WO1992000375A1 - Sex determining gene - Google Patents

Abstract

The invention provides a DNA encoding a gene which controls the sex of the embryos of eutherian (Placental) mammals. Nucleic acid fragments, proteins, polypeptides, antibodies and related products and their use in medicine and agriculture are provided. The invention may be used in diagnosis or for controlling the sex of the progeny of commercially important animals.

Description

Sex Determining Gene

The present invention relates to proteins, polypeptides, nucleic acid fragments, antibodies and related products and to their use in medicine and agriculture, for instance in diagnosis and therapy. More particularly the invention relates to a gene which controls the sex of the embryos of eutherian (placental) mammals and to associated products and their use in ascertaining the sex of cells, embryos and tissues and controlling the sex of the progeny of

commercially important animals.

In eutherian mammals evidence suggests that sex

determination is equivalent to testis determination

[McLaren, A., Trends Genet.. 4, 153-157 (1988)]. The gonad is composed of cells derived from four lineages: the supporting cells, steroid-producing cells, connective tissue cells and germ cells. In the developing testis, the differentiation of Sertoli cells from the supporting cell lineage is thought to result in cells of the other lineages following the male pathway [Burgoyne, P.S., Phil. Trans. R. Soc. Lond., 322. 63-72 (1988)]. The nature of this

influence is not known; it results, however, in the

differentiation of Leydig cells, the induction of mitotic arrest in the germ cells and the proliferation and organization of connective tissue and blood vessels into the testicular pattern. The Sertoli cells themselves proliferate rapidly and organize into the testis cords', apparent in the mouse at about 12.5 days post coitum

(d.p.c), and characteristic of the first signs of male development. In the female pathway, the first

morphologically apparent event other than failure to form cords is the entry of germ cells into meiosis [McLaren, A., Phil. Trans. R. Soc. Lond.. 322, 3-9 (1988), Borum, K.,

Exot. Cell Res.. 24. 495-507 (1961)], followed by the differentiation of follicle cells and their aggregation around the oogonia. The cells of the steroid-producing cell lineage give rise to theca cells, and there is little proliferation or organization of connective tissue cells. In the absence of germ cells the ovary fails to develop, which suggests that ovarian determination is dependent on gene action within the germ cells [Burgoyne, P.S., Nature, 342. 860-862 (1989)]. Germ cells are not required for testis development [Merchant, H., Devi. Biol.. 44. 1-21 (1975); McLaren. A., in The Origin and Evolution of Sex

(eds Halvorsen, H.O. & Monroy. A.) pp289-300 (Liss. New York, 1985)]. The ovarian pathway can be thought of as the normal or 'default* pathway, and the consequence of

expression of the testis determining gene(s) in the male is to switch the fate of the supporting cell precursors in the indifferent gonad (genital ridge) from that of follicle cells to that of Sertoli cells. It is assumed that the testis determining gene(s) must act within the context of other regulatory molecules that are initially responsible for the formation of the urogenital ridge and then for the development of the gonad itself.

For some time it has been suspected that there would be a single gene, normally found on the Y-chromosome, which would be responsible for initiating testis formation and the ensuing cascade of events resulting in the development of an embryo as a male. Much effort has been expended in identifying this gene and success was claimed by D. Page et al [Cell, 51: 1091-1104 (1987) ] with the discovery of a gene encoding a "zinc-finger" protein. Subsequent work has however shown that this protein is not the putative "testis determining factor" as originally believed since the gene is not present in all males containing Y sequences [Palmer et al., Nature. 342: 937-939 (1989)]. This is described in more detail below.

The mammalian Y chromosome plays a crucial role in male sex determination: an embryo that inherits a Y chromosome develops as a male; an embryo lacking a Y chromosome develops as a female [Goodfellow, P.N. and Darling, S.M., Development, 102. 251-258 (1988)]. The male sex

determining gene(s) on the Y chromosome induces testicular development and subsequent male sexual differentiation is a consequence of the hormonal products of the <testis

[Jost, A. et al., Recent Procr. Horm. Res.. 29. 1-41

(1973)]. The Y-encoded testis determining gene has been named testis determining factor (TDF) in humans and

generically and Tdv (testis determining Y chromosome) in mouse. Although it is likely that many different genes are required for both male and female sex determination and differentiation, understanding the mode of action of the testis determining gene may provide a general model for the genetic control of developmental decisions in mammals.

Attempts to identify and clone the testis determining gene have exploited detailed maps of the Y chromosome. Three types of map have been constructed: (a) deletion maps constructed by analysing the genomes of sex reversed XX males and XY females [Vergnaud, G. et al., Am. J. Hum.

Genet., 38 , 109-124 (1985) ] [Guellaen, G et al., Nature, 307. 172-173 (1984), Goodfellow, P.J. gt al., Science. 234. 740-743 (1986)], (b) a meiotic map of the pseudoautosomal region which is shared by the X and Y chromosomes

[Goodfellow, P J et al., loc. cit.; Weissenbach, J. et a l., Development. 101S. 67-74 (1987)] and (c) a long range restriction map which links the first two maps [Pritchard, C. A., Goodfellow, P. J. and Goodfellow, P. N., Nature. 328., 273-275 (1987)]. The majority of XX males have inherited Y derived

sequences, including the testis determining gene, by terminal exchange between the X and Y chromosomes

[Ferguson-Smith, M. A., Lancet. ii. 711 (1966)] [Petit, C. e t al. , Cell. 49, 595-602 (1987)]. A map based on the Y fragments present in different XX males placed the testis determining gene in the distal part of the Y chromosome adjacent to the pseudoautosomal region. The meiotic map of the pseudoautosomal region indicated that MIC2 was the pseudoautosomal locus closest to the sex specific part of the Y chromosome. A long range restriction map rooted in MIC2 identified a CpG rich region that marked a candidate for the testis determining gene. This gene, subsequently named ZFY. was cloned by Page et al. (loc. cit.) after a chromosome walk was initiated 130 kb proximal to the CpG rich region.

Analysis of the sequences present in a particular XX male (LGL203) and absent in an XY female (WHT1013) had defined the position of the testis determining gene to within an interval of 140 kb; ZFY was located within this same region. Other evidence consistent with identity between ZFY and the testis determining gene included the finding of ZFY related sequences on the Y chromosome of all eutherian mammals tested; the presence of a ZFY related gene, Zfy-1. in Sxr', the smallest part of the mouse Y chromosome known to be sex determining and the structure of the ZFY encoded protein which has many features in common with

transcription factors. However, a number of unexpected observations were made: ZFX. a homologue of ZFY. is present on the eutherian X chromosome and escapes

inactivation [Palmer et al., Proc. Natl. Acad. Sci. USA. 87. 1681-1685 (1990), Schneider-Gadicke et al. , Cell. 57. 1247-1258 (1989) ] and ZFY related sequences in metatherian (marsupial) mammals are not present on the Y or X

chromosome, but are autosomal [Sinclair, A. H. et al.,

Nature. 336. 780-782 (1988).

Two recent reports have further questioned the role of ZFY in male sex determination. In a study of the expression of Zfy-1 and Zfy-2. [Koopman, P. et al., Nature. 342. 940-942 (1989)] the mouse homologues of ZFY. it was found that their expression was limited to germ cells - a cell type not required for normal male development and in W^e/W^e mutant mice testicular development occurred in the absence of Zfy-1 and Zfy-2 expression. Moreover four XX males were found that had inherited Y derived sequences, but had not inherited ZFY. Assuming that these individuals are male because of their Y derived sequences, this mapped the testis determining gene to within 60 kb immediately

proximal to the pseudoautosomal boundary. This result is inconsistent with the published breakpoint of the XY female [Page et a l. , loc cit.] and formally excludes ZFY as a candidate for the testis determining gene [Palmer, M. S. et al.. Nature. 342. 937-939 (1989)]. (This result has subsequently been revised and the patients' breakpoints now agree with the present inventors' results for the location of the testis determining gene).

The present inventors have narrowed the search to an area of the Y-chromosome close to the pseudoautosomal boundary and, within this area, have located a coding sequence which is expressed in adult testis but no other adult human tissues, has a high degree of homology with a coding sequence in the mouse which is expressed in the genital ridge at the time that sex determination occurs and which cross hybridises with sequences found on the Y-chromosome of all eutherian mammals tested to date. It has also been shown that a portion of the coding sequence has a

remarkably high degree of homology with a part of the Mc gene which encodes the Mc protein and is located at the mating type- or mat-locus of Schizosaccharomyces pombe required for mating type determination and meiosis of this fission yeast.

The present invention will be defined and illustrated with reference to the figures of the accompanying drawings in which:

Fig. 1. shows sequences for human and rabbit genomic DNA and corresponding predicted polypeptide sequences.

Fig. 2. is a map of the distal region of the short arm of the human Y chromosome.

Fig. 3. shows a Southern blot analysis of HindIII digests of various human DNA samples. Fig. 4. shows a Southern blot analysis of Hindlll digests of DNA samples from various eutherian mammals.

Fig. 5. shows a Southern blot analysis of Hindlll digests of DNA samples from various cell lines.

Fig. 6. shows the nucleotide sequence of human and rabbit DNA segments.

Fig. 7. gives the amino acid sequences of homologous regions of S. pombe Mc Protein and the human and rabbit predicted polypeptides corresponding to the DNA sequences of Fig. 6. Fig. 8. shows a Northern blot analysis of poly(A⁺) RNA from various human tissues and cell lines.

Fig. 9. is a diagram of the probe pY53.3 (2.1 kb) subclone.

Fig. 10. shows the sequence of the 80 amino acid residue conserved mt-box of pY53.3 and corresponding sequences of other proteins.

Fig. 11 shows a Southern blot of mouse DNA digests probed with three different probes.

Fig. 12 shows a restriction map of phage L7.4.1. Fig. 13 shows a Southern blot of mouse DNA probed with various probes obtained from L7.4.1.

Fig. 14 gives the nucleotide sequences of mouse and human Y-linked regions containing an mt-box.

Fig. 15 shows the results of a polymerase chain reaction on mouse embryonic cDNA.

Fig. 16 gives the amino acid residue sequences of various mouse DNA regions compared with homologous human and east sequences. Fig. 17 shows a summary of the homology between the

sequences of Fig. 16.

Fig. 18 shows the coding sequences and corresponding amino acid residue sequences of human, rabbit and mouse mt proteins.

Fig. 19 shows the sequence of pY53.3 as deposited with the EMBL database and with NCIMB under accession number NCIMB 40308.

Fig. 20 shows the sequence of insert PKS 741, from phage clone L 7.4.1 of Example 3.

Fig. 21 shows a diagram of the open reading frame of pY53.3 and the sequence of a portion of SRY and two mutants thereof.

Fig. 22 shows portions of the sequence of SRY and related genes.

Fig. 23 shows genomic DNA fragments described in Example 5.

Fig. 24 shows results of analysis of sex-reversed embryos. Fig. 25 shows results of an analysis of an adult sex-reversed mouse.

Fig. 26 shows a Southern blot analysis of offspring from a transgenic female mouse. Fig. 27 shows the results of expression of human SRY in transgenic mouse embryos.

In Fig. 1. the top line shows the sequence of the nontranscribed strand of the human genomic DNA coding sequence identified by the present inventors. The second line, where present, is the corresponding predicted polypeptide sequence. The third and fourth lines, where present, are the corresponding non-transcribed strand of the rabbit genomic DNA coding sequence and predicted polypeptide sequence respectively. The 240 nucleotide base sequence (80 amino acid residue sequence) boxed corresponds to the Mc nucleic acid (and protein) sequence of S.pombe. The boxed bases AATAAA (line 1) and AAATAA (line 3) are the polyadenylation signals. The internationally recognised 1-letter codes are used for nucleotide base sequences in Fig. 1 and throughout this specification, and the

internationally recognised 3-letter codes are used for amino acid residue sequences in Fig. 1. and either the 3-letter code or the internationally recognised 1-letter code is used throughout this specification for amino acid sequences.

It is believed that the fission yeasts and all higher eukaryotic organisms, such as plants, invertebrates and vertebrates including avians such as chickens and

particularly mammals such as metatherian mammals and humans, domestic livestock including bovines, ovines, equines and porcines and other eutherian mammals of

commercial significance, will have a gene containing a sequence similar to the 240 nucleotide sequence boxed in Fig. 1. and that the gene product will be a protein which is crucial to the determination of the mating type or sex of the organism.

It will be understood that the exact sequence will vary between species and between individuals of the same species at least at the nucleic acid level and often also at the protein level. The sequence of the mouse gene is shown in Fig. 20. That portion of a gene or protein which

corresponds in sequence with the boxed protein in Fig. 1 will hereafter be referred to as an "mt-box" and proteins and fragments thereof, polypeptides, nucleic acids and fragments thereof and oligonucleotides containing an mt-box will hereafter be referred to as mt-proteins, mt-nucleic acids and so on. The present invention therefore provides an mt-protein or a fragment thereof or polypeptide comprising an mt-box or a part thereof, subject to the proviso below.

The present invention also provides a protein or a fragment thereof or a polypeptide containing a mimetope of an epitope of an mt-protein or fragment thereof of polypeptide containing an mt-box or a part thereof, subject to the proviso below. Such proteins, fragments and polypeptides are hereafter referred to as mt-mimetope proteins or fragments thereof and mt-mimetope polypeptides.

The present invention also provides an mt-nucleic acid or a fragment thereof or oligonucleotide comprising an mt-box, or a part thereof subject to the proviso below.

In a particular aspect the present invention provides a single or double stranded nucleic acid comprising the mt-box of an eutherian mammal or a part thereof of at least 17 contiguous nucleotide bases or base pairs, or a single or double stranded nucleic acid hybridisable with the mt-box of an eutherian mammal, or a part thereof of at least 17 contiguous nucleotide bases or base pairs, subject to the proviso below. The invention further provides a nucleic acid or a fragment thereof or an oligonucleotide encoding an mt-protein or fragment thereof or a polypeptide comprising an mt-box or a part thereof or an mt-mimetope protein or fragment thereof or mt-mimetope polypeptide, subject to the following proviso. These nucleic acids, fragments and

oligonucleotides may have sequences differing from the sequences of mt-nucleic acids, fragments and

oligonucleotides due to alternative codon usage and/or encoding alternative amino acids sequences of mimetopes.

The present invention does not, however, extend to any known protein or fragment thereof or polypeptide or nucleic acid or fragment thereof or oligonucleotide containing an mt-box, such as the S.pombe Mc protein and the S. pombe gene at the mat locus, or the HMG proteins described below, insofar as that protein or fragment, polypeptide, nucleic acid or fragment or oligonucleotide is known per se.

The amino acid sequence of the mt-box is similar to the DNA-binding motif of a number of known DNA-binding proteins and sequence specific DNA binding has been obtained with mt-proteins. This suggests that the mt-protein of the present invention may have a regulatory function requiring DNA binding. However there are residues in the amino acid sequence of the mt-box which are conserved at least between human, rabbit and mouse mt-proteins but which are not conserved in the sequences of DNA-binding proteins not associated with sex determination at least at the stage of testis formation. Any one or more of these conserved residues is therefore considered characteristic of the mt-box proteins of the present invention. A protein having the DNA-binding motif but lacking all of these

characteristic amino acid residues is therefore outside the scope of the present invention. The characteristic amino acid residues are shown in Fig. 16, which is described in more detail below, at positions 46,63,67,74,75,76 and 98.

The nucleotide base sequence of the mt-box includes bases which encode the DNA-binding motif of DNA-binding proteins as described above. However the base sequence of the mt-box of the mt-nucleic acids of the invention will also include one or more codons specifying one or more of the characteristic amino acid residues described above and/or will be hydridisable with a Y-chromosome specific sequence associated with the TDF-gene under conditions which

substantially prevent hybridisation with any autosomal or X-chromosomal sequence of a XX female or XY male eutherian mammal. Preferably the mt-nucleic acids of the invention encode one or more, preferably all, of the characteristic amino acid residues and meet the above hybridisation requirement.

Fragments of mt-nucleic acids according to the present invention will likewise contain codons specifying one or more of the characteristic amino acid residues discussed above together with at least a part of the DNA-binding motif or mt-box adjacent to the codons encoding the

characteristic residue or residues, and/or the fragments will be hybridisable with a Y-chromosome specific sequence associated with the TDF-gene preferably under conditions which substantially prevent hybridisation with any

autosomal or X-chromosomal sequence of a XX female or XY male eutherian mammal. Oligonucleotides containing the mt-box or a part thereof according to the present invention may contain codons specifying one or more of the characteristic amino acid residues discussed above together with neighbouring

residues of the DNA-binding motif or mt-box but this is not essential. However all such oligonucleotides of the invention must be capable of hybridisation with a Y- chromosome-specific sequence associated with the TDF-gene, preferably under conditions which substantially prevent hybridisation with any Y-chromosome sequence not associated with the TDF-gene. Preferably the mt-oligonucleotides will hybridise with a Y-chromosome specific sequence associated with the TDF-gene under conditions which substantially prevent hybridisation with any autosomal or X-chromosomal sequence of a XX female or XY male eutherian mammal.

As used herein the term "Y-chromosome specific sequence" refers to a DNA sequence found in the Y-chromosome of a XY male eutherian mammal which sequence is not found in the wild type X-chromosome nor in any of the wild type

autosomes of that mammal.

The TDF-gene referred to herein is that Y-chromosome specific sequence which contains the mt-box and which encodes a functional testis determining factor which when expressed at the appropriate stage of embryo development results in testis formation and subsequent growth of the embryo as a male.

As used herein the term "Y-chromosome specific sequence associated with the TDF gene" refers to a DNA sequence which is found in the region immediately adjacent the pseudoautosomal boundary and extending for 35 kb into the Y-specific sequence of the human Y-chromosome, or

corresponding region of the Y-chromosome of any other eutherian mammal. The hybridisation conditions referred to above which prevent unwanted hybridisations with Y-chromosome sequences not associated with the TDF gene or X-chromosomal and autosomal sequences will depend to some extent on the length of the nucleic acid, fragment or oligonucleotide of the invention being tested. Thus for instance lower stringency will be sufficient to secure hybridisation to the Y-chromosome sequence associated with the TDF-gene whilst preventing unwanted hybridisation when the nucleic acid or fragment is several thousand nucleotide base pairs in length, such as the probe pY53.3 described below (see for instance the moderate high stringency conditions described below), than for a fragment of only a few

hundreds of bases or an oligonucleotide of from 17 bases up to a few tens or hundreds of bases. With the smallest oligonucleotides and fragments of the invention

hybridisation conditions will be such that only complete complementarity between the oligonucleotide or fragment and the Y-chromosome specific sequence associated with the TDF gene will result in hybridisation.

Preferred nucleic acids and fragments of the invention will only hybridise selectively to the Y-chromosome specific sequence associated with the TDF-gene under conditions requiring at least 80%, for instance 85, 90 or even 95%, more preferably at least 99% complementarity. Yet more preferred nucleic acids and fragments of the invention are those having a sequence corresponding exactly to that of probe pY53.3 described hereafter or the 0.9 kilobase (kb) Hindi fragment thereof (pY53.3B) although the nucleic acids or fragments of the invention may be longer or shorter than pY53.3 or pY53.3B. Further preferred nucleic acids and fragments of the invention are those having a sequence corresponding exactly to the human, rabbit or mouse sequence as set out in Fig. 1, Fig. 6, Fig. 14 or Fig. 18.

In particular embodiments the present invention provides the nucleic acid deposited on 12th July 1990 with NCIMB under accession number NCIMB 40308 referred to as pY53.3 (2.2 kb) and the 0.9 kb fragment thereof resulting from HincII digestion, pY53.3B.

The sequence of pY53.3 has been deposited with the EMBL DNA database.

The invention particularly provides an oligonucleotide, polypeptide, nucleic acid or protein comprising the entire sequence of the mt-box of a eutherian mammal and more preferably comprising the entire amino acid or nucleotide sequence of the human, mouse or rabbit as set out in any one of Fig. 1 , Fig. 6 , Fig. 7 , Fig. 10 , Fig. 14 , Fig. 16 and Fig. 18.

The nucleic acids hybridisable with the mt-box of a

eutherian mammal are preferably hybridisable under

moderate, or more preferably, high stringency conditions as defined below:

Moderate stringency: buffer: 2 X SSC

temp: 50ºC

annealing period: 6-8 hours

High Stringency: buffer: 1 X SSC

temp: 65ºC

annealing period: 6-8 hours Moderate stringency as defined above corresponds with about 75% homology. High stringency as defined above corresponds with about 90% homology. 1 X SSC is 0.15 M sodium

chloride, 0.015 H sodium citrate, pH 7.0.

Preferably the portion of the nucleic acid corresponding to or hybridisable with the mt-box is at least 20, more preferably at least 30, 40 or 60 and most preferably 100 or more nucleotide bases in length. ,

The nucleic acids of the invention may be single or double stranded DNA or RNA. DNA's of the invention may comprise coding and/or non-coding sequences and/or transcriptional and/or translational start and/or stop signals and/or regulatory, signal and/or control sequences such as

promoters, enhancers and/or polyadenylation sites,

endonuclease restriction sites and/or splice donor and/or acceptor sites, in addition to the mt box sequence.

Included within the DNA's of the invention are genomic DNA's and complementary DNA's (cDNA's) including functional genes or at least an exon containing the mt box. They may also contain non-coding sequences such as one or more introns. Single stranded DNA may be the transcribed strand or the non-transcribed (complementary) strand. The nucleic acids may be present in a vector, for instance a cloning or expression vector, such as a plasmid or cosmid or a viral genomic nucleic acid. RNA's of the invention include unprocessed and processed transcripts of DNA, messenger RNA (mRNA) containing the mt-box and anti-sense RNA containing a sequence complementary to the mt-box.

In another aspect of the invention there is provided nucleic acid having a sequence the same as or homologous to at least a portion of the human, mouse or rabbit mt-nucleic acid sequence of Fig. 1 or Fig. 6 or Fig. 1 4 or Fig. 18 but not including the mt-box. Such nucleic acids will have at least 50% homology, more preferably at least 75% homology, for instance 80, 85 or 90%, 95 or even 99% homology with the sequence of human, mouse or rabbit mt-nucleic acid over a region at least 20, preferably at least 30, for instance 40, 60 or 100 or more contiguous nucleotide bases in length. These nucleic acids will further be hybridisable under conditions of moderate or high stringency as defined above with a mt-nucleic acid of a fission yeast or higher organism, preferably a eutherian mammal. They may be single or double stranded DNA or RNA as described above.

Nucleic acids of the present invention are particularly useful as primers for polymerase chain reactions (PCRs) conducted to ascertain the mating type or sex of an

organism as described below. They may also be used to express proteins or fragments or polypeptides corresponding to the whole or a part of an mt-protein (whether or not containing an mt-box) or as probes in hybridisation experiments. As used herein the term "fragments" used in connection with proteins is intended to refer to both chemically produced and recombinant portions of proteins. The mt-proteins and fragments thereof and polypeptides containing the mt-box or a part thereof and mt-mimetope proteins and fragments thereof and mt-mimetope polypeptides of the invention are useful in immunodiagnostic testing and for raising antibodies such as monoclonal antibodies for such uses. Antibodies against such proteins and fragments and polypeptides as well as fragments of such antibodies (which antibody fragments include at least one antigen binding site) including chemically derived and recombinant fragments of such antibodies, and cells, such as eukaryotic cells, for instance hybridomas and prokaryotic recombinant cells, capable of expressing and, preferably, secreting antibodies or fragments thereof against such proteins or fragments, also form part of the present invention. The nucleic acids of the invention may be obtained by conventional means such as by recovery from organisms using PCR technology or hybridisation probes, by de novo

synthesis or a combination thereof, by cloning pY53.3 or a fragment thereof or by other techniques well known in the art of recombinant DNA technology.

Proteins and fragments thereof and polypeptides of the invention may be recovered from cells of organisms

expressing an mt-gene or generated by expression of an mt-gene or coding sequence contained in a nucleic acid of the present invention in an appropriate expression system and host, or are obtained by de novo synthesis or a combination thereof, by techniques well known in the art of

biotechnology. The proteins, fragments thereof and

polypeptides of the invention will contain naturally occurring L-α-amino acids and may also contain one or more non-naturally occurring o-amino acids having the D- or L-configuration.

Antibodies may be obtained by immunisation of a suitable host animal and recovery of the antibodies, by culture of antibody-producing cells obtained from suitably immunised host animals or by in vitro stimulation of B-cells with a suitable mt-protein, fragment or polypeptide or mt-mimetope, protein, fragment or polypeptide and culture of the cells. Such cells may be immortalised as necessary for instance by fusion with myeloma cells. Antibody fragments may be obtained by well known chemical and biotechnological methods.

All these techniques are well known to practitioners of the arts of biotechnology. Reference may particularly be made to the well known text book "Molecular Cloning: A

Laboratory Manual" 2nd Edition (Eds Sambrook, J., Fritsch, E.F. and Maniatis, T.), (Cold Spring Harbour Laboratory, New York, 1989), hereafter referred to as "Maniatis". The invention further provides the use of a nucleic acid, protein, polypeptide, antibody or antibody producing cell as hereinbefore defined including the Mc protein and gene of S. pombe or other mt-nucleic acid or protein for

ascertaining the mating type or sex of a cell or organism of a fission yeast or higher eukaryote or for isolating nucleic acids useful in ascertaining the mating type or sex of a cell or organism and for instituting single sex breeding programmes. Now that the nucleic acid, proteins, polypeptides, antibodies and fragments thereof and antibody producing cells of the invention have been made available such uses may be conducted by routine techniques well known to practitioners of the arts of biotechnology.

Knowledge of the mouse or human testis determining gene can be used to isolate the equivalent gene from other mammals. Once isolated from a particular species, this gene and its sequence can be used in two types of application:

1. The construction of sequence-based sexing tests which can be applied to embryos, tissues and other

biological materials containing nucleic acids.

2. The genetic modification of the germ line of mammals to create breeding systems that produce offspring statistically biased towards one sex or of one sex only (single sex breeding systems) .

The feasibility of this approach is demonstrated in

Examples 1 and 2 below where the human SRY sequences are used to isolate the equivalent mouse and rabbit sequences. PCR based sexing methods for human, rabbit and mouse are also described below.

A particularly preferred technique for ascertaining the mating type or sex of a cell or an organism in accordance with the invention involves the use of oligonucleotides as primers in a PCR, for instance as follows:

A cell or cells are obtained, for instance by surgical removal from an embryo, and the DNA is released by a crude lysis procedure, for instance using a detergent or by heating. Primer oligonucleotides of the invention are used to initiate a conventional PCR in order to amplify mtrelated DNA from the cells. The products of the PCR are analysed by agarose gel electrophoresis and detected using labelled probes. The presence of amplified DNA indicates the presence of an mt-gene in the cells and thus, in eutherian mammals, that the cell(s) were male.

This technique may be applied for instance to identify human embryos likely to suffer a sex-related disease for termination, or to control the sex of the progeny of breeding stock for commercial exploitation (by selection of the breeding stock or by slaughter or termination of animals of undesired sex). The oligonucleotide primers for ascertaining or controlling sex in one species may also be used for developing primers for ascertaining or controlling sex in another species since hybridisation of the primers to the mt-gene of the other species will still serve to initiate a PCR and amplify the species-specific sequences.

Techniques for conducting such sex determinations are well known in the art of recombinant DNA technology.

In one aspect the present invention provides a process for isolating a Y-chromosome specific sequence associated with the TDF gene of an eutherian mammal which comprises probing a genomic library from a male of the species, preferably of Y-chromosome sequences, for instance of lambda phage, cosmid or YAC library or a cDNA library constructed from an RNA from an expressing tissue such as adult testis or foetal genital ridge tissue, with a probe comprising a nucleic acid, fragment or oligonucleotide of the invention as hereinbefore defined and a detectable label under high stringency conditions. Preferably the isolation is conducted using standard molecular biology techniques (as described in, for example "Maniatis") by plating out several genome-equivalents of the genomic or cDNA library and screening with a nucleic acid probe of the invention under conditions of moderate to low stringency. Positive clones are isolated and the sequences corresponding to the testis determining gene are subcloned.

Using the newly isolated subclone, Southern blots are performed on male and female DNA of the species of interest at high stringency to confirm that the correct clone has been isolated. The probe should give a strong male

specific signal (other male/female shared bands may also be present at lesser intensities). The subclone is sequenced using standard methods and primers suitable for PCR chosen from the sequence so identified.

Alternatively, other approaches to cloning the sequences related to the testis-determining gene could be used such as PCR methods using "degenerate" oligonucleotides. (For methods in PCR see, for example, "PCR Protocols - a Guide to Methods and Application"; edited by M.A. Innis, D.H. Gelfand, J.J. Sninsky, T.J. White; published by Academic Press, Inc.). Preferably the probe is pY53.3 or a fragment thereof or a nucleic acid or fragment or oligonucleotide having a sequence exactly as set out in Fig. 1 or Fig. 6 or Fig. 14 or Fig. 18 for the human, rabbit or mouse. Techniques for forming a genomic or cDNA library and for probing and detecting the detectable label and isolating the nucleic acid identified by the probe are well known in the art of biotechnology and recombinant DNA manipulation. The process may be conducted for instance using a probe having the human sequence such as the pY53.3 sequence to identify and isolate the corresponding sequence from another

eutherian mammal such as a bovine. The thus-identified sequence can then be used to generate primers for PCR which in turn can be used to ascertain the sex of an individual or of cells, tissues, embryos or sperm of the bovine or other mammal. This will permit experiments to ascertain sex to be conducted and controlled sex breeding of the bovine or other mammal as described below.

The isolated nucleic acid, fragment or oligonucleotide may thereafter be amplified, cloned or sub-cloned as necessary. The invention further provides a process for detecting the sex of an individual eutherian mammal or of cells, tissues, embryos, foetuses or sperm of a eutherian mammal comprising conducting a polymerase chain reaction using DNA from the individual, cell, tissue, embryo or sperm as template and a nucleic acid, fragment or oligonucleotide of the invention as primer. Preferably the nucleic acid, fragment or oligonucleotide of the invention used as primer is pY53.3 or a part thereof or has a sequence corresponding exactly to the human, rabbit or mouse sequence of any one of Fig. 1, Fig. 6 or Fig. 14 or Fig. 18 or a part thereof or is a nucleic acid, fragment or oligonucleotide which is a Y-chromosome specific sequence associated with the TDF gene of a eutherian mammal of the same species as the

individual, cell, tissue, embryo, foetus or sperm whose sex is to be ascertained. The Y-chromosome specific sequence associated with the TDF gene of the mammal involved may itself have been obtained by the process of isolation and amplification or cloning described above.

The identification of TDF genes according to the present invention raises the possibility of controlling the sex of progeny of commercially important animals such as bovines, ovines, equines, porcines and also avians. This will be valuable in many aspects of animal breeding and husbandry such as where one sex has more desirable characteristics, for instance only male progeny are desired from beefproducing strains or breeds of cattle whereas only female progeny are desired from dairy breeds of cattle and egglaying breeds of chicken. The economic advantages of single sex breeding programmes and strategies for

instituting these are described for instance in "Single Sex Beef Cattle Systems", Taylor, St. C.S., Thiessen, R.B., and Moore, A.J. in "Exploiting New Technologies in Animal

Breeding; Genetic Developments", (Eds. Smith, C., King, J.W.B. and McKay, J.C.), (Oxford University Press, Oxford, 1986) .

The nucleic acids making up all or part of the testis determining gene, from the same or different animal

species, can be introduced into any early embryo through established transgenic technology. This latter includes microinjection of DNA into pronuclei or nuclei of early embryos, the use of retroviral vectors with either early embryos or embryonic stem cells, or any transformation technique (including microinjection, electroporation or carrier techniques) into embryonic stem cells or other cells able to give rise to functional germ cells. These procedures will allow the derivation of individual

transgenic animals (founder transgenics) or chimeric animals composed in part of cells carrying the introduced DNA. Where the functional germ cells of the founder transgenic or chimeric animal carry the introduced DNA it will be possible to obtain transmission of the introduced DNA to offspring and to generate lines or strains of animals carrying these DNA sequences. The nucleic acids making up part or all of the coding sequence of the testis determining gene, or derivatives of it, may be introduced in combination with its own

regulatory sequences (promoter/enhancers etc.) or

regulatory sequences from another gene, the whole making the "construct", to give expression from the construct at an appropriate developmental stage and tissue location critical to sex determination in the animal species under consideration. For example, in the mouse this would be between 9.5 and 13.5 days post coitum in the urogenital ridge.

The following are examples of the types of procedures for providing altered sex ratios of offspring from anywhere within the range of 100 % male to 100 % female. It is important to note that males with an XX sex chromosome constitution are sterile, and females with an XY sex chromosome constitution are likely to have much reduced fertility. However, it may not be necessary to produce many animals of the desired genotype in cases where semen can be collected and stored and used in artificial insemination or in in vitro fertilisation programmes.

Repetition of transfections, screening and selection of transgenic animals may be required in order to identify suitable breeding stock able to pass the gene in effective form to offspring.

Scheme A. The testis determining gene is introduced by one of the techniques outlined above (through embryos or ES cells). This will give a proportion of founder animals of XY sex chromosome constitution carrying the

introduced copy of the gene, the "transgene", on an autosome or on the X chromosome, in addition to the endogenous gene on the Y chromosome. If the transgene is autosomal, then 75 % of the offspring of this founder animal will be male (Table 1, part 1) unless the transgene has been integrated into a site where local effects prevent or reduce its expression; individual animals are screened and only those where the transgene is

functionally integrated will be selected for use. If the transgene is X-linked then up to 100 % of offspring will be male (Table 1, part

2). If copies of the transgene have

integrated to more than one autosomal location then the proportion of male offspring from the founder will be greater than 75 % and may approach 100%. (Table 1, part 3).

Scheme B. With embryonic stem cells, targetted or site specific integration (e.g. via homologous recombination) may be used to introduce copies of the transgene at identical sites in

homologous chromosomes, such that these cells, and germ cells derived from them, will be homozygous for the transgene. 100% of

offspring derived from these cells will be male.

Scheme C. It is possible to use antisense RNA, or DNA constructs transcribing antisense RNA, to effectively reduce or abolish translatable mRNA within a cell. The antisense RNA can be produced from a DNA construct comprising homologous or heterologous regulatory

sequences linked to part or all of the

transcribed sequence placed in the reverse orientation compared to the normal gene.

Antisense RNA expressed from this antisense construct will complex with the "sense" or normal transcripts and reduce the amount of gene product (protein), causing sex reversal, such that XY individuals carrying an autosomal or X-linked copy of the antisense construct will be female. XX animals carrying copies of the construct at one locus will give 75% female offspring (Table 2, part 1). Multiple copies of the antisense construct integrated at different locations will give up to 100% female offspring (Table 2, part 2).

Scheme D. Expression of altered protein may interfere with the action of the normal protein leading to sex reversal. Constructs expressing an altered protein product will produce similar results to those coding for antisense RNAs, and their use will be essentially equivalent.

Scheme E. Combination of sense (testis determining) and antisense or altered protein producing

constructs may be used to obtain animals homozygous for one or the other. The

offspring of these animals will be all one sex (Table 2, part 3).

In a further aspect of the invention there is provided a process for producing a eutherian mammal whose progeny will be statistically biased in favour of, or only of a single sex which process comprises introducing a functional nucleic acid containing a coding sequence into the genome of the animal or a progenitor thereof which coding sequence encodes the TDF protein of that mammal or which encodes anti-sense RNA having complementary sequence to the TDF mRNA of that mammal. In accordance with one embodiment of this process,

insertion of a functional TDF gene into an autosomal, X- chromosomal or pseudoautomal locus will result in

expression of the TDF gene product in the embryo and development of that embryo as a male. Preferably several copies of the TDF gene are inserted at different loci in the genome, particularly in more than one autosomal locus, such that all gametes produced by that animal will contain copies of the TDF gene irrespective of whether they contain a X or a Y chromosome.

In accordance with another embodiment of the process the coding sequence is inserted together with appropriate regulatory sequences such as promoters and enhancers which ensure transcription of the gene into anti-sense RNA.

Again it is preferred that the coding sequence is inserted at autosomal, X-chromosomal or pseudoautosomal loci, preferably in several copies at different autosomal loci. When expressed in an embryo such a sequence will result in the production of anti-sense RNA which can anneal to and thereby prevent translation of the natural mRNA of

potential males of the species, thereby preventing

expression of the natural TDF gene and ensuring that the embryo will develop as a female. Techniques for producing such transgenic animals are available and conventionally used in the field of biotechnology. Such transgenic animals and single-sex breeding processes using animals form further aspects of the invention.

The following references give more details of transgenic techniques in general: Manipulating the Mouse Embryo. B. Hogan, F. Constantini & E. Lacy. Cold Spring Harbor Laboratory, (1986); Transgenic Animals, R. Jaenisch., Science. 240. 1468-1474, 1988;

Teratocarcinomas and Embryonic Stem Cells, a practical approach. edited by E.J. Robertson, IRL Press, Oxford

(1987) and Altering the Genome bv Homologous Recombination. M.R. Capecchi. Science, 244, 1288-1292, 1989, the third and fourth being particularly pertinent to manipulation of embryonic stem cells and homologous recombination

respectively. Antisense techniques are described further in Antisense RNA and DNA Edited by D.A. Melton. Cold Spring Harbor Laboratory (1988), and the application of these techniques is described in Conversion of Normal Behaviour to Shiverer by Mvelin Basic Protein Antisense cDNA in

Transgenic Mice M. Katsuki, M. Sato, M. Kimura, M.

Yokoyama, K. Kobayshi & T. Nomura. Science 241, 593-595, 1988. Other conventional genetic engineering techniques may be applied to inactivate the testis determining gene and thereby secure all female progeny. The invention will now be described with reference to the accompanying Figures in the following non-limiting

Examples.

Example 1 A Southern blot search of 40 kb of the Y chromosome

Probes from a previously described chromosome walk, comprising a series of overlapping lambda and cosmid clones, from the pseudoautosomal region, across the

pseudoautosomal boundary, to the sex specific region of the Y-chromosome [Ellis, N. A. et al., Nature 337. 81-84

(1989)], were used on the genomes of XX males to define the region in which TDF must lie [Palmer, M. S. et al ., Nature 342. 937-939 (1989)].

Three additional probes are used in this Example on the genomes of the same ZFY negative XX males, as shown in Fig. 2, which shows a map of the distal short arm of the human Y chromosome: the stippled region at the left is the

pseudoautosomal region, the broken line is the boundary between the pseudoautosomal and the Y-specific regions. At the top are the three overlapping lambda clones lambda 51, lambda 4, lambda 53 and the plasmid pNB obtained from walking along the Y chromosome. The breakpoints of the XX males are defined by the broken lines at 35 kb (see Fig. 3). The black boxes represent probes that detect single copy Y-specific human DNA fragments, indicated by (+).

When these probes were hybridised to male and female DNA from bovine and murine genomes, only pY53.3 detected Y- specific fragments (+). All of the probes except pYH8 hybridised to sequences in the XX males. pY4.1B which was positive with the XX males and the probe pYH8 which was negative. The third Y-specific probe, pYRO.4 is derived from sequences lying between pY4-1B and pYH8, and appears to define the break points in the XX males. pYRO.4 detects an 8.5 kb Hindlll fragment in normal males but only a 4 kb fragment in two related individuals: an XX male (TL) and an hermaphrodite (DL), while in a third, unrelated XX male (ZM) a 6kb fragment was detected as follows: Genomic DNA (10μg) was digested with Hindlll, separated on a 9.8% agarose gel, transferred to Hybond N⁺ (Amersham) and fixed in a 0.4 M NaOH (20 min). In order to suppress repeat sequences present in the probe pYRO.4 it was

labelled with ³²P, denatured along with 500 μl (5 mg/ml^-1) sonicated human placental DNA and prehybridised in 2 ml hybridisation buffer at 65ºC for 2 hours [Sealy, P. G., Whittaker, P. A. and Southern, E. M., Nucl. Acids. Res., 13. 1905-1922]. This probe mixture (2 ml) was added to the filter in a buffer of 5 × SSPE, 5 × Denhardt's solution, 0.5% SDS, 200 μg/ml denatured salmon sperm DNA, and 1%% sonicated denatured human placental DNA and hybridised for 16 hours at 65°C. The filter was extensively washed for 0.2 × SSC, 0.2% SDS at 65°C and autoradiography was for 6 days. The breakpoints in the XX males are clustered around a region which is approximately 35 kb from the boundary.

This result implies that TDF is located in sex specific sequences within 35 kb of the pseudoautosomal boundary. Further refinement of the positions of the breakpoints was not possible because of the highly repetitive nature of the sequences between pY4.1B, pYRO.4 and pYH8.

To find TDF within this 35 kb of DNA, the following

strategy was adopted. The DNA was subcloned into

convenient fragments of approximately 4 kb in size;

subsequently each 4 kb fragment was cleaved with a

"frequent cutter" restriction enzyme such as Rsal. to produce smaller fragments in the size range 500 bp to 1 kb. A total of 50 probes were tested. Each small fragment was radioactively labelled and used to probe Southern blots of DNA from: human males and females; murine males and females; bovine males and females; and human-hamster somatic cell hybrids containing the human X or human Y chromosome. Methods were as described in connection with Fig. 4 but autoradiography was overnight. Fig. 5 shows the Southern blot of

(male cell line PGF) Goodfellow, P. J. et al., Ann. Hum. Genet.. 53. 15-22

(1989)]; 4Y (49 XYYYY cell line, Oxen) [Bishop, C. E. et al. _f Nature. 303. 831-832 (1983)]; ♀ (female cell line WT49) [DeKretser, T. A. et al., Eur. J. Immunol.. 12. 600606 (1982)]; 4X (48 XXXX cell line, GM1416B) (Coriell

Institute for Medical Research Camden, N.J.); H (hamster parent cell line, W3GH) [Westerveld, A. et al., Nature New Biol.. 234. 20-24 (1971)]; X (hamster-human hybrid cell line containing the human X chromosome, CL2D) [Westerfeld et tl., loc.cit.]; Y (hamster-human hybrid cell line containing the human Y chromosome, 853) [Burk, R. D., Ma, P. and Smith, K. D., Mol Cell Biol.. 1, 576-581 (1985)], in which the filter was hybridised with the (0.9 kb) Hindi fragment of pY53.3 which detects a 2.1 kb Hindlll fragment in

4Y and Y; the intensity of the bands reflects the number of copies of the Y chromosome present.

All probes were tested with and without prehybridisation to total human DNA in an attempt to suppress the contribution of repeat sequences [Sealy et al. , loc.cit.]. Despite this latter precaution the majority of probes tested failed to detect unique sequences in the genome of humans but reacted with repetitive elements distributed throughout the genome. These repetitive probes frequently detected repeats in the bovine genome. Seven probes were found that detect single copy Y-specific bands in human DNA (see Fig. 2) :pYNB (0.7 kb), pY53.3 (2.1 kb), pY53.1 α(0.8 kb) , pY53.1B(0.8 kb), pY53.1 gamma (1.3 kb), pY53.2 (0.9 kb) and pY4.1B(0.2 kb). However, of the seven probes only pY53.3 reacts with Y-specific bands in the murine and bovine genomes.

Fig. 9 shows the pY53.3 (2.1 kb) subclone: the shaded region is the open reading frame (ORF); the black box is the region covered by the 80 amino-acid conserved motif, which shows homology with Mc protein of S. pombe and several non-histone proteins relating to HMG1 and HMG2. The numbers represent base pair numbering and Hindi sites define the 0.9-kb subclone used as a probe.

The 0.9 kb Hindi subclone of pY53.3 hybridised most strongly to Y-specific fragments in human, murine and bovine genomic DNA, consequently this subclone was used as a probe in subsequent experiments.

Conservation of pY53.3

An Ark blot containing DNA from males and females of eutherian mammals hybridized with the 0.9 kb Hindi

subclone of pY53.3 was prepared as follows: Genomic DNA (10 μg) was digested with Hindlll, separated on a 0.8% agarose gel, transferred to Hybond N⁺ (Amersham) and fixed in 0.4M NaOH. pY53.3 was labelled with ³²P and added to the filter in a buffer of 5 × SSPE, 5 × Denhardt's solution, 0.5% SDS, 10% dextran sulphate, 200 μg/ml denatured salmon sperm DNA and hybridised for 16 hours at 65°C. The filter was washed in 1 × SSC, 0.2% SDS at 65°C and

autoradiographed for 2 days. The results are shown in Fig. 4. The probe detects male-specific fragments in: human (2.1 kb), chimp (-18 kb), rabbit (4.2 kb), pig (-6.6 kb), horse (-10 kb), cattle (-1.6 kb) and tiger (-6.6 kb).

The sequences detected by the subclone of pY53.3 are conserved and male specific in a wide spectrum of mammals. At low stringency, additional fragments were found that are

shared between males and females, suggesting the existence of pY53.3 related X-linked or autosomal sequences.

However, these fragments were not detected in a humanrodent somatic cell hybrid which retained the human X chromosome as the only human contribution (Fig. 5).

It is unusual to find sequences that are conserved between the Y chromosomes of different mammalian species: the only other example known is ZFY [Page et al., loc. cit.]. This result is consistent with pY53.3 reacting with functional sequences on the eutherian Y-chromosome.

Sequence analysis of pY53.3 and other Y-unique sequences

The sequence of pY53.3 was determined by primer walking and is set out in the top line of Fig. 6.

Plasmids were subcloned into pUClδ vectors (NEB) and were sequenced as double-stranded DNA by the dideoxy termination method [Sanger, F, Nicklen, S and Coulson, A R, Proc Natl Acad Sci USA 74. 5463-5467 (1977)] using synthetic

oligonucleotide primers and Sequenase (USB). Nucleotide sequences were compared to the EMBL DNA data base.

A search of the EMBL DNA sequence database failed to find any sequence related to pY53.3. However, inspection of the pY53.3 (2.1 kb) sequence reveals two long open reading frames that overlap in different frames 5' → 3' from the centromere toward the pseudoautosomal boundary. The conceptual translation of these open reading frames was used to screen the PIR protein data-base using a similarity search algorithm [Smith, T F and Waterman, M. S., J. Mol. Biol.. 147. 195-197 (1981), Collins, J. F. Coulson, A. F. W. and Lyall, A., CABIOS, 4, 67-71 (1988)]. The second open reading frame encodes a region of 120 amino acids shared with the homologous rabbit Y-specific sequence which was found to be related to the Mc protein encoded by the sequence at the mat3-M locus of the fission yeast

Schizosaccharomvces pombe [Kelly, M. et al., EMBO J., 7 , 1537-1547 (1988)]. Over a region of 80 amino acids the two proteins share 47 residues which are either identical or represent conservative changes (Fig. 7, top and middle lines). The 80-residue conserved motif in pY53.3 also showed homology with the domain found in the nuclear non-histone proteins HMG1 and HMG2. High mobility group (HMG) proteins 1 and 2 are thought to play a part in chromosomal structure and gene activity, and some display enhanced DNA-binding to A+T-rich single-stranded sequences [Jantzen H.M. et al., Nature. 344. 830-836 (1990)]. HMG1 and HMG2 are known to be associated with regions of transcriptionally active chromatin. Within HMG1 and HMG2 there is a motif, the HMG box, which has been found in several proteins including the non-histone chromosomal protein NHP6 of Saccharomyces cerevisiae [Kolodrubetz D. & Burgum, A., J. Biol. Chem.. 265: 3234-3239 (1990)], the yeast ARS-binding protein, ABF2, and the human nucleolar transcription faccjr hUBF (human upstream binding factor) [Jantzen, loc. cit.]. The hUBF product is an RNA polymerase I transcription factor which interacts with sequence-specific DNA regions. This motif might represent a novel class of DNA-binding domains [Jantzen, loc. cit.]. The conserved binding motif seems to be present in a large family of sequences perhaps originating from an early HMG-like non-specific DNA-binding structure.

Fig. 10 compares the conserved 80 amino acids (single letter code) of pY53.3 (human) with the Mc protein of S.. pombe (Mc), human upstream binding factor (hUBF), non-histone chromosomal protein (NHP6) from S. cerevisiae and high mobility group protein 1 (HMG1). Boxed regions are identical amino acids, shaded regions show conservative amino acid changes with respect to the human pY53.3

sequence.

At the 3' end of the Mc-related open reading frame the PY53.3 sequence continues for 68 amino acids before

reaching a stop codon. No potential splice donor site was found in the DNA sequence of this region but a putative polyadenylation signal is present 133 base pairs downstream of the stop codon. In the 5' direction there is a

potential consensus splice acceptor signal in the pY53.3 DNA sequence at the point where homology between the

S.pombe Mc protein and the pY53.3 Mc-related sequence breaks down. The open reading frame continues in the 5' direction for another 75 amino acids and within this region two potential start codons are found in pY53.3.

To test the conservation and hence functional importance of the sequence motifs in pY53.3, the homologous, Y-specific, rabbit sequence was cloned and sequenced. Fig. 6 shows the nucleotide sequence of pY53.3 (human) (top line) and rabbit Y-specific homologous seguence (bottom line). Vertical lines indicate matching bases, lower case letters indicate base differences. The homology with the S pombe Mc region is boxed.

Fig 7 compares the amino acid sequence of the pY53.3

(human) (middle line) and rabbit Y-specific homologue

(bottom line) with the S pombe Mc protein (top line). The shaded boxes represent identical amino acids, while

unshaded boxed indicate conservative amino acid changes. The amino acid sequence comparison was performed on the PIR protein database. The rabbit sequence shares 45 amino acids out of 80 with the S.pombe Mc-protein. Within the conserved motif the human and rabbit Y-specific sequences share 64 out of 80 amino acids (80%) with a further 8 amino acids showing conservative changes (90% similarity overall) (Fig. 1). Outside the motif the human and rabbit show only 54% identity. This very high degree of homology strongly suggests pY53.3 is the coding sequence of a gene.

There is also partial homology between the human pY53.3 and rabbit amino acid sequence both 5' and 3' of the point where the Mc-S.Pombe homology is lost (Fig.6). The other Y-unique probes found in the original search were also sequenced. The pNB and pY53.2 probes did not reveal any potential open reading frames nor were they related to sequences in the EMBL or the PIR protein database. The probe pY4.1β is part of a larger 1.2 kb Rsal fragment which contains an open reading frame related to retroviral reverse transcriptase, commonly found in repetitive sequences. The probe pY53.1 encompasses a 5.6 kb region which contains several open reading frames, however none of these was predicted to encode a protein related to

sequences that are present in the EMBL sequence data base or the PIR protein database. In total 10.5 kb of the Y-chromosome were sequenced in the search for potential coding sequence. The sequence of pY53.3 has been deposited with EMBL.

Tissue distribution and expression

A Northern blot prepared with poly(A⁺) RNA human tissues [ovary, adult testis, lung (male) and kidney (male)] was hybridised with the 0.9 kb Hindi fragment of probe pY53.3 (Fig. 8).

RNA was prepared from each tissue [Goodfellow, P. J. et al.. Ann. Hum. Genet.. 53. 15-22 (1989)] and poly(A⁺) mRNA selected by polyATract isolation system (Promega). Poly(A⁺) RNA (8 μg) was separated on a 1% agarose gel containing 2.2M formaldehyde, transferred to Hybond N

(Amersham), UV fixed and hybridised with ³²P-labelled 0.9 kb Hindi fragment of PY53.3. Hybridisation was at 65°C in a buffer of 2 × SSC, 5 × Denhardt's solution, 200 μg/ml denatured salmon sperm DNA, 6% polyethylene glycol and 0.1% SDS. The filter was washed at 1 × SSC, 0.2% SDS at 65°C and autoradiography was for 6 days at -70ºC. Conditions for reprobing the filter with ß-actin were the same as above but the filter was washed at 0.1 x SSC, 0.2% SDS and 65°C and autoradiography was for 8 hours. The probe detects a fragment of approximately 1.1 kb in adult testis. No bands were detected in ovary, lung (male), kidney (male) nor in male and female lymphoblastoid cell lines.

Poly(A⁺) RNA was also prepared from three lymphoblastoid cell lines: (49 XYYYY cell line, Oxen) [Bishop et al , loc.cit.]: 46XY cell line, PGF) [Goodfellow, P. J. et al., Ann. Hum. Genet.. 53, 15-22 (1989)]; (46XX cell line, WT49) [DeKretser et al., loc.cit.] and probed as above. The probe detects a transcript of approximately 1.1 kb in adult testis and in no other tissue tested. Stripping and reprobing the filter with ß-actin confirmed the presence of poly(A⁺) RNA in the samples (Fig.8). This result is consistent with a testis specific transcript being encoded by the pY53.3 Y-specific sequence. In addition, 3' RACE (Rapid Amplification of cDNA Ends) PCR from adult testis poly(A⁺) RNA [Frohman, M. A., Dush, M. K. and Martin, G. R., Proc. Natl. Acad. Sci. USA. 85. 8998-9002 (1988)] showed the presence of a polyA tract 15 bases downstream from the potential polyadenylation signal, further indicating this as the 3' end of a Y-specific transcript.

Discussion

These experiments have defined a 35 kb region of Y-specific sequence immediately adjacent to the pseudoautosomal boundary in which TDF must reside. An extensive Southern blot analysis of this 35 kb of the Y chromosome revealed a Y-unique probe, pY53.3 which detects conserved Y-specific sequences in a wide range of eutherian mammals. In Example 2 below it is demonstrated that the equivalent murine sequence is present in Sxr' [McLaren, A. et al.. , Nature, 312. 552-555 (1984)] the smallest part of the mouse Y chromosome known to be male sex determining and is deleted from a mutant Y chromosome that has lost male sex

determining function. The conservation of pY53.3 related sequences between the Y chromosomes of eutherian mammals suggests that these sequences have a functional role. The location of the pY53.3 related sequences on the Y chromosomes of man and mouse is consistent with a role in male sex determination.

The nucleotide sequence of pY53.3 contains two open reading frames the second of which when translated is related to the Mc protein of the mating type locus of the yeast S. pombe. The Mc protein is 181 amino acids long, but the human sequence homology spans only the last 80 amino acids of the protein. Similarly, the rabbit and mouse Y-located sequences are homologous to Mc over this region as are the mouse autosomal cDNA sequences (Example 2). This striking homology is likely to represent a conserved protein motif referred to herein as the mt box (mating type box).

For purposes of discussion hereafter the human Y-located gene defined by pY53.3, is referred to as SRY (a gene from the Sex-determining Region of the χ-chromosome) and the equivalent mouse gene Sry (formerly the terms (mating type box v) MTY and Mtv were used).

The mating type locus, mat-1, in the fission yeast has two alternate alleles M and P. These alleles are transported during switch of mating type from either the donor loci mat3-M or mat2-P to the mat-1 locus. Both loci contain two transposable genes (Pc, Pi and Mc, Mi); none of the four genes are related to each other in sequence. The precise function of the four genes is not known, however Mc and Pc are required for mating and all four genes are needed for meiotic competence [Kelly et al., loc.cit.1. By analogy to the budding yeast it has been suggested that genes at the mat-1 locus may function as transcription factors. This suggestion has been supported by the finding of a diverged homeobox domain in Pi [Kelly et al., loc.cit. ] . It is tempting to speculate that Mc is also a transcription factor and that the mt box is a DNA binding domain. The only structural evidence to support this conjecture, however, is the high Arg/Lys content of both Mc and the mt box. A similar high (20%) Arg/Lys content is also found in the mt box of the SRY-related genes.

It is intriguing, but probably coincident, that Mc has a dual function in both mating and meiosis and that

transcripts of SRY have been found in adult testis, while in the mouse Sry is expressed in male genital ridge and adult testis.

The presence of a 3' stop codon and a polyA tract in cDNA from 3' RACE PCR implies that the open reading frame present in pY53.3 corresponds to the last exon of SRY. At the 5' end of the mt box sequence in SRY there is a

potential splice acceptor site however, as homology between the rabbit and human genomic sequences continues past this site, it may not represent an intron/exon boundary. This question will be resolved by isolating and sequencing transcripts of the SRY gene.

The 35 kb of Y-specific sequences immediately adjacent to the pseudoautosomal boundary are rich in repetitive

sequences and this has hampered analysis. The open reading frame present in pY53.3 is the only well conserved sequence detected. Example 2 Described here is the cloning of the sequence from the mouse Y chromosome corresponding to pY53.3 in humans.

The mouse gene contains an open reading frame homologous to that found in the human gene. The predicted protein product of these genes includes a domain characterized by a region of homology with several known or putative DNA-binding proteins, including human upstream binding factor (hUBF) [Janzen, loc.cit.] and the Mc protein [Kelly et al., loc.cit.], the product of one of the mating-type genes of the fission yeast Schizosaccharomvces pombe. Four

additional genes not linked to the Y chromosome, expressed at 8.5 d.p.c. in mouse, were found by their close homology within this protein domain. Genetic mapping places the Y- pecific sequence within the sex-determining region of the mouse Y chromosome. Further, the developmental time and tissue of expression of this gene is consistent with that expected for Sry.

Southern analysis

The sex-reversed mutation Sxr. has helped to define the position of Sry in the mouse. Sxr probably arose by translocation of the small short arm of the Y chromosome onto the pseudoautosomal region, located at the distal end of the long arm [Cattanach, B.M., Pollard, C.E. & Hawkes, S.G., Cvtoqenetics. 10. 318-337 (1971), Evans, E.P.,

Burtenshaw, M.D. & Cattanach, B.M. , Nature, 300. 443-445 (1982), McLaren, A. et al. , Proc. Natl. Acad. Sci. U.S.A.. 85, 6442-6445 (1988), Roberts, C. et al., Proc. Natl. Acad. Sci U.S.A.. 85, 6646-6649 (1988)]. Through obligatory crossover in the pseudoautosomal region during male meiosis this Sxr fragment may be transferred to the X chromosome, and resulting XX Sxr offspring are male. Apart from Sry. two other gene functions map to Sxr and consequently to the short arm: a gene involved in spermatogenesis termed Spy [Sutcliffe, M. J. & Burgoyne, P.S., Development. 107. 373- 380 (1989)], and the gene controlling expression of the H-Y minor histocompatibility antigen Hya [Simpson, E. et al., Immunogenetics. 13. 355-358 (1981)]. Sxr' is a variant of Sxr that retains Sry (XX Sxr' animals are male), but is deleted for Hva and Spy. Sxr' is therefore the minimum portion of the mouse Y chromosome known to contain Sry.

Mapping conserved sequences to the sex-determining region of the mouse Y chromosome. A Southern blot of EcoRI-digested genomic DNA from a XX Sxr' male (a mouse carrying the minimum portion of the Y chromosome that confers maleness), a XY* female (a mouse carrying a mutation in Sry), and a normal XY male and XX female mice, was hybridised consecutively to three separate probes as follows:

XY* female mice carry a strain 129 Y chromosome, the same strain as the XX and XY samples used here. Sxr' was maintained on a C57B6J background. Mouse clone 4.2.2 was isolated from a size selected library constructed from

EcoRI-digested strain 129 male spleen DNA. DNA in the size range 3.0-4.0 kb was recovered from an agarose gel on

Geneclean (Bio 101, Ine) ligated into lambda ZapII

(Stratagene), and packaged with Gigapack gold packaging extract (Stratagene) using manufacturer's protocols.

Plaque-forming units 0.5 × 10⁶ were directly screened without amplification, using a 0.9-kb Hindi fragment from pY53.3 as a probe and following previously described procedures (Maniatis). Positively hybridising lambda-clone 4.2.2 was plaque purified and the insert recovered in pBluescript by in vivo excision from the lambda ZapII vector as described by the manufacturer. A 380 bp Bg1ll-Pstl fragment was isolated from p4.2.2 and subcloned into a pBluescript (p422.04). This fragment contained all the homology to pY53.3 and was not repetitive.

For Southern blots, restricted DNA was electrophoresed through 0.7% agarose and transferred to Hybond N (Amersham) in 0.5 M sodium hydroxide, 1.5 M sodium chloride. Probes were labelled with ³²P added to the filter in 3 × SSC buffer, 0.1% sodium pyrophosphate, 5 × Denhardt's solution, 0.1% sodium dodecyl sulphate (SDS), 9% dextran sulphate and 50 μg ml^-1 denatured herring-testes DNA, and hybridised for 16 h at 65°C. The filter was washed in 0.1 × SSC, 0.1% SDS for mouse-specific probes, and 2 × SSC, 0.1% SDS for the human probe at 65°C and autoradiographed for 1-3 days then stripped in 0.1% SDS 100°C between experiments. The results are shown if Fig. 11 in which panel a shows the results of probing with a 1.0-kb Hindlll genomic fragment containing the zinc-finger domain of Zfy-1 [Koopman et al. , loc. cit.] which reveals bands corresponding to each of the four Zfy-related genes in XY male DNA. All of these genes are also seen in the XY female track. Zfy-1. but not Zfy- 2, maps to Sxr' in the XX Sxr' track but only the X-linked and autosomal members of this gene family (Zfa and Zfx) are present in a normal XX female.

In panel b, the probe was the sequence unique to the human sex-determining region, pY53.3 see [Example 1] , which hybridises strongly to a 3.5-kb band present in DNA from XY male and XX gxr' male mice but absent from DNA from XX female and XY* female mice. Additional weakly hybridising bands are present in all tracks, so cannot be Y-linked.

In panel c, a mouse Y chromosome-derived clone, p422.02 containing sequences homologous to pY53.3 was used and hybridised at high stringency only to the 3.5-kb Y-linked band. The numbers represent DNA size markers (kb).

When pY53.3, or a 0.9 kb Hindi fragment derived from it, was used as a probe on Southern blots of EcoRI-digested DNA from XY* male and XX female mice, a strongly hybridizing 3.5 kb male-specific band was detected, in addition to several weaker bands present in both sexes. The same 3.5-kb band was also present in DNA from XX Sxr' males (Fig. lib). Few sequences are known that map to the Sxr' region of the Y chromosome. Apart from Zfy-l [Nagamine, CM. et al. , Science, 243. 80-83 (1989), Mardon, G. et al..

Science. 243, 78-80 (1989)] (see Fig. 11a) that has a so far undefined role in male germ-cell development, only repetitive sequences, such as Bkm [Singh, L. Phillips, C & Jones K. W., Cell. 36., 111-120 (1984)] and Sx1 [Roberts et al.. loc.cit.1. map to this region. The probe pY53.3 therefore defines a new single-copy sequence in the minimum fragment of the Y chromosome shown to be sex-determining in both mice and humans.

Further data were obtained from a line of sex-reversed female mice carrying a mutant Y chromosome. A heritable mutation gives rise to such XY female mice. The mutation segregates exclusively with the Y chromosome amongst the offspring of these females. Moreover, because of frequent non-disjunction of the X and Y chromosome during female meiosis, the XY females often produce XXY daughters and XYY sons. The latter shows that the mutation can be

complemented by a normal Y chromosome. Further

characterization by karyotypic and Southern blot analysis and with a range of Y-specific DNA probes suggested that there had been no gross deletion or rearrangement of the Y chromosome carrying the mutation. There also seemed to be no loss of Y-specific gene functions apart from that of testis determination, as the mutation was able to be fully complemented by Sxr'. Because of the phenotype and deduced location of the mutation, it was concluded that it had occurred in Sry itself. The mutation is therefore referred to as Sry , and for simplicity we indicate the Y

chromosome carrying it as Y*.

It is therefore likely that there must be a molecular basis for the Sry^m1 mutation in any candidate sequence for Sry. Zfy-1, and its homologue Zfy-2. failed to satisfy this prediction as both genes have a normal structure and show a normal pattern of expression from the mutant Y*. These genes are present in XY female mice (Fig. 11a). When the same filter is probed with pY53.3, the 3.5 kb EcoRI male-specific band is absent (Fig. lib). Because there are no additional bands, which would be expected if this region were polymorphic, this result shows that the hybridizing sequences have been deleted in the XY* females. The deletion is, in fact, only 6kb in length and deletes all of the Sry open reading frame. The probe pY53.3 is the first probe capable of distinguishing the mutant Y* chromosome from the normal Y chromosome, and at least has to be considered as the closest marker to Sry. It is consistent with this that the sequences detected by pY53.3 are part of the testis-determining gene.

Cloning the mouse homologue of pY53.3

Several mouse genomic phage and cosmid libraries were screened with pY53.3. Although several clones were isolated, none of them contained the 3.5 kb EcoRI fragment or mapped to the Y chromosome and these probably correspond to some of the fainter bands seen on Southern blots in both male and female DNA (see below). The difficulty in cloning the Y* sequence may have been due to a general under-representation of Y-chromosomal DNA in genomic libraries or to a specific instability of the clone of interest. Sizeselected DNA was therefore used in a forced cloning

strategy to isolate the 3.5 kb EcoRI Y-specific fragment described above. Several clones of the expected size were obtained that hybridized strongly to pY53.3. One of these clones, p4.2.2, was analysed further and shown to contain a 380-base pair (bp) Bg1II-PstI restriction fragment (clone p422.04) that retained homology with pY53.3. When this 380-bp fragment was used as a probe at high stringency on genomic Southern blots it hybridized only to the Y-specific 3.5-kb EcoRI band (Fig. lie). Clone p4.2.2 therefore contains the mouse homologue of pY53.3.

Clone p4.2.2 was sequenced in the region of homology with pY53.3. Fig. 14 shows a comparison of the mouse and human Y-linked nucleotide sequences over 471-bp, in which the degree of homology is 62%.

Fig. 14 compares mouse and human Y-linked nucleotide sequences. The sequences of p4.2.2 (Mouse Y) and pY53.3 (Human Y) are shown on upper and lower lines respectively. Nucleotide homology is indicated by vertical bars. The single open reading frame in the mouse sequence is defined at the 5' end by a stop codon at nucleotides 14-16 (***). Within the open reading frame are two possible splice acceptor sites (###) and an in-frame translational start codon (Met). The region of amino-acid homology with the S. pombe gene encoding Mc is boxed. The nucleotide positions are numbered in blocks of 50. Of the six possible translation frames in p4.2.2, only one contained a long open reading frame in this region. This corresponds to one of the two open reading frames in pY53.3, and to an open reading frame in the corresponding Y-linked sequence from rabbit DNA [Example 1 above]. The high degree of evolutionary conservation between the open reading frame region of p4.2.2 and that of pY53.3 is a strong indication that it is part of a functional gene.

This gene is referred to as Sry.

The open reading frame in p4.2.2 contains near its 5' end two potential intron splice acceptor sites, and one inframe ATG codon in a reasonable context for translational initiation. [Kozak, M., Nucleic Acids Res.. 15. 8125-8148 (1987)]. Poor homology between the mouse and human

sequences 5' to the ATG codon is consistent with functional constraints being lost at this point, although no

corresponding ATG appears in the human sequence.

Similarly, the potential splice acceptor sites in the p4.2.2 sequence have no cognates in pY53.3 (Fig. 14).

There is evidence for the homologous region being the final exon of the human gene, [Example 1 above] and this is probably also the case in Sry. A gradual divergence of the mouse and human sequences is seen towards the 3' end of the region of homology, implying a diminished functional importance of this part of the gene. No possible splice donor motifs have been detected at this end, nor has the 3' limit of the open reading frame been defined.

The most striking feature of the open reading frame in p4.2.2 is a 237-bp sequence in which the degree of homology with pY53.3 rises to 80% (Fig. 14). The conceptual

translation of this sequence shows a similarity to the C+ terminal 80 amino-acid residues of the Mc protein [Kelly et al., loc.cit.]. This homology is remarkable given the evolutionary divergency between yeast and mammals, and indicates that these 237 bp define an important functional domain.

A deletion breakpoint in XY* females As the DNA fragment hybridizing to pY53.3, and the mouse homologue p4.2.2 are absent in XY* females,, it was

necessary to clone a larger genomic region from the mouse to search for the boundaries of this deletion. A library was constructed by partial digestion of unamplified genomic DNA of a strain 129 male mouse with Sau3a and ligation into a lambda FIX II vector (Stratagene) following

manufacturer's protocols. The library was plated using DL652, a bacterial strain that stabilises end to end repeats to overcome problems in cloning the mouse Y

homologue and screened with p422.04 as described

(Maniatis). From 10⁶ clones, one was obtained that showed very strong homology with p422.04. Restriction mapping confirmed that this clone, L7.4.1, contained the 3.5-kb

EcoRI band corresponding to 4.2.2 and to the mouse Y-linked locus. Fig. 12 shows an EcoRI (E) and Sad (S) restriction map of the insert from phage L7.4.1 using probes made from each of the EcoRI fragments contained within L7.4.1 and used to screen genomic blots of DNA from XY male, XY* female and XX female mice. In Fig. 12, L7.4.1. (solid line) spans 14 kb contiguous with genomic DNA detected by Southern Analysis but not present in L7.4.1 (dotted line). Open boxes show the location of the EcoRI fragments (A, B, C, D1, D2, E) used as probes and the small solid box is the conserved region. The stippled box indicates the limits of the region detectable in the XY* female and the large solid box, the colinear genomic region in XY males. The scale bar represents 1 kb.

Fig. 13 shows the results of probing Southern blots of

EcoRI-or Sad-digested genomic DNA from XY male, XY* female and XX female mice using the protocols described in

relation to Fig. 11. Probe A and a combination of probes Dl and D2 detect a Y-specific EcoRI band of the same size in an XY male and a XY* female. But a Sad digest reveals a difference in the size of the band detected by probe C, indicating a breakpoint within this genomic region in the XY* female.

Probe B, although found to be highly repetitive when used to probe a Southern blot of Sad-digested DNA, nevertheless gave the same result as probe C. The relevant bands are indicated in Fig. 13 by arrows and sizes are given in kb.

Three of the six probes (D1, D2 and A) failed to detect a Y-specific band in DNA derived from XY* females (Fig. 13). Note that probe A detects a 3.5-kb band in normal XY male in addition to the expected 1.5-kb band. This seems to be due to a cross-hybridizing sequence shared between probes A and E. Probe E also failed to detect either the 3.5-kb or the 1.5-kb band in DNA derived from XY* females. Unlike probes Dl, D2, A and E, probe C detects an EcoRI band of roughly the same size in DNA from both XY males and XY* females. The adjacent probe B, which corresponds to one end of the phage insert, similarly detects an EcoRI band of the same size in DNA from both XY males and XY* females. These results suggest that the breakpoint is close to the EcoRI site between probes C and Dl. This was confirmed by hybridizing Sacl-diαested DNA with probe C. This reveals a band of altered size in XY* females compared with XY males (also shown in Fig. 13). Expression of Sry in testis differentiation

Sry is thought to act around 11.5 d.p.c. to divert the supporting cell precursors in the urogenital ridge from the follicle cell pathway to the Sertoli cell pathway. A sensitive polymerase chain reaction (PCR)-based assay was used to look for the presence of transcripts from Sry in urogenital ridges isolated from male and female embryos. Oligonucleotide primers specific to Sry sequences in the conserved exon were used to amplify reverse transcribed RNA from 11.5 day urogenital ridges, adult testis and adult male liver. Total RNA (1μg) was added to a reverse transcription reaction in the presence of (+) or absence (-) of reverse transcriptase (RT). Subsequent PCR as

described [Koopman et al., loc.cit.] except using 30 cycles and an annealing temperature of 53°C used Sry-specific oligonucleotide primers (5' - 3') CTGTGTAAGATCTTCAATC and GTGGTGAGAGGCACAAGT and included Hprt primers as a control for the quality and quantity of RNA in each sample. The upper panel shows an ethidium bromide-stained agarose gel with 148 bp PCR products corresponding to Sry transcripts in adult testis and, less intensely, genital ridge (GR) of 11.5 d.p.c. male embryos. The band was absent from adult male liver and 11.5 d.p.c. female genital ridge samples. Sry primers were derived from a single exon and are

therefore capable of amplifying sequences from XY genomic DNA as well as from reverse transcripted Sry mRNA; the absence of visible product in the -RT samples confirms that any bands seen were due to the presence of Sry mRNA. Hprt primers spanning several exons do not amplify sequences from genomic DNA under the conditions used and yielded the expected 352-bp product in all +RT samples. The lower panel shows a corresponding Southern blot hybridised with probe p422.04. This confirms the Sry expression described above. In addition, a low level of genomic DNA

contamination was revealed in some RNA samples; this does not affect the interpretation of the results.

Independently isolated RNA samples have given similar results.

The results of the PCR are illustrated by Fig. 15 in which a Sry band is clearly seen in both adult testis and in male urogenital ridge, but not in liver or female urogenital ridge. The bands obtained are not from contaminating genomic or plasmid DNA in the samples, as control lanes in which reverse transcriptase was omitted show no signal.

These results confirm that Sry is indeed an expressed gene. Further, Sry is expressed in the urogenital ridge at a time consistent with it having a role in testis determination.

Isolation of homologous cDNAs

Several complementary DNA libraries were screened,

initially with pY53.3 and more recently with p422.04, to find cDNAs corresponding to Sry. In particular, the screening of two libraries prepared from 11.5 d.p.c. male genital ridges, a stage at which the testis-determining gene is thought to be expressed, failed to reveal any hybridizing phages; however, both libraries are very small [Cunliffe, V. et al., EMBO J.. 9, 197-205 (1990)]. The screen was therefore extended to 14.5 d.p.c. and adult testis libraries and to a 8.5 d.p.c. whole-mouse embryo library. A 8.5 d.p.c. mouse total-embryo cDNA library

[Fahrner et al., loc.cit.] was screened with the HincII fragment of pY53.3. Fourteen positive clones were purified and inserts subcloned into pBluescript (Stratagene).

Restriction maps and Southern analysis of these clones showed that they correspond to four different loci not linked to the Y chromosome. For each of these messengers the region that hybridised with pY53.3 was subcloned into pBluescript and sequenced as double stranded DNA using T7 polymerase (Pharmacia) according to the manufacturer's instructions. Sequence analysis and calculations were performed using the University of Wisconsin Genetics

Computer Group Sequence Analysis Software Package

[Devereux, J., Haeberli, P. & Smithies, O., Nucleic Acids Res., 12: 387-395 (1984)]. The amino acid sequences encoded by these four subclones as shown in Fig. 16

together with corresponding sequences of related proteins. Although the testis libraries failed, to yield any

recombinants with homology to the probe, the 8.5. d.p.c. library produced several positive clones. Detailed

restriction mapping indicated that these clones fall into several discrete classes, but their individual use as probes on genomic Southern blots of male and female DNA indicated that none of them were Y-linked. These cDNAs are provisionally referred to as corresponding to autosomal loci, although the possibility that they are X-linked has not been excluded. In each of these cDNAs a region was defined that showed strong hybridization to both the mouse and human probes and this region was therefore sequenced. Fig. 16 shows the sequence of various homologous protein domains (single-letter code) as follows: ubf-hmg3, the third HMG box of hUBF [Jantzen, loc.cit.], which shares the highest homology with this motif; pombe-mc, mating type protein Mc from S. pombe [Kelly et al., loc.cit. ] : human-y, SRY: mouse-y, Sry: and mouse-al, mouse-a2, mouse-a3, mouse- a4, four different mouse cDNAs from genes not linked to the Y chromosome. Residues that are absolutely or strongly conserved between the upper six sequences are enclosed by shaded boxes and where this conservation extends to the S. pombe Mc or hUBF-HMG proteins, residues of these are also marked by shaded boxes. Residues that are Y-specific in mouse and man (and rabbit per Example 1) are shown in open boxes. The amino-acid positions are numbered in blocks of 50 according to the human sequence. The region of

conservation defining a new protein motif extends from residue 11 to residue 90 in the upper sequences. Fig. 17 summarises the percentage of amino-acid homology, within the conserved motif, between the different

sequences. The unboxed area emphasises the homology between the Sry-related sequences and both the S. pombe Mc sequence and the HMG motif, using the third HMG box of hUBF as an example. Shaded boxes emphasise the homology between the mouse a-1, a-2 and a-3 sequences as well as the lower degree of homology of the mouse a-4 sequence to the other members of the gene family. Sequence comparison in this manner has results in a new 80 amino-acid motif being defined. This has a high ratio of basic residues (up to 25%) and a strong helical content (Fig. 16). Conservation is strong in the motif, with 42 amino acids identical between the Sry-related genes. The amino-acid comparison reveals homology of this motif not only to the S. pombe Mc protein, but also to a domain present in four copies in hUBF [Kelly et al., loc.cit.]. This domain, termed the HMG box because of its homology to the high mobility group proteins HMGl and HMG2 [Einck, L. & Bustin, M., Expl. Cell Res.. 156. 295-310 (1985)], has been associated with binding to the upstream control element of the ribosomal RNA gene promoter to activate transcription. Clearly the motif present in the Sry-related genes is similar to, but distinct from, the HMG-box, thus defining a new family of genes. Five members of this family have been isolated from the mouse but the number of bands seen when probing a Southern blot at low stringency with either pY53.3 or any of the cDNAs suggests that not all of the family has been cloned.

Certain amino-acid residues are identical throughout the Sry-related gene family and the Mc and HMG sequences (Fig. 16). This may be due to protein structural or sequence recognition constraints common to all genes of this type. Within the mouse gene family, at least three subfamilies can be distinguished (Fig. 17) . The sequences al, a2 and a3 are almost identical in the conserved motif; the Sry and a4 sequences seem to have diverged independently from this group. Several residues are common to only the Y-linked sequences of the human, mouse and rabbit. The Y-linked gene products may therefore be functionally distinct from the autosomal gene products.

Discussion

Any candidate sequence for the testis-determining gene has to satisfy several criteria. In this Example and Example 1 above the isolation and sequence of part of a gene that would fulfil these criteria is described. First, it should be a sequence conserved on the Y chromosomes of all mammals shown to have a Y-chromosomal sex-determining mechanism. Southern blotting revealed a homologue of this sequence was present on the Y chromosome of a wide range of eutherian mammals. Second, it must map to the minimal portion of the Y chromosome shown to confer maleness. The human gene, SRY. was found in a detailed search of the 3.5 kb adjacent to the pseudoautosomal boundary, which is sufficient to confer maleness on an XX chromosomal

background. It is shown here that the mouse homologue, Sry. maps to Sxr'-the smallest region of the Y chromosome known to be sex determining. Third, a candidate must satisfy the prediction that there is a molecular basis for the occurrence of XY* females shown genetically to have a mutation in the testis-determining gene. It is shown here that part of Sry_r including the highly conserved exon, is deleted in these Sry XY* female mice. The finding of one deletion breakpoint close to the conserved exon is again consistent with Sry being the testis determining gene although this argument would clearly be strengthened if the other breakpoint were also found to be in or near the gene. In a previous study, no alteration was found in either the structure or expression of Zfy-1. This is the only other gene known to map Sxr' . and had previously been considered a good candidate for the testis determining gene.

Finally, any candidate for testis determining gene should be expressed in a tissue and at a time consistent with what is known of its action. In mice the gene should be

expressed in the urogenital ridge at about 11.5 d.p.c, just before testis-cord formation, the first recognizable difference in male and female embryogenesis. The

expression of Sry conforms with this prediction. The level of transcripts found in 11.5 d.p.c. urogenital ridge is very low, and preliminary PCR data also suggests that Sry is not expressed before 10.5 d.p.c. or during late stages of fetal gonad development. This low level of expression is not surprising for a gene possibly operating as a switch in development. Sry was also found to be expressed in adult testis (see Example 1). Several genes suspected to have roles in development decisions in the embryo, for example, some homeobox-containing genes, also show expression in adult testis. Sry could have one role in the embryo in testis determination, and another postnatally in male germ-cell development. It is not yet clear what the mode of action of Sry might be in testis determination. Evidence suggests that Sry

promotes differentiation of Sertoli cells in a cellautonomous manner. Further steps in male development follow directly from the differentiation of Sertoli cells. Sry might, therefore, be expected to encode a protein capable of acting as a regulatory molecule in the cell.

The conserved protein domain in Sry is homologous to known DNA-binding proteins and, by analogy, the Sry protein could be a nuclear protein that binds DNA and acts as a

transcriptional switch. The homology of Sry with the gene encoding the Mc protein at the mat3-M locus of S. pombe is intriguing. It is tempting to draw parallels between mating type in yeast and sex determination in mammals, although there is no reason to suppose any evolutionary connection between the mechanisms is employed in the two cases. It is more likely that the conserved motif

represents a functional protein domain that is appropriate for the type of control necessary to achieve the switch. In attempts to isolate cDNAs corresponding to transcripts of Sry, at least four additional genes were discovered that contain the 80 amino-acid region of homology. The high degree of conservation of this protein box, both between the different cDNAs and between the Y-linked genes in different species, implies that it is a functional domain. All four of the autosomal genes were isolated from an 8.5 d.p.c. embryo library, and could be involved in other early developmental decisions.

In conclusion, a novel gene family in mice is described, linked by the presence of a conserved amino-acid domain also found in a gene involved in mating type of S. pombe. and in the DNA-binding protein hUBF. One member of this gene family maps to the sex-determining region of the Y chromosome and satisfies various predictions made for the testis-determining gene.

It is believed that the overall process of sex

determination is likely to be conserved in all vertebrates and that the testis determining gene will play a central role even in those species which do not use a XX/XY chromosome-based system for sex determination. In all species of vertebrates the testis determining gene will be a target for experiments to ascertain the sex of cells, tissues, embryos, foetuses, sperm and individuals and to control the sex ratio of the progeny of breeding animals.

EXAMPLE 3

There is compelling evidence to implicate SRY (or Sry in mice) in the process of testis determination. It satisfies all predictions that can be made concerning the location of the testis determining gene on the Y chromosome. Thus the gene maps to the smallest region of the human and mouse Y chromosomes known to be male determining, and is conserved on the Y chromosome of all other eutherian mammals tested. The gene encodes a putative DNA binding protein, consistent with a regulatory role. Sry also shows a pattern of expression in the mouse entirely consistent with a role in testis determination, being expressed for a short period just prior to overt testis differentiation, specifically in somatic cells of the genital ridge.

Data obtained from the study of cases of sex reversal involving XY females strongly argue that SRY/Sry is

necessary for normal testis determination. Thus, a line of mice shown genetically to be mutant in the testis

determining gene is deleted for Sry. Also, through

examination of the SRY DNA sequences of a random sample of human XY females, a number of mutations have been

identified that would alter the gene product within the highly conserved putative DNA binding domain.

This leaves one final question to be answered to provide formal proof that SRY/Sry is the testis determining gene, namely, is the gene sufficient? This question needs a qualification because it is clear that the gene cannot act alone. XY females or hermaphrodites are found in some situations despite the presence of a normal testis

determining gene. Some of these cases are clearly due to mutations in non-Y-linked genes, and others reflect an incompatibility between Y chromosome variants and

particular alleles at autosomal loci. The question can therefore be rephrased as: is SRY the only Y-linked gene necessary for testis determination?

The minimum region of the human Y chromosome that gives testicular development is the approximately 35 kb region found in the four XX individuals showing sex reversal, that were instrumental in finding SRY. However, even though no other conserved sequences were found, it is conceivable that there is another gene within this region required for testis determination. Furthermore, unlike other XX males with much larger fragments of the Y chromosome, none of the four individuals had a completely normal male phenotype, which might argue that an additional gene outside the 35 kb is also required.

The best test of the function of SRY/Sry is to introduce it alone into XX embryos, and to see if male development ensues, i.e. if it gives sex reversal. Ideally, to be certain that no other gene is present within the introduced DNA, this should be done using full length cDNA clones together with a defined heterologous promoter. However, the pattern of Sry expression observed during foetal gonad development in the mouse suggests that precise regulation of the gene may be critical for its action. Therefore, in the hope of having appropriate regulation we have initiated transgenic experiments using genomic DNA fragments carrying either the human or the mouse SRY/Sry gene. We find that while the human gene fails to give sex reversal in mice, the mouse gene contained within a 14 kb genomic fragment gives apparently normal testis development in chromosomally female transgenic mice, as can be seen at both embryonic and adult stages. MATERIALS AND METHODS

Production of transgenics

DNA constructs

SRY or Sry containing fragments were isolated from cosmid, phage or plasmid clones by digestion with appropriate restriction enzymes and then purified from other fragments and vector sequences by agarose gel electrophoresis. The isolated DNA fragments were cleaned either by "gene clean" according to manufacturers instructions, or by phenol extraction and by passing down a sephadex G50 column, or an elutip, followed by ethanol precipitation.

1. Human SRY

A 24.6 kb BamH1 - Sal1 fragment was isolated from the human cosmid cAMF. This contains 18.6 kb of Y-specific sequences adjacent to the pseudoautosomal region, plus 6 kb of sequences distal to the boundary. This fragment therefore includes SRY and 12 kb of sequences 5' to the known exon (see Fig. 20), and is subsequently referred to as HuSRY-A.

2. Mouse Sry

A 14 kb fragment containing Sry. with about 8 kb of sequences 5' to the known exon, was isolated either directly from the phage clone L7.4.1 or from a subclone in pBluescript, by digestion with Sail. This fragment is referred to as MoSry-741.

A 3.5 kb EcoRI fragment containing Sry, with about 1.7 kb 5' and 1 kb 3' to the known exon, was isolated from p4.2.2. Mice and micromanipulation

Three to five-week old (CBA × C57BL/10) FI females were superovulated and mated to FI stud males. The following day, fertilised eggs were collected from oviducts of females showing vaginal plugs. Pronuclei were

microinjected with 1-2 pl of DNA at a concentration of approximately 2 μg/ml, essentially as described by Hogan, Constantini & Lacy, "Manipulating the Mouse Embryo", Cold Spring Harbor, 1986. The eggs were cultured overnight in M16 or T6 medium, and 2-cell embryos implanted by oviduct transfer into day of plug pseudopregnant recipients.

Injected embryos were either examined at 14 days post transfer, or allowed to go to term. 14 day embryos were analysed in the following way: (i) Chromosomal sex was initially determined by staining for sex chromatin in amnion cells (Monk and McLaren, 1981). (ii) Gonadal sex was determined by examining for the presence or absence of testis cords, which are normally very distinct at this stage. Gonads were subsequently fixed, photographed and those of interest processed for histology. (iii) Embryos showing apparent sex reversal were karyotyped from cultures of skin fibroblasts. (iv) Genomic DNA was prepared from limbs. (v) the remainder of the carcass was frozen and stored at -70°C. Newborn animals were sexed by external examination and tail biopsies taken for DNA preparation at 2-3 weeks. Animals shown to be transgenic were set up to mate and/or sacrificed for internal examination. For RNA preparations, transgenic embryos were obtained from timed matings of transgenic males with F1 females.

Developmental stage was verified by examination of limb morphology.

RESULTS Transgenic mice with human SRY

1. One definite XX transgenic 14 day embryo was obtained. The gonads showed no signs of testis cord formation.

2. Three liveborn XY transgenics, HuSRY-A6, HuSRY-A9 and HuSRY-A12 were obtained. All were normal looking males. One of these, A12, failed to breed. The other two have both been fertile and have transmitted the transgene to approximately 50% of their offspring. At least 4 XX transgenics have been examined in detail from each line, but no evidence of sex reversal has been seen, either in 13.5 dpc embryos, or in adults. Transgenic mice with mouse Sry

(a) Sry isolated from phage clone L 7.4.1.

1. Two definite XX transgenic 14 day embryos, MoSry-741/7.22 and MoSry-741/10.2 were obtained. Both of these had normal looking testes, with testis cords throughout the gonads. No female embryos have been found that carry the transgene.

2. One liveborn XX transgenic, MoSry-741/33.13 was obtained which appeared to be a normal male by examination of external genitalia. 3. One liveborn XY transgenic, MoSry-741/AAA, normal male was obtained. This was a low grade mosaic for the

transgene, and although fertile, has so far failed to transmit the transgene to offspring.

4. One liveborn XY transgenic, MoSry-741/32.7 was obtained which was used in breeding studies, so far without

successful transmission of the transgene to the offspring.

5. Two liveborn XX transgenic mice, MoSry-741/32.10 and MoSry-741/33.2 were obtained which appeared to be normal females from examination of external genitalia, but seemed to be somewhat smaller than non-transgenic littermates.

The rate of transgenesis with Sry from phage clone L 7.4.1 has been very low compared with other DNA fragments, despite trying different methods of preparation. This may be due to the unusual structure of this genomic fragment which contains an inverted repeat flanking the Sry gene.

(b) Sry isolated from p4.2.2.

1. One definite XX transgenic 14 day embryo, MoSry-422/BBB was obtained but there was no evidence of sex reversal. 2. Two definite XX transgenic adults, MoSry-422/CCC and MoSry-422/DDD were obtained but there was no evidence of sex reversal in either case.

3. A number of XY transgenics have also been identified. These appear to be normal males. DISCUSSION

These experiments show that a 14kb genomic fragment containing Sry is sufficient to direct the formation of testes in XX transgenic embryos and subsequently to give rise to full phenotypic sex reversal in an XX transgenic adult. While previous data have shown that SRY/Sry is normally required for testis determination, the current study indicates that Sry is also sufficient to direct this process in an otherwise female genetic background. As Sry is the only Y-encoded sequence required to give rise to male development it is concluded that Sry is the testis determining gene.

Given that complete sex-reversal in mice is possible when Sry is the only Y-derived genetic material present, how do these results relate to the analogous situation in human XX individuals that have only inherited 35kb of Y-specific material containing SRY? In four such individuals

described by Palmer et al., the sex reversed phenotype was only partial and varied between individuals. This is in contrast to previously described XX males whose Y-derived components were much larger, which display a completely normal male phenotype. One interpretation of these

observations is that an additional gene (or genes) outside the 35kb region containing SRY may be required to attain the full male phenotype. This now seems unlikely

considering the fact that Sry, in the absence of any other Y-specific genes, can by itself cause complete sex reversal in chromosomally female mice. An alternative explanation for the partially sex-reversed phenotypes in these

individuals is that the expression of SRY is being affected by its new location. Through abnormal X-Y interchange during male meiosis, the 35kb of Y-unique DNA carrying SRY is translocated onto one of the X chromosomes. Here, SRY may be subjected to position effects, which may alter its normal level of expression in a manner which would depend on its exact position on the X chromosome and the amount of Y-specific DNA which had been translocated. The phenomenon of position effect variegation which has been described in detail in drosophila provides ample evidence for position dependant effects on gene expression. In addition to the effect described above, SRY in these individuals may be affected by a spreading of X-inactivation. In some cases involving X:autosome translocations it has been shown that X-inactivation can spread into adjacent sequences and affect expression of autosomal genes. There is a precedent for X-inactivation affecting testis determination in the mouse; when Sxr, a region of the Y chromosome including the testis determining gene is made to be carried only on the inactive X chromosome, resulting XX Sxr animals are

frequently hermaphrodites or females rather than males. In human cases, because X-inactivation is random, and occurs within few cells in the early embryo, the proportion of cells with an inactive X chromosome carrying SRY may vary between individuals. If most cells within the developing gonad contained an inactive X carrying SRY then an ovary or ovotestis may have formed. Amongst the four patients the two extremes of phenotype were represented by sibs despite the fact that they both have an identical portion of Y chromosome arising from the same abnormal X:Y interchange. This is most easily interpreted as a variation in SRY

activity due to X-inactivation. It can not be due to missing genes. As both position effects and spreading of X-inactivation may be more pronounced when SRY is carried on only a short piece of DNA, these effects explain why only XX males which have inherited a minimal portion of the Y chromosome show partial sex reversal.

The transgenic data presented here may represent a mouse model for the partial sex reversal of human XX males described above. Two of the XX transgenic mice described were apparently normal females. The simplest explanation for the failure of Sry to cause sex reversal in these cases is that the transgene had integrated into a chromosomal position in which its normal expression may have been altered or completely shut off. An analysis of Sry

expression in XX transgenic embryos of the two mouse lines involved may confirm this possibility. Alternatively, the transgenes in these cases may have integrated into an X chromosome and so be subject to random X-inactivation. If this is the case then a sex-reversed or partially sex-reversed phenotype may become apparent in offspring of the founder transgenics, in which the proportions of cells where Sry is carried on an inactive X chromosome are fewer, or when mice are bred which are homozygous for the

transgene. Another possibility for the failure of Sry to act in these particular mice may be due to a timing mismatch. The time of onset of expression of the testis determining gene is thought to be critical for its successful action. The failure of the Mus musculus poschiavinus Y chromosome to cause male development when it is on a C57BL background has been explained by the presence of a late acting allele of the testis determining gene on this particular domesticus type Y chromosome, coupled with an early acting signal for ovary determination associated with the C57BL background, which preempts the action of the testis determining gene in these cases. A timing mismatch could occur if a position effect were to cause a delay in the time of onset of Sry expression during embryogenesis. Thus it will be important to study the time course of Sry expression during

development in lines of these transgenics, if the gene is expressed at all. As the animals used in these experiments were from a cross between (CBA × C57BL)F1 mice, it is possible that the founder transgenics that failed to sex reverse may have been, by chance, homozygous for the particular autosomal alleles involved in the early acting ovary determination signal associated with the C57BL background. This effect coupled with a delay in expression of the transgene could explain the failure of male

development in these mice. If this is the case then progressive breeding of the transgene away from this background may reveal its activity. It is also possible that integration of the transgene may have disrupted a locus in such a way as to slow down the development of the somatic portion of the genital ridge, where the testis determining gene must act, relative to the development of the germ line, in which the ovary determining signal is thought to act. This would also have the effect of

creating a timing mismatch. The concept of sex reversal caused by such a retarding effect has been postulated by Cattanach. A final possibility is that the transgene no longer functions because it has become rearranged during integration. A detailed Southern analysis will reveal any gross rearrangements.

These experiments also indicate that human SRY may not be able to function in mice. There are at least three

possible explanations for the failure to promote testis formation: (i) SRY alone is not sufficient. (ii) The SRY gene is not being expressed at all or is expressed at an inappropriate time. This may be because all necessary control elements were not present in the DNA injected; or because endogenous factors required for Sry gene expression fail to interact with the human SRY promotor elements. An investigation is in progress looking for transcription from the human SRY gene to see if it is regulated correctly in the mouse genital ridge and in the adult testis. (iii) The differences between the human SRY and the mouse Sry

proteins are significant, such that the former fails to correctly interact with downstream genes. The data from Berta et al- suggests that even an apparently conservative single amino acid alteration in the SRY protein is

sufficient to prevent its action, giving rise to an XY female human individual. Thus it would not be surprising if the human sequence has diverged sufficiently to

compromise its function in mouse. It would be interesting to test this hypothesis by exchanging the human and mouse open reading frames.

The ability of a 14kb fragment containing Sry to cause sex reversal suggests that this fragment contains the entire Sry gene, including the regulatory elements required for appropriate embryonic expression. Koopman et al. have recently described that Sry expression occurs in a tight window around the time of genital ridge differentiation and is only expressed in the somatic component of the genital ridge. The transgenic system can be used to further localise Sry regulatory elements within the 14kb fragment, by microinjecting smaller Sry containing constructs.

Initial attempts to do this have begun with the use of the 3.5 kb genomic fragment from p4.2.2 (hereafter 422).

Assuming that it contains the whole structural portion of the gene (it is possible that it is missing an untranslated exon, either at the 5' or 3' end, as full length cDNA clones have not yet been obtained), the most likely

explanation for why this fragment does not cause testis development is that appropriate cis-acting elements are missing. Alternatively, such a short fragment as 422 may be very susceptible to integration site effects on

expression. If so, transgenics in which the insertion is in a favourable location may show sex reversal. It is also formally possible that Sry is not sufficient, and there is another gene present in 7.4.1. This is highly unlikely, given that it is only 14 kb long and given the lack of other conserved elements between human and mouse besides the Sry open reading frame. Sry expression from the transgene in 11.5 dpc XX embryos transgenic for 422 will be investigated. The only other site of Sry expression is in adult testis, probably in the germ cell component. It will be

interesting to see if this 14kb fragment also contains the regulatory information required for the switch from

expression in the somatic part of the embryonic gonad to expression in adult testis associated with germ cells. As XX transgenic males lack germ cells, these mice will not be useful for analysis of transgene expression in the adult. However it should be possible, by breeding, to obtain an XY transgenic. The Y chromosome appears to be normal in all respects except that it is deleted for Sry and is therefore no longer male determining. Thus assuming that the

transgene complements this defect, producing males, it will be possible to examine transgene expression in the adult testes of these mice, in which normal germ cell development should occur.

EXAMPLE 4

The testis determining factor is encoded by a Y chromosome gene responsible for initiating male sex determination. SRY is a transcribed gene located in the sex determining region of the human Y chromosome. If SRY is the testis determining gene, responsible for initiating male sex determination, it would be predicted that some sex-reversed XY females will have suffered mutations in this gene. Human XY females and normal XY males were tested for alterations in SRY using the single strand conformation polymorphism assay (SSCP) [Orita, M., Suzuki, Y., Sekiya, T. and Hayashi, K., Genomics. 5; 874-879 (1989); Orita, M., Iwahana, H., Kanazawa, H., Hayashi, K, and Sekiya, T.,

Proc. Natl. Acad. Sci. USA.. 86; 2766-2270 (1989)] and subsequent DNA sequencing. A de novo mutation was found in the SRY gene of one XY female: this mutation was not present in the patient's normal father and brother. A second variant was found in the SRY gene of another XY female, however, in this family the father shared the same alteration. The variant in the familial case may be fortuitously associated with the family or predisposing towards sex reversal; the de novo mutation associated with sex reversal provides compelling evidence that SRY is required for male sex determination.

The minimum portion of the human Y chromosome known to be sex determining is 35 kb of Y-specific sequence located immediately adjacent to the pseudoautosomal boundary.

Within this region several open reading frames have been found; the largest of these defines the gene SRY. An homologous gene, Sry. is present in the sex determining region of the mouse Y chromosome and is deleted from a mutant Y chromosome that is no longer sex determining. The SRY gene has, in addition, many of the properties expected for the testis determining gene including a Y chromosome location in all eutherian mammals tested and is expressed in the somatic cells of the mouse genital ridge immediately prior to testis formation. Formal proof that SRY is the testis determining gene can be obtained by showing that mutations in SRY affect sex determination. In this study we have investigated gRY in sex reversed human XY females.

Sex reversal in XY females results from the failure of the testis determination and differentiation pathways. XY females with gonadal dysgenesis can occur sporadically or, more rarely, in familial clusters [Nazareth, M.R.S. et al., Am. J. Med. Genet.. 2; 149-154 (1979); Simpson, J.L.

Blagowidow, N. and Martin, A., Hum. Genet.. 58: 91-97

(1981)]. Some XY females have lost the sex determining region by a terminal exchange between the sex chromosomes [Levilliers, K., Quack, B., Weissenbach, J. and Petit, C., Proc. Natl. Acad. Sci. USA.. 86; 2296-2300 (1989)] or by other deletions [Page, D.C., Fisher, E.M.C., McGillivary, B. and Brown, L.G., Nature. 346: 279-281 (1990)]. The sporadic, non-deletion cases have been interpreted as possible mutations at the testis determining gene and the familial cases as autosomal or X linked mutations in genes that interact directly or indirectly with the testis determining gene [German, J., et al., Science. 202: F3-56 (1978)]. SRY sequences were amplified by PCR from a collection of sporadic and familial cases of XY females, as well as normal male controls, using the primers XES7 and XES2 located within the SRY open reading frame, amplifying a 609 bp fragment. The primer sequences are:

XES7, 5'CCCGAATTCGACAATGCAATCATATGCTTCTGC-3'; XES2, 5'- CTGTAGCGGTCCCGTTGCTGCGGTG-3 ' . PCRs were performed with approximately 100 ng of genomic DNA, 200 μM each dNTP, 0.5 μM each primer, 1.5 mM MgCl₂, 10 mM Tris (pH 8.3), 50mM KCl, 0.01% (w/v) gelatin, 0.25 U of Taq polymerase and 0.5 μl of [α- 32P]dCTP (3000 Ci/mmol, 10mCi/ml) in a volume of

10 μl. Reactions were cycled for 1.2 min at 94°C, 1.2 min at 65ºC and 2 min at 72°C for 35 cycles.

1 μl of the product was digested with Hinfl and TaqI in the presence of 4 mM spermidine hydrochloride in a 10 μl volume. The digested DNA was diluted 1:10 in 0.1% SDS, 10mM EDTA, followed by a 1:2 dilution in 95% formamide, 20 mM EDTA, 0.05% bromophenol blue, 0.05% xylene cyanol.

Samples of 1-3 μl were heated at 80°C for 5 min to denature the DNA, then loaded onto 6% acrylamide, 10% glycerol nondenaturing gels using a sequencing-gel apparatus.

Electrophoresis was carried out at 25 mA, with a fan heater set on cold directed at the gel as a cooling device.

Autoradiography of the dried gels was for 3 days without an intensifying screen, to detect single strand conformational polymorphisms (SSCP) . In total 11 XY females were tested by SSCP. These individuals: PF, JA, AM, RB, GM 2598, AS, M, AA, ID, JN and KL all have gonadal dysgenesis, are karyotypically normal and positive both for the Y-pseudoautosomal boundary and ZFY. AA and JN both have pure gonadal dysgenesis. Fifty normal males were tested as controls and no variation detected. Thus three patterns were detected: one pattern common to all of 50 normal male controls and the majority of patient samples; a second pattern for XY female individual (AA) and a third pattern for XY female (JN). The father and normal brother of (AA) were tested and found to give the same pattern as the controls, implying that (AA) has suffered a de novo

mutation. In contrast, DNA from father (N) and his XY daughter (JN) showed the same pattern. The primers XES-7 and XES-2 were used to amplify a 609 bp region. The PCR products were subcloned in pUClδ vectors (NEB) and

sequenced as double-stranded DNA by the dideoxy chain termination method 17 using synthetic oligonucleotide primers and Sequenase (USB). The entire amplification product was sequenced, however, only the small region containing the changes in the XY females is displayed in

Fig 21 where the open reading frame of the genomic clone pY53.3 (SRY) is shown extending from 354 bp to 1,022 bp.

The conserved motif which encodes a potential DNA binding protein extends from 582 bp to 821 bp. The dotted line indicates the location within the open reading frame of the nucleotide sequence shown. The top sequence line of

Fig. 21 shows the nucleotide sequence for XY female (JN) with base changes from G>C indicated by arrow. The middle line shows the normal male SRY nucleotide sequence. The bottom line shows XY female (AA) with a base change from G>A indicated by arrow. In Fig. 22 the amino acid sequence of SRY for the human-Y, rabbit-Y, mouse-Y and mouse

autosomal SRY-like genes a₁, a₂, a₃, a₄ is shown. Boxed shaded regions show identical amino acids conserved across these species. In the XY female (JN) the variant causes a change from valine to leucine whereas in the XY female (AA) the mutation causes a change from methionine to isoleucine. (Fig. 21) . As predicted by the SSCP assay, (AA) has a de novo mutation (G>A) that is not shared with her father and brother. In the familial case, a variant (G>C) was shared by (JN) and her father. Other than these observed base changes, the XY females (JN) and (AA) had complete sequence identity to SRY over the amplified 609 bp. Paternity in both families was confirmed by Southern blotting with minisatellite probes [Wong, Z., Wilson, V., Patel, I., Povey, S. and Jeffreys, A.J., Ann. Hum. Genet.. 51: 269-288 (1978); Wong, Z., Wilson, V., Jeffreys, A.J. and Thein, S.L., Nuc. Acids Res.. 14: 4605-4616 (1986)]. The de novo mutation in (AA) causes a conservative change from methionine to isoleucine at a residue which lies within the putative DNA binding motif at SRY and is

identical in all SRY and SRY related genes (Fig. 22). The association of a de novo mutation with a new phenotype provides evidence that the phenotype and mutation are related, and by inference that SRY is required for male sex determination. Formally this does not exclude the

possibility that other genes on the Y chromosome are required for sex determination, however, previous work suggests that, if they exist, these genes must be located adjacent to the psuedautosomal boundary, close to SRY

[Palmer, M.S. et al., Nature. 342: 937-939 (1989)].

The variant found in the familial case causes a

conservative change from valine to leucine (Fig. 22) . This residue is conserved amongst SRY and SRY related sequences with the exception of a mouse autosomal gene (autosomal-4) which has a conservative change to isoleucine at this position. There are several possible explanations for the variant found in familial cases. First, the variant could cause conditional sex reversal depending on other genetic or environmental factors. A similar observation has been described in the mouse where the ability of some alleles of the testis determining gene to induce testis formation depends on the genetic background [Eicher, E.M. and Washburn, L.L., Ann. Rev. Genet.. 20; 327-360 (1986)].

Second, the variant could be fortuitously found in a family segregating for an autosomal or X-linked sex reversing gene. Finally, the variant could cause sex reversal and the father would be mosaic for wild type and variant sequences.

The SRY genes of the majority of the XY females tested appear normal by the SSCP assay. It is possible that these individuals have mutations in SRY that are not detected by the assay either because they do not cause a band shift or because they fall outside the region tested.

Alternatively, these individuals may have mutations in another part of the sex determining pathway.

In conclusion, a de novo mutation in the gene SRY is associated with sex reversal in an XY female. This

provides compelling evidence that SRY is required for testis formation and male sex determination.

EXAMPLE 5

Sex reversal of transgenic mouse embryos Fertilised eggs were microinjected with Sry gene sequences, transferred to pseudopregnant recipients, and a proportion of the resulting embryos were analysed at 14 days post transfer, rather than allowing them all to develop to term. The first visible sign of testis development from the genital ridge is the formation of testis cords at about 12.4 dpc in the mouse. This is due to the differentiation of Sertoli cells and their alignment into the epithelial structures surrounding the germ cells [Jost, A. & Marge, S. Phil. Trans. R. Soc. Lond.. 322. 55-61 (1988)]. Cord formation confers a characteristic striped appearance to the developing testis, distinguishing it from the fetal ovary. Other morphological changes characteristic of the testis are its rapid growth and prominent vasculature.

Examination of fetuses at about 14 days after oviduct transfer at a time when the testis cord formation is obvious even in cases of partial sex reversal [Eicher, E.M. Phil. Trans. R. Soc. Lond, 322. 109-118 (1988) and Eicher, E.M. et al., Cvtooenet. Cell. Genet.. 28. 104-115 (1980)], is a rapid assay of phenotypic sex. An indication of chromosomal sex was obtained by staining for sex chromatin in amnion cells [Monk, M & McLaren, A., J. Embryol. Exp. Morph.. 63. 75-84 (1981)]. Where necessary this was confirmed by Southern analysis or polymerase chain reaction (PCR) using a DNA probe or oligonucleotide primers derived from the Y-linked gene Zfy-l [Mardon, G. et al ., Science. 243. 78-80 (1989), Nagamine, CM. et al., Nature. 243, 80- 83 (1989) and Koopman P. et al., Nature. 342. 940-942 (1989)]. A 14kb fragment, derived from a phage clone L741 containing 8kb of sequence 5' and 5kb 3' to the putative DNA binding domain of Sry (Fig. 23) was purified from vector seguences and used to generate transgenic mice.

Fig 23. shows restriction maps of mouse Sry fragment 741 and human SRY fragment A (isolated from the cosmid cAMF, [Ellis, N.A. et al.. Nature. 337. 81-84 (1989)] are shown using the following restriction endonucleases: B, BamHI; E, EcoRI : H, Hindlll and S, Sall. Fragment sizes are

indicated in kb. The conserved Sry/SRY open reading frame is indicated by a shaded box. The direction of the open reading frame is shown above the two clones. The position of the human pseudoautosomal boundary is indicated by an arrow, the pseudoautosomal region being to the right of this point. The positions of oligonucleotide primers used for PCR analysis (described in relation to Figs. 25 and 27) are indicated by triangles.

SRY- or Sry-containing fragments were released from cosmid or phage vectors by digestion with appropriate restriction enzymes and then isolated by agarose gel electrophoresis and further purified by one of three methods: (i) Geneclean (Bio101) according to manufacturers' instructions; (ii) phenol extraction, Sephadex G50 column chromatography and ethanol precipitation; (iii) Geneclean followed by Elutip (Schleicher & Schuell) and ethanol precipitation. Transgenic mice were produced essentially as described in "Manipulating the Mouse Embryo" (Hogan B., Contantini, F. & Lacy, E.) (Cold Spring Harbour Laboratory, New York, 1986). In brief: three to five-week old (CBA × C57BL/10) F1 females were superovulated and mated to FI stud males. The following day, fertilised eggs were collected from oviducts of females showing vaginal plugs. Pronuclei were

microinjected with 1-2pl of DNA at a concentration of approximately 2 ug/ml. The eggs were cultured overnight in M16 or T6 medium, and 2-cell embryos implanted by oviduct transfer into day of plug pseudopregnant recipients. A total of 158 embryos were obtained after injection with this fragment (subsequently referred to as f741).

LEGEND TO TABLE 1

Injected embryos were examined at 14 days post transfer. Embryos were analysed in the following way: chromosomal sex (XX or XY/XO) was determined by staining for sex chromatin in amnion cells [Monk, M & McLaren, A., J. Embryol Exp. Morph. 63. 75-84 (1981)]. Transgenesis was assayed either by Southern blot or PCR detection of Sry. and the presence or absence of a Y chromosome judged from similar assays for Zfy gene sequences. Genomic DNA for Southern analysis was prepared from limbs. Southern analysis of EcoRI-digested DNA was performed as described in "Molecular Cloning" (Maniatis, R., Fritsch, E.F. & Sambrook, J) (Cold Spring Harbor Laboratory, New York, 1982). For Sry. blots were probed with clone 422.04 which contains Sry conserved motif. For Zfy genes, a 1.9kb Hindlll genomic fragment containing the Zfy-1 zinc finger domain [Koopman, P. et al., Nature. 342. 940-942 (1989)] was used as a probe. For PCR analysis proteinase K digestion was performed in ImM EDTA, the DNA was extracted once with phenol/chloroform and a small aliquot added directly to the PCR reaction mix.PCR analysis was performed as described in relation to Fig. 25. Phenotypic sex was determined by scoring for testis or ovary development. The results are shown in Table 1 below. The frequency of XO progeny was consistent with previous studies [Russell, L.B., in "Chemical mutagens, Principles and methods for their detection", Vol. 4 (ed. Hollaender, A.) 55-91 (Plenum Press, New York, London 1976)]. In four cases of XX transgenesis, comparison of the Sry signal to that of a control male indicated mosaicism for the

transgene. The last two entries show the transgenic embryos obtained.

Most of embryos were, in roughly egual proportion, XY males or XX females. However, in two cases (m7.22 and mlθ.2) testes were seen in embryos whose sex chromatin indicated an XX rather than XY sex chromosome constitution. By

Southern analysis, both these were found to lack Zfy

sequences and to be transgenic, with multiple copies of Sry (Fig. 24a). Lack of hybridisation to a probe recognising Zfv-1 and Zfy-2 shows the absence of a Y chromosome, while the intensity of hybridisation to the Sry probe

demonstrates that they carry multiple copies of the

transgene. XY male and XX female samples are included for comparison. The Zfy probe also detects Zfx and Zfa

[Ashworth, A. et al ., EMBO. J.. 9, 1529-1534 (1990)], providing a control for the amount of DNA in each lane.

Sizes of bands are shown in kb.

The gonads from these two embryos were photographed whole in phosphate-buffered saline (PBS), then fixed in 4% paraformaldehyde, dehydrated in ethanol, and embedded in paraffin. Sections (7μm) were stained in haematoxylin and eosin and both exhibited normal testis cord formation, and were indistinguishable from testes from normal XY sib embryos (Fig 24b, c).

Fig. 24b shows pairs of gonads, dissected from embryo m7.22 (upper panel, centre) and M10.2 (lower panel, centre), are shown between single testes (left) and ovaries (right) of nontransgenic sibs. The gonads of the transgenic embryos show the characteristic stripes associated with testis cord formation. Fig. 24c shows histology of m7.22 (upper panel) and m10.2 (lower panel) testis sections. The apparent difference in size is due to plane of section. Cord morphology was similar to that of littermates.

From these experiments it was concluded that a 14kb genomic fragment carrying Sry sequences is sufficient to initiate testis development in mice.

To determine the frequency with which f741 gives sex reversal all the embryos scored as females were examined for the presence of Sry sequences by PCR. Two were

unequivocally identified as transgenic, and a further 4 embryos gave weak signals suggesting that they were mosaics possessing the transgene in a low proportion of cells

(Table 1). The finding that not all XX transgenics show sex reversal is attributed to such mosaicism or to position effects on the level of Sry expression (see below).

Development of normal adult male phenotype in a sex

reversed transgenic mouse. To test the adult phenotype of Sry transgenic mice some of the embryos injected with f741 were allowed to develop to term. A total of 93 animals were born (49 males and 44 females) . Five of these were found to be transgenic by Southern blotting. Two were XY males that did not transmit the transgene and so were uninformative with respect to sex reversal.

One of the trangenics, M33.13 was analysed as follows:

For PCR analysis, 0.1 μg of genomic DNA was added to a 50 μl reaction mix containing 1.5mJJ each dNTP, 50mM Tris-HCl pH9, 15mM ammonium sulphate, 7mM MgCl₂, 0.05% Nonidet P-40, 0.5U Tag polymerase (Anglian Biotec) and 500ng of each oligonucleotide primer. Amplification consisted of 30 cycles of 94°C for 5s, 65ºC for 30s and 72°C for 30s in a Techne PHC-2 thermocycler. An 8μl aliquot was

electrophoresed on a 2% agarose-TBE gel. Primers for Sry were (5 '-3') TCA TGA GAC TGC CAA CCA CAG and CAT GAC CAC CAC CAC CAC CAA (indicated as triangles in Fig 23) and for Zfv-l. CCT ATT GCA TGG ACT GCA GCT TAT G and GAC TAG ACA TGT CTT AAC ATC TGT CC; myogenin primers corresponded to nucleotides 656-685 and 882-901 of the rat cDNA sequence [Wright, W.E., et al., Cell. 16, 607-617 (1989)]. PCR products were 441, 180 and 245bp respectively. The testes were processed for histology as described in relation to Fig. 24. The results of the PCR and examination of the testes are shown in Fig. 25 in which Fig. 25a, shows PCR analysis of genomic DNA from m33.13 (lane 3), showing Sry and control (myogenin) bands. No band corresponding to

Zfy-1 was seen, demonstrating the lack of a Y chromosome; (this result was confirmed by Southern blotting using Y-chromosome probes Y353B [Bishop, CE. et al ., Nature. 315. 70-72 (1985) and Sxl [Roberts, C, et al., Proc. Natl.

Acad. Sci. U.S.A.. 85. 6646-6649 (1988)]. Normal XX female and XY male littermates (33.9, lane 2 and 33.17, lane 1) are shown for comparison. M, marker bands (1081, 510, 396, 344, 298, 220, 201, 154 and 134bp). Fig. 25b shows external genitalia of mice 33.17 (left) and 33.13 (right) and Fig. 25c shows the histology of testis sections from mice 33.17 (left) and 33.13 (right) (Scale Bar, 90μm).

The mouse m33.13 was similar in size and weight to normal XY littermates. At about six weeks post partum m33.13 was caged with females (maximum of two per night). The mouse exhibited normal copulatory behaviour, mating four times in six days.

The presence of two X chromosomes in a male mouse always results in sterility, as germ cells are prevented from progressing beyond prospermatogonia. This phenomenon has been documented in XX Sxr and XX Sxr' mice which are male due to the presence of Y-derived sequences including Tdy on one of their X chromosomes [Cattanach, B.M., et al..

Cytogenetics. 10, 318-337 (1971), Burgoyne, P.S., et al.. Nature. 320 170-172 (1986) and Sutcliffe, M.J. & Byrgoyne, P.S., Development. 107. 373-380 (1989)]. It was therefore not surprising that the sex-reversed transgenic mouse m33.13 was also sterile. None of the four females with which he mated became pregnant. In three cases the vaginal plugs were examined for the presence of sperm, but none was found. The only difference between m33.13 and a normal XY sibling was in the size of the testes: m33.13 has a testis weight of 17mg (in the range expected for an XX Sxr' male), as opposed to 76 mg for an XY littermate. The testes were processed for histology and sections revealed the presence of tubules, with clearly defined and apparently normal populations of Leydig cells, peritubular myoid cells and Sertoli cells, but a complete absence of cells undergoing spermatogensis (Fig. 25c).

Internal examination of m33.13 revealed a normal male reproductive tract with no signs of hermaphroditism. This indicates that Sertoli cells must have been functionally normal in terms of anti-Mullerian hormone (AMH) production and Leydig cells in terms of testosterone production. AMH is required for the elimination of the female Mullerian duct system (oviducts, uterus and upper vagina) and testosterone for development of the Wolffian duct

derivatives (vas deferens and accessory glands such as the seminal vesicles) as well as for development of male secondary sexual characteristics [Jost, A & Magre, S. Phil. Trans. R. Soc. Lond. 322. 55-61 (1988) and Josso, N., in "Mechanisms of sex differentiation in animals and man" (eds Austin, C. R. & Edwards, R.G.) 165-203 (Academic Press, London 1981)].

A further two XX transgenics, m32.10 and m33.2, showed an external female phenotype, yet both carried multiple copies of Sry. These mice have produced offspring indicating that they have functional reproductive tract and ovaries. These animals provide further evidence, along with the transgenic XX female fetuses described above, that f741 does not always cause sex reversal. While it is

conveivable that there could be subtle rearrangements of the Sry gene making it non-functional, the possibility of this occurring in all these cases seems remote. There are two other, more probable explanations. First, these females could be mosaic for the transgene, with only a small proportion of the cells making up the somatic portion of the genital ridge carrying Sry gene copies. Analysis of XX↔XY chimeras suggests that females or hermaphrodites develop if there are less than about 30% XY cells [Eicher, E.M. et al ., Cvtooenet. Cell Genet. . 28 , 104-115 (1980) , McLaren A., in "Chimeras in Developmental Biology" (eds Le Douarin, N. & McLaren, A.) 381-399 (Academic Press, New York & London, 1984) and Burgoyne, P.S. et al .,

Development. 102, 443-450 (1988)]. Secondly, the

expression of the transgene could be subject to position effects due to the site of integration. Except for a few cases where locus controlling regions are present

expression of transgenes almost always depend on their chromosomal location [Grosveld, F. et al., Cell., 51. 975- 985 (1987)]. These two alternatives can be examined by breeding from the adult XX transgenic females.

Mouse m32.10 was mated with an FI (CBA × C57BL/10) male and resulting offspring were tail biopsied at 3 weeks. Genomic DNA preparation and Southern analysis were as described in relation to Table 1 and results are shown in Fig. 26.

Lack of hydridisation to a probe recognising Zfy-1 and Zfy- 2 indicates the absence of a Y chromosome. Founder female m32.10 shows intense hybridisation to the probe recognising Sry demonstrating the presence of multiple copies of the transgene. The same band pattern is shown by offspring number 7. Examination of the external genitalia of this animal revealed a normal female phenotype. DNA from the male offspring nos. 1-5 shows hybridisation to both Zfy and Sry probes. Three of these have multiple copies of Sry and are therefore transgenic. Band sizes are shown in kb. 33.2 has not yet given transgenic offspring. However, 32.10 has transmitted the transgene to female offspring (Fig. 26) , suggesting that mosaicism is not an explanation in this case. Human SRY does not function in mice

Mouse gry and human SRY show a high degree of similarly in primary structure in the 79 amino acid putative DNA binding domain. The specificity of the protein's interaction with other genes in the sex determination pathway presumably depends on the sequence of this domain. However, the two genes differ in 23 of these amino acid residues, with only two of the differences being conservative. The finding of only a single amino acid substitution leading to failure of testicular development in an XY woman suggests that in some cases even an apparently conservative alteration in the SRY protein can disrupt its action. In addition, the

nucleotide sequences of the mouse and human genes diverge strikingly outside the putative DNA-binding domain. In view of these observations it was interesting to test whether the human SRY gene is as effective as its murine counterpart in causing sex reversal in mice.

The DNA used for pronuclear injection was a 25kb BamHI -SalI fragment representing human Y chromosomal DNA around SRY. is isolated from a cosmid clone cAMF [Ellis, N.A. et al.. Nature. 337. 81-84 (1989)]. This fragment includes approximately 12.5kb of Y-unique sequence 5', and 5kb 3' to the SRY conserved domain. The remaining 6kb represents sequences from the pseudoautosomal region that is common to the X and Y chromosomes (Fig. 23).

The ability of human SRY to cause sex reversal was assessed from animals representing three independent integrations of the transgene. Two of these were lines derived from XY founder transgenics, hA6 and hA9. These founders

transmitted SRY to about half their offspring: 8 out of 17 hA6 and 13 out of 36 hA9 embryos assayed were transgenic. When progeny of these line were analysed at 14.5dpc there was no evidence of testis cord formation in XX transgenic fetuses, demonstrating that neither integration event was able to cause sex reversal. A third integration was represented by a single XX transgenic founder embryo; this too was phenotypically female.

Having transgenic lines that transmit human SRY to their offspring allowed examination of expression of the

transgene in developing gonads, and in the only other known site of SRY expression, the adult testis. Genital ridges were dissected from hA6 and hA9 fetuses at 11.5-12 dpc, a time at which mouse Sry is known to be expressed. RNA was extracted from pairs of urogenital ridges and reverse transcribed in small-scale reactions [Koopman, P. et al.. Nature. 342. 940-942 (1989)] using half the yield from each pair of ridges in reactions with and half without reverse transcriptase. Reverse transcription was performed on each RNA sample with (+) and without (-) reverse transcriptase (RT), to demonstrate that the observed bands were due to the presence of SRY transcripts and not contaminating

DNA.Genomic DNA or reverse transcription products were added to PCR reactions and amplified as described in relation to Fig. 25. For DNA analysis, SRY primers used were (5 '-3') GAT CAG CAA GCA GCT GGG ATA CCA GTG and CTG TAG CGG TCC CGT TGC TGC GGT G (336bp product, solid triangles in Fig 23); for RNA analysis, CAG GAG GCA CAG AAA TTA CAG GGC ATG C and ACA GTC ATC CCT GTA CAA CCT GTT GTC C (174bp product, open triangles in Fig 23) were used. PCR analysis of genomic DNA (Fig. 27 upper panel) showed that all embryos (lanes 1-12) were transgenic for SRY. Samples 4-6 and 10-12 showed bands corresponding to Zfy-1.

confirming that they were XY, unlike samples 1-3 and 7-9 which were XX. Non-transgenic controls (non TG) are shown on the right. Oligonucelotide primers for myogenin were included in each analysis as a control. Below each lane is shown a PCR analysis of reverse-transcribed RNA

extracted from the urogenital ridges of the same embryos (lower panel). Bands corresponding to SRY expression were seen in samples 1-9 and 11. PCR analysis of RNA extracted from these samples shows that SRY transcripts were present in transgenic XX fetuses that were not sex-reversed (Fig. 27). Curiously, not all hA6 XY fetuses expressed the transgene (Fig. 27); this may be due to variation in developmental stage, and is being

investigated. The level of SRY expression in the genital ridges was estimated to be several times that of the endogenous Sry gene, and was greater than that seen in transgenic XY adult testis material. Clearly lack of transcription in the genital ridge cannot account for the failure of SRY to give sex reversal in mice. It is formally possible that the SRY mRNA is not correctly processed or translated. Alternatively the protein product could be unstable in mouse cells. However, it is more likely that differences in sequence incapacitate the human SRY protein in mouse cells, due to a failure to interact with other regulatory proteins or target genes. It would be possible to test this hypothesis by exchanging the human and mouse open reading frames. Discussion

These experiments demonstrate that a 14kb mouse genomic DNA fragment containing Sry is sufficient to direct the formation of testes in XX transgenic embryos and

subsequently to give rise to full phenotypic sex reversal in an XX transgenic adult. Complete sequencing of f741, as well as cross hybridisation experiments to human DNA, have failed to detect any gene sequences other than Sry. While previous data provided compelling evidence to implicate Sry in the process of testis determination, the current study suggests that Sry is the only Y-linked gene required to give rise to male development, and therefore Sry is the testis determining gene.

The ability of a 14kb fragment containing Sry to cause sex reversal suggests that this fragment contains the entire Sry gene, including all the regulatory elements required for appropriate embryonic expression. The transgenic system can be used to further localise these regulatory elements within the 14kb fragment, by microinjecting progressively smaller constructs. The only other site of Sry expression is in adult testis, probably in the germ cell component. It will be interesting to see if the 14kb fragment represented by f741 also contains the regulatory information required for the switch from expression in the somatic part of the embryonic gonad expression in adult testis associated with germ cells. The finding that the human SRY gene is expressed at both stages in transgenic mice is consistent with the appropriate regulatory

sequences being present within the 25 kb of DNA flanking the gene. It also indicates that the mechanisms regulating Sry/SR/ expression have been conserved between humans and mice.

While it has been shown that Sry alone can promote

testicular development in the absence of other Y-linked genes, sex reversal does not always occur. The most likely explanation for this is sensitivity of the Sry transgene to position effects. This variability may mirror the

situation in human XX individuals carrying small portions of the Y chromosome including SRY. Four such cases

[Palmer, M.S. et al., Nature. 342. 937-939 (1989)], each inherited about 35 kb of Y-unique sequences on their paternal X chromosome but only two manifested only partial sex reversal. An XX hermaphrodite is known with a similar small portion of the Y. In these cases SRY may be

sensitive not only to the effect of adjacent DNA sequences in its new chromosomal location, but also to spreading of X-inactivation. There are a number of precedents for the latter in the mouse, for example females or hermaphrodites frequently develop instead of males when the Sxr fragment is present on the inactive X chromosome [McLaren, A. & Monk, M., Nature. 300. 446-448 (1982)]. Similar cases of partial sex reversal due to position effects will be seen when additional transgenic mice are analysed, where the level of Sry expression is close to a critical threshold level.

Sry acts over a short time period to initiate testis development. It must do this through interaction with other genes, some of which will be involved in the

regulation of Sry. others will be downstream targets of

Sry. These other genes must map elsewhere in the genome as it has been shown that Sry is the only Y-linked gene required to bring about male development in mice.

Mutations in some of these genes could explain cases of male development in XX individuals lacking SRY [Palmer, M.S. et al ., Nature. 342. 937-939 (1989)], and XY females where SRY is intact [Scherer, G. et al ., Hum. Genet.. 81.

291-294 (1989), Fredga. K., Phil. Trans. R. Soc. Lond.. 322. 83-95 (1988)]. Using molecular genetic techniques to work stepwise from Sry it should now be possible to

identify these other genes.

Claims

1. pY53.3 as deposited on 12th July 1990 with NCIMB under accession number NCIMB 40308.

2. A clone or subclone of pY53.3 as defined in claim 1.

3. A fragment of pY53.3 capable of giving a Y-specific signal on hybridisation to genomic DNA of a eutherian mammal.

4. A fragment of pY53.3 obtainable by restriction endonuclease digestion thereof and being capable of giving a Y-specific signal on hybridisation to genomic DNA of a eutherian mammal.

5. A fragment according to claim 4 which is the 0.9 kb product of HincII digestion of pY53.3.

6. A clone or subclone of a fragment according to any one of claims 3 to 5.

7. A nucleic acid or fragment having substantially the sequence of pY53.3 as set out in Fig. 19.

8. A nucleic acid or fragment having substantially the sequence of the 0.9 kb HincII digestion product of pY53.3.

9. A clone or subclone of a nucleic acid or fragment according to claim 7 or claim 8.

10. A nucleic acid or fragment or oligonucleotide having substantially the sequence of the human, mouse or rabbit mt-box as set out in Fig. 1 or Fig. 6 or Fig. 14 or Fig. 18 or Fig. 19.

11. A nucleic acid or fragment or oligonucleotide being capable of giving a Y-chromosome specific signal on

hybridisation to the genomic DNA of eutherian mammal.

12. A nucleic acid or fragment or oligonucleotide according to claim 4 or claim 11 capable of giving a Y-chromosome specific signal on hybridisation to the genomic DNA of a human, bovine, ovine, equine or porcine mammal.

13. A nucleic acid or fragment or oligonucleotide according to any one of claims 4, 11 and 12 capable of giving a Y-chromosome specific signal on hybridisation to the genomic DNA of a eutherian mammal under conditions of high stringency.

14. A nucleic acid or fragment or oligonucleotide according to any one of claims 4, 11 and 12 capable of giving a Y-chromosome specific signal on hybridisation to the genomic DNA of a eutherian mammal under conditions of low stringency.

15. A nucleic acid or fragment or oligonucleotide capable, on hybridisation to the genomic DNA of an XY human male of giving a specific signal for a region of 35 kb extending from the pseudoautosomal boundary of the Y-chromosome into the Y-specific region.

16. A nucleic acid or fragment or oligonucleotide according to claim 15 capable of giving the specific signal on hybridisation under low stringency conditions.

17. A nucleic acid or fragment or oligonucleotide according to claim 15 capable of giving the specific signal on hybridisation under high stringency conditions.

18. A nucleic acid or fragment or oligonucleotide according to any one of claims 11 to 17 containing

substantially the sequence of the human, mouse or rabbit mt-box as set out in Fig. 1 or Fig. 6 or Fig. 14 or Fig. 18 or Fig. 19.

19. A nucleic acid or fragment or oligonucleotide encoding an mt-protein, fragment thereof or polypeptide containing an mt-box or a part thereof or encoding an mt-mimetope protein or fragment thereof or mt-mimetope polypeptide.

20. A process for ascertaining the sex of an embryo, foetus, cell, tissue or organism comprising hybridising a nucleic acid or fragment or oligonucleotide according to any one of claims 1 to 19 with DNA or RNA of the embryo, foetus, cell, tissue or organism or with cDNA reverse transcribed from RNA of the embryo, cell, tissue or organism or with DNA or cDNA amplified by cloning or polymerase chain reaction from DNA or RNA of the embryo, foetus, cell, tissue or organism.

21. Use of a nucleic acid or fragment or

oligonucleotide of any one of claims 1 to 19 in ascertaining the sex of an embryo, foetus, cell, tissue or organism.

22. A process for controlling the sex of the progeny of an organism comprising inserting a nucleic acid or fragment or oligonucleotide of any one of claims 1 to 19 into the genome of the organism or a progenitor thereof.

23. Use of a nucleic acid or fragment or

oligonucleotide of any one of claims 1 to 19 in controlling the sex of the progeny of an organism.

24. An mt-protein, fragment thereof or polypeptide containing an mt-box or a part-thereof or an mt-mimetope protein, fragment thereof or mt-mimetope polypeptide.

25. A protein or fragment thereof or a polypeptide containing an mt-box sequence including at least one of the characteristic amino acid residues at position 46, 63, 67, 74, 75, 76 and 98 as set out in Fig. 16.

26. A protein or fragment thereof or a polypeptide encoded by a nucleic acid or fragment or oligonucleotide according to any one of claims 1 to 19 and containing an mt-box.

27. A process for controlling the sex of the progeny of an organism comprising supplying exogenously to a cell of the organism or a progenitor of the organism a protein or fragment thereof or a polypeptide according to any one of claims 24 to 26.

28. A process according to claim 27 wherein the protein or fragment thereof or polypeptide is supplied and

activates a sry target gene.

29. An antibody or fragment thereof against a protein or fragment thereof or polypeptide according to any one of claims 24 to 26.

30. An antibody-producing cell capable of expressing an antibody or fragment thereof according to claim 29.

31. Use of a protein or fragment thereof or polypeptide according to any one of claims 24 to 26 or antibody or fragment thereof or cell according to claim 29 or claim 30 in ascertaining the sex of an embryo, cell, tissue or organism.

32. A nucleic acid or fragment, oligonucleotide, protein or fragment thereof, polypeptide, antibody or fragment thereof or cell according to any one of claims 1 to 19, 24 to 26 and 29 or 30 for use in a method of treatment or diagnosis practised on the human or animal body including embryos and foetuses.

33. Use of a nucleic acid or fragment, oligonucleotide, protein or fragment thereof, polypeptide, antibody or fragment thereof or cell according to any one of claims 1 to 19, 24 to 26, 29 and 30 in the production of a

medicament for use in a method of treatment or diagnosis practised on the human or animal body including embryos and foetuses.

34. A transgenic or chimeric animal having a

heterologous nucleic acid or fragment or oligonucleotide according to any one of claims 1 to 19 in the genome of at least the germ cells of the animal.

35. Gametes of an animal according to claim 34.

36. Progeny of an animal according to claim 34.

37. Progeny according to claim 36 which are transgenic or chimeric and have a heterologous nucleic acid or fragment or oligonucleotide according to any one of claims 1 to 19 in the genome of at least the germ cells of the progeny.

38. A method of controlling a population of a species of eutherian mammal which comprises introducing a male member of the species into the population, said male having a copy of a nucleic acid fragment or oligonucleotide according to any one of claims 1 to 19 integrated on an autosome (carrier autosome), whereby when the male breeds with females of the population the following types of progeny will be produced:

(i) Normal males (XY) without the carrier

autosome;

(ii) Carrier males (XY) with the carrier autosome; (iii) Normal females (XX) without the carrier

automsome; and

(iv) Sterile (XX) males with the carrier autosome.

39. A method according to claim 38 where the nucleic acid integrated into the carrier autosome is homologous to the native SRY gene of the mammals.

40. A method according to claim 38 or 39 for

controlling a population of mice, rats, racoons, rabbits, kangaroos or deer.