Coordinated activation of the secretory pathway during notochord formation in the Xenopus embryo

Abstract

We compared the transcriptome in the developing notochord of Xenopus laevis embryos with that of other embryonic regions. A coordinated and intense activation of a large set of secretory pathway genes was observed in the notochord, but not in notochord precursors in the axial mesoderm at early gastrula stage. The genes encoding Xbp1 and Creb3l2 were also activated in the notochord. These two transcription factors are implicated in the activation of secretory pathway genes during the unfolded protein response, where cells react to the stress of a build-up of unfolded proteins in their endoplasmic reticulum. Xbp1 and Creb3l2 are differentially expressed but not differentially activated in the notochord. Reduction of expression of Xbp1 or Creb3l2 by injection of antisense morpholinos led to strong deficits in notochord but not somitic muscle development. In addition, the expression of some, but not all, genes encoding secretory proteins was inhibited by injection of xbp1 morpholinos. Furthermore, expression of activated forms of Xbp1 or Creb3l2 in animal explants could activate a similar subset of secretory pathway genes. We conclude that coordinated activation of a battery of secretory pathway genes mediated by Xbp1 and Creb/ATF factors is a characteristic and necessary feature of notochord formation.

Fig. 1.

High representation of secretory pathway genes in the notochord transcriptome. (A) The results of DNA microarray analysis on dissected Xenopus embryos at stage 12 (Tanegashima et al., 2008) (see also Materials and methods) were subjected to analysis of gene ontology (GO) terms. In the notochord sample, the 12 most over-represented GO terms refer to secretory pathway genes, but no such excess is seen in the somite/presomitic mesoderm or in anterior mesoderm (prechordal plate). The column labeled `count' gives the number of genes in the notochord sample that are members of the class of the GO term listed. `PValue' gives the probability of finding the observed GO term representation by chance. (B) Relative abundance of secretory pathway transcripts in dissected regions of stage 10 and stage 12 embryos, based on microarray analyses of dissected embryos (see Materials and methods) (Tanegashima et al., 2008; Zhao et al., 2008). Am, anterior mesoderm; Pm, posterior mesoderm including the notochord; noto, notochord; Ps, presomitic mesoderm; Ani, animal quadrant; DMZ, dorsal marginal zone; VMZ, ventral marginal zone; Vg, vegetal quadrant. Red represents high, and green low relative abundance. Gene names and processes in which these genes act are indicated.

Fig. 2.

In situ hybridization of selected secretory pathway genes to Xenopus embryos. (A-K″) Embryos at stage 12.5 (A-K), stages 15-17 (A′-K′) and stages 28-32 (A″-K″). All of these genes show preferential expression in the notochord, and some also show expression in the pronephros (red triangle).

Fig. 3.

XPB1 and Creb3l2 are involved in notochord formation. (A,B) In situ hybridization at stage 17 shows the preferential notochord expression of xbp1 (A) and creb3l2 (B). (C-J) Embryos were injected with control MO (C,E,G,I), a mixture of xbp1(n) and xbp1(s) MOs (D,H), or a mixture of creb3l2 A and creb3l2 B MOs (F,J). Embryos were stained with the notochord-specific antibody MZ15 (C-F), or the muscle-specific antibody 12-101 (G-J).

Fig. 4.

Response of secretory pathway genes to manipulations of the levels of Xbp1 or Creb3l2. (A,B) Reduction of expression of some secretory pathway genes by inhibiting Xbp1 and Creb3l2 levels. MOs against xbp1, creb3l2, or both, as indicated by color coding, were injected into the equatorial region of two-cell-stage embryos, the embryos were harvested at stage 12 (A) or stage 18 (B), and RNA was extracted. The levels of various transcripts were assayed by PCR. The graphs are the direct output of the Q-PCR program, after color coding, of ratios in experimental samples compared with control-MO-injected samples. The data represent three repeat assays in one of two similar experiments. (C) Induction of secretory pathway genes in animal caps. RNAs were injected, and transcripts in animal caps were assayed as above, and the data represented as fold induction on a linear scale. The level of each transcript in control animal caps was set as one (dashed line). Triplicate assays of one of at least two experiments with similar results are shown.