Search

Environmental stressors come in many different forms and affect every level of ecological organization. In a natural environment, organisms experience mixtures of stressors at various doses or strengths of exposure that could be constantly changing, both temporally and spatially. Complex environmental regimes and interactions among multiple stressors can have a profound effect on both the short-term fitness of organisms and their long-term evolutionary dynamics. This dissertation uses a bacterium-drug system to study drug-drug, drug-temperature, and gene-gene interactions. The first question we ask is how bacteria evolve to combat new stressors. Through an interaction network clustering approach followed by transcriptomic analysis, we show that Escherichia coli may have co-opted its cellular response to temperature stress for antibiotic stress because these stressors share similar physiological effects. We also found that antibiotic stress modifies the thermal response of E. coli by altering both its optimal growth temperature and temperature breadth. Next, we ask how we can identify and quantify dose-dependent drug interactions, including interactions among more than two components, i.e., higher-order interactions. To do this, we introduce a novel visual representation termed interaction landscape to directly map local dose-dependent interactions and the transitions between different interaction classes. Finally, we ask how gene-gene interactions change in fluctuating environments. We showed that changes in the type and magnitude of environmental fluctuations could affect fitness due to the differences in epistatic interactions of mutations. We quantify structural features of fitness landscapes by calculating the ruggedness across broad concentration gradients of various antibiotics. We show that fluctuating environments frequently lead to epistasis sign switches (from negative to positive or vice versa) on the pairwise level, as a potential mechanism to either promote specialization or maintain genetic variation. Overall, this dissertation combines experimental, mathematical, and computational biology to identify and understand the structures and patterns of interactions at different scales and their effects on the fitness and eco-evolutionary dynamics of bacterial populations.

BackgroundAlthough antipsychotics can relieve or prevent many psychotic disorders, it also has risks. They are subject to drug-drug interactions (DDIs) amongst themselves and with other agents used in the treatment of various physical conditions. Given the significant clinical effect of DDIs with antipsychotics on patient’s health and potential economic impact of DDIs on healthcare expenditure, little information is available about the prevalence, risk factors, and cost of DDIs with antipsychotics and is based on small studies with uncertain generalizability.Research PurposesThe purposes of this research were, first, to develop a list of clinically important DDIs likely to be encountered by U.S. adult population who treated with any antipsychotics and then determine the prevalence and identify predictors of DDIs. The last purpose of this study is to investigate the relationship between the health expenditures and DDIs.MethodsIn this research, I first conducted a systemic review of four compendia and published literature to identify all potential important DDIs. Then, the list was reviewed and approved by a panel of experts to be used as criteria in our analyses. A retrospective database analysis was conducted, and individuals who were exposed to any DDI were identified from the 2010-2014 Medical Expenditure Panel Survey (MEPS). The prevalence of DDI was evaluated using the developed list of clinically important DDIs. Next, I developed a theoretical framework based on Anderson Behavioral Model to test relationships among the latent constructs of predisposing characteristics, enabling resources, and need for healthcare, health behaviors, and medication use factor; and, their influence on DDIs exposure. The model also was used in investigating the relationship between the total health expenditures and DDIs exposure by employing multiple regression models and the propensity score method.ResultsFrom the process of identifying clinically important DDIs pairs, a total of 277 DDIs were identified. From 2010 to 2014, the national prevalence of DDIs was 4.9 million (36.12%). Exposure of an individual to a DDIs was influenced by patient’s poor perceived physical health status (adjusted odds ratio [OR], 2.23; 95% confidence interval [CI], 1.58-2.36), obesity (OR, 2.89; 95% CI, 1.47- 5.68), higher number of physical conditions, seeing internist as a primary care provider (OR, 3.43; 95% CI, 1.41-8.38), and polypharmacy (OR, 11.39; 95% CI, 4.01-32.42).In the adjusted models, adults exposed to DDIs cost 35% more annual total health care expenditures than those are not exposed (RR = 1.35, 95% CI [1.23, 1.48]). Likewise, cost of office-based visits (RR = 1.36, 95% CI [1.24 to 1.50]), and prescription drugs (RR = 1.50, 95% CI [1.36 to 1.66]) were significantly associated with exposure to DDIs. ConclusionThe prevalence of DDIs substantially high among adults treated with antipsychotics. It was also concluded that factors such as polypharmacy influenced the occurrence of DDIs. A significant relationship between the exposure to the DDI and higher total health care expenditures were found. This identification of high-risk adults can be used by policy makers in implement intervention strategies to lower the DDI incidence and in reducing the overall cost of care.

Pediatric patients are at an increased risk for medication errors and can benefit from processes that facilitate and promote medication safety (Stone et al., 2010). Medication reconciliation (Med Rec) is a valuable tool in improving patients’ medication safety and reducing adverse drug events (The Joint Commission, 2015). The purpose of this evidence-based practice (EBP) project was to improve the accuracy of the Med Rec process in a Midwestern pediatric hematology/oncology outpatient clinic by developing, promoting, and evaluating a standardized, collaborative Med Rec process. The Stetler EBP model guided the implementation of the intervention, with the goal of integrating current evidence into current practice. Kotter’s Model of Change laid the theoretical foundation for successful implementation of a current practice change. This EBP project intervention included a patient and team member component. The patient component consisted of a verbal call reminder to bring medications to the visit, a patient handout emphasizing the importance of medication safety and reconciliation, and patient education regarding Med Rec process. The team member component included education regarding the importance of the Med Rec process and updates regarding Med Rec accuracy. The outcomes measured included the number, type, and severity of medication discrepancies and the number of voluntarily reported medication errors. Data were collected during Phase 1 (pre-intervention) and Phase 2 (post-intervention) by the physicians and the project leader (PL). These data were analyzed using chi-square tests. The intervention lead to a significant increase in the number of accurate Med Recs reported by the physicians between Phase 1 (n = 50, 70%) and Phase 2 (n = 65, 90.8%) (X2 = 8.167, df = 1, p = .004). An insignificant decrease in the number of accurate Med Recs was reported by the PL between Phase 1 (73.1%) and 2 (72.5%) (X2 = .003, df = 1, p = 0.959). Physicians reported more incorrectness errors in Phase 1 (73.3%) and Phase 2 (83.3%) than incompleteness errors (X2 = .481, df = 1, p = .786). PL reported more incompleteness errors in Phase 1 (71.4%) and Phase 2 than incorrectness errors (81.8%) (X2 = 1.670, df = 2, p = .434). The majority of Med Rec inaccuracies were classified as minor during Phase 1 and 2 by the physicians (X2 = .827, df = 2, p = .363) and the PL (X2 = 1.039, df = 1, p = .308). No inaccurate Med Rec was classified as severe by physicians or the PL. Finally, there were no voluntary medication errors were reported during the duration of the EBP project. Revision and replication of this EBP project would be helpful in further improving Med Rec accuracy in this setting.

Background: Bexarotene, a novel RXR agonist, is used in the treatment of cutaneous T cell lymphoma and is under investigation for the treatment of solid tumors in combination with other chemotherapeutic agents including docetaxel. Both docetaxel and bexarotene are metabolized by CYP3A4 and in animal studies bexarotene has been shown to enhance the activity of CYP3A enzymes and increase its content. Thus, there exists a potential for interactions between these two agents, which was evaluated in this study. Method: Bexarotene from commercial capsules was dissolved in DMSO. Employing pooled human liver microsomes, we assessed the influence of bexarotene on the metabolism of docetaxel. Microsomal fractions (1mg protein) were incubated with bexarotene (0.1 – 10µM) and docetaxel (0.5 – 5µM) in the presence of NADPH (1 mM). Using liquid-liquid extraction, unreacted docetaxel was extracted and levels were measured using a validated HPLC method. Next, we employed primary human hepatocytes to assess the inductive effects of bexarotene. Hepatocytes were treated with bexarotene (1 – 50µM) for 72 hours. Docetaxel metabolism by microsomal fractions was then evaluated. Testosterone 6â-hydroxylation and CYP3A4- specific protein and mRNA levels were also measured to evaluate whether bexarotene acts as an inducer of CYP3A4. Results: Bexarotene, at clinically relevant concentrations (1-10µM) did not inhibit docetaxel metabolism. Also, incubation of primary human hepatocytes with bexarotene did not enhance docetaxel metabolism, which was consistent with the observation that the testosterone 6â-hydroxylation and CYP3A4 expression in bexarotene – treated cells were not statistically different than those from untreated cells. Conclusions: Our findings suggest that bexarotene is unlikely to alter hepatic metabolism of docetaxel. Bexarotene appears to exhibit species – specific differences in the induction of CYP3A and further studies are required to understand the mechanistic basis of these differences.