Nephrotoxicity biomarkers in vivo and in vitro: Identification and validation of biomarkers for the detection of substance-induced acute and sub-acute renal damage
The major aims of this work included the evaluation of urinary protein markers, transcriptional analysis of the underlying mechanism of Vancomycin-induced nephrotoxicity based on gene expression analysis, the evaluation and generation of reliable transcriptional biomarkers. In addition, the utility...
| Summary: | The major aims of this work included the evaluation of urinary protein markers, transcriptional analysis of the underlying mechanism of Vancomycin-induced nephrotoxicity based on gene expression analysis, the evaluation and generation of reliable transcriptional biomarkers. In addition, the utility of an in vitro test system to produce in vivo like responses to nephrotoxin treatment was tested. Therefore, several questions, important for the further enhancement of preclinical safety assessment as well as the understanding of the underlying mechanism of drug-induced nephrotoxicity, were addressed.
For urinary biomarker analysis three well known nephrotoxins (Vancomycin, Cisplatin and Puromycin) were used. The urinary proteins Kim-1, Clusterin and Osteopontin were clearly identified as the most predictive biomarkers for tubular degeneration, while 2-microglobulin and Cystatin C can be used to discriminate glomerular induced proteinuria. In addition, exploratory urinary biomarkers like Timp-1, NGAL, Calbindin and VEGF showed great potential to deliver further information about the underlying mechanism of toxicity, the area of damage or potential gender differences. However, there are several limitations associated with these new markers and therefore, further investigations and improvements are still needed. The urine dipstick assay for Kim-1, as a quick and easy to use method, was compared to the multiplex Luminex assay. The data showed a clear limitation in sensitivity when there are only slight alterations in urinary Kim-1 levels. In cases where there are strong changes, the results showed a good correlation and can therefore be recommended as general screening tool.
More traditional detection systems, in this case immunohistochemical analysis, can help the identification of early renal changes in a reliable way. This could be especially shown for Kim-1 and Clusterin, while the interpretation of renal expression of Osteopontin was hindered because of strong basal expression and therefore background staining.
The transcriptomic analysis was performed primarily to obtain further insights into the underlying mechanism of Vancomycin-induced nephrotoxicity. Several potential key pathways and processes, which can be considered to be a good starting point for further investigations, were identified. For example, the inflammatory processes, characterized by an activation of the complement system, leukocyte extravasations and several processes involving the transcription factor NF-B, seems to play an important role in the nephritis induced by Vancomycin. In addition, a Fas-induced and caspase-dependent apoptosis was identified, which overlaps to the already mentioned inflammatory processes. The regenerative properties of the kidney, also reflected by the urinary protein biomarkers as well as almost all histo- and clinical-pathological observations, can be due to the important position of integrin-signaling and Rho-GTPase-induced actin remodeling. These processes lead to cellular shape restoration and cell migration.
The hypothesis that transcriptional changes reflect renal (and other organs) pathological changes is well accepted. Therefore, a commercial test system provided by Compugen Ltd. (based on the measurement by SYBR® green detection of several specific genes, in combination with a related software tool) was tested for its predictive power. Three independent rat studies were chosen for this approach, namely Gentamycin, 3-pyrrolidineacetic acid, 5-[[[4_-[imino[(methoxycarbonyl) amino]methyl] [1,1_-biphenyl]-4-yl]oxy]methyl]-2-oxo-, methyl ester,(3S-trans) (FP007SE/ BI-3), and aristolochia acid. The results indicated a good reflection of Gentamycin induced nephrotxicity, while FP007SE and aristolochic acid delivered unclear results. Several animals within the FP007SE and aristolochic acid studies were identified as positive for renal insult, but without any histopathological correlation. Because of the time- and/or dose-dependent increase in incidence, it was hypothesized that a real prediction, i.e. identification of substance induced alteration with no manifested morphological changes, could be possible.
Three published transcriptional biomarker lists (containing, 4 genes, 19 genes, and 45 genes) were evaluated for their potential to predict the outcome of our Vancomycin study. In addition, the methodological independence of such publicly available biomarker lists could be assessed. From these three gene lists a new “best predictor” biomarker list was created, by also taking the observations from the mechanistic gene expression analysis into account. This final list contains 40 individual transcripts which were grouped into 10 functional categories and reflect the biological reality in more detail. These can now be used in further exploratory or pre-clinical routine studies for the prediction of rodent specific nephrotoxicity.
Finally, a proof-of-principal investigation was conducted with the aim to study whether or not a renal cell line (NRK-52E) can reflect transcriptional changes in a comparable way to those observed in rat in vivo after treatment with nephrotoxins. 192 Significantly altered genes were identified over all positive substances (AmphotericinB, Doxorubicin, Puromycin, and Paracetamol); where 44 genes (~23%) have been frequently reported in the literature to be associated with renal cell damage. Nine of these genes, including Cd44, Cyr61 and Cdkn1a, were of special interest since they were some of the most prominently deregulated in vivo genes and a possible mechanistic relation was identified within this study. |
|---|