Fig. S1. disc1 expression in the developing zebrafish embryo. (A) Semi-quantitative RT-PCR analysis of disc1 expression in the developing zebrafish embryo and larva demonstrated maternal expression at 1 hpf but downregulation of the transcript by 12 hpf. disc1 expression then gradually increased through 120 hpf. Levels were compared with an actin internal control. (B-G) disc1 expression likely overlaps with foxd3 and sox10 at 10s. Dorsal views of in situ hybridizations with anterior to the left illustrated that disc1 is not expressed at 8s but is expressed in at least a subset of foxd3- and

Fig. S2. Pharyngeal arch CNC populations were reduced in Disc1 morphants. (A,B,D,E) dlx2a expression, marking the neural crest population in these structures, is reduced in the pharyngeal arches of Disc1 morphants (double-headed arrows). (G-L) Similarly, hand2 expression was dramatically reduced or lost in the pharyngeal arches at 26 hpf, 30 hpf, and 32 hpf in Disc1 morphants (arrow marks first arch; arrowhead marks posterior arches). (C,F) Pharyngeal arch (P1-P7) development was also reduced at 32 hpf in Tg(sox10:egfp) zebrafish. All views are dorsal with anterior to the left, except for C and F, which are lateral views.

Fig. S3. Disc1 morphant CNC cells were induced normally. (A-D) At 6s, expression of foxd3 and sox10 were normal in Disc1 morphants. (E-H) Similarly, lateral (E,F) and dorsal (G,H) views of wild-type and Disc1 morphant embryos demonstrated identical expression of foxd3 at 8s. Anterior is to the left in all images. Arrows mark foxd3 expression in CNC cells of wild-type and Disc1 morphant embryos.

Fig. S4. Expansion of foxd3 and sox10 expression was not due to increased CNC proliferation. (A-D) Phosphohistone H3 immunohistochemical staining was done in Disc1 morphant and control Tg(sox10:egfp) zebrafish at 8s (not shown) and 10s (A-D). The proportion of CNC cells undergoing cell division was determined in the area dorsal to the developing neural rod between the posterior eyes and the anterior ear. (E) There was no difference in mitotic index ratio of phosphohistone H3-positive CNC cells (yellow) to total CNC cells (green and yellow) between control and Disc1 morphant siblings. All views are dorsal with anterior to the left. (F) In addition, total CNC cell number was not altered in Disc1 morphants, indicating no change in the size of the CNC population in this region.

Fig. S5. Total CNC cell death was not changed in Disc1 morphants. (A) Disc1 morphant and control Tg(sox10:egfp) zebrafish were flat mounted and confocal z-stacks were taken through the CNC population from the posterior eye to just posterior to the developing ear. The ratio of TUNEL-positive CNC cells to total CNC cell number was determined from 14s to 22s. Although there are slight variations in CNC cell death at the various developmental stages, overall (B) there is no change in the proportion of CNC cells undergoing apoptosis in Disc1 morphants.

Fig. S6. crestin and snai1b were expressed normally in Disc1 morphants. (A,B) Dorsal views of wild-type and Disc1 morphant embryos subjected to in situ hybridization for crestin demonstrate normal expression of this marker at 10s. (C,D) Similarly, expression of snai1b, another CNC marker, was normal in Disc1 morphants at 10s. Anterior is to the left.

Fig. S7. Disc1 morphants had normal peripheral neuronal development but slightly delayed pigmentation. (A,B) HuC/D immunohistochemistry at 48 hpf in wild type and Disc1 morphants demonstrated identical staining, indicating that Disc1 knockdown does not affect peripheral neuronal development. Lateral views with anterior to the left. (C) There is no difference in HuC/D-positive cell number between wild-type and Disc1 morphant zebrafish in the trigeminal ganglion (gV), which is partially CNC derived. For comparison, the number of HuC/D-positive cells in the anterior lateral line ganglion (gAD), which is derived from placodal tissue, was also counted. In addition, the melanocyte population in the head (anterior to the midbrain-hindbrain boundary) was slightly developmentally delayed in Disc1 morphants but remained comparable to that of wild-type counterparts.

Movie 1. Medial migration of CNC cells in Tg(sox10:egfp) zebrafish. Control Tg(sox10:egfp) zebrafish were imaged every 8 minutes for 4 hours. A z-stack was taken through the entire depth of the CNC from the eye fields to the otic vesicle and projected into one plane to make a movie of CNC cell movement from 6s to the last time-point measured (12s). In the control embryos, CNC cells show a dramatic increase in medial CNC cell movement during the first hour of imaging. This medial movement then ceases and the CNC cells begin to migrate away from their site of origin, lateral to the midbrain. Depicted first is a DIC overlay of the embryo at the start of imaging with dorsal up and anterior to the left. The second DIC image illustrates the region of interest (ROI), which is the medial third of the area dorsal to the midbrain. This region moves anteriorly (arrow) as the embryo ages. Next, the movie shows the movement of the CNC cells across the dorsal midline. Finally, the track of a single CNC cell is shown (pink), as well as the point where migration initiated (start). The movement of the cell is then tracked across time (pink dot) and the path end shown (end). This movie illustrates the movement of the CNC cell into the midline and then back to the lateral edge of the neural rod.