High throughput microfluidic system with multiple oxygen levels for the study of hypoxia in tumor spheroids.

Replication of physiological oxygen levels is fundamental for modeling human physiology and pathology inmodels. Environmental oxygen levels, applied in mostmodels, poorly imitate the oxygen conditions cells experience, where oxygen levels average ∼5%. Most solid tumors exhibit regions of hypoxic levels, promoting tumor progression and resistance to therapy. Though this phenomenon offers a specific target for cancer therapy, appropriateplatforms are still lacking. Microfluidic models offer advanced spatio-temporal control of physico-chemical parameters. However, most of the systems described to date control a single oxygen level per chip, thus offering limited experimental throughput. Here, we developed a multi-layer microfluidic device coupling the high throughput generation of 3D tumor spheroids with a linear gradient of five oxygen levels, thus enabling multiple conditions and hundreds of replicates on a single chip. We showed how the applied oxygen gradient affects the generation of reactive oxygen species (ROS) and the cytotoxicity of Doxorubicin and Tirapazamine in breast tumor spheroids. Our results aligned with previous reports of increased ROS production under hypoxia and provide new insights on drug cytotoxicity levels that are closer to previously reportedfindings, demonstrating the predictive potential of our system.