Molecular diagnosis of Y chromosome microdeletions in Europe: state-of-the-art and quality control

Manuela Simoni

1Institute of Reproductive Medicine of the University, Domagkstrasse 11, 48129 Münster, Germany. E-mail: simoni{at}uni-muenster.de

Abstract

The polymerase chain reaction (PCR) screening of microdeletions of the Y chromosome has become an important diagnostic step in the work-up of male infertility. However, there is no agreement about how this diagnosis should be performed. There are suggestions that the large variation in deletion frequency reported in the literature could be due to the various selection criteria of the patients analysed, although methodological aspects may play a role as well. As for other genetic diseases, molecular diagnosis of Y chromosome microdeletions should be controlled by adopting strict internal quality control measures and by participating in external quality assessment schemes. Such an external quality assessment project is presently being organized jointly by the European Academy of Andrology and the European Molecular Genetics Quality Network. Three preliminary trials have given a state-of-the-art picture of the diagnostic performance in various European laboratories, showing an overall rate of misdiagnosis of ~5% for both AZFb and AZFc regions, and providing data useful in the generation of guidelines for the molecular diagnosis of Y chromosome microdeletions.

Key words: assisted reproduction/azoospermia factor/male infertility/quality control/Y chromosome deletion

Introduction

The molecular analysis of microdeletions of the Y chromosome is currently performed in many diagnostic and research laboratories as part of work-up of male infertility. Following the discovery of the DAZ gene (Reijo et al., 1995Go) and the description of the three AZF regions (Vogt et al., 1996Go), a large number of studies have shown that AZF deletions are responsible for male infertility characterized by azoospermia or severe oligozoospermia and that such deletions can be transmitted to the offspring via intracytoplasmic sperm injection (ICSI) and testicular sperm extraction (TESE) (for reviews, see Kostiner et al., 1998; Simoni et al., 1998; Vogt, 1998; Kausz and McElreavey, 1999; Ma et al., 2000; Maurer and Simoni, 2000).

With the aim of providing an additional diagnostic tool for male infertility, many laboratories have started offering polymerase chain reaction (PCR)-based deletion analysis using in-house methods. This occurred in the absence of consensus on the procedure and on which loci on the Y chromosome should be analysed, as reflected by the lack of standardized diagnostic kits on the market. Based on the results of quality control trials involving other diagnostic procedures in molecular genetics, it is reasonable to expect that a portion of the diagnoses concerning the Y chromosome microdeletions is wrong, leading to false positive (when non-existent deletions are diagnosed) or false negative (when deletion are not detected) results (Losekoot et al., 1999Go; Dequeker and Cassiman, 2000Go).

In 1997, on the initiative of the European Academy of Andrology (EAA) training centre of Münster, a pilot trial was started in which DNA samples were exchanged between laboratories in several European countries (Simoni, 1998Go). The aims of the trial were: (i) to gain information about the molecular diagnosis of Y-chromosomal microdeletions; (ii) to cross-check the reliability of the various diagnostic protocols; and (iii) to collect data useful for a minimal standardization of this procedure.

Three trials have been completed in 2 years. Two DNA samples were sent to the participating laboratories, which were simply requested to perform the diagnosis according to their current protocol. The results of these trials, presented here, formed the basis for the generation of guidelines for the molecular diagnosis of Y chromosome microdeletions (Simoni et al., 1999Go).

Diagnosis of Y chromosome microdeletions: state of the art

In all, 20 laboratories participated in the first trial, 25 in second trial and 29 in the third trial (Table IGo). Six DNA samples were used in this study. DNA was extracted from EDTA blood leukocytes from patients attending the infertility clinic of the Institute of Reproductive Medicine in Münster (n = 5) using the Nucleon BACC1 kit (Amersham, Braunschweig, Germany) and stored in aliquots of ~100–200 µg/ml. One sample was provided by a participating laboratory. The patients gave their informed, written consent to donate a blood sample for this study. The characteristics of the six DNA samples are summarized in Table IIGo. Two DNA samples were sent by mail to the individual laboratories in each trial. All samples were received and PCR analysis was possible in each case.


View this table:
[in this window]
[in a new window]
 
Table I. List of participants
 

View this table:
[in this window]
[in a new window]
 
Table II. DNA samples distributed in the three trials
 
Each laboratory was requested to perform PCR diagnosis of the Y chromosome according to its usual protocol. The individual protocols were reported to the centre only in part. Seven laboratories used a multiplex PCR approach, five simplex PCR, while the format adopted by the remaining 17 laboratories was not specified.

Although reporting is a very important component of the diagnostic procedure which is considered in quality assessment schemes of genetic diseases, the participating laboratories were not requested to return official reports in English at this stage. Results were returned to the centre by fax and/or mail (n = 21) or by e-mail (n = 8). Official reports written in English were returned by nine centres, all laboratories belonging to human genetics institutions already involved in other quality control programmes for genetic diseases.

At the beginning of the study, no specific indications were given concerning internal quality control. After termination of the second trial it was suggested that the laboratories should adopt minimal quality control rules, i.e. to use a blank, a female DNA sample and a DNA sample from a male with proven fertility in each PCR reaction. Moreover, the laboratories were advised to amplify a locus on the short arm of the Y chromosome. At the end of the third trial, 18 of 29 laboratories (62%) used a marker specific for the short arm of the Y chromosome. Information about compliance with the other indications was available only from some laboratories: 13 laboratories regularly ran a blank, 12 laboratories ran a female DNA sample and 12 laboratories ran a fertile male DNA sample. The remaining laboratories did not provide similar information, thus making accurate assessment of internal quality control impossible. From the rate of misdiagnosis (see below) it was clear that internal quality control was insufficient in some laboratories. However, a spontaneous reduction of the rate of wrong results was observed from trial to trial, suggesting that the laboratories improved their internal quality control as a result of their participation.

The sequence-tagged site (STS) primers used by the participants in each trial are listed in Table IIIGo. No reason for the choice of primers was indicated. Overall, 90 different primers were employed in trial 1 (nine of which amplified STSs in the AZFa region, 23 in AZFb, 36 in AZFc and 22 other STSs), 77 primers in trial 2 (eight in AZFa, 24 in AZFb, 30 in AZFc and 15 in other regions) and 80 primers in trial 3 (nine in AZFa, 24 in AZFb, 36 in AZFc and 11 in other regions). Figure 1Go shows the number of primers used by the individual laboratories in each trial. The number of STS primers varied from 3 to 29 (median = 8.5) and varied within 14 individual laboratories between the trials. The reason for changing the primer choice derived in part from the results of the first and the second trial, which showed inter-laboratory inconsistencies. A large number of STS primers reported in Table IIIGo are known to amplify sequences belonging to the DAZ gene cluster. These are: sY152, sY232, sY233, sY155, sY147, sY242, sY148, sY149, sY208, sY254, sY255, sY277, sY279, sY283, SGPY (Kostiner et al., 1998Go). The number of PCR markers amplifying DAZ varied between the laboratories from 1 to 9 (not shown).


View this table:
[in this window]
[in a new window]
 
Table III. Number of laboratories analysing the individual loci
 


View larger version (52K):
[in this window]
[in a new window]
 
Figure 1. Number of sequence-tagged site (STS) primers used in AZF diagnostics by the laboratories participating in the external quality control assessment scheme.

 
The results returned by the participants were summarized in tables and distributed to the participants. Figure 2Go show the results obtained by the individual laboratories from the analysis of sample 1b, and Figure 3Go shows the results of sample 3a. The heterogeneity of the primer sets used is evident. Remarkably, none of the primers was common to all participants.



View larger version (76K):
[in this window]
[in a new window]
 
Figure 2. Results of the first trial, sample 1b (proven father). Black box = locus analysed and present; white box = locus analysed and absent; – = not analysed.

 


View larger version (93K):
[in this window]
[in a new window]
 
Figure 3. Results of the third trial, sample 3a (patient with AZFb deletion). Black box = locus analysed and present; white box = locus analysed and absent; – = not analysed.

 
In the first trial all participants correctly diagnosed the deletion in AZFc of sample 1a, while two laboratories diagnosed a deletion incorrectly (i.e. obtained a false positive result) in sample 1b (Figure 2Go). One of these two laboratories (no. 10) was unable to amplify as many as seven STSs in sample 1b, a result clearly at odds with the data obtained by the other participants. The second laboratory had a false positive result in only one STS. These data correspond to a rate of misdiagnosis for trial 1 of 5% (2/40 diagnoses). In trial 2, all laboratories diagnosed a deletion in AZFc correctly. However, one laboratory incorrectly diagnosed an additional microdeletion involving only one and the same STS in AZFa in both DNA samples (not shown) (rate of misdiagnosis: 2/50, 4%). In trial 3, all laboratories correctly diagnosed the deletion in AZFc in sample 3b, while three laboratories completely failed to diagnose the deletion in AZFb in sample 3a (Figure 3Go). This corresponds to a rate of misdiagnosis of 5.2% (three out of 58 diagnoses). Such false negative results were obtained using two STS primers by two laboratories (nos. 7 and 18 in Figure 3Go) and only one STS by one laboratory (no. 6).

Laboratory 4 produced `uncertain' results for two STSs in trial 1 and for four STS in trial 3. Laboratory 15 returned `uncertain' results for one STS and laboratory 26 for two STSs in trial two. The overall ratio of incorrect and dubious results is summarized in Figure 4Go. The percentage of wrong and dubious PCR results decreased progressively, partly because the use of some primers which gave inconsistent inter-laboratory results was abandoned. The use of STS primers sY55, sY75, sY231, sY272 and sY167 was spontaneously abandoned by the laboratories using them after the first trial (Table IIIGo). Among these primers, only sY272 had given inconsistent inter-laboratory results in trial 1. Further inter-laboratory discrepancies in trial 1 were found for the following primers: sY152, sY153, sY236, sY166 and sY147. After trial 2, the use of the following primers was abandoned: Y6HP35PR, sY150, sY138, sY279, sY273 and sY247 (Table IIIGo). Among them, sY150 and sY273 had shown poor inter-laboratory consistency.



View larger version (27K):
[in this window]
[in a new window]
 
Figure 4. Error rate = percentage of wrong or dubious polymerase chain reaction (PCR) results in the three trials. The total number of sequence-tagged site (STS) primers used in each trial is given over the bars.

 
Necessity for standardization and quality control

The first obvious issue resulting from this survey is the extreme heterogeneity of the panels of PCR markers (number and type) used by the various laboratories. This reflects perfectly the heterogeneity of the protocol adopted in the published scientific papers. The criteria for choosing a given protocol are unknown and, in the absence of guidelines, each laboratory arbitrarily chooses and sets up its own method. The heterogeneity of the methods has been useful to identify unreliable STS markers, the use of which has been spontaneously abandoned by many laboratories, but renders the interpretation of the performance of the individual laboratories impossible.

Secondly, the choice of primers is unbalanced towards the AZFc region and, in particular, the DAZ gene. At least 15 AZF candidate genes have been described so far (Lahn and Page, 1997Go; Kostiner et al., 1998Go; Vogt, 1998Go) but their role in spermatogenesis is basically unknown. The large majority of microdeletions in infertile patients occurs in the AZFc regions, where the DAZ gene cluster is located. Moreover, DAZ was the first discrete gene described in this region for which a causal role in infertility was proposed (Rejio et al., 1995). These are probably the two main reasons for the preference of PCR markers in this area for diagnostic purposes. Although, in principle, the choice of STS primers amplifying discrete genes would be preferable, the amplification of anonymous STS loci in diagnostic protocols is fully acceptable, provided that they cover clinically relevant regions of the Y chromosome and give reproducible results. Nevertheless, some form of standardization should be implemented, so that inter-laboratory variability can be better controlled.

The overall diagnostic accuracy, expressed in terms of sensitivity and specificity as calculated from the results of the three trials, is shown in Table IVGo. Although the methods employed are non-homogeneous, the overall sensitivity in detecting deletions in AZFb and AZFc is 97%, while, due to false positive results, the specificity is lower (91%). It is interesting to notice that, although no samples with AZFa deletion were distributed in these three trials, one laboratory misdiagnosed two deletions in AZFa (false positive) in the second trial.


View this table:
[in this window]
[in a new window]
 
Table IV. Current diagnostic accuracy of AZF deletions
 
The present data show that the diagnosis of samples with AZFc deletions has a good sensitivity since no false negative results were reported. On the other hand, microdeletions of the AZFc region can be occasionally diagnosed in normal samples, as suggested by the two cases where false positive results were reported (Table IIIGo). Obviously, the diagnostic specificity of AZFc deletions is independent of the number of primers used. In fact, one of the two false positive diagnoses of deletion in AZFc was generated from a large number of primers. In principle, the use of many primers in the same region prevents neither false negative results nor overestimation of the rate of deletions. It has been reported that when PCR results showing deletions were checked by Southern blotting, the percentage of microdeletions in a group of patients fell from 13.2 to 3.2% (Ma et al., 2000Go). Again, this stresses the necessity for rigorous internal quality control. Extrapolating these results to the clinical arena, it can be concluded that patients with deletions involving the DAZ gene are not expected to escape diagnosis, but deletions of the AZFc area can be incorrectly diagnosed in ~3–5% of men with a normal Y chromosome. This is particularly important in the case of azoo/oligozoospermic candidates for TESE/ICSI, where the genetic counselling and the decision whether to undergo assisted reproduction may depend upon the outcome of the genetic test (Annemiek et al., 1999Go).

The diagnosis of samples with AZFb deletions is even more problematic. In fact, as many as three laboratories failed to make the correct diagnosis. Two laboratories failed by using the three most commonly employed primers, which were not found to present analytical problems to other laboratories. These results suggest poor diagnostic performance of the PCR protocol adopted by those laboratories for AZFb deletions and stress the need for a more stringent control of assay conditions by using appropriate control samples. This underlines the necessity of running negative controls (female DNA, blank) and a positive control (normal male DNA) in each assay in order to detect contamination or amplification of spurious bands promptly (Neumaier et al., 1998Go). Internal controls are particularly important in avoiding false negative results, because if a band of the expected size is found to be present, the PCR reaction is usually not repeated a second time, while samples showing deletions are usually reanalysed. Extrapolating these results to routine clinical use, one might expect a clinical underestimation of deletions in the AZFb regions of ~5%. Unfortunately, no sample with AZFa deletion was available at the time when these trials were organized, so that the diagnostic accuracy of AZFa deletions cannot be calculated.

The overall rate of misdiagnosis of AZFb and AZFc was 5%, a value within the range of misdiagnoses for other genetic tests (Dequeker and Cassiman, 1998Go; Losekoot et al., 1999Go). However, this value is probably an underestimate because other important aspects contributing to the accuracy of the diagnosis, e.g. sample drawing and storage, DNA extraction and, especially, clerical errors in reporting the results, have not been considered so far. As for the analytical procedure, in addition to unsatisfactory internal quality control, the use of STSs with poor inter-laboratory consistency could play a role in misdiagnosis. Moreover, some STS loci analysed by the participants are known to be polymorphic, i.e. are also found to be deleted in the normal population (Pryor et al., 1997Go; Vogt, 1998Go; Kent-First et al., 1999Go), and some primers amplify repetitive sequences on the Y chromosome. Far from contributing to an accurate and meaningful diagnosis, the use of such STS primers should be abandoned. It is expected that the adoption of strict internal quality control measures (Neumaier et al., 1998Go) and participation in an external quality control assessment scheme will reduce and should (eventually) eliminate misdiagnoses.

The EAA/EMQN laboratory guidelines for molecular diagnosis of Y-chromosomal microdeletions

After the completion of this study the participating laboratories were invited to evaluate the results and to contribute to the generation of laboratory guidelines for the molecular diagnosis of Y chromosome microdeletions. Such guidelines (Simoni et al., 1999Go) have been compiled with the valuable contribution of both andrologists and geneticists of several European countries and represent a first step towards the standardization and improvement of the quality of this procedure. The guidelines have been designed keeping in mind that a diagnostic test should be accurate, easy to perform, reproducible and, possibly, inexpensive. It is not to be expected that the proposed protocol will detect all deletions, but no diagnostic test does that in molecular genetics. For instance, the current molecular diagnosis of cystic fibrosis detects `only' 80% of mutations. These guidelines are now officially recognized and promoted by both the EAA and the European Molecular Genetics Quality Network (EMQN) and can be downloaded free from the respective home pages (http://www.blackwell-science.com/uk/society/eaa/ and http://www.emqn.org/). The guidelines are by no means an immutable precept. On the contrary, the rapid expansion of the information in this field demands frequent revisions, in the light of the upcoming knowledge of the complete sequence of the Y chromosome and of the function of its genes. For instance, the recent description of the genomic organization of the DAZ gene cluster revealed four individual genes arranged in two clusters each comprising an inverted pair (Saxena et al., 2000Go). The four genes differ in their length. In the near future it will be possible to design amplification strategies detecting the individual genes to test the hypothesis that some cases of infertility are associated with partial DAZ gene deletions. Obviously, future diagnostic protocols will have to consider these developments and include the amplification of discrete genes whenever it is clinically relevant. For the time being the EAA and the EMQN are jointly organizing an external quality control assessment scheme, based on the distribution of well-characterized DNA samples of good quality and on the evaluation of the results by a panel of expert assessors according to the rules followed by the EMQN for the external quality control schemes of other genetic diseases. It should be pointed out that the persons involved in this effort are doing this on a voluntary basis and have no personal interest in its promotion other than that of providing a service to the scientific community and to the diagnostic laboratories.

Perspectives and conclusions

While the option for an infertile couple of whether to undergo ICSI or TESE/ICSI can remain uninfluenced by the diagnosis of a microdeletion of the Y chromosome (Annamiek et al., 1999; Meschede et al, 2000Go), Y deletion analysis is assuming an important prognostic value about the probability of recovering sperm upon TESE in azoospermic men (Krausz et al., 2000Go). It can be easily foreseen that this diagnostic procedure will be continued in the future. In the meantime the first diagnostic kits have appeared on the market and many laboratories might choose to adopt such commercial methods. Similar to the in-house methods, commercial kits should be carefully validated before routine clinical use. Moreover, each diagnostic laboratory should participate in an external quality control scheme. For many other genetic diseases, in many countries, diagnostic laboratories have to undergo a certification procedure or at least demonstrate that they participate in a programme for proficiency testing. Soon this will probably also be true for the molecular diagnosis of the Y chromosome. The experience in external quality control of other genetic diseases shows that the error rate improves during quality control trials but only less than half of the laboratories make no mistakes (Dequeker and Cassiman, 2000Go). The adoption of well-validated, reliable and reproducible diagnostic protocols and the continuous, qualitative improvement deriving from taking part in external assessment schemes are necessary for diagnosis of the Y chromosome as well.

Acknowledgments

Thanks to Professor E.Nieschlag for his continuous, stimulating support. Evaluation of the results was performed with the help of Dr Renee Reijo, Department of Obstetrics, Gynecology and Reproductive Sciences, University of California, San Francisco, USA and Dr Peter H.Vogt, Institute of Human Genetics, University of Heidelberg, Germany. Thanks to Dr Ken McElreavey, Immunogenetique Humaine, Institut Pasteur, Paris, France, for help in ordering the STS loci according to most recent information available from the internet. The technical assistance of N.Terwort and the language editing of S.Nieschlag and T.G.Cooper, are gratefully acknowledged.

References

Annemiek, W.N., Van Golde, R.J.T., Tuerlings, J.H.A.M. et al. (1999) Reproductive decisions in men with microdeletions of the Y chromosome: the role of genetic counselling. Hum. Reprod., 14, 2166–2169.[Abstract/Free Full Text]

Dequeker, E. and Cassiman, J.J. (1998) Evaluation of CFTR gene mutation testing methods in 136 diagnostic laboratories: report of a large European external quality assessment. Eur. J. Hum. Genet., 6, 165–175.[Medline]

Dequeker, E. and Cassiman, J.J. (2000) Genetic testing and quality control in diagbostic laboratories. Nature Genet., 25, 259–260.[ISI][Medline]

Kent-First, M., Muallem, A., Schultz, J. et al. (1999) Defining regions of the Y-chromosome responsible for male infertility and identification of a fourth AZF region (AZFd) by Y-chromosome microdeletion detection. Mol. Reprod. Dev., 53, 27–41.[ISI][Medline]

Kostiner, D.R., Turek, P.J. and Reijo,R.A. (1998) Male infertility: analysis of the markers and genes on the human Y chromosome. Hum. Reprod., 13, 3032–3038.[Abstract]

Krausz, C. and McElreavey, K. (1999) Y chromosome and male infertility Front. Biosci., 4, E1–E8.

Krausz, C., Quintana-Murci, L. and McElreavey, K. (2000) Prognostic value of the Y deletion analysis. What is the clinical prognostic value of Y chromosome microdeletion analysis? Hum. Reprod., 15, 1431–1434.[Abstract/Free Full Text]

Lahn, B.T. and Page, D.C. (1997) Functional coherence of the human Y chromosome. Science, 278, 675–680.[Abstract/Free Full Text]

Losekoot, M., Bakker, B., Laccone, F. et al. (1999) A European pilot quality assessment scheme for molecular diagnosis of Huntington's disease. Eur. J. Hum. Genet., 7, 217–222.[ISI][Medline]

Ma, K., Mallidis, C. and Bashin, S, (2000) The role of Y chromosome deletions in male infertility. Eur. J. Endocrinol., 142, 418–430.[ISI][Medline]

Maurer, B. and Simoni, M. (2000) Y chromosome microdeletion screening in infertile men. J. Endocrinol Invest., 23, 664–670.[ISI][Medline]

Meschede, D., Lemke, B., Behre, H.M. et al. (2000) Non-reproducible heritable disorders in infertile couples and their first degree relatives. Hum. Reprod., 15, 1609–1612.[Abstract/Free Full Text]

Neumaier, M., Braun, A. and Wagner, C. (1998) Fundamentals of quality assessment of molecular amplification methods in clinical diagnostics. Clin. Chem., 44, 12–36.[Abstract/Free Full Text]

Pryor, J.L., Kent-First, M., Muallem, A. et al. (1997) Microdeletion of the Y chromosome of infertile men. N. Engl J. Med., 336, 534–539.[Abstract/Free Full Text]

Reijo, R., Lee, T.-Y., Salo, P. et al. (1995) Diverse spermatogenic defects in humans caused by Y chromosome deletions encompassing a novel RNA-binding protein gene. Nature Genet., 10, 383–393.[ISI][Medline]

Saxena, R., de Vries, J.W.A., Repping, S. et al. (2000) Four DAZ genes in two clusters in the AZFc region of the human Y chromosome. Genomics, 67, 256–267.[ISI][Medline]

Simoni, M. (1998) News from the European Academy of Andrology (EAA). The EAA international quality control programme for Y-chromosomal microdeletions. Int. J. Androl., 21, 315–316.[ISI][Medline]

Simoni, M., Kamischke, A. and Nieschlag, E. (1998) Current status of the molecular diagnosis of Y-chromosomal microdeletions in the work-up of male infertility. Hum. Reprod., 13, 1764–1768.[Free Full Text]

Simoni, M., Bakker E., Eurlings, M.C.M. et al. (1999) Laboratory guidelines for molecular diagnosis of Y-chromosomal microdeletions. Int. J. Androl., 22, 292–299.[ISI][Medline]

Vogt, P.H. (1998) Human chromosome deletions in Yq11, AZF candidate genes and male infertility:history and update. Mol. Hum. Reprod., 4, 739–744.[Abstract]

Vogt, P.H., Edelmann, A., Kirsch, S. et al. (1996) Human Y chromosome azoospermia factors (AZF) mapped to different subregions in Yq11. Hum. Mol. Genet., 5, 933–943.[Abstract/Free Full Text]