The mRNA vaccines we are currently using are not sufficient to stop the ongoing Covid-19 pandemic. Although they have been critical in the fight against SARS-CoV-2, they suffer from a glaring limitation: poor longevity. Wait long enough between booster shots, and you’re back to square one — once again susceptible to infection, with variable protection against serious disease, hospitalization, and death.
Even against the same strain of the virus, immunity wanes. Factor in viral mutation and the constant rise of new variants, and this trend is supercharged. Just within the Omicron family of variants, for example, many of the newer offshoots manage to evade immunity built up against their parent lineages (Figure 1). Generally, this happens through structural changes to the spike (S) protein that the virus uses to bind and enter host cells. If these changes are significant enough, our immune system —and in particular antibodies— simply can’t recognize the virus.
Can we do better? Yes.
Slow and Steady Wins the Race
Adenovirus-vector vaccines, despite lower initial antibody titers, manage to retain their efficacy for a longer period of time than do mRNA vaccines (Figure 2). They may also elicit stronger CD8+ T cell responses, which are a crucial part of the adaptive immune response and a key player in protecting us during reinfection.
Are the adenovirus vector vaccines better than the mRNA vaccines? Epidemiological evidence suggests no. But they do hint at a potential way of strengthening immune memory: prolonging antigen availability.
This is an area in which mRNA vaccines fall short. Despite seeding messenger RNA with synthetic nucleotides that stave off degradation, the mRNA lasts only one or two days following injection. By extension, the antigen proteins that the mRNA encodes are also of short duration, persisting no longer than three days. By contrast, adenovirus vector vaccines can continue to express the target antigen for a much longer time. This may be one of the reasons they induce longer-lasting immunity.
A similar phenomenon can be observed with influenza, where data indicate the antigen must be present for at least seven to eight days for a robust memory response to develop. This is evidenced by the fact that germinal center activity doesn’t begin until around this time (Figure 3). Germinal centers act as a kind of boot camp for B cells, training them to recognize diverse regions of an antigen and to bind the antigen more tightly. B cells, in turn, secrete antibodies that bind to the virus, making it harder for the virus to bind host cells and, if present in large enough quantities, inhibiting viral entry wholesale.
Another series of studies showed that primates immunized according to a slow-delivery strategy —one in which vaccine administration is spread out across multiple days— developed enhanced neutralizing antibody responses against human immunodeficiency virus (HIV) compared to primates following a traditional, single-dose immunization protocol —where the entire vaccine dose is administered in one sitting. Simply extending antigen availability boosted antibody production 10-fold.
The same group of scientists has now published a new study investigating this process in more detail. Their work shows that a slow-delivery approach, with dose-escalation over the span of 12 days, can lead to broader, higher quality, and longer-lived immune memory. In particular, Lee et al. discovered that germinal centers stay active for far longer and produce a more diverse set of memory B cells when vaccine administration is extended over the span of a week or more.
Here, we take a closer look at their findings and discuss the implications for future vaccine design.
Slow-Delivery Vaccination Extends Germinal Center Activation
To compare the impact of different vaccination strategies on germinal center activation and durability, Lee et al. separated rhesus monkeys into three groups. The first group received two average-sized doses of recombinant HIV envelope trimer MD39 protein, one in each arm. This was combined with a common adjuvant, called alum, to mimic conventional vaccination. The other two groups (groups 2 and 3) received the same recombinant protein dose, but formulated with an immune-stimulating adjuvant saponin/MPLA nanoparticle (SMNP). Instead of injecting the entire dose in one sitting, they split it up over a span of seven smaller doses, injected every other day for a 12-day period. As before, injections were administered in both arms.
Monkeys immunized according to the traditional, single-dose approach showed a spike in total germinal center B cells three weeks post-inoculation. Only a small portion of these B cells effectively bound the envelope protein, but from week three onwards, their numbers began to fall quickly.
Monkeys vaccinated according to the dose-escalation protocol —groups 2 and 3— displayed a steeper spike in total germinal center B cells than their counterparts. Of those, the frequency of envelope-specific B cells was also significantly greater, with a 7.8 fold increase over group 1. And most strikingly, the number of germinal center B cells continued to rise for many weeks. By week 10, the gap had increased to a 186-fold difference.
Groups 1 and 2 received an additional booster dose 10 weeks after initial vaccination. Group 3 monkeys, on the other hand, did not receive an additional booster until 29 weeks after initial vaccination, giving the researchers a chance to study the impact of slow-delivery over a longer interval.
Indeed, germinal center activity in group 3 monkeys was recorded for a total of 191 days, a full 27 weeks after the end of the initial priming period. Even at 191 days, the frequency of envelope-specific germinal center B cells was 27-fold higher than the peak frequency following conventional, bolus vaccination.
Continued Germinal Center Activity Produces Better B Cells
Extended germinal center activity overlapped with continued B cell evolution throughout.
To retain efficacy in the face of viral mutation, B cells need to be able to recognize a broad variety of binding sites. They also need to bind to these sites tightly. Both diversity and affinity are improved through a process called somatic hypermutation (SHM). In a nutshell, B cells undergo rapid mutation and the resultant lineages are put into competition with one another. Those that best bind the antigen survive, the others are got rid of.
In group-3 monkeys, germinal center B cell lineages that were initially unable to bind the HIV envelope trimer protein managed to do so several weeks later, indicating continued evolution. In some cell lines, affinity improved 1000-fold over a six month period.
Similar somatic hypermutation was witnessed in memory B cells, improving both affinity as well as diversity.
What About Antibodies?
Germinal center activity is an important indicator of B cell maturation, but antibodies are the ultimate metric of vaccine efficacy. Having established that slow-delivery supercharges germinal centers, the researchers next analyzed how this prolonged activation shapes the antibody response.
Only one monkey in the conventional vaccine group developed neutralizing antibodies towards the HIV envelope trimer protein, and this only after receiving the booster dose. Even then, titers were low.
All of the monkeys in the slow-delivery vaccine groups developed neutralizing antibodies. Crucially, all of the monkeys in group 3 developed these antibodies after only primary immunization, before they even received a booster dose. The number of antibodies remained stable from week three all the way to week 29. Following administration of a booster dose, peak titers were 50 times as high as those in group 1.
Clearly the slow-prime approach boosts antibody production, but what about antibody quality? Antibodies are only useful if they can properly bind to the circulating antigen, after all.
A key feature of antibody efficacy is antibody diversity; a single antigen can have multiple binding sites, or “epitopes”, and the more diverse the antibodies the better the chances of neutralizing the antigen. This is especially true given a phenomenon called “immunodominance”, which is the tendency for antibodies to target only certain binding sites, those deemed most important. This lets the virus escape the antibody response by mutating in just one area, something it cannot do as easily if it is targeted at multiple sites.
Again, the slow-delivery approach proved superior: monkeys in group 2 and group 3 developed a more diverse pool of antibodies compared to those in group 1, targeting a larger array of antigen binding sites (Figure 4). Monkeys in group 3 displayed the greatest breadth of antibodies.
Vaccine development is a complex field with lots of moving parts; the seemingly smallest of differences —are you injecting into the muscle or into the skin?— can potentially change how well the final product protects people from disease. There are some clear factors that impact immune response: the size of the dose, what part of the virus a vaccine targets, what platform is used to present the viral antigen —inactivated or live-attenuated or mRNA or any of the other myriad options— and whether anything is added to stimulate the immune system, called an adjuvant. The work by Lee et al. adds to a growing list of research suggesting that speed of delivery and duration of antigen availability are another fundamental part of the equation.
Their work also has important implications for our current vaccines against Covid-19. Practically speaking, few people are going to want to sit through seven “micro” vaccinations spread across 12 days. Given the clear benefits, this is an unfortunate limitation. However, there are several technologies that may be able to mimic this extended delivery schedule by prolonging antigen expression.
Adenovirus vector-based vaccines, as mentioned above, are one such technology. But they come with a few drawbacks. Since adenoviruses are common, cold-causing viruses, many people have already been exposed to them during their lifetime. This means they have built up immune memory towards the viruses, making it difficult for the viral vector to ferry the target gene into our cells — the vector is quickly recognized, targeted, and cleared as soon as it enters the body. Second, there exists the serious, albeit rare, risk of blood-clotting, called vaccine-induced thrombocytopenia (VIT). This can happen when the vaccine is accidentally injected into the bloodstream, rather than muscle tissue.
Self-amplifying RNA vaccines represent an alternative option. Where conventional mRNA vaccines only include the genetic information for the antigen protein, self-amplifying RNA vaccines include an additional sequence of RNA that encodes four non-structural proteins. These nonstructural proteins come together once in the cell to form an enzyme called RNA replicase — think of this as a genetic photocopier. The RNA replicase makes many copies of the target mRNA, each of which can then go on to be translated into the antigen. This massively extends the time of antigen expression, potentially up to 26 days. Some manufacturers have already begun incorporating self-amplifying RNA into their Covid-19 vaccine design.
Although not a perfect stand-in for the dose-escalation strategy explored by Lee and colleagues, these technologies harness the benefits of extended antigen expression while maintaining real-world practicality.