Delta-Notch signaling and lateral inhibition in zebrafish spinal cord development

Abstract

Background: Vertebrate neural development requires precise coordination of cell proliferation and cell specification to guide orderly transition of mitotically active precursor cells into different types of post-mitotic neurons and glia. Lateral inhibition, mediated by the Delta-Notch signaling pathway, may provide a mechanism to regulate proliferation and specification in the vertebrate nervous system. We examined delta and notch gene expression in zebrafish embryos and tested the role of lateral inhibition in spinal cord patterning by ablating cells and genetically disrupting Delta-Notch signaling.

Results: Zebrafish embryos express multiple delta and notch genes throughout the developing nervous system. All or most proliferative precursors appeared to express notch genes whereas subsets of precursors and post-mitotic neurons expressed delta genes. When we ablated identified primary motor neurons soon after they were born, they were replaced, indicating that specified neurons laterally inhibit neighboring precursors. Mutation of a delta gene caused precursor cells of the trunk neural tube to cease dividing prematurely and develop as neurons. Additionally, mutant embryos had excess early specified neurons, with fates appropriate for their normal positions within the neural tube, and a concomitant deficit of late specified cells.

Conclusions: Our results are consistent with the idea that zebrafish Delta proteins, expressed by newly specified neurons, promote Notch activity in neighboring precursors. This signaling is required to maintain a proliferative precursor population and generate late-born neurons and glia. Thus, Delta-Notch signaling may diversify vertebrate neural cell fates by coordinating cell cycle control and cell specification.

Figure 1

notch and delta gene expression in zebrafish trunk neural tube. All images are of transverse sections through trunk neural tube. Images in each row are of the same section. First column: fluorescent detection of RNA hybridization using notch and delta probes. Second column: localization of post-mitotic neurons using anti-Hu antibody. Third column: composite images of RNA and Hu localization. Both Hu-positive and negative cells expressed n1a (A-C), dlA (J-L) and dlB (M-O) and dlD (Q-R) whereas predominantly Hu-negative cells appear to have expressed n1b (D-F) and n5 (G-I). Scale bar, 10 μm.

Figure 2

Proliferative neural cells express delta genes. Labeling to detect incorporation of BrdU during S phase of the cell cycle (brown staining) and RNA expression (blue staining). Images are of transverse sections through trunk neural tube. BrdU-labeled cells expressed dlA (A), dlD (B) and dlB (C). Scale bar, 10 μm.

Figure 3

Notch signaling is disrupted in dlA ^dx2 mutant embryos. Dorsal views, anterior to left, of neural plate stage embryos probed for expression of ngn1. (A) Wild-type embryo showing three longitudinal domains of ngn1 expression. (B) dlA^dx2 mutant embryo showing uniformly high level expression of ngn1 in each longitudinal domain. This mutant embryo was slightly older than the wild-type; consequently, the lateral longitudinal ngn1 domains are closer to the embryonic midline. Scale bar, 50 μm.

Figure 4

Neural precursor cells prematurely exit the cell cycle and develop as neurons in dlA^dx2 mutant embryos. Labeling to detect proliferative cells by BrdU incorporation (brown) and post-mitotic neurons by huC expression (blue) on transverse sections through trunk neural tube. Both proliferative cells and post-mitotic neurons are evident in a wild-type embryo (A) whereas a mutant embryo had many more neurons and fewer proliferative cells (B). Scale bar, 10 μm.

Figure 5

dlA^dx2 mutant embryos develop excess early born neurons and fewer late-born cell types. (A, B) Dorsal views, anterior to left, of neural plate stage embryos probed for isl1 expression to reveal presumptive primary motor neurons (pmn) in medial neural plate and presumptive Rohon Beard sensory neurons (RB) in lateral neural plate. Excess primary motor neurons and Rohon Beard neurons developed in dlA^dx2 mutant embryos. (C-L) Lateral views, anterior left, dorsal up. (C, D) Approximately 36 h embryos labeled with zn-8 antibody, which identifies secondary motor neuron cell bodies (bracket) and ventral nerve roots (arrows). The mutant embryo had fewer cell bodies and only two ventral nerve roots, which appear to be smaller than normal, were evident in this region of the embryo. (E, F)18 somite stage embryos probed to detect isl2 expression, which marks CaP and VaP motor neurons in ventral spinal cord (asterisks) and Rohon Beard neurons in dorsal neural tube. One or two cells per hemisegment expressed isl2 in ventral spinal cord of a wild-type embryo (E). Clusters of 3–5 cells per hemisegment expressed isl2 in ventral spinal cord of the mutant embryo (F). (G, H) Approximately 30 h embryos probed for lim1 RNA expression. The mutant embryo had many more cells that expressed lim1 relative to wild type. Note, also, the lack of neural-crest derived melanophores, which appear brown in wild-type, in the mutant embryos shown in H and J. (I, J) Approximately 30 h embryos probed for lim2 expression. More cells expressed lim2 in the mutant than in wild type. (K, L) Approximately 26 h embryos labeled with antibody specific to Pax2. The mutant embryo had more Pax2-expressing cells than the wild-type. (M, N) Transverse sections of approximately 36 h embryos labeled with zrf-1 antibody, which identifies radial glial fibers. The mutant embryo completely lacked zrf-1 labeling, except for occasional cells in the ventral neural tube, which may be remnants of floorplate (arrows). Note, again, the absence of black melanophores in the mutant embryo. Scale bar, A, B 60 μm, E-L 30 μm, M, N 40 μm.