Cellular and molecular insights into the individual difference in COVID-19 mRNA vaccine responses
Article ID: 2023-008
Details
Article ID: 2023-008
The effectiveness of vaccines in preventing infection from specific pathogens is closely related to the magnitude of the antigen-specific T cell and antibody responses induced by the vaccine. These responses depend on the immune states of the individual, which is shaped by genetic and environmental factors. Recent studies using omics technologies identified immune cells, genes, and gut microbial species and metabolic pathways at baseline or during early vaccine responses as correlates of vaccine responses. These findings shed light on the molecular and cellular mechanisms underlying the inter-individual differences in vaccine responses. In this review, we provide an overview of correlates of recently developed COVID-19 mRNA vaccine-induced adaptive immune responses.
Recent studies focusing on individual differences in human vaccine responses revealed that multiple parameters at baseline and in early vaccine responses are associated with antibody and T cell responses induced by COVID-19 mRNA vaccination. Based on these findings, in this review, we discuss the potential molecular and cellular mechanisms underlying individual differences in COVID-19 mRNA vaccine responses.
Vaccines can provide protection against infection of specific pathogens by inducing adaptive immune responses, including long-lasting antibody and T cell memory responses. Studies using knockout mouse models demonstrated that vaccine-induced adaptive immune responses depend on genes involved in critical events in innate and adaptive immune responses, including recognition of vaccine adjuvants, inflammation, antigen presentation and effector T and B cell differentiation and functions [1,2,3,4]. In human, levels of vaccine-induced adaptive immune responses are remarkably different between individuals even under the same vaccination conditions [5]. Understanding the factors underlying individual differences in vaccine-induced adaptive immune responses may allow us to predict vaccine efficacy and therapeutically improve vaccine responses in low responders [5].
Studies over a decade have revealed that adaptive immune responses induced by various vaccines, including the recently developed COVID-19 mRNA vaccines, are influenced by baseline frequency of antigen-specific T and B cells, age, underlying diseases like diabetes, and sex differences [6,7,8,9,10]. In addition, recent studies, primarily using omics approaches, have shown that specific immune cell types, genes, and gut microbial species and functions at baseline or during early vaccine responses are associated with vaccine-induced adaptive immune responses. Those factors associated with responses to conventional vaccines, such influenza vaccines, were discussed in previous reviews [5, 11, 12]. In this review, we outline recent findings, including ours, on molecular, cellular, and microbial factors associated with adaptive immune responses to COVID-19 mRNA vaccine (Table 1).
| Categories | Factors | Correlation |
|---|---|---|
| Immune cell subsets | Early classical monocyte frequency | Positive correlation with antibody responses [16] |
| Late classical monocyte frequency | Positive correlation with T cell responses [17] | |
| Baseline MAIT cell frequency | Positive correlation with antibody and T cell responses [19] | |
| Cytokines | IFN-γ and IL-15 in the serum | Positive correlation with antibody responses [20] |
| CCL10 in the serum | Positive correlation with T-cell responses [20] | |
| Levels of IFNB1 mRNA induction induced by mRNA vaccine ex vivo | Positive correlation with T cell responses [17] | |
| Transcriptome | Baseline AP-1 transcription network module | Positive correlation with antibody responses [17] |
| Negative correlation with T cell responses [17] | ||
| Baseline expression of FOS, FOSB, and ATF3 and some AP-1 target genes | Negative correlation with T cell responses [17] | |
| Negative correlation with induction of IFNB1 expression ex vivo [17] | ||
| Gut Microbiome | alpha-diversity of gut microbiota | Positive correlation with antibody responses [28,29,30] |
| Baseline abundance of specific bacterial species in healthy people | Positive correlation with antibody responses [28,29,30] | |
| BCFA | Negative correlation with antibody responses [30] | |
| [taxon] Streptococcus; [metabolites] trimethylamine, isobutyrate, and omega-muricholic acid) in IBD patients | Positive correlation with baseline gut microbial parameters [31] | |
| [taxon] Staphylococcus; [metabolites] succinate, phenylalanine, taurolithocholate, and taurodeoxycholate in IBD patients | Negative correlation with baseline gut microbial parameters [31] | |
| Gut microbial fucose/rhamnose degradation activity | Negative correlation with T cell responses [17] | |
| Gut microbial fucose/rhamnose degradation and AP-1 expression in PBMCs | Positive correlation with expression of PTGS2 [17] |
MAIT: mucosal-associated invariant T; BCFA: branched chain fatty acids; IBD: inflammatory bowel disease; PBMC: peripheral blood mononuclear cell.
Human monocytes can be categorized into three distinct subsets based on their surface expression of CD14 and CD16: classical monocytes (CD14+ CD16−), non-classical monocytes (CD14− CD16+), and intermediate monocytes (CD14+ CD16+) [13]. Classical monocytes have higher capacity for phagocytosis and inflammatory response than other subsets [13]. Studies on murine counterpart subsets suggested that classical monocytes migrate to various tissues with or without inflammation, while non-classical monocytes crawl on the luminal side of endothelium to respond to tissue damage [14, 15]. In a systems biological assessment of BNT162b2 COVID-19 mRNA vaccine responses, Arunachalam et al. found that the vaccine-induced rapid increase of classical monocytes [16]. Importantly, the peak levels in classical monocyte frequency detected on day 2 after second booster dose were correlated with the levels of vaccine-induced neutralizing antibodies against SARS-CoV-2 spike antigen [16]. Additionally, our independent data showed that the frequency of classical monocyte frequency on day 40 after the second dose was positively associated with vaccine-induced T-cell responses [17]. Taken together, in response to COVID-19 mRNA vaccines, classical monocyte frequency dynamically changes, and its levels in early and late vaccine responses are positively associated with vaccine-induced antibody and T cell responses, respectively.
Mucosal-associated invariant T (MAIT) cells express invariant T cell receptors (TCRs) that recognize microbial metabolites, such as vitamin B2, presented by major histocompatibility complex (MHC)-related molecule 1 (MR1) and play a role in innate immune responses against bacterial infection [18]. Moreover, MAIT cells are involved in innate immune responses against viruses independently of antigen presentation with MR1 [18]. Boulouis et al. assessed relationship between MAIT cells and COVID-19 mRNA vaccine responses and found that baseline MAIT cell frequency was correlated with both antibody and T cell responses induced by BNT162b2 [19]. Using single cell (sc) RNA-seq analysis, another study revealed that vaccine-induced activation of MAIT cells marked by increased expression of genes involved in cell proliferation and tumor necrosis factor (TNF), which may promote B cell activation [19].
COVID-19 mRNA vaccination induces expression of a variety of cytokines, including interferons (IFNs), and chemokines which play roles in inflammation and adaptive immune responses [16]. BNT162b2-induced adaptive immune responses were positively associated with specific cytokines and chemokines in the serum, such as IFN-γ and interleukin (IL)-15 for antibody responses and CXC motif chemokine 10 (CXCL10) for T-cell responses [20]. Furthermore, our analysis of peripheral blood mononuclear cells (PBMCs) stimulated with BNT162b2 mRNA ex vivo for 6 hr revealed that levels of IFNB1 mRNA induction were positively associated with vaccine-induced T cell responses [17]. This finding is consistent with the observation that knockout mice of a cytosolic RNA sensor, melanoma differentiation-associated protein 5 (MDA5), induce neither type I IFNs (IFN-α and IFN-β) production nor antigen-specific CD8 T cells responses in response to BNT162b2 vaccination [21]. Thus, differences in the expression levels of specific cytokines and chemokines in the early stages of vaccine response likely contribute to individual variations in vaccine-induced adaptive immune responses [21].
Activator protein 1 (AP-1) transcription factors, including FOS, JUN, activating transcription factor (ATF), and musculoaponeurotic fibrosarcoma (MAF) family proteins, form homo- or hetero-dimers to regulate expression of various immune-related genes in accordance with the dimer composition [22]. Some of AP-1 factors have been demonstrated to play essential or modulatory roles in adaptive immune responses against infection and vaccination in animal models [23, 24]. Our transcriptome analysis of PBMCs revealed that the baseline AP-1 transcription network module is positively associated with vaccine-induced antibody responses and negatively associated with T cell responses (antigen-specific IFN-γ production) [17]. We also observed that baseline expression of FOS, FOSB, and ATF3 and some AP-1 target genes were negatively associated with T cell responses [17]. The AP-1 transcription network module appears to be downregulated upon BNT162b2 vaccination [16], and there was no association between AP-1 expression levels post vaccination and vaccine-induced T cell responses [17]. We also found that mRNA stimulation-induced induction of IFNB1 expression ex vivo is also inversely correlated with baseline expression of these AP-1 factors [17]. scRNA-seq data showed that FOS is expressed in all major immune cell populations in PBMCs and appeared to be negatively associated with early T cell response to ex vivo BNT162b2 mRNA stimulation [17].
Interaction between the immune system and gut microbiota can affect not only mucosal but also systemic immune responses [25,26,27]. It has been observed that diversity of gut microbiota in healthy people before and after COVID-19 mRNA vaccination is positively associated with vaccine-induced antibody responses, although the diversity is reduced by vaccination [28,29,30]. Furthermore, in healthy people, baseline abundance of specific bacterial species (i.e. Eubacterium rectale, Roseburia faces, Bacteroides thetaiotaomicron, Prevotella spp., Haemophilus spp., Veillonella spp. and Ruminococcus gnavus) were positively associated with COVID-19 mRNA vaccine-induced antibody responses [28, 30]. Furthermore, levels of gut microbial metabolites, branched chain fatty acids (BCFA), were negatively associated with the antibody responses [30]. In inflammatory bowel disease (IBD) patients treated with anti-TNF therapy, known to have abnormal gut microbiome, several other baseline gut microbial parameters were identified as positive correlates (i.e. [taxon] Streptococcus; [metabolites] trimethylamine, isobutyrate, and omega-muricholic acid) or negative correlates (i.e. [taxon] Staphylococcus; [metabolites] succinate, phenylalanine, taurolithocholate, and taurodeoxycholate) of COVID-19 mRNA vaccine-induced antibody responses [31].
In analysis of microbiome in our cohort, we discovered that gut microbial fucose/rhamnose degradation activity is negatively associated with COVID-19 mRNA-induced T cell responses [17]. Various bacteria species degrade mucosal fucose/rhamnose and serve as cross-feeders to other species in the microbiota [32]. This pathway produces short chain fatty acids (SCFAs), such as butylate, acetate, and propionate, which have immunoregulatory functions [32]. Furthermore, SFCA is known to promote expression of prostaglandin E2 (PGE2) [33], which in turn promotes expression of FOS [34]. Indeed, we demonstrated that stimulation with SCFAs and PGE2 induces expression of PGE2 and FOS, respectively, in PBMCs [17]. We also found that expression levels of cyclooxygenase 2 (COX2), a gene encoding an enzyme generating PGE2, were positively associated with fucose/rhamnose degradation in gut microbiota and AP-1 expression in PBMCs and were negatively associated with vaccine-induced T cell responses [17]. Thus, gut microbial fucose/rhamnose degradation may promote production of PGE2, which in turn promotes expression of AP-1 in PBMCs, thereby inhibiting vaccine-induced T cell responses.
The discovery of the factors associated with COVID-19 mRNA vaccine responses mentioned above has led to significant progress in understanding the cellular and molecular basis of individual differences in vaccine-induced adaptive immune responses. To utilize these factors as biomarkers to predict vaccine responses, validation of the findings in larger and diverse cohorts is needed. Another important issue is to understand whether and how these factors control vaccine-induced adaptive immune responses. Functional characterization of these factors in animal models of mRNA vaccination can be the first step in addressing this challenge. mRNA vaccine technology would be extensively utilized not only for protection against various infectious diseases but also for treatment of cancers. Better understanding of cellular, molecular, and gut microbial mechanisms underlying individual differences in mRNA vaccine responses would provide accurate prediction of vaccine responses and therapeutic targets for personalized vaccine strategies.
There are no conflicts of interest associated with this study.
We thank OIST Graduate University for its generous funding of the Immune Signal Unit. We also thank our laboratory members for valuable discussions.
