Although human papillomavirus (HPV) vaccines effectively prevent cervical cancer, the HPV vaccination rates in Japan remain low because of concerns about alleged neurological adverse events. Darja Kanduc proposed a flawed hypothesis that molecular mimicry between HPV and human proteins could induce cross-reactive antibodies, causing autoimmune organ damage, even when only the portions of amino acid (AA)-sequences of the epitopes were identical between HPV and human proteins.
In this study, we conducted the same computational data analysis as Kanduc, using 22 linear epitopes (9–23 AA-length) of the HPV type 16 L1 protein (HPV16L1) registered in the database.
We found that no human epitopes had identical AA-sequences to any HPV16L1 epitopes, demonstrating that HPV16L1 had no molecular mimicry with linear epitopes that have the potential to induce cross-reactive autoantibodies. On the other hand, we identified various numbers of human protein epitopes whose AA-sequences were partially identical with epitopes of HPV16L1, hepatitis B virus (HBV), and respiratory syncytial virus (RSV). We found that HPV16L1 had a smaller number of such proteins having “partial molecular mimicry” than HBV and RSV.
Our current in silico analysis provided no evidence that HPV vaccinations could induce cross-reactive autoantibodies. The flawed molecular mimicry data should not be used as a scientific basis for alleged HPV vaccine-induced adverse events.