Bioinformatics analysis of potential biological effects of herbicides on human health

JUAN CRUZ RODRIGUEZ ELGASSI; ANTONELLA VILA; MATÍAS BLAUSTEIN; MERCEDES GARCÍA CARRILLO

Congreso; Reunión Anual de la Sociedad Argentina de Investigación Bioquímica y Biología Molecular - SAIB; 2022

There is great controversy concerning the potential toxicological effects of several herbicides commonly used in agriculture. A growing body of scientific evidence suggests a negative impact on human health. The objective of this work is to evaluate -in a comparative manner- the potential toxicological effects on human health of four herbicides commonly used in Argentina: glyphosate, atrazine, 2,4-D, and paraquat. A bioinformatic analysis was performed to identify which proteins could be affected by their interaction with these herbicides, identifying in turn those biological processes and diseases that could be altered as a result of the interactions between these herbicides and proteins. To determine which proteins could be affected in their regulation by the use of the mentioned herbicides in humans, the STITCH platform was used. This platform integrates databases, text bases and prediction methods, generating interaction networks between a given chemical compound and those proteins involved in different biological processes. The STRING platform was used in turn to obtain a second shell of protein interactors. For both platforms, only those interactions from curated data and/or experimental data ("Database" and "Experiments", respectively) were selected. Only those interactions with a confidence score of medium or higher (≥0.4 points) were considered and a maximum of 500 interactors was set. To identify those biological processes mostly affected as a consequence of the interaction between each of the herbicides under study and those proteins identified in the interaction networks, overrepresented terms associated with the "Biological Process´´ category of Gene Ontology were extracted from the protein list from the interaction networks by performing an enrichment analysis. Moreover, an enrichment analysis of overrepresented terms in the protein list for those diseases annotated in the DisGeNET database was performed. Firstly, interaction networks for each one of the herbicides under study revealed the presence of proteins involved in two signaling pathways linked to cell proliferation, cell death, stress response and cell survival: the PI3K/Akt pathway and the Unfolded Protein Response (UPR). Our results also showed that all herbicides interact with proteins involved in key biological processes such as proliferation, cell growth, cytoskeletal organization, and axogenesis, among others. For all the herbicides under analysis, we could observe an enrichment in terms associated with cancer such as carcinoma, sarcoma, lymphoma, and immune tumors, and in terms associated with neuronal pathologies such as the Alzheimer’s family of diseases. These results are consistent with past reports on the toxicological effects of these herbicides on human health and their impact on key biological processes. Furthermore, our results indicate that the regulation of both the PI3K/Akt pathway and the UPR could be disrupted by exposure to these herbicides.