‍Why SARS CoV 2 Is Currently Trapped in a Prolonged Metastable State Rather Than Becoming Endemic

Why SARS-CoV-2 Is Currently Trapped in a Prolonged Metastable State Rather Than Becoming Endemic

For several years now, SARS-CoV-2 (SC-2) has continued to evolve despite the remarkable breadth of adaptive immune responses that repeated vaccine-breakthrough infections (VBTIs) have generated in highly Covid-19 (C-19)-vaccinated populations. Yet something equally remarkable has happened during this period: although countless new variants have emerged, none has fundamentally changed the biological behavior of the virus. Instead, we observe a continuous succession of variants carrying modest collections of amino acid substitutions. Several occur in Spike (S), whereas others occur in non-S proteins. Newly emerging lineages repeatedly combine mutations that have already proven successful elsewhere. The overall evolutionary pattern is one of continuous fine-tuning rather than biological innovation.

To many observers this suggests that the virus is gradually approaching endemicity. I would argue, though, that it suggests precisely the opposite!

In my view, this prolonged period of incremental adaptation reflects an increasingly metastable evolutionary state in which the virus continues to evolve because immune selection pressure persists, yet the available adaptive options capable of substantially improving viral fitness are becoming progressively exhausted.

The obvious question is therefore not why SC-2 continues to mutate.

The real question is why evolution continues to favor numerous mutations that individually provide only marginal improvements instead of selecting a fundamentally different adaptive solution capable of restoring efficient viral transmission in a population that has become increasingly hostile to viral spread.

The answer may lie in one of evolution's oldest principles. Natural selection generally prioritizes the safest adaptive route, not necessarily the best one.

Every amino acid substitution represents a relatively inexpensive evolutionary experiment. A single nucleotide change modifies one amino acid while usually preserving the overall architecture of the protein. Most such mutations have little effect, some are harmful, and a few provide a slight competitive advantage. Because they rarely disrupt essential protein function, evolution can continuously sample enormous numbers of these mutations with relatively little risk.

The situation is fundamentally different for glycosylation. Adding, removing or repositioning glycans is much more impactful than a simple additional amino acid substitution. It modifies the three-dimensional organization of the S protein, influences its folding, affects receptor accessibility, protease cleavage, membrane fusion, and interactions with the host glycosylation machinery. Most such changes are therefore expected to be incompatible with efficient viral replication and are likely to be eliminated immediately by purifying selection.

From an evolutionary perspective, amino acid substitutions are therefore cheap whereas glycan remodeling is expensive. This explains why evolution preferentially exhausts the ‘cheap’ adaptive options first, despite their diminishing returns, before eventually being forced into a much riskier but potentially much more rewarding adaptive solution. Evolution consequently continues to exploit every remaining protein-based adaptive opportunity, even when each additional mutation provides only a minute improvement in fitness.

This may explain why today's mutation trackers continue to document predominantly incremental amino acid substitutions and recombinations rather than evidence of a fundamentally new evolutionary strategy. S continues to accumulate convergent substitutions at familiar positions, while non-S proteins acquire additional mutations that may subtly influence replication, innate immune antagonism or viral fitness [1].

None of these changes appears to transform the biological phenotype of the virus. Instead, they represent repeated exploration of an increasingly narrow adaptive landscape.

At first glance, the growing number of mutations in viral proteins other than S appears paradoxical. Many of these proteins play no direct role in receptor binding or membrane fusion. Consequently, their capacity to substantially improve viral entry or transmission is inherently limited. They may modestly optimize replication kinetics, antagonize innate immunity, stabilize viral RNA or improve intracellular efficiency, but none can fundamentally overcome the increasingly broad adaptive immune pressure directed against viral transmission!

Why, then, does evolution continue to select such apparently modest improvements?

Because every small, low-risk gain postpones the need for a far riskier evolutionary innovation. As long as another amino acid substitution can still increase fitness by even a fraction of a percent without compromising protein function, natural selection is expected to favor that safe solution over a high-risk structural innovation that has a substantial probability of destroying viral fitness altogether.

Ironically, this very conservatism may explain the extraordinary duration of the current evolutionary regime.

The virus may remain trapped on what evolutionary biologists call a local fitness peak.

Countless incremental mutations continue to produce diminishing ─ but still positive ─ returns, thereby delaying exploration of a much steeper adaptive pathway whose initial steps carry considerable fitness costs.

Eventually, however, the evolutionary arithmetic changes.

As the cumulative adaptive immune landscape broadens through repeated VBTIs, the selective value of one additional amino acid substitution is expected to become progressively smaller. Conventional protein-based adaptation may eventually reach a point where it can no longer restore efficient viral transmission. At that stage, evolution may increasingly reward adaptive solutions that are currently too risky to compete.

Within the framework of sustained population-level immune pressure, extensive remodeling of the S glycan shield represents the most plausible candidate. Unlike conventional amino acid substitutions, glycan remodeling could simultaneously shield multiple adaptive immune targets, alter receptor accessibility, promote lectin-mediated attachment, facilitate trans infection and trans fusion (cell-to-cell spread), and thereby bypass rather than merely fine-tune the existing vaccine-conditioned adaptive immune landscape.

Such changes would undoubtedly impose substantial structural constraints and would likely require several coordinated mutations before becoming viable. Precisely because they are so difficult to achieve, they are expected to emerge only after conventional adaptive routes have become largely exhausted.

This is why I regard glycan remodeling as the virus's solution of last resort. This is also why it becomes so difficult to predict when exactly a qualitatively different adaptive strategy is going to occur.

Because of the metastable evolutionary situation, the prolonged coexistence of numerous lineages with increasingly modest phenotypic differences should therefore not be interpreted as evidence that SC-2 is settling into endemic equilibrium.

Instead, it may represent the evolutionary signature of a virus that continues to postpone a high-risk but potentially transformative innovation by exploiting every remaining low-risk adaptive opportunity.

The apparent calm does, therefore, not reflect evolutionary stability. It reflects a virus approaching the limits of conventional adaptation.

If SC-2 continues to experience persistent population-level immune selection while conventional amino acid-mediated adaptation yields progressively smaller returns, evolutionary theory suggests that the relative attractiveness of a fundamentally game-changing phase transition will continue to increase. In my view, that is precisely why the current metastable phase deserves far more attention than the comforting narrative that the virus is simply ‘running out of options.’

[1] Proteins such as N, NSP1, NSP3, NSP6, ORF6, ORF8, and ORF9b can influence replication efficiency, innate immune antagonism, RNA production, and viral load. Those effects can indirectly affect transmission even though they do not participate directly in receptor binding or membrane fusion.

‍