(D-150) Elucidating Oxygen-Dependent Crosstalk in Human Cardiac Fibroblasts Post-Infarction Using Microfluidic Devices from 3D Printed Templates

Introduction:: A heart attack blocks downstream tissue from receiving vital oxygen, leading to localized hypoxia (0% O₂) and subsequent cell death. To repair injured tissue, cardiac fibroblasts activate into a myofibroblast phenotype, which deposits matrix proteins that form a scar. In many cases, excessive scarring occurs which leads to heart failure. We hypothesize that crosstalk between hypoxic, injured tissue and normoxic (10% O₂), healthy tissue may regulate excessive scarring after infarction, but this has been unexplored due to limitations of existing research tools, which are unable to coculture hypoxic and normoxic cells in real-time. Our goal was to engineer modular, microfluidic devices from 3D-printed templates that enable real-time crosstalk between hypoxic and normoxic cells. We then measured human cardiac fibroblast phenotypes at the gene and protein level after oxygen-dependent cell-cell signaling.


Materials and Methods:: Primary human cardiac fibroblasts were first cultured on polydimethylsiloxane (PDMS)-coated coverslips, serum-arrested, and then exposed to uniform hypoxia treatment in a hypoxia chamber (0% O₂) for various timepoints (0-24h) with TGF-β1 as a positive control for myofibroblast activation, to identify timepoints for microfluidic device experiments. Then, microfluidic devices with channels for hypoxic and normoxic gas flow were fabricated through Digital Light Processing 3D printing and polydimethylsiloxane (PDMS) soft lithography and were validated with an optical O₂ sensor. Next, fibroblasts were cultured on PDMS devices that enable hypoxic-normoxic cell crosstalk for 4 or 24h and compared with uniform hypoxia or normoxia control conditions. Cells were stained for hypoxia inducible factor-1α, a transcription factor upregulated in low O₂; α-smooth muscle actin (α-SMA), a myofibroblast marker; proliferation (Ki67); and DAPI. Cells were also lysed and RNA and protein were isolated for quantitative PCR and proteomic analysis. For qPCR, target genes included VEGFA, an angiogenic growth factor and known target of HIF.


Results, Conclusions, and Discussions:: In human cardiac fibroblasts exposed to uniform hypoxia, HIF-1α accumulates in the nuclei by 4h (Fig. 1A) and was followed by VEGFA gene upregulation by 24h (data not shown). Proliferation (Ki67) is inhibited by 24h hypoxia to levels comparable to TGF-β1 treatment (Fig. 1A). Finally, activation into myofibroblasts (α-SMA coverage) is upregulated by TGF-β1 but not hypoxia durations less than 24h (data not shown). Proteomic analysis demonstrates that both 24h hypoxia treatment or TGF- β1 treatment enrich various fibrotic proteins, including type I and type II collagens (CO1A1, CO1A2) (Fig. 1B). Microfluidic devices demonstrate stable and adjacent regions of hypoxia and normoxia (Fig. 1C). In the device that enables real-time crosstalk between hypoxic and normoxic cells, the same phenotypes were observed in hypoxic fibroblasts regardless of coculture with normoxic fibroblasts, indicating that increased HIF activity by 4h, decreased proliferation by 24h, and unchanged α-SMA expression after 24h are not affected by crosstalk with normoxic cells (Fig. 1D). Tissues cultured on microfluidic devices have been lysed and both RNA and protein have been isolated for sequencing and proteomic analysis after crosstalk between hypoxic and normoxic cells. By looking at total gene and protein expression, these studies may identify novel changes in RNA or protein levels in hypoxic fibroblasts after crosstalk with normoxic fibroblasts, or vice versa, and identify molecular targets for anti-fibrotic strategies. In the future, the device will be implemented to explore oxygen-dependent crosstalk between cardiac myocytes and fibroblasts.


Acknowledgements (Optional): :

References (Optional): :