Comprehensive Gene Expression Profiling in Diabetic Nephropathy: Inflammation as a Prominent Event of Progressive Disease.

Objective: Diabetic nephropathy (DN) has developed into the leading cause of terminal renal failure. To gain insight into molecular programs activated in DN genome wide gene expression profiling of microdissected renal biopsies was performed in a disease stage specific manner. Methods: Human renal biopsies were collected in the ERCB multicenter study, microdissected into tubulo-interstitial and glomerular segments and RNA hybridized to Affimetrix oligonucleotide array HGU133A, containing 22.283 probe sets. Biopsies with well-defined stages of early and progressive diabetic nephropathy (n = 13) were compared with pre-transplant biopsies (n = 7) minimal change disease (MCD, n = 4) and thin basement membrane disease (TBM, n = 6). Expression data were processed with RMA, SAM and dChip algorithm. Significantly differentially regulated mRNAs (p < 0.01) were functionally categorized using the Gene Ontology database. Specific inflammatory pathways were further analyzed in an extend cohort of biopsies by real-time RT-PCR. Results: Gene expression profiles allowed effective segregation of the different patient populations according to disease state and stage. Compared to living related donor kidney biopsies, early diabetic nephropathy showed significant differential regulation of 311 (1.4%), progressive diabetic nephropathy of 1,735 (7.8%) and minimal change disease of 240 (1.1%) genes. Functional categorization of the differentially regulated mRNAs allowed to match the observed expression changes to specific pathways, showing again a striking difference between progressive DN to controls, early DN, and MCD. This difference was most pronounced in the gene ontology category of 'inflammation -- stress response', with 347 regulated transcripts in progressive DN compared to 25 in early DN and 26 in MCD. Chemokines were selected for further analysis, as they showed a prominent induction in the array based assay in DN and are functionally considered to be key elements in leukocyte infiltration. Using an extended sample set of progressive tubulointerstitial damage in DN, MCD and other glomerulopathies the induction of CC/CXC chemokines and their receptors could be confirmed by qt RT-PCR with maximal mRNA levels in progressive DN compared to all other diseases analyzed. Conclusions: Unbiased gene expression profiling identified a pronounced inflammatory response in the tubulo-interstitium in progressive DN. Focused analysis of CC chemokines showed maximal levels in tubulo-interstitial damage of progressive DN compared to all other categories analyzed. In conclusion, inflammatory pathways may play an underestimated role for the progression of diabetic nephropathy and warrant further functional analysis. [ABSTRACT FROM AUTHOR]