Neural tube closure requires the endocytic receptor Lrp2 and its functional interaction with intracellular scaffolds

Abstract

Pathogenic mutations in the endocytic receptor LRP2 in humans are associated with severe neural tube closure defects (NTDs) such as anencephaly and spina bifida. Here, we have combined analysis of neural tube closure in mouse and in the African Clawed Frog Xenopus laevis to elucidate the etiology of Lrp2-related NTDs. Lrp2 loss of function impaired neuroepithelial morphogenesis, culminating in NTDs that impeded anterior neural plate folding and neural tube closure in both model organisms. Loss of Lrp2 severely affected apical constriction as well as proper localization of the core planar cell polarity (PCP) protein Vangl2, demonstrating a highly conserved role of the receptor in these processes, which are essential for neural tube formation. In addition, we identified a novel functional interaction of Lrp2 with the intracellular adaptor proteins Shroom3 and Gipc1 in the developing forebrain. Our data suggest that, during neurulation, motifs within the intracellular domain of Lrp2 function as a hub that orchestrates endocytic membrane removal for efficient apical constriction, as well as PCP component trafficking in a temporospatial manner.

Keywords: Gipc1; Lrp2; Mouse; Neural tube closure; Vangl2; Xenopus.

Fig. 1.

Lrp2 is expressed in the neuroepithelium and increased in constricting cells. (A-D) lrp2 mRNA (A) and protein (B-D) expression analyzed by in situ hybridization and immunofluorescence, respectively. Neurula stage (st.) embryos (frontal views, dorsal upwards). (A) Neural lrp2 expression. (B) Stage 15 forebrain region; Lrp2 is expressed in most cells (outlined by F-actin); single cells with high Lrp2 levels are located along the anterior rim of neural folds (NFs; arrowheads). (B′) Magnification of boxed region in B; increased Lrp2 levels are found in cells with small apical surface (circled in single channels B′₁ and B′₂). (C) LRP2 is detected throughout E8.5 anterior NFs, concentrated in areas undergoing apical constriction. Compare C′ (constricted cells in optic evagination) with C″ (dorsolateral cells with larger cell surface). ZO-1 marks cell boundaries. (D) STED imaging; LRP2 is condensed around neuroepithelial primary cilia (ARL13b⁺) at E9.5. Scale bars: 10 µm in B′,C′,C″; 1 µm in D.

Fig. 2.

Lrp2 is required for proper neuroepithelial morphogenesis and neural tube closure. Neural plate (NP) morphology. (A-F) Scanning electron micrographs of neural folds (NFs) from E8.5 wild-type (WT) and Lrp2^−/− mouse embryos at the 6-, 7- and 8-somite (som) stages (st.), frontal views. (A-C) Wild-type NFs are progressively elevated and optic evagination is initiated (arrowheads in B,C). (D-F) Narrower NFs have delayed elevation and there is impaired optic evagination in mutants (arrowheads in E,F). (G,H) Immunofluorescence staining detecting acetylated α-tubulin (ac. α-tub.) and DAPI-stained nuclei on coronal sections of 10-somite wild-type and Lrp2^−/− mouse NFs. Scale bars: 100 µm in A-F; 50 µm in G,H. (I,J) Dotted lines indicate normal (green) and abnormal (red) positioning of the border between neural and non-neural ectoderm. (I) Morpholino oligomer targeting lrp2.L (lrp2 MO) injected as indicated. Impaired hingepoint formation and NF convergence on the injected side are indicated (asterisk). (J) Closely apposed NFs in control at stage 19. Open anterior and posterior NFs, and a short anteroposterior axis are formed upon bilateral injection (asterisks); embryos were photographed at the same magnification. (J′,J″,J′₁,J″₁) Transverse sections and magnifications thereof; levels are as indicated in J. Green markers indicate normal floor plate width (J′) and apical cell surfaces (J′₁); red markers indicate wide floor plate (J″) and abnormally wide apical cell surfaces (J″₁). (K-M) In situ hybridization for sox3; normal NP width (green bar) is found in control (K); lrp2 MO-impaired NP narrowing (red bar, L) was partially rescued (yellow bar) by re-introduction of lrp2 (M). (N) Graphical representation of results from K-M (box plot, Wilcoxon rank sum test).

Fig. 3.

Lrp2 is cell-autonomously required for efficient apical constriction. Apical constriction (AC) in forebrain neural plate (NP) cells. (A,A₁) F-actin revealed a larger apical cell surface in lrp2 morpholino oligomer (MO)-injected cells (identified by lineage tracer fluorescence) compared with the uninjected side; there was a lack of AC in the optic evagination (ev.) area on the injected side. (B-E) Quantification of cell surface areas in unilaterally injected lineage tracer-only controls (B,B₁), lrp2 morphants (C,C₁) and morphants with re-introduced lrp2 (D,D₁). (E) Cell surface area ratios calculated between injected and uninjected sides; box plot and Wilcoxon rank sum test. (F) The intermingling of constricting NP cells with lrp2 morphant cells demonstrates cell autonomy of AC failure. (G,H) Frontal views of the forebrain area of wild-type (WT; G) and Lrp2^−/− (H) 7-somite stage mouse embryos at E8.5; ZO-1 delineates cell borders. (G′,H′) Magnification of optic evagination area, indicated in G,H; cell surface area was increased in mutants. (G″,H″) Color-coded maps (areas indicated in G,H) visualize cell surface area. (I) Graphical representation of results from G″,H″; four areas from each embryo were analyzed; box plot and Student's t-test. (J) Live imaging (Movie 2) using LifeAct indicates a failure of AC in morphant cells, despite actin dynamics. Cell surface area measurements revealed size fluctuations; final AC occured in control cells only. Scale bars: 25 µm in A-D,F; 20 µm in G,H.

Fig. 4.

No efficient remodeling of the apical surface in cells with apical constriction failure. (A-D) Scanning electron micrographs of mouse neuroepithelial cells at E8.5. Prominent filamentous, microvilli-like protrusions were present on wild-type (WT; A) and LRP2-deficient (Lrp2^−/−; B) cells at the 7-somite (som) stage. Reduced filamentous protrusions are seen at the 9-somite stage in the wild type (C), but are persistent in Lrp2^−/− (D). (E) Transverse section of Xenopus forebrain area at stage 19; lrp2 morpholino oligomer (MO)-positive cells are identified by their enlarged apical surface and by cytosolic lineage tracer fluorescence. Embryos were incubated in fluorescent dextran from stages 14 to 19. Signal is present in constricted cells (filled arrowheads) but absent from MO-targeted cells (empty arrowheads). F-actin staining indicates cell borders. (F,G) Transmission electron micrographs of E9.5 coronal ultrathin sections. Normal apical cell diameters (green lines) and moderate bulging are present in the wild type (F); increased cell diameter (red lines) and excessive bulging are found in Lrp2^−/− cells (G). (H) Quantification and statistical analysis of cell diameters. Student's t-test. Data are mean±s.e.m. (I) En face view of neural plate showing apically enlarged lrp2 MO-targeted cells (lineage tracer⁺) bulged outwards, see orthogonal view (I′; optical section at level indicated in I). Scale bars: 2 µm in A-D; 20 µm in E; 2 µm in F,G; 10 µm in I.

Fig. 5.

Lrp2 is required for planar cell polarity and regulates subcellular localization of Vangl2. Analysis of planar cell polarity in embryos en face (A-E) and on sections (F-H). (A) Scheme for injection of morpholino oligomer (MO) targeting lrp2; area for analysis at stage (st.) 16 in B is indicated by a box. (B) Dorsal view of mid-/hindbrain area; the width of the neural plate is greater on the injected side, F-actin delineates cell borders. (B₁) Bright-field image of same embryo shows pigment granule localization towards the anterior in uninjected control cells (B_1′); circumferential distribution is seen in apically wide morphant cells (B_1″). (C) Beginning of asymmetric Lrp2 distribution; F-actin delineates cell borders. (C′) Magnification of area indicated in C; there is medio-anterior distribution of Lrp2 (arrowheads in C′₁), which is also seen in apically constricted cells (circle in C′₁). (C′₂) F-actin single channel. (D,E) eYFP-vangl2 injected into the A1 lineage in 4- to 8-cell embryos, detected at stage 15 using immunofluorescence for GFP and F-actin staining. Dotted pattern of eYFP-Vangl2 expression (6/6 embryos; D,D′; magnification of inset in D), which is located subapically in vesicle-like structures (green arrowheads in D″; optical section indicated in D′; magenta arrowheads indicate f-Actin belt). Upon lrp2 MO injection, Vangl2 localized at cell borders (7/9 embryos, two independent experiments; E,E′) and subapically in the basolateral membrane (green arrowheads in E″, optical section as indicated in E′; magenta arrowheads indicate f-Actin belt). (D′₁,D′₂,E′₁,E′₂) Single channels. (F) Schematic illustrating the level of transverse sections in G,H. (G,H) LRP2 and VANGL2 distribution at the 9-somite stage. (G′,G′_1-3,H′,H′_1-3) Higher magnifications of the areas indicated by the boxes in G,H, and single channels thereof. Apical colocalization (dashed circles) of LRP2 and VANGL2 in RAB11-positive compartments occurs in wild type (WT; G,G′) compared with absence of VANGL2 from RAB11-positive compartments in Lrp2^−/− cells (H,H′); arrowheads in H′₃ indicate relocalization of VANGL2 to the basolateral membrane. l, left; p, posterior; r, right; v, ventral. Scale bars: 100 µm in B; 10 µm in C′,D′,E′; 5 µm in G′,H′.

Fig. 6.

lrp2 functionally interacts with shroom3 in mediating apical constriction. (A) Transverse section through neural plate; unilateral injection of shroom3 morpholino oligomer (MO) produces a failure of apical constriction (α-tubulin highlights wide cell surfaces; A₁); apical Lrp2 accumulation in constricting hinge point cells (arrowheads in A₂) and lack of apical Lrp2 recruitment in targeted cells. (B) Section through animal cap; ectopic apical constriction is induced in cells injected with shroom3-myc (detected by anti-Myc antibody; B₁) coinciding with apical Lrp2 accumulation (arrowheads in B₂). (C-E) shroom3-induced ectopic apical constriction in animal cap cells at stage (st.) 10.5 (C) is abrogated by co-injection of lrp2 MO (D) and restored by co-injection of lrp2. (F) Quantification and statistical analysis of experiments in C-E: no, mild or strong ectopic constriction are quantified. N=number of experiments; n=number of embryos, χ² test.

Fig. 7.

Lrp2 mediates apical constriction by functional interaction with Gipc1. (A) Frontal view of wild-type (WT) mouse forebrain area at E8.5 (7 somites). Immunofluorescence (IF) staining reveals localization of GIPC1 and LRP2; ZO-1 marks cell borders. (A₁-A₃) Single channels. (A_1′-A_3″) Magnification of areas indicated in A₁-A₃; differential GIPC1 intensities exist between large and constricted cells. (B,C) Frontal views of stage (st.) 16 (B) and stage 15 (C) embryos; dashed line indicates the anterior limit of the neural plate; immunofluorescence reveals the spatially dynamic localization of Gipc1 (B) and spatially dynamic colocalization of Lrp2 and Gipc1 (C). (B′) Single channel of Gipc1; higher magnification of the area indicated in B. Boxed areas indicated in C are shown at higher magnification in C′-C″″. (C′₁,C″₁,C′₂,C″₂) Single channels. Gipc1 is present in areas with low (C′₁,C′₂) or high (C″₁,C″₂) amounts of Lrp2. Dispersed distribution of Lrp2 and Gipc1 in a cell with a large apical surface (C″′), also visible in an orthogonal optical section (C‴_a; indicated in C‴); sites of Lrp2/Gipc1 co-localization are indicated (white arrowheads). (C″″) Cell with a constricted surface showing Gipc1 accumulation. (C⁗_a,C⁗_b) Colocalization (white arrowheads) or separate localization (green or magenta arrowheads) are indicated. Blue arrowheads indicate the level of circumferential actin; dashed line indicates apical surface in C″′_a,C⁗_a,C⁗_b. (D-F) Functional interaction of Lrp2 and Gipc1 demonstrated by individual (D,E) or combined (F) injection of low-dose lrp2/gipc1 morpholino oligomer (MO); targeted cells with lineage tracer (LT) fluorescence are shown. (G) Graphical representation of results from D-F; box plot and Wilcoxon rank sum test. (H) Frontal view of LRP2-deficient (Lrp2^−/−) mouse forebrain area at E8.5 (seven somites). Immunofluorescence reveals localization of Gipc1; ZO-1 outlines cells. (H₁,H₂) Single channels. There is homogenous GIPC1 signal throughout the neuroepithelium. (I) Mislocalization of Gipc1 in lrp2 MO-injected cells in stage 16 embryos; the magnified forebrain area is marked by a box in the schematic. Targeted cells are identified by LT; dashed line delineates targeted/non-targeted areas. (I₁) Without LT. (I_1′,I_1″) Higher magnification of single cells indicated by the boxes in I₁; Gipc1 increases in hindbrain and decreases in the forebrain area. (J-M) Embryos for CRISPR/Cas9 experiments selected at the one-cell stage and incubated until uninjected controls (J) reached stage 18/19. (K) Injection of Cas9 ribonucleic particles (CRNP) containing sgRNA1. (L,M) Co-injection of CRNP together with the lrp2 construct (L) or with lrp2 ΔPBD (M). (N) Graphical representation of results from J-M, χ² test. Cell borders in D-F,I are visualized using F-actin staining. Scale bars: 50 µm in A,C,H,I; 100 µm in B; 20 µm in D-F.

Fig. 8.

Hypothetical model of Lrp2 functional interactions in neural tube closure. Apical constriction is seen as a stepwise process with repetitive modules of actomyosin-mediated constriction and Lrp2-mediated membrane removal (left column). Apical constriction of Lrp2-positive neuroepithelial cells facilitates neural tube closure (middle column). Lrp2 mediates endocytic removal of apical membrane (1) as well as correct temporospatial localization of Vangl2 (2) via recycling endosomes. The Lrp2/Vangl2 interaction is likely facilitated by PDZ/PBD-mediated intracellular scaffolding via dimerized Gipc1, connecting to Myo6 and the actin cytoskeleton. Lack of Lrp2 (right column) entails defective neural tube closure due to impaired apical constriction. Removal of apical membrane fails and proper subcellular sorting of Vangl2 and Gipc1 is disturbed, leading to their mislocalization.