Longevity & AgingResearch PaperOpen Access

Backbone-Modified SIINFEKL Peptides Activate T Cells in Dishes but Fail as Cancer Vaccines in Mice

Swapping single α-amino acids for β-amino acids in a model cancer antigen didn't protect it from breakdown or produce anti-tumor immunity in mice.

Sunday, November 15, 2026 0 views
Published in Cancer Biol Ther
Molecular model of a peptide chain with an extended β-amino acid backbone docked in an MHC groove, a T cell receptor nearby

Summary

Peptide cancer vaccines are quickly degraded in the body, so researchers have tried swapping single α-amino acids for β-amino acids to slow proteolysis. This study re-tested seven such analogues of SIINFEKL, the classic ovalbumin epitope used in mouse models. The analogues were no more stable in mouse serum than the native peptide. They activated OT-1 T cells in vitro, and the Phe5 analogue was nearly as potent as the native peptide. In vivo, however, no peptide, native or modified, generated detectable antigen-specific CD8 T cells after direct immunization. In tumor-bearing mice, only the native peptide combined with TLR3 and TLR9 agonists suppressed tumor growth and extended survival. The β-modified analogues did not. The authors conclude that this class of backbone-modified peptides is not promising for antitumor vaccines.

Detailed Summary

Peptide vaccines are an attractive way to train T cells against cancer-specific antigens. They are cheap to make and easy to tailor to a patient's mutations. Their weakness is that short peptides are rapidly chewed up by proteases at the injection site and are often poorly immunogenic without adjuvants. One proposed fix is replacing a single α-amino acid with its β-amino acid counterpart (same side chain, backbone extended by one CH2), which can block nearby proteolysis. An earlier group applied this to SIINFEKL, the dominant H-2Kb-restricted epitope of ovalbumin, but never tested whether the analogues could drive an antitumor effect.

The researchers re-examined seven SIINFEKL analogues, each with a single β3 substitution at positions 1 through 7. They measured H-2Kb binding, stability in mouse serum, and activation of OT-1 T cells (which carry a TCR specific for SIINFEKL) in vitro. They then immunized naïve C57BL/6 mice and tested selected analogues in mice bearing OVA-expressing E.G7-OVA tumors that had received adoptively transferred OT-1 cells.

Binding to H-2Kb depended on the substitution site. Changes at Asn4 and Lys7 had little effect (about 2-fold). Changes at Ser1, Ile3, Phe5 and Glu6 reduced affinity roughly 9- to 40-fold, and the Ile2 anchor substitution cut affinity by more than 1000-fold. None of the analogues improved serum stability, and half-lives were indistinguishable from the native peptide. All analogues activated OT-1 cells in vitro, with the Phe5 analogue the most potent. Its EC50 was within about 2-fold of the native peptide for most markers, though about 100-fold higher for PD-1. Others were generally more than 10-fold less potent. Short-pulse experiments suggested some analogues were less efficient at activating T cells without prolonged exposure.

In vivo, direct subcutaneous peptide immunization, native or modified, produced no detectable tetramer-positive CD8 T cells, whereas an OVA-encoding DNA plasmid did. In tumor-bearing mice, peptide immunization alone did not slow tumor growth or improve survival. Adding TLR3 and TLR9 agonists (poly I:C and CpG) made the native peptide effective, with significant tumor suppression and survival benefit. The Asn4, Phe5 and Glu6 analogues gave no benefit. The abstract also reports that T cells induced against the analogues did not cross-react with native SIINFEKL.

The findings show that α-to-β substitution can preserve some in vitro T cell activation but does not deliver the intended benefits of improved stability or functional mimicry of the native antigen. Caveats include the single model antigen, a limited set of substitutions (β3 only, one site at a time), and the adoptive-transfer tumor design. The full text provided ends partway through the TLR experiments, so some details are drawn from the abstract. These results are preclinical mouse data and do not directly predict human outcomes.

Key Findings

  • None of the seven single α-to-β SIINFEKL analogues resisted proteolysis in mouse serum better than the native peptide.
  • All analogues activated OT-1 CD8 T cells in vitro; the Phe5 analogue was nearly as potent as native SIINFEKL.
  • Direct peptide immunization, native or β-modified, produced no detectable antigen-specific CD8 T cells in naïve mice.
  • Native SIINFEKL plus TLR3/TLR9 agonists suppressed E.G7-OVA tumors and extended survival; β-modified analogues did not.
  • T cells induced by α/β analogues in vivo were not cross-reactive to native antigen, undermining backbone modification as a vaccine strategy.

Methodology

Preclinical study using mouse models. Seven SIINFEKL analogues with single β3-amino acid substitutions were tested for H-2Kb binding, serum stability by UPLC, and OT-1 splenocyte activation in vitro. Selected analogues were then tested in naïve C57BL/6 mice and in E.G7-OVA tumor-bearing mice given adoptive OT-1 transfer, with or without poly I:C and CpG, using groups of about 4-5 mice.

Study Limitations

The study used a single model antigen (SIINFEKL) in mice, only β3 single-site substitutions, and a small number of animals per group, so the findings may not generalize to other epitopes, human HLA, or other backbone modifications. Peptides alone failed even in the native form, which limited the comparisons that could be made in vivo. The provided full text was truncated, so some results come from the abstract.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: