SARS-CoV-2 subvariants BQ.1.1 and XBB.1 have been circulating globally with superior growth advantages over most omicron mutants (appendix p 5). However, XBB.1.5, a subvariant of the recombinant mutant XBB, has shown a substantial growth advantage compared with BQ.1.1 and XBB.1. Because of its enhanced transmissibility, XBB.1.5 has rapidly become the dominant SARS-CoV-2 strain in the USA and is highly likely to cause the next global wave of COVID-19 (appendix p 5). 1 XBB and XBB.1 has already been shown to be extremely evasive against the neutralisation of plasma and serum from vaccinated or convalescent individuals and monoclonal antibodies (mAbs), with a greater evasive ability than the BQ.1.1 variant.2, 3, 4, 5 Compared with XBB.1, XBB.1.5 carries a Ser486Pro mutation on the spike protein, a rare two nucleotide substitution compared with the ancestral strain (appendix p 5). The mechanism behind the rapid transmission of XBB.1.5, especially the effect of Ser486Pro, requires immediate investigation. We used vesicular stomatitis virus-based pseudovirus neutralisation assays to evaluate the neutralisation titres against XBB.1.5 of convalescent plasma from individuals who had received three doses of CoronaVac (Sinovac) before BA.1 (n=50), BA.5 (n=36), or BF.7 (n=30)breakthrough infection. A cohort of patients with convalescence from BA.5 breakthrough infection who had received at least two doses of BNT162b2 (Pfizer-BioNtech) or mRNA-1273 (Moderna) is also included in the analysis (n=10). Human ACE2 (hACE2)-binding affinity of XBB.1.5 receptor-binding domain was compared with that of XBB.1, BQ.1.1, and BA.2.75 using surface plasmon resonance. Plasma samples associated with CoronaVac were collected on average 27 days (SD 8) after hospital discharge (appendix pp 7–8). Plasma samples associated with the mRNA vaccine were collected within 2–3 weeks after hospital admission (appendix pp 7–8). The absence of BQ.1.1 breakthrough infection in individuals who were convalescent is a limitation of the ability of this study to estimate the scale of immune evasion of XBB.1.5 for this group. Plasma samples from individuals who had received three doses of CoronaVac and had a BA.1, BA.5, or BF.7 breakthrough infection showed a substantial decrease in plasma 50% neutralisation titre (NT50) against XBB.1 and XBB.1.5 compared with that against B.1 (ASP614Gly) variant (figure A ). Plasma from patients who received CoronaVac and had a BA.5 breakthrough infection showed a 44-times lower NT50 against XBB.1 compared with the NT50 after B.1. The decrease was 40-times lower for XBB.1.5. For patients who received CoronaVac and had a BF.7 breakthrough infection, the plasma NT50 against XBB.1 was 31-times lower and XBB.1.5 was 27-times lower compared with the NT50 for B.1. A similar trend was also observed in plasma from patients who received two doses of an mRNA vaccine and had a BA.5 breakthrough infection and patients who received CoronaVac and had a BA.1 breakthrough infection. These findings suggest that Pro486 is also a strong neutralising antibody evading mutation, and that the humoral immune escape ability of XBB.1.5 is similar to that of XBB.1. Figure Comparison of antibody evasiveness and hACE2 binding affinity of XBB.1 and XBB.1.5 (A) NT50 against SARS-CoV-2 B.1 (Asp614Gly), XBB.1, and XBB.1.5 pseudovirus using plasma from patients with BA.1 (n=50), BA.5 (n=36), or BF.7 (n=30) breakthrough infection convalescents who had received three doses of CoronaVac, and those with a BA.5 breakthrough infection convalescents who had received three or four vaccinations, including at least two doses of mRNA vaccines (BNT162b2 or mRNA-1273; n=10). p values were calculated using two-tailed Wilcoxon signed rank tests. (B) Pseudovirus IC50 of therapeutic neutralising antibodies. (C) Surface plasmon resonance sensorgrams measuring the hACE2-binding affinity of SARS-CoV-2 BQ.1.1, XBB and XBB.1, and XBB.1.5 receptor-binding domain. Surface plasmon resonance data were fitted to a 1:1 binding model using Biacore 8K Evaluation Software (version 3.0.12; Cytiva, Uppsala, Sweden). All neutralisation assays were done in at least two independent experiments. hACE2=human ACE2. IC50=50% inhibition concentration. ka=fitted association rate constant. kd=fitted dissociation rate constant. KD=dissociation equilibrium constant. *10 000 was the upper limit of detect; these analyses gave values more than 10 000. Compared with XBB.1, XBB.1.5 had similar evasion against therapeutic mAbs (figure B); Evusheld and bebtelovimab did not neutralise XBB.1.5 pseudovirus. Sotrovimab is still active but weak against XBB.1.5. Another BA.5-effective mAb, SA58, is escaped by both XBB.1 and XBB.1.5. However, SA55 remains highly effective against XBB.1.5.2, 6 Previous deep mutational scanning studies have shown that Pro486 might enhance the affinity to hACE2 compared with Ser486. 7 The binding affinity of the XBB.1.5 receptor-binding domain to hACE2 (dissociation constant [KD] 3·4 nM) was similar to that of BA.2.75 (KD 1·8 nM) and much stronger than that of XBB.1 (KD 19 nM) and BQ.1.1 (KD 8.1 nM; figure C; appendix p 6). These results suggest that the probable reason for the significant growth advantage of XBB.1.5 over XBB.1 is that it gained substantially higher ACE2 binding affinity through the Ser486Pro mutation, while retaining an extremely high immune evasion capability. With stronger immune escape ability but weaker ACE2 binding affinity than BQ.1.1, XBB and XBB.1 have only prevailed in a few countries, such as Singapore and India, since September, 2022. Whereas BQ.1.1 has quickly become the dominant global strain. Because of its enhanced hACE2-binding affinity and similar ability to evade the immune system, the prevalence of XBB.1.5 shows that receptor-binding affinity will substantially affect the transmissibility of the strain. The underlying mechanism needs to be investigated. Also, whether the increased receptor-binding affinity would cause a difference in pathogenicity compared with XBB is unclear and requires immediate research. 8 Moreover, the strong affinity to hACE2 might allow XBB.1.5 to acquire additional immune-escape mutations, similar to the evolution trend of BA.2.75, when met with substantial immune pressure. 9 Therefore, the circulation of XBB.1.5 needs to be closely monitored, and the development of effective neutralising antibodies and vaccines against XBB.1.5 is urgently needed. YC is a cofounder of Singlomics Biopharmaceuticals and inventor of provisional patents associated with SARS-CoV-2 neutralising antibodies, including SA55 and SA58. All other authors declare no competing interests. CY and WS contributed equally.
