Proc. Nati. Acad. Sci. USA Vol. 91, pp. 1927-1931, March 1994 Genetics An SRY-related sequence on the marsupial X chromosome: Implications for the evolution of the mammalian testis- determining gene (sex chromosomes/sex dUeterminaion/hro e evoltion) JAMIE W. FOSTER*t AND JENNIFER A. MARSHALL GRAVES Department of Genetics and Human Variation, LaTrobe University, Bundoora, Victoria 3083, Australia Communicated by Mary F. Lyon, November 8, 1993 (receivedfor review September 27, 1993) ABSTRACT The SRY gene on the human, mouse, and and SOX genes are obviously of interest in considering the marsupial Y chromosomes is the testis-determining gene that origin and evolution of the SRY gene in mammals. iitiates male development in mammals. The SRY protein has Marsupials diverged from eutherian (placental) mammals a DNA-binding domain (high mobility group or HMG box) 120-150 million years ago and monotremes (egg-laying mam- similar tothosefound in the high-mobility-group proteins. SRY mals) diverged even earlier, so that comparisons between is speific for the Y chromosome, but many a tl genes these three major mammalian groups may provide informa- have be identified that possess a similar HMG box region; tion about the function and early evolution ofmammalian sex those with the most closely SRY-related box regions form a gene chromosomes and sex-determining genes. Eutherian, mar- family now referred to as SOX genes. We have identified a supial, and monotreme sex chromosomes have been found to sequence on the marsupial X chromosome that shares homol- differ in size and gene content, enabling the different evolu- ogy with SRY. Sequence comparisons show near-identity with tionary origins of regions of the human sex chromosomes to the mouse and human SOX3 gene (formerly called a3), the SOX be deduced. The genes on the long arm and proximal short gene which is the most closely related to SRY. We suggest here arm of the human X chromosome are present on the X that the highly conserved X chromosome-linked SOX3 repre- chromosome in marsupial and monotreme mammals, and this sents the ancestral SOX gene from which the sex-determining region, therefore, represents a conserved, probably original, gene SRY was derived. In this model SOX3/SRY divergence mammalian X chromosome (10, 11). The marsupial and and the acquisition of a testis-determining role by SRY might monotreme X chromosome lacks genes borne on the short have preceded (and initiated) sex chromosome differentiation arm of the human X chromosome, suggesting that this region or, alternatively, might have been a consequence of X chro- was originally autosomal and was added later to the eutherian mosome-Y chromosome differentiation initiated at the locus of X chromosome. The presence of several genes in this region an original sex-determining gene(s), later superseded by SRY. with homologues on the Y chromosome implies that the region was added to both X and Y chromosomes, probably by Mammals have an XX female/XY male sex chromosome recombination within an original pseudoautosomal region mechanism, in which a gene on the Y chromosome deter- (10, 12). In marsupials, as in eutherian mammals, the Y mines testis development, the first step in the male develop- chromosome is testis determining, but at least some sexual mental pathway. SRY has been cloned from the sex- dimorphisms are sex hormone independent and seem to be a determining region of the human Y chromosome (1), and function of X chromosome dosage, rather than the presence evidence from mutational analysis (2, 3) and transgenesis (4) or absence of a Y chromosome (13). confirms that it acts as the testis-determining gene. The We have isolated an SR Y-related sequence from the mar- predicted SRYgene product contains an HMG (high mobility supial X chromosome, which is closely homologous to the group) box that binds to double-stranded DNA in a sequence- mouse and human SOX3 gene. This raises the possibility that dependent manner and to cruciform DNA without sequence X chromosome inactivation or gene dosage may play a role specificity (5, 6). Its similarity to a number of proteins that in sex determination in marsupials. However, we suggest that affect transcription suggests that the gene functions by acti- a more likely explanation for the presence of SRY homo- vating or repressing other genes in the testis-determining logues on marsupial, mouse, and human X chromosomes is pathway. that SOX3 and SRYwere originally alleles of a developmen- As would be expected ofthe mammalian testis-determining tally important gene shared by partly differentiated ancestral gene, homologues have been detected on the Y chromosomes X and Y chromosomes. of all eutherian and marsupial mammals tested (1, 7). The SR Y gene is unique to the Y chromosome and is thus MATERIALS AND METHODS restricted to males. However, when used in Southern hy- We used two marsupial species, representing the two major bridization experiments, human SRY probes also detect Australian orders that diverged about 50 million years ago, sequences shared by males and females of many eutherian the striped-faced dunnart Sminthopsis macroura (Order and marsupial species (1, 7, 8). These SRY-related sequences Polyprotodonta, Family Dasyuridae) and the Tammar wall- were thought to be members ofa large family ofrelated genes aby Macropus eugenii (Order Diprotodonta, Family (8), originally termed al, a2, etc. but now known as SOX], Macropodidae). Tissue was originally provided by D. W. SOX2, etc. (for SRY-related HMG box-containing). These Cooper(School ofBiological Sciences, Macqurie University) SOXgenes are highly conserved and have been demonstrated and L. Selwood (Department of Genetics and Human Vari- in other vertebrates (9). The relationships between the SRY Abbreviation: HPRT, hypoxanthine phosphoribosyltransferase. The publication costs ofthis article were defrayed in part by page charge *resent address: Department of Genetics, Cambridge University, payment. This article must therefore be hereby marked "advertisement" Tennis Court Road, Cambridge CB2 3EH, United Kingdom. in accordance with 18 U.S.C. §1734 solely to indicate this fact. tTo whom reprint requests should be addressed. 1927 1928 Genetics: Foster and Graves Proc. Natl. Acad. Sci. USA 91 (1994) ation, Latrobe University) and was retained under the pro- a h visions ofpermit RP 90-177 from the Victorian Department of Conservation and the Environment. Rodent-marsupial cell b 0o* O 0- - - - hybrids containing a marsupial X chromosome were obtained - f iS li0 by fusing hypoxanthine phosphoribosyltransferase (HPRT)- _i- r]> ofj _._ deficient Chinese hamster cells with fibroblasts from Plani- gale maculata, a dasyurid marsupial related to S. macroura (14). Marsupial fibroblast and hybrid lines were cultured under standard conditions in Dulbecco's modified Eagle's medium/10%o fetal calf serum. DNA extraction and Southern blotting procedures were adapted from Reed and Mann (15). Ten micrograms of restricted DNA was electrophoresed through 0.8% agarose, transferred to Hybond-N+ (Amersham) with 20x standard saline citrate (SSC) and fixed with 0.4 M NaOH. A 0.9-kb HincII subclone of pY53.3, containing the human SRY gene (1), was labeled with [32P]dCTP by random priming. The .l2 probe was hybridized in 5x SSPE [1x SSPE is 0.5% SDS/5x ..d fi Denhardt's solution (0.18 M NaCl/10 mM phosphate, pH | ; 7.4/1 mM EDTA)/10%o (wt/vol) dextran sulfate/denatured salmon sperm at 100 pg/ml at 65°CJ. Membranes were ; | washed in 2x SSC/0.1% SDS at 65°C and autoradiographed E E for6-10 days at -70°C. Fragment sizes are estimated relative - W~u- f., to A HindIII molecular-weight markers. Genomic libraries were constructed from S. macroura liver DNA, partially restricted with Sau3AI and size-selected in en. 10-40% glycerol gradients. DNA was ligated to EMBL3A '#'1u'' (BRL) and packaged with A in vitro packaging reactions (Amersham). A total of 1.0 x 106 plaque-forming units were screened without amplification, using the 0.9-kb HincII frag- ment of pY53.3 as a probe. Positive clones were plaque- purified and subcloned into pUC vectors. Subclones were F: sequenced by the dideoxynucleotide chain-termination II (United States Biochemical) Fio. 1. Southern blot analysis ofSRY-related genes in marsupials method (16) using Sequenase and cell hybrids. (a) Southern analysis of EcoRI-restricted DNA and Taq Track (Promega). from male and female S. macroura (S. ma., lanes 1 and 2) and M. In situ hybridization was done as described (17). Probe was eugenii (M. eu., lanes 3 and 4), hybridized with a 0.9-kb Hincd nick-translated with [3HldATP, [3H]dCTP, and [3H]dTTP to fragment of p53.3, containing the human SRY open reading frame. activities of 2-6 x 107 cpm/pg and was hybridized to The probe detects a male-specific fragment (7) in both species, as formamide-denatured metaphase spreads over a range of well as several fragments shared between males (d) and females (v). concentrations. Slides were autoradiographed by using Am- The shared bands at 2.3kb and 7.5kb inS. macroura andM. eugenii, ersham nuclear track emulsion, and autoradiographs with the respectively (arrows), show a 1:2 male/female intensity difference, highest signal/noise ratio were selected for scoring. The suggesting X chromosome linkage. (b) Southern analysis of EcoRI- least 100 spreads was recorded by restricted DNA from Chinese hamster-dasyurid marsupial somatic location of grains over at cell hybrid. Lanes: 1 and 2, S. macroura (S. ma.) female and male; using a Leitz Dialux microscope fitted with a Sony video 3, P. maculata (P. ma.) marsupial parent cell line; 4, V2C5 HPRT-- recorder. The grain distribution was analyzed by using the deficient revertant cell line lacking the marsupial X chromosome; 5, Z statistic, which tests the overall distribution for depar- V2C5 HPRT+ hybrid retaining only the marsupial X chromosome; 6, tures from randomness, detects overlabeled sites, and at- Chinese hamster parent cell line.