FUSOBACTERIUM NUCLEATUM FACILITATES ANTI-PD-1 THERAPY IN MICROSATELLITE STABLE COLORECTAL CANCER

Background: Around 80-85% of colorectal cancer (CRC) belongs to ‘immune cold’ microsatellite stable (MSS) subtype without responsive to anti-PD-1 mAb therapy. Here, we aim to decipher the function of Fusobacterium nucleatum (Fn), a CRC-promoting gut pathogen, in modulating anti-PD-1 response in MSS CRC and its underlying molecular mechanism.
Methods: Correlation between PD1 expression and Fn abundance was determined in patients with MSS CRC (N=80). To mimic human immune responses, we utilized CD34⁺ humanized NOG mice implanted with orthotopic HT29 (MSS) tumors. Germ-free CD34⁺ humanized mice were constructed to determine if Fn mono-colonization is sufficient to provoke systemic CD8⁺ T cell responses. To explore the role of butyrate identified as a functional metabolite produced by Fn, we generated a Fn strain with mutant enoyl-CoA hydratase (FN0271) that is deficient in butyrate biosynthesis. To assess if Fn can boost anti-PD-1 therapy efficacy in human immune system, we generated autologous co-culture of human primary CRC organoids with PBMCs-derived CD8⁺ T cells from the same patient.
Results: Fn abundance negatively correlated with PD1 expression in MSS CRC patients (N=80, R=-0.302, P=0.0065), implying that it may alleviate CD8⁺ T cell exhaustion. In CD34⁺ humanized mice bearing orthotopic HT29 tumors, Fnpotentiated the antitumor effect of anti-PD-1 by repressing PD-1 expression in CD8⁺ TILs, leading to elevated IFN-γ⁺, TNF-α⁺ and GZMB⁺ CD8⁺ TILs. The potentiation of anti-PD1 efficacy by Fn was validated in germ-free CD34⁺humanized mice bearing HT29 tumors, CT26 syngeneic tumors in BALB/c mice, and autologous CD8⁺TILs-primary CRC organoids. We revealed that Fn conditioned medium (CM) recapitulated the effect of Fn in vitro and in vivo, suggesting the involvement of Fn metabolites. Metabolomic analysis of Fn CM identified butyrate as a top enriched metabolite. Concordantly, Fn gavage increased tumoral butyrate in mice in connection with decreased PD-1 expression in CD8⁺ TILs. Moreover, abolishing butyrate production in mutant FN0271 Fn attenuated its effect on repressing PD-1⁺CD8⁺ TILs and potentiation of anti-PD-1 efficacy, inferring butyrate as a critical metabolite in Fn-driven activation of CD8⁺TILs. Mechanistically, butyrate inhibits HDAC3/8 to induce Tbx21 promoter H3K27 acetylation, leading to increased expression of TBX21, a transcriptional repressor of PD-1. Knockout of Tbx21 in CD8⁺ TILs validated its role in inhibiting PD-1 expression and reactivating CD8⁺TILs by Fn or butyrate. Together, these results indicate a Fn-metabolite-immune crosstalk underlies Fn-driven anti-PD-1 efficacy in MSS CRC.
Conclusion: Fn boosts anti-PD-1 efficacy in MSS CRC by repressing PD-1 in CD8⁺ TILs via a butyrate-HDAC3/8-Tbx21 axis, leading to the activation antitumor immunity. Fn is a potential predictive marker of anti-PD-1 response in MSS CRC patients.