Disruption of Thyroid Hormone Receptor Thrab Leads to Female Infertility in Zebrafish

Abstract

Thyroid hormones (THs) T4 and T3 are vital for development, growth, and metabolism. Thyroid dysfunction can also cause problems in fertility, suggesting involvement of THs in reproduction. In zebrafish, there exist 2 forms of TH receptor alpha gene (thraa and thrab). Disruption of these genes by CRISPR/Cas9 showed no reproductive irregularities in the thraa mutant; however, inactivation of the thrab gene resulted in female infertility. Although young female mutants (thrab^m/m) showed normal ovarian development and folliculogenesis before sexual maturation, they failed to release eggs during oviposition after sexual maturation. This spawning failure was due to oviductal blockage at the genital papilla. The obstruction of the oviduct subsequently caused an accumulation of the eggs in the ovary, resulting in severe ovarian hypertrophy, abdominal distention, and disruption of folliculogenesis. Gene expression analysis showed expression of both TH receptors and estrogen receptors in the genital papilla, suggesting a direct TH action and potential interactions between thyroid and estrogen signaling pathways in controlling genital papilla development and function. In addition to their actions in the reproductive tracts, THs may also have direct effects in the ovary, as suggested by follicle atresia and cessation of folliculogenesis in the heterozygous mutant (thrab^+/m), which was normal in all aspects of female reproduction in young and sexually mature fish but exhibited premature ovarian failure in aged females. In summary, this study provides substantial evidence for roles of THs in controlling the development and functions of both reproductive tract and ovary.

The thyroid axis plays important roles in vertebrate development, growth, and metabolism (1). In mammals, the axis is controlled by the hypothalamic TRH, which stimulates the secretion of TSH from the pituitary (2). However, in various species of nonmammalian vertebrates, TSH is primarily controlled by corticotropin-releasing hormone, whereas in fish, TRH acts as pleiotropic hormone to regulate a variety of pituitary hormones, including GH, prolactin, ACTH, and melanocyte stimulating hormone α, but not TSH. In all vertebrate groups, TSH is the primary hormone to stimulate the biosynthesis and secretion of the thyroid hormones (THs), including T3 and T4 (2). Similar to other neuroendocrine and endocrine axes, THs exert a strong negative feedback in both the hypothalamus and pituitary to control the activity of the thyroid axis (3), and they act through specific thyroid hormone nuclear receptors (THRs), which work as transcription factors in the target cells upon activation (4). Two major types of THRs (TRα and TRβ) have been identified in vertebrates from fish to mammals, and they are encoded by 2 separate genes (THRA/Thra and THRB/Thrb), which can produce different isoforms through differential splicing of transcripts or alternative use of promoters (2, 4, 5). Genetic studies in mice have provided substantial evidence for differential roles of THRA and THRB and their isoforms (4, 6).

In addition to growth and metabolism, THs have also been implicated in controlling reproduction (7, 8) and their involvement occurs at all levels of the reproductive axis. In the hypothalamus, the GNRH neurons have been shown to express TH receptors in mammals (9), suggesting a direct action for THs on GNRH neurons. At the pituitary level, both T3 and T4 suppressed GNRH-stimulated synthesis and secretion of FSH and LH in cultured rat pituitary cells (10). There are also lines of evidence for TH actions in the gonads. In the ovary, both THRs (THRα1, THRα2, and THRβ1) and TSH receptor are widely expressed in different cells including ovarian epithelial cells, granulosa cells, and oocytes (11), suggesting widespread actions of THs in the ovary. In vitro treatment of ovarian follicles and granulosa cells with T3 promoted follicle growth and reduced apoptosis in granulosa cells (12). In the testis, THRs are also widely expressed, and THs have widespread actions on functions of all type of testicular cells (13).

Clinical observations in humans and experimental data in animal models have provided further evidence for roles of THs in reproduction. Both hyperthyroidism and hypothyroidism have widespread and profound effects on reproductive development and function (8), which often leads to subfertility or infertility due to various functional defects including follicle atresia, anovulation, uterine disorders, and abnormal sperm morphology and motility (13-15). In females, hypothyroidism can cause ovarian disorders including delayed puberty onset (16), anovulation, menstrual and/or estral irregularities (14), infertility, and increased morbidity during pregnancy (3). In rats, hypothyroidism significantly reduced ovarian follicle reserve (17).

The thyroid axis has also been well studied in teleosts (5, 18, 19), and THs are known to regulate key aspects of reproductive development, gametogenesis, and spawning behavior in fish (5, 20-22). Despite these studies, the crosstalk between the thyroid and reproductive axes is not as well defined as that in mammals, partly due to the biodiversity of fish species (5, 20). Although the plasma levels of THs vary during reproductive cycle in fish, their roles remain largely unknown (5). THRs (TRα1, TRα2, and TRβ) are expressed in the gonads of fish such as goldfish (23) and zebrafish (24), suggesting direct actions of THs in the gonads. The expression of THRs in fish gonads (eg, goldfish) showed significant seasonal variation in both ovary and testis (23), and they are subject to autoregulation by THs themselves (25). THs also act at higher levels of the reproductive axis. In tilapia, treatment with T3 suppressed expression of salmon GNRH in the brain (26). In goldfish, treatment with T3 reduced expression of pituitary LH (27) and attenuated GNRH-induced LH secretion (28), similar to the effects in mammals. At the gonadal level, TH treatment reduced aromatase cyp19a1a expression in both ovary and testis of goldfish (27), whereas suppression of TH biosynthesis increased its expression in the testis of goldfish (29) and ovary of catfish (30).

In recent years, the zebrafish has emerged as a popular model for studying endocrine control of reproduction (31). All major components of the thyroid axis have been characterized in zebrafish with conserved structures and functions (32), making it an increasingly popular model for thyroid disorders and diseases (33). Disruption of TH levels in zebrafish through induced hypothyroidism using goitrogens (methimazole) and hyperthyroidism using T4 has been shown to result in female-skewed and male-skewed sex ratios, respectively, suggesting involvement of THs in sex differentiation (34). In male zebrafish, treatment with T3 stimulated spermatogenesis by increasing proliferation of both spermatogonia and Sertoli cells (35). Similar to mammals, zebrafish also have 2 THRs encoded by 3 genes (thraa and thrab for THRA and thrb for THRB) (14, 36). Recently, both THRA genes (thraa and thrab) were inactivated in zebrafish using the CRISPR/Cas9 method (37). Phenotype analysis showed that the thrab mutant [thrab 1-bp ins (m/m), thrab  ^{+  1}] (thrab^m/m) exhibited more severe retardation in body growth than the thraa mutant [thraa 8-bp ins (m/m), thraa  ^{+  8}] (thraa^m/m) although both displayed heart defects (37, 38).

To explore functional importance of TH signaling in fish reproduction, we analyzed reproductive performance of the zebrafish thrab mutant (thrab^m/m). We demonstrated that the thrab^m/m females were infertile due to failure of spawning. Further investigations revealed that the primary cause of female infertility was oviductal blockade at the genital papilla, which prevented release of the mature eggs. Additionally, we also observed other defects in female reproduction, including impaired follicle activation in young females and follicle degeneration in adults. Interestingly, we discovered that the thrab mutant displayed growth retardation and weight gain in mixed and separated culture, respectively, which had profound effects on ovarian growth and folliculogenesis. This study provides crucial genetic evidence for physiological importance of THs in vertebrate reproduction, offering valuable insights from a comparative perspective.

Materials and Methods

Zebrafish Maintenance and Husbandry

The mutant zebrafish [thrab 1-bp ins (m/m), thrab⁺¹] (thrab^m/m) were obtained from the National Institutes of Health, USA (37). The wild-type (WT; AB strain) and mutant zebrafish were maintained under controlled conditions in the ZebTEC Multilinking Zebrafish system (28 °C, pH 7.3, 400 mS/cm conductivity, and 14-hour light/10-hour dark photoperiod) (Tecniplast, Buguggiate, Italy). The fish were fed the Otohime fish diet (Marubeni Nisshin Feed, Tokyo, Japan), delivered by the Tritone automatic feeder system (Tecniplast). From 5 to 10 days post-fertilization (dpf), we fed zebrafish larvae with paramecium twice a day and subsequently artemia in nursery tanks until they were moved to the main aquarium system around 15 dpf. All procedures were conducted in line with the Animal Protection Act enacted by the Legislative Council of Macao Special Administrative Region under Article 71 (1) of the Basic Law, and the experimental protocols were approved by the Research Ethics Committee of the University of Macau (AEC-13-002).

Mutation Screening and Genotyping by High-Resolution Melting Analysis and Heteroduplex Mobility Assay

The mutation screening and genotyping were conducted using high-resolution melting analysis (HRMA) and heteroduplex mobility assay (HMA) based on our earlier work (39). Briefly, the genomic DNA was extracted from a single embryo or a piece of caudal fin using the sodium hydroxide method (40). HRMA was performed on the DNA with specific primers (Table 1) flanking the target site of zebrafish thrab. The reaction was carried out on the Bio-Rad CFX96 real-time system, and the data were analyzed with the Precision Melt Analysis software (Bio-Rad Laboratories, Hercules, CA). We used HMA to confirm the HRMA screening results (39, 41). Briefly, 5 µL of PCR product from HRMA was subjected to electrophoresis in 20% PAGE gel, and the signals were visualized on the ChemiDoc imaging system (Bio-Rad).

Table 1.

Primers used for HRMA analysis and PCR

Primer name	Primer sequence (5' →3’)
5922_thrab F	CTACATCAACTATCGCAAGC
5923_thrab R	ACGTTCCTGATTCTCTTGCC
3088_esr1 F	CAGGATTCCAGGATCTGTCTCT
3089_esr1 R	CCTCAACACATTCTCCTTCACTC
1224_esr2a F	TGTTGGAGTGCTGCTGGTTAGAG
1225_esr2a R	TGTTGGAGTTCAGGAGGATCATGG
1228_esr2b F	AAGGTGCTGAATCTGCTGGACTC
2966_esr2b R	GTCCTGGACTCCAGCAAGGCCGATCG
728_ef1a F	GGCTGACTGTGCTGTGCTGATTG
729_ef1a R	CTTGTCGGTGGGACGGCTAGG
7691_thraa F	CCATGGACTTGGTGCTGGAT
7692_thraa R	TTGTAGCAGTGCCACCTCAG
7693_thrb F	GGGTCATTTCAGGCCACGTA
7693_thrb R	TCAGCTTCCGCTTGGCTAAA

Abbreviations: HRMA, high-resolution melting analysis.

Oocyte Maturation Assay In Vitro

Full-grown (FG) follicles (>650 µm) were isolated from sexually mature control and thrab^m/m females maintained in a group of mixed males and females (10-50 fish in total). Briefly, the ovaries were dissected out, placed in 60% L-15 medium (Thermo Fisher; Waltham, MA), and rinsed a few times with the same medium. Immature FG follicles were isolated using fine forceps and distributed in a 24-well plate with 25 follicles per well. They were incubated at room temperature in 1 mL 60% L-15 medium per well with or without 5 nM 17α,20β-dihydroxy-4-pregnen-3-one (DHP), the maturation-inducing hormone in zebrafish. After a 4-hour treatment, the follicles were examined under the SMZ18 dissecting microscope (Nikon, Tokyo, Japan) to observe germinal vesicle breakdown (GVBD), which is characterized by the oocytes becoming translucent (42).

Assessment of Fertility and Fecundity of Thrab Mutant

After sexual maturation, we assessed the fertility of each genotype. We tested each fish at least 3 times at 3-day intervals by natural mating with fertile WT fish of the opposite sex, and at least 3 individuals of each genotype were tested for statistical analysis. After a 24-hour incubation at 28 ± 0.5 °C, the total and fertilized embryos were counted for each genotype.

Tissue Processing and Histological Analysis

For each fish sampled for histological analysis, we measured the morphometric parameters such as standard body length (BL) and body weight (BW) before processing. After measurements, the whole fish was fixed in Bouin's solution for a minimum of 24 hours at ambient temperature. The fixed samples were then processed on the ASP200S Automatic Tissue Processor (Leica Biosystems, Deer Park, IL) as follows: initial dehydration in 50%, 70%, and 85% ethanol for 1.5 hours each; 95% ethanol for 2 hours; and 100% ethanol 3 times for 30 minutes each; sequential clearance in xylene and ethanol mixture (1:1) for 30 minutes and xylene twice for 30 minutes each; and overnight infiltration in paraffin. The processed samples were then embedded in paraffin and sectioned serially at a thickness of 5 µm. Standard haematoxylin and eosin staining and slide mounting were performed on the ST5020 Stainer Integrated Workstation and the CV5030 Glass Coverslipper (Leica Biosystems), respectively, at the Histopathology Core of the Faculty of Health Sciences. The sections were examined on the Eclipse Ni-U upright microscope (Nikon), and micrographs were captured with the Digital Sight DS-i2 camera (Nikon). To assess the impact of thrab disruption on folliculogenesis, we measured diameters of all follicles with visible nucleus (germinal vesicle) in serial longitudinal sections as described (41, 43).

Terminal Deoxynucleotidyl-Mediated dUTP Nick Labeling Assay

To identify apoptotic cells in the ovary, the terminal deoxynucleotidyl-mediated dUTP nick labeling assay was performed with In Situ Cell Death Detection Kit (Roche, Basel, Switzerland), which detects apoptotic DNA fragmentation. First, the paraffin-embedded tissue sections were rehydrated through a graded series of ethanol solutions, followed by permeabilization using proteinase K (20 μg/mL). Then, the tissue sections were incubated with a mixture of biotinylated nucleotides and recombinant terminal deoxynucleotidyl transferase in an equilibration buffer for 1 hour at 37 °C. After incubation, the tissue sections were washed 3 times in phosphate-buffered saline for 5 minutes each. Finally, the sections were stained with 4',6-diamidino-2-phenylindole (DAPI) for 10 minutes at room temperature to visualize the nuclei (44).

RNA Extraction and Quantitative Real-time qPCR

Total RNA was isolated from the ovarian tissues using TRIzol (Invitrogen, Carlsbad, CA), following the manufacture's protocol. The extracted RNA was then reverse transcribed into cDNA using M-MLV reverse transcriptase (Invitrogen) as previously reported (45). The primers used for real-time qPCR are listed in Table 1. Gene expression levels were normalized to those of the housekeeping gene ef1a, and the data are presented as fold changes relative to the control group.

Statistical Analysis

All values used for statistical analysis were obtained from independent experiments and/or biological repeats (n ≥ 3) and expressed as mean ± SEM. The data were analyzed by Student's t-test or one-way ANOVA followed by a Tukey multiple comparison test using Prism (GraphPad, La Jolla, CA).

Results

Effects of Thrab Disruption on Growth and Thyroid Disorders

It was reported recently that disruption of thrab gene caused growth retardation in juveniles (28 dpf) (37). We also observed similar phenotype in this study at 30 dpf, with thrab mutants slightly smaller than the control. To explore this further, we raised the fish under 2 different conditions from 30 dpf: mixed culture and separated culture. In mixed culture, the thrab mutants consistently exhibited a significantly smaller body size (BL and BW) at 40 and 50 dpf. This disparity in body size between mutants and controls not only persisted but also became more pronounced at 120 and 225 dpf, with growth of the mutants nearly ceasing or even regressing (Fig. 1A and 1B). Clonversely, when the mutants were reared in isolation from 30 dpf, they showed little difference in body size compared to control fish at 38, 47, and 60 dpf, indicating a recuperation in growth after separation from the control fish. Notably, the growth of the mutants began to surpass that of the controls from 70 dpf onwards, showing a remarkable divergence at 120 dpf as the growth of mutants surged dramatically compared to the controls, displaying a significant weight gain (Fig. 1A and 1B). Furthermore, the female mutants reared separately exhibited obvious belly distension at 120 dpf, a condition absent in the mutants from the mixed culture group (Fig. 1B). We obtained 8 mutant fish in the mixed group at 120 dpf and identified only 1 female (2.25 cm BL and 175.5 mg BW). Histological analysis of this individual revealed a marked suppression of ovarian growth, with follicle development stalling at the early previtellogenic (PV) stage (Fig. 1D), in sharp contrast to the mutant females reared in separation (see later for details).

Figure 1.

Effects of thrab mutation on zebrafish development. (A) Growth performance of the mutant (thrab^m/m) and control (thrab^+/m) fish. The fish were reared in mixed or separated cultures, and the standard body length and body weight of the fish were measured at different time points. (B) Morphology of representative fish at different times in mixed or separated cultures. (C) Morphological and histological analysis of mutant (thrab^m/m) and control (thrab^+/m) fish at 120 dpf. Values are presented as means ± SEM (n = 7-13). *P < .05, **P < .01 and ***P < .001 vs control group.

Abbreviations: FG, full-grown; ns, no significance; PG, primary growth; PV, previtellogenic.

In mammals, THs are well known to exert strong negative feedback on both TRH in the hypothalamus and TSH in the pituitary (2). Diminished output of THs, often due to conditions such as iodide deficiency, can lead to thyroid hypertrophy, a pathological condition commonly referred to as goiter (1). In agreement with a recent study on thyroglobulin (tg) mutant zebrafish (46), we also observed occurrence of goiter in about 40% thrab mutant fish after 12 months postfertilization (mpf). The hypertrophied thyroid glands in these mutant fish were voluminous with large cavities and amorphous tissues. The periphery of the glands contained numerous thyroid follicles, some of which were extended with abundant colloidal material (Fig. 2). We only observed goiter development in male mutants in the separated group because all female mutants in the separated group died before 12 mpf due to severe ovarian hypertrophy (see later for details). In the mixed group, none of the mutants could survive beyond 12 mpf due to severe growth retardation (data not shown).

Figure 2.

Development of goiter in aged thrab^m/m zebrafish. (A) Occurrence of goiter (arrow) under the jaw in mutant zebrafish at 12 mpf. (B) Histology of the goiters in the mutant fish (males). The enlarged goiter-like structure contained a large cavity with amorphous tissues and thyroid follicles at the periphery. (C) Frequency of goiter occurrence in the mutant fish.

Normal Folliculogenesis in Thrab Mutant Prior to Sexual Maturation

To evaluate the role of THs in female reproduction of zebrafish, we focused on the fish reared in separation because the mutants in the mixed group could not grow normally as described previously. We first examined folliculogenesis in thrab^m/m ovaries in young fish before sexual maturation (60-80 dpf). The follicle development was largely normal in the mutant ovary compared to the control fish (Fig. 3). In the control ovary at 60 dpf, the follicles were primarily at the primary growth (PG; stage I) and PV (stage II) stages. The PV stage is characterized by the formation of cortical alveoli in the oocyte, indicating successful PG-PV transition or follicle activation (puberty onset) (47). Similarly, the mutant females (thrab^m/m) also showed normal ovarian development with both PG and PV follicles present (Fig. 3A). However, quantification of follicle composition showed more PG follicles and fewer PV follicles in the mutant fish compared to the controls (Fig. 3B and 3C). At 70 dpf, both control and mutant fish had early vitellogenic and mid-vitellogenic follicles in the ovary (Fig. 3D and 3E); however, the mutant fish continued to contain more PG follicles but less PV follicle (Fig. 3F), indicating a delayed PG-PV transition. This developmental delay appeared to persist at 80 dpf, though the differences were not stastistically significant (Fig. 3G-3I). Together, these data indicate a suppression of follicle progression in thrab^m/m females before sexual maturation.

Figure 3.

Normal folliculogenesis in thrab mutant prior to sexual maturation. (A, D, and G) Morphological and histological analysis of mutant (thrab^m/m) and control (thrab^+/+) fish at 60, 70, and 80 dpf. (B, E, and H) Diameters of follicles present in the ovary at 60, 70, and 80 dpf. (C, F, and I) Follicle composition in the ovary at 60, 70, and 80 dpf. The follicles were grouped into 3 categories: PG, PV, and EV-FG. Values are presented as means ± SEM (n = 3). *P < .05 and **P < .01 vs control group.

Abbreviations: EV, early vitellogenic; FG, full-grown; LV, late vitellogenic; MV, mid- vitellogenic; ns, no significance; PG, primary growth; PV, previtellogenic.

Ovarian Hypertrophy and Female Infertility in Thrab Mutant After Sexual Maturation

Interestingly, when the thrab^m/m females became sexually mature at 90 dpf, the fish abdomen became obviously enlarged compared to the control group, suggesting ovarian hypertrophy (Fig. 4A). The same phenotype was observed at 120 dpf (Fig. 4B). The ovarian hypertrophy was confirmed by dissection and analysis of the gonadosomatic index (gonadal weight/BW), which revealed significantly bigger ovaries in the mutant as compared to those in the control (Fig. 4C). In addition to hypertrophy, the mutant ovaries also displayed some other abnormalities. First, the follicles displayed clear signs of degeneration or atresia at 90 dpf, indicating a disruption of folliculogenesis. Second, a proliferation of somatic cells was observed, infiltrating the interfollicular spaces within the mutant ovaries (Fig. 4A). The deterioration of the mutant ovary was supported by terminal deoxynucleotidyl-mediated dUTP nick labeling assay, which demonstrated a significantly increased apoptotic signal in mutant ovary at 90 dpf compared to the control (Fig. 4D). The ovarian degeneration and follicle atresia continued as the fish aged. At 120 dpf, the mutant females showed severely disintegrated ovaries with widespread follicle atresia and a reduction in vitellogenic follicles (early vitellogenic EV-FG) (Fig. 4B and 4E). Despite ovarian hypertrophy, the number of follicles per unit area (mm²) decreased significantly at both 90 and 120 dpf, partly due to the increased presence of somatic cells and ascites in the interfollicular spaces (Fig. 4E).

Figure 4.

Ovarian hypertrophy and female infertility in thrab^m/m mutants after sexual maturation. (A and B) Morphology and histology of control (thrab^+/m) and mutant fish (thrab^m/m) at 90 and 120 dpf. The mutant ovary showed hypertrophy with follicle atresia and tissue degeneration or fibrosis. (C) gonadosomatic index of controls (thrab^+/+ and thrab^+/m) and mutant fish (thrab^m/m) at 90 dpf. Values are presented as means ± SEM (n = 3). (D) Terminal deoxynucleotidyl-mediated dUTP nick labeling assay at 90 dpf on control and mutant ovaries. (E) Diameters of follicles and follicle composition in the ovary at 90 and 120 dpf. The follicles were grouped into 3 categories: PG, PV, and EV-FG. The relative number of follicles indicates the density of follicles per square millimeter in ovarian sections. This serves as an indicator of follicle atresia in mutant ovaries. Values are presented as means ± SEM (n = 3). (F) Fecundity assessment of females of different genotypes (+/+, +/m, and m/m) at 150 dpf. In each test, 3 females of each genotype were paired with 3 males in the same breeding tank. Each data point represents the mean output of each female. The testing was repeated 5 times at 3-day intervals for each female for statistical analysis. Values are presented as means ± SEM (n = 5). The mutant fish (m/m) were infertile as shown by the lack of spawned eggs when breeding with the control males (+/+). (G) Morphological comparison and histological analysis of gonadal development among thrab^+/+, thrab^+/m, and thrab^m/m females at 150 dpf. *P < .05 and ***P < .001 vs control group.

Abbreviations: EV, early vitellogenic; FG, full-grown; ns, no significance; PV, previtellogenic; SC, somatic cells.

A fertility test at 150 dpf (5 mpf) showed that in sharp contrast to the WT control fish (+/+) and the heterozygote (+/m), the homozygous mutant females (m/m) were unable to spawn when bred with WT males (Fig. 4F), indicating female infertility. Dissection of the fish at 150 dpf revealed a substantial accumulation of ascites in the abdominal cavity (not visible in the figure). Histological analysis showed that the mutant ovaries continued to deteriorate, displaying serious ovarian edema with large amount of serous fluid (ascites) accumulated in the interfollicular spaces. Interestingly, the heterozygous mutant (thrab^+/m) also started to show signs of ovarian edema (Fig. 4G). Despite this, the heterozygous females could spawn normally, in contrast to the homozygous mutants. This phenotype provoked intriguing questions regarding the potential causes of spawning failure in the mutant. Possible mechanisms could include defects in oocyte maturation, ovulation, or oviposition.

Normal Response of Thrab^m/m Oocytes to Maturation-inducing Hormone

To investigate the mechanism underlying female infertility observed in thrab^m/m zebrafish, we first examined the final oocyte maturation, which is characterized by GVBD, and its response to stimulation by DHP, the known maturation-inducing hormone in zebrafish (48, 49). Dissection of the ovaries at 120 dpf revealed presence of mature FG follicles with GVBD in both control (thrab^+/m) and mutant females (thrab^m/m) (Fig. 5A). When incubated in vitro, the immature FG follicles from both controls and mutants could undergo spontaneous maturation (∼10%). In response to DHP, the maturational rate increased to 30% in both groups after a 4-hour incubation, indicating normal maturational competence (responsiveness to DHP) (Fig. 5B and 5C). These results negate the possibility that the infertility observed in the female mutant (thrab^m/m) was due to impaired oocyte maturation. This also suggests that the observed infertility could stem from problems occurring during the oviposition stage.

Figure 5.

Oocyte maturation and maturation competency in thrab^m/m mutant. (A) Control (thrab^+/m) and mutant (thrab^m/m) females with normal ovarian growth at 120 dpf. Mature follicles could be found in the ovaries of both control and mutant fish. Arrow, follicles that had undergone GVBD; arrowhead, immature follicles. (B and C) In vitro response of follicles to maturation-inducing hormone DHP in 4-h incubation. Values are presented as means ± SEM (n = 7). *P < .05; **P < .01 vs control group.

Abbreviations: DHP, 17α,20β-dihydroxy-4-pregnen-3-one; ns, no significance.

Oviductal Blockade in Sexually Mature Thrab^m/m Females

To investigate potential defects in oviposition, we undertook a detailed morphometric examination of the genital papilla by both dissection and histology. In zebrafish, genital papilla is a fleshy tube that extends from the reproductive (oviduct) and urinary (ureter) tracts, protruding posteriorly behind the anus (Fig. 6A). Interestingly, measurement of the dissected mutant genital papilla demonstrated that they were significantly smaller than those of the control fish (Fig. 6B and 6C). We then performed histological analysis through both transverse and longitudinal serial sections across the area of genital papilla. In the control fish, 3 tubular structures could be identified from anterior to posterior, representing intestine (anus), oviduct, and ureter, respectively. The intestine was lined with villi while the oviduct was encircled by a thick layer of muscular tissues (Fig. 6D). In some fish samples examined, we observed ovulated eggs in the oviduct or developing follicles nearby (Fig. 6E). In contrast to the control fish that displayed 3 tracts in the area (intestine, oviduct, and ureter), the oviduct in the thrab^m/m females was either completely absent (type 1) or partially open as a honeycomb-like structure (type 2) (Fig. 6F). This observation strongly suggests that the oviduct in mutant females was blocked, which was likely one of the major factors for female infertility in the thrab^m/m mutant. This blockade prevented the passage of mature and ovulated eggs at oviposition, resulting in abdominal distention and ovarian hypertrophy.

Figure 6.

Oviductal blockade in thrab^m/m female zebrafish. (A) Location of genital papilla (arrow), a female secondary sex characteristic. The region of the genital papilla was isolated for histological examination. The signs of scissors indicate the cutting sites for isolation. (B) Isolation of genital papilla from the control (thrab^+/m) and mutant (thrab^m/m) females. (C) Relative size of the genital papilla from the control and mutant females. The sizes of the papilla were determined with the software ImageJ. Values are presented as means ± SEM (n = 3). *P < .05 vs control group. (D) Serial longitudinal sections of the genital papilla from the control and mutant females (4 sections per fish). Three tubes could be seen at the papilla: IT, OD, and U from anterior to posterior. The asterisks show the opening of the oviduct in the control fish and the arrowheads show the location of the oviductal blockade in the mutant. (E) A representative transverse and longitudinal section of control fish with diagrams showing 3 tubes (IT, OD, and U). The transverse section shows a few mature oocytes in the oviduct, ready to be released upon spawning, whereas a group of developing follicles was visible next to the oviduct in the longitudinal section. (F) Transverse sections of the control and mutant females highlighting oviduct in the control and oviductal blockade in the mutant. Two types of oviductal blockade were observed in the mutant. Type 1 shows a complete absence of oviduct, whereas type 2 shows partial obstruction with honeycomb-like structure.

Abbreviations: IT, intestinal tube; OD, oviduct; U, ureter.

Potential Interaction of TH and Estrogen Signaling in the Development of Genital Papilla

The aforementioned discovery strongly indicates the significant involvement of THs in the development of genital papilla, acting through Thrab. Our recent studies showed that the female-specific enlargement and extension of the genital papilla was estrogen-dependent (50). This observation prompted us to investigate possible interactions between estrogen and thyroid hormone signaling pathways. To provide evidence for this, we removed the genital papilla by surgery as shown in Fig. 6 and extracted RNA from each to analyze gene expression by RT-qPCR. As shown in Fig. 7A, the expression of all 3 nuclear estrogen receptors (esr1, esr2a, and esr2b) decreased dramatically in the mutant genital papilla, while the expression of thrab remained unchanged. To demonstrate tissue specificity of this response, we also examined the expression of these genes in muscle tissues collected from the adjacent sites, and they showed no discernible differences between the control and mutant fish (Fig. 7B). RT-PCR analysis showed that of the 3 TH receptors (thraa, thrab, and thrb), only thrab was expressed in the genital papilla, while all 3 receptors were expressed in the ovary (Fig. 7C).

Figure 7.

Expression of estrogen receptors and thrab in the genital papilla. (A) Expression of estrogen receptor (esr1, esr2a, and esr2b) and thrab in the genital papilla. All 3 estrogen receptors were expressed in the genital papilla with a significant reduction in expression in the thrab mutant (m/m). The expression data were normalized to the housekeeping gene ef1a and presented as relative levels compared to the control. Values are presented as means ± SEM (n = 3). ***P < .001 vs control group. (B) Gene expression in the muscle. The expression of esr1, esr2a, esr2b, and thrab was determined in skeletal muscle sampled from the nearby region as a control to demonstrate the specificity of the responses in the genital papilla. (C) RT-PCR detection of thyroid hormone receptor expression in the genital papilla. Of the 3 thyroid hormone receptors (thraa, thrab, and thrb), only thrab was expressed in the genital papilla, in contrast to the expression of all 3 receptors in the ovary.

Abbreviations: ns, no significance.

Premature Ovarian Failure in Aged Heterozygous Females (Thrab^+/m)

Contrary to the homozygous mutant (thrab^m/m), the heterozygotes (thrab^+/m) exhibited no evident abnormalities in either structure or function during early adulthood. As described earlier, the heterozygotes showed normal ovarian development with typical folliculogenesis and spawned normally with WT males at 150 dpf (5 mpf) with fecundity comparable to that of the WT females (thrab^+/+) (Fig. 4F and 4G).

However, a striking change was observed at later stage around 12 mpf, when the heterozygous females (thrab^+/m) stopped spawning with WT males, a stark contrast to their normal fecundity at 5 mpf (Fig. 8A). Histological examination revealed a significant regression in the ovaries of the heterozygotes, which contained only PG follicles. In addition, the oocytes in many follicles had been supplanted by amorphous tissues encased by an intact follicle layer (Fig. 8B and 8C). This phenotype is reminiscent of premature ovarian failure or primary ovarian insufficiency in humans (51, 52). Given that no homozygous mutant females survived beyond 12 mpf and only some male mutants did, we tested the fertility of these surviving males at 12 mpf. The results showed that none were able to induce spawning in WT females (Fig. 8A).

Figure 8.

Premature ovarian failure in aged heterozygous thrab^+/m females. (A) Fertility test at 5 and 12 mpf. The homozygous mutant females (F, m/m) showed no fertility at both time points. The heterozygotes (+/m) showed normal fertility at 5 mpf but were infertile at 12 mpf when crossing with wildtype males (M, +/+). In each test, 3 females were paired with 3 males in a breeding tank, and the data are expressed as the average per female. The testing was repeated 4 times at 3-day intervals for statistical analysis. Three mutant males (M, m/m) were also tested at 12 mpf with wildtype females (F, +/+), showing no spawning. (B) Histology of the wildtype control (thrab^+/+) and heterozygous mutant (thrab^+/m) ovaries at 12 mpf. In all the 12 mutant fish examined, the ovaries contained PG follicles only, indicating a complete cessation of folliculogenesis. In 4 mutant fish (4/12), the ovaries were almost absent with only small numbers of PG follicles. In majority of the mutant fish (8/12), the PG follicles were abundant; however, many follicles had lost the oocytes (arrow) and the intrafollicular spaces were occupied by amorphous somatic tissues (asterisk). (C) Follicle composition in the control and heterozygous mutant ovary. Values are presented as means ± SEM (n = 3-5). **P < .01 and ***P < .005 vs control group.

Abbreviations: EV-FG, vitellogenic stages; PG, primary growth; PV, previtellogenic.

Discussion

As pleiotropic hormones, THs play fundamental roles in vertebrate development, growth, and metabolism (1). The most notable example of their roles in development is the control of amphibian metamorphosis (53). In humans, an early-life deficiency of TH signaling or congenital hypothyroidism can lead to severe mental and growth retardation, a condition known as cretinism (54, 55). The hypothyroidism in adults can also result in a variety of symptoms, including decreased basal metabolic rate and thermogenesis and increased body mass index (56, 57). In addition to development, growth, and metabolism, THs are also increasingly recognized for their roles in reproduction across vertebrates (14, 58).

In mammals and humans, TH involvement in female reproduction has been extensively reported and reviewed. Thyroid dysfunctions can lead to severe reproductive disorders (14). TH deficiency or hypothyroidism is often caused by self-destruction of the thyroid gland by the immune system, which leads to a series of reproductive disorders in females, including reduced folliculogenesis and ovulation, increased follicle atresia and oxidative stress in the ovary, and decreased uterine weight due to a reduced proliferation of the uterine cells (14).

Using a loss-of-function approach, the present study investigated the roles of Thrab (thrab), one of the two paralogues of the TH receptor alpha (Thra, thraa, and thrab) in zebrafish, in female reproduction. We focused on thrab because its mutation generated much more severe phenotypes in growth and gene expression than that of thraa (37). The zebrafish is an excellent model for studying TH functions because all components of the thyroid axis are present in this model species and function similarly to those in other vertebrates (32). The high conservation of the thyroid axis suggests fundamental roles for THs in animal life and species survival. Our findings showed clearly that TH signaling through Thrab plays critical roles in controlling zebrafish female reproduction.

Growth Performance of Thrab Mutant and its Impact on Ovarian Development

An intriguing observation from this study is that the thrab mutants (thrab^m/m) displayed contrasting growth trajectories under different rearing conditions. In the mixed culture with control fish, the growth of mutants was significantly suppressed. Consequently, their ovarian development was also impaired, with follicles arrested at the PV stage, failing to initiate vitellogenesis due to the mutants not reaching the necessary body size threshold (59). In contrast, when raised in isolation from the controls, the mutants exhibited an accelerated growth rate, eventually surpassing the body size of control fish post-sexual maturation. This increase in body weight is consistent with a classic symptom of hypothyroidism: weight gain (57). Surprisingly, the mutant ovaries underwent marked hypertrophy, in stark contrast to the hypotrophy observed in mixed culture. These disparate phenotypes under different rearing conditions underscore the profound influence of social and environmental factors on the expression of mutant traits. In either condition, the mutant females were infertile but due to different etiology: ovarian hypotrophy with incomplete folliculogenesis in mixed culture vs ovarian hypertrophy with failed oviposition in separated culture.

Oviductal Obstruction in the Thrab Mutant

To elucidate the mechanisms behind ovarian hypertrophy and oviposition failure, we conducted systematic analyses of mutant females (thrab^m/m) cultured separately from the control fish. In young females prior to sexual maturation (<90 dpf), the loss of Thrab function did not significantly affect ovarian development and folliculogenesis compared to the control females except for a slight delay in PG-PV transition. However, the mutant started to exhibit severe reproductive defects after about 90 dpf when zebrafish become sexually mature, including abdominal distention due to ovarian hypertrophy, ovarian tissue deterioration with follicle atresia, and lack of spawning due to oviposition failure. Further investigations led us to discover that the mutant females had significantly smaller genital papilla and their oviducts were blocked, which prevented release of the mature eggs.

The reproductive or genital tracts in zebrafish (oviduct in females and sperm duct in males) extend to the outside as a fleshy genital papilla behind the anus. As a secondary sexual characteristic, the genital papilla is more prominent in females, and it can be used to determine fish sex. Since the development of genital papilla in females is estrogen-dependent in zebrafish (50), we examined the expression of estrogen receptors in the genital papilla to see if the oviductal blockage in the thrab mutant might involve impaired estrogen signaling. Interestingly, all 3 estrogen nuclear receptors (esr1, esr2a, and esr2b) decreased their expression dramatically in the genital papilla of the thrab mutant compared to the control, suggesting that deficient estrogen signaling could be part of the mechanism underlying the defective development of the oviduct and genital papilla. This agrees very well with a previous study in medaka fish. The female medaka mutant lacking Esr2a/esr2a showed normal spawning behavior; however, the mutant was infertile because of failed oviposition. Further histological analysis revealed that the mutant had normal ovarian cavity, but its oviduct was blocked and atretic as compared to that in the control (esr2a+/−) (60). Interestingly, the loss of the TSH βa subunit gene (tshba) in zebrafish also resulted in infertility because the homozygous males and females failed to produce offspring. The cause for the infertility remains unknown. The male mutant lacked breeding tubercles on the pectoral fins (male secondary sex characteristic), and female mutants also exhibited smaller genital papilla (46). Our finding that the expression of estrogen receptors was diminished in the mutant genital papilla strongly suggests a regulation of these receptors by THs through Thrab, which acts as a transcription factor upon TH binding (37). This interplay agrees well with previous studies showing various levels of interaction between TH and estrogen signaling pathways, including direct protein-protein interactions between receptors and shared gene promoter elements responsive to both hormones (61, 62). In fish, for instance, all 3 estrogen receptors have been shown to be downregulated in the ovary in response to T3 treatment (27). The regulation of estrogen receptors by TH in the zebrafish genital papilla may involve Thrab only as the other 2 forms of TH receptors (thraa and thrb) were not expressed in the genital papilla, in contrast to the expression of all 3 forms in the ovary.

The involvement of TH signaling in the development and function of female reproductive tracts has been suggested by studies in mammalian species. THRs (TRα1-2 and TRß1) are expressed throughout rat female reproductive tracks, including the uterus, oviduct, and vagina (both abdominal and perineal vagina) (63), suggesting functional roles of THs in controlling their development and function. In the oviduct, the THRs (TRα1-2) were primarily expressed in the epithelial cells of the tube and smooth muscle cells of the wall (64). Thyroidectomy in rats resulted in a reduction of the smooth muscle wall, a morphological change of epithelial cells from columnar to cuboidal shape, and a flattening of the mucosal folds. Functionally, both spontaneous contraction of the oviductal muscle and its response to oxytocin were contingent on plasma levels of THs (65). Although zebrafish oviduct is not developmentally homologous to that in mammals, it is functionally analogous. Our discovery in zebrafish provides further evidence for conservative roles of THs in controlling the development and function of reproductive tracks across vertebrates.

Impaired Ovarian Homeostasis in Thrab Mutant

One of the most prominent phenotypes displayed by the thrab^m/m mutant was the abdominal distention due to ovarian hypertrophy. Since the onset of these phenotypes correlated well with the start of sexual maturation and spawning, we believe that they might be induced by the oviductal blockade, which prevented the release of mature eggs. The causal relationship between ovarian disorders and oviductal blockade requires further investigation.

In addition to hypertrophy, the mutant ovary also exhibited severe tissue disintegration, characterized by follicle atresia and ovarian edema. It is unclear whether these abnormalities were the primary response to impaired TH signaling through Thrab in the ovary or secondary to the accumulation of the deteriorated eggs in the ovarian cavity due to the oviductal blockage. The exact mechanism for the ovarian tissue destruction in the thrab mutant is unknown.

The severe female reproductive disorders in the thrab mutant are in sharp contrast to the phenotypes of the zebrafish tshba mutant. Although the tshba-/− mutant also exhibited female infertility, indicating the importance of THs in reproduction, its folliculogenesis was normal without abdominal distention up to 8 mpf (46). The discrepancy between the 2 studies could be attributed to the fact that the mutation of tshba did not completely eliminate the production of TSH because of the presence of the second TSH β gene (tshbb), which might have compensated for the loss of tshba to some extent, resulting in a partial reduction in TSH level.

Premature Ovarian Failure in Thrab Mutant

As described earlier, the young females of the thrab^m/m mutant displayed normal ovarian development and folliculogenesis prior to sexual maturation (<90 dpf) despite a slight delay in PG-PV transition. This did not seem to support a significant role for THs in controlling follicle development in zebrafish. As discussed earlier, the deterioration of the ovarian tissues and disruption of folliculogenesis exhibited by the thrab mutant post-sexual maturation could possibly be attributed to the retention of mature eggs in the ovary, a result of the oviductal obstruction.

The supportive evidence for direct roles of THs in folliculogenesis was obtained from the heterozygous mutant (thrab^+/m). The heterozygous females functioned normally in all aspects of female reproduction, including follicle growth, oocyte maturation, ovulation, oviposition, and fertilization before and after sexual maturation. However, these females gradually lost their fertility by 12 mpf due to a complete cessation of folliculogenesis in the ovary. The follicles of advanced stages gradually disappeared, and the ovary became atrophic containing only early follicles at PG or PV stages, a sign reminiscent of premature ovarian failure. Since the oviduct functioned normally in the heterozygote, this observation strongly implies a role for THs in regulating the hypothalamic-pituitary-ovary axis, including direct actions in the ovary. The appearance of the phenotype in older, but not younger, heterozygous females also suggests that the defects in folliculogenesis might be potentially tied to aging-related factors such as reduced energy metabolism and mitochondrial dysfunction. The delayed onset of the defective phenotypes in the heterozygote (thrab^+/m) relative to the homozygous mutant (thrab^m/m) was likely due to the difference in gene dosage. Alternatively, it could also be due to the dominant negative interference of the WT receptor Thrab by the mutant form. As reported recently, the Thrab mutant receptor lacked transcriptional response to T3 and also suppressed the activity of the WT form of Thrab when cotransfected (37).

In summary, using the loss-of-function approach, this study demonstrated that the loss of TH signaling through Thrab receptor had profound effects on female reproductive function. First, the mutant females displayed impaired development of the oviduct, which subsequently led to spawning failure and female infertility. Second, the oviductal obstruction resulted in ovarian hypertrophy, follicle atresia, and ovarian tissue degeneration. Third, THs are also essential for long-term maintenance of ovarian homeostasis and folliculogenesis. This study provides a strong support to the notion that THs are not merely important for development, growth, and metabolism but also indispensable for reproduction.

Funding

This study was supported by grants from the University of Macau and the University of Macau Development Foundation (MYRG2022-00219-FHS, MYRG-GRG2023-00144-FHS-UMDF, and CPG2024-00030-FHS) and The Science and Technology Development Fund (FDCT), Macao SAR (FDCT173/2017/A3 and NSFC-FDCT0086/2022/AFJ) to W.G. N.A. was supported by the FDCT Funding Scheme for Postdoctoral Researchers (0029/APD/2021).

Disclosures

The authors declare no conflicts of interest.

Data Availability

Original data generated and analyzed during this study are included in this published article.