UC Irvine

Intercellular communication is a fundamental driver of tissue function across developing, mature, as well as diseased organ systems. We have developed a variety of tools for controlling the patterning of cells at the microscale, each enabling experiments targeting different aspects of cell communication.

Organ-on-a-chip models often reconstitute tissue interfaces across a porous membrane that is intended to maintain compartmentalization of different cell layers while preserving cell- cell interactions. However, such membranes are typically substantially thicker and exhibit lower porosity than natural basement membrane. We optimized a photolithographic process in 1002F epoxy resin to produce thin microporous membranes with good biocompatibility, optical transparency, and low auto-fluorescence. Characterization of fabrication limits by profilometry and SEM show we are able to produce membranes under 1 μm in thickness with pores as small as 0.8 μm. Cells seeded on either side of the membrane can interact through the pores but do not migrate across. 1002F membranes were employed to divide a two- chambered microfluidic model system of perivascular endometrial stroma, which successfully recapitulates decidualization in response to combined progesterone and oestrogen treatment. This model could potentially be applied to screening for drug toxicity effects on endometrial tissue, which remains an under explored area of women’s health.

Secondly, micropatterning of planar co-cultures can achieve precise organization suitable for optimized organ models or the study of developmental processes. We devised a reconfig- urable elastomeric substrate to pattern a very clean boundary interface between two cell populations. This system has been employed to recapitulate morphogen gradients in vitro. A significant limitation shared across most cell micropatterning approaches is the inability to pattern more than two populations. We adopted a DNA-programmed cell adhesion tech- nique to extend patterning to three or more cell populations. This approach combines the incorporation of short DNA oligomers onto cell surfaces with glass substrates patterned with complimentary DNA. Using two different pairs of DNA compliments allowed patterning of two different cell types onto different regions of a single substrate, with less than 5% cross contamination. We also demonstrate the feasibility of a simple microfluidic approach to immobilizing DNA on glass with μm-scale resolution over cm-scale areas without the use of expensive micropatterning instruments. Expanding the number of patterned cell types may improve the physiological relevance of in vitro models of complex organ systems such as the liver.

Finally, we establish a platform for engineering the photoswitchable protein interaction of PhyB and PIF for optogenetic control of gene expression, with the goal of patterning cell phenotype using light. Exposure to 650nm red light induces PhyB and PIF to interact, while 750nm far-red light drives their dissociation. We tied PhyB-PIF interaction to reporter gene expression in a yeast two-hybrid assay and studied the relationship between red light pulse duration and induced gene expression using a programmable LED array. We also present work towards engineering this response through altering subcellular localization of the PIF component, which could allow encoding distinct responses by tuning pulse length.