The Zfy-1 and Zfy-2 genes, which arose by gene duplication, map to the mouse Y chromosome and encode nearly identical zinc-finger proteins. Zfy-1 is expressed in the genital ridge and adult testis and likely encodes a transcription activator. Although potential roles in sex determination and spermatogenesis have been hotly debated, the biological functions ofZfy-1remain unknown. To study the gene’s regulation, transgenes with 21-28 kb of Zfy-1 5′ flanking DNA placed upstream of lacZ were constructed in plasmids or created by homologous recombination of coinjected DNA molecules. The resulting transgenic mice expressed β-galactosidase in the genital ridge of both males and females starting between embryonic day 10 and 11 (E10-E11), peaking at E12-E13 and then declining to low levels by E15, a pattern that matches Zfy-1 mRNA as detected by RT-PCR. This lacZ expression in genital ridge was confined to somatic cells as demonstrated by its absence from the alkaline phosphatase-positive germ cells. It had been reported previously that Zfy-1 mRNA was absent from the embryonic gonad of homozygous Weembryos, which virtually lack germ cells. By contrast, we observed normal expression of the Zfy-1/lacZ transgene when introduced into the We background, suggesting that germ cells are not necessary for expression. In the adult, the Zfy-1/lacZtransgene is expressed abundantly in developing germ cells. Extragonadal (kidney, meninges, arteries, choroid plexus) expression of the transgene was also observed in embryos. A smaller transgene with only 4.3 kb of Zfy-1 5′ flanking DNA was expressed only in germ cells of adult mice. These results suggest that an enhancer for germ cell expression in the adult lies near the Zfy-1 promoter and that an enhancer for expression in the somatic cells of the embryonic gonad is located further 5′.

ZFY (zinc-finger gene on the Y chromosome) was identified because of its location within a genetically defined sex-determining region of the human Y chromosome (Page et al., 1987). Subsequent identification of SRY (gene from the sex-determining region of the Y chromosome) as the testis-determining factor eliminated ZFY as a candidate for that function (Sinclair et al., 1990; Gubbay et al., 1990; Koopman et al., 1991a). Zfy-1, a mouse homologue of human ZFY(Mardon and Page, 1989; Mardon et al., 1989; Ashworth et al., 1989), is a member of a multigene family including Zfy-2, also found on the Y chromosome, as well as Zfx and Zfa which are X-linked and autosomal, respectively (reviewed by Koopman et al., 1991b). All members of the family contain 13 zinc fingers, although differences in the third finger in Zfy-2 and Zfa may prevent them from coordinating zinc. The zinc finger motif is found in many eukaryotic DNA- and RNA-binding proteins and these fingers interact with the nucleic acids (Miller et al., 1985; Pavletich and Pabo, 1991). Hence, these zinc-finger genes are thought to encode transcription factors; however, no target genes for any member of the family have been identified.

The expression patterns of the Zfy-like genes have been partially determined by northern blot hybridization and RT-PCR. Zfx is expressed in every organ examined (Mardon et al., 1990) while the other family members have more restricted patterns of expression. Zfa is expressed only in the adult testis (Ashworth et al., 1990). Zfy-1, but not Zfy-2, is expressed in mouse embryonic stem (ES) cells and blastocysts, while both Zfy-1 and 2 are expressed in the germ cells of the adult testis (Koopman et al., 1989; Nagamine et al., 1989, 1990; Zwingman et al., 1993). In one study, both Zfy-1 and Zfy-2 mRNAs were detected by RT-PCR in the embryonic gonad (Nagamine et al., 1990) while in another study Zfy-1mRNA was detected but Zfy-2 mRNA was below the level of detection of the RT-PCR assay (Koopman et al., 1989). The developmental timing of Zfy-1 expression in the genital ridge is coincident with arrival of the germ cells, which occurs between E10 and E12 (Eddy et al., 1981; Tam and Snow, 1981). Expression peaks at E13.5, a time when the primordial germ cell (PGC) population is also peaking in the embryonic gonad. Zfy-1gene expression was not detected in the genital ridge of We homozygous mice (Koopman et al., 1989) in which the testes form, but are virtually devoid of germ cells (Cattanach, 1978; Koopman et al., 1989). These data were interpreted to indicate that the early gonadal expression of Zfy-1 is germ cell-dependent (Koopman et al., 1989, 1991b). Message levels for Zfy-1and Zfy-2increase in the testis between postnatal days 7 and 14 and high levels of transcript are found in isolated round spermatids (Nagamine et al., 1989, 1990). Because of the very low level of Zfy-1expression outside of the adult testis, embryonic expression studies have relied upon RT-PCR precluding determination of the cell types that express the gene. There are also indications from RT-PCR and immunohistochemistry, using a Zfy antibody, that Zfy-1is expressed in tissues other than the gonad during mouse development including the kidney and brain (Nagamine et al., 1990; Su and Lau, 1992). To characterize the expression pattern of Zfy-1 during mouse development, including the cell types expressing the gene, we have identified the Zfy-1 promoter and used it along with the 5′ flanking region to direct the expression of a lacZ reporter gene in transgenic mice. We report here that the Zfy-1/lacZtransgene is expressed in somatic cells of the genital ridge in both males and females.

Constructs and transgenic mice

Constructs were made using standard cloning techniques (Sambrook et al., 1989). All Zfy family member DNAs were isolated previously (Simpson and Page, 1991). Zfy-1 sequences were obtained from cEMS132. cEMS177 sequences were subcloned into plasmid pEMS406.1, and Zfy-2 sequences near exon 1A were obtained from plasmid pEMS501.1, a subclone of cosmid cEMS212, while Zfy-2 sequences around exon 2 were obtained from plasmid pEMS283, a subclone of cosmid cEMS142. Restriction fragments containing the DNA for microinjection were separated from other fragments by electrophoresis on agarose gels; they were isolated by electroelution into a dialysis bag, followed by phenol chloroform extraction and ethanol precipitation. Transgenic mice were generated as described (Brinster et al., 1985), and the presence of the transgene was identified by DNA hybridization. Coinjections of more than one DNA molecule were done with approximately equivalent numbers of the two DNA molecules. The structure of the integrated DNA was analyzed by probing a Southern blot of Dra III-digested liver DNA with a radio-labelled 1.1 kb XbaI-EcoRV fragment from placF (Mercer et al., 1991) containing the E. coli lacZ gene. Animals were examined by staining for β-galactosidase activity to ascertain transgene expression in various organs. Embryonic time points were obtained by mating mice for 1-2 hour periods and checking for vaginal plugs. The time of appearance of the vaginal plug was denoted as time E0.

β-galactosidase staining

Embryos or tissues were prepared for the staining of E. coli β-galactosidase by fixation for 1 hour in 4% paraformaldehyde in PBS (pH 7.3). They were rinsed 3× with gentle shaking in 0.1 M NaH2PO4 (pH 7.3), 2 mM MgCl2, 0.01% sodium deoxycholate and 0.02% NP-40. Staining was performed overnight at 37°C in the same buffer supplemented with 1 mg/ml X-Gal (Molecular Probes), 5 mM potassium ferricyanide and 5 mM potassium ferrocyanide. After staining, tissues were stored in 70% ethanol. No embryonic gonad staining was observed in any of the hundreds of nontransgenic embryos examined.

Alkaline phosphatase staining

Alkaline phosphatase staining was done using a leukocyte alkaline phosphatase kit (Sigma). Embryos were frozen on dry ice in OCT embedding medium (Miles Scientific) and frozen sections were stained according to the Sigma protocol. Costaining for β-galac-tosidase activity and alkaline phosphatase activity was also done on the embryonic gonads from dissected embryos. Embryonic gonads were exposed by removal of the internal organs and subsequently stained for 5 hours at 37°C for β-galactosidase activity followed immediately by staining for 1 hour at room temperature in alkalinedye mix (Sigma) with gentle agitation. Embryos were then stored in 70% ethanol. Litters contained transgenic and nontransgenic embryos.

Breeding of the transgene into the We background

Two C57B6/SJL hybrid transgenic mice were mated to produce a male mouse homozygous for the Zfy-1 transgene (line 4592) and it was mated with 6 We heterozygous females. The resulting We heterozygous mice were identified by the white patches of fur on their heads and stomachs. All the offspring carried the transgene. These We heterozygous, Zfy-1/lacZ transgene-positive mice were bred to produce We homozygous mice carrying the transgene. The offspring of the matings were killed at E13 and stained for β-galactosidase activity. Placenta DNA was obtained from all mice and used for DNA dot hybridizations to identify the transgene carrying mice and for Southern blot hybridizations to identify the We homozygotes. Radioactivity was quantified and compared for DNA dot blots probed with a transgene probe verses an endogenous gene probe to determine hemiversus homozygosity of the transgene. Southern blot hybridizations were carried out on BglII-digested placenta DNA, and were probed with a radiolabelled 4.3 kb EcoRI fragment of the murine c-kit cDNA (Koopman et al., 1991c).

Zfy-1/lacZ expression in transgenic mice

Cosmid clones were identified that contain the 5′ untranslated exons of Zfy-1 (cEMS132 and cEMS189) and Zfy-2 (cEMS212 and cEMS142) (Simpson and Page, 1991). The locations of the exon-intron boundaries were determined by comparison of the genomic sequences obtained from these cosmids with that of the published cDNAs (Mardon and Page, 1989; Ashworth et al., 1989). Zfy-1 has several, alternatively spliced 5′ non-coding exons (Koopman et al., 1989 and our unpublished results) which complicates any nomenclature. In our nomenclature, exon 1A encodes the 5′ end of the Zfy-2 transcript (Fig. 1A). This exon is spliced to exon 2, which is the same as exon A described by Koopman et al. (1989). Mapping data indicates that transcription of Zfy-1 begins with exon 1B, which is then spliced to exon 2 (unpublished data). The unique EagI site in exon 1B of Zfy-1 was used to fuse approximately 4.3 kb of 5′ flanking sequence to a lacZ reporter gene (Fig. 1B, construct 2). This construct was either injected alone or coinjected with larger fragments obtained from the Zfy-1 cosmid (cEMS132) in hopes that they would cointegrate or recombine to provide additional regulatory elements. The coinjected fragment was either (a) an approximately 28 kb portion of the genomic clone containing Zfy-1 sequences 5′ of the unique EagI site (Fig. 1B, construct 5), (b) an approximately 13 kb portion of the genomic clone containing sequences 3′ of the EagI site (Fig. 1B, construct 4) or (c) the entire cosmid linearized with SalI (Fig. 1B, construct 3). Founder animals were killed at E12-E13 and stained for β-galactosidase activity. Table 1 summarizes the results. When injected alone, the −4.3 kb Zfy-1/lacZ construct was expressed only in the adult testis as will be described later. Four of ten transgenic founder mice, produced as a result of coinjection with the 5′ cosmid fragment, stained for β-galactosidase activity in the embryonic gonad and four of seven transgenic mice produced using the linearized cosmid, stained in the embryonic gonad. However, none of the twelve transgenic mice produced using the 3′ cosmid fragment stained in the embryonic gonad. This suggests that sequences necessary for embryonic gonad expression lie 5′ of −4.3 kb. Therefore, stable transgenic mouse lines were produced by coinjection of the 5′ genomic fragment with the −4.3 kb Zfy-1/ lacZ transgene. Three lines of mice were produced, two of which (4572 [TgN(ZFY1lacZ)217Bri] and 4592 [TgN(ZFY1lacZ)218Bri]) express β-galactosidase in the embryonic gonad.

Table 1.

Staining patterns of transgenic lines

Staining patterns of transgenic lines
Staining patterns of transgenic lines
Fig. 1.

(A) 5′ Splicing of Zfy-1and Zfy-2. The splicing patterns of the 5′ exons of the Zfy-1and Zfy-2genes are based on the comparison of cDNA sequences (Mardon and Page, 1989; Ashworth et al., 1989) and genomic sequences for the two genes. Sequences of Zfy-1and Zfy-2 are nearly identical around exons 1B and 2, but diverge at some point greater than 2 kb 5′ of exon 1B. Exons are depicted as rectangles and splicing by thin lines. Arrows indicate the transcription start site. The drawing is not to scale. (B) Transgenes tested in transgenic mice. All fusions were made to an E. coli β-galactosidase reporter gene (dotted rectangle) that contains an intron and two exons (open rectangles and arrows) of the mouse Protamine-1 gene including the Protamine-1 polyadenylation signal. Closed rectangles depict exons of Zfy-1and are labelled as either exon 1B or 2. Construct 1 contains 3.5 kb of DNA 5′ of the cEMS177 exon 1A like sequences fused at the XbaI site to lacZ (the hatched sequences are the exon 1A-like sequences of cEMS177). Construct 2 contains 4.3 kb of Zfy-15′ flanking sequence fused at the unique EagI site in exon 1B to lacZ. Constructs 3, 4 and 5 were coinjected with construct 2. Construct 3 is cosmid cEMS 132 linearized with SalI. Construct 4 is the 3′ portion (EagI-SalI fragment) of the cosmid containing sequences 3′ of the EagI site including Zfy-1exons 1B and 2 and intron 1B and part of intron 2. Construct 5 is the 5′ portion of the cosmid (SalI-EagI fragment) containing sequences 5′ of the EagI site. Construct 6 contains 21 kb of Zfy-15′ flanking sequence fused at the EagI site to lacZ. The lacZ probe used to analyze the structure of the coinjected transgenes is depicted below construct 6 (dotted rectangle) and contains the first 1.1 kb of lacZ sequence.

Fig. 1.

(A) 5′ Splicing of Zfy-1and Zfy-2. The splicing patterns of the 5′ exons of the Zfy-1and Zfy-2genes are based on the comparison of cDNA sequences (Mardon and Page, 1989; Ashworth et al., 1989) and genomic sequences for the two genes. Sequences of Zfy-1and Zfy-2 are nearly identical around exons 1B and 2, but diverge at some point greater than 2 kb 5′ of exon 1B. Exons are depicted as rectangles and splicing by thin lines. Arrows indicate the transcription start site. The drawing is not to scale. (B) Transgenes tested in transgenic mice. All fusions were made to an E. coli β-galactosidase reporter gene (dotted rectangle) that contains an intron and two exons (open rectangles and arrows) of the mouse Protamine-1 gene including the Protamine-1 polyadenylation signal. Closed rectangles depict exons of Zfy-1and are labelled as either exon 1B or 2. Construct 1 contains 3.5 kb of DNA 5′ of the cEMS177 exon 1A like sequences fused at the XbaI site to lacZ (the hatched sequences are the exon 1A-like sequences of cEMS177). Construct 2 contains 4.3 kb of Zfy-15′ flanking sequence fused at the unique EagI site in exon 1B to lacZ. Constructs 3, 4 and 5 were coinjected with construct 2. Construct 3 is cosmid cEMS 132 linearized with SalI. Construct 4 is the 3′ portion (EagI-SalI fragment) of the cosmid containing sequences 3′ of the EagI site including Zfy-1exons 1B and 2 and intron 1B and part of intron 2. Construct 5 is the 5′ portion of the cosmid (SalI-EagI fragment) containing sequences 5′ of the EagI site. Construct 6 contains 21 kb of Zfy-15′ flanking sequence fused at the EagI site to lacZ. The lacZ probe used to analyze the structure of the coinjected transgenes is depicted below construct 6 (dotted rectangle) and contains the first 1.1 kb of lacZ sequence.

Southern blot hybridization was carried out on genomic DNA obtained from these two lines of mice to examine the structure of the integrated transgenes produced by DNA coinjection. Fig. 2 shows the results obtained when DNA was digested with Dra III (unless otherwise stated) and hybridized with a probe specific for lacZ. Dra III cuts once within the −4.3 kb Zfy-1/lacZ construct (Fig. 1B, construct 2) and also cuts approximately 12 kb 5′ of the EagI site in the Zfy-1 gene. Lanes 4 and 5 contain the −4.3 kb Zfy-1/lacZ transgene that was used in the coinjections as a size reference (This reference DNA was not digested with Dra III but was released from plasmid sequences by EcoRI and HdIII digestion) and genomic DNA from a line of mice carrying only the −4.3 kb Zfy-1/lacZ transgene, respectively. The 8 kb band in both these lanes indicates that multiple copies of the −4.3 kb Zfy-1/lacZ transgene integrated in head-to-tail tandem array. Lane 6 contains the Dra III digestion of a reference plasmid that contains 21 kb of Zfy-1 5′ flanking sequence fused at the Eag I site to lacZ (Fig. 1B, construct 6). Lanes 7 and 8 contain Dra III-digested genomic DNA obtained from animals in lines 4572 and 4592, respectively. In both of these lines, a 13 kb band was observed that is the same size as that generated from the reference plasmid demonstrating that homologous recombination of the coinjected DNA molecules had occurred. Both lines contain multiple copies of the transgene and it appears that the vast majority of the copies have undergone the recombination event. Both lines contain the 8 kb band which would be expected if some of the −4.3 kb Zfy-1/lacZ fragment also integrated as head-to-tail repeats. Digestion of DNA from the two lines with MscI and Cla I confirmed that homologous recombination had occurred (data not shown).

Fig. 2.

Structure of the coinjected transgenes. Restriction enzyme (Dra III except for lane 4) digested DNA was subjected to Southern blot analysis with a lacZ probe (Fig. 1B, probe). The HdIII digested λ molecular mass standards are in lanes 1 and 9 (23.1, 9.4, 6.6, 4.3, 2.3 and 2.0 kb). Genomic DNA was obtained from a nontransgenic male mouse (lane 2), a nontransgenic female mouse (lane 3), a female mouse carrying just the −4.3 kb Zfy-1/lacZconstruct (lane 5), and female mice from lines 4572 (lane 7) and 4592 (lane 8) both of which carry coinjections of construct 5 and construct 2. Lane 4 contains the 8 kb construct 2 that was used for microinjection. Lane 6 contains the DraIII digestion of a reference plasmid that contains 21 kb of Zfy-15′ flanking sequence fused directly to lacZ at the EagI site (Fig. 1B, construct 6).

Fig. 2.

Structure of the coinjected transgenes. Restriction enzyme (Dra III except for lane 4) digested DNA was subjected to Southern blot analysis with a lacZ probe (Fig. 1B, probe). The HdIII digested λ molecular mass standards are in lanes 1 and 9 (23.1, 9.4, 6.6, 4.3, 2.3 and 2.0 kb). Genomic DNA was obtained from a nontransgenic male mouse (lane 2), a nontransgenic female mouse (lane 3), a female mouse carrying just the −4.3 kb Zfy-1/lacZconstruct (lane 5), and female mice from lines 4572 (lane 7) and 4592 (lane 8) both of which carry coinjections of construct 5 and construct 2. Lane 4 contains the 8 kb construct 2 that was used for microinjection. Lane 6 contains the DraIII digestion of a reference plasmid that contains 21 kb of Zfy-15′ flanking sequence fused directly to lacZ at the EagI site (Fig. 1B, construct 6).

To define a Zfy-1 promoter that could be used as a tool for directing expression of transgenes to organs of endogenous Zfy-1 expression without relying on recombination, a construct with approximately 21 kb of Zfy-1 5′ flanking sequence fused directly to lacZ was tested (Fig. 1B, construct 6). This transgene was used to produce 20 transgenic founder mice that were killed at E11.5-E12 and stained for β-galactosidase activity (Table 1). Five of the 20 founders expressed the lacZ gene in the genital ridge indicating that sequences that direct expression to the genital ridge reside within this 21 kb 5′ of Zfy-1.

Developmental expression of Zfy-1/lacZ in the gonad

We analyzed the developmental expression of the transgene in mouse lines 4572 and 4592, but because both lines revealed nearly identical developmental expression pattern in all organs, we only show data for the 4592 line that expresses at a higher level. The only difference in expression between the two lines was that the lower level of expression in the 4572 line resulted in a loss of detectable staining in the embryonic gonad after E14 in male mice. Fig. 3 shows the profile for β-galactosidase staining of the developing gonad. Staining was seen in both males and females, but females expressed at a higher level and expression persisted longer, whereas it was repressed in males shortly after the formation of testis cords. This demonstrates that other genes on the Y chromosome are not necessary for Zfy-1/lacZtransgene transcription and that testis differentiation may inhibit Zfy-1 expression. Previously it was reported that Zfy-1 mRNA was detected in the embryonic gonads of XY mice, which develop as females due to a mutation in the Sry gene (Gubbay et al., 1990). The staining we observed in the genital ridge was not detected at E10, but was faintly detected at E11 and increased and peaked at E12-E13 followed by a decrease to low levels by E15. This is the same pattern that was observed for Zfy-1 mRNA as detected by RT-PCR (Koopman et al., 1989) and a Zfy antibody detected only minimal amounts of protein in E13 testis and was unable to detect any Zfy protein by E15 (Su and Lau, 1992). The similarity between transgene expression and Zfy-1 mRNA and protein expression suggests that all the correct regulatory elements are present within this construct.

Fig. 3.

Expression of the Zfy-1/lacZtransgene in the developing embryonic gonad. Embryos were obtained from line 4592 mice at E10 through E15 and stained with X-Gal to reveal β-galactosidase activity. Staining is compared for both males and females. All gonads were photographed at the same magnification.

Fig. 3.

Expression of the Zfy-1/lacZtransgene in the developing embryonic gonad. Embryos were obtained from line 4592 mice at E10 through E15 and stained with X-Gal to reveal β-galactosidase activity. Staining is compared for both males and females. All gonads were photographed at the same magnification.

When RT-PCR was carried out on mRNA derived from the embryonic gonad of We homozygotes, no Zfy-1 message was detected (Koopman et al., 1989). We homozygotes are virtually devoid of germ cells in the gonad due to a mutation in the c-kit receptor; therefore, Koopman et al. (1989) suggested that Zfy-1 expression is germ cell dependent. Surprisingly, when we examined cross-sections of the stained genital ridges, we found the transgene was expressed in the somatic cells rather than the germ cells (Fig. 4A). To confirm this, E12 gonads were either stained just for alkaline phosphatase activity, which reveals germ cells (Chiquone et al., 1954) (Fig. 4B) or costained with X-Gal, to detect the transgene-expressing cells (Fig. 4C). The two stains clearly identify separate populations of cells, demonstrating that the transgene is expressed in the somatic cells and not the germ cells. Staining was carried out at E12 because there is less mixing of the somatic and germ cells prior to the formation of the testis cords. The gonadal staining pattern at this time is indistinguishable in males and females.

Fig. 4.

Cellular localization of Zfy-1/lacZ transgene expression in the embryonic gonad. Female embryos were obtained at E12 from line 4592 and stained with X-Gal alone (A; pink staining under dark-field illumination) to reveal transgene expressing cells, for alkaline phosphatase activity alone (B; orange-red stain under dark-field illumination) to identify primordial germ cells, or stained for both cell types (C; blue for the transgene expressing cells and red for cells with high alkaline phosphatase activity under bright-field illumination). Male E12 gonad is identical in staining pattern, but stains less intensely for lacZ activity. Gonads in A,B, and C were photographed at the same magnification (bar in A indicates 200 μm). m and g denote the mesonephros and gonads, respectively. (D-F) X-Gal staining was done on E13 littermates resulting from the crossing of a We/+, Zfy-1/lacZ+ (line 4592) male with a We/+, Zfy-1/lacZ+ (line 4592) female. D-F depict the stained gonads of +/+ homozygous, We/+ heterozygous, and We/We homozygous littermates, respectively. All three animals are hemizygous for the transgene.Gonads in D,E, and F werephotographed at the same magnification (bar in D indicates 1.0 mm).

Fig. 4.

Cellular localization of Zfy-1/lacZ transgene expression in the embryonic gonad. Female embryos were obtained at E12 from line 4592 and stained with X-Gal alone (A; pink staining under dark-field illumination) to reveal transgene expressing cells, for alkaline phosphatase activity alone (B; orange-red stain under dark-field illumination) to identify primordial germ cells, or stained for both cell types (C; blue for the transgene expressing cells and red for cells with high alkaline phosphatase activity under bright-field illumination). Male E12 gonad is identical in staining pattern, but stains less intensely for lacZ activity. Gonads in A,B, and C were photographed at the same magnification (bar in A indicates 200 μm). m and g denote the mesonephros and gonads, respectively. (D-F) X-Gal staining was done on E13 littermates resulting from the crossing of a We/+, Zfy-1/lacZ+ (line 4592) male with a We/+, Zfy-1/lacZ+ (line 4592) female. D-F depict the stained gonads of +/+ homozygous, We/+ heterozygous, and We/We homozygous littermates, respectively. All three animals are hemizygous for the transgene.Gonads in D,E, and F werephotographed at the same magnification (bar in D indicates 1.0 mm).

Because we observed somatic cell expression of the Zfy-1 transgene, but Koopman et al. (1989) were unable to detect Zfy-1message by RT-PCR in We gonads, we hypothesized that Zfy-1 is normally expressed in the somatic cells of the embryonic gonad, but requires the presence of germ cells to be activated. To test this hypothesis, the Zfy-1/lacZtransgene (line 4592) was introduced into the We homozygous background and the resulting embryos were stained for β-galactosidase activity at E13. The transgenic mice were identified by placental DNA dot hybridization and placental DNA was also used to identify We homozygotes using RFLP analysis. The We allele cosegregates with an RFLP of approximately 13.5 kb that is detected by digestion with BglII and hybridization of Southern blots with a mouse c-kit probe. Male and female heterozygous We mice that carried the transgene were mated and the RFLP analysis of their offspring allowed us to identify 9 We homozygous mice that carried the Zfy-1/lacZ transgene. These 9 We homozygous embryos all expressed the transgene at a normal level in the genital ridge (Fig. 4D-F for homozygous wild type, heterozygous, and homozygous We littermates, respectively, stained for β-galactosidase activity). These results demonstrate that the transgene is expressed in the developing gonad of We mice.

Zfy-1/lacZ expression in adult testis

The developmental time when each spermatogenic cell type first appears in the prepuberal mouse has been well defined (Bellve’ et al., 1977). Thus, for example, spermatocytes first appear in the mouse testis 9-10 days after birth and haploid round spermatids first appear after day 19. By examining both the timing of a gene’s expression in the prepuberal mouse testis and the position of the expressing cells in the seminiferous tubule, it is possible to establish what cell types express the gene. When the −4.3 kb Zfy-1/lacZtransgene was tested alone, expression was not detected in the genital ridge in any of 28 founder mice (Table 1), but expression was observed in the adult testis. There was no expression at day 1 (data not shown) or day 7 (Fig. 5C) after birth, but pachytene spermatocytes clearly stained at day 17 (Fig. 5D) and both spermatocytes and spermatids stained in the adult (Fig. 5E). The −4.3 kb Zfy-1/lacZ testis expression was examined in 6 lines including line 3287 [TgN(ZflacZ)219Bri] and 3283 [TgN(ZflacZ)220Bri].

Fig. 5.

Developmental expression of Zfy/lacZtransgenes in the testis after birth. Testes were isolated from animals containing the −3.5 kb cEMS177/lacZ construct (A,B). The −4.3 kb Zfy-1/lacZ transgene (C-E), or the coinjection of the 5 ′ cosmid fragment with the −4.3 kb Zfy-1/lacZ construct (line 4592) (F-I) at the times after birth indicated and stained with X-Gal. Cross sections were 10 μm and all sections were photographed at the same magnification (bar in I indicates 100 μm). All sections were counterstained with nuclear fast red. The −3.5 kb cEMS177/lacZ testis does not stain at day 17 (A), but intense blue staining is clearly seen in round and elongating spermatids in the adult animal (B, arrow). The -4.3 kb Zfy-1/lacZtestis does not stain at day 7 (C), but staining is clearly present in spermatocytes at day 17 (D, arrow) and persists in both spermatocytes (E, larger arrow) and spermatids (E, smaller arrow) in the adult. The line 4592 testis stains well in the interstitial cells (F, arrow) and extremely faintly in Sertoli cells at day 1. Sertoli cell (G, larger arrow) and lower level interstitial cell staining (G, smaller arrow) remains at day 7. At day 17 staining is seen clearly in spermatocytes (H, arrow) and staining persists in spermatocytes (I, larger arrow) and spermatids (I, smaller arrow) in the adult.

Fig. 5.

Developmental expression of Zfy/lacZtransgenes in the testis after birth. Testes were isolated from animals containing the −3.5 kb cEMS177/lacZ construct (A,B). The −4.3 kb Zfy-1/lacZ transgene (C-E), or the coinjection of the 5 ′ cosmid fragment with the −4.3 kb Zfy-1/lacZ construct (line 4592) (F-I) at the times after birth indicated and stained with X-Gal. Cross sections were 10 μm and all sections were photographed at the same magnification (bar in I indicates 100 μm). All sections were counterstained with nuclear fast red. The −3.5 kb cEMS177/lacZ testis does not stain at day 17 (A), but intense blue staining is clearly seen in round and elongating spermatids in the adult animal (B, arrow). The -4.3 kb Zfy-1/lacZtestis does not stain at day 7 (C), but staining is clearly present in spermatocytes at day 17 (D, arrow) and persists in both spermatocytes (E, larger arrow) and spermatids (E, smaller arrow) in the adult. The line 4592 testis stains well in the interstitial cells (F, arrow) and extremely faintly in Sertoli cells at day 1. Sertoli cell (G, larger arrow) and lower level interstitial cell staining (G, smaller arrow) remains at day 7. At day 17 staining is seen clearly in spermatocytes (H, arrow) and staining persists in spermatocytes (I, larger arrow) and spermatids (I, smaller arrow) in the adult.

In contrast, staining was observed at all developmental time points in the 4592 line that contains 28 kb of 5′ flanking DNA. Staining was restricted to Sertoli and Leydig cells at days 1 and 7 (Fig. 5F,G). Staining was abundant in spermatocytes at day 17 (Fig. 5H) and persisted at similar levels in spermatids of the adult testis (Fig. 5I). In the 4572 line, no staining was detected between E14 and the onset of meiosis (day 10 after birth), but staining resembled that of the 4592 line and the −4.3 kb Zfy-1/lacZ line at later stages (data not shown). This difference probably reflects the generally lower level of expression in the somatic cells of the 4572 line compared to the 4592 line. These results indicate that the enhancer responsible for adult germ cell expression resides within 4.3 kb upstream of exon 1B, whereas the enhancer for genital ridge expression lies 5′ of −4.3 kb. In all of the Zfy-1/lacZtransgenic mice, staining in round spermatids did not appear to increase in intensity relative to the staining seen in spermatocytes. This is in contrast to the large increase in Zfy (the probe used could not distinguish between Zfy-1 and Zfy-2 message) mRNA in spermatids as detected by northern blot (Nagamine et al., 1990). This difference will be discussed later. The 4592 line of mice also expressed the transgene in theca and granulosa cells of the ovary (data not shown). It is noteworthy that these cells derive from the same genital ridge precursors as the Sertoli and Leydig cells.

In the process of looking for exon 1A of Zfy-1 (using probes complementary to Zfy-2), cosmid cEMS177 was isolated. This cosmid contains a sequence that is clearly homologous to exon 1A of Zfy-2. There is 83% sequence identity over exon 1A and this identity extends at least 250 bp 5′ of exon 1A(unpublished data). This exon 1A-like sequence from cEMS177 and 3.5 kb of 5′ flanking DNA was fused to lacZ (Fig. 1B, construct 1) and the resulting construct was used to generate transgenic mice. When these mice were analyzed for expression, X-Gal staining was observed exclusively in round and elongating spermatids of the adult testis. Hence, staining was observed at day 25 and in adults (Fig. 5B) but not at day 17 (Fig. 5A). We postulate that expression of the cEMS177/lacZ transgene resembles that of Zfy-2.

Extragonadal expression of Zfy-1/lacZ

Lines 4572 and 4592 (containing 28 kb of coinjected Zfy-1 5′ flanking DNA) also express the transgene in the kidney, brain and spine prior to birth. The expression in the brain and spine was lower than that seen in the gonad as shown by the whole E12 embryo stained for 6 hours (Fig. 6A). The expression in E14 kidney was high in the parietal cells and the Bowmans capsule of glomeruli (Fig. 6B, larger arrow). Faint staining was also detected in cells lining the ducts (smaller arrow). Zfy protein was also detected in the ducts of the embryonic kidney (Su and Lau, 1992). Mature glomeruli are not present in the mouse kidney until about E14 and we did not detect strong kidney expression until about this time. Expression in the brain was found in the meninges (Fig. 6C, largest and smallest arrows) as well as in the choroid plexus (middle-sized arrow). Spine expression was localized at E14 to the meninges and the walls of some arteries (Fig. 6D). RT-PCR had detected Zfy-1 transcript in E12 brain and kidney and immunohistochemistry identified Zfyprotein in the kidney (Nagamine et al., 1990; Su and Lau, 1992). Ten of 20 founder mice carrying the −21 kb Zfy-1/lacZtransgene, 4 of 10 mice carrying the 5′ cosmid coinjection and 3 of 7 mice carrying the entire cosmid coinjection also stained in the brain and spine with the same pattern described for lines 4572 and 4592 (Table 1). All founders or lines except one that stained in the embryonic gonad also stained in the brain and spine.

Fig. 6.

Meninges and kidney expression of the Zfy-1/lacZtransgene. (A) An E12 mouse embryo (line 4592) was stained for 6 hours with X-Gal, revealing that the neural expression (A, arrow) is of a lower level than that seen in the embryonic gonad. Line 4592 transgenic mice were obtained at E14 (B-D) and stained overnight with X-Gal to reveal β-galactosidase activity. Kidney expression is found in the parietal cells and Bowman’s capsule of glomeruli (B, larger arrow) and very faintly in the cells surrounding the ducts (B, smaller arrow). k and m denote kidney and mesonephros, respectively. Expression in the brain (sagittal section) is found in the meninges (C, largest and smallest arrow), in the walls of some arteries and in the choroid plexus (C, middle-sized arrow). Spine expression is found in the meninges (D, arrow) and in the walls of some arteries. The bars in A and C indicate 1 mm and in B and D indicate 200 μm. B-D are under dark-field illumination (β-galactosidase staining appears pink).

Fig. 6.

Meninges and kidney expression of the Zfy-1/lacZtransgene. (A) An E12 mouse embryo (line 4592) was stained for 6 hours with X-Gal, revealing that the neural expression (A, arrow) is of a lower level than that seen in the embryonic gonad. Line 4592 transgenic mice were obtained at E14 (B-D) and stained overnight with X-Gal to reveal β-galactosidase activity. Kidney expression is found in the parietal cells and Bowman’s capsule of glomeruli (B, larger arrow) and very faintly in the cells surrounding the ducts (B, smaller arrow). k and m denote kidney and mesonephros, respectively. Expression in the brain (sagittal section) is found in the meninges (C, largest and smallest arrow), in the walls of some arteries and in the choroid plexus (C, middle-sized arrow). Spine expression is found in the meninges (D, arrow) and in the walls of some arteries. The bars in A and C indicate 1 mm and in B and D indicate 200 μm. B-D are under dark-field illumination (β-galactosidase staining appears pink).

The developmental timing of Zfy-1/lacZ fusion gene expression in both the embryonic gonad and in the testis after birth mimics the mRNA levels as detected by RT-PCR and northern blot analysis (Koopman et al., 1989; Nagamine et al., 1990). RT-PCR had also detected Zfy-1 mRNA in the E12 kidney and brain (Nagamine et al., 1990), organs where we detect the expression of our Zfy-1/lacZ transgenes. The only tissue in which Zfy-1 mRNA was detected in which we failed to demonstrate a corresponding expression of the transgene was the E12 liver (Nagamine et al., 1990). The remarkable resemblance of mRNA localization and transgene expression suggest that most if not all of the regulatory elements required for the normal expression of Zfy-1 lie within 28 kb 5′ of the start of Zfy-1 transcription.

The restriction of transgene expression to the somatic cells of the genital ridge was surprising as it had been suggested that Zfy-1 expression in the developing gonad is germ cell-dependent (Koopman et al., 1989, 1991b; Ashworth et al., 1991). In fact, this was one of the pieces of data suggesting that Zfy-1could not be the sex-determining gene on the Y chromosome. Simultaneous staining of germ cells for alkaline phosphatase and of transgene-expressing cells for β-galactosidase leaves little doubt that transgene expression is confined to the somatic cells in the genital ridge. Expression of the transgene in the E13 gonad of We homozygotes indicates that the transgene can be expressed in gonads deficient in germ cells. The question remains as to why the Zfy-1 mRNA was not detected by RT-PCR in gonads of We mice (Koopman et al., 1989). There are several possibilities. Koopman et al. (1989) identified We homozygotes by paleness of the liver at E14.5, which is a manifestation of acute anemia. As many We homozygotes die before birth, it is reasonable to assume that those with overtly pale livers at E14.5 are already very sick. Consequently, Zfy-1 transcription may already have ceased or transcription in general may be lower. The only control for this was the amplification of HPRT as a positive control and it is noticeably lower in We homozygotes compared to wild-type littermates. We stained at E13 because by E14.5 transgene expression level falls significantly in normal animals and in line 4572, expression is no longer detectable. In addition, Su and Lau (1992), using an antibody directed against Zfyprotein, detected only minimal protein in the E13 testis and were unable to detect any Zfyprotein by E15. Perhaps Zfy-1transcript levels are so low at E14.5 that even a small reduction in Zfy-1 transcript levels could place Zfy-1mRNA below a level detectable by RT-PCR. Another possibility is that the transgene expression in the somatic cells is ectopic and germ cells actually do express the endogenous Zfy-1 gene. This possibility seems unlikely because of the close resemblance of the transgene expression profile to the endogenous mRNA profile not only in the developing gonad but also in extragonadal tissues. The evidence suggests that Zfy-1 is expressed in the somatic compartment of the genital ridge regardless of the presence or absence of germ cells.

The expression of Zfy-1 in the somatic cells of the genital ridge is probably in the same cells that express Sry. These cells are the precursors for the Sertoli cells of the adult testis and granulosa cells of the adult ovary. The enhancer for somatic expression of Zfy-1 apparently lies between −4.3 and −21 kb, whereas the enhancer for expression in spermatocytes and ES cells is located within the proximal 4.3 kb (our unpublished observations). Sry is also expressed in germ cells of the adult including spermatocytes and spermatids (Rossi et al., 1993), and blastocysts (Zwingman et al., 1993) which emphasizes the similarity in expression of the two genes. Perhaps the enhancers for Zfy-1 and Sry may be similar and bind at least some of the same transcription factors. None of the factors required for embryonic gonad expression appear to be encoded by the Y chromosome because expression of the Zfy-1/lacZ and Sry transgenes (Koopman et al., 1991a) occurs in both males and females.

The cEMS177/lacZ transgenic mice may provide data concerning the regulation of the Zfy-2 gene because cEMS177 and Zfy-2 have homologous exon 1A and 5′ flanking sequences (83% sequence identity over 500 bp). We now know that cEMS177 is derived from an autosome (E.M. Simpson, unpublished observations). Considering that there are no sequences resembling exon 1A upstream of Zfy-1, further mapping of cEMS177 may provide clues regarding the absence of this exon and suggest how Zfy-1 and Zfy-2 evolved. cEMS177/lacZ transgene expression was restricted to round and elongating spermatids; hence we predict that regulatory elements in the corresponding region of the Zfy-2 promoter may direct expression to round spermatids. This would explain why the Zfy-1/acZ transgenes do not demonstrate increased expression levels in round spermatids relative to spermatocytes, whereas there is a dramatic increase in total Zfy mRNA detected between these two stages, by northern blot hybridization (their probe could not differentiate between Zfy-1 and Zfy-2 mRNAs; Nagamine et al., 1990). We hypothesize that before birth, Zfy-1 and possibly Zfy-2 are expressed at low levels in the somatic cells of the gonad, Zfy-1, and likely Zfy-2 (Nagamine et al., 1990), are expressed in germ cells at the onset of meiosis, and Zfy-2 subsequently is expressed at high levels in round spermatids. Verification of this hypothesis requires identification and testing of the Zfy-2 promoter. Interestingly, both cEMS177 and Zfy-2 contain multiple CRE-like half-site sequences within 200 bp 5′ of exon 1A (unpublished data). CRE-like sequences have been implicated in the spermatid-specific transcription regulation of the mouse protamine 1 gene (Zambrowicz et al., 1993) and the gene encoding the mouse testis-specific isoform of the angiotensin-converting enzyme (Howard et al., 1993). CRE half-sites can be bound by CREB in vitro and split CREs are capable of activating transcription with the orientation and distance between half-sites having little effect upon this activation (Fink et al., 1988). A testis-specific protein has been identified that appears after day 12 and binds to the CRE-like sequence of the mouse protamine 1 gene (Zambrowicz and Palmiter, 1994). Hence, the CRE-like sequences in cEMS177 and the Zfy-2 gene may also serve to regulate spermatid transcription of these genes.

Our data also provide clues to the potential function of Zfy-1. It has been suggested that Zfy-1 could be the Spy gene (Simpson and Page, 1991; Koopman et al., 1990b), a gene required for normal spermatogenesis, which is lost in the deletion from the Sxra region that produces Sxrb (Burgoyne et al., 1986; Sutcliffe and Burgoyne, 1989). Spy was identified genetically because XO Sxrb mice have a more severe phenotype than XO Sxra mice. In contrast to XO Sxra mice, XO Sxrb mice almost totally lack meiotic and post meiotic germ cells and Spy is therefore thought to be required for the proliferation of differentiating type A spermatogonia (Sutcliffe and Burgoyne, 1989). Levy and Burgoyne (1986) have demonstrated that XO germ cells in XO/XY chimeras show a similar if not identical defect in spermatogonial divisions to that seen in XO Sxrb mice. If we assume that this phenotypic similarity is entirely due to the absence of Spy in both cases, then that suggests Spy functions in a cell autonomous fashion within germ cells. The lacZ expression described here indicates that the only gonadal expression of Zfy-1 prior to the Spy arrest is in somatic cells. The germ cell autonomy of Spy expression and the lack of Zfy-1 expression in germ cells (prior to Spy arrest) argue against their being one and the same. If Zfy-1 has a function in the embryonic gonad, it must be subtle, as germ cells appear to develop normally to the spermatogonial stem cell stages in XO Sxrb mice and testes appear to form normally in XX SRYtransgenic male mice (Koopman et al., 1991a), both of which lack an intact Zfy-1 gene. It is difficult to determine the function of Zfy-1 in the germ cells during spermatogenesis because no mutations have been identified that affect only the Zfy-1 gene. All known mutations in Zfy-1 are part of larger chromosomal deletions. Targeted disruption of the Zfy-1 gene will be required to obtain a clear understanding of Zfy-1 function.

We thank Robert Braun for constructive discussions and support during the course of this work. We also thank Glenda Froelick for her histology work, Raj Kapur and Carol Quaife for helping analyze expression, and Mary Avarbock for DNA microinjection. We are also grateful for the We mice provided by Robin Lovell-Badge and the mouse c-kit probe provided by Peter Besmer. This work was supported in part by NIH grants HD-09172, HD-23657, and HD-22532. B. Z. was supported by PHS NRSA T32 GM-07270 from NIGMS. E. M. S. was the recipient of a Medical Foundation/Charles A. King Trust Fellowship.

Ashworth
,
A.
,
Swift
,
S.
and
Affara
,
N.
(
1989
).
Sequence of cDNA for murine Zfy-1, a candidate for Tdy
.
Nucl. Acids Res
.
17
,
2864
.
Ashworth
,
A.
,
Skene
,
B.
,
Swift
,
S.
and
Lovell-Badge
,
R
(
1990
).
Zfa is an expressed retroposon derived from an alternative transcript of the Zfx gene. EMBO J
.
9
,
1529
1534
.
Ashworth
,
A.
,
Sohaila
,
R.
,
Lovell-Badge
,
R.
and
Kay
,
G.
(
1991
).
X-chromosome inactivation may explain the difference in viability of XO humans and mice
.
Nature
351
,
406
408
.
Bellve’
,
A. R.
,
Cavicchia
,
J. C.
,
Millette
,
C. F.
,
O’Brien
D. A.
,
Bhatnager
,
Y. M.
and
Dym
M.
(
1977
).
Spermatogenic cells of the prepuberal mouse
.
J. Cell. Biol
.
74
,
68
85
.
Brinster
,
R. L.
,
Chen
,
H. Y.
,
Trumbauer
,
M.
,
Yagle
,
M. K.
and
Palmiter
,
R. D.
(
1985
).
Factors affecting the efficiency of introducing foreign DNA into mice by microinjecting eggs
.
Proc. Nat. Acad. Sci. USA
82
,
4438
4442
.
Burgoyne
,
P. S.
,
Levy
,
E. R.
and
McLaren
A.
(
1986
).
Spermatogenic failure in male mice lacking H-Y antigen
.
Nature
320
,
170
172
.
Cattanach
,
B. M.
(
1978
).
Another mutation at the W locus
.
Mouse News Letter
59
,
18
.
Eddy
,
E. M.
,
Clark
,
J. M.
,
Gong
,
D.
and
Fenderson
,
B. A.
(
1981
).
Origin and migration of primordial germ cells in mammals
.
Gamete Res
.
4
,
333
362
.
Fink
,
J. S.
,
Verhave
,
M.
,
Kasper
,
S.
,
Tsukada
,
T.
,
Mandel
,
G.
and
Goodman
,
R. H.
(
1988
).
The CGTCA motif is essential for biological activity of the vasoactive intestinal peptide gene cAMP-regulated enhancer
.
Proc. Nat. Acad. Sci. USA
85
,
6662
6666
.
Gubbay
,
J.
,
Koopman
,
P.
,
Collignon
,
J.
,
Burgoyne
,
P.
and
Lovell-Badge
,
R.
(
1990
).
Normal structure and expression of ZFY genes in XY female mice mutant in TDY
.
Development
109
,
647
653
.
Howard
,
T.
,
Balogh
,
R.
,
Overbeek
,
P.
and
Bernstein
,
K. E.
(
1993
).
Sperm-specific expression of Angiotensin-converting enzyme (ACE) is mediated by a 91-base-pair promoter containing a CRE-like element
.
Mol. Cell. Biol
.
13
,
18
27
.
Koopman
,
P.
,
Gubbay
,
J.
,
Collignon
,
J.
and
Lovell-Badge
,
R.
(
1989
).
Zfy gene expression patterns are not compatible with a primary role in mouse sex determination
.
Nature
342
,
940
942
.
Koopman
,
P.
,
Gubbay
,
J.
,
Vivian
,
N.
,
Goodfellow
,
P.
and
Lovell-Badge
,
R.
(
1991a
).
Male development of chromosomally female mice transgenic for Sry
.
Nature
351
,
117
121
.
Koopman
,
P.
,
Ashworth
,
A.
and
Lovell-Badge
,
R.
(
1991b
).
The Zfy-1 gene family in humans and mice
.
Trends Genet
.
7
,
132
136
.
Koopman
,
P.
,
Munsterberg
,
A.
,
Capel
,
B.
,
Vivian
,
N.
and
Lovell-Badge
,
R.
(
1991c
).
Expression of a candidate sex-determining gene during mouse testis differentiation
.
Nature
348
,
450
452
.
Levy
,
E.
and
Burgoyne
,
P. S.
(
1986
).
The fate of XO germ cells in the testes of XO/XY and XO/XY/XYY mouse mosaics: evidence for a spermatogenesis gene on the mouse Y chromosome
.
Cytogenet. Cell Genet
.
42
,
208
213
.
Mardon
,
G.
and
Page
,
D. C.
(
1989
).
The sex-determining region of the mouse Y chromosome encodes a protein with a highly acidic domain and 13 zinc fingers
.
Cell
56
,
765
770
.
Mardon
,
G.
,
Mosher
,
R.
,
Disteche
,
C. M.
,
Nishioka
,
Y.
,
McLaren
,
A.
and
Page
,
D. C.
(
1989
).
Duplication, deletion, and polymorphism in the sex-determining region of the mouse Y chromosome
.
Science
243
,
78
80
.
Mardon
,
G.
,
Luoh
,
S. W.
,
Simpson
,
E. M.
,
Gill
,
G.
,
Brown
,
L. G.
and
Page
,
D. C.
(
1990
).
Mouse Zfx protein is similar to Zfy-2: each contains an acidic activating domain and 13 zinc fingers
.
Mol. Cell. Biol
.
10
,
681
688
.
Mercer
,
E.H.
,
Hoyle
,
G.W.
,
Kapur
,
R.P.
,
Brinster
,
R.L.
and
Palmiter
,
R.D.
(
1991
).
The Dopamine β-hydroxylase gene promoter directs expression of E. coli lacZ to sympathetic neurons in adult transgenic mice
.
Neuron
7
,
703
716
.
Miller
,
J.
,
McClachlan
,
A. D.
and
Klug
,
A.
(
1985
).
Repetitive zinc binding domains in the protein transcription factor IIIa from xenopus oocytes
.
EMBO J
.
4
,
1609
1614
.
Nagamine
,
C. M.
,
Chan
,
K.
,
Kozak
,
C. A.
and
Lau
,
Y.-F.
(
1989
).
Chromosome mapping and expression of a putative testis-determining gene in the mouse
.
Science
243
,
80
83
.
Nagamine
,
C. M.
,
Chan
,
K.
,
Hake
,
L. E.
and
Lau
,
Y.-F.
(
1990
).
The two candidate testis-determining Y genes ( Zfy-1 and Zfy-2 ) are differentially expressed in fetal and adult mouse tissues
.
Genes Dev
.
4
,
63
74
.
Page
,
D. C.
,
Mosher
,
R.
,
Simpson
,
E. M.
,
Fischer
,
E. M.
,
Mardon
,
G.
,
Pollack
,
J.
,
McGillivray
,
B.
,
de la Chapelle
,
A.
and
Brown
,
L. G.
(
1987
).
The sex-determining region of the human Y chromosome encodes a finger protein
.
Cell
51
,
1091
1104
.
Pavletich
,
N. P.
and
Pabo
,
C. O.
(
1991
).
Zinc finger-DNA recognition: crystal structure of a Zif-268-DNA complex at 2.1 A
.
Science
252
,
809
817
.
Rossi
,
P.
,
Dolc
,
S.
,
Abanesi
,
C.
,
Grimaldi
,
P.
and
Geremia
,
R.
(
1993
).
Direct evidence that the mouse sex-determining gene Sry is expressed in the somatic cells of male fetal gonads and in the germ cell line in adult testis
.
Molec. Reprod. Dev
.
34
,
369
373
.
Sambrook
,
J.
,
Fritsch
,
E. F.
and
Maniatis
,
T.
(
1989
).
In Molecular Cloning; A Laboratory Manual
.
Cold Spring Harbor, New York
:
Cold Spring Harbor Laboratory Press
.
Simpson
,
E. M.
and
Page
,
D. C.
(
1991
).
An interstitial deletion in mouse Y chromosomal DNA created a transcribed Zfyfusion gene
.
Genomics
11
,
601
608
.
Sinclair
,
A. H.
,
Berta
,
P.
,
Palmer
,
M. S.
,
Hawkins
,
R.
,
Griffiths
,
B. L.
,
Smith
,
M. J.
,
Foster
,
J. W.
,
Frischauf
,
A. M.
,
Lovell-Badge
,
R.
and
Goodfellow
,
P. N.
(
1990
).
A gene from the human sex-determining region encodes a protein with homology to a conserved DNA-binding motif
.
Nature
346
,
239
244
.
Su
,
H.
and
Lau
,
Y.-F. C.
(
1992
).
Demonstration of a stage-specific expression of the Zfy protein in fetal mouse testis using anti-peptide antibodies
.
Molec. Reprod. Dev
.
33
,
252
257
.
Sutcliffe
,
M. J.
and
Burgoyne
,
P. S.
(
1989
).
Analysis of the testes of H-Y negative XOSxrb mice suggests that the spermatogenesis gene (Spy ) acts during the differentiation of the A spermatogonia
.
Development
107
,
373
380
.
Tam
,
P. P. L.
and
Snow
,
M. H. L.
(
1981
).
Proliferation and migration of primordial germ cells during compensatory growth in mouse embryos
.
J. Embryol. exp. Morph
.
64
,
133
147
.
Zambrowicz
,
B. P.
,
Harendza
,
C. J.
,
Zimmermann
,
J. W.
,
Brinster
,
R. L.
and
Palmiter
,
R. D.
(
1993
).
Analysis of the mouse protamine 1 promoter in transgenic mice
.
Proc. Nat. Acad. Sci. USA
90
,
5071
5075
.
Zambrowicz
,
B. P.
and
Palmiter
,
R. D.
(
1994
).
Testis specific and ubiquitous factors bind to functionally important regions of the mouse protamine-1 gene
.
Biol. Reprod
.
50
,
65
72
.
Zwingman
,
T.
,
Erickson
,
R. P.
,
Boyer
,
T.
and
Asangla
,
A.
(
1993
).
Transcription of the sex-determining region genes Sryand Zfy in the mouse preimplantation embryo
.
Proc. Nat. Acad. Sci. USA
90
,
814
817
.