Omicron Subvariants Infection Kinetics and Nirmatrelvir Efficacy in Transgenic K18-hACE2 Mice

Abstract

The persistent evolution of SARS-CoV-2 has led to the emergence of antigenically distinct Omicron subvariants exhibiting increased transmissibility, immune evasion, and altered pathogenicity. Among these, recent subvariants such as JN.1, KP.3.1.1, and LB.1 possess unique antigenic and virological features, underscoring the need for continued surveillance and therapeutic evaluation. As vaccines and commercial monoclonal antibodies show reduced effectiveness against these variants, the role of direct-acting antivirals, such as Nirmatrelvir, targeting conserved viral elements like the main protease inhibitor, becomes increasingly crucial. In this study, we investigated the replication kinetics, host immune responses, and therapeutic susceptibility of three recently circulating Omicron subvariants in the K18-hACE2 transgenic mouse model, using the SARS-CoV-2 parent WA1/2020 strain as a reference. Omicron subvariants exhibited a marked temporal shift in viral infection kinetics characterized by an early lung viral titer peak (~7–8 Log PFU) at 2 days post-infection (dpi), followed by a decline (1–3 Log PFU) by 4 dpi. Pulmonary cytokine and chemokine responses (GM-CSF, TNF-α, IL-1β, IL-6) showed an earlier increase in subvariant-infected mice compared to a gradual response in WA1/2020 infection. Notably, Nirmatrelvir treatment led to significant reductions in lung viral titers in subvariant-infected mice compared to WA1/2020, surpassing its efficacy against the parent strain. These findings highlight that infection with Omicron subvariants yields a broad dynamic range in viral burden with minimum variability, while retaining a prominent therapeutic response to Nirmatrelvir. This study provides insights into the emerging subvariants’ pathogenesis and therapeutic responsiveness, reinforcing the importance of continued variant monitoring and the development of effective countermeasures.

1. Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) continues to evolve as one of the most significant public health threats of the 21st century. Since its emergence, it has caused widespread morbidity and mortality, with over 7.78 million deaths globally reported as of December 2024 (WHO report 2024) [1]. A key driver of the virus’s persistence has been its remarkable evolutionary capacity, fueled by the intrinsic infidelity of its RNA-dependent RNA polymerase, leading to the continuous emergence of novel variants with enhanced transmissibility and immune escape properties [2,3].

Among these, the Omicron lineage has proven particularly adept at evading immunity, rapidly diversifying into a multitude of subvariants [4]. Several of these subvariants, such as JN.1, KP3.1.1, and LB.1, have risen to dominance between August 2023 and March 2025, displacing earlier strains and posing fresh challenges for both natural and vaccine-induced immunity [5,6,7]. These subvariants carry extensive spike protein mutations, some of which confer characteristics reminiscent of earlier variants like Delta, including increased fitness and altered tissue tropism [4]. JN.1, a descendant of BA.2.86, is characterized by the L455S mutation in the receptor-binding domain (RBD) of the spike protein, enhancing its binding affinity to the ACE2 receptor and contributing to its rapid global spread [8]. JN.1 contains more than 70 mutations in the spike protein compared to the ancestral Wuhan-Hu-1 strain [9]. As of early 2024, JN.1 accounted for approximately 62% of circulating variants in the United States and was designated a variant of interest by the World Health Organization [10]. In mid-2024, the emergence of new subvariants KP.3 and LB.1, both derived from JN.1, was followed by the evolution of KP.3.1.1 from KP.3 through the accumulation of additional mutations [5,6]. KP.3.1.1 became the predominant variant in the United States, representing over 50% of cases by September 2024. This subvariant harbors mutations such as F456L and Q493E in the RBD, and a deletion at position S31 in the N-terminal domain (NTD), which may alter spike protein conformation and enhance immune escape [11,12]. LB.1, another emerging subvariant, shares similarities with the FLiRT variants and includes a deletion at residue 31 outside the RBD [12]. In August 2024, LB.1 accounted for approximately 14.9% of new cases in the United States. The rapid antigenic shift is thought to be driven by selection pressures from widespread vaccination and repeated natural infections, prompting the virus to find new pathways of immune evasion. While much of the global response has focused on neutralizing antibody-based strategies targeting the spike protein, the pace of antigenic drift has significantly reduced the effectiveness of such countermeasures [12,13,14,15].

Importantly, all three subvariants share a conserved P132H mutation in the main protease (Mpro), a mutation located at the interface between domains II and III, a region critical for Mpro structural stability but distal to its catalytic dyad and active site. Biochemical and structural studies have shown that, while P132H reduces thermal stability of the enzyme slightly, it does not impair catalytic activity nor compromise binding affinity for selective inhibitors such as Nirmatrelvir [12,16]. Nirmatrelvir binds to a highly conserved active site unaffected by the P132H mutation, and has emerged as a cornerstone of outpatient COVID-19 treatment due to its broad-spectrum activity and clinical effectiveness [16,17]. Given the declining efficacy of commercial monoclonal antibodies and neutralizing vaccines against these subvariants, small-molecule antivirals directed at conserved viral targets, including Mpro, methyltransferases, and RNA polymerase, offer a more resilient therapeutic approach [16,17,18]. However, with the virus in a state of constant evolution, there remains an urgent need to evaluate the effectiveness of such antivirals against newly emerging subvariants. Understanding the infection dynamics and therapeutic susceptibilities of these subvariants is crucial for informing public health strategies and treatment approaches.

In this study, we investigate the viral replication dynamics, immune response, and antiviral susceptibility of the currently circulating Omicron subvariants JN.1, KP3.1.1, and LB.1 isolated from New Jersey clinical isolates, using the K18-hACE2 transgenic mouse model, a well-established platform for preclinical SARS-CoV-2 pathogenesis and intervention studies [19,20]. Specifically, we investigated whether these subvariants exhibit altered infection kinetics, including early viral peaks and altered host immune response, relative to the parent WA1/2020 strain. Furthermore, we assessed the in vivo efficacy of Nirmatrelvir against these subvariants to test their therapeutic susceptibility and the hypothesis that the K18-hACE2 mouse model provides a robust platform for preclinical antiviral evaluation across emerging variants, integrating virological, immunological, and therapeutic assessments [21,22]. This work provides valuable insights into the ongoing evolution of SARS-CoV-2, the continued utility of the transgenic K18-hACE-2 mouse model for efficacy evaluation, and the importance of developing antivirals with novel targets in the face of continued viral diversification.

2. Results

2.1. Altered Lung Infection Kinetics of Omicron Subvariants in K18-hACE2 Mice Compared to Parent WA1/2020

To evaluate the infection dynamics of emerging Omicron subvariants, we intranasally inoculated K18-hACE2 transgenic mice with JN.1, LB.1, KP3.1.1, or the parent WA1/2020 strain at doses shown in

Table 1. Infection dose given per mouse.

Strain	Infection Dose PFU/Mouse
WA1/2020	2.85 × 105
JN.1	1.2 × 105
LB.1	1.25 × 104
KP.3.1.1	4.6 × 104
WA1/2020	2.85 × 105
JN.1	1.2 × 105

(Figure 1a). To assess infection kinetics during early disease, we focused on days 2 through 4 post-infection, a window selected based on prior studies and preliminary data indicating that WA1/2020 viral replication peaks at day 4, while Omicron subvariants peak earlier, between days 2 and 3. This time frame allowed us to capture both the peak viral loads and underlying early innate immune responses associated with each strain. To assess clinical disease severity, we monitored body weight in K18-hACE2 mice over 4 days post-infection (dpi) with WA1/2020 or Omicron subvariants JN.1, LB.1, and KP.3.1.1. Mice infected with WA1/2020 showed a progressive decline in body weight, culminating in an average loss of ~10% by 4 dpi. In contrast, mice infected with JN.1 maintained baseline weights throughout the observation period, with no net loss observed. LB.1 and KP.3.1.1 infected mice exhibited modest fluctuations, with average body weights remaining above 95% of baseline on all days, with 2–3% net weight loss by day 4. These data indicate that Omicron subvariants induce milder clinical phenotypes compared to the parent WA1/2020 strain (Figure 1b).

To evaluate viral replication kinetics in the lower respiratory tract, lung viral load was assessed as Log PFU/g of tissue by plaque-forming unit (PFU) assay and Log GE (genomic equivalents)/g of tissue by RT-PCR at days 2, 3, and 4 post-infection in K18-hACE2 mice (Figure 1c,d). WA1/2020-infected mice showed high lung titers across all days, with mean Log PFU values ranging from ~7.0 to 8.5, attaining peak titer at 4 dpi (Figure 1c). Mice infected with JN.1 exhibited moderately lower titers at all time points, with mean Log PFU levels 6.7–7 across days 2–4. While stable, the day 4 titers were significantly lower than those observed in WA1/2020, indicating a degree of attenuation in lung viral replication. LB.1-infected mice showed a reduction in viral titers. On day 2, mean lung viral titers were Log 7.0 PFU, but decreased gradually to mean Log PFU values of 5.9 and 5.5 by 3 and 4 dpi, respectively. KP.3.1.1 displayed similar kinetics, peaking at ~7.0 Log PFU at day 2, dropping to 6.6 and 5.9 Log PFU at 3 and 4 dpi, respectively. The lung titers revealed no significant differences between WA1/2020 and any Omicron subvariants at 2 dpi. However, by 3 dpi, LB.1 and KP.3.1.1-infected mice had significantly reduced titers compared to WA1/2020, with mean differences of 1.7 log (p = 0.0032) and 0.935 log (p = 0.0489), respectively, while JN.1 showed a non-significant trend. At 4 dpi, the differences became more pronounced: WA1/2020 vs. JN.1 (1.5 log, p = 0.0012), vs. LB.1 (2.767 log, p < 0.0001), and vs. KP.3.1.1 (2.285 log, p < 0.0001), indicating robust clearance in Omicron-infected lungs. While WA1-infected mice exhibited peak lung viral titers at 4 dpi, all Omicron subvariants exhibited a temporal shift, peaking at 2 dpi in LB.1 and KP.3.1.1 infected mice, and 3 dpi in JN.1 infected mice (Figure 1c). This shift suggests accelerated lung viral clearance in Omicron-infected mice. These data indicate that parent WA1/2020 replicates more efficiently in the lower respiratory tract, followed by JN.1, while LB.1 and KP.3.1.1 exhibit attenuated viral replication in lungs consistent with reduced pathogenicity.

To complement the replicating viral quantification by PFU, lung viral RNA levels were also measured by RT-PCR analysis at days 2, 3, and 4 dpi (Figure 1d). WA1/2020-infected mice exhibited the highest mean RT-PCR log values across all time points, ranging from mean Log GE values of ~10.0 on day 2 to ~12.0 by day 4, indicating persistent and escalating viral RNA burden in the lungs. JN.1-infected mice showed slightly lower yet sustained levels of viral RNA, with Log GE means ranging from 9.4 to 10.2 across the three days, suggesting moderately efficient replication compared to WA1/2020. LB.1-infected mice demonstrated a more attenuated viral RNA profile. Mean Log GE values on day 2 were ~9.7, followed by significantly lower ~9.6 by day 3 (p = 0.0029) than WA1/2020, and remained relatively stable by day 4 (~9.7), indicating limited viral amplification. Similarly, KP.3.1.1-infected mice exhibited intermediate replication. Early RNA levels, with mean Log GE values of 9 at 2 dpi, stabilized to 9.6 by 3 dpi, which was lower than WA1/2020 (p = 0.0378), and remained steady through day 4 to 9.8, consistent with a milder viral replication phenotype.

Notably, while WA1/2020 showed a continued increase in viral RNA through 4 dpi, Omicron subvariants (JN.1, LB.1, KP.3.1.1) demonstrated plateauing RT-PCR curves after 2–3 dpi, further supporting the trend of more rapid viral clearance observed in PFU assays. These findings underscore that WA1/2020 exhibits the most robust replication kinetics in the lower respiratory tract by 4 dpi, while Omicron subvariants, especially LB.1 and KP.3.1.1, are associated with attenuated replication, as evidenced by both PFU and RT-PCR measures.

2.2. Divergent Temporal Patterns of Lung Viral Replication Among Omicron Subvariants and Parent WA1/2020

To investigate the relationships in lung viral replication kinetics among SARS-CoV-2 variants, we performed Spearman correlation analyses using plaque-forming unit (PFU) titers at 2, 3, and 4 dpi in K18-hACE2 mice (Figure 1e). At 2 dpi, WA1/2020 and JN.1 showed a moderate positive correlation (ρ = 0.59), suggesting aligned replication trends early in infection. In contrast, WA1/2020 was moderately negatively correlated with LB.1 (ρ = −0.50) and weakly negatively correlated with KP.3.1.1 (ρ = −0.29), indicating that this Omicron subvariants may follow divergent replication trajectories from the parent strain. A striking finding was the very strong negative correlation between JN.1 and LB.1 (ρ = −0.96), indicating inverse viral replication trends. Meanwhile, LB.1 and KP.3.1.1 were moderately positively correlated (ρ = 0.61), suggesting partial alignment in viral dynamics among Omicron lineages. By 3 dpi, the relationships shifted substantially. WA1/2020 and JN.1 displayed a strong positive correlation (ρ = 0.79), indicating synchronous viral replication during peak infection. WA1/2020 also correlated moderately with LB.1 (ρ = 0.40) and weakly with KP.3.1.1 (ρ = 0.28). JN.1 and LB.1 maintained a moderate positive correlation (ρ = 0.66), in stark contrast to their inverse trend at 2 dpi, suggesting that replication kinetics may converge at this intermediate stage. Relationships between JN.1 and KP.3.1.1 were weakly negative (ρ = −0.18), and LB.1 and KP.3.1.1 were weakly correlated (ρ = 0.04). At 4 dpi, WA1/2020 showed negative correlations with all other variants, including JN.1 (ρ = −0.41), LB.1 (ρ = −0.31), and KP.3.1.1 (ρ = −0.17), indicating a divergence in viral persistence, as WA1 continued to replicate, while Omicron subvariants declined. In contrast, JN.1 and LB.1 showed a strong positive correlation (ρ = 0.75), indicating convergence in viral burden reduction. Similarly, KP.3.1.1 showed a positive, though weaker, correlation with JN.1 (ρ = 0.09) and LB.1 (ρ = 0.17).

Overall, the correlation analysis revealed dynamic and temporally variable relationships among SARS-CoV-2 variants in lung replication. At early timepoints (2 dpi), inverse replication trends were apparent between JN.1 and LB.1, while WA1 aligned most closely with JN.1. As infection progressed, JN.1 KP.3.1.1 and LB.1 became increasingly synchronized, while WA1/2020 began to diverge, maintaining high titers as Omicron subvariants declined. These results underscore the distinct temporal patterns of viral replication among Omicron subvariants relative to the parent strain, highlighting the importance of temporal resolution in assessing viral pathogenesis. A similar analysis was performed for viral RNA levels (Figure S1).

2.3. Broad In Vivo Efficacy of Nirmatrelvir Across WA1/2020 and Omicron Subvariants

To evaluate the antiviral activity of Nirmatrelvir treatment against the parent WA1/2020 strain, and the Omicron subvariants JN.1, LB.1, and KP.3.1, K18-hACE2 mice were infected and treated with vehicle (10% Tween 80) or 1000 mg/kg of Nirmatrelvir orally twice daily for four days, starting 12 h post-infection (hpi) (Figure 2a). No significant weight change was observed in Nirmatrelvir-treated groups in comparison to vehicle control at all days post-administration (Figure S2).

Viral load in the lungs was quantified as Log PFU/g and Log GE/g of tissue at the study endpoint (Figure 2b,c). Plaque assay analysis revealed a significant reduction in viral load in the Nirmatrelvir-treated groups compared to the vehicle control across all tested SARS-CoV-2 strains (Figure 2b). Mice infected with the parent WA1/2020 strain showed a significant reduction in mean viral titers from ~7.8 Log PFU in the vehicle-treated group to 6.53 Log PFU in the Nirmatrelvir-treated group (mean difference = 1.413, p < 0.0001).

Mice infected with Omicron subvariant JN.1 demonstrated a substantial response to Nirmatrelvir treatment. The vehicle control group infected with JN.1 showed a mean viral titer of 6.73 Log PFU, which was significantly decreased to 3.22 Log PFU in the Nirmatrelvir-treated group (mean difference = 4.267, p = 0.0001). Similarly, Nirmatrelvir treatment showed significant efficacy against the LB.1 and KP.3.1.1 Omicron subvariants. In LB.1 treated mice, the vehicle-treated group showed a mean viral titer of 5.47 Log PFU, which was significantly reduced to 2.50 Log PFU in the Nirmatrelvir-treated group (mean difference = 3.467, p < 0.0001). Whereas in KP.3.1.1 infected mice, the vehicle control group showed a mean viral titer of 5.95 Log PFU, which was significantly reduced to 3.83 Log PFU in the Nirmatrelvir—treated group (mean difference = 2.892, p = 0.0099).

Consistent with the infectious viral data (PFU), RT-PCR analysis revealed a significant reduction in lung viral RNA levels in the Nirmatrelvir-treated groups compared to the vehicle control across all tested SARS-CoV-2 strains (Figure 2c). In parent WA1/2020 infected mice, the mean viral RNA level in the vehicle-treated group was 10.45 Log GE, which was significantly reduced to 9.85 Log GE in the Nirmatrelvir-treated group (mean difference = 0.72, p = 0.0097). The Omicron subvariant JN.1 infected mice demonstrated a marked reduction in lung viral RNA following Nirmatrelvir treatment. The vehicle control group infected with JN.1 exhibited a mean viral RNA level of 8.53 Log GE, which was significantly decreased to 7.13 Log GE in the Nirmatrelvir-treated group (mean difference = 1.59, p < 0.0001). Similarly, Nirmatrelvir treatment showed a significant decrease in lung viral RNA levels against the LB.1 and KP.3.1.1 infected mice. In LB.1 infected mice, the vehicle-treated group had a mean viral RNA level of 9.74 Log GE, which was significantly reduced to 8.95 Log GE in the Nirmatrelvir-treated group (mean difference = 0.99, p = 0.0012). Finally, in KP.3.1 infected mice, the vehicle control group showed a mean viral RNA level of 9.77 Log GE, while the Nirmatrelvir-treated group showed a mean viral RNA level of 8.80 Log GE (mean difference = 1.15, p = 0.0027). These findings underscore that Nirmatrelvir retained potent efficacy against Omicron subvariants.

2.4. Comparative Nirmatrelvir Susceptibility Reveals Enhanced Antiviral Response in Omicron Subvariants Infected Mice in Comparison to WA1/2020

Nirmatrelvir treatment in K18-hACE2 mice demonstrated significant antiviral activity, as measured by an infectious virus as PFU, against both the parent WA1/2020 strain and all Omicron subvariants (JN.1, LB.1, and KP.3.1.1). Nirmatrelvir nearly completely suppressed infectious virus in the Omicron subvariants JN.1 and LB.1, reaching the limit of detection (p < 0.0001). KP.3.1.1 also showed a substantial reduction, as evident by a decrease in PFU (Figure 2d, grey bars), which was significantly lower than parent WA1/2020 (p < 0.0001). Nirmatrelvir treatment also significantly reduced viral RNA levels in JN.1 (p < 0.0001) treated mice in comparison to WA1/2020 infected mice. However, no significant viral RNA level reduction was observed in the LB.1 and KP.3.1.1 infected mice following treatment, as compared to the WA1/2020 infected group (Figure 2d, blue bars). Nevertheless, in vitro evaluation of Nirmatrelvir efficacy revealed EC₅₀ values of 0.0158 ± 0.0105 μM for JN.1, 0.060 ± 0.011 μM for LB.1, and 0.090 ± 0.051 μM for KP.3.1.1, compared to 0.0301 ± 0.0114 μM for the WA1/2020 strain (Figure S3). Overall, these findings further support the continued susceptibility of Omicron subvariants to Nirmatrelvir treatment, surpassing that shown in WA1/2020.

2.5. Early Pulmonary Immune Response Analysis Reveals Distinct Pattern in Omicron Subvariants Infected Mice

To determine the early pulmonary immune activation in response to SARS-CoV-2 infection, cytokine and chemokine levels were quantified in lung homogenates of K18-hACE2 mice at days 2, 3, and 4 post-infection (Figure 3a and Figure S4a). The panel of cytokines and chemokines assessed in this study was chosen based on clinical observations from severe SARS-CoV-2-infected patients, where excessive and dysregulated immune responses, often termed “cytokine storm”, have been implicated in severe disease progression. Previous clinical reports identified several cytokines, such as GM-CSF, TNF-α, MCP-1 (CCL2), IL-1β, IL-6, IL-10, CXCL10 (IP-10), and CCL3 (MIP-1α), as key mediators consistently elevated in severe cases, contributing to hyperinflammation, immune-mediated tissue damage, and poor clinical outcomes [23,24,25,26]. In this study, these cytokines and chemokines were analyzed in lung homogenates of infected mice to provide a comprehensive understanding of the early immune activation elicited by SARS-CoV-2 infection in the K18-hACE2 mouse model. The levels of the selected cytokines and chemokines were higher than baseline in all the groups, suggesting a strong inflammatory response to infection. However, the dynamic range of levels shows the heterogeneous nature of the immune response as seen in the patients.

At 2 dpi, WA1/2020-infected mice showed moderate GM-CSF induction with a mean of 28 pg/mL, while JN.1 and LB.1 infected mice exhibited a mean value of ~47 pg/mL, respectively, which was significantly higher (p = 0.0087) in KP.3.1.1 infected mice with a mean of 53 pg/mL, reflecting an accelerated early immune activation (Figure 3a). At 3 and 4 dpi, there was a slight increase in the levels with no significant difference in the levels in Omicron-infected mice in comparison to WA1/2020.

At 2 dpi, WA1/2020 induced mild TNF-α (mean 6.4 pg/mL), which is a proinflammatory cytokine, while JN.1, KP.3.1.1 and LB.1 showed a mean of ~10, 19and 20 pg/mL though not significantly different from the WA1/2020 infection. By 4 dpi, all variants showed a similar trend in TNF-α levels (Figure S4a).

MCP-1 levels followed a similar trend in all the groups in the range of 2000–4000 pg/mL at 2 dpi, which was significantly decreased (p = 0.04) in LB.1 infected mice at 3 dpi in comparison to WA1/2020 (Figure 3a). MCP-1 levels showed a gradual increase from a mean of 4529 to 8151 pg/mL in WA1/2020, while Omicron-infected mice showed a sustained trend of mean 2200–5333 pg/mL in JN.1, 2800–3994 pg/mL in LB.1, and 3272–4085 pg/mL in KP.3.1.1 over a 2–4-day period. The increasing trend in MCP-1 levels in WA1/2020 infection is indicative of strong monocyte recruitment in lungs, which was mild in Omicron-infected mice.

WA1/2020-infected mice had moderate IL-1β levels of mean ~827 pg/mL at 2 dpi, while JN.1, KP.3.1.1, and LB.1 infected mice had higher means of 1454, 1564, and 1484 pg/mL, respectively (Figure 3a). At 3 dpi, JN.1 showed significantly higher levels with a mean difference of 1379 pg/mL (p = 0.0003) than WA1/2020. There was no significant change in the levels at 4 dpi in all subvariant-infected groups compared to WA1/2020, showing a more sustained IL-1β response, reflecting a steady immune activation pattern.

IL-6, a central pro-inflammatory cytokine, showed a mean of 275, 459, 443, and 390 pg/mL in WA1/2020, JN.1, LB.1, and KP.3.1.1 infected mice, respectively, at 2 dpi (Figure 3a). By 3 and 4 dpi, IL-6 levels further increased across all groups, indicating persistent immune activation. Nonetheless, JN.1 showed a significant increase to a mean of 580 pg/mL (p = 0.0181) relative to WA1/2020 at 3 dpi. At 4 dpi, KP.3.1.1 and JN.1 showed a significant increase to a mean of 351 pg/mL (p = 0.0093) and 691 pg/mL (p = 0.0028). Overall, Omicron subvariant infection induced a more pronounced early response than the parent strain.

IL-10 levels showed a minimal increase from baseline IL-10 in all groups at 2, 3, and 4 dpi, suggesting a balanced anti-inflammatory response (Figure 3a). This increase was significant in JN.1 infected mice at 3 dpi (p = 0.0087) and 4 dpi (p = 0.0119) than WA1/2020, which implies JN.1 induces an early anti-inflammatory response in lungs.

CXCL10 was markedly elevated in all variants from baseline, which was highest in WA1/2020 infected mice, indicating strong interferon-driven responses across 2–4 dpi. However, LB.1 showed a significant decrease (mean difference = 2297 pg/mL; p = 0.047) compared to WA1/2020 at 3 dpi. Another chemokine, CCL3, had a similar trend in all the groups through 2–4 dpi, maintaining strong chemokine production (from baseline), which is indicative of macrophage activation in lungs (Figure S4a). These results were further analyzed by the heatmap, which reveals distinct temporal and variant-specific immune profiles (Figure 3b). The fold change pattern reveals a distinct cytokine and chemokine profile in Omicron subvariant-infected mice, where an early immune response was observed, while WA1/2020-infected mice displayed a more gradual and steady immune response.

The observed differences in cytokine and chemokine levels between WA1/2020 and Omicron subvariant-infected mice can be attributed to the distinct replication dynamics and immune response of these viral strains. Elevated levels of GM-CSF, IL-6, IL-1β, and CXCL10 in Omicron-infected mice suggest enhanced early immune activation. Conversely, WA1/2020 maintained a more balanced cytokine and chemokine profile, consistent with its robust lung replication. These findings highlight the distinct immunopathological profiles of Omicron subvariants.

2.6. Nirmatrelvir Treatment Attenuates Inflammatory Cytokine and Chemokine Responses in Omicron-Infected Mice

To assess the immunomodulatory impact of Nirmatrelvir, we quantified cytokines and chemokines levels in lung homogenates of K18-hACE2 mice infected with Omicron subvariants (JN.1, LB.1, KP.3.1.1) and treated with Nirmatrelvir or vehicle. Results revealed distinct suppression of key inflammatory mediators with treatment. At day 4 post-infection, several key differences in lung cytokine and chemokine levels were observed between the vehicle-treated and Nirmatrelvir-treated groups (Figure 3c and Figure S4b). Nirmatrelvir treatment noticeably suppressed the pulmonary inflammatory response. The levels of GM-CSF, TNF-α, IL-6, MCP-1, CCL3, CXCL10, and IL-1β showed a marked decrease in Nirmatrelvir-treated mice. Among these, TNF-α was significantly reduced in JN.1-infected mice by a mean of ~10.1 pg/mL following Nirmatrelvir treatment (p = 0.011), while KP.3.1.1 and LB.1 showed no significant change (Figure 3c). A mean 3723 pg/mL reduction in MCP-1 levels was observed in JN.1-infected mice (p = 0.0017) upon Nirmatrelvir treatment. Though low, changes in MCP-1 levels in LB.1 and KP.3.1.1 were not statistically significant with mean differences of 1632 pg/mL and 2131 pg/mL, respectively (Figure 3c). CXCL10 was significantly suppressed in all subvariant-infected groups with Nirmatrelvir treatment, which was reduced by a mean of 3335 pg/mL in JN.1 (p = 0.0169), 2396 pg/mL in KP.3.1.1 (p = 0.0207), and 2362 pg/mL in LB.1 (p = 0.0326) in comparison to vehicle-treated infected controls. Similarly, Nirmatrelvir treatment robustly suppressed CCL3 across all groups, showing the strongest downregulation in JN.1-infected mice with a mean difference of 162.3 pg/mL (p < 0.0001), which was 81.88 pg/mL (p = 0.0364) in KP.3.1.1 and 90.57 pg/mL (p = 0.0133) in LB.1.

While GM-CSF and IL-6 displayed numerically lower levels in Nirmatrelvir-treated groups as compared to vehicle-treated, the differences were not statistically significant across any subvariant (Figure S4b). Other cytokines, including IL-1β and IL-10, also showed no significant change following treatment, suggesting their expression may be less sensitive to early Nirmatrelvir-mediated modulation (Figure S4b). The heatmap visualization (Figure 3d) further corroborated these findings, illustrating a general downregulation of the fold change in these inflammatory mediators in JN.1, KP.3.1.1, and LB.1 infected mice treated with Nirmatrelvir relative to vehicle-treated infected controls. These findings further support the antiviral potential of Nirmatrelvir and its potential to mitigate the inflammatory response associated with Omicron subvariant infection.

3. Discussion

The rapid emergence of SARS-CoV-2 Omicron subvariants has posed ongoing challenges for disease control due to their antigenic drift, altered pathogenesis, and immune evasion capabilities [4,5,6,13]. Our study demonstrates that the recent Omicron subvariants JN.1, LB.1, and KP.3.1.1 exhibit distinct infection kinetics and immune responses in the K18-hACE2 mouse model, highlighting variant-specific temporal patterns of viral replication. Importantly, our results show that these subvariants, despite their altered infection kinetics relative to the WA1/2020 strain, retain robust susceptibility to main protease-targeting antivirals such as Nirmatrelvir [16,17,18,19,20,21]. A key observation in our study is the accelerated viral kinetics of Omicron subvariants, which displayed peak replication in the lungs at 2–3 dpi, followed by a marked decline by 4 dpi. This contrasts with the WA1/2020 strain, which exhibited delayed peak replication and persistent viral RNA accumulation through 4 dpi. Additionally, a prior ex vivo study from our group demonstrated that Omicron subvariants showed tropism for bronchiolar epithelial cells in an air-liquid interface (ALI) model using human bronchial airway epithelial cells (hBAECs), where more robust infection was observed at 2–3 dpi compared with WA1/2020 [22].

Omicron-infected mice mounted an early and transient inflammatory response, specifically characterized by elevated GM-CSF, MCP-1, TNF-α, IL-6, and IL-1β levels [23,24,25,26]. Omicron subvariants exhibited early high levels of proinflammatory cytokine and chemokine levels at 2–3 dpi. This rapid immune activation is consistent with the reported upper respiratory tract tropism of Omicron, where robust innate immune sensing occurs early, leading to rapid cytokine production [11]. A similar trend was observed in the ALI model using hBAECs, where a more pronounced immune response was observed in Omicron-infected cells compared to the WA1/2020 strain at 2 and 3 dpi [22]. The early peak suggests that Omicron efficiently triggers pattern recognition receptors (PRRs) in airway epithelial cells and alveolar macrophages, leading to a surge in pro-inflammatory mediators [26,27]. However, the faster viral clearance in Omicron-infected mice is accompanied by a decline in immune activation, reflecting the transient nature of this response.

Infected mice exhibited early induction of select inflammatory mediators, including GM-CSF, MCP-1, TNF-α, IL-6, and IL-1β [23,24,25,26]. While this may reflect a more rapid activation of innate immune pathways by Omicron subvariants, the observed cytokine and chemokine elevations were transient and generally subsided by 4 dpi. The rapid immune kinetics may be attributable to the upper airway tropism of Omicron and the early triggering of pattern recognition receptors (PRRs) in epithelial cells and alveolar macrophages [11,26,27]. A similar trend was observed in the ALI model, where early immune gene expression was more pronounced in Omicron-infected hBAECs than WA1/2020-infected cells [22]. However, it is important to note that our study focused on early timepoints and cytokine measurements in lung homogenates. More comprehensive immunophenotyping, including cell-specific profiling and histopathological analysis, is needed to better understand the spatial and cellular dynamics of the immune response. Further studies incorporating lung tissue histology and flow cytometry will help clarify whether the early cytokine signals correspond to localized inflammation, immune cell infiltration, or tissue injury.

In contrast, WA1/2020-infected mice displayed a gradual increase in cytokine and chemokine levels, peaking at approximately 4 dpi. This delayed response aligns with the stronger lower respiratory tract tropism of WA1/2020, where viral replication in the lungs triggers a slower, more sustained inflammatory response. As WA1/2020 efficiently replicates in lung tissues, immune activation is progressively amplified, leading to peak cytokine levels later in the infection. Additionally, the slower replication dynamics of WA1/2020 may delay antigen presentation and immunity [9,14]. Thus, the temporal differences in immune activation reflect the inherent differences in viral replication kinetics, tissue tropism, and immune activation pathways between the parent strain (WA1/2020) and the rapidly replicating, immune-evasive Omicron subvariants.

The therapeutic arm of our study affirms that Nirmatrelvir remains broadly effective against contemporary Omicron subvariants. Early treatment with Nirmatrelvir (12 hpi) resulted in significant reductions in both viral titers and RNA levels in the lungs of infected mice, with particularly strong antiviral effects observed in JN.1 and LB.1 infections, reaching near-complete suppression of infectious virus. Interestingly, the antiviral efficacy was more pronounced in subvariant infections than in WA1/2020, potentially reflecting lower baseline replication or potential altered pharmacodynamics in the context of variant tropism [16,17]. Importantly, despite their extensive divergence in spike protein mutations, the Omicron subvariants JN.1, KP.3.1.1, and LB.1 retain the conserved P132H substitution within the main protease, a defining feature of the broader Omicron lineage. This mutation, though it changes the thermal stability of the enzyme, does not affect the binding activity of selective inhibitors against Mpro like Nirmatrelvir, MK-7845 [17,28,29]. Consistent with these findings, our in vivo data show that Nirmatrelvir retains potent antiviral efficacy against all three subvariants.

In addition to reducing viral burden, Nirmatrelvir treatment was also associated with a reduction in select pro-inflammatory mediators, including CXCL10, CCL3, TNF-α, and MCP-1, suggesting a possible immunomodulatory effect suggesting a possible immunomodulatory; however, it is important to note that this effect may be indirect and secondary to the observed reduction in viral replication. Our data do not distinguish between a direct suppression of inflammatory pathways and an immunologic response to lower viral burden. Therefore, while the reduction in inflammatory signaling may contribute to improved disease outcomes, this interpretation remains correlative and warrants further investigation. A more detailed mechanistic analysis, such as immune profiling and tissue-level evaluation, will be necessary to determine whether Nirmatrelvir exerts any direct immunomodulatory effects in vivo.

Critically, our findings validate the continued relevance of the K18-hACE2 mouse model as a preclinical platform for evaluating therapeutic efficacy across emerging SARS-CoV-2 variants. Despite varying disease manifestations, Omicron subvariants induced a sufficiently broad dynamic range in viral burden, allowing discrimination of therapeutic outcomes. The consistency in infection kinetics and therapeutic response among Omicron subvariants supports the utility of the transgenic mouse model for antiviral efficacy testing.

4. Materials and Methods

4.1. Cell and Virus Culture

Vero E6 (African green monkey kidney cells) expressing TMPRSS2 (transmembrane serine protease 2) were obtained from XenoTech, Okayama, Japan (JCRB1819, Lot no. 2222020). The cell monolayers were maintained at 37 °C with 5% CO₂ and 90% relative humidity, in high-glucose Dulbecco’s Modified Eagle Medium (DMEM, ATCC Cat. no. 30-2002, Manassas, VA, USA), supplemented with 10% fetal bovine serum (FBS USDA sourced, Cat. no. FB5002, Thomas Scientific, Swedesboro, NJ, USA) and antibiotic-antimycotic (AB/AM, Gibco, Grand Island, NY, USA; Cat. no. 15240062). The SARS-CoV-2 parent strain (Isolate USA-WA1/2020, Cat. no. NR-52281) was sourced from BEI Resources (NIH-supported program managed by ATCC). The SARS-CoV-2 Omicron strains were obtained from BEI Resources and through the Hackensack Meridian Health Bio-Repository (HMH-BioR). All HMH-BioR strains were recovered from nasopharyngeal swabs collected by the New Jersey Department of Health (NJDOH) surveillance program for COVID-19 and were confirmed as SARS-CoV-2 by whole-genome sequencing performed at the New York Genome Center for their unique mutations. Virus SARS-CoV-2 WA1/2020 and Omicron subvariants were propagated in VeroE6/TMPRSS2 cells to prepare stocks. The virus stock titers were determined by plaque assay as plaque-forming units per mL of culture as described previously [30]. The stocks used had virus titers of (i) WA1/2020: 5.7 × 10⁶ PFU/mL, (ii) JN.1: 2.4 × 10⁶ PFU/mL, (iii) LB.1: 2.5 × 10⁶ PFU/mL, and (iv) KP.3.1.1: 9.2 × 10⁵ PFU/mL.

4.2. In Vitro Antiviral Assay

Nirmatrelvir, used for in vitro and in vivo studies, was purchased from BOC Biosciences. (CAS. No. 2628280-40-8, Batch no. B22LN0.131). The compound in vitro potency was evaluated in VeroE6/TMPRSS2 cells as described previously [28,29]. For the cell-based assay, 10,000 cells per well were seeded in 96-well plates and treated with 3-fold dilutions of Nirmatrelvir 10 mM DMSO stock (36 µM to 0.0018 µM) for 2 h at 37 °C and 5% CO_2, where media only and uninfected cells served as controls. Following incubation with the compound, the cells were infected with each corresponding virus (WA1/20250, JN.1, LB.1, or KP.3.1.1) at 0.5 multiplicity of infection (MOI) and incubated immediately for 3 days under the conditions described above. At 72 h, 100 µL of CellTiter-Glo^® 2.0 reagent (Cat. no. G9243, Promega Corporation, Madison, WI, USA) was added to each well for ATP detection using the TECAN D300e digital dispenser. The half maximal effective concentration (EC₅₀) was calculated using the relative luminescence values with respect to uninfected controls.

4.3. In Vivo Studies

All animal procedures described in this study have been approved by the Institutional Animal Care and Use Committee (IACUC) at HMH-Center for Discovery and Innovation. Male and female transgenic B6.Cg-Tg(K18-ACE2)2Prlmn/J mice, 8–10 weeks old, were obtained from The Jackson Laboratories, Bar Harbor, ME, USA [16]. The animals were housed in individual ventilated caging (IVC) units and handled under sterile conditions for a minimum of 72 h for acclimation with food and water ad libitum in the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC) accredited the Center for Discovery and Innovation (CDI) Research Animal Facility.

Mice (n = 6; 3 male and 3 female) from each group were anesthetized by inhalation of vaporized isoflurane and intranasally infected with 50 µL of the virus stock at day 0, to a final infection dose per mouse of (i) WA1/2020: 2.85 × 10⁵ PFU, (ii) JN.1: 1.2 × 10⁵ PFU, (iii) LB.1: 1.25 × 10⁵ PFU and (iv) KP.3.1.1: 4.6 × 10⁴ PFU. The inoculation doses selected were based on the pilot infection studies performed to identify doses that reliably produced infection in mice, within the maximum intranasal volume limit of 50 µL from the available virus stock titer. Mice were monitored, and weights were recorded daily. At necropsy on days 2, 3, 4 post-infection, mice from each group were euthanized by carbon dioxide inhalation, and lung, brain, and nasal tissues were collected. Each infection study in mice was conducted independently.

For the Nirmatrelvir treatment study, mice were administered Nirmatrelvir at 1000 mg/kg b.w. (body weight) or vehicle (10 mL/kg) via oral gavage twice daily starting 12 hpi until day 3. On day 4 post-infection, mice from each group were euthanized by carbon dioxide inhalation, and lungs were collected. The left lung was weighed and homogenized in a GentleMACS M tube containing 2.5 mL of DMEM containing 2% FBS + 1× AB/AM. A clear supernatant was collected from homogenates upon centrifugation at 4000–5000 rpm for 10 min until clear. Cleared supernatants (0.5 mL) were used for PFU assay, and the remaining 0.25 mL was inactivated by Proteinase K for 45 min at 65 °C for qRT-PCR.

4.4. Viral Load Assessment

4.4.1. Plaque Assay

A total of 0.1 mL of the clarified homogenates suspended in DMEM was added to a 24-well plate containing 250,000 VeroE6 + TMPRSS2 cells per well, directly or 10-fold serial dilutions up to 10-6 prepared in DMEM. After inoculation, the 24-well plate is gently rocked for 1h at 37 °C, followed by the addition of 0.5 mL of a pre-warmed overlay mixture [2× MEM + 2.5% Cellulose (1:1)]. The plate was incubated for 72 h at 37 °C. The media was removed, and plates were fixed with 10% neutral buffered formalin and stained with 0.5% crystal violet and washed with water and enumerated.

4.4.2. RNA Extraction and RT-qPCR

Total RNA isolation of Proteinase K and heat-inactivated lung homogenate supernatants was performed using the Qiagen Qiacube HT automated mid- to high-throughput nucleic acid purification instrument with the QIAamp 96 Virus QIAcube HT Kit. qRT-PCR was performed on the samples using the E gene (from Charité/Berlin (WHO) protocol primer and probe panel) and RNase P gene (CDC kit) [31]. The primer and probe sequences used are: E gene forward 5′ ACAGGTACGTTAATAGTTAATAGCGT 3′; E gene reverse primer 5′ ATATTGCAGCAGTACGCACACA 3′; E gene probe 5′ FAM-ACACTAGCCATCCTTACTGCGCTTCG-BHQ-1 3′; RNAse P Forward 5′ AGA TTT GGA CCT GCG AGC G 3′; RNAse P Reverse 5′GAG CGG CTG TCT CCA CAA G 3′; and RNAse P Probe 5′ RNAse P Probe 3′.

4.5. Chemokine and Cytokine Analysis

Chemokine and cytokine levels of GM-CSF, TNF-α, MCP-1 (CCL2), IL-1β, IL-6, IL-10, CXCL10, and CCL3 were analyzed in collected lung tissue homogenates using the magnetic bead-based Mouse Luminex^® Discovery Multiplex Assay (Cat. no. LXSAMSM, RnD Systems, Biotechne, MN, USA). Homogenates were mixed with cell lysis buffer 2 (RnD systems, Minneapolis, MN, USA, Cat. no. 895347) in a 1:1 ratio for 30 min, and samples were centrifuged at 4500× g for 10 min at 4 °C. The assay was performed using the manufacturer’s instructions and read on a Luminex 200 xMAP platform (Austin, TX, USA). Cytokine/chemokine concentrations were calculated as picograms per milligram (pg/mL) of tissue using Belysa^® immunoassay curve fitting software version 1.2.

4.6. Statistical Analysis

The normalized data obtained from the in vitro assays were plotted as effective response vs. concentration. The EC₅₀ values were determined using a non-linear regression (curve fit), applying the equation for inhibitor vs. response variable slope (four parameters) and 95% confidence interval. One-way or two-way Analysis of Variance (ANOVA) with Tukey’s, Sidak’s, Dunnett’s, or Bonferroni post-hoc multiple comparisons was performed to compare the means of weight change, viral burden, and cytokine-chemokine levels among different experimental groups. To account for variability in data, all individual biological replicates were included in the analysis, as presented. Statistical analysis was performed using GraphPad Prism v10.4.1 software (GraphPad, San Diego, CA, USA). Spearman’s correlation and normality testing were conducted using R (version 4.4.3) with the corrplot and MVN packages. p values of <0.05 were considered statistically significant.

5. Conclusions

Collectively, this study advances our understanding of subvariant-specific infection dynamics and therapeutic susceptibility, emphasizing the importance of adaptable preclinical platforms and the continued susceptibility of direct-acting antivirals targeting conserved viral proteins. In the context of rapid viral diversification and declining effectiveness of neutralizing antibody strategies, small-molecule antivirals such as Nirmatrelvir remain critical pillars of SARS-CoV-2 management. Continued genomic surveillance, coupled with timely preclinical evaluation, is essential for informing therapeutic strategies against the evolving viral threats.