Assessment of the frequency of sperm annulus defects in a large cohort of patients presenting asthenozoospermia

The annulus is a ring-shaped structure located beneath the plasma membrane that connects the midpiece and the principal piece of mature sperm flagellum. It is formed in the round spermatids at very early stages of flagellum assembly, concomitant with nucleus condensation and the development of the acrosome (Fig. 1). It was long thought, based on the spatiotemporal distribution of this structure during sperm flagellum elongation, that it might behave as a morphological organizer, guiding the growth of the flagellum and the alignment of the mitochondria along the axoneme [1]. In addition, its position between the two major compartments of the flagellum in mature spermatozoa suggested that it might act as a diffusion barrier confining proteins to particular compartments of the sperm tail [2, 3]. Strong evidence to support these two hypotheses has been provided by the recent identification of protein components of the annulus and the characterization of mice in which the corresponding genes have been knocked out [4–8]. However, the precise molecular mechanisms underlying the biogenesis and function of the annulus remain unclear.

Only a few proteins have been shown to localize to the annulus. The first to be identified were proteins of the Septin family (Septins 1, 4, 6, 7 and 12), small G proteins forming homo- and heterofilaments in various cell types and compartments [4, 5, 7]. Septin 4 invalidation in the mouse resulted in male infertility due to total asthenozoospermia, and morphological defects of the flagellum, including an absence of the annulus, a bending of the flagellum and an abnormal arrangement of the mitochondria in the midpiece. Annulus defects were also observed in Septin 12 mutant mice [4, 5, 7, 9]. DNAJB13, a member of the HSP40 co-chaperone family, and SLC26A8 (TAT1; testis anion transporter 1), a sperm-specific anion transporter, were subsequently also localized to the annulus [6, 10]. DNAJB13 is located in the radial spokes of the (9 + 2) axoneme in mouse sperm and in the Chlamydomonas reinhardtii flagellum [10–13]. It has been reported to interact physically with Septin 4 at the annulus during spermiogenesis. SLC26A8 is a sperm-specific member of the SLC26 (solute-linked carrier 26) family of anion exchangers that has been shown to be restricted to the sperm annulus and head equatorial segment in humans and mice [6, 14]. Interestingly, Slc26a8 inactivation in the mouse results in a near-phenocopy of the Septin 4-null mouse, suggesting that this protein is also essential for annulus integrity, sperm flagellum morphology and motility [6]. In Slc26a8 mice, electron microscopy studies have shown the sperm annulus to be present but markedly atrophic and detached from the plasma membrane. This transmembrane protein is thought to play a structural role, anchoring the cytoplasmic components of the annulus to the plasma membrane, in addition to its regulatory function in mediating the anion fluxes required to induce sperm motility and capacitation [15]. A few enzymes, such as the soluble adenylate cyclase (sAC) activated during sperm motility, and the vitamin D-metabolizing enzyme CYP24A1 have been found to localize at the annulus [16, 17]. The reasons for this restriction to the annulus, as opposed to the location along the length of the principal piece observed for other enzymes required for sperm motility and function, remain unclear.

Despite these recent findings in mice, the importance of the annulus for mammalian sperm flagellum structure and motility has not been clearly established. Several studies have attempted to resolve these questions by analysing the correlation between annulus defects and the occurrence of human asthenozoospermia. A first study on a cohort of 20 Japanese patients with asthenozoospermia indicated that 15 % of these patients had no annulus, leading the authors to claim that an absence of the annulus could be used as a diagnostic marker for a subset of patients with asthenozoospermia [5]. A larger cohort of 108 Japanese men with normal and low sperm motility was then analysed and a similar frequency (13 %) was reported for the absence of the annulus [18]. We studied a cohort of 75 asthenozoospermic patients but observed an absence of the annulus in only one of these patients, resulting in a much lower frequency (1.2 %) [19]. We assumed that this discrepancy between our results and those of previous studies reflected ethnic differences or the small sizes of the cohorts studied to date. In this study, we analysed a new, larger cohort of asthenozoospermic patients with precisely characterized phenotypes, to determine the frequency of annulus defects in asthenozoospermia more accurately. We screened sperm preparations from 254 additional asthenozoospermic patients for the presence of the annulus, by an immunodetection method, with specific antibodies directed against the SLC26A8 protein, a previously described component of the annulus that can be used as a marker of this structure.

Patient I1, aged 40, was consulting for medically assisted procreation because his wife was infected with HIV and presented bilateral tubal obstruction. Patient I1 had no specific history of genital disease, but he displayed bilateral varicocele and obesity. Semen analysis showed a moderately low progressive motility (PR 20 %) and percentage of typical forms, following no particular pattern. Selection by centrifugation through a density gradient increased PR to 70 %, with 40 % rapid spermatozoa, and increased the percentage of typical forms to 19 %. A pregnancy was obtained by in vitro fertilization and the couple had a healthy child.

Patient I2 was 38 years old and had no previous medical or genital health history. He and his partner had been trying to have children for six years and had suffered a miscarriage two years previously. Semen analysis showed PR to be almost normal (PR 30 %). Only 5 % of the spermatozoa were considered to be rapid, but semen viscosity was high and, after selection, 70 % of the selected spermatozoa were rapidly motile. The percentage of typical spermatozoa was very low, but most of the abnormalities observed concerned the acrosomal region of the sperm head.

Patient I3 was 35 years old. His partner had had two natural pregnancies in the two years preceding semen analysis at our center: the first was an ectopic pregnancy, and the second ended in miscarriage. No other information about fertility or medical history was available. The couple was followed elsewhere, but we were informed that they achieved a natural pregnancy, with a healthy child born the following year. The semen parameters were close to normal values (PR 30 %), except for sperm morphology, but most of the morphological abnormalities observed concerned the sperm head and followed no particular pattern.

An analysis of sperm from individual I1 with antibody SE5362 showed that the SLC26A8 signal was spread out along the flagellum, rather than being located in the annulus as in control sperm (Fig. 3a, b), whereas detection in the equatorial segment was normal. This result was confirmed by staining with the L2CL4 antibody, which showed that SLC26A8 was absent from the annulus (Fig. 3c, d). A signal for SEPTIN 4 (SEPT4) was observed at the annulus in immunodetection experiments, suggesting that the annulus was incomplete rather than absent (Fig. 3e, f). Sperm from individuals I2 and I3 displayed no SLC26A8 signal at the annulus or the equatorial segment (data not shown). It was not possible to perform immunodetection with SEPTIN 4 (SEPT4) for these individuals. The identification of an abnormal pattern at the annulus in immunodetection studies for these individuals suggested that the annulus might be modified or missing from their spermatozoa. Unfortunately, too small volume of semen remained for electron microscopy and we were therefore unable to carry out an ultrastructure analysis.

We had previously screened this cohort of patients by genomic amplification and sequencing of sperm DNA. We found mutations of SLC26A8 (MIM 608480) responsible for SLC26A8 protein instability and proteasomal degradation in three different patients from those we present here [22]. We therefore investigated whether the observed absence of the SLC26A8 protein at the sperm annulus was due to mutations of SLC26A8. An analysis of the coding regions of SLC26A8 in individuals I1 and I2 identified only the most frequently described sequence polymorphisms (I2: p.M73V and p.V639I). We therefore concluded that the absence of the SLC26A8 protein at the annulus was probably due to a defective structure of the annulus, preventing the correct localization of the protein, consistent with the staining for SLC26A8 observed along the flagellum.

In total we found that only three of the 254 asthenozoospermic individuals in the cohort studied here (1.18 %) had a potentially incomplete or absent annulus. In a previous study of 75 asthenozoospermic patients, only one individual with a complete lack of the sperm annulus was identified by the immunodetection of SLC26A8 and SEPTIN, with confirmation by electron microscopy [19]. If we consider all the individuals included in these two studies together, 329 patients, then the overall frequency of annulus defects in our population of asthenozoospermic patients can be estimated at 1.21 %.

In this study, we investigated the integrity of the annulus on sperm preparations from 254 patients presenting proven asthenozoospermia, according to the current WHO criteria (progressive motility < 32 %). By combining the data obtained in this study with those for 75 asthenozoospermic patients from a previous study [19] we were able to analyse 329 individuals in total. This cohort is the largest cohort of individuals studied to date in investigations of the association between annulus abnormalities and the occurrence of asthenozoospermia. Our data indicate that defects or an absence of the annulus are rarely detected, with an estimated frequency of only 1.2 % in human patients with asthenozoospermia and a progressive motility of more than 10 %. Our findings differ from the results initially published for the analysis of smaller cohorts of Japanese asthenozoospermic individuals, which indicated a high frequency of annulus defects in this population (13 to 15 %). The difference between our results and those of these previous studies may reflect differences in ethnic origin, as the patients of the cohort studied here were mostly Caucasian men; in addition, SLC26A8 defects may constitute only a subset of the annulus defects present, as SEPTIN proteins form the core of the annulus. However, our data and published findings do not support these two hypotheses. Hence Hosseinifar and colleagues recently analyzed sperm from a cohort of 100 asthenozoospermic Iranian men. They carried out immunodetection for the SEPTIN 4 and SEPTIN 7 proteins on sperm samples from their patients and found only one individual lacking the annulus, as confirmed by electron microscopy. Their conclusion for the Iranian population was therefore similar to that reported here: a very low frequency of annulus defects in the population studied [23]. In addition, our unpublished data for analyses of Slc26a8 and Septin4 knockout mice (provided by H. Kissel & H. Steller) indicate that the lack of Septin 4 prevents the localization of Slc26a8 to the annulus. Thus any annulus defect due to the absence of Septin 4 should have been identified by staining with the SLC26A8 antibody. Based on all these findings, we conclude that the integrity of the annulus is not a relevant diagnostic endpoint for screening and classifying asthenozoospermic patients.

Overall, caution is currently required when drawing conclusions about the possible function and causality/pathogenicity to be attributed to the annulus. The molecular mechanisms by which this intriguing structure is assembled and functions remain unclear, in both humans and mice; the possible role of annulus abnormalities in human asthenozoospermia also remains to be determined. As several lines of evidence suggest that the annulus is involved in guiding flagellum assembly (in particular, in midpiece arrangement during sperm terminal differentiation), annulus defects may be be the cause of asthenozoospermia due to abnormal flagellum biogenesis in humans, a phenotype, which would be however distinct from that of Septin4 and Slc26a8 invalidation in mice. This phenotype, previously described as dysplasia of the fibrous sheath [24–27] and, more recently, as multiple morphological abnormalities of the flagella (MMAF) [28], consists of a lack of flagellum assembly and the presence of short, bent and coiled flagella of irregular thickness. Further studies of the annulus in sperm from MMAF patients are required to address this aspect.