Ca2+ channel-independent requirement for MAGUK family CACNB4 genes in initiation of zebrafish epiboly

Abstract

CACNB genes encode membrane-associated guanylate kinase (MAGUK) proteins once thought to function exclusively as auxiliary beta subunits in assembly and gating of voltage-gated Ca(2+) channels. Here, we report that zygotic deficiency of zebrafish beta4 protein blocks initiation of epiboly, the first morphogenetic movement of teleost embryos. Reduced beta4 function in the yolk syncytial layer (YSL) leads to abnormal division and dispersal of yolk syncytial nuclei, blastoderm retraction, and death, effects highly similar to microtubule disruption by nocodazole. Epiboly is restored by coinjection of human beta4 cRNA or, surprisingly, by mutant cRNA encoding beta4 subunits incapable of binding to Ca(2+) channel alpha1 subunits. This study defines a YSL-driven zygotic mechanism essential for epiboly initiation and reveals a Ca(2+) channel-independent beta4 protein function potentially involving the cytoskeleton.

Fig. 1.

Cloning and early expression of zebrafish β4.1 and β4.2 MAGUK proteins. (A) The shorter N-terminal variants (orange exons spliced to exon 3) resemble human β4a variants. Longer variants (blue exons spliced to exon 3) resemble human β4b variants. Alternative splicing of β4.1 in the HOOK domain leads to additional β4.1 variants that lack exon 6, thus introducing an in-frame, early termination codon (*). ×, Target sites of splice donor MOs. ^, Location of three residues mutated in the human β4-mut construct. Blue arrows indicate RT-PCR primers. (B–G) Whole-mount in situ hybridization using 3′ antisense probes to β4.1 and β4.2. The arrowhead marks the margin between yolk/YSL and the blastoderm. (B and C) β4.1 and (D and E) β4.2 embryonic expression at 30% epiboly in the blastoderm (bl). (F and G) β4.2 expression at the five-somite stage (11.7 hpf) in the nuclei (ysn) of the yolk syncytial layer (ysl). β4.1 expression is similar (data not shown). Tailbud (tb) indicates the embryo posterior. (G) Nine-fold enlargement of a portion of F. (H) RT-PCR-amplified β4.1 or β4.2 fragments from RNA extracted from control or MO-injected embryos at 50% epiboly, with EF1α loading control.

Fig. 2.

Amino acid alignment of zebrafish and human (HS) β4 subunit proteins. (A) β4a isoforms. (B) The N-terminal domain for the β4b isoform. The light and dark gray areas depict SH3 and GK domains. ▾, Exon boundaries. The plus signs (+) indicate the α binding pocket residues that contribute side-chain contacts to the AID-binding domain, and the carats (^) mark side chains with direct hydrogen bonds to the AID (–3). Human residues underlined in white (M204, L208, and L350) were mutated in the β4a-mut construct.

Fig. 3.

Morpholino-mediated reduction of β4.1 and β4.2 mRNA. Embryos were injected with rhodamine dextran dye only (A–C and J–L) or dye plus β4.1MO or β4.2MO (D–I and M–R). Arrows indicate cells detached from the embryo, and arrowheads indicate dying tissue. n, notochord.

Fig. 4.

Human cRNA injection rescue of β4 morphants. Embryos were injected with β4.1MO, β4.2MO, or fluorescent tracer alone. Some embryos also received 200 ng of human β4a, β4b, or β4a-mut mRNA. Note human cRNA does not contain the MO target sites.

Fig. 5.

Site-directed alanine mutagenesis of β4 subunit GK domain residues M203, L208, and L350 disables Ca²⁺ channel expression. (A) Agarose gel showing RNA (α1, α2δ, and β4 cRNAs) used for electrophysiological comparison of Ca²⁺ channel expression when either β4a or β4a-mut was included in the mixture. Note the 10-fold equivalent amounts of β4 cRNA used for injection. (B) Representative current traces overlaid to show the robust Ca²⁺ current recorded when wild-type β4a is included in the mixture (β4 WT, blue) compared with no current when the three principal residues in the GK domain of β4a responsible for α1:β subunit interaction (M203, L208, L350) are mutated to alanines (β4a mut, red). Currents were recorded 34–36 h after cRNA injection. The average current from five β4a WT oocytes was 1.6 ± 0.22 μA.

Fig. 6.

Phenotypes of the YSL nuclei in β4-deficient or microtubule-disrupted embryos. Brightfield (A–J) or Sytox Green fluorescent (A′–J′) images. (A–D and F–I) Embryos were injected at the one-cell stage with dye (A and F), β4.1MO (B and G), β4.2MO (C and H), and β4.1MO plus 200 pg of human β4 cRNA (D and I). (E and J) Embryos were injected a second time into the yolk at the 1,000-cell stage with Sytox Green, with some embryos also receiving nocodazole. (A″–J″) Nine-fold magnification of the corresponding portions of A′–J′. (B″, C″, and E″) The arrowheads indicate nuclei or nuclear fragments substantially smaller or larger than wild-type YSL nuclei (see A″). The brackets indicate clumps of multiple Sytox-positive bodies. (F′–J′) The dotted lines indicate the width of the YSL layer [averaging two to four rows of nuclei for F′ (68.5 μm ± 3.1 SE) and I′ (66.5 μm ± 3.4 SE), but 6–10 rows thick for G′ (155.9 μm ± 7.1 SE), H′ (150.9 μm ± 9.0 SE), and J′ (92 μm ± 11.1 SE); n ≈ 10 embryos each]. vp, vegetal pole. (Scale bars: A–J and A′–J′, 200 μm; A″–J″, 600 μm.)