Date of Award


Level of Access Assigned by Author

Campus-Only Dissertation

Degree Name

Doctor of Philosophy (PhD)


Functional Genomics/Interdisciplinary


Peter Brooks

Second Committee Member

Calvin Vary

Third Committee Member

Robert Friesel


Accumulating evidence indicates that malignant tumor progression depends not only on tumor cells themselves, but also on stromal cells comprising the malignant lesion, including tumor-associated macrophages (TAMs). TAMs have been recognized for contributions to structural changes in the extracellular matrix (ECM) through remodeling of architectural proteins like collagen type-I. Previous studies have shown that structural remodeling of the ECM results in localized triggering of what has been termed “biomechanical ECM switches”, resulting in exposure of matrix immobilized cryptic ECM epitopes. Collagen type-I has a number of Arg-Gly-Asp (RGD) containing cryptic sites and soluble RGD-containing reagents are currently being developed as angiogenic antagonist for the treatment of malignant tumors. However, it is not known if the distinct cryptic RGD elements present in collagen are functionally redundant or if specific flanking sequences convey differential control of cellular behavior. To address this question, a series of monoclonal antibodies (Mabs) that target distinct RGD- containing cryptic sites of collagen type-I have been generated. One Mab, designated XL313, exhibits selective recognition of the conserved RGDKGE cryptic element and significantly (P<0.05) inhibits B16F10 melanoma tumor growth as well as growth factor induced angiogenesis and inflammation in vivo. Interestingly, Mab XL 166, selective for alternative less conserved RGD sites of collagen type-I, showed no effect on tumor growth, angiogenesis or inflammation. To evaluate the role of the RGDKGE motif in angiogenesis and inflammation in more detail, I examined exposure of this epitope within inflamed tissues. Surprisingly, while the RGDKGE epitope was detected in close proximity to angiogenic blood vessels, the majority of the epitope localized to what appeared to be inflammatory infiltrates, suggesting a role for inflammatory cells in generating the Mab XL313 reactive collagen element. Importantly, my data also indicate that the soluble form of the RGDKGE-containing epitope, but not the Mab XL 166 reactive epitopes, significantly (P<0.05) induced both angiogenesis and inflammation in vivo.

The RGDKGE-containing peptide, but not the RGDAPG peptide, induced phosphorylation of P38MAPK in vivo, suggesting that generation of different RGD collagen peptides may result in distinct signaling cascades. Similar results, indicating activation of P38MAPK were observed in human endothelial cells attached to the RGDKGE epitope in vitro. Given these new findings along with previous data indicating a role for the RGDKGE epitope in angiogenesis, I investigated whether P38MAPK phosphorylation is required for RGDKGE peptide-stimulated angiogenesis. A significant inhibition of new vessel recruitment in RGDKGE peptide-stimulated tissues treated with the p38MAPK inhibitor, SB202190 was observed. Due to emerging information that has proposed a link between integrin signaling and phosphorylation of P38MAPK, I sought to identify potential receptors and downstream signaling cascades driving differential responses to the distinct RGD-containing epitopes. Endothelial cell adhesion to the RGDKGE containing peptide induced integrin P3 phosphorylation in vitro, indicating an important role for αvβ3 in functioning as a receptor for the RGDKGE containing peptide in endothelial cells.

Endothelial cell adhesion to the RGDKGE peptide, but not the RGDAPG containing peptide, induced phosphorylation of Src, and P38MAPK as well as an observed nuclear accumulation of YAP. I observed that blocking Src activity with a small molecule inhibitor, PP2, resulted in impaired P3 and P38MAPK phosphorylation. Furthermore, the RGDKGE containing collagen peptide induced endothelial cell proliferation that was dependent on YAP.

Collectively, these findings are consistent with a novel mechanism by which generation of a specific highly conserved cryptic RGD collagen peptide enhances rather than inhibits angiogenesis in part through an αvβ3/P38MAPK/YAP-dependent mechanism. These results provide novel insight into the specific and distinct mechanistic regulators of tumor growth, angiogenesis and inflammation through RGD-containing epitopes with distinct flanking sequences.