Date of Award


Document Type

Campus Access Dissertation


Biomedical Science

First Advisor

Kim E Creek


Disruption of the transforming growth factor-beta (TGF-β) signaling pathway is observed in many cancers, including cervical cancer, resulting in TGF-β resistance. While normal human keratinocytes (HKc) and human papillomavirus type 16 immortalized HKc (HKc/HPV16) are sensitive to the growth inhibitory effects of TGF-β, HKc/HPV16 gradually develop resistance to TGF-β as the cells progress in vitro to a differentiation resistant phenotype (HKc/DR). Previous studies in our laboratory have demonstrated that the complete loss of TGF-β cytostatic effects in HKc/DR is attributed to decreased expression of TGF-β receptor type I, while the levels of TGF-β receptor type II remain unchanged. Despite the complete resistance to TGF-β elicited cytostasis in HKc/DR, we have previously shown using a Smad-driven luciferase reporter construct that TGF-β signaling is only reduced about 50% in HKc/DR as compared to HKc/HPV16. In the first section of the present thesis we studied whether alterations in the expression of the Smad components of the TGF-β signaling pathway could be involved in the development of resistance to the growth inhibitory effects of TGF-β during in vitro progression of HKc/HPV16. In addition we investigated how TGF-β induced nuclear trafficking of Smad3 and Smad4, and phosphorylation of Smad2 are affected during in vitro progression. We did not find any consistent decrease in protein expression of Smad2, Smad3 or Smad4; or increase in protein expression of inhibitory Smad7 that paralleled the increasing loss of sensitivity to the growth inhibitory response to TGF-β observed in our in vitro model of HPV16-mediated transformation. We did observe delayed TGF-β signaling, as assessed by Smad3 nuclear accumulation and Smad2 phosphorylation experiments, and a diminished TGF-β-mediated phosphorylation of Smad2 in HKc/DR compared to normal HKc and TGF-β sensitive HKc/HPV16. These results reinforce previous findings in our laboratory which demonstrated that, although TGF-β signaling is impaired in HKc/DR, there is still some signaling left but it is not sufficient to elicit the TGF-β antiproliferative effects.

In the second section of this thesis we present studies aiming at identifying molecular biomarkers that can predict what women are at higher risk of developing cervical cancer following human papillomavirus (HPV) infection. The main and necessary etiologic agent in the development of cervical cancer is the infection of cervical cells with sexually-transmitted, high-risk HPV (HR-HPVs), with HPV16 being the most prevalent type. Although most women will come in contact and get infected with HPV at some point during their lifetime, cervical cancer is not a frequent outcome of HR-HPV infection and only women with persistent HPV infection are truly at risk of developing cervical cancer. In spite of the fact that cervical cancer incidence has steadily decreased in the U.S. over the past fifty years due to broadly available screening and follow-up programs that detect and treat abnormal cervical lesions, respectively, cervical cancer remains a considerable burden worldwide with nearly 300,000 deaths each year. Nonetheless, since a significant proportion of abnormal cervical lesions would spontaneously regress if not treated, the current screening and follow-up protocols result in expensive, unnecessary and painful interventions, some of which have detrimental consequences for future pregnancies. We performed microarray experiments on RNA samples isolated from exfoliated cervical cells to search for biomarkers that could predict persistent HPV16 infections, the ones at higher risk of progressing into cervical cancer, which would constitute a cheaper, non-invasive and more accurate tool for the triage of women for further treatment. In order to be able to perform such experiments we first optimized the RNA isolation method since previous reported protocols yielded low quality RNA from exfoliated cervical cells. We identified 39 differentially expressed genes in HPV16 persistent infections as compared to HPV16 non-persistent infections. In addition, we identified a molecular signature composed of 9 genes that has an immense potential clinical importance since it can perfectly predict if HPV16 infections will be persistent or will be cleared. Once these results are validated and translated into clinical diagnostic tools, women will be able to be triaged more accurately for further follow up and treatment. In addition, we also compared cytologically normal and abnormal samples and found 11 genes that are able to perfectly classify the studied samples in their corresponding cytological groups. An accurate cytologic classification in conjunction with HR-HPV persistence assessment could subcategorize persister patients into immediate treatment or in "closely follow-up" groups. However, additional studies are required to validate and expand the presented findings.