Balashova OA et al. (2017), Folate receptor 1 is necessary for neural plate...

XB-ART-53082

Development 2017 Apr 15;1448:1518-1530. doi: 10.1242/dev.137315.

Show Gene links Show Anatomy links

Folate receptor 1 is necessary for neural plate cell apical constriction during Xenopus neural tube formation.

Balashova OA , Visina O , Borodinsky LN .

???displayArticle.abstract???
Folate supplementation prevents up to 70% of neural tube defects (NTDs), which result from a failure of neural tube closure during embryogenesis. The elucidation of the mechanisms underlying folate action has been challenging. This study introduces Xenopus laevis as a model to determine the cellular and molecular mechanisms involved in folate action during neural tube formation. We show that knockdown of folate receptor 1 (Folr1; also known as FRα) impairs neural tube formation and leads to NTDs. Folr1 knockdown in neural plate cells only is necessary and sufficient to induce NTDs. Folr1-deficient neural plate cells fail to constrict, resulting in widening of the neural plate midline and defective neural tube closure. Pharmacological inhibition of folate action by methotrexate during neurulation induces NTDs by inhibiting folate interaction with its uptake systems. Our findings support a model in which the folate receptor interacts with cell adhesion molecules, thus regulating the apical cell membrane remodeling and cytoskeletal dynamics necessary for neural plate folding. Further studies in this organism could unveil novel cellular and molecular events mediated by folate and lead to new ways of preventing NTDs.

???displayArticle.pubmedLink??? 28255006
???displayArticle.pmcLink??? PMC5399658
???displayArticle.link??? Development
???displayArticle.grants??? [+]

Species referenced: Xenopus laevis
Genes referenced: cdh1 cdh3 ctnnb1 eea1 folr1 lat sox2 tubb
???displayArticle.antibodies??? Cdh1 Ab1 Cdh3 Ab1 Ctnnb1 Ab13 Sox2 Ab6 Tubb Ab1
???displayArticle.morpholinos??? folr1 MO1 folr1 MO2

???attribute.lit??? ???displayArticles.show???

	Fig. 1. Folate receptor 1 expression during Xenopus laevis neural tube formation. (A) Time course of Xenopus laevis neurulation. hpf, hours post-fertilization. (B) Western blot assays from homogenates of wild-type (WT) or Xenopus laevis folate receptor 1 (Folr1)-overexpressing embryos at the indicated developmental stages. (C) Folate is present during Xenopus neural tube formation and folate levels do not change upon Folr1 knockdown. Data show folate levels measured by ELISA assay in homogenates from stage 13-17 (14.5-19 hpf) WT, folate receptor 1 morpholino 1 (Folr1-MO1)- or standard control morpholino (CMO)-injected embryos. Mean±s.e.m.; n=8 independent measurements from n>40 embryos/group; average embryo weight: 2±0.4 mg. (D,E) Neural plate stage embryos were processed for Folr1 (green) and β-tubulin (D, red) or C-cadherin (E, red) immunostaining. Shown are representative transverse sections of immunostained samples at neural plate stages 15 (D) and 17 (E). (D) Arrow in top panel indicates Folr1 localization at the apical neural plate surface and arrowheads indicate localization in extracellular spaces. Bottom panels show higher magnifications of the boxed area in the panel above and illustrate Folr1 localization to extracellular spaces. L, notochord lumen; Not, notochord. (E) Arrow indicates that Folr1 localizes to the apical neural plate surface along with adherens junction protein C-cadherin. Bottom panels show higher magnifications of the boxed area in the panel above. Scale bars: 20 μm.
	Fig. 2. Efficacy and specificity of folate receptor 1 knockdown by targeted translation-blocking morpholino. (A-G) Two- (A,C,D) or sixteen- (E-G) cell-stage embryos were unilaterally (A,E-G) or bilaterally (C,D) microinjected with standard control morpholino (CMO), morpholino against Folr1 (Folr1-MO) and/or with 600 pg Folr1-MO-sensitive (wt-folr1 mRNA or sense-flag-folr1 mRNA) or -insensitive folr1 mRNA (flag-folr1 mRNA or resistant-folr1 mRNA). (A) Two non-overlapping Folr1 translation blocking morpholinos (10 pmol/blastomere) elicit similar defects in neural plate folding. Shown are representative examples of unilaterally injected CMO, Folr1-MO1 and Folr1-MO2 embryos. Red indicates morpholino-containing tissue. (B) Schematics of construct design for folr1 mRNA sensitive and resistant to Folr1-MO. (C,D) Specificity of Folr1-MO. Representative western blot assays from neural plate stage embryos bilaterally injected (10 pmol MO/blastomere) with the indicated constructs and probed with anti-flag (C) or anti-Folr1 (D). GAPDH and Na+/K+ ATPase α1 subunit were used as loading controls. (E,F) Representative examples of transverse sections from a neurulating embryo unilaterally microinjected at the 16-cell stage with 1.8 pmol Folr1-MO (E) or CMO (F) per blastomere along with GFP mRNA in the dorsal medial and lateral animal blastomeres, as indicated in drawing on the right, sectioned and processed for Folr1 (green), GFP (red) and β-tubulin (white) immunostaining. Red indicates Folr1-MO- (E) or CMO- (F) containing cells. Scale bars: 20 µm. (E) Folr1-MO downregulates expression of endogenous Folr1 in the neural plate apical surface. Brackets indicate same number of Folr1-MO-containing (red) and wild-type (white) apical neural plate cells. (F) Presence of CMO in medial neural plate cells does not affect Folr1 expression. (G) Graph shows area labeled by apical Folr1 immunostaining in three wild-type and morpholino-containing medial neural plate cells (indicated with brackets in E,F). Mean±s.e.m.; n=42 sections from ten Folr1-MO and CMO unilaterally injected embryos; ****P<0.0001; ns, not significant; paired and unpaired Student's t-tests.
	Fig. 3. Folate receptor 1 knockdown in medial neural plate induces neural tube defects. (A,B) Two-cell-stage embryos were bilaterally microinjected with standard control morpholino (CMO) or morpholino against Folr1 (Folr1-MO). Rescue experiments were performed by either microinjecting along with Folr1-MO, 250 pg folr1 mRNA resistant to Folr1-MO (resistant-folr1 mRNA) or by incubating Folr1-MO-injected embryos with 150 μM folinic acid (FA). (A) Representative examples of control and experimental sibling embryos during neural tube closure (20 hpf) from four out of six experiments in which the severe NTD phenotype prevails in Folr1-MO-injected embryos. Arrowheads indicate open neural tube. Graph shows incidence of severe and nonsevere defective neural tube. Mean±s.e.m.; (number of embryos): 20 pmol CMO (89), 20 pmol Folr1-MO (68), 20 pmol Folr1-MO+150 μM FA (54), Folr1-MO+folr1 mRNA (46); **P<0.0001; ns, not significant; two-tailed Mann–Whitney U-test (Wilcoxon rank-sum test). (B) Folr1-MO-induced moderate NTD phenotype results in open neural tube. Shown are representative examples of whole embryos at stage 21 and transverse sections of the neural tissue from Folr1-MO-injected embryos exhibiting a moderate NTD phenotype (arrowheads; two out of six experiments) and their siblings injected with CMO or Folr1-MO and FA. Embryos were then processed for β-tubulin (green) and E-cadherin (cadherin 1; red) immunostaining and nuclear labeling (blue, DAPI). Graph shows percentage of embryos with open neural tube; number of embryos: CMO (21), Folr1-MO (21), Folr1-MO+FA (20); P<0.0001, P<0.0005, P<0.01; Mann–Whitney two-tailed U-test (Wilcoxon rank-sum test). (C) Dorsal medial and lateral (medial+lateral neural plate, NP) or only dorsal medial (medial neural plate, Med NP) animal blastomeres from 16-cell-stage embryos were microinjected with 3 pmol/blastomere CMO or Folr1-MO along with Alexa 594-dextran conjugate (in red). Embryos were fixed and photomicrographed under a macroscope. Red indicates CMO- or Folr1-MO-containing tissue. Graph shows incidence of severe NTD phenotype (%). Number of embryos with severe neural tube defects out of total in each group was: 0/33 CMO-NP, 42/42 Folr1-MO-NP and 44/44 Folr1-MO-Med NP; ****P<0.0001; Mann–Whitney two-tailed U-test (Wilcoxon rank-sum test).
	Fig. 4. Folate receptor 1 in medial neural plate cells is necessary for changes in cell morphology required for neural plate folding. (A-D) Representative examples of neurulating embryos unilaterally microinjected at the 16-cell stage with 1.8 pmol Folr1-MO along with GFP mRNA in the dorsal medial (A-C) or dorsal lateral (D) animal blastomeres, as indicated in drawings on the right, sectioned and processed for Sox2, neural progenitor marker (blue), Folr1 (green), GFP (red) and β-tubulin (white) immunostaining. Red indicates Folr1-MO-containing cells. (A) Folr1 knockdown in medial neural plate (Med NP) cells impedes neural plate folding. Asterisks indicate same number of medial cells in wild-type (blue) and Folr1-deficient (pink) sides of the neural plate counted from the midline to represent the impact on the extent of the medial apical surface (white and red double arrows). Curves indicate the bending of the neural plate in wild-type (white) and Folr1-deficient (red) tissues. Arrows in A indicate cells shown in B (white, wild type) and C (red, Folr1 deficient). (B) Wild-type apically constricted (bracket) cell (second from the midline) from the superficial layer of the medial neural plate. Arrowhead indicates apically localized Folr1. (C) Cell (second from the midline) from the superficial layer of the Folr1-deficient side (red) of the medial neural plate fails to constrict (bracket). Arrowhead indicates lack of apically localized Folr1. (D) Folr1 knockdown in lateral neural plate (Lat NP) cells does not induce any apparent morphogenic phenotype. Scale bars: 20 µm.
	Fig. 5. Medial neural plate cells deficient in folate receptor 1 fail to constrict apically during neural plate folding. Two-cell-stage embryos were unilaterally microinjected with 10 pmol morpholino against Folr1 (Folr1-MO) or control morpholino (CMO) along with Alexa Fluor 594-dextran conjugate and bilaterally injected with membrane-GFP. Apical surface of superficial neural plate was time-lapse imaged from whole embryos at a rate of 0.2 min−1. Regions of interest were selected to contour cells close to the midline that remained visible during the 1-h recording and did not divide during this period. (A) Shown are representative examples for the indicated time points of Folr1-MO- and CMO-unilaterally injected and imaged embryos. Outlined is one wild-type (dashed) and one MO-containing (solid) cell for which the apical surface was measured over time. Numbers indicate apical cell surface for the same outlined cells at 0 and 60 min of recording. Scale bar: 10 μm. (B) Graph shows apical cell surface (as percentage of initial cell surface at 0 min, 100%) at the indicated time points. Mean±s.e.m.; n=30 Folr1-MO-, 28 CMO-containing cells, and 30 contralateral wild-type (WT) cells per group; ****P<0.0001; ns, not significant; two-way ANOVA.
	Fig. 6. Folate receptor interacts with C-cadherin and is necessary for its endocytosis in medial neural plate cells during neural plate folding. (A) Neural plate stage embryos were processed for Folr1 (green), C-cadherin (red) and β-catenin (blue) immunostaining. Shown is a representative transverse single z-section of immunostained stage 15 medial neural plate cell with orthogonal view through the indicated planes to demonstrate partial apical colocalization of Folr1, C-cadherin and β-catenin in the neural plate. Scale bar: 5 µm. (B) C-cadherin and β-catenin co-immunoprecipitate with Folr1. Lysates from wild-type neural plate stage embryos were incubated with anti-Folr1-crosslinked beads or naked beads. Co-immunoprecipitated proteins were dissociated from beads and run in SDS-PAGE gels for western blot assays with anti-Folr1, C-cadherin and β-catenin antibodies. Shown are representative western blot assays, which were performed five times with similar results. (C) The number of C-cadherin-containing vesicles including C-cadherin-containing early endosomes is reduced in Folr1-deficient medial neural plate cells. Two-cell-stage embryos were unilaterally microinjected with CMO and Folr1-MO in contralateral blastomeres. Shown is a representative example of transverse section from a neurulating embryo immunostained for C-cadherin (green) and Eea1 (red). Schematic indicates the cells from which the number of immunopositive vesicles were counted. Yellow arrow points to a C-cadherin+/Eea1+ vesicle and white arrow points to a C-cadherin+/ Eea1− vesicle. Scale bar: 10 µm. Graph shows mean±s.e.m. C-cadherin- and EEA1-labeled vesicles/cell measured in 54 CMO- and 54 Folr1-MO-containing medial neural plate cells in 18 sections from five embryos; **P<0.005; paired Student's t-test.
	Fig. 7. Pharmacological inhibition of folate uptake systems affects Xenopus laevis neural tube formation. Embryos at stage 13 (14.75 hpf) were incubated with 100 nM-1.5 mM methotrexate (MTX), 1 nM-300 μM methotrexate ester, 1 nM-100 μM pyrimethamine, 100-500 μM trimetrexate or vehicle (control, 0.3% DMSO or saline). Embryos were fixed and photographed under a macroscope when controls reached neural tube closure, at stage 20 (22 hpf). Shown are representative examples. Arrowheads indicate defective neural tube closure. Rescue of 1.5 mM MTX-induced NTD phenotype was performed by preincubating two- to four-cell-stage embryos with 10 mM folinic acid (FA). Graph shows mean±s.e.m. percentage defective embryos; n≥6 independent experiments with n≥10 embryos per group, per experiment; *P<0.01; two-way ANOVA.
	Fig. S1. Specificity of Xenopus folate receptor 1 antibody. (A,B) Two-cell-stage embryos were unilaterally injected with 700 pg xfolr1 mRNA+GFP mRNA (A) or Alexa Fluor 594-dextran conjugate (B), paraffin-sectioned and processed for xFolr1 immunostaining with (A) or without (B) Folr1 primary antibody added. Scale bar: 20 μm. (C) Schematic of construct to demonstrate specificity of xFolr1 antibody and representative Western blot from whole cell homogenates from embryos injected at the 2- cell stage with the indicated constructs. xFolr1: Xenopus laevis folate receptor 1, Folr1- MO: xFolr1-targeted translation blocking-morpholino.
	Fig. S2. Folr1-MO or methotrexate does not induce apoptosis of neural plate cells during neurulation. (A) Two-cell-stage embryos were unilaterally injected with 10 pmol folate receptor 1 morpholino (Folr1-MO) and Alexa 594-conjugated dextran and grown until neural plate stage 17 (18.75 hpf). (B) Early neural plate stage embryos (stage 13, 14.75 hpf) were incubated in the absence (Control) or presence of 1-1.5 mM methotrexate (MTX) until closure of the neural tube in controls (stage 20, 21.75 hpf). (A,B) Embryos were then processed for TUNEL assay, immunostaining for β-tubulin and nuclear labeling with DAPI. Shown are representative transverse sections of neural tissue from aforementioned embryos. Positive control is a sample treated with DNase; negative control is a sample in which no labeling enzyme (Terminal Deoxynuclotidyl Transferase) was added. Scale bars: 20 μm.
	Fig. S3. Knockdown of folate receptor 1 in mesoderm or non-neural ectoderm does not induce neural tube defects. Targeted blastomeres of 16-cell-stage embryos were unilaterally microinjected with 3 pmol/cell morpholino against the Xenopus laevis Folr1 (Folr1-MO) along with Alexa 594-dextran conjugate in indicated blastomeres (left). Neural plate stage embryos were then sectioned and processed for immunostaining with anti-β-tubulin (grayscale) and Sox2 (green). Red indicates MO-containing cells. Scale bars: 20 μm.
	Fig. S4. Folr1 knockdown does not induce changes in actin dynamics or in overall apicobasal polarity in the folding neural plate. (A) Two-cell-stage embryos were unilaterally microinjected with Folr1-MO along with Alexa Fluor 594-dextran conjugate (red) and bilaterally injected with mCherry-UtropinCH (F-actin reporter, green). Superficial medial neural plate was time-lapse imaged from whole embryos at a rate of 1- 2 min−1. Fluorescence intensity profiles (arrows) across cell-cell borders were measured among wild-type (WT, 1) and Folr1-MO (2) containing cells during 15 min imaging. Scale bar, 10 μm. Intensity profiles were fitted with Gauss function using R software after background substraction. Bar graph shows mean±SEM peak width at 50% maximum intensity (left) and standard deviation (SD) of normalized maximum intensity during 15 min recording (right), n: 28 cell-cell borders per group, t-test. (B) Unilateral Folr1-MO-containing neural plate stage embryos were sectioned and processed for immunostaining with anti-ZO-1 (green). Scale bar, 20 μm.
	Fig. S5. Neural plate cell apical constriction is concurrent with endocytosis of Ccadherin. Neural plate stage embryos were sectioned and processed for immunostaining with anti-C-cadherin (green) and anti-Eea1 (red). Scale bar, 20 μm. Arrow points to a Ccadherin- containing endosome.

References [+] :

Afman, Inhibition of transmethylation disturbs neurulation in chick embryos. 2005, Pubmed