MALS-3 regulates polarity and early neurogenesis in the developing cerebral cortex

Abstract

Apicobasal polarity plays an important role in regulating asymmetric cell divisions by neural progenitor cells (NPCs) in invertebrates, but the role of polarity in mammalian NPCs is poorly understood. Here, we characterize the function of the PDZ domain protein MALS-3 in the developing cerebral cortex. We find that MALS-3 is localized to the apical domain of NPCs. Mice lacking all three MALS genes fail to localize the polarity proteins PATJ and PALS1 apically in NPCs, whereas the formation and maintenance of adherens junctions appears normal. In the absence of MALS proteins, early NPCs progressed more slowly through the cell cycle, and their daughter cells were more likely to exit the cell cycle and differentiate into neurons. Interestingly, these effects were transient; NPCs recovered normal cell cycle properties during late neurogenesis. Experiments in which MALS-3 was targeted to the entire membrane resulted in a breakdown of apicobasal polarity, loss of adherens junctions, and a slowing of the cell cycle. Our results suggest that MALS-3 plays a role in maintaining apicobasal polarity and is required for normal neurogenesis in the developing cortex.

Figure 1. MALS-3 is expressed by and localizes to the apical surface of neural progenitors during neurogenesis

(A-C) Immunostaining of E14 coronal sections of rat telencephalon with antibodies against MALS-3 (green) and ZO-2, F-actin or β-catenin (red) reveal supra-apical localization of MALS-3 relative to ZO-2 (A), β-catenin (B) and F-actin (C). Nuclei are stained with SYTO-11 (blue). The en face view in panel C shows the lumen of the ventricle at the bottom, then (from bottom to top) passes through the apical-most ends of VZ cells and into the domain occupied by adherens junctions, which appear as circles of F-actin staining. (D-L) Immunostaining of coronal sections of rat telencephalon at different times during corticogenesis with antibodies against MALS (green) and ZO-2, F-actin or β-catenin (red). Nuclei are stained with SYTO-11 (blue). MALS-3 is initially localized diffusely in VZ cells (D-F), but appears to adopt an apical localization at the onset of neurogenesis (~E11.5) (G-I). This apical localization is maintained throughout the rest of embryonic life (J-L). Scale bar 5 μm.

Figure 2. Localization of MALS immunoreactivity compared with other proteins in the VZ

(A-C) Immunostaining of E14.5 coronal sections of rat telencephalon reveals that MALS is localized supra-apical to F-Actin. (D-F) CASK antibody did not reveal specific immunostaining in the VZ. (G-I) PALS1 is also localized supra-apical to F-Actin. in a pattern similar to that of MALS. (J-L) Crb3 localization is similar to that of PALS1 and MALS. (M-O) Mint shows a diffuse yet apically biased localization in the VZ relative. (P-R) Dlg, a protein localized basolaterally in many cell types, does not localize to the apical surface in NPCs. (S-U). F-Actin colocalized with staining for pan- cadherin in the apical domains of VZ cells. Scale bar 5 μm.

Figure 3. Distribution of proteins between membrane and cytosol in postnuclear supernatant from the embryonic telencephalon

(A) Triton soluble membranes (TX100), triton insoluble membranes (SDS) and cytosol (S100) from E14 rat telencephalon homogenate postnuclear supernatant (PNS) were prepared with either Magnesium ion (Mg⁺² group) or chelating agent (EDTA group), separated by SDS-PAGE, and visualized by Western blot. The presence of either Mg⁺² or EDTA resulted in no reproducible differences in molecular distributions between cytosol, Triton soluble membranes, and Triton insoluble membranes. (B, C) Integrated density plots of the distribution of MALS, CASK and Mint proteins in an iodixanol density gradient (B), with the corresponding Western blots of the different fractions (C). The distributions of MALS and CASK showed a strong similarity. (D, E) Integrated density plots of the distribution of known cell polarity proteins (D) together with corresponding western blots (E) reveal the distribution of several cell polarity proteins found in NPCs. Taken together, the data suggest that MALS might interact with CASK and PALS1 in NPCs based on their similar cellular localization. (F) Bars represent the cellular distribution of proteins known to associate with different subcellular compartments (based on data shown in Suppl. Fig. 4).

Figure 4. MALS is required for the maintenance but not establishment of apical localization of PALS1 and PATJ during corticogenesis

Coronal sections of control and MALS^TKO mutant brains at E13.5 immunostained for MALS (A,B), PALS1 (C,D), Crb (E,F) and PATJ (G,H) reveal loss of MALS (B) and apical PATJ (H) relative to control. Coronal sections of control and MALS^TKO mutant brains at E18.5 immunostained for aPKCζ (I, J), PALS1 (K,L), Crb (M,N) and PATJ (O,P) reveal loss of apical PALS1 (L) and PATJ (P) in MALS^TKO mutants. Scale bar 10 μm. (Q) No changes were observed in total protein levels for any of the above proteins in MALS^TKO mutant brains relative to control.

Figure 5. Adherens junction integrity is not compromised in MALS^TKO embryos

(A,B) Coronal sections of E18.5 embryos immunostained for ZO-1 to identify adherens junctions reveal normal staining in MALS^TKO mutants. (C-F) Antibodies against β-catenin (C,D) and pan-cadherin (E,F), both of which localize to the apical surface at the adherens junctions, reveal no change between MALS^TKO mice relative to controls. (G-L) Coronal sections of E17.5 embryos stained with antibodies against nestin (G,H) to identify neural progenitors and GFAP (K,L) to reveal glial cells reveal no change between MALS^TKO mice and controls. (I,J) However, antibodies against the neuronal marker Tuj1 reveals a broader Tuj1 staining in the cortex of MALS^TKO embryos, suggesting that MALS^TKO progenitors quit cycling prematurely and differentiate into neurons. Scale bars 10 μm.

Figure 6. MALS^TKO embryos display significant differences in labeling index and quit fraction during early neurogenesis

Coronal sections of E11.5 (A,B) and E12.5 (D,E) MALS^TKO mutants and matching littermate control brains injected with BrdU 2 hours (A,B,G,H) or 24 hours (D,E) prior to sacrifice. Sections were immunostained with antibodies to BrdU (A,C,D,E,G,H) and Ki67 (D,E). (C) At E11.5, MALS^TKO progenitors have a lower labeling index (total number of BrdU⁺ cells / total number of cells) relative to controls. (F) A larger number of MALS^TKO mutant progenitors at E12.5 have quit the cell cycle relative to control littermates (BrdU⁺Ki67⁻/ total number of BrdU⁺ cells). (I) MALS^TKO mutant progenitor cells recover normal cell cycle dynamics just a day later at E13.5, at which time no differences in labeling index are apparent. Scale bar 10 μm.

Figure 7. The mislocalization of MALS-3 disrupts polarity

Analyses of brains electroporated with FL_MALS-3/EGFP (A-C), Myr_MALS-3 (E,F,H,I,K,L,N-R) or Myr_Crb3 (D,G,J,M) at E13.5 and sacrificed at E15.5 (D-O) or E18.5 (P-R). Brains electroporated with FL-MALS-3 show no changes in the apical localization of Crb3 (A), PALS1 (B) or aPKCζ (C). Brains electroporated with Myr_Crb3 show no change in apical localization of ZO-1 (red, D), PATJ (red, G), or PALS1 (red, J) in GFP-positive cells (green). (M) Crb protein (red) is apparent in the basolateral regions of GFP-positive cells (green) following electroporation with Myr_Crb3, as expected. Myr_MALS-3 electroporated, GFP-positive cells (green) show loss of apical localization of ZO-1 (E,F), PATJ (H,I), and PALS1 (K,L). Crb staining (red) appears reduced in some GFP-positive cells (N,O). (P-Q) At E18.5, Myr_MALS-3 electroporated brains reveal misplaced cells in the lateral ventricles, breaks in the VZ, and MALS immunostaining is lost at those breaks. (P) Larger breaks and loss of PALS1 staining (red) are also observed. (Q) The ventricles are populated by delaminated cells that are Tuj1-positive (R). Scale bar 10 μm.