Formation of Different Polyploids Through Disrupting Meiotic Crossover Frequencies Based on cntd1 Knockout in Zebrafish

Abstract

Polyploidy, a significant catalyst for speciation and evolutionary processes in both plant and animal kingdoms, has been recognized for a long time. However, the exact molecular mechanism that leads to polyploid formation, especially in vertebrates, is not fully understood. Our study aimed to elucidate this phenomenon using the zebrafish model. We successfully achieved an effective knockout of the cyclin N-terminal domain containing 1 (cntd1) using CRISPR/Cas9 technology. This resulted in impaired formation of meiotic crossovers, leading to cell-cycle arrest during meiotic metaphase and triggering apoptosis of spermatocytes in the testes. Despite these defects, the mutant (cntd1^−/−) males were still able to produce a limited amount of sperm with normal ploidy and function. Interestingly, in the mutant females, it was the ploidy not the capacity of egg production that was altered. This resulted in the production of haploid, aneuploid, and unreduced gametes. This alteration enabled us to successfully obtain triploid and tetraploid zebrafish from cntd1^−/− and cntd1^−/−/− females, respectively. Furthermore, the tetraploid-heterozygous zebrafish produced reduced-diploid gametes and yielded all-triploid or all-tetraploid offspring when crossed with wild-type (WT) or tetraploid zebrafish, respectively. Collectively, our findings provide direct evidence supporting the crucial role of meiotic crossover defects in the process of polyploidization. This is particularly evident in the generation of unreduced eggs in fish and, potentially, other vertebrate species.

Introduction

Polyploid organisms, defined as those having three or more sets of chromosomes, typically arise as a result of whole-genome duplication (Comai 2005; Van de Peer et al. 2017). A wide variety of crucial crop species, including potatoes, coffee, bananas, peanuts, tobacco, wheat, oats, sugarcane, and various fruits, are known to exhibit polyploidy (Renny-Byfield and Wendel 2014; Akagi et al. 2022). Conservative estimates suggest that all angiosperm species in natural ecosystems are polyploid (Soltis 2005; Jiao et al. 2011), and polyploid characteristics are widespread in higher plant lineages (Soltis et al. 2015; Alix et al. 2017). While polyploidy was once thought to be rare in animals, emerging evidence indicates its extensive occurrence across a diverse range of taxa, including insects, amphibians, reptiles, and fish (Otto and Whitton 2000; Gregory and Mable 2005; Mable et al. 2011; Song et al. 2012; Wertheim et al. 2013). This suggests that polyploidy is more prevalent among animal groups than previously believed.

Polyploidy is widely acknowledged as a prevalent mechanism in speciation, giving rise to novel phenotypes, ecological diversification, and the exploration of new niches. It offers advantages such as genome buffering, fixation of heterosis, and evolving plasticity (Adams and Wendel 2005; Comai 2005; Burgess 2015; Van de Peer et al. 2017; Shimizu 2022). Despite potential drawbacks on fertility, fitness, and the occurrence of mitotic and meiotic abnormalities (Comai 2005; Stenberg and Saura 2013), the genetic variation facilitated by alterations in gene expression and epigenetic remodeling contributes significantly to the adaptive potential of polyploids. This enhances their capacity to withstand environmental challenges (Comai 2005; Madlung and Wendel 2013; Burgess 2015; Scarrow et al. 2021). This adaptive potential not only ensures the survival of polyploids in the short term but also positions them to potentially thrive while reducing the relative risk of extinction in more demanding environments over the long term (Van de Peer et al. 2017).

Polyploids can be categorized as allopolyploids or autopolyploids. Natural allopolyploids arise through hybridization (Stebbins 1950; Soltis and Soltis 2009; Deb et al. 2023). Examples of plant species exhibiting allopolyploidy include hexaploid bread wheat (Triticum aestivum L.) (Haider 2013), tetraploid cotton species Gossypium hirsutum L. and Gossypium barbadense L. (Jiang et al. 1998), Spartina anglica (Ayres and Strong 2001; Baumel et al. 2001), Senecio cambrensis and S. eboracensis (Abbott and Lowe 2004), Cardamine schultzii (Urbanska et al. 1997; Mandáková et al. 2013), and Tragopogon mirus and T. miscellus (Buggs 2012). In fish species, examples include Leuciscus alburnoides (Alves et al. 1997, 2001), Phoxinus eos-neogaeus (Angers and Schlosser 2007), and Poecilia formosa (Lampert and Schartl 2008). The formation of autopolyploids is a more intricate process that can result from the production of unreduced gametes through premeiotic doubling of chromosome number or meiotic defects (Brownfield and Köhler 2011; Mason and Pires 2015) as well as polyspermy (Toda and Okamoto 2016).

While the molecular mechanisms driving polyploid formation remain incompletely elucidated, there is widespread consensus that the generation of unreduced gametes represents a common pathway to polyploidy (Brownfield and Köhler 2011). Investigations into meiotic mutants across various plant species have disclosed that meiotic defects, including anomalies such as the omission of meiotic divisions, aberrant spindle morphology, or disrupted cytokinesis, can culminate in the production of viable, unreduced gametes (Bretagnolle and Thompson 1995). The use of Arabidopsis models has led to the identification of several genes crucial for maintaining ploidy in plants. Disruptions in genes such as swi1 (switch1), tam (tardy asynchronous meiosis), osd1 (omission of the second division 1), atps1 (Arabidopsis thaliana parallel spindles1), jason (aspartyl/glutamyl-tRNA amidotransferase subunit), and tes (tetraspore) can result in meiotic defects related to sister chromatid cohesion, recombination, entry into the second meiotic division, spindle orientation, and meiotic cytokinesis, respectively. These insights shed light on the molecular intricacies underlying polyploid formation (Spielman et al. 1997; Mercier et al. 2001; Mercier et al. 2003; d'Erfurth et al. 2008; d'Erfurth et al. 2010; De Storme and Geelen 2011). Consequently, these findings significantly contribute to our comprehension of polyploidy formation at the molecular level in plants.

In vertebrates, polyploidy is most frequently observed in fish, especially among teleost fish groups, with recent evidence supporting the occurrence of polyploidy (Gregory and Mable 2005). Fish groups, including the Salmonid and Catostomidae families, nearly all species of Carassius and Cyprinus and various special complex polyploid systems, exhibit polyploidy (Gregory and Mable 2005; Song et al. 2012). Polyploidy is also prevalent in hybrid taxa, with four bisexual-fertile tetraploid lineages derived from hybridizations between Carassius auratus red var and Cyprinus carpio (Liu et al. 2001), Carassius auratus red var and Megalobrama amblycephala (Qin et al. 2014), Carassius auratus cuvieri and Megalobrama amblycephala (Hu et al. 2019), and Cyprinus carpio and Megalobrama amblycephala (Wang et al. 2020). Despite this prevalence, the molecular mechanism underlying the formation of polyploid fish remains poorly understood.

The disruption of meiotic crossovers (COs) is of particular interest due to their critical role in ensuring the precise segregation of homologous chromosomes during meiosis I (Pazhayam et al. 2021). The investigation focuses on whether such disruption can lead to polyploidization. Cntd1, a member of the cyclin superfamily, encodes a conserved cyclin-related protein found in metazoans. Cntd1 orthologs are found throughout the metazoan lineage from worms to mammals but are absent in the genus Drosophila. CO site-associated 1, the C. elegans Cntd1 ortholog, has been proposed to function in conjunction with MSH-4/MSH-5 and ZHP-3 in a self-amplifying mechanism to sequester CO-promoting factors at specific CO sites (Yokoo et al. 2012). In mice, CNTD1 has been identified as a factor in meiotic crossover maturation and the deselection of excess precrossover sites, influencing CO sites alongside MutL and HEI10 (Holloway et al. 2014). The observations in nematodes and mammals suggest a conserved function of Cntd1 in influencing meiotic crossover.

The zebrafish Cntd1 ortholog, located on chromosome 12, comprises seven exons encoding a 254 amino acid protein; however, its functional properties were unexplored. This study involved generating a zebrafish line with a cntd1 knockout, revealing impaired crossover formation and random segregation of homologous chromosomes during meiosis. This defect triggered apoptosis in missegregated spermatocytes, resulting in the production of a limited number of normal sperm. Unevenly divided oocytes developed into mature follicles, producing a majority of aneuploid eggs and a smaller proportion of euploid eggs with haploidy and diploidy in mutant females. Interestingly, mutant females exhibited the ability to produce unreduced eggs, providing an opportunity to increase the ploidy of their offspring. Consequently, triploid and tetraploid zebrafish were successfully generated from diploid and triploid mutant females, respectively. By crossing tetraploid and diploid zebrafish, all-triploid zebrafish were obtained. This study represents the first evidence elucidating the potential molecular mechanism underlying the formation of polyploids in fish and establishes a new feasible strategy for ploidy regulation through gene knockout from a breeding perspective.

Results

Generation of cntd1 Knockout Zebrafish

A mutant strain was successfully generated by introducing a G base into exon 2, leading to a frameshift mutation. This mutation is anticipated to cause premature translation termination, resulting in a truncated protein comprising only 36 amino acids with 19 correctly translated amino acids, in contrast to the wild-type (WT) Cntd1, which consists of 256 amino acids (Fig. 1A). WT, cntd1^+/−, and cntd1^−/− populations of both female and male individuals were maintained under identical feeding conditions. The different genotypic populations exhibited similar appearances without discernible growth differences (Fig. 1B–G). All male zebrafish from the three genotypic populations displayed spikelike breeding tubercles distributed along the central fin rays on the pectoral fins, absent from the dorsal surface of the pectoral fins of all female zebrafish (Fig. 1H–M). Reverse transcription polymerase chain reaction (PCR) detection in WT zebrafish demonstrated that cntd1 was expressed in nearly all tissues examined, including the eye, brain, gill, spleen, heart, liver, intestine, testis, ovary, and muscle, with notable enrichment in the testis (Fig. 1N). Cntd1 expression was also traced from the early embryonic stage to the juvenile stage in WT zebrafish, revealing cntd1 transcripts at all examined stages from 0 hpf (hours postfertilization) to 40 dpf (days postfertilization), with an enrichment in the early stages of embryonic development (Fig. 1O). Further investigation uncovered a significant decrease in cntd1 expression levels in the testis of cntd1^−/− zebrafish, confirming the compromised function of Cntd1 (Fig. 1P).

Fig. 1.

Generation of cntd1 knockout zebrafish. A, Schematic representation of the genomic structures of the zebrafish cntd1 gene and the corresponding putative peptides and the target sites of engineered CRISPR-Cas9 in exon 2 (exons shown in red). The inserted base is highlighted in the blue box. The correct amino acids are indicated by white boxes, while the incorrect amino acids in cntd1 mutants are presented in a dark box. B–G, General morphological observations of zebrafish with different genotypes: cntd1^+/+ (B), cntd1^+/− (C), and cntd1^−/− (D) females; cntd1^+/+ (E), cntd1^+/− (F), and cntd1^−/− (G) males. H–M, Microscopic observation of the central fin rays on the pectoral fins. Absence of breeding tubercles in cntd1^+/+ (H), cntd1^+/− (I), and cntd1^−/− (J) females; the presence of spike-like breeding tubercle clusters in cntd1^+/+ (K), cntd1^+/− (L), and cntd1^−/− (M) males. Scale bar: 200 μm. (N) Tissue distribution of endogenous cntd1 transcripts with RT-PCR in WT zebrafish. (O) Expression levels of endogenous cntd1 at different developmental stages checked by RT-PCR in WT zebrafish. (P) Relative expressions of cntd1 in the testis of WT and cntd1^−/− zebrafish. Six individuals were used, and ef1α was selected as the internal reference in this experiment. The error bars represent the means ± SDs; “***” above the error bar indicates statistically significant differences at P < 0.001 by two-tailed Student's t-test.

cntd1 Knockout Results in the Meiotic Arrest and Apoptosis of Spermatocytes in Zebrafish

Anatomical observation revealed distinct differences between the testes of cntd1^−/− and WT zebrafish. WT zebrafish exhibited well-developed, milky white testes (Fig. 2A), while cntd1^−/− testes were nearly transparent and poorly developed (Fig. 2B). Histological sections with HE staining demonstrated that the testes of cntd1^−/− zebrafish had significantly fewer spermatids (ST) and spermatozoa (SZ) compared to WT testes (Fig. 2C–F), indicating defective sperm production in cntd1^−/− zebrafish. In the testes of cntd1^−/− zebrafish, a considerable number of spermatocytes with highly condensed chromatin arranged on the equatorial plates were observed, accompanied by the conspicuous presence of spindles on both sides. Some of these arrested spermatocytes exhibited aberrant chromosome segregation (Fig. 2F). In contrast, spermatocytes at meiotic metaphase were not commonly observed in WT testes (Fig. 2E). These results suggested that a significant portion of spermatocytes in cntd1^−/− testes were arrested at meiotic metaphase. Based on chromosomal morphology and cell size, observed germ cells were classified into nine stages, including spermatogonia, leptotene (supplementary fig. S1A, Supplementary Material online), and pachytene of prophase I (supplementary fig. S1B and C, Supplementary Material online), metaphase I (supplementary fig. S1D, Supplementary Material online), anaphase I (supplementary fig. S1E, Supplementary Material online), prophase II (supplementary fig. S1F, Supplementary Material online), metaphase II (supplementary fig. S1G, Supplementary Material online), spermatid stage (supplementary fig. S1H, Supplementary Material online), and spermatozoon stage (supplementary fig. S1I, Supplementary Material online), with primary spermatocytes (PSCs) and secondary spermatocytes (SSCs) identified as those at meiosis I and meiosis II, respectively (supplementary fig. S1, Supplementary Material online). Statistical analysis showed that the germ cell composition in cntd1^−/− testes was significantly altered, with decreased amounts of ST and SZ and increased amounts of PSC and SSC (Fig. 2G). The number of spermatocytes arrested at meiosis metaphase was also notably higher in cntd1^−/− zebrafish (Fig. 2H). Male cntd1^−/− zebrafish could induce female zebrafish to spawn but could only fertilize a small number of eggs due to sperm insufficiency. The fertilization rate of WT males was significantly higher than that of cntd1^−/− males (Fig. 2I). These findings demonstrate that Cntd1 deficiency leads to the arrest of spermatocytes in the metaphase of meiosis, resulting in only a small number of spermatocytes completing meiosis and forming spermatozoa.

Fig. 2.

Characterization of testis development and spermatogenesis. A and B, Anatomical views of the testes from wild-type (WT) (A) and cntd1^−/− (B) males, with the red dotted line highlighting the mutant testis. The scale bar represents 1 mm. C–F, Histological analysis of testes from WT (C and E) and cntd1^−/− (D and F) zebrafish, showing both a macroscopic view (C and D) and a magnified view (E and F) of the testes structure in WT and cntd1^−/− zebrafish. The arrows indicate spermatogonia, spermatids, and spermatozoa. Black circles mark cells arrested at meiotic metaphase I, red boxes highlight arrested cells with abnormal chromosome segregation, and the white box indicates cells arrested at meiotic metaphase II. Abbreviations: ST, spermatid; SZ, spermatozoon; SG, spermatogonia; PSC, primary spermatocyte; SSC, secondary spermatocyte; MI, meiotic metaphase I; MII, meiotic metaphase II. Scale bars in C–D and E–F represent 100 and 20 μm, respectively. (G) Statistical analysis of testicular cell composition in WT and cntd1^−/− zebrafish. (H) Statistical analysis of spermatocytes arrested at meiotic metaphases I and II. Each group in G and H contains three individuals for statistical analysis. (I) Statistical analysis of the fertilization rates in WT and cntd1^−/− males, with each group containing more than ten individuals for statistical purposes. Error bars represent the means ± standard deviations (SDs); “*,” “**,” and “***” asterisks above the error bar indicate statistically significant differences at 0.01 < P < 0.05, 0.001 < P < 0.01, and P < 0.001, respectively, as determined by a two-tailed Student's t-test.

Disruptions in meiosis have the potential to trigger apoptosis induced by checkpoints. Therefore, tunnel analysis was employed to detect apoptosis in the testes, revealing a higher level of apoptotic signals in cntd1^−/− zebrafish testes (Fig. 3A–E). Upon closer examination, apoptosis in WT zebrafish occurred exclusively in peripheric Sertoli cells surrounding the germ cells (Fig. 3A and C). In contrast, cntd1^−/− testes exhibited a significant number of apoptotic spermatocytes arrested in metaphase and early stages of anaphase, with apoptotic chromatin appearing in pairs (Fig. 3B,D and E). These findings suggest that spermatocytes arrested at metaphase or missegregated at anaphase ultimately undergo apoptosis. Despite the limited sperm production, flow cytometric analysis was employed to examine the ploidy of sperm cells in cntd1^−/− zebrafish. The produced spermatozoa were found to be reduced haploid, similar to WT zebrafish (Fig. 3F and G). Additionally, tetraploid and diploid primary and secondary spermatocytes were detected in the testicular suspension. Interestingly, a small peak before the haploid peak occasionally occurred in cntd1^−/− zebrafish (Fig. 3G), possibly attributed to the formation of apoptotic bodies resulting from spermatocytes undergoing apoptosis. When cntd1^−/− males were paired with WT females, a limited number of eggs were successfully fertilized, and the subsequent embryonic development of the progeny closely resembled that of WT embryos (Fig. 3H and J). Notably, the progeny originating from cntd1^−/− males exhibited a diploid chromosome configuration, akin to WT embryos (Fig. 3I and K). This suggests that even with the compromised spermatogenesis in cntd1^−/− males, there is still a capacity for the formation of small amounts of spermatozoa with chromatin correctly segregated at random, resulting in viable and diploid offspring.

Fig. 3.

Observations of spermatocyte apoptosis. A–D, TUNEL staining of testes from wild-type (WT) (A and C) and cntd1^−/− zebrafish (B and D). Each group consists of four individuals. White boxes highlight the magnified areas. Scale bars in A and B and C and D represent 100 and 10 μm, respectively. (E) Magnification of the area indicated by the white box. F and G, Representative cytometric histograms of testes from WT (F) and cntd1^−/− (G) zebrafish, with each group comprising three individuals. The terms 1C, 2C, and 4C denote haploidy, diploidy, and tetraploidy, respectively. H and J, Representative images of embryonic development in offspring obtained from cntd1^+/+ females crossed with cntd1^+/+ (H) and cntd1^−/− (J) males, with a scale bar representing 250 μm. I and K, Cytometric histograms of embryos obtained from cntd1^+/+ females crossed with cntd1^+/+ (I) and cntd1^−/− (K) males. Each group includes 20 individuals. The term 2C indicates diploidy.

These findings illustrate that the deficiency of Cntd1 leads to meiotic arrest and apoptosis in spermatocytes, culminating in a significant reduction in sperm production. Despite this, male cntd1^−/− individuals exhibit the capacity to generate a diminished but notable quantity of normal, haploid spermatozoa.

cntd1 Knockout Disrupts Meiotic Crossover Maturation and Chiasma Formation During Spermatogenesis

To investigate meiotic defects in cntd1^−/− zebrafish, testicular chromosome spreads were meticulously prepared to scrutinize the behavior of spermatocyte chromosomes during meiosis. The chromatin morphology of cntd1^−/− spermatocytes mirrored that of WT zebrafish from zygotene to pachytene (Fig. 4A–F). However, distinctive chiasma structures, such as “∝” and “∞” at the diplotene stage, and cross-linked homologous chromosomes forming shapes like “十,” “○,” “—,” and “V” at the diakinesis stage were usually absent in cntd1^−/− zebrafish, while evident in their WT counterparts (Fig. 4G–J). Consequently, spermatocyte chromosomes typically manifested as bivalents during metaphase in WT zebrafish, whereas metaphasic chromosomes in cntd1^−/− zebrafish displayed numerous univalents (Fig. 4K and L). Statistical analysis revealed a significant increase in the number of univalents at metaphase I in cntd1^−/− zebrafish compared to WT zebrafish (Fig. 4M). Given that chiasma serves as the cellular basis for meiotic crossovers, it was inferred that cntd1 knockout disrupts the process of meiotic crossover maturation in meiosis I. Subsequently, immunofluorescence was conducted with antibodies against Sycp3 and Mlh1 to analyze the formation of crossovers at the pachytene stage during meiosis. The entire length of homologous chromosomes was indicated by the Sycp3 signals (Fig. 4N and P). The location of MutLγ at designated crossover sites was delineated using an anti-Mlh1 antibody. Notably, the chromosomes of WT zebrafish showed an enrichment of the Mlh1 signals (Fig. 4O and Q). Statistical analysis revealed a significant decrease in the accumulation of Mlh1 signals on chromosomes at the pachytene stage in cntd1^−/− zebrafish (Fig. 4R). These results strongly suggest that Cntd1 plays a pivotal role in the generation of meiotic crossovers and the formation of chiasma during meiosis I in zebrafish.

Fig. 4.

Observations of meiotic chromosome behavior during spermatogenesis. A and B, Spermatocytes at the zygotene stage in wild-type (WT) (A) and cntd1^−/− (B) zebrafish. C and D, Spermatocytes at the early pachytene stage in WT (C) and cntd1^−/− (D) zebrafish with “E” indicating “early.” E and F, Spermatocytes at the late pachytene stage in WT (E) and cntd1^−/− (F) zebrafish with “L” indicating “late.” G and H, Spermatocytes at the diplotene stage in WT (G) and cntd1^−/− (H) zebrafish. I and J, Spermatocytes at the diakinesis stage in WT (I) and cntd1^−/− (J) zebrafish. K and L, Spermatocytes at the metaphase I stage in WT (K) and cntd1^−/− (L) zebrafish. The scale bar in A–L represents 10 μm. (M) Statistical analysis of univalent at the metaphase I stage in WT and cntd1^−/− zebrafish. For this analysis, two testes isolated from the WT and cntd1^−/− males were collected as one sample, and three such samples were used. (N and O) Immunofluorescent staining of WT zebrafish spermatocytes with antibodies against Sycp3 (N) and Mlh1 (O). P and Q, Immunofluorescent staining of cntd1^−/− zebrafish spermatocytes with antibodies against Sycp3 (P) and Mlh1 (Q). The scale bar in N–Q represents 10 μm. R, Statistical analysis of Mlh1 signals in WT and cntd1^−/− zebrafish, with two testes collected as one sample and three such samples used for the analysis. (S) Hierarchical clustering of meiosis-related genes based on transcriptome analysis. The error bars represent the means ± standard deviations (SDs); “**” and “***” above the error bar indicate statistically significant differences at 0.001 < P < 0.01 and P < 0.001, respectively, as determined by a two-tailed Student's t-test.

To further explore this, the transcriptional levels of specific genes associated with meiosis were scrutinized in cntd1^−/− and WT males. These genes encompassed those involved in the formation of DNA double-strand breaks (spo11 and dmc1), synaptonemal complexes (sycp1, sycp2, and sycp3), and meiotic crossovers (mlh1, mlh3, and mus81). The results of qPCR revealed no statistically significant variance in the expression levels of dmc1, sycp1, sycp2, and sycp3, as well as mlh1 and mlh3, between WT and cntd1^−/− zebrafish (supplementary fig. S2A–F, Supplementary Material online). However, the expression of spo11 and mus81 was significantly downregulated in cntd1^−/− zebrafish compared to WT zebrafish (supplementary fig. S2A,C and F, Supplementary Material online). The genes associated with meiosis were categorized into three groups based on transcriptome analysis (supplementary fig. S3A–C, Supplementary Material online). Analysis of the whole testis transcriptome revealed that 9 genes associated with homolog pairing and synapsis in prophase I and 12 genes involved in recombination and crossover formation exhibited differential expression between WT and cntd1^−/− zebrafish (supplementary fig. S4S and A–C, Supplementary Material online). Among these differentially expressed genes, six members, namely, fkbp6, scml2, prdm9, cntd1, mus81, and slx4, demonstrated downregulation, while the expression of other genes appeared to be upregulated (Fig. 4S). Given the observed increase in the number of spermatocytes in cntd1^−/−, it is plausible to assume that these downregulated genes are more reliable as a direct result of Cntd1 deficiency. The specific information pertaining to the mentioned relevant genes can be found in supplementary table S1, Supplementary Material online.

cntd1 Knockout Alters the Ploidy of Eggs Rather than Impairs Ovary Development in Female Zebrafish

To explore the consequences of Cntd1 deficiency on female zebrafish, a thorough histological study of ovarian development was conducted. At 30 dpf, both WT and cntd1^−/− zebrafish exhibited undifferentiated gonads, bipotential gonads, and early ovaries (supplementary fig. S4A–F, Supplementary Material online). By 50 dpf, both WT and cntd1^−/− zebrafish showcased the presence of testes and ovaries (supplementary fig. S4G–J, Supplementary Material online). Notably, the early-stage ovaries observed in cntd1^−/− zebrafish displayed a morphology indistinguishable from that of WT zebrafish. Upon reaching sexual maturity, anatomical examinations of mature ovaries revealed no discernible disparities in appearance among WT, cntd1^+/−, and cntd1^−/− zebrafish, with all ovaries being well-developed and filled with follicles (Fig. 5A–C). Histological sections stained with HE further elucidated a comparable composition of ovarian cells in the three different genotypes of zebrafish, encompassing oocytes at various stages, including primary growth (PG), previtellogenic (PV), early vitellogenic (EV), midvitellogenic (MV), and full-grown (FG) stages (Fig. 5D–I). Statistical analysis revealed a prevalence of oocytes at the PG and PV stages, with other oocytes in the minority across the ovaries of the three different female zebrafish genotypes (supplementary fig. S4K, Supplementary Material online). Egg production among the three genotypic zebrafish displayed no significant differences, with each female fish producing an average of 300 eggs per spawning event (supplementary fig. S4L, Supplementary Material online). When mated with WT males, eggs from the three genotypic female zebrafish exhibited similar fertilization rates, averaging around 80% (supplementary fig. S4M, Supplementary Material online). These results unequivocally demonstrate that cntd1 knockout does not impair ovarian development or oogenesis.

Fig. 5.

Observations of ovarian development and reproduction. A–C, Anatomical views of the ovaries from wild-type (WT) (A), cntd1^+/− (B), and cntd1^−/− (C) females, with each group consisting of three individuals. The scale bar in A–C represents 1 mm. D–I, Histological observations of the ovaries from WT (D and G), cntd1^+/− (E and H), and cntd1^−/− (F and I) females. Both a holistic view (D–F) and a magnified view (G–I) of the ovarian section are provided. White boxes highlight the magnified areas. Scale bars in D–F and G–I represent 0.5 and 200 μm, respectively. J–M, Embryos obtained from self-crossing of WT zebrafish (J) and cntd1^−/− females crossed with cntd1^−/− (K), cntd1^+/− (L), and cntd1^+/+ (M) males at 6 h postfertilization (hpf). N–Q, Embryos obtained from self-crossing of WT zebrafish (N) and cntd1^−/− females crossed with cntd1^−/− (O), cntd1^+/− (P), and cntd1^+/+ (Q) males at 24 hpf. R–U, Embryos obtained from self-crossing of WT zebrafish (R) and cntd1^−/− females crossed with cntd1^−/− (S), cntd1^+/− (T), and cntd1^+/+ (U) males at 72 hpf. The scale bar in J–U represents 1 mm.

However, when cntd1^−/− females were mated with WT, cntd1^+/−, and cntd1^−/− males, respectively, the survival rates of all offspring were extremely low. Most embryos resulting from these crosses exhibited malformed and lethal development (Fig. 5J–U). Offspring from cntd1^−/− female zebrafish typically showed no defects in development at 6 hpf (Fig. 5K–M) but exhibited obvious malformations at 24 hpf (Fig. 5O–Q). Upon hatching, various malformations, such as tail loss, small head, lack of eyes, edema, and curved body, were observed in the offspring at 72 hpf (Fig. 5S–U). These malformed embryos gradually died during larval development, whereas the WT offspring remained consistent and robust (Fig. 5R). According to the survival curve, the majority of WT embryos could survive, with only a few experiencing mortality during the period of feed conversion (supplementary fig. S5A, Supplementary Material online). Conversely, embryos produced by cntd1^−/− females had significantly lower survival rates, with only a few managing to survive beyond 10 dpf (supplementary fig. S5B–D, Supplementary Material online). When cntd1^−/− females mated with cntd1^−/− males, both the fertilization rate of eggs and the survival rate of embryos were notably low (supplementary fig. S5D, Supplementary Material online). When cntd1^−/− females mated with WT or cntd1^+/− males, the offspring exhibited a high fertilization rate but low survival rates (supplementary fig. S5B and C, Supplementary Material online). Typically, the embryos thrived well before 4 dpf, after which a substantial number of malformed embryos perished at 4 or 5 dpf, leading to a final survival rate of less than 10% (supplementary fig. S5B–D, Supplementary Material online). These findings indicate the presence of maternal defects in the eggs produced by cntd1^−/− females.

Given the observed homologous chromosome missegregation in male zebrafish due to Cntd1 deficiency, it was hypothesized that the ploidy of eggs produced by cntd1^−/− females might be altered. As determining the ploidy of eggs directly proved challenging, the ploidy of the offspring from cntd1^−/− females was examined instead. Results indicated that WT embryos exhibited diploid ploidy, with a DNA fluorescence value per cell of approximately 100 (supplementary fig. S6A–C and M, Supplementary Material online). However, the fluorescence values of embryos derived from cntd1^−/− females mating with different genotypes of male zebrafish fluctuated widely, ranging from about 50 to around 150, suggesting the production of diploid, triploid, and other aneuploid offspring by cntd1^−/− females (supplementary fig. S6D–L and N–P, Supplementary Material online). Additionally, chromosome spreads of embryos resulting from crosses between cntd1^−/− females and WT males revealed different chromosome numbers in different individuals, indicating the presence of haploid, diploid, triploid, and aneuploid offspring (supplementary fig. S6Q–U, Supplementary Material online). Consequently, gametes produced by cntd1^−/− females appeared to be haploid, diploid, aneuploid, or even lacking chromatin entirely. These observations underscored that cntd1 knockout disrupts the process of sorting chromosomes during meiotic division, leading to the frequent occurrence of unnatural ploidies and, consequently, a drastic reduction in the proportion of normal, haploid sperm cells and oocytes as well as the occurrence of malformations and low survival rates in zebrafish.

Cntd1 Deficiency Results in Diploid and Triploid Progeny in Diploid Females

Despite the high mortality in the progeny from cntd1^−/− females, a few offspring managed to survive to adulthood. These surviving progenies comprised diploid and triploid individuals. When cntd1^−/− females mated with cntd1^−/− males, both diploid and triploid homozygous progenies included male and female individuals (Fig. 6A–D). Flow cytometry analysis of 122 offspring demonstrated that 72 individuals were diploid, while the remaining 50 individuals were triploid, with somatic fluorescence values of 100 and 150 per cell, respectively (Fig. 6E and F). Furthermore, 20 out of 72 diploid- (cntd1^−/−) and 6 out of 50 triploid (cntd1^−/−/−) homozygotes were female (Fig. 6A and B). When cntd1^−/− females mated with cntd1^+/+ males, the diploid progenies also included both male and female populations, while the triploid progenies were exclusively male. Collectively, 10 out of 34 diploid heterozygotes (cntd1^+/−) were female, whereas none of the 59 heterozygous triploids (cntd1^−/−/+) developed into females (Fig. 6G–I). Among the 93 tails of progenies resulting from crosses between cntd1^−/− females and cntd1^+/+ males, 34 and 59 tails were identified as diploid and triploid heterozygotes, respectively (Fig. 6J and K). Upon closer examination, it became evident that diploid males exhibited haploid sperm production, while triploid males showed a reduction in aneuploid sperm production (Fig. 6L–O). Remarkably, all these male progenies retained the ability to induce cntd1^+/+ females to spawn and fertilize eggs. The cntd1^−/− males could only fertilize a tiny number of eggs, mostly less than 20%, resulting in low survival rates (Fig. 6P). The cntd1^−/−/− males fertilized the fewest eggs, which could be attributed to the combined impact of triploidy and homozygous mutation. The obtained offspring gradually died during embryonic development, with none surviving beyond 11 d (Fig. 6Q). In contrast, the progeny from cntd1^+/− males displayed relatively high survival rates, ranging from about 80% to 90% (Fig. 6R). The fertilization rates of cntd1^−/−/+ males were similar to those of cntd1^−/− males, and the resulting offspring also experienced gradual demise during embryonic development, and none of them survived beyond 12 dpf (supplementary fig. S6S, Supplementary Material online). The observed inviability of the progeny generated by triploid males underscores the inherent challenges faced by aneuploid offspring in achieving successful development.

Fig. 6.

Generation of diploid and triploid progeny by cntd1^−/− females. A–D, Adult offspring resulting from the self-crossing of cntd1^−/− zebrafish, including diploid females (A), males (C), and triploid females (B) and males (D). E and F, Representative flow cytometry results of red blood cells from cntd1^−/− (E) and cntd1^−/−/− (F) zebrafish. G–I, Adult offspring obtained from the cross between female cntd1^−/− and cntd1^+/+ males, including diploid females (G), males (H), and triploid males (I). J and K, Representative flow cytometry results of red blood cells from cntd1^+/− (J) and cntd1^−/−/+ (K) zebrafish. L–O, Representative flow cytometry results of testicular cells from cntd1^−/− (L) and cntd1^−/−/− (M) zebrafish, as well as cntd1^+/− (N) and cntd1^−/−/+ (O) zebrafish. The terms 1C, 1.5C, 2C, 3C, and 6C denote haploidy, aneuploidy, diploidy, triploidy, and hexaploidy, respectively. P–S, Survival rates of the offspring normalized to the total number of spawning eggs obtained from cntd1^+/+ females crossed with cntd1^−/− (P) and cntd1^−/−/− (Q) males, as well as cntd1^+/− (R) and cntd1^−/−/+ (S) males.

Cntd1 Deficiency Results in Diploid and Tetraploid Progeny in Triploid Females

Upon further observation, it was noted that cntd1^−/−/− females retained the ability to spawn, and their eggs exhibited normal fertilization capabilities. However, similar to the cntd1^−/− females’ offspring, the embryos produced by cntd1^−/−/− females displayed various malformations during embryonic development (supplementary fig. S7A–D, Supplementary Material online). These malformed embryos exhibited continuous mortality until 10 dpf, resulting in minimal surviving embryos and a survival rate of less than 5% (supplementary fig. S7E–H, Supplementary Material online). Flow cytometry analysis revealed wide variations in the ploidies of offspring from cntd1^−/−/− females, with DNA fluorescence values per cell distributed mainly from about 50 to 200, while diploid embryos exhibited concentrated values around 100 (supplementary fig. S7I–L, Supplementary Material online). Additionally, chromosome spreads were prepared to observe the chromosome numbers of the progenies resulting from the cross between cntd1^−/−/− females and diploid WT males, revealing a range of chromosome numbers from 43 to 88 in the checked population (supplementary fig. S7M–R, Supplementary Material online). These findings suggest a significant alteration in the ploidy of eggs from cntd1^−/−/− females.

Given the limited number of cntd1^−/−/− females and the low survival rate of their offspring, only 15 adult survivors have been obtained so far. Among these survivors, only one was identified as diploid, while the others were confirmed to be tetraploid males (Fig. 7A and B). The DNA fluorescence value per cell was 100 for diploid progeny and 200 for tetraploid progeny (Fig. 7C and D). Consequently, the fluorescence values of their gametes detected by flow cytometry were 50 and 100, respectively (Fig. 7E and F). When tetraploid-heterozygous (cntd1^{−/−/−/+}) males mated with WT females, they successfully induced the females to spawn. The resulting offspring developed normally without embryonic malformations or mortality and were indistinguishable in appearance from diploid progenies (Fig. 7G and H). Flow cytometry analysis revealed that these progenies were all triploid, with a DNA fluorescence value of 150 (Fig. 7I and J). Observation of chromosome spreads further confirmed these triploid progenies with 75 pairs of chromosomes (Fig. 7K and L). As adults, all these triploid progenies turned out to be male (supplementary fig. S8A, Supplementary Material online). Considering the special sexual determination dependent on the numbers of primordial germ cells (PGCs) at the period of gender differentiation in zebrafish, a p53 mutation was introduced in the tetraploids, resulting in the generation of female cntd1^{−/−/−/+}P53^−/+/+/+ zebrafish. These females could produce tetraploid offspring when mated with tetraploid males (supplementary fig. S8B, Supplementary Material online), indicating that cntd1^{−/−/−/+}P53^−/+/+/+ females could produce reduced-diploid eggs. The obtained tetraploid offspring survived well and exhibited normal development (supplementary fig. S8C–F, Supplementary Material online). These results strongly demonstrate that both female and male cntd1^{−/−/−/+} zebrafish can produce reduced-diploid gametes.

Fig. 7.

Generation of tetraploid zebrafish by cntd1^−/−/− females. A and B, Adult progeny obtained from cntd1^−/−/− females crossed with cntd1^+/+ males, showcasing both diploid (A) and tetraploid (B) offspring. C and D, Representative flow cytometry results of red blood cells from diploid (C) and tetraploid (D) offspring. E and F, Representative flow cytometry results of testicular cells from diploid (E) and tetraploid (F) zebrafish. The terms 1C, 2C, and 4C indicate haploidy, diploidy, and tetraploidy, respectively. G and H, Images of progeny obtained from wild-type (WT) females crossed with diploid (G) and tetraploid (H) males at 72 h postfertilization (hpf), with a scale bar representing 1 mm. I and J, Representative flow cytometry results of progeny from diploid (I) and tetraploid (J) males crossed with WT females. The terms 2C and 3C indicate diploidy and triploidy, respectively. K and L, Chromosome spreads of embryos from WT females crossed with diploid (K) and tetraploid (L) males, with a scale bar representing 10 μm. M, Formation of different polyploids in zebrafish based on cntd1 knockout. N, A potential model of polyploidization in vertebrates due to naturally induced defects in crossover formation.

In summary, we have developed a novel method for generating different polyploids in zebrafish by disrupting the cntd1 gene, resulting in the capacity to produce unreduced eggs (Fig. 7M). Based on the findings of this study, we propose a potential model for polyploidy formation in fish that may have broader implications for all vertebrates (Fig. 7N). In this proposed model, mutant homozygotes with crossover defects arising from natural mutations can produce unreduced eggs, leading to the sequential production of triploids and tetraploids. Once cntd1^{−/−/−/+} individuals are established, they can develop into a tetraploid population through self-crossing or produce a triploid population through mating with diploids (Fig. 7N).

Discussion

Polyploidization is recognized for its potential to facilitate specific adaptation in response to changing environmental conditions or mitigate the risk of extinction during environmental upheavals (Van de Peer et al. 2017). In the context of aquaculture, polyploids often exhibit economically advantageous traits, including accelerated growth, improved quality, and enhanced disease resistance (Liu 2010). Although the production of unreduced gametes is a well-documented route for polyploid formation, even in hybrids (Ramsey and Schemske 2002; Brownfield and Köhler 2011; Mason and Pires 2015), the molecular mechanisms underlying this process, particularly in animals, remain poorly understood. Polyploidy is prevalent in both plants and fish, often arising from hybridization between different species (Liu 2010; Stull et al. 2023). Notably, polyploid offspring commonly emerge from hybrid fish with compromised fecundity rather than those with normal reproductive capacity (Liu et al. 2001; Qin et al. 2014; Hu et al. 2019; Liu et al. 2020; Wang et al. 2020; Gong et al. 2022), suggesting a potential role of meiotic defects in fish polyploidization. In this study, the effective knockout of cntd1 was confirmed by the substantial downregulation of cntd1 transcripts (Fig. 1). This transcript downregulation is often associated with nonsense-mediated mRNA decay (Kurosaki et al. 2019; Cheruiyot et al. 2021). Remarkably, both female and male cntd1^−/− zebrafish were not entirely sterile, but sex-specific differences in the effects of Cntd1 deficiency were observed. Similar sexual dimorphism in tolerance for meiotic defects has been documented in both plants and mammals, with females typically exhibiting fewer defects (Brownfield and Köhler 2011; Hua and Liu 2021). Comparing our findings with studies on cntd1 knockout in mice (Holloway et al. 2014; Gray et al. 2020), a crucial revelation in this study is the acquisition of the ability to produce unreduced eggs in cntd1 mutant females due to meiotic crossover defects resulting from Cntd1 deficiency (Figs. 6 and 7). This phenomenon serves as the foundation for polyploidy formation. The study presents direct evidence that a single gene mutation can trigger the sequential generation of triploid and tetraploid individuals as well as triploid and tetraploid populations through natural mating. This discovery holds significant implications for evolutionary studies and fisheries research.

Meiotic defects in fish typically lead to infertility due to the absence of normal gametes (Gervai et al. 1980; Carrasco et al. 1998; Saito et al. 2011; Takemoto et al. 2020; Dai et al. 2021). Specifically, disruptions in meiotic crossovers (COs), vital for accurate segregation of homologous chromosomes during meiosis I (Pazhayam et al. 2021), can result in aneuploid gametes and unreduced eggs in zebrafish (Feitsma et al. 2007). The increased number of univalents, reduced Mlh1 signals, fewer chiasmas during diplotene and diakinesis, and a significant rise in univalents at metaphase in cntd1^−/− zebrafish (Fig. 4) strongly indicate impaired crossover formation. Notably, unlike zebrafish mutants lacking key proteins in meiotic synapsis, such as Sycp2 and Sycp3, causing spermatocyte arrest before the diplotene stage (Takemoto et al. 2020; Pan et al. 2022), Cntd1-deficient zebrafish exhibited enriched metaphase-stage spermatocytes (Fig. 2), consistent with mlh1 mutants showing crossover formation defects (Feitsma et al. 2007). These findings suggest that defective crossover formation in cntd1^−/− zebrafish led to the production of unreduced eggs, setting the stage for polyploidy formation. To delve deeper into the role of Cntd1 in meiotic crossover formation, we performed a transcriptional analysis, given the unavailability of specific antibodies for direct protein detection. However, it is crucial to consider the increased number of spermatocytes in cntd1^−/− males. The higher transcriptional levels observed for meiosis-related genes in the transcriptome analysis may be attributed to the increased number of spermatocytes rather than genuine upregulation. This is supported by the disparity between sycp1 levels in the transcriptome and qPCR analyses. Nevertheless, the downregulated genes identified are deemed credible although the degree of downregulation may be mitigated. This is further substantiated by the consistent reduction in the expression of mus81 observed in both transcriptome and qPCR analyses (Fig. 4 and supplementary fig. S3, Supplementary Material online). Intriguingly, both cntd1 and mus81 have been confirmed to regulate meiotic crossover formation in yeast and mammals (Hollingsworth and Brill 2004; Holloway et al. 2008; Yokoo et al. 2012; Holloway et al. 2014). However, whether reciprocal relationships exist between them warrants further investigation. To address this knowledge gap, more advanced techniques such as single-cell transcriptome sequencing, co-immunoprecipitation, and immunofluorescence, coupled with additional meiosis-related antibodies, are warranted for a comprehensive exploration of the potential interaction of Cntd1 in fish meiosis.

Polyploidy is generally considered less common in animals compared to plants, especially in mammals and birds, where it is relatively rare (Gregory and Mable 2005). The prevailing explanation for the rarity of animal polyploids revolves around their interference with the development and reproductive abilities of one sex, often due to chromosome-dependent sex determination (Wertheim et al. 2013). However, contrasting findings regarding meiotic defects in females between mice and zebrafish introduce an alternative perspective on this phenomenon. Despite shared features such as defective crossover formation, poorly developed testes, and impaired sperm production in Cntd1 mutant mice and zebrafish (Fig. 2), they exhibit different outcomes in females concerning meiotic defects. While meiotic defects in mammal females are known to cause infertility, little is understood about why these females fail to produce litters despite having well-developed ovaries indistinguishable from those of WT animals (Edelmann et al. 1996; Holloway et al. 2014). The preservation of fertility in cntd1 mutants and the viability of polyploid offspring in zebrafish appear to stem from their high capacities for gametes and offspring production. Consequently, the extensive polyploidy observed in fish may depend on their high fecundity, a trait that is not shared by birds or mammals.

In fish, polyploidization can be induced through artificial means, employing techniques such as temperature shock and hydrostatic pressure. This method of artificial induction yields triploid or tetraploid fish by interfering with processes like the extrusion of the second polar body or the initiation of the first zygotic mitosis. This approach has been explored in various species (Peruzzi and Chatain 2003; Nam et al. 2004; Hamasaki et al. 2013; Goo et al. 2015; Baars et al. 2016; Zhu et al. 2017; Hassan et al. 2018; Glover et al. 2020; Káldy et al. 2021). Although stressors such as high or low temperatures and hydrostatic pressure can occur naturally, achieving the precise timing required for polyploid induction under natural conditions presents challenges and demands meticulous control. For instance, in zebrafish, it necessitates approximately 2 min around the extrusion of the second polar body at 2 min postfertilization or the commencement of the first zygotic mitosis at 18 min postfertilization (Menon and Nair 2019). While polyploid individuals induced by meiotic stressors may opportunistically arise in nature, their ability to establish a stable lineage in the short term is limited due to potential mismatches between inducing environmental conditions and the survival requirements of polyploids. These neopolyploids may even be susceptible to extinction due to deleterious effects triggered by polyploidy, such as disruptions in nuclear and cell enlargement, difficulties in mitosis and meiosis, as well as epigenetic and genomic instability (Comai 2005). In contrast, the proposed polyploidization in this study can potentially benefit from a continuous genetic basis, thereby increasing the likelihood of stabilizing the polyploidy state.

The emergence of newly formed polyploids often encounter with challenges in meiotic chromosome segregation, as homologs tend to form multivalents during meiosis (Comai 2005; Lloyd and Bomblies 2016). In contrast, natural polyploids seem to have evolved effective mechanisms, showcasing high fertility. In plants, the early stages of prophase I witness the intriguing phenomenon of parallel pairing of four homologs, followed by the formation of synaptic connections between multiple homologs (Khazanehdari et al. 1995; Stack and Roelofs 1996; Higgins et al. 2014). Despite these initial associations, they tend to disassemble before reaching metaphase I, leading to the formation of pairs of bivalents in the natural tetraploid A. arenosa. This disassembly is attributed to strong crossover interference, resulting in the dissolution of associated recombination and the suppression of recombination through robust selection on genes influencing crossover rates (Jones et al. 1996; Carvalho et al. 2010; Yant et al. 2013). Consequently, reduced crossover rates are prevalent in evolved tetraploids compared to their diploid progenitors across various natural species (Yant et al. 2013).

Therefore, a strategic approach for neopolyploids to mitigate the prevalence of multivalents is to reduce crossover rates to approximately one per chromosome. Intriguingly, the polyploidization process in this study benefits from the diminished crossover rates due to Cntd1 deficiency (Fig. 4). Thus, the genetic background of cntd1 knockout may confer advantages in preventing multivalent formation and enhancing fecundity in this type of tetraploid zebrafish. Notably, both female and male cntd1^{−/−/−/+} tetraploids produced reduced-diploid gametes, and the crosses between them yielded tetraploid offspring with normal development (supplementary fig. S8, Supplementary Material online), thereby contributing to the maintenance of tetraploidy. The gametogenesis of these tetraploid zebrafish aligns with that of tetraploid fish lineages formed through hybridization (Wang et al. 2019).

In comparison to diploid organisms, tetraploids confer a significant advantage in terms of gene redundancy, acting as a protective shield against the deleterious effects of mutations by masking recessive alleles through dominant WT alleles. During the gametophytic stage, the vulnerability of haploid gametes from diploids to detrimental loss-of-function mutations is mitigated by diploid gametes from tetraploids, which can camouflage the impact of deleterious recessive alleles with additional wild-type alleles (Immler 2019; Spoelhof et al. 2021). Furthermore, at the zygotic stage, polyploidy reduces the occurrence of homozygous recessive alleles (Stadler 1929; Mable and Otto 2001). This gene buffering effect serves as a defense against deleterious recessive mutations and genotoxicity, particularly in situations where isolated and bottlenecked populations resort to inbreeding. Moreover, gene redundancy empowers the diversification of gene function by modifying redundant copies of crucial or essential genes, acting as a driving force for evolution in response to rapidly changing environments (Comai 2005). In contrast, the corresponding ability in diploids relies on the fortuitous occurrence of rare segmental duplication events (Prince and Pickett 2002; Moore and Purugganan 2005). Additionally, the heightened genetic variation and genomic alterations facilitated by changes in gene expression and epigenetic remodeling have the potential to influence the morphology, physiology, and ecology of newly formed polyploids, contributing to their survival and eventual thriving in the short term (Soltis et al. 2014; Schoenfelder and Fox 2015; Van de Peer et al. 2017).

In summary, our current study employing the cntd1 knockout model provides compelling evidence that defective formation of meiotic crossovers can lead to distinct polyploidization in fish. Notably, our research introduces a novel model of polyploidization with potential applicability across vertebrates.

Materials and Methods

Zebrafish

Zebrafish (AB strain) were housed in a controlled environment with a 14 h light and 10 h dark cycle at a temperature of 28.5 °C, utilizing a circulated water system. Their diet consisted of newly hatched brine shrimp. Before dissection, all fish underwent deep anesthesia with 0.016% tricaine (Sigma-Aldrich, St. Louis, MO), adhering to the Guiding Principles for the Care and Use of Laboratory Animals. Approval for all experiments was obtained from the Animal Care Committee of Hunan Normal University.

Target Knockout of cntd1

The knockout of cntd1 was executed through the CRISPR/Cas9 strategy. The sequence of zebrafish cntd1 was retrieved from the NCBI database, and, following the CRISPR/Cas9 target design rules, the knockout target was specifically chosen to target the second exon of cntd1. The target sequence within the second exon of cntd1 was as follows: GGCCTGATATGCGACCACA. To synthesize the gRNA, the TranscriptAid T7 High-Yield Transcription Kits (K0441, Thermo Scientific Fermentas, Waltham, MA) were utilized. Similarly, the mMESSAGE mMACHINE T3 Transcription Kit (AM1348 Ambion, Austin, TX, USA) was employed to synthesize Cas9 mRNA. Prior to microinjection, gRNA and Cas9 mRNA were mixed at final concentrations of 50 and 100 ng/μL, respectively, and then injected into one-cell stage embryos. The targeted regions were subsequently amplified using the primers listed in supplementary table S2, Supplementary Material online and 2× Rapid Taq Master Mix (P222-01, Vazyme, China) with a conventional PCR program, resulting in PCR products of 460 base pairs. The accuracy of the target sequence was confirmed by sequencing the PCR products with the forward primer. To identify effective F0 populations, the genotypes of injected embryos were detected. Heterozygotes were discerned from the F1 population generated by crosses between F0 mutants and WT zebrafish, while homozygotes were identified through the inbreeding of F1 heterozygotes.

RNA Extraction and RT-PCR

Total RNA was extracted from various developmental stages, including embryos and juveniles, as well as diverse adult tissues such as eyes, brain, gill, spleen, kidney, heart, liver, intestine, testis, ovary, and muscle tissues. The TRIzol reagent (Invitrogen, Carlsbad, CA) was employed for the isolation of RNA. The tissues were fully ground with ceramic beads in 1 mL TRIzol reagent, and the total RNA was extracted using the phenol/chloroform extraction method. Subsequently, a cDNA synthesis kit (MR05201S, Monard) was utilized to synthesize cDNA following the standard procedure with RNA templates at a concentration of 1 μg. The primers utilized for reverse transcription polymerase chain reaction (RT-PCR) are detailed in supplementary table S2, Supplementary Material online.

Quantitative PCR

Quantitative real-time PCR primers were devised using the primer blast tool from the National Biotechnology Information Center (NCBI) website. Details for cntd1, dmc1, spo11, sycp1, sycp2, sycp3, mlh1, mlh3, and mus81 are outlined in supplementary table S1, Supplementary Material online, with the specific quantitative PCR primers for these genes provided in supplementary table S2, Supplementary Material online. The SYBR Green-based quantitative PCR (qPCR) assay was conducted in a 20 μL amplification reaction system, including 10 μL of SYBR Green real-time PCR master mix, 1 μL each of forward and reverse primers, and an additional 2 μL of cDNA template. Referring to a prior study (McCurley and Callard 2008), the general reference gene ef1a was selected as the internal reference in one set of quantitative experiments. Acknowledging the distinct composition of germ cells in WT and cntd1^−/− testes, the spermatocyte-specific gene sycp2 was specifically chosen as the internal reference in another qPCR experiment. The relative expression of each gene in cntd1^−/− zebrafish was calculated relative to WT zebrafish. All qPCR experiments were executed on an ABI real-time thermocycler (QuantStudio 5, USA).

Transcriptome Analyses

Four testes, isolated from males, were combined as one sample and promptly placed on dry ice after isolation. For both the WT and cntd1 mutant groups, three samples were utilized for transcriptome analysis. Total RNA extraction was carried out using TRIzol reagent (Ambion, 15596, USA). RNA-seq was performed using a DNBSEQ-T7 system. The resulting clean reads in WT_1-3 and cntd1_1-3 ranged from 40 932 124 to 81 752 212. These clean reads were aligned to the Danio rerio genome (GRCz11) using HISAT2 (version 2.2.4, http://daehwankimlab.github.io/hisat2/). The heatmap for candidate genes was generated with GraphPad Prism 8 based on the RPKM values. Details for relevant genes are outlined in supplementary table S1, Supplementary Material online. Additional details regarding the transcriptome analysis can be found in supplementary tables S3 and S4, Supplementary Material online.

HE Staining

The gonads were isolated intact, fixed overnight in Bouin's solution, and subsequently stored in 70% ethanol. Following dehydration in an ascending ethanol gradient (70%, 80%, 90%, 95%, and 100%) in water, the samples underwent treatment with various ratios of xylene/ethanol until achieving transparency. Subsequently, they were fully immersed in wax, embedded in paraffin, and sliced into 5-μm-thick sections. These sections were dewaxed by immersion in xylene, rehydrated in a decreasing ethanol gradient (100%, 95%, 90%, 80%, and 70%) in water, and then stained with hematoxylin and eosin. After another round of dehydration, the sections were promptly sealed. Finally, the prepared sections were observed using a Leica DM2500 microscope (Leica, Germany).

TUNEL Staining

The testes were meticulously isolated from WT and cntd1^−/− males, fixed in 4% polyformaldehyde, and subsequently embedded in an optimal cutting temperature compound (OCT). Cryosections of 5 μm thickness were obtained. TUNEL staining was carried out using the In Situ Cell Death Detection Kit, TMR red (12156792910, Roche), following the standard instructions provided with kit. The observations were made using a fluorescence microscope (Leica, Germany).

Flow Cytometry Analysis

Testes preparation: The testes of both WT and cntd1^−/− zebrafish were dissected and placed in a 1.5 mL Eppendorf tube containing 100 μL PBS. After mincing with scissors, the resulting testicular cell suspension was supplemented with 300 μL PBS and stained with 300 μL DAPI solution for 10 min in the dark.

Embryos preparation: Embryos at 4 dpf were placed into 1.5 mL EP tubes, with one embryo per tube. The embryo was then added to 50 μL PBS and 5 μL ACD, fully minced. Subsequently, the mince was added with 300 μL PBS and 300 μL DAPI staining solution, well-mixed, and stained in the dark for 10 min.

Tail fin preparation: The tail fin was amputated from the zebrafish and immersed in a 1.5 mL EP tube containing 100 μL ACD. After mincing with scissors, the mince was rinsed with 300 μL PBS and 300 μL DAPI staining solution, well-mixed, and stained for 10 min in the dark. All cell suspensions were filtered through a 30 μm mesh to remove large debris before being assayed by flow cytometry (Partec, Germany).

Spermatocyte Chromosome Spreads

The testes of both WT and cntd1^−/− zebrafish were isolated and placed in glass dishes with a small amount of PBS. Each testis was carefully divided into three sections and rinsed with PBS. Subsequently, the testis sections were transferred to 1.5 mL EP tubes, treated with 1 mL of 0.075 mol/L KCL solution, and maintained in a 28 °C thermostat for 4 h. Following this, the testis sections were transferred to a new EP tube, treated with 1 mL of precooled Carnot's fixative (methanol:acetic acid = 3:1), and fixed in a −20 °C refrigerator for 1 h. Postfixation, the sample was supplemented with 100 μL of precooled, newly prepared fixative (methanol: acetic acid = 1:1) and finely minced with scissors for approximately 5 min. The resulting testicular suspension was then dropped vertically from a height of about 40 cm onto frozen slides, and the slides were promptly flame-baked. Once dry, the slides were stained with Giemsa stain (liquid A:liquid B = 1:10) for 30 min, followed by rinsing with running water from the back of the slides. Finally, after the slides were dried, the chromosomes were observed under a microscope (Leica, Germany) at 100× magnification.

Embryonic Chromosome Spreads

Embryos at 4 dpf were cultured in a Petri dish with colchicine (100 μg/mL) at 28 °C for 3 h in the dark. Following colchicine treatment, the Petri dish was supplemented with 500 μL of prewarmed hypotonic solution (1.1% sodium citrate solution) and incubated in a thermostat for 1.5 h. Posthypotonic treatment, the embryos were immersed in a precooled, newly prepared fixative and kept in the refrigerator at −20 °C for 1 h. After fixation, a single embryo was placed into each EP tube containing 100 μL of precooled, newly prepared fixative (methanol:acetic acid = 1:1). The embryo was fully minced with scissors until a homogenate formed. Subsequently, the embryonic suspension was dropped onto a frozen slide and quickly dried using an alcohol lamp. Following drying, the slide was stained with Giemsa for 30 min, rinsed on the back with running water, and observed under a microscope. Finally, the slides were air-dried, mounted with an antifade mounting medium, and examined under a fluorescence microscope (Leica, Germany).

Immunostaining of Spermatocyte Chromosomes

Spermatocyte chromosomal spreads were meticulously prepared using the dry-down method, following established protocols (Peters et al. 1997). The resulting spread slides underwent a 20 min blocking step with a solution of PBS containing 1% goat serum, 0.3% bovine serum albumin, and 0.005% Triton. Subsequently, the slides were incubated at 4 °C overnight with primary antibodies against anti-Sycp3 (Abcam, ab150292) and anti-Mlh1 (BD Pharmingen, 554073). This was followed by a 1 h incubation at 37 °C with secondary antibodies. After each antibody incubation step, the slides underwent thorough washing for 10 min each in PBS containing 0.1% Tween 20, PBS containing 0.3% Triton X-100, and PBS containing 0.1% Tween 20, respectively. Finally, the slides were sealed with an antiquenching agent and stored at −20 °C until observation.

Statistics Analysis

The data are presented as mean ± standard deviation. Two-group comparisons were conducted using the unpaired, two-tailed Student's t-test. For comparisons involving more than two groups, the analysis of variance was employed.

Supplementary Material

Supplementary material is available at Molecular Biology and Evolution online.

Author Contributions

Y.S., S.L., and Y.O. designed the study. Y.O., H.L., X.D., J.H., S.W., and Q.L. prepared the samples and carried out the experiments. C.Y., J.W., R.Z., and Z.Y. analyzed and discussed the results. Y.S., Y.O., and S.L. wrote the paper.

Funding

This work was supported by the National Natural Science Foundation of China (32293252, 32293251, and 31802291), The Higher Education Discipline Innovation Project (D20007), the Natural Science Foundation of Hunan Province (2021JJ40342 and 2021JJ40343), the Training Program for Excellent Young Innovators of Changsha (kq2107006), and the Hunan Province College Students Research Learning and Innovative Experiment project (S202210542154).

Data Availability

Data are available on request.