Do mRNA-based COVID Vaccines Have an Achilles Heel?

By Kevin E. Noonan --

The COVID-19 pandemic has spread throughout the globe, infecting more than 90 million people and causing almost two million deaths (see "Tracking coronavirus' global spread").  SARS-CoV-2 infection is the cause of the COVID-19 pandemic; this virus is recognized as the latest viral infection in humans of zoonotic origins, in this case bats first arising in Wuhan, China.

But the world's hope against the pandemic has been raised by the development of effective vaccines against virus infection in record time, using technology based on the mRNA encoding a particular viral protein called Spike.  The rapidity of vaccine development is one of the advantages of using modern RNA technology, and this is a direct consequence of two aspects of this technology.  First, RNA sequencing technology has permitted quick determination of the genetic sequence of the viral RNA.  In particular, this technology identified the sequence of the viral Spike protein, which is major antigenic determinant and required for virus entry into infected cells, attaching to the angiotensin 1 converting enzyme 2 (ACE-2) receptor.  Second, because this protein is necessary for infectivity, immunological responses directed at the encoded protein were believed to be (and turned out to be) particularly effective in provoking immunity to infection.  And because just this specific protein can be produced in the laboratory (and then in industrial quantities) for producing the vaccine used to raise immunity, many of the deleterious effects, both biological and temporal, of using "whole" viruses can be avoided.  It is a remarkable demonstration of applying technology developed over the past 40 years to solve a critical problem in global health.  And it would have been almost impossible to have achieved these vaccines "the old fashioned way," i.e., using virus samples and either genetically attenuating them or inactivating them with heat, chemicals, and like methods.

But the capacity of the virus to mutate and change portions of the sequence of the Spike protein has raised concerns that the current vaccines, which were all raised with specific Spike protein-encoded RNAs, might be ineffective for such variants.  This potential deficiency raises the possibility if not the likelihood that vaccination against the known ("original") Spike protein might in fact select for the variants, which could spread even in an inoculated population.  The only way to forestall if not prevent such an outcome would be a massive inoculation effort that could prevent sufficient replication of the virus in naïve members of the population to not give the virus replication and mutation opportunities to outrun inoculation efforts.

But such a strategy is impractical in a global epidemic, if only due to the logistics of inoculating a large fraction of the world's population when transmission across national boundaries, oceans, and continents is facilitated by global travel.  And the most draconian lockdown, quarantine, and travel ban regimes are unlikely to be effective unless aided by population and geography, for example, in New Zealand which is almost literally at the end of the world.

The variant problem is one that arises precisely because of the way the vaccine has been made.  Viruses like all living things are genetically heterogenous, and the SARS-CoV-2 virus "in the wild" is likely comprised of thousands if not millions of slightly different genetic sequences, some of which are found in the Spike protein.  In addition, the biologically active portion of the Spike protein, i.e., the portion that interacts with the ACE-2 receptor and facilitates virus entry into infected cells, is not necessarily the most antigenic portion of the protein, albeit because the vaccines are "neutralizing" antibodies (which means just what it sounds like) antibodies produced by the vaccine must interfere with such binding.  When vaccines are prepared from populations of viruses they can be expected to comprise a somewhat representative population of these slightly variable proteins; for SARS-CoV-2 these would be not only the Spike protein but any protein antigenically available to the immune system.

As a result, any vaccines produced using mRNA technology cannot have the plurality of viral proteins and the populations of variant viral proteins that are provided by more conventional immunological methods.  The immunity thereby provided is specific but its very specificity renders the immunological protection narrow compared with more traditional vaccines.  This narrowness provides the virus with an opportunity for immune avoidance and thus infection even in individuals immunized against the current Spike protein.  And paradoxically, the vaccine may in fact select (in the Darwin, "survival of the fittest" meaning) for such variants, which can propagate through the human population in spite of even robust vaccination efforts.

Fortunately, there may be two silver linings in this viral cloud.  First, the very rapidness of the production of these first mRNA vaccines may permit equally (or perhaps even faster) development of mRNA vaccines specific for these variants, and the provision of immunological "cocktails" of vaccines directed against the most prevalent or infectious variants.  Second, many variants may not be capable of immune evasion, so despite the genetic changes created during infection some portion of the new variants will not be capable of infecting vaccinated individuals.  Third, there is a balance between such variants and their effectiveness in infection that could select for less infectious or perhaps less deleterious strains of the virus; the thought in classical immunology was that pathogens that were limited to human infections over time attenuated because the more virulent strains burned themselves out while the milder forms permitted their hosts to live long enough to infect more, other individuals.  This is one explanation for why influenza infection is constant, because those viruses can be passed through fowl (ducks and chickens) and swine (pigs and hogs) as well as humans, and thus selection in these animal populations can produce virus populations more virulent in people.

There may be a grandeur in this biological view of life, but when it comes to pandemic infection a perhaps more apt meme is that nature truly is red in tooth and claw (and Spike proteins), and pathological infections are almost impossible to avoid in the long run.  In the short term, however, the hope is that they can be tamped down enough to put out the fire this time, and for us to learn the lesson of the importance of eternal vigilance and preparedness against the next potential pandemic.

Comments

[Greg E.]

This article seems to imply that the target mRNA is exogenously translated into protein, and that resulting protein is what is injected as the vaccine antigen. However, mRNA vaccines work by injecting the mRNA itself as the antigen. Then the injected mRNA is then translated by the host's immune cells to create the protein antigen in vivo, which results in the immune response.

[Derek Freyberg]

The article seems to suggest that this is a problem peculiar to mRNA vaccines, and yet we know that this is not so. A killed virus vaccine will only have the antigens of the virus population from which it is taken - so new variants are no more addressed by a killed virus vaccine than by an mRNA vaccine. Would a vaccine be better with more than one antigen? - seems reasonable, but doesn't work for the flu vaccine.

[Kevin E Noonan]

Derek: Flu is a special case, for at least two reasons. One I allude to in the post: the frequent passage of the virus through animals (swine flu, for example), as opposed to viruses like SARS-CoV-2, which is a "jump" from the species it is adapted to (and in which it is not particularly lethal) to man. The other reason is that because the influenza virus genome is segmented so there is the possibility for reassortment of the viral segments to provide another level of variability.

The problem with mRNA vaccines is that when the mRNA is synthesized every molecule encodes exactly the same protein, which does not occur in a population of viruses any more than it occurs in a human population. I'm not saying the vaccine is not remarkable for how quickly it was developed and how effective is has been. But it is a good thing to recognize its limitations.

Thanks for the comment

[Kevin E Noonan]

Greg: sorry you got that impression; I thought it was evident that an mRNA virus was made in immune cells.

But the point is that the mRNA encodes one protein that is invariant in then encoded amino acid sequence. Thus, naturally occurring variants can arise that lack the immunological effectiveness that the vaccine has against SARS-Cov-2 virus having the same Spike protein from which the vaccine was made.

Thanks for the comment.

[Ken Dow]

Nice article Kevin. One point worth mentioning is that the antibodies produced by the body upon injection of the mRNA and the resultant expression of the protein are polyclonal in nature and bind to different regions of the spike protein. So even if the virus mutates the protein, presumably we would have antibodies that might be effective against the mutated virus. So in order for the virus to mutate in a way to avoid the antibodies it has to find the right mutation that avoids antibodies but still retains the ability to bind to the receptor. While natural selection is pretty good at doing that, it's still a needle in a haystack. What do you think?

[Kevin Noonan]

Dear Ken:

Agreed, the protein mutates at random and so it becomes combinatorial pretty quickly. But because RNA viruses lack access to most of the sequence correction mechanisms in the cell that DNA viruses benefit from there is a lot of variant production (one reason there is no HIV vaccine for example). And then we get to the economies of scale: estimates of there being 10 to the 23rd power (more than here are stars in the universe; https://www.airspacemag.com/daily-planet/there-are-more-viruses-earth-there-are-stars-universe-180974433/) viruses worldwide, and all you need it one to infect someone and start spread of the resistant variant. What makes viruses so successful.

Hope you are well.