Project description:The aim of this study is to compare the transcriptomic profiles of various ex vivo models of pancreatic cancer. First, primary tumours and their corresponding xenografts in nude mice were analysed by RNA-seq. Monocultures of pancreatic cancer cells derived from the xenografts were then prepared as 2D monolayers, Matrigel-embedded organoids, spheres in suspension and 3D cultures in self-assembling peptide amphiphile (PA) hydrogels. Seven-day cultures were then analysed by RNA-seq. The results suggest that all ex vivo monocultures retain patient-specific transcriptional profiles, especially cancer stem cell signatures, while being deficient in the expression of stromal components such as the core matrisome. Correlations with the primary tumours were generally higher in PA hydrogels than organoids. Biased gene expression signatures were identified in certain models. This is the first study to explore the transcriptomic signatures of four different ex vivo models matched to their primary tumours of origin.
Project description:The ex vivo modelling of pancreatic ductal adenocarcinoma (PDAC) using patient-derived cells is a promising tool to predict treatment responses. Matrigel-based organoid and organotypic approaches are limited by their undefined molecular composition, hindering the recapitulation of the tumour’s characteristic desmoplasia, which is known to promote drug resistance. To overcome these limitations, we used self-assembling peptide amphiphiles (PAs) gelled in a minimal extracellular matrix to model the pancreatic tumour microenvironment and to establish 3D multicellular cultures of patient-derived PDAC cells, pancreatic stellate cells and macrophages. Matrisome analysis of 3D cultures demonstrated consistent ECM protein deposition, which was highly reminiscent of the corresponding primary PDAC tissues. The proteomic data obtained was also compared to the corresponding patient-derived xenografts (PDX) in nude mice and Matrigel-based cultures. Characterisation of the chemosensitivity of the cultures revealed realistic treatment responses by PDAC cells in PA hydrogels based on the responses of their corresponding PDX tumours. Histological, transcriptional and functional techniques confirmed these similarities, which were not observed in Matrigel-based cultures. These findings demonstrate the biomimetic nature of PA hydrogels, which enable cultured cells to recreate the PDAC matrisome ex vivo and to respond to chemotherapeutic agents in a predictive manner.
Project description:Prospective, open labelled, multicenter trial to evaluate the feasibility of ex vivo culture 3D (chemogram obtaining) on biopsies in order to estimate the predictive value of this technique for treatment response in patients treated by two different chemotherapies (FOLFOX or FOLFIRI) for colorectal cancer.
Project description:Ascites or solid tumour from patients with ovarian cancer was collected and grown in culture as ex vivo models. Each sample has a tumour component and some samples have matched stromal cells, which were separated into individual cultures. RNA-seq was performed on these models to establish gene expression profiles, which allow the assessment of the separation protocol and identification of genes that are differentially expressed. The histological subtype from which the models were collected includes majorly high-grade serous, but also low-grade serous, clear cell and mucinous ovarian cancer. The sample subtypes have been assessed using a machine-learning based transcriptional classifier. These samples are processed in the same manner as a previous study, “E-MTAB-7223 - RNA-seq of human ex vivo ovarian cancer models with matched stromal cells”
Project description:Ascites or solid tumour from patients with ovarian cancer was collected and grown in culture as ex vivo models. Each sample has a mixture of tumour and stromal cells which were separated into individual cultures. Therefore each patient has tumour and stromal cultures originating from the same tissue collection. RNA-seq was performed on these matched models to establish gene expression profiles which allow the assessment of the separation protocol and establish genes that are differentially expressed for use in future validation processes.