Fig. 4: BNT162b2-based cancer therapy depends on T cell responses but not B cell responses, and induces potent antigen spreading by tumor cell killing and the release of tumor-derived exosomes.

a Experimental design for investigating the role of B and T cells in BNT162b2-based cancer therapy in μMT or Rag1−/− mice (n = 5 per group). b Tumor growth curves of the intratumoral BNT162b2 treatment group vs the intratumoral PBS control group for the experiment described in a. c Photo of tumors at the endpoint in the treatment and control groups from the experiment described in a. d Tumor growth curves of the intratumoral BNT162b2 treatment group vs the intratumoral PBS control group in μMT mice. e Tumor growth curves of the intratumoral BNT162b2 treatment group vs the PBS control group in Rag1−/− mice. f Expression of SARS-CoV-2 spike protein on the surface of tumor cells. g Expression of SARS-CoV-2 spike protein inside tumor cells. h Representative IFN-γ ELISpot assays show tumor antigen-specific T cell responses post intratumoral BNT162b2 therapy (n = 4 per group). i Titers of anti-spike protein and anti-B16F10 surface protein IgG antibody in sera post intratumoral BNT162b2 injections or PBS injections (n = 5 per group). In h and i, X i.m.-PBS i.t.: sera or splenocytes from mice without any intramuscular vaccination but with intratumoral PBS injections. BNT i.m.-BNT i.t.: sera or splenocytes from mice with intramuscular and intratumoral BNT162b2 injections. BNT i.m.-PBS i.t.: sera or splenocytes from mice with intramuscular BNT162b2 injections and intratumoral PBS injections. j Experimental design of the detection of intracellular CD63+ exosomes in the B16F10 cell line post BNT162b2 treatments or DMSO treatments (n = 5 per group). k Experimental design for investigating the characteristics of exosomes derived from BNT162b2 treated B16F10-OVA cells. l Transmission electron microscopy (TEM) images of isolated B16F10-OVA-derived exosomes. m Identification of B16F10-OVA-derived exosomes by western blotting assay, and detecting the expression of spike protein, OVA protein, and PD-L1 in B16F10-OVA-derived exosomes. n Representative IFN-γ ELISpot assays show tumor antigen-specific T cell responses induced by BNT162b2 transfected B16F10-OVA-derived exosomes (n = 8 per group, full data are shown in Supplementary Fig. S6e). o Immunofluorescent staining of CD63 (marker of exosomes) on tumor sections from mice described in Fig. 1a. p Comparing the mean fluorescence intensity (MFI) of intracellular CD63+ exosomes in B16F10 cells treated with BNT162b2 to that in B16F10 cells treated with DMSO, as determined by flow cytometry. The experimental procedure is described in j. BNT i.m.-PBS i.t. (control group): the mice with BNT162b2 intramuscular injections and PBS intratumoral injections. BNT i.m.-BNT i.t. (treatment group): the mice with BNT162b2 intramuscular injections and BNT162b2 intratumoral injections. Scale bars, 20 μm in f and g, 1 μm in l, and 10 μm in o. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. ns, not statistically significant.