In advanced prostate cancer (PCa), cancer cells preferentially metastasise to bone. At this stage, survival rates drop from 56% to 3%1 and most treatment options for patients remain palliative. The bone metastatic niche (BMN) is indeed a complex microenvironment with various cell populations that have different roles in cancer cell migration, survival, and therapy resistance2. While osteoblasts and osteoclasts fuel metastastic establishment and progression in the bone microenvironemnt, bone marrow adipocytes (BMAs) account for 70% of the marrow and have been identified as a potential contributor3. To date, there is however a lack of relevant models to study the complex relationship between PCa cells and adipocytes within the bone microenvironment, as most models either employ murine cell lines, high fat diet mice4, and white adipocytes instead of bone marrow adipocytes, impairing to truly dissociate the role of human BMAs in PCa progression in bone. Therefore, we have developed sophisticated bioengineered humanised models with a high human adipose content that better replicate the complexity of the BMN, in vitro and in vivo.
We developed a humanised fatty BMN by co-culturing primary human osteoblasts, human bone-marrow derived mesenchymal stem cells differentiated into adipocytes and androgen-sensitive human PCa cells (LNCaP and C4-2B), encapsulated into Gelatin Methacryloyl (GelMA) hydrogels5 for in vitro and in vivo analysis. Construct characterization was done by evaluating cell viability, proliferation and differentiation, spheroid formation and osteoblast mineralization. After four weeks of osteogenic differentiation and mineralization in vitro, the bioengineered mineralized tissues were subcutaneously implanted into male NSG mice. Six weeks later, adipose/cancer constructs were implanted in the same subcutaneous pocket to create a fatty BMN, as to assess the effects of human adipocytes on cancer progression. In vivo bone formation was followed by computed tomography (CT) and cancer progression by bioluminescence imaging. After five weeks of co-culture, the humanised niche was characterized ex vivo by microCT and histology.
Construct optimization allowed to culture each cell type independently or concomitantly in GelMA hydrogels for up to six weeks in vitro, resulting in appropriate adipogenic differentiation, cancer spheroids formation and osteoblast mineralization. In vivo data showed bone formation with high mineralization overtime and the morphology of human native bone, including a cortical shell and trabecular centre. In vitro and in vivo, cancer progression was observed for the most aggressive cancer cell line (C4-2B) and was enhanced by the presence of human adipocytes.
The bioengineered models presented here offer an advanced platform for bone metastatic cancer research, such as the study of the effects of conventional therapies as well as for the screening of novel therapies on the bone tumour microenvironment, which is the next step of this study.
1. Nogaard, M. et al., J. Urol. 184-1, 162-167 (2010)
2. Croucher, P. et al., Nat. Rev. Cancer. 16-6, 373-386 (2016)
3. Hardaway, A. et al., Cancer Metastasis Rev. 33-0, 527-543 (2014)
4. Herroon, M. et al., Oncotarget. 4-11, 2108-2123 (2013)
5. Ravichandran, A. et al., Mat. Sci. Eng. C. 128, 112313 (2021)