CD137L is a co-stimulator borrowed from cancer therapy
CD137 is a costimulatory molecule expressed on activated T cells and NK cells. When those cells meet cognate CD137 ligand (CD137L, also called 4-1BBL or TNFSF9) on antigen-presenting cells, they receive a co-stimulatory signal. Signalling through CD137 is potent enough that agonistic anti-CD137 antibodies were developed for tumour immunotherapy, and the authors reasoned that supplying CD137L alongside a vaccine antigen might strengthen T-cell immunity.
The study tested that idea in two vaccine formats. The first was an mRNA-LNP construct expressing influenza A M2 and NP antigens, with or without murine CD137L co-expressed on the same RNA. The second was a recombinant SARS-CoV-2 RBD protein co-injected with the soluble extracellular domain of mouse CD137L and formulated in Addavax, a research mimetic of the squalene emulsion adjuvant MF59. The protein format was added because it fixes antigen dose and localization, which the mRNA construct cannot guarantee.
The mRNA construct raised CD4 and CD8 T-cell responses
Adult female BALB/c mice received two intramuscular doses three weeks apart, with equimolar amounts of M2-NP mRNA-LNP (5 μg) or CD137L-M2-NP mRNA-LNP (7.5 μg). Two weeks after the boost, splenocytes were restimulated with the M2e peptide for CD4 T cells or the immunodominant NP peptide for CD8 T cells, then stained for intracellular TNFα, IFNγ and IL-2.
The CD137L group showed increased CD4 and CD8 T-cell responses at that time point. The effect on CD8 cells persisted three months after the boost, and the authors report increased percentages of CD8 T cells positive for IFNγ and for CD107a, a degranulation marker that correlates with cytotoxic activity but is not itself a direct cytotoxicity assay. This is the central positive result: co-expressing CD137L on the mRNA construct improved the magnitude of antigen-specific T-cell responses, including a durable functional CD8 population.
Antibody loss was antigen-dependent, not uniform
The humoral side moved the other way, but not for every antigen. Mice given CD137L-M2-NP mRNA had significantly lower anti-M2e IgG titres after the boost than mice given M2-NP mRNA, alongside a reduced percentage of M2e-specific germinal centre B cells in the spleen. Anti-NP IgG titres, by contrast, were comparable between the two groups.
A longitudinal follow-up to three months post-boost showed that CD137L changed the magnitude of the M2-specific response but not the NP-specific one, and did not change the durability of either antibody response. The authors note that both M2e- and NP-specific responses are expected to be T-cell dependent, so the split cannot be explained simply by T-cell dependence. They suggest the intracellular localization of NP may mean NP-specific B cells encounter antigen after NP-producing cells die or release it, possibly minimizing interference with surface-expressed CD137L. That explanation is offered as a hypothesis, not a demonstrated mechanism.
The protein vaccine gained no T-cell benefit
In the protein format, adult female or male C57BL/6J mice were vaccinated intramuscularly in a prime-boost regimen three weeks apart with RBD (20 μg) alone or with soluble CD137L (20 μg), both in Addavax. RBD-specific T-cell responses were read after restimulation with a commercial RBD peptide pool.
The differences were minor. There was a slight increase in CD4 T cells secreting IL-5 in the CD137L-treated group, with no change in other CD4 groups; follicular helper T-cell numbers and tissue-resident-phenotype CD4 T cells were comparable between groups. Among CD8 T cells, cytokine-secreting subsets showed no significant differences, while Tc17 cells rose slightly. The authors attribute the absence of a clear T-cell effect to Addavax itself, which is a strong inducer of cytotoxic, helper and follicular helper T-cell formation, and suggest any CD137L effect could be masked by it. They also note that in this format CD137L was supplied as a separate soluble protein, whereas CD137-CD137L co-stimulation has been reported to be more potent in cis than in trans.
Antibody suppression appeared in both formats
The protein format reproduced the antibody effect. Anti-RBD total IgG titres and neutralizing antibody responses, measured with a surrogate neutralization assay on 1:10 diluted individual sera, were lower in the CD137L group. Unlike the mRNA experiment, however, germinal centre, memory and plasma B-cell numbers, both total and antigen-specific, were comparable between the two immunized groups, although plasmablasts showed a lower trend.
That contrast matters for interpretation. In the mRNA format the antibody drop came with fewer M2e-specific germinal centre B cells; in the protein format it did not come with a measurable germinal centre B-cell loss. Because the protein format controls antigen dose and localization, the authors argue the suppressive effect on antibodies cannot be explained by antigen dose and appears independent of vaccine format. They propose that vaccine-included CD137L might compete with endogenous CD137L on antigen-specific B cells, limiting B–T interactions through the CD137-CD137L axis, but they also note that murine B cells do not express CD137, ruling out direct interaction with B cells in these mice. The mechanism therefore remains unresolved.
A T-cell-first strategy with an antibody cost
The authors conclude that incorporating CD137L within an mRNA-LNP platform may be a way to tune vaccine immunity toward strong, sustained T-cell responses while limiting antibody production against a defined T-cell-dependent antigen. They frame the potential advantage for pathogens where T cells are established immune correlates of protection, or where antibody responses are undesirable, citing antibody-dependent enhancement as seen in Dengue vaccination and early anti-RSV vaccine designs.
That framing is a hypothesis about future use, not a demonstrated outcome. The study did not run challenge experiments in a murine influenza model, so protective efficacy is untested, and the authors call for additional disease models that rely predominantly on T-cell-mediated immunity. A second stated limitation is that M2 or NP expression may differ in vivo or be affected by CD137L co-expression on the same RNA; delivering CD137L from a separate mRNA might behave differently.
The practical reading is a trade-off rather than a free upgrade. CD137L co-expression on the mRNA construct bought larger and more durable CD8 T-cell responses, but the same construct significantly reduced anti-M2e IgG and germinal centre B cells while leaving anti-NP antibodies intact. The protein experiment shows the antibody reduction can occur without any T-cell gain, which means the two effects are not a single coupled switch. For a T-cell-focused mRNA vaccine against an antigen like M2e, the relevant question is whether the T-cell benefit outweighs the antibody loss; the paper establishes that both happen but does not establish which matters more for protection.
Sources used in this answer
CD137L adjuvanted vaccine increases antigen-specific T-cell responses but impacts antigen-specific humoral responses
In mice, CD137L co-expressed on an mRNA-LNP vaccine boosted antigen-specific CD4 and CD8 T-cell responses, including durable CD8 responses at three months, but significantly lowered anti-M2e IgG and germinal centre B cells while leaving anti-NP IgG unchanged; soluble CD137L with an Addavax-adjuvanted RBD protein vaccine gave no clear T-cell benefit and also lowered anti-RBD IgG and neutralizing antibodies.