Human papillomavirus causes an angiogenic switch in keratinocytes which is sufficient to alter endothelial cell behavior

Abstract

One of the requirements for tumor growth is the ability to recruit a blood supply, a process known as angiogenesis. Angiogenesis begins early in the progression of cervical disease from mild to severe dysplasia and on to invasive cancer. We have previously reported that expression of human papillomavirus type 16 E6 and E7 (HPV16 E6E7) proteins in primary foreskin keratinocytes (HFKs) decreases expression of two inhibitors and increases expression of two angiogenic inducers [Toussaint-Smith, E., Donner, D.B., Roman, A., 2004. Expression of human papillomavirus type 16 E6 and E7 oncoproteins in primary foreskin keratinocytes is sufficient to alter the expression of angiogenic factors. Oncogene 23, 2988-2995]. Here we report that HPV-induced early changes in the keratinocyte phenotype are sufficient to alter endothelial cell behavior both in vitro and in vivo. Conditioned media from HPV16 E6E7 expressing HFKs as well as from human cervical keratinocytes containing the intact HPV16 were able to stimulate proliferation and migration of human microvascular endothelial cells. In addition, introduction of the conditioned media into immunocompetent mice using a Matrigel plug model resulted in a clear angiogenic response. These novel data support the hypothesis that HPV proteins contribute not only to the uncontrolled keratinocyte growth seen following HPV infection but also to the angiogenic response needed for tumor formation.

Figure 1

(A) Proliferation of ECs in response to conditioned media from cells expressing HPV 16 proteins. HMVECs were seeded at 5×10⁴ cells/well onto collagen coated 6-well plates. After serum starvation, conditioned media from cells transduced with LXSN or L(16E6E7)SN or from HPV16P7AK or CIN612 cells were added overnight and then replaced with media containing ³H-TdR for five hours. Cells were lysed and incorporated radiolabel counted in a beta-counter. (B) Migration of ECs in response to conditioned media from cells expressing HPV 16 proteins. Serum starved HMVECs were plated on a collagen-coated transwell membrane. Conditioned media were placed in the bottom chamber. After 3 hrs, at least 5 high power fields (HPF) of cells on the bottom of the membrane were counted. For (A,B) the results (average ± standard deviation) are representative of three experiments conducted in triplicate. Statistical analysis combining the results of three independent experiments indicated that, in both the proliferation and the migration assays, the difference between LXSN and 16E6E7 and between C4302 and HPV16P7AK/CIN612 is significant (**p<0.01).

Figure 2

In vivo angiogenesis assays using conditioned media from keratinocytes expressing HPV 16 proteins. Conditioned media were mixed with Matrigel (BD Matrigel™) and injected into the ventral groin of mice. Seven days later the mice were sacrificed, the Matrigel plug was removed and photographed, and then hemoglobin was extracted and quantitated. Conditioned media from C4302 and LXSN serve as negative controls for CIN612 and HPV16P7AK, and 16E6E7, respectively. The results shown are representative of three independent experiments, each conducted in triplicate. The average ± standard deviation for the hemoglobin measured in one such experiment is shown (**p = <0.01).

Figure 3

(A) Semi-quantitative RT-PCR analysis of thrombospondin-1, maspin, VEGF, and IL-8. RNA extracts from HFKs transduced with empty vector (LXSN) or HPV 16 E6E7 or from control human cervical keratinocytes (C4302 cells) or cervical keratinocytes containing the intact HPV 16 (HPV16P7AK) or HPV 31 (CIN612) were subjected to RT-PCR and analyzed on 2% agarose gels. GAPDH amplification was used as an internal control. The multiple bands seen in the VEGF panel represent different isoforms (Kodama et al., 1999). (B) Western analysis of thrombospondin-1 and maspin in conditioned media from cells transduced with LXSN or L(16E6E7)SN or from HPV16P7AK or CIN612 cells. Proteins were separated on 10% SDS-PAGE, transferred to nitrocellulose membrane and probed with antibodies to the indicated proteins. (C) ELISA for VEGF and IL-8 in conditioned media. The quantity of VEGF and IL-8 was determined and normalized to total protein in the conditioned media. The results shown are representative of three independent experiments conducted in triplicate (*p<0.05; **p<0.01).