Identification and dissection of the Nrf2 mediated oxidative stress pathway in human renal proximal tubule toxicity.

The identification and dissection of cellular stress mechanisms is fundamental to understanding the susceptibility of the kidney to chemicals and pharmaceuticals and for the development of renal biomarkers indicative of sub lethal injury. Here, we utilised whole genome DNA microarrays in an attempt to uncover molecular mechanisms of response to nephrotoxin exposure. Human renal proximal tubular cells (HK-2) were treated for 12h and 48 h with 5 µM Cadmium (Cd), 30 µM Diquat Dibromide (Diq), and 5 µM Cyclosporine A (CsA). Nephrotoxin treatment resulted in an alteration of a total of 4608 transcripts. Ingenuity Pathways Analysis™ revealed the anti-oxidant and detoxification Nrf2 pathway as the most significantly enriched signaling pathway in the selected dataset. Activation of this transcription factor was confirmed as nuclear translocation and paralleled the temporal alterations of compound induced H(2)O(2) production. Transcriptomics, western blot and immunofluorescence showed an induction of both HO-1 and NQO1 with Cd and Diq exposure, but not with CsA treatment. Knockdown of Nrf2 by siRNA, reduced compound induced NQO1 mRNA to basal levels and attenuated toxin induced HO-1 mRNA expression. siRNA knock down of HO-1, but not NQO1, enhanced Cd induced H(2)O(2) production and Cd induced toxicity. Using an un-biased transcriptomic approach we have identified the Nrf2 pathway as the most significant signaling response in renal epithelial cells challenged with nephrotoxin. This study highlights the importance of this pathway and particularly HO-1 in renal epithelial adaptation to oxidative stress.

Inter-laboratory comparison of human renal proximal tubule (HK-2) transcriptome alterations due to Cyclosporine A exposure and medium exhaustion.

There is an acknowledged need to promote and further develop in vitro techniques in order to achieve the goal of improved risk assessment of chemicals and pharmaceuticals to humans. The EU 6th framework project "PREDICTOMICS" was established in order to contribute to the further development of in vitro toxicology, with a particular focus on emerging techniques including toxicogenomics. DNA microarray technology is being used more frequently in the in vitro field, however, only very few studies have assessed the reproducibility of this technique with respect to in vitro toxicology. To this end we conducted an interlaboratory comparison to test the reproducibility of transcriptomic changes induced by the immunosuppressive agent, Cyclosporine A (CsA) on the human renal proximal tubular cell line, HK-2 cell. Four European laboratories took part in this study. Under standardised conditions, each laboratory treated HK-2 cells with 5microM CsA for 12 and 48h. RNA was isolated and hybridised to Affymetrix HGU-133 plus two arrays at three different sites. Analysis of the transcription profiles demonstrated that one laboratory clustered away from the other laboratories, potentially due to an inclusion of a trypsinisation step by this laboratory. Once the genes responsible for this separate clustering were removed all laboratories showed similar expression profiles. There was a major impact of time since feed, due to medium exhaustion in the 48h arrays compared to the 12h arrays, regardless of CsA treatment. Biological processes including general vesicle transport, amino acid metabolism, amino acid transport and amino acid biosynthesis were over-represented due to time since feed, while cell cycle, DNA replication, mitosis and DNA metabolism were under-represented. CsA responsive genes were involved in cell cycle, the p53 pathway and Wnt signaling. Additionally there was an overlap of differentially expressed genes due to CsA and medium exhaustion which is most likely due to CsA induced glycolysis. The glucose deprivation dependent genes HspA5 and GP96 and the Hsp70 chaperones DNAJ/Hsp40, DNAJ/HspB9, DNAJ/HspC3 DNAJ/HspC10 were induced by both CsA and medium exhaustion. We conclude that under standardised conditions the application of Affymetrix DNA microarrays to in vitro toxiciological studies are satisfactorily reproducible. However, confounding factors such as medium exhaustion must also be considered in such analyses.

Influence of microvascular endothelial cells on transcriptional regulation of proximal tubular epithelial cells.

In the renal cortex the peritubular capillary network and the proximal tubular epithelium cooperate in solute and water reabsorption, secretion, and inflammation. However, the mechanisms by which these two cell types coordinate such diverse functions remain to be characterized. Here we investigated the influence of microvascular endothelial cells on proximal tubule cells, using a filter-based, noncontact, close-proximity coculture of the human microvascular endothelial cell line HMEC-1 and the human proximal tubular epithelial cell line HK-2. With the use of DNA microarrays the transcriptomes of HK-2 cells cultured in mono- and coculture were compared. HK-2 cells in coculture exhibited a differential expression of 99 genes involved in pathways such as extracellular matrix (e.g., lysyl oxidase), cell-cell communication (e.g., IL-6 and IL-1 beta), and transport (e.g., GLUT3 and lipocalin 2). HK-2 cells also exhibited an enhanced paracellular gating function in coculture, which was dependent on HMEC-1-derived extracellular matrix. We identified a number of HMEC-1-enriched genes that are potential regulators of epithelial cell function such as extracellular matrix proteins (e.g., collagen I, III, IV, and V, laminin-alpha IV) and cytokines/growth factors (e.g., hepatocyte growth factor, endothelin-1, VEGF-C). This study demonstrates a complex network of communication between microvascular endothelial cells and proximal tubular epithelial cells that ultimately affects proximal tubular cell function. This coculture model and the data described will be important in the further elucidation of microvascular endothelial and proximal tubular epithelial cross talk mechanisms.

Cyclosporine A induces senescence in renal tubular epithelial cells.

The nephrotoxic potential of the widely used immunosuppressive agent cyclosporine A (CsA) is well recognized. However, the mechanism of renal tubular toxicity is not yet fully elucidated. Chronic CsA nephropathy and renal organ aging share some clinical features, such as renal fibrosis and tubular atrophy, raising the possibility that CsA may exert some of its deleterious effects via induction of a stress-induced senescent phenotype. We investigated this hypothesis in HK-2 cells and primary proximal tubular cells in vitro. CsA induced the production of H2O2, caused cell cycle arrest in the G0/G1 phase, and inhibited DNA synthesis. Furthermore, CsA exposure lead to a reduction of telomere length, increased p53 serine 15 phosphorylation, and caused an upregulation of the cell cycle inhibitor p21(Kip1) (CDKN1A) mRNA levels. CsA caused an increase in p16(INK4a) (CDKN2A) expression after a 13-day exposure in primary proximal tubular cells but not in HK-2 cells. Coincubation of cells with CsA and catalase was able to prevent telomere shortening and partially restored DNA synthesis. In summary, CsA induces cellular senescence in human renal tubular epithelial cells, which can be attenuated by scavenging reactive oxygen species.

Migration of leukocytes across an endothelium-epithelium bilayer as a model of renal interstitial inflammation.

Interstitial inflammation has emerged as a key event in the development of acute renal failure. To gain better insight into the nature of these inflammatory processes, the interplay between tubular epithelial cells, endothelial cells, and neutrophils (PMN) was investigated. A coculture transmigration model was developed, composed of human dermal microvascular endothelial (HDMEC) and human renal proximal tubular cells (HK-2) cultured on opposite sides of Transwell growth supports. Correct formation of an endoepithelial bilayer was verified by light and electron microscopy. The model was used to study the effects of endotoxin (LPS), tumor necrosis factor (TNF)-alpha, and alpha-melanocyte-stimulating hormone (alpha-MSH) by measuring PMN migration and cytokine release. To distinguish between individual roles of microvascular endothelial and epithelial cells in transmigration processes, migration of PMN was investigated separately in HK-2 and HDMEC monolayers. Sequential migration of PMN through endothelium and epithelium could be observed and was significantly increased after proinflammatory stimulation with either TNF-alpha or LPS (3.5 +/- 0.58 and 2.76 +/- 0.64-fold vs. control, respectively). Coincubation with alpha-MSH inhibited the transmigration of PMN through the bilayer after proinflammatory stimulation with LPS but not after TNF-alpha. The bilayers produced significant amounts of IL-8 and IL-6 mostly released from the epithelial cells. Furthermore, alpha-MSH decreased LPS-induced IL-6 secretion by 30% but had no significant effect on IL-8 secretion. We established a transmigration model showing sequential migration of PMN across microvascular endothelial and renal tubular epithelial cells stimulated by TNF-alpha and LPS. Anti-inflammatory effects of alpha-MSH in this bilayer model are demonstrated by inhibition on PMN transmigration and IL-6 secretion.

Membrane targeting and secretion of mutant uromodulin in familial juvenile hyperuricemic nephropathy.

Familial juvenile hyperuricemic nephropathy (FJHN) is an autosomal dominant genetic disorder that is characterized by hyperuricemia, gout, and tubulointerstitial nephritis. FJHN is caused by mutations in the UMOD gene, which encodes for uromodulin, the most abundant urinary protein. Herein is demonstrated that patients with FJHN and renal insufficiency exhibit a profound reduction in urinary uromodulin together with either elevated or decreased plasma uromodulin. One young patient with FJHN, however, had normal serum creatinine and normal urinary uromodulin with elevated plasma uromodulin. These observations suggest that there are different urinary and plasma uromodulin profiles in early and late disease and that there may be an altered direction of uromodulin secretion in the course of FJHN as a result of improper intracellular sorting of the mutated protein in the thick ascending limb. With the use of immunohistochemistry and a quantitative immunoassay, targeting and secretion of wild-type and mutant (C77Y and N128S) uromodulin were investigated in the polarized renal epithelial cell line LLC-PK1. In transfected cells, uromodulin mutants were targeted properly to the apical membrane but were secreted less efficiently to the apical compartment than wild-type protein. The expression of mutant uromodulin had no effect on caspase 3 activity. These results indicate that the mutations studied do not impair glycosyl-phosphatidylinositol-mediated apical targeting of the protein but do affect apical secretion. Because the mutant proteins are secreted as efficiently as wild type to the basolateral compartment, the possibility arises that interactions with the immune system at the site of secretion are a contributing factor to the development of tubulointerstitial nephritis in FJHN.