FERTILITY AND STERILITY VOL. 73, NO.

2, FEBRUARY 2000
Copyright 2000 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A.

A novel, rapid, and accurate method for detecting microdeletion involving the DAZ gene in infertile men
Herve Lucas, M.D.,* Catherine Patrat, M.D.,* Pierre Jouannet, M.D.,* Cherif Beldjord, Ph.D., and Thierry Bienvenu, Ph.D.*
Hopital Cochin, Paris, France

Objective: To report on a novel, accurate method for detecting microdeletion involving the DAZ gene in infertile men. Design: Retrospective clinical study. Setting: University Infertility Center of Cochin Hospital, Paris, France. Patient(s): Infertile patients (n 25) consulting our infertility department during 1998. The patient cohort included subjects with nonobstructive azoospermia and oligoasthenospermia. Intervention(s): Blood samples were collected from each subject. Main Outcome Measure(s): DNA analysis using polymerase chain reaction (PCR) denaturing gradient gel electrophoresis (DGGE). Result(s): We used a new molecular genetic strategy to rapidly identify deletions of the Y chromosome that include the DAZ locus. The experiment consists of amplifying simultaneously exon 4 of the DAZ and DAZLA genes with the use of specic primers that are complementary to intronic sequences of these genes. DGGE was used to separate the two PCR products, with good resolution. In infertile men with a microdeletion of the DAZ gene, this method allows amplication of an internal control when a deletion of that portion of the Yq chromosome is observed on a single amplication. Conclusion(s): This PCR-DGGE method for detection of DAZ gene deletion is simple and fast and does not require the use of radioactive elements. Compared with the classic PCR approach, this new method allows the amplication of the DAZLA copy to be used as an effective internal control in infertile men with microdeletion of the DAZ locus. This procedure could be particularly useful in screening for the DAZ locus in the diagnostic workup of nonobstructive azoospermia and severe oligoasthenoteratozoospermia. (Fertil Steril 2000;73: 2427. 2000 by American Society for Reproductive Medicine.) Key Words: Infertility, human DAZ gene, denaturing gradient gel electrophoresis

Received June 24, 1998; revised and accepted August 25, 1999. Supported by a grant from Assistance PubliqueHopitaux de Paris (Contrat Recherche Clinique #96053). Reprint requests: Thierry Bienvenu, Ph.D., Laboratoire de Biochimie et Genetique Moleculaire, Hopital Cochin, 123 boulevard de Port-Royal, 75014 Paris, France (FAX: 01-44-41-15-22; E-mail: bienvenu@cochin.inserm.fr). * Service dHisto-Embryologie, Biologie de la Reproduction. Service de Biochimie et Genetique Moleculaire.
0015-0282/00/$20.00 PII S0015-0282(99)00504-X

In 1976, Tiepolo and Zuffardi (1) reported de novo, microscopically detectable deletions of the distal half of Yq in four men with azoospermia, and on this basis they postulated the existence of one or more critical Yq genes for spermatogenesis. In recent years, this azoospermia factor (AZF) hypothesis has been widely validated. Exploiting the availability of comprehensive DNA probe based physical maps of the Y chromosome, investigators have reported many interstitial Yq deletions in infertile men with normal karyotypes. In particular, overlapping de novo deletions within intervals 6D 6E of the Y chromosome have been shown to cause at least 13% of cases of nonobstructive

azoospermia and some cases of severe oligospermia as well (2 4). Three nonoverlapping regions of microdeletions in Yq11 (AZFa, AZFb, and AZFc) have been identied that are probably responsible for azoospermia or oligozoospermia. The candidate gene for the expression of AZFc is called DAZ (Deleted In AZoospermia). The DAZ gene is absent in some men with azoospermia or severe oligospermia. The DAZ gene is an attractive candidate for AZF. In previous reports, the screening program was based mainly on a multiplex polymerase chain reaction (PCR) approach with use of a series of Y-specic primers amplifying single DNA loci in

242

FIGURE 1 Comparison of nucleic acids for exon 4 and anking sequences of the human DAZLA and DAZ genes (9). The sequences of the genomic fragments of DAZ and DAZLA have been aligned with the GAP program by Seboun et al. (9). Intron sequences are shown in lowercase letters and transcribed sequences in uppercase letters. Dots indicate gaps introduced to maximize the alignments. The two primers used in the PCR amplication are underlined. Nucleotide sequence differences between the DAZ and DAZLA genes are shown as letters with double underlines.

crodeletion of the DAZ locus, 2 infertile patients with a microdeletion that involved the DAZ gene, and 10 healthy controls (5 fertile men and 5 fertile women). On the basis of semen analysis (World Health Organization criteria of 1992) (7), 2 patients had a diagnosis of nonobstructive azoospermia and 20 had oligoasthenozoospermia ( 20 million sperm per millimeter of ejaculate; 40% a b motility). Among the latter group, 10 had severe oligospermia ( 5 million sperm per millimeter of ejaculate) and 3 had isolated asthenospermia. Patients with obstructive azoospermia were excluded from the study. Cytogenetic analysis was performed for all patients on peripheral blood cells cultured for 72 hours in the presence of phytohemagglutinin. We also obtained blood samples from the fathers of the two infertile men who were found to have Y-chromosome microdeletions. The control group consisted of ve men who had normal characteristics on the spermogram (7, 8) and who had fathered at least one child in the last 5 years. Written consent for use of the remaining samples for research was obtained from the patients and from their fathers according to the guidelines for research on human subjects established by the national and local committees of ethics.

Lucas. Detecting microdeletion. Fertil Steril 2000.

Yq11 (i.e., sY254 and sY255) (2 4). A sequence-tagged site was considered absent only after at least three amplication failures in the presence of successful amplication of the control. This strategy requires many precautions to minimize the number of false-negative or false-positive results. It becomes limiting in a clinical context in which service, speed, accuracy, and cost are major concerns. An autosomal homologue of the DAZ gene has been identied on human chromosome 3 and named DAZLA (DAZ-like autosomal) or DAZH (DAZ homologue) (5). This homologue is not identical to the DAZ gene on the Y chromosome because it displays only one sequence repeat of 24 amino acids, whereas the DAZ gene on the Y chromosome has 716 sequence repeats (6). Using the high degree of homology between the DAZ gene and the autosomally equivalent DAZLA gene (Fig. 1) and the denaturing gradient gel electrophoresis (DGGE) system, we show the effectiveness of a single amplication in a clinical setting for the screening of microdeletions that include the DAZ gene. The major advantage of this method is that it allows amplication of a coamplied product of the DAZLA gene as an internal control when a deletion of that portion of the Yq chromosome is observed after a single amplication.

The sY254 and sY255 sequence-tagged sites, specic for the AZFc locus, were used to screen for microdeletion of the DAZ gene in all infertile men (2 4). In previous reports, it was shown that sY254 and sY255 were specic for the DAZ gene (4). SY254 is a product of 330 base pairs (bp) containing the exonic sequences that correspond to nucleotide positions 89 150 and 151194 of the DAZ complementary DNA and one intron. The rst infertile patient who presented with a Y microdeletion had nonobstructive azoospermia. He underwent a testis biopsy, which revealed few spermatozoa. The semen biochemical markers from the epididymis, seminal, or prostatic glands were normal. The plasma FSH level was increased. His father had conceived eight children without any delay to fertilize. The second infertile patient with a Y microdeletion had severe oligoasthenozoospermia. He had a sperm count at 0.33 106 spermatozoa per ejaculate with 5% progressive motility, 20% vitality, and 15% typical forms. The semen biochemical markers from the epididymis, seminal gland, or prostatic gland were normal. The plasma FSH level was subnormal (8.6 IU/L). He was an only child, and his parents had attempted pregnancy for 12 years before his birth. His mother was not infertile. Y-chromosome microdeletions have been conrmed by Southern blot analysis using the pHT4.1 clone as the DNA probe (9). Electrophoresis (10 g of EcoRI-digested DNA), transfer, and hybridization of lters were performed using standard procedures. PCR amplication was performed using 200 ng of 243

MATERIALS AND METHODS

Genomic DNA was isolated by a standard method from peripheral leukocytes of 23 infertile patients without miFERTILITY & STERILITY

FIGURE 2 PCR-DGGE of exon 4 of the DAZLA gene. The samples were amplied with the use of chemiclamped primers. Unclamped incompletely cross-linked DNA fragment. Lanes 1 and 2 fertile male subjects; lane 3 control female subject; lane 4 azoospermic patient with microdeletion involving sY254 and sY255; lane 5 infertile man without microdeletion of the DAZ locus. The lower band corresponds to exon 4 of the DAZ gene and the upper band to exon 4 of the DAZLA gene.

Lucas. Detecting microdeletion. Fertil Steril 2000.

genomic DNA, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 2.5 mM MgCl2, 400 mM of each dNTP, 10 pmol of each primer (4F 5 -ATGATGGTGATTTTATACTT-3 and 4R 5 -psoralen-5 -TAAGAGATTTTTCTATTAGATTA-3 ), and 2 U of Taq DNA polymerase in a total volume of 50 L. Under experimental conditions, DGGE cannot resolve DNA fragments that differ by base variants located within the highest melting temperature (Tm) domains because of the loss of sequence-dependent migration upon complete strand dissociation. This problem is overcome by introducing into the fragment to be analyzed a G-Crich domain (GC clamp) or a psoralen derivative the high Tm of which will prevent complete dissociation. In this study, a psoralen derivative was attached to the 5 end of the forward primer by the manufacturer (Appligene-Oncor, Strasbourg, France). ` The DNA thermal cycler device (Genamp PCR 9600; Perkin Elmer Cetus, Foster City, CA) was programmed to perform 40 cycles of 94C for 20 seconds, 47C for 30 seconds, and 72C for 30 seconds. A 5-minute incubation step at 94C was performed initially, as well as a nal step of 7 minutes at 72C. The amplied DNA was then irradiated with a 365-nm ultraviolet lamp for 15 minutes. After photo-induced cross-linking, the PCR products were analyzed by DGGE. The gel apparatus is the same as initially described (10). One fth of the amplied product was loaded onto a 6% polyacrylamide gel with a parallel gradient increasing linearly from 20% to 70% at 160 V for 7 hours. The fragments were visualized by ultraviolet transillumination after ethidium bromide staining (10). Sequence analysis of the PCR products was performed with 100 ng of puried DNA, 20 ng of primer, and uo244
Lucas et al. Molecular screening of DAZ gene

rescently labeled Taq Dyedeoxy terminator reaction mix (Applied Biosystems, Foster City, CA) according to the manufacturers instructions. The DNA sequence was determined using a 377 automated DNA sequencer (Applied Biosystems).

RESULTS
DNA from healthy controls (male and female) and from infertile men was amplied with primers 4F and 4R. Genomic DNA amplication was accomplished with forward and reverse primers that are complementary to intronic sequences of the DAZLA and DAZ genes. DGGE is a proved method to differentiate subtle DNA sequence differences between DNA fragments of identical length and a nearly identical sequence, as is the case in the DAZ gene family. Thus, analysis of the 117-bp fragment on a 20 70% denaturing gradient gel showed two different bands with good separation and resolution in normal male and female subjects (Fig. 2). Direct sequencing of PCR products showed that the upper band was the result of the amplication of exon 4 of the DAZLA gene and the lower band was the result of the amplication of exon 4 of the DAZ gene (Fig. 3). No band corresponding to the DAZ gene was observed in female subjects. In fertile male subjects and in infertile men without microdeletion of the DAZ locus, exon 4 of the DAZ gene was preferentially amplied (5). In the two unrelated infertile men with a microdeletion of the DAZ locus, only the upper band corresponding to the DAZLA gene (Fig. 2) was observed. In these patients, the PCR-DGGE method allowed amplication of an internal control (exon 4 of the DAZLA gene) when a deletion of that portion of the Yq chromosome (exon 4 of the DAZ gene) was observed on a single amplication.
Vol. 73, No. 2, February 2000

FIGURE 3 Direct sequencing of exon-4 PCR products. (A) Fertile male subject: The PCR product is the result of amplication of the DAZ gene. (B), Azoospermic patient with microdeletion involving sY254 and sY255: The PCR product is the result of amplication of the DAZLA gene (see Fig. 1). Sequence variations between DAZ and DAZLA are indicated (arrows).

Lucas. Detecting microdeletion. Fertil Steril 2000.

Microdeletions of the DAZ locus detected by our novel approach were conrmed by PCR with sequence-tagged sites and by Southern blot analysis (Fig. 4).

DISCUSSION
It is likely that the Y chromosome plays an important role in male germ cell development, as evidenced by deletions in 13% of azoospermic men (2). In previous reports, the Ychromosome screening program was based mainly on a multiplex PCR approach with a series of Y-specic primers amplifying single DNA loci in Yq11 (2 4). A sequencetagged site was considered absent only after at least three amplication failures in the presence of successful amplication of a control. This strategy necessitates many precautions to minimize the number of false-negative or falsepositive results and becomes limiting in a clinical context in which service, speed, accuracy, and cost are major concerns. In our study, we detected DAZ gene deletion with the use of DGGE. This method is unique among mutation scanning methods because the optimal conditions for mutation detection can be predicted using the computer algorithms Melt and SQHTX. This method is based on the principle that the
FERTILITY & STERILITY

mobility of double-stranded DNA in polyacrylamide gel electrophoresis is delayed by denaturation or melting. When double-stranded DNA migrates through a gel containing an increasing denaturant concentration, its mobility will depend on the base composition of the rst domain that melts. A single base pair change in this domain will modify the mobility of the fragment and make it detectable. With the use of this approach, we amplied simultaneously exon 4 of the DAZ gene and exon 4 of the DAZLA gene with a unique pair of primers. The PCR fragments that differ by some nucleotide substitutions migrate at different levels in the polyacrylamide gel. In all cases, a characteristic pattern was observed with good separation and resolution. The ultraviolet cross-linking is never 100% complete, and a small amount of the uncross-linked fragment (20% 40%, depending on the size of the PCR fragment) may add an extraneous band to the gel pattern. The band is easily distinguished; it is rarely focused, is not retarded, and is likely to disappear in a long run. In fertile men and in infertile men without microdeletion of the DAZ gene, we observed that exon 4 of the DAZ gene was preferentially amplied. This nding is probably due to the existence of at least three copies of DAZ exhibiting 245

FIGURE 4 The pHT4.1 probe was hybridized to Southern blot of EcoRIdigested genomic DNA (10 g/lane) from men with microdeletion of the DAZ locus, normal men, and normal women. Lane 1 man with microdeletion of the DAZ locus; lane 2 control female; lanes 3 and 4 control normal males. All Y-specic fragments are absent in the AZFc-deleted man and correspond to the DAZ gene cluster. Fragment H, present in all individuals tested, corresponds to DAZLA. Scale is shown at right.

only one child, suggesting that in his case, the deletion did affect fertility. The other patient with DAZ deletion was azoospermic, and this DAZ deletion was de novo. His father does not have the DAZ deletion and fathered eight children, but we did not examine his sperm. Our system has limitations because deletions other than those specically tested may go undetected. Many of the deletions in interval 6 that were associated with azoospermia were outside the region of the DAZ gene (13). It seems, however, that the DAZ locus is most frequently involved (14). Our novel PCR-DGGE method for detection of DAZ gene deletion is simple and fast and does not involve the use of radioelements. Moreover, the results can be visualized after a single electrophoresis, and a deletion of the DAZ gene allows amplication of the DAZLA copy to be used as an effective internal control. This procedure could be particularly useful for screening of the DAZ locus in the diagnostic workup of nonobstructive azoospermia and severe oligoasthenoteratozoospermia in infertility clinics.

Lucas. Detecting microdeletion. Fertil Steril 2000.

Acknowledgments: The authors thank Catherine Poirot, M.D., Martha De Almeida, M.D., and Jean Marie Kunstmann, M.D., for their contributions. They also thank Dr. Ken McElreavey and coworkers for providing the pHT4.1 clone and Oana Zamr for his help in DNA sequencing.

99.9% sequence identity in the Y chromosome. This is a minimum estimate of a gene copy number; the real number of DAZ copies may be greater (11). We do not know whether the number of DAZ copies is identical in all men. Moreover, the sequence of some DAZ gene copies may vary, at least in some individuals. In the two patients with microdeletion of the Y chromosome, DAZ gene sequences (detected with sequence-tagged sites sY254 and sY255) were absent, whereas PCR analysis of sequence-tagged sites corresponding to the previously described AZFa and AZFb regions was positive (data not shown). For analysis of the more proximal AZFa and AZFb regions, we used the primer pairs sY84 and sY143 which, in the cohort of Vogt et al. (12), were constantly nonampliable in patients with deletions of the loci. Although the precise extension of the deletions was not investigated in detail, it did not involve the AZFa and AZFb loci. In our study, the absence of the DAZ fragment in these infertile patients suggests that all copies of DAZ map in the distal part of the Y-chromosome euchromatin in the AZFc region. Of the fathers of both infertile men whose Y chromosomes we studied, one had microdeletion identical to that of his son. The father of this patient attempted procreation for 12 years before the birth of his son and was able to father 246
Lucas et al. Molecular screening of DAZ gene

References
1. Tiepolo L, Zuffardi O. Localization of factors controlling spermatogenesis in the nonuorescent portion of the human Y chromosome. Hum Genet 1976;34:119 24. 2. 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;10:38392. 3. Reijo R, Alagappan RK, Patrizio P, Page DC. Severe oligozoospermia resulting from deletions of azoospermia factor gene on Y chromosome. Lancet 1996;347:1290 3. 4. Simoni M, Gromoll J, Dworniczak B, Rolf C, Abshagen K, Kamischke A, et al. Screening for deletions of the Y chromosome involving the DAZ (Deleted in AZoospermia) gene in azoospermia and severe oligozoospermia. Fertil Steril 1997;67:5427. 5. 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 amplied and pruned. Nat Genet 1996;14:2929. 6. Patrizio P. Mapping of the Y chromosome and clinical consequence. In: Barratt C, De Jonge C, Mortimer D, Parinaud J, eds. Genetics of human male infertility. Paris: Editions EDK, 1997:25 42. 7. World Health Organization. WHO laboratory manual for the examination of human semen and semen-cervical mucus interaction. 3rd ed. Cambridge: Cambridge University Press, 1992:327. 8. David G, Bisson JP, Czyglik F, Jouannet P, Gernigon C. Anomalies morphologiques du spermatozode humain et proposition pour un systeme de classication. J Gynecol Obstet Biol Reprod 1975;4:1736. ` 9. Seboun E, Barbaux S, Bourgeron T, Nishi S, Algonik A, Egashira M, et al. Gene sequence, localization, and evolutionary conservation of DAZLA, a candidate male sterility gene. Genomics 1997;41:22735. 10. Bienvenu T, Cazeneuve C, Kaplan JC, Beldjord C. Mutation heterogeneity of cystic brosis in France: screening by denaturing gradient gel electrophoresis using psoralen-modied oligonucleotide. Hum Mutat 1995;6:239. 11. Yen PH. A long-range restriction map of deletion interval 6 of the

Vol. 73, No. 2, February 2000

human Y chromosome: a region frequently deleted in azoospermic males. Genomics 1998;54:512. 12. Vogt PH, Edelmann A, Kirsch S, Heneguriu O, Hirschmann P, Kiesewetter F, et al. Human Y chromosome azoospermia factor (AZF) mapped to different subregions in Yq11. Hum Mol Genet 1996;5:933 43. 13. Stuppia L, Gatta V, Mastroprimiano G, Pompetti F, Calabrese G,

Franchi F, et al. Clustering of Y chromosome deletions in subinternal E of internal 6 supports the existence of an oligozoospermia critical region outside the DAZ gene. J Med Genet 1997;34:8813. 14. Hoelfsloot LH, Tuerlings JHAM, Kremer JAM, Meuleman EJH. PCR analysis of Y-chromosome deletions in subinfertile men [letter]. Lancet 1997;349:1400.

FERTILITY & STERILITY

247