Protein interaction of angiocidin that modulates its anti-tumor activity and the prognostic significance of blood levels of angiocidin in tumor progression

My laboratory in 1993 characterized an anti-tumor protein termed angiocidin. I evaluated the clinical significance of serum angiocidin levels in hepatocellular carcinoma patients. I developed a sensitive capture ELISA to detect angiocidin serum level as low as 10 pg/ml and with a working range of 15--500 pg/ml. I found that angiocidin was expressed in serum but undetectable in healthy controls. HCC patients showed significantly elevated levels of angiocidin ranging from 15.09--195.73 pg/ml. Patients with stages III--IV had statistically higher levels of angiocidin compared to those with stages I--II, p<0.043. Similarly, patients with microsatellite tumor nodules had statistically higher average levels of angiocidin compared to those with no tumor nodules, p<0.032. No statistically significant differences in angiocidin serum levels were found with age, sex, presence of liver cirrhosis, hepatitis B-surface antigen, serum alpha fetoprotein (AFP) levels, tumor size, encapsulation, and venous invasion. Angiocidin was likely secreted by tumor cells since a HCC cell line (Hep3B) secreted an average of 22 ng/10 6 cells after 48 hours in culture. My studies suggest that serum levels of angiocidin may predict worsening stage and intra-hepatic metastasis of HCC. I also report that angiocidin may inhibit angiogenesis by binding collagen and its receptors. Angiocidin bound purified type I collagen and alpha2beta1 with high affinity. alpha2beta1-expressing cells specifically bound and adhered to angiocidin in comparison to wildtype K562 cells. Binding was specific since a neutralizing antibody against alpha2beta1 inhibited binding of alpha2beta1-expressing cells. Angiocidin showed surface binding with alpha2beta1 on K562 alpha2beta1 transfected cells and breast cancer MB-231 cells. In a alpha2beta1-dependent collagen gel angiogenesis assay, angiocidin showed potent inhibitory activity. I also identified a 20 amino acid amino terminal peptide of angiocidin that bound both alpha2beta1 and type I collagen. This peptide promoted alpha2beta1 dependent cell adhesion and inhibited tumor growth. These results are consistent with the conclusion that the anti-tumor activity of angiocidin arises from its ability to ligate collagen and alpha2beta1 on endothelial cell and tumor cells. My results provide support for the concept that targeting matrix-cell interactions is a viable strategy for the development of anti-cancer therapeutics.