Disrupting MDSC-fibroblast interactions to overcome chemoimmunotherapy resistance in pancreatic cancer

PROJECT SUMMARY Pancreatic ductal adenocarcinoma (PDAC) is a lethal malignancy projected to be the 2nd leading cause of cancer-related deaths in the United States by 2030, with a dismal 5-year survival of 12%. PDAC is refractory to standard chemotherapy and immunotherapy due to intrinsic and acquired resistance. Two key culprits

associated with therapeutic resistance are frequent infiltration of tumors by immunosuppressive innate immune cells, particularly neutrophilic/granulocytic myeloid-derived suppressor cells (gMDSC), as well as pro- inflammatory signaling by cancer-associated fibroblasts (CAF), which dominate the non-tumor stroma and act

as cellular antennae to transmit inflammatory cues that further beckon gMDSCs to the tumor microenvironment (TME). In dissecting the unifying mechanisms that underpin the relationship between these key constituents in the TME and chemoimmunotherapy resistance, our data uncover novel signaling interactions between gMDSC-

derived transmembrane TNF (tmTNF) and TNFR2-expressing inflammatory CAFs in driving tumor-permissive effects. Moreover, overexpression of TNF-TNFR2 signaling in human PDAC tumor transcriptomes is associated with chemoimmunotherapy resistance across multiple datasets spanning diverse clinical settings.

Building on these observations, the overall mission of this proposal is: (1) to mechanistically dissect the pathogenic signaling and spatial interactions between gMDSC-tmTNF and CAF-TNFR2 governing chemoimmunotherapy resistance in PDAC; and (2) catalyze preclinical development of a highly selective TNFR2

antagonistic antibody TY-101—which traps TNFR2 monomers and prevents assembly of signaling trimers—to disrupt gMDSC-CAF crosstalk and overcome chemoimmunotherapy resistance in preclinical models. Aim 1 will define the precise contributions of gMDSC-tmTNF and CAF-TNFR2 signaling communication and spatial interactions to chemoimmunotherapy resistance using innovative preclinical mouse knock-in/knock-

out modeling (murine) and single-cell spatial methodologies (human). Aim 2 will elucidate the molecular mechanism by which gMDSC-CAF crosstalk regulates gMDSC-TNF production via cooperative MAP kinase and STAT3 signaling. Moreover, we will also determine the functional significance of this cooperative gMDSC-TNF

mechanism on CAF polarization: a) in vivo using a tandem gMDSC depletion-adoptive transfer model; and b) in human patients’ peripheral blood gMDSCs from subjects enrolled in a phase 1 clinical trial investigating MEK and STAT3 inhibition in advanced PDAC. Finally, Aim 3 will determine if selective targeting of TNFR2 using TY-

101 improves sensitivity to chemoimmunotherapy by remodeling gMDSC-CAF spatial habitats and CAF plasticity to prolong survival in vivo. Together, successful completion of these aims will uncover novel insights into the tumor-permissive and tolerogenic role of myeloid-CAF crosstalk in the TME, as well as generate compelling

preclinical rationale and a compendium of novel predictive biomarkers for an upcoming clinical trial combining neoadjuvant TNFR2 inhibition with chemoimmunotherapy in patients with operable PDAC at our institution.