Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Medical Sciences

Major Professor

Patricia Kruk, Ph.D.

Committee Member

Paula C. Bickford, Ph.D.

Committee Member

Meera Nanjundan, Ph.D.

Committee Member

Marzenna Wiranowska, Ph.D.

Committee Member

Lauri Wright, Ph.D.


Ovarian cancer, amylase, computational analysis, glycocalyx, cellular invasion


Ovarian cancer (OC) accounts for 4% of all cancer cases and 4.2% of all cancer deaths worldwide. OC is the most lethal gynecological cancer because it lacks early disease symptoms and does not have a specific diagnostic marker. As a result, more than 70% of OC patients are diagnosed in later stages when the disease has already metastasized and the 5-year survival rate has decreased to less than 20% compared with approximately 90% survival for women diagnosed with early stage disease. Therefore, I initiated my studies with a computational analysis of the 27 most commonly reported literature-derived ovarian cancer (LDOC) protein biomarkers. I found that LDOC protein biomarkers share many biochemical features including a preponderance for a stable protein structure, the ability to be secreted, and functionality related to extracellular matrix (ECM) modification, immune response and/or energy production. Subsequently, I analyzed the human proteome to identify proteins that also share these biochemical features. Of the 70,616 proteins in the human proteome, 683 proteins were found to have similar biochemical features to the 27 LDOC proteins. I also identified a subset of 21 potential additional protein regulators of ovarian cancer (APROC) that interact with LDOCs. Three of the APROCs identified were amylase proteins AMY1A, AMY2A, and AMY2B which cleaves alpha 1, 4-glycosidic bonds in polysaccharides. Amylase is reportedly overexpressed in and secreted by ovarian tumors but its functional contribution to OC remains unknown[1]. In this thesis, I posit that amylase contributes to OC invasion. I initiated my studies by computational characterizing the different amylase isozymes to predict which amylase isozyme(s) is most likely overexpressed in and contributory to OC invasion. I found that AMY1 and AMY2B have unique regions of disorder and unique phosphorylation sites indicating that AMY1 and AMY2B would be more likely to interact with other proteins, and to be easily secreted. Using OC patient serum samples, I was able to validate AMY1 and AMY2B overexpression by western immunoblotting.

I then developed an in vitro model system to study the molecular contribution of amylase to OC invasion using normal ovarian surface epithelial (IOSE) and OC cell lines. I showed that OC cells generally overexpress and secrete metabolically active amylase isozymes AMY1 and AMY2B. Abrogating amylase activity using siRNA silencing technology decreased the capacity of OC cells to invade collagen coated Boyden chambers and increased sulfated glycosaminoglycans (sGAG) production. Since a survey of OC cell lines indicated that cancer cells have a bulkier glycocalyx compared to IOSE cells and immunogold labeling studies indicated the presence of amylase within the immediate OC microenvironment, my data suggest that, by cleaving alpha 1, 4-glycosidic bonds in glycoconjugates present within ECM, amylase may remodel the ECM to promote an invasive cancer phenotype. Amylase is therefore a target for therapeutic intervention in OC patients with hyperamylasemia. I established Spirulina, a dietary supplement, as a novel transcriptional inhibitor of amylase. Spirulina inhibited amylase expression in OC cell lines at both the message and protein levels. Spirulina reduced OC cell invasion and migration in vitro, putatively by decreasing amylase expression.

Included in

Pathology Commons