Background:  The processes that permitted a few SIV strains to emerge epidemically as HIV-1 and HIV-2 groups are unknown. Paradigmatic theories leave several loose ends unexplained, in particular regarding timing and geographical origin of HIV groups.

Methods:  We performed new molecular clock analyses to time SIV/HIV divergence. We reviewed HIV-1 parenteral and sexual transmissibilities, and the co-factors genital ulcer disease (GUD) and lack of male circumcision. For all relevant geographic areas, we surveyed medical colonial literature, for incidences of GUD and injection campaigns, and primary ethnographic literature, for circumcision data, at the ethnic group level. We built epidemiological simulations of SIV spread in promiscuous cities with various sizes, GUD incidences, and circumcision rates.

Results:  Molecular clock analysis on the newly reported SIV narrows the time intervals (roughly 1880 to 1940) for all transfers to humans, in contrast to human T lymphotropic virus (HTLV). Apart from unsafe injections, other factors seem to be necessary to permit SIV adaptation through serial transmission in this time frame only. While HIV-1 parenteral transmissibility (roughly 1% per reuse) is higher than its common per-act sexual transmissibility, GUD involvement raises the latter to 4 to 32%. We found that, coinciding in time with most recent common ancestor (MRCA) of all HIV epidemic groups, intense GUD outbreaks (e.g., with ulcerative syphilis incidences 100 to 300 times higher than in mid-20^th century) occurred at the colonial cities. Moreover, among 12 relevant colonies, only in 3 (Cameroon, Belgian Congo, and Côte d’Ivoire) circumcision rate in major cities was low (50 to -80% vs nearly 100% in the others) and HIV groups emerged only in them. Our simulations suggest that the odds of SIV initial spread were much higher in these cities at their intense GUD period (1920 to 1945), than with other conditions (rarified GUD, universal circumcision, or small towns).

Conclusions:  The timing of HIV emergence fits better GUD-intensive periods in nascent cities, than the periods of higher injection, and urbanization intensities. Early HIV epicenters arose only in cities that had relatively low circumcision rates. These results suggest that intense GUD, and non-circumcision, in promiscuous urban centers together contributed to HIV emergence and adaptation. Our results imply a zoonotic danger posed by GUD in Africa, and recommend its monitoring, particularly in areas where male circumcision is less common.