High-throughput platform for predicting kidney toxicity is part of new international collaboration between Singapore researchers and US EPA to advance non-animal approaches to chemical safety testing

Home / Journal Watch / High-throughput platform for predicting kidney toxicity is part of new international collaboration between Singapore researchers and US EPA to advance non-animal approaches to chemical safety testing

Journal Watch

High-throughput platform for predicting kidney toxicity is part of new international collaboration between Singapore researchers and US EPA to advance non-animal approaches to chemical safety testing

Researchers from the Institute of Bioengineering and Nanotechnology and the Bioinformatics Institute at the Agency for Science, Technology and Research (A*STAR) in Singapore have reported significant breakthroughs in the development of an animal-free approach for predicting the toxicity of drugs and other substances to the human kidney (i.e., nephrotoxicity).

The proximal tubule cells (PTCs) of the kidney are a major target of toxic injury. Researchers from A*STAR’s Zink laboratory originally used embryonic stem cells to establish cultures of human PTCs for toxicity testing (Narayanan et al., 2013), but have now developed a more efficient method using reprogrammed induced pluripotent stem cells (iPSCs) (Kandasamy et al., 2015).

Another group of A*STAR researchers, the Loo laboratory, developed an automated testing platform that was used to screen the toxicity of 30 substances using the iPSC-derived PTCs. The expression of interleukin-6 (IL-6) and interleukin-8 (IL-8) had previously been identified as endpoints predictive of toxicity in cultured human PTCs (Li et al., 2013), and these endpoints were also predictive of kidney toxicity in the iPSC-derived PTCs (Kandasamy et al., 2015). This screening platform was also useful in identifying mechanisms of toxicity/injury to the kidney cells.

In a subsequent publication, these researchers used microscopic cellular imaging of primary human PTCs, also called phenotypic profiling, to predict human kidney toxicants (Su et al., 2015). In these experiments, 129 image-based phenotypic features were quantified, and the cellular features most predictive of in vivo PTC toxicity were identified for 44 reference compounds. A clustering analysis based on the phenotypic features revealed two major clusters, one enriched in PTC toxic substances and one in non-PTC toxic substances. The researchers concluded that their high-throughput kidney cell-based imaging platform can be used to accurately predict human nephrotoxic substances with diverse chemical structures.

A surprising result was the DNA damage response caused by most of the PTC-toxic substances (Su et al., 2015). Two DNA phenotypic features indicating changes to the chromatin structure and suggesting a DNA damage response were observed with most of the PTC-toxic substances. Furthermore, γH2AX, “a known marker for genotoxicity and carcinogenesis, was also induced by many compounds with diverse chemical structures.”

A*STAR and US EPA Collaboration

On January 25, 2016, an A*STAR press release announced a collaboration between A*STAR and the US Environmental Protection Agency (EPA) on several new approaches to chemical safety testing without the use of animals.

A*STAR will draw on its multidisciplinary capabilities in stem cell research and tissue models, genomics, high throughput bioimaging, and computational sciences. The collaboration will build on EPA’s ToxCast program which has generated high-throughput screening data on over 1,800 chemicals.”

Researchers from both organizations will be collaborating on three areas of research:

Kidney toxicity – This project will use the first and only predictive kidney technologies that were developed by [A*STAR scientists] to predict the effects of environmental toxicants on the human kidney efficiently and accurately. Their innovative technologies include stem cell-based models and a powerful high-throughput platform.

Liver toxicity – This project will use 3D liver models developed at [A*STAR] and computational tools at the NCCT to identify novel predictive biomarkers of human liver toxicity to overcome limitations in existing 2D model tests, which limit their sensitivity, especially over extended periods. Machine learning approaches will be used to analyse and improve existing predictive models of acute and sub-acute liver toxicity.

Developmental toxicity – This project aims to investigate the potential of certain chemicals to disrupt the development of blood vessels and the blood-brain-barrier during prenatal development – a key process during one of the most important life stages.

Articles featured:

Kandasamy, K., Chuah, J. K. C, Su, R., Huang, P., Eng, K. G., Xiong, S., Li, Y., Chia, C.S., Loo, L.H., & Zink, D. (2015). Prediction of drug-induced nephrotoxicity and injury mechanisms with human induced pluripotent stem cell-derived cells and machine learning methods. Sci. Rep. 5, 12337. Available at: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4515747/

Su, R., Xiong, S., Zink, D., & Loo, L.H. (2015). High-throughput imaging-based nephrotoxicity prediction for xenobiotics with diverse chemical structures. Arch. Toxicol. [Epub ahead of print]. Available at: http://link.springer.com/article/10.1007/s00204-015-1638-y/fulltext.html

Posted: February 22, 2016

Leave a Comment


*