A multivalent mRNA-LNP vaccine protects against Clostridioides difficile infection

Abstract

Clostridioides difficile infection (CDI) is an urgent public health threat with limited preventative options. In this work, we developed a messenger RNA (mRNA)-lipid nanoparticle (LNP) vaccine targeting C. difficile toxins and virulence factors. This multivalent vaccine elicited robust and long-lived systemic and mucosal antigen-specific humoral and cellular immune responses across animal models and independent of changes to the intestinal microbiota. Vaccination protected mice from lethal CDI in both primary and recurrent infection models, and inclusion of non-toxin cellular and spore antigens improved decolonization of toxigenic C. difficile from the gastrointestinal tract. Our studies demonstrate mRNA-LNP vaccine technology as a promising platform for the development of novel C. difficile therapeutics with potential for limiting acute disease and promoting bacterial decolonization.

One-Sentence Summary:

mRNA-LNP vaccines targeting Clostridioides difficile virulence factors across its lifecycle elicit systemic and mucosal immunity in multiple animal models and protect against lethal infection.

Main Text:

Clostridioides difficile causes a wide range of gastrointestinal disorders varying from mild diarrhea to toxic megacolon and death (1–3). In the US, C. difficile is one of the most reported nosocomial pathogens, and C. difficile infection (CDI) is a major public health threat worldwide (4). Previous vaccine trials for CDI prevention focused on adjuvant-supplemented toxoid or recombinant vaccines targeting the combined repetitive oligopeptide (CROP) domains or receptor binding domains (RBDs) of two potent C. difficile toxins, TcdA and TcdB, to inhibit toxin binding to intestinal epithelial cells (5). A recent Pfizer phase 3 clinical trial showed promise in reducing the severity and duration of CDI but failed to prevent initial infection, highlighting the challenges in developing effective C. difficile vaccines (NCT03090191). This setback, along with Sanofi discontinuing its C. difficile vaccine (NCT01887912), underscores the need to explore alternative vaccination strategies (6).

Nucleoside-modified messenger RNA (mRNA) vaccines have emerged as a leading platform against a myriad of infectious pathogens, including lethal bacterial pathogens (7–9). We hypothesized that a multivalent mRNA strategy targeting multiple virulence factors would prevent C. difficile-associated disease. Our initial vaccine strategy included RBDs and CROP domains of both toxins, and the metalloprotease virulence factor Pro-Pro endopeptidase 1 (PPEP-1/Zmp1) (fig. S1A). Since PPEP-1 is a highly conserved factor that modulates pathogen motility and adhesion through the cleavage of multiple factors on the C. difficile cell surface, we reasoned that targeting a fitness and virulence regulator would promote decolonization after disease onset (10).

Genetically diverse strains of C. difficile circulate in both hospital and community settings (11). To explore the conservation of our vaccine targets, we conducted a comparative genomic analysis of shared amino acid sequence identity between the TcdA, TcdB, and PPEP-1 in our mRNA constructs and 137 unique representative C. difficile strains (Fig. 1A) (12). We observed that our construct for PPEP-1 has a minimum of 97.4%, a median of 99.5%, and an average of 99.4% amino acid similarity across the C. difficile strains examined (Fig. 1A). Among toxigenic strains, TcdA has a minimum of 39.4%, a median of 97.1%, and an average of 90.9% identity, and TcdB has a minimum of 87.0%, a median of 100%, and an average of 97.8% amino acid identity (Fig. 1A). Together, our constructs cover the diversity of clinically relevant and potentially emergent strains and represent strong candidates for further development. Resulting mRNA sequences were synthesized and encapsulated into LNPs (13–15). All mRNA-LNPs passed quality control parameters (Table S1, S2) and all three proteins were expressed and secreted following transfection of Neuro2A cells (fig. S1B). PPEP-1 expression was notably reduced compared to the toxins (fig. S1B). In vitro expression of TcdA and TcdB mRNA-LNPs did not reduce cell viability, showing safety of engineered immunogens for intramuscular administration into animals (fig. S1C).

Fig. 1. mRNA-LNP vaccines elicit robust immune responses against C. difficile virulence factors.

(A) Maximum likelihood phylogenetic tree illustrating the amino acid sequence identity of TcdA, TcdB, and PPEP-1 between mRNA constructs and 137 C. difficile strains. (B-E) Mice were immunized intramuscularly (i.m.) once (C) or twice (B), (D), (E) with 1 or 5μg of bivalent mRNA-LNPs (TcdA/TcdB) (green), trivalent mRNA-LNPs (TcdA/TcdB/PPEP-1) (purple), or trivalent recombinant protein with alum (grey). Age-matched naïve mice served as unvaccinated controls. Two weeks after boost, TcdA (left), TcdB (middle), and PPEP-1 (right) specific antibodies in sera were measured by ELISA. Two weeks after prime, total number of T follicular helper (Tfh) cells in draining lymph nodes (dLN) (C) were measured by flow cytometry. (D, E) Two weeks after boost, total germinal center (GC) B cells (D) and antigen-specific B cells (E) were measured in spleen by flow cytometry. (F and G) Two weeks after immunization, mice were challenged intraperitoneally (i.p.) with 625ng recombinant TcdA (F) or 125ng recombinant TcdB (G) and monitored for survival. (B-G) n = 5–14 mice per group, 2 independent experiments. Data are represented as mean ± SEM with fold change (B), mean ± SD (C-E), or percent survival (F, G). Statistics by one-way ANOVA with Tukey’s multiple comparisons (C-E) or Log-rank (Mantel-Cox) test (F, G). Stats in (F) and (G) shown between 5μg mRNA-LNP vaccine against other conditions. *p <0.05, ** p<0.01, *** p<0.001, **** p<0.0001.

mRNA-LNP vaccination induces robust immune responses

Multivalent mRNA-LNP vaccines elicit high antibody titers against multiple antigens simultaneously (16, 17). To examine the immunogenicity of our vaccines, mice were vaccinated with TcdA/TcdB (bivalent) or TcdA/TcdB/PPEP-1 (trivalent) mRNA-LNPs, trivalent recombinant protein with alum as an adjuvant, or luciferase mRNA-LNP, and serum antigen-specific antibody titers were measured by endpoint ELISA (Fig. 1B). mRNA-LNP vaccines induced ~2–4-fold higher anti-toxin immunoglobulin G (IgG) responses compared to recombinant protein + alum (Fig. 1B). IgG against PPEP-1 was also detected at lower levels compared to anti-toxin IgG (Fig. 1B). Notably, anti-PPEP-1 IgG was significantly higher with mRNA-LNP vaccine compared to recombinant protein + alum (Fig. 1B). We observed a modest dose-dependency (fig. S2A), and IgG titers significantly increased following a second vaccination (fig. S2B, C). IgA responses in circulation were significantly lower for all immunogens and required at least two immunizations (fig. S3). Monovalent, bivalent, and trivalent mRNA-LNP vaccines showed no reduction in antibody titers with increased valency, corroborating previous results with multivalent influenza vaccines (fig. S4) (16, 17). Finally, antibody responses were independent of mouse genetic background (fig. S5). Together, these data demonstrate the robust capacity of mRNA-LNP vaccines to elicit antigen-specific immune responses to C. difficile virulence factors.

T follicular helper (Tfh) cells drive germinal center (GC) reactions and are important for potent antibody induction (18, 19). Vaccination with mRNA-LNP vaccines led to a significant increase in total Tfh cells (Fig. 1C and fig. S6A) and, given their role in supporting B cell responses in the GC, we hypothesized this would correlate with strong GC B cell responses in the spleen and the draining lymph nodes (dLNs) (Fig. 1D, E and fig. S6 and S7). Indeed, mRNA-LNP vaccines induced a dose-dependent, antigen-specific GC B cell response (Fig. 1D, E and fig. S7). In accordance with antibody titers, trivalent mRNA-LNP vaccine induced PPEP-1-specific B cells to a lesser extent compared to anti-toxin-specific B cells (Fig. 1E and fig. S7B). Despite modest increases in IgG antibodies compared to alum-adjuvanted recombinant proteins, GC B cells and antigen-specific B cells were markedly higher in the mRNA-LNP groups (Fig. 1D, E and fig. S7). These data demonstrate that mRNA-LNP vaccination against C. difficile virulence factors induces robust Tfh and GC B cell responses that are associated with potent and long-lived responses (18, 19).

Additionally, immunogenicity studies in hamsters showed that the vaccine was well tolerated, with no observed weight loss or adverse effects (fig. S8A, B). A single immunization induced robust anti-toxin systemic IgG levels (fig. S8C). A second immunization was necessary to elicit anti-PPEP-1 IgG response (fig. S8B), supporting our observations in mice (fig. S2C). In contrast, a single immunization with recombinant protein + alum was insufficient to induce anti-TcdB IgG response (fig. S8C). These data demonstrate the capacity of mRNA-LNP vaccines to drive robust antigen-specific immune responses to C. difficile virulence factors in multiple relevant animal models.

mRNA-LNP vaccination activates all arms of adaptive immunity, including antigen-specific CD4⁺ and CD8⁺ T cells. CD4⁺ T cell responses and skewing are essential for highly neutralizing antigen-specific antibody responses (20). To assess T cell responses, we stimulated splenocytes with overlapping peptide pools covering TcdA, TcdB, or PPEP-1 (fig. S9, 10). T cell responses increased with vaccine dose and were significantly higher in mRNA-LNP vaccines compared to alum-adjuvanted recombinant protein vaccines (fig. S9B, C). Interestingly, we observed higher frequencies of antigen-responsive T cells with increasing vaccine valency (fig. S9B, C). TcdA and TcdB peptide-responsive CD4⁺ T cells were higher than CD8⁺ responses, characterized by increased IL-2 and TNFα expression (fig. S9B, C). In contrast, PPEP-1 showed poor CD4⁺ responses but higher CD8⁺ responses (fig. S9B, C). Similar results were obtained with full recombinant proteins (fig. S9D, E). Polyfunctional T cells are associated with improved protection during infection and can support a more robust immune response (21–23). Following multivalent mRNA-LNP vaccination, the frequency of polyfunctional T cells was high, with double-positive being the most abundant followed by triple-positive cells (fig. S10). These data show that mRNA-LNP vaccines against C. difficile virulence factors provide improved immune responses compared to alum-adjuvanted recombinant vaccines, further supporting their viability for clinical development.

mRNA-LNP vaccines protect against C. difficile toxins

C. difficile-associated disease is primarily driven by the effects of the toxins (24, 25). To assess whether anti-toxin antibodies elicited by mRNA-LNP vaccines would effectively neutralize C. difficile toxins in vivo, we immunized mice with monovalent TcdA or TcdB mRNA-LNP vaccines and challenged them with five-times the LD₁₀₀ of intraperitoneal recombinant TcdA TcdA or TcdB and monitored survival (Fig. 1F, G and fig. S11A, B). Mice vaccinated with recombinant TcdA or TcdB, without adjuvant, all died within 1 day-post toxin challenge (Fig. 1F, G). The inclusion of alum as an adjuvant improved survival by only 20%; however, immunization with monovalent TcdA or TcdB mRNA-LNP vaccines improved survival to 100% (Fig. 1F, G). Notably, serum transfer from mice vaccinated with mRNA-LNPs was sufficient to neutralize toxin and protect from lethal challenge (fig. S11C, D). Collectively, these data show the efficacy and potency of mRNA-LNP-stimulated circulating, neutralizing antibodies and highlight this platform’s value in targeting bacterial proteins and toxins.

mRNA-LNP vaccines do not compromise intestinal microbiota

The gut microbiota is important for health and providing resistance to invading pathogens, including C. difficile, and antibiotics are the primary risk factor for CDI (26–28). Since the effect of vaccines targeting enteric pathogens on the resident gut microbiota has not been explored, we examined the composition of the gut microbiota following vaccination and antibiotic treatment (fig. S12). α and β diversity of the fecal microbiota was not affected, showing that immunization does not enhance antibiotic-mediated perturbation or impact microbiome stability (fig. S12A–G).

Vaccination did not lead to significant shifts in relative abundances of any taxa, including those known to antagonize C. difficile, such as the Clostridiales family (fig. S12H). These data show the safety of our approach and vaccine targets in a mouse model; however, additional testing in dirty mice and other models is warranted to confirm our findings and recapitulate the immune responses we report.

mRNA-LNP vaccines protect from severe CDI

We next investigated whether mRNA-LNP vaccination was protective in a murine model of infection (Fig. 2A–E) (27–30). Following infection with 20-times the lethal dose of C. difficile (strain VPI10463), vaccinated mice were fully protected from mortality, while all unvaccinated mice were moribund by day 2 post-infection (Fig. 2B). Vaccinated mice presented with mild disease, lost an average of 5% body weight, and all mice were alert and active following infection (Fig. 2C, D). Notably, we found that a single low dose vaccination was sufficient to protect mice from lethal CDI (fig. S13A–D). Additionally, vaccination against C. difficile toxin was necessary for protection, as immunization with PPEP-1 mRNA-LNP alone did not protect mice (fig. S13E–F). We assessed disease pathology and found that CDI induces edema, inflammation, and epithelial damage that is independent of vaccination status, suggesting that while vaccination reduces the effects of acute infection, it is insufficient to protect the intestinal epithelium from the local cytotoxic effects of TcdA and TcdB in this mouse model (Fig. 2E). Furthermore, this suggests that translocation and systemic effects of the toxins may play a key role in lethal disease and that vaccine-mediated systemic immunity plays an important role in protection against lethality during CDI.

Fig. 2. mRNA-LNP vaccines protect against lethal CDI and induce mucosal immunity.

(A) Schematic of experimental design. Mice were immunized i.m. with 1μg bivalent or trivalent mRNA-LNPs, treated with cefoperazone (abx) in their drinking water, and challenged with 10,000 spores of C. difficile (VPI 10463). (B - D) Mice were monitored for survival (B), weight loss (C), and clinical sickness (D) over the course of infection. (E) Ceca pathology scores on day 2 post-infection. (F and G) C. difficile colony-forming units (CFUs) (F) and C. difficile toxin titers (G) in stool over time. Dotted lines indicate assay limit of detection (LoD). (H and I) Mucosal IgA (H) and IgG (I) titers 2 weeks after boost or infection. For C.d only, mice received cefoperazone followed by 100,000 spores of CD196. Vaccine only received two 1μg trivalent mRNA-LNP immunizations. Vaccine + C.d, immunized then infected with VPI 10463. C.d + vaccine, infected with CD196 then immunized twice with 1μg trivalent mRNA-LNP. ND, not detectable. n = 4–5 mice per group, 2 independent experiments. Data are represented as mean ± SD (C-E), mean ± SEM (F-G), or box and whiskers (H-I). Statistics by Log-rank (Mantel-Cox) (B), two-way ANOVA with Tukey’s multiple comparisons (C-D), Mann-Whitney (F-G) or Kruskal-Wallis multiple comparisons test. *p <0.05, ** p<0.01, *** p<0.001, **** p<0.0001 vs unvaccinated.

To define the effects of vaccination on C. difficile persistence and virulence, we evaluated pathogen burden and toxin titers in the stool of mice during infection (Fig. 2F, G). All vaccinated mice cleared C. difficile from the GI tract at similar rates, as measured by colony forming units (CFUs) (Fig. 2F). However, trivalent vaccination resulted in increased clearance of toxin titers in the stool compared to bivalent vaccination (Fig. 2G). This indicates that despite low expression (fig. S1B) and low immunogenicity (Fig. 1B, E), immunization against PPEP-1 has an added benefit in reducing C. difficile pathogenesis (Fig. 2G). Together, these data show that a multivalent strategy can promote enhanced clearance from the gut, and we postulate that this response can be improved with greater immunogenicity of PPEP-1. We also observed that bivalent and trivalent vaccines protected mice with a phylogenetically divergent ribotype 027 strain of C. difficile (fig. S14), further validating high conservation of our vaccine targets and broad protection conferred by vaccination.

mRNA-LNP vaccines against C. difficile prompt mucosal immunity

Generation of antigen-specific mucosal immunity remains a primary challenge for developing an effective C. difficile vaccine. Recent reports have shown that TcdB aids C. difficile in subverting the immune system (31). Indeed, we find that naïve mice mount a poor natural humoral immune response to C. difficile toxins (Cd only) at day 14 post-infection (Fig. 2H, I and fig. S15), whereas vaccination elicited significant anti-toxin IgG and IgA titers in stool 14 days post-boost (vaccine only) (Fig. 2H, I and fig. S8D). Notably, vaccination alone was sufficient to mount robust anti-toxin mucosal antibodies but insufficient to elicit anti-PPEP-1 stool antibodies (Fig. 2H, I and fig. S8D). However, when we examined antibody titers in the stool of vaccinated mice 14 days post-infection (vaccination + Cd) we observed a 4–10 fold increase in anti-toxin antibodies compared to vaccination alone and observed anti-PPEP-1 IgA (Fig. 2H, I), suggesting that hybrid immunity elicited by vaccination and infection enhances mucosal immunity against PPEP-1.

Individuals with prior CDI are at an elevated risk of relapsing or recurrent infections, making this population ideal targets for vaccination. To determine the ability of mRNA-LNP vaccines to elicit mucosal antibodies when administered post-infection, we vaccinated mice after recovery from a sublethal infection (Cd + vaccination) (Fig. 2H, I). At day 14 post-boost, these mice had ~100–300-fold increase in anti-toxin IgA compared to infection alone, and a 1–4-fold increase compared to vaccination alone (Fig. 2H). Infection prior to vaccination also led to a 60-fold increase in anti-PPEP-1 IgA in stool (Fig. 2H). Similar trends were also observed in mucosal anti-PPEP-1 IgG (Fig. 2I), and vaccination also elicited antigen-specific mucosal antibody responses in hamsters (fig. S8D). Together, these data show that mRNA-LNP vaccination leads to mucosal immunity to C. difficile toxins and this vaccine strategy primes the immune response for mucosal antibodies against surface proteins like PPEP-1 during a subsequent infection. Importantly, this effect is also observed when infection occurs prior to vaccination; vaccination can overcome dampened immunity to natural infection and lead to strong mucosal immunity in patients with recent CDI. Notably, mucosal antibody titers are reduced compared to systemic titers and induction of a potent mucosal IgA response against each immunogen would improve the efficacy of vaccines against C. difficile significantly.

Durable immunity protects against CDI

Long-term durability of anti-C. difficile vaccine immune responses are essential for successful protection against acute as well as relapsing and recurrent infections. IgG titers in the serum of immunized mice remained stable 7 weeks post-boost (Fig. 3A). Mucosal TcdA and TcdB antibodies also remained stable over time, while PPEP-1 antibodies were not detected (Fig. 3B). Vaccination induced similar levels of TcdA- and TcdB-specific splenic memory B cells (MBCs), irrespective of vaccine valency (Fig. 3C). PPEP-1 specific MBCs were significantly higher compared to anti-toxin MBCs (Fig. 3C) but with lower overall antibody responses, indicating a lack of correlation between MBCs and circulating antibodies for PPEP-1, a phenomenon that has been described for other antigens (32). We next addressed whether long-lived vaccine responses could protect mice from severe CDI using a delayed infection model (Fig. 3D). Long-lived immunity protected mice from lethal CDI (Fig. 3E). Mice lost 6% of their original body weight during acute infection and presented with mild disease (Fig. 3F–G).

Fig. 3. mRNA-LNP vaccines provide long-term protection against CDI.

(A - B) Antigen-specific antibodies in sera (A) and feces (B) at 14 days and 40+ days after the last immunization with 1μg of bivalent or trivalent mRNA-LNPs. (C) Antigen-specific memory B cells in the spleen 40+ days post-immunization. (D to G) Mice were infected with 10,000 spores of C. difficile (VPI 10463) 40 days after last immunization. (D) Schematic of experimental design. (E) Survival, (F) weight loss, and (G) clinical sickness were monitored over course of infection. (H to K) Mice were infected with 10,000 VPI spores 2 weeks after the last immunization, and reinfected with 10,000 VPI spores 26 weeks after initial infection. (H) Schematic. (I) Survival, (J) weight loss, and (K) clinical scores. ND, not detectable. n = 4–10 mice per group. Data are represented as box and whiskers (A-B), mean ± SD (C, F-G, J-K), or percent survival (E, I). Statistics by Mann-Whitney (A-B), Log-rank (Mantel-Cox) (E, I), or two-way ANOVA with Tukey’s multiple comparisons test (F-G, J-K). *p <0.05, ** p<0.01, *** p<0.001, **** p<0.0001.

The effect of mRNA-LNP vaccination on recurrent infection was examined using a long-term reinfection model. More than 6 months after vaccination and a primary infection, we re-challenged immunized mice who had recovered and completely cleared C. difficile with a second CDI (Fig. 3H). Like primary infection, vaccination protected mice from mortality (Fig. 3I). All mice presented with mild disease, lost an average of 8% body weight during secondary infection, and remained alert and active (Fig. 3J, K). These data show the capacity of the mRNA-LNP vaccine platform to induce long-lived memory responses that confer durable protection against CDI.

Targeting C. difficile spores reduces pathogenesis

So far we have shown that the mRNA-LNP vaccine platform induces robust anti-toxin immune responses to prevent morbidity and mortality associated with severe CDI. Moreover, by targeting non-toxin virulence factors we can advance towards decolonization of toxin-producing C. difficile. However, C. difficile is a highly complex organism that alternates between two main phases in its life cycle: the vegetative cell and spore. C. difficile spores – the primary mode of transmission for this pathogen – are highly resistant to traditional chemical disinfectants, phagocytosis, and antimicrobials (33, 34), presenting a major challenge in treating CDI. We postulated that addition of a spore-specific immunogen could further improve our vaccine. We selected the C. difficile exosporium morphogenic protein (CdeM), a cysteine-rich protein expressed on the outer spore coat unique to C. difficile among spore-formers in the gut (35). Comparative genomic analysis of the CdeM sequence in our mRNA constructs showed high conservation across C. difficile phylogeny (Fig. 4A and fig. S16) (12).

Fig. 4. mRNA-LNP vaccine targeting vegetative and spore proteins protects against CDI and elicits antibodies in non-human primates.

(A) Amino acid sequence identity of CdeM from our mRNA construct compared to 137 C. difficile strains across 5 clades. (B) Antigen-specific antibodies in sera 14 days after the last immunization with 1μg of mRNA-LNPs. (C to I) Mice were immunized twice i.m. with 1μg of tetravalent (red), spore trivalent (TcdA/TcdB/CdeM) (blue), vegetative trivalent (TcdA/TcdB/PPEP-1) (purple), or CdeM monovalent (orange) mRNA-LNPs. Two weeks after last immunization, mice were treated with cefoperazone and infected with 10,000 VPI spores. (C) Survival curve. (D) Weight loss. (E) Clinical scores. (F to I) C. difficile CFUs (F, H) and toxin titers (G, I) in feces over time. (J) Antigen-specific IgG titers in a macaque at baseline (naïve), after prime (day 21), and boost (day 35) immunizations with 200μg of tetravalent mRNA-LNP. ND, not detectable. n = 5 mice/group, 2 independent experiments. Data represented as mean ± SD (A, D-E), percent survival (C), or mean ± SEM (B, F-I). Data in (J) show fold change compared to control macaque. Statistics by Log-rank (Mantel-Cox) (C), two-way ANOVA with Tukey’s multiple comparisons test (D, E), or Kruskal-Wallis multiple comparisons test (B, F-I). *p <0.05, ** p<0.01, *** p<0.001, **** p<0.0001.

Immunization with a tetravalent vaccine elicited antigen-specific systemic and mucosal antibodies against all immunogens without compromised immunogenicity (Fig. 4B and fig. S16B). To experimentally determine whether a multivalent approach targeting C. difficile spore and vegetative cells influences acute CDI, we immunized mice with either CdeM/PPEP-1/TcdA/TcdB (tetravalent), CdeM/TcdA/TcdB (spore trivalent), PPEP-1/TcdA/TcdB (vegetative trivalent) or CdeM monovalent mRNA-LNPs. CdeM alone was insufficient to protect mice from lethal infection, while mice immunized with multivalent vaccines presented with mild disease and were fully protected (Fig. 4C–E). Notably, inclusion of CdeM (tetravalent and spore trivalent) immunization resulted in lower C. difficile CFUs and toxin in stool on day 1 post-infection compared to vegetative trivalent immunization (Fig. 4F, G), suggesting that vaccination against spores can limit initial colonization even when challenged with a high spore inoculum. At day 7 post-infection, tetravalent immunization resulted in significantly reduced toxin titers in stool compared to spore or vegetative trivalent mRNA-LNPs vaccines (Fig. 4H, I). Together these data indicate a multivalent approach can protect from disease, reduce colonization, and promote decolonization of toxigenic C. difficile from the GI tract.

Finally, to model human immune responses to vaccination, we performed an immunogenicity study in relatively aged non-human primates. While a single immunization was sufficient to induce anti-TcdA and anti-CdeM IgG, two immunizations were necessary for antigen specific IgG against all immunogens (Fig. 4J), further highlighting the capacity of mRNA-LNP vaccines to elicit robust immunity against bacterial toxins and virulence factors.

This study provides evidence for the value of the mRNA-LNP vaccine platform for the treatment and prevention of CDI. We have shown that clinically relevant doses of mRNA-LNP vaccines elicit robust systemic and mucosal immunity against several C. difficile virulence factors simultaneously in multiple clinically relevant animal models. Through a combination of vaccine targets, we have demonstrated that toxin-specific antibodies protect mice from lethal toxin challenges and CDI, while immunization against vegetative cell and spore proteins reduce C. difficile colonization and promote decolonization during severe infection (fig. S17). We hypothesize that clearance depends on a vaccine-induced humoral immune response, although further work is needed to optimize expression and immunogenicity of vaccine targets including PPEP-1 and to improve mucosal responses to vaccination.

Data and materials availability:

The C. difficile genome sequences are publicly available in the NCBI SRA under BioProject ID PRJNA524299. The data from 16S rRNA gene sequencing are publicly available in the NCBI SRA under BioProject ID PRJNA104103. All data are available in the main text or the supplementary materials. All materials used in this manuscript are available from the authors upon reasonable request.