N1MΨU-modified mRNA vaccines break the mold in fish by enhancing innate immune activation
Dean Porter, Luc Jouneau, Mathilde Peruzzi, Bertrand Collet, Bernard Verrier, Pierre Boudinot
TL;DR
Modified mRNA vaccines in fish trigger strong immune responses and may work differently than in mammals.
Contribution
N1MΨU-modified mRNA vaccines in fish do not suppress innate immunity and activate unique immune pathways.
Findings
N1MΨU-modified mRNA vaccines induce robust type I interferon responses in rainbow trout.
Modified mRNA vaccines trigger autophagy, ubiquitination, and transcription pathways in fish.
Modified mRNA retains strong immunostimulatory capacity in fish, unlike in mammals.
Abstract
Messenger RNA (mRNA) vaccines have revolutionized immunization strategies in mammals, with chemical modifications such as N1-methyl-pseudouridine (N1MΨU) enhancing stability, translation, and reducing innate immune activation. However, the immunological implications of such modifications in non-mammalian species, particularly teleost fish, remain unclear. In this study, we evaluated the innate immune responses elicited by four vaccine platforms: a DNA vaccine, a live attenuated viral hemorrhagic septicemia virus (VHSV), an unmodified mRNA, and an N1MΨU-modified mRNA. All four vaccines encode the G protein of VHSV (GVHSV) in rainbow trout. Following intramuscular injection, both mRNA vaccine formats induced robust type I interferon (IFN) responses, comparable to those induced by the DNA and attenuated virus vaccines. Notably, the N1MΨU-modified mRNA vaccine did not suppress IFN…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsInvertebrate Immune Response Mechanisms · Aquaculture disease management and microbiota · RNA Interference and Gene Delivery
Introduction
In recent years, messenger RNA (mRNA) vaccines have emerged as an important vaccine platform, with the development of a successful vaccine against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) during the coronavirus pandemic. Vaccine mRNA encoding the SARS-CoV-2 spike protein or other mRNA vaccines developed in mammals are typically modified by incorporating pseudo-uridine (ΨU) or N1-methyl-pseudouridine (N1MΨU), which dampens innate immune sensing and increases mRNA translation in vivo.1^,^2^,^3^,^4^,^5 Modified nucleosides are used to improve the stability of the vaccine and to reduce mRNA interaction with cellular sensors, such as Toll-like receptor (TLR) 3, TLR7, TLR8, and RNA-sensing cytoplasmic helicases like Retinoic acid-inducible gene I RIG-I, leading to low immunogenicity.6^,^7^,^8 Importantly, studies in mice have shown that type I IFN-β responses at the site of injection of the mRNA vaccine were associated with the mRNA rather than the liponanoparticles (LNPs) themselves.9 Notably, while uridine-containing transcripts produced in vitro activate RNA-dependent protein kinase (PKR), which phosphorylates eIF-2α and blocks translation, ΨU mRNAs do not efficiently activate PKR, leading to better translation of the target protein.10
Modified mRNA vaccination in mice and non-human primates led to efficient induction of antigen-specific T-follicular helper cells and to high numbers of germinal centers and plasma cells.11 The germinal center reaction did not seem to be affected in mammalian models, with similar class-switching and somatic hypermutation rates observed in immunoglobulin genes in response to normal uridine and N1MΨU mRNA vaccine.2 The only difference revealed by single immune cell transcriptome sequencing was an additional cluster of B cells showing an interferon (IFN) response in mice vaccinated with normal uridine mRNA vaccine but not with the N1MΨU-vaccine. Various 5′-end and internal mRNA modifications have been tested in mammalian models, including ΨU, N1MΨU, 5-methoxyuridine, and cytidine modifications, with different impacts.7^,^12 However, high efficiency of translation has been reported for ΨU and N1MΨU in several mammalian models, and they remain the most commonly used modifications of mRNAs in mammalian vaccines.
A recent study by Ayad et al. has reported a proof-of-concept mRNA vaccine for fish using a non-modified mRNA encoding the G protein of viral hemorrhagic septicemia virus (G_VHSV_) encapsulated in LNPs.13 Immunization with this vaccine led to expression of the G protein, production of neutralizing antibodies (Abs) and offered almost complete protection of rainbow trout against a lethal viral challenge. N1MψU-modified mRNA encoding EGFP, encapsulated in lipid nanoparticles, has been demonstrated to successfully drive protein expression in Atlantic salmon cells for up to 11 days, and in muscle tissue.14 The expression of Infectious Hematopoietic Necrosis Virus (IHNV) glycoprotein after administration of a modified mRNA vaccine was also shown by the observed protective efficacy in rainbow trout.15 While these observations show promise for the future of fish mRNA vaccines, their composition and mechanisms of action need to be explored in the context of the fish immune system. The sites of initiation of B cell responses in fish have been recently identified in lymphoid aggregates associated with melano-macrophage centers;16^,^17 however, the mechanisms of the reaction leading to these structures, which are analogous to germinal centers, remain poorly understood and likely differ from those described in mammals. One key factor that may influence this process is the type I IFN response, which not only plays a central role in antiviral defense,18^,^19^,^20 but also enhances antibody responses by acting as an adjuvant in fish when co-expressed alongside DNA vaccines.21^,^22 These observations suggested that fish type I IFN might be involved in B cell co-stimulation via mechanisms different from those in mammals. While it has been well established that DNA vaccines against fish viruses induce robust type I IFN responses,23^,^24^,^25^,^26 the capacity of mRNA vaccines to induce such responses in these species, and the effects of type I IFN co-expression with the antigen, have not yet been explored. This raises important questions about whether mRNA vaccines can similarly induce type I IFN responses in fish and how the use of chemically modified mRNA might affect their immunogenicity across vertebrates.
In this study, we compared the innate reaction induced by uridine- and N1MΨU-modified G_VHSV_ mRNA vaccines formulated in liponanoparticles (LNP) in rainbow trout. We selected the VHSV G protein for this comparison because it has been shown to be the target of the protective neutralizing Ab response.23^,^27^,^28^,^29 We also included DNA and attenuated live vaccines against VHSV in the comparison. We analyzed the transcriptional responses to these four different vaccine platforms, all leading to expression of the same G_VHSV_ protein following intramuscular injection: a DNA vaccine, a non-modified mRNA vaccine, an N1MΨU-modified mRNA vaccine, and an attenuated VHSV strain. Interestingly, both unmodified mRNA and N1MΨU-modified mRNA vaccines triggered strong IFN responses at 3 and 7 days post-vaccination but with distinct kinetics. Transcriptomic analysis confirmed that N1MΨU modification does not reduce the capacity of mRNA to induce a type I IFN response in this fish model, as it drove robust induction of many IFN-stimulated genes (ISGs), similar to what was observed with the attenuated VHSV and non-modified mRNA vaccines. Unlike in mammalian systems, where N1mΨU modification typically dampens innate immune activation by RNA, our data reveal enhanced enrichment of autophagy, ubiquitination, and transcriptional pathways, surpassing the enrichment observed with the unmodified mRNA vaccine. These findings highlight the need for a deeper understanding of how mRNA vaccine modifications influence innate immune activation in fish models.
Results
Both unmodified and N1MΨU-modified mRNA vaccines lead to similar levels of VHSV G transcripts at 3 and 7 dpi
The G_VHSV_ protein sequence used for our nucleic acid (NA) vaccines13^,^23 was cloned from the attenuated VHSV strain, 25-111 ensuring all vaccines encoded the exact same epitope. Both mRNA vaccines were encapsulated in parallel in LNPs, which are currently the gold standard in mRNA vaccine delivery, as used in the coronavirus disease 2019 (COVID-19) vaccines in humans. The G_VHSV_ mRNA level was analyzed by quantitative reverse-transcription PCR (RT-qPCR) in the spleen, gut, and site of injection at two time points, 3 and 7 days post-injection (dpi) (Figure 1A). At three dpi, G_VHSV_ mRNA levels were highest in the fish vaccinated with the attenuated VHSV and both mRNA vaccines across all analyzed tissues, compared to fish injected with the DNA vaccine (Figure 1B). LNP-mRNA:G_VHSV_ (unmodified) and LNP-N1MΨUmRNA:G_VHSV_ (N1mΨU-modified) vaccines led to similar mRNA levels in all tissues, with the site of injection and spleen showing significantly higher levels than the gut. At 7 dpi, G_VHSV_ mRNA was detectable in all tissues, with the site of injection showing the highest expression levels and minimal levels in the gut. Injection of the DNA vaccine led to lower G_VHSV_ mRNA levels than the other vaccines. Again, mRNA vaccines led to comparable levels in all three tissues. The GVHSV mRNA levels detected at 7 dpi in the gut of fish vaccinated with the live-attenuated VHSV were higher than at 3 dpi, possibly reflecting systemic dissemination of the attenuated virus to this tissue. No G_VHSV_ mRNA was detected in the PBS control group at 3 or 7 dpi. Overall, our data show that both mRNA-based vaccines maintained similar mRNA levels between 3 and 7 dpi in central and mucosal lymphoid tissues, as well as at the site of injection.Figure 1. Experimental design and expression of viral glycoprotein mRNA(A) Experimental design. All groups were assessed at both 3 and 7 days post-immunization. (B) Bar plots showing the expression of VHSV G mRNA (strain 07.71) determined by RT-qPCR. The data represent fold change compared to DNA:G_VHSV_, as no expression was seen in the PBS controls. mRNA expression was analyzed in three tissues: spleen, gut, and the injection site, at two time points: 3 and 7 days post-injection (dpi). Statistical significance was assessed using a two-way ANOVA, with significant differences denoted by an asterisk (∗). n = 4 per group per time point. Data are shown as mean ± SEM. (C) Schematic overview of the prime-boost vaccination strategy for the serum neutralization experiment. (D) Serum neutralization comparing D90 against D0 for each group. The bar plot shows mean ± SEM. n = 6 per group, run in duplicates.
A further experiment to assess the levels of neutralizing antibodies after vaccination was carried out with serum taken 90 days post-prime vaccination (Figure 1C). Serum neutralization results showed partial neutralization in both unmodified mRNA- and N1mΨU-vaccinated fish, both significantly lower than PBS-injected fish (Figure 1D). Interestingly, unmodified mRNA showed significantly more neutralization 90 days after prime vaccination. As shown previously, AttVHSV offers almost complete neutralization with just a prime vaccination.13 These results demonstrate that both vaccines lead to VHSV glycoprotein expression.
Comparable sensing of N1MΨU and unmodified mRNA delivered in LNP to fish cells
We assessed whether N1MΨU- and unmodified single-stranded RNA (ssRNA) delivered in LNP are recognized differently by fish NA sensors. Following 24 h of treatment of the Chinook salmon epithelial cell line (CHSE-EC) with LNP-formulated mRNA:G_VHSV_ or N1mΨU-modified mRNA: G_VHSV_, several ISGs, including mx1-2, ifit5, and ch25h, were significantly upregulated to a comparable extent (Figure 2). Interestingly, irf3 was not induced, indicating that this ISG is likely not upregulated by ssRNA stimulation in CHSE-EC cells. No induction was seen after treatment with empty LNPs. While the mRNA vaccines lead to G expression at the protein level, which may induce IFN, this pathway cannot explain the upregulation of ISGs as early as 24 h post-treatment with LNPs. These findings therefore suggest that fish can detect both non-modified uridine and N1MΨU-modified mRNAs, likely via TLR7/TLR8 or by RIG-I and melanoma differentiation-associated protein 5 (MDA5). In contrast, polyinosinic:polycytidylic acid (Poly I:C) stimulation did not induce the expression of these genes in CHSE-EC cells, in line with previous reports showing that extracellular double-stranded RNA (dsRNA)30^,^31 does not lead to a significant type I IFN response in these cells.Figure 2ISG response to mRNA vaccines after 24 h in the CHSE-EC cell lineCHSE-EC cells are unable to respond to extracellular dsRNA. Fold change for each group is compared to the PBS controls. Data are shown as mean ± SEM. ^a^ and ^b^ denote a significant difference where p < 0.05.
N1MΨU and unmodified mRNA LNP vaccines induce divergent type I IFN responses in vivo in fish
To evaluate in vivo short-term innate and inflammatory responses induced by our different vaccine platforms, we first quantified the expression of several ISGs (mx1-3, rsad2, isg15, irf11, sifn, ifn1a3, and stat1a3) and key proinflammatory cytokines (tnfα and il1β) by qRT-PCR at 3 and 7 dpi with LNP (mRNA: G_VHSV_), LNP (N1MΨUmRNA: G_VHSV_), DNA, and attenuated VHSV vaccines. The same tissues as for G_VHSV_ transcript levels were analyzed, i.e., the site of injection, the spleen (secondary lymphoid tissue), and the distal intestine (mucosal tissue).
All tissues showed a clear upregulation of ISGs in fish vaccinated with attenuated VHSV, with a peak at 3 dpi and a slightly lower response at 7 dpi (Figure 3A). Similarly, the LNP (mRNA: G_VHSV_) vaccine elicited a strong ISG response, with peak expression at 3 dpi and a marginally reduced response at 7 dpi. In both cases, ISG mRNA levels remained highly upregulated compared to PBS controls. This upregulation was most pronounced in the spleen. In this tissue, the N1MΨU-modified RNA vaccine induced delayed ISG responses compared to the attenuated virus and the LNP (mRNA: G_VHSV_). Several ISGs, including isg15, stat1a3, mx1-3, and irf11, exhibited a marked increase between 3 and 7 dpi, with irf11 and stat1a3 shifting from a 1.9-fold change at 3 dpi to a 23.3-fold change and 7.2-fold changes, respectively, at 7 dpi. Other ISGs exhibited sustained levels of expression between 3 and 7 dpi, as shown in Figure 3B. In contrast, DNA-vaccinated fish showed the weakest ISG response, although an increase in response was seen between 3 and 7 dpi. The site of injection and gut exhibited similar overall patterns of induction, although responses were weaker in the gut except after injection of the attenuated vaccine. The proinflammatory cytokines showed limited or no upregulation in any tissue, other than in the gut of LNP (mRNA: G_VHSV_)-vaccinated fish at 3 dpi. This confirms that the vaccine-induced responses are driven by type I IFN, even in the case of the N1MΨU-modified mRNA vaccine, rather than by proinflammatory signaling. These observations therefore reveal that N1MΨU modification did not abolish recognition of the G_VHSV_ mRNA by fish ssRNA and, hence, did not block activation of the type I IFN pathway. The N1MΨU modification influenced the kinetics of the IFN response, but not its amplitude, within the analyzed time window. Although much lower, the type I IFN response induced by the DNA vaccine paralleled the response to the N1MΨU-modified vaccine.Figure 3. Expression of ISGs and innate cytokines after vaccination with different vaccine platforms at two time points(A) Heatmap of ISGs and innate cytokines. Data represent fold change compared to PBS controls at each time point. (B) Comparison of LNP (mRNA:G_VHSV_)- and N1MΨU-modified LNP (mRNA:G_VHSV_)-vaccinated fish at 3 and 7 dpi. N = 4 per group per time point. Data are shown as mean ± SEM. Significant differences are denoted by ^a, b,^ and ^c^ where p < 0.05.
As the fish belonged to the same genetically homogeneous (doubled haploid) strain, variations in the expression levels of ISGs induced by LNP-mRNA:GVHSV (unmodified) and LNP-N1MΨUmRNA:GVHSV were somewhat unexpected and may be due to a divergent distribution of cell types at the time of immunization, influenced by the particular immunological history of each individual.
Attenuated virus, DNA, and RNA vaccines induce well-distinct transcriptome changes: The N1MΨU modification significantly affects the innate response
To get a wider view of the effects of the different vaccine platforms, we employed a whole transcriptome sequencing approach to compare gene expression at 7 dpi in the spleen of vaccinated and control fish. 7 dpi was used due to the gene expression differences observed between mRNA vaccines in the qPCR analysis. Results from RNA sequencing (RNA-seq) and RT-qPCR were compared for selected genes, indicating that both methods revealed similar expression patterns (Figure S1). Results from differential analysis are provided in Tables S1–S4. Principal component analysis (PCA) showed a clear distinction between control, attenuated virus, and NA vaccine groups (Figure 4A). The use of a live attenuated vaccine compared with PBS explained 35.38% of the variance (Dim1), whereas the type and structure of NA vaccines explained 26.26% of the variance (Dim2). There was clear overlap of NA vaccines in both dimensions of the PCA, with mRNA-based vaccines clustering together, suggesting that they drive more similar, but potentially unique, transcriptomic responses, distinct from those induced by the DNA vaccine.Figure 4. Comparison of DEGs in the spleen after vaccination with different vaccine platforms at 7 dpi(A) Principal component analysis showing the distribution of all samples (n = 4 per condition). (B) Venn diagram showing upregulated DEGs after vaccination compared to PBS-vaccinated fish. Genes were considered as DEGs if they met the following criteria: adjusted p < 0.05 and log2foldchange (FC) > 1 or < −1. (C) Fold change distribution of DEGs in LNP (mRNA:G_VHSV_)-vaccinated fish (x = Log2(FC_RNAgVHSV)) and N1MΨU-modified LNP (mRNA:G_VHSV_)-vaccinated fish (y = Log2(FC_pseudoU-RNAgVHSV)). Average values are shown.
Differentially expressed genes (DEGs) were identified by comparing each vaccinated group against the PBS control group (Figure 4B; Tables S1–S4). Figure 4C shows the transcriptional response to the two mRNA vaccines, with a globally similar distribution of fold-change values for genes affected by both, such as typical IFN-induced genes rsad2, isg15 (three paralogs), mx, cmpk2, aste1, and herc3, as well as the trim gene btr26 and the long non-coding RNA ENSOMYG00000075651. Besides, significant numbers of genes specific to LNP (mRNA: G_VHSV_) or to N1MΨU-modified LNP (mRNA: G_VHSV_) were identified, including genes of the type I IFN response such as samd9l. Paralogs of calcoco2 or ifit5 were induced by LNP (mRNA: G_VHSV_), while other paralogs of calcoco2 and ifit5 were induced by both vaccines, illustrating possible differential responses of salmonid paralogs. Curiously, the gene encoding the DNA repair lyase enzyme apex1 was induced only by the N1MΨU-modified mRNA vaccine.
Immunization with NA vaccines drives a shared transcriptional response centered on antiviral innate immunity
A set of 2,504 ISGs were previously found to be significantly upregulated in rainbow trout following exposure to Poly I:C, in vitro or in vivo.32 As a synthetic analog of dsRNA, Poly I:C is a potent agonist of TLR3 and MDA5, and it triggers a typical type I IFN response that mimics viral infection. This makes it a relevant comparator for mRNA vaccine-induced immune responses.
Across all vaccinated groups, 450 genes were commonly upregulated, with 262 overlapping with the ISG list upregulated in response to Poly I:C (Figures 4B and 5A). Heatmap analysis of the top 50 highly upregulated genes shows key ISGs, including isg15, ifit5, aste1, cmpk2, herc3, mx1, and rsad2 (Figure 5B). Gene Ontology (GO) analysis indicated that these 450 common DEGs were associated with significant enrichment of terms referring to innate immune activation (GO:0045087) and antiviral defense (GO:0051607), as well as ubiquitination (GO:0000209, GO:0016567, GO:0006511) and protein phosphorylation (GO:0006468) (Figure 5C). These findings emphasize that the shared transcriptome response elicited by all vaccine platforms is largely IFN-driven.Figure 5. Convergent ISG response to vaccination(A) Venn diagram comparing common DEGs in all vaccinated groups with upregulated DEGs in response to Poly I:C. (B) Heatmap of the top 50 DEGs identified across all groups. Colors show Log2FC. The top 50 genes were selected on the average across all gene sets. (C) GO terms related to the DEG list shared among all conditions. (D) Influenza A pathway with log2FC expression mapped onto the pathway for each group. Red represents upregulated DEGs, while blue represents downregulation DEGs. Squares are shown in the following order: DNA:G_VHSV_, LNP (mRNA:G_VHSV_), LNP(N1MΨU:G_VHSV_), and attenuated VHSV.
Additionally, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis identified Influenza A (hsa05164) as a significantly enriched pathway, further underscoring the convergent antiviral innate response induced by vaccination (Figure 5D). Figure 5D shows the Influenza A pathway with log2FC gene expression mapped for each condition. Attenuated VHSV modulates the Influenza A pathway the most, followed by N1MΨU-modified LNP (mRNA:G_VHSV_), LNP (mRNA:G_VHSV_), and DNA:G_VHSV_, respectively.
A remarkable feature of this transcriptome analysis is the large number of genes modulated by the N1MΨU-modified vaccine compared to the unmodified mRNA vaccine. In contrast, a large majority of genes modulated by DNA vaccination are shared with both the mRNA and N1MΨU mRNA vaccines.
NA vaccines induce a specific set of genes involved in RNA detection, processing, and inflammation
There were 299 DEGs shared between (and restricted to) NA vaccine groups, with only 4 DEGs found in the Poly I:C induced ISG set.
Interestingly, we found upregulation of tlr7 within the spleen of all three NA-vaccinated fish (Figure 6). TLR7 is involved in the detection of ssRNA molecules and is important in detecting viruses. The attenuated vaccine upregulated tlr8 instead of tlr7 and tlr3, indicating potential differences in how the immune system perceives live-attenuated virus versus synthetic DNA/RNA-based vaccines, and how this affects the expression of RNA sensors. NA vaccines can activate multiple RNA-sensing pathways through receptors related to the immune response (TLR3, TLR7, TLR22, MDA5, CGAS, DICER1) or through NA receptors with direct antiviral activity (ADAR, PKR, IFIT1), leading to strong ISG induction.33 Our results suggest that different vaccine platforms may upregulate, and likely engage, distinct pattern recognition receptor (PRR) pathways but ultimately converge on a shared antiviral response.Figure 6RNA sensing receptor gene expression after vaccination with different vaccine platformsData shows log2FC.
Analysis of DEGs specifically induced by NA vaccination revealed upregulation of several genes involved in innate immune responses, particularly those associated with antiviral defense (but not necessarily the IFN pathway). Notably, ago2, an important component of the RNA interference pathway, was upregulated by all three NA vaccines. Similarly, samhd1, a known viral restriction factor, exhibited increased expression, potentially contributing to limiting pathways related to viral replication induced by NA vaccines. The glucocorticoid receptor (GR) nr3c1, which modulates inflammatory responses and can rapidly degrade mRNA when GR is bound to glucocorticoid, was upregulated, suggesting its involvement in immune regulation post-vaccination.34 Several genes related to the mitogen-activated protein kinase (MAPK) signaling pathway, including dusp4, dapk2, and dapk3, were upregulated. The expression of apol6, linked to cell death pathways during viral infections, further suggests an active antiviral response after vaccination. Several genes of the immunoglobulin superfamily, such as ENSOMYG00000067672 and ENSOMYG00000066228, were also upregulated.
Transcriptional responses specific to mRNA vaccines
When comparing the DEGs in the spleen of fish vaccinated with LNP(mRNA:G_VHSV_) or N1MΨU-modified LNP(mRNA:G_VHSV_) vaccines, we identified 606 shared DEGs, with 9 of these found in the Poly I:C-induced ISG gene list (Figure 7A). Among these, several antiviral genes were significantly upregulated and were found within the top 50 induced genes (Figure 7B), encoding irf1, which is a key transcription factor in type I IFN signaling and antiviral responses; ax1, a receptor tyrosine kinase involved in IFN modulation; sqstm1, involved in autophagy and immune signaling; and scarb1, a scavenger receptor involved in lipid metabolism. SCARB1 detects cholesterol in high-density lipoproteins and is known to play a role in modulating virus replication in Hepatitis C.35 GO analysis showed upregulation of terms related to ubiquitination (GO:0016567, GO:0006511) and transcription related to RNA polymerase II (GO:0045944, GO:0006357, GO:0000122) (Figure 7C). KEGG pathway analysis further highlighted activation of autophagy (hsa04140, hsa04137) and ubiquitination (hsa04120) pathways, reinforcing the idea that mRNA vaccines trigger antiviral defense mechanisms (Figure 7D).Figure 7DEGs specific to mRNA vaccines(A) Venn diagram comparing DEGs identified in both mRNA vaccines with the Poly I:C ISG list. (B) Heatmap of the top 50 induced DEGs in both mRNA vaccines. Data show log2FC. (C) GO analysis showing significantly enriched terms after vaccination with either mRNA vaccine. (D) Significantly enriched KEGG pathways following vaccination with either mRNA vaccine. For GO terms or KEGG pathways to be considered significantly enriched, they had to meet the following criteria: gene count >10, Benjamini value < 0.05, and fold enrichment > 1.5.
Analysis of DEGs unique to N1MΨU-modified mRNA vaccination pointed to a distinct transcriptional response compared to unmodified mRNA vaccines. A total of 1,674 DEGs were exclusive to N1MΨU-modified LNP (mRNA:G_VHSV_)-vaccinated fish, with 8 found in the Poly I:C induced DEGs, encoding important regulators of immune signaling, autophagy, and cellular stress responses (Figure 8A). Among these, stat5a and stat5b, key transducers of cytokine signaling, were significantly upregulated, suggesting enhanced immune activation and potential effects on T cell responses. Additionally, ULK1, a master regulator of autophagy, was enriched, indicating that N1MΨU vaccination may promote antigen processing and viral clearance via autophagic pathways. Several ISGs (znf395, t53inp1, ddx17, ifit2, pik3r3) were also upregulated, as well as effectors of the NF-κB signaling pathway (nlrP12, tnfaip, tnfsf21), with several of these found within the top 50 induced genes (Figure 8B). GO analysis revealed the upregulation of several pathways uniquely associated with N1MΨU-modified mRNA vaccination, including those related to autophagy (GO:0006914, GO:0000045, GO:0000422), ubiquitination and protein degradation (GO:0016567, GO:0043161, GO:0000209, GO:0006511, GO:0070936), and transcriptional regulation (GO:0045944, GO:0006357, GO:0000122, GO:0045892, GO:0006355, GO:0045893) (Figure 8D). Additionally, chromatin remodeling (GO:0006338), vesicle-mediated transport (GO:0016192), and circadian rhythm regulation (GO:0042752) were also significantly enriched, suggesting a broader impact of N1MΨU modification on cellular homeostasis and immune activation. KEGG pathway analysis showed significant enrichment of “upregulation of autophagy” (hsa04140) and “mitophagy” (hsa04137) in N1MΨU-modified LNP (mRNA:G_VHSV_)-vaccinated fish, suggesting enhanced antigen processing and viral clearance (Figure 8E). Autophagy is essential for immune activation, while mitophagy regulates mitochondrial quality control, aiding immune cell function. Additionally, the ubiquitin-mediated proteolysis pathway (hsa04120) was enriched, indicating that N1MΨU vaccination may optimize protein turnover, enhancing the regulation of immune signaling and improving antiviral responses. This shift toward auto/mitophagy and ubiquitination is further confirmed by the comparison of the 305 DEGs shared between N1MΨU-modified LNP (mRNA:gVHSV) and the attenuated VHSV, with GO terms relating to autophagy (GO:0006914) and ubiquitin-dependent protein catabolic processes (GO:0006511).Figure 8DEGs specific to N1MΨU-modified LNP mRNA:gVHSV and non-modified LNP mRNA:gVHSV(A) Venn diagram comparing DEGs uniquely found in N1MΨU-vaccinated or non-modified mRNA-vaccinated fish with the Poly I:C-induced ISG list. (B) Heatmap of the top 50 induced DEGs in non-modified mRNA-vaccinated fish. Data show log2FC. (C) Heatmap of the top 50 induced DEGs in N1MΨU-vaccinated fish. Data show log2FC. (D) GO analysis showing significantly enriched terms after vaccination with N1MΨU-modified LNP (mRNA:gVHSV). (E) Significantly enriched KEGG pathways following vaccination with N1MΨU-modified LNP (mRNA:gVHSV). For GO terms or KEGG pathways to be considered significantly enriched, they had to meet the following criteria: gene count >10, Benjamini value < 0.05, and fold enrichment >1.5.
The non-modified mRNA vaccine led to a lesser transcriptome modulation, with only 108 DEGs specific to the LNP (mRNA:gVHSV)-vaccinated group, 5 of which were found in the Poly I:C-induced ISG DEGs (Figure 8A). The top 50 most induced genes from this set of DEGs are shown in Figure 8C. Fish vaccinated with the non-modified mRNA vaccine showed upregulation of genes encoding nt5e, involved in adenosine metabolism; pcyt1b, a regulator of phosphatidylcholine biosynthesis; and upf3a, a component of the nonsense-mediated mRNA decay (NMD) pathway, suggesting potential impacts on mRNA stability and processing. Additionally, raver2 and ssbp4 were upregulated, which are implicated in RNA splicing and mitochondrial DNA stability, respectively, indicating altered RNA processing and mitochondrial function in response to the non-modified vaccine. No GO or KEGG pathways were found to be significantly enriched using these DEG sets. In addition, we identified 47 ISGs from the 151 DEGs shared between groups vaccinated with LNP (mRNA:gVHSV) and the attenuated VHSV (but not with the N1MΨU-modified mRNA vaccine), which also confirm different involvement of genes of the type I IFN response and immune activation (saa1, nod2, and ifit5), as well as differences in lipid metabolism (ffar2, apoc1, and p2rx5). These data suggest that there are subtle differences in genes encoding factors of RNA metabolism between N1MΨu- and uridine-based mRNA vaccines.
Discussion
Our findings challenge the notion that N1MΨU modification reduces the capacity of the RNA molecule to trigger a type I IFN response in fish, a well-accepted assumption in mammalian models.6^,^36^,^37 Although we do not have molecular evidence to demonstrate this, these results raise questions about differences in RNA sensing, both in terms of individuals, as seen by the variability of responses to the COVID vaccine, and in terms of differences between species. Further research is needed to understand the extent of these differences among mammals, fish, and avian species.
In mammalian models, mRNA vaccines have been shown to induce IFN and inflammatory responses. In mice, N1MΨU-modified mRNA vaccines trigger rapid and localized infiltration of leukocytes (neutrophils, monocytes, and dendritic cells) to the lymph nodes and site of injection, accompanied by upregulation of type I IFN-related genes38 Similar responses have been observed in humans, where vaccination with the N1MΨU-modified Pfizer-BioNTech mRNA vaccine (BNT162b2) led to a clear induction of IFN responses.39 Both modified and unmodified mRNA vaccines have been shown to activate innate RNA sensors such as TLR7/8 and RIG-I-like receptors (RLRs), resulting in the expression of IFN-stimulated genes and inflammatory cytokines that promote the recruitment and activation of innate immune cells.40 Although few studies directly compare unmodified and N1MΨU-modified mRNA vaccines, available data suggest that while both formulations can activate IFN and inflammatory pathways, these responses tend to be reduced following N1MΨU modification.41
Although N1MΨU modification is known to dampen innate immune detection of synthetic mRNA in mammals by reducing activation of TLR3, TLR7, TLR8, and RLRs,3 our results demonstrate that while the N1MΨU-modified mRNA vaccine alters the kinetics of the immune response, it does not dampen its overall magnitude. An important consideration is that these nucleoside modifications may not silence RNA sensing as effectively in fish as they do in humans and mice. Teleosts possess a distinct repertoire of pattern recognition receptors, including RIG-I, MDA5, and LGP2, whose ligand specificities may differ from those of their mammalian counterparts.39^,^40 In several fish species LGP2 functions as a positive regulator of antiviral signaling, amplifying IFN responses rather than suppressing them, which may increase sensitivity to exogenous RNA regardless of nucleoside modification, particularly during early responses.41^,^42 Furthermore, teleost TLR7 retains broad recognition of uridine-rich ssRNA and may not be fully inhibited by N1MΨU substitution.43 These factors suggest that RIG-I, MDA5, and TLR7 are the most likely sensors capable of detecting N1MΨU-modified mRNA in fish. A difference in TLR3 sensing is not likely to explain our observations, since this TLR detects dsRNA. In addition, CHSE cells, in which dsRNA sensing is defective,30 still respond to both LNP mRNA vaccines, with modified and unmodififed mRNA. Indeed, while extracellular poly I:C stimulation failed to elicit a typical ISG response in CHSE-EC cells, as expected in this cell line, both unmodified and N1MΨU-modified mRNA vaccines led to a clear upregulation of ISGs in these cells. This observation reinforced the specificity of ssRNA sensing in ultimately driving a robust type I IFN response but indicated that N1MΨU modification did not hamper the induction of this response. Similarly, in rainbow trout tissues, we detected, after administration of the N1MΨU-modified mRNA vaccine, a delayed yet sustained activation of ISGs such as mx1-3, isg15, and stat1a3, with the response intensifying between 3 and 7 dpi. This delayed activation contrasts with the rapid peak of ISG responses seen in fish vaccinated with unmodified mRNA, suggesting that the N1MΨU modification likely influences the kinetics of the immune response rather than reducing its strength. While unmodified mRNA vaccines triggered strong initial ISG responses, the N1MΨU-modified vaccine led to a more sustained immune activation, particularly in the spleen and site-of-injection tissues, further supporting the idea that N1MΨU modification does not impede antiviral responses in fish. Moreover, the lack of significant changes in proinflammatory cytokines (tnfα, il1β) across most tissues further suggests that the immune response triggered by mRNA vaccines is primarily mediated through a type I IFN signaling rather than inflammation, as observed with the innate reaction induced by the attenuated virus or the DNA vaccine. Many more experiments will be required to clarify this question, since fish are highly diverse and their repertoire of sensors remains largely uncharacterized. In fact, variations in mRNA sensing in fish, as well as in other vertebrate species, will need to be explored to better understand how mRNA vaccines can be developed in farmed species.
Differences between responses induced by modified and unmodified mRNA might be explained by differential degradation of the modified mRNA. However, such degradation should have been detected by our PCR quantification. Our previous experiment in zebrafish with unmodified mRNA showed long-term expression of EGFP (>7 dpi), suggesting that mRNA in fish is relatively long-lasting,13 since the protein half-life is about 24 h. N1MψU-modified mRNA should be less prone to degradation, and such mRNAs encoding EGFP encapsulated in LNP have been shown to drive protein expression in Atlantic salmon cells for up to 11 days.14 We therefore think that the kinetics of mRNA degradation are unlikely to explain our observations.
Interestingly, after transcriptional analysis, we observed that N1MΨU modification led to increased expression of genes involved in autophagy, ubiquitination, and transcriptional regulation, highlighting that N1MΨU modification may augment the translation of the vaccine mRNA, leading to a potent antiviral response at later time points. This is consistent with studies suggesting that N1MΨU-modified mRNA vaccines can enhance antigen expression, allowing for sustained translation.6^,^41 In mammalian models, translation of both modified and unmodified mRNA does not result in differences in the number of miscoded peptides produced.42 Our data support the idea that the N1MΨU modification in mRNA vaccines does not reduce the innate IFN response in fish, and in fact, may promote a more sustained reaction compared to unmodified mRNA vaccines during the first week after vaccination.
As rhabdovirus glycoproteins can induce IFN-dependent responses, it could be argued that the glycoprotein itself could be responsible for a fraction of the IFN response observed; thus, differences between responses induced by modified and unmodified mRNA vaccines could be masked by IFN induced by the VHSV G. However, this hypothesis can be ruled out, as we observed similar induction of several genes, with high fold changes, in CHSE cells as early as 24 h post-transfection with both modified and unmodified mRNA LNP. In salmonid cells at 20°C, this observation indicates that the ISG inductions observed are due to mRNA sensing and that both modified and unmodified mRNA LNP can mediate this response.
Another potential effect that was not studied in this experiment was the role of LNPs in modulating the immune response. It has previously been shown in non-human primates that LNPs play a key role in dampening innate immune responses and that N1MΨU modification alone may not reduce innate immune responses.43 In our experiment, we checked only CHSE cell lines and observed no significant increase in any antiviral genes after stimulation with LNPs alone for 24 h, indicating that under these conditions the impact of LNPs is minimal. For the in vivo studies, we chose to keep the formulation of LNPs the same; however, future experiments are needed to determine whether different LNP formulations can improve or affect both the innate and adaptive immune responses driven by vaccination. Fish have unique cell membranes that can alter the saturation/unsaturation ratio of their membrane lipids in response to temperature and stress.44 As such, it may be of interest for future studies to determine whether alternative LNP compositions can reduce any potential adjuvant effect of the LNPs themselves45 and whether they are better suited for fish models.
The comparison of N1MΨU-modified and non-modified mRNA vaccines revealed key differences in gene expression, suggesting distinct impacts on innate immune activation. While N1MΨU modification is known to reduce direct innate immune sensing of synthetic mRNA in mammalian models,3^,^7^,^46 we observed a broad upregulation of genes involved in autophagy (ulk**1, gaa), ubiquitination (ncoa**4, usp**13), and transcriptional regulation (tef*,* clock*,* tp53inp1) in N1MΨU-vaccinated fish. This suggests that, rather than suppressing immune activation, N1MΨU modification might enable more efficient antigen expression, leading to a strong and coordinated adaptive antiviral response. In contrast, fish vaccinated with non-modified mRNA showed increased expression of upf3a, a key factor in the NMD pathway, suggesting that it may reduce mRNA stability and increase turnover.47 The upregulation of raver2 and ssbp4 further suggests alterations in RNA splicing and mitochondrial stability, which may indicate a stronger cellular stress response due to innate immune detection of the unmodified mRNA.48^,^49 Additionally, the upregulation of nt5e and pcyt1b suggests metabolic adjustments, potentially as a compensatory response to dampen the immune activation triggered by non-modified mRNA.50^,^51 These results show clear differences in the innate transcriptional responses driven by modified and non-modified vaccines. Future studies are needed to evaluate the impact and mechanisms of such differences on the adaptive immune response induced by mRNA vaccines.
While our primary focus is on the innate immune responses induced by mRNA vaccines, it is important to contextualize these findings alongside responses elicited by the DNA vaccine and the attenuated VHSV. The live attenuated virus provokes a broad and robust innate response due to active replication and presentation of multiple viral Pathogen-associated molecular patterns (PAMPs). The DNA vaccine delivers antigen-encoding plasmid DNA that must enter the nucleus for expression, typically inducing a distinct innate profile characterized by cytosolic DNA-sensing pathways. In contrast, mRNA vaccines provide a defined dose of synthetic RNA without replication, leading to an IFN response mediated primarily via ssRNA sensors. These differences in vaccine type, antigen load, and replication potential inherently shape their unique innate immune activation patterns and should be considered when interpreting comparative results.
In our study, both non-modified and N1MΨU-modified mRNA vaccines against VHSV generated measurable neutralizing antibodies that are typically associated with host protection against rhabdovirus infections.13^,^52^,^53 These data are consistent with previous findings showing associated protection of rainbow trout against live VHSV13 for non-modified mRNA vaccines and protection against IHNV15 for N1MΨU-modified vaccines. However, it is essential to examine the innate transcriptional signatures unique to each vaccine platform to fully understand how these upstream mechanisms may shape the subsequent adaptive immune response.
Building on these findings, the next step will be to explore how the observed innate immune responses translate into adaptive immunity, particularly in terms of T cell and B cell activation, clonal expansion, and antibody repertoire development. The delayed but sustained IFN response suggests that mechanisms affecting antigen availability and presentation are affected by N1MΨU modification, with a possible impact on germinal center reactions and the diversity of the neutralizing antibody response. In mammalian models, the incorporation of N1MΨU into vaccines resulted in similar but slightly increased germinal center responses, in terms of class switching and somatic hypermutation rates, compared to non-modified mRNA vaccines.2 Investigating whether N1MΨU vaccines influence TCR clonotype distribution or lead to differences in the B cell repertoire in fish will be essential to determine their long-term efficacy.
Another interesting avenue would be to characterize the innate immune landscape at the time of booster vaccination. Few studies, particularly in non-mammalian models, have examined how pre-existing adaptive immunity influences early innate responses upon re-exposure to antigen. Understanding the kinetics of IFN signaling, antigen-presenting cell (APC) activation, and inflammatory cytokine profiles during booster vaccination could provide insight into how the immune system integrates primary and secondary immune responses in fish. This is particularly relevant for optimizing vaccine regimens in aquaculture, where temperature can dramatically influence the immune response, potentially affecting the development of immune memory. Future studies combining longitudinal transcriptomic analysis with functional assays of the antibody response will be invaluable for understanding the interplay between innate priming and the adaptive immune response.
Materials and methods
Fish experiments
Adult rainbow trout were reared and vaccinated at the fish facilities at the Institute National de la Recherche en Agriculture et Environnement (INRAE, Jouy-en-Josas, France). All experiments were carried out in accordance with the European Union guidelines for the handling and welfare of laboratory animals (https://ec.europa.eu/environment/chemicals/lab_animals/index_en.htm). The experimental protocols were approved by the INRAE institutionalethics committee “COMETHEA” (DAP 18-37). Fish were kept in separate groups within the same tanks at 15°C. Adult rainbow trout (300 g) were vaccinated by intramuscular injection below the dorsal fin using a variety of vaccine platforms. For the negative control, fish were vaccinated with 1× PBS (Gibco). For the NA vaccines, each fish was vaccinated with 100 μL containing 20 μg of vaccine, whereas fish vaccinated with the attenuated VHSV were injected with 100 μL containing 1 × 10^5^ plaque-forming units. The doses were determined based on previous experiments and conditions offering protection13^,^23^,^24^,^28. At each time point (3 and 7 dpi), tissue samples (spleen, distal intestine [gut] and site of injection) were taken from each fish. Samples were stored in RNAlater (Sigma) and kept at −80°C until further processing. To ensure that samples were taken from the site of injection rather than surrounding tissue, it was determined as the area 1 cm above the lateral line and in line with the point at which the distal end of the dorsal fin meets the rest of the body. When sampling the site of injection, a 1 cm × 1 cm area around this point was excised. At 3 or 7 dpi, this site was still visible. The tissue was then placed in RNAlater until further use. Half of each site of injection sample was used for RNA extraction.
A separate experiment was conducted using the same setup; however, rainbow trout of a similar age were primed and boosted 30 days later, with blood samples collected for serum at 90 days post-initial vaccination (Figure 1C). Blood was allowed to coagulate at 4°C overnight before being centrifuged at 1500 ×g to allow collection of serum, which was then stored at −80°C until further use.
Serum neutralization assays
Serum neutralization assays were carried out as previously described by Ayad et al. and Castro et al.13^,^54
Cell culture
The CHSE-EC cell line was cultured at 20°C as described previously.55 Cells were stimulated in 6-well plates in 2 mL of medium for 24 h with the following concentrations: PBS 1× (Gibco), Poly I:C 33 ng/well (Sigma), LNP (mRNA:G_VHSV_), and LNP (N1MΨUmRNA:G_VHSV_) at 20 μg per well. Cells were plated at 2 million cells per well.
Vaccine preparation
The DNA vaccine comprised the expression plasmid pcDNA3.1 encoding the G protein of VHSV.23 The attenuated vaccine contains the sequence encoding the glycoprotein of the thermoresistant, attenuated VHSV strain 25.111, which does not replicate well at 16°C.56 To design modified and unmodified mRNA vaccines, we cloned the G-protein-encoding sequence into a master DNA vector optimized for In vitro transcription, as extensively described in Linares et al.57 Briefly, it uses a 5′ UTR human beta-globin sequence and 2 copies of the same beta-globin as the 3′ UTR. It also permits the addition of a 148-poly(A) tail during the in vitro transcription process. The mRNAs used throughout this study were prepared and purified by Quantoom (Nivelles, Belgium) using the CleanCap strategy. They were either VHSV glycoprotein G_VHSV_ mRNA or N1MΨU-modified G_VHSV_ mRNA using N1MΨU. Importantly, full substitution was used, and the in vitro transcription was performed in the presence of ATP, guanosine triphosphate (GTP), CTP, and modified N1MΨU triphosphate (without UTP). Hence, we could exclude the presence of non-modified UTP bases in our modified mRNA. This full substitution does not impair in vitro transcription. Formulations of these mRNAs were carried out as described by Ayad et al.,13 using a mixture of four lipids with a molar ratio of 50/10/38.5/1.5 (DLin/DSPC/cholesterol/DMG-PEG2000), and mRNA in 100 mM sodium acetate, pH 5.5, at the desired concentration. LNP were formulated using a NanoAssemblrTM benchtop instrument equipped with a microfluidic cartridge (Precision NanoSystems Inc., Vancouver, BC, CA). The recovered LNP dispersion was stored at 4°C until use. mRNA encapsulation efficiency was quantified using the Quant-IT RiboGreen RNA Assay Kit (InvitroGen by Thermo Fisher Scientific, France) according to the manufacturer’s instruction. The particles were characterized by dynamic light scattering (DLS) using a Zetasizer Nano ZS427 plus (Malvern Instruments, UK). The complexation of mRNA within the formulations was assessed using a gel retardation assay by electrophoresis.
RNA extraction
RNA from tissue samples was extracted using TRI reagent (also sold as TRIzol) (Sigma) following the manufacturer’s instructions. For cell culture experiments, RNA was extracted using the Qiagen RNAeasy Mini Kit according to the manufactures’ instructions. Quality control of RNA samples was performed using a Nanodrop 2000 machine. RNA was stored at −80°C until further use for RNA-seq and cDNA synthesis for targeted gene expression.
Real-time quantitative PCR
Total RNA (1 ng) was used as a template for reverse transcription and cDNA synthesis using the iScript Adv cDNA kit for RT-qPCR (Biorad), following the manufacturer’s instructions. The cDNA was diluted 10× with RNAse-free water (Sigma) and stored at −20°C until further use.
The cDNA was mixed with forward and reverse primers (Table S5) and iTaq universal SYBR Green Supermix (Biorad) in real time-qPCR plates (Eppendorf).58^,^59^,^60^,^61^,^62^,^63 Amplification was performed using the CFX connect Real-time PCR machine (Biorad) with the following cycling program: initial denaturation at 95°C for 3 min, followed by 40 cycles of 10 s at 95°C and 30 s at 60°C. For each biological replicate, the mean threshold cycles (Ct) values for each target gene were calculated from duplicate reactions and normalized against two housekeeping genes (elf1α and rps29). Gene expression and fold changes relative to PBS controls were calculated using the Pfaffl method in Excel. For each primer set, efficiencies were calculated by linear regression using a 5-fold serial dilution prepared from a pooled cDNA sample generated from all experimental samples.
RNA-seq and mapping of reads
Library preparation and sequencing were carried out by the Institute for Integrative Biology of the Cell (I2BC), Gif-sur-Yvette, using the NextSeq 2000 sequencer (Illumina). Raw data were processed using bcl2fastq v4.1.5 (for demultiplexing), cutadapt v1.8.3 (adapter and sequence quality trimming), and FastQC v0.12.1 (quality control), with an average of 31 million reads per sample.
Sequences were aligned to the rainbow trout genome (USDA OmykA 1.1 assembly – Ensembl v113 annotation) using STAR v2.7.10b, with 88% mapping to the rainbow trout genome and 48% of reads assigned to genes.
DEGs
Identification of possible outliers was performed through graphical analysis, including PCA plots in R using the FactoMineR package (v2.11). DEGs were identified using the DESeq2 package64 (v1.38.3) in R by comparing vaccinated conditions against PBS-vaccinated fish as negative controls. Genes were considered differentially expressed if they had a log2fold change of >1 (upregulated) or < −1 (downregulated) and a Benjamini-Hochberg adjusted p value of <0.05. Annotations of gene models predicted in the latest rainbow trout genome assembly were downloaded from the Ensembl portal (Ensembl 113) using Biomart and complemented with the results of a BLAST-based approach to generate a list of best-matching hits against various databases. Proteomes from each species were retrieved, filtered as described in Clark et al., 2023, and subjected to a BLASTp search against the human and zebrafish proteomes.32
Gene set enrichment analysis
Gene set enrichment analysis was carried out using GO and KEGG pathway analysis via the DAVID web interface.65^,^66 The predicted GO terms and KEGG pathways were based on lists of official gene symbols corresponding to human orthologs, without expression or fold change values. The Pathview web interface was used to plot gene expression data, including official gene symbols and log2fold change values, onto target KEGG pathways.
Data analysis
Gene expression data from qPCR were displayed and analyzed using one- or two-way ANOVA in GraphPad Prism 9.
Data and code availability
Sequence data supporting the findings of this study have been deposited in the BioProject database at the National Center for Biotechnology Information under accession number Genbank - BioProject: PRJNA1309644.
Acknowledgments
We would like to thank the IERP team at INRAE for their assistance with the fish experiments. We would also like to thank Quantoom and Pierre Libeau for their help in preparing the LNP mRNA vaccine formulations. This research was funded by the 10.13039/501100001665Agence Nationale de la Recherche ANR-21-35–0019CE (LipofishVac), by the ERANET project Nucnanofish (ANR-21-ICRD-0009), and by institutional grants from 10.13039/501100022077INRAE.
Author contributions
Conceptualization, D.P., B.C., P.B., and B.V.; funding acquisition, B.V. and P.B.; methodology, D.P., L.J., M.P., B.C., B.V., and P.B.; investigation, D.P., M.P., B.C., and P.B.; resources, L.J., B.C., and B.V.; visualization, D.P. and L.J.; software, D.P. and L.J.; writing - original draft, D.P. and P.B.; writing - review & editing, D.P., L.J., M.P., B.C., B.V., and P.B.
Declaration of interests
All authors declare no conflict of interests.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Jackson L.A.Anderson E.J.Rouphael N.G.Roberts P.C.Makhene M.Coler R.N.Mc Cullough M.P.Chappell J.D.Denison M.R.Stevens L.J.An m RNA Vaccine against SARS-Co V-2—Preliminary Report N. Engl. J. Med.38320201920193110.1056/NEJ Moa 202248332663912 PMC 7377258 · doi ↗ · pubmed ↗
- 2Kuzmin I.V.Soto Acosta R.Pruitt L.Wasdin P.T.Kedarinath K.Hernandez K.R.Gonzales K.A.Hill K.Weidner N.G.Mire C.Comparison of uridine and N 1-methylpseudouridine m RNA platforms in development of an Andes virus vaccine Nat. Commun.152024642110.1038/s 41467-024-50774-339080316 PMC 11289437 · doi ↗ · pubmed ↗
- 3Morais P.Adachi H.Yu Y.-T.The Critical Contribution of Pseudouridine to m RNA COVID-19 Vaccines Front. Cell Dev. Biol.9202178942710.3389/fcell.2021.789427 PMC 860007134805188 · doi ↗ · pubmed ↗
- 4Mulligan M.J.Lyke K.E.Kitchin N.Absalon J.Gurtman A.Lockhart S.Neuzil K.Raabe V.Bailey R.Swanson K.A.Phase I/II study of COVID-19 RNA vaccine BNT 162b 1 in adults Nature 586202058959310.1038/s 41586-020-2639-432785213 · doi ↗ · pubmed ↗
- 5Pardi N.Hogan M.J.Pelc R.S.Muramatsu H.Andersen H.De Maso C.R.Dowd K.A.Sutherland L.L.Scearce R.M.Parks R.Zika virus protection by a single low-dose nucleoside-modified m RNA vaccination Nature 543201724825110.1038/nature 2142828151488 PMC 5344708 · doi ↗ · pubmed ↗
- 6Andries O.Mc Cafferty S.De Smedt S.C.Weiss R.Sanders N.N.Kitada T.N 1-methylpseudouridine-incorporated m RNA outperforms pseudouridine-incorporated m RNA by providing enhanced protein expression and reduced immunogenicity in mammalian cell lines and mice J. Control. Release 217201533734410.1016/j.jconrel.2015.08.05126342664 · doi ↗ · pubmed ↗
- 7KarikóK.Buckstein M.Ni H.Weissman D.Suppression of RNA Recognition by Toll-like Receptors: The Impact of Nucleoside Modification and the Evolutionary Origin of RNA Immunity 23200516517510.1016/j.immuni.2005.06.00816111635 · doi ↗ · pubmed ↗
- 8Kormann M.S.D.Hasenpusch G.Aneja M.K.Nica G.Flemmer A.W.Herber-Jonat S.Huppmann M.Mays L.E.Illenyi M.Schams A.Expression of therapeutic proteins after delivery of chemically modified m RNA in mice Nat. Biotechnol.29201115415710.1038/nbt.173321217696 · doi ↗ · pubmed ↗