Diagnosis of drug-induced renal tubular toxicity using global gene expression profiles

Abstract

Toxicogenomics can measure the expression of thousands of genes to identify changes associated with drug induced toxicities. It is expected that toxicogenomics can be an alternative or complementary approach in preclinical drug safety evaluation to identify or predict drug induced toxicities. One of the major concerns in applying toxicogenomics to diagnose or predict drug induced organ toxicity, is how generalizable the statistical classification model is when derived from small datasets? Here we presented that a diagnosis of kidney proximal tubule toxicity, measured by pathology, can successfully be achieved even with a study design of limited number of training studies or samples. We selected a total of ten kidney toxicants, designed the in life study with multiple dose and multiple time points to cover samples at doses and time points with or without concurrent toxicity. We employed SVM (Support Vector Machine) as the classification algorithm for the toxicogenomic diagnosis of kidney proximal tubule toxicity. Instead of applying cross validation methods, we used an independent testing set by dividing the studies or samples into independent training and testing sets to evaluate the diagnostic performance. We achieved a Sn (sensitivity) = 88% and a Sp (specificity) = 91%. The diagnosis performance underscores the potential application of toxicogenomics in a preclinical lead optimization process of drugs entering into development.

Identification of genomic biomarkers for concurrent diagnosis of drug-induced renal tubular injury using a large-scale toxicogenomics database

“…In a separate study that assessed the expression profiling endpoints in parallel with the histopathological diagnosis of concurrent renal tubular toxicity, the performance was improved and a sensitivity of 82% was achieved with 100% of selectivity (Thukral et al, 2005). Furthermore, Jiang et al (2007) achieved a sensitivity of 88% and a specificity of 91% using the expression profiling endpoints in parallel with the histopathological diagnosis of concurrent renal tubular toxicity. It is thought that concurrent diagnosis is easier than the prediction of future onset because the early stage toxic gene expression changes are heterogenic between different compounds, but the gene expression changes concurrent with the same toxic endpoints are comparatively homogenous among different compounds.…”

Modulation of Porcine β-Defensins 1 and 2 upon Individual and Combined Fusarium Toxin Exposure in a Swine Jejunal Epithelial Cell Line

“…Despite advances in the technologies for measuring protein abundance in recent decades, the experimental techniques for protein identification and quantification still lag considerably behind the highly sensitive methods available for quantifying mRNA transcript levels (45). However, while mRNA expression values are useful in various applications, such as classification, identification, and prediction of drug-induced toxicities or cancers (46,47), the results are correlative rather than causative. It is generally recognized that changes in protein levels, even when subtle (i.e., less than what is deemed statistically significant), may have significant biological effects, and this is currently an area in which analytical techniques with increased sensitivity are required (45).…”

Using transcriptomics to guide lead optimization in drug discovery projects: Lessons learned from the QSTAR project

“…The use of transcriptomic data for characterizing biological effects of small molecules has become increasingly popular since the advent of the Connectivity Map [15]. Several applications ranging from pathway elucidation [16], toxicity models [17,18] and toxicogenomic classifications [19] to tool discovery and drug repurposing [20][21][22][23] have been developed based on drug-induced gene expression profiling [9]. However, whereas these studies certainly have significant scientific value, they do not address the utility of gene expression profiling for decision making during the lead optimization phase of a typical drug discovery project.…”