Loss of TSLC1 Causes Male Infertility Due to a Defect at the Spermatid Stage of Spermatogenesis

Construction of Tslc1 targeting vector and generation of Tslc1 mutant mice.To generate the Tslc1 targeting vector (pAL1), DNA fragments for the 5′, conditional, and 3′ homology arms were amplified from RPCI-21 PAC “456b9” DNA by PCR using Platinum PCR High Fidelity supermix (Invitrogen, San Diego, CA) for 10 to 18 cycles. The homology arms consisted of a 4-kb 5′ arm (containing a portion of intron 8 flanked by AscI sites), a 1-kb conditional arm (containing exon 9 flanked by HindIII sites), and a 4.5-kb 3′ homology arm (containing exon 10 and the 3′ untranslated region flanked by NotI sites), which were cloned into pGEM-T Easy (Promega, Madison, WI), sequenced, and then subcloned into the AscI, HindIII, and NotI sites of pFlexible (45), respectively. A negatively selectable marker, DT-A (50), was then cloned into the PmeI site of pFlexible. Ten micrograms of AvrII-linearized pAL1 was electroporated into 1 × 10⁷ AB2.2 embryonic stem (ES) cells (from mouse strain 129S5/SvEvBrd) (36) and grown with 3 μg/ml puromycin selection. Cells were cultured on a lethally irradiated SNL76/7 feeder layer (30). ES clones were picked into 96-well plates after 9 days of drug selection and expanded, and targeted clones were identified by Southern blot analysis using PCR-amplified probes on StuI- or EcoRV-digested genomic DNA with a 5′ or 3′ external probe, respectively. pAI3 is the 540-bp 5′ external probe (forward, 5′-GGG GGT TGA AGG AGG CTT AGA AGT GCC CCA CTT TAA ATG A-3′; reverse, 5′-CTT TTT CTC CTT GCA TTC ACT CCT TAA ACG CTT AC-3′), and pAJ10 is the 214-bp 3′ external probe (forward, 5′-ACA AAA AAA CTA CCA TTG TTC ACA GAA TTG GCA TCT CAT T-3′; reverse, 5′-ATT ATT ATT TTT CCA ATG TGT GGG GCC TAC ACT TT-3′). The correctly targeted allele was termed Tslc1^Brdm1 (m1). To generate null and conditional alleles, a cell line with the m1 allele (LW1) was electroporated with 10 μg of a Cre-expressing plasmid (pTURBO-Cre; GenBank accession no. AF334827) or an Flpe-expressing plasmid (pFLPe) (38), respectively. After 3 days of culture, the cells were trypsinized, reseeded at 3 × 10⁵ cells/10-cm-diameter plate, and cultured in 0.2 μM FIAU for 5 days. After this time, the cells were grown for a further 3 days in drug-free medium before the resulting colonies were trypsinized, pooled, and analyzed by Southern blotting using the pAJ10 probe on PacI-digested genomic DNA. The allele generated with Flpe, which showed a 7.5-kb fragment, was termed Tslc1^Brdc1 (c1; conditional allele), and the allele generated with Cre, which showed a 9-kb fragment, was termed Tslc1^Brdm2 (m2; null allele). ES cell clones for the null and conditional alleles were transmitted through the germ line and bred to homozygosity (maintained on a mixed 129/Sv-C57BL/6 background). Mice were housed in accordance with Home Office regulations (United Kingdom).

Genotyping of Tslc1 mutant mice by genomic PCR.Genomic DNA was amplified in 50-μl reaction mixtures using 45 μl Platinum PCR supermix (Invitrogen) and 100 ng of each primer pair (TSL-5/TSL-3 or TSL-C/TSL-3). The primer sequences were as follows: TSL-3 (reverse primer), 5′-TAC CAG GAG GGG AGA AGA GGC CCA GAG C-3′; TSL-5 (forward primer), 5′-AGC ATC CCT TTC CAC CAT AGT TTT CTC TCT-3′; and TSL-C (forward primer), 5′-AGT CTA CAT TTG GGA ATC AAA ATG GTA AGG-3′. The PCR cycle profile was as follows: 1 cycle at 94°C for 2 min followed by 30 cycles at 94°C for 30 s, 65°C for 1 min, and 72°C for 30 s, with a final cycle of 72°C for 10 min.

Reverse transcription-PCR (RT-PCR).RNAs were extracted from +/+, c1/c1, and m2/m2 mouse brains (RNAqueous; Ambion, Austin, TX), and 1.5 μg of total RNA was used to generate cDNAs (RETROscript; Ambion). The cDNAs were amplified in 50-μl reaction mixtures using 45 μl Platinum PCR supermix (Invitrogen) and 100 ng of each primer pair with the following PCR cycle profile: 1 cycle at 94°C for 2 min followed by 30 cycles at 94°C for 30 s, 65°C for 1 min, and 72°C for 30 s, with a final cycle of 72°C for 10 min. All PCR products were cloned into pGEM-T Easy and sequenced to confirm their identity. The specific primers used were as follows: for Tslc1 exons 1 to 3, 5′-GAT ATC CAG AAA GAC ACG GCA GTT GAA GGG G-3′ and 5′-CTT CCC GGG TTA GGC CTT GCA GAG GGT A-3′; for exons 5 to 7, 5′-TCA GAG GTG GAG GAG TGG TCG GAC ATG TA-3′ and 5′-CAT ACA GCA TAT AGT CCG AAT GAG CCT T-3′; for exons 9 and 10, 5′-CTC GAG CAG GTG AAG AGG GGA CCA TTG GG-3′ and 5′-GAT GAA GTA CTC TTT CTT TTC TTC GGA GT-3′; for exons 7 to 10, 5′-AAG GCT CAT TCG GAC TAT ATG CTG TAT GTA T-3′ and 5′-TGC TGC GTC ATC GGC TCC TTT GGC TTC ATG A-3′; and for β-actin gene exons 2 and 3, 5′-ACC AAC TGG GAC GAT ATG GAG AAG A-3′ and 5′-TAC GAC CAG AGG CAT ACA GGG ACA A-3′. For PCR using an oocyte (germinal vesicle stage) cDNA library (a gift from Joanna Maldonado-Saldivia), primers for Tslc1 exons 1 to 3 and Oct4 (5′-GAG GAT CAC CTT GGG GTA CA-3′ and 5′-CTC ATT GTT GTC GGC TTC CT-3′) were used with the PCR cycle profile shown above.

Quantitative real-time PCR.The expression of Cklf1, Xmr, and Ssty was examined in testes and epididymides from 3-month-old +/+, +/m2, and m2/m2 mice, with the β-actin gene used as a control gene (n = 2 for each genotype; the assay was performed in triplicate). Total RNAs were extracted with TRIreagent (Sigma) and cleaned using MinElute columns (QIAGEN, Valencia, CA). RNAs were treated with DNase I before RT-PCR, and 500 ng total RNA was used for each RT-PCR. Real-time RT-PCR was performed in 96-well plates using a QuantiTect SYBR green RT-PCR kit (QIAGEN) according to the manufacturer's protocols, quantifying the resulting fluorescence using an iCycler system (Bio-Rad, United Kingdom). The RT step was done at 50°C for 30 min, followed by incubation at 95°C for 15 min to activate the PCR enzyme. The PCR cycling conditions were 95°C for 15 s, 59°C for 30 s, 72°C for 30 s, and 77°C for 10 s. Real-time fluorescence data were captured during the 77°C step of the cycle (the extra step was necessary to avoid interference from primer dimer signals). Melt curve data were obtained to confirm amplification of the correct product in each well. A crossing-point threshold cycle (C_T) value was obtained for each well as the fractional cycle number at which the measured fluorescence crossed the threshold of 60 units, a value chosen such that all reactions in all plates were in log phase. The average C_T was calculated for each gene in each sample. Data were normalized by reference to β-actin, with ΔC_T calculated as follows: ΔC_T = C_T(test) − C_T(actin). ΔΔC_T values were then calculated as the change in ΔC_T for the m2/m2 genotype relative to the wild type. The specific primers used were as follows: for Cklf1, 5′-TTTX GAA ACA ATG CTC CCA CA-3′ and 5′-ACC TCC AAG GAA CAC ACA GG-3′; for Xmr, 5′-TTC AGA TGA AGA AGA AGA GCA GG-3′ and 5′-TCC ATA TCA AAC TTC TGC TCA CAC-3′; for Ssty, 5′-AGA AGG ATC CAG CTC TCT ATG CT-3′ and 5′-CCA GTT ACC AAT CAA CAC ATC AC-3′); and for β-actin, 5′-ATC ATG TTT GAG ACC TTC AAC AC′ and 5′-TCT GCG CAA GTT AGG TTT TGT-3′.

Microarray analysis.RNAs for microarray analysis were obtained from the testes of 3-month-old +/+ and m2/m2 mice (n = 2 for each genotype). Array hybridizations were performed on the MmcDNAv1 chip, a cDNA microarray representing 11,000 different mouse genes. The clone collection was composed of the NIA 7.4k cDNA set (44), 400 hand-picked clones from the NIA 15k mouse cDNA set (22, 42), and ∼8,500 hand-selected clones from eight testis cell-specific and subtracted mouse cDNA libraries (8, 29) (Pathology Department, Centre for Microarray Resources, Cambridge, United Kingdom). cDNA labeling and hybridization to arrays were performed as previously described (11). Technical reproducibility was assessed at two levels, namely, within each slide (duplicate of each clone) and between slides (four technical replicates of the experiment). Signal intensities were quantified using BlueFuse software (BlueGnome Limited, Cambridge, United Kingdom), and the data from duplicates on the slides were fused. This method averages spots from the same array on the basis of the estimated confidence in each ratio and the level of agreement between duplicates. The resulting data were imported into Knowledge Discovery Environment (InforSense Limited, London, United Kingdom), and the data were normalized to the median of overall intensities for the slide. Data from the four technical replicates allowed us to calculate a coefficient of variation for each clone (the quality of the data are inversely related to the coefficient of variation). Data were sorted into 10 equal-width bins based on absolute fluorescence intensities, and the most variable clones (the 10% with the highest coefficients of variation) in each bin were excluded from further analysis (43). To identify differentially expressed genes, we used an intensity-dependent Z score. This measures the number of standard deviations a particular data point is from the mean relative to other clones expressed at a similar absolute level (35). Clones with ratios of >2 standard deviations from the mean of the window were considered differentially expressed (95% confidence level). Onto-Express (10, 23) was used to classify the differentially expressed genes according to their gene ontology categories.

Histological analysis.For histology, tissues were fixed in 10% neutral buffered formalin and embedded in paraffin. Tissue sections were stained with either hematoxylin and eosin or periodic acid-Schiff stain and examined by light microscopy. For immunohistochemistry, tissues were fixed in 4% paraformaldehyde in phosphate-buffered saline (PBS), embedded in paraffin, sectioned, and stained for Tslc1 using either CC2 (a rabbit polyclonal antibody against the C-terminal 18 amino acids of TSLC1 [24]) or EC2 (a rabbit polyclonal antibody against the extracellular domain of TSLC1 [amino acids 159 to 223] [23]). For cryosections, tissues were fixed in 4% paraformaldehyde in PBS, embedded in optimal-cutting-temperature compound, sectioned, stained for actin using Alexa fluor 568-phalloidin (Molecular Probes) or for cell death using an in situ cell death detection kit (Roche Diagnostics, Indianapolis, IN), and imaged by confocal microscopy. Data from terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling (TUNEL) counts are based on a total count of 290 tubules per testis. For electron microscopy, tissues were fixed and processed in resin as previously described (1), and transmission electron microscopy was carried out on a Phillips 410LS instrument at 80 kV.

Flow cytometry.Testes were excised from 3-month-old +/+ and m2/m2 mice and gently crushed through 100-μm cell strainers (BD Falcon, Bedford, MA) in PBS before being fixed in 100% ethanol overnight at 4°C. The cells (1.5 × 10⁶) were then treated with RNase A (2 μg/ml; Sigma, St. Louis, MO) for 30 min at room temperature and stained with propidium iodide (50 μg/ml; Sigma), and the DNA content was analyzed by flow cytometry. All fluorescence-activated cell sorting data analysis was performed using WinMDI, version 2.8, software.

Fertility studies.To assess male fertility, +/+ (n = 6), +/m2 (n = 15), and m2/m2 (n = 15) male mice at 2 to 4 months of age were mated with either +/+ or +/m2 females of different ages at a ratio of 1:1 or 1:2. To assess female fertility, +/+ (n = 6), +/m2 (n = 18), and m2/m2 (n = 13) female mice at 2 to 6 months of age were mated with either +/+ or +/m2 males of different ages at a ratio of 2:1. In both cases, the breeding pairs were maintained for a minimum of 2 months or until a litter was born.

Analysis of spermatozoa.Epididymides were removed from four or five (from each group) +/+, +/m2, and m2/m2 male mice at 3 months of age. The cauda region from one epididymis was cut open and incubated in a defined volume of Dulbecco modified Eagle medium with 10% fetal calf serum at 37°C for 10 min to allow dispersion of the contents. A homogeneous suspension was then diluted and spread over a hemacytometer, and the number of mature spermatozoa (those with a head and tail) was counted and expressed as spermatozoa (10⁶)/cauda. Motility was defined as the percentage of motile spermatozoa per defined volume.

Measurement of testosterone levels.To assess testosterone levels, whole blood was collected from terminally anesthetized +/+ (n = 9), +/m2 (n = 4), and m2/m2 (n = 10) male mice at 32 to 36 weeks of age, and sera were isolated by centrifugation at 2,000 × g for 20 min. Serum testosterone levels were quantified using a Coat-A-Count free testosterone in vitro diagnostic test kit (Diagnostic Products Corporation, Los Angeles, CA) according to the manufacturer's instructions.

Statistics.For all comparisons, Student's t test (unpaired and two-tailed) was used to generate P values.

Generation of Tslc1 null and conditional mice.To assess the physiological role of TSLC1, we have generated null and conditional alleles of the Tslc1 gene by gene targeting. The targeted locus has loxP sites flanking exon 9 and FRT sites flanking the selection cassette (Fig. 1A). Exon 9 encodes the transmembrane domain. We reasoned that the exon 9-deleted form of Tslc1 would not be membrane bound and would lack a functional cytoplasmic domain (splicing of exon 8 with exon 10 would cause a shift in the reading frame). The cytoplasmic domain of TSLC1 (exon 10), containing key protein interaction domains, has been shown to be responsible for its tumor suppressor activities both in vitro and in vivo (27).

• Open in new tab
• Download powerpoint

FIG. 1.

Generation of null and conditional alleles of Tslc1. (A) Tumor suppressor of lung cancer 1 (Tslc1) is a 10-exon gene. loxP (L) sites were positioned to allow Cre-mediated excision of the transmembrane domain (exon 9) to generate a knockout (null) allele. The loxP-flanked genomic segment is located between nucleotide positions 47,952,177 and 47,953,164 on chromosome 9 of Ensembl (NCBI Mouse Genome Assembly m33). The puroΔtk cassette, flanked by FRT (F) sites, can be excised by using Flp to generate a conditional allele. Southern blot analysis of the Tslc1 targeted (m1), null (m2), and conditional alleles (c1) was performed using two probes, pAI3 and pAJ10, which are 5′ and 3′ of the targeted allele, respectively. Restriction enzyme sites and fragment sizes are indicated. E, EcoRV; P, PacI; S, StuI. (B) Genotyping of Tslc1 null and conditional tail DNAs was performed by PCR using a combination of two primer pairs (either TSL-5/TSL-3 or TSL-C/TSL-3) which can distinguish between all five genotypes (+/+, +/c1, c1/c1, +/m2, and m2/m2) based on the presence and size of the product(s), which can be resolved in a 2% agarose gel. (C) RT-PCR was performed to confirm deletion of the loxP-flanked exons in Tslc1 null mice. RT-PCR performed on RNAs extracted from wild-type mice showed a 245-bp product corresponding to Tslc1 exons 9 and 10. In contrast, no such product was seen for RNAs from m2/m2 mice. RNA from wild-type mice showed 291- and 207-bp products corresponding to exons 7 to 10 and 8 to 10 (reflecting the alternative isoform of murine Tslc1 that differs by the 84 bp that constitute exon 8), whereas RNA from m2/m2 mice produced 159- and 75-bp products, reflecting the deletion of exon 9. β-actin was used as a positive control for cDNA production. (D) Immunohistochemistry performed on testis sections from 3-month-old wild-type and m2/m2 mice, using antibodies against the extracellular domain (EC2) and the C-terminal tail (CC2) of TSLC1.

The conditional targeting vector was introduced into ES cells, and correctly targeted clones with the genotype Tslc1^+/Brdm1 (+/m1) were identified by Southern blotting (Fig. 1A). +/m1 ES cell clones were then exposed to either Flpe (to delete the selection cassette) or Cre (to delete both exon 9 and the selection cassette) to generate ES cell lines with the Tslc1^+/Brdc1 (+/c1) and Tslc1^+/Brdm2 (+/m2) alleles, respectively (Fig. 1A). These ES cell clones were used to generate chimeric mice that transmitted the mutated alleles through the germ line. All offspring were genotyped by PCR using the strategy shown in Fig. 1B, which can distinguish between the five possible genotypes (+/+, +/c1, c1/c1, +/m2, and m2/m2).

Interbreeding of heterozygous (+/c1 or +/m2) mice produced homozygous conditional (c1/c1) and null (m2/m2) offspring at the expected Mendelian ratios. These mice were indistinguishable from littermate controls in terms of growth and development and did not show decreased survival or an increased incidence of spontaneous tumor development compared to wild-type mice (by 1 year of age). To confirm that the mutant allele was null, RT-PCR was performed on RNAs extracted from the brains of wild-type and homozygous (m2/m2) mutant mice. Using primers designed against exons 9 and 10, the expected 245-bp product was produced from RNAs from wild-type and c1/c1 mice but not from m2/m2 mice (Fig. 1C and data not shown). Using primers designed against exons 7 to 10, RNAs from wild-type mice produced the expected 291- and 207-bp products (reflecting the alternative isoform of murine Tslc1 that differs by the 84 bp that constitute exon 8), whereas RNAs from m2/m2 mice produced 159- and 75-bp products (Fig. 1C) which, when sequenced, showed splicing of exon 8 (or exon 7) with exon 10. The skipping of exon 9 in the m2 allele results in a reading frame shift which is predicted to result in a nonsense transcript. Immunohistochemistry using anti-Tslc1 antibodies on m2/m2 testis sections showed the absence of any Tslc1 protein, in contrast to wild-type testis sections (Fig. 1D). Thus, we concluded that the Tslc1^Brdm2 allele is null.

Tslc1 is essential for normal male fertility.Adult c1/c1 mice were viable, healthy, and fertile, and thus the c1 allele was not characterized further. However, the conditional inactivation of Tslc1 using c1/c1 mice will make it possible to evaluate the role of Tslc1 in specific tissues.

In contrast, although adult Tslc1 null (m2/m2) mice appeared normal, no offspring ever resulted from matings between 2- to 3-month-old males and wild-type or Tslc1 heterozygous females (Table 1). Older Tslc1 null males (6 to 12 months of age) also showed no evidence of fertility (data not shown), suggesting that a delay in sexual maturation was not the cause of the earlier infertility. A defect in fertility was not observed in Tslc1 heterozygous males (Table 1), and despite Tslc1 expression being detected in oocytes (see Fig. S1A in the supplemental material), Tslc1 null females showed no overt abnormalities in their ovaries (see Fig. S1B in the supplemental material) and possessed fecundity comparable to that of wild-type mice (data not shown). To further investigate the basis for infertility in Tslc1 null males, spermatozoa were isolated from the caudal regions of Tslc1 mice at 3 months of age. As shown in Table 1, Tslc1 null males showed significantly reduced sperm counts compared with their wild-type and heterozygous littermates, and the majority of Tslc1 null sperm were static. Thus, Tslc1 is essential for normal spermatogenesis and male fertility. Interestingly, adult Tslc1 null mice did not show a significantly altered level of testosterone (6.3 ± 2.5 nmol/liter; n = 10) compared with their wild-type (9.5 ± 4.2 nmol/liter; n = 9) or Tslc1 heterozygous (8.3 ± 5.6 nmol/liter; n = 4) littermates; however, this was most likely due to the fact that testosterone is produced by the Leydig cells, which do not express Tslc1 (47) and do not show any abnormalities in Tslc1 null mice.

Loss of Tslc1 causes a maturation defect at the spermatid stage of spermatogenesis.Spermatogenesis is a highly organized series of events, with germ cells progressing through multiple phases of development. Histological analysis of sectioned testes and epididymides from Tslc1 null mice revealed major abnormalities in spermatogenesis. As shown in Fig. 2A, adult testes from wild-type mice at 3 months of age showed spermatogenic cells in all of the different stages of maturation, from spermatogonia to spermatocyte, spermatid, and, finally, spermatozoa, which then passed into the epididymis (Fig. 2C). In contrast, however, adult Tslc1 null testes showed a nearly complete block in maturation at the spermatid stage of spermatogenesis, with round and elongated spermatids degenerating, dying, and sloughing off into the lumen (Fig. 2B). The few surviving elongated spermatids that underwent the next step of maturation to form spermatozoa did not complete the normal maturation process; the heads of the sperm were fatter than those seen in wild-type mice, had abnormal shapes, and sometimes had no or only rudimentary tails. These degenerating spermatids and immature/malformed spermatozoa that sloughed off into the lumen were subsequently passively propelled into the epididymis (Fig. 2D). We also noted that several Tslc1 null testes from mice at 3 months of age showed whole tubules undergoing degeneration, with large vacuoles present (Fig. 2E), and in 5-month-old mice, many tubules showed a complete loss of architecture (Fig. 2F), in contrast to the case in wild-type littermates (Fig. 2G).

• Open in new tab
• Download powerpoint

FIG. 2.

Analysis of wild-type and Tslc1 null testes and epididymides. Hematoxylin- and eosin-stained sections of (A) wild-type testis, (B) Tslc1 null testis, (C) wild-type epididymis, (D) Tslc1 null epididymis, and (E) Tslc1 null testis from mice at 3 months of age are shown. Hematoxylin- and eosin-stained testis sections from (F) Tslc1 null and (G) wild-type mice at 5 months of age are also shown. Magnification, ×400 for panels A to D, ×100 for panels E to G, and ×400 for the inset of panel E. Abbreviations: LC, Leydig cells; Sg, spermatogonia; Sc, spermatocytes; St, spermatids (round and elongated); Sz, spermatozoa; iSz, immature spermatozoa; dSt, degenerated spermatids. (H) DNA flow cytometry analysis of testicular cell suspensions from wild-type and Tslc1 null (m2/m2) mice at 3 months of age (n = 4 or 5 for each genotype). Arrows highlight the differences between the wild-type and Tslc1 null samples. Abbreviations: HC, elongated spermatids; 1C, round spermatids; 2C, spermatogonia; S-ph, spermatogonia synthesizing DNA; 4C, pachytene spermatocytes and G₂ spermatogonia. (I) Real-time quantitative RT-PCR of three spermatid genes, Cklf1, Xmr, and Ssty, on RNAs extracted from the testes of Tslc1 null mice at 3 months of age. The results were normalized to β-actin and shown as relative changes in expression level compared with the wild type. Asterisks indicate statistical significance in the relative changes in transcript levels between the wild-type and m2/m2 samples (P < 0.05).

To further characterize the stage of the maturation defect in Tslc1 null testes, we determined the frequency distributions of propidium iodide-stained testicular cells from wild-type and Tslc1 null mice at 3 months of age (Fig. 2H). Compared to wild-type mice, Tslc1 null mice showed a lower HC cell peak (elongated spermatids), an increased 4C peak (most pachytene spermatocytes and a few G₁ spermatogonia), and a small increase in the 1C peak (round spermatids). It is possible that failure to complete all stages of maturation may have an effect (via a feedback mechanism) leading to an increase in cell proliferation of spermatogonia (increased 4C) and then to more spermatocytes (increased 4C) and more round spermatids (increased 1C). Taken together, these observations are consistent with the histological findings of a maturation defect at the spermatid stage with an accumulation of spermatids (mostly round spermatids, with some elongated spermatids) and subsequent degeneration and death of the elongated spermatids with sloughing off into the lumen.

To confirm the origin of the sloughed cells, we carried out real-time quantitative RT-PCR on the transcripts for three known spermatid antigens, namely, Cklf1 (11), Xmr (11), and Ssty (8). Both Cklf1 and Xmr were upregulated severalfold in the epididymides of Tscl1 null males; however, Ssty showed a more moderate decrease (Fig. 2I). Since the sloughed cells were positive for Cklf1 and Xmr, this indicates a spermatid origin for these cells. Interestingly, the changes in expression levels of these genes in the testis, though lower in magnitude, are the reverse of the situation in the epididymis, with Cklf1 and Xmr being downregulated and Ssty being upregulated. This is consistent with the sloughing of a particular subset of Cklf1⁺ Xmr⁺ Ssty⁻ spermatids into the lumina of the tubules of the testis, with passive propulsion into the epididymis, potentially reflecting stage-specific effects on spermatogenic cell adhesion in Tscl1 null mice.

Tslc1 null testes show increased numbers of apoptotic cells.Adult Tslc1 null testes from mice at 3 months of age weighed significantly less (0.080 ± 0.003 g; n = 25) than testes from their wild-type (0.112 ± 0.004 g; n = 45; P < 0.005) and Tslc1 heterozygous (0.105 ± 0.004 g; n = 27; P < 0.005) littermates, most likely due to the maturation defect at the spermatid stage of spermatogenesis and the degeneration and death of spermatids with sloughing into the lumina of the seminiferous tubules in these mice. To further examine the nature of the spermatid death, we performed TUNEL analysis (Fig. 3). As shown in Fig. 3A, Tslc1 null testes showed a significantly increased number of TUNEL-positive tubules (tubules containing TUNEL-positive cells; P < 0.01), confirming apoptosis as the major mode of cell death. The number of TUNEL-positive cells per TUNEL-positive tubule was also higher in Tslc1 null testes (Fig. 3B), but this difference was not statistically significant. Interestingly, in contrast with wild-type testes, in which apoptosis occurred in a stage-specific manner and was almost exclusively limited to the spermatogonia (3, 4, 17), Tslc1 null testes also showed apoptosis occurring in the spermatids (Fig. 3C), thus demonstrating that the increased apoptosis was occurring among the maturation-delayed cells in a non-stage-specific manner. The apoptotic death of cells was also confirmed by electron microscopic examination of the Tslc1 null testes (data not shown).

• Open in new tab
• Download powerpoint

FIG.3.

Tslc1 null testes show increased numbers of apoptotic cells. Tslc1 null (m2/m2) mice show a statistically significant increase in the percentage of TUNEL-positive tubules compared with wild-type (+/+) mice (as indicated by the asterisk; P < 0.05) (A) but not in the number of TUNEL-positive cells per TUNEL-positive tubule (B). (C) Detection of apoptosis (TUNEL-positive cells are indicated by arrows) in testis sections from wild-type mice at 3 months of age, with apoptosis occurring in the spermatogonia, in contrast to Tslc1 null mice, which show apoptosis occurring in the spermatids as well as the spermatogonia. Slides for three mice of each genotype were examined, and representative slides are shown. Magnification, ×480.

Delayed maturation in the first wave of spermatogenesis in juvenile Tslc1 null mice.In order to unveil the mechanism of maturation arrest of spermatogenesis in Tslc1 null mice, the progression of the first wave of spermatogenesis was examined in wild-type and Tslc1 null (m2/m2) mice (Fig. 4). At postnatal day 7, the histological appearances of the testes from both wild-type and m2/m2 mice were similar, with seminiferous tubules containing only Sertoli cells and spermatogonia. On day 14, when the premeiosis phase of spermatogenesis begins, germ cell differentiation to spermatocytes (primary and secondary) in m2/m2 mice was similar to that seen in wild-type mice. On day 22, the tubules from wild-type and +/m2 mice that had completed the first and second rounds of meiosis showed the presence of spermatids (mostly round spermatids). However, in m2/m2 mice, progression was delayed, and the most advanced germ cells were primarily still in the late pachytene spermatocyte stage (occasional tubules showed the presence of round spermatids). On day 28, the tubules in most wild-type and +/m2 mice had completed meiosis, and numerous round and elongated spermatids could be seen, as well as spermatozoa. However, while m2/m2 tubules showed differentiation of the spermatocytes into round spermatids and some elongated spermatids, there was evidence of spermatids degenerating and sloughing off into the lumen. On day 35, tubules from wild-type mice showed the presence of numerous fully matured spermatozoa. In contrast, although the m2/m2 tubules did show the presence of some spermatozoa, they were immature and malformed, with fat or deformed heads and no or only rudimentary tails. Interestingly, many testes from mice on days 22, 28, and 35 showed the presence of multinucleate giant cells (MNC), containing multiple nuclei within a single cytoplasm, suggesting that one feature of degeneration could be the opening up of cytoplasmic bridges between sister spermatids.

• Open in new tab
• Download powerpoint

FIG. 4.

Delayed and disrupted first wave of spermatogenesis in Tslc1 null mice. Hematoxylin- and eosin-stained sections of testes from wild-type (+/+) and Tslc1 null (m2/m2) mice at postnatal days 7, 14, 22, 28, and 35 are shown. Magnification for first and second columns, ×200 for day 7 images and ×400 for day 14 to 35 images; magnification for third column, ×1,000. Abbreviations: Sg, spermatogonia; 1° Sp, primary spermatocytes; 2° Sp, secondary spermatocytes; St, spermatids (round and elongated); dSt, degenerated spermatids; Sz, spermatozoa; iSz, immature spermatozoa; MNC, multinucleate giant cells (seen in m2/m2 mice on days 22, 28, and 35; shown at a ×400 magnification).

Tslc1 null testes show some features of differentiation.We next examined Tslc1-deficient spermatogenic cells for evidence of certain features of differentiation. Staining of spermatids for glycoproteins showed the presence of an acrosome cap in both wild-type and Tslc1 null testes in round and elongated spermatids as well as on immature spermatozoon-like head structures in the Tslc1 null testes (Fig. 5A and B); this finding was confirmed by electron microscopy (data not shown). Since TSLC1 has been shown to directly associate with the F-actin binding protein DAL-1 (49), and since cytoskeletal elements at the Sertoli cell-spermatid interface play critical roles during spermatogenesis (24, 34), we wanted to examine whether the loss of Tslc1 would result in abnormal actin bundling. However, as shown in Fig. 5C and D, there was no detectable difference in the polarization of F-actin in spermatids from wild-type or Tscl1 null mice.

• Open in new tab
• Download powerpoint

FIG. 5.

Tslc1 null testes show some features of differentiation. Histochemical analysis of testis sections from (A) wild-type and (B) Tslc1 null (m2/m2) mice by periodic acid and Schiff staining showed the presence of acrosomal caps (stained pink) in the spermatids (St) and spermatozoa (Sz). Slides for three mice of each genotype were examined, and representative images are shown. Magnification, ×400. An analysis of testis sections from (C) wild-type and (D) Tslc1 null (m2/m2) mice by phalloidin staining showed normal F-actin (red) polarization in the spermatids of both. Slides for three mice of each genotype were examined, and representative images are shown. Magnification, ×1,600.

Tslc1 null spermatozoa show major structural abnormalities.Analysis of the contents of the epididymides from wild-type and Tslc1 heterozygous (+/m2) mice showed mature spermatozoa with their distinctive nuclear morphology (head region), very little cytoplasm, and a tail (Fig. 6A and B, respectively). In stark contrast, the epididymal contents of Tslc1 null (m2/m2) mice showed degenerating round spermatids with very few spermatozoa, all of which showed major structural abnormalities, including deformed and irregularly shaped heads, often with a large amount of residual cytoplasm (Fig. 6C to J). Consistent with this, electron microscopy performed on testis sections from wild-type mice showed sperm heads with narrow nuclei and very little surrounding cytoplasm (Fig. 6K). In contrast, sperm heads from m2/m2 mice showed multiple defects, including enlarged nuclei, abnormally shaped nuclei (many of which were degenerated), and large amounts of residual cytoplasm, often with residual body-like changes with dilated mitochondria and vacuoles, suggesting a lack of separation of the residual body from the developing spermatozoa (Fig. 6L to P). MNC with multiple, often abnormally shaped and enlarged, nuclei within a single cellular cytoplasm were also noted (Fig. 6Q and R).

• Open in new tab
• Download powerpoint

FIG. 6.

Analysis of wild-type and Tslc1 null spermatozoa. Phase-contrast light microscopy showed that in contrast to wild-type (A) or Tslc1 heterozygous (B) spermatozoa, Tslc1 null spermatozoa showed morphological abnormalities, as indicated by the arrows (C to J). Images shown are representative of spermatozoa from the caudal epididymides of three to five mice of each genotype. Magnification, ×400. Electron microscopy analysis showed normal sperm head nuclei from wild-type mice (K), in contrast to sperm heads from Tslc1 null mice, which showed abnormally shaped nuclei (L) and enlarged nuclei with abnormal residual cytoplasm, often showing residual body-like changes (M and O), degenerated nuclei (N), dilated mitochondria and vacuoles (P), and multiple malformed nuclei within a single cellular cytoplasm (Q and R). Abbreviations: N, nucleus; C, cytoplasm; RB, residual body-like changes. Sections from two mice of each genotype were examined, and representative images are shown. Magnification, ×4,500 for panels K to N, Q, and R and ×10,000 for panels O and P.

Tslc1 null testes show upregulated expression of genes involved in apoptosis, adhesion, and the cytoskeleton.To understand the molecular consequences of the loss of Tslc1 expression in the testis, we analyzed RNAs from Tslc1 null testes on a germ cell-specific microarray. The data set was normalized and filtered to select differentially expressed genes, with 136 genes being selected as significantly altered in expression (see Materials and Methods). Interestingly, there were many more upregulated genes (103/136) than downregulated genes, which may reflect the accumulation of transcriptionally active round spermatids and the loss of comparatively transcriptionally inactive elongating spermatids and spermatozoa (Fig. 7; see Table S1 in the supplemental material). Differentially expressed genes were then classified according to their likely biological functions, using Onto-Express (10, 23). As shown in Fig. 7, there was an upregulation of a number of proapoptotic genes, again in agreement with the histological findings of increased spermatid apoptosis. There was also an upregulation of several genes with functions related to the cytoskeleton, cell/cell adhesion, and/or maintaining cell junctions. There was no clear functional bias among downregulated genes. When the differentially expressed genes were compared to previous data (11; unpublished data from the same laboratory [E. Clemente, personal communication]) in order to determine their likely sites of expression, it was found that the majority of these genes appeared to be germ cell specific, with only 10 showing expression profiles suggestive of Sertoli cell expression. This may indicate that the majority of downstream effects of Tscl1 loss are germ cell specific or may simply reflect the lower abundance of Sertoli cells in whole testicular tissue. One-third of the upregulated genes were expressed in spermatids, consistent with the histological findings of a maturation defect with an accumulation of spermatids (although this upregulation involved only a subset of spermatid-specific genes), whereas downregulated genes showed no clear bias towards different cell types.

• Open in new tab
• Download powerpoint

FIG. 7.

Microarray analysis of testis RNAs from Tslc1 null mice. Germ cell-specific microarray analysis was used to identify differentially expressed genes in the testes of Tslc1 null mice compared with those of wild-type mice. The data set was normalized and filtered to select differentially expressed genes, with 136 genes being selected as significantly altered in expression in Tslc1 null testes relative to wild-type testes (see Materials and Methods). There were many more upregulated genes (103/136) than downregulated genes, and the differentially expressed genes are shown by classification according to their likely biological function in Onto-Express.