Eir nonprogressor counterparts (Figure 5). Striking differences were observed in the canonical pathways get PNPP over-represented by the differentially expressed genes in progressor and nonprogressor patients at the second time point biopsy post-KT. Biopsies from patients classified as progressors showed activation of immune pathways such as CD28 and iCOS-iCOSL signaling in T helper cells, B cell and T cell receptor signaling, and CTLA4 signaling in cytotoxic T lymphocytes (Supplemental Figure 2). Interestingly, the analysis of genes involved in kidney injury / damage showed overrepresentation of genes associated with damage of kidney (p = 6.6E-03) with increased predicted activation (z-score=2.63) (17 genes including ALOX5, C1QA, CCR2, among others); reperfusion injury of kidney (p= 1.6E-02), increased predicted activation (z-score= 2.38) (12 genes including THBS1, TLR2 CP, FGB, LCN2, among others) and Olumacostat glasaretil site proximal tubular toxicity (p= 2.3E-02), increased predicted activation (z-score=2.46) (12 up-regulated genes) in the progressor group at 12 months post-KT. Also, analysis of renal apoptosis/cell death showed a higher number of genes inducing kidney graft injury in the same group of patients (Supplemental Figure 3).Author Manuscript Author Manuscript Author Manuscript Author ManuscriptDiscussionEarly identification of patients at risk for developing renal fibrosis as a result of CNIT could lead to timely implementation of effective strategies to prevent long-term deterioration of renal function. Identification of novel markers of CNIT depends on the discovery of underlying biological mechanisms (12-14, 31-34). A better understanding of the biological response of stressed kidney tissue could facilitate discovery of biomarkers of injury before this response leads to irreversible damage. The present study aimed to identify molecular pathways associated with CNIT, common and unique genes associated with graft injury among CNIT, AR, and IF/TA, and to explore the role of the identified CNIT molecular signatures in progression to CAD. Molecular signatures that associated only with CNIT after filtering first by NA and second by AR and IF/TA were established. For sample selection, a histological diagnosis of `pure CNIT’ was used. Even when histological allograft evaluation is the accepted gold standard for CNIT diagnosis, the limitations and variability likely associated with this approach are well recognized (12, 33, 34). Consequently, this issue might also affect the study sample selection. The histological lesions classically associated with CNIT lack specificity, and thus diagnosis is based on the finding of characteristic histological criteria in the appropriate clinical setting. However, the present approach represents an appropriate start-point for the proposed analyses. In addition to the described criteria for biopsy selection, additional stepsAm J Transplant. Author manuscript; available in PMC 2015 May 01.Maluf et al.Pageto minimize the risk of unspecific diagnosis by avoiding confounding variables were included. Molecular signatures were obtained for each evaluated condition, identifying unique and overlapped genes; and specific and common pathways of tissue allograft injury. For each comparison, background in gene expression that might associate with the use of CNI was avoided by using the same set of NA biopsies (undergoing same immunosuppressive treatment). The `clean-up’ strategy used in our approach is supported by results from several st.Eir nonprogressor counterparts (Figure 5). Striking differences were observed in the canonical pathways over-represented by the differentially expressed genes in progressor and nonprogressor patients at the second time point biopsy post-KT. Biopsies from patients classified as progressors showed activation of immune pathways such as CD28 and iCOS-iCOSL signaling in T helper cells, B cell and T cell receptor signaling, and CTLA4 signaling in cytotoxic T lymphocytes (Supplemental Figure 2). Interestingly, the analysis of genes involved in kidney injury / damage showed overrepresentation of genes associated with damage of kidney (p = 6.6E-03) with increased predicted activation (z-score=2.63) (17 genes including ALOX5, C1QA, CCR2, among others); reperfusion injury of kidney (p= 1.6E-02), increased predicted activation (z-score= 2.38) (12 genes including THBS1, TLR2 CP, FGB, LCN2, among others) and proximal tubular toxicity (p= 2.3E-02), increased predicted activation (z-score=2.46) (12 up-regulated genes) in the progressor group at 12 months post-KT. Also, analysis of renal apoptosis/cell death showed a higher number of genes inducing kidney graft injury in the same group of patients (Supplemental Figure 3).Author Manuscript Author Manuscript Author Manuscript Author ManuscriptDiscussionEarly identification of patients at risk for developing renal fibrosis as a result of CNIT could lead to timely implementation of effective strategies to prevent long-term deterioration of renal function. Identification of novel markers of CNIT depends on the discovery of underlying biological mechanisms (12-14, 31-34). A better understanding of the biological response of stressed kidney tissue could facilitate discovery of biomarkers of injury before this response leads to irreversible damage. The present study aimed to identify molecular pathways associated with CNIT, common and unique genes associated with graft injury among CNIT, AR, and IF/TA, and to explore the role of the identified CNIT molecular signatures in progression to CAD. Molecular signatures that associated only with CNIT after filtering first by NA and second by AR and IF/TA were established. For sample selection, a histological diagnosis of `pure CNIT’ was used. Even when histological allograft evaluation is the accepted gold standard for CNIT diagnosis, the limitations and variability likely associated with this approach are well recognized (12, 33, 34). Consequently, this issue might also affect the study sample selection. The histological lesions classically associated with CNIT lack specificity, and thus diagnosis is based on the finding of characteristic histological criteria in the appropriate clinical setting. However, the present approach represents an appropriate start-point for the proposed analyses. In addition to the described criteria for biopsy selection, additional stepsAm J Transplant. Author manuscript; available in PMC 2015 May 01.Maluf et al.Pageto minimize the risk of unspecific diagnosis by avoiding confounding variables were included. Molecular signatures were obtained for each evaluated condition, identifying unique and overlapped genes; and specific and common pathways of tissue allograft injury. For each comparison, background in gene expression that might associate with the use of CNI was avoided by using the same set of NA biopsies (undergoing same immunosuppressive treatment). The `clean-up’ strategy used in our approach is supported by results from several st.