Poster T6 - Optimizing a 3D tissue engineered cancer model using different fractions of cancer-conditioned medium

Introduction: Both in vivo and in vitro cancer models are essential to assess the effectiveness of new potential cancer treatments. However, 95% of treatments validated on existing models fail during clinical trials due to lack of effectiveness or nonspecific adverse effects. The models currently in use do not reflect the complexity of the tumor microenvironment (TME) and/or are limited in their ability to recreate human physiological conditions, highlighting the need for a new generation of cancer models. Notably, the absence of in vitro models taking extracellular matric (ECM) stiffness into account greatly contributes to the high failure rate of translational cancer research. During cancer progression, the TME gradually stiffens resulting in poor clinical outcomes as stiffness regulates various biological processes such as cell proliferation, epithelial-to-mesenchymal transition, drug resistance and metastasis formation. The deposition and cross-linking of collagen fibers, mediated in part by cancer associated fibroblasts (CAFs), are the main causes of this TME stiffening. Thus, integrating CAFs into engineered models could provide the means to finely control the physical properties of such systems and cover the stiffness range observed in tumors. Our team developed a 3D tissue engineered bladder cancer model based on the use of conditioned media from aggressive bladder cancer cells to activate healthy fibroblasts into CAFs. Here, we investigated how different cancer-conditioned media fractions impact the induction of healthy fibroblasts into CAFs, and in turn, how this strategy can be used to optimize and control the properties of the resulting engineered tissues.

Materials and

Methods: Extracellular vesicles (EVs) and soluble proteins contained in conditioned media from aggressive T24 bladder cancer cells were separated based on their size by centrifugation and filtrations. Firstly, the content of each fraction was characterized using dynamic light scattering (DLS) to measure particle size distribution. Then, two different population of primary vesical fibroblasts (FHU5 and FHU6) were activated by the different fractions for 10 days to allow stable activation. Expression of different CAF markers such as αSMA, FAP and FSP1 were then measured using immunofluorescence and immunoblotting techniques. Furthermore, the contractile properties and matrix metalloproteinase activity of our induced CAFs were also measured using traction force microscopy (TFM) and fluorescent gelatin degradation assay respectively. Subsequently, the stroma portions of our bioengineered bladder cancer model were produced following the self-assembly method for 30 days using healthy vesical fibroblasts (FHU5) induced into CAFs by the different fractions. The biophysical properties such as tissue thickness and collagen deposition and organisation were measured on picrosirius red stained histology slides using brightfield and quantitative polarized light microscopy (QPOL). The stiffness was finally measured by atomic force microscopy.

Results, Conclusions, and Discussions: Considering the critical roles of EVs in CAFs activation, we first assessed if we could isolate different EV population using sequential filtration. Using DLS, we observed that we could separate EVs into microvesicles (over 0.1µm) and exosomes (under 0.1µm) fractions (Fig. 1A). The complete conditioned medium (CM) and the exosomes and soluble factors (EXs+SF) fractions successfully activated healthy vesical fibroblasts into CAFs as seen by the elevated expression of α-Smooth Muscle Actin (αSMA), whereas the microvesicles (MVs) fraction did not (Fig. 1B-C). In contrast to other fibroblast cell lines, microvesicles did not seem to induce CAF activation in bladder fibroblast when looking solely at αSMA expression. We, however, did not see any significant differences in the tissue thickness measured by confocal reflectance microscopy (Fig. 1D). It is now well accepted that CAFs are heterogeneous and can be classified into different sub-populations. Some sub-populations do not express high quantity of aSMA, but still have an important impact on the TME. Measuring other general markers of CAFs such as FAP and FSP1 and sub-population related markers like PDGFRα and RGS5 will allow us to better conclude on the fibroblast activation mediated by our different conditioned media fractions. We also need to further study the tissues to confirm CAF activation in 3D and measure the biomechanical properties. We are also looking to concentrate the different conditioned media fraction in order to amplify the CAF activation. Overall, our preliminary results show it is possible to modulate cancer associated fibroblasts (CAFs) activation by fractioning the extracellular vesicles contained in the cancer-conditioned media. The difference in activation can than modulate their capacity of producing and remodelling the ECM allowing for an engineered cancer model better resembling different clinical stages of cancer.

Acknowledgements (Optional):