Physiological map to study kidney toxicity in the ONTOX project

Background and Objectives:
Continuous improvements of computational approaches also increase the predictive performances of toxicological in silico models [1]. However, being mainly based on animal test data, these computational models lack a good correlation with human toxicity, and, being often based uniquely on chemical structures, they are unable to explain toxicological processes. To overcome these limitations, we have developed a new semi-automated strategy to build a Physiological Map (PM), a framework to study human toxicological mechanisms.
Materials and Methods:
Our method is useful to build a PM or to validate an existing PM. To retrieve information, a manual literature review was accompanied by computational interrogation of ontologies (e.g. Gene Ontology), thus creating a network of genes, proteins, molecules and phenotypes [2]. The network was converted manually into a PM using the CellDesigner software and visualized on the web using the MINERVA platform. The entire procedure was supported and revised by field experts.
Results:
We present here the human kidney PM, developed in the framework of ONTOX, a European project aimed at improving risk assessment avoiding the use of animal tests [3]. With the purpose to better understand tubular necrosis and nephrolithiasis, the PM represents the normal physiology in proximal tubule, the loop of Henle, distal tubule, and collecting duct cells, displaying the vitamin D metabolism and the urine production processes: filtration, reabsorption and secretion.
Discussion and Conclusions:
Our method assists the user to build a PM even starting from limited data. The PM is initially a static representation of physiological processes, also useful to study and develop new adverse outcome pathways. Subsequently, we could add kinetic parameters, transforming the PM into a dynamic model able to represent cellular perturbations. This approach presents an opportunity to investigate human toxicities, improving the toxicological predictions from a qualitative and quantitative perspective.
References:
[1] Manganelli S, Gamba A, Colombo E, Benfenati E (2022) 'Using VEGAHUB Within a
Weight-of-Evidence Strategy'. In: Benfenati E. (eds) In Silico Methods for Predicting Drug
Toxicity. Methods in Molecular Biology, vol 2425. Humana, New York, NY.
https://doi.org/10.1007/978-1-0716-1960-5_18
[2] Gamba A, Salmona M, Bazzoni G (2020) 'Quantitative analysis of proteins which are
members of the same protein complex but cause locus heterogeneity in disease', Sci Rep
10, 10423. https://doi.org/10.1038/s41598-020-66836-7
[3] Vinken M., et al. (2021) 'Safer chemicals using less animals: kick-off of the European
ONTOX project', Toxicology 458, 152846. https://doi.org/10.1016/j.tox.2021.152846