Introduction

Ataxia oculomotor apraxia type 2 (AOA2, also known as SCAR1, MIM# 606002) is a member of the autosomal recessive spinocerebellar ataxias (SCAR, SCA recessive), a class of progressive neurological disorders that result from cerebellar atrophy [1]. First described in 2000, AOA2 was subsequently mapped to chromosome 9q34 and the gene mutated in AOA2, SETX, was found to encode the protein senataxin, a DNA/RNA helicase involved in transcription regulation, RNA processing, and protection of the genome against oxidative damage [2,3,4,5,6,7,8,9,10,11,12,13]. The major clinical features of AOA2 include onset in the second decade, progressive cerebellar ataxia with axonal sensorimotor peripheral neuropathy, diffuse cerebellar atrophy, frequent oculomotor apraxia, early loss of reflexes, loss of Purkinje cells in the cerebellar cortex, and elevated alpha-fetoprotein (AFP) serum level [14, 15]. Disruption of the SETX gene in mice failed to recapitulate the neurological phenotype seen in human patients [16]; however, it uncovered a new aspect of senataxin function in germ cell maturation by showing that Setx−/− male mice were infertile [16]. Senataxin is essential for spermatogenesis since in its absence, germ cells failed to progress past the pachytene stage of meiosis prophase I [16]. Furthermore, inefficient repair of Spo11-mediated DNA double-strand breaks (DSB) was observed in Setx−/− spermatocytes resulting in the accumulation of Rad51 protein on autosomes at the pachytene stage and a failure to complete meiotic recombination and initiate crossovers [16]. A defect in meiotic sex chromosome inactivation (MSCI) was also found in Setx−/− germ cells due to reduced recruitment of the chromatin remodeler CHD4 to the XY chromosomes, preventing the silencing of XY-linked genes at pachytene stage [17]. Finally, senataxin has been shown to attenuate the activity of RNA polymerase II at genes stimulated by viral infection to regulate the magnitude of the host response [16]. Altogether, these animal studies revealed an important role for senataxin at the interface of transcription regulation, gene silencing, chromatin remodeling, RNA processing, and protection of the genome during germ cell maturation.

Several ataxic disorders have been associated with reproductive conditions [18]. Gordon Holmes syndrome, for example, due to mutations in the PNPLA6, RNF216, or STUB1 genes, causes hypogonadotropic hypogonadism through effects on the endocrine system [18]. Women with expansions in the FMR1 gene associated with Fragile X associated tremor ataxia syndrome may experience primary ovarian insufficiency [19]. While there is currently no information on male fertility in AOA2, there are a few reports of problems in females. In two studies, two AOA2 female patients were reported to have entered menopause in their mid-20s providing evidence of premature ovarian failure [15, 20]. In addition, ovarian failure [21, 22], polycystic ovarian syndrome [23], and amenorrhoea secondary to hypogonadotrophic hypogonadism [14] have also been reported in patients with AOA2. These findings suggest a role for senataxin in reproductive function and perhaps, as shown in the Setx−/− mouse model, in germ cell maturation and survival. However, cellular gonadal investigation was not performed in these patients so it is unknown whether there was a primary ovarian defect. Moreover, the contrasting clinical features ranging from high ovarian reserve to suspected ovarian failure leave it unclear as to whether loss of senataxin compromises germ cell integrity.

Here, we report the first clinical and cellular evidence of impaired spermatogenesis and the first cellular evidence of impaired gonadal function in AOA2 patients. Analyses of testicular biopsies from an AOA2 patient recapitulate features of the histology seen in Setx−/− mice. These biopsies revealed the presence of immature germ cells with high levels of DNA damage, an increased number of R-loop-containing germ cells, and elevated germ cell apoptosis. Together, this provides the first clinical and cellular evidence that an infertility phenotype exists in male AOA2 patients similar to that observed in Setx−/− male mice.

Patients and Methods

Patient Recruitment

All patients were initially seen at a tertiary referral ataxia center. For enrollment in this study, patients were required to have positive SETX genetic testing and/or a clinical phenotype consistent with a diagnosis of AOA2. Patients were provided genetic counseling both before and after completion of the study. For semen analysis, men were advised to abstain from ejaculation for 48 h prior to testing, to provide the sample to the lab within 30–45 min of production, and to keep the sample warm. Written informed consent was obtained to collect biological samples for further analysis. All study methods were approved by the Institutional Review Board of the University of California, Los Angeles (UCLA).

Mouse Tissue Samples

All animal work and experiments have been approved by The University of Queensland Animal Ethics Committee (HMRC/UQCCR/155/15/ARC). Testes from adult (35-day-old) mice were collected and fixed in PBS buffered 10% formalin, embedded in paraffin block, and sectioned at 4 μm as previously described [16, 24].

Molecular Analysis

Testicular tissue was collected using standard procedures and immediately processed for paraffin embedding followed by histological sectioning and immunostaining. Briefly, testis biopsies were collected and fixed in PBS-buffered 10% formalin, embedded in paraffin block, and sectioned at 4 μm. Control testes sample was obtained from a tissue bank repository (Aquesta Pathology, QLD, Australia). Sections were stained with hematoxylin and eosin (H&E). Slides were examined under light microscope (Olympus CKX41 and objective CAchN × 10/0.25 PhP), and images were captured by QIMAGING MicroPublisher 3.3 RTV camera and QCapture Pro 6.0 software.

Apoptosis (TUNEL) and Immunostaining

Paraffin sections from control, AOA2 patient, and Setx+/+ and Setx−/− testes were dewaxed and rehydrated with Shandon Varistain Gemini ES (Thermo Scientific, USA). Apoptotic cells were detected by Terminal deoxynucleotidyl transferase UTP Nick End Labelling (TUNEL) assay using the Fluorescence in situ Cell Death Detection Kit (Roche, Switzerland) following the manufacturer’s instructions. TUNEL is a method for detecting DNA fragmentation by labeling the terminal end of nucleic acids and a common method for detecting DNA fragmentation that results from apoptotic signaling cascades. DNA was stained with DAPI (1:10,000; Sigma-Aldrich) for 10 min at room temperature, and slides were mounted in Prolong Gold medium (Life Technologies, USA). Images were captured at room temperature using a digital camera (AxioCam 503 mono, Carl Zeiss Microimaging Inc., Germany) attached to a fluorescent microscope (AXIO Imager.M1, Carl Zeiss Microimaging Inc., Germany) and the Zen 2012 (Blue Edition) software (Carl Zeiss, Microimaging Inc. Germany). The objectives employed were Zeiss APOCHROMAT × 10/0.45 Ph1; Zeiss EC Neofluar × 20/0.5; and Zeiss EC Neofluar × 40/0.75 (Carl Zeiss, Germany). Images were subsequently assembled in Adobe Photoshop 10 (Adobe Systems Inc., USA), and contrast and brightness were adjusted on the whole image panel at the same time. For immunostaining, slides with tissue sections were dewaxed and enzymatic antigen retrieval was performed by incubating the sections with 1:10 Trypsin dilution in PBS for 20 min at 37 °C. Slides were washed three times for 5 min with PBS at room temperature for 5 min each. Tissues sections were blocked in (20% FCS, 2% BSA, 0.2% Triton X-100) for 1 h at room temperature. Slides were incubated with anti-R-loop (1:100, S9.6), anti-γH2AX (1:100, Y-P1016, Millipore), or anti-VASA (DDX4/MVH, 1:100, ab13840, Abcam) antibodies overnight at 4 °C in a humidified chamber. Slides were washed five times with 1× PBS containing 0.5% Triton X-100 for 5 min each at room temperature. Alexa-Dye488 or Alexa-Dye594-conjugated secondary antibody (Molecular Probes, Life Technologies) was added for 1 h at 37 °C in a humidified chamber. Subsequently, slides were washed three times as before and DAPI (1:10,000; Sigma-Aldrich) was added for 10 min to staining nuclei. Slides were finally washed twice, and glass coverslips were mounted in Prolong Gold (Life Technologies) medium for imaging. Images were subsequently assembled in Adobe Photoshop 10 (Adobe Systems, San Jose, CA, USA), and contrast and brightness were adjusted on the whole image panel at the same time. R-loop and apoptosis quantification was performed by examining a random field of view under the microscope and counting all the seminiferous tubules within it. The presence of one or more positive cells within a tubule defined that tubule as positive.

Results

Sperm Abnormalities Are Observed in AOA2 Patients

To investigate possible abnormalities in spermatogenesis and fertility in AOA2 patients, we assessed three subjects (Table 1). The first patient (AOA2-P1), reported in more detail below, demonstrated a complete absence of sperm production. Another adult subject (AOA2-P2) was found to have oligospermia with 100% of his sperm non-motile. A third subject, AOA2-P3, also showed oligospermia with 73% non-motility. In total, sperm abnormalities were seen in all three subjects assessed (3/3, 100%) (Table 1). We assessed hormonal levels in these three individuals as well as an additional two other male AOA2 patients and no clear correlation was observed (data not shown). Overall these findings are consistent with a reproductive defect in a subset of male patients with AOA2. In order to pursue these abnormalities in greater detail, we obtained a testicular biopsy for analysis from patient AOA2-P1.

Table 1 Semen testing in AOA2 patients
Full size table

Patient AOA2-P1 Clinical Presentation

The patient is a 34-year-old man of Eastern European descent with no other past medical history who developed slowly progressive ataxia starting at age 15. Neurological examination was notable for cerebellar ataxia characterized by abnormal eye movements (saccadic smooth pursuit, direction-changing nystagmus, overshoot and ocular dysmetria on tracking, but no oculomotor apraxia), moderately impaired limb coordination, severe gait and truncal ataxia, distal sensorimotor polyneuropathy with amyotrophy, and severe impairment of posterior column function. At the time of evaluation, his Scale for the Assessment and Rating of Ataxia (SARA-5) score was 22/40. Genetic testing demonstrated two novel compound heterozygous variations in the SETX gene: c.2747_2748insAT (p.Met917Leufs*2) and c.6689T>C (p.Met2230Thr). His sister is also affected with a similar clinical presentation and carries the identical mutations. The patient and his wife underwent genetic counseling and decided to proceed with childbearing. Unfortunately, pregnancy did not occur so a fertility workup was initiated during which the patient was noted to have absence of sperm in the ejaculate (azoospermia) which prompted our further evaluation including a testicular biopsy.

Sterility in an AOA2 Patient Due to the Absence of Mature Germ Cells in Seminiferous Tubules

We have previously shown that Setx−/− male mice have severe disruption of seminiferous tubules and lack mature germ cells due to a block at the pachytene stage of meiotic prophase I [16]. To investigate the effect of senataxin deficiency on reproductive function in AOA2, we obtained a testicular biopsy from the infertile AOA2 man described above. Semen analysis performed on this patient revealed azoospermia. A small testicular biopsy was subsequently collected, fixed, and embedded in paraffin to further characterize the histology of the germ cells or lack thereof defect in this patient. H&E staining of testicular sections showed the presence of immature germ cells ranging from spermatogonia to spermatocytes in the seminiferous tubules but an absence of mature sperm in the AOA2 patient compared to a control sample (Fig. 1a). These findings are reminiscent of those obtained in adult Setx−/− mice which show a total absence of mature germ cells in contrast to wildtype littermates, which exhibit all stages of spermatogenic development (Fig. 1b). Vacuolated seminiferous tubules in which both spermatozoa and mature spermatids are absent are observed in both AOA2 and Setx−/−. Taken together, this patient’s infertility is consistent with early maturation arrest. To confirm the presence of germ cells in the seminiferous tubules, sections were immunostained for the VASA (DDX4/MVH) protein (Fig. 2), an RNA-dependent helicase essential for germ cell development that is expressed in fetal and adult gonadal germ cells in both males and females [25, 26]. A normal cytoplasmic staining pattern was observed in both human (control and AOA2 patient) (Fig. 2a) and in mouse seminiferous tubules (Fig. 2b) as previously described [26].

Fig. 1
figure 1

Spermatogenesis is disrupted in an AOA2 patient as shown by the absence of mature germ cells in seminiferous tubules. Hematoxylin and eosin (H&E)–stained sections of testis from control and AOA2 patient samples and adult Setx+/+ and Setx−/− mice. Black arrows indicate mature germ cells in the control and Setx+/+ sections while vacuolated seminiferous tubules in which both spermatozoa and mature spermatids are absent are observed in both AOA2 patient and Setx−/− seminiferous tubules. Scale bar, 20 μm

Full size image
Fig. 2
figure 2

Presence of germs cells in control, AOA2, and Setx mice seminiferous tubules. a Immunostaining of germ cells in human AOA2 and control testis biopsy sections, and b mouse Setx+/+ and Setx−/− seminiferous tubules. Germ cells were stained for VASA (DDX4/MVH), an RNA-binding protein with an RNA-dependent helicase that is essential for germ cell development [25]. Nuclei were counterstained with DAPI. Scale bar, 20 μm

Full size image

Elevated Levels of DNA Damage, R-Loops and Apoptosis in AOA2 Germ Cells Consistent with Infertility Secondary to Maturation Arrest

We have previously shown that the persistence of unrepaired DNA damage in Setx−/− mouse spermatocytes prevented cross-over formation and triggered the mid-pachytene checkpoint resulting in the elimination of DNA-damage-containing cells [16, 24]. To assess whether DNA repair is also compromised in AOA2 patient germ cells, we carried out immunostaining for pSer139 of histone H2AX (γH2AX), a well-characterized maker of DNA double-strand breaks [27]. AOA2 spermatocytes exhibited increased levels of γH2AX staining compared to the spermatocytes from a healthy control patient (Fig. 3) suggesting that defective senataxin function hinders meiotic DSB repair and leads to the accumulation of unrepaired DNA breaks that can threaten genomic integrity in these cells. These findings mirror those of Setx−/− mice (Fig. 3) confirming that senataxin is essential for the effective repair of programmed meiotic DSBs [16].

Fig. 3
figure 3

Elevated levels of DNA damage in AOA2 patient germ cells compared to control. Increased levels of DNA breaks are shown by immunostaining with anti-γH2AX (pSer139 H2AX) antibody were detected in AOA2 patient germ cells and Setx−/− seminiferous tubules as compared to the control and Setx+/+. Both low and high magnification images are shown with scale bars of 100 μm and 20 μm, respectively. Nuclei were stained with DAPI

Full size image

Recent studies have demonstrated that senataxin, through its DNA/RNA helicase activity, resolves DNA:RNA hybrid structures known as R-loops that form during normal DNA transcription when the RNA transcription machinery encounters a GC rich region and stalls, allowing the nascent RNA strand to pair with the DNA template [12]. Both in vitro and in vivo studies have provided compelling evidence for the formation and/or accumulation of these structures in Setx-defective proliferating cells [4, 14, 16, 24]. Indeed, increased levels of R-loops were detected in Setx−/− spermatocytes compared to their wild type counterpart, and this was not only limited to germ cells but was also found in other tissues such as the spleen and intestine of Setx−/− mice, two tissues characterized by a high proportion of proliferating cells [24]. To determine whether this would also be the case in AOA2 male patient germ cells, we looked for the presence of R-loops by immunostaining using the well-characterized R-loop-specific (S9.6) antibody [28, 29]. A greater proportion of tubules containing R-loop-positive cells (57%) were detected in the AOA2 patient testis compared to that of the control (8%) in agreement with findings in the Setx−/− mouse model [16, 24] (Fig. 4a). However, the number of R-loop-positive cells per tubule in the AOA2 patient was less than that observed in Setx−/− mouse testis (Fig. 4b and c). For each positive seminiferous tubule, only one to two R-loop-positive cells were detected in the AOA2 patient, while multiple R-loop positive cells (4 to 40; mean 18 ± 9) were detected in Setx−/− mouse tubules (Fig. 4b and c). This difference in the number of R-loop-positive germ cells per tubule may possibly reflect differences in phenotypic expressivity in patients, disease variance across tissues, a milder phenotype of the patient’s AOA2 mutation(s) as compared to the complete ablation of the Setx gene in the mouse, or an inherent difference between humans and rodents.

Fig. 4
figure 4

Detection of R-loops (DNA:RNA hybrid structures) in AOA2, control, Setx+/+, and Setx−/− testis. a Percentage of R-loop-positive tubules in the control, AOA2, Setx+/+, and Setx−/− samples. Tubules containing ≥ 1 R-loop-positive germ cell were scored as R-loop-positive tubules. b Example of R-loop staining in control and AOA2 patient testis biopsy sections. c R-loop staining in Setx+/+ and Setx−/− mouse testis seminiferous tubules. R-loops were detected in germ cells as shown by the white arrow. While many R-loop-positive germs cells were detected per tubule in Setx−/− mice as shown in (c), only one to two R-loop-positive germ cells were detected per tubule in the AOA2 patient sample (b)

Full size image

The accumulation of DNA breaks, the increased number of R-loop-containing cells, and the absence of mature germ cells in the AOA2 patient testis prompted us to next investigate the level of germ cell apoptosis. As expected, elevated levels of apoptosis were detected in the AOA2 patient (29%) compared to the control (5%) using TUNEL assay (Fig. 5a and b, upper panel) and these levels were comparable to those observed in Setx−/− (34%) and Setx+/+ (6%) mice (Fig. 5b, lower panel). Overall, these data suggest that in humans, as in mice, persistent DNA breaks secondary to defective senataxin function leads to the induction of a stage IV mid-pachytene checkpoint that results in male germ cell depletion and severe infertility.

Fig. 5
figure 5

Elevated levels of germ cell apoptosis in an AOA2 patient. a TUNEL-stained testis sections from the control, AOA2 patient, and adult Setx+/+ and Setx−/− mice. White arrow indicates TUNEL-positive germ cells. DAPI was used to stain nuclei. Scale bar, 20 μm. b Quantitation of the number of apoptotic tubules (tubules containing ≥ 1 TUNEL-positive cell). More than 200 tubules were scored for each sample

Full size image

Discussion

Our earlier work identified a severe male infertility phenotype in transgenic mice lacking the Setx gene and characterized its underlying mechanism. In this report, we describe the observation of a similar phenotype characterized by defective spermatogenesis and lack of mature germ cells in male AOA2 patients, confirming the essential role of senataxin in male reproductive function. The similarities between the mouse model and the AOA2 patient phenotype are striking and strongly suggest that similar molecular mechanisms apply to both. The absence of mature germ cells in the AOA2 patient suggests a blockage in meiosis and perhaps the elimination of aberrant germ cells at the mid-pachytene checkpoint as previously observed in the Setx−/− mouse model [16]. This observation identifies a protein with a significant role in protecting the brain as also playing a role in spermatogenesis. This is consistent with the association of gonadal dysfunction with other genes involved in protecting the genome, such as premature ovarian failure seen with mutation of ATM in ataxia-telangiectasia [30], another ataxia disorder, or NBN in Nijmegen breakage syndrome [31]. Our observation is also reminiscent of how the generation of a mouse model of pantothenate kinase deficiency revealed the presence of azoospermia, which was not previously appreciated in the human disorder [32], emphasizing the importance of such models in identifying associated phenotypes or pathologies in rare neurological disorders such as cerebellar ataxia. Furthermore, this is one of the few examples of male infertility in humans in which the cellular defect underpinning spermatogenic failure has been elucidated. These findings are the first evidence of male infertility in male patients with AOA2 and support essential extra-neurological roles for senataxin in germ cell maturation and survival in humans. As indicated in Table 1, the phenotype may be variable among male AOA2 patients so future studies will be necessary to examine the role of senataxin in normal human spermatogenesis and connections to this and potentially other forms of azoospermia. Interestingly, during this study, we identified one male AOA2 patient who had successfully fathered a child at age 30, suggesting either variability to the phenotype, similar to what has been reported with female fertility issues [15, 20,21,22,23], or a successful pregnancy in the setting of oligospermia. This raises an important reproductive counseling issue for clinicians to address in their male patients with AOA2.

These findings have major implications for the investigation and management of male infertility due to spermatogenic failure. Genetic causes of severe spermatogenic defects have typically been linked with sex chromosome abnormalities such as Y chromosome micro-deletions or sex chromosome aneuploidies such as Klinefelter syndrome (47, XXY males). Our data uncover a novel cause of spermatogenic impairment in humans that can be ascribed to an autosomal mutation. It is possible that other SETX variants that do not incur significant neurological sequelae could affect male infertility and future study will be needed to address this.

Conclusion

In this study, we report the first clinical and cellular evidence of impaired spermatogenesis in AOA2 patients, a finding not previously observed in this patient population, and consistent with the phenotype seen in Setx knockout mice, both clinically and histologically. This supports a novel role for senataxin in human germ cell development and reproductive function and extends its extra-neurological roles. Fertility specialists should be made aware of SETX mutations as a possible diagnosis in young male patients who present with severe oligozoospermia or azoospermia since infertility may potentially presage the later onset of neurological manifestations in some individuals.