Despite widespread annual vaccination campaigns, influenza viruses continue to impose a significant public health burden, causing an estimated 290,000 to 650,000 deaths globally each year due to seasonal infections (1, 2). Multiple influenza vaccine strategies have been developed and routinely updated to match emerging viral strains; however, their overall effectiveness remains below 40% in the general population (3–5). The effectiveness is further diminished in susceptible populations, including children, older adults, and other immunocompromised individuals with chronic diseases, who are at greater risk of developing severe illnesses and complications due to influenza infection (6, 7). The emergence of novel pandemic influenza strains, generated through antigenic reassortment, poses a further substantial threat to global health due to the absence of pre‐existing antibodies or memory T cells (8). The COVID-19 pandemic heightened the global urgency of enhanced vaccine development, invigorating efforts to develop broadly protective next-generation influenza vaccines.

Influenza vaccine platforms have evolved substantially from traditional egg-based, inactivated virus formulations to highly adaptable recombinant protein and mRNA-based technologies, each offering distinct advantages in pandemic preparedness and public health impact. Next-generation mRNA-based influenza vaccines are currently undergoing clinical trials and have demonstrated superior efficacy compared to licensed vaccines (9–11). Nevertheless, recombinant subunit vaccines remain an important complementary approach due to their well-established, predictable safety profiles with less systemic reactogenicity and rare adverse events compared to mRNA/LNP vaccines, which can occasionally provoke stronger innate responses, fever, and rarely myocarditis (11–14). Most recombinant subunit influenza vaccines utilize the trimeric surface glycoprotein hemagglutinin (HA), by which the virus attaches to sialic acid-containing proteins on the host cell surface, and therefore a key determinant of virulence (15, 16). However, purified subunit vaccines are inherently less immunogenic, lacking pathogen-associated molecular patterns (PAMPs), and therefore, adjuvants have become an inevitable component in their formulations.

In addition to lowering the required vaccine antigen doses, adjuvants have a crucial role in shaping the resulting immune response by generating a mixed T helper cell response (Th1/Th2), thereby driving potent cellular and humoral responses. Currently licensed influenza vaccines typically employ alum or squalene-in-water (S/W) emulsion adjuvants such as MF59 and AS03 (3, 17, 18). Although alum is widely used, its limited efficacy and potential reactogenicity have been documented in pandemic influenza vaccines against H1N1 and H5N1 strains (19, 20). In contrast, currently approved S/W emulsions have demonstrated favorable safety profiles alongside enhanced immunogenicity; MF59 and AS03 have been successfully deployed in both seasonal and pandemic vaccines, eliciting stronger and broader immune responses against both homologous and heterologous influenza strains across general and high-risk populations (21, 22).

A squalane-in-water emulsion LiteVax™ adjuvant incorporating a synthetic carbohydrate fatty acid monosulphate ester (CMS) represents a promising candidate, currently advancing through phase I clinical trials for a licensed seasonal influenza vaccine (23–25). Beyond serving as an antigen delivery vehicle, the immunostimulatory properties of its components, including CMS, have been evaluated. Our previous studies have demonstrated that CMS robustly activates human antigen-presenting cells (APCs), with CMS-treated dendritic cells (DCs) capable of polarizing robust T helper responses (26). However, the nature of antigen-specific cellular responses generated by the adjuvant to influenza antigens and the precise receptors and molecular pathways engaged by CMS to activate APCs remain to be investigated.

Classically, only H1, H2, and H3 influenza A subtypes have been responsible for human pandemics. However, the emergence and spread of highly pathogenic H5N1 and H7N9 avian influenza viruses, both associated with severe human disease and high case-fatality rates, have raised substantial concern about their pandemic potential. Although recent H5N1 infections, particularly those caused by Clade 2.3.4.4b viruses, are receiving increased attention (27, 28), H7N9 has historically ranked among the highest-risk viruses in the United States Centers for Disease Control and Prevention’s Influenza Risk Assessment Tool because of its capacity to infect humans, cause severe illness, and persist in avian reservoirs (29). To date, there is no universally licensed, widely deployed H7N9 vaccine for humans, and available pre-pandemic candidates are generally poorly immunogenic without strong adjuvants, often requiring high antigen doses and multiple administrations while offering limited breadth against antigenically drifted variants and suboptimal durability of protection (30, 31). These limitations underscore an urgent need for improved, adjuvanted H7N9 vaccine formulations capable of inducing robust, long-lasting, and broadly protective immunity.

In this study, we investigate the immunomodulatory effects of the CMS adjuvant in human immune cells, with an emphasis on its capacity to enhance the immunogenicity of influenza HA antigens. We report that CMS robustly increases antigen-specific T cell responses to H1N1 HA peptide pools in human peripheral blood mononuclear cells (PBMCs). Although recombinant H7N9 HA antigens alone induced weak immunogenicity in DCs, co-formulation with CMS substantially boosted their immunogenic potential. We also examined the T-cell polarizing properties of this adjuvant formulation. Mechanistically, our data show that CMS triggers innate immunity via toll-like receptor (TLR)-mediated inflammatory signaling and promotes improved antigen presentation to T cells. Together, our findings indicate that combining recombinant HA proteins with the CMS platform may represent an effective future vaccine approach for emerging pandemic influenza strains.

The evolution of vaccination platforms from potent live attenuated/inactivated whole pathogens to safer and more economical subunit vaccines has taken a toll on their immunogenic potential, leading to the incorporation of adjuvants as an essential component of modern vaccines (38). The CMS adjuvant in the pipeline is a promising candidate to be added to the limited repertoire of safe and effective adjuvant systems approved for human use. The LiteVax™ formulation is a nano-emulsion composed of CMS, squalane, and Polysorbate 80. It forms droplets of approximately 80 nm. Its composition closely resembles widely used emulsion-based adjuvants such as MF59, which is a squalene-in-water emulsion stabilized by two non-ionic surfactants, Polysorbate 80 and Span 85.

However, it has been reported that MF59 adjuvant activity arises only from the complete formulation, as no single component exhibits intrinsic activity in isolation, either in an in vivo or in vitro setting (39). In contrast, CMS is a glycolipid similar to lipid A and monophosphoryl lipid A, and it shows inherent immunostimulatory properties. The LiteVax™ formulation combines these properties, where the emulsion serves as the delivery system of CMS, and CMS acts as the immunostimulatory agent, resulting in strong synergistic collaboration (25). Incorporating TLR agonists into oil-in-water emulsions has been demonstrated to enhance humoral and cellular immunity in preclinical studies (40–43), underscoring the translational potential of the concept.

CMS adjuvant formulation has been shown to augment immune responses to inactivated whole virus A/H5N1 and A/H7N9 vaccines in ferrets, toward a recombinant malaria antigen in rabbits (44), and more recently against the Receptor Binding Domain (RBD) spike protein antigens derived from SARS-CoV-2 in mice, with higher efficacy than comparator adjuvants tested in parallel (45). Preclinical toxicity assessments in ferrets and Phase I clinical trials have established the safety and tolerability of CMS with licensed inactivated seasonal influenza vaccines (46–48). CMS-adjuvanted low-dose vaccine (1/5^th antigen dose) gave comparable humoral and cell-mediated immunogenicity as the standard full-dose unadjuvanted vaccine, demonstrating efficient dose-sparing potential (46). A Phase Ib clinical trial with full doses of vaccine with or without CMS in younger and older, healthy adults revealed an acceptable safety profile in both cohorts and enhanced humoral immune responses in older adults, highlighting the potential of this adjuvant to counteract immune senescence (48).

In this report, we have investigated the immunological and molecular mechanisms underlying CMS adjuvanticity in primary human immune cells using various in vitro systems. The ability of CMS to enhance antigen-specific CD4⁺ and CD8⁺ T cell responses to H1N1 HA peptide pools in human PBMC assays recapitulates the strong humoral and cellular responses observed in clinical trials with low-dose adjuvanted influenza vaccine (46). While antigenic peptides alone failed to significantly expand CD8⁺ T cells, combining with the adjuvant substantially increased their frequency, pointing to the ability of CMS to promote cross-presentation and cytotoxic T lymphocyte activity. For a long time, the primary goal of vaccination strategies for respiratory viruses focused on the induction of neutralizing antibody responses. The recent severe SARS-CoV-2 variants shifted our attention to CD8⁺ T cells, with growing evidence showing that even the variants that escaped from neutralizing antibodies could not escape CD8⁺ T cells induced by mRNA and adenoviral vector vaccines (49). Influenza viruses are also typically known to escape from neutralizing antibodies via rapid mutations, and several studies have provided evidence for the protective role of cross-reactive CD8⁺ T cells against pandemic flu (50).

These observations suggest that adjuvants capable of eliciting robust CD8⁺ T cell responses might be particularly valuable for vaccines targeting highly variable respiratory pathogens. Our findings show that CMS drives strong antigen-specific CD4⁺ and CD8⁺ T cell responses in vitro, raising the possibility that CMS-adjuvanted recombinant influenza vaccines could enhance the magnitude and breadth of T cell immunity compared with antigen alone.

However, we did not assess in vivo protection or neutralizing antibody titers in this study. Therefore, any effect of CMS on vaccine efficacy, durability, or clinical benefit remains speculative. Future in vivo challenge studies, including detailed evaluation of neutralizing antibodies and cross-reactive CD8⁺ T cells, will be necessary to determine whether CMS adjuvantation can enhance and sustain protective immunity against seasonal and potentially pandemic influenza strains.

Further, the adjuvant capacity of CMS has been tested against recombinant flu HA protein antigens of the H7N9 strain in human APCs in vitro. Though an earlier study reported that recombinant HA proteins of H1N1 and H5N1 Influenza A viruses effectively promoted the activation and maturation of mouse myeloid DCs (42), experiments with human DC-based systems (DC-cell line MUTZ-3 and monocyte-derived DCs) (51) revealed the inefficacy of subunit vaccines of H5N1 to induce DC maturation, in contrast to whole inactivated virus (52). Consistent with these data, we demonstrated that H7N9 HA alone failed to activate human DCs. Combined with CMS, however, H7N9 HA elicited potent activation of human APCs, as evidenced by the enhanced expression of co-stimulatory molecules and release of cytokines. Fluorescence microscopy confirmed active endocytosis of the antigen-adjuvant complex by DCs, although CMS did not directly increase HA uptake. Though some adjuvants have the capacity to enhance antigen uptake by APCs, others, including CMS, mainly function through stimulating innate immunity and optimizing antigen presentation to generate robust adaptive immune responses (53).

Next-generation vaccine platforms utilize various advanced technologies to design antigens with precise epitope selection, aiming to elicit targeted cellular responses and to shape the T-cell repertoire. The choice of adjuvant also plays a pivotal role in this aspect, as different adjuvants have distinct effects on the polarization of T-cell responses. For example, adjuvants such as CpG, Adjuvant System (AS; AS01, AS03, AS04), or mRNA-lipid nanoparticles tend to promote strong Th1-oriented immunity, whereas alum or MF59 more commonly induce Th2-biased responses (54, 55). In our experiments, DCs stimulated with CMS supported the polarization of Th1, Th2, Th17, and FoxP3⁺ CD4⁺ T cell subsets using autologous total CD4⁺ T cells. This broad induction of T helper lineages demonstrates the capacity of CMS to engage multiple CD4⁺ differentiation pathways, potentially supporting both cytotoxic and humoral immunity. The induction of FoxP3⁺ regulatory T cells alongside effector T cells could be beneficial by limiting excessive inflammation and thereby reducing reactogenicity, which may enhance the safety profile of adjuvanted vaccines (56). On the other hand, vaccine antigen–specific Tregs can dampen responses to conserved epitopes (57), potentially complicating heterosubtypic protection. However, FoxP3 expression in human CD4⁺ T cells at early stages of activation can also reflect transient activation of conventional T cells rather than stable regulatory T cells (58), and functional suppressive activity of the FoxP3⁺ CD4⁺ T cells population was not assessed in this study. Thus, while CMS clearly promotes a diverse phenotypic T helper response, the precise regulatory capacity of the FoxP3⁺ CD4⁺ T cell-and its biological significance in shaping vaccine-induced immunity-will require further functional and longitudinal investigation.

Despite strong evidence for the induction of immune responses by CMS in in vivo and in vitro systems, the receptors and the downstream signaling pathways involved in this process were not described. TLRs play a major role in innate immunity by responding to various pathogen-derived ligands, thereby activating various pro-inflammatory signaling pathways (59). Our transcriptomic profiling showed that stimulation of human DCs with CMS led to strong activation of innate immune signaling pathways that result in the upregulation of NF-κB, IL-1, MAPK, JAK-STAT, and Interferon pathways. Engagement of TLR-2/4 triggers MyD88/TRIF pathways that activate NF-κB and MAPKs, leading to rapid transcription of pro-inflammatory cytokines and co-stimulatory molecules as observed in our analysis. The concurrent induction of JAK-STAT and interferon pathways suggests that CMS also engages in the induction of interferons (Supplementary Figure 4). Thus, CMS appears to mimic key infection signals and play a key role in DC maturation by activating pro-inflammatory cytokines and co-stimulatory molecules essential for initiating adaptive immune responses.

Our RNA-seq analysis is based on a relatively small cohort (n = 3 donors per group), which represents a limitation of the study. This dataset was however generated as a mechanistic discovery cohort rather than a population-level study, and the transcriptomic results are therefore interpreted as hypothesis-generating and supported by complementary experimental and functional evidence presented in this report. Furthermore, principal component analysis (PCA) revealed clear separation between CMS and control samples along PC1, which explained most of the variance, indicating that biological condition was the dominant driver (Supplementary Figure 5). The within-group dispersion observed along PC2 likely reflects inter-donor variability rather than technical batch effects, particularly given the small sample size and uniform experimental processing. Consequently, no batch correction was applied to avoid removal of biological signal. Future studies incorporating larger and independent experimental cohorts would be valuable to further validate and extend these observations.

Key hub genes identified in the protein-protein interaction network (Figure 6A) further highlighted the immunostimulatory effects of CMS. These genes play central role in DC functions: NFKB1 - a master regulator of inflammatory genes (60); IL-1B encodes for IL-1β, a potent pro-inflammatory cytokine produced by activated DCs (61); CXCL10, an interferon inducible chemokine recruits CXCR3⁺ effector T cells and is characteristic of Th1-type responses (62); IRF1 drives the antigen presentation by amplifying the MHC expression (63), which is consistent with our observation of increased MHC-1/II expression in CMS-treated DCs. Upregulation of both the MHC classes substantiates our in vitro data showing that CMS enhances the ability of DCs to present antigens to both CD4⁺ and CD8⁺ T cells. Collectively, the enhanced antigen presentation signature underscores CMS’s ability to strengthen the innate and adaptive interface, which is a prerequisite for robust vaccine responses (Figure 8).

We have previously shown that the monosulphate group in CMS is indispensable for the adjuvanticity of CMS in DCs (26). Some endogenous ligands containing a sulfate group, namely heparan sulfate and sulphatides, are reported to act as TLR4 agonists (64, 65). The binding of the sulfate group with TLR4 is also confirmed by the structural analysis of the binding of sulfatide with TLR4 (65), which activates MyD88 and TRIF-dependent signaling (66). Our experiments with THP-1 reporter assays demonstrated the activation of the NF-κB pathway in response to CMS, especially with the overexpression of the MD2-CD14-TLR4 receptor complex. Similar experiments were carried out previously with HEK reporter cells overexpressing the TLR4 receptor complex, which did not show any reporter activity with CMS stimulation (44), indicating that endocytic processing by APCs might be a prerequisite for TLR engagement by CMS. Functional blocking studies with TLR antagonists confirmed the role of TLR2 alongside TLR4 in CMS-induced activation of DCs. Although transcriptomic analysis revealed upregulation of TLR1, TLR2, and TLR6, the functional involvement of specific TLR2 heterodimers (TLR1/2 versus TLR2/6) remains unresolved. Heterodimer-specific neutralization assays will be required to address this question in future studies.

TLR4 and TLR2 both promote DC maturation by upregulating expression of CD80, CD86, and MHC-I/II. However, due to the TRIF pathway activated by TLR4, it gives a strong antiviral response, pushing DC toward a Th1 immune profile and enhancing cross presentation with CD8⁺ T cells (67). While TLR2 gives a balanced immune response without strongly polarizing between Th1 and Th2. Experimental data have shown that compared to LPS-treated DC, TLR2 agonists such as SUP1 boost DC antigen presentation to CD4⁺ T cells and enhance memory cell formation against the antigen (68). Moreover, TLR4 leads to sustained IL-10 production as the TRIF pathway has been shown to stabilize the IL-10 mRNA (69). Thus, in vaccine terms, concomitant engagement of TLR4 and 2 becomes more beneficial. These findings position CMS as a promising, compatible adjuvant that efficiently activates DCs via TLR-signaling and interferon pathways, augmenting cytokine and co-stimulatory molecule expression alongside antigen processing gene upregulation to mount a robust interface between innate sensing and adaptive immunity.

Our current study is limited to evaluating CMS-induced responses in PBMC-derived immune cell populations from healthy donors, which does not fully recapitulate the spatial and cellular complexity of immune activation in vivo. While these in vitro data provide initial mechanistic insight into early innate sensing and downstream activation, validation in relevant in vivo challenge models, such as mice or ferrets, will be essential to determine whether the CMS adjuvant enhances protective immunity and promotes robust neutralizing antibody responses. Such models will enable assessment of antigen-presenting cell activation, trafficking to draining lymph nodes, and the induction of germinal center reactions that support durable humoral immunity. Furthermore, single-cell transcriptomic profiling of PBMCs from clinical cohorts could provide an unbiased view of responsive innate subsets and T-cell polarization states. Together, these complementary approaches will enable validation of the primary target cell populations and elucidation of the receptor-mediated mechanisms of CMS within a more physiologically relevant setting with larger sample sizes, thereby strengthening the translational relevance of our findings and informing the rational optimization of CMS-based vaccine formulations.

In conclusion, this study provides proof of concept for the adjuvant activity of CMS against influenza HA antigens in in vitro immune cell systems. CMS induces strong antigen-specific T cell responses, especially CD8⁺ T cell responses to HA antigens in human PBMCs. Furthermore, CMS enhances the immunostimulatory potential of recombinant H7N9 HA antigen on APCs and induces a mixed effector T cell polarization. Mechanistically, CMS activates the TLR4 and TLR2-associated signaling, leading to cytokine/chemokine secretion that promotes innate immune cell recruitment, activation, and antigen presentation to T cells, thereby supporting strong humoral and cellular immune responses. Further evaluation of the adjuvant antigen formulation in appropriate animal models of various age groups (70) is warranted to assess the generation of neutralizing antibodies against the recombinant vaccine antigens and protection against influenza infection. Our results underscore the potential of the CMS recombinant HA antigen formulation for further clinical development and help in the design of effective novel vaccine adjuvant formulations to protect the vulnerable population from influenza-related complications.