Abstract Respiratory syncytial virus (RSV) causes severe respiratory disease in infants and the elderly. However, natural infection fails to induce durable immune protection, and existing mRNA vaccines for older adults exhibit limited long-term efficacy. We developed an antigen engineering strategy inserting ESCRT/ALIX-binding region (EABR) into truncated RSV prefusion F (PreF) cytoplasmic tails to form enveloped virus-like particles (eVLPs). In murine models, PreF-EABR mRNA vaccines elicited higher, more persistent neutralizing antibodies than conventional PreF mRNA, correlating with enhanced germinal center B cell and memory B cell responses. A lower dose of PreF-EABR mRNA (1 μg) suppressed viral load and pathology comparable to higher-dose PreF mRNA (2.5 μg). Transcriptomic analysis showed PreF-EABR mRNA activated toll-like receptor and chemokine signaling pathways, enhancing antibody longevity via platelet-associated signatures. This study explores the development and possible mechanism of long-lasting RSV mRNA vaccines by eVLPs technology, which also suggest its potential application in other vaccines. Subject terms: Viral infection, Protein vaccines, RNA vaccines Introduction Respiratory syncytial virus (RSV), a negative-sense single-stranded RNA virus, is one of the main cause of serious respiratory disease in young infants and the elderly^[62]1.Nearly all children experience an infection by 2 years of age^[63]2. It is estimated that nearly 32 million children under 5 years of age are infected by RSV worldwide each year, causing acute lower respiratory tract infection and about 60,000 deaths. Elderly and immunocompromised individuals are at higher risk of experiencing of symptomatic RSV infection, with appreciable mortality^[64]3. However, natural infections by RSV do not induce persistent immune protection^[65]4.Although vaccination is commonly considered as one of the most cost-effective strategies for preventing infectious diseases, RSV vaccines development has progressed slowly, partly due to the emergence of vaccine-enhanced disease (VED) in infants who received the first experimental vaccine, formalin-inactivated RSV (FI-RSV) vaccine in the 1960s^[66]5,[67]6. Nearly 50 years later, a strategy based on a stabilized version of the prefusion (PreF) conformation of the RSV F protein led to the development of several clinically safe and efficacious RSV vaccines^[68]7,[69]8. Currently, two recombinant protein vaccines (Arexvy and Abrysvo) and one mRNA vaccine (mRESVIA), have been approved for use in adults aged 60 and older, demonstrating both efficacy and safety^[70]9–[71]11. During the COVID-19 pandemic, the safety and efficacy of these two mRNA vaccines for COVID-19 developed by Pfizer/BioNTech and Moderna, respectively, were rigorously validated. They showed remarkable efficacy with an effectiveness rate exceeding 90% and were approved for human use for the first in class vaccines^[72]12. Moreover, compared to traditional protein vaccines, mRNA vaccines elicit more robust CD4^+ and CD8^+ T-cell responses^[73]13, increasing their potential advantages by activating cellular immune responses. The currently approved RSV mRNA vaccines, encoding the PreF protein of RSV have exhibited high efficacy, preventing over 83.7% of RSV-related lower respiratory tract diseases characterized by at least two signs or symptoms in clinical trials^[74]14. However, the latest data showed the durability of this protective efficacy is limited, as their effectiveness in preventing RSV declines to only 50% after 18 months post-immunization^[75]15. Further design modifications are necessary to enhance long-term protection of these mRNA-based RSV vaccines. Virus - like particles (VLPs) are protein structures that self-assemble and possess multiple subunit capabilities. This technology can present viral antigens on the cell surface or in virus with assembled nanoparticles^[76]16. VLPs mimic the viruses in terms of size (20-200 nm), geometry, and the ability to activate T-helper cells^[77]17,[78]18. VLPs naturally encode T-helper cell epitopes, which are presented to T-helper cells by antigen-presenting cells (APCs) in association with major histocompatibility complex (MHC) class II. This is more conducive to the activation of B cells, thereby inducing a strong B - cell immune response to achieve long-term protection^[79]19,[80]20. Many VLP-based vaccines induce a highly efficacious, long-lasting immune response. They have been approved for protection against hepatitis B virus (HBV), human papillomaviruses (HPV), and hepatitis E (HEV)^[81]18. Thus, the combined advantages of both mRNA and nanoparticle-based vaccines may further improve mRNA vaccine potency. In 2023, a groundbreaking approach (EABR technology) was developed. EABR involves engineering membrane proteins to self-assemble into enveloped virus-like particles (eVLPs) by incorporating an EABR (ESCRT- and ALIX-binding region) motif at the C terminus of the viral membrane protein, which can recruit host proteins from the endosomal sorting complex required for transport (ESCRT) pathway^[82]21. This technology leverages dual antigen display-on cell surfaces and secreted eVLPs to prolong immune exposure and amplify B cell responses. Such an approach has been successfully in improving immunological responses in development of novel SARS - CoV 2 vaccines. However, this technology has not been reported in developing other vaccines. Moreover, the mechanism underlying of this enhanced immunological response remains poorly understood. Here, we pioneer the application of this EABR technology to RSV vaccine design by engineering a PreF with a modified cytoplasmic tail. We found the expressed PreF-EABR forms self - assembled eVLPs with typical molecular maker of exosomes. Most importantly, we demonstrated that PreF-EABR mRNA vaccine outperform conventional PreF mRNA by eliciting durable neutralizing antibody titers, a robust germinal center (GC) B cell response, F-specific memory B cells (MBCs) responses and long-lived plasma cells (LLPCs) responses with enhanced protection. Moreover, we investigated the early innate immune response and explored the underlying mechanism of the strong immune response by PreF-EABR mRNA vaccine. Results PreF-eVLPs is achieved by inserting EABR into the modified cytoplasmic tail To optimize the surface expression of PreF on the cell membrane and facilitate the self - assembly of eVLPs, we truncated the intracellular region of PreF (Fig. [83]1a). Previous studies have shown that retaining at least one amino acid residue in the intracellular region is necessary to maintain the surface expression of the F protein on the membrane^[84]22. Our results demonstrated truncating the C-terminal (CT) region while retaining the K amino acid residue increased the expression level in the cell pellet, but no expression was detected in the supernatant (Fig.[85]1b,c). Then we incorporated the EABR motif downstream of PreF[△CT]. This modification resulted in decreased expression of PreF[△CT]-EABR/no EPM (Endocytosis prevention motif, EPM) in the cell pellet, while detectable expression was observed in the supernatant, consistent with previous findings^[86]21. EPM can inhibit coated pit localization and endocytosis^[87]23. Therefore, combining the EPM and EABR sequences can further enhance eVLPs self-assembly^[88]21. Further, we added the EPM sequence to the deleted CT region, creating the PreF-EABR construct. The inclusion of EPM significantly increased the expression of PreF in both the cell pellet and supernatant. (Fig.[89]1b, c). To confirm that our modifications did not alter the localization of the PreF protein, particularly after truncating the CT region, we performed confocal microscopy. Both PreF and PreF-EABR were successfully expressed on the cell membrane surface (Fig. [90]1d). Additionally, we verified the formation of eVLPs using nanoparticle tracking analysis (NTA) and immuno-electron microscopy (IEM). NTA revealed that the concentration of PreF- eVLPs was 2.4 × 10¹¹ particles/mL, with a median diameter of 123.6 nm (Fig. [91]1e). IEM analysis indicated that PreF proteins were present on the surface of each eVLPs (Fig. [92]1f). These results confirmed modified PreF successfully forms eVLPs by inserting EABR into the cytoplasmic tail. Fig. 1. Designing and characterizing the PreF-EABR. [93]Fig. 1 [94]Open in a new tab a Schematic of PreF[△CT]-EABR/no EPM and PreF-EABR construction. SP, Signal peptide; TM, Transmembrane domain; CT, e C-terminal region; GS, GGGGS linker; EPM, an endocytosis prevention motif; EABR, ESCRT- and ALIX-binding region; (b and c) Western blot (WB) analysis of PreF expression in cell pellet (b) and supernatant (c). d Confocal microscopy of the PreF and PreF-EABR. e, f Nanoparticle tracking analysis (e) and immuno-electron microscopy (f) were used to verify the formation of PreF-eVLPs. g Western blot analysis of PreF-eVLPs using the CD9, CD63, CD81 and Calnexin marker. PreF-eVLPs has typical molecular markers of exosomes The ESCRT pathway also plays a central role in mediating the formation of extracellular vesicles (EVs)^[95]24–[96]26. Therefore, we further characterized PreF-eVLPs using Western blot (WB). Interestingly, these eVLPs expressed typical extracellular vesicle markers, including CD9⁺, CD63⁺, and CD81⁺, while Calnexin was absent (Fig. [97]1g). Considering its diameter, we confirmed that PreF-eVLPs exhibits typical molecular markers of exosomes (30−150 nm EVs), indicating that PreF-eVLPs has a similar formation mechanism to that of exosomes, rather than budding from the surface of the cells. PreF-EABR mRNA vaccine elicits superior antibody response To investigate whether genetic encoding of PreF-EABR enhances the immune potency of PreF, we developed two nucleoside-modified mRNA vaccines encoding PreF and PreF-EABR. The immunization scheme of experiments was as shown in Fig. [98]2a. Following the prime immunization, PreF-EABR mRNA elicited significantly higher PreF-specific antibody responses compared to PreF mRNA, in a dose-dependent manner (Fig. [99]2b). Notably, neutralizing antibodies against RSV Long and RSV B1 were detected in mice vaccinated with PreF-EABR mRNA at all tested doses (0.5 μg, 1 μg, and 2.5 μg). In contrast, no significant neutralizing antibodies was observed in the PreF mRNA group at the lower doses (0.5 μg and 1 μg), except at the highest dose (2.5 μg) (Fig. [100]2c,d). Fig. 2. PreF mRNA and PreF-EABR mRNA induce superior antibody response and potent T cell response in mice. [101]Fig. 2 [102]Open in a new tab a The immunization scheme of experiments. Serum was harvested at 0, 14, 21 and 28 days post-vaccination. b anti-PreF IgG. Serum neutralizing antibody against RSV Long (c) and RSV B1 (d). Frequency of PBMCs effector CD4^+ T cells (e−g) and effector CD8^+ T cells (h−j) following restimulation with RSV F peptide pools (A total of 130 peptides spanning the full-length F glycoprotein of RSV were synthesized as 15-mers, with 11 overlapping amino acids) and soluble co-stimulatory molecules CD28 and CD49d. Data are represented as mean ± SEM and analyzed by one-way ANOVA with Tukey correction. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, and ****P ≤ 0.0001. After the booster immunization, PreF-EABR mRNA induced significantly higher neutralizing antibody titers against both RSV Long and RSV B1 compared to PreF mRNA (Fig. [103]2c,d). The geometric mean of neutralizing antibody titers against RSV Long and RSV B1 elicited by PreF-EABR mRNA were 5.3-fold and 2.6-fold higher, respectively, than those elicited by PreF mRNA at the 2.5 μg dose. These results demonstrate that PreF-EABR mRNA elicits a superior immune response compared to PreF mRNA, highlighting the enhanced immunogenicity conferred by the PreF-EABR construct. The PreF-EABR mRNA vaccine elicits potent but distinct T cell responses compared with PreF mRNA vaccine We also evaluated the T cell response following the booster immunization (Fig. [104]2a). Gating strategy for measuring effector T cell response was shown in Fig. [105]S1. Both PreF mRNA and PreF-EABR mRNA vaccines elicited Th1-type cytokine-producing peripheral blood mononuclear cells (PBMCs) compared to the LNP control group, including cytokines such as IL-2, IFN-γ, and TNF-α, with the responses exhibiting dose-dependence (Fig. [106]2e–j). Notably, PreF-EABR mRNA induced a higher IFN-γ^+CD4^+ T cell response compared to PreF mRNA (Fig. [107]2f), while eliciting a lower IFN-γ^+CD8^+ T cell response than PreF mRNA (Fig. [108]2i). These findings suggest that PreF-EABR mRNA and PreF mRNA may engage distinct immune response mechanisms, further highlighting the unique immunogenic properties of the PreF-EABR construct. PreF-EABR mRNA vaccine provided better protection than PreF mRNA against RSV in vivo To evaluate the protection of PreF mRNA and PreF-EABR mRNA, the immunization-challenge schedule was designed as illustrated in Fig. [109]3a. All vaccines elicited neutralizing antibody titers against both against RSV Long and RSV B1 except for the naïve and LNP groups (Fig.[110] 3b and c). The medium of RSV Long viral load in the lungs of LNP groups was 6.6 log10 copies/g, while all vaccine groups showed no significant difference compared to the naïve groups, except for the FI-RSV group and PreF mRNA group at a dose of 0.5 μg (Fig. [111]3d). With the exception of the FI - RSV groups, the copy numbers of genomic RNA (gRNA) were significantly reduced in all vaccine groups compared to the LNP groups. Notably, while RSV Long gRNA was detectable in the nasal turbinates of all vaccine groups, the levels were significantly lower than those in the LNP and FI-RSV groups (Fig. [112]3e). A strong negative correlation was between RSV Long-neutralizing antibody titers and gRNA loads in both the lungs (R = -0.8531, P < 0.0001) and nasal turbinates (R = -0.6594, P < 0.0001) (Fig. [113]3f, g). Histopathological analysis revealed that mice immunized with LNP and FI-RSV experienced moderate pulmonary damage at 5 days post-infection (dpi), including interstitial pneumonia, peribronchiolitis, and perivasculitis (Fig. [114]3h). Interestingly, PreF mRNA-immunized mice exhibited more severe pulmonary damage compared to PreF-EABR mRNA-immunized mice, particularly at the 0.5 μg dose (Fig. [115]3i). Further analysis of mucus production in lung tissues using Periodic Acid-Schiff (PAS) staining demonstrated that, except for the LNP and FI-RSV groups, which showed evident mucus secretion, no significant mucus secretion was observed in the other vaccine groups (Fig. [116]3j). These results indicate that a 2.5 μg dose of PreF mRNA is sufficient to elicit robust protection against RSV Long. In contrast, PreF-EABR mRNA achieved comparable protection at lower doses of 0.5 μg or 1 μg, highlighting its superior efficacy. Fig. 3. PreF-EABR mRNA provided enhanced protection against RSV Long in vivo. [117]Fig. 3 [118]Open in a new tab a Scheme of vaccination and challenge. Balb/C mice (n = 5) were immunized with LNP or FI-RSV or PreF mRNA with 0.5 μg, 1 μg or 2.5 μg or PreF-EABR mRNA with 0.5 μg, 1 μg or 2.5 μg, respectively, with an interval of 2 weeks. At 4 weeks post-last vaccination, mice were challenged with 2 × 10^6 PFU of RSV Long. Naïve group is the negative control group. At 5 days post-challenge, mice were euthanized, and lung tissue and nasal concha were collected. Serum RSV Long (b) and RSV B1(c) neutralizing antibody titers at 42 days. RSV Long viral burden in lungs (d) and nasal turbinates (e) measured by copies of viral gRNA. Spearman correlation of viral copies in lung (f) or nasal turbinates (g) and RSV Long neutralizing titers. h Inflammation in the lungs post-challenge was scored by immunohistology. i, j Hematoxylin and eosin (i) and periodic acid-Schiff (j) for pathological evaluation and mucus secretion. Arrows indicate pathological damage and mucus secretion. Scale bars: 100 µm. All data are shown as means ± SEM. P-values were analyzed with one-way ANOVA (*P  < 0.05; **P  < 0.01; ***P  < 0.001; ****P  < 0.0001). PreF-EABR mRNA induce stronger GC B and Tfh responses than PreF mRNA GC B and follicular helper T cell (Tfh) response are known to play a critical role in the development of a high-quality antibody response by antigen-activated B cells^[119]27–[120]29.We hypothesized that PreF-EABR mRNA might induce a greater GC B cell response than PreF mRNA. Balb/C mice were immunized with LNP, PreF mRNA or PreF-EABR mRNA. LNs and spleen were harvested at 7 and 14 days post-last vaccination (Fig.[121]4a). Gating strategy for measuring total GC B (CD4^-CD19^+Fas^+ GL7^+), F-specific (CD4^-CD19^+Fas^+ GL7^+ F^+) and Tfh (CD19^-B220^-CD4^+CD62L^- CXCR5^+ PD-1^+) was shown in Fig.[122]S2. All vaccine groups stimulated robust total GC B cell response in draining LNs and spleen at 7 and 14 days post boost immunization compared to PBS groups, as well as F- specific GC B cell response (Fig. [123]4d, e). PreF-EABR-mRNA elicited higher total GC B cell responses and F-specific GC B cell responses than PreF mRNA groups in LNs (Fig. [124]4b, d and e). A similar trend was observed in the spleen (Fig. [125]4g, h). Animals immunized with PreF mRNA or PreF-EABR mRNA presented a higher frequency of Tfh cells in LNs and spleen. Notably, there was a significant difference the frequency of these cells in LNs between PreF mRNA and PreF-EABR mRNA groups (Fig. [126]4f and i). These results indicate that PreF-EABR mRNA induce stronger GC B and Tfh response than PreF mRNA. Fig. 4. PreF-EABR mRNA induce stronger GC B and Tfh response than PreF mRNA. [127]Fig. 4 [128]Open in a new tab a Scheme of experiments. Two sets of mouse experiments were performed simultaneously. Balb/C mice (n = 5) were euthanized at day 7 and 14 post- last vaccination, and LNs and spleen were harvested. b, c Schematic diagram of gating for GC B cells (b) and Tfh cells (c) by flow cytometry. d, e Frequency of GC B cells or F specific GC B cells in LNs. f Frequency of Tfh cell response in LNs (g, h) Frequency of GC B cells or F specific GC B cells in spleen. i Frequency of Tfh cell response in Spleen. All data are shown as means ± SEM. P-values were analyzed with one-way ANOVA (*P  < 0.05; **P  < 0.01; ***P  < 0.001; ****P  < 0.0001). 1 µg of PreF mRNA and PreF-EABR mRNA were not compared with Naïve and LNP. PreF-EABR mRNA vaccine induce better long-term immunity than PreF mRNA Subsequently, we assessed the long-term immunity in mice immunized with PreF mRNA and PreF-EABR mRNA (Fig. [129]5A). Serum binding antibody levels and neutralization activity against various RSV strains were monitored up to 20 weeks post-immunization (Fig. [130]5B–D). Neutralizing antibody titers in both the PreF mRNA and PreF-EABR mRNA groups declined over time but remained significantly higher than those in the control group from weeks 1−20. Furthermore, the reduction in neutralizing antibody titers was less pronounced in the PreF-EABR mRNA group than in the PreF mRNA group (Fig. [131]5C, D). Fig. 5. Long-term immune response in immunized Balb/C mice. [132]Fig. 5 [133]Open in a new tab A Schedule of vaccine immunization and bleeding strategies in Balb/c mice (6–8 week-old). B Kinetics of PreF-specific total IgG endpoint titers (log10) were measured within 20 weeks of initial vaccination. C Kinetics of RSV Long neutralizing antibody NT[50] titers (log2) were measured within 20 weeks after the first vaccination. D Kinetics of RSV B1 neutralizing antibody NT[50] titers (log2) were measured within 20 weeks after the first vaccination. E and f MBCs (E) and (F) specific MBCs (F) were measured in spleen at 20 weeks after the first vaccination. g, H LLPCs (G) and F specific LLPCs (H) were measured in BM at 20 weeks after the first vaccination. All data were shown as means ± SEM. P-values were analyzed with one-way ANOVA (*P  < 0.05; **P  < 0.01; ***P  < 0.001; ****P  < 0.0001). We next evaluated the MBCs and LLPCs immune responses at 20 weeks post-immunization. Gating strategy for measuring MBCs and LLPCs was shown in Fig. [134]S3. Both vaccines induced robust MBCs responses in the spleen and LLPCs responses in the bone marrow (BM) compared to the control group. The frequency of F-specific MBCs was significantly higher in the PreF-EABR mRNA group than in the PreF mRNA group (Fig. [135]5F, G). Additionally, although the frequency of LLPCs in the PreF-EABR mRNA group was higher than in the PreF mRNA group, but no significant difference. (Fig. [136]5H, I). Compared with PreF mRNA, enhanced innate immune responses were induced by PreF-EABR mRNA Effective activation of innate immunity through vaccination is critical for eliciting robust, antigen-specific adaptive immune responses^[137]30. To gain a more in-depth understanding of the innate immune activation, we performed transcriptomic analyses on PBMCs isolated 24 h post the prime immunization. RNA-Seq revealed that a large number of different expressed genes (DEGs) compared with naïve mice were induced in the LNP, PreF mRNA, and PreF-EABR mRNA groups (Fig. [138]6a, b). We used Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of the DEGs (Fig. [139]6c, d). PreF mRNA and PreF-EABR mRNA vaccine induce similar KEGG pathways such as the Toll-like receptor signaling pathway (such as Tlr3, Tlr9), TNF signaling pathway and antigen processing/presentation (such as Tap1, Tap2, H2-Q6, Hspa1b) (Fig. [140]6e–h). However, key divergences emerged in pathway prioritization, especially the JAK-STAT signaling pathway and the platelet activation pathway (such as Gp1ba, Itga2b, Fermt3) (Fig. [141]6c, d). Furthermore, we analyzed the DEGs between PreF mRNA - EABR and PreF mRNA (|log₂ (fold change) | > 1, q < 0.05), and conducted KEGG pathway enrichment analysis. Cytokine-cytokine receptor interaction, IL-17 signaling pathway, Chemokine signaling pathway, and TLR signaling pathway were significantly enriched (Fig. [142]6i), suggesting PreF-EABR mRNA induced a much stronger innate immune response than PreF mRNA. Finally, we measured the cytokine levels in the serum of mice 24 h after vaccination using a multiplex assay to validate the RNA–Seq results. The results were consistent with those of the RNA–Seq analysis. Elevated levels of pro-inflammatory analytes such as IL-6, TNFα, IFN-γ, IL-17 and analytes involved in immune cell trafficking/modulation such as CXCL1, CCL-3 and IL-12-p70 were observed (Fig. [143]6j–w). Fig. 6. RNA-seq analysis and cytokine expression of serum in vaccinated mice. [144]Fig. 6 [145]Open in a new tab BALB/c mice (n = 8/group) were vaccinated i.m. with LNP, PreF mRNA, and PreF-EABR mRNA at the dose of 2.5 μg. Naïve mice were set as the control. Mouse PBMC were isolated 24 h post-injection for bulk RNA-seq (n = 3) and mouse sera were collected for a cytokine array (n = 5). a RNA-seq analysis of hierarchical clustering of significant DEGs in different groups. b DEG numbers of each group are indicated. c, d KEGG enrichment of upregulated DEGs in the PreF mRNA group (c) or PreF-EABR mRNA group (d). e–h Heatmaps of selected DEGs are classified by their immune function, such as Toll-like receptor signaling pathway (e), antigen processing/presentation (f), platelet activation pathway (g) and chemokine signaling pathway (h). (i) KEGG enrichment of upregulated DEGs in the PreF-EABR mRNA group compared to the PreF mRNA- group. j–w cytokines in sera of vaccinated mice. Discussion Given the substantial disease burden associated with annual respiratory syncytial virus (RSV) epidemics in infants and the elderly, and considering the limitation in immune durability of currently approved RSV mRNA vaccines, we aimed to design an RSV mRNA vaccine that can provide long-lasting protection by taking the advantages of both mRNA and VLPs vaccines. Magnus A.G et al. reported a self – assembling eVLPs technology for forming eVLPs that mimics natural infection^[146]21.Here, we initially applied this new technology for RSV vaccine design. We found PreF-EABR is not only be expressed at the cell surface in vitro, but also forms eVLPs that escape from the cell into the supernatant. We also confirmed that the EPM is crucial for enhancing the expression of eVLPs^[147]21. To the best of our knowledge, this is the first study that has applied the EABR eVLPs technology to develop an mRNA vaccine for RSV. We further discovered that the PreF-EABR exhibits typical molecular markers of exosomes, such as CD9⁺, CD63⁺, CD81⁺ with Calnexin^−. As expected, we did not detect exosome markers in the supernatant of cells expressing anchorless PreF produced under the same conditions as PreF-EABR. These results rule out the possibility of autocrine exosome formation. Exosomes are naturally formed within multivesicular bodies (MVBs) in the endosomal pathway and released into the extracellular space upon fusion of MVBs with the plasma membrane through the ESCRT-dependent pathway and ESCRT-independent pathway^[148]31,[149]32. It was reported that the EABR technology is designed to enhance the budding and release of eVLPs by recruiting the host cell’s ESCRT machinery, but the characteristics and mechanism of eVLPs formation was not explored^[150]21. Because exosomes can also be formed through budding, we hypothesized that eVLPs technology may share the same or similar formation mechanism to that of exosomes. The formation mechanism of eVLPs requires further study to ascertain whether it is specific to PreF-eVLPs, universal, or similar to that of exosomes. Consistent with previous studies, SARS-CoV-2 spike-EABR mRNA vaccines induced a 10-fold increase in neutralizing antibody titers compared to conventional spike mRNA vaccine^[151]21. In this study, PreF-EABR mRNA also induced significantly higher neutralizing antibody titers (5.3-fold and 2.6-fold higher) against both RSV Long and RSV B1 compared to PreF mRNA. Protection experiments demonstrate that the eVLPs can provide optimal protection with a reduced dosage compared to PreF (0.5-1 μg vs 2.5 μg). This characteristic could be an advantage to vaccine production. Additionally, it may contribute to minimizing adverse reactions within the human body, thereby enhancing the vaccine’s safety profile and increasing its acceptability for administration. Overall, these data demonstrate that PreF-EABR mRNA can potently stimulate more superior antibody response and provide better protection than PreF mRNA. Chandrav De et al. demonstrated human CD8^+ T cells or CD4^+ T cells effectively and independently control RSV replication in human lung tissue in the absence of an RSV-specific antibody response^[152]33. Compared to traditional protein vaccines, mRNA vaccines elicit robust CD4^+ and CD8^+ T-cell responses^[153]13.In this study, both PreF mRNA and PreF-EABR mRNA vaccines elicited CD4^+ and CD8^+T cell response with a dose-dependent manner. Notably, PreF-EABR mRNA induced a higher IFN-γ^+CD4 + T cell response compared to PreF mRNA. This may be due to the formation of eVLPs, which present antigens in an organized and highly repetitive manner, thereby triggering efficient humoral and cellular immune responses^[154]34. Additionally, VLPs naturally encode T-helper cell epitopes presented by APCs via MHC class II to trigger B cell activation^[155]18. These findings are consistent with the enhanced B-cell immunity observed in our mouse studies. In addition, we observed PreF-EABR mRNA induced a lower IFN-γ^+CD8^+ T cell response than PreF mRNA. This discrepancy may be attributed to the deletion of the CT region, which contains CD8^+ T cell epitopes^[156]35,[157]36. In the future, we will determine what caused the decrease in frequency of CD8⁺ T cells induced by PreF EABR mRNA compared with traditional mRNA vaccines affect the vaccine’s efficacy. The enhancement of this vaccine’s efficacy may be attributed to the promotion of B - cell activation by PreF-EABR, due to PreF not only be expressed at the cell surface, but also forming extracellular eVLPs to activate immune cells^[158]21. Durable vaccine efficacy hinges on the generation of LLPCs and MBCs, which arise from GC reactions in secondary lymphoid organs^[159]27,[160]37. The GC reaction is tightly regulated by Tfh cells, which enable proliferation, survival, and differentiation of GC B cells through the delivery of costimulatory molecules and cytokines. In this study, both PreF-EABR mRNA and PreF mRNA groups elicited robust total GC B cell responses and F specific GC B cell response, which consist with other mRNA vaccines^[161]38,[162]39. Importantly, we observed that PreF-EABR mRNA elicited higher total GC B cell responses and F specific GC B cell response than PreF mRNA groups in LNs and spleen, and elicited a higher frequency of Tfh cells in LNs and spleen, suggesting PreF-EABR mRNA elicited potent stimulation of B cell immune responses than PreF mRNA. However, previous studies have indicated that while mRNA vaccines against SARS-Cov2 effectively elicit a robust GC response, their capacity to induce MBCs and LLPCs is limited^[163]15, which may explain why the Moderna RSV mRNA vaccine is less persistent. This longer-term immunity in mice demonstrated that the PreF-EABR mRNA could induce higher PreF binding antibody levels and neutralization antibody against the RSV Long and B1 strains compared to PreF mRNA, and remain durable for at least 20 weeks in mice. We found that the frequency of F-specific MBCs was significantly higher in the PreF-EABR mRNA group than in the PreF mRNA group, which means that the PreF-EABR mRNA stimulates stronger and more persistent immunity compared to traditional mRNA vaccines. Collectively, these data demonstrate that PreF-EABR mRNA can potently stimulate B cell immune responses more effectively than PreF mRNA, which further illustrates the advantages of PreF-EABR mRNA in persistent immune responses. Effective activation of innate immunity through vaccination is critical for eliciting robust, antigen-specific adaptive immune responses^[164]40,[165]41. Both PreF mRNA and PreF-EABR mRNA vaccine induce similar KEGG pathways, suggesting the stimulation of innate immune responses is closely related to the types of vaccines. Moreover, it is consistent with a previous report that mRNA vaccine activates the type I IFN signaling pathway^[166]42. Activation of type I IFN signaling pathway is relies on the JAK-STAT signaling pathway and non-receptor tyrosine kinase 2 pathway^[167]43. We found that both PreF-EABR mRNA and PreF mRNA upregulated numerous JAK-STAT pathway-related genes. Notably, the magnitude of upregulation in both JAK-STAT signaling pathway were significantly greater for PreF-EABR mRNA, suggesting that PreF-eVLPs mRNA induce a stronger innate immune response than PreF mRNA. This finding was also confirmed by multiplex cytokine profiling. Platelets are produced from megakaryocytes precursor cells that reside in the bone marrow, interact with the megakaryocytes, and induce produce the prominent plasma cell survival factors APRIL and IL-6^[168]44. Platelet-associated signature can predict antibody response longevity through megakaryocyte activation via thrombopoietin receptor signaling which increases plasma cell survival and antibody response durability^[169]45. We observed the platelet activation pathways elicited in PreF-EABR mRNA were significantly greater for PreF mRNA. These results further suggest that the PreF-EABR mRNA induces a more persistent immune response compared to PreF mRNA, which is consistent with the antibody persistence and memory B cells response that we observed in mice. Additionally, we observed that an antigen processing/presentation pathway, involving the H2 - related genes (H2-Q3, H2-Q4, H2-Q6 and H2-T22) were upregulated. H2 genes (MHC II in mice), bound antigen peptide can activate CD4 T cell response^[170]46. These results are consistent with the CD4^+ T cell immune response induced by the PreF-EABR mRNA in mice. Overall, transcriptomic analyses further confirmed that PreF-eVLPs mRNA induces a stronger innate immune response, thereby triggering the immune responses of CD4 + T cells and long-lived memory B cells. Our research has some limitations: immunization studies were performed only in mice, future studies will be needed to evaluate the immune response in cotton rats and humans. In addition, we only analyzed the innate immune response 24 h post immunization in the present study. The immune responses at the other time points should be further studied to clarify the patterns of how they elicit the innate immune response. Moreover, we introduced additional amino acid sequences, the EABR motif and the EPM motif in the RSV PreF protein. Whether these two motifs can stimulate the body to generate an immune response remains unclear. These issues should be addressed before the clinical applications of PreF-EABR mRNA vaccines in future. To summarize, we have developed a novel PreF-EABR based RSV mRNA vaccine with superior immunogenicity and protection against RSV Long in mice. Furthermore, our study found this self-assembled eVLPs hass typical molecular markers of exosomes, indicating the possibility of a similar to formation mechanism of these eVLPs with that of exosomes though more studies are needed. Importantly, we explored the early innate immune activation mechanisms, particularly platelet-mediated pathways, which may be the reasons for triggering a persistent immune response by this PreF-EABR based RSV mRNA vaccine. These findings may be helpful for the research and development of the next generation of long-lasting RSV vaccines and other vaccines. Methods Cell and viruses HEK293T cells (CRL-3216, ATCC) and HEp-2 cells (CCL-23, ATCC) were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) and 1× penicillin-streptomycin (Thermo, USA) at 37 °C, 5% CO[2]. RSV Long (VR-26) and RSV B1 were passaged in HEp-2 cells in DMEM/2% FBS, purified via ultrafiltration and stored in DMEM/3% sucrose at −80 °C as the previous reported^[171]47. Design of PreF-EABR constructs The amino acid sequence of the RSV fusion protein used was RSV A CC14-10 (GenBank: [172]APW77972). PreF construct was codon-optimized and modified by introducing amino acid substitutions (N67I, R106Q, F137S, S215P, and E487Q), the P27 (109-136aa) was substituted with the linker (GSGSGR)^[173]47,[174]48, and the native signal peptide was replaced with tPA signal peptide. The C terminal 23 residues was truncated to optimized cell surface expression to form PreF[△CT] construct^[175]22.PreF[△CT]-EABR/no EPM construct was generated by inserting the EABR domain (160-217aa) of the human CEP55 protein in C terminal to the PreF[△CT] construct separated by the GGGGS linker. An endocytosis prevention motif (EPM), a 47-residue insertion from the murine Fc gamma receptor FcgRII-B1 cytoplasmic tail was used to prevent coated pit localization and endocytosis, inserting the upstream of the GS linker and the EABR domain and behind the PreF[△CT] to form PreF-EABR construct. All constructs were cloned into the pCDNA3.1 expression plasmid. Production of PreF-EABR PreF-EABR was generated by transfecting HEK293F cells through the X-treme GENE^TM HP transfection reagent (Roche, Switzerland), 48 h post-transfection, cells were centrifuged at 800 × g for 10 min to separate the cell pellet and supernatant. The cell supernatant was processed according to the previously reported method to obtain eVLPs^[176]21. In brief, supernatant was filtered through 0.45 μm filter to further remove the cell debris, and then, eVLPs was precipitated by 20% sucrose density gradient centrifugation at 50000 rpm for 2 h at 4 °C. The pellet was resuspended in PBS at 4 °C overnight, and then samples were centrifuged at 10000 × g for 10 min to remove residual cell debris. Western blot (WB) Cell pellet was lysed by the RIPA lysis buffer and eVLPs were prepared for SDS-PAGE, then transferred onto PVDF membranes for WB analysis. After blocking with 5% skim milk. For F-specific WB, the membranes were incubated with mouse anti-F monoclonal antibody (1:1000 dilution, Abcam, USA) overnight at 4 °C followed by HRP-conjugated goat anti-mouse IgG antibody (1:5000 dilution, Abcam, USA). For identification of extracellular vesicle, membranes were separately incubated with anti-CD63 antibody (1:1000 dilution, Abcam, USA), anti-CD9 antibody (1:1000 dilution, Abcam, USA), anti-CD81 antibody (1:1000 dilution, Abcam, USA), anti-Calnexin antibody (1:1000 dilution, Abcam, USA) overnight at 4 °C followed by HRP-conjugated goat anti-mouse IgG antibody (1:5000 dilution, Abcam, USA). Immunofluorescence Assay (IFA) The slides of cells were sterilized by 75% ethanol and PBS rinsing and air-dried in 6-well plates. Then, HEK293 cells were seeded at 2 × 10⁵ cells/well and cultured in DMEM with 10% FBS at 37 °C, 5% CO[2]. When the cell density reached 50-70% confluence, transfection complexes were prepared, which consist of 2.5 μg plasmid DNA (PreF or PreF-EABR construct) and 5 μl X-treme GENE^TM HP transfection (Roche, Switzerland) in 180 μl Opti-MEM (Gibco, USA) were then added to the cells. 24 h post-transfection, cells were subjected to a series of processing steps: fixed with 4% paraformaldehyde (Sigma, Germany) for 20 min. After PBS washed, cells were stained with 5 μg/ml WGA-633 (Thermo Fisher, USA) for 5 min, then cells were permeabilization using 0.2% Triton X-100 (Sigma), and nuclear counterstaining with 1 μg/ml DAPI (Sigma, Germany). Subsequently, the cells were stained with FITC-anti F antibody (1:400, Vazyme, China) for 30 min. After PBS washed, the images of the targeting proteins were viewed and analyzed using confocal microscopy (LEICA TCS SP8, Germany). Characterization of PreF-EABR The number and size of eVLPs were measured by nanoparticle tracking analysis (NTA) using a ZetaView_Particle Metrix (PMX-120, Germany). Immunotransmission electron microscopy (IEM) was used to identify the particle. eVLPs were fixed with an equal volume of 4% paraformaldehyde, and then pipette the liquid onto a carbon-coated nickel grid, block with 0.05 M glycine and then block with 1% BSA (Sigma, Germany), incubate with rabbit anti F antibody overnight at 4 °C. After washed, incubated with 6-nm colloidal gold Goat anti rabbit antibody for 3 h. Fixed with 2.5% glutaraldehyde, then stained with 2% uranyl acetate. IEM was performed using a JEOL–1200 electron microscope (JEOL, Japan) operated at 80 KV to analysis. mRNA vaccine preparation PreF and PreF-EABR gene were codon optimized through LinearDesign and insert in-vitro transcription vector, which flank with the 5’ and 3’ untranslated regions and a 120 nt poly(A) tail. After the plasmid was linearized by digestion with Bsa I, T7 RNA polymerase-mediated transcription was employed to synthesize the mRNA using the T7 High Yield RNA Transcription Kit 2.0 (N1-Me-pUTP) (Novoprotein, China), After this, transcription products were incubated at 37 °C for 1 h by treatment with DNase I. RNA was capped using Cap 1 Capping System Kit (Novoprotein, China) and purified using oligo dT (BIA, Germany). Quality of mRNA was analyzed by agarose gel electrophoresis and stored at -80 °C. For mRNA encapsulation into LNP, the LNP were prepared using Moderna’ SM102 LNP formulation and encapsulated by Novoprotein (Shanghai, China). Vaccine formulation was characterized for particle diameter, polymer dispersity index (PDI), and zeta potentials using NanoBrook Omni ZetaPlus (Brookhaven, Germany). Immunization and virus challenge Six- to eight-week-old female Balb/c mice were purchased from Beijing Vital River Laboratory Animal Technology Co. (Beijing, China) and housed in appropriate animal facilities at the laboratory animal center of the Chinese Center for Disease Control and Prevention. All animal experiments were conducted under the approval of the ethics committee of the Chinese Center for Disease Control and Prevention (approval NO.20221107115). Mice (n = 5/group) were immunized intramuscularly (i.m.) on days 0 and 14 with PreF mRNA or PreF-EABR mRNA at doses of 0.5 μg, 1 μg, or 2.5 μg in a 100 μl volume (50 μl injected into each thigh muscle). LNP was used as control. Serum samples were collected for Ab response and RSV-specific neutralization assay analysis at different time points. For the challenge experiments, Balb/C mice (n = 5/group) were vaccinated i.m. with LNP, PreF mRNA, and PreF-EABR mRNA at the dose of 2.5 μg in a 100 μl volume (50 μl injected into each thigh muscle) with an interval of 2 weeks. FI-RSV were used as positive control at the dose of 20 μg. At the same time, mice that were neither immunized nor challenged were used as the negative control group (naïve group). Four weeks after the last vaccination, mice were anesthetized with isoflurane and then intranasally challenged with 2 × 10^6 PFU of RSV Long in a 30 μl volume. Five days post - challenge, the mice were euthanized by cervical dislocation, and their lung tissues and nasal turbinates were collected and stored at −80 °C until processed. For GC B and Tfh immune response, Balb/C mice (n = 5/group) were immunized with LNP, PreF mRNA or PreF-EABR mRNA at the dose of 2.5 μg in a 100 μl volume (50 μl injected into each thigh muscle). At the end point of experiments, mice were euthanized by cervical dislocation and spleens and draining lymph nodes (dLN) were collected and processed to obtain single-cell suspension for analysis at 7 and 14 days post-last vaccination. For the RNA-Seq experiments, BALB/c mice (n = 8/group) were vaccinated i.m. with LNP, PreF mRNA, and PreF-EABR mRNA at the dose of 2.5 μg in a 100 μl volume (50 μl injected into each thigh muscle). Naïve mice were set as the control. PBMCs were isolated 24 h post-injection for bulk RNA-seq (n = 3) and mouse serum were collected for cytokine array (n = 5). Measurement of PreF-specific IgG PreF protein (SinoBiological, China) were coated into 96-well plates (Corning, China) at a concentration of 50 ng/ well and incubated overnight at 4 °C. The plates were washed with PBS containing 0.01% Tween-20 (PBST) and blocked for 2 h at 37 °C with 5% skim milk. Sera samples serially diluted were added and incubated for 1 h at 37 °C. binding IgG were determined using HRP-conjugated goat-anti-mouse IgG (1:30,000, Abcam, USA) for 1 h at 37 °C, respectively. TMB substrate (Solarbio, China) was used for development and the absorbance was read at 450 nm and 630 nm wavelength. The endpoint titer was calculated as previous reported^[177]47. RSV-specific neutralization assay A micro-neutralization assay was used to determine the anti-RSV Long and RSV B1neutralizing antibody titers, as our previously described^[178]47. The neutralization titers 50 (NT[50]) were defined as the equivalent of the serum dilution required for 50% neutralization of viral infection. Histopathological analysis After the RSV challenge, lungs were immersed in 4% paraformaldehyde and stained with hematoxylin and eosin (H&E) and periodic acid-Schiff (PAS) for pathological evaluation and mucus secretion. Four parameters associated with lung pathological changes were evaluated. Each of these parameters was scored separately as previously described^[179]49. RSV titer in lungs and nasal turbinates The RSV Long loads in the lung and nasal turbinates were quantified using qRT-PCR as previous reported^[180]47. After homogenization of lung tissue and nasal turbinates, the virus RAN was isolated and reverse transcription PCR were performed using HiScript II One Step qRT-PCR Probe Kit (Vazyme, China). Viral loads were calculated using a standard curve. The primer and probe sequences were as follows. L qPCR Forward: 5′-GAACTCAGTGTAGGTAGAATGTTTGCA-3′. L qPCR Reverse: 5′- TTCAGCTATCATTTTCTCTGCCAAT- 3′. L-probe: FAM-ATTTGAACCTGTCTGAACATTCCCGGTTGCAT- BHQ1. PBMCs, Splenocyte, dLN and LLPC cell preparation PBMCs were isolated using Lymphoprep^TM in SepMate™ PBMC Isolation Tubes according to the manufacturer’s instructions. Blood collection involved 400−500 μl of peripheral blood sampled via the buccal vein. Spleens were homogenized and passed through 40 μm cell strainers to produce single cells. LNs (Inguinal LNs) were harvested after immunization, then homogenized and passed through 40 μm cell strainers to produce single cells. LLPCs were isolate from bone marrow (BM) of mouse femur using a syringe. Folw cytometry detection Erythrocytes were lysed in RBC lysis buffer (eBioscience, USA). Cells were washed with DPBS, then blocked with anti-CD16/CD32 antibody (BD, USA), and stained with Live/Dead staining Kit (Invitrogen, USA) in DPBS. Effector T cell response PBMCs (1 × 10^6 cells/100 μl) were plated in 96-well V tissue culture plates and stimulated with RSV F peptide pools (A total of 130 peptides spanning the full-length F protein of RSV were synthesized as 15-mers, with 11 overlapping amino acids) and soluble co-stimulatory molecules CD28 and CD49d at a final concentration of 2 μg/mL and 1 μg/mL, and cultured at 37 °C in 5% CO[2] for 5 h, followed by the addition of Golgi-Plugs (BD Biosciences) at 37 °C with 5% CO[2] for another 6 h. After blocking and staining with the LIVE/DEAD, cells were stained with Percp/Cy5.5-CD3 (1:400, BD, USA), FITC-CD4 (1:400, BD, USA), and PE/Cy7-CD8a (1:400, BD, USA) at 4 °C for 0.5 h, then permeabilized with fixation/permeabilization solution (Invitrogen, USA). Samples were then intracellularly stained by incubating with BV421- TNF-α (1:250, BD, USA), PE-IL-2 (1:250, BD USA), APC-IFN-γ (1:250, BD, USA) or BV421-IL-4 (1:250, BD, USA), PE-IL-13 (1:250, BD, USA), and APC-IL-5 (1:250, BD, USA) for 1 h at 4 °C. Cells were washed and resuspended in 200 μL of PBS buffer for acquisition using BD Arial II (BD Bioscience), and analyzed using FlowJo v10. The gating strategy is shown in Fig. [181]S1. GC B and Tfh staining LNs and splenocytes were stained with BV421-CD4 (1:250, BD, USA), PE/Cy7-Fas (1:250, BD, USA), APC-GL7 (1:250, BD, USA), BV786-B220 (1:250, BD, USA), FITC-RSV F protein (1:100, Vazyme, China), PerCP/Cy5.5-CXCR5 (1:250, BD, USA), PE-PD-1 (1:250, BD, USA), PE/Dazzle594-CD19 (1:250, BD, USA), and APC-Cy7-IgD (1:250, BD, USA) for 1 h. The gating strategy is shown in Fig. [182]S2. MBC and LLPC staining Splenocytes and LLPC were stained with BV421-CD138 (1:250, BD, USA), PE-B220 (1:250, BD, USA), FITC-RSV F protein (1:100, Vazyme, China), PerCP/Cy5.5-CD19 (1:250, BD, USA), PE/Cy7-CD38 (1:250, BD, USA), APC-GL7 (1:250, BD, USA) and APC/Cy7-IgD (1:250, BD, USA) for 1 h. The gating strategy is shown in Fig. [183]S3 RNA-seq analysis RNA-seq and differential expression analysis were performed at Novogene (Beijing, China). After quality control, differentially expressed genes (DEGs) were defined as those exhibiting ≥2-fold up- or downregulation (|log₂ (fold change) | > 1, q < 0.05) in vaccinated versus naïve mice or versus vaccinated PreF mRNA group. Pathway analysis of significantly differentially expressed genes identified from the RNA-seq results was performed using KEGG in R Studio v.3.0.3. Pathways and biological processes with adjusted P-value ≤ 0.05 were considered significant. Multi-cytokine assay Mice serum was collected at 24 h post prime immunization for a multi-cytokine assay. Cytokine levels were evaluated with a Luminex 200 system using a panel of 23 mouse cytokines (LX-MultiDTM-23) according to the manufacturer’s instruction. The selected cytokines included Eotaxin/CCL11, G-CSF, GM-CSF, IFN-γ, IL-10, IL-12(p40), IL-12(p70), IL-13, IL-17A, IL-1α, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-9, GRO-α(Gro-a/KC/CXCL1), MCP-1/CCL2, MIP-1α/CCL3, MIP-1β, RANTES, TNF-α. Supplementary information [184]Supplementary materials^ (1.7MB, pdf) Acknowledgements