Centre for Cellular and Molecular Biology, Uppal Road, Hyderabad, 500 007, India (e-mail: email@example.com).
ABSTRACT: Genetic factors cause about 10% of male infertility. Azoospermia factors (AZFa, AZFb, AZFc) are considered to be the most important for spermatogenesis. We therefore made an attempt to evaluate the genetic cause of azoospermia, Y chromosome deletion in particular, in Indian men. We have analyzed a total of 570 men, including 340 azoospermic men and 230 normal control subjects. DNA samples were initially screened with 30 sequence-tagged site (STS) markers representing AZF regions (AZFa, AZFb, AZFc). Samples, with deletion in the above regions were mapped by STS walking. Further, the deletions were confirmed by Southern hybridization using the probes from both euchromatic and heterochromatic regions. Of the total 340 azoospermic men analyzed, 29 individuals (8.5%) showed Y chromosome deletion, of which deletion in AZFc region was the most common (82.8%) followed by AZFb (55.2%) and AZFa (24.1%). Microdeletions were observed in AZFa, whereas macrodeletions were observed in AZFb and AZFc regions. Deletion of heterochromatic and azoospermic regions was detected in 20.7% of the azoospermic men. In 7 azoospermic men, deletion was found in more than 8.0 Mb spanning AZFb and AZFc regions. Sequence analysis at the break points on the Y chromosome revealed the presence of L1, ERV, and other retroviral repeat elements. We also identified a ∼240-kb region consisting of 125 bp tandem repeats predominantly comprised of ERV elements in the AZFb region. Histological study of the testicular tissue of the azoospermic men, who showed Y chromosome deletion, revealed complete absence of germ cells and presence of only Sertoli cells.
About 15% of couples are infertile because of several reasons, of which the male factor is responsible for approximately 50% of the cases. Various causative factors have been identified for male infertility, including varicocele, obstruction of spermatic ducts, agglutination of sperms, impotency, hormonal imbalance, and genetic defects. Based on the semen profile, male infertility is classified into 4 major categories, namely oligozoospermia (low sperm count), asthenozoospermia (low motility), teratozoospermia (abnormal shape and size), and azoospermia (complete absence of sperm). Genetic factors have been found to play a role in about 10% of male infertility (Pryor et al, 1997).
Large numbers of sequence-tagged sites (STS) markers have been mapped to AZFa, AZFb, and AZFc regions. Incidence of Y chromosome deletion is about 1 in 1000, and these deletions are often associated with male infertility. Deletion of the growth control gene on the Y chromosome (GCY) interval in the AZFa region was found to be associated with short stature (Kirsch et al, 2000). However, some of the Y chromosome deletions are neutral, which do not affect the fertility status and phenotype of the individual (Thangaraj et al, 2002b). Several studies have demonstrated that microdeletion in AZF regions causes male infertility. Deletion of each AZF region has been found to have a different phenotypic effect. Sargent et al (1999) studied the AZFa region of 4 infertile men and found a deletion of AZFa. DFFRY and DBY were associated with Sertoli cell—only syndrome type 1 (SCOS) phenotype, whereas the patient, who retained DBY, seemed a mild oligozoospermia. However, genotype-phenotype correlation has not been fully understood. Deletions in the AZFb region have been found to be associated with azoospermia, oligozoospermia, and normozoospermia. Deletion of the AZFc region has been found to be associated with azoospermia and severe to mild oligozoospermia (Affara, 2001). It has been found in many cases that similar deletions of AZF regions cause quantitative loss in spermatogenesis. Among the 3 AZF regions, deletion of AZFc has been found to be the most frequent abnormality, followed by AZFb (Layman, 2002).
Most of the STS-based studies on male infertility have been carried out with a few markers for each AZF region (Martinez et al, 2000). Hence they fail to detect the Y chromosome deletion in many cases. Therefore, there is no collective opinion about the marker to be used for Y chromosome microdeletion analysis. In this study, for the first time, we have made an attempt to evaluate the association of Y chromosome microdeletion among Indian men with azoospermia using several STS makers from each AZF region. This study will help generate epidemiological data, which in turn, will be useful for infertility clinics for adopting an appropriate strategy to provide genetic counseling to the affected couples.
Materials and Methods
Azoospermic and Control Men
Three hundred forty azoospermic men, selected from 870 azoospermic men, attending the infertility clinic at the Institute of Reproductive Medicine (IRM), Kolkata, India, were included in the present study. The age groups of these azoospermic men ranged from 26 to 42 years. An experienced urologist at IRM carried out a detailed case history and clinical examination of every patient. Endocrinological assays (follicle stimulating hormone [FSH], luteinizing hormone [LH], testosterone, prolactin [PRL], thyroid stimulating hormone [TSH], HIV, and venereal disease research laboratory [VDRL] test); vasogram; and karyotyping were performed for each patient by respective experts. Testicular pathology was carried out wherever it was possible. Patients who showed any abnormality such as obstruction; endocrinological defect; obvious karyotypic abnormalities (eg, Klinefelter syndrome, mosaicism, translocations, marker chromosome, etc); pelvic injuries; and major illness were excluded in the present study. Two hundred thirty random fertile Indian men were included in this study as control. Blood sample (5.0 mL) from each azoospermic and fertile control man were collected with informed written consent. DNA of the above samples was isolated using the protocol described by Thangaraj et al (2002a).
Polymerase Chain Reaction Assay
Polymerase chain reaction (PCR) based studies for microdeletion on both azoospermic and control men were carried out using STS markers on the long arm of the Y chromosome. Initially the screening for AZF regions was done using 30 STS markers. The AZFa region was analyzed with sY83, sY746, sY740, sY86, sY2320, sY741, sY84, DFFRY, sY742, sY615, and sY743. The AZFb region was analyzed with sY98, sY100, sY110, sY80, sY113, sY118, sY124, sY127, sY1211, sY134, and sY143. The AZFc region was analyzed with sY152, sY148, sY156, sY255, sY581, sY254, sY247, and sY158. To detect the heterochromatic region, sY160 was used in all cases. Individuals who had deletions in any of these STS markers were further analyzed by using the flanking STS markers (Ma et al, 1993; Reijo et al, 1995) in order to map the deletion breakpoints. SRY was analyzed in all infertile men (Thangaraj et al, 2002b).
Primers were synthesized using ABI392 Oligosynthesizer (Perkin Elmer, Foster City, Calif). Each marker was amplified separately in a 0.2 mL thin wall tube using an MJ Research Thermal Cycler (Waltham, Mass) with a female negative control sample. PCR conditions used for STS markers were as follows: initial denaturation (94°C for 5 minutes) and subsequent denaturations (94°C for 45 seconds) was the same for all the samples. Extension was 65°C for 1 minute for sY127, sY134, and sY255, whereas for the other STS markers extension was 72°C for 2 minutes. Different annealing temperatures that were used for different STS markers were as follows: 60°C for 15 seconds for sY83, sY254, and sY581; 60°C for 1 minutes for sY740, sY746, sY741, DFFRY, sY742, sY615, and sY743; 55°C for 30 seconds for sY86; 53°C for 30 seconds for sY84 and sY158; 57.4°C for 1 minute for sY2320, sY113, sY100, sY98, and sY121; 58°C for 30 seconds for sY127, sY134, and sY255; and 55°C for 1 minute for sY110, sY143, sY124, sY156, sY148, sY247, sY118, and sY152. PCR products were analyzed by 2% agarose gel electrophoresis.
Southern hybridization was carried out wherever a sufficient amount of DNA of the azoospermic men, who showed deletion of AZF regions, was available. Two Y chromosome—specific probes, such as 50f2 and 102(d)2, were used for Southern hybridization. The probe 50f2 detects 5 different loci in the euchromatic region of the Y chromosome (Jobling, 1994), whereas 102(d)2 is a Y chromosome heterochromatin-specific (DYZ) probe (Singh et al, 1994). Five and 2.0 μg of DNA was digested with EcoRI separately to hybridize with 50f2 and 102(d)2, respectively. After restriction digestion, DNA samples were size fractionated in 1.0% agarose gel and transferred onto a Hybond N+ Nylon membrane (Amersham Pharmacia, Buckinghamshire, United Kingdom) by capillary transfer using 0.4N NaOH. Southern hybridization was carried out with α32P-labeled (BARC, Jonaki, India) probe. Hybridization and posthybridization washing conditions were as given in the above references.
To evaluate the nature of the DNA sequence present at the Y chromosome breakpoints of the azoospermic men, we have analyzed the human Y chromosome sequence, downloaded from GeneBank (available at http:ncbi.nlm.nih.govgenbankgenomesH_sapienschr_Y). The STS markers, which showed deletion in the ends of the flanking STS markers, were mapped on the Y chromosome with the help of the STS database and electronic PCR from the National Resource for Molecular Biology Information (NCBI). In the proximal region of the Yq, the deleted STS marker and its flanking STS markers (deletion interval 5C) were mapped to the contig NT_011875. Sequences that were present in AZFb and AZFc were identified on the contigs NT_011875 and NT_011903. DNA sequences of these contigs were subjected to various sequence analysis tools such as the tandem repeat finder (Benson, 1999), BLAST (Altschul et al, 1997), and alignment tools. Since the sequence information for the Yq heterochromatic region was not available in the database, we were unable to analyze the sequences from the distal heterochromatic region. The sequence and the accession numbers of the corresponding break point regions were noted, and the repeats flanking the particular break point were marked by blasting with the repeat masker program. Also, annotations that were given for each accession identification number were manually checked for the presence of repeats flanking these breakpoints. Distance between the break points was also obtained by calculating the differences between each marker position on the sequence of both the contigs and then analyzed. Approximate size of the deletion of each individual was estimated using the sequence information from the database.
Screening of 340 azoospermic men with the STS markers specific to AZF regions showed deletion in 29 individuals (Figures 1 and 2), which accounts for 8.5% of the total azoospermic men analyzed. Of these, deletion of the AZFa region alone was detected in 5 individuals (A145, A146, A162, A221, and A250), which accounted for 17.2% of the total deletion (Figure 3). One azoospermic man (A118) showed deletion in the AZFa, AZFb, and AZFc regions (3.5%). In this azoospermic man, deletion in the AZFa region was interstitial, whereas deletions of AZFb and AZFc were continuous (Figure 2). Another azoospermic man (A224) showed deletion in both the AZFa and AZFc regions (Figure 2), but deletion observed in this individual was microdeletion in both the regions (Figure 2). Interstitial deletion in the AZFa region was observed in another azoospermic man (A146); interestingly, this individual showed large deletion in the AZFa region (Figure 2). None of the azoospermic men showed deletion, which is unique to the AZFb region. Deletion of AZFc alone was detected in 7 individuals (A113, A115, A123, A133, A263, A326, and A328), which accounted for 24.1% of the total deletion (Figure 3). Deletion of the AZFb and AZFc region occurred for 51.7% of the total deletions observed (Figure 3). Eight individuals (A16, A25, A56, A66, A142, A151, A234, A305) showed the deletion of the AZFb and AZFc regions (Figure 2). Of these, deletions in 6 individuals were continuous for these regions. However, in 2 individuals (A142 and A151), deletions in both the AZFb and AZFc regions were independent. A total of 8 individuals showed the deletion of the heterochromatic region (DYZ), of which 7 azoospermic men (A18, A58, A85, A254, A267, A294, A302) showed the deletion of the AZFb and AZFc regions, and 1 individual (A123) showed the deletion of AZFc in addition to the heterochromatic region (Figure 2). None of the control men showed deletion for the 30 STS markers, which were used for the primary screening for the deletion of AZFa, AZFb, and AZFc regions.
Our STS analyses have shown that in the majority of the cases, the deletion in both the AZFb and AZFc regions were very large (macrodeletion), spanning the AZFb, AZFc, and heterochromatic regions. We therefore carried out Southern hybridization to confirm the macrodeletions. Probe 50f2 detects 3 loci (A, B, and D) on the p arm and 2 loci (C and E) on the q arm of the Y chromosome. We analyzed a total of 12 azoospermia samples representing different deletion events. Of the 12 samples, 3 (A113, A263, A326) were with the deletion of the AZFc region. The locus 50f2/C corresponds to the AZFc region, and the absence of the 50f2/C signal in all 3 individuals confirms the microdeletion in the AZFc region (Figure 4). Three azoospermic men (A16, A66, A118) who showed deletion of the AZFb and AZFc regions by STS analysis were also confirmed by Southern hybridization. Four azoospermic men (A18, A85, A254, A302) who were detected as having the deletion of the AZFb, AZFc, and heterochromatic regions by STS analysis showed the absence of a hybridization signal from 50f2/C, 50f2/E, and 102(d)2 loci (Figure 4), confirming that these 4 azoospermic men had microdeletions involving the AZFb, AZFc, and heterochromatic regions. Two azoospermic men (A145, A162) who showed deletion of only the AZFa region were also included in the Southern analysis and revealed no deletion in the AZFb, AZFc, and heterochromatic regions (Figure 4). Approximate size of the deletions in the AZF regions, based on STS and Southern analysis, are given in Figure 5.
All the deletion breakpoints on the human Y chromosome were mapped onto 2 contigs, namely NT_011875.7 and NT_011903.8. The accession numbers for the sequences, which showed the deletions, are given in the Table. It was interesting to note that the deletions that occurred in the AZFa region were mostly microdeletions compared with that of the deletions in the AZFb and AFZc regions. In the AZFa region, most the breakpoints were flanked by human endogenous retroviral (HERV) and L1 (LINE) type of repeat elements. Also, the breakpoints of the AZFb and AZFc regions predominantly contained L1 and retroviral sequences. It was interesting to note that there was ∼240-kb region containing 125 bp tandem repeats that was identified between sY124 and sY127. These repeats spanned the sequences of GenBank accession numbers AC079157, AC079261, AC079156, AC024250, and AC009240. These tandem repeats were identified as ERVs. This region was deleted in about 50% of the azoospermic individuals (Figure 2). The markers, which failed to amplify the distal end of the Y euchromatic region, were found to be flanked by the clusters of L1-type repeat elements. However, this region did not contain 125bp tandem repeats as found above.
Comparison of the sequences of the deleted regions revealed that there were significant levels of homology, which ranged between 81% to 93% in the AZFa region; however, the size of these homologous sequences are about 1000 bp only. Similarly, the AZFb regions also showed high levels of sequence homology between the repeat sequences. Analysis of AC053490 and AC006338, which are present in the AZFc region, revealed that these regions had segmental duplication. A maximum of about a 40-kb sequence was duplicated with 99% homology. Approximate size of deletion in the euchromatic region of the Y chromosome of azoospermic men is shown in Figure 5. Of the 29 azoospermic men who showed deletion, 7 had a total deletion of more than 8.0 Mb, and 5 individuals had a deletion of ∼4.5 Mb (Figure 5).
Testicular volumes of the infertile men who showed Y chromosome deletion ranged between 1.0 mL and 12.0 mL. However, hormone profiles of these individuals showed normal ranges, except 2 azoospermic men; one of them (A18) had high LH (27.5 mIU/mL), and the other one (A250) had low testosterone (1.5 ng/mL). Testicular biopsy of these individuals showed the presence of only Sertoli cells. There was no correlation between testicular size and quantity of sperm production, as all the infertile men analyzed were azoospermic.
We, for the first time, made an attempt to study the genetic causes of male infertility in Indian populations. PCR-based STS analysis of 340 azoospermic men revealed micro- and macrodeletions on the Y chromosome in 29 individuals (Figures 1 and 2), accounting for 8.5% of the total azoospermic men analyzed. Earlier studies revealed that the Y chromosome microdeletions were responsible for 7% to 13% of the infertile men (Reijo et al, 1995; Pryor et al, 1997; Martinez et al, 2000), leading to azoospermia, severe oligozoospermia, and oligozoospermia. Size of the deletion on the Y chromosome, in our study, did not show any significant correlation with the amount of sperm production, as all the samples analyzed were azoospermic.
In our study, approximately one fourth (24.2%) of the total Y chromosome deletion was in the AZFa region (Figure 3). The size of the deletion in the AZFa region was very narrow (microdeletion) compared with the AZFb and AZFc regions. One of the azoospermic men (A162) showed deletion of just 1 STS marker (DFFRY), but the majority of them (6 individuals) showed the deletion of sY746. Sun et al (1999) reported that point mutation in the genes present on the AZFa region can cause azoospermia. This is in agreement with the present observation that although the deletion of the AZFa region was very small, all the individuals with a deletion in the AZFa region were azoospermic. A similar type of abnormality was also detected by Kamp et al (2001). It was interesting that none of the azoospermic men studied showed the deletion of either sY84, sY86, or both. It is indeed surprising that these 2 markers were recommended for the diagnosis of the AZFa region (Simoni et al, 1999). Using only sY84 and sY86, we would have diagnosed as normal all 7 individuals who possess deletion in the AZFa region. Martinez et al (2000) have analyzed 128 infertile men with sY84, sY85, and sY86 and found none of them have shown deletion. Deletions would have been noted in those cases if more markers were used. Dohle et al (2002) also did not see any deletion in the AZFa region during their screening of 37 azoospermic and 113 severely oligozoospermic individuals with 2 STS markers for each AZF region. It may be possible that the deletion of sY84 and sY86 was predominantly seen in 1 population but not in the other populations. For instance, we found deletion of sY746 in 6 out of 7 individuals who showed deletion in the AZFa region, which may not be true for other populations. Considering the fact that some deletions are more predominant in certain populations (Thangaraj et al, 2002b), one has to use a large number of STS markers from each AZF region to detect Y chromosome deletions. In general, frequency of deletion in the AZFa region is less compared with the AZFb and AZFc regions (Fujisawa et al, 2001; Layman, 2002; Peterlin et al, 2002). In most of the studies, deletion of the complete AZFa was found to be associated with SCOS (Simoni et al, 1999; Fujisawa et al, 2001; Kamp et al, 2001; Dohle et al, 2002; Peterlin et al, 2002; Sun et al, 2000). Our study demonstrates that even deletion of a narrow region on the AZFa region can cause SCOS. Therefore, we suggest that sequencing of the entire AZFa region would help in identifying point mutations, implicating azoospermia/male infertility.
Very recently, Ferlin et al (2003) have reported the deletion of 5 single copies and 2 duplicated genes, but they found intact RBMY1 in 4 infertile men. Our study did not show the deletion of the AZFb region in any of the azoospermic men analyzed. The reason that none of the infertile individuals shows deletion of only the AZFb region might be because of the nature of DNA sequence present in the distal end of the AZFb region. Altogether, deletion of the AZFb region was involved in 55.2% of the total deletions, of which 51.7% were in association with AZFc and 3.5% with AZFa and AZFc (Figure 3).
Deletion of the AZFc region was detected in 24 individuals (Figure 2), accounting for 82.8% of the total deletion of azoospermic men (Figure 3). This is in agreement with the earlier studies showing that the incidence of deletion in the AZFc region was high compared with the AZFa and AZFb regions (Martinez et al, 2000; Peterlin et al, 2002). Reduction in the copy number of DAZ genes (AZFc) was reported to have increased risk in subfertility and infertility (de Vries et al, 2002). Of the total 29 deletions, AZFc alone (in addition to DYZ) was observed in 7 individuals (24.1%). It was interesting to note that a total of 8 Y chromosomes (27.6%) showed the deletion of the heterochromatic region (DYZ), which was diagnosed as having a short Y chromosome. Although the Yq heterochromatic region predominantly consists of DYZ1 and DYZ2 classes of repeat sequences (Manz et al, 1992), a few studies have shown its association with infertility and reproductive failure. Quantitative variation in heterochromatin has been found to be more in couples with fetal wastage (Buertic-Tomljanovic et al, 1997). Deletion of the entire heterochromatin and part of euchromatin was detected in an infertile man (Hartung et al, 1988). In our recent study, we observed the deletion of the entire heterochromatin, along with very little euchromatic region and duplication of the Yq euchromatic region of the Y chromosome in an individual having ambiguous genitalia and undescended testes (Thangaraj et al, 2003). We could not obtain either semen profile or assess the fertility status of the individual, as he was only 5 years old when we made the examination. But in the present study, we did not see any case showing the deletion of only the heterochromatic region of the Y chromosome.
Southern analysis with 50f2 confirmed the macrodeletion in both AZFb (50f2/E) and AZFc (50f2/C) regions (Figure 4). Deletion of only the 50f2/C locus in infertile men was earlier reported by Ma et al (1993). We are reporting for the first time the deletion of the 50f2/E region, which corresponds to the AZFb region and suggests that the probe 50f2 can be used to confirm the macrodeletions in both the AZFb and AZFc regions.
Sequence analysis revealed the presence of repeat elements (HERV, L1, and other retroviral sequences) at the breakpoints. Presence of a homologous repeat sequences block in this region (deletion intervals D3 and D6) suggests that the deletion in the above cases might be because of intrachromosomal recombination of homologous repetitional blocks (Sun et al, 2000; Kamp et al, 2000). As the deletion in the AZFa region was very small (microdeletions), it is possible that only certain repeat sequences, which are highly homologous to themselves in this region, undergo intrachromosomal recombination leading to deletion. Based on the fact that the repeat sequences are prone to deletion, the probable reason for the frequent deletion in the AZFc region might be due to the crossing over between the repetitive sequences in the AZFc region. Kuroda-Kawaguchi et al (2001) have reported that the AZFc region contains massive palindromes, which leads to recurrent deletion in this region. Repping et al (2002) also reported that the palindromes P5 and P1 were responsible for large deletions, up to 7.7 Mb, involving AZFb and AZFc regions of the infertile men. It can be speculated that most of the deletions involving AZFb and AZFc are mediated through intrachromosomal recombination between the ERVs and L1 repetitive elements present in both AZFb and distal to the AZFc regions. In fact, a previous study showed the presence of HERV at the heterochromatic region (Kjellman et al, 1995). Our present study does not show a single case of Y deletion involving only the AZFb region. Similar observation has been reported in other populations as well (Brandell et al, 1998), but Repping et al (2002) detected the deletion of only AZFb in 3 infertile men. If every Y chromosome consists of similar sequences, then why are the deletions occurring on Y chromosomes unique to certain individuals? This may be explained on the assumption that the retroviral or L1 repeats are involved in the deletion process, as discussed above, and their repeat blocks may be differently organized in different populations. It is possible that the 125-bp tandem repeats present in the AZFb region predominantly contains ERVs could contribute more to the intrachromosomal crossing over, since this region spans about 240 Kb.
Histological analysis of some of the azoospermic men with Y chromosome deletion showed SCOS. Kamp et al (2001) reported that the deletion of the AZFa region was associated with SCOS. Sargent et al (1999) reported that the deletion of DFFRY and DBY in the AZFa region was associated with SCOS. However, Fujisawa et al (2001) reported that SCOS was associated with deletion of all 3 azoospermic factors (AZFa, AZFb, and AZFc). Our observation of SCOS in all AZF-deleted individuals is in agreement with the study of Fujisawa et al (2001).
Male infertility is caused by various genetic and non-genetic factors. We have earlier studied the CAG repeat length polymorphism in the androgen receptor gene of infertile and fertile men. Although several studies found that the expanded CAG repeat was associated with male infertility in certain populations (Yong et al, 1998; Dowsing et al, 1999), we have not observed repeat length differences between infertile and fertile groups (Thangaraj et al, 2002b). The present study showed that the Y chromosome deletion was responsible for 8.5% among azoospermic men, of which deletion of the AZFc region was found to be of a very high frequency. We found that the deletions were up to 11.0 Mb involving the AZFb and AZFc regions. HERVs and ERVs were found to be responsible for the deletions. For the first time we have identified 125-bp tandem repeats in the AZFb region. However, etiologies of a large number of azoospermic men are still unknown. Analyzing the remaining azoospermic men with additional Y chromosome STS, × chromosome, and autosomal markers would help in identifying the etiology of the remaining azoospermic individuals (Brandell et al, 1998; Wang et al, 2001). In light of the above, we believe that the etiology of male infertility may differ between ethnic populations. Therefore, researchers need to keep this in mind and define the strategies for analyzing infertile samples. Our data will be useful for infertility clinics for genetic counseling by advising them to choose a female child in case of Y chromosome deletion and to adopt appropriate methods for assisted reproduction.
Supported by a grant from the Indian Council of Medical Research (ICMR), Government of India, and the Council of Scientific and Industrial Research (CSIR), Government of India.