WO1995011300A2 - Azoospermia identification and treatment - Google Patents

Abstract

The present invention provides an AZF gene (seq ID Nos 1 & 2), and polynucleotide probes specifically binding to this gene or to naturally occurring variants thereof, as well as polypeptide probes specifically binding to the AZF polypeptide or to naturally occurring variants thereof. Expression of the gene in a host is also useful for providing means for use in the evaluation of drugs for use in increasing or decreasing expression of the AZF gene or AZF polypeptide.

Description

AZOOSPERMIA IDENTIFICATION AND TREATMENT Background of invention Field of invention

This invention relates to the discovery of cDNA sequences corresponding to a newly identified gene from the Y chromosome whose deletion is linked to azoospermia, and the application of this discovery to inter alia 1) the development and use of polynucleotide sequences derived from the azoospermia factor (AZF) gene(s) for the measurement (qualitative and quantitative) and/or regulation of gene expression and 2) the development and use of polypeptides and other materials for use in monitoring and/or regulating expression of the gene (including translation thereof) . Description of the prior art

About 1 in 10 couples in the UK are childless through infertility 1. Of all men who seek help at infertility clinics, about 20 per cent are diagnosed to have oligo- or azoospermia of unknown aetiology 2.

On the one hand the processing of such cases can be very time consuming and expensive due to various different possible causes that have to be investigated before the physician can admit defeat. On the other hand such an inconclusive diagnosis is particularly unsatisfactory to the couples concerned in view of inter alia the considerable emotional and psychological trauma that can be associated with this problem.

It is likely that a gene controlling spermatogenesis, the "azoospermia factor" or "AZF", mapping within Interval 63 of the human Y chromosome long arm4 (band Yqll.23) , is defective in a proportion of male infertility cases. Phenotype-karyotype correlations indicate that loss of the most distal segment of Yq including all fluorescent heterochromatin (Yqh) , is associated with severe spermatogenic impairment, the testis showing absent or severely reduced germ cell development in cases of Yq deletion or structural rearrangements,6. Similar phenotypes in oligo- or azoosper ic men with a cytologically normal Y chromosome might indicate microdeletion or mutation in AZF, and screening of such individuals in our laboratory, has already revealed selected cases to have microdeletions in Interval 6 where disruption or loss of the AZF gene could have occurred. Amongst these, one individual showed deletion of two probes in proximal Sub-Interval I of Interval 6; three others each showed deletion of the same eleven probes, mapping across distal Sub-Intervals XII, XIII and XIV7. The proximal and distal microdeletions in these four patients did not, however, overlap perhaps indicating, either that the AZF locus is very large, or, that a family of Y chromosome long arm genes spanning interval 6, is involved in the process of spermatogenesis.

Thus to date the AZF gene has not been identified and as a consequence it has not been possible to provide a basis for effective diagnostic means to provide positive identification of this type of infertility, let alone for possible treatments therefor. There is accordingly a need for reliable and accurate diagnostic means and methods for oligo- or azoospermia of genetic origin, and for means and methods for the treatment of such cases.

There is also a need for improved and/or alternative means and methods for population control for various purposes, including a need for a safe and reliable human male contraceptive and a need for pest control systems based on fertility control e.g. for rodents which present major health hazards and loss of food resources in many countries.

It is an object of the present invention to avoid or minimise one or more of the above problems or disadvantages.

Summary of Invention We have recently isolated cos ids and Yeast Artificial Chromosomes (YACs) derived from the region of the Y chromosome that corresponds to the distal interval deleted in the abovementioned three patients and, using the cosmid clones and restriction enzymes, we have searched for and identified a potential CpG island, a strong indicator for the presence of a gene. From a testis cDNA library, a 2kb cDNA clone has been identified which maps only to the Yq distal euchromatin by in situ hybridisation. We believe, for the following reasons, that this constitutes, an excellent candidate gene for the AZF locus:- 1. The cDNA contains an open reading frame which would putatively code for a protein. Three distinct domains are present, one of which has homology to previously known RNA binding proteins.

2 . The gene maps to the distal deletion interval (sub- intervals XII-XIV) of the Y chromosome, a region of approximately 200kb associated with azoospermia or severe oligospermia in three patients. 3 . At least part of the gene is contained within a microdeletion of an azoospermic and an oligospermic patient. None of the probes used previously to map Interval 6 on our deletion map7 are deleted in these patients. 4. Multiple copies of the gene are present, consistent with the conclusions previously drawn from mapping data.

Thus a cDNA sequence has now been discovered for an AZF gene, which sequence is presented as the basis of this invention. The sequence is presented in interleaved format (see Fig. 2) .

In more detail, the present invention provides: an AZF gene; and preferably an AZF gene having the nucleotide sequence SEQ ID No:l or 2 disclosed herewith and includes AZF genes which have substantial nucleotide sequence homology with the nucleotide sequence SEQ ID No:l and/or 2 disclosed herein; and in particular the genes of the present invention in a form substantially free from other genes. It will be understood that the genes of the present invention may include nucleic acid sequences (upstream and/or downstream of the protein coding sequence) which are utilized in the expression of the gene such as promoter, operator, and terminator sequences as well as other sequences which do not inhibit its expression. Thus the expression "gene" includes DNA (including cDNA) and/or RNA sequences as well as plasmid or viral "genes" containing the receptor gene and expression vectors for the gene especially cosmid and yeast artificial chromosome (YAC) types.

In one aspect therefore the present invention provides new methods and means based upon the newly discovered AZF nucleotide sequence for use in the clinical diagnosis and therapeutic management of male infertility and abnormalities thereof.

In general the AZF gene sequences are also useful for the design of oligonucleotide probes capable of specifically hybridising with the AZF genes of the present invention, and for the synthesis of polypeptides which may be used in immunoassays. Such oligonucleotide probes may be used in in vitro assays for determining qualitativly and/or quantitatively the presence of AZF DNA. A quentitative assay may be used to determine the number of copies of an AZF gene present in an individual. Alternatively antisense oligonuclestide probes may be designed to hybridize to AZF DNA in vivo. Such hydridisation may be designed to prevent AZF gene expression. This antisense technique may be developed for use as a contreceptive technique whereby the lack of AZF polypeptide may result in the loss of sperm production. In addition parts of the cDNA sequence may be used to design and/or provide oligonucleotide probes for use in identifying human and other mammalian AZF genes. Both oligonucleotide probes and the polypeptides may be useful for the diagnosis of AZF gene abnormalities. Polypeptides encoded within the cDNA sequences may also be used to raise antibodies against selected regions of proteins expressed by the AZF genes, and for the purification of antibodies directed against such regions. These antibodies may be useful in immunoassays for detecting normal or abnormal (including absent or deleted AZF) in individuals. The proteins expressed by the AZF genes are believed to find RNA to a greater or lesser extent and may accordingly be referred to herein as RNA binding proteins (RNP) for convenience.

Thus the present invention further includes a method of producing AZF polypeptides (including polypeptides corresponding to the full length of the protein expressed by the AZF gene as well as to a lesser portion thereof) which method includes the step of expressing the AZF genes of the present invention in a host, as well as AZF polypeptides produced by such a method. Various suitable hosts are known in the art though eukaryotic hosts are generally preferred, e.g. Xenopus oocytes and COS-7 cells. Fungi e.g. yeast may also be used. Prokaryotic hosts that may be used include E. coli. and B. Subtilis.

Once a suitable host expressing the AZF gene as well as the AZF polypeptide is obtained in vitro, it may be possible to assay what drugs or chemicals have an effect on such expression. This assay may be directed at the level of gene expression or alternatively at polypeptide expression.

The present invention also includes products and processes utilizing, directly or indirectly, human AZF polypeptides obtained in this way. One preferred method of restriction enzyme analysis of male genes in this invention depends on Restriction Fragment Length Polymorphisms (RFLPs) . A sample is taken from any suitable tissue such as blood. DNA is extracted from the cells in any conventional way. It is then digested with an appropriate restriction enzyme e.g. one which cuts in CG- rich sequence. The fragments of different length are separated by gel electrophoresis in any conventional way. A restriction fragment pattern is generated. Probing of the fragments will generally be necessary for clearer detection of the pattern and of the fragment(s) of interest, e.g. a fragment which extends from restriction sites "n" to "n + 2" (where "n" denotes any arbitrary number) , seemingly not being restricted at the normal site "n + 1" lying between "n" and "n + 2" due to an abnormality at the "normal" restriction site. Alternatively, a polymorphism might generate restriction enzyme sites and thereby give rise to a plurality of shorter fragments where the normal DNA provides longer ones. Whether it is appropriate to probe for long or short fragments will therefore depend on the circumstances of the polymorphism. In some instances, the probe will extend outside the region designated.

Direct hybridisation of polynucleotide probes to the genomic region may also be used. Thus suitable biopsy or other samples can be subjected to cloning techniques, to isolate a library of genomic DNA. Clones containing the gene can be amplified by Polymerase Chain Reaction (PCR) and probes complementary to the said region used directly on PCR products, which need not be first restricted by enzymes. Alternatively sequencing of the amplified DNA can be carried out to determine any DNA alteration.

It will be appreciated, therefore, that the cDNA of the invention also has uses in assays which are not of the RFLP type. Accordingly, the polynucleotides per se are part of this invention, as 'intermediates' suitable (when labelled) for use as probes. Both double-stranded and single- stranded polynucleotides are included as well as sense and anti-sense forms. Suitable polynucleotide probes may be oligonucleotides of from 10 to 50, preferably from 16 to 30 nucleotides in length. Shorter probes are unlikely to be sufficiently specific for the sequence of interest. Longer polynucleotide probes of from 100 to 500 nucleotides or more may also be used. The probe will usually be of DNA or RNA and labelled in any suitable manner e.g. by labelling with an enzyme, radioisotope, fluorescent, luminescent, or chemiluminescent labels or biotinylation.

The fragments are probed under any appropriate conventional hybridisation conditions, the fragments being conveniently first transferred to a filter. The complexes thus formed are detected by autoradiography or other detection means appropriate to the particular kind of label used.

Abnormalities in the polynucleotide sequence of restriction fragments of the AZF gene which are as small as single- point mutations can also be detected by means of Temperature Gradient Gel Electrophoresis in which a temperature gradient is superimposed, parallel to or transversely of, the electrical field in gel electrophoresis. The method is based on the fact that the temperature of denaturation of double stranded (ds) DNA is altered by changes in polynucleotide sequence. Furthermore, partial denaturation of a DNA duplex causes a change in electrophoretic mobility. Further details of this technique are described in the literature by Reisner et al, 1989, Birmse et al, 1990, and Wartell et al.

An alternative method to detect a mutation or mutations in an AZF gene involves the use of hydroxylamine osmium tetroxide. DNA from the AZF gene in question is hybridised to a "normal" AZF gene. If there are any base changes along the length of the DNA in question, then hydridisation will not occur to the "normal" AZF DNA at these points. Reacution with hydroxlamine osmium tetroxide cleaves the DNA at these unhybridised sites. Separation by conventional gel electrophoresis and detection of the DNA then identifies any such alterations by the fragment patterns obtained. Further details are described in the literature (Condie et al, 1993) .

The AZF cDNA sequences that have been cloned and sequenced in the present invention are shown in Fig. 2 along with the putative amino acid sequence translations thereof.

It will be appreciated that abnormality in human AZF RNP and/or its expression may be "assayed" in a number of ways. Thus the DNA encoding the AZF gene may itself be assayed for the presence or absence of abnormalities or the AZF RNP which is normally expressed by the AZF gene may be assayed for such purposes, to determine whether it is actually expressed at all, and if it is expressed how and to what extent.

The former case generally involves the use of labelled polynucleotide probes to hybridise with DNA within the AZF for the purposes of indicating the presence or absence of particular polynucleotide sequences. In the latter case antibody probes are used to form antigen-antibody complexes with regions of the expressed AZF RNP polypeptide for the purposes of indicating the presence or absence of particular polypeptide sequences.

It will be understood that once more or less common or typical abnormalities have been specifically identified e.g. by initially probing with 'normal' polynucleotide and then sequencing, polynucleotide probes can be synthesized or otherwise produced with sequences corresponding to or complementary to the "abnormal" sequences, to allow screening of tissue samples for specific AZF gene abnormalities.

In the case of assays of the polypeptide itself, suitable stretches of amino acids based on the cDNA sequence information provided by the present information, may be synthesised on a peptide synthesiser. These peptides would generally have a length of from 10 to 50, preferably 15 to 30, amino acids but could be even shorter or longer. Alternatively the complete AZF RNP polypeptide or fragements thereof may be expressed in a suitable eukaryotic or prokaryotic host such as E. Coli using an appropriate vector. Polyclonal antibodies to these peptides may be produced by conventional approaches such as the immunisation of host animals (rabbit, goat etc.) with said peptides, optionally conjugated to a protein carrier such as thyroglobulin, and recovery of the desired antibody material therefrom. Monoclonal antibodies could also be raised using conventional monoclonal antibody production procedures.

It will also be understood that "assay" may be qualitative and or quantitative (e.g. where detection of under or over- "expression" of the AZF gene or the AZF RNP is required) .

For the avoidance of doubt it is confirmed that the term expression is used herein in accordance with contemporary practice in the art to indicate inter alia inclusion within the genomic DNA or RNA of the AZF gene; and/or production of RNP encoded by the AZF gene.

Further preferred features and advantages of the invention will appear from the following detailed description given by way of example.

Detailed Description Isolation and Sequencing of AZF Gene Isolation and Restriction Analysis of Cosmids

Cosmid DNAs from a Y-Chromosome specific cosmid library (Taylor et al., 1991) were obtained as gridded arrays on filters and hybridised to 32P-dCTP labelled probes from the KLARD deletion (Ma et al., 1992). Individual clones were grown up in 200ml cultures, and DNA prepared by standard methods (Sambrook et al., 1991). Cosmid DNA (3 μg) was then digested with a variety of CpG-specific restriction enzymes, and electophoresed in 0.8% agarose gels to determine the numbers and locations of specific restriction sites.

Using probes deleted in patient KLARD (and other patients with similar deletions, NIKEI and KUPAU; Figure 1) we have isolated a series of cosmids from a Y-chromosome cosmid library (Taylor et al., 1991). As the first stage of characterisation of these cosmids they were scanned for the possible presence of a CpG island using restriction enzymes with one or more CpG oligonucleotides in their recognition site (Bickmore and Bird, 1987) . One cosmid A5F, was found to have clustered sites for the restriction enzymes Asc I (1 site) , Eag I (4 sites) , and Ksp I (l site) and was considered potentially to contain a CpG island and hence perhaps a gene.

Isolation and Sequencing of cDNAs

An adult human testis library in λgtll (Clontech) was plated at a density of 2500pfu/cm2, and probed with the A5F cosmid DNA by standard procedures (Sambrook et al. , 1991). Probe was labelled by random priming with 32P-dCTP, and preannealed with human Cotl DNA (Gibco BRL) to remove repeated sequences.

Areas corresponding to positive signals obtained in the primary screen were plaque-purified in subsequent rounds of screening. EcoRI inserts from plaque-purified phage were subcloned into pBluescriptSKII+ (Stratagene) and sequenced using dideoxy chemistry (Sanger et al., 1977) and Sequenase (USB) . A combination of deletion cloning and internal oligonucleotide strategies was used to obtain the sequences of both cDNAs on both strands. Sequences were analysed using the computer resources of the UK Human Genome Mapping Project Resource Centre, Harrow. DNA and RNA analysis Preparation of genomic DNAs, and Southern Blot procedures were essentially as previously described (Ma et al., 1992). Northern blots of various human mRNAs (Clontech, 2 μg per track) were hybridised to a 32P-dCTP labelled 3 'end probe from pMK5 (nt 1080-nt 1878) . Hybridisation in 2X SSPE was carried out at 65°C for 16 hours, according to manufacturers instructions. Filters were washed at moderate stringency (IX SSC, 65°C for 1 hour and analysed with the Molecular Dynamics ImageQuant phosphoimaging system following overnight exposure. PCR analysis of cDNAs

Replica filter lifts from a testis cDNA library in lλgtll plated at 1250 pfu/cm2 were hybridised with the insert from pMK5 by random priming using 32p dCTP. Hybridisation was in 0.5M NaHP04 7% sodium dodecyl sulphate (SDS)at 65°C and filters were washed at 65°C in 0.1 x SSC 0.1% SDS. After overnight exposure agar circles 5mm in diameter were punched from the agar into SM buffer. Two ml aliquots were used for PCR amplification using primers e593 and e355. PCR reactions were carried our in TAPS buffer with 2u polymerase and an annealing temperature of 60°C. For sequencing, one primer was pre-biotinylated and the PCR product captured onto 20cl of streptavidin-linked Dynabeads. Sequencing was carried out after alkali denaturation and washing with Sequenase and S35-dATP according to the manufacturers instructions. In Vitro Translation

MK5 and MK29 cDNAs, and two derivatives MK5-delXba, MK29- delXba, were translated in vitro using the TNT kit (Promega) following manufacturers recommendations. Following coupled transcription/translation for 90 minutes with T7polymerase/35S-Methionine (.Amersham) , samples were electrophoresed in a 12% SDS-PAGE gel system. After fixing and drying, gels were autoradiographed and images analysed using the Molecular Dynamics ImageQuant system. In Situ Hybridisation Normal male peripheral lymphocytes were incubated for 48hrs in RPMI 1640 with 10%FCS and 1%PHA followed by synchronisation with methotrexate (0.045cg/ml) for 16 hours before release with BudR (30cg/ml) for 4.5. hours. The cells were finally arrested with colcemid (0.leg/ml) for 10 mins before harvest. Probe labelling and in situ hybridisation was as described (Farr et al., 1991). The EcoRI insert from pMK5 was biotin labelled, hybridised at a concentration of 30ng/ml and visualised with AvidinTexas Red (4gg/ml) . Chromosomes were counterstained with DAPI (lOcg/ml) and banded with anti-BudR FITC conjugate

(lOcg/ml) . Images were collected on a Zeiss Axioscop with a Photometries CCD camera.

Multiplex PCR analysis of Genomic DNA

PCR was carried out on 200 ng of genomic DNAs in 50cl reaction volumes in accordance with manufacturers instructions (Perkin Elmer Cetus) . Four primers were used in each reaction, to co-amplify a YRMM sequence and a control Y-chromosome (Sry) or autosomal sequence (CENP-C; Saitoh et al., 1992). Primer sequences were E355 (GGAAAAGGAATTGTTTTCAAAG, YRRM-non-specific) , F19

(ATGCACTTCAGAGTCGG, YRRMl-specific) , F20 (GAGATGCACTTCAG AGGG, YRRM2 -specific) , E76 (ACACAGAGCAAGGCCAGAAT, CENP-C) , E77 (CTTCATGGGCCTGAACTGAT, CENP-C) . All primers worked well in multiplex reactions, and gave reproducible results. Autosomal controls were chosen since the Sry product (778 bp) obscured the YRRM product (approx. 800bp) on gels.

Total A5F DNA, competed with total human DNA to reduce repeat sequence concentration, was used as a probe to an adult human testis cDNA library in λgtll. In total 2 x 106 reco binants were screened and 33 primary positives detected. Following further rounds of screening, two 1.9Kb cDNA clones, MK5 and MK29, were selected for sub cloning and sequencing. Rehybridisation of the primary library screening filter with the MK5 insert, identified 45 positives, including all those detected by the A5F probe. This suggests that no other testis-specific expressed sequences occur in the A5F cosmid.

Inserts from the two clones MK5 and MK29 were sequenced by a combination of deletion and primer walking strategies. The MK5 insert is 1878 bp long and the complete nucleotide and predicted peptide sequence is shown in Fig.l. There is an AUG codon (at 118 bp) within a suitable context for the initiation of translation (Kozak 1991) . There then follows an open reading frame (ORF) of 1491 bp. Downstream sequence contains a poly [A]+ tail and a polyadenylation signal. Conceptual translation of this ORF predicts a novel peptide of 496 residues. MK29 is 1874 bp long and has an ORF of 1260 bp and also contains a downstream poly [A]+ tail and a polyadenylation signal.

The MK29 cDNA displays seven nucleotide substitutions and a 5 bp deletion, relative to MK5 and the nucleotide and amino acid sequences are shown in Fig. 2. The 5 bp deletion causes a translation fra eshift which truncates the MK29 ORF, giving a predicted peptide of 419 amino acids. In addition, 3 of the 7 substitutions alter the predicted amino acid residue for the MK29 sequence. Neither DNA sequence detects close homologues in the GenBank or EMBL nucleic acid databases.

Northern blot hybridisation of mRNA from adult testis using a MK5 probe reveals a single transcript of about 2kb with no detectable signal in male brain, heart, kidney, liver lung, prostate gland, or skeletal muscle. Thus both MK5 and MK29 cDNAs are likely to represent near full-length transcripts from their cognate genes.

BLAST (Altschul et al., 1990) searching of the GenBank, NBRF and Swissprot sequence databases with the conceptual MK5 and MK29 translations reveals a clear 5' homology to a superfamily of RNA-binding proteins which contain a 90 amino acid RNA recognition motif (RRM) , often in a tandemly repeated array (Dreyfuss et al., 1988; Kenan et al., 1991). MK5 and MK29 are unusual in that only one of these motifs is present. The MK5 and MK29 RRMs are most closely related to RRMs from the polyadenylate-binding protein (PABP: Adam et al., 1986; Sachs et al., 1986) family, which possess four copies of RRM. The highest similarity being found to the third domain. Lower similarities to many other RRM- containing proteins are detected, some with high local regions of similarity. Because of the RRM homology, and because of their Y chromosome location (see below) , we have called these genes YRRM1 (MK5) and YRRM2 (MK29) . Adjacent to the RRM domain is a 139-residue segment with no overt homology to previously published sequences. Following this are four tandem repeats of a 37-residue peptide, all of which show high conservation at both the nucleotide and predicted amino acid level. This repeated domain is notable for its high Arginine (20%), Serine (15%), Tyrosine (14%) , and Glycine (9%) content, with no aliphatic Leucine, Isoleucine, Methionine and Valine residues. The repeat shows no pronounced homology to other database sequences, but because of its bias towards the amino acids described above and the occurence of the Ser-Arg Gly-Tyr (SRGY) tetrapeptide, or a closely related sequence, twice in each repeat, we propose to call this domain the SRGY box. Moderate similarities (25-30% identity) to repetitive sequence stretches of several proteins are detected, most notably the RNA-binding translation initiation factor elF- 4B (Milburn et al., 1990) which contains an extended repetitive hydrophilic peptide adjacent to a consensus RRM domain. This similarity is due to coincident amino acid bias rather than specific conservation of the SRGY motif. This is also true of a second and possibly more siginificant sequence similarity to the mammalian proteins ASF/SF2 (Ge et al. , 1991; Krainer et al., 1991) and U1-70K (Query et al., 1989), and the Drosophila tra (Boggs et al. , 1987) and tra-2 (Amrein et al., 1988) splicing factors. In these proteins, three of which also contain a single consensus RRM domain, an Arginine-Serine rich (RS) region is considered likely to mediate protein-protein interactions following RNA binding. The SRGY repeat appears to represent a more complex motif, but may merely indicate an alternative strategy for evolution of an RS type domain. The final 122 residues of YRRMl exhibit no obvious database homologies, although a recurring Tyrosine residue (6 to 7 residue spacing) is apparent in parts of the tail region, a featuer reminiscent of several hnRNP proteins (Matunis et al., 1992;). The YRRM2 deletion occurs shortly after the final SRGY box, yielding a carboxy1 tail of 45 residues, with no Tyrosine residues.

Using oligonucleotide primers derived from the MK5 sequence we have attempted to get an estimate of transcript variability. Primary positive areas from high density (Lambda) λgt 11 cDNA library plates were picked and used as a DNA source for PCR with primers which span the SRGY region. Four size classes of PCR product were produced with one member of class a, 6 members of class b, 13 members of class c, and one member of class d.

Sequencing of PCR products confirmed that two (class a) , three (class b) or four (class c or d) SRGY repeats may occur in a cDNA, strongly suggesting an alternative splicing mechanism. The most 5' pair of SRGY boxes were always present in the PCR products, suggesting that the 3' pair may be subject to omission by splicing. PCR analysis of pMK5 confirm it as a class c member.

An additional non-SRGY sequence is also detected in the class d isolate. This 450-500 bp of additional sequence occurs after the SRGY repeats and appears likely to represent the unspliced intron between the final SRGY box and the final coding exon downstream, since we have evidence that this intron is about 500 bp in length. In addition, PCR sequence data suggests that at least one other (closely related) YRRM sequence is expressed in testis. Thus multiple YRRM products may arise by differential processing of individual genes and by transcription of different genes.

Partial sequencing of cosmid A5F was carried out in order to determine the intron/exon boundaries of the human AZF gene. This was achieved by comparing the sequence obtained from AZF with the cDNA sequence of MK5 (Figure 3) . The results show that the AZF gene is made up of 12 exons stretching over approxiately 14.5kb of DNA. Using the sequence obtained from A5F it was possible to design oligonucleotide primers from intron 1 and intron 2. These primers were designed such that they could be used in a polymerase chain reaction to amplify from genomic DNA, exon 2 of the AZF gene from any indicidual. This indeed was carried out on a number of individuals and the PCR products obtained were subjected to single strand conformational polymorphism (SSCP) analysis. This revealed differences between the PCR products. Various PCR products have been sequenced and four different classes of exon 2 have been identifed (Figure 4) of which class A appears to be normal, Class B a significant mutation, and classes C and D other mutations.

It was also of interest to attempt to clone the mouse AZF gene using the human A5F cDNA. Clone MK5 was radio - labelled by random priming and this nucleic and probe was used to probe a A2001 genomic library constructed from the DNA of a male C129 mouse. Positive clones were isolated, labelled by nick translation with biotinylated nucleotide triphosphates and mupped by in situ hybridisation to mouse chromosones. One clone mapping to

Yp, 3,2, was resriction mapped by single and double enzyme digests and a fragment hybridising to pMK5 subcloned into a plasmid vector, pBluescript. Sequence analysis of this fragment revealed homology to the sequence of pMK5 (Figure 5) . A PCR primer with the sequence GCAGGGCGTCGGAAAGTAAGG was synthesised on the basis of this genomic sequence and used in conjunction with a further primer based on a vector sequence AGCGGATAACAATTTCACACAGGA, in PRC reactions on mouse cDNA clones which hybridised with pMK5. One positive clone was purified and its insert sequenced to give the cDNA sequence for the mouse YRRM transcript (Figure 6) .

Figure Legends

Figure 1. Patients and probes defining AZF deletions on Y- Chromosome Interval 6 on Yq shown subdivided into subintervals I to

XV. Patients KLARD, NIKEI, KUPAU (all deleted XII to XIV), and JOLAR (deleted I) shown to the right as heavy lines. The positions of YAC 12S, and cosmid A5F (defined by probes 1/2/B and GMGY1, and Y367/B respectively) are indicated by heavy lines. The ends of YAC 12S (225 kb) have not been defined by more proximal or distal probes and are shown as dashed lines. The position of cosmid A5F relative to 12S, and within the KLARD/NIKEI/KUPAU deletion is not known. Selected probes, with their approximate location relative to deletions and clones, are shown on the right. All probes except 1/2/B (O'Reilly et al., 1992) have been described previously (Ma et al., 1992) the contents of these papers including in particular the restriction enzy mes and fragment sizes defining the probes, being incorporated herein by reference thereto.

Figure 2. DNA and predicted protein sequence of MK5 and

MK29 cDNAs

The complete nucleotide (bp) sequence of MK5 cDNA is shown numbered at the left, with the predicted protein translation below, numbered at the right. MK29 substitutions (7) relative to MK5 are shown above the MK5 bp sequence; where these cause amino acid (aa) substitutions (3) , the variant aa is shown below the MK5 protein sequence. The 5 bp deleted in MK29 relative to MK5 (bp 1304-1308) are indicated by dashed overline, and the alternative MK29 aa sequence shown in full from this point, numbered (in brackets) at right. The location of the RRM is indicated by square brackets, and a bar at right, with the consensus RNP-1 (aa 48-55) and RNP-2 (aa 10-15) underlined. The SRGY box domain is indicated by a bar at right, and the boundaries of each repeat shown as stippled boxes. The periodic Tyrosine residues in the COOH region of the MK5-encoded aa sequence are circled.

Figure 3. Intron-Exon Boundaries from human A5F

The intron/exon boundaries of a human YRRM gene were determined by comparison of the sequences found in cosmid A5F and the cDNA clone MK5 (both described above) . In the sequence shown in Figure 3 the uppercase type corresponds to MK5 sequence and the lowercase to genomic sequence. Approximate intron sizes are given in brackets.

Figure 4. Gene sequences from Exon 2

Using the information of Figure 3 we have designed primers using sequence information from the neighbouring introns to amplify exon 2 from genomic DNA using appropriate PCR conditions.

These primers are 5'TTTCAGCACAGTATGTAAT and

5'ACGGTTTTAAGAGTTA used with a 50 degree anealing temperature. Radioactive dCTP is included in the reaction to label the products. One part of the reaction mix is mixed with 5 parts denaturing dye (95% forma ide 5mM NaOH 0.1% bromophenol blue and 0.1% xylene cyanol) heated to 95 degrees C for three minutes and on ice for two minutes. Electrophoresis is on an 6% acrylamide 2% bis acrylamide gel in standard tris/borate/EDTA buffer at 10 watts for five hours followed by drying and autoradiography.

We have subjected these exon 2 PCR products to single strand conformational polymorphism analysis and find that the products are separated into multiple bands. Re- amplification and sequencing of these bands gives four different sequence classes as shown in Figure 4.

This SSCP analysis when used on patients reveals differences from the normal patterns in a number of cases. The aim is to use these sequences, primers and methodology to discriminate between YRRM genes and to detect minor changes in patients. Various other techniques known in the art may also be used for detecting minor changes in nucleic acid sequences including the HOT (hydroxylamine osmium tetoxide) technique as described hereinabove (A.Condie et al, 1993) .

Figure 5. Partial mouse gene DNA sequence

Figure 5 shows part of the sequence of mouse gene homology to human YRRM gene obtained from a mouse genomic phage clone M3.2 which is isolated by screening a mouse genomic library with MK5 (YRRM 1). M3.2, cloned in Lambda 2001, has been mapped to the short arm of mouse Y chromosome by in situ hybridisation. When translated, the sequence contains an RNP-1 protein box and further homology to human YRRM sequence. REFERENCES

1 HMSO (1960) . Report of the Departmental Committee on Human Artificial Intelligence C nd 1105.

2 Chandley AC, Ed ond P, Christie S, Gowans L, Fletcher J, Frackiewicz A, Newton M (1975) . Ann Hum Genet. Lond. 39: 231-254.

3 Vergnaud G, Page DC, Simmler MC, Brown L, Rouyer F,

Noel B, Botstein D, de la Chapelle A, Weissenbach J (1986) . A deletion map of the human Y chromosome based on DNA hybridisation. Am J Hum Genet 38: 109-124.

4 Andersson M, Page DC, Pettay D, Subrt I, Turleau C, de Grouchy J, de la Chapelle A (1988) . Y; autosome translocations and mosaicism in the aetiology of 45, maleness: assignment of fertility factor to distal Yqll. Hum Genet 79: 2-7.

5 Tiepolo L, Zuffardi 0 (1976) . Localization of factors controlling spermatogenesis in the nonfluorescent portion of the human Y chromosome long arm. Hum Genet 34: 119- 124.

6 Chandley AC, Gosden JR, Hargreave TB, Spowart G, Speed RM, McBeath S (1989) . Deleted Yq in the sterile son of a man with a satellited Y chromosome (Yqs) . J Med Genet 26: 145-153.

7 Ma K, Sharkey A, Kirsch S, Vogt P, Keil R, Hargreave TB, McBeath S, Chandley AC (1992). Towards the molecular localisation of the AZF locus: mapping of microdeletions in azoospermic men within 14 subintervals of interval 6 of the human Y chromosome. Hum Molec Genet 1: 29-33.

8 Sutcliffe MJ, Burgoyne PS (1989). Analysis in the testes of H-Y negative XO Sxrb mice suggests that the spermatogenesis gene (Spy) acts during the differentiation of the A spermatogonia. Development 107: 373-380.

9 Arnemann J, Jakubiczka S, Thuring S, Schmidtke J (1991) . Cloning and sequence analysis of a human Y- chromosome- derived, testicular cDNA, TSPY. Genomics 11: 108-114.

Taylor, K. , Ulinowski, Z., and Wolfe, J. (1991). Contig

assembly of cosmid clones from the Y chromosome. Cytog. Cell Genet. 58, 2096.

Bickmore, W.A. , and Bird, A.P. (1992). Use of restriction endonucleases to detect and isolate genes from mammalian cells. Meth. Anzymol. 216, 224-244.

Kozak, M. (1991) . Structural features in eukaryotic mRNAs that modulate the initiation of translation. J. Biol. Chem. 266, 19867-19870. Altschul, S.F., Gish. W. , Miller, W. , Myers, E.W. , and Lip an. D.J. (1990) . Basic local alignment search tool. J.Mol.Biol. 215, 403-410.

Dreyfuss, G. , Swanson, M.S., and Pinol-Roma, S. (1988). Heterogeneous nuclear ribonucleoprotein particles and the pathway of mRNA formation. Trends Biochem. 13, 86-91.

Kenan, D.J., Query, C.C., and Keene, J.D. (1991). RNA recognition: towards identifying determinants of specificity. Trends Biochem. 16, 214-220.

Adam, S.A., Nakagawa, T.Y., Swanson, M.S., Woodruff, T., and Dreyfuss, G. (1986) . mRNA polyadentlate-binding protein: gene isolation and sequencing and indentification of a ribonucleoprotein consensus sequence. Mol. Cell. Biol. 6, 2932-2943.

Sachs, A.B., Bond, M.W. , and Kormberg, R.D. (1986). A single gene from yeast for both nuclear and cytoplasmic polyadenylate-binding proteins: domain structure and expression. Cell 45, 827-835.

Milburn, S.C., Herchey, J.W.B., Davies M.V. , Kelleher,

K. , and Kaufmann, R.J. (1990). Cloning and expression of eukaryotic initiation factor 4B cDNA: sequence determination identifies a common RNA recognition motif. EMBO J. 9, 2783-2790.

Ge, H., Zuo, P., and Maley, J.L. (1991). Primary structure of the human plicing factor ASF reveals similarities with Drosophila regulators. Cell 66, 373-382.

Krainer, A.R. , Mayeda, A., Kozak, D. , and Binns, G. (1991).

Functional expression of cloned human splicing factor SF2: homology to RNA-binding proteins, U1-70K, and

SUBSTITUTESHEET(RULE2) Dropophila splicing regulators. Cell 66, 383-394.

Query, C.C., Bentley, R.C., and Keene, J.D. (1989). A common RNA recognition motif identified within a defined Ul RNA bindng domain of the 7OK Ul snRNP protein. Cell 57, 89-101.

Boggs, R.T., Gregor, P., Idriss, S., Belote, J.M. and McKeown, M. (1987) . Regulation of sexual differentiation in D. melanogaster via alternative splicing of RNA from the transformer gener. Cell 50, 739-747.

Amrein, H. , Gorman, M. , and Nothinger, R. (1988). The sex-determining gene tra-2 of Drosophila encodes a putative RNA-binding domain. Cell 55, 1025-103.

Matunis, E.L. , Matunis M.J. , and Dreyfuss, G. (1992). Characterisation of the major hnRNP proteins from Drosophila melanogaster. J. Cell Biol. 116, 257-269.

Reisner, D. , Steger, G. , Simmat R. et al, (1989). Electrophoresis 10 (5-6) :377-389.

Birmse, A., Sattler, A., Maurer, K.H. & Reisner D. (1990). Electrophoresis 11, 795-801.

O'Reilly, A.J. Affara, N.A. Simpson, E. , Chandler, P., Goulmy E. , and Ferguson-Smith, M.A. (1992) . A molecular deletion map of the Y-chromosome long arm defining X and autosomal homologous regions and the localisation of the HYA locus to the proximal region of the Yq euchromatin. Hum. Mol Genet. 1.379-385.

Wartell et al

A. Condie et al (1993)

Detection of point mutations in the P53 Gene: Compararison of SSCP, CDGE, and HOT Techniques. Human Mutation 2 58-66.

Claims

1.Nucleic acid sequence encoding human AZF gene.

2.Nucleic acid sequence according to claim 1 characterised in that said sequence encodes a polypeptide having an amino acid sequence shown in SEQ ID NO:l or SEQ ID NO:2 or derivatives of said polypeptides.

3.Nucleic acid sequence according to claim 2 characterised in that said sequence corresponds with the deoxyribonucleic acid sequence shown in SEQ ID NO:l or SEQ ID NO:2 or derivatives of said deoxyribonucleic acid.

4. An isolated nucleic acid sequence according to any one of claims 1 to 3.

5. Recombinant nucleic acid molecule comprising a nucleic acid sequence according to any one of claims 1 to 4, operably linked to an expression control system.

6. A yeast artificial chromosome (YAC) vector containing a recombinant nucleic acid molecule according to claim 5.

7. A cosmid vector containing a recombinant nucleic acid molecule according to claim 5.

8. A prokaryotic or eukaryotic host cell transformed or transfected with a gene or DNA sequence according to any one of the proceding claims.

9. A polynucleotide probe comprising a labelled nucleic acid sequence capable of specifically hybridizing to a gene according to any one of claims 1 to 4 or a naturally occurring variant thereof.

10. A polynucleotide probe comprising a labelled nucleic acid sequence capable of specifically hybridizing to a mammalian AZF gene.

11. A polynucleotide probe comprising a labelled nucleic acid sequence capable of specifically hybrising to a mouse AZF gene.

12. A method of detecting a mammalian AZF gene which method comprises hybridizing a probe according to any one of claims 9 to 11 with said gene, and detecting bound labelled probe.

13. Human AZF polypeptide or an antigenic fragment thereof.

14. Polypeptide comprising at least part of the amino acid sequence shown in SEQ ID N0:1 or SEQ ID NO:2 or derivatives of said polypeptide.

15. AZF polypeptide encoded by a nucleic acid sequence according to any one of claims 1 to 3.

16. An isolated polypeptide or fragment according to any one of claims 13 to 15.

17. Antibody probe immuno-reactive with a polypeptide or fragment according to any one of claims 11 to 13.

18. An antibody probe comprising a labelled antibody accroding to any one of claims 13 to 17.

19. Use of a polynucleotide probe according to any one of claims 9 to 11 for detecting an AZF gene or an AZF gene abnormality.

20. Use of an antisense polynucleotide probe to disrupt expression of an AZF gene, wherein the antisense probe hybridizes to the nucleic acid sequence according to any of of claims 1 to 4 and prevents or decreases expression of said sequence.

21. Use of an antibody according to claim 14 for detecting an AZF polypeptide, a defective AZF polypeptide or a lack of AZF polypeptide.

22. A method of diagnosing an AZF gene abnormality or a condition associated therewith which comprises restricting the AZF gene and monitoring for any variation in restriction fragments obtained.

23. A method of diagnosing an AZF gene abnormality which comprises hybridizing a sample AZF gene with a normal gene, selective cleavage of the hybridized nucleic acid at any unhybridized sites, and monitoring for any resulting cleavage products.

24. A method of diagnosing an AZF gene abnormality or a condition associated therewith which method comprises amplification by polymerase chain reaction using polynucleotide primers from substantially spaced apart portions of a nucleic acid sequence encoding AZF, and monitoring for absence or alteration of amplified polynucleotide sequence product.

25. A method of diaprosing an AZF gene abnormality or a condition associated therewith which method comprises hybridizing a polynucleotide probe according to any one of claims 9 to 11 with the Interval 6 Y chromosomal region, and monitoring for absence or reduced level of bound labelled probe.

26. A method of evaluating an effect of a drug on AZF gene enhancement or suppression comprising the steps of: providing the drug to a host cell according to claim 8, and monitoring expression of the nucleic acid.

27. A method of evaluating an effect of a drug on AZF polypeptide enhancement or suppression comprising the steps of: providing the drug to a host cell according to claim 8; effecting expression of the nucleic acid; and monitoring a level of expressed polypeptide.

28. A method of detecting AZF polypeptide expressed by a mammalian AZF gene which method comprises allowing an antibody probe according to claim 17 or claim 18 to bind with said polypeptide, and detecting bound probe.