Cadherin-2 Is Required Cell Autonomously for Collective Migration of Facial Branchiomotor Neurons

Abstract

Collective migration depends on cell-cell interactions between neighbors that contribute to their overall directionality, yet the mechanisms that control the coordinated migration of neurons remains to be elucidated. During hindbrain development, facial branchiomotor neurons (FBMNs) undergo a stereotypic tangential caudal migration from their place of birth in rhombomere (r)4 to their final location in r6/7. FBMNs engage in collective cell migration that depends on neuron-to-neuron interactions to facilitate caudal directionality. Here, we demonstrate that Cadherin-2-mediated neuron-to-neuron adhesion is necessary for directional and collective migration of FBMNs. We generated stable transgenic zebrafish expressing dominant-negative Cadherin-2 (Cdh2ΔEC) driven by the islet1 promoter. Cell-autonomous inactivation of Cadherin-2 function led to non-directional migration of FBMNs and a defect in caudal tangential migration. Additionally, mosaic analysis revealed that Cdh2ΔEC-expressing FBMNs are not influenced to migrate caudally by neighboring wild-type FBMNs due to a defect in collective cell migration. Taken together, our data suggest that Cadherin-2 plays an essential cell-autonomous role in mediating the collective migration of FBMNs.

Fig 1. Generation of stable transgenic fish expressing dominant-negative Cadherin-2 in cranial branchiomotor neurons.

(A) Schematic representation of full length Cadherin-2 (N-cadherin) and extracellular domain-deleted Cadherin-2 (Cdh2ΔEC) fused with mCherry that functions as a dominant-negative protein. E1-E5 cadherin ectodomains; TD, transmembrane domain; CD, cytoplasmic domain. (B). Schematic representation of plasmid used to generate stable transgenic fish expressing cdh2ΔEC-mCherry driven by the zCrest1 enhancer element of the islet1 promoter (islet1) upstream of the minimal promoter (hsp70l). (C,D) Photographs of wild-type Tg(isl1:GFP) and Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos at 24hpf, showing normal morphology. (E-H) Lateral images at 24 hpf of the green and red channels showing GFP-expressing and Cdh2ΔEC-mCherry-expressing cranial branchiomotor neurons in Tg(isl1:GFP) fish and Tg(isl1:cdh2ΔEC-mCherry) transgenic fish. Arrowheads point to FBMNs. (I,J) Lateral images of Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos at 24hpf with and without heat shock. Arrowheads denote defects at midbrain-hindbrain region after heat shock. (K,L) Cross sections through hindbrain neuroepithelium of Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos before and after heat shock labeled with DAPI and Alexa-488-phalliodin. Dotted lines outline the neural tube. (M-S) Confocal micrographs of immunostained embryos showing low-magnification (M-O) and high magnification (P-S) dorsal views of wild-type Tg(isl1:GFP) (M,P) and Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos (N,O,Q-S) at 26 hpf. Embryos were labeled with α-GFP and α-mCherry showing that Cdh2ΔEC-mCherry is only expressed in cranial branchiomotor neurons and not the surrounding neuroepithelium. Arrow points to the abnormal position of FBMNs in Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos. Rhomobomeres (r2-r6) are indicated. White asterisk denotes r6-derived PLL efferent neurons, which differ from r4-derived FBMN populations.

Fig 2. Cadherin-2 is required cell autonomously for caudal migration of FBMNs.

(A-I) Whole-mount immunocytochemistry showing dorsal views of Tg(isl1:GFP) (A-C) and Tg(isl1:cdh2ΔEC-mCherry)vc25 transgenic embryos (D-I) at 38 hpf embryos. Embryos are labeled with α-GFP (green) (A,D,G) and α-mCherry (red) (B,E,H) antibodies. (A-C) Wild-type Tg(isl1:GFP) embryos with FBMNs fully migrated into r6. (D-I) Defective caudal migration of FBMNs in Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos carrying one copy of the transgene (hemizygous) or two copies (homozygous). (J) Histograms indicate the percent of FBMNS at 38 hpf that failed to migrate (r4), migrated partially (r5), or migrated fully (r6). Each histogram corresponds to the genetic condition in the image to its left and numbers indicate the number of FBMNs counted. White asterisk denotes r6-derived PLL efferent neurons, which differ from r4-derived FBMN populations.

Fig 3. Inactivation of Cadherin-2 affects directionality of FBMN migration.

(A,B) Live imaging of lateral views of the hindbrain showing the positioning of CBMNs and their peripheral axonal projections at 48 hpf. (A) In wild-type Tg(isl1:GFP) embryos, FBMNs (VII) migration is complete and their axons (asterisk) project into the second branchial arch. (B) In Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos, the facial axons (asterisk) and trigeminal (Va,Vp), and vagus (X) axons can be seen to project normally. However, the FBMNs (VII) remain in r4 and lie in an abnormal dorsal position. (C,D) Coronal sections of 48hpf hindbrain from Tg(isl1:GFP) embryos at the level of r6 and Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos at the level of r4. FBMNs (α-GFP, green) in control Tg(isl1:GFP) embryos occupy a ventral position within r6. Cdh2ΔEC-mCherry-expressing FBMNs (α-mCherry, red) are found ectopically in a dorsal portion within r4. Nuclei are labeled with DAPI (blue). (E,F) Tracings of migratory paths of FBMNs captured from time-lapse images between 20–24 hpf from Tg(isl1:GFP) and Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos. Each time-lapse lasted 35 minutes with one frame every 5 minutes. Each trace is oriented so that caudal is to the bottom and medial is to the right. Arrowheads indicate the starting point for each cell. (G) Plot of the migratory tracks from start to endpoint shows a highly directional caudal migration of wild-type FBMNs (green arrows) in comparison to the random paths taken by Cdh2ΔEC-mCherry-expressing FBMNs (red arrows). C, caudal; R, rostral; L, lateral; M, medial (H) Quantitation of average distance traveled along the rostral-caudal axis by FBMNs in Tg(isl1:GFP) and Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos during the time-lapse sequences. (I) Quantitation of average instantaneous speed of FBMN movements in Tg(isl1:GFP) and Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryos. (Mean values ± SD are shown; p < 0.05).

Fig 4. Cadherin-2 is required for collective migration of FBMNs.

(A-C) Confocal images showing dorsal views of Tg(isl1:GFP) embryos at 38 hpf injected with plasmids driving expression of mCherry alone or Cdh2ΔEC-mCherry mosaically in CBMNs. Embryos are labeled with α-GFP (green) and α-mCherry (red). Expression of mCherry alone has no effect on the caudal migration of FBMNs. (D-F) In contrast, FBMNs expressing Cdh2ΔEC-mCherry do not migrate caudally even though neighboring wild-type FBMNs that do not express the transgene migrate appropriately towards r6. (G) Quantitation of the percent of mCherry- or Cdh2ΔEC-mCherry-expressing FBMNs that failed to migrate (r4), migrated partially (r5), or migrated fully (r6). Each histogram corresponds to the condition in the image to its left and numbers indicate the number of FBMNs counted. White asterisk denotes PLL efferent neurons, which differ from r4-derived FBMN populations.

Fig 5. Expression of dominant-negative Cadherin-2 in trigeminal and vagus branchiomotor neurons leads to aberrant neuron positioning.

(A) Dorsal view of live Tg(isl1:GFP) embryo at 48hpf shows positioning of anterior and posterior clusters of trigeminal neurons (Va,Vp) found in r2 and r3, respectively. Note the lateral positioning of the Va cluster of trigeminal motor neurons at 48 hpf. (B) Dorsal view of live Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryo at 48 hpf shows that trigeminal neurons (Va; asterisk) remain in a medial location. Green is the GFP signal, whereas red is the Cdh2ΔEC-mCherry signal. (C) Dorsal view of live Tg(isl1:GFP) embryo at 48hpf shows correct positioning of vagus motor neurons in dorsolateral motor nucleus (dlX) and medial motor nucleus (mmX). (D) Dorsal view of live Tg(isl1:GFP)/Tg(isl1:cdh2ΔEC-mCherry)vc25 embryo shows that vagus neurons (X) do not migrate and coalesce into discrete dorsolateral nuclei. Scale Bars = 20 μm.