In this issue of the JCI, Bear et al. (11) add to the expanding library of promising TCRs targeting oncogenic KRAS mutations (12–15). The authors further characterize two TCRs from their past study (16) derived from healthy donors’ peripheral blood T cells. These TCRs targeted KRAS G12V in the context of HLA-A*03:01 or HLA-A*11:01, which are common HLA alleles found in individuals identified as White, Black, and Asian in the US. Bear and authors also reported an additional KRAS G12V-reactive, HLA-A*11:01–restricted TCR from one of the healthy donors. Notably, the authors explored the potential of isolating mutant KRAS–reactive (mKRAS-reactive) TCRs from patients who received a mKRAS vaccine in a phase 1 trial. The vaccine comprised an autologous mature DC vaccine given intravenously in the adjuvant setting in patients with PDAC. Nine patients received two doses of the vaccine (prime and boost), which consisted of DCs exposed to (aka pulsed with) various long and/or short mKRAS peptides. Some patients received a vaccine that contained other KRAS G12 mutations in addition to the KRAS G12 mutation expressed by the autologous tumor. The vaccine was safe, with no grade 3 adverse events observed, and five of nine patients were alive at the median follow-up of about 25 months. The vaccine elicited T cell responses against mKRAS in six of nine patients as determined by IFN-γ ELISPOT assay, and the inclusion of long mKRAS peptides appeared to enhance immunogenicity of the vaccine, likely through stimulation of CD4+ T cells.

In an exemplary case, a patient with the HLA-A*03:01 and HLA-A*11:01 genotype was vaccinated with mature DCs pulsed with a nonamer (nine amino acid) and decamer (10 amino acid) KRAS G12V peptide. Two weeks later, the authors detected an elevated frequency of T cells in the peripheral blood that specifically recognized mKRAS G12V and not wild-type KRAS in the context of HLA-A*11:01. A mKRAS-reactive TCR sequence was isolated from an oligoclonal population of T cells in this patient (designated A11Vc).

This TCR alongside three other KRAS G12V–reactive TCRs derived from peripheral blood T cells of two healthy donors underwent in vitro testing for specificity, cross-reactivity, functional avidity, function in CD4+ T cells (CD8 coreceptor dependency), and tumor cell-line recognition. Three of the four KRAS G12V-reactive TCRs were restricted by HLA-A*11:01 (A11Va, A11Vb, and A11Vc), while one TCR was restricted by HLA-A*03:01 (A3V). All four TCRs did not recognize wild-type KRAS, but interestingly, the three HLA-A*11:01–restricted TCRs recognized the KRAS G12C peptide, albeit at 10-to-100-fold lower potency compared with the KRAS G12V peptide. The authors did not identify any concerning cross-reactivity of the TCRs against a select panel of peptides derived from the human proteome that were structurally related to the G12V peptide.

Coculture experiments of the TCR-engineered T cells with peptide-pulsed antigen-presenting cells and various tumor cell lines as targets revealed that T cells with TCR A11Va had high functional avidity and were the most potent at killing tumor cell lines in vitro. In CD4+ T cells, the three HLA-A*11:01 restricted TCRs were partially functional while the A3V TCR was not functional, which has implications if the TCR is to be introduced into CD4+ T cells for therapy. Interestingly, the TCR A11Vc derived from the vaccinated patient recognized a nonamer KRAS peptide, while the other three TCRs recognized a decamer KRAS peptide. This finding is important because targeted mass spectrometry revealed that the decamer was more abundantly presented than the nonamer at the surface of the cancer cell lines tested, suggesting that TCRs targeting the decamer KRAS G12V peptide may be more effective than nonamer targeting TCRs due to a higher density of target antigen.