Skip to main content
. 2020 Sep 23;9:e57779. doi: 10.7554/eLife.57779

Figure 6. Hmga1 activates Wnt signaling pathway in delaminating neural crest cells.

(A) Plasmid construct used as a readout for Wnt activity (after Ferrer-Vaquer et al., 2010). Six TCF/Lef-binding sites together with a minimal promoter regulate the expression of nuclear GFP in transfected cells in response to Wnt signaling. (B) Transverse section through the midbrain of a representative HH9+ embryo immunostained for Pax7, GFP, RFP, and DAPI. (C–C’’) Individual channels of image in B focused on the dorsal neural tube. In the absence of Hmga1, Wnt reporter activity was downregulated, resulting in fewer cells that expressed nuclear GFP (arrowheads) on the right side (C), even though cells were uniformly transfected on both experimental and control sides (C’), and the thickness of the neural tube remained unaffected (C’’). (D) The reduction in Wnt reporter output was quantified as a ratio of number of cells that expressed nuclear GFP, and the number of cells that were successfully transfected and therefore expressed nuclear RFP. The observed difference in GFP+/RFP+ ratio between the knockout and control sides was statistically significant (**p<0.01, student’s t-test). (E–F) In situ hybridization against Wnt1 in an Hmga1-knockout background. Transverse section through the midbrain (E) and hindbrain (F) shows reduced and unchanged levels of Wnt1 ligand in the dorsal neural tube on the experimental (arrow) versus control neural tubes, respectively. (E’–F’) The number of Pax7-positive cells in the midbrain appeared unchanged (E’), while a reduction was observed in the hindbrain (F’). (G–H) Transverse section through a representative embryo where Hmga1 was knocked out using CRISPR plasmids (G), and an embryo where Draxin was ectopically expressed (H) on the right side, immunostained for Laminin. Similar to Draxin overexpression, Hmga1 loss resulted in a failure of basement membrane remodeling due to reduced canonical Wnt signaling in neural crest cells, causing the laminin channel to remain blocked on the experimental side (arrows). (I–J) Expression of stabilized ß-catenin (NC1-∆90ß-cat) in delaminating cranial neural crest of Hmga1-knockout embryos was sufficient to rescue the migration defect. (K) Quantification of area covered by cranial neural crest cells on the experimental versus control sides. In the absence of Hmga1, cranial neural crest cells fail to migrate properly, a defect that can be separately rescued in Hmga1-knockout background by overexpression (OE) of stabilized ß-catenin in delaminating cranial neural crest, or exogenous expression of Hmga1 coding sequence ectopically (Rescue). nt, neural tube; nc, notochord; KO, knockout; OE, overexpression. See also Figure 6—figure supplement 1, Figure 6—source data 1, Figure 6—source data 2.

Figure 6—source data 1. Ratio of Tcf/Lef:GFP-positive and RFP-positive cell counts in transverse sections through the midbrain compared between Control and Hmga1-knockout sides in panel D.
Figure 6—source data 2. Ratio of neural crest cell migration area between experimental and control sides in whole mount embryos compared between Hmga1-knockout, ß-catenin-overexpression, and Hmga1-rescue conditions in panel K.

Figure 6.

Figure 6—figure supplement 1. Hmga1 regulates Wnt1 expression in premigratory cranial neural crest cells.

Figure 6—figure supplement 1.

(A) Genomic locus of Wnt1 obtained from the UCSC genome browser (Karolchik et al., 2003). ATAC-seq profile of premigratory cranial neural crest cells (from Williams et al., 2019) shows a putative Wnt1 enhancer upstream of the gene (highlighted in red). (B) Genomic sequence of the putative enhancer identified in (A) contains a predicted AT-rich Hmga1 binding motif (red box). (C) A representative Hmga1-knockout embryo immunostained for Pax7 shows a migration defect on the experimental side (right) compared to the control side (left). (D) The area covered by migratory cranial neural crest cells, marked by the position of the lateral-most leader cell, is reduced following Hmga1 knockout (green – control; 53.3 a.u.2, magenta – knockout; 40.5 a.u.2). An overlay of the two areas highlights this reduction. (E) Transverse section through the embryo shown in (C). The number of dorsal neural tube cells positive for canonical Wnt activity, as indicated by the nuclear GFP reporter expression (E), was significantly decreased following the loss of Hmga1 (shown in Figure 6C). However, the number of Pax7+ cells (E’) on either side of the dorsal neural tube were relatively similar, despite the observed migration defect in (C). No difference in the thickness of the neural tube (E’’) was observed (shown in Figure 6C’’).