Identification and validation of blood biomarkers to develop effective diagnostic tools for coronary microvascular disease

Arredondo Eve, Alicia

Permalink

https://hdl.handle.net/2142/117537 Copy

Description

Title

Identification and validation of blood biomarkers to develop effective diagnostic tools for coronary microvascular disease

Author(s)

Arredondo Eve, Alicia

Issue Date

2022-10-25

Director of Research (if dissertation) or Advisor (if thesis)

Madak Erdogan, Zeynep

Doctoral Committee Chair(s)

Chen, Hong

Committee Member(s)

• Miller, Michael J.
• Arthur, Anna

Department of Study

Food Science & Human Nutrition

Discipline

Food Science & Human Nutrition

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Cardiovascular disease
• postmenopausal women
• coronary microvascular dysfunction
• metabolic circulating markers
• human serum
• PFAS

Abstract

Cardiovascular disease (CVD) is a major public health concern since it is the leading cause of death among men and women in the US. In addition to traditional risk factors such as unhealthy lifestyles (smoking, obesity, sedentary) and genetics, common environmental exposures, including persistent environmental contaminants (PFAS), may also influence cardiovascular disease risk. Additionally, postmenopausal state in women is accompanied by a higher risk of metabolic and cardiovascular diseases. This phenomenon is thought to be caused by the changes in sex hormone levels in the menopausal transition period, where endogenous estrogen levels decrease. Women have a higher prevalence of coronary microvascular disease (CMD); men tend to have atheroma and epicardial endothelial dysfunction or coronary artery disease (CAD). Most female patients with CMD are postmenopausal women, and hormone-replacement therapies (HRT) decrease CMD risk by up to 30% in this population, suggesting a role for estrogens in the development and progression of CMD. Since there are no standard and specific diagnostic tests for CMD, patients do not receive a positive diagnosis when tested. Therefore, there is a clinical need for novel ways to diagnose, treat and prevent CMD in postmenopausal women. In the current study, we focused on demonstrating that metabolic profiling is a promising strategy for developing a reliable biomarker, that in combination with computer learning techniques can be applied to the clinic by identifying novel circulating biomarkers of CMD (coronary microvascular disease), that could potentially be used as a diagnostic test. To assess the characterization of circulating biomarkers, we focused on plasma metabolite profile and composition of the gut microbiome in rats, to determine if changes in the plasma metabolite profiles caused by long-term dietary broccoli relate to molecular changes in liver. We aimed to identify plasma indicators that reflect how liver health is impacted by dietary broccoli. Another hypothesis we assessed was if, in our cardiovascular cohort studies, postmenopausal women with CMD have a distinct plasma metabolite profile compared to healthy women and women with CAD. Likewise, we also looked at environmental toxicants perfluoroalkyl substances (PFAS) to assess if, in our cardiovascular (CVD) cohort study, PFAS levels were significantly higher in any of the groups. We hypothesize that postmenopausal women with cardiovascular diseases will have higher plasma levels of PFAS due to the cessation of menstruation, which is an important route of PFAS elimination. We performed full metabolite and PFAS profiling of plasma samples from animals and individuals and, identified and classified circulating biomarkers using machine learning approaches for our CVD study. We identify twenty-five plasma metabolites that change with broccoli consumption, including metabolites associated with hepatic glutathione synthesis (glutamine, s-methyl-L-cysteine) and de novo fatty acid synthesis (stearic acid) in rats fed broccoli long-term. In our CVD study, stearic acid and ornithine levels were significantly higher in postmenopausal women with CMD. In contrast, valine levels were higher for women with CAD. When we included fifteen metabolites that were identified by the algorithm using machine learning techniques, our mean area under the curve (AUC) increased, suggesting nonlinear methods may provide us with a better set of biomarkers to develop as a diagnostic tool for CMD. We also observed that perfluorooctane sulfonate (PFOS) levels were significantly higher in postmenopausal women with CAD when compared to control and CMD group, which was supported by machine learning techniques where CAD and PFOS had moderate AUC. Overall, our studies in this project indicated that metabolic profiling is a suitable method to assess how the body response to internal and external stimuli such as lifestyle, genetics, age, and disease in a presently manner. Furthermore, when combined with computer learning techniques it provides a broader view of biology and integrative pathways which help us to develop a more reliable biomarker that can be applied to the clinic.

Graduation Semester

2022-12

Type of Resource

Thesis

Copyright and License Information

Copyright 2022 Alicia Arredondo Eve

Owning Collections