In a remarkable feat of scientific ingenuity, vaccines were developed within weeks after SARS-CoV-2 was identified as the etiological agent of COVID-19. This was facilitated by earlier outbreaks with related viruses such as SARS-CoV-1 in 2002-200442,43 or Middle East respiratory syndrome-related coronavirus (MERS-CoV), which appeared in 2012 in Saudi Arabia and has been circulating at low levels ever since44. Experimental vaccines had been developed and tested pre-clinically against these coronaviruses45,47,49 and provided blueprints for the COVID-19 vaccines based on mRNAs or Ad vectors.
RNA-based COVID vaccines have been used most widely followed by Ad vectors. RNA vaccines although initially highly efficacious have several disadvantages. Adverse events such as flu-like symptoms and fatigue are common and serious adverse events such as anaphylactic shock or myocarditis have been reported46,48. The vaccines are relatively costly and not thermostable50. Most importantly antibody titers are not sustained51–53, and booster immunization only improve responses for a short time21 and cause a switch of antibodies to IgG428, an isotype that lacks crucial effector functions30. The main disadvantage of Ad vectors is that on rare occasions they have been linked to VIITT36 and that their immunogenicity is blunted by pre-existing neutralizing antibodies induced by a natural infection or previous immunization with an Ad vector-based vaccine54–56. The advantages of Ad vectors include that they are relatively inexpensive and thermostable57,58. Also results thus far indicate that antibody responses are more sustained than upon mRNA vaccination59,60. This mirrors previous studies using the two platforms expressing a rabies virus antigen61,62 thus suggesting that differences in durability of antibody responses may relate to the vaccine platforms.
Data presented here indicate that there are ways to further improve the magnitude and longevity of Ad vector-induced antibody responses, which may assist to deflect future outbreaks. The J&J vaccine, which is no longer available in the US, is based on a single dose of an HAdV-26 vector that is considered a rare serotype in the US63. It is very similar to Sputnik Light but for the dose, which is two-fold higher for the latter and the design of the insert, which is stabilized in the J&J vaccine64 and left in its wild-type form in Sputnik V or Light65. Efficacy of the two vaccines was also similar19,66 and, as was shown with Sputnik V, could be improved by a heterologous HAdV-5 boost that was given one month after the prime5. This boost is unlikely to be optimal for two reasons – neutralizing antibodies to HAdV-5 are common and may blunt the effectiveness of the recall response and giving the 2nd immunization so soon after the prime is likely to be relatively ineffective according to results obtained with AstraZeneca vaccine. This vaccine is based on an AdC vector, which uses two doses of the same construct expressing the native S protein67; the vaccine had an efficacy of 55.1% if the 2nd dose is given less than 6 weeks after the prime while delaying the second dose to equal or more than 12 weeks increases the efficacy to 81.3%. According to our data vaccination with two distinct AdC vectors given at a high dose in an 8-week interval results in higher and more sustained antibody titers than two immunizations with the same AdC vector given at the same doses and in the same interval. Delaying the boost blunts the negative effects of pre-existing immunity but considering that rapid protection is essential during a pandemic outbreak our data do not support a 4-month interval between an Ad vector prime and boost. Interesting a boost given approximately 6 months after the prime failed to increase VNA titers after the heterologous high or low dose boost and only marginal increased antibody titers measured by ELISA. This contrasts the results obtained by the J&J COVID-19 vaccine that was reported to achieve a 9-fold increase after a boost given at 6 months ((https://www.jnj.com/johnson-johnson-announces-real-world-evidence-and-phase-3-data-confirming-strong-and-long-lasting-protection-of-single-shot-covid-19-vaccine-in-the-u-s). The caveat should be pointed out though that this study compared post-boost titers to those obtained 4 weeks after the prime; antibody titers to the vaccine do not peak as early as week 460 so that this comparison distorts the actual effectiveness of the boost.
The relatively short-lived protective antibody responses to the COVID-19 vaccines in part relates to the rapid accumulation of mutations within the S1 subunit, which is the target of more than 90% of antibodies induced by the mRNA vaccines68. The pre-fusion stabilization within the mRNA’s vaccine insert not only mutates antibody regions within the S2 protein but also affects the dynamics that drive exposure of the fusion peptide within S2 during viral infection, which may explain lack of a more cross-reactive S2 specific VNA response upon mRNA vaccination. In contrast one dose of AZD1222 or two doses of the AdC-S vector described here, which both express a native form of the S protein, induce responses to S1 and S2 that are largely comparable69. Antibodies induced by the mRNA vaccines that express a pre-fusion stabilized S protein show an ~ 40-fold reduction in reactivity against the omicron variant70,71 compared to the Wuhan isolate while those induced by Sputnik V that expresses the wild-type S only show an 8.8–11.6 fold reduction72,73. Our vaccines after a 2nd immunization showed a 3-fold reduction in antibody titers and a very small decline in response rates (82% vs 100%) to omicron. Nevertheless, AZD1222, which also expresses the native form of S, was shown to induce barely detectable antibody responses against omicron74,75 thus not allowing the conclusion that vaccine-induced antibodies to S2 are more suited to cross-neutralize viral variants.
Outbreaks by emerging viruses with pandemic potential are becoming more common and necessitate the development of platforms that allow for rapid deployment of effective vaccines against these new threats. Ad vectors meet the requirements for pandemic vaccines; procedures to develop, perform pre-clinical testing and generate billions of quality-controlled doses in a short time are available. They are effective and safe, stable at ambient temperature, and affordable. Pre-existing VNAs to common human serotypes, which can dampen immune responses to an Ad vector’s transgene product, can be circumvented using rare human serotypes or Ad vectors that originated from non-human primates and by using different serotypes for prime boost regimens. Ad vectors induce potent B and T cell responses and immune responses are sustained. The problem of VIITT that surfaced after millions of Ad vector-based COVID-19 vaccines had been distributed can, if it is indeed triggered by electrostatic binding of platelet factor 4 to the negatively charged hexon of Ad vectors76, most likely be circumvented by changing part of the sequences of the hypervariable loops of hexon77,78.