Optogenetic‐mediated cardiovascular differentiation and patterning of human pluripotent stem cells

Abstract Precise spatial and temporal regulation of dynamic morphogen signals during human development governs the processes of cell proliferation, migration, and differentiation to form organized tissues and organs. Tissue patterns spontaneously emerge in various human pluripotent stem cell (hPSC) models. However, the lack of molecular methods for precise control over signal dynamics limits the reproducible production of tissue patterns and a mechanistic understanding of self‐organization. We recently implemented an optogenetic‐based OptoWnt platform for light‐controllable regulation of Wnt/β‐catenin signaling in hPSCs for in vitro studies. Using engineered illumination devices to generate light patterns and thus precise spatiotemporal control over Wnt activation, here we triggered spatially organized transcriptional changes and mesoderm differentiation of hPSCs. In this way, the OptoWnt system enabled robust endothelial cell differentiation and cardiac tissue patterning in vitro. Our results demonstrate that spatiotemporal regulation of signaling pathways via synthetic OptoWnt enables instructive stem cell fate engineering and tissue patterning.


| INTRODUCTION
Human pluripotent stem cells (hPSCs) possess great potential in developmental biology studies, regenerative medicine, disease modeling, and drug screening. For example, functional cardiomyocytes 1,2 (CMs) and cardiac organoids 3,4 have been recently produced from hPSCs via temporal regulation of Wnt signaling for the elucidation of human heart development. However, a broader application of hPSCs is currently limited by inability to create more physiologically relevant models that recapitulate the cell diversity and spatial organization of tissues. Precise spatial and temporal control of developmental pathways within an initially homogenous cell population would direct differentiation to Peter B. Hellwarth and Yun Chang contributed equally to this study. generate organized patterns of heterogeneous cell populations and better recapitulate in vivo tissue structures. These heterogeneous cultures would permit cross cell-type interactions that do not occur in mono-cultures, allowing for synergistic maturation in cell phenotypes and more translatable models. 5 For instance, micropatterning technology has been developed to construct spatially-patterned cardiac organoids for developmental biology study and toxicity testing. 4,6 The light-responsive optogenetic system OptoWnt 7,8 attains the desired precise spatial and temporal control representative of an ideal synthetic signaling construct. OptoWnt is composed of the photolyase homology domain of the Arabidopsis thaliana blue-light photoreceptor cryptochrome 2 (Cry2) conjugated to the cytoplasmic fragment of the Wnt co-receptor LRP6 (LRP6c). 9 The complexity of differentiation obtainable using current differentiation protocols based on dif-

| Design of a genetically engineered hPSC line for optogenetic Wnt activation
The canonical Wnt signaling pathway is initiated when Wnt ligands bind to the transmembrane receptor Frizzled, leading to the clustering of LRP6, a Wnt co-receptor. 11 The formation of clusters triggers the phosphorylation of LRP6 and induces the downstream signaling cascade to stabilize β-catenin, which is translocated to the nucleus and activates Wnt target genes, including mesoderm marker brachyury (or T). 12  To achieve stable expression of OptoWnt in hPSCs for lightinduced activation of Wnt, we firstly knocked the Cry2-LRP6c-mCherry transgene into the endogenous AAVS1 safe harbor locus 13 ( Figure 1A) via CRISPR/Cas9-mediated homologous recombination. The employment of P2A-linked mCherry (2A-mCherry) enabled the rapid identification and isolation of successfully targeted clones ( Figure 1B), which were subjected for further homozygosity analysis ( Figure 1C). The resulting OptoWnt hPSCs retained a pluripotent phenotype with expression of OCT4 and SSEA4 ( Figure 1D). Since Wnt activation induces hPSC differentiation into brachyury + (or T + ) mesoderm, 14,15 we measured the percentage of T + cells post-illumination to optimize the blue light intensity using a LAVA illumination device 8 ( Figure S1A,B). After 48 hours of light illumination, the percentages of T + cells were proportional to blue light intensity and approached saturation above 1 μW/mm 2 , yielding $90% T + cells and comparable to the positive control using a Wnt agonist CHIR99021 (CHIR, 6 μM) ( Figure 1E, Figure S2). Our results demonstrated the feasibility of using light illumination to manipulate Wnt signaling in hPSCs for cell fate engineering and patterning.

| Light-induced EPC differentiation of hPSCs
We further differentiated hPSCs into EPCs by replacing our previous CHIR-based protocol 16

| Light-induced cardiac differentiation in hPSCs
To further demonstrate the utility and versatility of the OptoWnt system, we differentiated hPSCs into another mesoderm-derived lineage À CMs À after light-induced activation of Wnt signaling in hPSCs. We previously developed a small molecule-based GiWi protocol to induce cardiac differentiation in hPSCs by temporal regulation of Wnt signaling, in which Gsk3β inhibitor (Gi) CHIR was used for mesoderm induction followed by treatment with Wnt inhibitor IWP2 or Wnt-C59 to produce functional CMs. Here we  10 To reduce long-term illumination toxicity, we applied our previous GiWi protocol to obtain day 6 cardiac progenitors from OptoWnt hPSCs, which were then seeded onto plates with vinyl photomasks for blue light illumination from day 7 to 9 and subjected for immunostaining at day 15. (B) Patterned light illumination enabled proof-of-concept co-differentiation of hPSC-derived day 6 cardiac progenitors into cardiac and epicardial cells in a specified ring geometry of epicardium-myocardium-epicardium (right), as compared to a homogeneous myocardium layer with a full mask (left). Full mask was removed right before imaging. Scale bars, 2 mm dimensions (3D), 19,20 we illuminated day 6 cardiac progenitor aggregates for 24 hours, leading to self-organized cardiac organoids with a cavity ( Figure S3B). Our results highlight the potential of our OptoWnt platform for light-controllable tissue patterning in hPSCs.

| DISCUSSION
In this study, we used an OptoWnt hPSC line to show that optogenetic-mediated activation of Wnt signaling could be applied to

| Light-activated Wnt activation
Utilizing a LAVA device developed by Repina et al,7,18 the intensity, timing, and uniformity of light exposure could be controlled for hPSC differentiation. The LAVA board was programmed to an intensity of 1 μW/mm 2 , which achieves a similar mesoderm induction to exposure to 6 μM CHIR99021. To activate Wnt, the OptoWnt cells were seeded onto Matrigel-coated plates and placed on top of the LAVA device. If a patterned signaling was desired, the regionally transparent and regionally opaque mask was printed and placed underneath the culture dish, blocking specified regions of the cell culture dish from being exposed to the light. Both the cells and the LAVA board were placed in the tissue culture incubator at 37 C and 5% CO 2 .

| EPC differentiation
Endothelial progenitors were obtained via a GSK-3β inhibition or light exposure followed by an exposure to VEGF as previously reported. 16 Briefly, mesoderm induction was achieved in OptoWnt cells either by 1-to 2-day illumination of 470 nm light or by culturing with 6 μM CHIR99021 in LaSR basal medium. Starting on day 2, cells were exposed to 25 to 50 ng/mL of VEGF with medium change every 24 hours. On day 5, the cells were analyzed for flow cytometry and immunostaining of CD31, CD34, and VECAD.

| Cardiomyocyte differentiation
In order to directly differentiate hPSCs into CMs, we followed a modified version of our previous GiWi protocol. 1,2 Mesoderm induction was achieved in OptoWnt hPSCs either by 470 nm light exposure or by culturing with 6 μM CHIR99021 in RPMI medium on day 0. On day 3, 2 μM Wnt-C59 was used to induce cardiac differentiation in RPMI medium supplemented with 200 μg/mL ascorbic acid (AA) and 0.1% human serum albumin (HSA). On day 5, medium change with RPMI/AA/HSA. On day 7 and every 3 days afterward, medium change with RPMI/B27 medium. For co-differentiation of cardiac and epicardial cells in Figure 4, GiWi protocol was used to obtain day 6 cardiac progenitor cells, which were then seeded onto Matrigelcoated plates with vinyl photomasks for blue light illumination.

| Immunostaining and flow cytometry analysis
For immunostaining analysis, cells were fixed in PBSÀ/À with 4% paraformaldehyde for 15 minutes and washed twice with room temperature PBSÀ/À. Fixed cells were then stained with appropriate primary and secondary antibodies (Table S1) in PBSÀ/À solution with 5% nonfat dry milk and 0.4% Triton X-100 followed by nuclei staining.
The stained cells were then imaged and processed with Leica DMi-8 fluorescent microscope and ImageJ, respectively. Image threshold settings were held constant. For flow cytometry analysis, endothelial and cardiac cells were dissociated with TrypLE and filtered through a strainer. Singularized CMs were then fixed in 1% paraformaldehyde PBSÀ/À solution for 20 minutes and permeabilized in 90% cold methanol for at least an hour. After washing twice in FlowBuffer-2 (PBS with 0.5% BSA and 0.1% Triton X-100), CMs were stained with cTnT primary and secondary antibodies. Singularized endothelial cells were washed once with FlowBuffer-1 (PBS with 0.5% BSA) and stained with conjugated CD31 and CD34 antibodies for 30 minutes at room temperature. Flow data was collected in a BD Accuri C6 plus machine.