Publication Date

5-2023

Date of Final Oral Examination (Defense)

March 2023

Type of Culminating Activity

Thesis

Degree Title

Master of Science in Biology

Department

Biology

Supervisory Committee Chair

Jennifer S. Forbey, Ph.D.

Supervisory Committee Member

Sven Buerki, Ph.D.

Supervisory Committee Member

Lisa Warner, Ph.D.

Abstract

Pharmacokinetics (PK) is the time course of a compound in the body that is dependent on mechanisms of absorption, distribution, metabolism, and excretion or ADME. A thorough understanding of PK is essential to predict the consequences of organisms exposed to chemicals. In medicine, predictions of PK of drugs allows us to properly prescribe drug treatments. In toxicology, PK allows us to predict the potential exposure of environmental contaminants and how they may affect organisms at the time of exposure or in the future. Chemical ecology could benefit from computational predictions of PK to better understand which plants are consumed or avoided by wild herbivores. A limitation in computational predictions of PK in chemical ecology is the large quantities of biodiverse natural products involved in complex plant-herbivore-microbial interactions compared to biomedical and environmental toxicology studies that focus on a select number of chemicals. The objective of this research was to automate the process of mining predicted PK of known chemical structures in plants consumed by herbivores and to use predicted PK output to test hypotheses. The first hypothesis is that because monoterpenes are smaller in molecular weight and have relatively high lipophilicity when compared to phenolics and sesquiterpenes, they would have higher absorption, be more likely to be substrates for efflux transporters that regulate absorption, and be more likely to inhibit metabolizing enzymes than phenolics and sesquiterpenes. The second hypothesis is that monoterpenes that are induced or avoided by foraging herbivores would have higher absorption, be less likely to be substates for efflux transporters, and be more likely to inhibit metabolizing enzymes compared to the individual monoterpenes that are not induced or avoided by herbivores. This automated approach used Python packages to obtain chemical notations from the PubChem website and mine predicted PK information for chemical input from the SwissADME website. The PK output from SwissADME was analyzed using ANOVAs to test for differences in molecular weight and lipophilicity among chemical classes (monoterpenes, phenolics, and sesquiterpenes). Chi-squared tests were used to assess if chemical groups had high or low absorption, were substrates of efflux transporters, or inhibited metabolizing enzymes. Mined PK data for chemicals can be used to understand drug-drug interactions in pharmacology, predict exposure to environmental contaminants in toxicology, and identify mechanisms mediating plant-microbe-herbivore interactions. However, the broad benefits of mining predicted PK across disciplines requires a workforce with competency in chemistry, physiology, and computing who can validate the automation process and test hypotheses relative to different disciplines. Course-based and Lab-based Undergraduate Research Experiences (CUREs and LUREs) have been proven to not only improve grades but also increase engagement diversity and inclusion. As a graduate teaching assistant, I created and taught a PK LURE module in an undergraduate Animal Physiology and Nutrition course to create a sustainable quality control step to validate input of chemical structures and PK output generated from the automated process. The course simultaneously provided students with an authentic research experience where they integrated chemistry, pharmacology, computing, public databases, and literature searches to propose and test new hypotheses. Students gained indispensable interdisciplinary research skills that can be transferred to jobs in veterinary and human medicine, pharmaceutics, and natural sciences. Moreover, undergraduates used existing and new PK data to generate and test novel hypotheses that go beyond the work of any single graduate student or discipline. Overall, the integration of computing and authentic research experiences has advanced the research capacity of a diverse workforce who can predict exposure and consequences of chemicals in organisms.

DOI

https://doi.org/10.18122/td.2088.boisestate

Download

Included in

Biology Commons

Share

COinS