Regulation of vertebrate embryogenesis by the exon junction complex core component Eif4a3

Abstract

The establishment and maintenance of cellular identity are ultimately dependent upon the accurate regulation of gene expression, the process by which genetic information is used to synthesize functional gene products. The post-transcriptional, pre-translational regulation of RNA constitutes RNA processing, which plays a prominent role in the modulation of gene expression in differentiated animal cells. The multi-protein Exon Junction Complex (EJC) serves as a critical signaling hub within the network that underlies many RNA processing events. Here, we identify a requirement for the EJC during early vertebrate embryogenesis. Knockdown of the EJC component Eukaryotic initiation factor 4a3 (Eif4a3) in embryos of the frog Xenopus laevis results in full-body paralysis, with defects in sensory neuron, pigment cell, and cardiac development; similar phenotypes are seen following knockdown of other "core" EJC protein constituents. Our studies point to an essential role for the EJC in the development of neural plate border derivatives.

Fig. 1

Activity of Eif4a3 translation-blocking morpholinos. Western blot analysis of exogenous (lanes 1-8) and endogenous (lanes 9 and 10) Eif4a3 protein, visualized using a mouse monoclonal anti-Eif4a3 antibody. Eif4a3MO inhibits translation of eif4a3 RNA containing the Eif4a3 morpholino binding site (5′UTR-4a3); translation of an RNA that lacks the Eif4a3MO binding site (4a3) is not inhibited. 21ng, 10.5ng, or 5.25ng Eif4a3 morpholino (MO) and/or 2ng 5′UTR-4a3 or 4a3 RNA were injected, as listed; higher morpholino doses did not lead to an appreciable increase in the degree of translational block (data not shown). Longer exposure times and 21ng morpholino (MO or 5 base pair mismatch, 5MM) were used for lanes 9 and 10. Whole-embryo lysates from stage 14 embryos were used for this study; β-Tubulin levels were used as a loading control.

Fig. 2

The Eif4a3 morphant phenotype. A: Loss of pigment-forming cells following Eif4a3 knockdown. 21ng of each morpholino (“4a3MO”: Eif4a3 morpholino; “5MM”: 5 base pair mismatch Eif4a3 morpholino) was injected, as listed, in this and subsequent figures. “Rescue”: Eif4a3MO+2ng eif4a3 RNA. Embryos were derived from albino eggs fertilized with sperm from wild-type males. B: Loss of touch response behavior in Eif4a3 morphants. Graph depicting percentage of touch-responsive embryos injected with listed doses of eif4a3 RNA and/or morpholino (4a3MO), as shown. C: Defects of heart formation in Eif4a3 morphants. Troponin T staining of cardiac tissue in stage 37/38 embryos injected with 4a3MO, 5MM, or 4a3MO+Eif4a3 RNA (rescue). Note lack of heart looping in 4a3MO-injected embryo. D: Whole-mount in situ hybridization with antisense probes against Xnkx-2.5 in stage 28 embryos injected with Eif4a3MO (4a3MO) or Eif4a3MM (5MM), as listed. No significant differences were seen in the expression of this marker at neurula or tailbud stages (data not shown).

Fig. 3

The Y14 and Magoh morphant phenotypes. Loss of pigment-forming cells following Y14 or Magoh knockdown. 2.6ng of Y14 morpholino (“Y14MO”) or 10ng of Magoh morpholino (“Magoh MO”) was injected, as listed. “Rescue”: MO+2ng corresponding RNA. Embryos were derived from albino eggs fertilized with sperm from wild-type males.

Fig. 4

Eif4a3 is required by neurula stages for movement. A: Glucocorticoid receptor-Eif4a3 fusion construct. B: Addition of dexamethasone (DEX) allows nuclear entry of GR-Eif4a3. C: Addition of DEX at or before stage 17 to embryos injected with Eif4a3MO+GR-Eif4a3 RNA rescues the embryonic touch response, scored at stage 32; diminished rescue or lack of rescue is seen with DEX addition at later stages. DEX addition prior to stage 8 showed similar rescue to that seen from stages 8-17 (data not shown).

Fig. 5

Somite and neuronal development in Eif4a3 morphants. A: 12/101 antibody staining of somitic mesoderm in embryos injected with Eif4a3MO (4a3MO) or a control (scrambled) morpholino (CMO). No differences in somite staining were observed between the two populations. B: Whole-mount in situ hybridization with antisense probes against elrC (left panels) or sox3 (right panels) in embryos injected with Eif4a3MO (4a3MO) or Eif4a3MM (5MM). Note slight expansion of anterior neural tube in Eif4a3 morphants. C: Left panels: 3A10 antibody staining of neurons in embryos injected with Eif4a3MO (4a3MO) or a control (scrambled) morpholino (CMO). Right panels: high magnification of trunk detail of embryos at left. Note disorganization of “Y”-shaped motor neuron bundles in Eif4a3 morphants.

Fig. 6

Eif4a3 morphants display late defects in sensory neuron and neural crest development. A: Islet-1 antibody staining of Rohon-Beard (top band of nuclei in top panel, arrows) and motor (bottom band of nuclei in top panel, arrowheads) neurons. Increased dorsoventral spread of Islet-1 positive Rohon-Beard cells, as well as a reduction in intensity of Islet-1 staining in motor neuron nuclei, was observed in Eif4a3 morphants; both defects were partially rescued by co-injection of eif4a3 RNA. B: Whole-mount in situ hybridization with an antisense probe against hox11L2 in embryos injected with Eif4a3MO (4a3MO) or a control (scrambled) morpholino (CMO). Note disorganization of hox11L2-positive Rohon-Beard cells (arrows) in the Eif4a3 morphant embryo. C, D, E, F: Whole-mount in situ hybridization with antisense probes against slug/snail2 (C), sox10 (D), Xtwist (E), or Xsix-1 (F) in embryos injected with Eif4a3MO (4a3MO) or Eif4a3MM (5MM) morpholinos, as listed. No consistent differences were seen in the expression of these markers at neurula or tailbud stages.

Fig. 7

Eif4a3 is required for expression of tyrosinase genes. Whole-mount in situ hybridization with antisense probes against tyrosinase (tyr) (top panels) or tyrosinase-related protein-2 (dct) (bottom panels) in embryos injected with Eif4a3MO (4a3MO), Eif4a3MM (5MM), or Eif4a3MO+2ng eif4a3 RNA (Rescue). With both probes, intensity of signal and number of positive cells were reduced in Eif4a3 morphants at tailbud stages; this effect was rescued by co-expression of eif4a3 RNA.

Fig. 8

Apoptotic cell death is increased in Eif4a3 morphants. A: TUNEL assay of embryos injected with Eif4a3MO (4a3MO), Eif4a3MM (5MM), or Eif4a3MO+2ng eif4a3 RNA (Rescue) at stage 32. Strong staining observed in head only. B: Immunohistochemistry using Caspase 3 antibodies in embryos injected with Eif4a3MO (4a3MO), Eif4a3MM (5MM), or Eif4a3MO+2ng eif4a3 RNA (Rescue); signal was observed in the eyes of Eif4a3 morphants. C: Phosphohistone H3 antibody staining (blue) of embryos injected on one side (arrowhead) with Eif4a3MO (4a3MO) and 500pg β-gal RNA as a lineage trace (red) at stage 32. Phosphohistone H3 signal was not significantly affected by Eif4a3 knockdown.