| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Biomedical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, New York 14853
ABSTRACT
Poor sperm quality is the major cause of infertility in humans. Other than sex-linked factors, the genetic basis for male infertility is poorly defined, largely due to practical difficulties in studying the inheritance of this trait in humans. As an alternative, we have conducted forward genetic screens in mice to generate relevant models. We report on the identification and characterization of a chemically-induced mutation, Ste5Jcs1, which causes affected male mice to be sterile or subfertile. Mutant sperm exhibited depressed progressive motility associated with a rigid flagellar midpiece (but not principal piece) segment, which could not be rescued by treatment with agents that stimulate cAMP or calcium signaling pathways. Overall mutant sperm ultrastructure appeared normal, including the axoneme, although the midpiece mitochondrial sheath showed abnormal electron density patterns. Positional cloning of Ste5Jcs1 led to the identification of a mutation in a novel gene called Nsun7, which encodes a protein with a Sun domain that is homologous to tRNA and rRNA cytosine methyltransferases. Therefore, Ste5Jcs1 mutation uncovers a previously unrecognized biological process in sperm that underscores the functional compartmentalization of the midpiece and principal piece of the flagellum.
gamete biology,, gametogenesis,, mitochondria,, mouse,, mutagenesis,, sperm midpiece,, sperm motility and transport,, spermiogenesis,, testis
It is estimated that 10–15% of couples of childbearing age in the United States are infertile, and of these, approximately 40% have male-related infertility [1]. Some studies have suggested that at least half of the cases of idiopathic infertility in men have a genetic basis [2]. The genes underlying human infertility have been difficult to elucidate because of the nature of the phenotype, which confounds genetic analysis. An exception is interstitial deletion of the Y chromosome, which contains several genes required for fertility [3].
Most of the progress in understanding the genetics of human infertility has come from analysis of the mouse. Using gene knockout technology, more than 300 genes causing abnormal spermatogenesis have been identified (according to Mouse Genome Database, MGD). These mutations affect a variety of spermatogenic stages, and many have been identified in studies that were originally intended to evaluate somatic defects [4]. In addition, mouse models have been generated by forward genetic approaches [5–7], although full exploitation of these important resources requires positional cloning of the underlying genes.
Reduced sperm quantity and/or motility are primary causes of infertility in men. Therefore, an understanding of the mechanisms that regulate flagellar movement is crucial. Regulation of flagellar movement includes the activation of motility when sperm leave the epididymis and the hyperactivation of motility in the oviduct. Activation is primarily stimulated by cAMP signaling, whereas hyperactivation is primarily stimulated by calcium (Ca) signaling [8]. Proper flagellar movement is dependent upon the axoneme, which is subjected to different regulatory systems in the midpiece and principal piece of the flagellum. For example, the soluble adenylyl cyclase (sAC), which is required for the activation of motility, is restricted to the midpiece [9], while the plasma membrane calcium channels CatSper1, CatSper2, and CatSper4, which are required for the hyperactivation of motility, are restricted to the principal piece [10, 11]. Little is known about how the smooth waves of axonemal contractions are coordinated along the entire length of the flagellum, across its distinct subregions.
We present the genetic and phenotypic characterization of a male infertility mutant, Ste5Jcs1, which was isolated in a forward genetic mutagenesis screen [12]. Ste5Jcs1/Ste5Jcs1 males are either infertile or subfertile. The sperm from these animals are characterized by poor progressive motility, linked to rigidity of the midpiece (but not the principal piece) and apparent defects of the mitochondrial sheath therein. The Ste5Jcs1 mutation is unique in that it provides a relevant model for the most common cause of human male infertility (poor sperm quality), and provides insight into the functional compartmentalization of the midpiece and principal piece of the sperm flagellum.
All animal experiments were approved by the Institutional Animal Care and Use Committee of Cornell University and were consistent with National Institutes of Health Guidelines.
The Ste5Jcs1 mutation was generated as described previously [12]. For positional cloning of the mutation, C3HeB/FeJ-Ste5Jcs1/Ste5Jcs1 N5 females were outcrossed to males that carried the CAST/Ei chromosome 5 semicongenic (N4) on the same strain background. Non-Rw F1-hybrid sibs were intercrossed, and the resulting F2 offspring were screened for recombination along the Rw region by PCR of microsatellite markers polymorphic between C57BL/6J and CAST/EiJ. Epididymal sperm from recombinant males were extracted at 6–8 wk of age and assessed for the presence or absence of the mutant motility pattern by viewing the sperm in a slide chamber (50-µm depth) under phase-contrast microscopy.
Sperm were analyzed using the media and procedures described previously [13]. The cauda epididymides were removed from adult males and punctured four times at the distal end with a 27-gauge needle in a prewarmed 35-mm polystyrene culture dish that contained 3 ml of medium. Sperm were allowed to swim out for 10 min before 1 ml of the sperm suspension was drawn off from the far side of the dish. The sperm concentration was adjusted to 5 x 106 sperm/ml by the addition of medium.
Ste5Jcs1/Ste5Jcs1 and Ste5Jcs1/Rw males at the N5 backcross generation into strain C3HeB/FeJ were placed with fertile females of 2–6 mo of age. The females were checked for the presence of a copulatory plug over the first 4 days, and either removed to a holding cage or left with the male, then checked again at 3 wk for pregnancy (Supplementary Table 1, available online at www.biolreprod.org).
|
Testes and cauda epididymides from adult male Ste5Jcs1/Ste5Jcs1 mutants and control Ste5Jcs1/Rw littermates were perfused with Bouin fixative, dehydrated, and embedded in paraffin wax. The embedded tissues were sectioned at 4-µm thickness and stained with hematoxylin and eosin.
Sperm motility was assessed by video capture of the swimming pattern, and by computer-assisted semen analysis (CASA) of path velocity (VAP), straight line velocity (VSL), curvilinear velocity (VCL), amplitude of lateral head displacement (ALH), and linearity (LIN). Sperm aliquots were placed in slide chambers on a 37°C stage and the swimming pattern was recorded with stroboscopic illumination at 60 Hz from a 75 W xenon flash tube. CASA analysis of sperm was conducted using a 4x Olympus negative-phase objective. Video images were digitized (30 frames at 60 Hz) and analyzed using HTM-IVOS ver. 10 (Hamilton Thorne Biosciences).
To attempt the rescue of the mutant phenotype, sperm samples from Ste5Jcs1/Ste5Jcs1 and control Ste5Jcs1/Rw males (n = 3 of each genotype) were treated individually with the following chemicals: 60 µM cAMP-AM (at t = 0 and t = 0.5 h); 5 mM procaine (at t = 0 h and t = 1.5 h); 10 mM caffeine (t = 0 h); 25 mM ammonium chloride (t = 0 h); 3 µM ionomycin (t = 0 h); 50 µM thimerosal (t = 0 h); and 10 µM thapsigargin (t = 0 h). Sperm were incubated at 37°C with 5% CO2 for 90 min under capacitating conditions, to induce hyperactivated motility.
The mitochondrial membrane potentials of Ste3Jcs1 mutant sperm were evaluated using the cationic dye JC-1 (5,5',6,6'-tetrachloro-1,1',3,3'-tetraethylbenzimidazolylcarbocyanine iodide; Molecular Probes), which changes from green to orange fluorescence upon aggregation inside active mitochondria. Sperm from Ste5Jcs1/Ste5Jcs1 males and their Ste5Jcs1/Rw littermates were prepared as described above. The JC-1 probe was diluted 15-fold from the stock (1.92 mM in DMSO) in medium, and diluted again 20-fold upon loading into sperm to a final concentration of 6.4 µM. After incubating for 20 min at 37°C, the sperm were centrifuged at 2000 rpm for 30 s in a bench-top centrifuge and the sperm pellet was resuspended in 100 µl of BSA-free medium. Sperm fluorescence was viewed through an Endow GFP Long-pass Emission Filter set (Spectra Services Inc.).
Transmission Electron Microscopy
Sperm were washed by centrifugation at 4000 x g for 2 min in PBS. The pellet was fixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer (pH 7.4), postfixed with 1% OsO4, dehydrated through a graded series of ethanol solutions to 100% ethanol, and embedded in LR white embedding medium. Thin sections (80-nm thickness) were cut and stained with uranyl acetate and lead citrate. Specimens were examined in a Philips 201 Transmission electron microscope at 80 kV.
Total RNA samples were isolated from the testes of 12-, 14-, and 18-day-old wild-type males as well as from sexually mature wild-type and Ste5Jcs1/Ste5Jcs1 males using RNA Miniprep columns (Qiagen). Poly(A) cDNA was generated from 1 µg of RNA using oligo(dT12–18) primer (Invitrogen) and Superscript III (Invitrogen). The 384-bp Nsun7 product that spans exons 5 to 7 was amplified using the following primers: forward, 5'-CGCCCTCTCGATTTACCATA-3', and reverse, 5'-GCAACTGTGTACCACGAACC-3'. The 784-bp Gapdh product was amplified as a control using the following primer pair: forward, 5'-TGAAGGTCGGTGTGAACGGATTTGGC-3', and reverse, 5'-GCCTGCTTCACCACCTTC-3'. Templates were amplified from 2 µl of cDNA using standard PCR reaction conditions.
Experimental data generated during CASA analysis of sperm were analyzed using the Minitab statistical software. Treatment effects were detected using ANOVA, followed by the Tukey test for individual post-hoc comparisons, and differences were considered significant for P < 0.05. The sperm concentrations were compared between Ste5Jcs1/Ste5Jcs1 males and fertile control males using the Welch approximate t-test, and differences were considered significant for P < 0.05.
Isolation of Ste5Jcs1, a Male Infertility Mutant, by N-ethyl-N-nitrosourea (ENU) Mutagenesis
In a previous study, a forward genetic screen was conducted to isolate ENU-induced mutations that mapped to the Rw inversion region of proximal mouse Chr 5 [12]. Ste5Jcs1 was one of two male-specific infertility mutations recovered in the screen. In the original experiments, the male homozygotes were found to be sterile, as assessed by failing to impregnate a wild-type partner after about 1 mo of cohabitation, and sperm from the affected males were reported to show abnormal motility [12].
In the course of testing animals for infertility over longer periods of time, we found that 8/15 (53%) homozygous males could sire litters, compared to 10/10 (100%) Ste5Jcs1/Rw littermates (Supplementary Table 1, available online at www.biolreprod.org). Given time (often several weeks after sexual maturity), some males produced small litters (average of 3.7 pups vs. 5.9 for the controls) after an average mating period of 5.25 wk (vs. 1.6 wk for the controls). We observed no cases of such males impregnating more than one female in a 7-day period. This indicates that a small percentage of the sperm in the ejaculate (although possibly not every ejaculate) is capable of reaching the oviduct and penetrating the outer layers of the oocyte to achieve fertilization. Thus, the mutation results in subfertility, rather than sterility.
The subfertility phenotype was not attributable to decreased sperm counts, as similar concentrations of sperm were extracted from mutant and control males (Table 1). The Ste5Jcs1/Rw group had a median of 11.88 sperm (range, 6.51–31; SD, 7.21; 95% CI, 9.78–18.11). The Ste5Jcs1/Ste5Jcs1 group had a median of 15.9 sperm (range, 2.7–48; SD, 13.3; 95% CI, 9.86–28.89). Histological sectioning through the cauda epididymides confirmed that mutant males had high concentrations of sperm (Fig. 1, A and B), consistent with normal seminiferous tubule morphology (Fig. 1, C and D). However, when examined microscopically, sperm from mutant males displayed an obviously aberrant motility pattern, which was quantified and characterized as described below.
|
Ste5Jcs1 Mutant Sperm Have Decreased Motility and Aberrant Swimming Behavior
To explore further the basis of the subfertility, video tracking and CASA of sperm from Ste5Jcs1/Ste5Jcs1 males were conducted. Mutant males had a lower percentage of motile sperm (Table 1). In addition, mutant sperm displayed slow swimming velocities and failed to display the forward progressive motility typical of activated sperm (49 ± 4.6% vs. 76 ± 4.8% in controls; Table 1). Most notable was that the midpiece section of the flagellum appeared abnormally rigid and did not contribute to flagellar bending (see video in Supplemental material, available online at www.biolreprod.org; Fig. 2, A and B). Ste5Jcs1 sperm swam in a slow circular trajectory or erratically with frequent changes of direction due to the asymmetrical beating of the principal piece and the immobility of the midpiece. Incubation under conditions that resulted in a hyperactivated motility pattern for the wild-type sperm increased the straightness of the swimming pattern but not the midpiece flexibility of Ste5Jcs1 sperm (Table 1).
|
Mutant sperm were treated with various pharmacological agents, to assess whether any of these agents could rescue the phenotype and thereby indicate a possible basis for the motility defect. Incubation with cAMP-AM [14] or caffeine [15], which stimulates the soluble adenylyl cyclase (SACY) signaling pathway to activate flagellar bending, did not ameliorate midpiece rigidity. The nonprogressive motility phenotype was partially rescued when sperm were treated with thimerosal (Table 1), an agent that stimulates an increase in intracellular calcium in sperm [16]; however, the midpiece rigidity remained. Other agents that raise the level of cytoplasmic calcium in sperm, such as thapsigargin [16], ionomycin [17], procaine [18], and caffeine [16], also failed to rescue midpiece rigidity (data not shown). Lastly, NH4Cl, which elevates intracellular pH and calcium [15, 19, 20], did not rescue the motility defect (data not shown).
Ste5Jcs1 Sperm Exhibit Abnormal Electron Densities in the Mitochondrial Sheath
To uncover the ultrastructural defects in mutant sperm, sections through the middle and principal pieces were examined by transmission electron microscopy (TEM). This revealed a normal flagellar axoneme structure (with the typical 9 + 2 microtubule arrangement) and normal outer dense fibers (Fig. 2, C and D; principal piece not shown). However, the mitochondrial region of the midpiece section appeared vacuolated in the mutant compared to the wild-type (Fig. 2D). We also assessed the mitochondrial membrane potential and respiratory activity in Ste5Jcs1 sperm using the JC-1 cationic dye. The mutant sperm stained normally (Fig. 2, E–H).
Although the vacuolated appearance of Ste5Jcs1 mutant mitochondria bears close resemblance to that of the mitochondria of PMCA4-null sperm [21], the results of the single-sperm calcium imaging suggest that these sperm do not experience calcium overload and that they have lower basal levels of calcium than the wild-type (Fig. 2, I and J).
Genetic Mapping and Positional Cloning of Ste5Jcs1
We mapped Ste5Jcs1 to a 2.71-Mb region between microsatellite markers D5Mit290 and D5Mit234. This region contains 24 candidate genes (Table 2), only one of which is expressed predominantly in testis (according to the available microarray and EST expression data), i.e., the RefSeq-annotated gene of unknown function named Nsun7. According to the Stanford SOURCE database [22], about 50% of all existing Nsun7 ESTs are testis-derived (on a normalized basis, compared to all other tissues and developmental time-points combined). Microarray analysis of the mouse testicular transcriptome (presented in the GermOnline database at www.germonline.org, which represents published data [23]) indicates that Nsun7 (formerly 4921525L17Rik) RNA is highest in spermatocytes and haploid spermatids, but is below background in spermatogonia. This is roughly concordant with the results obtained with one Affymetrix probe set [24] indicating that transcription initiates on Day 21 after birth, at which time round spermatids are the most advanced male germ cells.
|
Given the consistency between the timing of Nsun7 expression and the phenotype, we considered this defect to be a strong candidate for Ste5Jcs1 and thus, we sequenced the allele in mutant mice. This revealed a single C to T base substitution within exon 7 at nucleotide 1282 of the transcript (NM 027602, UCSC, NCBI mouse build 36), which results in the conversion of codon 333, encoding glutamine, into a premature (UAA) stop codon (Fig. 3A). This mutation truncates the predicted 700-amino acid NSUN7 protein by more than half (Fig. 3B). Hereinafter, we refer to this allele as Nsun7Ste5Jcs1.
|
NSUN7 contains a subregion homologous to the Sun domain present in tRNA and rRNA cytosine-C5-methylases. Within this domain lies a smaller eukaryotic nucleolar NOL1/NOP2/Sun subfamily motif (Fig. 3B). Eukaryotic genomes contain multiple RNA cytosine methyltransferases, and the seven members of the NOL1/NOP2/Sun family have been annotated in the mouse genome as Nsun2 through Nsun7, plus Nol1. Orthologs (or homologs) of NSUN7 are present in vertebrates (zebrafish) to yeast, with most of the amino acid sequence conservation centered on the Sun domain (Fig. 3B). The location of the mutation within the protein would potentially remove completely the smaller NOL1/NOP2/Sun domain and truncate the larger Sun domain by half.
To confirm the published microarray data (alluded to above) and to assess the transcript levels in mutant testes, RT-PCR analysis was performed on RNA samples obtained from prepubertal and adult mice. Confirming the microarray studies, Nsun7 transcription was low in 12-day-old testes when the majority of advanced cells are in zygonema, but increased by 14–18 days of age when most spermatocytes are in pachynema (Fig. 3C) [25]. The Nsun7Ste5Jcs1 transcript was present at approximately wild-type levels, which indicates that the nonsense mutation does not trigger nonsense-mediated decay (NMD) of the mRNA (Fig. 3C). It is therefore possible that truncated NSUN7 is produced in mutant animals.
The Ste5Jcs1 mutation led to the discovery of a novel gene that was not previously suspected to play a role in spermatogenesis. Moreover, the phenotype is unique in terms of its effects on the sperm midpiece, which in turn affects progressive motility. Although we do not know whether mutations in Nsun7 are responsible for any cases of human infertility or subfertility, the general phenotype of poor sperm quality is a relevant one for idiopathic fertility defects, and it is possible that mutations in other genes that affect the same or related processes occur in humans.
Sperm acquire motility in the epididymis, and this motility is mediated primarily by cAMP signaling. Motility is manifested by proper function of the axoneme, which is comprised of microtubules and ATP-dependent dynein motors. ATP, long thought to be provided by oxidative phosphorylation in the mitochondria, can nevertheless by adequately supplied by glycolytic activity in the flagellum of the mouse sperm [26]. This observation, coupled with our finding that mitochondrial function appears to be normal in Nsun7Ste5Jcs1 sperm, indicates that something other that mitochondrial respiration is responsible for the midpiece-related motility defects and abnormal mitochondrial electron density.
Perhaps the most important aspect of the phenotype is that the flagellar beating defect is compartmentalized; the principal piece, but not the midpiece, of the mutant sperm is able to beat vigorously. Unless the rigidity of the sperm midpiece is purely structural, it would seem that the midpiece region of the axoneme is deficient in components required for normal flagellar function. We have previously speculated that this may be explained by a differential compromise of distinct signaling activities in these two regions of the sperm [27], citing the curious findings that the CatSper1 and CatSper2 ion channels, which are required for hyperactivated motility, are present in the principal piece [10], while the soluble adenyl cyclase (sAC), which is required for activated motility, is restricted to the midpiece [9].
The identification of a mutation that is apparently responsible for Ste5Jcs1 is an important step towards elucidating the actual cause of the sperm defects. However, comparative analyses do not yield obvious clues. Homology studies revealed that orthologs of Nsun7 are in organisms as distant as yeast (S. cerevisiae), which contains a predicted gene (an ORF dubbed YNL022C) that encodes a nuclear-localized protein with 22% similarity. Yeast that contain a deletion of this ORF (generated in a yeast knockout project) are viable, according to the Saccharomyces database (SDB). Protein-protein BLAST searching also revealed the presence of an ortholog in Drosophila (CG13035). According to Flybase, a transposon insertion allele is viable and fertile. This predicted protein is more highly similar to NSUN7 than the closest relative in zebrafish (Fig. 3B).
It is possible that more insight into the function of Nsun7 can be derived from the functions of other NOL1/NOP2/Sun-containing homologs. The Sun part of the domain acronym derives from a bacterial gene with RNA methyltransferase activity. Little is known about the functions of tRNA and rRNA methylation. Interestingly, it has recently been shown that the presumed DNA methyltransferase DNMT2, which is highly conserved throughout evolution but not a member of the NOL1/NOP2/Sun family, actually methylates tRNAAsp, although the mutant mice have no deleterious phenotype [28].
Since NSUN7 is not absolutely required for fertility (some males are partially fertile) and there are no other apparent deleterious effects of the mutation on somatic tissues (despite evidence that there is at least some level of transcription in nontesticular cell types), some degree of functional redundancy may exist both in the testis and somatic cells. Indeed, this gene is one of at least seven NOL1/NOP2/Sun-domain family genes in the genome. By far the most similar paralog is Nsun5 (27% identity over a 358-amino acid stretch), which is weakly expressed in the testis (only 2% of the ESTs are of male genital origin, on a normalized basis). It is possible that NSUN5 and NSUN7 are functionally similar, although their primary roles are in somatic and male germ cells, respectively.
One of the more intriguing potential functions of NSUN7 is in spermatozoal protein translation. The NOP2 in the protein motif acronym refers to a yeast protein Nop2p, the ortholog of which is NOL1 in humans. NOL1 is a nucleolar protein, also referred to as p120, which is associated with various proliferating cancer cells [29]. Studies of S. cerevisiae NOP2 mutations demonstrate a role for the encoded protein in the processing of ribosomal pre-RNAs into the mature subunits [30–32]. Interestingly, there is evidence that rRNA genes are expressed in postmeiotic spermatids of mammals and diverse vertebrates [33].
Despite the conventional dogma that spermatozoa are translationally silent, it has recently been reported that nuclear-encoded proteins are translated by mitochondrial ribosomes, either in or near the mitochondria [34]. If true, it is plausible that NSUN7 is involved in mitochondrial rRNA processing in postmeiotic sperm, and that disruption of the gene causes a general deficiency of mRNA translation and of the proteins required for optimal sperm function. We conjecture that there is a redundant activity (possibly Nsun5) in somatic cells, and to a lesser degree in premeiotic spermatocytes. This would enable progression through meiosis but not postmeiotically, when Nsun7 becomes most highly expressed and provides the primary source of this activity. Nevertheless, residual ribosomes that are produced premeiotically in the mutant persist and are able to synthesize sufficient protein to yield marginally adequate sperm function (hence partial fertility). The abnormal appearance of the midpiece mitochondria in TEM may reflect abnormalities in the ribosomes and/or deficiencies of protein production.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the contributions of Suzanne Hartford, Kerry Schimenti, and Larry Wilson to the initial phenotypic characterization, genetic mapping, and electron microscopy, respectively.
FOOTNOTES
1Supported by a grant (HD 35984) from the National Institute of Child Health and Human Development to J.S., and by a grant (MCB-0421855) from the National Science Foundation to S.S. 
Correspondence: 2FAX: 607 253 3789; e-mail: jcs92{at}cornell.edu
Received: 3 November 2006.
First decision: 18 December 2006.
Accepted: 11 April 2007.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
