Fig. 1

Expression of pax2 and pax8 genes in zebrafish spinal cord. pax2a, pax2b and pax8 are co-expressed by many cells but pax5 is not expressed in zebrafish spinal cord. Lateral views of pax2a (A), pax2b (B), pax2a and pax2b (C and D; probes were mixed), pax8 (E and F), pax5 (H and L) and pax2a and pax2b (J; pax2a in green and pax2b in red) expression in 24 h wild-type (WT) trunks (A–C, E and H) and triple-knock-down (TKD) embryos (pax2a mutants injected with pax2b and pax8 MOs) (D, F and L). pax8 staining is weaker than pax2a and pax2b and hard to photograph. This means that several weak pax8-expressing cells are hard to see in (E). However, when pax8-expressing cells were counted, similar numbers were obtained to when pax2a- or pax2b-expressing cells were counted (see I). (G) shows a dorsal view of the midbrain–hindbrain boundary of a 24 h wild-type embryo (a region of the embryo where pax5 is expressed, to show that this in situ hybridisation probe works). (J) shows merged and single channel images of a 2 colour fluorescent double in situ with pax2a in green and pax2b in red. In all cases, rostral is left. In all cases except (G), dorsal is top. Scale bar = 25 μm (A–H and L); 18 μm (J). (I) Average number of cells expressing pax2a, pax2b, pax8 and pax2a, pax2b or pax8 (probes were mixed) at 24 h in 5 different WT embryos. As the number of cells expressing pax2a, pax2b, pax8 or (pax2a, pax2b or pax8) are not statistically different from each other, these genes must be co-expressed in spinal cord cells. (K) Average number of cells expressing pax2a or pax2b (probes were mixed) in 12 uninjected noi sibs and 12 TKD embryos. In this and all other figures, cells were counted in a 5-somite length of spinal cord adjacent to somites 6–10. Cell row numbers are counted ventral to dorsal (e.g. cells directly above the notochord are in row 1, see Supp. Data Fig. 1F). Rows not shown have a cell count of zero. None of the differences in (I) and (K) are statistically significant. Error bars denote standard deviation.