Microfluidic 3D Cell Culture Systems to Model Breast Cancer and the Bone Microenvironment In Vitro

The presence of bone metastasis significantly reduces survival in breast cancer patients, largely due to the absence of targeted therapies, which is linked to a limited understanding of the metastatic cascade’s molecular mechanisms. This knowledge gap arises from inadequate models: traditional 2D in vitro tissue culture oversimplifies, and in vivo mammalian studies face challenges in human application. Recent developments employ 3D cell culture systems in microfluidic devices to enhance physiological relevance, incorporating mechanobiological cues, such as matrix stiffness, topography, chemistry, and shear stress, that shed light on breast cancer metastasis to bone.

Three organs-on-a-chip were developed to understand the role of mechanobiology and delivery of mechanobiological cues within microfluidic devices.

A bone-on-a-chip model was designed using polymeric microparticles (made from poly(lactic acid), 40.5 ± 5.0 μm in diameter) to support mesenchymal stem cell (MSC) culture and differentiation down an osteogenic lineage. Hydroxyapatite coating on these microparticles, visualised with SEM and confirmed with Raman spectroscopy, further promoted MSC differentiation, while 3D cell-microparticle cultures maintained viability for 5 days in an optimised microfluidic device.
Breast cancer cell spheroids were generated, ranging from 1,000 to 100,000 cells per spheroid, with growth assessed by brightfield imaging and viability assessed by fluorescent imaging on day five in both well plates and custom microfluidic devices. The complexity of optimising spheroid sizing for use within microfluidic devices was demonstrated.
An endothelial microfluidic device was also designed to support endothelial cell growth. Human umbilical vein cell adhesion, viability, and stability were observed on poly(lacticco glycolic) microparticles and fibrin hydrogels. While network formation was noted in fibrin hydrogels over three days in culture, it was not confirmed in the microfluidic device.

The work described herein demonstrates the use of biomaterials to deliver mechanobiological cues and shows how microfluidic devices offer a potential solution for modelling different tissue types.