Low seroprevalence of neutralizing antibodies to gorilla adenovirus 32 (GRAd32) in southern African populations supports evaluation of this vector platform for HIV vaccine development

Adenoviruses (Ads) are widely used as vaccine vectors. However, pre-existing immunity to commonly used serotypes, like Ad5, can reduce vaccine immunogenicity, with neutralizing antibody titers >200 previously shown to impact vaccine efficacy. Gorilla adenovirus (GRAd) vectors have been developed to evade pre-existing anti-vector responses, but their seroprevalence in southern Africa is poorly defined. Here, we assessed seroprevalence to GRAd32, Ad26 and Ad5 before (baseline) and after COVID-19 vaccination, in cohorts from South Africa and Zimbabwe.

Sera from South African participants enrolled in the Sisonke sub-study (n=100, prior to Ad26.COV2.S vaccination) and the follow-up “Booster after Sisonke” (BaSiS) study (n=226) were tested for neutralizing antibodies to Ad5, Ad26, and GRAd32. These samples included paired pre/post boost samples for 27 donors. We also tested sera from the Zimbabwean Mutala cohort (n=131, of which 44 were unvaccinated, and 87 vaccinated with inactivated vaccines). Participants living with HIV (PLWH) comprised 30-50% of each cohort.

In the pre-vaccination samples from the Sisonke cohort, geometric mean titers (GMT) for anti-GRAd32, Ad26, and Ad5 antibodies were 78, 142, and 459, with neutralization titers >200 observed in 14%, 32%, and 68% of participants, respectively. Similarly, in the unvaccinated participants in the Mutala cohort, GMTs for GRAd32, Ad26, and Ad5 were 117, 245, and 536, with neutralizing titers >200 in 22%, 42%, and 69% of participants. We observed no significant difference in Ad antibody titers between PLWH and those living without HIV. We next assessed the impact of COVID-19 vaccination on titers. Vaccination with inactivated COVID-19 vaccines (Sinopharm/Sinovac) did not significantly affect Ad5, Ad26 or GRAd32 titers in an unpaired analysis. In contrast, ∼9 months after Ad26.COV2.S vaccination, anti-Ad26 titers for longitudinally sampled participants (n=27) increased 10-fold from a GMT of 141 to 1,426. By comparison, GRAd32 responses were not significantly altered by Ad26.COV2.S vaccination, while anti-Ad5 responses showed a modest <2-fold increase.

Our data support previous findings that, whereas anti-Ad5 neutralizing antibody responses are commonly detected globally, GRAd32 responses are less frequent. Importantly, GRAd32 neutralizing responses remained unchanged after Ad26.COV2.S vaccination. HIV status had no significant effect on antibody titers. These results support the use of the GRAd32 vector in upcoming HIV vaccine trials, including in regions where Ad26-based COVID-19 vaccines were widely deployed.