Biomimetic platforms for modeling in vitro the functional unit of the exocrine pancreas

The functional unit of exocrine pancreas is responsible for the secretion of digestive enzymes and it is the region where the first lesions of the most lethal pancreatic cancers (i.e., pancreatic ductal adenocarcinoma, PDAC) develop. PDAC is a type of exocrine pancreas tumor which currently represents one of the leading causes of cancer-related death worldwide (5-year survival rate below 9%). Current research is focusing on better understanding of this pathology through creating effective in vitro models that can be used to recapitulate the key mechanisms involved in pancreatic cancer. However, although recent studies have shown the possibility of modeling the PDAC microenvironment in vitro, the microanatomy (in terms of 3D architecture and cellular composition) of the human pancreatic gland remains extremely challenging to be reproduced and monitored in functionally reliable models. The main purpose of my PhD research is to reproduce the functional unit of the exocrine pancreas, constituted by epithelial and stromal cells. Specifically, this work aims at developing human in vitro models that allow to analyze the PDAC-stroma interplay and the mechanisms implicated during the initial stage of PDAC progression. To achieve this goal, different biofabrication strategies were explored to obtain models which can be classified as two-dimensional (2D), two-and-a-half-dimensional (2.5D) and three-dimensional (3D). Each one, although being a simplified model showing both advantages and limitations, represented an important step in the process toward the development of a valuable and effective in vitro platform for the study of pancreatic cancer. Moreover, all the here designed and fabricated models constitute, for different reasons, innovative engineering strategies that go beyond the state-of-the-art in cancer research. The 2D model, composed by transwell inserts including a polycaprolactone/gelatin (PCL/Gel) electrospun membrane, allowed to preliminary study the reciprocal influence of different cell types (i.e., stromal and epithelial cells) and it was able to reassemble the highest cytokines release and changes in cell morphology by fibroblasts co-cultured with epithelial cells overexpressing the KRAS oncogene which are also reported in vivo. The information acquired using this simplified model were then transferred to a more complex 2.5D model, represented by a multilayer PDAC-on-chip system. This microfluidic device was designed to incorporate PDAC cells and a stromal cell-laden type I collagen hydrogel in the top and bottom layers, respectively. The use of a nanofibrous and biomimetic electrospun membrane, the same that has been integrated in the 2D model, allowed to compartmentalize the microfluidic device and thus separate the cancer component from the stromal tissue. In this way, the effect of the inflammation stimuli on stromal cells was studied in a controlled and specific way. This 2.5D model permitted to perform tests (e.g., evaluation of cell resistance to chemotherapy) and analyses in a fast, medium-throughput and accessible manner. Finally, the 3D models were obtained by two different approaches, i.e. layer-by-layer approaches (FDM and MEW models) and tomographic volumetric bioprinting (VBP model). The layer-by-layer techniques used in this thesis project allowed to obtain macro- and microscale models replicating the half-structure of the complex gland morphology. Specifically, the FDM model was used to preliminary assess the feasibility of reproducing the glandular structure by using a layer-by-layer approach and to monitor the fibroblasts viability on PCL printed structures over several weeks. Nevertheless, melt electrowriting (MEW) permitted to achieve better resolutions of the printed structures, that have dimensions about four times smaller than those of FDM constructs. These studies led us to proceed with the implementation of co-culture conditions only in MEW scaffolds. The biomimicry of this model was demonstrated in terms of (i) capability to recreate the compartmentalization of stroma and epithelium found in PDAC microenvironment and (ii) ability to mirror the fibroblasts inflammation process occurring during pathology development. Finally, the innovative technique of volumetric bioprinting was here adopted to develop a 3D in vitro model at the microscale, resembling the physiological “closed” structure typical of the pancreatic gland. In particular, a gelatin metacrylate (GelMA) hydrogel was ad hoc prepared and loaded with human fibroblasts to mimic the stromal compartment. Healthy or KRAS-mutated human pancreatic ductal epithelial cells were then introduced inside the construct’s cavity to reproduce the exocrine tissue that evolves to neoplastic lesions during pancreatic carcinogenesis. The ability of VBP model in recapitulating the tumor-stroma interplay occurring in pancreatic cancer while also accurately reproducing the microanatomy of the exocrine gland was proved. In conclusion, the in vitro models developed in this PhD work represent attractive and powerful tools for the establishment of new diagnostic approaches and for the screening and testing of drugs. Therefore, they can be fundamental to improve the knowledge of the complex mechanisms implicated in PDAC and find innovative therapeutical strategies to fight pancreatic cancer.