Central topography of cranial motor nuclei controlled by differential cadherin expression

Abstract

Neuronal nuclei are prominent, evolutionarily conserved features of vertebrate central nervous system (CNS) organization. Nuclei are clusters of soma of functionally related neurons and are located in highly stereotyped positions. Establishment of this CNS topography is critical to neural circuit assembly. However, little is known of either the cellular or molecular mechanisms that drive nucleus formation during development, a process termed nucleogenesis. Brainstem motor neurons, which contribute axons to distinct cranial nerves and whose functions are essential to vertebrate survival, are organized exclusively as nuclei. Cranial motor nuclei are composed of two main classes, termed branchiomotor/visceromotor and somatomotor. Each of these classes innervates evolutionarily distinct structures, for example, the branchial arches and eyes, respectively. Additionally, each class is generated by distinct progenitor cell populations and is defined by differential transcription factor expression; for example, Hb9 distinguishes somatomotor from branchiomotor neurons. We characterized the time course of cranial motornucleogenesis, finding that despite differences in cellular origin, segregation of branchiomotor and somatomotor nuclei occurs actively, passing through a phase of each being intermingled. We also found that differential expression of cadherin cell adhesion family members uniquely defines each motor nucleus. We show that cadherin expression is critical to nucleogenesis as its perturbation degrades nucleus topography predictably.

Figure 1

Development of Motor Nucleus Formation at Rhombomere 5 (A–D) Branchiomotor (Hb9⁻/Islet-1⁺) and Somatomotor (Hb9⁺Islet-1⁺) neurons in r5 at st20 (A), st26 (B), st29 (C), and st31 (D). Arrows show accessory abducens (AcAb), abducens (Ab) or facial motorneurons. Abducens neurons form initially a relatively undefined cluster (A and B) that becomes more coherent by st31 (D). Accessory abducens cells and facial neurons have distinct migration paths which converge and then segregate and cluster by st31. (E) Summary. VZ is ventricular zone, FM is facial motor neurons, and d and v are dorsal and ventral, respectively. See also Figure S1.

Figure 2

Cadherin and Catenin Expression in r5 at stage 35 (A) Islet-1 in situ hybridization shows the topography of nucleus positioning of abducens (Ab), accessory abducens (AcAb), dorsal facial nucleus (dFM), and ventral facial motor nucleus (vFM). (B–H) cad-6b (B), cad–20 (C), cad-8 (D), cad-13 (E), cad–22 (F), and cad-11 (G) expression on adjacent sections. A summary of differential cadherin expression in r5 motor nuclei is shown in (H). Note that no two nuclei share the same cadherin combination and that the accessory abducens and dorsal facial nuclei differ by the expression of cad-20 in the dorsal facial nucleus. (I) γ-catenin expression in all four motor nuclei at r5. See also Figure S2.

Figure 3

Developmental Time Course of cadherin-20 Expression at r5 (A–D) cadherin-20 in situ hybridization with islet-1 immunohistochemistry expression at st20 (A) and st30 (B–D). For clarity, higher-magnification images of facial (B), abducens (C), and accessory abducens (D) nuclei are shown. Arrows indicate Cadherin 20⁺ islet-1⁺ neurons as examples. (E) Quantification of the percentage of islet-1 motor neurons at r5 that express cadherin-20 at st20, st24, and st29. n = 4 embryos at each stage. Student’s t test p values shown above the bar graphs. Error bars indicate the SEM. (F) Summary of this expression. cadherin-20 is expressed in the majority of motor neurons at r5 at st20, and this expression is refined to the mature pattern by st30. VZ, ventricular zone. See also Figure S3.

Figure 4

Manipulation of Cadherin Gene Function or Expression Perturbs Cranial Motor Nucleus Topography at r5 (A–E) NΔ390-GFP expression disrupts nucleus clustering at r5 as assayed by Hb9 (B and C) and islet-1 (B and D) immunoreactivity. (A) shows a schematic of the experiment, (B) shows the internal control side of the brainstem, (C) and (D) show the results of NΔ390 expression, and (E) shows quantification of nucleus coalescence using a nucleus coalescence index (see the Supplemental Experimental Procedures). The brackets in (B) and (C) show the spatial extent of facial (FM) and abducens (Ab) nuclei in the control and experimental sides. (F–J) cad-20/GFP coexpression results in the mixing of accessory abducens (AcAb) and facial motor (FM) nuclei assayed by Hb9 (G and H) and islet-1 (G–I) immunoreactivity. (F) shows a schematic of the experiment, (G) shows internal control, and (H) and (I) show results of cad-20 expression, marked by GFP in (I). Quantification of the nucleus mixing (see Supplemental Experimental Procedures) is shown in (J). p values for a Student’s t test of each bin are shown above the graph bars. The chi-squared p value for the entire distribution is p < 0.0001, with two degrees of freedom. (K–O) DNcad-20/GFP coexpression results in a similar mixing of nuclei to cad-20 expression. (K) shows a schematic of the experiment, (L) shows the internal control, and (M) and (N) show the effect of DNcad-20 expression assessed by Hb9 (L and M) and islet-1 (L and N) expression. Electroporation is marked by GFP immunofluorescence in (N). Quantification of nucleus mixing is shown in (O). p values for a Student’s t test of each bin is shown above the graph bars. The chi-squared p value for the entire distribution is p < 0.0001, with two degrees of freedom. (P–T) Cadherin-6b (whose expression is found in both accessory abducens and facial motor nuclei) has no effect on AcAb and FMN segregation at st30 when misexpressed. (P) shows a summary of experiment, (Q) shows the control side of the brainstem, and (R) and (S) show the experimental side of the brainstem. Electroporation is marked by GFP in (S). Quantification of neuronal mixing index is shown in (T). (U–Y) N-cadherin (whose expression is found in neither accessory abducens and facial motor nuclei) has no effect on AcAb and FMN segregation at st30 when misexpressed. (U) shows a summary of experiment, (V) shows the control side of the brainstem, and (W) and (X) show the experimental side of the brainstem. Electroporation is marked by GFP in (X). Quantification of neuronal mixing index is shown in (Y). Error bars indicate the SEM. See also Figure S4.