A common rejection module (CRM) for acute rejection across multiple organs identifies novel therapeutics for organ transplantation

Abstract

Using meta-analysis of eight independent transplant datasets (236 graft biopsy samples) from four organs, we identified a common rejection module (CRM) consisting of 11 genes that were significantly overexpressed in acute rejection (AR) across all transplanted organs. The CRM genes could diagnose AR with high specificity and sensitivity in three additional independent cohorts (794 samples). In another two independent cohorts (151 renal transplant biopsies), the CRM genes correlated with the extent of graft injury and predicted future injury to a graft using protocol biopsies. Inferred drug mechanisms from the literature suggested that two FDA-approved drugs (atorvastatin and dasatinib), approved for nontransplant indications, could regulate specific CRM genes and reduce the number of graft-infiltrating cells during AR. We treated mice with HLA-mismatched mouse cardiac transplant with atorvastatin and dasatinib and showed reduction of the CRM genes, significant reduction of graft-infiltrating cells, and extended graft survival. We further validated the beneficial effect of atorvastatin on graft survival by retrospective analysis of electronic medical records of a single-center cohort of 2,515 renal transplant patients followed for up to 22 yr. In conclusion, we identified a CRM in transplantation that provides new opportunities for diagnosis, drug repositioning, and rational drug design.

Figure 1.

IPA regulatory network using 96 of the 102 genes. 102 significantly overexpressed genes by both meta-analysis methods were used as input to IPA to create a gene–gene interaction network. We chose the direct relationship option in the IPA to create the interaction networks. Nodes highlighted in blue represent the 12 genes identified as the CRM, and nodes in solid salmon represent rejection by leave-one-organ-out meta-analysis.

Figure 2.

Discovery of a CRM consisting of 12 genes by leave-one-organ-out analysis. (A–L) 12 genes were significantly overexpressed during AR in all transplanted organs analyzed in this study, although they may not be significantly overexpressed in individual datasets. The x axes represent standardized mean difference between AR and STA, computed as Hedges’ g, in log₂ scale. The size of the blue rectangles is proportional to the SEM difference in the study. Whiskers represent the 95% confidence interval. The yellow diamonds represent overall, combined mean difference for a given gene. Width of the yellow diamonds represents the 95% confidence interval of overall mean difference. (M) Network analysis using MetaCore showed that 10 out of the 12 genes are part of a single regulatory network, with NF-κB and STAT1 forming the central axis of regulation.

Figure 3.

Validation of the CRM genes in three independent cohorts consisting of 794 renal allograft biopsies. Data are presented as in Fig. 2 (A–L) in three independent cohorts consisting of 794 renal transplant biopsies.

Figure 4.

Intragraft expression of the CRM genes can distinguish AR and STA samples with high specificity and sensitivity. CRM score was defined as geometric mean of the CRM genes expression. (A–F) Distribution of CRM scores in AR and STA groups and receiver operating characteristic (ROC) curve for GSE21374 (A and B), GSE36059 (C and D), and the Stanford cohort (E and F). The x axes represent false positive rate, and the y axes represents true positive rate when using the CRM scores for predicting AR. Error bars indicate SEM.

Figure 5.

CRM score correlates significantly with extent of graft injury in two independent renal transplant cohorts consisting of 151 graft biopsies. (A) Comparison of CRM scores (y axis) and serum creatinine (x axis) from healthy donors, STA renal transplant patients with and without renal dysfunction, and patients with AR. (B) Comparison of CRM scores in renal biopsies from nonprogressors, progressors, and AR transplant patients in GSE25902. (C and D) Comparison of CRM scores (y axis) with Banff tubulitis (t-score) and interstitial inflammation (i-score) scores in renal transplant patients, respectively. (E and F) Comparison of CRM scores with ct-score and ci-score in STA renal transplant patients who did not have any AR episode in the first 2 yr after transplantation. (G) Comparison of change in CRM scores over time in STA renal transplant patients. The patients were divided into two groups: progressors (n = 12, CADI ≥ 6) and nonprogressors (n = 12, CADI < 6) 2 yr after transplantation. Protocol biopsies were obtained from each patient at three time points: (1) at the time of transplant, (2) 6 mo after transplant, and (3) 24 mo after transplant. Repeated measures analysis of variance was used for analysis of the CRM scores between progressor and nonprogressors over three time points. (H) CRM scores of 6-mo protocol biopsy were used to predict patients with severe histological damage after 2 yr after transplantation. JT trend test was used to compute p-values for correlation between the CRM scores and Banff t- and i-scores as well as ct- and ci-scores. Error bars indicate SEM.

Figure 6.

Atorvastatin and dasatinib treatment reduced infiltrating cells in completely mismatched mouse cardiac allografts. C57BL/6 mice were transplanted with hearts from FVB mice, and mice were treated with no drug, cyclosporine (20 mg/kg/day), atorvastatin (75 mg/kg/day), or dasatinib (25 mg/kg/day). Each treatment group used six pairs of mice. In total, we used 53 mice (24 pairs of FVB-to-C57BL/6 cardiac transplant and 5 mice without cardiac transplant). (A–E) Immunohistochemistry at postoperative day 7 showed that the number of infiltrating cells in the cyclosporine, atorvastatin, and dasatinib treatment groups was significantly reduced compared with untreated AR. The pictures were taken at 40× using a Nikon E600 on postoperative day 7. (F–M) Number of infiltrating cells in cardiac allografts (×10⁶) in each group. *, statistically significant (P < 0.05) reduction in the number of infiltrating cells compared with untreated AR group; +, statistically significant (P < 0.05) reduction in the number of infiltrating cells compared with the cyclosporine group. Error bars indicate SEM.

Figure 7.

Atorvastatin and dasatinib treatment significantly extended allograft survival in mice and humans. (A) C57BL/6 mice were transplanted with hearts from FVB mice, and mice were treated with no drug, cyclosporine (20 mg/kg/day), atorvastatin (75 mg/kg/day), or dasatinib (25 mg/kg/day). Each treatment group used six pairs of mice. In total, we used 48 mice (24 pairs of FVB-to-C57BL/6 cardiac transplant). Mice were treated for up to 30 d, and graft survival was monitored. (B) Electronic medical records of 2,515 renal transplant patients, transplanted between January 1989 and March 2012 at the University Hospitals Leuven, were divided into two groups: (1) 1,566 patients who received statin within the first 180 d after transplantation, with grafts surviving at least 180 d, and (2) 949 patients who did not receive statin. None of the patients started or stopped statin because of renal function evolution or intragraft phenomena. Patients were censored when a patient stopped taking statin, graft failed, or recipient death. Cox proportional hazard analysis was used to associate statin use with graft survival while adjusting for donor and recipient age, repeat transplantation, and calendar year.