Umeå University's logo

In Sweden 1200-1300 people are diagnosed with pancreatic ductal adenocarcinoma (PDAC) every year. Late diagnosis, together with poor treatment response and resistance to checkpoint inhibitor immunotherapy, contributes to the poor prognosis of the disease. 

PDAC is characterized by abundant deposits of extracellular matrix, which mainly includes structural proteins including collagens, proteoglycans, cell-binding glycoproteins, carbohydrates, and secreted proteins, all constituting the matrisome of the tumor. The matrisome protects cancer cells and affects the outcome. Several highly expressed matrisome proteins are involved in oncogenesis, including the processes of immunosuppression and metastasis formation, therefore contributing to the poor prognosis. In this thesis the pathophysiological role of several matrisome proteins in PDAC tumor progression was studied. Unbiased analysis of matrisome proteins in PDAC tumors revealed increased levels of cancer cell-derived secreted proteins compared to normal healthy control tissue. Subsequently, differentially expressed candidate proteins, with known cellular functions in other disease but hitherto uncharacterized role in PDAC progression, were selected.

Serine protease inhibitor clade B member 5 (SERPINB5), agrin, and cystatin B (CSTB), were selected for the study described in paper I based on their known roles in the metastasis formation process in other types of cancers. SERPINB5, agrin and CSTB were found to increase metastasis in models of PDAC by affecting epithelial to mesenchymal transition, ECM degradation and extravasation. In PDAC tumors, high levels of extracellular SERPINB5 correlated to reduced overall survival.  

Galectin 4 (Gal 4) was selected for the study described in paper II based on its known immunosuppressive effects. Gal 4 is highly expressed in PDAC and was found to inhibit T cell infiltration and induce apoptosis in CD8+ T cells by binding to CD3 on the surface of T cells. Gal 4 was associated with better survival in PDAC patients and correlated to higher activation and cytolytic effect of CD8+ T cells.

The relation between gal 4 and other immunosuppressive proteins was studied in paper III. Analysis of available datasets revealed that gal 4 expression correlates with other cancer cell-derived immunosuppressive proteins of the galectin family, galectin 3 and galectin 9, while negatively correlating with the stroma-derived factors galectin-1 and TGFBI.

Findings in this thesis show that targeting of matrisome proteins in PDAC can be a promising therapy strategy. Blocking extracellular SERPINB5 could result in reduced metastasis and increased survival. Blocking intracellular gal 4 could increase anti-tumor immunity and synergize with checkpoint inhibition therapy.

The identified co-expression and coregulation of different immunosuppressive proteins indicate that different tumors can be classified based on their predominant immunosuppressive mechanisms. Following this classification in individual patients, combinations of therapies against different immunosuppressive mechanisms could represent a promising strategy to introduce effective immunotherapies for PDAC patients.  

Pancreatic ductal adenocarcinoma (PDAC) is a highly lethal malignancy characterized by an extensive desmoplastic stroma. The stroma is the site of intricate communication between malignant cells and their surrounding environment. This tissue microenvironment (TME) is populated by a heterogenous mixture of cell types and extracellular matrix proteins. Distinct stromal elements confer tumour-restraining or tumour-promoting influences on tumorigenesis. Characterizing stromal composition therefore represents an opportunity to identify candidates for therapeutic intervention to facilitate improved clinical outcomes. In this thesis we identify galectin-4 as an extracellular matrix protein which is upregulated in PDAC. We find that galectin-4 exerts a pro-tumorigenic influence in PDAC through promoting immune suppression, highlighting its potential as a novel therapeutic target. We subsequently provide a comprehensive characterization of mesenchymal cell diversity in PDAC including cancer-associated fibroblasts (CAFs) which represent one of the dominant stromal cellular components. We identify inflammatory CAF (iCAF) and myofibroblastic CAF (myCAF) subtypes in addition to defining a novel interferon-response CAF (ifCAF) subtype. In addition, we demonstrate that pancreatic stellate cells (PSCs) are capable of forming iCAFs, myCAFs and ifCAFs in response to tumour-derived signals using an organoid-based co-culture model and define biological pathways regulating CAF subtype formation. We then perform a high-throughput drug-screen using this co-culture model to identify compounds which can suppress tumour growth indirectly through modifying CAFs. One such compound is GNF-5 which we show can suppress cancer cell proliferation indirectly through manipulating CAF phenotype. Taken together, this thesis augments our understanding of the composition of the PDAC stroma and identifies potential therapeutic targets as well as developing an approach to discover drugs which yield a therapeutic benefit through targeting the PDAC stroma.