Isolation of tail DNA and sperm head DNA

To obtain tail DNA for genotyping, tail snips of 0.5 cm were cut and digested using Viagen DirectPCR lysis reagent (Viagen Biotechnology, Los Angeles, CA) with proteinase K (Life Technologies, Carlsbad, CA) at 55°C for 4 to 6 hours or digested with MyTaq extract-PCR fast digestion kit (Bioline, Luckenwalde, Germany), following the manufacturers’ protocols. Crude DNA lysates were obtained after the digestion and deactivation of proteinase K at 85°C for 45 minutes.

For isolation of sperm head DNA, fresh epididymal sperm from caudal epididymides were collected and suspended in nucleus isolation medium (NIM) containing 121.6 mM KCl, 7.8 mM Na₂HPO₄, 1.4 mM KH₂PO₄, 0.1% polyvinyl alcohol, and 10 mM EDTA. Sperm heads and tails were detached by sonication (Branson Sonifier 250; Emerson, St. Louis, MO) and centrifuged to collect the pellets. The sperm pellet was then resuspended in one part fresh NIM buffer layered over two parts 70% sucrose and centrifuged for 1 hour to separate sperm heads from tails. Separation of heads from tails was confirmed under an inverted phase-contrast Olympus IMT-2 microscope (Olympus, Center Valley, PA).

The sperm head pellet was suspended in NIM buffer, containing 0.1 M dithiothreitol and DNA extraction buffer [1 M Tris-HCl (pH 8), 0.5 M EDTA, and 10% SDS], and incubated for 20 minutes at 65°C to release nuclear content. Proteins were digested with proteinase K for 1 hour at 55°C, precipitated on ice for 15 minutes in protein precipitation solution (Promega Corporation, Madison, WI), and centrifuged at 4°C for 20 minutes. Equal volumes of ice-cold isopropanol and Glycoblue (Invitrogen, Carlsbad, CA) were added and mixed by inversion, incubated at −20°C for 2 hours, and centrifuged at 13,500 rpm at 4°C for 30 minutes to precipitate DNA. The DNA pellet was washed with 70% cold ethanol twice, dried, and resuspended with distilled water for use in genotyping.

Genotyping

For genotyping, the typical PCR reaction mix was 25 μL, composed of 1× GoTaq PCR buffer (Promega Corporation), 0.25 mM 2′-deoxynucleoside 5′-triphosphate (Invitrogen), 0.5 μM forward and reverse primers (Invitrogen), 0.25 U of GoTaq (Promega Corporation), and 1 μL of tail or sperm head DNA solution. Forward and reverse primer pairs for genotyping the Stra8-iCre transgene and the Rara⁺, Rara^fl, and Rara⁻ alleles are listed in Table 1. Rara-specific primer positions are illustrated on the Rara gene, shown in Fig. 1(b).

Testicular tissue collection

Male mice at postnatal day (P)1 to P365 were humanely euthanized and testes were collected, fixed in 4% paraformaldehyde (PFA) or Bouin solution for 30 minutes to 6 hours, dependent on the size of testis, embedded in paraffin wax, cut into 4-μm-thick sections, and mounted on HistoBond slides (VWR, Radnor, PA).

Germ cell enrichment, RNA isolation, and reverse transcription and real-time quantitative PCR

Germ cells were enriched from testes using a two-step digestion process, as described (31). Briefly, testes were removed from P4 and P8 animals, detunicated, and tubules were placed in a digestion media of Hanks balanced salt solution (Sigma-Aldrich, St. Louis, MO) and trypsin (Gibco, Thermo Fisher Scientific, Waltham, MA). After shearing of tubules via pipette mixing, cells were incubated at 37°C, centrifuged, resuspended in a second digestion media composed of trypsin (Gibco), Hanks balanced salt solution, collagenase, DNase, and hyaluronidase (Sigma-Aldrich), and incubated for 15 to 20 minutes at 37°C. After fetal bovine serum (Sigma-Aldrich) was added to cells and incubated for another 5 minutes at 37°C to stop enzyme activities, cells were passed through a 40-μm mesh sieve screen (Corning, Corning, NY) and centrifuged. Enriched cells were then used for RNA isolation using TRIzol reagent (Invitrogen), following the manufacturer’s protocol.

cDNA was synthesized (reverse transcribed) from 500 ng of total RNA isolated from germ cells using an iScript cDNA synthesis kit (Bio-Rad Laboratories, Hercules, CA). A reaction mixture of 25 μL contained 12.5 μL of 2× Power SYBR Green PCR master mix (Applied Biosystems Technology, Foster City, CA), 500 nM forward and reverse primers, 5 μL of diluted cDNA (1:20 dilution), and distilled water. Real-time quantitative PCR (qPCR) was performed as follows: 1 cycle at 50°C for 2 minutes, 1 cycle at 95°C for 10 minutes, 40 cycles at 95°C for 15 seconds, and 1 cycle at 60°C for 1 minute (Applied Biosystems 7500 Fast real-time PCR system). Forward and reverse primers for real-time qPCR were designed using Primer Express version 2.0 (Applied Biosystems Technology) and are listed in Table 2. The real-time RT-qPCR was conducted on at least three animal samples with each sample in a technical duplicate or triplicate. The abundance of mRNA was evaluated using the 2^−ΔCT method (32). Cycle threshold values for Rara and the two housekeeping genes for ribosomal S2 protein (Rps2) and glyceraldehyde-3-phosphate dehydrogenase (Gapdh) were determined using the 7500 software version 2.0.1 (Applied Biosystems Technology). Cycle threshold values for Rara and Stra8 were normalized to those of either Rps2 or Gapdh values in each sample, and the abundance of Rara and Stra8 were calculated relative to the level in the reference sample (P8 WT value for each target).

Immunohistochemistry and immunofluorescence of tissue sections

Testicular sections fixed in 4% PFA were deparaffinized, and antigens were heat retrieved via microwaving slide sections in 0.01 M sodium citrate solution. Antibody incubations were conducted in a humidified chamber. For immunohistochemistry (IHC), tissue sections were blocked with 10% normal goat serum (Vector Laboratories, Burlingame, CA) at room temperature and incubated with a primary antibody made in rabbit at 4°C overnight, followed by biotinylated goat anti-rabbit IgG secondary antibody (Vector Laboratories) at room temperature. Sections were further washed in PBS, treated with streptavidin-peroxidase (Vector Laboratories) and DAB-Plus substrate kit (Invitrogen), counterstained with hematoxylin, and sealed with HistoMount (Invitrogen).

For immunofluorescence (IF) using mouse monoclonal antibody, the Mouse on Mouse basic kit (Vector Laboratories) was used following the manufacturer’s protocol. Briefly, tissue sections were first blocked with blocking reagent from the kit, incubated with primary antibody overnight at 4°C, washed in PBS, and then incubated with secondary antibody supplied in the kit. For IF using rabbit polyclonal antibody, tissue sections were treated with 0.2% Triton X-100, blocked with 10% normal goat serum, incubated with primary antibody overnight at 4°C, washed in PBS, incubated with biotinylated anti-rabbit IgG secondary antibody (Vector Laboratories), followed by streptavidin-conjugated Alexa Fluor 488 or 555 (Life Technologies), and mounted with VectaShield reagent (Vector Laboratories). If needed, 4′,6-diamidino-2-phenylindole (DAPI) (1:10,000; Life Technologies) was included with streptavidin-conjugated Alexa Fluor to counterstain DNA. IHC and IF sections were visualized using a Leica DMRB microscope (Leica, Wetzlar, Germany). Digital images were obtained using a Leica DFC 310 FX digital camera.

Sperm count and fertility analysis

Sperm were collected from the caudal epididymis of adult mice at P75, P120, P180, and P365. Specifically, epididymides were cut into pieces and placed in PBS at 37°C for 5 minutes and gently pipette mixed to facilitate sperm release. Fresh sperm were counted on a hemocytometer under an inverted phase-contrast Olympus IMT-2 microscope. For fertility analysis, Rara cKO males at P120, P180, and P300 were mated to C57BL/6J WT females aged between P50 and P126. Breeding pairs were kept together for 1 week, and vaginal plugs were confirmed before separation. Number and sex of live pups before P5 were recorded for multiple litters per male breeder age.

Biotin permeability assay

The integrity of Sertoli cell tight junctions was examined, as previously described with some modifications (33). Male mice at P60 to P75 were humanely euthanized, and testes from WT and Rara cKO mice were collected. Then, 10 mg/mL water-soluble EZ-Link Sulfo-NHS-LC-Biotin reagent (Thermo Fisher Scientific, Rockford, IL) in PBS with 1 mM calcium chloride (PBS-C) was injected into testes, followed by submersion of testes in biotin solution, and incubation at room temperature for 30 minutes. Negative controls were injected with and submerged in PBS-C. Testes were then fixed in 4% PFA solution overnight at 4°C, embedded in paraffin wax, cut into 4-μm-thick sections, and mounted on HistoBond slides (VWR). The slides were further incubated with streptavidin-conjugated Alexa Fluor 555 (Life Technologies) and mounted with VectaShield reagent (Vector Laboratories). Sections were visualized using a Leica DMRB microscope. Digital images were obtained using a Leica DFC 310 FX digital camera.

Spermatocyte chromosomal spreads and immunostaining

Chromosomes of spermatocytes from P15, P17, P19, and P21 pups were spread on glass slides, as previously described (34). Briefly, seminiferous tubules were isolated and submerged in a defined water-based hypotonic buffer for the swelling and release of spermatocytes from tubules. Spermatocyte suspension was transferred to a precleaned glass slide coated with 1.1% PFA (pH 9.2) containing Triton X-100. Spermatocytes were dispersed across the slide by manual rocking. Slides were dried overnight in a humid chamber, cleaned with 0.04% Photo-Flo (Kodak, Rochester, NY), and air-dried before staining.

For immunostaining, chromosomal spreads were blocked in 10% normal donkey serum solution (Jackson ImmunoResearch Laboratories, West Grove, PA) in PBS containing 3% BSA (Sigma-Aldrich) and 0.05% Triton X-100 for 60 to 90 minutes at room temperature. Antibody incubations were conducted at 37°C in a humidified chamber with plastic or glass coverslips. When the rabbit polyclonal primary anti- synaptonemal complex protein 3 (SYCP3) antibody purchased from Abcam (Cambridge, MA) (35) was used at 1:100 dilution, the secondary antibody used was rhodamine donkey anti-rabbit (1:50; Jackson ImmunoResearch Laboratories). When the mouse monoclonal primary anti-γH2AX antibody purchased from MilliporeSigma (Burlington, MA) (36) was used at 1:750 dilution, the secondary antibody used was FITC donkey anti-mouse (1:50; Jackson ImmunoResearch Laboratories). After antibody incubations, spreads were washed in PBS, counterstained with DAPI, rinsed with distilled water, and sealed with VectaShield (Vector Laboratories) and rubber cement. Imaging for meiotic stage progression and meiotic defects was performed using Leica DM6B and Zeiss Axiovert 200M microscopes (Zeiss, Oberkochen, Germany).

TUNEL assay

Testis sections fixed in 4% PFA from P15, P17, and P19 were deparaffinized in preparation for the TUNEL apoptotic assay. The TUNEL assay labels DNA ends indicating DNA breaks, a signature of cells undergoing apoptosis (37). Apoptotic cells were detected according to the manufacturer’s instructions (Promega Corporation). As a positive control, tissue sections were pretreated with deoxyribonuclease I (Promega Corporation). Conversely, tissue sections were incubated without terminal deoxynucleotidyl transferase recombinant enzyme for negative control. Sections were visualized using a Leitz DMRB microscope, and digital images were obtained using a Leica DFC 310 FX digital camera.

Busulfan treatment and restoration of spermatogenesis in the seminiferous tubules

Animals aged 31 to 43 days were injected with 35 mg/kg body weight with busulfan (Sigma-Aldrich) and left to recover for either 4 or 8 weeks (38, 39). Testes were collected after recovery periods, fixed in 4% PFA, mounted in paraffin wax, cut into 4-μm-thick sections, and mounted on HistoBond slides (VWR). To examine the recovery of spermatogenesis in the seminiferous tubules, testicular sections from mice recovered for 4 weeks or 8 weeks were deparaffinized and stained with hematoxylin and eosin. Sections were examined using a Leica DMRB microscope, and digital images were obtained using a Leica DFC 310 FX digital camera.

Germ cell–specific deletion of Rara

It was known that Stra8-iCre transgenic mice express the iCre transgene as early as P3 in male germ cells (41). Previously, the morphology of the seminiferous tubules from global heterozygous Rara mutant mice was examined and found that histological abnormalities were minor, compared with the dramatic disruption in germ cell organization observed in Rara-null seminiferous tubules, with both Rara alleles deleted (19). Thus, to achieve an in-frame excision of both Rara^fl alleles in germ cells, a two-step breeding scheme was conducted to generate germ cell–specific Rara cKO pups with Stra8^iCre/+;Rara^−/− germ cell genotype [Fig. 1(b)]. In the first step of the breeding scheme, Rara^fl/fl females (C57BL/6J genetic background) were mated to Stra8-iCre males (FVB genetic background) to generate Stra8^iCre/+;Rara^fl/+ F₁ generation heterozygotes. In the second step, F₁ males were bred to Rara^fl/fl females to generate F₂ pups. DNA from tail snips of the F₁ and F₂ generations was used to genotype, and only F₂ pups with Stra8^iCre/+;Rara^fl/− tail genotype, which is homozygous knockout in germ cells and heterozygous in somatic cells, were used for Rara cKO sample collection [Fig. 1(b)]. WT control animals were generated in a similar way as the two-step breeding scheme except that C57BL/6J-Rara^+/+ females were used in place of Rara^fl/fl females. In this way, the F₂ generation WT males have the same strain genetic composition (75% C57BL/6J and 25% FVB) as do the Rara cKO mice, but they have not been exposed to the Rara^fl gene. Animal genotypes were confirmed using PCR, as described in the “Materials and Methods” section. Locations of primer pairs are shown in Fig. 1(b) and primer sequences are listed in Table 1.

Expression of Stra8-iCre using R26R-EYFP mice

The earliest expression of Stra8-iCre during mouse development was investigated using R26R-EYFP mice (The Jackson Laboratory), which have a loxP-flanked STOP sequence following the enhanced yellow fluorescent protein (Eyfp) gene inserted into the Gt(ROSA)26Sor locus [Fig. 2(a)]. To determine the expression of iCRE between P0 and P7 in detail, R26R-EYFP females were mated to Stra8-iCre males to generate Eyfp;Stra8-iCre F₁ (Eyfp F₁) pups that express EYFP in the Stra8-iCre–expressing germ cells. In Eyfp F₁ mice, iCRE recombinase in Stra8-iCre–expressing germ cells recombines the two loxP sites, deleting the STOP sequence and allowing the expression of EYFP [Fig. 2(a)].

Open in a separate window

Figure 2.

Verification of Rara inactivation in germ cells. (a) Mating between the R26R-Eyfp females carrying an iCRE-activity reporter construct and Stra8-iCre males leads to EYFP expression in germ cells. CCAG, chicken β-actin promoter. (b–e) Green fluorescent cells are germ cells with iCRE activity (white arrowheads) in testis cross-sections from animals at P1, P3, P5, and P7. (f) Gonocytes with iCRE activity are EYFP- and ZBTB16-positive germ cells (white pointed arrowheads) in P1 testis cross-sections. Scale bars, 50 µm. (g) Percentage of tubules containing at least one EYFP-positive cell in animals at P1, P3, P5, and P7. n = 3 to 4 for P1, P3, and P7; n = 2 for P5. Error bars are SD. (h and i) IHC for RARA on testicular sections from WT and Rara cKO mice at P8. Black arrowheads indicate Sertoli cells; black arrows indicate germ cells. n = 3. Scale bars, 50 µm. (j) PCR products obtained using DNA template isolated from epididymal sperm heads in WT and cKO animals at P75. A 487-bp fragment represents the deleted form of Rara allele, and a 554-bp fragment represents the WT Rara allele. n = 3 to 5. (k) Real-time RT-qPCR analysis on RNA collected from enriched germ cell populations from WT (open bar) and Rara cKO (cross-hatched bar) animals at P4 and P8 for Rara (***P < 0.001; WT and cKO n = 3) and Stra8 (**P < 0.01, ***P < 0.001; WT n = 6, cKO n = 4). Error bars are SEM. L, DNA ladder.

Fresh testicular tubules were isolated from mice at P0, P1, P2, P3, P4, and P5 and examined for EYFP expression under fluorescent light. iCRE recombinase was active as early as P0 with 20% of F₁ pups showing EYFP signal, and from P1 to P5 with 100% of F₁ pups showing EYFP expression in the tubules (n = 5 to 7; data not shown). To further investigate the expression of EYFP in germ cells, testicular sections were immunostained with rabbit primary polyclonal anti-GFP antibody purchased from Abcam (42) at 1:500 dilution, which detects the protein product of Eyfp, and then were scored for tubules containing one or more EYFP-positive cells [Fig. 2(b)–2(e), arrowheads). A mean total number of 50 tubules per animal were scored. At P1 and P3, 52.2% and 58.2% of tubules had at least one EYFP-positive germ cell, respectively [Fig. 2(g)]. At P5 and P7, virtually all tubules contained EYFP-positive germ cells (92.6% and 98.6%, respectively). The earliest germ cells expressing EYFP were gonocytes, as evidenced by the positive coimmunolocalization of EYFP and zinc finger and BTB domain–containing protein 16 (ZBTB16) in gonocytes from P1 animals [Fig. 2(f), pointed arrowheads] using anti-GFP antibody (42) and mouse monoclonal anti-ZBTB16 antibody (43) at 1:500 dilution (MilliporeSigma). ZBTB16 is expressed in gonocytes, SSCs, and progenitor spermatogonia (44).

Efficiency of Stra8-iCre expression in germ cells

IHC assays were conducted on testicular sections from P8 animals using rabbit polyclonal anti-RARA antibody purchased from Santa Cruz Biotechnology (Dallas, TX) (45) at 1:400 dilution. RARA immunostaining in WT testis cross-sections [Fig. 2(h)] was uniform for both germ cells (arrows) and Sertoli cells (arrowheads), indicating that they both express RARA. In contrast, there was a markedly reduced level of RARA immunostaining in germ cells (arrows) of Rara cKO testis cross-sections [Fig. 2(i)] compared with Sertoli cells (arrowheads).

To further characterize the efficiency of Stra8-iCre expression in germ cells, without any potential contribution of expression from somatic cells, epididymal sperm were collected from adult WT and Rara cKO animals at P75 and DNA was isolated from the sperm heads. Epididymal sperm preparation is virtually free of somatic cells. The WT Rara allele was identified with the TD200/VZ200 primer set (Table 1), generating a 554-bp fragment, and the deleted form of the Rara allele was detected with the 6799/7263 primer set (Table 1), producing a 487-bp fragment. Genotyping of epididymal sperm DNA from Rara cKO animals showed that whereas the WT Rara allele was absent, the DNA fragment that is indicative of the deleted form of the Rara allele was present with 100% efficiency [Fig. 2(j)].

To determine the amount of Rara mRNAs in the enriched population of germ cells, germ cells were isolated from WT and Rara cKO testes. Enriched germ cells from Rara cKO testes had significantly reduced levels of Rara mRNA in animals at P4 (P < 0.001) and P8 (P < 0.0001) vs WT controls [Fig. 2(k)]. To control for the level of somatic cell contamination in the enriched germ cell population, real-time RT-qPCR analyses were conducted to evaluate mRNA levels of the genes Gata4 (a marker for the Sertoli cell) and Hsd3β1 (a marker for the Leydig cell). There were no statistically significant differences in Gata4 or Hsd3β1 mRNA levels between WT and cKO samples (data not shown), indicating similar somatic cell contamination in both samples. These experiments collectively show that Rara is inactivated (exon 4 deleted) in germ cells of Rara cKO mice. Additionally, the expression of Stra8, an RA-responsive gene (46), was determined using real-time RT-qPCR to verify that RARA signaling in the animal model was impaired. Indeed, there were 1.6-fold (P < 0.01) and 4.3-fold (P < 0.0001) declines in the level of Stra8 mRNA in Rara cKO mice compare with the WT controls at P4 and P8, respectively [Fig. 2(k)].

Morphology of seminiferous tubules in germ cell–specific Rara cKO mice was disrupted throughout development

Bouin-fixed testicular sections of WT and Rara cKO animals throughout development were deparaffinized and stained with hematoxylin and eosin (Fig. 3). The WT control seminiferous tubules exhibited the typical, healthy germ cell associations in the seminiferous epithelium showing strictly organized and defined germ cell layers [Fig. 3(a), 3(c), 3(e), 3(g), 3(i), 3(k), 3(m), 3(o), and 3(q)]. In contrast, abnormalities were varied and abundant in the tubules of Rara cKO mice throughout development [Fig. 3(b), 3(d), 3(f), 3(h), 3(j), 3(l), 3(n), 3(p), and 3(r)]. Morphological abnormalities were first evident at P15 [Fig. 3(b)] and almost all testes in Rara cKO animals contained a mix of normal tubules with an assortment of abnormal ones after P15. Germ cell sloughing, virtually blocking the lumen [Fig. 3(f), 3(j), and 3(l)], was the most common phenotype, and other tubule abnormalities included the presence of vacuoles of varying sizes and multinucleated cells as well as missing germ cell layers [Fig. 3(j), 3(n), 3(p), and 3(r)]. The degeneration of tubules was progressively more severe with aging, and by P180 and P365, most tubules were highly vacuolated with layers of germ cells missing [Fig. 3(n), 3(p), 3(r)], and in one animal (out of seven examined) at P365, all tubules were Sertoli cells only. Consistently, the epididymides of WT animals [Fig. 3(s)] contained only spermatozoa, whereas epididymides of Rara cKO mice [Fig. 3(t)] displayed immature germ cells, both round spermatids and pachytene spermatocytes, as shown in the inset.

Open in a separate window

Figure 3.

Histological analyses show increased morphological abnormalities in testicular and epididymal cross-sections of Rara cKO animals. Hematoxylin and eosin–stained testis cross-sections of (a, c, e, g, i, k, m, o, and q) WT and (b, d, f, h, j, l, n, p, and r) Rara cKO animals at various ages from P15 to P365. (a, c, and g) Black arrowheads in WT animals show most advanced cell types at P15, P20, and P30, which are pachytene spermatocytes, round spermatids, and elongated spermatids, respectively. (s and t) Hematoxylin and eosin–stained epididymis cross-sections of (s) WT and (t) Rara cKO mice at P75. Insets in (t) show immature cells at a higher magnification. Seminiferous tubules in Rara cKO animals showed (f, j, and l) sloughing cells blocking the lumen, vacuolization, and (n, p, and r) missing germ cell layers. Scale bars, 50 µm in (a)–(r). Scale bars, 100 µm in (s) and (t).

A mean total of 150 tubules per animal were scored for abnormalities that included germ cell sloughing, vacuolization, multinucleated cells, tubules lacking lumen, and missing germ cell layers. From P25 onward, Rara cKO animals had a significantly increased number of abnormal tubules (P < 0.05 or P < 0.01), an average 40% of the total tubules, vs WT controls [Fig. 4(a)]. Additionally, a mean total of 150 tubules per animal were scored for germ cell sloughing specifically. Relative to WT, Rara cKO testes had a significantly higher level of sloughing germ cells at P45 (P < 0.05) and P75 (P < 0.01) [Fig. 4(b)]. Nearly all abnormal tubules at P75 had the massive cell sloughing phenotype, which consisted of 33% of the seminiferous tubules from Rara cKO mice at P75. Although the mean level of sloughing was higher in P120, the P value was not significant due to a large animal-to-animal variation. By P180, massive germ cell sloughing had subsided.

Open in a separate window

Figure 4.

Quantitation of increased seminiferous tubule abnormalities, delayed progression of germ cell development, decreased sperm count and testis weights, and compromised BTB in Rara cKO testes. (a) Percentage of tubules with abnormalities was present in Rara cKO mice (cross-hatched bar) compared with WT mice (open bar) at P15 and onward, up to P365 (*P < 0.05, **P < 0.01). n = 5 to 10. Error bars are SEM. (b) Tubules containing sloughed germ cells were quantified in mice at P45, P75, P120, and P180, presented as fold change relative to the WT level (*P < 0.05, **P < 0.01). n = 4 to 5. Error bars are SEM. (c) Percentage of tubules containing the most advanced cell types in the testis of Rara cKO mice compared with WT mice at P15, P20, and P30 (***P < 0.001). n = 3. Error bars are SEM. (d) Sperm counts, shown in millions per milliliter, in mice at P75, P120, P180, and P365 (**P < 0.01, ***P < 0.001). n = 5 to 14. Error bars are SEM. (e) Testis weights, measured in milligrams (*P < 0.05, **P < 0.01). n = 9 to 29. Error bars are SEM. (f–h) Tubules were scored for the presence of biotin in the adluminal compartment (red), presented as fold change relative to the WT level (*P < 0.05). n = 3 to 4. Error bars are SEM. Scale bars, 50 µm.

Reduced number of spermatocytes and spermatids in Rara cKO animals

To characterize the relative efficiency of spermatogenesis in the testis, seminiferous tubules containing the most advanced germ cell types at P15, P20, and P30 were counted. Large pachytene spermatocytes [Fig. 3(a), arrowheads], round spermatids [Fig. 3(c), arrowheads], and step 8 to 10 elongating spermatids [Fig. 3(g), arrowheads] were reported to be the most advanced cell types in P15, P20, and P30 animals, respectively (47). A mean total number of 200 tubules per animal were scored. In testes of P15 animals, 39.8% of Rara cKO tubules contained large pachytene spermatocytes compared with 70.8% in WT testes [Fig. 4(c)]. The decline in the tubules with the most advanced germ cell types was persistent at P20 and P30 in Rara cKO animals. There were only 9.6% of tubules containing round spermatids at P20, and 7.1% tubules containing step 8 to 10 elongating spermatids at P30 in Rara cKO animals, whereas in the WT animals, 27.5% consisted of round spermatids at P20 and 14.6% consisted of step 8 to 10 elongating spermatids at P30 [Fig. 4(c)]. At each age observed, germ cell development in Rara cKO animals was significantly impeded compared with the WT animals (P < 0.001).

A reduction in sperm count of germ cell–specific Rara cKO males did not affect fertility

To gain an overall view of spermatogenesis competence in germ cell–specific Rara cKO animals, testis weights were recorded, and epididymal sperm were isolated and counted. There was on average a significant reduction of 43% for the epididymal sperm counts in Rara cKO males at P75, P120, and P365 (P < 0.01), and at P180 (P < 0.001), relative to WT controls [Fig. 4(d)]. These data were reflected by significant decreases in testis weight of Rara cKO animals compared with WT at P30 and P120 (P < 0.05), as well as at P180 and P365 (P < 0.01) [Fig. 4(e)]. However, the 43% reduction in sperm yield was not enough to affect fertility in germ cell–specific Rara cKO males, as there were no differences in the number of pups sired between Rara cKO and WT males at ages P120, P180, and P300 (Table 3). This was not surprising because as little as 10% of control sperm production was enough to restore fertility (48) when other parameters were standard. Previously, an infertility phenotype was observed in global Rara-null mice, where sperm counts dropped to 1.7% of the WT counterpart (19). These findings together highlight the fact that both germ cell RARA and Sertoli cell RARA must be abrogated to have an infertility phenotype.

BTB is perturbed in Rara cKO animals

The disorganization of germ cells in Rara cKO testes suggested possible disruption of the integrity of the BTB. BTB is composed of inter–Sertoli cell tight junctions, ectoplasmic specializations, desmosome-like junctions, and gap junctions between Sertoli cells and germ cells, and together form an impermeable barrier. The BTB separates the tubule into the basal compartment containing spermatogonia and the adluminal compartment containing spermatocytes and spermatids toward the center of the tubule in the lumen (1). A biotin permeability assay was used to assess the integrity of BTB. Testes were collected from WT and germ cell–specific Rara cKO adult mice (P60 to P75) and were injected and incubated with a solution containing biotin. Biotin was prohibited from entering the adluminal compartment of seminiferous tubules when the BTB was intact, as shown in Fig. 4(f). For each animal, the percentage of tubules with biotin in the adluminal compartment was calculated, based on a mean total of 150 tubules per animal scored. In the Rara cKO testes, there was a twofold increase in tubules with biotin leakage to the adluminal compartment compared with WT (P < 0.05) [Fig. 4(g) and 4(h)]. The results suggest that more of the BTB was breached in the Rara cKO testes.

Deletion of Rara in germ cells interferes with the progression of spermatocytes through meiosis

A sizable decrease in the number of pachytene spermatocytes in Rara cKO animals compared with WT animals [Fig. 4(c)] suggests an alteration in the progression of meiosis. Spermatocytes from the P10 to P20 ages include the four stages of meiotic prophase during the first wave of spermatogenesis: leptotene, zygotene, pachytene, and diplotene (49). These stages carry out the stepwise chromosome pairing, synapsis, and genetic recombination that occur during meiotic prophase.

The status of meiotic progression was assessed using two methods. In the first immunological method, testis cross-sections from animals at P11, P15, and P20 were incubated with an anti-γH2AX antibody (36), counterstained with DAPI, and processed for IF detection. Anti-γH2AX antibody (36) immunostains the phosphorylated histone 2AX, localized to DNA double-stranded breaks that occur at the beginning of meiosis. Positive immunostaining for γH2AX in leptotene and zygotene spermatocytes was diffused throughout the cell, whereas in pachytene spermatocytes, immunostaining was limited to the sex chromosomes and is seen as punctate stains (50). At P11, the percentage of tubules with leptotene and zygotene spermatocytes was higher in the WT animals, albeit not statistically significant [Fig. 5(a), 5(b), and 5(e)]. By P15 (P < 0.01) and P20 (P < 0.05), the balance shifted, and the percentage of tubules with leptotene and zygotene spermatocytes was significantly higher in the Rara cKO animals compared with the WT animals [Fig. 5(c)–5(e)]. A mean total of 200 tubules per animal were scored.

Open in a separate window

Figure 5.

Meiotic progression delayed in Rara cKO mice. (a–d) Testicular cross-sections of WT and Rara cKO mice at P11 and P15; γH2AX-positive cells are in red. L/Z indicates tubules containing leptotene and zygotene spermatocytes; P indicates tubules containing pachytene spermatocytes. Scale bars, 100 µm. (e) Quantification of the percentage of tubules with γH2AX-positive leptotene and zygotene spermatocytes for WT and Rara cKO animals at P11, P15, and P20 ages (*P < 0.05, **P < 0.01). n = 3 to 8. Error bars are SEM. (f and g) From chromosomal spreads, the percentages of meiotic prophase cells in leptotene/zygotene and pachytene stages are shown for WT (blue circles) and Rara cKO (orange squares) mice at P15, P17, P19, and P21 ages (*P < 0.05). n = 3 to 5. Error bars are SEM.

In the second method, meiotic stage progression was analyzed from chromosomal spreads of spermatocytes prepared from P15, P17, P19, and P21 animals and immunostained with anti-SYCP3 antibody (35). Homologous chromosomes are paired, or synapsed, along a proteinaceous complex called the synaptonemal complex (SC) in spermatocytes, and the protein SYCP3 is located in the lateral element of SCs (51). Based on the immunostaining pattern of SYCP3 on SCs, cells were classified into one of three meiotic prophase groups: leptotene/zygotene, pachytene, and diplotene stages. The percentage of cells in each meiotic prophase stage was calculated from the total number of 100 spermatocytes per animal that were randomly selected. Rara cKO animals had a significant increase of leptotene and zygotene spermatocytes at P15 (P < 0.05) and had a significantly lower percentage of pachytene spermatocytes at P15 (P < 0.05) compared with WT controls [Fig. 5(f) and 5(g)]. The trend of accumulation of leptotene and zygotene spermatocytes, as well as the decline of pachytene spermatocytes in Rara cKO animals compared with WT controls, continued through P17, P19, and P21. There was a low percentage, ~5%, of spermatocytes in diplotene stages that did not differ between WT and cKO animals at each age examined (data not shown). The results from both immunological staining and chromosomal spread methods attest that the meiotic progression was indeed disrupted in the Rara cKO testes.

Deletion of Rara in germ cells induces SC defects in pachytene spermatocytes

Unexpectedly, examination of WT and Rara cKO chromosomal spreads from P15, P17, P19, and P21 animals revealed increased SC defects in Rara cKO pachytene spermatocytes [Fig. 6(b) and 6(c)]. In pachytene spermatocytes, homologous chromosomes are completely synapsed along the SC. Using chromosomal spreads immunostained with anti-SYCP3 antibody (35) and anti-γH2AX antibody (36), 50 cells per animal were scored for the following SC defects in pachytene spermatocytes: fragmentation, forks, gaps, and partial asynapsis. Fragmentation was defined as segments of SC that measured at least one-third the length of the nearest intact SC, existing separately from the 19 autosomal SCs and the XY body. Forks were defined as asynapsis at the telomere end of the SC, with each separated segment less than one-third the length of the SC. Partial asynapsis was defined as separated segments equal to or greater than one-third the length of the SC. Gaps were defined as an absence of SYCP3 staining in the SC that was a distance of at least twice the width of the SC. The most common chromosomal defect observed was chromosomal fragmentation [Fig. 6(b) and (c), arrows; Fig. 6(f)], which was increased in Rara cKO animals at P15 (P = 0.0673), P17 (P = 0.0532), and P21 (P = 0.0869) compared with WT controls [Fig. 6(d), white bars]. Other defects included the persistence of γH2AX foci on autosomal chromosomes marking unpaired double-stranded breaks [Fig. 6(b) and 6(c), arrowheads; Fig. 6(g)], partial asynapsis, forks, and gaps in the SC of the pachytene spermatocytes. Accordingly, there was a significant decrease in perfect synapsis [Fig. 6(d), gray bars; Fig. 6(e)] in pachytene spermatocytes of Rara cKO animals at P15 and P17 (P < 0.05) compared with WT controls.

Open in a separate window

Figure 6.

Defective synapsis and increased SC fragmentation in pachytene spermatocytes, followed by increased apoptosis in Rara cKO mice. (a–c and e–g) Chromosomal spreads of pachytene spermatocytes from WT and Rara cKO animals at P15. SYCP3 immunostain showing SC (red), γH2AX immunostain showing DNA double-stranded breaks (green), and DAPI DNA stain (blue) were used. White box in (a) outlines perfect synapsis enlarged in (e). White arrows in (b) and (c) indicate chromosomal fragmentation; a representative outlined by a white box in (b) is enlarged in (f). White arrowheads in (b) and (c) indicate autosomal γH2AX staining; a representative outlined by a white box in (c) is enlarged in (g). Scale bars, 10 μm in (a–c); 2.5 µm in (e–g). (d) From chromosomal spreads, the percentage of cells with perfect synapsis (gray), containing fragmented SC segments (white), and containing other defects (forks, gaps, and partial asynapsis) (black) are shown (^‡P < 0.1, *P < 0.05). n = 3 to 5. (h–k) Apoptotic cells visualized by TUNEL assay (green) and DNA visualized by DAPI stain (blue) in seminiferous tubules from WT and Rara cKO mice at P17 and P19. Scale bars, 100µm. (l) Percentage of tubules containing three or more apoptotic cells (^‡P < 0.1, *P < 0.05). n = 5 to 6. Error bars are SEM.

Germ cell apoptosis is enhanced in Rara cKO germ cells

The prevalence of chromosomal defects in Rara cKO spermatocytes suggested that these germ cells could be undergoing apoptosis. Apoptosis is a normal process during spermatogenesis, with the highest numbers of apoptotic cells in mice from P8 to P22 (52). TUNEL assays were performed on testicular sections from Rara cKO and WT mice at P15, P17, and P19 to detect apoptotic cells [Fig. 6(h)–6(k)]. The mean total number of 200 tubules was scored. The percentage of tubules containing three or more TUNEL-positive cells was determined. There was an increase in cells undergoing apoptosis in Rara cKO testes from animals at P17 (P = 0.0722) and at P19 (P < 0.05) compared with WT controls [Fig. 6(l)]. Many of the germ cells undergoing apoptosis were spermatocytes, from their DAPI-stained morphology and the location in the tubules. TUNEL-positive spermatocytes were increased in Rara cKO animals at P17 compared with WT controls (P = 0.0536; data not shown), supporting the notion that SC defects are monitored by the meiotic checkpoints, sending Rara cKO spermatocytes with SC defects to the apoptotic program.

Germ cell RARA crosstalks with Sertoli cells

It was no surprise that spatial disturbance of germ cell arrangements and the chronological disturbance in germ cell development in the Rara mutant testis models could be originating from issues with germ cells crossing the BTB, which involves both inter–Sertoli cell interactions and germ–Sertoli cell interactions via various junctional molecules that make up the BTB (55). Normally, preleptotene spermatocytes transmigrate through the BTB by breaking the inter–Sertoli tight junctions ahead of them and reforming the junctions behind them, without leaving the BTB open at any time, forming a transient compartment [Fig. 8(b)]. The process involves germ cells breaking inter–Sertoli tight junction proteins by transiently interfacing with Sertoli cells via their own germ cell junctional molecules (56). Interestingly, the highest expression of Rara mRNA and RARA in germ cells was in preleptotene and leptotene spermatocytes (10, 20–23). We propose that RARA in preleptotene spermatocytes participates in this transmigration process [Fig. 8(b)]. Moreover, massive sloughing of germ cells also argues for problems with germ cell–Sertoli cell adhesive interactions in Rara cKO testes. Because it has been shown that RARs modulate transcription of genes associated with cell adhesion and cell–extracellular matrix interactions (57), it is possible that ablation of RARA in germ cells could disturb the regulation of some proteins involved in the germ and Sertoli cell interactions, which could then lead to compromised function of the BTB. Indeed, biotin leaked more readily from Rara cKO testes than from WT controls (Fig. 4). Reports of BTB breach in adult testis on a vitamin A–deficient condition as well as a restricted RA condition via metabolic inhibition are supporting our findings (58, 59). Collectively, these results strongly suggest that the germ cell RARA is the molecular player participating in the crosstalk between germ cells and Sertoli cells for regulating transmigration of preleptotene spermatocytes through the BTB [Fig. 8(b), double-end arrows]. Furthermore, it is also likely that the inefficiency in the crosstalk is responsible for the premature release of a massive number of germ cells in Rara cKO testis.

Open in a separate window

Figure 8.

Schematic drawing showing three potential functions of germ cell RARA evident from the characterization of germ cell–specific Rara cKO mouse model. (a) Germ cell RARA has a direct role in meiotic prophase spermatocytes (dark purple), keeping SC integrity. (b) Preleptotene spermatocyte (PL; light purple) expresses RARA at the highest level in the testis. We propose that germ cell RARA in PL regulates crosstalk (double-ended arrows) with Sertoli cells (yellow) for PL to traverse BTB. PL breaks the inter–Sertoli tight junction (TJ1; blue) and enters into the transient compartment where the TJ2 (blue) ahead is still intact. In the transient compartment, germ cells interact with Sertoli cells via junctional molecules (orange ovals), which also mediate Sertoli–germ cell contact in the adluminal compartment. When TJ1 reforms behind PL, PL breaks TJ2 to enter into the adluminal compartment, past the BTB. (c) Germ cell RARA has a function in progenitor spermatogonia (PS, SSC; light green), similar to germ cell RARG, for the spermatogonial differentiation, producing differentiated spermatogonia (DS; dark green) and the initiation of meiosis. White arrows throughout indicate germ cell differentiation steps.

A novel role of RARA in SC integrity and normal progression in meiosis

In addition to being integral in germ cell differentiation, we show, to our knowledge for the first time, that RA via RARA supports meiotic progression by monitoring the SCs. Significantly impeded germ cell development during the first wave of spermatogenesis (Fig. 4) provided a sufficient purpose to examine the meiotic progression in detail between P11 and P20 ages. As suspected, a meiotic progression traffic jam was detected with an accumulation of leptotene/zygotene spermatocytes between P13 to P21, with a concomitant decrease in pachytene spermatocytes in the conditional mutant mice (Fig. 5). These results, obtained through two methods of immunological study of the testis cross-sections and chromosomal spreads, demonstrate that germ cell RARA may be a culpable player in the regulation of meiotic progression.

Moreover, unexpectedly, meiotic germ cells with a deletion of Rara in germ cells revealed prominent phenotypes of fragmented meiotic chromosomes and significant decreases in numbers of cells with perfect synapsis (Fig. 6). The evidence suggests that germ cell RARA is directly regulating the SC integrity in the meiotic prophase [Fig. 8(a)]. Our exciting finding couples well with a recent study showing a decreased RA ligand level associated with increased meiotic defects, including varying degrees of unsynapsed homologous chromosomes and a decrease in cells with perfect synapsis (59). However, to our knowledge, the current study is the first to demonstrate that it is the germ cell RARA that mediates the RA effects on SC integrity. These SC defects seen in pachytene spermatocytes undoubtedly disrupted meiotic progression in Rara cKO testes. Spermatocytes must pass several checkpoints to progress through meiosis properly, and a failure to pass through these checkpoints could cause spermatocytes to undergo apoptosis (60–62). Previously, meiotic prophase germ cells were revealed to undergo apoptosis in Rara-null mice (19). In the current study, Rara cKO spermatocytes experienced a higher rate of apoptosis, indicating that germ cell Rara ablation may contribute to more failures in passing the meiotic checkpoints. Similar results are observed in mice with a knockout of Hsp70-2 gene, which is required for desynapsis of SCs (52). Hsp70-2 knockout mice displayed an increased level of apoptosis at P17 in pachytene spermatocytes.

Germ cell RARA plays a role in the initiation of spermatogenesis from SSCs and progenitors

After busulfan treatment and recovery, there were about fivefold fewer recovered tubules in Rara cKO mice compared with the WT controls (Fig. 7). These results demonstrate that germ cell RARA has a role in the initiation of spermatogenesis from remaining SSCs and progenitor population. Supportive of this, RARA is expressed in the inhibitor of DNA binding 4-eGFP–positive spermatogonial population enriched in SSCs, as well as in the inhibitor of DNA binding 4-eGFP–negative spermatogonial population enriched in progenitor spermatogonia, known to be RA responsive (24). Moreover, the onset of RA signaling is known to start around P3 and P4 (63), around the time of SSC establishment in the germline niche, the proliferation of progenitor spermatogonia, and the beginning phase of spermatogonial differentiation (47). These results together strongly implicate the potential function of RARA in SSCs and progenitor spermatogonia. Because floxed Rara can be deleted as early as in gonocytes, the mutated Rara could also be poised to affect the early gonocyte population in germ cell–specific Rara cKO animals. Future studies are necessary to decipher the exact cellular location of RARA function in neonatal ages.

Intriguingly, the weak recovery of spermatogenesis from the endogenous stem cells in Rara cKO mice after busulfan treatment was still less severe than virtually no recovery observed when Rara-null germ cells were injected into the Rara^+/+ recipient testes in spermatogonial transplantation experiments (19). Contrasting the two methods, it is clear that injected donor SSCs in the spermatogonial transplantation method have an additional step of transmigration through the BTB to establish in the germline niche located along the basal lamina near the blood supply (64, 65). In the busulfan method (39), the endogenous stem cells are already established in the germline niche to reinitiate spermatogenesis, eliminating a potential challenging complication rising from SSC transmigration through the BTB. These results support the notion that germ cell RARA regulates crosstalks with Sertoli cells in modulating the junctional integrity of the BTB [Fig. 8(b)] for the reverse transmigration of SSCs on the way to the germline niche. Future studies are necessary to investigate the junctional proteins, or any of their upstream regulators, by which germ cell RARA could crosstalk with Sertoli cells for regulating the BTB integrity.

We thank Dr. Hui Li for contributing to the production of the Rara^fl construct and William D. Willis for contributing to the production of Rara floxed mice. We also thank Kylie Nelson, Jennifer Onken, Sophie Ascaso, and Ciera Sitton for technical assistance. We are grateful to Dr. Timothy Doyle for critical reading of the manuscript.

Financial Support:This work was supported by National Institutes of Health Grant T32GM008336 (to N.R.P.) from the National Institute of General Medical Sciences, National Institutes of Health Grant T32GM083864 (to S.M.L.) from the National Institute of General Medical Sciences, and by National Institutes of Health Grant HD44569 (to K.K.) from the National Institute of Child Health and Human Development.

Disclosure Summary: The authors have nothing to disclose.