Sea urchins of the species Strongylocentrotus purpuratus were obtained from Marinus Scientific (Lakewood, CA) or collected from Slip Point, a wave-exposed intertidal site in Clallam Bay, WA, United States (48.26260, −124.2532) under the auspices of a series of Washington Department of Fish and Wildlife-approved scientific collecting permits. Animals collected in WA were maintained in subtidal cages year-round at the University of Washington Friday Harbor Laboratories (FHL, Friday Harbor, WA) or in running natural sea water or closed artificial sea water tanks at 10°C–15°C. Animals kept at FHL were fed ad libitum with bullwhip kelp, Nereocystis luetkeana, whereas animals in closed artificial sea water tanks were fed dried seaweed from Omega One Seaweed (Blacksburg, VA). LLC-PK1 cells expressing LifeAct-EGFP were obtained courtesy of Dr. Patricia Wadsworth (University of Massachusetts) and cultured according to the methods of Beach et al. (2014).

Primary antibodies used included a rabbit polyclonal antibody raised against sea urchin egg myosin II heavy chain isolated via ATP-based precipitation of actomyosin from S. purpuratus egg extracts and electrophoretically purified (Henson et al., 1999), an affinity purified rabbit polyclonal antibody raised against the PH domain of S. purpuratus anillin (Garno et al., 2021), a rabbit polyclonal antibody against the C-terminal region of human α-actinin from Santa Cruz Biotechnology (Dallas, TX), a mouse monoclonal antibody against the Ser19 phosphorylated form of myosin II regulatory light chain (P-Myo) from Cell Signaling Technology (Danvers, MA), and a rabbit monoclonal antibody against a peptide from human septin 2 from Abcam, Inc. (Cambridge, MA). Appropriate secondary antibodies conjugated to Alexa Fluor 488, 555, 568, 633, 647, or Oregon Green as well as Alexa Fluor 488, 555, 633, and 647 conjugated phalloidin were obtained from Thermo Fisher Scientific (Waltham, MA). In order to generate secondary antibodies conjugated to the fluorophore Cy3B suitable for iPALM imaging, Cy3b-NHS ester from Amersham Biosciences (Piscataway, NJ) was resuspended to 10 mg/mL in a solvent comprised of a 1:1mixture of dimethyl sulfoxide (DMSO) and dimethyl formamide. Goat anti-mouse IgG and anti-rabbi IgG was resuspended at 1 mg/mL in 0.1 M NaHCO₃, pH 8.3. Dye and antibody were combined at a molar ratio of 3:1 for an hour at room temperature in the dark. Labeled antibody was separated from free dye using spin columns from an Alexa Fluor Labeling Kit (Thermo Fisher Scientific). Dye labeling was determined using spectrophotometry and working dilutions were determined empirically. Para-aminoblebbistatin (PAB) and Latrunculin A (LatA) were obtained from Cayman Chemical (Ann Arbor, MI). Unless otherwise indicated, the majority of reagents were purchased from either Sigma-Aldrich (St. Louis, MO) or Thermo Fisher Scientific.

Sea urchin gametes were collected via intracoelomic injection with 0.5 M KCl, with sperm collected dry and eggs spawned in either natural sea water or MBL artificial sea water (ASW: 423 mM NaCl, 9 mM KCl, 9.27 mM CaCl₂, 22.94 mM MgCl₂, 25.5 mM MgSO₄, 2.14 mM NaHCO₃, pH 8.0) and subsequently dejellied by multiple washing with ASW. Sea urchin coelomocytes were isolated from the perivisceral fluid of adult animals and maintained in coelomocyte culture media (0.5 M NaCl, 5 mM MgCl₂, 1 mM EGTA, and 20 mM HEPES, pH 7.2) as described in Smith et al. (2019).

Eggs were fertilized by addition of dilute sperm, the fertilization envelopes removed using 1 M urea (pH 8.0), and then washed into and reared in MBL calcium free sea water (CFSW: MBL ASW minus CaCl₂ and plus 1 mM EGTA) at 10°C–15°C. Cleavage cortices were generated as described in Henson et al. (2019). In brief, embryos at the appropriate stage of cell division were allowed to quickly settle onto poly-L-lysine (2 mg/mL in 250 mM borate buffer, pH 8.0) coated coverslips and then exposed to fluid shear force from a pipette containing an isotonic cortex isolation buffer (CIB: 0.8 M mannitol, 5 mM MgCl₂, 10 mM EGTA, 100 mM HEPES, pH 7.4). Isolated cortices were rinsed twice in CIB prior to further processing for fluorescent staining.

For disruption of actin filaments, embryos were treated with either 100 nM LatA or an equivalent dilution of the drug carrier ethanol in CFSW starting at 30 min prior to first division. For myosin II inhibition, embryos 60 min prior to first division were treated with 50 or 100 µM of PAB or the equivalent dilution of the drug carrier DMSO in CFSW. PAB and DMSO were added to CFSW coincident with rapid vortexing in order to ensure dissolution and treated embryos were protected from light exposure given the photosensitivity of PAB. For imaging live embryos treated with PAB, ∼200–300 µL of embryos in CFSW plus PAB were placed on a slide in a well created by nail polish and observed in brightfield or phase contrast microscopy using either a 4X/0.1 NA or a 10X/0.25 NA objective lens on a Nikon (Tokyo, Japan) TS100 microscope. Images were captured using a Photometrics (Tuscon, AZ) QImaging Retiga 2000R cooled CCD camera. For PAB photoinactivation, embryos + PAB were imaged and then exposed to light from four fiber optic light pipes of two Nikon NI-150 high intensity illuminators set at the maximum intensity. After 5–7 min the lights were turned off and the embryos were imaged again to monitor the extent of the reactivation of cytokinesis in the embryos prior to fixation.

For widefield or laser scanning confocal fluorescence microscopy, whole embryos, either attached to poly-L-lysine coated coverslips or kept in suspension, were fixed in Millonig’s fixative (0.2 M NaH₂PO₄, 0.136 M NaCl, pH 7.0) containing 3.7% formaldehyde for 20 min, washed 3X and then left 2 h to overnight in 0.1% Triton X-100 in phosphate buffered saline (PBST), and then blocked 2 h to overnight in 3% BSA diluted in PBST. Isolated cortices were fixed in 2%–3% formaldehyde in CIB for 5 min followed by blocking in 2% goat serum and 1% BSA in PBS for at least 30 min. Coelomocytes were settled and fixed as described in Smith et al. (2019), and LLC-PK1 cultured cells were fixed in 3.7% formaldehyde in PBS prior to permeabilization in PBST. Immunostaining of all samples was performed with appropriate primary and secondary antibodies diluted in blocking buffer with embryos typically being stained in each antibody step for >1.5 h whereas cortices, coelomocytes, and cultured cells were stained for 45 min in each stage. Fluorescent phalloidin was added to the secondary antibody staining step, and nuclei in embryos and coelomocytes samples were stained in a tertiary step using DAPI in PBST for 10 min. Nuclei in live embryos were fluorescently labeled by incubating cells for 10 min in CFSW containing the membrane permeant DNA dye Hoechst-33342. Fixed samples were typically mounted in nonhardening Vectashield antifade mounting media (Vector Laboratories, Newark, CA) prior to imaging.

For fluorescent staining of isolated cortices for iPALM imaging we followed the methods detailed in Herron et al. (2022). Briefly, cortices were isolated attached to coverslips coated with gold nanorod fiducial markers used for drift correction and calibrations, fixed, quenched using 0.1 M glycine and lysine in PBS, extensively washed in large volumes of PBS, stained, and then mounted in STORM buffer using a second coverslip and this coverslip chamber was sealed with epoxy prior to imaging. iPALM imaging was restricted to secondary antibodies or phalloidin conjugated with either Alexa Flour 647 or Cy3B because of the superior photoactivation properties of these two fluorophores.

Two channel iPALM super-resolution microscopy was performed according to Herron et al. (2022) on a custom-built, noncommercial instrument designed by Hess and colleagues (Shtengel et al., 2009) and housed in the Advanced Imaging Center at the Janelia Research Campus of the Howard Hughes Medical Institute (HHMI-JRC). Samples contained in coverslip chambers were excited with either a 560 nm or a 640 nm laser and emission viewed using opposed, dual Nikon 60X/1.49 NA Apo TIRF objective lenses, with images captured using three iXon3-DU897E EMCCD cameras (Andor Technologies, Abingdon, United Kingdom). iPALM imaging of samples was very low throughput due to the requirement for high staining intensity/density, the slow search process for occasional specimens, and the requirement that tens of thousands of images of a single sample had to be captured over several hours. Images were processed and rendered as 3D tiff stacks using open source PeakSelector software developed at HHMI-JRC. The 3D stacks were rendered using a XY pixel size of 20 nm and a Z step interval of 10 nm and processed to allow for two channel color overlay and alignment, and the fluorescent signatures of the coverslip-associated nanorod fiducial markers were removed.

Conventional widefield epifluorescence microscopy of samples was performed on either a Nikon 80i microscope using either a 40X/0.75 NA Plan Fluorite or 60X/1.4 NA Plan Apo phase contrast objective lens with digital images captured using a Photometrics CoolSnap Cf cooled CCD camera or a Nikon TE2000 microscope using either a 40X/0.95 NA or 60X/1.4 NA Plan Apo DIC lens coupled with a Photometrics QImaging Retiga 2000R cooled CCD camera. Laser scanning confocal microscopy was performed on a Leica (Wetzlar, Germany) SP8 TCS instrument using either a 40X/1.3 NA or a 63X/1.4 NA Plan Apo objective lens.

For all forms of microscopic images, final processing and analysis was performed using Fiji/ImageJ (Bethesda, MD) and/or Imaris Viewer (Oxford Instruments, Abingdon, United Kingdom). Graphs were prepared and statistical analysis carried out using GraphPad Prism 8 (San Diego, CA), with box and whiskers plots generated having the following attributes: all data points are shown, boxes extend from the 25th to the 75th percentile, the horizontal line within the box marks the median, and whiskers extend to the minimum and maximum values. Final figures were prepared using Adobe Photoshop (San Jose, CA).

Gel sample lysates of either sea urchin eggs/embryos or LLC-PK1 cells were generated by pelleting cells and then adding hot 2X sample buffer (625 mM Tris-HCl pH 6.8, 25% glycerol, 2% SDS, 0.01% bromophenol blue, 5% β-mercaptoethanol) at four times the volume of packed cells followed by boiling for 5 min. Gel samples were loaded into a 4%–15% Mini-PROTEAN TGX precast protein gel (Bio-Rad, Hercules, CA) and transferred onto nitrocellulose for blotting. Total protein staining of nitrocellulose blots was performed using 0.1% Ponceau S in 5% acetic acid. Blocking and antibody incubation was performed in TBS-Tween supplemented with 3% instant milk. Bound antibody detection was performed for immunoblots using alkaline phosphatase conjugated secondary antibodies and a BCIP + NBT substrate.

Figure 1 shows widefield imaging of isolated cortices and confocal imaging of embryos immunofluorescently (IMF) stained for myosin II (P-Myo) and either septin2 or anillin illustrating the structural progression of the CR from clusters/nodes (Figures 1A, D, E, J) - to patches (Figures 1B, F, G, K) - to linear arrays (Figures 1C, H, I, L). Immunolabeling of whole embryos (Figures 1J–L) fixed and stained in suspension revealed a similar pattern of node to ring structural evolution, suggesting that the observed structures in isolated cortices were not an artifact of the cell adherence to poly-L-lysine coated coverslips.

Our earlier work employed 3D-SIM and STED super-resolution microscopy of isolated cortices to characterize the nanostructure of the mature ring as well as the organization of the early clusters (Henson et al., 2017; Garno et al., 2021). However, we needed higher resolution imaging to better visualize the potential existence of CR-associated septin filaments, as well as to define the precise Z axial positioning of septin and anillin in the context of CR organization. Therefore, we employed iPALM super-resolution microscopy developed by Hess and colleagues (Shtengel et al., 2009), which can achieve upwards of 10–20 nm 3D resolution (Shtengel et al., 2009). As a proof of principle, iPALM imaging was performed on phalloidin-labeled actin filaments in isolated first division cortices (Figure 2). These images highlighted the complexity of the actin cortical cytoskeleton which includes the core bundles of microvilli (most evident on the edge of the cortex in Figure 2A), and were similar to the cortical actin cytoskeleton visualized in our previous platinum replica TEM images of cortices (Henson et al., 2017). It also reveals that the CR region, as identified by superposition of near-TIRF imaging of activated myosin II staining (P-Myo, Figure 2B), appears as a ridge of increased relative actin staining intensity as well as relative Z height (Figures 2C, D).

We were able to capture a limited number of iPALM images of CRs in isolated cortices double-labeled for activated myosin II and either septin2 or anillin (three-four cortices per dual labeling from four separate experiments were imaged). This imaging confirmed the close association between activated myosin II and septin2 or anillin in the early clusters (Figures 3A–D, 4A–D), with septin2 and anillin generally located in the cluster center as we showed previously with SIM imaging. In the early-stage clusters (Figures 3A–D) and mid-stage CR patches (Figures 3E–H), Z axial imaging indicated that septin2 was consistently found near the expected plane of the plasma membrane, and between the membrane and staining for myosin II (Figures 3B–D, F–H). Within the mid stage patches the XY images of septin2 staining often appeared as an array of aligned filaments (Figures 3E–H) and image analysis of these patches support this identification in two ways. First, the widths of the septin2-stained filamentous structures are statistically similar between patches (Figure 3I, ANOVA, p = 0.16), and they also are similar in value to the widths of contractile stress fiber-associated septin filaments in the human U2OS cell line measured using SIM in the recent study by Martins et al. (2023). Second, the alignment of these filament-like structures is supported by the measurement of a statistically significant increase (t-test, p < 0.0001) in anisotropy index in patches versus non-patch organizations (Figures 3J, K). Figure 3J shows the output of this fibrillar analysis on a patch (lower box) and a non-patch septin2 staining organization in which the designated ROI appears as a yellow box and the program analysis appears as a red line within this box, the angle of which represents the orientation of the fibers whereas the length is proportional to the measured anisotropy index (Boudaoud et al., 2014).

iPALM imaging of an anillin and activated myosin II stained cortex shows that early clusters consist of close associations between these proteins in the XY plane (Figure 4), whereas in the Z-axis anillin, unlike septin, appears to be spread over a wider distribution within the CR structure (Figures 4B–D). In patch-like accumulations the Z axial imaging shows that anillin appears to be more restricted to a submembranous distribution (Figures 4E, F), similar to what is apparent in the septin2 images (Figures 3F–H). This suggests that anillin localization changes over the course of the structural evolution of the CR which may have relevance for the scaffolding role this protein plays in organizing the CR structure.

The organization of myosin II, anillin and septins into clusters in the sea urchin CR (Figure 1; Henson et al., 2017; Garno et al., 2021) is reminiscent of the CR precursor nodes in fission yeast (Pollard and O’Shaughnessy, 2019). To further explore how these clusters coalesce into a mature ring, we investigated the importance of the actin crosslinker α-actinin given that in fission yeast CR development, proper node condensation depends on the actin crosslinking proteins α-actinin and fimbrin (Laporte et al., 2012; Li et al., 2016). We were interested in testing the hypothesis that α-actinin only associates with the late-stage sea urchin embryo CR that is undergoing actomyosin-based contraction. Sea urchin egg α-actinin has a molecular mass of ∼95 kDa, resembles mammalian α-actinin in rotary shadow TEM, cross-links actin filaments into large bundles, and localizes to the cleavage furrow when fluorescent conjugates are microinjected into fertilized eggs (Mabuchi et al., 1985). Immunoblotting with a commercial rabbit polyclonal antibody specific for the conserved carboxy terminus region of human α-actinin showed the expected immunoreactive band at ∼100 kDa in lysates of cultured mammalian LLC-PK1 cells as well as a similar molecular weight band in S. purpuratus sea urchin eggs and first division embryos (Supplementary Figure S1A). Immunolabeling of porcine LLC-PK1 cells with the anti-human α-actinin gives the expected staining of cell margins, the CR in dividing cells, and actomyosin stress fibers (Supplementary Figures S1B–SE) – in which the staining is a characteristic pattern of periodic foci (Katoh et al., 1998). Staining of isolated coelomocytes from adult S. purpuratus sea urchins shows that the anti-human α-actinin also codistributes with actin and myosin II and periodically stains actomyosin stress fibers in the cytoskeleton of these somatic cells (Supplementary Figures S1E–SH).

In cortices isolated from first division sea urchin embryos, α-actinin does not stain the myosin II clusters, or patches present in early to mid-stage CRs (Figures 5A–F, arrows) but does associate with the more aligned myosin II linear arrays of later stage CRs (Figures 5A–F, arrowheads). In these later stage CRs, the concentration of α-actinin labeling is associated with a concentration of F-actin staining (Figures 5G–J) suggesting a role in the bundling of CR-associated actin filaments and/or serving as a mechanosensor for actomyosin tension. In select cortices the structure of the late stage, linearized CR is stretched during the isolation process which leads to a significant broadening of its structure (Figures 5K–P). In these cortices, the linear periodic pattern of α-actinin staining of the actomyosin cytoskeleton is particularly clear (Figures 5K–P) and reminiscent of the staining pattern seen in stress fibers of coelomocytes and LLC-PK1 cells (Supplementary Figures S1B–SH). We have shown previously in platinum replica TEM images that the architecture of the actomyosin cytoskeleton is similar in the stress fibers of LLC-PK1 cells and the late-stage CR of dividing sea urchin embryos (Henson et al., 2017).

The lack of association of α-actinin with the early CR clusters is also observed in whole embryos fixed and stained in suspension (Figures 6A–H). In these embryos, α-actinin only localizes to the mature CR in embryos undergoing furrowing (Figures 6E–H) and not in the myosin II clusters present in unconstricted, late anaphase embryos (Figures 6A–D). Settling embryos onto poly-L-lysine-coated coverslips can slow CR formation on the attached side of the embryo relative to the unattached side, and in these embryos, the α-actinin still only associates with the linearized late-stage CR structure and not with the CR clusters on the attached side (Figures 6I–L). In these same experiments, complete linearized CRs with associated myosin II and α-actinin ring-staining are present in dividing embryos in which the region of the developing CR is not attached to the coverslip (Figures 6M–O).

Tracking cytokinesis completion in first cell cycle S. purpuratus embryos clearly shows that treatment with 100 µM of the myosin II inhibitor PAB prevents cytokinesis, although embryos can exhibit anaphase B elongation along with flattening in the equatorial region, typically in a time frame that is 20–30 min later than cytokinesis in the control embryos (Figure 7). The uninhibited progression of karyokinesis means that the PAB treated embryos eventually become binucleate (Figure 7B), and that they continue with additional nuclear divisions that are delayed relative to the timing of cell divisions in control embryos. Further support of the specificity of PAB on the process of actomyosin-mediated cytokinesis is provided by additional experiments showing that PAB displays a dose-dependent impact on cytokinesis inhibition (Supplementary Figure S2). Embryos treated with 50 µM PAB develop significant furrowing relative to the 100 µM treatment, although this lower dose treatment still inhibits full cytokinesis in the majority of embryos, with many embryos becoming binucleate (Supplementary Figure S2).

As an additional functional test of the specificity of PAB as an inhibitor of myosin II contraction in sea urchin cells, we turned to the actomyosin-dependent process of centripetal flow in sea urchin coelomocytes. We have previously shown that discoidal coelomocytes settled on glass exhibit an extensive type of actin-based centripetal flow in which movement at the cell edge is driven by Arp2/3-dependent actin polymerization, whereas movement within the cell center is based on myosin II contraction as assayed by treatment with myosin II light chain kinase and general phosphatase inhibitors (Henson et al., 1999; Henson et al., 2003). Therefore, treatment of discoidal coelomocytes with PAB should arrest central but not peripheral centripetal flow and disrupt the central actomyosin structure. This is the precise result we obtained when treating these cells as demonstrated by live cell phase contrast microscopy and actin and myosin II fluorescent staining (Supplementary Figure S3).

In order to examine the impact of PAB inhibition of myosin II contraction on cluster aggregation and ring organization, cortices (Figure 8) and whole embryos (Figure 9) derived from PAB and DMSO treated embryos were probed for CR constituents. Cortices isolated from +PAB embryos 140–160 min post-fertilization showed early CR-like structures with broad bands of myosin II nodes which also contained septin2, anillin, and actin (Figure 8). The linearized arrays of CR proteins characteristic of late-stage dividing cortices were not observed in the PAB treated cortices (Figure 8). Similarly, in whole embryos stained in suspension, wide bands of myosin II and actin clusters were present in the often elongated embryos following PAB treatment in the 140–160 min post-fertilization time frame (Figures 9A–L). These PAB treated embryos underwent microtubule-mediated karyokinesis with embryos displaying an early CR-structure and a dense array of microtubules resembling that seen in control telophase embryos (Figures 9M–X). This is important to demonstrate given that one possible off target effect of PAB might be an impact on the cytokinesis-essential interaction of astral microtubules with the equatorial cortex (Su et al., 2014), however the staining suggests that this was not the case. Finally, as a test of the requirement for any actomyosin interaction in the formation of the early CR clusters, we treated developing embryos simultaneously with 100 µM PAB to inhibit myosin II and 100 nM Latrunculin A to eliminate actin filaments prior to first division. Staining of these embryos revealed that the myosin II clusters of the early CR still assembled and microtubule arrays still accomplished karyokinesis despite the absence of actin filaments and myosin II function (Supplementary Figure S4).

The combination of the results of the PAB treatment experiments indicate that inhibition of myosin II contraction does not interfere with the initial assembly of the early CR myosin II clusters, however it does prohibit the coalescence of these structures into a linearized array. In order to test whether the band of myosin II clusters present in PAB treated embryos was capable of transforming into a functional CR, we photoinactivated PAB in treated embryos 140–160 min post fertilization using high intensity white light from fiber optic illuminators (Supplementary Figure S5). Immediately upon photoinactivation embryos commenced cytokinesis-like constriction, including complete abscission (Supplementary Figure S5), suggesting that the nascent CR cluster structure in PAB treated embryos can serve as a template for a functional CR. Immunofluorescent staining shows that embryos undergoing cytokinesis following PAB photoinactivation have the expected concentration of myosin II and actin in the CR along with a microtubule array and decondensed nuclei that are characteristic of PAB treated binucleated embryos that have undergone karyokinesis (Supplementary Figure S5).