INTRODUCTION
Since the beginning of the coronavirus disease 2019 (COVID-19) pandemic, multiple severe acute respiratory syndrome 2 (SARS-CoV-2) variants emerged, raising concerns about the effectiveness of infection and/or vaccine-elicited immunity (
1–5). The emergence of a SARS-CoV-2 variant harboring 33 mutations in its Spike glycoprotein (S), Omicron (BA.1), reduced vaccine efficacy against infection due to its improved antibody escape (
6–12). While modified versions of mRNA vaccines were produced to induce an immune response against the Omicron Spike (BA.1) (
13,
14) and its subvariants (
15–18), the persistent evolution of SARS-CoV-2 gave rise to various subvariants across the world (
Fig. 1A and 2A) (
19). BA.1 was rapidly surpassed by BA.2 (
20), and since then, several of its progenies have emerged and demonstrated improved transmission (
21). Notably, BA.2.75, which surfaced in May 2022, further mutated into CH.1.1 (
22,
23). On the other hand, BA.4 and BA.5, which harbor the same Spike, further evolved to BQ.1.1 in late 2022, showing improved immune escape ability (
14,
16,
24). In that same period, a recombinant sublineage, XBB, emerged and showed enhanced immune escape (
25,
26). Since then, most newly occurring Omicron subvariants are derived from XBB and carry the Ser486Pro mutation known to enhance the affinity for the human receptor angiotensin converting enzyme 2 (ACE2) (
27), such as XBB.1.5 which quickly dominated over XBB in January 2023 (
28,
29). Recently, the main XBB subvariants were XBB.1.5, XBB.1.9.1, XBB.1.16, XBB.2.3, and EG.5.1, representing around 80% of reported viral sequences (
30) in addition to EG.5.1 sublineages HK.3 and HV.1, which are growing rapidly around the globe (
31,
32). In August 2023, the emergence of BA.2.86, a highly divergent BA.2 subvariant, caused great concerns regarding infection- and vaccine-elicited immune responses (
30,
33,
34). Although the global number of infections related to BA.2.86 were relatively limited, its fast expansion and diversification in various countries has been noted, with the emergence of BA.2.86 sublineages (i.e., JN.1, JN.2, and JN.3) showing enhanced transmission globally (
32,
35). As of January 2024, JN.1 and its derivatives represent around 70% of reported viral sequences according to Nextstrain (
https://nextstrain.org/ncov/gisaid/global/6m).
The Spike glycoprotein is a metastable fusion protein composed of a trimer of heterodimers expressed at the surface of viral particles and can also be detected at the surface of infected cells (
36–39). Its interaction with the receptor ACE2 on host cells enables S cleavage by host proteases, thus exposing the fusion peptide leading to viral entry (
40–45). Given that the Spike glycoprotein is one of the main targets of humoral responses elicited by SARS-CoV-2 infection and vaccines, a strong selective immune pressure against this crucial protein led to the current evolution of emerging Omicron subvariants (
21,
46–49). Each of these subvariants has acquired mutations in Spike that help evade humoral responses, resulting, in some cases, in increased binding affinity for ACE2 (
50–52). Interestingly, multiple Omicron lineages gained identical or similar Spike mutations in key antigenic sites in the receptor binding domain (RBD) and in the N-terminal domain (NTD), suggesting a convergent evolution (
53,
54). It is now well established that Spike evolution is intimately associated with viral fitness and transmission in humans (
5,
50).
While viral transmission is a multifactorial phenomenon, Spike-ACE2 interaction appears to play an predominant role (
19,
40,
50). As such, parameters influencing the strength of this interaction may influence viral growth of emerging Omicron subvariant. We previously demonstrated that temperature modulates the interaction between SARS-CoV-2 Spike and ACE2, with temperatures lower than 37°C (i.e., 4°C and 25°C) increasing ACE2-binding affinity and viral entry (
55). We showed that this modulation was explained by favorable thermodynamic changes leading to the stabilization of the RBD-ACE2 interface and by triggering a more “open’’ conformation of the Spike trimer. Subsequent work on early Omicron subvariants (BA.1, BA.2, BA.2.12.1, BA.4/5) also showed an impact of temperature on Spike-ACE2 interaction (
56). This is of particular interest because it has been suggested that the optimal air temperature for SARS-CoV-2 transmission ranges from 5°C to 15°C (
57,
58). Furthermore, SARS-CoV-2 airway transmission is influenced by external factors, such as the temperature gradient that exists in human airways (from 30°C to around 36°C) (
59,
60). Hence, external temperatures reaching sub-zero degrees Celsius might prime SARS-CoV-2 Spike for ACE2 binding, thus improving its adsorption onto the epithelial airway cells (
55,
61). Interestingly, several studies have observed a link between colder temperatures and a higher incidence of COVID-19 cases (
62,
63), and this could also be another factor underlying SARS-CoV-2 seasonality (
64,
65).
To better understand parameters associated with viral transmission, we functionally characterized the Spike glycoprotein of emerging Omicron subvariants, which possess enhanced immune escape and increased binding affinity to ACE2 compared to early Omicron strains. We tested the capacity of plasma from individuals who received a fifth dose of bivalent (BA.1 or BA.4/5) mRNA vaccine to recognize and neutralize several Spikes from recent Omicron subvariants. We next determined how temperature affects the interaction between Spike and ACE2 by combining an array of biochemical and biological assays, including biolayer interferometry, flow cytometry, and virus capture assay. We also evaluated the susceptibility of Omicron subvariants to cold inactivation, a property that may reflect the propensity of Spikes to ensue conformational transitions (
66–68). The associations between these parameters and the viral growth rate of each Omicron subvariant in the population between early 2022 and early 2024 were evaluated.
DISCUSSION
The continued evolution of SARS-CoV-2 requires constant monitoring of its new variants (
21). Current circulating strains are derived from the Omicron variant, with each newly emerging subvariant showing improved transmission, which can mostly be explained by their enhanced antibody escape and ACE2 binding affinity (
54,
115). To determine if other Spike characteristics are involved in viral growth rates, we characterized the functional properties of recent Omicron subvariant Spikes (
Fig. 8), including their capacity to evade humoral responses.
We observed significant differences in the capacity of individuals receiving a fifth dose of bivalent (BA.1 or BA.4/5) mRNA vaccine to recognize and neutralize recent Omicron subvariants (
Fig. 1B and C). The evolutions of CH.1.1 to DV.7.1, EG.5.1 to HK.3, and BA.2.86 to JN.1 led to an improvement in their capacity to evade humoral responses, which could likely be explained by the acquisition of mutations on either S:L455 or S:F456 (
30–32,
97,
98). Consistent with other reports, we found that BA.2.86 possesses slightly better escape from plasma-mediated neutralization compared to XBB.1.5, and this was further pronounced with JN.1 (
30,
32,
35,
98,
100,
102,
103). These results demonstrate that the high levels of preexisting immune pressure against SARS-CoV-2 Spike glycoprotein are driving its evolution and transmission (
48).
By using various approaches, we characterized the ACE2-binding properties of recent Omicron subvariants Spikes. We found that among all Omicron subvariants tested, BA.2.86 was the most susceptible to ACE2-Fc neutralization and exhibited the highest RBD affinity, protomer cooperativity, and ACE2 binding at the surface of transfected cells and viral particles (
Fig. 2 to
4). These results are in line with previous studies demonstrating the marked improvement in ACE2 binding affinity and explained notably through the acquisition of the S:R403K mutation (
33,
101). Furthermore, this increased binding affinity could also be linked to the intrinsic charge properties of the region of ACE2 targeted by the Omicron RBD, which are negatively and positively charged, respectively (
116–118). Notably, the additional positive charges within the RBD associated with mutations V445H, N460K, N481K, and A484K might contribute to its improved ACE2 binding (
101). Recently, a BA.2.86 sublineage, JN.1, has shown improved growth rate due to one substitution S:L455S, improving its neutralization escape at the expense of ACE2 binding (
32). In this study, we confirm these results by which JN.1 possesses improved immune escape concomitant with its lower ACE2 binding. This further demonstrates the delicate balance existing between neutralization escape and ACE2 affinity in SARS-CoV-2 transmission (
5).
It has been suggested that the optimal air temperature for SARS-CoV-2 transmission ranges from 5°C to 15°C (
57,
58). Within upper airways, lower temperature creates a gradient of temperature from the nasal cavity to the trachea, where it reaches around 33°C (
119–122). We previously found that lower temperatures improve RBD-ACE2 interaction and enhance the Spike propensity to sample the “up’’ conformation leading to higher ACE2 binding, fusogenicity, and viral replication (
52,
55,
56). By priming SARS-CoV-2 Spike, lower temperatures could enhance its ability to bind ACE2 in the upper airways, favoring the initial adsorption (
61). In this study, we found that all Omicron subvariants Spikes tested remained sensitive to the impact of low temperatures, further enhancing their protomer cooperativity and ACE2 binding at the surface of transfected cells and viral particles (
Fig. 3 and 4). More importantly, we found that most Omicron subvariants reached similar levels of binding at 25°C than that of D614G at 4°C at the cell surface and with the RBD, suggesting a lesser reliability on cold temperatures for their improved ACE2 binding. This impact was most notable with DV.7.1, HK.3, and BA.2.86. While multiple factors are involved in SARS-CoV-2 transmission, several lines of evidence demonstrated an association between the climate and increased disease transmission, with lower temperature and humidity being associated with higher COVID-19 incidence (
62,
63,
123–129). Whether this is due to ACE2 binding at lower temperatures remains to be determined. Notably, the rapid spread of BA.2.86 and expansion of its sublineage JN.1 in North America and in “colder’’ Northern European countries such as the United Kingdom, Denmark, Sweden, Iceland, and France raises the intriguing possibility that lower temperatures could play a role in transmission through improved ACE2 binding (
30,
32,
33,
100,
130,
131).
The Spike glycoprotein requires the adoption of the RBD “up’’ conformation to interact with its receptor ACE2 (
44,
45,
132). One parameter affecting the propensity to adopt the “up’’ conformation, and thus ACE2 interaction, is the degree of inter-protomer cooperativity upon ACE2 binding within Spike trimers (
72,
133). We and others have demonstrated that SARS-CoV-2 Spike from early variants such as Alpha, Delta, and Omicron (BA.1 and BA.4/5) have an improved inter-protomer cooperativity compared to D614G, which was further enhanced at low temperatures (
55,
56,
133). Here, we show that Omicron evolution continues to follow this trend. We observe a marked improvement in cooperativity among recently emerging Omicron subvariants, with DV.7.1, HK.3, BA.2.86, and JN.1 showing the highest levels of cooperativity at 37°C compared to D614G, with these differences being further pronounced at low temperatures (
Fig. 4). Interestingly, the convergent evolution of emerging Omicron sublineages harboring mutations at either S:L455 or S:F456 illustrates the importance of these residues in transmission (
30,
97). Our results show that the acquisition of the “FLip’’ mutations strongly enhanced inter-protomer cooperativity, concomitant with higher ACE2 binding, with DV.7.1 and HK.3 showing a remarked improvement compared to their respective parental lineages (
31,
97). Thus, the ongoing surveillance of emerging SARS-CoV-2 variants is most likely due to their concomitant improvement in neutralization escape, enhanced protomer cooperativity, and increased ACE2 binding, which could inform future COVID-19 vaccine design.
We also observed differences in S cleavage to S2 of cells expressing emerging Omicron subvariant Spikes (
Fig. 5A). In agreement with previous reports, we found that BQ.1.1 and CH.1.1 were processed more efficiently than their respective parental lineages, which could impact their fusogenicity and intrinsic viral pathogenicity (
24,
94,
134). We also found that BA.2.86 and JN.1 were more processed than XBB.1.5 and EG.5.1, likely explained by the S:P681R mutation known to enhance Spike processing, fusogenicity, and pathogenicity (
33,
41,
134,
135). Whether this enhanced processing will affect pathogenicity in humans remains to be known.
We also found a correlation between S2 processing and susceptibility to cold inactivation (
Fig. 5). In contrast to the early Omicron BA.1, recent Omicron subvariants possess enhanced processing while remaining remarkedly sensitive to cold inactivation (
66). Interestingly, it was found that susceptibility to cold inactivation may reflect the propensity to sample more “open’’ Spike conformations, thus facilitating conformational transitions (
66,
68). As such, cold inactivation might be a helpful tool to determine which variants sample more “open’’ conformations which may lead to better viral replication. Our results suggest that, at least for Omicron subvariants, S2 processing could modulate Spike conformation to adopt more “open’’ conformations, rendering them more susceptible to cold inactivation. Interestingly, DV.7.1 and HK.3, harboring the “Flip’’ mutations, had less S2 processing and were less sensitive to cold inactivation compared to their parental lineages, suggesting that “Flip’’ mutations might improve stability through decreased processing, while also strongly enhancing ACE2 binding and the effect of low temperatures. Of note, EG.5.1 was more resistant to cold inactivation than FD.1.1 and XBB.1.5, which could possibly be attributed to its S:Q52H mutation. Interestingly, it was found that EG.5.1 had an increased transmissibility and altered tropism from that of XBB.1.5 (
136). Whether this is linked to its higher susceptibility to ACE2-Fc neutralization and improved stability remains to be established.
Finally, we also observed that plasma-mediated recognition, neutralization, and ACE2-Fc binding at the surface of cells or pseudoviral particles was associated with emerging Omicron subvariants growth rate (
Fig. 7). The combination of both escape from plasma and ACE2 binding at low temperatures enhanced these associations. These observations further support that ACE2 interaction and immune escape is associated with SARS-CoV-2 transmission and evolution (
54,
115). Intriguingly, we observed a higher correlation coefficient for ACE2 binding and virus capture with growth rates at lower temperatures, thus suggesting that temperature modulation of Spike-ACE2 interaction plays a role in viral transmission.
The continued evolution of SARS-CoV-2 requires constant monitoring of its ability to evade immune responses elicited by previous infections and/or vaccination. Our results suggest that the capacity of new emerging variants to interact with ACE2, particularly at low temperatures, is another parameter that deserves to be closely monitored. The growth advantage of XBB.1.5 and its sublineages compared to XBB demonstrate the importance of monitoring ACE2 interaction (
28,
29,
40,
44,
110,
115). Although the exact mechanisms through which temperature affects SARS-CoV-2 transmission remain unclear, our findings indicate that Omicron subvariants have undergone mutations that enhance resistance to neutralization by plasma, improve S processing, and increase affinity for ACE2 at both low and high temperatures. Of note, we found that measurement of Spike-ACE2 interaction of viral particles at low temperatures is strongly associated with Omicron subvariants growth rates. Such measures can readily be performed upon the emergence of new variants and could help define which variants have the potential to rapidly expand. In summary, our study underscores the necessity for ongoing surveillance of emerging subvariants and their characteristic mutations, as this information is likely to inform which variants have the potential to become predominant and inform the development of vaccines and other interventions.
ACKNOWLEDGMENTS
The authors are grateful to the donors who participated in this study. The authors thank the CRCHUM BSL3 and Flow Cytometry Platforms for technical assistance. We gratefully acknowledge the GISAID Initiative and the generous contribution of all data contributors, including the laboratories that collect the specimens and generated the genetic sequence and metadata on which part of this research is based.
Panel A of
Fig. 1 and panel A of
Fig. 2 were prepared using illustrations from BioRender.
This work was supported by le Ministère de l’Économie et de l’Innovation du Québec, Programme de soutien aux organismes de recherche et d’innovation, to A.F. and by the Fondation du CHUM. This work was also supported by CIHR foundation grant #352417, by CIHR operating Pandemic and Health Emergencies Research grant #177958, and by Exceptional Fund COVID-19 from the Canada Foundation for Innovation (CFI) #41027 to A.F. This work was also supported by CIHR Project grant #174924 to J.H. Work on variants presented was also supported by the Sentinelle COVID Quebec network led by the LSPQ in collaboration with Fonds de Recherche du Québec Santé (FRQS) to M.C. and A.F. A.F. is the recipient of a Canada Research Chair on Retroviral Entry, no. RCHS0235. M.C. is a Tier II Canada Research Chair in Molecular Virology and Antiviral Therapeutics no. 950-232424., and J.H. is FRQS Junior 2 research scholar. A.T. was supported by a MITACS Elevation postdoctoral fellowship. M.B. is the recipient of a CIHR master’s scholarship award. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. We declare no competing interests.
M.B. and A.F. conceived the study. M.B., S.D., E.B., A.T., H.M., J.F., M.P., I.L., and C.A. performed, analyzed, and interpreted the experiments. H.M., O.E.F., Y.B., C.B., J.F., M.P., M.C., and A.F. contributed unique reagents. R.P. and J.H. performed lineage growth rate estimation analyses. M.B. and A.F. wrote the manuscript with input from others. All authors have read and agreed to the published version of the manuscript.
The views expressed in this paper are those of the authors and do not reflect the official policy or position of the Uniformed Services University, US Army, the Department of Defense, or the US Government.