Eno et al., 2016 - aura/mid1ip1L regulates the cytoskeleton at the zebrafish egg-to-embryo transition. Development (Cambridge, England)   143(9):1585-99 Full text @ Development

Fig. 1

Developmental timecourse of zebrafish aura mutant embryos. (A,B) Embryo clutches from wild-type (A) and aura mutant (B) mothers at 10mpf showing egg lysis in mutants. (C-G) Wild-type (C,G) and aura mutant (D-F) timecourse. aura embryos do not display normal yolk coalescence (red arrow), instead resembling inactive wild-type mature oocytes (G). At the 8-cell stage, a septum is apparent in wild type (black arrow), whereas aura mutants lack this septum (blue arrow). aura mutants subsequently exhibit rounded, non-adhesive cells (purple arrow). In the cleavage stages, aura mutant embryos are partially or fully syncytial (orange arrows). Other phenotypes include discontinuities between the egg plasma membrane and the yolk along any animal-vegetal position (E, dark red arrow). Syncytia in aura mutants undergo an epiboly-like movement (F, green arrows indicate migrating edge).

Fig. 2

aura mutant embryos exhibit reduced accumulation of pericleavage and adhesive junction F-actin. (A,A′) In wild type (25/25), F-actin becomes recruited to the contractile ring (arrows) and pericleavage regions (arrowheads). (B,B′) In aura mutants, the contractile ring pericleavage F-actin is reduced (30/30). (C,C′) At the 4-cell stage, the furrow corresponding to the first cell cycle has formed adhesion junctions containing F-actin (white arrow). (D,D′) At the same stage, aura embryos do not exhibit adhesion junction F-actin cables (white arrow). (E-F′) Defects continue to be observed in subsequent furrows. White and blue arrows indicate furrows for the first and second cell cycles, respectively. (A′-F′) Higher magnification views of boxed regions in A-F. Scale bars: 20µm in A,B; 100µm in C-F; 10µm in A′-F′.

Fig. 3

aura mutant embryos do not recruit cell adhesion components to furrows and exhibit aberrant furrow microtubule dynamics. (A-E) In wild-type (12/12), the FMA forms as tubules become arranged in a parallel fashion (A-B′). As the furrow matures, FMA tubules become tilted and accumulate at the furrow distal end (E,E′). During furrow maturation, cell adhesion components such as β-Catenin accumulate at the furrow (C-E). (F-J) In aura mutants, FMA tubules form (F-H, brackets in H,H′) but remain in a parallel array arrangement throughout the length of the furrow (J,J′, brackets), failing to aggregate at the furrow distal ends (assayed at 49-63mpf; 15/15). All aura mutant embryos examined also fail to recruit β-Catenin to the furrow (H-J, arrows). Instead, β-Catenin localizes to ectopic CGs (F′-J′, asterisks). (A′-J′) Higher magnification views of A-J. Scale bars: 100µm in A-J; 10µm in A′-J′.

Fig. 4

aura mutants retain CGs. (A,B) During egg activation (2mpf), CGs appear embedded in an F-actin network in both wild type (A) and aura mutants (B). (C,D) By 10mpf, wild-type eggs have extruded nearly all CGs (C), whereas a large fraction of CGs are retained in aura mutants (D) (wild type, 0 mpf n=152, 10 mpf n=3; mutant, 0 mpf n=245, 10 mpf n=130; chi-square, P<0.0001). By 10mpf, the cortical F-actin cytoskeleton appears as a network of short F-actin fibers in aura mutants (D′′), in contrast to being largely disassembled as in wild type (C′′). (A′-D′) Higher magnification views of A-D; the boxed regions are further magnified in A′′-D′′. Scale bars: 100µm in A-D; 10µm in A′-D′,A′′-D′′.

Fig. 5

Wound healing defect in aura mutant embryos. (A,B,D,E) Embryos from wild-type (A,B) and aura mutant (D,E) mothers were pricked with a glass injection needle at 10-15mpf and allowed to recover until 30mpf. After recovery, most wild-type embryos have resealed their membrane, whereas a majority of aura mutant embryos show continued leakage of yolk and ooplasm (D,E, asterisks). (C,F) Wild-type (C) and aura mutant (F) embryos were fixed 1min after wounding and labeled with phalloidin. Wild-type embryos recruit F-actin to the closing wound edge (C, arrows; 7/7). All aura embryos examined show reduced F-actin enrichment at the wound edge and a larger wound diameter (F, arrows; 13/13). Scale bar: 100µm in C,F.

Fig. 6

Distribution of exocytic and endocytic factors in aura mutants. (A,B) The exocytic factor Rab11 properly localizes to the tips of microtubules at the furrow in wild-type (5/5) and aura mutant (6/6) embryos. (C,D) The exocytic factor Vamp2 localizes to mature furrows in wild type (5/5) but not in mutants (7/7). (E-H) The endocytic factors Clathrin and Dynamin 2 become enriched in ectopic CGs in aura mutants (F,F′,H,H′), whereas they exhibit a disperse cortical distribution in wild type (E,E′,G,G′). Clathrin appears localized to the CG surface in ectopic CGs in aura mutants (5/5 embryos, compared with 10/10 wild-type embryos with diffuse cortical labeling). Dynamin 2 also localizes to ectopic CGs (15/15 embryos, compared with 13/13 wild-type embryos with diffuse cortical labeling), which can appear as ring-like structures (3/15 embryos as shown in H′) or throughout the granule surface (12/15 embryos as shown in Fig. S5M,M′). Ectopic localization to CGs was confirmed by colabeling with WGA (not shown). (A′-H′) Higher magnifications of boxed regions in A-H. Scale bars: 100µm in A-H; 10µm in A′-H′.

Fig. 7

aura mutant embryos do not undergo dynamic actin rearrangement. (A-E) Wild-type embryos undergo dramatic actin rearrangement at the cortex leading to organized actin arcs (arrows), whereas aura mutant embryos fail in cortical actin rearrangement resulting in punctate F-actin aggregation (F-J, arrowheads). (A′-J′) Higher magnifications of boxed regions in A-J. At least three embryos were imaged per time point. Scale bars: 10µm.

Fig. 8

The aura mutant phenotype is enhanced by actin polymerization inhibitor and ameliorated by actin stabilizer. In wild type, cytochalasin D leads to a lack of F-actin arcs (B,B′; 17/17) and phalloidin does not have an effect (C,C′; 23/23). In aura mutant embryos, cytochalasin D leads to strongly punctate F-actin (G,G′; 17/17), and phalloidin reverses the phenotype, stabilizing F-actin arcs (H,H′, arrows; 11/11). Microtubule inhibiting (nocodazole, D,D′,I,I′) and stabilizing (taxol, E,E′,J,J′) drugs do not change the dynamic F-actin reorganization in wild-type (16/16 and 15/15, respectively) or aura mutant (15/15 and 10/10, respectively) embryos. (A′-J′) Higher magnifications of boxed regions in A-J. Embryos are at 35mpf. Scale bars: 10µm.

Fig. 9

aura encodes Mid1ip1l. (A) Mid1ip1l protein in wild-type and aura mutant alleles. Red boxes indicate predicted alpha helices. The aurt9792 mutation generates a premature stop that truncates the last conserved helix. The CRISPR/Cas9-generated mid1ip1luw39 allele results in a frameshift translated region (dark red box) followed by an early stop. (B) DNA sequencing trace of the aurt9792 allele, which creates a stop codon in amino acid 142. (C) Amino acid sequence comparison between various mid1ip1 homologs in the aurt9792 mutation site region. The mutation in aurt9792 occurs at a highly conserved lysine. Red stars (A,C) indicate amino acid directly affected by the mutations. D.r., Danio rerio; X.t., Xenopus tropicalis; H.s., Homo sapiens; G.g., Gallus gallus; M.m., Mus Musculus. Bottom row numbers are the Consistency Score according to the PRALINE protein alignment program. (D-G′) The CRISPR/Cas9-generated mid1ip1luw39 allele does not complement aurt9792. (D,E) All embryos from aurt9792/mid1ip1luw39 transheterozygous females (E) exhibit reduced membrane integrity (asterisk), reduced yolk coalescence (arrowhead) and regressed furrows (arrow; see also Fig. S1A,B), whereas control embryos from heterozygous siblings (D) are wild type in appearance. (F,G) 8-cell embryos from transheterozygous females exhibit reduced β-Catenin accumulation in mature furrows (27/27; G,G′, arrowheads), ectopic CGs throughout the cortex (G) and apparently stabilized FMA (G′, arrow), compared with control siblings (0/25; F,F′). (F′,G′) Higher magnifications of boxed regions in F,G. Scale bars: 100µm in F,G; 10µm in F′,G′.

Fig. 10

Expression of mid1ip1l and gene phylogeny. (A) RT-PCR analysis over developmental time (h or days post fertilization) of mid1ip1l RNA, co-amplifying β-actin RNA as an internal control. mid1ip1l RNA is maternally expressed, with levels subsiding at later stages. (B-G) In situ hybridization analysis. mid1ip1l antisense probes label early cleavage stage embryos (B, 25/25; E,F), with a reduction by dome stage (20/20; G). Levels and distribution of mid1ip1l RNA are similar in aura mutants (from aurt9792/t9792 females, D). Control sense RNA (C). (H-J) Mid1ip1l antibodies label within cortical F-actin in wild type (26/26; H). Reduced labeling is observed in aurt9792 embryos (20/20; I) and no localization in mid1ip1l uw39 embryos (21/21; J). (K) Higher resolution imaging of wild type (from H) shows Mid1ip1l protein localizing as puncta within cortical F-actin aggregates (8/8). Insert (K) shows a 2× magnification of the boxed region. In wild type, Mid1ip1l protein becomes enriched at the forming furrow (4/4; L) and induced wounds (4/4; M). (H-L) 45 mpf; (M) 15mpf. (N) RT-PCR analysis of mid1ip1a, mid1ip1b and mid1ip1l. RT-PCR data (A,N) are representative of two trials. (O) Phylogenetic tree for Mid1ip1 protein shows zebrafish mid1ip1b as most closely related to other vertebrate mid1ip1 genes, and mid1ip1a and mid1ip1l as products of gene duplication. Ixodes scapularis (deer tick) is an outgroup. Numbers in red indicate branch support values; number in black indicates branch scale bar. Scale bars: 10µm in H-J,L,M; 5µm in K.

Fig. S1 ZFIN is incorporating published figure images and captions as part of an ongoing project. Figures from some publications have not yet been curated, or are not available for display because of copyright restrictions.

Fig. S2

Blastodisc height and chorion expansion in wild-type and aura mutant embryos. (A,B) Blastodisc height in wild-type (n=39) and aura mutant embryos (n=42) (A). Height (orange lines, B) was measured as a percentage of the whole embryo length (red lines) at 30 mpf. Average blastodisc height in mutants is mildly yet statistically significantly reduced compared to wild-type (t-test, p-value=0.020). (C,D) Average chorion expansion in mutants is mildly yet statistically significantly reduced compared to wild-type in two of three measured time points (t-test type, p- values: 5 mpf: 0.010, 10 mpf: 0.16; 30 mpf: 0.0099). (C). Chorion lifting was measured by subtracting the area of the embryo (orange circle in D) from the area of the chorion (red circle in D). A fraction of aura mutant embryos (60%, n=1096) exhibit chorion fragility in which it separates as sublayers (E, arrows). Although we can not rule out other interpretations, this chorion fragility phenotype that can be interpreted as a consequence of reduced chorionic protein modification of chorion sublayers (Ulrich, 1969; Hart and Collins, 1991; Selman et al., 1993) due to decreased CG release in aura mutants. Embryos were a random collection from 3 females per genotype. Bars represent standard error values. Brackets reflect degree of statistical significance, with p-values < 0.01 (**), 0.01 - 0.05 (*), and > 0.05 (n.s.). Ovary sections of wild-type (pre- and post-CG (cg) accumulation at the cortex, just below the vitelline envelope (ve). F,G) and mutant (pre- and post-CG accumulation). H,I) Ovaries showing that CGs (arrowheads) are aligned at the cortex, and yolk (y) morphology is normal in aura oocytes.

PHENOTYPE:
Fish:
Observed In:
Stage: 1-cell

Fig. S3

Overview of landmarks of oogenesis and early cleavage in aura mutants. (A-D) Landmarks during oogenesis appear normal in aura oocytes. (A,B) Accumulation of F-actin appears unaffected normal in stage I aura mutant oocytes (wt= 5/5, aura= 4/4). Accumulation and localization of mitochondria and ER, marking the Balbiani body (Gupta et al. 2010), appears normal in stage I aura mutant oocytes (wt= 23/23, aura= 20/20). (E-H) Loss of blastomere cellular adhesion in aura mutant embryos. Live imaging of F-actin in embryos using the Life Act transgene in wild-type (3/3 E,G) and aura mutant line (3/3 F,H). The cellular pattern is relatively normal up to the 8-cell stage, as expected since cell adhesion is a property of mature furrows that become fully formed for the first cell cycle at this stage. By late 16-cell stage, aura embryos begin to display rounded and loose cells, which can be of varying sizes, consistent with defects in late cytokinesis involving reduced adhesive membrane deposition and membrane regression. The normal cleavage pattern up to the 8-cell stage indicates that the spindle orientation that results in the typically invariant cleavage pattern is normal. Images were taken in one minute intervals from 60 mpf to 85 mpf (E,F) and two minute intervals from 90 mpf to 120 mpf (G,H). (I-L) In aura mutants, DNA masses undergo epiboly-like vegetal movement in the absence of cellular layers. The DNA masses are uneven, likely due to the lack of cell membranes in these embryos and consequent mis-segregation of chromosomes by neighboring asters. Top panels (I,J, animal view) are pre-epiboly stages (4 hpf), bottom panels (K,L, side view) are 50% epiboly stages (5.5 hpf). DNA masses in aura mutants (J,L) appear at vegetal levels consistent with an epibolic-like movement, similar to migration in wild-type (I,K) except with reduced cell membrane formation. β-catenin (green, highlights membrane, which is reduced in mutants), DNA (blue). Scale bars: A-D: 25 µm; I-L: 100 µm

Fig. S4

Lack of FMA reorganization in aura embryos. Live imaging of microtubules in embryos using the EMTB transgene in wild-type and aura mutant line. (A) In wild-type, the FMA forms along the furrow (double arrows), compacts distally (arrows) and disappears (arrowheads). (B) In aura mutant embryos, the FMA also appears along the furrow (double arrows), but does not experience a distal shift and remains in an apparently stabilized condition until it eventually becomes undetectable (dotted line arrows). Images were taken in one minute intervals from 60 mpf to 85 mpf. (C,D) Microtubule reorganization defect occurs independently of the number of ectopic CGs in aura mutants. In aura mutants, the FMA appears stabilized in all (100%) embryos, regardless of the number of ectopic CGs at the furrow: (C) furrow with ≥ 5 GCs (6/16 examined furrows); (D) furrow with ≤ 4 CGs (10/16 examined furrows). Furrows shown are examples of the 1st furrow of a single embryo at 75 mpf, when the FMA has normally become either distally enriched or disassembled (Urven et al, 2006). Similar results are obtained with the 2nd furrow (not shown) Scale bar: C,D, 10 µm

EXPRESSION / LABELING:
Gene:
Fish:
Anatomical Term:
Stage: 1-cell

Fig. S5

Membrane dynamics of wild-type and aura mutant embryos. (A-D) Side view of wild-type (A,B) and aura mutant (C,D) embryos at 2 mpf and 10 mpf, labeled to detect F-actin and CGs, showing that mutants exhibits a similar localization of CGs at the cortex during egg activation (A,C) and retains a fraction of CGs (B,D). (E-F) In wild-type at 45 mpf, β-catenin localizes to the furrow but Vamp2 does not yet localize to this structure (4/4 embryos) (E, E′) In aura mutants, Vamp2 appears localize to ectopic CGs (8/9 embryos) (F,F′). Vamp2 also localizes to small cortical particles in both wild-type and mutant embryos (E′, F′). (G-L) Localization of Clathrin (G, G′), Dynamin2 (H,H′), Rab11 (I,I′), Vamp2 (J,J′), Vamp4 (K,K′), and ß-catenin (L,L′) to a subset of CGs in wild-type at 2 mpf (G-H) or 0 mpf (I-L). (M, M′) Dynamin2 surrounding CGs in aura mutant embryos as seen by a face-on view (M) and a reconstructed orthogonal view (M′). Scale bars: A-D (bar in A): 10 µm; E-H (bar in E): 100 µm; E′-H′(bar in E′): 10 µm; I-M′ (bar in I,I′,M): 10 µm.

Acknowledgments:
ZFIN wishes to thank the journal Development (Cambridge, England) for permission to reproduce figures from this article. Please note that this material may be protected by copyright. Full text @ Development