Calcium signaling mediates proliferation of the precursor cells that give rise to the ciliated left-right organizer in the zebrafish embryo

Several of our internal organs, including heart, lungs, stomach, and spleen, develop asymmetrically along the left-right (LR) body axis. Errors in establishing LR asymmetry, or laterality, of internal organs during early embryonic development can result in birth defects. In several vertebrates—including humans, mice, frogs, and fish—cilia play a central role in establishing organ laterality. Motile cilia in a transient embryonic structure called the “left-right organizer” (LRO) generate a directional fluid flow that has been proposed to be detected by mechanosensory cilia to trigger asymmetric signaling pathways that orient the LR axis. However, the mechanisms that control the form and function of the ciliated LRO remain poorly understood. In the zebrafish embryo, precursor cells called dorsal forerunner cells (DFCs) develop into a transient ciliated structure called Kupffer’s vesicle (KV) that functions as the LRO. DFCs can be visualized and tracked in the embryo, thereby providing an opportunity to investigate mechanisms that control LRO development. Previous work revealed that proliferation of DFCs via mitosis is a critical step for developing a functional KV. Here, we conducted a targeted pharmacological screen to identify mechanisms that control DFC proliferation. Small molecule inhibitors of the sarcoplasmic/endoplasmic reticulum Ca2+-ATPase (SERCA) were found to reduce DFC mitosis. The SERCA pump is involved in regulating intracellular calcium ion (Ca2+) concentration. To visualize Ca2+ in living embryos, we generated transgenic zebrafish using the fluorescent Ca2+ biosensor GCaMP6f. Live imaging identified dynamic cytoplasmic Ca2+ transients (“flux”) that occur unambiguously in DFCs. In addition, we report Ca2+ flux events that occur in the nucleus of DFCs. Nuclear Ca2+ flux occurred in DFCs that were about to undergo mitosis. We find that SERCA inhibitor treatments during DFC proliferation stages alters Ca2+ dynamics, reduces the number of ciliated cells in KV, and alters embryo laterality. Mechanistically, SERCA inhibitor treatments eliminated both cytoplasmic and nuclear Ca2+ flux events, and reduced progression of DFCs through the S/G2 phases of the cell cycle. These results identify SERCA-mediated Ca2+ signaling as a mitotic regulator of the precursor cells that give rise to the ciliated LRO.


Supplemental Figure and Movie Legends
Movie S4.Visualization of Ca 2+ dynamics in DFCs.Time-lapse confocal imaging of Ca 2+ flux events in wild-type Tg(act2b:GCaMP6f); Tg(sox17:EGFP-caax) embryos starting at the 60% epiboly stage.DFC cell membranes are labeled by Tg(sox17:EGFPcaax) expression, and changes in Ca 2+ concentration (GCaMP6f fluorescence intensity) are visualized using the cyan hot lookup table (FIJI software).This is a maximum projection of the entire DFC cluster.Timestamp= hours:minutes:seconds:milliseconds.Movie S5.Visualization of Ca 2+ dynamics in DMSO control treated DFCs.Timelapse confocal imaging of Ca 2+ flux events in Tg(act2b:GCaMP6f); Tg(sox17:EGFPcaax) embryos treated with 1% DMSO at the 60% epiboly stage for 60 min.DFCs were imaged at at the 70% epiboly stage.DFC cell membranes are labeled by Tg(sox17:EGFP-caax) expression, and changes in Ca 2+ concentration (GCaMP6f fluorescence intensity) are visualized using the cyan hot lookup table (FIJI software).This is a maximum projection of the entire DFC cluster.Timestamp= hours:minutes:seconds:milliseconds.Movie S6.Visualization of Ca 2+ dynamics in thapsigargin treated DFCs.Time-lapse confocal imaging of Ca 2+ flux events in Tg(act2b:GCaMP6f); Tg(sox17:EGFP-caax) embryos treated with 1 μM thapsigargin at the 60% epiboly stage for 60 min.DFCs were imaged at at the 70% epiboly stage.DFC cell membranes are labeled by Tg(sox17:EGFP-caax) expression, and changes in Ca 2+ concentration (GCaMP6f fluorescence intensity) are visualized using the cyan hot lookup table (FIJI software).This is a maximum projection of the entire DFC cluster.Timestamp= hours:minutes:seconds:milliseconds.Movie S7.Visualization of Ca 2+ dynamics in cyclopiazonic acid treated DFCs.Time-lapse confocal imaging of Ca 2+ flux events in Tg(act2b:GCaMP6f); Tg(sox17:EGFP-caax) embryos treated with 100 μM cyclopiazonic acid at the 60% epiboly stage for 60 min.DFCs were imaged at at the 70% epiboly stage.DFC cell membranes are labeled by Tg(sox17:EGFP-caax) expression, and changes in Ca 2+ concentration (GCaMP6f fluorescence intensity) are visualized using the cyan hot lookup table (FIJI software).This is a maximum projection of the entire DFC cluster.Timestamp= hours:minutes:seconds:milliseconds.Movie S8.DFC behavior following cytoplasmic Ca 2+ fluxes.The behavior of a DFC (arrow) in a Tg(act2b:GCaMP6f); Tg(sox17:EGFP-caax) embryo that undergoes multiple cytoplasmic Ca 2+ flux events (asterisks) was tracked for 20 min.These cytoplasmic Ca 2+ flux events did not result in gross changes in DFC morphology, behavior, or position in this timeframe.This is a single focal plane within the DFC cluster.Timestamp= minutes:seconds.Movie S9.DFC mitosis following nuclear Ca 2+ flux.The behavior of a DFC (arrow) in a Tg(act2b:GCaMP6f); Tg(sox17:EGFP-caax) embryo that undergoes a nuclear Ca 2+ flux event (asterisk) was tracked for 20 min.The DFC rounded up and divided into two daughter cells.This is a single focal plane within the DFC cluster.Timestamp= minutes:seconds.

Fig. S1 .
Fig. S1.SERCA inhibitor treatments do not alter the mitotic index of non-DFCs at the dorsal margin.(A) Representative merged images of pHH3 staining in Tg(sox17:EGFP-caax) embryos treated with 1% DMSO (control) or 1 μM Thaps at 60% epiboly for 60 min.(B) Bar graphs indicate average mitotic index of neighboring non-DFCs in DMSO (control) and 1 μM Thaps treated embryos and error bars represent one standard deviation.Each circle on the graph represents results from an individual embryo.An unpaired two-tailed t-test with Welch's correction was used for statistical analysis.ns=not significant (p=0.4446).

Fig. S2 .
Fig. S2.SERCA inhibitor treatments during epiboly alters heart looping.(A-B) Embryos at 2 days post-fertilization that were treated with 1% DMSO (control) (A) or 1 μM thapsigargin (B) at 60% epiboly for 60 min.(C-E) Representative images of normal rightward looping of the heart (arrow) in a control embryo (C), and midline (D) or reversed (E) looping in thapsigargin treated embryos.The heart was labeled by EGFP expression in cardiomyocytes.V=ventricle; A=atrium.(F) Rightward heart looping was observed in most control embryos, whereas the heart often remained along the midline or looped to the left in thapsigargin treated embryos.n=number of embryos analyzed.

Fig. S3 .
Fig. S3.Spatial location of DFC cytoplasmic Ca 2+ flux events in individual embryos.Solid colored lines indicate the boundary of the DFC cluster in an individual embryo as determined by confocal images (see Fig. 6B).The DFC cluster was divided into quadrants based on the overall length and height of the cluster.Colored ovals indicate the location of cytoplasmic Ca 2+ flux events.Dashed lines group clusters of DFC cytoplasmic Ca 2+ flux events.The DFC cluster for all 5 embryos were overlayed to create Fig. 6C.L=left, R=right, A=anterior, P=posterior.

Fig. S4 .
Fig. S4.Spatial location of DFC nuclear Ca 2+ fluxes in individual embryos.(A) Solid colored lines indicate the boundary of the DFC cluster in an individual embryo as determined by confocal images (see Fig. 7B).The DFC cluster was divided into quadrants based on the overall length and height of the cluster.Colored ovals indicate the location of nuclear Ca 2+ flux events.The DFC cluster for all 5 embryos were overlayed to create Fig. 7C.(B-D) Location of DFC nuclear Ca 2+ flux events in DFC quadrants (B), along the LR axis (C), and along the AP axis (D).L=left, R=right, A=anterior, P=posterior.