Y-chromosome AZFc structural architecture and relationship to male fertility

Transcription

1 Y-chromosome AZFc structural architecture and relationship to male fertility Celia Ravel, M.D., Ph.D., a,b,c Sandra Chantot-Bastaraud, M.D., Ph.D., a,b,d Brahim El Houate, Ph.D., e Hassan Rouba, Ph.D., e Marie Legendre, M.D., Ph.D., d Diana Lorencxo, Ph.D., a,b Jacqueline Mandelbaum, M.D., Ph.D., b,c Jean Pierre Siffroi, M.D., Ph.D., b,d and Ken McElreavey, Ph.D. a a Human Developmental Genetics, Institut Pasteur, Paris; b UPMC University of Paris, Paris; c Service d Histologie Biologie de la Reproduction et Cytogenetique, AP-HP, H^opital Tenon, Paris; d Service de Genetique Medicale, AP-HP, H^opital Trousseau, Paris, France; and e Human Genetics Unit, Institut Pasteur of Morocco, Casablanca, Morocco Objective: To determine if there is a relationship between various forms of partial AZFc deletions and spermatogenic failure. Design: Case-control study. Setting: Infertility clinic (Tenon Hospital, Paris). Patient(s): 557 men, comprising 364 infertile men from mixed ethnic backgrounds, and 193 men with known fertility (n ¼ 84) and/or normospermic (n ¼ 109). Intervention(s): Characterization of 32 partial AZFc deletions. Main Outcome Measure(s): DAZ gene cluster divided into two families (DAZ1/2 and DAZ3/4), CDY1 gene, and Y-chromosome haplogroups. Result(s): We observed 18 partial AZFc deletions in 364 (4.95%) infertile men compared with 14 out of 193 (7.25%) in the control normospermic/fertile group. Conclusion(s): The analysis of informative Y-chromosome single nucleotide variants combined with Y-chromosome haplogroup definition enabled us to infer seven deletion classes that occur on a minimum of six Y-chromosome parental architectures. We found no relationship between either the presence or the absence of DAZ1/2, DAZ3/4, CDY1a, orcdy1b with spermatogenic failure at least on one Y-chromosome lineage. The DAZ dosage and Southern blot analyses indicated that the majority of individuals tested carried two copies of the DAZ gene, indicating a partial AZFc deletion. Our data are consistent with the hypothesis that, at least in our study populations, partial AZFc deletions may have a limited impact on fertility. (Fertil Steril Ò 2009;92: Ó2009 by American Society for Reproductive Medicine.) Key Words: Y chromosome, male infertility, AZFc region, partial deletion Approximately 2% of human males are infertile because they produce few or no sperm (1). In many of these cases the cause is unknown. The male-specific region of the human Y chromosome (MSY) is inherited in a uniparental manner and contains 78 genes and gene families that play critical roles in male germ cell development (2). Around 10% of men with severe spermatogenic failure lack portions of the MSY (3). Much of the MSY consists of long, Y-specific repeats termed amplicons, and homologous recombination between amplicons generates deletions that cause spermatogenic failure. Indeed, many amplicons themselves contain genes that have testes-restricted expression patterns and may have critical testis functions (2). The precise contribution of each Received June 25, 2008; revised August 26, 2008; accepted August 29, 2008; published online November 6, C.R. has nothing to disclose. S.C-B. has nothing to disclose. B.E. has nothing to disclose. H.R. has nothing to disclose. M.L. has nothing to disclose. D.L. has nothing to disclose. J.M. has nothing to disclose. J.P.S. has nothing to disclose. K.M. is a recipient of a grant from the Association pour la Recherche contre le Cancer (ARC). Reprint requests: Ken McElreavey, Ph.D., Human Developmental Genetics, Institut Pasteur, 25 rue du Dr Roux, Paris Cedex 15, France (FAX: ; kenmce@pasteur.fr). Y-chromosome gene or gene family in achieving male fertility is still the subject of debate (4). Various types of deletions of the Y chromosome have been described, and some of these appear to be specifically associated with spermatogenic failure whereas others may be inconsequential polymorphic variants. Distinguishing pathogenic deletions from those that are not remains a key objective in our understanding of the genetic basis of male infertility. The great majority of Y-chromosome microdeletions fall within the long arm of the Y chromosome, comprising AZoospermia Factor (AZF), three regions associated with severe spermatogenic failure: AZFa, AZFb, and AZFc. The AZFa deletions are located in proximal Yq, and are they are caused by recombination that takes place between retroviral homologous sequences. These deletions account for less than 1% of all microdeletions of the Y chromosome reported to be associated with spermatogenic failure. Clinically, AZFa deletions are associated with complete absence of germ cells in the testes (5). Complete deletions of AZFb have a size of 6.23 Mb and extend within 1.5 Mb of the proximal portion of AZFc. The AZFbþc deletions are the result of recombination between palindromic sequences. These deletions, 1924 Fertility and Sterility â Vol. 92, No. 6, December /09/$36.00 Copyright ª2009 American Society for Reproductive Medicine, Published by Elsevier Inc. doi: /j.fertnstert

2 although large, do not include the distal part of AZFc (6). Clinically, complete AZFb deletions are associated with meiotic arrest or Sertoli cell only syndrome (7). Recurrent deletions of AZFc are relatively common, affecting around 1:4000 men (8 10). The AZFc region contains 21 genes, including four copies of the Deleted in AZoospermia gene (DAZ) that has been implicated in the regulation of messenger RNA (mrna) translation. The four DAZ copies are organized in two distinct clusters: DAZ1/2 and DAZ3/4 (11, 12). There are also two copies of the CDY1 gene, termed CDY1a and CDY1b. Absence of the entire AZFc region is caused by homologous recombination between repeat elements termed b2/b4. Small deletions within the AZFc region can occur by intrachromosomal recombination, either within a single chromatid or between sister chromatids (13, 14). Several different types of partial AZFc regions have been described in humans. Although some of these are considered to be polymorphisms with no known consequences for fertility, one partial AZFc deletion, termed gr/gr, has been shown to be a risk factor for infertility in some studies; but in other studies, it appears to be a polymorphism that is unrelated to the phenotype (13, 15 28). Of the 15 studies that have been published investigating the association of gr/gr deletions with male infertility or testicular cancer, eight studies have found an association with the phenotype, and seven studies have failed to find such a link. Understanding the contribution of these deletions to infertility is important, not only for diagnostic purposes but also to determine if there is one key fertility gene in the AZFc region or if the AZFc phenotype is due to the cumulative effect of the absence of several genes, each contributing to spermatogenesis (29). This is becoming of increased importance because the gr/gr deletion has also been suggested to be a risk factor for the development of testicular cancer (22). Several of the problems associated with the characterization of partial AZFc deletions include the lack of informative markers in the region that can distinguish between the gene families and no simple system for quantifying gene copy number. However, recently several new markers within the AZFc region have been described that have the capacity to distinguish between DAZ clusters and the CDY1a/CDY1b gene (16). In addition, a simple dosage assay based on polymerase chain reaction (PCR) has been described that enables the DAZ copy number to be estimated, and a worldwide survey of AZFc architecture has been reported (30, 31). Using these approaches, we investigated the relationship between spermatogenic failure, AZFc deletion classes, and the Y-chromosome haplogroup. Our data suggest no relationship between spermatogenic failure and partial AZFc deletions, at least on the E3b2 background. MATERIALS AND METHODS Study Population Our preliminary analysis of the gr/gr deletion interval in 192 men who presented with idiopathic infertility and a control population of 181 men was previously reported elsewhere (24). We extended that study to obtain a population composed of infertile men (n ¼ 364) from mixed ethnic backgrounds who sought infertility treatment at the Tenon Hospital in Paris. They presented with a complete absence of spermatozoa in the ejaculate (azoospermia, n ¼ 115) or reduced sperm counts (< /ml, severe oligozoospermia, n ¼ 72; < /ml, moderate oligozoospermia, n ¼ 159; < / ml, mild oligozoospermia, n ¼ 18). Informed consent was obtained for karyotype and molecular investigations, and approval was given by the local ethics committee in France. Individuals with known causes of infertility such as karyotype anomalies, absence of vas deferens, chemotherapy or radiotherapy, trauma, or cryptorchidism were excluded from the study. Each of these men had been previously screened for Y-chromosome microdeletions according to the European Academy of Andrology (EAA) guidelines (32), and men harboring Y AZF microdeletions were excluded from further studies. The control population consisted of 193 men who were of known fertility (n ¼ 84) and/or normospermic (n ¼ 109). None of these men carried an AZFa, AZFb, or AZFc deletion. The control population was of mixed ethnic origin. Detection and Characterization of Partial AZFc Deletions Partial AZFc deletions were detected using the markers sy1191 and sy1291, as previously described elsewhere (24). When a partial AZFc deletion was identified, the Y-chromosome haplogroup was defined using the binary markers TAT, M9, 92R7, SRY 1532, YAP, 12f2, M172, LLY22g, P2F, SRY 4064, M174, and M81 (33). The two DAZ gene clusters (DAZ1/2 and DAZ3/4) were differentiated by PCR amplification using the primer pairs 912/913, followed by digestion of the PCR product using the enzyme DraI (16). The CDY1a and CDY1b genes were differentiated by amplification of DNA using the primer pairs 1025 and 1026, followed by digestion with the restriction enzyme PvuII (16). A dosage PCR assay was performed to determine the DAZ copy number using a coamplification of the autosomal DAZ- Like (DAZL) gene as an internal dosage control (30). Both DAZ and DAZL share 89% sequence similarity (11, 34), and the PCR primers used for the dosage assay will recognize both of these genes. The primers were designed from the 3 0 UTR of DAZL; PrDAZ109 contained a mismatch with the DAZ sequence, whereas PrDAZ110 is a perfect match. The PCR amplification was performed as previously described elsewhere (30). The DAZL and DAZ amplified fragments are 331 base pairs (bp) and 245 bp in size, respectively. Dosage of each fragment was measured by the intensity of each band using Quantity One Basic software (Bio- Rad Laboratories, Hercules, CA). The DAZ/DAZL signal ratios were calculated, and knowing that the number of DAZL copies is constant (two copies), we can infer the Fertility and Sterility â 1925

3 number of DAZ copies for each individual. As a control, we used three DNA samples with a known number of DAZ copies (gift from P. Yen, Institute of Biomedical Sciences, Taiwan). These were termed C2, C4, and C6, and each contained two, four, and six DAZ copies, respectively. Where sufficient DNA was available, Southern blot analysis was performed. Briefly 5 mg of genomic DNA from lymphocytes was digested with NsiI and migrated on a 0.8% agarose gel for 6 hours, then was transferred by capillary blotting to a nylon member (Hybond Nþ; GE Healthcare, Piscataway, NJ). Membranes were hybridized with a 32 P-labeled probe consisting of amplified portions of the 3 0 UTR region of DAZ, as previously described elsewhere (30). Hybridization signals were detected using a phosphoimager, and copy numbers were determined using Quantity One Basic software (Bio-Rad). RESULTS Incidence of Partial AZFc Deletions in Case and Control Groups The men who did not carry a deletion for AZFa, AZFb, or AZFc according to EAA guidelines were screened for partial AZFc deletions. This was a total of 364 men with idiopathic infertility associated with reduced sperm counts. Among the patients and controls studied, a total of 32 men were found to have partial deletions for the AZFc region. A summary of the results is shown in Table 1. Eighteen out of 364 (4.95%) partial AZFc deletions were observed in the infertile group compared with 14 out of 193 (7.25%) in the control group. In each group of men, the frequency of deletions were azoospermic 5 out of 115 (4.3%), severe oligozoospermic 6 out of 72 (8.3%), moderate oligozoospermia 4 out of 159 (2.5%), and mild oligozoospermia 3 out of 18 (1.16%). A brief clinical description of the men carrying partial AZFc deletions is given in Table 1. Y-Chromosome Lineages Defined by Binary Markers A number of different Y-chromosome lineages was observed, indicating that these deletions have occurred multiple times in different Y-chromosome lineages (see Table 1). The haplogroup E3b2 predominated in both the control and patient groups. This haplogroup is present in approximately 40% of the North African population (35). The relatively high frequency in our study populations simply reflects the ethnic origin of the individuals who were recruited for our study at the Tenon Hospital in Paris. More than 50% of our patients were of North African orgin, and, in particular, of Moroccan origin. DAZ Gene Dosage The DAZ gene dosage assay was performed when sufficient DNA was available for multiple testing. In all samples, two fragments corresponding to the DAZ and DAZL genes were amplified. In each assay (>10) the control samples with known DAZ copy number consistently showed a DAZ/DAZL ratio of 1:1 (two copies), 2:1 (four copies), or 3:1 (six copies) (Table 2, Fig. 1). All the gr/gr-deleted patients revealed a DAZ/DAZL ratio of 1; the exception was two individuals, Y248 and Y566, who each had a DAZ/DAZL ratio of 2, indicating the presence of four DAZ copies. This suggests that both individuals carry a partial DAZ deletion associated with a duplication of the AZFc region. This observation has been previously reported in other studies (13, 16, 31). Because the two individuals, Y248 and Y566, carrying four DAZ copies were fertile and oligozoospermic, respectively, and all the other gr/ gr samples assayed for DAZ copy number indicated the presence of only two DAZ genes, there does not seem to be a simple association between the DAZ copy number and the fertility status of the individual. When sufficient DNA was available, these results were confirmed by Southern blot analysis (see Table 2), which confirmed the presence of four DAZ copies in individual Y248 and two copies in the other samples tested. Taken together, these data indicate that gene conversion events do not seem to explain the single-nucleotide variant (SNV)/sequence-tagged site (STS) profiles that we observed. Partial AZFc Deletion Variants The use of informative sequence family variants (SFVs) revealed a number of deletion variants that resulted in the apparent copy loss of different DAZ or CDY1 gene variants. The most parsimonious interpretation of these data is shown in Figure 2 and summarized in Table 3. The parental AZFc architectural structure can be deduced for the majority of these deletion classes based on recent data in a worldwide survey of Y-chromosome architecture (31). The first two deletion classes (I and II) occur on the parental Y-chromosome reference sequence. The first class of deletion is represented by the absence of the unique marker sy1291, the DAZ1/2 cluster, and the CDY1a gene. This deletion class can be formed by a recombination between g1 and g2 on a parental Y chromosome carrying the AZFc reference sequence (see Fig. 2). The second, closely related class is characterized by the loss of sy1291, DAZ3/4, and CDY1b and may represent a deletion event resulting from ectopic recombination between the r2 and r4 blocks of homology, again on the background of the reference Y-chromosome sequence. These deletions occurred in infertile as well as fertile/normospermic individuals belonging to different Y-chromosome lineages, and they were the most frequent deletion classes observed in our study (see Table 3). The origin of the third class of deletions can be deduced by assuming that the deletions occurred on a parental Y-chromosome AZFc architecture termed c10. The c10 AZFc region arose by a gr/rg inversion on the Y-reference sequence. This results in a chromosome structure that places DAZ3/4 between the amplicons b2/b3 and DAZ1/2 between the amplicons b3/b4. Class III deletions are caused by ectopic recombination between b3/b4 amplicons. Class III deletions were observed more frequently in infertile men than the class IV 1926 Ravel et al. AZFc architecture and male fertility Vol. 92, No. 6, December 2009

4 TABLE 1 Case and control samples with partial AZFc deletions. Phenotype Markers Sample Fertility status Sperm count (10 6 /ml) sy1191 sy / / / /1026 Haplogroup Control C77 Fertile D 0 DAZ1/2 CDY1a 0 0 E3b2 C83 Fertile D 0 DAZ1/2 CDY1a 0 0 E3b2 C93 Fertile D 0 DAZ1/2 CDY1a 0 0 E3b2 Y248* Fertile D DAZ3/4 CDY1b E3a Y D DAZ3/4 CDY1b R1b Y D 0 0 CDY1a DAZ3/4 0 E3b2 C5 56 D 0 DAZ1/2 0 0 CDY1b E3b2 C4 102 D 0 DAZ1/2 CDY1a 0 0 E3b2 C2 Fertile D 0 DAZ1/2 0 0 CDY1b E3b2 C9 111 D 0 DAZ1/2 0 0 CDY1b R1a1 Y D DAZ3/4 CDY1b P(xR1a1) Y574 Fertile 30 D 0 DAZ1/2 0 0 CDY1b I Y D 0 CDY1a DAZ3/4 0 E3b2 Y602 Fertile 100 D 0 0 CDY1a DAZ3/4 0 J Infertile D1 0 D 0 0 CDY1a DAZ3/4 0 I O7 0.9 D 0 0 CDY1a DAZ3/4 0 I S D 0 0 CDY1a DAZ3/4 0 I Y117 0 D 0 0 CDY1a DAZ3/4 0 K(xP) Y D 0 DAZ1/2 0 0 CDY1b E3b2 Y D 0 0 CDY1a DAZ3/4 0 R1b Y D 0 CDY1a DAZ3/4 0 E3b2 Y D 0 DAZ1/2 CDY1a 0 0 I Y D 0 DAZ1/2 0 0 CDY1b E3b2 Y D DAZ1/2 CDY1a 0 0 N3 Y206 0 D 0 DAZ1/2 0 0 CDY1b E3b2 Y D DAZ3/4 CDY1b R1a1 Y566* 0.25 D 0 0 CDY1a DAZ3/4 CDY1b E3b2 Y D DAZ3/4 CDY1b P(xR1a1) Y592 0 D DAZ3/4 CDY1b E3a Y D 0 CDY1a DAZ3/4 0 E3b2 Y612 2 D 0 DAZ1/2 0 0 CDY1b E3b2 Y D DAZ3/4 CDY1b E3a Notes: 0 ¼ absence, þ ¼ presence. * Four copies of DAZ, sperm count values of both case and control individuals according to World Health Organization guidelines and the corresponding Y haplogroup. Ravel. AZFc architecture and male fertility. Fertil Steril deletion although the difference in frequencies is not statistically significant. Class I deletions can also arise on this parental chromosome structure via an ectopic g2/g1 deletion. Class IV deletions are characterized by the absence of the marker sy1291, and the DAZ3/4 and CDY1b variants. This deletion could arise on a parental chromosome carrying the AZFc reference structure on which a b3/b4 inversion (shown) or a P1 inversion (not shown) has occurred. Such an event has been previously hypothesized, and ectopic recombination between r2 and r3 amplicons can generate this deletion pattern. This Y-chromosome architecture can also generate class III deletions. The parental Y-chromosome architecture on which these deletions have been generated would be difficult to detect by two-color interphase fluorescence in situ hybridization (FISH) techniques, and this may explain why it has not Fertility and Sterility â 1927

5 TABLE 2 DAZ dosage and Southern blot analyses of DAZ copy number in samples with the detected partial AZFc deletions. Sample DAZ dosage PCR Average ratio ± standard error Ratio DAZ/DAZL copies by Southern blot Inferred DAZ copies Control Control Control C C C Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Note: The inferred copy number is indicated. Dashes indicate not determined. PCR, polymerase chain reaction. Ravel. AZFc architecture and male fertility. Fertil Steril been observed previously although it has been hypothesized to exist (16). Two classes of deletion are shown in Figure 2 that were not observed in this study group although the parental architecture has been previously described. These are variants of the b2/b3 deletion that may be a polymorphic variant (14). Both of these classes, V and VI, occur on a parental background of a b2/b3 inversion. No individuals were observed with these deletion classes; however, a b2/b3 inversion event is necessary to explain in part the deletion classes VIII and IX. Case Y566, a man of West African geographic origin, had an E3a Y-chromosome haplogroup. The SNV analysis indicated that this individual lacked the marker sy1291 and DAZ1/2 gene copies, and yet the DAZ dosage analysis has indicated that he carries four copies of the DAZ gene. The AZFc architecture termed c10 (see Fig. 2) is characterized by a gr/rg inversion that localizes DAZ1/2 copies adjacent to the CDY1b gene and distal to sy1291 (31). An individual with a complex AZFc structure (AZFc architecture c36) that was hypothesized to have arisen by a b2/b3 duplication of an ancestral c10 chromosome has been reported (see Fig. 2) (31). The resulting chromosome contains two copies of CDY1a, one copy of CDY1b, four copies of DAZ3/4, and two copies of DAZ1/2. Therefore, a deletion (class VII), on this parental background, involving recombination between distal g2 and g1 amplicons will result in a chromosome lacking DAZ1/2 and sy1291 while retaining both CDY1 copies, the marker sy1191, and four copies of DAZ (DAZ3/4). Deletion classes VIII and IX are both variants of the b2/b3 deletion (13). Deletion class VIII can be deduced by assuming that the deletion occurred on a parental Y-chromosome architecture containing two inversions: a b3/b4 inversion and a b1/b3 inversion. This is essentially a combination of the two types of AZFc structure described previously, although the order of the inversion events that generated this structural variation is known. It could have been generated by a b3/b4 inversion followed by a b2/b4 inversion (as shown) or a b2/b3 inversion followed by a b2/b4 inversion (not shown). In either scenario, both the marker sy1191 and the CDY1b gene are localized between the gr/gr amplicons. A g1/g2 deletion will lead to the absence of sy1191, CDY1b, and DAZ3/4. This b2/b3 deletion variant was observed in one infertile individual (Y446) with the N3 haplogroup. Deletion class IX is characterized by the absence of sy1191, DAZ1/2, and CDY1b, and it can 1928 Ravel et al. AZFc architecture and male fertility Vol. 92, No. 6, December 2009

6 FIGURE 1 Representative polymerase chain reaction amplifications for evaluation DAZ copy number. DAZL and DAZ specific fragments are indicated. Control samples C2, C4, and C6 are shown together with several examples of amplifications from case and control samples. M, molecular-weight ladder. times, respectively. In three lineages, a more complex architecture was observed: C6 consists of a b2/b4 duplication, c7 a b2/b3 inversion, c8 a gr/gr deletion, and c10 a gr/rg inversion. Also, c36 and c38 are generated by duplication on a c10 parental background, and c35 is a b2/b3 deletion that can occur on a c10 parental background. In our study, the majority of cases can be explained by a simple deletion occurring on either the background of the reference Y-chromosome structure (14 cases) or a Y chromosome carrying an inversion as a result of recombination between either b4/ b3 repeat amplicons or between the arms of the P1 palindrome (13 cases). Of the remaining five cases, one is a gr/ gr deletion occurring on the c7 paternal background, three are gr/gr deletions occurring on a c7 architecture that also contained a b3/b4 inversion, and one is a gr/gr deletion on the parental c36 architecture. Partial AZFc Deletions on Multiple Y-Chromosome Haplogroups All of the samples that were analyzed in this study were from a Parisian clinic. The majority of samples were from men of northern or western African origin, and this was reflected in the predominance of the haplogroup E associated with partial AZFc deletions (17 out of 32). The incidence of the haplogroup therefore reflects the recruitment of the study population rather than a bias for deletions in this haplogroup. The E3b2 haplogroup exhibits the seven parental AZFc structures inferred in this study. This highlights the multitude of different AZFc structural variants that occur even within restricted lineages. Ravel. AZFc architecture and male fertility. Fertil Steril occur on the same parental chromosome architecture as deletion class VIII. Two individuals with the E3b2 haplogroup carried this b2/b3 deletion variant (Y595 and Y252). DISCUSSION We have demonstrated that the frequency of partial AZFc deletions is similar between infertile and fertile/normospermic men. The presence of an associated DAZ duplication does not appear to explain the similar partial AZFc deletion frequencies in our patients and controls. Additionally, the DAZ dosage analysis excludes the possibility that gene conversion events may explain the SNV/STS deletion assay. Partial AZFc Deletions on Multiple Parental AZFc Chromosome Architectures The Y-chromosome architecture has been deduced by interphase FISH on 47 Y chromosomes, representing each of the major branches of the genealogical tree (31). The Y-chromosome reference structure was observed in 27 lineages, and structures referred to as c6, c7, c8, c10, c35, c36, and c38 were observed one, three, one, seven, three, one, and one b2/b3 Deletion Variants In a survey of the Y chromosome in 1563 men, Repping et al. (14) found a deletion in 25 individuals that removed the sy1191 marker but all flanking markers were present. Of the 25 men, 21 belonged to Y-chromosome haplogroup N3, where this deletion appears to be fixed and may have a common founder Y chromosome. The other men carrying the b2/ b3 deletion belonged to haplogroups N*(xN3), F*, I, and Q3. The postulated mechanism to generate this deletion involves either a gr/rg inversion followed by a b2/b3 deletion, or a b2/ b3 inversion followed by an rg/rg deletion. In the former scenario, it is predicted that the deletion would be associated with the loss of DAZ3/4 and CDY1a, and in the latter scenario CDY1a is deleted with either DAZ1/2 or DAZ3/4. We observed two individuals with the absence of sy1191, DAZ1/ 2, and CDY1a, suggesting that the latter pathway can generate the b2/b3 deletion. The other individuals with this deletion, including one individual with the N3 haplogroup, did not show these deletion profiles. Further investigation is necessary to determine whether the other N3 chromosomes that are present in high frequencies in northern Eurasia also carry this deletion profile. Fernandes et al. (36), in their study of 15 men with haplogroup N, deduced that each was deleted for DAZ3/4, Fertility and Sterility â 1929

7 FIGURE 2 Schematic representations of the AZFc region structural architecture. Seven AZFc chromosome structures are shown that can explain all the deletion classes observed in the case and control samples. The reference AZFc sequence is indicated, and the organization of amplicons belonging to various families are shown in colors (b1 to b4, blue; r1 to r4, red; g1 to g3, green). Two small palindromes, P1.1 and P1.2 (yellow), lie within a larger P1 palindrome. Sequences with the same color code exhibit >99.9% sequence identity. The relative location of CDY1 and DAZ gene copies and the STS markers used in this study are indicated underneath their respective amplicons. Ravel. AZFc architecture and male fertility. Fertil Steril Ravel et al. AZFc architecture and male fertility Vol. 92, No. 6, December 2009

8 TABLE 3 Partial AZFc deletions and associated phenotypes. Partial AZFc class Deleted genes Infertile (Hg) Fertile/normospermic (Hg) I DAZ1/2, CDY1a Y549 (R1a1), Y583 (P(xR1a1)), Y592 (E3a), Y728 (E3a) H2 (E3b2), Y248* (E3a), Y315 (P(xR1a1)), Y516 (P(xR1a1)) II DAZ3/4, CDY1a Y206 (E3b2), Y324 (E3b2), Y455 N5 (E3b2), Y574 (I) (E3b2), Y612 (E3b 2) III DAZ1/2, CDY1b D1 (I), O7 (I), S6 (I), Y117 (K(xP)), Y359 (E3b2), Y602 (J) Y389 (Px(R1a1)) IV DAZ3/4, CDY1a Y454 (I) T4 (E3b2), N9 (R1a1), C77 (E3b2), C83 (E3b2), C93 (E3b2) V (b2/b3) DAZ1/2, CDY1a VI (b2/b3) DAZ3/4, CDY1a VII DAZ1/2 Y566* (E3b2) VIII (b2/b3) DAZ3/4, CDY1b Y446 (N3) IX (b2/b3) DAZ1/2, CDY1b Y595 (E3b2), Y252 (E3b2) Y529 (E3b2) Notes: Genes that are absent as a consequence of the deletions deduced by single-nucleotide variant analysis are indicated. The associated phenotypes with Y-chromosome haplogroups are indicated. * Denotes four DAZ copies. Ravel. AZFc architecture and male fertility. Fertil Steril although their CDY status was not investigated. The effect of the absence of DAZ3/4 and CDY1b on sperm numbers may be minimal (37). Lack of Association between DAZ Gene Copy Number and Fertility Status We could find no evidence to support a relationship between the presence or absence of either DAZ1/2 or DAZ3/4 gene copies and the fertility status at least on the E3b2 background. The relatively small number of other haplogroups carrying partial AZFc deletions in our study limits the interpretation of the DAZ data. Several studies have suggested that the absence of DAZ1/2 gene copies is associated with spermatogenic failure, whereas the absence of DAZ3/4 has little effect on fertility status (15, 19). The fact that haplogroup N chromosomes lack the DAZ3/4 gene cluster while retaining the DAZ1/2 cluster also supports this hypothesis (36). In our study, the possibility that a gene conversion event may remove the variant that distinguishes between the DAZ1/2 and DAZ3/4 clusters was excluded by the gene dosage approaches. An association between the absence of the DAZ1/2 cluster and male infertility has not been supported by other studies. Machev et al. (16) found an association between the absence of DAZ3/4 CDY1a and spermatogenic failure but not with absence of the DAZ1/2 cluster. Other studies have specifically tested the hypothesis that the absence of DAZ1/2 is a cause of male infertility, and in each case they have failed to detect such an association (25, 26). Indeed, in some populations such as East Asians, the DAZ1/2 deletion is present at frequencies of up to 8% in the general population. Zhang et al. (26) observed that gr/gr deletions are present at a frequency of approximately 15% of the Northern Han and Tujia populations and that these deletions are present on a number of different Y-chromosome haplogroups, suggesting that the gr/gr deletion is under limited selective pressure. In the same study, the mean sperm concentrations of the donors who carried gr/gr deletions was not significantly different from the mean sperm concentrations of wild-type individuals. There also was no significant difference in Y-chromosome haplogroup frequencies between the case and control groups, and most gr/gr deleted DNA samples from both case and control cohorts carried two DAZ gene copies (26). Therefore, the variation of fertility status between gr/gr-deleted individuals was not related to DAZ copy number. Most of the gr/gr deletions, from both infertile men and sperm donors, lacked the DAZ1/DAZ2 cluster and carried the DAZ3/DAZ4 gene cluster. CDY1 Variants and Fertility We did not detect an association between either CDY1a or CDY1b and male infertility. Investigators have postulated that CDY is involved in the hyperacetylation of histones during the maturation of spermatids, which is a prerequisite step for the displacement of most of the histone proteins by protamines during the final stages of spermatogenesis (38). Machev et al. (16) suggested a relationship between the absence of DAZ3/4 together with CDY1a and fertility status; however, in our study we observed the absence of CDY1a Fertility and Sterility â 1931

9 associated with a gr/gr deletion in both case and control populations at similar frequencies. Other Y-Chromosome Genes and Male Infertility A number of other Y-chromosome gene families are located within the AZFc region, and it could be hypothesized that the absence of some of these genes may contribute to infertility (2). Indeed, an analysis of the chromosome architectures that are illustrated in Figure 2 shows that several of these genes are predicted to be lost in the various classes of partial AZFc deletions that are described. These gene families include BPY (also known as VCY) is expressed only in the testis, and encodes a small, positively charged protein of unknown function (39). These characteristics of both a small size and a high charge have led to the suggestion that BPY proteins may functionally interact with DNA in a manner that resembles chromatin-associated proteins such as histones and HMG proteins. Two copies of BPY2 are located at the boundaries of the gr/gr deletion, flanking the DAZ gene clusters (see Fig. 2). In close physical association with BPY2 are the testis-specific transcripts TTY3, TTY4, and TTY17, which are transcribed but apparently do not encode proteins. Palindrome P1 also contains an 80-kb fragment with 95% nucleotide identity to human chromosome 15 and the GOLGA2-like, Y-linked (GOLGA2LY), and chondroitin sulphate proteoglycan 4-like Y-linked (CSPG4LY) transcripts are located within this region (2). The function, if any, of these two gene families is unknown. We did not observe a relationship between partial AZFc deletions and spermatogenic failure. Some rare deletion events were detected in individuals with male factor infertility and not in the control panel, and in these examples we cannot exclude that the deletion is responsible for the phenotype. We found no evidence of an association between the absence of specific DAZ or CDY1 gene copy variants and fertility status. Clinically, it may be premature to screen for partial AZFc deletions as a diagnostic test, but it is also premature to exclude the possibility that partial deletions are a minor risk factor that could contribute to a reduction in total sperm counts in a given individual. Acknowledgments: The authors thank Florence Dastot, Nathalie Collot, and Jo elle Bignon-Topalovic for their technical help. REFERENCES 1. McElreavey K, Krausz C, Bishop CE. The human Y chromosome and male infertility. Results Probl Cell Differ 2000;28: Skaletsky H, Kuroda-Kawaguchi T, Minx PJ, Cordum HS, Hillier L, Brown LG, et al. The male-specific region of the human Y chromosome is a mosaic of discrete sequence classes. Nature 2003;423: Krausz C, Degl Innocenti S. Y chromosome and male infertility: update. Front Biosci 2006;11: Carrell DT. The genetics of male infertility. Totowa, NJ: Humana Press, Vogt PH. Azoospermia factor (AZF) in Yq11: towards a molecular understanding of its function for human male fertility and spermatogenesis. Reprod Biomed Online 2005;10: Repping S, Skaletsky H, Lange J, Silber S, Van Der Veen F, Oates RD, et al. Recombination between palindromes P5 and P1 on the human Y chromosome causes massive deletions and spermatogenic failure. Am J Hum Genet 2002;71: Ferlin A, Moro E, Rossi A, Dallapiccola B, Foresta C. The human Y chromosome s azoospermia factor b (AZFb) region: sequence, structure, and deletion analysis in infertile men. J Med Genet 2003;40: Reijo R, Lee TY, Salo P, Alagappan R, Brown LG, Rosenberg M, et al. Diverse spermatogenic defects in humans caused by Y chromosome deletions encompassing a novel RNA-binding protein gene. Nat Genet 1995;10: Vogt PH, Edelmann A, Kirsch S, Henegariu O, Hirschmann P, Kiesewetter F, et al. Human Y chromosome azoospermia factors (AZF) mapped to different subregions in Yq11. Hum Mol Genet 1996;5: Vogt PH. AZF deletions and Y chromosomal haplogroups: history and update based on sequence. Hum Reprod Update 2005;11: Saxena R, de Vries JW, Repping S, Alagappan RK, Skaletsky H, Brown LG, et al. Four DAZ genes in two clusters found in the AZFc region of the human Y chromosome. Genomics 2000;67: Kuroda-Kawaguchi T, Skaletsky H, Brown LG, Minx PJ, Cordum HS, Waterston RH, et al. The AZFc region of the Y chromosome features massive palindromes and uniform recurrent deletions in infertile men. Nat Genet 2001;29: Repping S, Skaletsky H, Brown L, van Daalen SK, Korver CM, Pyntikova T, et al. Polymorphism for a 1.6-Mb deletion of the human Y chromosome persists through balance between recurrent mutation and haploid selection. Nat Genet 2003;35: Repping S, van Daalen SK, Korver CM, Brown LG, Marszalek JD, Gianotten J, et al. A family of human Y chromosomes has dispersed throughout northern Eurasia despite a 1.8-Mb deletion in the azoospermia factor c region. Genomics 2004;83: Fernandes S, Huellen K, Goncalves J, Dukal H, Zeisler J, Rajpert De Meyts E, et al. High frequency of DAZ1/DAZ2 gene deletions in patients with severe oligozoospermia. Mol Hum Reprod 2002;8: Machev N, Saut N, Longepied G, Terriou P, Navarro A, Levy N, et al. Sequence family variant loss from the AZFc interval of the human Y chromosome, but not gene copy loss, is strongly associated with male infertility. J Med Genet 2004;41: Hucklenbroich K, Gromoll J, Heinrich M, Hohoff C, Nieschlag E, Simoni M. Partial deletions in the AZFc region of the Y chromosome occur in men with impaired as well as normal spermatogenesis. Hum Reprod 2005;20: de Llanos M, Ballesca JL, Gazquez C, Margarit E, Oliva R. High frequency of gr/gr chromosome Y deletions in consecutive oligospermic ICSI candidates. Hum Reprod 2005;20: Ferlin A, Tessari A, Ganz F, Marchina E, Barlati S, Garolla A, et al. Association of partial AZFc region deletions with spermatogenic impairment and male infertility. J Med Genet 2005;42: Lynch M, Cram DS, Reilly A, O Bryan MK, Baker HW, de Kretser DM, et al. The Y chromosome gr/gr subdeletion is associated with male infertility. Mol Hum Reprod 2005;11: Giachini C, Guarducci E, Longepied G, Degl Innocenti S, Becherini L, Forti G, et al. The gr/gr deletion(s): a new genetic test in male infertility? J Med Genet 2005;42: Nathanson KL, Kanetsky PA, Hawes R, Vaughn DJ, Letrero R, Tucker K, et al. The Y deletion gr/gr and susceptibility to testicular germ cell tumor. Am J Hum Genet 2005;77: Fernando L, Gromoll J, Weerasooriya TR, Nieschlag E, Simoni M. Y-chromosomal microdeletions and partial deletions of the Azoospermia Factor c (AZFc) region in normozoospermic, severe oligozoospermic and azoospermic men in Sri Lanka. Asian J Androl 2006;8: Ravel C, Chantot-Bastaraud S, El Houate B, Mandelbaum J, Siffroi JP, McElreavey K. GR/GR deletions within the azoospermia factor c region on the Y chromosome might not be associated with spermatogenic failure. Fertil Steril 2006;85: Yang Y, Xiao Cy, A ZC, Zhang SZ, Li X, Zhang SX. DAZ1/DAZ2 cluster deletion mediated by gr/gr recombination per se may not be sufficient for spermatogenesis impairment: a study of Chinese normozoospermic men. Asian J Androl 2006;8: Ravel et al. AZFc architecture and male fertility Vol. 92, No. 6, December 2009

10 26. Zhang F, Li Z, Wen B, Jiang J, Shao M, Zhao Y, et al. A frequent partial AZFc deletion does not render an increased risk of spermatogenic impairment in East Asians. Ann Hum Genet 2006;70: Carvalho CM, Zuccherato LW, Bastos-Rodrigues L, Santos FR, Pena SD. No association found between gr/gr deletions and infertility in Brazilian males. Mol Hum Reprod 2006;12: de Carvalho CM, Zuccherato LW, Fujisawa M, Shirakawa T, Ribeirodos-Santos AK, Santos SE, et al. Study of AZFc partial deletion gr/gr in fertile and infertile Japanese males. J Hum Genet 2006;51: Tyler-Smith C. An evolutionary perspective on Y-chromosomal variation and male infertility. Int J Androl 2008;31: Lin YW, Hsu CL, Yen PH. A two-step protocol for the detection of rearrangements at the AZFc region on the human Y chromosome. Mol Hum Reprod 2006;12: Repping S, van Daalen SK, Brown LG, Korver CM, Lange J, Marszalek JD, et al. High mutation rates have driven extensive structural polymorphism among human Y chromosomes. Nat Genet 2006;38: Simoni M, Bakker E, Krausz C. EAA/EMQN best practice guidelines for molecular diagnosis of y-chromosomal microdeletions. State of the art Int J Androl 2004;27: Y Chromosome Consortium. A nomenclature system for the tree of human Y-chromosomal binary haplogroups. Genome Res 2002;12: Saxena R, Brown LG, Hawkins T, Alagappan RK, Skaletsky H, Reeve MP, et al. The DAZ gene cluster on the human Y chromosome arose from an autosomal gene that was transposed, repeatedly amplified and pruned. Nat Genet 1996;14: Arredi B, Poloni ES, Paracchini S, Zerjal T, Fathallah DM, Makrelouf M, et al. A predominantly neolithic origin for Y-chromosomal DNA variation in North Africa. Am J Hum Genet 2004;75: Fernandes S, Paracchini S, Meyer LH, Floridia G, Tyler-Smith C, Vogt PH. A large AZFc deletion removes DAZ3/DAZ4 and nearby genes from men in Y haplogroup N. Am J Hum Genet 2004;74: Jorgensen N, Asklund C, Carlsen E, Skakkebaek NE. Coordinated European investigations of semen quality: results from studies of Scandinavian young men is a matter of concern. Int J Androl 2006;29: Lahn BT, Tang ZL, Zhou J, Barndt RJ, Parvinen M, Allis CD, et al. Previously uncharacterized histone acetyltransferases implicated in mammalian spermatogenesis. Proc Natl Acad Sci USA 2002;99: Lahn BT, Page DC. A human sex-chromosomal gene family expressed in male germ cells and encoding variably charged proteins. Hum Mol Genet 2000;9: Fertility and Sterility â 1933