2019 Volume 44 Issue 4 Pages 257-271
Vitamin A and its derivatives contribute to many physiological processes, including vision, neural differentiation, and reproduction. Vitamin A deficiency causes early cessation of spermatogenesis, characterized by a marked depletion of germ cells. However, there has been no clear understanding about the role of chronic intake of vitamin A excess (VAE) in spermatogenesis. The objective of this study was to investigate whether chronic intake of VAE diet causes arrest of spermatogenesis. To examine the effects of VAE on spermatogenesis, we used ICR male mice fed with control (AIN-93G purified diet: 4 IU/g) diet or VAE (modified AIN-93G diet with VAE: 1,000 IU/g) diet for 7 weeks (from 3 to 10 weeks of age). At 10 weeks of age, the retinol concentration in the testes of VAE mice was significantly higher than that of control mice. Testicular cross sections from control mice contained a normal array of germ cells, while the seminiferous tubules from VAE mice exhibited varying degrees of testicular degeneration. Daily sperm production in VAE testes was dramatically decreased compared to that in control testes. Sperm viability, motility, and morphology were also impaired in VAE mice. Furthermore, we examined the effects of VAE on the expression of genes involved in retinoid signaling and spermatogenesis to determine the underlying molecular mechanisms. Therefore, we are the first to present results describing the long-term dietary intake of VAE impairs spermatogenesis using a mouse model.
The biological effects of vitamin A are mediated through binding of their active metabolites (retinoic acid) to retinoic acid receptors (RARs) and/or retinoid X receptors (RXRs). Both receptors include at least three isotypes designated α, β, and γ, encoded by distinct genes (Chambon, 1996; Amann et al., 2011). These receptors belong to the steroid/thyroid hormone nuclear receptor superfamily and function as ligand-dependent transcription factors binding to retinoic acid response elements in the promoters of their target genes (Mangelsdorf and Evans, 1995; Kastner et al., 1997). More than five hundreds of genes can be regulated directly through the classical retinoic acid pathway, which indicates that retinoid metabolism is important for regulation of cell development (Balmer and Blomhoff, 2002).
Spermatogenesis is the process of sperm cell development from spermatogonia to spermatozoa, which involves many genes and molecular mechanisms (Jan et al., 2012; Mark et al., 2015). Spermatogenesis is a very complex, highly organized, and regulated process that takes place in the seminiferous epithelium of the testis, and involves three major biological processes: renewal of stem cells and production and expansion of progenitor cells (mitosis), reduction (by one-half) of the number of chromosomes in each progenitor cell (meiosis), and unique differentiation of haploid cells (spermiogenesis). These steps are described as a cycle of cellular changes, referred to as stages of the seminiferous epithelial cycle, and they occur within defined regions of the epithelium (Lie et al., 2009).
Vitamin A (in the form of retinol or retinoic acid) is required for fertility and normal spermatogenesis, and recently the mechanisms that drive the retinoic acid-mediated regulation of germ cell development have begun to be understood (Mark et al., 2015). Since 1925, when spermatogenesis was found to be arrested in rats fed with a vitamin A-deficient diet, it has been known that vitamin A is essential for sperm production in mammals (Wolbach and Howe, 1925). Subsequent studies demonstrated that vitamin A deficiency in rats induces a progressive loss of differentiating germ cells, ultimately yielding seminiferous tubules containing only Sertoli cells and pre-meiotic germ cells (Mitranond et al., 1979; Sobhon et al., 1979; Huang et al., 1988; Morales and Cavicchia, 2002), and administration of vitamin A to vitamin A-deficient rats resumes spermatogenesis (van Pelt and de Rooij, 1990b, 1991). The deficient state that impaired spermatogenesis in mice was also produced by a vitamin A-deficient diet (van Pelt and de Rooij, 1990a; Sugimoto et al., 2012; Boucheron-Houston et al., 2013; Ikami et al., 2015), in a manner similar to that in rats, as described above.
While most studies describe the effects of vitamin A deficiency on spermatogenesis in adult rodents, the role of long-term intake of an excess-vitamin A diet (VAE) in mouse spermatogenesis has not been well investigated. In fact, there were few studies about the long-term effects of VAE on spermatogenesis using rat models, but they only showed the effects of VAE on body and reproductive organ weights (Lamano Carvalho et al., 1978; Bosakowski et al., 1988). Further characterization is needed toward understanding the precise mechanism of male reproductive toxicity of long-term VAE consumption. Therefore, the present study was designed to determine whether long-term intake of VAE is associated with failure of testicular differentiation and spermatogenesis in male mice.
Five pregnant ICR mice were obtained from CLEA Japan, Inc. (Tokyo, Japan). On the day of birth (day 0), the litter size was standardized to eight pups to avoid fluctuations in the pups’ dietary intake. From gestation until weaning, all dams and their pups were fed an AIN-93G semi-purified diet containing sufficient vitamin A (4 IU/g diet) in the form of retinyl palmitate (Oriental Yeast Co., Ltd., Osaka, Japan). After weaning on postnatal day 22, male mice were randomly assigned to AIN-93G (control) or modified AIN-93G with VAE diet group. The control group (two litters per dam) was maintained on the AIN-93G semi-purified (control) diet. The diet of VAE group (two litters per dam) was changed to modified AIN-93G semi-purified diet containing VAE (1,000 IU/g). The doses used were part of a study aimed to observe the effects of high vitamin A intake on leptin expression in the mice epididymal white adipose tissue samples (Felipe et al., 2005) or on fatty acid composition of phospholipids in mice (Weiss et al., 2014), as previously reported. Both groups were fed each diet from 3 to 10 weeks of age in home cages at 23 ± 2°C on a 12-hr light/dark cycle (lights on from 8:00 to 20:00). Food and water were provided ad libitum during the experiment. Body weight was measured once a week from 3 to 10 weeks of age. Feed intake was estimated on a per-cage basis (2 mice/cage) from the actual amount of food consumed by the animals and its caloric equivalence (two weekly measurements, covering 3 and 4 days, respectively).
At 10 weeks of age, reproductive organs of male mice were obtained under sodium pentobarbital euthanasia. All efforts were made to minimize suffering. At the time of euthanasia, each mouse was weighed, blood was taken from the inferior vena cava, and the tissue was removed under a dim yellow safety light to prevent photoisomerization and photodegradation of retinoid (Yokota and Oshio, 2018). Dissected tissues except for the epididymis were rapidly frozen in liquid nitrogen and stored at –80°C. A detailed list of the respective analyses is shown in Supplemental Table 1.
All experiments were performed in accordance with the National Institute of Health (USA) guidelines for animal experiments and were approved by the Animal Care Committee of Ohu University (approval no. 2016-40).
Spermatozoa were collected based on our previous method (Oshio et al., 1990). Briefly, the cauda epididymis was dissected free of the connective tissue, fat pad, and vas deferens. For sperm sampling, both cauda epididymis from each animal were cut with a surgical blade. The excised cauda epididymis was minced with small scissors in 1 mL of 10 mM HEPES buffered-TYH culture medium (pH 7.4: 119 mM NaCl, 4.8 mM KCl, 1.7 mM CaCl2, 1.2 mM KH2PO4, 1.2 mM MgSO4, 25 mM NaHCO3, 5.6 mM glucose, 1.0 mM sodium pyruvate, and 4 mg/mL bovine serum albumin) in a sterile 1.5-mL tube. The sperm suspensions were allowed to disperse for 15 min on a warming tray at 37°C. The suspension was then filtered using high quality 40-μm nylon filter (PP-40N, Kyoshin Rikoh Inc., Tokyo, Japan) to remove any undigested tissue fragments, and the sperm was collected for evaluation of sperm count, sperm motility, sperm viability, and sperm morphology.
Sperm count, viability, and percent motility were analyzed under a phase contrast microscope (BX51, Olympus Co., Tokyo, Japan) using a computer-assisted sperm analysis (CASA) system (Sperm Class Analyzer, Microptic SL, Barcelona, Spain). Briefly, 20 μL of sperm suspension was placed in a counting chamber for the Standard Count 2 Chamber 20 micron Slide (Leja, Nieuw-Vennep, Netherlands). The filled chamber was maintained on a 37°C heated stage with a constant-temperature unit MP-10 (Kitazato Supply Co. Ltd., Shizuoka, Japan). Using a 10 × objective in phase contrast, the image was automatically relayed, digitized, and analyzed. The movement of at least 500 sperm cells was recorded from at least four random fields for each sample. Percentage of motile sperms was defined as Nm/Ns × 100 (%), where Nm is the number of fast and slow progressive motile sperms and Ns is the total number of sperms in the view area. Viability was also evaluated as previously described, with minor modifications (Moskovtsev and Librach, 2013). Briefly, an aliquot of the sperm suspension and 1% eosin Y solution (Merck & Co., Inc., Whitehouse station, NJ, USA) were mixed (1:1), and sperm viability was then determined by light microscopy. A minimum of 200 spermatozoa was counted per sample. Sperm morphology was evaluated under a phase contrast microscope. The percentage of broken spermatozoa was evaluated by counting at least 200 cells per slide (Oshio et al., 1990).
The method of DSP assesses the number of sperms produced in the testicles per day, and therefore provides a measure of the male reproductive function (Amann,1981). In short, a half of right testes were dissected and homogenized using a Polytron homogenizer (PT 1300D Homogenizer; KINEMATICA AG, Luzern, Switzerland) in homogenization buffer containing saline, 0.05% Triton X-100 (Nacalai Tesque, Inc., Kyoto, Japan), and 0.2% eosin-Y (Merck & Co., Inc.). The concentration of sperm nuclei in each suspension was determined using a hemocytometer under a light microscope. The counts served as a basis for the calculation of the number of homogenization-resistant spermatids per gram of testicular tissue (sperm content per gram of testicular parenchyma, SC/g). The total number of sperm nuclei in the left testicle was then calculated by multiplication with the weight of the left testicle. DSP was calculated using the following formula as previously described (Yoshida et al., 2010):
sperm count/mL × volume of lysis buffer = testicular sperm count
testicular sperm count / 4.84 = (sperm produced / day) (4.84 is a coefficient for calculating sperm)
(sperm produced / day) /testis weight = DSP
For histopathological analysis, the left testes were removed and immersed in a fixative within 24 hr at 4°C. The fixative used was modified Davidson’s fluid (mDF). We prepared mDF as described previously (Latendresse et al., 2002) and used it within a day. The composition of mDF is 30% of 37-40% formaldehyde, 15% ethanol, 5% glacial acid, and 50% distilled water. Following fixation, the testes were processed through graded alcohols, cleared in xylene, and embedded in paraffin. Four micrometer-sections were obtained, de-paraffinized, and stained with hematoxylin and eosin (HE) using our routine method for overall morphological evaluation. Slides were mounted using Entellan (Nacalai Tesque, Inc.) and air-dried prior to microscopic examination. The area of the seminiferous tubule was measured for 246 seminiferous tubules (78-85 seminiferous tubules/testis; n = 3). The seminiferous tubule area (μm2) was determined by measuring the length and width of the seminiferous tubule: calculating the seminiferous tubule area as the area of an ellipse = π × tubule length/2 (semi-major axis) × tubule width/2 (semi-minor axis). The ratio of seminiferous tubules containing elongated spermatids to those lacking elongated spermatids was assessed. At least five sections from each testis (three mice per group) were observed using light microscopy (Axio Imager A1, Carl Zeiss, Oberkochen, Germany) with Axio Vision software.
Total RNA of a half of right testes was extracted with ISOGEN (Nippon Gene Co., Ltd., Tokyo, Japan) according to the manufacturer’s protocol and suspended in diethyl-pyrocarbonate-treated water. RNA quantity and quality were confirmed using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). RNA samples with OD 260/280 ratio > 1.9 were used for further analysis. Total RNA (1 μg) with DNase treatment for each sample was used as a template to synthesize cDNA using M-MLV reverse transcriptase (Invitrogen Co., Carlsbad, CA, USA) according to the manufacturer’s instructions.
Real time PCR was performed with the SYBR Green Universal PCR Master Mix protocol (Applied Biosystems, Cheshire, U.K.) using an ABI Prism 7500 apparatus (Applied Biosystems) with an initial holding step (95°C for 10 min), and 40 cycles (95°C for 15 sec and 60°C for 60 sec) of a two-step PCR. At each cycle, the fluorescence intensity of each sample was measured to monitor target gene amplification. Relative target gene expression levels were calculated for each sample after normalization against glyceraldehyde-3-phosphate dehydrogenase (Gapdh). We found no significant differences in the Gapdh expression between the AIN-93G and VAE groups (data not shown). The primer sequences are listed in Table 1.
The serum testosterone concentration was measured by enzyme-linked immunosorbent assay (ELISA) (testosterone ELISA kit, Enzo Life Sciences, Inc., Farmingdale, NY, USA). Samples were diluted (1:60) and run in duplicate. This kit has a lower limit of detection of 5.67 pg/mL (range 7.81-2,000 pg/mL), an intra-assay coefficient of variation of 10.0%, and an inter-assay coefficient of variation of 11.3%.
Retinol and retinyl palmitate were extracted from the left testes under a dim yellow safety light. Retinol and retinyl palmitate quantification in the testes (pmol/g tissue) was performed by reverse-phase high performance liquid chromatography (HPLC) using a previously described procedure (Yokota and Oshio, 2018). Briefly, the testes were thawed, weighed, kept on ice, and homogenized in phosphate buffered saline (pH 7.4) (Thermo Fisher Scientific) using a disposable homogenizer. An equal volume of absolute ethanol containing a known amount of internal standard was added to an aliquot of testis homogenate. This mixture was briefly vortexed. Hexane (4 mL) was then added, and the mixture was vortexed twice so that the internal standard and endogenous retinol and retinyl palmitate were extracted into hexane. Solutions were centrifuged at 1500 × g for 5 min at 4°C. Thereafter, supernatants were treated with 200 μL sterile water and centrifuged at 1500 × g for 5 min at 4°C for purification of analytes. Hexane extracts (supernatants) were evaporated until dry under a gentle stream of nitrogen gas. Immediately upon drying completely, the analytes were dissolved in ethanol for injection into an HPLC/UV apparatus (PU-2080 plus chromatography pump, UV-2075 plus ultraviolet detector, and 807-IT integrator; JASCO, Tokyo, Japan) equipped with a Mightysil RP-18 GP column (Cat. No. 25416-96, 150 × 4.6 mm, 5-µm particle size; Kanto Chemical Co., Inc., Tokyo, Japan). The column was protected by a guard column (Cat. No. 25418-96; Kanto Chemical Co., Inc.). The injection volume was 20 µL for all samples using a universal loop injector (Rheodyne 7725i, Rheodyne L.P., Cotati, CA, USA). The mobile phase was 100% methanol delivered at a flow rate of 1.0 mL/min. Evaluation of chromatograms was performed using a detection wavelength of 325 nm. For peak assignment, each retinoid alone and/or mixtures were injected to determine an analysis sequence as the retention time-control sample (Yokota and Oshio, 2018). Based on the calibration lines obtained for each retinoid standard solution, intrinsic retinol and retinyl palmitate levels in the mice testes were determined.
Data are presented as the mean ± S.E. The non-parametric Mann-Whitney-Wilcoxon test was used to detect significant differences between the AIN-93G and VAE groups. Significance was set at P < 0.05.
Male ICR mice received either AIN-93G (control) or VAE diet from 3 to 10 weeks of age. With age, the gain in body weight was gradually decreased by long-term dietary consumption of VAE (Fig. 1A). Food intake (g) per day of VAE mice during the entire experimental period was also decreased compared with that of control mice (Fig. 1B).
Mean (A) food consumption (g) and (B) body weights of the control (AIN-93G: n = 10) and the VAE (n = 7) mice. Dietary intake was estimated on a per-cage basis (2 animals/cage) from the actual amount of food consumed by the animals (two weekly measurements, covering 3 and 4 days, respectively). Error bars represent S.E. Asterisks indicate significant differences between the control and VAE groups (*P < 0.05, ***P < 0.001). Abbreviations: VAE, vitamin A excess.
Absolute weight of testis relative to the body weight of VAE mice was significantly decreased compared with that of the control group (Table 2). However, VAE did not alter the accessory gland containing prostate and epididymis absolute weight relative to the body weight (Table 2).
Each value represents the mean ± S.E. p < 0.001 AIN-93G vs. VAE diet.
The percentage of motile sperm in VAE mice was significantly decreased (control: 40.0 ± 3.0%, VAE: 2.4 ± 0.8%; Fig. 2A; P < 0.01). The percentage of viable sperm in VAE mice also showed a significant decrease (control: 50.9 ± 1.3%, VAE: 9.0 ± 2.3%; Fig. 2B; P < 0.01) as compared with that in the control. Importantly, the sperm in two out of six VAE mice was not motile at all (sperm motility: 0%). The proportion of abnormal sperms was significantly increased in VAE mice compared to that in the control (control: 44.8 ± 1.1%, VAE: 82.4 ± 4.8%; Fig. 2C; P < 0.01). Sperm counts indicating the levels of DSP were (12.3 ± 3.9) × 107 (sperm counts/day/g testis) in the control group, but decreased to almost 2.1% of this value in the VAE group (Fig. 2D; P < 0.01). No sperm was found in four out of six VAE mice.
Sperm parameters in the control and VAE mice. (A) Sperm motility, (B) viability, (C) abnormal sperm morphology, and (D) DSP of the control (AIN-93G: n = 10) and VAE (n = 7) mice are shown. Data are presented as mean ± S.E. ***P < 0.001 vs control. Abbreviations: VAE, vitamin A excess; DSP, daily sperm production.
Overall histopathological examination of the testis tissue section revealed alterations in the point of atrophy between control and VAE mice (Fig. 3A, B). Testicular cross sections from control mice contained a normal array of germ cells at different developmental stages, namely, mitosis, meiosis, and spermiogenesis (Fig. 3C). However, the seminiferous tubules from VAE mice exhibited varying degrees of testicular degeneration, including impairment of the differentiation process and presence of vacuoles (Fig. 3D). Some tubules displayed substantially round/elongated spermatids, and sloughing of immature germ cells into the lumen was observed (Fig. 3D). In addition, histopathological findings, such as microcyst and atypical mitotic figures were observed (Fig. 3D), which indicated an increase in the number of abnormal tubules in VAE mice. The area of the seminiferous tubule was significantly decreased in the VAE group compared with that in the control group (Fig. 3E). The percentage of seminiferous tubules containing elongated spermatids was lower in the VAE group than in the control group (Fig. 3F).
Photographs of HE-stained testicular cross sections from control and VAE mice. The testicular morphology was examined by light microscopy. (A, C) Sections from the control mice testes. (B, D) Sections from the VAE mice testes. Black arrow: Atypical mitotic figure; Star: microcyst; V: Vacuole. (E) Measurements of the seminiferous tubule cross-sectional area from control (AIN-93G: 246 tubules, n = 3) and VAE (192 tubules, n = 3) mice. (F) The percentage of seminiferous tubules with elongated spermatids was measured. Each column is expressed as the mean [or relative values to the percentage of control (100%)] ± S.E.M. Asterisks indicate significant differences between the control and VAE groups (*P < 0.05, **P < 0.01). For details on how the measurements were taken, please refer to the ‘Materials and Methods’ section. Scale bars: 500 μm (A, B), 20 μm (C, D). Abbreviations: VAE, vitamin A excess.
We measured the gene expression levels of the germ cell markers: stimulated by retinoic acid gene 8 (Stra8), zinc finger and BTB domain containing 16 (Zbtb16), neurogenin 3 (Ngn3), spermatogenesis and oogenesis specific basic helix-loop-helix 2 (Sohlh2), KIT proto-oncogene receptor tyrosine kinase (C-kit), DNA meiotic recombinase 1 (Dmc1), synaptonemal complex protein 3 (Sycp3), and protamine 1 (Prm1). Stra8 and Zbtb16 mRNA expression levels were significantly increased as a result of the VAE diet (Fig. 4A, B). The expression of Ngn3, Sycp3, and Prm1 was significantly decreased as a result of the VAE diet (Fig. 4C, G, H). There were no significant differences in Sohlh2, C-kit, and Dmc1 mRNA expression levels between the control and VAE groups (Fig. 4D-F).
Impact of long-term consumption of VAE on the mRNA expression of germ cell-specific marker genes. (A) Stra8, (B) Zbtb16, (C) Ngn3, (D) Soh1h2, (E) C-kit, (F) Dmc1, (G) Sycp3, and (H) Prm1 mRNA levels in the testes. Fold-change of expression of genes in each group has been shown. The data are expressed as relative target gene expression compared with Gapdh expression. Each column represents the mean ± S.E. Asterisks indicate significant differences between the AIN-93G and VAE groups (AIN-93G: n = 10, VAE: n = 7). *P < 0.05, **P < 0.01. Abbreviations: VAE, vitamin A excess; Gapdh, glyceraldehyde-3-phosphate dehydrogenase; Stra8, retinoic acid gene 8; Zbtb16, zinc finger and BTB domain containing 16; Ngn3, neurogenin 3; Sohlh2, spermatogenesis and oogenesis specific basic helix-loop-helix 2; C-kit, KIT proto-oncogene receptor tyrosine kinase; Dmc1, DNA meiotic recombinase 1; Sycp3, synaptonemal complex protein 3; Prm1, protamine 1.
We measured the expression levels of genes involved in the retinoid signaling pathway: retinoic acid receptor alpha (Rarα), retinoic acid receptor beta (Rarβ), retinoic acid receptor gamma (Rarγ), retinoid X receptor alpha (Rxrα), retinoid X receptor gamma (Rxrγ), lecithin-retinol acyltransferase (Lrat), aldehyde dehydrogenase family 1 subfamily a2 (Aldh1a2), and cytochrome P450 family 26 subfamily a1 (Cyp26a1). Quantitative PCR results showed a significant increase in the expression levels of Rarα and Cyp26a1 in VAE mice testes (Fig. 5A, H), and a significant decrease in the expression levels of Rarβ, Rxrγ, Lrat, and Aldh1a2 (Fig. 5B, E, F, G). In addition, the mRNA expression level of Rxrα) in the testis of VAE mice appeared to decrease compared with that in the control (Fig. 5D; 0.05 < P < 0.1). There was no significant difference in Rarγ mRNA expression between the control and VAE groups (Fig. 5C).
Impact of long-term consumption of VAE on the mRNA expression of genes involved in retinoid signaling. (A) Rarα, (B) Rarβ, (C) Rarγ, (D) Rxrα, (E) Rxrγ, (F) Lrat, (G) Aldh1a2, and (H) Cyp26a1 mRNA levels in the testes. Fold-change of expression of genes in each group has been shown. The data are expressed as relative target gene expression compared with Gapdh expression. Each column represents the mean ± S.E. Asterisks indicate significant differences between the AIN-93G and VAE groups (AIN-93G: n = 10, VAE: n = 7). *P < 0.05, **P < 0.01. Abbreviations: VAE, vitamin A excess; Rarα, retinoic acid receptor alpha; Rarβ, retinoic acid receptor beta; Rarγ, retinoic acid receptor gamma; Rxrα, retinoid X receptor alpha; Rxrγ, retinoid X receptor gamma; Lrat, lecithin-retinol acyltransferase; Aldh1a2, aldehyde dehydrogenase family 1 subfamily a2; Cyp26a1, cytochrome P450 family 26 subfamily a1.
Serum testosterone levels were significantly higher in VAE males (mean ± S.E.: 35.1 ± 16.4 ng/mL) than in control males (mean ± S.E.: 5.2 ± 1.8 ng/mL; P < 0.05; Fig. 6).
Measurement of serum testosterone levels. Each column shows the concentration of serum testosterone (ng/mL) in the serum sample. The white column shows data from mice that had ingested the AIN-93G purified diet (n = 10). The black column shows data from mice that had ingested the VAE diet (n = 7). The asterisk indicates significant differences in serum testosterone levels between the AIN-93G and VAE diet groups. *P < 0.05. Abbreviations: VAE, vitamin A excess.
The retinol concentration in the testes of VAE mice was significantly higher than that of the control (95.0 ± 11.5 vs. 12143.9 ± 1860.4 pmol/g tissue) (Fig. 7A). The retinyl palmitate concentration in the testes of VAE mice also significantly increased compared with that of the control (2.9 ± 0.1 vs. 1453.1 ± 285.6 nmol/g tissue) (Fig. 7B).
Retinoid accumulation in the control and VAE testes. (A) Each column shows the concentration of retinol in the testes (pmol/g tissue). The white column shows data from mice that had ingested the AIN-93G purified diet (n = 5). The black column shows data from mice that had ingested the VAE diet (n = 3). (B) Each column shows the concentration of retinyl palmitate in the testes (nmol/g tissue). The white column shows data from mice that had ingested the AIN-93G purified diet (n = 5). The black column shows data from mice that had ingested the VAE diet (n = 3). Values are expressed as mean ± S.E. The asterisk indicates significant differences in each retinoid concentration between the AIN-93G and VAE diet groups. ***P < 0.001. Abbreviations: VAE, vitamin A excess.
Vitamin A has long been recognized to be essential for spermatogenesis using vitamin A-deficient rodent models (Wolbach and Howe, 1925; Mitranond et al., 1979; Sobhon et al., 1979; Huang et al., 1988; Morales and Cavicchia, 2002), and acute administration of retinol or retinoic acid has been shown to re-initiate spermatogenesis in vitamin A-deficient rodents (van Pelt and de Rooij, 1990b, 1991). However, although there is a small amount of evidences about the long-term effects of VAE on rat spermatogenesis, previous studies focused on the effects on body and reproductive organ weights, which have not been previously well characterized (Lamano Carvalho et al., 1978; Bosakowski et al., 1988). The present study was conducted to investigate whether long-term intake of VAE arrested spermatogenesis in mice, and it demonstrated for the first time the molecular mechanism of spermatogenic arrest, which was partly different from that in vitamin A-deficient mice as reported previously (Boucheron-Houston et al., 2013). The vitamin A intake (1,000 IU/g diet) in the present study was lesser than in previous reports (Bosakowski et al., 1988) and could not cause teratogenicity (10,000 IU/day) (Miller et al., 1998).
We showed for the first time that long-term intake of VAE resulted in failure of testicular differentiation and spermatogenesis in mice. Both the percentage of epididymal sperm motility and viability in VAE mice was less than 10%. Abnormal sperm morphology in VAE mice was significantly enhanced compared to that in the control. These results indicated that long-term VAE intake affected the process of spermatogenesis. In addition, the results of DSP also revealed that differentiation to spermatids in VAE mice significantly decreased to almost 0%, which indicated the possibility of VAE-induced testicular toxicity. Indeed, the absolute weight of the testis, but not of the other reproductive glands, relative to the body weight of VAE mice decreased significantly compared to that of control mice. These results demonstrated severe and selective testicular toxicity in VAE mice. These results are in contrast to those from a previous study (Bosakowski et al., 1988) in which rodents on VAE diet, receiving a two-fold higher dose of vitamin A than that used in the present study, showed only reduction in body weight and no reduction in absolute weight of testis relative to the reduction in body weight. Additionally, in the previous study (Bosakowski et al., 1988), 5 to 18-week-old rodents were fed VAE diet for a period of 13 weeks, whereas in our study 3 to 10-week-old mice were fed VAE diet for a period of 7 weeks. It was likely that the effect of VAE toxicity on spermatogenesis was more severe when exposed to VAE at a younger age than that at an older age.
Besides evaluation of reproductive parameters, we focused on the detection of changes in gene expressions involved in the germ cell markers: Stra8, Zbtb16, Ngn3, Sohlh2, C-kit, Dmc1, Sycp3, and Prm1. Zbtb16 is expressed by As-Aal spermatogonia, and loss of Zbtb16 causes a reduction in the number of undifferentiated spermatogonia and an increase in the number of differentiating spermatogonia (Buaas et al., 2004; Hobbs et al., 2010). In addition, recent studies have revealed that the gene expression profiles of undifferentiated spermatogonia are heterogeneous (Meng et al., 2000; Yoshida et al., 2004; Hofmann et al., 2005). In particular, glial cell line-derived neurotrophic factor receptor α1 (Gfrα1), and Ngn3, a basic helix-loop-helix transcription factor, were reciprocally expressed in undifferentiated spermatogonia. GFRα1-positive cells mainly comprised of As and Apr, whereas NGN3-positive cells included a majority of Aal and a smaller number of As and Apr (Hofmann et al., 2005, Yoshida et al., 2004). There were two processes involved in germ cell development. In the first one, Apr produced two new As spermatogonia (self-renew), and in the other, Apr produced a chain of four Aal spermatogonia at the next division (Phillips et al., 2010). Our results of increased Zbtb16 mRNA expression, but decreased Ngn3 mRNA expression in the VAE testes strongly suggested that long-term VAE intake directly affected undifferentiated early spermatogonia, facilitating stem cell self-renewal.
To confirm the possibility, we also tested the gene expression levels of post-meiotic germ cell markers. Sycp3, a specific marker for spermatocytes and a major component of the synaptonemal complex, plays a crucial role in synapsis and recombination during meiosis (Miyamoto et al., 2003; Schramm et al., 2011; Snyder et al., 2011). Decreased expression of Sycp3 mRNA in the VAE mice testes indicated decreased differentiation to spermatocytes, probably due to inhibited production of A1 spermatogonia and meiosis initiation in the VAE testes (as discussed above). Recent studies have shown that Sycp3 expression is required for centromere pairing in mice, which is important for proper chromosome segregation (Bisig et al., 2012). Errors in meiotic chromosome segregation are the main cause of human aneuploidy, and therefore lack of Sycp3 mRNA expression could have a negative effect on the sperm quality of VAE mice. Moreover, a dramatic reduction in Prm1 mRNA expression level in the VAE testes, compared to that in control mice, has been noted. Prm1 is expressed exclusively in post-meiotic haploid spermatids (Kleene et al., 1988). The Prm1 transcript was first detected in round spermatids and then translated in elongated spermatids (Esakky et al., 2013). The data of decreased Sycp3 and Prm1 mRNA expression in the VAE testes were correlated with that of decreased DSP, and also with the data from abnormal histopathological findings in the seminiferous tubules. Indeed, quantitative histopathological analysis revealed a significant reduction in the number of the seminiferous tubules containing elongated spermatids. Stra8 is a marker of spermatogonia, spermatocyte, and spermatids, and a known direct target of retinoic acid that initiates meiosis (Koubova et al., 2006; Anderson et al., 2008). Our results revealed that Stra8 expression in the VAE testes was increased more than 2-fold compared to that in control mice, which likely caused meiosis induction by long-term VAE intake. However, as discussed above, the number of spermatocytes and round/elongated spermatids in the VAE testes were remarkably decreased, preventing them from entering meiosis. The significant increase in Stra8 expression in the VAE testes might be due to the dramatic increase in the number of undifferentiated spermatogonia, which is consistent with the result of Zbtb16 mRNA expression. Another potential mechanism was the significant increase (100-fold) in retinol accumulation in the VAE testes, which could stimulate the expression of Stra8, a known direct target of retinoic acid. Therefore, it was necessary to determine the gene expression changes involved in retinoid signaling and metabolism. The mRNA expression levels of Lrat and Aldh1a2 were significantly decreased, while that of Cyp26a1 was significantly increased compared to that in the control, suggesting that excessive accumulation of retinoid in the VAE testes induced an increase in metabolic rate. In addition, Aldh1a2 was expressed in late spermatocytes (Vernet et al., 2006). Decreased expression of Aldh1a2 mRNA in the VAE mice testes also indicated decreased differentiation to spermatocyte.
Retinol in the Sertoli cells was transported from the circulating blood, following which it was oxidized into retinoic acid in the Sertoli cells, leading to an observation that Sertoli cells were the main sites of retinoic acid synthesis (Cavazzini et al., 1996). The two major active isoforms of retinoic acid, all trans retinoic acid (ATRA) and 9-cis retinoic acid (9-cis RA), both exert pleiotropic effects of vitamin A by transcriptional regulation of target genes via a class of nuclear receptors comprised of two subfamilies: the RARs and RXRs (Chambon, 1996). In normal mice testes, the cellular localization of these retinoid receptors has been exclusively studied by in situ hybridization and immunohistochemical analysis (Vernet et al., 2006). Rarβ and Rxrγ mRNA expressed in step 7 and 8 round spermatids (Vernet et al., 2006) in the testes were decreased by VAE intake, identical to the changes induced by vitamin A deficiency (Boucheron-Houston et al., 2013). These results suggested that excess of vitamin A or its deficiency in mice resulted in decreased differentiation of step 7 and 8 round spermatids.
More than a 24-week vitamin A deficiency leads to reduction of Rarα expression in the mice testes (Boucheron-Houston et al., 2013). Therefore, we expected that long-term VAE intake may also lead to decreased Rarα expression in the testes because testes of both vitamin A deficient and excess mice showed the same changes in the expression of genes related to retinoid signals such as Rarβ and Rxrγ. Contrary to our expectation, Rarα expression was significantly increased in the VAE testes, which seemingly confounded the interpretation of the results. Rarα was expressed only in Sertoli cells, but not in germ cells (Vernet et al., 2006). Histopathological observations confirmed the presence of Sertoli cells in the seminiferous tubules of both VAE and control mice. We speculated that the increase in Rarα expression in the VAE testes was due to an increased response to retinoid signaling in the Sertoli cells.
In the cross sections of the seminiferous tubules of VAE mice, Leydig cells were observed similar to that of control mice. Leydig cells produce testosterone to maintain spermatogenesis (Walker, 2011). There was a significant increase in serum testosterone levels in VAE mice compared to that in the control. It has been reported that steroidogenesis, such as testosterone and estrogen synthesis, is stimulated by retinol and retinoic acid in the peripheral steroidogenic organs. Using rat Leydig cells in primary culture, both retinoic acid and retinol have been shown to stimulate testosterone production (Chaudhary et al., 1989). In addition, serum testosterone levels in retinol-deficient rats have been shown to be significantly lower than that in control rats (Appling and Chytil, 1981). Administration of retinoic acid stimulates serum (Appling and Chytil, 1981) and Leydig cell (Chaudhary et al., 1989) testosterone production. In addition, vitamin A is also involved in estrogen activity. Vitamin A insufficiency exacerbates the decrease in the number of sperms induced by exposure to endocrine disruptors, most of which are categorized as estrogenic substances interacting with estrogen receptors, and their effect has been shown to be suppressed by retinoic acid administration (Nakahashi et al., 2001). Notably, estrogen receptor α knockout male infertile mice, exhibiting disrupted spermatogenesis and seminiferous epithelium disorganization, display elevated serum testosterone levels (Eddy et al., 1996). Dysfunction of testes in these knockout mice was similar to that observed in the VAE testes. These reports suggested the possibility that long-term VAE intake could affect spermatogenesis via excessively elevated steroidogenesis, like testosterone biosynthesis, which cause spermatogenesis defect. Further investigations are needed to examine the precise mechanisms of elevated steroidogenesis in VAE mice.
Finally, it is also important to consider the possibility of indirect mechanisms of spermatogenic arrest by VAE. Because our results showed body weight reduction in VAE mice, we were concerned with the possibility that this reduction in body weight could affect spermatogenesis. A previous study showed that in utero protein restriction caused body weight reduction and reduced sperm count in Wistar rat male offspring (Toledo et al., 2011). However, testicular histological examination revealed that histology of seminiferous tubules in the male offspring of protein-restricted mothers was similar to that of standard chow-fed mothers. Another study also revealed that male mice subjected to undernutrition from birth to weaning by separating pups from their mothers showed body weight reduction and subsequently induced delay in their growth (Jean-Faucher et al., 1982). Although the establishment of spermatogenesis and the appearance of mature Leydig cells in undernourished males were delayed, testicular histological examination demonstrated that histology of the seminiferous tubules in the undernourished group was similar to that of the control group. In contrast to these, our results showed that the seminiferous tubules from VAE mice exhibited varying degrees of testicular degeneration. A comparison of the histological findings of testes from malnourished mice with those of testes from VAE mice, suggested that the degenerated seminiferous tubules were caused by VAE diet. However, reduction in sperm count and motility could be induced not only by VAE, but also by body weight reduction and malnutrition.
Our observations provided important information on the impact of long-term VAE diet on spermatogenesis arrest and its underlying molecular mechanisms. The transition of Aal into A1 spermatogonia marks the start of the strictly time-regulated process of spermatogenesis (Jan et al., 2012). A block in differentiation into A1 spermatogonia was observed in vitamin A-deficient rodents, demonstrating that this step was dependent on retinoic acid (van Pelt and de Rooij, 1990a, 1990b, 1991). However, how various vitamin A conditions, such as low, moderate, or high retinol diet intake, decide germ cell fate in vivo is still largely unknown. The present findings gave novel insights into the effects of excessive consumption of vitamin A on the male reproductive system. Further work in this area of reproductive toxicology is necessary to discern the molecular mechanisms responsible for the physiological and pathological roles of vitamin A in male reproductive physiology.
We thank Mr. Takao Sekine and Shiori Suzuki (Ohu University) for technical assistance. We thank Editage (https://www.editage.jp) for their English language editing service. This work was supported by a JSPS KAKENHI Grant Number 17K15854 to Satoshi Yokota (https://kaken.nii.ac.jp/ja/grant/KAKENHI-PROJECT-17K15854/).
The authors declare that there is no conflict of interest.