Mouse but not zebrafish requires retinoic acid for control of neuromesodermal progenitors and body axis extension

Abstract

In mouse, retinoic acid (RA) is required for the early phase of body axis extension controlled by a population of neuromesodermal progenitors (NMPs) in the trunk called expanding-NMPs, but not for the later phase of body axis extension controlled by a population of NMPs in the tail called depleting-NMPs. Recent observations suggest that zebrafish utilize depleting-NMPs but not expanding-NMPs for body axis extension. In zebrafish, a role for RA in body axis extension was not supported by previous studies on aldh1a2 (raldh2) mutants lacking RA synthesis. Here, by treating zebrafish embryos with an RA synthesis inhibitor, we also found that body axis extension and somitogenesis was not perturbed, although loss of pectoral fin and cardiac edema were observed consistent with previous studies. The conclusion that zebrafish diverges from mouse in not requiring RA for body axis extension is consistent with zebrafish lacking early expanding-NMPs to generate the trunk. We suggest that RA control of body axis extension was added to higher vertebrates during evolution of expanding-NMPs.

Keywords: Body axis extension; NMPs; Neuromesodermal progenitors; Retinoic acid; Somitogenesis.

Fig. 1

Loss of RA synthesis does not perturb zebrafish body axis extension. (A) Embryos treated with DMSO (control), 4 μM DEAB, or 10 μM DEAB from 5–15 hpf, followed by analysis of myoD expression. Brackets mark an 8 somite region in each embryo that is unchanged in length along the anteroposterior axis following DEAB treatment used to inhibit RA synthesis. (B) Embryos treated as described above from 5–15 hpf followed by a return to normal growth medium until 32 hpf. Arrows point to the heart region showing cardiac edema in embryos treated with DEAB. (C) Shown are embryos treated with DMSO (control) or 4 μM DEAB from 9.5–15 hpf followed by a return to normal growth medium until 72 hpf when it can be observed that DEAB treatment results in failure of pectoral fin outgrowth (arrows). (D) Comparison of body axis length along an 8-somite region in embryos at 15 hpf across the indicated treatment conditions; data are expressed as mean ± SD. For all comparisons, p > 0.05 (not significant difference) using one-way ANOVA non-parametric test (DMSO control, n = 62; 4 μM DEAB n = 108; 10 μM DEAB n = 101; each are biological replicates). AU, arbitrary units.

Fig. 2

Model comparing mechanisms of body axis extension in zebrafish and mouse. Mouse and zebrafish body axis extension occurs differently in that mouse utilizes expanding-NMPs that extend the trunk body axis (spinal cord and somites), whereas the zebrafish trunk is formed by continued gastrulation convergence and extension movements after the head forms; both mouse and zebrafish utilize depleting-NMPs to generate the tail spinal cord and somites (Steventon and Martinez Arias, 2017). Our studies here show that zebrafish body axis extension does not require RA, whereas previous studies demonstrated that the mouse RA requirement for body axis extension is limited to the trunk somites (Cunningham et al., 2011), suggesting that RA controls expanding-NMPs but not depleting-NMPs.