Identification of nephropathy candidate genes by comparing sclerosis-prone and sclerosis-resistant mouse strain kidney transcriptomes

Background The genetic architecture responsible for chronic kidney disease (CKD) remains incompletely described. The Oligosyndactyly (Os) mouse models focal and segmental glomerulosclerosis (FSGS), which is associated with reduced nephron number caused by the Os mutation. The Os mutation leads to FSGS in multiple strains including the ROP-Os/+. However, on the C57Bl/6J background the mutation does not cause FSGS, although nephron number in these mice are equivalent to those in ROP-Os/+ mice. We exploited this phenotypic variation to identify genes that potentially contribute to glomerulosclerosis. Methods To identify such novel genes, which regulate susceptibility or resistance to renal disease progression, we generated and compared the renal transcriptomes using serial analysis of gene expression (SAGE) from the sclerosis-prone ROP-Os/+ and sclerosis resistant C57-Os/+ mouse kidneys. We confirmed the validity of the differential gene expression using multiple approaches. We also used an Ingenuity Pathway Analysis engine to assemble differentially regulated molecular networks. Cell culture techniques were employed to confirm functional relevance of selected genes. Results A comparative analysis of the kidney transcriptomes revealed multiple genes, with expression levels that were statistically different. These novel, candidate, renal disease susceptibility/resistance genes included neuropilin2 (Nrp2), glutathione-S-transferase theta (Gstt1) and itchy (Itch). Of 34 genes with the most robust statistical difference in expression levels between ROP-Os/+ and C57-Os/+ mice, 13 and 3 transcripts localized to glomerular and tubulointerstitial compartments, respectively, from micro-dissected human FSGS biopsies. Network analysis of all significantly differentially expressed genes identified 13 connectivity networks. The most highly scored network highlighted the roles for oxidative stress and mitochondrial dysfunction pathways. Functional analyses of these networks provided evidence for activation of transforming growth factor beta (TGFβ) signaling in ROP-Os/+ kidneys despite similar expression of the TGFβ ligand between the tested strains. Conclusions These data demonstrate the complex dysregulation of normal cellular functions in this animal model of FSGS and suggest that therapies directed at multiple levels will be needed to effectively treat human kidney diseases.

In order to verify that the sclerotic phenotype previously described in the ROP-Os mice is still expressed in our animal colony, we phenotyped the mice using biochemical and histological measures. Urinary albumin measurement. Urinary albumin concentration was determined using a mouse albumin ELISA kit (Bethyl Laboratories, Montgomery, TX) according to the manufacturer's protocol. In brief, microtiter ELISA plates were coated with an affinity-purified goat anti-mouse albumin antibody, then washed and blocked. Subsequently, protein standards and unknown test samples were added to the plates and incubated for 1 hour. The plates were vigorously washed multiple times and then incubated with a different goat anti-mouse albumin antibody conjugated to HRP.
After the incubation step, the colorimetric reaction was initiated using the TMB color substrate (Kirkegaard & Perry Labs, Gaithersburg, MD 20879). The color reaction was stopped by adding 2 M sulfuric acid and the color depth determination was measured at a wavelength of 450 nm.
The standard curve was plotted using a semi-logarithmic scale and sample concentration was determined by comparison to the standard curve. Samples with concentration outside the range of the standard curve were diluted and re-assayed.
Urinary and serum creatinine measurements: Urine and serum creatinine were assayed by the enzymatic creatininase method using CREA-Plus reagent according to the manufacturer's protocol (Roche diagnostics; Indianapolis, IN). This assay has been shown to be more accurate in mice and correlates with results obtained by HPLC [1,2]. The ROP-Os/+ mice had higher albumin to creatinine ratio as early as 6 weeks.
Sclerosis score: Mesangial sclerosis was scored by three blinded observers and averaged using a previously described method. [3,4]. ROP-Os mice displayed a higher sclerosis score as previously described.
Confirmation of SAGE tag differential expression. To validate differential gene expression, we confirmed the relative mRNA concentrations of a selected number of genes using kidney RNA derived from different animals (average, n=3). We used multiple methods for confirmation based on abundance of the expression of the target cDNA, the presence of different isofroms or similar cDNAs, and the possibility of designing accurate real-time PCR primers.
RNase protection assays were carried out using the Direct Protect and Max-Script kit (Ambion, Austin, TX). Briefly, cDNA fragments were PCR-generated using intron-spanning, exonspecific primers. The PCR products were cloned into pCR-II-topo (Invitrogen, Carlsbad, CA).
The recombinant plasmids were sequenced to ensure identity. [ 32 P] labeled cRNA probes were generated using the Max-Script kit and the probes were gel-purified according to manufacturer recommendation. The labeled probes were then annealed to the mRNA, followed by RNAse digestion. Protected probe fragments were resolved by electrophoresis and subjected to autoradiography using a phosphorimager, followed by quantification of the protected fragment intensity using ImageQuant software (Molecular Dynamics, Sunnyvale, CA). The signal from the protected probe is directly proportional to the abundance of the corresponding mRNA in the assayed pool. A Gapdh probe was used as a control to ensure equal input of mRNA across samples and normalize expression level.
Real-time quantitative RT-PCR was done using a Roche LightCycler PCR machine in combination with SYBR green reagent supplied by Roche (Roche, Indianapolis, IN) according to the manufacturer's protocol. The data obtained were normalized to either the Gapdh housekeeping gene or 18S rRNA. Northern Blot analysis: Northern blots were performed as previously described [5]. Briefly, [ 32 P] labeled cDNA probes generated by PCR were used to probe Northern blots using RNA obtained from both ROP-Os/+ and C57-Os/+ kidney cortices.
Equal loading of RNA was ascertained by hybridization with a Gapdh cDNA probe. Signal intensity was quantified using densitometry in combination with ImageQuant software (Molecular Dynamics).
We selected the genes for confirmatory experiments based on the possible pathophysiologic and/or statistical significance. Among the genes we chose to confirm were Gpx3 (glutathione peroxidase 3), Nrp2 (neuropilin), and Itch (itchy E3 ubiquitin protein ligase homolog). All confirmatory approaches revealed relative expression levels in agreement with SAGE analysis.
Primers used in Real-time quantitative RT-PCR are depicted in table 1s.