardiac zyme inhibitors and acute kidney injury after coronary artery surgery. Kidney Int 70: 199–203, 2006 bypass grafting. Ann Thorac Surg 86: 1160–1165, 2008 14. Haase M, Bellomo R, Story D, Davenport P, Haase-Fielitz A: 33. Plataki M, Kashani K, Cabello-Garza J, Maldonado F, Kashyap R, Urinary interleukin-18 does not predict acute kidney injury after Kor DJ, Gajic O, Cartin-Ceba R: Predictors of acute kidney injury adult cardiac surgery: A prospective observational cohort study. in septic shock patients: An observational cohort study. Clin J Am Crit Care 12: R96, 2008 Soc Nephrol 6: 1744–1751, 2011 15. Koyner JL, Bennett MR, Worcester EM, Ma Q, Raman J, 34. Yoo YC, Youn YN, Shim JK, Kim JC, Kim NY, Kwak YL: Effects of Jeevanandam V, Kasza KE, O’Connor MF, Konczal DJ, Trevino S, renin-angiotensin system inhibitors on the occurrence of acute Devarajan P, Murray PT: Urinary cystatin C as an early biomarker kidney injury following off-pump coronary artery bypass grafting. of acute kidney injury following adult cardiothoracic surgery. Circ J 74: 1852–1858, 2010 Kidney Int 74: 1059–1069, 2008 35. du Cheyron D, Fradin S, Ramakers M, Terzi N, Guillotin D, 16. Parikh CR, Devarajan P,Zappitelli M, Sint K, Thiessen-Philbrook H, Bouchet B, Daubin C, Charbonneau P: Angiotensin converting Li S, Kim RW, Koyner JL, Coca SG, Edelstein CL, Shlipak MG, Garg enzyme insertion/deletion genetic polymorphism: Its impact on AX, Krawczeski CD; TRIBE-AKI Consortium; : Postoperative bio- renal function in critically ill patients. Crit Care Med 36: 3178– markers predict acute kidney injury and poor outcomes after pe- 3183, 2008 diatric cardiac surgery. J Am Soc Nephrol 22: 1737–1747, 2011 36. Allred AJ, Chappell MC, Ferrario CM, Diz DI: Differential actions 17. Parikh CR, Coca SG, Thiessen-Philbrook H, Shlipak MG, Koyner of renal ischemic injury on the intrarenal angiotensin system. Am JL, Wang Z, Edelstein CL, Devarajan P, Patel UD, Zappitelli M, J Physiol Renal Physiol 279: F636–F645, 2000 Krawczeski CD, Passik CS, Swaminathan M, Garg AX; TRIBE-AKI Consortium; : Postoperative biomarkers predict acute kidney Received: June 22, 2012 Accepted: October 17, 2012 injury and poor outcomes after adult cardiac surgery. JAmSoc Nephrol 22: 1748–1757, 2011 18. Hall IE, Coca SG, Perazella MA, Eko UU, Luciano RL, Peter PR, Published online ahead of print. Publication date available at www. Han WK, Parikh CR: Risk of poor outcomes with novel and tra- cjasn.org. ditional biomarkers at clinical AKI diagnosis. Clin J Am Soc Nephrol 6: 2740–2749, 2011 This article contains supplemental material online at http://cjasn. 19. Koyner JL, Garg AX, Coca SG, Sint K, Thiessen-Philbrook H, Patel asnjournals.org/lookup/suppl/doi:10.2215/CJN.06280612/-/ UD, Shlipak MG, Parikh CR; TRIBE-AKI Consortium: Biomarkers DCSupplemental.

Supplemental Figure 4. Sequence coverage of angiotensinogen identified by LC-MS/MS. 7 unique peptides, 63/485 amino acids (13% sequence coverage). Identified sequences are highlighted in yellow. Renin cleaves angiotensinogen at amino acid 43 (denoted by asterisk), to release angio- tensin I. Supplemental Figure 5. Mass spectrum of representative angiotensinogen peptide with a parent ion mass of 1267.75 AMU (ALQDQLVLVAAK) identified in the urine of a patient who developed acute kidney injury after cardiac surgery and later required renal replacement therapy. Supplemental Table 1. Characteristics of patients used in a discovery phase proteomics study to identify candidate biomarkers of severe AKI No RRT RRT p-value n 6 6 Male 67% (4) 67% (4) 1 Caucasian 100% (6) 100% (6) 1 a Age (yrs) 63.8 +/- 7.9 72.5 +/- 17.1 0.29 a Weight (kg) 75.9 +/- 40.2 85.3 +/- 37.5 0.33 a Sample Collection Time (hrs post-op) 29.2 +/- 14.4 38.0 +/- 12.0 0.3 Operative Variables CABG 50% (3) 33% (2) 1 Valve Replacement 17 % (1) 17% (1) 1 CABG+Valve Replacement 33% (2) 17% 1) 1 Other Surgery 0% (0) 33% (2) 0.46 Bypass 67% (4) 67% (4) 1 a Bypass Time (min) 160.8 +/- 67.0 165.0 +/- 88.2 1 Serum Creatinine (mg/dL) a Pre-op Value 1.3 +/- 0.4 1.4 +/- 0.3 0.76 a At Sample Collection 1.9 +/- 0.4 2.6 +/- 0.6 0.37 a Maximum Post-Op Value 2.1 +/- 0.5 4.2 +/- 1.4 0.006 Outcomes a Days to Max sCr (from surgery) 1.7 +/- 1.0 4.8 +/- 2.4 0.02 RRT 0% (0) 100% (6) 0.002 Death 0% (0) 67% (4) 0.06 aMean +/- SD Supplemental Table 2. Complete list of protein identi cations

Wilcoxon Mean No RRT No RRT No RRT No RRT No RRT No RRT RRT RRT RRT RRT RRT RRT Rank-Sum Up or Fold Identified Proteins (349) 1 3 7 5 9 11 10 4 2 8 6 12 p-val Down Change Angiotensinogen P01019 0.00 0.00 0.00 2.13 1.25 0.00 2.95 17.00 3.36 4.90 44.63 25.12 0.002 Up 9.67 Serum albumin P02768 577 796 479 419 918 734 972 5874 1038 1167 5840 2465 0.002 Up 4.42 Apolipoprotein A-IV P06727 0.00 0.00 0.00 0.00 0.00 2.36 1.18 19.13 6.71 1.63 80.76 19.32 0.006 Up 9.09 Complement C3 P01024 0.00 9.66 17.00 0.00 6.25 2.36 21.84 93.51 40.27 14.71 114.76 15.46 0.009 Up 5.68 Vitamin D-binding protein P02774 1.77 0.00 4.25 0.00 7.50 0.00 21.84 91.38 10.07 3.27 148.76 52.16 0.009 Up 12.11 Complement C4-B P0C0L5 0.00 0.00 0.85 6.38 1.25 2.36 41.91 36.13 16.78 4.90 34.00 5.80 0.009 Up 8.58 Superoxide dismutase [Cu-Zn] P00441 0.00 0.00 0.43 14.88 15.00 0.00 4.13 42.50 19.01 16.35 38.25 23.18 0.009 Up 2.37 Epididymal secretory protein E1 P61916 1.77 0.00 1.28 2.13 7.50 0.00 3.54 8.50 10.07 14.71 4.25 21.25 0.009 Up 3.28 Phosphatidylethanolamine-binding protein 1 P30086 0.00 0.00 0.00 0.00 0.00 0.00 4.72 2.13 14.54 0.00 4.25 3.86 0.015 Up N/A Complement factor D P00746 0.00 0.00 0.00 0.00 0.00 0.00 1.18 2.13 8.95 0.00 14.88 1.93 0.015 Up N/A Coactosin-like protein Q14019 0.00 0.00 0.00 0.00 0.00 0.00 2.36 0.00 6.71 3.27 4.25 1.93 0.015 Up N/A Serotransferrin P02787 20 100 59 81 161 14 106 1165 140 121 910 392 0.015 Up 6.50 Profilin-1 P07737 0.00 0.00 0.85 4.25 12.50 0.00 7.08 10.63 21.25 42.50 53.13 9.66 0.015 Up 4.10 Cystatin-B P04080 0.00 0.00 0.43 0.00 0.00 0.00 3.54 2.13 2.24 0.00 14.88 1.93 0.015 Up 11.63 Fibrinogen alpha chain P02671 4.43 7.73 5.95 8.50 15.00 2.36 23.02 25.50 44.74 11.44 99.88 7.73 0.017 Up 4.83 Brain acid soluble protein 1 P80723 0.89 0.00 0.85 0.00 0.00 0.00 2.95 0.00 2.24 3.27 14.88 1.93 0.022 Up 5.82 Zinc-alpha-2-glycoprotein P25311 54.90 0.00 145.79 82.88 196.27 80.28 115 300 110 366 215 267 0.026 Up 2.04 Alpha-1-antitrypsin P01009 1.77 1.93 48.45 2.13 53.75 21.25 65 578 18 52 576 147 0.026 Up 11.11 Alpha-1-acid glycoprotein 1 P02763 30.11 5.80 47.60 51.00 95.01 40.14 104.49 167.89 30.20 137.32 99.88 137.17 0.026 Up 2.51 Hemopexin P02790 1.77 5.80 6.38 12.75 16.25 4.72 25.38 38.25 24.61 4.90 27.63 59.89 0.026 Up 3.79 Fibrinogen beta chain P02675 1.77 3.86 4.25 2.13 7.50 2.36 24.20 6.38 6.71 6.54 36.13 5.80 0.026 Up 3.92

Pigment epithelium-derived factor P36955 0.00 0.00 0.00 0.00 2.50 0.00 0.00 4.25 8.95 1.63 44.63 54.10 0.028 Up 9.08 Fatty acid-binding protein, adipocyte P15090 0.00 0.00 0.00 0.00 6.25 0.00 0.00 6.38 15.66 1.63 12.75 13.52 0.028 Up 1.60 Alpha-1-acid glycoprotein 2 P19652 3.54 3.86 10.63 23.38 46.25 21.25 49.59 80.76 6.71 34.33 63.76 50.23 0.041 Up 2.62 Metallothionein-2 P02795 0.89 0.00 1.28 2.13 0.00 0.00 4.13 2.13 22.37 0.00 63.76 3.86 0.041 Up 13.47 Apolipoprotein A-I P02647 0.89 0.00 1.28 0.00 5.00 0.00 5.90 6.38 8.95 0.00 78.63 1.93 0.043 Up 8.53 Thymosin beta-4-like protein 3 A8MW06 0.00 0.00 0.00 0.00 2.50 0.00 0.59 2.13 14.54 1.63 76.51 0.00 0.054 Up 7.63 Apolipoprotein C-III P02656 0.00 0.00 0.00 0.00 0.00 0.00 0.00 8.50 40.27 3.27 25.50 0.00 0.061 Up N/A Dermcidin P81605 0.00 0.00 0.00 0.00 0.00 0.00 0.59 0.00 3.36 53.95 14.88 0.00 0.061 Up N/A Neutrophil gelatinase-associated lipocalin P80188 0.00 0.00 0.00 0.00 0.00 0.00 2.95 0.00 11.19 0.00 10.63 3.86 0.061 Up N/A Insulin-like growth factor-binding protein 1 P08833 0.00 0.00 0.00 0.00 0.00 0.00 0.59 0.00 8.95 8.17 12.75 0.00 0.061 Up N/A Carbonic anhydrase 3 P07451 0.00 0.00 0.00 0.00 0.00 0.00 0.00 17.00 0.00 3.27 17.00 3.86 0.061 Up N/A Ig heavy chain V-III region BUT P01767 0.00 0.00 0.00 0.00 0.00 0.00 1.77 12.75 0.00 1.63 4.25 0.00 0.061 Up N/A Heme-binding protein 2 Q9Y5Z4 0.00 0.00 0.00 0.00 0.00 0.00 1.18 2.13 0.00 3.27 0.00 1.93 0.061 Up N/A Myoglobin P02144 0.00 0.00 0.00 0.00 1.25 0.00 0.00 82.88 5.59 9.81 44.63 0.00 0.061 Up 28.58 Ig heavy chain V-III region BRO P01766 0.00 0.00 0.85 0.00 0.00 0.00 16.53 38.25 0.00 1.63 8.50 0.00 0.061 Up 19.09 Ribosome-binding protein 1 Q9P2E9 0.00 0.00 0.00 0.00 1.25 0.00 6.49 34.00 0.00 0.00 2.13 3.86 0.061 Up 9.30 Ig lambda chain V-III region SH P01714 0.00 0.00 1.28 0.00 0.00 0.00 2.95 8.50 0.00 0.00 12.75 7.73 0.061 Up 6.26 Retinol-binding protein 4 P02753 0.89 1.93 25.50 108.38 25.00 7.08 2.95 53.13 136.46 173.28 184.89 34.78 0.065 Up 3.47 Monocyte differentiation antigen CD14 P08571 1.77 0.00 4.25 17.00 22.50 11.81 41.91 21.25 11.19 19.62 19.13 30.91 0.065 Up 2.09 Ig mu chain C region P01871 0.00 0.00 13.18 0.00 17.50 0.00 17.12 12.75 3.36 1.63 19.13 23.18 0.076 Up 1.19

Fatty acid-binding protein, heart P05413 0.00 0.00 0.43 21.25 26.25 23.61 12.40 121.13 14.54 68.66 61.63 25.12 0.089 Up 2.83 Complement factor B P00751 0.00 0.00 2.55 12.75 0.00 40.14 5.31 21.25 17.90 19.62 46.75 3.86 0.091 Up 1.03 Alpha-1B-glycoprotein P04217 0.89 13.52 11.90 29.75 20.00 7.08 33.65 167.89 2.24 22.89 65.88 25.12 0.093 Up 3.82 Ceruloplasmin P00450 0.89 3.86 8.08 4.25 7.50 0.00 23.61 240.14 1.12 6.54 95.63 7.73 0.093 Up 12.71 Clusterin P10909 1.77 3.86 7.23 2.13 26.25 4.72 38.37 10.63 2.24 42.50 8.50 19.32 0.093 Up 2.64 Beta-2-glycoprotein 1 P02749 2.66 21.25 13.18 8.50 8.75 4.72 18.89 19.13 10.07 13.08 42.50 17.39 0.093 Up 2.05 Glutathione peroxidase 3 P22352 0.00 0.00 0.85 0.00 2.50 0.00 4.72 31.88 4.47 0.00 19.13 0.00 0.100 Up 8.98 Plasma protease C1 inhibitor P05155 0.00 0.00 2.55 0.00 2.50 0.00 3.54 4.25 0.00 0.00 8.50 13.52 0.100 Up 2.95 Fatty acid-binding protein, liver P07148 0.00 0.00 0.00 0.00 3.75 0.00 0.00 2.13 89.48 0.00 29.75 34.78 0.106 Up 10.41 SH3 domain-binding glutamic acid- rich-like protein O75368 0.00 0.00 0.00 0.00 1.25 0.00 1.18 0.00 2.24 8.17 12.75 0.00 0.106 Up 4.87 Apolipoprotein A-II P02652 0.00 0.00 0.00 0.00 1.25 0.00 1.18 4.25 3.36 0.00 14.88 0.00 0.106 Up 4.73 Ganglioside GM2 activator P17900 4.43 0.00 12.75 0.00 8.75 2.36 21.25 6.38 11.19 0.00 14.88 17.39 0.128 Up 2.01 Alpha-2-HS-glycoprotein P02765 0.89 5.80 11.05 63.76 52.50 49.59 59.62 65.88 19.01 112.80 53.13 36.71 0.132 Up 1.89 Afamin P43652 0.00 0.00 0.00 0.00 5.00 0.00 0.59 53.13 0.00 0.00 21.25 1.93 0.152 Up 3.84 Serine protease inhibitor Kazal-type 5 Q9NQ38 0.00 0.00 0.00 0.00 1.25 0.00 1.18 0.00 1.12 4.90 2.13 0.00 0.152 Up 1.87 Carbonic anhydrase 1 P00915 48.70 0.00 1.70 199.77 312.53 0.00 83.24 99.88 162.18 119.34 306.02 473.33 0.169 Up 1.47 Ig lambda chain V-III region LOI P80748 0.00 0.00 0.85 14.88 32.50 2.36 2.95 57.38 1.12 8.17 23.38 17.39 0.169 Up 1.45

Peptidyl-prolyl cis-trans isomerase A P62937 0.00 0.00 0.00 2.13 3.75 0.00 0.59 0.00 2.24 3.27 2.13 3.86 0.177 Up 1.22 Ig kappa chain V-I region AU P01594 10.63 0.00 7.65 0.00 33.75 0.00 22.43 40.38 3.36 4.90 12.75 15.46 0.177 Up 1.05 Ig gamma-2 chain C region P01859 5.31 50.23 27.20 19.13 40.00 33.06 72.61 242.27 11.19 21.25 229.52 63.76 0.180 Up 3.66 N(G),N(G)-dimethylarginine dimethylaminohydrolase 2 O95865 0.00 0.00 0.00 0.00 0.00 0.00 6.49 2.13 1.12 0.00 0.00 0.00 0.182 Up N/A Triosephosphate isomerase P60174 0.00 0.00 0.00 0.00 0.00 0.00 1.18 0.00 0.00 3.27 0.00 11.59 0.182 Up N/A Fatty acid-binding protein, epidermal Q01469 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 5.59 1.63 2.13 0.00 0.182 Up N/A Phosphatidylethanolamine-binding protein 4 Q96S96 0.00 0.00 0.00 0.00 0.00 0.00 1.18 0.00 2.24 0.00 2.13 0.00 0.182 Up N/A DNA repair protein REV1 Q9UBZ9 0.00 0.00 0.00 0.00 0.00 0.00 0.59 4.25 0.00 0.00 2.13 0.00 0.182 Up N/A Metalloproteinase inhibitor 1 P01033 0.00 0.00 0.00 0.00 0.00 0.00 0.59 0.00 2.24 0.00 2.13 0.00 0.182 Up N/A Complement component C8 gamma chain P07360 0.00 0.00 0.43 0.00 0.00 0.00 0.00 0.00 1.12 8.17 34.00 0.00 0.182 Up 33.95 Transthyretin P02766 0.00 0.00 0.00 0.00 1.25 0.00 0.00 0.00 5.59 0.00 10.63 1.93 0.182 Up 4.84 Vesicular integral-membrane protein VIP36 Q12907 0.89 1.93 8.50 21.25 28.75 11.81 30.70 12.75 3.36 16.35 21.25 28.98 0.193 Up 1.55 Alpha-1-antichymotrypsin P01011 4.43 7.73 7.23 19.13 2.50 23.61 54.90 131.76 2.24 19.62 85.01 7.73 0.197 Up 4.66 N-acetylmuramoyl-L-alanine amidase Q96PD5 0.00 0.00 0.00 0.00 7.50 0.00 2.95 23.38 0.00 0.00 6.38 3.86 0.197 Up 1.22 Abhydrolase domain-containing protein 14B Q96IU4 0.00 0.00 0.00 2.13 0.00 0.00 5.90 0.00 1.12 0.00 2.13 1.93 0.197 Up 1.30 Neutrophil defensin 1 P59665 0.00 0.00 1.28 4.25 5.00 0.00 2.36 8.50 2.24 14.71 4.25 0.00 0.203 Up 1.83 Uteroglobin P11684 0.00 0.00 1.28 0.00 5.00 0.00 17.71 2.13 3.36 0.00 6.38 0.00 0.210 Up 2.36 Trypsin-2 P07478 0.00 0.00 0.00 14.88 11.25 0.00 0.59 0.00 6.71 34.33 21.25 3.86 0.221 Up 1.02 Lysozyme C P61626 0.89 0.00 0.43 0.00 2.50 0.00 2.36 6.38 23.49 0.00 31.88 0.00 0.221 Up 12.62 Prothymosin alpha P06454 0.00 0.00 0.43 4.25 1.25 0.00 11.22 0.00 6.71 1.63 38.25 0.00 0.221 Up 7.32 Cathepsin Z Q9UBR2 0.00 0.00 0.85 12.75 8.75 0.00 3.54 6.38 1.12 14.71 6.38 1.93 0.221 Up 1.31 Pancreatic secretory trypsin inhibitor P00995 0.00 0.00 1.70 2.13 3.75 9.45 12.40 0.00 3.36 94.82 10.63 1.93 0.234 Up 5.79 Ig kappa chain V-I region Scw P01609 13.28 0.00 7.23 0.00 35.00 0.00 25.97 59.50 3.36 4.90 12.75 15.46 0.234 Up 1.10 Ig kappa chain C region P01834 294.87 142.97 230.37 507.92 461.29 453.37 322.91 733.18 278.51 315.51 558.92 562.20 0.240 Up 1.33 Ig alpha-1 chain C region P01876 1.77 27.05 22.10 36.13 62.51 37.78 48.41 116.88 10.07 42.50 97.76 28.98 0.240 Up 1.84 Ig gamma-3 chain C region P01860 13.28 44.44 29.33 25.50 50.00 25.97 68.48 221.02 21.25 19.62 284.77 79.21 0.240 Up 3.68 Cystatin-C P01034 0.00 3.86 3.40 4.25 6.25 7.08 2.95 14.88 23.49 17.98 38.25 1.93 0.240 Up 3.34 Fibrinogen gamma chain P02679 0.89 1.93 1.28 2.13 12.50 0.00 17.12 2.13 0.00 6.54 40.38 1.93 0.260 Up 3.64

N-acetylglucosamine-6-sulfatase P15586 1.77 0.00 5.53 2.13 0.00 0.00 4.13 4.25 0.00 8.17 0.00 9.66 0.286 Up 2.09 Ubiquitin carboxyl-terminal hydrolase 36 Q9P275 0.89 5.80 0.85 2.13 1.25 0.00 1.77 12.75 1.12 0.00 6.38 5.80 0.290 Up 2.55 Apolipoprotein C-II P02655 0.00 0.00 0.00 0.00 0.00 2.36 0.00 0.00 6.71 1.63 10.63 0.00 0.303 Up 2.68 Polyubiquitin-B P0CG47 2.66 0.00 4.25 17.00 35.00 0.00 14.17 0.00 19.01 6.54 63.76 27.05 0.307 Up 1.77 Protein FAM3C Q92520 0.00 0.00 0.00 2.13 6.25 0.00 2.36 0.00 3.36 8.17 2.13 0.00 0.307 Up 1.05 Ig gamma-1 chain C region P01857 25.68 69.55 39.95 70.13 65.01 40.14 73.79 233.77 20.13 35.96 299.65 98.53 0.310 Up 2.45 Ig lambda-2 chain C regions P0CG05 35.42 17.39 31.45 89.26 56.25 110.98 79.69 104.13 54.81 57.22 180.64 40.57 0.310 Up 1.52 SH3 domain-binding glutamic acid- rich-like protein 3 Q9H299 7.97 7.73 8.08 40.38 11.25 61.39 35.42 17.00 16.78 16.35 42.50 17.39 0.310 Up 1.06 Ig heavy chain V-III region GAL P01781 0.00 0.00 0.43 0.00 1.25 0.00 2.36 10.63 0.00 0.00 6.38 0.00 0.318 Up 7.71 Chromogranin-A P10645 0.89 0.00 0.43 0.00 0.00 0.00 1.18 0.00 4.47 0.00 4.25 0.00 0.318 Up 5.04 Ig kappa chain V-I region Wes P01611 0.00 0.00 0.43 0.00 6.25 7.08 2.95 12.75 2.24 0.00 6.38 3.86 0.327 Up 1.23 Antithrombin-III P01008 0.00 0.00 0.43 2.13 10.00 0.00 0.00 17.00 1.12 0.00 14.88 7.73 0.372 Up 2.43 Alpha-2-macroglobulin P01023 0.00 5.80 2.98 2.13 8.75 4.72 19.48 40.38 0.00 8.17 38.25 0.00 0.387 Up 5.45 Protein AMBP P02760 177.98 218.31 250.35 461.16 413.78 998.83 402.01 465.41 304.24 953.06 612.05 347.76 0.394 Up 1.22

Leucine-rich alpha-2-glycoprotein P02750 15.94 0.00 59.08 48.88 77.51 54.31 107.44 80.76 33.56 168.38 34.00 30.91 0.394 Up 1.48 Immunoglobulin lambda-like polypeptide 5 B9A064 24.79 17.39 26.35 61.63 40.00 110.98 47.23 82.88 23.49 75.20 106.26 48.30 0.394 Up 1.36 Ig kappa chain V-III region VG (Fragment) P04433 2.66 5.80 5.53 23.38 5.00 9.45 3.54 10.63 7.83 8.17 25.50 5.80 0.416 Up 1.19 Beta-2-microglobulin P61769 16.82 3.86 8.50 74.38 18.75 37.78 24.20 19.13 30.20 191.27 40.38 3.86 0.418 Up 1.93 3-mercaptopyruvate sulfurtransferase P25325 0.00 3.86 0.00 0.00 0.00 0.00 1.18 0.00 0.00 1.63 0.00 5.80 0.424 Up 1.35 Myc box-dependent-interacting protein 1 O00499 0.00 0.00 0.00 0.00 0.00 2.36 1.18 0.00 0.00 3.27 2.13 0.00 0.424 Up 1.08 Glutaredoxin-1 P35754 0.00 0.00 0.00 4.25 0.00 2.36 4.13 0.00 1.12 1.63 6.38 0.00 0.437 Up 1.00 Gastrotropin P51161 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 17.90 0.00 0.00 1.93 0.455 Up N/A

Fructose-bisphosphate aldolase B P05062 0.00 0.00 0.00 0.00 1.25 0.00 7.67 0.00 0.00 0.00 4.25 0.00 0.455 Up 4.77 Thyroxine-binding globulin P05543 0.00 0.00 0.43 0.00 0.00 0.00 1.18 12.75 0.00 0.00 0.00 0.00 0.455 Up 16.39 Complement component C9 P02748 0.00 0.00 0.00 0.00 0.00 0.00 0.59 0.00 0.00 0.00 12.75 0.00 0.455 Up N/A Phosphoglycerate kinase 1 P00558 0.00 0.00 0.00 0.00 1.25 0.00 1.77 0.00 0.00 0.00 0.00 5.80 0.455 Up 3.03 Thioredoxin P10599 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 10.63 3.86 0.455 Up N/A Alpha-2-antiplasmin P08697 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2.13 0.00 0.00 12.75 0.00 0.455 Up N/A Junctional adhesion molecule A Q9Y624 0.00 0.00 0.43 0.00 0.00 0.00 1.77 0.00 0.00 0.00 4.25 0.00 0.455 Up 7.08 C-reactive protein P02741 0.00 0.00 0.43 0.00 0.00 0.00 1.18 0.00 0.00 0.00 6.38 0.00 0.455 Up 8.89 Complement factor H P08603 0.00 0.00 0.00 0.00 0.00 0.00 2.95 0.00 0.00 0.00 2.13 0.00 0.455 Up N/A Dermokine Q6E0U4 0.00 0.00 0.43 0.00 0.00 0.00 0.00 0.00 0.00 1.63 4.25 0.00 0.455 Up 6.92

Intercellular adhesion molecule 2 P13598 0.00 0.00 0.00 0.00 0.00 0.00 2.95 0.00 0.00 1.63 0.00 0.00 0.455 Up N/A Protein NOV homolog P48745 0.00 0.00 0.43 0.00 0.00 0.00 0.59 0.00 0.00 4.90 0.00 0.00 0.455 Up 6.46 Connective tissue growth factor P29279 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.47 0.00 2.13 0.00 0.455 Up N/A Inter-alpha-trypsin inhibitor heavy chain H2 P19823 0.00 0.00 0.00 0.00 0.00 0.00 0.59 0.00 0.00 0.00 8.50 0.00 0.455 Up N/A Protein S100-A6 P06703 0.00 0.00 0.43 0.00 0.00 0.00 1.18 0.00 0.00 0.00 0.00 1.93 0.455 Up 3.66 Alpha-hemoglobin-stabilizing protein Q9NZD4 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.12 0.00 6.38 0.00 0.455 Up N/A Complement factor H-related protein 2 P36980 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.90 2.13 0.00 0.455 Up N/A Insulin-like growth factor-binding protein 4 P22692 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2.24 0.00 2.13 0.00 0.455 Up N/A Macrophage colony-stimulating factor 1 P09603 0.00 0.00 1.28 6.38 0.00 7.08 10.63 0.00 4.47 14.71 2.13 0.00 0.470 Up 1.63 Complement component C7 P10643 0.00 0.00 1.28 2.13 1.25 0.00 2.36 4.25 0.00 0.00 21.25 0.00 0.481 Up 5.99 Cadherin-13 P55290 0.89 0.00 3.83 0.00 10.00 2.36 4.13 4.25 0.00 3.27 8.50 1.93 0.483 Up 1.03 Ig kappa chain V-I region EU P01598 0.89 0.00 4.68 4.25 8.75 7.08 12.40 17.00 2.24 1.63 8.50 3.86 0.485 Up 1.48 Ig alpha-2 chain C region P01877 0.89 7.73 10.20 10.63 12.50 11.81 14.17 40.38 10.07 3.27 61.63 9.66 0.485 Up 2.59 Immunoglobulin J chain P01591 0.00 0.00 2.98 8.50 2.50 0.00 4.72 2.13 3.36 3.27 0.00 1.93 0.498 Up 1.51 Guanylin Q02747 2.66 0.00 0.43 12.75 18.75 16.53 2.36 6.38 17.90 27.79 12.75 9.66 0.509 Up 1.25 Thrombospondin-1 P07996 2.66 0.00 0.00 0.00 5.00 0.00 8.26 2.13 0.00 1.63 0.00 1.93 0.545 Up 1.10 Endothelial protein C receptor Q9UNN8 1.77 0.00 2.55 0.00 1.25 0.00 9.45 0.00 0.00 0.00 0.00 0.00 0.545 Up 5.09 Fibulin-1 P23142 0.00 0.00 0.00 2.13 2.50 0.00 9.45 0.00 0.00 4.90 0.00 1.93 0.545 Up 2.35 SPARC-like protein 1 Q14515 0.00 3.86 0.00 2.13 0.00 0.00 6.49 0.00 1.12 0.00 10.63 0.00 0.545 Up 2.03 Nuclear transport factor 2 P61970 0.00 0.00 0.00 2.13 3.75 0.00 6.49 0.00 1.12 0.00 0.00 3.86 0.545 Up 1.30

L-lactate dehydrogenase B chain P07195 0.00 0.00 0.43 0.00 2.50 0.00 2.36 0.00 0.00 3.27 0.00 1.93 0.545 Up 1.72 Lithostathine-1-beta P48304 0.00 0.00 1.28 21.25 28.75 49.59 16.53 10.63 2.24 45.77 25.50 25.12 0.554 Up 1.20 Haptoglobin P00738 0.00 7.73 21.25 14.88 13.75 11.81 16.53 2.13 20.13 17.98 6.38 17.39 0.589 Up 1.03 Nidogen-1 P14543 2.66 1.93 1.28 14.88 23.75 7.08 27.15 8.50 0.00 8.17 4.25 19.32 0.589 Up 1.57 Tumor necrosis factor receptor superfamily member 12A Q9NP84 0.00 0.00 0.85 2.13 1.25 0.00 2.95 0.00 0.00 1.63 21.25 0.00 0.589 Up 6.12 Proactivator polypeptide P07602 0.89 0.00 0.85 0.00 1.25 0.00 1.18 0.00 2.24 0.00 2.13 0.00 0.589 Up 1.86 6-phosphogluconolactonase O95336 0.00 0.00 0.00 4.25 3.75 2.36 15.94 0.00 1.12 1.63 2.13 3.86 0.606 Up 1.43 Ig lambda chain V-I region HA P01700 0.00 1.93 0.00 8.50 8.75 0.00 2.36 6.38 0.00 4.90 2.13 3.86 0.606 Up 1.63 Fibulin-5 Q9UBX5 1.77 0.00 5.95 2.13 15.00 37.78 17.71 6.38 0.00 14.71 14.88 9.66 0.623 Up 1.01 Syndecan-1 P18827 0.00 0.00 0.00 2.13 1.25 0.00 4.72 0.00 0.00 4.90 0.00 0.00 0.697 Up 2.85 Cathepsin L1 P07711 0.00 0.00 0.43 0.00 2.50 0.00 1.18 0.00 1.12 1.63 0.00 0.00 0.697 Up 1.12 Apolipoprotein D P05090 1.77 13.52 149.19 25.50 40.00 99.17 58.44 238.02 0.00 91.55 46.75 27.05 0.699 Up 1.68 Kininogen-1 P01042 15.94 77.28 30.18 68.01 72.51 61.39 64.94 76.51 43.62 39.23 97.76 61.82 0.699 Up 1.18 Prothrombin P00734 0.89 67.62 23.80 34.00 12.50 40.14 42.50 14.88 24.61 53.95 38.25 17.39 0.699 Up 1.07 Ig kappa chain V-I region AG P01593 12.40 0.00 13.60 19.13 37.50 7.08 27.15 40.38 3.36 6.54 14.88 17.39 0.699 Up 1.02 EGF-containing fibulin-like extracellular matrix protein 1 Q12805 0.89 9.66 11.48 12.75 16.25 14.17 28.34 6.38 3.36 45.77 38.25 3.86 0.699 Up 1.93 Liver-expressed antimicrobial peptide 2 Q969E1 7.97 9.66 2.55 6.38 6.25 28.34 7.08 2.13 5.59 13.08 38.25 1.93 0.699 Up 1.11 Peptidase inhibitor 16 Q6UXB8 0.00 0.00 0.85 2.13 1.25 0.00 4.72 0.00 1.12 0.00 4.25 0.00 0.708 Up 2.39 Vascular cell adhesion protein 1 P19320 0.00 0.00 2.98 2.13 15.00 0.00 9.45 0.00 0.00 19.62 2.13 1.93 0.719 Up 1.24 Carbonic anhydrase 2 P00918 0.00 0.00 0.00 0.00 7.50 0.00 0.59 0.00 0.00 0.00 0.00 52.16 0.727 Up 3.52 Prostate-specific antigen P07288 0.00 0.00 0.00 0.00 10.00 0.00 0.00 0.00 11.19 9.81 0.00 0.00 0.727 Up 1.05 Acyl-CoA-binding protein P07108 0.00 0.00 0.00 0.00 1.25 0.00 0.00 0.00 1.12 0.00 21.25 0.00 0.727 Up 8.95 V-set and immunoglobulin domain- containing protein 4 Q9Y279 0.00 0.00 0.00 0.00 0.00 2.36 1.18 0.00 0.00 0.00 14.88 0.00 0.727 Up 3.40 Lumican P51884 0.00 0.00 0.00 0.00 2.50 0.00 2.36 0.00 0.00 0.00 0.00 3.86 0.727 Up 1.24 ADM P35318 0.00 0.00 0.00 0.00 0.00 2.36 0.00 0.00 4.47 0.00 2.13 0.00 0.727 Up 1.40 Tyrosine-protein phosphatase non- receptor type substrate 1 P78324 0.00 1.93 0.00 0.00 0.00 0.00 2.36 0.00 0.00 1.63 0.00 0.00 0.727 Up 1.03 Lymphatic vessel endothelial hyaluronic acid receptor 1 Q9Y5Y7 0.00 0.00 11.05 8.50 16.25 2.36 13.58 8.50 1.12 1.63 12.75 5.80 0.729 Up 1.32 Actin, cytoplasmic 1 P60709 0.00 3.86 4.25 12.75 17.50 33.06 13.58 0.00 7.83 21.25 19.13 9.66 0.738 Up 1.00 Vitronectin P04004 0.89 1.93 6.38 0.00 3.75 9.45 8.85 2.13 0.00 4.90 4.25 3.86 0.738 Up 1.07 Complement decay-accelerating factor P08174 0.00 1.93 2.13 8.50 3.75 0.00 14.17 0.00 0.00 3.27 0.00 1.93 0.740 Up 1.58 Lithostathine-1-alpha P05451 0.00 0.00 2.13 31.88 47.50 49.59 24.79 10.63 4.47 44.14 40.38 25.12 0.777 Up 1.32 Plasminogen P00747 0.00 0.00 2.55 6.38 3.75 4.72 4.13 0.00 17.90 1.63 65.88 0.00 0.797 Up 5.15 Glyceraldehyde-3-phosphate dehydrogenase P04406 0.00 0.00 0.43 0.00 2.50 0.00 8.85 0.00 0.00 0.00 0.00 1.93 0.848 Up 3.69 Tripeptidyl-peptidase 1 O14773 0.00 0.00 1.28 0.00 3.75 0.00 2.36 0.00 0.00 0.00 0.00 3.86 0.848 Up 1.24 Complement factor I P05156 0.00 0.00 0.85 0.00 0.00 2.36 3.54 0.00 0.00 0.00 0.00 1.93 0.848 Up 1.70 Insulin-like growth factor-binding protein 3 P17936 0.00 0.00 0.43 0.00 1.25 0.00 2.36 0.00 1.12 0.00 0.00 0.00 0.848 Up 2.08 Hemoglobin subunit gamma-1 P69891 0.00 0.00 0.43 0.00 3.75 0.00 1.18 0.00 0.00 0.00 0.00 27.05 0.848 Up 6.76 WNT1-inducible-signaling pathway protein 2 O76076 0.00 0.00 1.28 2.13 0.00 2.36 6.49 0.00 0.00 1.63 0.00 0.00 0.848 Up 2.12 Lactotransferrin P02788 1.77 0.00 14.03 0.00 1.25 0.00 11.22 0.00 0.00 13.08 0.00 1.93 0.870 Up 1.54 Endonuclease domain-containing 1 protein O94919 0.00 0.00 0.85 2.13 1.25 0.00 6.49 0.00 0.00 1.63 0.00 0.00 0.924 Up 2.89 Secretogranin-1 P05060 1.77 0.00 0.85 0.00 2.50 0.00 4.13 0.00 2.24 0.00 0.00 0.00 0.924 Up 1.87 Hemoglobin subunit beta P68871 4.43 23.18 8.93 29.75 108.76 9.45 12.40 2.13 8.95 4.90 42.50 481.06 0.937 Up 2.99 Hemoglobin subunit alpha P69905 1.77 21.25 3.40 51.00 85.01 7.08 9.45 0.00 4.47 6.54 25.50 548.68 0.937 Up 4.21 Gelsolin P06396 4.43 3.86 10.20 48.88 25.00 59.03 36.60 12.75 3.36 58.85 44.63 21.25 0.937 Up 1.17 Cathepsin D P07339 0.89 1.93 9.78 12.75 21.25 16.53 17.71 10.63 3.36 9.81 14.88 7.73 0.937 Up 1.02 Lysosomal alpha-glucosidase P10253 1.77 9.66 2.98 0.00 5.00 4.72 7.08 2.13 2.24 6.54 4.25 1.93 0.937 Up 1.20 Latent-transforming growth factor beta-binding protein 2 Q14767 0.00 0.00 1.28 2.13 1.25 2.36 1.77 0.00 0.00 13.08 2.13 0.00 0.957 Up 3.23 Collagen alpha-3(VI) chain P12111 0.00 0.00 1.70 4.25 2.50 2.36 2.36 0.00 1.12 9.81 2.13 1.93 0.972 Up 1.28 Collagen alpha-1(I) chain P02452 3.54 0.00 4.68 0.00 6.25 4.72 4.72 4.25 3.36 4.90 2.13 0.00 0.974 Up 1.24 Hemoglobin subunit delta P02042 1.77 3.86 3.40 8.50 61.25 0.00 2.95 0.00 7.83 0.00 8.50 260.82 0.974 Up 4.44 Latent-transforming growth factor beta-binding protein 1 Q14766 0.00 0.00 0.00 2.13 1.25 2.36 1.18 0.00 1.12 1.63 2.13 0.00 0.978 Up 1.26 Alpha-amylase 1 P04745 0.00 1.93 2.13 27.63 5.00 18.89 183.00 0.00 3.36 13.08 0.00 5.80 1.000 Up 4.62 Insulin-like growth factor-binding protein 7 Q16270 3.54 3.86 6.38 8.50 27.50 33.06 29.52 31.88 1.12 19.62 2.13 9.66 1.000 Up 1.13 Semenogelin-2 Q02383 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 147.13 0.00 0.00 1.000 Up N/A Granulins P28799 0.89 5.80 4.68 4.25 1.25 21.25 27.15 6.38 0.00 8.17 0.00 1.93 1.000 Up 1.72 Fibronectin P02751 0.00 0.00 2.13 0.00 1.25 7.08 17.71 0.00 0.00 14.71 0.00 0.00 1.000 Up 4.65 Vasorin Q6EMK4 1.77 5.80 2.55 0.00 3.75 0.00 11.22 0.00 0.00 1.63 2.13 3.86 1.000 Up 1.36 Low-density lipoprotein receptor- related protein 2 P98164 0.89 0.00 2.98 4.25 0.00 2.36 7.08 2.13 0.00 0.00 0.00 7.73 1.000 Up 2.16

Cathelicidin antimicrobial peptide P49913 0.00 0.00 0.00 4.25 0.00 7.08 12.40 0.00 0.00 3.27 0.00 0.00 1.000 Up 1.38 Mucin-like protein 1 Q96DR8 0.00 0.00 0.00 2.13 5.00 0.00 1.18 0.00 0.00 27.79 0.00 0.00 1.000 Up 4.07 Semenogelin-1 P04279 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 73.56 0.00 0.00 1.000 Up N/A Deleted in malignant brain tumors 1 protein Q9UGM3 0.00 0.00 0.00 0.00 0.00 0.00 10.04 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A

Pyruvate kinase isozymes M1/M2 P14618 0.00 0.00 0.00 0.00 1.25 0.00 5.31 0.00 0.00 0.00 0.00 0.00 1.000 Up 4.25 Tumor necrosis factor receptor superfamily member 16 P08138 0.00 0.00 0.00 0.00 0.00 2.36 4.72 0.00 0.00 0.00 0.00 0.00 1.000 Up 2.00 Zymogen granule protein 16 homolog B Q96DA0 0.00 0.00 0.00 0.00 2.50 0.00 1.77 0.00 1.12 0.00 0.00 0.00 1.000 Up 1.73 Calreticulin P27797 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 14.88 0.00 1.000 Up N/A Flavin reductase P30043 0.00 0.00 0.00 0.00 3.75 0.00 1.77 0.00 0.00 0.00 2.13 0.00 1.000 Up 1.93 Bisphosphoglycerate mutase P07738 0.00 0.00 0.00 0.00 0.00 2.36 0.00 0.00 0.00 0.00 0.00 9.66 1.000 Up 4.09

Cadherin-related family member 2 Q9BYE9 0.00 0.00 0.43 0.00 1.25 0.00 1.77 0.00 0.00 0.00 0.00 0.00 1.000 Up 2.11 Cadherin-11 P55287 0.00 0.00 0.43 0.00 0.00 0.00 1.77 0.00 0.00 0.00 0.00 0.00 1.000 Up 4.17 Prostatic acid phosphatase P15309 0.00 0.00 0.00 0.00 2.50 0.00 0.00 0.00 3.36 0.00 0.00 0.00 1.000 Up 1.34 Beta-microseminoprotein P08118 0.00 0.00 0.00 0.00 2.50 0.00 0.00 0.00 4.47 0.00 0.00 0.00 1.000 Up 1.79 Ubiquitin carboxyl-terminal hydrolase 8 P40818 0.00 0.00 0.00 0.00 0.00 0.00 0.00 8.50 0.00 0.00 0.00 0.00 1.000 Up N/A Signal-regulatory protein beta-1 O00241 0.00 0.00 0.00 0.00 0.00 2.36 1.77 0.00 0.00 1.63 0.00 0.00 1.000 Up 1.39 Chitinase-3-like protein 1 P36222 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2.24 0.00 0.00 0.00 1.000 Up N/A Heparin cofactor 2 P05546 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 8.50 0.00 1.000 Up N/A Complement C2 P06681 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.25 0.00 0.00 0.00 0.00 1.000 Up N/A Aminoacylase-1 Q03154 0.00 0.00 0.00 0.00 1.25 0.00 1.77 0.00 0.00 0.00 0.00 0.00 1.000 Up 1.42 Thioredoxin domain-containing protein 5 Q8NBS9 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 6.38 0.00 1.000 Up N/A Mesothelin Q13421 0.00 0.00 0.00 0.00 0.00 0.00 1.77 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A

Glutaminyl-peptide cyclotransferase Q16769 0.00 0.00 0.00 0.00 0.00 0.00 2.36 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Cystatin-S P01036 0.00 0.00 0.00 0.00 0.00 0.00 2.36 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Di-N-acetylchitobiase Q01459 0.00 0.00 0.00 0.00 0.00 0.00 2.36 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Matrix-remodeling-associated protein 8 Q9BRK3 0.00 0.00 0.00 0.00 0.00 0.00 1.77 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Mucin-5B Q9HC84 0.00 0.00 0.00 0.00 0.00 0.00 2.36 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Copper transport protein ATOX1 O00244 0.00 0.00 0.00 0.00 0.00 0.00 2.36 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A CMRF35-like molecule 1 Q8TDQ1 0.00 0.00 0.00 0.00 0.00 0.00 2.36 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Neprilysin P08473 0.00 0.00 0.00 0.00 0.00 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A

Cytosolic non-specific dipeptidase Q96KP4 0.00 0.00 0.00 0.00 0.00 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Ephrin type-B receptor 4 P54760 0.00 0.00 0.00 0.00 0.00 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Fructose-1,6-bisphosphatase 1 P09467 0.00 0.00 0.00 0.00 0.00 0.00 1.77 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Peroxiredoxin-6 P30041 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 5.80 1.000 Up N/A 5'(3')-deoxyribonucleotidase, cytosolic type Q8TCD5 0.00 0.00 0.00 0.00 0.00 0.00 1.77 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Inter-alpha-trypsin inhibitor heavy chain H1 P19827 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.25 0.00 1.000 Up N/A D-dopachrome decarboxylase P30046 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.25 0.00 1.000 Up N/A Collagen alpha-1(XV) chain P39059 0.00 0.00 0.00 0.00 0.00 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Folate receptor gamma P41439 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.25 0.00 1.000 Up N/A Elongation factor 1-alpha 1 P68104 0.00 0.00 0.00 0.00 0.00 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Gamma-glutamyl hydrolase Q92820 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 3.86 1.000 Up N/A Heat shock protein beta-11 Q9Y547 0.00 0.00 0.00 0.00 0.00 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Trans-Golgi network integral membrane protein 2 O43493 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2.24 0.00 0.00 0.00 1.000 Up N/A

Tetratricopeptide repeat protein 17 Q96AE7 0.00 0.00 0.00 0.00 0.00 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Up N/A Keratin, type II cytoskeletal 5 P13647 2.66 1.93 7.65 4.25 1.25 4.72 1.18 0.00 3.36 0.00 0.00 1.93 0.028 Down 1.74 Secreted Ly-6/uPAR-related protein 1 P55000 1.77 1.93 3.83 6.38 1.25 11.81 4.72 0.00 0.00 1.63 0.00 0.00 0.035 Down 1.41 Non-secretory ribonuclease P10153 7.08 17.39 11.48 27.63 6.25 33.06 21.84 6.38 3.36 4.90 2.13 0.00 0.041 Down 2.22 Keratin, type II cytoskeletal 1 P04264 17.71 17.39 36.98 31.88 23.75 87.37 15.35 10.63 10.07 49.04 10.63 17.39 0.048 Down 1.90 Leukocyte-associated immunoglobulin-like receptor 1 Q6GTX8 0.00 5.80 0.85 4.25 1.25 2.36 4.13 0.00 0.00 0.00 0.00 0.00 0.054 Down 1.42 Keratin, type I cytoskeletal 10 P13645 1.77 13.52 48.88 25.50 8.75 21.25 12.40 0.00 5.59 14.71 0.00 5.80 0.089 Down 2.07 Keratin, type I cytoskeletal 9 P35527 6.20 5.80 13.18 29.75 15.00 28.34 5.90 4.25 3.36 8.17 2.13 21.25 0.093 Down 2.18 Immunoglobulin superfamily member 8 Q969P0 0.00 3.86 0.43 6.38 2.50 40.14 3.54 0.00 0.00 3.27 0.00 0.00 0.113 Down 3.13 CD27 antigen P26842 0.89 1.93 1.70 2.13 0.00 2.36 4.13 0.00 0.00 0.00 0.00 0.00 0.113 Down 2.29 CD44 antigen P16070 3.54 13.52 5.95 4.25 6.25 11.81 9.45 0.00 0.00 4.90 0.00 5.80 0.128 Down 1.12 Extracellular sulfatase Sulf-2 Q8IWU5 3.54 19.32 5.10 17.00 21.25 61.39 6.49 0.00 1.12 27.79 2.13 3.86 0.132 Down 2.57 Keratin, type I cytoskeletal 14 P02533 0.00 0.00 5.10 4.25 0.00 4.72 0.00 0.00 1.12 0.00 0.00 0.00 0.182 Down 4.19 Kallikrein-1 P06870 0.00 11.59 0.85 0.00 13.75 11.81 11.22 0.00 0.00 1.63 0.00 0.00 0.210 Down 1.48 Pepsin A P00790 17.71 3.86 8.08 8.50 10.00 54.31 7.67 2.13 4.47 19.62 2.13 9.66 0.225 Down 2.24 Keratin, type II cytoskeletal 8 P05787 0.89 3.86 2.13 14.88 2.50 2.36 3.54 0.00 3.36 1.63 0.00 1.93 0.225 Down 1.70 CMRF35-like molecule 9 Q6UXG3 0.00 5.80 0.43 4.25 0.00 7.08 1.18 4.25 0.00 0.00 0.00 0.00 0.242 Down 1.62 Cubilin O60494 1.77 5.80 1.70 0.00 5.00 7.08 8.85 4.25 0.00 0.00 0.00 0.00 0.286 Down 1.53 Tumor necrosis factor receptor superfamily member 14 Q92956 0.89 0.00 0.00 2.13 1.25 0.00 1.18 0.00 0.00 0.00 0.00 0.00 0.303 Down 1.20 Secreted and transmembrane protein 1 Q8WVN6 2.66 0.00 0.00 2.13 3.75 4.72 2.95 0.00 1.12 0.00 2.13 0.00 0.307 Down 1.60 Uromodulin P07911 80.58 177.74 51.43 463.29 207.52 443.93 183.59 144.51 34.67 349.84 78.63 54.10 0.310 Down 1.69 CD59 glycoprotein P13987 18.60 40.57 40.38 61.63 32.50 181.82 45.46 27.63 11.19 94.82 14.88 27.05 0.310 Down 1.70 Ig kappa chain V-III region NG9 (Fragment) P01621 1.77 5.80 5.10 17.00 8.75 16.53 6.49 4.25 5.59 4.90 14.88 0.00 0.310 Down 1.27 Protein shisa-5 Q8N114 0.00 0.00 2.13 14.88 3.75 14.17 3.54 0.00 1.12 9.81 0.00 0.00 0.372 Down 1.81 Carboxypeptidase N subunit 2 P22792 0.00 1.93 0.43 0.00 2.50 4.72 4.13 0.00 0.00 1.63 0.00 0.00 0.372 Down 1.20 Pro-epidermal growth factor P01133 0.89 11.59 2.13 0.00 0.00 0.00 2.95 2.13 1.12 0.00 4.25 1.93 0.381 Down 1.97 Ig kappa chain V-IV region Len P01625 3.54 5.80 2.98 2.13 16.25 4.72 12.99 0.00 10.07 3.27 0.00 0.00 0.387 Down 1.49 Keratin, type II cytoskeletal 2 epidermal P35908 7.97 1.93 50.15 6.38 7.50 2.36 5.90 8.50 6.71 1.63 0.00 5.80 0.394 Down 2.23 Trefoil factor 1 P04155 0.00 0.00 0.43 0.00 2.50 25.97 8.26 2.13 1.12 3.27 2.13 0.00 0.411 Down 2.85 Phosphoinositide-3-kinase- interacting protein 1 Q96FE7 7.08 9.66 12.33 2.13 13.75 18.89 33.06 6.38 0.00 8.17 2.13 11.59 0.420 Down 1.15 Platelet glycoprotein VI Q9HCN6 0.00 1.93 0.85 0.00 1.25 0.00 1.77 0.00 0.00 0.00 0.00 0.00 0.424 Down 1.32 Alpha-enolase P06733 0.00 1.93 0.43 2.13 1.25 0.00 3.54 0.00 1.12 0.00 0.00 0.00 0.437 Down 1.63 Cathepsin B P07858 0.00 1.93 0.43 6.38 2.50 2.36 2.36 2.13 0.00 3.27 0.00 0.00 0.450 Down 1.05 Src substrate cortactin Q14247 0.00 3.86 0.00 6.38 0.00 0.00 1.77 0.00 0.00 0.00 0.00 0.00 0.455 Down 2.89 Matrix metalloproteinase-9 P14780 0.00 0.00 0.85 0.00 0.00 2.36 0.00 0.00 0.00 0.00 0.00 0.00 0.455 Down N/A Protein YIPF3 Q9GZM5 0.89 3.86 4.25 8.50 6.25 25.97 8.26 0.00 0.00 11.44 0.00 7.73 0.483 Down 1.10 Basement membrane-specific heparan sulfate proteoglycan core protein P98160 10.63 21.25 27.20 80.76 42.50 247.94 46.64 40.38 26.84 170.01 99.88 44.44 0.485 Down 1.00 Trefoil factor 2 Q03403 0.00 3.86 3.83 4.25 1.25 18.89 6.49 2.13 0.00 9.81 0.00 0.00 0.498 Down 1.04 Ig kappa chain V-II region MIL P01616 1.77 7.73 8.50 14.88 6.25 9.45 11.22 4.25 1.12 8.17 8.50 3.86 0.513 Down 1.31 Agrin O00468 0.00 0.00 0.43 6.38 12.50 7.08 7.08 0.00 0.00 3.27 0.00 3.86 0.524 Down 1.39

Growth/differentiation factor 15 Q99988 0.00 0.00 0.00 0.00 0.00 7.08 2.95 0.00 0.00 0.00 2.13 1.93 0.545 Down 3.03 Biotinidase P43251 0.00 1.93 0.00 2.13 2.50 0.00 4.72 0.00 0.00 0.00 0.00 0.00 0.545 Down 2.16 Amyloid beta A4 protein P05067 0.89 0.00 0.00 2.13 0.00 2.36 2.95 0.00 0.00 0.00 0.00 0.00 0.545 Down 1.65 Ig kappa chain V-II region TEW P01617 5.31 7.73 14.03 21.25 21.25 11.81 20.07 8.50 4.47 14.71 10.63 9.66 0.554 Down 1.20 Mannan-binding lectin serine protease 2 O00187 0.89 1.93 13.60 12.75 2.50 18.89 17.71 6.38 0.00 9.81 0.00 11.59 0.554 Down 1.35 Aminopeptidase N P15144 0.89 0.00 1.70 4.25 3.75 9.45 5.31 4.25 0.00 0.00 0.00 1.93 0.567 Down 1.05 Beta-defensin 1 P60022 0.89 7.73 1.28 2.13 3.75 18.89 4.72 0.00 0.00 3.27 2.13 7.73 0.571 Down 1.29 Cadherin-1 P12830 2.66 1.93 7.23 2.13 2.50 18.89 21.25 0.00 0.00 3.27 0.00 3.86 0.582 Down 1.61 Acid ceramidase Q13510 0.00 1.93 2.55 4.25 2.50 2.36 1.77 2.13 0.00 3.27 0.00 3.86 0.582 Down 1.01 Galectin-3-binding protein Q08380 7.08 9.66 9.78 4.25 6.25 14.17 20.07 10.63 4.47 3.27 6.38 1.93 0.589 Down 1.09

Maltase-glucoamylase, intestinal O43451 2.66 0.00 0.43 2.13 0.00 0.00 0.59 2.13 0.00 0.00 0.00 0.00 0.621 Down 1.28 Collagen alpha-1(VI) chain P12109 0.00 3.86 0.85 0.00 0.00 2.36 2.95 0.00 0.00 1.63 0.00 0.00 0.697 Down 1.03 Peroxiredoxin-2 P32119 0.00 0.00 0.00 4.25 5.00 0.00 0.59 0.00 0.00 0.00 0.00 3.86 0.697 Down 2.08 Tyrosine-protein kinase receptor UFO P30530 0.00 1.93 0.00 4.25 0.00 0.00 1.18 0.00 0.00 1.63 0.00 0.00 0.697 Down 2.20 Osteopontin P10451 0.00 63.76 25.50 119.01 147.51 231.41 97.40 106.26 14.54 130.78 27.63 52.16 0.699 Down 1.64 WAP four-disulfide core domain protein 2 Q14508 47.82 9.66 19.55 53.13 42.50 61.39 67.89 40.38 15.66 37.60 31.88 28.98 0.699 Down 1.05 Interleukin-18-binding protein O95998 0.00 0.00 1.28 0.00 0.00 21.25 2.36 0.00 0.00 0.00 0.00 0.00 0.727 Down 4.77 Colipase P04118 0.00 0.00 0.00 23.38 0.00 2.36 0.00 0.00 0.00 0.00 6.38 0.00 0.727 Down 2.02 Hepatitis A virus cellular receptor 2 Q8TDQ0 0.00 0.00 1.70 0.00 0.00 7.08 3.54 0.00 0.00 0.00 0.00 0.00 0.727 Down 1.24 Low affinity immunoglobulin gamma Fc region receptor III-A P08637 0.00 0.00 0.43 0.00 0.00 7.08 1.18 0.00 0.00 0.00 0.00 0.00 0.727 Down 3.18 Hemicentin-1 Q96RW7 0.00 1.93 0.43 0.00 0.00 0.00 1.77 0.00 0.00 0.00 0.00 0.00 0.727 Down 1.50 Pappalysin-2 Q9BXP8 0.00 0.00 0.43 0.00 0.00 2.36 1.77 0.00 0.00 0.00 0.00 0.00 0.727 Down 1.27 Titin Q8WZ42 0.00 0.00 1.28 0.00 3.75 14.17 2.36 10.63 0.00 3.27 2.13 0.00 0.805 Down 1.39 Resistin Q9HD89 0.00 1.93 0.85 14.88 2.50 14.17 5.31 8.50 0.00 9.81 2.13 0.00 0.814 Down 1.07

Polymeric immunoglobulin receptor P01833 41.62 67.62 24.23 68.01 43.75 85.01 64.35 48.88 26.84 78.47 53.13 25.12 0.818 Down 1.11 Ribonuclease pancreatic P07998 5.31 30.91 10.63 21.25 10.00 35.42 17.71 6.38 6.71 39.23 29.75 5.80 0.818 Down 1.08

Peptidoglycan recognition protein 1 O75594 1.77 9.66 2.98 29.75 10.00 30.70 16.53 0.00 3.36 31.06 6.38 9.66 0.848 Down 1.06

Plasma serine protease inhibitor P05154 0.00 1.93 0.00 2.13 0.00 18.89 10.04 0.00 0.00 4.90 0.00 0.00 0.848 Down 1.02 Vitelline membrane outer layer protein 1 homolog Q7Z5L0 0.00 5.80 2.55 2.13 0.00 35.42 18.30 2.13 0.00 9.81 0.00 1.93 0.857 Down 1.43 Desmocollin-2 Q02487 0.00 0.00 0.85 2.13 0.00 2.36 1.18 0.00 1.12 1.63 0.00 0.00 0.870 Down 1.36 Fibrillin-1 P35555 0.00 7.73 2.55 0.00 2.50 11.81 7.67 6.38 0.00 8.17 2.13 0.00 0.922 Down 1.01 Lipocalin-1 P31025 0.00 0.00 0.00 6.38 18.75 0.00 12.40 0.00 3.36 0.00 0.00 1.93 0.924 Down 2.13 Dipeptidyl peptidase 1 P53634 2.66 0.00 1.70 4.25 1.25 0.00 2.95 2.13 0.00 0.00 0.00 3.86 0.935 Down 1.21 Ig kappa chain V-III region CLL P04207 1.77 3.86 8.50 21.25 5.00 16.53 2.95 6.38 5.59 6.54 12.75 5.80 0.937 Down 1.42 Ig kappa chain V-III region HAH P18135 7.97 7.73 14.88 25.50 22.50 30.70 30.70 12.75 8.95 8.17 19.13 17.39 0.981 Down 1.13 Proteoglycan 4 Q92954 0.89 0.00 2.55 10.63 3.75 21.25 2.95 2.13 2.24 22.89 0.00 5.80 0.981 Down 1.09 Prostaglandin-H2 D-isomerase P41222 64.64 81.14 40.38 99.88 108.76 328.22 115.11 89.26 33.56 219.06 48.88 104.33 1.000 Down 1.18 Cystatin-M Q15828 2.66 5.80 8.93 42.50 22.50 56.67 11.81 2.13 25.73 52.31 10.63 17.39 1.000 Down 1.16 Inter-alpha-trypsin inhibitor heavy chain H4 Q14624 2.66 5.80 6.80 34.00 13.75 54.31 22.43 4.25 10.07 11.44 23.38 9.66 1.000 Down 1.44 Interleukin-1 receptor antagonist protein P18510 0.00 0.00 0.00 8.50 13.75 0.00 15.35 0.00 0.00 6.54 0.00 0.00 1.000 Down 1.02 Trefoil factor 3 Q07654 0.00 0.00 0.43 0.00 3.75 23.61 4.13 0.00 0.00 0.00 6.38 1.93 1.000 Down 2.23 Hepcidin P81172 0.00 1.93 1.70 0.00 6.25 0.00 3.54 0.00 0.00 1.63 0.00 3.86 1.000 Down 1.09 Dipeptidyl peptidase 4 P27487 0.00 0.00 6.80 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.000 Down N/A Submaxillary gland androgen- regulated protein 3B P02814 0.00 3.86 0.00 0.00 2.50 0.00 5.90 0.00 0.00 0.00 0.00 0.00 1.000 Down 1.86 Olfactomedin-4 Q6UX06 0.00 0.00 2.55 0.00 2.50 0.00 3.54 0.00 0.00 0.00 0.00 0.00 1.000 Down 1.40 Platelet factor 4 P02776 (+1) 0.00 0.00 0.00 0.00 6.25 0.00 4.13 0.00 1.12 0.00 0.00 0.00 1.000 Down 2.38 Eukaryotic translation initiation factor 6 P56537 0.00 5.80 0.00 0.00 0.00 0.00 4.72 0.00 0.00 0.00 0.00 0.00 1.000 Down 1.23 Poliovirus receptor-related protein 2 Q92692 0.00 0.00 0.43 0.00 0.00 7.08 1.18 0.00 0.00 4.90 0.00 0.00 1.000 Down 1.23 Glutamyl aminopeptidase Q07075 0.00 0.00 2.55 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.000 Down N/A Na(+)/H(+) exchange regulatory cofactor NHE-RF1 O14745 0.89 0.00 0.00 0.00 5.00 0.00 1.18 0.00 1.12 0.00 0.00 0.00 1.000 Down 2.56 Lymphocyte antigen 6D Q14210 0.00 0.00 0.00 0.00 0.00 7.08 0.59 0.00 0.00 4.90 0.00 0.00 1.000 Down 2.58 Collagen alpha-1(XII) chain Q99715 0.00 0.00 1.70 0.00 0.00 0.00 0.59 0.00 0.00 0.00 0.00 0.00 1.000 Down 2.88 Extracellular superoxide dismutase [Cu-Zn] P08294 0.00 0.00 0.00 12.75 0.00 0.00 0.59 0.00 0.00 0.00 0.00 0.00 1.000 Down 21.60 Bone marrow proteoglycan P13727 0.00 0.00 0.43 4.25 0.00 0.00 1.18 0.00 0.00 0.00 2.13 0.00 1.000 Down 1.41 Arylsulfatase A P15289 0.00 0.00 1.70 0.00 0.00 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Down 1.44 Sodium/potassium-transporting ATPase subunit gamma P54710 0.00 1.93 0.00 0.00 0.00 0.00 1.77 0.00 0.00 0.00 0.00 0.00 1.000 Down 1.09 Tenascin-X P22105 0.00 0.00 0.00 0.00 0.00 2.36 1.18 0.00 1.12 0.00 0.00 0.00 1.000 Down 2.05 Folate receptor alpha P15328 0.00 0.00 0.00 0.00 2.50 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Down 2.12 Histone H2B type 1-K O60814 (+13 0.00 0.00 0.00 2.13 0.00 0.00 1.77 0.00 0.00 0.00 0.00 0.00 1.000 Down 1.20 G-protein coupled receptor family C group 5 member C Q9NQ84 0.00 0.00 0.00 0.00 1.25 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Down 1.06 Gamma-interferon-inducible lysosomal thiol reductase P13284 0.00 0.00 0.85 0.00 0.00 0.00 0.59 0.00 0.00 0.00 0.00 0.00 1.000 Down 1.44 Syntenin-1 O00560 0.00 0.00 0.00 0.00 1.25 0.00 1.18 0.00 0.00 0.00 0.00 0.00 1.000 Down 1.06 Myosin-Vb Q9ULV0 0.00 0.00 0.85 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.000 Down N/A Lysosomal acid phosphatase P11117 0.00 0.00 0.85 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.000 Down N/A Supplemental Table 3. Risk of Outcomes by uAnCR Quartile Patients who were any stage AKI at the time of sample collection (n=97) Worsening of AKI AKIN stage 3 RRT Yes No OR Yes No OR Yes No OR 1st Quartile 13% (3) 87% (21) 5.0 4% (1) 96% (21) 13.8 4% (1) 96% (23) 6.1 (1.16-21.46) (1.58-120.37) (0.65-56.37) 2nd Quartile 16% (4) 84% (21) 16% (4) 84% (21) 4% (1) 96% (24) 3rd Quartile 29% (7) 71% (17) 21% (5) 79% (19) 8% (2) 92% (22) 4th Quartile 42% (10) 58% (14) 38% (9) 62% (15) 21% (5) 79% (19) LOS RRT or Death AKIN stage 3 or Death Yes No OR Yes No OR Yes No OR 1st Quartile 25% (6) 75% (18) 21.0 4% (1) 96% (23) 9.5 4% (1) 96% (23) 13.8 (4.58-96.23) (1.06-84.37) (1.58-120.38) 2nd Quartile 56% (14) 44% (11) 12% (3) 88% (22) 16% (4) 84% (21) 3rd Quartile 62% (15) 38% (9) 8% (2) 92% (22) 21% (5) 79% (19) 4th Quartile 87% (21) 13% (3) 29% (7) 71% 17) 38% (9) 62% (15) AKIN stage 2 or 3 Yes No OR 1st Quartile 29% (7) 71% (17) 2.9 (0.87-9.45) 2nd Quartile 28% (7) 72% (18) 3rd Quartile 46% (11) 54% (13) 4th Quartile 54% (13) 46% (11) Patients who were classified as AKIN stage 1 at the time of sample collection (n=79) Worsening of AKI AKIN stage 3 RRT Yes No OR Yes No OR Yes No OR 1st Quartile 15% (3) 85% (17) 4.64 5% (1) 95% (19) 8.14 5% (1) 95% (19) 4.75 (1.02-21.0) (0.88-75.48) (0.48-46.91) 2nd Quartile 10% (2) 90% (18) 5% (1) 95% (19) 5% (1) 95% (19) 3rd Quartile 32% (6) 68% (13) 11%(2) 89% (17) 11% (2) 89% (17) 4th Quartile 45% (9) 55% (11) 30% (6) 70% (14) 20% (4) 80% (16) LOS RRT or Death AKIN stage 3 or Death Yes No OR Yes No OR Yes No OR 1st Quartile 25% (5) 75% (15) 17.0 5% (1) 95% (19) 8.14 5% (1) 95% (19) 8.0 (3.46-83.44) (0.88-75.48) (0.91-70.34) 2nd Quartile 50% (10) 50% (10) 5% (1) 95% (19) 5% (1) 95% (19) 3rd Quartile 63% (12) 37% (7) 11% (2) 89% (17) 11% (2) 89 (17) 4th Quartile 85% (17) 15% (3) 30% (6) 70% (14) 30% (8) 70% (12) The percentage and number of patients who met each outcome are shown by quartile of uAnCR. The odds ratio of 4th quartile to 1st quartile of patients and 95% CI are shown for each outcome. Supplemental Table 4. Comparison of uAnCR in patients who developed AKI after cardiac surgery compared to those who did not AKI No AKI p-value n 6 7 Male 67% (4) 71% (5) 1 Caucasian 100% (6) 100% (7) 1 a Age (yrs) 75.8 +/- 9.5 62.3 +/- 15.3 0.09 a Weight (kg) 77.7 +/- 12.7 73 +/- 35.2 0.77 a Sample Collection Time (hrs post-op) 20.3 +/- 3.4 21.3 +/- 3.8 0.6 b uAnCR 66.6 20.1 0.21 Operative Variables Bypass 100% (6) 86% (6) 1 a Bypass Time (min) 94.9 +/- 48.5 148 +/- 68.9 0.13 Serum Creatinine (mg/dL) a Pre-Op 1.33 +/- 0.4 1.2 +/- 0.3 0.6 a sCr at Collection 1.45 +/- 0.4 1.1 +/- 0.4 0.18 a Maximum Post-Op sCr 1.70 +/- 0.4 1.29 +/- 0.3 0.07 Outcomes RRT 0% (0) 0% (0) 1 Death 0% (0) 0% (0) 1 aMean +/- SD; bMedian Comprehensive Methods Urine Samples The Southern Acute Kidney Injury Network (SAKInet) was formed in 2007 to collect samples from patients who developed AKI after cardiac surgery with the goal of testing the diagnostic and prognostic accuracy of previously described AKI biomarkers and identifying novel ones. Urine samples were obtained between August 1, 2008 and October 6, 2011 from patients who had cardiac surgery at one of the SAKInet institutions (the Medical University of South Carolina, Duke University, George Washington University or University of Tennessee College of Medicine in Chattanooga). Prior to collection, informed consent was obtained in accordance with the Institutional Review Board approved protocol at each member institution. Samples were collected and stored using a rigorous standard operating procedure (SOP). Most patients were catheterized and urine was collected preferentially from the Foley tube or the urometer and processed immediately. Urine specimens were treated with a reversible, serine and cysteine protease inhibitor cocktail tablet (Roche, Complete mini , EDTA-free) at a concentration of 1 tablet per 50 ml of urine. The urine was centrifuged for 10 min at 1,000 x g and the supernatant was immediately stored at -80°C in polypropylene tubes that had been washed with 100% acetonitrile in order to minimize sample contamination with plastic polymer.

Patient Selection. The SAKInet SOP for urine collection is primarily focused on collection of urine samples from patients who have developed AKI after cardiac surgery. The goal is to collect urine samples as early as possible after AKIN serum creatinine criteria are met (increase in serum creatinine ≥0.3 mg/dL or ≥50% from baseline). Collections are made in the surgical ICUs. Inclusion criteria are consent by the patient or appropriate surrogate, surgery of the heart or ascending aorta and development of AKI (defined by the AKIN serum creatinine criteria) within 2 days of surgery. The only exclusion criterion is a baseline serum creatinine greater than 3 mg/dL. Only subjects who had collection of urine within 48 hours were used in order to conform to the AKIN staging criteria and to attempt to eliminate confounding effects of events that were not directly related to the cardiac surgery. Twelve samples were used in the proteomic studies, 10 of which were also used in a validation set that included samples collected from the remaining 87 subjects in the study. Of the samples used in the validation set, 79 were from patients classified as AKIN stage 1 at the time of sample collection. Urine samples were stored at -80°C and shipped to MUSC on dry ice. Samples used in this study were collected within the first 48 post-operative hours and were selected from among the stored samples to fit the criteria described in the results section.

The discovery phase analysis compared the urine proteome of 12 patients who had AKIN stage 1 AKI at the time of urine collection. Six patients required renal replacement therapy within 10 days and six did not. Patients for the discovery phase analysis were selected from among all patients in the cohort who had AKI within 48 hours of surgery. Subjects were matched first on bypass status, then on bypass time and preoperative serum creatinine. Finally, we attempted to match patients based on the type of surgery. Both groups had two patients who had off pump surgery and four that used cardiopulmonary bypass. Mean bypass time was 165 minutes in both groups and mean baseline serum creatinine was 1.3 mg/dl in both groups. The group that did not require RRT consisted of three patients with CABG alone, one with valve replacement and two with CABG/valve. The group that required RRT consisted of two patients with CABG alone, one septal myectomy, one valve replacement, one CABG/valve replacements and one CABG with repair of ascending aorta. Supplemental table 1 shows the characteristics of the subjects in the discovery analysis.

Patients in the follow-up study to confirm the ability of uAnCR to predict worsening of AKI were selected from all subjects in the cohort for whom data was available at the time the assay was run. Exclusion criteria were subjects who did not have AKI, samples that were collected greater than 48 hours after conclusion of the cardiac surgery, subjects with a baseline serum creatinine of 3.0 or greater, surgeries other than CABG and valve replacement. We excluded patients with samples collected more than 48 hours after surgery in order to conform to the AKIN criteria and to limit confounding effects of events that were not related to the surgery. All subjects with samples and data available at the time of analysis who met the inclusion and exclusion criteria were included in the analysis.

Proteomic Analyses Trypsin Digestion. Urine (supernatant from the 1,000 x g centrifugation) was thawed in a 37°C water bath and digested in-solution with trypsin using the following protocol. One hundred μL of each sample was diluted with 100 μL of 0.2% Rapigest SF surfactant (Waters) in 100mM ammonium bicarbonate. To account for technical variability in the digestion and liquid chromatography-tandem mass spectrometry (LC-MS/MS) protocols, 200 ng of the internal standard recombinant HIV protein gp160 (Bioclone, Inc.) was spiked into each sample. Proteins were denatured by the addition of 5mM dithiothreitol and heated to 60°C for 30 min. After cooling to room temperature, proteins were alkylated by the addition 12mM iodoacetamide and incubation at room temperature in the dark for 30 min. Proteins were digested with 10 µg of trypsin (Applied Biosystems,

TPCK treated with CaCl2) overnight at 37°C.

LC-MS/MS. Each digested sample was pre-fractionated using offline reversed phase solid phase extraction (SPE). The Strata-X SPE cartridge (Phenomenex; 30 mg/mL) was activated and equilibrated by application of 1 mL methanol followed by 1 mL of 0.1% formic acid in water. The sample was loaded onto the SPE column, and a series of elutions containing progressively higher concentrations of acetonitrile (10%, 15%, 20%, 25%, 30%, 35%, 40%, 50%, and 60%) in 0.1% formic acid were performed to separate the sample into fractions of increasing hydrophobicity. The 10% and 15% eluates were combined, as were the 50% and 60% eluates. Sample fractions were completely dried in a centrifugal vacuum concentrator, and each fraction was reconstituted in 50 μL of mobile phase A (98% H20, 0.1% formic acid; 2% acetonitrile). Fractions from each elution were analyzed by LC-MS/MS. Five μL of each fraction was injected onto an Acclaim PepMap100 trap column (100 μm ID x 2 cm, C18, 5 μm, 100 Å; Thermo Scientific) and washed with 100% mobile phase A for 10 minutes at 2 μL per minute. The fraction was then separated on an Acclaim PepMap100 analytical column (75 μm ID x 15 cm, C18, 3 μm, 100 Å; Thermo Scientific). The combined 50% and 60% elution fractions were separated using a 40 minute 2-step continuous gradient of increasing mobile phase B (MPB; 98% acetonitrile, 2% H20, 0.1% formic acid). The first step increased from 10% MPB to 40% MPB at 1.5% per minute. The second step increased from 40% MPB to 60% MPB at 1% per minute. All other elution fractions were separated using a 45 minute 2-step gradient. The first step increased from 10% to 40% MPB at 1% per minute, and the second step increased from 40% to 60% at 2% per minute. Tandem mass spectrometry was performed using an AB SCIEX Triple TOF 5600 mass spectrometer. This instrument was run in information dependent acquisition mode with the following parameters: 250 ms MS accumulation time, 50 ms MS/MS accumulation time, 20 ions selected per cycle, total cycle time of 1.3 s, 4 s dynamic exclusion time after one occurrence, and rolling collision energy. The scanning windows for the TOF-MS and MS/MS were 300 to 1250 and 55 to 2000 m/z, respectively.

Protein Identification and Quantification. Acquired spectra (.wiff files) were converted to the MGF format using an AB SCIEX converter (version 1.1 beta). MGF files from all the fractions of each sample were merged and searched against the 2011_6 release of the Human UniProtKB/Swiss-Prot database (20,127 entries) with addition of the common contaminants (112 entries) using the Mascot search engine with trypsin as the specified enzyme. Carbamidomethyl (C) was selected as a fixed modification, and oxidation (M) and deamidation (NQ) were selected as variable modifications. Monoisotopic masses were used, and the error tolerances were 10 ppm and 0.5 Da for peptides and MS/MS fragments, respectively. Mascot search results were loaded into Scaffold (Proteome Software, Inc), which used the Peptide Prophet and Protein Prophet algorithms to validate peptide and protein identifications. The Scaffold quantitative values of identified proteins were normalized to the internal standard recombinant HIV protein present in each biological sample, and the relative abundance of each protein is reported in normalized spectral counts.

Angiotensinogen ELISA. The Human Total Angiotensinogen Assay Kit (Immuno- Biological Laboratories Co., Ltd.), a solid phase sandwich ELISA, was used according the manufacturer’s protocol to measure urinary angiotensinogen. Urine samples were diluted 1:8 in EIA buffer. One hundred μL of diluted sample was added to the appropriate well and incubated for 60 min at 37°C. The plate was then washed 7 times by pipetting 250 µL of the provided wash buffer into each well using a multichannel, repeating pipet. After drying the plate, 100 µL of 30x diluted HRP-conjugated anti- angiotensinogen antibody was added to each well and incubated for 30 minutes at 37°C. The plate was washed 9 times as before and dried. 100 μL of chromogen (TMB) was added to each well, and the plate was incubated for 30 min in the dark at room temperature. One hundred μL of stop solution was added to each well, and the absorbance was measured at 450 nm using a SpectraMAX 340PC 96-well plate reader. The linear range of the assay is 20 to 0.31 ng/mL. Intra- and inter-assay variability (coefficient of variation) were calculated by measuring the standards and three selected biological samples in quadruplicate once, and in duplicate on all remaining plates. Values for intra- and interassay variability were 2.4% and 9.9%, respectively. Data were analyzed using Softmax Pro3.1.2. Samples whose values were above the upper limit of quantification for the assay were diluted 1:10 in EIA buffer and re-run on a separate plate.

Urine Creatinine Determination. Urine creatinine was used to correct the urine angiotensinogen concentration and values were reported as the ratio of angiotensinogen in ng/ml to creatinine in mg/ml (uAnCR, ng/mg). Urine creatinine was measured using the Jaffe assay. Three μL of sample was combined with 100 μL of 1% picric acid (Sigma-Aldrich), 100 μL of 0.75M NaOH (Genomic Solutions), and 300 μL distilled deionized H2O. Samples were incubated at room temperature for 15 min and absorbance at 490 nm was measured using a SpectraMAX 340PC 96-well plate reader. Data were analyzed using Softmax Pro 3.1.2.

Outcomes. The primary outcome was worsening of AKI, defined as progression to a higher AKIN stage after the time of sample collection. Secondary outcomes were progression to AKIN stage 3, the need for renal replacement therapy (RRT) within 10 days of sample collection, progression to AKIN stage 2 or 3, progression to AKIN stage 3 or death, RRT or death, and discharge >7 days from the time of sample collection or in-hospital mortality. Outcomes were tested using the entire cohort and in the subset of patients classified as AKIN stage 1 at the time of sample collection.

Statistical Analysis. Differential abundance of proteins identified and quantified by LC- MS/MS were selected using the Wilcoxon Rank-Sum test with a significance threshold of p≤ 0.05. This test was used because it has been previously shown to be a robust test for the identification of candidate biomarkers in proteomics studies with small sample sizes. Candidate biomarkers were selected based upon the combination of p-value from the Wilcoxon Rank-Sum test and mean fold-change between the experimental groups. The relationship between these two measures was visualized by “volcano plot”. In verification studies, count data were analyzed using the χ2 or Fisher’s exact test as appropriate. Continuous variables were analyzed using the t test or Mann Whitney U test when comparing two groups. ANOVA or Kruskal-Wallis ANOVA on Ranks test and the post- hoc Dunn’s test for pairwise comparison were used to evaluate continuous variables when more than two groups were compared. Odds ratios (OR) were used to test the association of uAnCR with selected outcomes. Patients were stratified by uAnCR into quartiles, the effect of uAnCR on the risk of developing an outcome was tested by calculating the OR of the upper and lower quartiles and estimating the 95% confidence interval of the OR. Receiver operator characteristic curves were constructed to determine the predictive power of uAnCR. The area under the ROC curve (AUC) was used as an estimate of an overall accuracy of the biomarker. An AUC of 1.0 represents 100% accuracy, whereas an AUC of 0.5 indicates 50% accuracy, which is no better than random chance. Univariate ROC curves were considered statistically significant if the AUC differed from 0.5, as determined by the z-test. Optimal cut-offs were determined by selecting the data point that minimized the geometric distance from 100% sensitivity and 100% specificity on the ROC curve.1 To visualize the relationship between uAnCR and length of stay, patients were stratified into tertiles by uAnCR. Additionally, cut-offs that maximized the positive likelihood ratio and minimized the negative likelihood ratio were reported since they could be useful in assigning high or low risk to a patient. Likelihood ratios of positive and negative predictive value were used since they are insensitive to changes in prevalence (unlike PPV and NPV) and can be used to infer post-test probability. Kaplan-Meier curves were used to visualize the relationship between uAnCR and length of stay. Patients who died were censored. The log-rank test was used to compare the curves, and the Holm-Sidak test was used for post-hoc pairwise comparison. Category free net reclassification improvement was used to determine if addition of uAnCR to a multivariate logistic regression model for prediction of risk increased the ability of the model to predict worsening of AKI. First, multiple logistic regression model (reference model) was created using the variables percent change in serum creatinine from baseline and Cleveland Score, a perioperative risk score that has been demonstrated to predict AKI outcomes after cardiac surgery.2, 3Then, a new model was created with included uAnCR, in addition to these two variables. Each patient’s probability (risk) of experiencing worsening of AKI after sample collection was calculated with each model. The category free net reclassification index was calculated as previously described, and was used to quantify the improved prognostic predictive power gained by including uAnCR in the model. 4, 5 Statistical tests were performed in either Matlab or SigmaPlot.

References for On-line methods.

1. Pepe M. The statistical evaluation of medical tests for classification and prediction. Oxford University Press; 2004. .

2. Thakar CV, Arrigain S, Worley S, Yared JP, Paganini EP. A clinical score to predict acute renal failure after cardiac surgery. J Am Soc Nephrol 2005 Jan;16(1):162-8.

3. Englberger L, Suri RM, Li Z, Dearani JA, Park SJ, Sundt TM,3rd, Schaff HV. Validation of clinical scores predicting severe acute kidney injury after cardiac surgery. Am J Kidney Dis 2010 Oct;56(4):623-31.

4. Pencina MJ, D'Agostino RB S, D'Agostino RB,Jr, Vasan RS. Evaluating the added predictive ability of a new marker: From area under the ROC curve to reclassification and beyond. Stat Med 2008 Jan 30;27(2):157,72; discussion 207-12.

5. Pickering JW, Endre ZH. New metrics for assessing diagnostic potential of candidate biomarkers. Clin J Am Soc Nephrol 2012 Jun 7.