Breast cancer, viruses, and human leukocyte antigen (HLA).

Several viruses have been implicated in breast cancer, including human herpes virus 4 (HHV4), human herpes virus 5 (HHV5), human papilloma virus (HPV), human JC polyoma virus (JCV), human endogenous retrovirus group K (HERVK), bovine leukemia virus (BLV) and mouse mammary tumor virus (MMTV). Human leukocyte antigen (HLA) is involved in virus elimination and has been shown to influence breast cancer protection/susceptibility. Here we investigated the hypothesis that the contribution of a virus to development of breast cancer would depend on the presence of the virus, which, in turn, would be inversely related to the success of its elimination. For that purpose, we estimated in silico predicted binding affinities (PBA) of proteins of the 7 viruses above to 127 common HLA alleles (69 Class I [HLA-I] and 58 Class II HLA-II]) and investigated the association of these binding affinities to the breast cancer-HLA (BC-HLA) immunogenetic profile of the same alleles. Using hierarchical tree clustering, we found that, for HLA-I, viruses BLV, JCV and MMTV were grouped with the BC-HLA, whereas, for HLA-II, viruses BLV, HERVK, HPV, JCV, and MMTV were grouped with BC-HLA. Finally, for both HLA classes, the average PBAs of the viruses grouped with the BC-HLA profile were significantly lower than those of the other, non BC-HLA associated viruses. Assuming that low PBAs are likely associated with slower viral elimination, these findings support the hypothesis that a defective/slower elimination and, hence, longer persistence and inefficient/delayed production of antibodies against them underlies the observed association of the low-PBA group with breast cancer.