February 15, 2015;
Snail2/Slug cooperates with Polycomb repressive complex 2 (PRC2) to regulate neural crest development.
Neural crest cells arise from the border of the neural plate and epidermal ectoderm
, migrate extensively and differentiate into diverse cell types during vertebrate embryogenesis. Although much has been learnt about growth factor signals and gene regulatory networks that regulate neural crest development, limited information is available on how epigenetic mechanisms control this process. In this study, we show that Polycomb repressive complex 2 (PRC2) cooperates with the transcription factor Snail2
/Slug to modulate neural crest development in Xenopus. The PRC2 core components Eed
are expressed in the neural crest cells and are required for neural crest marker expression. Knockdown of Ezh2
, the catalytic subunit of PRC2 for histone H3K27 methylation, results in defects in neural crest specification, migration and craniofacial cartilage
interacts directly with Snail2
, and Snail2
fails to expand the neural crest domains in the absence of Ezh2
. Chromatin immunoprecipitation analysis shows that Snail2
occupancy and histone H3K27 trimethylation levels at the promoter region of the Snail2
target E-cadherin. Our results indicate that Snail2
cooperates with EZH2
and PRC2 to control expression of the genes important for neural crest specification and migration during neural crest development.
[+] show captions
Fig. 1. Expression of the PRC2 core components during early Xenopus development. Eed, Ezh2 and Suz12 are expressed similarly in early Xenopus embryos. Their transcripts are detected in the pan-mesoderm in late blastulae and early gastrulae embryos, but are shifted to dorsal mesodermal and dorsal ectodermal cells at late gastrula stages. During neurulation, their RNAs are located in neural and neural crest cells. At tailbud and tadpole stages, these genes are expressed in neural, sensory organs and migrating neural crest cells.
Fig. 2. PRC2 regulates neural and neural crest development. MOs (20 ng) against Eed, Ezh2 or Suz12 were injected unilaterally into two-cell-stage embryos. The embryos were examined at neurula stages by ISH for neural and neural crest markers. Knockdown of either one of the PRC2 proteins reduced the expression of the late neural marker NRP1 and the neural crest genes Snail2 and Sox9, but did not drastically decrease the expression of the early neural marker Sox3 (open neural plate stages, first column from left; late neurula stages, second column from left) or the neural plate border genes AP2α and Msx1. The mesodermal marker MyoD was not inhibited either. Numbers indicate embryos showing the displayed patterns (no reduction for Sox3, AP2α, Msx1 and MyoD, and decreased expression for NRP1, Snail2 and Sox9) over the total number of embryos analyzed.
Fig. 3. EZH2 interacts directly with Snail2. RNAs (1-2 ng) encoding tagged and untagged PRC2 proteins and Flag-Snail2 were injected alone or together, and proteins were extracted at gastrula stages 10-11 for IP and western blot analyses. (A) Reciprocal co-IP assay demonstrated that HA-EZH2 interacts with Flag-Snail2 in Xenopus embryos. (B) HA-EED and Flag-Snail2 do not co-precipitate. However, in the presence of EZH2 and SUZ12, HA-EED was pulled down by anti-Flag antibody, indicating that Snail2 recruited the entire PRC2 complex. (C) Purified GST-Snail2, but not GST, pulled down commercially available purified EZH2 directly.
Fig. 4. Snail2 requires EZH2 to regulate neural crest markers. Snail2 RNA (0.15 ng) was injected alone or with 20 ng EZH2-MO unilaterally into two-cell-stage embryos. ISH performed at neurula stages 17-18 revealed that ectopic expression of Snail2 expanded domains of neural crest genes Snail2, Sox9, Sox10 and Twist. However, Snail2 failed to expand the domains of these genes in the presence of EZH2-MO, implying that neural crest expansion by Snail2 required EZH2 function. The numbers of embryos showing the displayed patterns over total embryos analyzed are shown.
Fig. 5. EZH2 and Snail2 regulate overlapping and distinct markers in Xenopus. (A) 10 pg Noggin and 25 pg Wnt8 RNAs, with or without 40 ng EZH2-MO or Snail2-MO, were injected into the animal regions of two-cell-stage embryos. Animal caps were dissected at blastula stage 9 and cultured until neurula stages 19-20. RNA was then extracted for RT-PCR analysis of marker expression. Co-expression of Noggin and Wnt8 induced a panel of neural crest markers in animal caps that were reduced by either EZH2 or Snail2 knockdown. Notably, however, depletion of EZH2 affected more markers (Pax3 and Zic1) than depletion of Snail2, indicating that EZH2 regulates a wider array of gene expression than Snail2. (B) 20 ng EZH2-MO or Snail2-MO was injected unilaterally into two-cell-stage embryos. The embryos were analyzed at neurula stages 16-17 by ISH for expression of Zic1 and Pax3. Knockdown of EZH2 reduced the posterior (arrow), but not the anterior (arrowhead), domain of Zic1 and the expression of Pax3; however, Snail2-MO was ineffective in reducing either Zic1 or Pax3. The numbers of embryos displaying the shown patterns over the total number of embryos analyzed are shown.
Fig. 6. EZH2 regulates neural crest migration. (A) Wild-type embryos were cultured in medium with DMSO or 50 µM EZH2 inhibitor GSK126 from stages 13-14 onward. The embryos were analyzed by ISH for Sox10 and Twist expression at tailbud stages 25-26. In DMSO-treated embryos, Sox10- and Twist-expressing cells migrated in distinct streams in the head. Treatment with GSK126 resulted in reduced migration of these cells. In addition, Sox10 and Twist expression in dorsal midline of the trunk (arrows) was eliminated by GSK126 treatment. (B) Neural crest tissues from fluorescein dextran-labeled donor embryos were transplanted into unlabeled control embryos at neurula stages 15-16. The embryos were cultured in medium with DMSO or 50 µM GSK126 until tailbud stages 26-28. Migration of the neural crest was examined by fluorescence microscopy. Whereas neural crest from DMSO-treated embryos migrated efficiently in the head, GSK126-treated embryos displayed defects in neural crest migration. (C) Neural crest explants were dissected from stage 15-17 embryos and cultured on fibronectin-coated dishes for about 16 h in the presence of DMSO or 8-10 µM GSK126. Migration of the neural crest was imaged between 2 and 16 h of culture. Whereas neural crest cells dissociated from the core explants and migrated efficiently in DMSO-containing medium, treatment with GSK126 reduced the migration of the neural crest on fibronectin. (D) Tracking of cell migration in time-lapse movies revealed that the velocity of neural crest cell migration was significantly reduced by treatment with GSK126. Four movies each for DMSO- and GSK126-treated samples were analyzed, with 60-170 cells tracked in each movie.
Fig. 7. EZH2 regulates formation of craniofacial cartilages. 20 ng EZH2-MO or Snail2-MO was injected unilaterally into two-cell-stage embryos. The embryos were cultured to the tadpole stages for Alcian Blue staining of the craniofacial cartilages. (A) Knockdown of either EZH2 or Snail2 led to malformation of craniofacial structures on the injected side (labeled by Red-Gal staining). (B) Alcian Blue staining revealed that the neural crest-derived craniofacial cartilages (Meckel's, ceratohyal and ceratobranchial cartilages) were defective on the side injected with EZH2-MO or Snail2-MO (right side of the panels).
Fig. 8. Snail2 regulates EZH2 occupancy and H3K27me3 levels at the promoter of the Snail2 direct target E-cadherin during neural crest EMT. (A) Knockdown of Snail2 or EZH2 did not appreciably affect E-cad expression at early neurula stages 13-14, but resulted in increased E-cad expression at late neurula stages 19-20, when the neural crest underwent EMT to emigrate from the neural tube. Student's t-test revealed statistically significant changes in E-cad expression in morphant versus control samples at stages 19-20 (P<0.05). (B) Genomic structure around the E-cad promoter and the regions assayed in the ChIP experiments are shown at the top. Snail2-MO (20 ng) was injected into each blastomere of two-cell-stage embryos. Control and injected embryos were harvested at late neurula stages 19-20 for ChIP assay of EZH2 binding and H3K27me3 association with the E-cad promoter. Anti-GFP antibody was used as negative control. Knockdown of Snail2 reduced EZH2 occupancy and decreased H3K27me3 levels at the E-cad promoter, implying that Snail2 recruited EZH2/PRC2 to the E-cad promoter to deposit the repressive chromatin marks during EMT.