The transcriptional dynamics occurring at the self-organizing glandular epithelial during branching morphogenesis are pivotal for deciphering various processes ranging from normal tissue growth to cancer formation.
For the past decade matrigel© has been widely used as extracellular matrix for the propagation of organoids, here we report that only in the presence of a higher than 70% collagen matrix symmetry breaking can be observed. Comparing the transcriptomic profile of collagen and matrigel grown-organoids we identified that matrigel organoids are more basal-like, have a higher EMT signature (GSEA) compared to the more classical collagen grown organoids.
In this study we used single primary murine pancreatic ductal adenocarcinoma cells (PDAC) cells embedded in floating collagen gels and adapted the media supplementation, thus allowing organoids to self-organize into highly complex branched structures, replicating the in vivo tumour architecture. During organoid development we identified four distinct morphogenetic phases, each characterized by a unique pattern of cell proliferation, invasion, matrix deformation and protein expression. In our search for novel molecular drivers of branching morphogenesis we compared the transcriptome of two distinct phases, an early development (day 7) vs a late mature time point (day 13). We compared these two distinct developmental time points in using PCA, GSEA and respective heat maps of major processes illustrating the overexpression of genes associated with: proliferation, ECM interaction, signalling by Rho GTPases and EMT in day-7 organoids and genes associated with ion channel transport at day-13 organoids.
An important aspect raised during our investigation was the effect of different genotypes (Ptf1aCre/+; KrasG12D/+, Pdx1Cre/+; KrasG12D/+; TP53fl/fl) on the branch formation of organoids. To this end we generated organoids derived from KC and KPC tumours, which interestingly gave rise to strikingly similar branching organoids. Although the morphology was not significantly altered the transcriptional profiling showed higher proliferation (Myc targets, E2F targets), EMT score for the KPC organoids.
We propose that the spatiotemporal synchronized processes of cell proliferation, matrix remodeling, contraction and ion channel flux are key-events of the different morphogenic phases and lead to the formation of these complex structures. Importantly, these dynamic processes are accompanied by strong transcriptional profile changes, with the organoids undergoing de-differentiation during branch elongation and later on activating an epithelial gene expression program upon maturation. Taken together, these results illustrate the ability of tumour cells to self-organize in multicellular complex structures and provide with a novel system to study branching morphogenesis and tumour biology in vitro.
In the updated submission we include new data sets from our recent publication. More specifically we focus on how we can capture inter- and intratumoral heterogeneity using our previously published organoid system (Randriamanantsoa, Papargyriou et al. 2022) of pancreatic cancer and how these morphologically diverse organoids respond to targeted therapy. We aim to understand the following four biological questions: First, we aim to address intertumoral heterogeneity by culturing classical and basal-like PDAC cells in 2D culture conditions and 3D floating collagen gels. Indeed, in both cases we observe a clear separation between mesenchymal and epithelial organoids and describe pivotal pathways (f.e. TGF pathway) that are differentially regulated between the two culture conditions (2D vs 3D). In detail, we culture 3 epithelial (ID: 8442, 9591, 53631) and 3 mesenchymal (ID: 8028, 9091, 16992) murine pancreatic cancer lines in established 2D conditions and in 3D floating collagen gels. Next, we focus on understanding the transcriptional background of distinct organoid morphologies derived from the same parental cell lines of either epithelial or mesenchymal origin. In more detail we transcriptionally analyze specific epithelial (line ID 9591) as a bulk population and their subclones: TEBBO, cystic branched, thick-branched and tree-like and mesenchymal pancreatic cancer line (line ID 16992) as a bulk population and their subclones: branched mesenchymal, firework and star-like. After performing a 102-drug screening on these clones we identified potent target therapy drugs acting in specific phenotypes. To validate the effects on the transcriptional level we treated all the epithelial organoid phenotypes with the combination of AZD5153 (BET/BRD4 inhibitor)+Poziotinib (pan-HER inhibitor) and the mesenchymal phenotypes with monotreatments of the following drugs: Birinapant (SMAC antagonist) Poziotinib (pan-HER inhibitor), Saracatinib (Src inhibitor) and RO5126766 (RAF/MEK inhibitor). Furthermore, to translate our findings in a human setting we developed a new media composition which enables the formation of branching organoids in pancreatic cancer patient-derived organoids (PDOs). We examined the effect of our newly developed media, the Basal Branching PDO Media on the transcriptome of 3 PDOs (PDO line ID: B211, B250, B320) compared to previously established media conditions such as the Full PDO Media.
Overall design: Study 1. The batch 332AP contains: bulk 2D cell lines vs. 3D organoids grown in collagen (for the line 9591 use the previously submitted samples: S40-S45 & S49-S56). Study 2. The batch 432AP: Understanding intra-cell line heterogeneity in 3D Organoids (note that for the 9591 bulk population samples were included in the first submission. Study 3. The batch 448AP: PDO lines comparison grown in different Media conditions (Full PDO Media vs Basal Branching PDO Media). Study 4. The batch 536APm: Mesenchymal organoids comparison control vs. targeted therapy treated. Study 5. The batch 536APe: Epithelial organoids comparison control vs. combinatory targeted therapy treated organoids.
Less...