Rho-kinase-dependent actin turnover and actomyosin disassembly are necessary for mouse spinal neural tube closure

ABSTRACT The cytoskeleton is widely considered essential for neurulation, yet the mouse spinal neural tube can close despite genetic and non-genetic disruption of the cytoskeleton. To investigate this apparent contradiction, we applied cytoskeletal inhibitors to mouse embryos in culture. Preventing actomyosin cross-linking, F-actin assembly or myosin II contractile activity did not disrupt spinal closure. In contrast, inhibiting Rho kinase (ROCK, for which there are two isoforms ROCK1 and ROCK2) or blocking F-actin disassembly prevented closure, with apical F-actin accumulation and adherens junction disturbance in the neuroepithelium. Cofilin-1-null embryos yielded a similar phenotype, supporting the hypothesis that there is a key role for actin turnover. Co-exposure to Blebbistatin rescued the neurulation defects caused by RhoA inhibition, whereas an inhibitor of myosin light chain kinase, ML-7, had no such effect. We conclude that regulation of RhoA, Rho kinase, LIM kinase and cofilin signalling is necessary for spinal neural tube closure through precise control of neuroepithelial actin turnover and actomyosin disassembly. In contrast, actomyosin assembly and myosin ATPase activity are not limiting for closure.

(C, D) Intensity profile scans of Phalloidin staining (C) and MHCB (D) along the basal-toapical axis of the neuroepithelium. Note the very low intensity of Phalloidin and MHCB in embryos treated with Blebbistatin + CytD or Blebbistatin + ML-7, consistent with severe reduction of neuroepithelial actomyosin by these inhibitors.
(E) PNP length (mm ± SEM) is significantly increased after treatment with Jasp for 18-20 h to the 15-19 or 20-24 somite stage. P-values are from t-tests compared with DMSO controls.
(G) Intensity profile scans of Phalloidin staining along the basal-to-apical axis of the neuroepithelium in embryos treated with Jasp or DMSO. The maximum intensity is normalised to 100%. Note the reproducibility of the basally extended Phalloidin staining in individual Jasp-treated embryos, compared with DMSO controls.
(H, I) Quantitation of cell proliferation by % phospho-histone H3 positive nuclei (H) and programmed cell death by % cleaved caspase 3 positive cells (I) in embryos cultured in Y27632 or Jasp for 18-20 h. There is no statistical difference between either treatment group and DMSO control embryos. There is no significant difference between ML-7 and DMSO-treated embryos at either 15-19 or 20-24 somites. P-values are from t-tests compared with DMSO controls.
(B) Embryos exposed to varying concentrations of ML-7 for 5-6 h undergo PNP closure at a similar rate as DMSO-treated controls (best fit linear regression lines plotted).
(F, H) Mean PNP length (mm ± SEM) after culture for 5-6 h (Short Term Culture, STC) or 15-18 h (Long Term Culture, LTC), shown graphically (F) and in table form (H, lower). Y27632 produces enlarged PNPs (pink symbols) but this effect is reversed if embryos are cultured in rat serum alone after removal ('washout') of Y27632 (STC Y27632 then RS).
However washout with rat serum after longer exposure to Y27632 does not restore PNP length (LTC Y27632 then RS) whereas subsequent culture in Blebbistatin rescues closure (LTC Y27632 then Blebb). P-values are from t-tests compared with DMSO controls.
(G, H) Mean PNP length (mm ± SEM) after culture in Y27632 followed by Blebbistatin, shown graphically (G) and in table form (H, upper). The effect of Y27632 in producing enlarged PNPs (pink symbols) is reversed when Blebbistatin is added to either short-term (STC Y27632 then Blebb) or long-term cultures (LTC Y27632 then Blebb). The same result is observed in embryos reaching three different somite stages.
(I) Biochemical fractionation shows similar a F/G ratio in embryos treated with a mixture of Y27632 and Blebbistatin after STC or LTC in Y27632 (as in G). There are no statistical differences in F/G ratio from DMSO control embryos.   (G) Immunohistochemistry (Phalloidin, red; anti-MHCB, green, middle; anti-β-catenin, green, right) reveals a less severely disrupted actomyosin distribution in Cofilin 1 mutant embryos exposed to Blebbistatin (G) compared with untreated Cofilin 1 mutants (compare with Figs 6C, S4B). Blebbistatin treatment is also able to restore normal β-catenin staining (compare with Fig. 7A).
b Concentrations in red are those used in the experimental studies.
c Overall morphology was scored with the following defects considered 'abnormal': non-smooth and round yolk sac, non-well formed branchial arches and maxillary, abnormal shaped somites. d The heart was bigger and the beating slower in Blebbistatin-treated embryos at all concentrations tested e Ssome embryos treated with Y27632 or jasplakinolide did not complete axial rotation Table S2.