Apical polarity proteins localize or enrich in a stepwise fashion at the apical surfaces in rhombomeres 5 and 6 during neurulation

A. As revealed by immunohistochemistry, E-Cadherin, -catenin, and F-actin localized broadly to the entire cell membranes at 5-ss, and then enriched apically at 14-ss and 26-ss. At 26-ss, E Cadherin signals in the dorsal region of the neural tube and the otic vesicle were stronger than those in the mid-ventral regions. ZO-1 initially scattered as small punctate sites at 5-ss and later enriched apically (14-ss, 26-ss). Arrowheads point at the apical surfaces; arrows otic v. point at the otic vesicles. Note that it is unknown to what extent the E-Cadherin immunostaining signals represented P-Cadherin because the E-Cadherin antibody cross-reacts with P-cadherin (http://www.bdbiosciences.com/ds/pm/tds/610182.pdf). B. Simultaneous immunostaining of Factin bundles, aPKC, and Na+/K+ ATPase (with the a6F antibody) at 14-ss, 18-ss, and 26-ss. Note the lack of aPKC staining at 14-ss as well as Na+/K+ ATPase staining at 14-ss and at 18-ss. C. Apical enrichment of Na+/K+ ATPase α (with weak lateral membrane staining) at 26-ss was also confirmed with the a5 antibody. Note that all other Na+/K+ ATPase α images in this study were visualized with the a6F antibody. D. Simultaneous immunostaining of ZO-1, F-actin, and Crb1 at 14-ss, 18-ss, and 26-ss. Note the lack of Crb1 staining at 14-ss, except at the very ventral end (arrowhead). E. Simultaneous immunostaining of ZO-1 and Pard3, counterstained with nuclear dye DAPI (blue). Note the lack of Pard3 signals at 13-ss. Arrow C indicates a cell going through a C-division at 13-ss when Pard3 was not detectable; arrow P indicates a cell going through a P-division at 18-ss when Pard3 became detectable at the apical surfaces. F. Simultaneous immunostaining of F-actin bundles, ZO-1, and Crb2a. Note the lack of Crb2a at 14-ss. G. A TEM image at a lower magnification illustrates the apical surfaces (marked by blue dots) juxtaposed at the midline. The boxed area was magnified in Figure 1E.

The development of the PAAs and OAAs and their dynamic molecular compositions

A. GFP-assisted sagittal serial microscopy revealed that N-Cad and -catenin but not ZO-1 or actin localized to the opposing apical surfaces. The boxed areas are presented in Fig. 2A at higher magnifications. B. An RT-PCR revealed that the mRNA products of the N-Cadwt-GFP transgene reached its highest level 30 min after a heat shock and disappeared completely by 10 hours after the heat shock. C. The N-Cadwt-GFP fluorescence intensities of whole embryos reached the highest level 2 hours after the heat shock, and fell by 77% 10 hours after the heat shock. The fluorescence intensities were measured with Q-imaging under a stereo-fluorescence scope (ten embryos; means ± SEM). D. Single channel and merged images of Fig. 2B are presented to better illustrate the distributions of N-Cadwt-GFP to the PAAs and OAAs; N-Cadwt GFP was transiently-induced by heat shock from a transgenic construct injected into the embryos. Note that N-Cadwt-GFP enriched at both the PAAs and OAAs at 14-ss, but only at the PAAs at 26-ss. Weaker N-Cadwt-GFP signals persisted on the lateral membranes. The drawings summarize the distributional dynamics of ZO-1 and N-Cad-GFP. E. Nok, a cytoplasmic partner of Crb proteins, closely associated with the apical side of ZO-1 sites. F. Na+/K+ ATPase  colocalized with aPKC and Nok at 26 ss (arrows), although Nok and aPKC distributed more broadly than Na+/K+ ATPase  did. G. The table presents the subcellular distributions of immuno-EM gold particles in the early neural rods (at 14-ss) of N-Cad-GFPwt-expressing transgenic embryos and non-transgenic wildtype control embryos: The numbers of gold particles (as well as their corresponding percentages in total particles) distributed at the PAAs, OAAs, opposing apical surface, and cytoplasm are tabulated. Note the enrichment of gold particles at the OAAs and PAAs in N-Cad-GFPwt-expressing embryos but not in wildtype control embryos. H. An immuno-EM micrograph of the midline region of a wildtype neural rod revealed that the PAAs and OAAs survived the Triton X-100 extraction of the staining procedure and that no gold particles were observed at the PAAs and OAAs.

Loss of pioneer protein N-Cad and intermediate proteins Nok and Pard6b prevented neural tube lumen formation and affected the mirror symmetry of the neural tissue A. Dorsal views of the neural tissues of wildtype, N-Cadm117, nokm520, pard6γbfh266, and nokm520/pard6γbfh266 embryos during neurulation. At 26-ss and 34 hpf (hours post-fertilization), the lumen was apparent in wildtype (white arrowhead) but not in mutants. B. Bar graphs illustrate that the neural tissue of N-Cadm117 mutants was 170% to 200% wider than wildtype and other mutants (measured from 5-ss to 18-ss at the otic vesicle regions marked by the opposing black arrows in A; means ± SEM; N numbers, measured embryos). C. At 34 hpf, apical marker ZO-1 localized ectopically in nokm520 single (6 embryos) and nokm520/pard6γbfh266 double mutations (7 embryos), although the basal marker GFAP appeared to localize properly; in pard6γbfh266 (6 embryos), apical marker ZO-1 continued localizing to the midline region and demarcated the boundaries of miniature lumens (black arrow); finally, in N-Cadm117 (20 embryos), the mirror symmetry of the neural tissue remained disrupted at 34 hpf, and cells aggregated into rosettes (two rosettes are marked with dashed lines). The drawings summarize the distributions of the apical and basal markers at 34 hpf. D. Effects of inhibiting cell division with DNA synthesis inhibitors (from 1-ss to 18-ss) on the distribution of apical markers ZO-1 and N-Cad-GPF in wildtype and N-Cadm117 at 18-ss. Note that in wildtype, apical markers localized in a jaggy line at the midline region at 14-ss and then separated into two parallel lines at 18-ss; by contrast, in N-Cadm117, apical markers scattered throughout the tissue at both 14-ss and 18-ss, whether treated with the inhibitors or not. N numbers are numbers of embryos examined for each condition. Left panels are lower magnifications, right panels are higher magnifications.

ZO-1 and N-Cadwt-GFP translocate apically during neural keel-neural rod transition in Tg(HSP70:N-Cadwt-GFP)pt137

A. Distributions of N-Cadwt-GFP and ZO-1 in the neural rod at 18-ss were revealed by immunohistochemistry of the transverse tissue sections. Note the apical enrichment of both proteins (arrows). N-Cadwt-GFP was briefly induced by a heat shock at 10-ss. B. Transverse imaging at 5-ss, 8-ss, 11-ss, and 14-ss: ZO-1 first localized to small punctate sites and then enriched at the midline region, whereas N-Cadwt-GFP distributed broadly on the cell membranes and then enriched at the midline region. Diagrams summarize the dynamics of ZO-1 distributions. C. A higher magnification to better reveal that ZO-1 foci, which were often closely associated with N-Cadwt-GFP, became more concentrated in the apical half of the cells at 8-ss and localized to larger and more elongated spots than at earlier stages. The bottom diagram depicts the distribution of ZO-1 and N-Cadwt-GFP on the cell membrane. D. Individual-value bar graphs (means ± SEM) show that the cycloheximide (CHX) treatment (see the protocol in Fig. 4E) did not affect the nuclear ratios between the left and right halves of the neural tissue. Numbers of embryos analyzed: 6 DMSO-treated; 8 CHX-treated. P values by two-tailed t test.

Endocytosis is required for OAA dissolution

A-C. Endocytosis inhibitors CPZ and MCD blocked lumen formation and caused the neural tissue to twist to various degrees, as classified into Type I – IV categories. The cause of the phenotypic variations is unclear, but we speculate that embryos responded to the drugs differently due to variations in their genetic background, despite the fact that they were all TU fish. Histograms represent the frequency distribution of the four categories (from three experiments of a total of 113 embryos treated with PCZ, 82 embryos treated with MCD, and 50 control embryos treated with DMSO; means ± SEM). D. The average numbers of dividing nuclei per section (by counting M-phase nuclei revealed by DAPI staining) did not vary among embryos treated with DMSO (12 embryos), CPZ (13 embryos), and MβCD (20 embryos). E. The average total numbers of nuclei per section did not vary among embryos treated with DMSO (12 embryos), CPZ (13 embryos), and MβCD (20 embryos). F. The effects of endocytosis inhibitors CPZ and MβCD (from 11-ss to 18-ss) on the apical adhesions in Tg(HSP70:N-Cadwt-GFP)pt137 embryos at 18-ss. Note that the inhibition of endocytosis resulted in jaggy midline alignment of ZO-1 and N-Cad-GFP, which were sometimes disrupted by cellular bridges; however, in DMSO controls, the PAAs segregated into two parallel planes flanking the midline, where the OAAs dissolved. Insets, magnifications of the boxed regions in the top and middle panels; the boxed regions in the bottom panels were presented at higher magnifications in Figure 5A. G. The images show the tissue regions from which local apical areas were magnified and presented in Figure 5G, which illustrates the fate of N-Cadwt-GFP that was briefly induced at 18-ss. Again, note that N-Cadwt-GFP first concentrated in some intracellular regions (30 min), then localized to the lateral membrane (80 min), and finally enriched at the PAAs (arrows; 180 min); however, at no time did N-Cadwt-GFP localize to the OAA regions (arrowheads).

Intermediate proteins maintain PAA integrity and specificity

A. Single channel images of Figure 6A illustrate the distribution of ZO-1 and N-Cadwt-GFP in wildtype, nokm520, pard6γbfh266, and nokm520/pard6γbfh266 mutant embryos at 26-ss. B. Transverse confocal imaging of 21 siblings of nokm520 heterozygous parents revealed no difference in the apical distributions of ZO-1 and N-Cad-GFP at 14-ss. (Please note that unlike at 18-ss or later, nokm520mutants do not display apparent phenotypes at 14-ss for easy genotyping. Thus, we decided to infer whether nok mutation affects the OAAs and PAAs at 14-ss by examining over 20 embryos, of which about five embryos were expected to be nokm520 mutants.) C. Sagittal confocal imaging of 22 siblings of nokm520 heterozygous parents confirmed the punctate N-Cad- GFP-positive OAAs (arrowheads) in the midline plane (SM) at 14-ss in all embryos examined. D, E. A TEM image shows the distributions of cell-cell adhesions at the midline region. Cell-cell adhesions assembled cell protrusions in plywood-like clusters (double arrowheads). Two neuroepithelial cells, highlighted with blue and magenta, opposed at their apical ends, which are indicated by the apical marker basal bodies (arrows BB). The lower boxed region in B is magnified in Figure 6D; the upper boxed region is enlarged in C to better illustrate the plywoodlike clustering of cell protrusions that were adhered together by cell-cell adhesions (double arrowheads).

Terminal protein Na+/K+ ATPase is required to inflate the neural tube lumen, and the gut development also follows a rod-tube transition

A. Western blotting revealed that Na+/K+ ATPase α1 was expressed at 26-ss in nokm520, N-Cad, and pard6γbfh266 mutants as in wildtype. B. Na+/K+ ATPase α co-localized with Nok at the apical surfaces in wildtype at 26-ss. C-D. Na+/K+ ATPase α localized apically in pard6γbfh266 mutants at 28 hpf but not yet at 26-ss. E. In N-Cad mutants, Na+/K+ ATPase α localized to the central miniature lumen of neural cellular rosettes at 26-ss. F. Ouabain treatment did not prevent Na+/K+ ATPase α from localizing apically but did prevent the lumen from inflating. G. Dorsal view of the brain ventricles (arrow) in 34-hpf wildtype but not in ouabain-treated embryos. H. Ouabain treatment from 1-ss to 34 hpf: Crb1 and actin localized normally at the apical ends of the neural tubes (transverse sections) in wildtype (9 embryos) and in ouabain-treated embryos (30 embryos); however, ouabain-induced three defects in the lumen: small lumen (6 out of 30), stuffed lumen (stuffed with cells, inset; 4 out of 30), and no lumen (20 out of 30). I. At 36 hpf, the gut epithelial cells organized into a rod. The red-dashed line demarcates the outer boundary of the gut rod. J. A boxed region in panel I is enlarged to better show the hook-like interlocking apical membrane protrusions of gut rod cells (red and blue asterisks). These hook-like apical protrusions are presumably the microvilli, and they were adhered together with adhesions (arrowheads) to join opposing cells together. In addition, the parallel apical adhesion complexes (arrows) adhered together neighboring cells of the same orientations. K. At 56 hpf, the gut tissue developed into a hollow tube. Microvilli were no longer interlocked with each other via opposing adhesions, and a lumen emerged. However, the parallel apical adhesion persisted (inset, arrow).