Therapeutic strategies to target cancer-associated fibroblasts (CAFs) in pancreatic ductal adenocarcinoma

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is characterized by abundant desmoplastic stroma that is primarily composed of cancer-associated fibroblasts (CAFs). It is widely accepted that CAFs stimulate tumor progression and might be also involved in drug resistance and immunosuppression. Previous results in Barbacid’s lab from bulk RNAseq identified targets that are differentially expressed between PDGFRα+ CAFs and PDGFRα+ normal pancreatic fibroblasts (NPFs). Therefore, I focused my Thesis on the in vivo functional validation of some of these genes: Lumican (Lum), Haptoglobin (Hp), Mesothelin (Msln) and Hyaluronic acid synthase 1 (Has1). My thesis also included the characterization of the CAF heterogeneity of mouse PDACs by single-cell RNA sequencing (scRNA-seq), with the aim of learning about the function of the different CAF subpopulations. Our study could identify 16 clusters but it was not possible to isolate them by the expression of specific genes to perform functional studies. To study the role of Lum, Msln, Hp and Has genes, we have generated knock-out (KO) alleles by CRISPR/Cas9 gene-editing technology. Importantly, with the idea of reprograming the CAFs to a less protumorigenic phenotype, we have studied the effect of the absence of these genes alone (single KOs) or in combination (triple KOs). Interestingly, single KO of Lum, Hp and Msln or TKO (Lum/Hp/Msln) CAFs, all can inhibit tumor growth in vitro or in orthotopic studies when tumor cells are competent for these genes. Unfortunately, in genetically engineered mouse models (single KO or TKO), germline elimination of these genes did not improve survival, and mice developed tumors with the same latency. However, there is stroma remodeling in the PDACs of these KO models, mainly characterized by less collagen content and increased macrophages infiltration. Although this stroma remodeling did not help to improve the efficiency of gemcitabine treatment in allograft studies, we cannot rule out that elimination of these genes could represent any vulnerability that can be used in other therapeutic strategies. In addition, Has proteins are responsible for the production of hyaluronic acid (HA) and it has been shown in preclinical trials that HA depletion improves drug delivery. Elimination of Has1 did not prevent PDAC development, did not improve survival and did not reduce the HA content of PDACs. To achieve this reduction it was necessary to develop a TKO model of Has1/2/3. PDAC of TKO mice exhibited a dramatic increase in macrophage infiltration, as well as infiltration of CD8+ cytotoxic T cells. Furthermore, in allograft studies, the TKO (Has1/2/3) mice environment significantly inhibited tumor growth of tumor cells competent for these genes. All these findings may help to design future therapeutic strategies