Maternal regulation of the vertebrate oocyte-to-embryo transition

Abstract

Maternally-loaded factors in the egg accumulate during oogenesis and are essential for the acquisition of oocyte and egg developmental competence to ensure the production of viable embryos. However, their molecular nature and functional importance remain poorly understood. Here, we present a collection of 9 recessive maternal-effect mutants identified in a zebrafish forward genetic screen that reveal unique molecular insights into the mechanisms controlling the vertebrate oocyte-to-embryo transition. Four genes, over easy, p33bjta, poached and black caviar, were found to control initial steps in yolk globule sizing and protein cleavage during oocyte maturation that act independently of nuclear maturation. The krang, kazukuram, p28tabj, and spotty genes play distinct roles in egg activation, including cortical granule biology, cytoplasmic segregation, the regulation of microtubule organizing center assembly and microtubule nucleation, and establishing the basic body plan. Furthermore, we cloned two of the mutant genes, identifying the over easy gene as a subunit of the Adaptor Protein complex 5, Ap5m1, which implicates it in regulating intracellular trafficking and yolk vesicle formation. The novel maternal protein Krang/Kiaa0513, highly conserved in metazoans, was discovered and linked to the function of cortical granules during egg activation. These mutant genes represent novel genetic entry points to decipher the molecular mechanisms functioning in the oocyte-to-embryo transition, fertility, and human disease. Additionally, our genetic adult screen not only contributes to the existing knowledge in the field but also sets the basis for future investigations. Thus, the identified maternal genes represent key players in the coordination and execution of events prior to fertilization.

Fig 1. Zebrafish oocyte maturation mutants.

A. Wild-type stage IV oocyte and egg in incident light. Eggs of p33bjta (1570 from 21 females), poac^p26ahubb (1620 from 35 females), and p35aluc (2088 from 25 females) resemble wild-type oocytes in their opacity, while eggs of p26ahubg (546 from 8 females) and blac^p25bdth (1020 from 17 females) were degenerated. B. Images of 5 minutes post-egg activation (mpa) wild-type (n = 68), p26ahubg (n = 56) and blac^p25bdth (n = 61) eggs, showing lysed mutant eggs. bd, blastodisc. C. Non-complementation test showed that females carrying the ovy^p37ad mutation in trans to the p35aluc mutation produced opaque mutant eggs (n = 1382 from 9 females), showing that p35aluc is a new ovy allele. D. Coomassie stained gel of major yolk proteins in wild-type and mutant eggs showing that the cleavage of yolk protein is impaired. kD, kilodaltons. E. Low (top row) and high (bottom row) magnification histological sections of wild-type and mutant stage III oocytes stained with H&E revealing YG sizes. Scale bar = 96 μm (top row) and 16 μm (bottom row). F. Scatter plots of YG diameters measured in a defined region of the oocyte. The mean diameter of the YGs in the mutants was significantly different to that of wild-type oocytes (WT = 14.6 μm, p33bjta = 3.87 μm, p35aluc = 5.95 μm, p25bdth = 9.25 μm). The number of YGs analyzed in this study was: wild type (n = 188 from 4 oocytes), p33bjta (n = 320 from 2 oocytes), ovy^p35aluc (n = 503 from 2 oocytes), and blac^p25bdth (n = 177 from 2 oocytes). Data are means ± SEM. ****p<0.0001 and **p = 0.0066 in parametric statistical Tukey’s test. ns, not significant. G. Confocal z-projections of isolated acid fixed oocytes analyzed by fluorescent in situ hybridization and also DiOC₆ staining (green) in lower panels. In stage II oocytes, cycB1 transcript localized to the animal pole in wild-type (n = 32), p33bjta (n = 22), poac^p26ahub (n = 47), ovy^p35aluc (n = 38), and blac^p25bdth (n = 28) oocytes. In stage I oocytes, dazl mRNA co-localized with DiOC₆ to the vegetal Balbiani body (Bb) in wild-type (n = 32), p33bjta (n = 23), poac^p26ahub (n = 63), ovy^p35aluc (n = 31), and blac^p25bdth (n = 29) oocytes. Scale bar = 35 μm.

Fig 2. Meiosis II completion and egg activation defect in dark eggs mutants.

A. Confocal micrographs of the animal pole show anaphase of meiosis II at 5 mpa and the extruded polar body at 10 mpa. Top row: DAPI-stained wild-type (n = 24), p33bjta (n = 14), poac^p26ahubb (n = 22) and ovy^p35aluc (n = 21) sister chromatids during meiotic anaphase II. Bottom row: Phalloidin-stained actin (red) and DAPI staining (green) of second polar bodies in wild-type (n = 14), p33bjta (n = 9), poac^p26ahubb (n = 12) and ovy^p35aluc (n = 10) activated eggs. Scale bar = 4 μm. B. Egg-to-chorion index (ECI) is the ratio of egg area (A₁) to chorion area (A₂) as a measure of chorion elevation. Scatter plots of the ECI measured in activated eggs at 30 mpa. The wild-type egg ECI value (ECI = 0.61; SD = 0.02; n = 65) defines the numerical value for normal chorion elevation. The ECI for the eggs from p33bjta (ECI = 0.64; SD = 0.05; n = 44, 2 females) and ovy^p35aluc (ECI = 0.63; SD = 0.03; n = 20, 2 females) mutant females are significantly different to wild type, while the eggs of poac^p26ahubb (ECI = 0.59; SD = 0.02; n = 25, 2 females) mutant females show a mean value similar to wild type. SD = standard deviation. Data are means ± SD. *p = 0.0131 and ***p = 0.0004 in parametric statistical Tukey’s test. C. Confocal z-projections (5 μm depth) of acid fixed and MPA stained wild-type and mutant activated eggs. In wild-type (n = 60), p33bjta (n = 55), ovy^p35aluc (n = 25), and poac^p26ahubb (n = 27) eggs, numerous large and small CGs were distributed throughout the cortex at 0 mpa. At 30 mpa, wild-type (n = 21), p33bjta (n = 28), poac^p26ahubb (n = 20) and ovy^p35aluc (n = 18), eggs displayed fewer CGs, indicating that they were released following activation. However, some CGs persisted in p33bjta (5/28), poac (11/20) and ovy^p35aluc (6/18) mutant eggs. D. Scatter plots of CG retention measured in a determined lateral region (DLR) of activated wild-type and opaque mutant eggs. The total area of retained CGs per total DLR area in all dark egg mutants was significantly increased compared to wild type at 30 mpa. n, number of eggs used in this analysis from 3 females. Data are means ± SEM. *p = 0.0249, **p = 0.0485 between wild type and poac, and ***p = 0.0042 between wild type and p33bjta in a nonparametric statistical Kruskal-Wallis test. ns, not significant; mpa, minutes post-activation. E. Scatter plots of CG number (top graph) and median size (bottom graph) measured in the DLR of the wild-type and dark activated eggs. The average number of CGs in the mutants was comparable to wild type at 0 mpa but significantly different in the p33bjta and poac^p26ahubb mutants at 30 mpa. Unactivated (wild type (n = 7), p33bjta (n = 7), poac^p26ahubb (n = 7) and ovy^p35aluc (n = 5)) and activated (wild type (n = 8), p33bjta (n = 5), poac^p26ahubb (n = 4) and ovy^p35aluc (n = 6)) eggs from 3 females per genotype were used in this analysis. Data are means ± SEM. Top graph: *p = 0.0353 and **p = 0.0012 in a nonparametric statistical Kruskal-Wallis test. Bottom graph: *p = 0.0250 in a nonparametric statistical Kruskal-Wallis test, and **p = 0.0011 in a parametric statistical Tukey’s test. F. Top row: Animal views of DAPI stained p33bjta/+ (n = 34, 2 females), and p33bjta (n = 36, 2 mutant females) 4 hpf embryos, showing abnormal chromatin organization. Middle row: Lateral views of 4-cell stage embryos from p33bjta/+ (n = 77/111, N = 2 females), and p33bjta (n = 29/59, N = 3) females. Bottom row: Lateral views of 4 hpf embryos from p33bjta/+ (n = 77/111, N = 2) and p33bjta (n = 29/59, N = 3) females. G. Top row: Animal views of DAPI stained 4 hpf embryos from ovy^p35aluc/+ (n = 19, N = 2), and ovy^p35aluc (n = 9, N = 2 females), showing abnormal chromatin organization. Middle row: Lateral views of 4- and 8-cell stage embryos from two heterozygous (n = 110), and one homozygous (n = 69) ovy^p35aluc females. Bottom row: Lateral views of 4 hpf embryos from 2 heterozygous (n = 110), and one homozygous (n = 10) ovy^p35aluc females. hpf, hours post fertilization. Scale bar = 4 μm (A), 40 μm (C), 210 μm (F, G).

Fig 3. Molecular nature of the ovy gene.

A. Genetic and physical map of the ovy locus and schematic representation of the zebrafish ap5m1 gene, which consists of 8 coding exons and 7 introns. z7170 and z8980 are SSLP markers flanking the ovy mutation. In parenthesis, the number of recombinants/total analyzed meioses defining the interval of ovy. Exons are shown as red boxes and introns as black lines. Sizes are not to scale. B. DNA sequencing analysis of the ovy^p37ad and ovy^p35aluc mutant alleles. Left: genomic DNA sequence of the wild-type and ovy^p37ad allele indicates a single point mutation in the splice donor site of intron 1 of the ap5m1 gene (red rectangle). Right: cDNA sequence of the wild-type and p35aluc allele indicates a single point mutation in exon 2 of the ap5m1 gene (red rectangle). C. Left: predicted tertiary structure of the zebrafish Ap5m1 protein. The α- and β-helixes are colored in purple and pink, respectively, and the connecting loops in cyan. N- and C-terminal domains are indicated. The boundary of the two domains is at residue 205 (arrow head), which is found in the flexible loop comprised by residues 204–208. The approximate volumetric density map of the protein is shown in transparent gray. Right: predicted domain architecture of the Ap5m1 N-terminal portion containing residues Arginine 24 (R24) and Threonine 27 (T27). The amino acid change and premature stop codon caused by the ovy^p35aluc and ovy^p37ad mutations, respectively, are shown. D. Multiple amino acid alignments of the Ap5m1 protein and representative members of the vertebrate lineage. Top: Schematic of the protein alignment. The overall percentage (%) identity decreases from top to bottom. The red box indicates the first 49 amino acid residues of Ap5m1. Bottom: Detailed amino acid alignment showing high conservation in fish, amphibians and mammals. Note the high conservation of the Threonine (T) amino acid (red asterisk), which is mutated to Lysine (K) in ovy^p35aluc (indicated in red in the lower panel, along with the splice site mutation, causing a premature stop codon). E. Structural superposition between zebrafish Ap5m1 (pink, amino acids 205–475) and the crystal structure of human Ap4m1 C-terminal domains (purple, amino acids 185–453). The RMSD value corresponds to 2.45 Å, thus demonstrating structural identity in overall 3D structure.

Fig 4. Egg activation mutants.

A. Bright-field images showing early developmental stages of live wild-type and mutant eggs and embryos. All unactivated eggs collected after gently squeezing gravid females showed no detectable defects. After egg activation, however, the krang mutant egg displays a pronounced chorion elevation defect (black arrows). As development proceeds, the cytoplasm is abnormally distributed in the yolk cell (yc) of spotty mutant eggs and early embryos (black arrowheads) and a defective blastoderm (bt) is formed. krang (1887 eggs from 50 females), spotty (1273 eggs from 40 females), and p28tabj (2506 eggs from 34 females) and kazu (1608 eggs from 32 females). B. Lateral view of acid-fixed wild-type and mutant eggs and early embryos showing the organization of the central cytoplasm (bright). In wild type, the blastodisc (bd) grows by animal-ward transport of cytoplasm from the yolk cell (yc) and by the 4-cell stage, the cytoplasm in the yolk cell is no longer evident. In contrast, spotty, p28tabj and kazu mutant eggs exhibited an abnormal distribution and severe retention of cytoplasm in the yolk cell (black asterisks). Wild type (n = 22), spotty (n = 17), and p28tabj (n = 19) and kazu (n = 12). C. Lateral view of acid-fixed and DiOC₆-stained wild-type and mutant eggs and early embryos showing the organization of the cortical cytoplasm (green) and yolk (dark). Wild type (n = 20), spotty (n = 20), and p28tabj (n = 14) and kazu (n = 18). Scale bar = 830 μm (A, columns 1 and 2 from left to right), 150 μm (A, columns 3 and 4 from left to right), 190 μm (B, C).

Fig 5. Characterization of the maternal-effect krang mutant phenotype.

A. Quantification of the chorion elevation phenotype. Scatter plots of the measured ECI value show a penetrant chorion elevation defect in the krang mutant egg at 30 mpa (ECI = 0.64 and 0.83; SD = 0.006 and 0.005; for the wild-type (n = 48, 2 females) and mutant (n = 48, 2 females) egg, respectively). SD = standard deviation. Data are means ± SD. ****p<0.0001 in statistical clustering analysis in a nonparametric unpaired t test of chorion elevation measurements. B. Confocal z-projections (5 μm depth) of MPA stained CGs showing their distribution in wild-type (top row, n = 68 eggs from 3 females) and krang mutant (bottom row, n = 55 eggs from 3 females) eggs at two different time points after activation. Images were taken in a determined lateral region (DLR) of each activated egg. Notice that smaller, possibly CGs lacking MPA-staining lectin accumulate in the mutant egg at 20 mpf. C. Scatter plots of CG exocytosis timing measured in the DLR of the activated egg. The average number of CGs in the krang mutant was not reduced compared to wild type at 0 min but is significantly higher at later time points. Normalized total area curves (upper right graph) showing CG exocytosis rate in wild-type and krang mutant activated eggs. Five eggs from 3 females per genotype and per time point were used in this analysis. Data are means ± SEM. *p = 0.0131 and **p = 0.0252 in a parametric statistical two-way ANOVA test. ns, not significant. D. Hematoxylin and Eosin stained sections of intact stage II (left column) and stage III (right column) oocytes. Wild-type (n = 2 ovaries), and krang (n = 2 ovaries) oocytes, were phenotypically comparable. CGs are formed (stage II) and accumulated at the cortex (stage III). E. Confocal micrographs of actin- and DAPI-stained fertilized eggs showing normal meiosis completion and second polar body formation (wild type (n = 6, 2 females) and mutant (n = 15, 2 females). F. Most krang mutant embryos failed to undergo cell divisions. Top row: wild-type embryo with a typical symmetric blastoderm at 2 hpf (left panel, n = 87/93 embryos from 2 females). In contrast, krang mutant embryos display relatively normal (middle panel, n = 78/227) or abnormal (right panel, n = 147/227) blastoderm formation (n = 3 mutant females). Bottom row: high magnification confocal micrographs showing the cell architecture in wild-type and mutant blastoderms. cg, cortical granule; yg, yolk globules. Scale bar = 40 μm (B), 25 μm (D, left column), 95 μm (right column), 50 μm (E, left column), 12 μm (E, right column), 160 μm (F, top row), 85 μm (F, bottom row).

Fig 6. Molecular characterization of the Krang maternal factor.

A. Schematic representation of the molecular lesion and generation of a premature donor splice site 35 nt upstream of the wild type (wt) donor splice site in the krang/kiaa0513 (mutant) transcript. B. Predicted tertiary structure of the full-length zebrafish Krang protein (left) and its truncated version (right). The α-helixes and the connecting loops are colored in pink and green, respectively. N- and C-terminal domains are indicated. The SBF1/2 functional domain of the protein is shown in yellow. The approximate volumetric density map of the protein is shown in translucent gray. C. Krang protein sequence alignment of multiple species represented in E. Top: Schematic representation of the protein sequence alignment. The overall percentage (%) similarity among sequences decreases from top to bottom. The yellow box indicates the SBF1/2 functional domain in Krang. Bottom: Detailed sequence alignment, showing high conservation of this domain in fungi, insects, fish, birds and mammals. Percent similarity is color coded by the scale bar at the right. D. Reduced neighbor-joining phylogenetic tree of Krang homologs found in public databases indicating its evolutionary history from invertebrate to vertebrate species. The scale bar in the middle represents evolutionary distances based on residue substitutions per site. E. In situ hybridization showing the cortically-restricted distribution of krang mRNA in a cryosectioned stage III oocyte (n = 18). Notice that the expression of the transcript (arrowheads) is associated with MPA-stained CGs. FCs, follicle cells. Right panels show selected areas (arrowheads) in separate channels. F. Whole-mount in situ hybridization showing krang transcript localization. Top row: wild-type stage III oocyte (n = 43), where krang mRNA is peripherally distributed in the cell, presumably overlapping with cortical granules (white asterisks in high magnification images of the same oocyte). Bottom row: stage II oocyte (n = 19), where the krang transcript is associated with nascent CGs (bright field and fluorescent high magnification images). Control sense probe did not show transcript signal in stage III oocytes (n = 23, top row, left). Boxes show selected magnified areas. Scale bar = 20 μm (E), 140 μm (F, stage III oocyte), 9 μm (F, top row high magnification images), 55 μm (F, stage II oocyte), 20 μm (F, bottom row high magnification images).

Fig 7. Phenotypic characterization of the kazu^p26thbd mutant early embryo.

A. Bright field images showing early developmental stages of whole-mount live wild-type (n = 99/99 from 2 females) and kazu^p26thbd (n = 185/265 from 2 females) embryos. The mutant phenotype reveals a smaller blastodisc (bd) with smaller cells during cleavage. These abnormalities in cell size acquisition, together with asynchronous and unequal cell divisions, leads to the formation of a defective blastula with many detached blastomeres. B. Higher magnification images showing a cellularized wild-type blastoderm (n = 125/125). In contrast, kazu^p26thbd embryos exhibit a less compacted blastoderm containing smaller cells (n = 62/91). C. Lateral and animal views of whole-mount wild-type (n = 31/35) and mutant (n = 39/47) early embryo that show the distribution of α-Tubulin at 2 hpf. In wild-type embryos, α-Tubulin is mainly distributed in the blastoderm (bt) and dividing blastomeres. In kazu^p26thbd mutants, α-Tubulin is distributed in different sized cells and throughout the yolk cell (yc, asterisks). D. Animal views of whole-mount wild-type (n = 26) and mutant (n = 8) early embryos that show the localization of β-Catenin at 2 hpf. In wild-type embryos, β-Catenin localized to the blastomere boundaries. In kazu^p26thbd mutants, β-Catenin is aberrantly localized in different cytoplasmic domains. E. Animal views of DAPI, α-tubulin, and centrosome staining showing abnormal blastoderm formation at the sphere stage. Cell boundaries and mitotic figures are not observed in most of the mutant (n = 25/33) compared to wild-type (n = 30/35) embryos examined. Notice the formation of a syncytial nuclei layer in the 5 hpf kazu^p26thbd embryo. Scale bar = 180 μm (A), 110 μm (B), 230 μm (C), 165 μm (D), 100 μm (E).

Fig 8. Chromosome, γ-Tubulin and microtubule organization in the spotty^p08bdth mutant egg and zygote.

A. Ectopic cytoplasmic domain formation in the spotty mutant. Instead of a regular organization of cytoplasm intermingled with YGs in the wild-type yolk cell (n = 74 embryos from 2 females), the spotty mutant (n = 65 embryos from 2 females) shows numerous, peripheral cytoplasmic domains (black arrows). B. Animal viewed confocal images of DiOC₆- stained zygotes reveal that YG arrangement is affected, coinciding with peripheral patches of cytoplasm (white asterisks). Higher magnification DAPI- and γ-Tubulin-stained images show perturbed nucleus formation and γ-Tubulin assembly in spotty (n = 37 from 3 females) compared to wild-type (n = 33 from 3 females) zygotes at 25 mpf. Boxes show magnified areas below. C. DAPI-stained wild-type (n = 20 from 2 females) and mutant (n = 48 from 2 females) unactivated eggs showing chromosome organization during metaphase II of meiosis. D. α-Tubulin- and DiOC₆-stained confocal z-projections showing MT and YG distribution at the cortex of wild-type (n = 34) and spotty (n = 41) eggs. Higher magnification images reveal the magnitude of such MT structures and YG organization in the mutant egg. E. α- and γ-Tubulin co-staining indicates that MTs emanate from numerous MTOCs across the mutant egg (n = 14). Image 4 was selected from a z-stack containing 31 x 1 μm-thick optical sections across the mutant egg. Boxes show magnified (B, D and E) areas. bd, blastodisc; yc, yolk cell; yg, yolk globules; cp, cytoplasmic pocket; A, animal pole; V, vegetal pole; mpf, minutes post fertilization. Scale bar: 190 μm (A), 200 μm (B, top row), 50 μm (B, bottom row), 5 μm (C), 180 μm (D, low magnification), 38 μm (D, high magnification), 180 μm (E, top row), 60 μm (E, bottom row).