Background: Corneal stromal cells (keratocytes) are responsible for developing and maintaining normal corneal structure and transparency, and for repairing the tissue after injury. Corneal keratocytes reside between highly aligned collagen lamellae in vivo. In addition to growth factors and other soluble biochemical factors, feedback from the extracellular matrix (ECM) itself has been shown to modulate corneal keratocyte behavior. Methods: In this study, we fabricate aligned collagen substrates using a microfluidics approach and assess their impact on corneal keratocyte morphology, cytoskeletal organization, and patterning after stimulation with platelet derived growth factor (PDGF) or transforming growth factor beta 1 (TGFβ). We also use time-lapse imaging to visualize the dynamic interactions between cells and fibrillar collagen during wound repopulation following an in vitro freeze injury. Results: Significant co-alignment between keratocytes and aligned collagen fibrils was detected, and the degree of cell/ECM co-alignment further increased in the presence of PDGF or TGFβ. Freeze injury produced an area of cell death without disrupting the collagen. High magnification, time-lapse differential interference contrast (DIC) imaging allowed cell movement and subcellular interactions with the underlying collagen fibrils to be directly visualized. Conclusions: With continued development, this experimental model could be an important tool for accessing how the integration of multiple biophysical and biochemical signals regulate corneal keratocyte differentiation.

No abstract

Efficient intracellular cargo delivery is a key hurdle for the translation of many emerging stem cell and cellular reprogramming therapies. Recently, a microfluidic-based device constructed from silicon was shown to transduce macromolecules into cells via shear-induced formation of plasma membrane pores. However, the scalability and widespread application of the current platform is limited since physical deformation-mediated delivery must be optimized for each therapeutic application. Therefore, we sought to create a low-cost, versatile device that could facilitate rapid prototyping and application-specific optimization in most academic research labs. Here we describe the design and implementation of a microfluidic device constructed from Polydimethylsiloxane (PDMS) that we call Cyto-PDMS (Cytoplasmic PDMS-based Delivery and Modification System). Using a systematic Cyto-PDMS workflow, we demonstrate intracellular cargo delivery with minimal effects on cellular viability. We identify specific flow rates at which a wide range of cargo sizes (1-70 kDa) can be delivered to the cell interior. As a proof-of-principle for the biological utility of Cyto-PDMS, we show (i) F-actin labeling in live human fibroblasts and (ii) intracellular delivery of recombinant Cre protein with appropriate genomic recombination in recipient fibroblasts. Taken together, our results demonstrate that Cyto-PDMS can deliver small-molecules to the cytoplasm and biologically active cargo to the nucleus without major effects on viability. We anticipate that the cost and versatility of PDMS can be leveraged to optimize delivery to a broad array of possible cell types and thus expand the potential impact of cellular therapies.

Traditional leukocyte adhesion assays have provided significant insight into the mechanisms of leukocyte rolling in part through the use of homogeneously coated surfaces. These assays typically involve protein coating of glass coverslips or plastic petri dishes applied via a static drop of protein solution. With this approach, it is difficult to spatially control the location of proteins to fabricate surface-bound protein gradients that mimic in vivo situations. Microfluidic patterning of proteins with microfluidic devices has become a popular technique due to the ability to spatially pattern proteins on a cellular scale. Despite the advantages of microfluidic patterning, few studies have systematically investigated the effects of perfusion time, protein concentration, and perfusion shear stress on protein deposition. Herein, we demonstrated the fabrication of both line and step gradients of P-selectin on glass substrates that support cell rolling and adhesion assays. Investigation of the flow conditions during the microfluidic patterning led to several significant findings. We observed that the protein deposition time of 5 min was sufficient to deposit adequate P-selectin to support neutrophil rolling. We demonstrated that the amount of membrane P-selectin (mP-selectin) or recombinant P-selectin (rP-selectin) deposited showed a dependence on the perfusion shear stress between 4.0 and 32.0 dyn/cm(2), while similar studies with fibronectin or fibrinogen showed no shear stress dependence. Finally, we also created step changes in surface adherent protein concentration of P-selectin to characterize leukocyte-rolling behavior in response to sudden changes in ligand density.

In vitro systems comprised of wells interconnected by microchannels have emerged as a platform for the study of cell migration or multicellular models. In the present study, we systematically evaluated the effect of microchannel width on spontaneous myoblast migration across these microchannels—from the proximal to the distal chamber. Myoblast migration was examined in microfluidic devices with varying microchannel widths of 1.5–20 µm, and in chips with uniform microchannel widths over time spans that are relevant for myoblast-to-myofiber differentiation in vitro. We found that the likelihood of spontaneous myoblast migration was microchannel width dependent and that a width of 3 µm was necessary to limit spontaneous migration below 5% of cells in the seeded well after 48 h. These results inform the future design of Polydimethylsiloxane (PDMS) microchannel-based co-culture platforms as well as future in vitro studies of myoblast migration.

During corneal wound healing, corneal keratocytes are exposed to both biophysical and soluble cues that cause them to transform from a quiescent state to a repair phenotype. How keratocytes integrate these multiple cues simultaneously is not well understood. To investigate this process, primary rabbit corneal keratocytes were cultured on substrates patterned with aligned collagen fibrils and coated with adsorbed fibronectin. After 2 or 5 days of culture, keratocytes were fixed and stained to assess changes in cell morphology and markers of myofibroblastic activation by fluorescence microscopy. Initially, adsorbed fibronectin had an activating effect on the keratocytes as evidenced by changes in cell shape, stress fiber formation, and expression of alpha-smooth muscle actin (α-SMA). The magnitude of these effects depended upon substrate topography (i.e., flat substrate vs aligned collagen fibrils) and decreased with culture time. When keratocytes were simultaneously exposed to adsorbed fibronectin and soluble platelet-derived growth factor-BB (PDGF-BB), the cells elongated and had reduced expression of stress fibers and α-SMA. In the presence of PDGF-BB, keratocytes plated on the aligned collagen fibrils elongated in the direction of the fibrils. These results provide new information on how keratocytes respond to multiple simultaneous cues and how the anisotropic topography of aligned collagen fibrils influences keratocyte behavior.