Same RSV antigen, two vaccine platforms, two different immune maps in mice

Two experimental respiratory syncytial virus vaccines built on exactly the same prefusion F antigen do not simply produce stronger or weaker immunity depending on the platform — they produce immunity in different places, according to a head-to-head mouse study published in EMBO Molecular Medicine. When researchers at Sichuan University compared a lipid nanoparticle-formulated self-amplifying RNA vaccine with an adenovirus type 5-vectored vaccine, both delivered by intramuscular injection, the adenoviral vector preferentially built systemic immunity, while the self-amplifying RNA preferentially built immunity within the lung itself.

The study, led by Yao Zhang, Yu Qin and colleagues, was designed to answer a question that has become increasingly pressing as RNA-based and adenoviral RSV vaccines both advance: does the vaccine platform itself, independent of the antigen and the delivery route, shape where in the body antiviral immunity is organized? To find out, the team constructed two vaccines encoding the identical stabilized prefusion F protein, based on the DS-Cav1 design that carries four stabilizing amino acid substitutions. Female BALB/c mice received either 2 micrograms of LNP-RSV-preF saRNA or 2.5 × 10⁹ viral particles of Ad5-RSV-preF intramuscularly on days 0 and 21, followed by immune profiling at multiple time points extending to 120 days after immunization.

Quality control confirmed that both platforms were functional. The self-amplifying RNA was verified as intact by gel electrophoresis, and the lipid nanoparticles, built from SM-102, DOPE, cholesterol and a PEG-lipid, showed particle sizes around 150 nanometers with an encapsulation efficiency of approximately 90 percent. The adenoviral particles measured roughly 123 nanometers. Both vaccines drove detectable prefusion F protein expression in transfected HEK293T cells in vitro, confirming that each could deliver and express the antigen, although the authors note that the denaturing immunoblot conditions used could not assess whether the expressed antigen retained its prefusion conformation.

The earliest and most striking divergence appeared in serum antibodies. Fourteen days after the prime immunization, Ad5-RSV-preF induced serum prefusion-specific IgG titers approximately nine-fold higher than those produced by the RNA vaccine, indicating markedly faster systemic humoral priming by the adenoviral vector. After the booster dose on day 21, the RNA vaccine’s titers rose rapidly to levels comparable to the adenoviral vaccine, but by day 28 after the boost the adenoviral platform again held higher serum IgG levels. Both vaccines produced an IgG2a to IgG1 ratio greater than one, a profile the authors describe as Th1-skewed, a pattern considered favorable in RSV vaccine development because Th2-skewed responses have historically been associated with the risk of vaccine-associated enhanced respiratory disease.

Neutralizing antibody measurements reinforced the systemic advantage of the adenoviral platform. In live-virus neutralization tests using serum collected 28 days after the booster, the geometric mean titer of the Ad5 vaccine against RSV strain A2 was 4898, compared with 1840 for the RNA vaccine; against RSV strain B the corresponding values were 1558 and 565. Yet the picture reversed in the airways. In bronchoalveolar lavage fluid collected on day 49, the RNA vaccine elicited significantly higher levels of IgG binding antibodies, while IgA levels in the lavage fluid were similar between the two groups.

Antibody-secreting cell analysis by ELISpot revealed where these antibody responses were being generated. Seven days after the boost, Ad5-RSV-preF had induced more IgG-positive antibody-secreting cells in the spleen, whereas LNP-RSV-preF saRNA had induced more in the lungs; counts in the bone marrow did not differ significantly between the groups. Germinal center responses followed the same compartmental logic: in the draining lymph nodes, the adenoviral vaccine produced higher proportions of T follicular helper cells, plasma cells and germinal center B cells, while the RNA vaccine drove higher proportions of plasma cells and germinal center B cells in the spleen. The authors interpret this as the adenoviral platform favoring a classical, lymph node-centered priming pattern, with the RNA platform distributing B-cell activation differently across systemic and pulmonary compartments.

T-cell responses showed a parallel split. In IFN-γ ELISpot assays of splenocytes, the adenoviral vaccine generated nearly twice as many IFN-γ-secreting cells as the RNA vaccine, and intracellular cytokine staining showed that Ad5-RSV-preF drove greater increases in TNF-α-producing CD4 and CD8 T cells in the spleen and greater increases in IFN-γ- and TNF-α-producing T cells in the lungs, along with stronger IL-2-associated recall potential after antigen restimulation. In the lung, however, the RNA vaccine produced more pronounced increases in CD69-expressing CD4 and CD8 T cells and higher proportions of both CD4 and CD8 tissue-resident memory T cells. In the authors’ framing, the adenoviral vector builds a readily expandable systemic effector reservoir, while the self-amplifying RNA establishes an activated, tissue-resident defensive population at the pulmonary site where RSV first encounters the immune system.

Single-cell RNA sequencing of lung cells collected three days after the booster added a transcriptional dimension to this divide. The RNA vaccine reshaped the lung immune landscape toward a T cell-dominant state marked by proliferating, cycling T cells, and induced 1265 upregulated genes in pulmonary T cells compared with the unimmunized group, versus 486 for the adenoviral vaccine. Direct comparison showed the RNA platform associated with higher expression of genes linked to T cell differentiation and memory formation, including Tcf7 and Ccr7, while the adenoviral platform was associated with increased expression of cytotoxic and inflammatory genes including Gzmb, Gzma and Ifng, along with chemokine-related genes, and a more heterogeneous composition with increased NK cell representation. Broader flow cytometric profiling three days after the boost showed that both vaccines increased T cell proportions in spleen and lung, but with distinct biases: the adenoviral vaccine more prominently increased CD8 T cells in the spleen, while the RNA vaccine’s changes were more consistent with pulmonary activation and memory-associated redistribution.

Long-term follow-up at 120 days preserved this compartmentalized pattern. Both vaccines maintained substantial serum IgG, with the adenoviral vaccine sustaining higher levels, and serum IgA did not differ significantly between groups. Ad5-RSV-preF elicited higher frequencies of long-lived IgG-positive antibody-secreting cells in the spleen and also in the lung, and roughly twice as many IFN-γ-secreting splenocytes upon antigen restimulation, indicating stronger durable systemic T cell memory. The RNA vaccine, by contrast, induced a greater frequency of long-lived IgA-positive antibody-secreting cells in the spleen — although the two platforms generated comparable numbers of long-lived IgA-positive cells in the lungs, a result the authors highlight as showing that the RNA platform did not provide a measurable long-term pulmonary IgA memory advantage in this model.

The authors are explicit about the limits of these findings. No viral challenge experiments were performed, so the protective efficacy of either vaccine in vivo remains untested; the study used only female BALB/c mice and only the intramuscular route, leaving intranasal and other delivery routes unexplored, and the authors suggest that combined saRNA-adenoviral regimens merit future evaluation. What the study does establish, within its preclinical scope, is a framework: with antigen and route held constant, the choice between an adenoviral vector and a lipid nanoparticle-delivered self-amplifying RNA biases immunity toward systemic durability on one side or pulmonary-localized immune readiness on the other, a distinction the authors argue should inform how next-generation RSV vaccines are compared and selected.

Subject of Research: How vaccine platform choice shapes systemic versus respiratory mucosal immunity to RSV in mice

Article Title: Intramuscular self-amplifying RNA and adenoviral RSV vaccines program distinct systemic and respiratory mucosal immune architectures

Article References: Zhang, Y., Qin, Y., Qi, Z., Yang, Q., Zhang, X., Hu, S., Li, Y., Xiao, B., Hu, J., & Cheng, P. (2026). Intramuscular self-amplifying RNA and adenoviral RSV vaccines program distinct systemic and respiratory mucosal immune architectures. EMBO Molecular Medicine. https://doi.org/10.1038/s44321-026-00537-y

Image Credits: AI Generated

DOI: 10.1038/s44321-026-00537-y

Keywords: RSV, vaccine platforms, self-amplifying RNA, adenoviral vector, mucosal immunity, prefusion F protein, tissue-resident memory T cells, antibody-secreting cells, germinal center, single-cell RNA sequencing, lipid nanoparticles, preclinical study