Identification of novel targets of diabetic nephropathy and PEDF peptide treatment using RNA-seq.

Background: Diabetic nephropathy (DN) is a major complication of type1 and type 2 diabetes. Understanding how diabetes regulate transcriptome dynamics in DN is important for understanding the biology of the disease and for guiding development of new treatments.

Results: We analyzed the kidney transcriptome of a DN mouse model, D2.B6-Ins2 /MatbJ, before/after treatment with P78-PEDF. Age, weight, and gender-matched mice and wild-type (wt) littermates were treated at 6 weeks (early treatment) or 12 weeks (late treatment) of age for the duration of 6 weeks. Animals were implanted with an osmotic mini pump delivering 0.3 ug/g/day P78-PEDF or vehicle. Using RNA-seq, we identified14,316 transcripts (12,328 coding;1,988 non-coding) that were significant and reliably expressed (FPKM > =1) in diabetic kidneys. Expression of 1,129 (7.9%) including 901 coding genes was altered by diabetes with log2 fold changes (FC) between -86.2 and +86.0 (q < 0.05) compared to wt. Of these, 164 (14.5%) showed increased and 965 (85.5%) decreased expression with FC > 1.5. Coding genes with highest FC in diabetic kidneys include Nhej1 (32.04), Ept1 (8.6), Srd5a2 (-6.55), Aif1 (-6.05), and Angptl7 (-4.71). Early and late stage diabetic groups receiving continuous infusion of P78 showed altered expression of 316/14,316 (2.2%) transcripts, including 121 coding genes compared to non-treated diabetic controls. Of these, 183 were upregulated and 133 downregulated with FC +50.9--93.3 (q < 0.05). P78 reversed diabetes-induced changes in 138/1129 (12.2%) transcripts, including 49/901 (5.44%) coding genes. Nhej1 (-37.94), Tceanc2 (5.76), Ept1 (-4.45), Ugt1a2 (3.03), and Tmsb15l (-3.0) showed the highest FC with treatment. The DNA repair gene, Nhej1 with the greatest FC in diabetic kidneys was completely restored to control levels by both early and late P78 treatments. Expression of other coding genes regulated by diabetes with FC > =(+/-) 1.5 and completely reversed by P78 include Mamdc4, Kdm4b, Tmem252, Selm, and Hpd. RT and QRT-PCR validated expression of gene with FC > (+/-)2.0. Transcriptome changes were also observed between early and late-stage treatments. Precursor non-coding miRNAs showed the highest fold changes in expression in the diabetic and P78 treatment groups. Several diabetic-induced changes were reversed in direction of expression by treatment including Gm24083, GM25953, miR1905, Gm25535, Gm27903, and miR196a1 with FC > =(+/-)20. From Ingenuity pathway analysis (IPA), mitochondrial dysfunction, Nrf-2- mediated oxidative stress and renal injury pathways emerged as key mechanisms in DN. DN-enriching genes in these pathways were reduced in number or regulated in the opposite direction by treatment.

Conclusions: Unique biomarkers and canonical pathways identified in this study may hold the key to understanding mechanisms of DN pathobiology with value for clinical translation. Our data suggest that mitochondrial dysfunction, genotoxicity and oxidative stress are principal events in DN and that P78-PEDF holds promise for its management.