Dissecting the molecular mechanisms underlying the metastasis of the colorectal cancer cell line CMT93

Colorectal cancer (CRC) is the third most commonly diagnosed cancer entity worldwide. The majority of cancer-related death occurs due to the malignant spread to distant organs and represents a major challenge in the clinics. The colonization of distant organs is the rate-limiting step in the metastatic cascade, and requires tumor cell adaptation to the new environment. Previous results from our group had identified two variants of the murine colorectal cancer cell line CMT93, which significantly differed in their ability for metastatic colonization upon injection into the brain of syngeneic mice. In this study, we now aimed to elucidate the mechanisms responsible for this difference. Besides the adaption to the new environment, tumor cells facilitate metastatic growth by influencing the resident cells of the metastatic tissue. In particular, tissue-resident macrophages have been described to play a critical role in creating a permissive environment. Focusing on CRC bone metastases, we were interested to investigate the role of bone-resident macrophages, termed osteoclasts. To this end, we aimed to establish an optimized in vitro differentiation model for human osteoclasts. Analysing the two CMT93 variants regarding their basic tumorigenic characteristics (proliferation, migration and invasion), did not reveal any differences. Instead, we identified major differences in the cargo and functionality of the secreted membrane-enclosed particles, referred to as extracellular vesicles (EVs). The large EVs secreted by the aggressive CMT93 cell variant (CMTHigh) were taken up more efficiently and induced an increase in tumor cell invasion in both CMT93 cell variants, which was measured via Boyden chamber assays. This induction in malignant behavior was not observed for EVs secreted from the less aggressive CMT93 variant (CMTLow). The pro-invasive effect was not observed after EV-depletion from the conditioned medium or with protein-denatured large EVs, suggesting that mainly the proteomic cargo of large EVs is responsible for the effect. Characterization of the large EVs from CMTHigh cells by mass spectrometry showed an enrichment of adhesion proteins, including α3 integrin and active β1 integrin (ITGB1). ITGB1 also seemed to have diagnostic relevance in CRC patients as flow cytometry analysis revealed that enhanced levels of ITGB1+ large EVs were found in peripheral blood of CRC patients compared to healthy and non-cancer controls. To study whether this EV-based crosstalk also plays a role in the tumor microenvironment of CRC-bone metastases, we established an optimized model for human osteoclast differentiation. Here we used blood-derived monocytes as precursors for osteoclastogenesis. We identified that in comparison to standard differentiation protocols, osteoclast differentiation and functionality was significantly increased when performing the initial macrophage differentiation in Teflon-coated bags. An intermediate cell number and high concentrations of macrophage-colony stimulating factor and receptor activator of NF-κB ligand resulted in mature and functional osteoclasts. Conclusively, our results point towards differentially loaded large EVs as one of the main factors dictating the difference in metastatic outgrowth between both CMT93 cell variants. The lEVs, which promote a pro-invasive phenotype, show an enhanced expression of adhesion- and migration related proteins. Furthermore, the establishment of a novel, optimized workflow for the generation of functionally highly active primary human osteoclasts enables future studies on the heterologous cell communication between osteoclasts and CRC cells.