Continuous evolution of Omicron has led to numerous subvariants that exhibit growth advantage over BA.5. Such rapid and simultaneous emergence of variants with enormous advantages is unprecedented.
Despite their rapidly divergent evolutionary courses, mutations on their receptor-binding domain (RBD) converge on several hotspots, including R346, R356, K444, L452, N460K and F486.
The driving force and destination of such convergent evolution and its impact on humoral immunity established by vaccination and infection remain unclear.
The researchers initially thought that the BA.2.3.20 and BQ.1.1 variants would only gain dominance in circulation and become a key player in the late fall and winter waves, but unexpectedly both are now rapidly replacing BA.5, BA. 4.6 and BA .2.75 variants.
In the United Kingdom, in the last 24 hours, it was estimated that there were about more than 170000 COVID-19 infections, bringing a total of approximately 1.95 million people with symptomatic infections in the last 7 days.

https://health-study.joinzoe.com/data

Despite cover-up attempts by the British government, media reports last week indicated that hospitalization rates are rising for the first time since July 2022.
https://www.theguardian.com/world/2022/sep/22/covid-hospitalisations-rise-by-nearly-20-in-a-week-in-england
https://www.hsj.co.uk/coronavirus/fourth-covid-wave-of-2022-picks-up-speed/7033202.article
It is the first time since late July that an increase had been seen in England. There was also a rise in Wales, although infection levels have dropped slightly in Northern Ireland and Scotland in the most recent week, after the latter showed a rise the week before.
An increase in cases has also been seen in UK data collected by the Zoe health study, while the latest NHS figures show a 17% increase in the number of Covid patients admitted to hospital in England – from 3,434 in the week ending 12 September to 4,015 in the week ending 19 September – with larger percentage rises in some regions.
As Dr Thomas Peacock, of Imperial College London, points out, recent data suggest the latter two each account for less than 0.5% of Covid genetic sequences in the UK – but they are growing fast. “It’s entirely possible an autumn/winter wave is driven by a mixture of variants,” Peacock said.
Prof Tom Wenseleers, an evolutionary biologist at the Catholic University of Leuven in Belgium, said BA.2.75.2 and BQ1.1 have mutations in their spike protein that help them to partly escape from BA.5-induced immunity.
“Combined with the fact that Covid hospitalisations have already started rising again in the UK, and that the full effect of these variants still isn’t felt, I would say this is not such great news,” he said.
Wenseleers said: “Most scientists believe that our high population immunity will cause the infection fatality rate to keep on declining. But any new infection wave will of course add to the toll of the pandemic.”
But deaths are not the only concern. Peacock said: “Even a small wave is going to put massive additional strain on the health service, particularly if paired with other respiratory viruses making a comeback this winter,” such as flu.
But let’s try to understand what is happening …
SARS-CoV-2 Omicron BA.1, BA.2, and BA.5 have demonstrated strong neutralization evasion capability, posing severe challenges to the efficacy of existing humoral immunity established through vaccination and infection 1-15. Nevertheless, Omicron is continuously evolving, leading to various new subvariants, including BA.2.75 and BA.4.6 16-22. Importantly, a high proportion of these emerging variants display high growth advantages over BA.5, such as BA.2.3.20, BA.2.75.2, BR.1, BN.1, BJ.1, and BQ.1.1 (Fig. 1a) 23.
Such rapid and simultaneous emergence of multiple variants with enormous growth advantages is unprecedented. Notably, although these derivative subvariants appear to diverge along the evolutionary course, the mutations they carry on the receptor-binding domain (RBD) converge on the same sites, including R346, K444, V445, G446, N450, L452, N460, F486, F490, and R493 (Fig. 1b, Extended Data Fig. 1).
Most mutations on these residues are known to be antibody-evasive, as revealed by deep mutational scanning (DMS) 1,2,24-26. It’s crucial to examine the impact of these convergent mutations on antibody-escaping capability, receptor binding affinity, and the efficacy of vaccines and antibody therapeutics.
It’s also important to investigate the driving force behind this accelerated emergence of RBD mutations, what such mutational convergence would lead to, and how we can prepare for such convergent RBD evolution.

Convergent evolution of Omicron RBD with growth advantage over BA.5.
a, Whole-genome maximum likelihood phylogenetic analysis of Omicron subvariants. Variants with a growth advantage over the original BA.5 are colored. Relative growth advantage values are calculated using the CoV-Spectrum website. b, Key RBD mutations in emerging SARS-CoV-2 BA.5 and BA.2.75 subvariants exhibit convergent patterns.
Convergent RBD evolution causes extensive immune evasion
First, we tested the antibody evasion capability of these convergent variants. We constructed the VSV-based spike-pseudotyped virus of Omicron BA.2, BA.2.75, and BA.4/5 sublineages carrying those convergent mutations and examined the neutralizing activities of therapeutic neutralizing antibodies (NAbs) against them (Fig. 2a and Extended Data Fig. 2a) 13,27-32.
The COV2-2196+COV2-2130 (Evusheld) 28,33 is vulnerable to F486, R346, and K444 mutations, evaded or highly impaired by BA.2.38.1 (K444N), BJ.1 (R346T), BR.1 (L452R+K444M), BA.5.2.7 (K444M), BA.5.6.2 (K444T), and BQ.1.1 (R346T+K444T). LY-CoV1404 (Bebtelovimab) remains potent against BF.16 (K444R) and BA.5.1.12 (V445A) but shows reduced potency against BA.5.5.1 (N450D) 31.
Importantly, LY-CoV1404 was escaped by BJ.1, BR.1, and BQ.1.1 while exhibiting strongly reduced activity against BA.2.38.1, BA.5.2.7, and BA.5.6.2 due to K444 N/M/T mutations 31. SA55+SA58 is a pair of broad NAbs isolated from vaccinated SARS convalescents that target non-competing conserved epitopes 2,32. SA58 is weak to G339H and R346T mutations and showed reduced neutralization efficacy against BJ.1 and BA.2.75 sublineages. SA55 is the only NAb demonstrating high potency against all tested Omicron subvariants.
Among the tested variants, BQ.1.1 exhibited the strongest resistance to therapeutic mAbs and cocktails (Fig. 2a). Since the SA55+SA58 cocktail is still in preclinical development, the efficacy of available antibody drugs, including the BA.2.75/BA.5-effective Evusheld and Bebtelovimab, are extensively affected by the emerging subvariants with convergent mutations.

Emerging Omicron subvariants further evade neutralizing antibodies.
a, IC50 of therapeutic NAbs against VSV-based pseudoviruses with spike glycoproteins of emerging SARS-CoV-2 BA.2/BA.5/BA.2.75 subvariants. green, IC50 ≤ 100ng/mL; white, 100ng/mL < IC50 < 1,000ng/mL; red, IC50 ≥ 1,000ng/mL; *, IC50 ≥ 10,000ng/mL. b, Relative hACE2-binding affinity measured by IC50 of hACE2 against pseudoviruses of variants. Error bars indicate mean±s.d. P-values were calculated using two-tailed Wilcoxon’s rank-sum test. *, p < 0.05; **, p < 0.01; ***, p < 0.001. No label on variants with p > 0.05. Variants with significantly higher affinity are colored blue, while those with lower affinity are colored red. c-f, Pseudovirus-neutralizing titers against SARS-CoV-2 D614G and Omicron subvariants of plasma from vaccinated individuals or convalescents of breakthrough infection. c, Individuals who had received 3 doses of CoronaVac (n = 40). d, Convalescents who had been infected with BA.1 after receiving 3 doses of CoronaVac (n = 50). e, Convalescents who had been infected with BA.2 after receiving 3 doses of CoronaVac (n = 39). f, Convalescents who had been infected with BA.5 after receiving 3 doses of CoronaVac (n = 10). The geometric mean titers are labeled. Statistical tests are performed using two-tailed Wilcoxon signed-rank tests of paired samples. *, p < 0.05; **, p < 0.01; ***, p < 0.001; NS, not significant, p > 0.05.
Sufficient ACE2-binding affinity is essential for SARS-CoV-2 variants to gain advantages in transmission. Thus, we examined the relative hACE2 binding affinity of these variants by evaluating hACE2 neutralizing potency against the pseudoviruses. The higher neutralizing efficacy of soluble hACE2 indicates a higher ACE2-binding affinity 18. Overall, these convergent variants all demonstrate sufficient ACE2-binding affinity, at least higher than D614G, allowing their prevalence, including BA.2.75.2 and BQ.1.1 (Fig. 2b and Extended Data Fig. 2b). Specifically, R493Q reversion increases the binding affinity to hACE2, which is consistent with previous reports 4,18,20. K417T shows a moderate increase in binding affinity to hACE2. In contrast, F486S, K444M, and K444N have a negative impact on binding affinity, while K444T does not cause significant impairment of ACE2 binding (Fig. 2b). These observations are also in line with previous DMS results 34.
Most importantly, we investigated how these variants escape the neutralization of plasma samples from individuals with various immune histories. We recruited cohorts of individuals who received 3 doses of CoronaVac 35 with or without breakthrough infection by BA.1, BA.2, or BA.5. Convalescent plasma samples were collected around 4 weeks after hospital discharge (Supplementary Table 1). Plasma from CoronaVac vaccinees was obtained two weeks after the third dose.
There was a significant reduction in the NT50 against most tested BA.2, BA.2.75, or BA.5 subvariants, compared to that against corresponding ancestral BA.2, BA.2.75, or BA.5, respectively (Fig. 2c-f and Extended Data Fig. 2c-f). BJ.1, BA.2.10.4, and BA.2.3.20 are highly immune evasive. Plasma of vaccinees and BA.1/BA.2 convalescents neutralize these three subvariants with similar NT50 (Fig. 2c-e). However, BA.2.10.4 exhibited the strongest evasion to BA.5 convalescent plasma (Fig. 2d). Considering the relatively high ACE2-binding affinity of BA.2.10.4 and BA.2.3.20 due to R493Q reversion, they may tolerate even more escape mutations (Fig. 2b).
Newly emerging BA.2.75 and BA.5 subvariants also demonstrated strong antibody-evading capability. BA.5 subvariants with R346T, K444T/R, and V445A exhibit a similar reduction in NT50, while N450D is slightly less evasive than others. BA.2.75 subvariants with L452R, R346T, K356T, K444M, and F490S exhibit further immune evasion as well. Importantly, BA.2.75.2 (BA.2.75+R346T+F486S) and BQ.1.1 (BA.5+R346T+K444T+N460K) are the most humoral immune evasive strain tested. BA.2.75.2 causes a reduction in NT50 of plasma from 3-dose vaccinees, BA.1 breakthrough infection, BA.2 breakthrough infection, and BA.5 breakthrough infection by 4.2, 5.9, 6.2, and 2.7 folds compared to BA.2.75, respectively. BQ.1.1 reduces the NT50 of the plasma from the four cohorts by 3.0, 3.9, 4.4 and 6.4 folds compared to BA.5, respectively.
NT50 of BA.5 convalescents against BQ.1.1 is higher than that against BA.2.75.2, which is probably contributed by the strong capability of evading NTD-targeting antibodies of BA.2.75 subvariants (Fig. 2f). We also found that in the three breakthrough infection cohorts, NT50 against D614G and previous VOCs (Alpha, Beta, Gamma and Delta) are higher than that against Omicron, even the corresponding infected strain, which indicates significant immune imprinting (Extended Data Fig. 2d-f).
These observations demonstrate that convergent RBD evolution could cause significant immune escape at a scale never seen before. Such rapid and convergent emergence of antibody-escaping variants, especially when they can evade convalescent plasma from BA.5 breakthrough infection, suggests that vaccine boosters designed based on BA.5 may not achieve broad-spectrum protection against infection.
reference link : https://www.biorxiv.org/content/10.1101/2022.09.15.507787v2.full