Maternal Groucho and bHLH repressors amplify the dose-sensitive X chromosome signal in Drosophila sex determination

Abstract

In Drosophila, XX embryos are fated to develop as females, and XY embryos as males, because the diplo-X dose of four X-linked signal element genes, XSEs, activates the Sex-lethal establishment promoter, SxlPe, whereas the haplo-X XSE dose leaves SxlPe off. The threshold response of SxlPe to XSE concentrations depends in part on the bHLH repressor, Deadpan, present in equal amounts in XX and XY embryos. We identified canonical and non-canonical DNA-binding sites for Dpn at SxlPe and found that cis-acting mutations in the Dpn-binding sites caused stronger and earlier Sxl expression than did deletion of dpn implicating other bHLH repressors in Sxl regulation. Maternal Hey encodes one such bHLH regulator but the E(spl) locus does not. Elimination of the maternal corepressor Groucho also caused strong ectopic Sxl expression in XY, and premature Sxl activation in XX embryos, but Sxl was still expressed differently in the sexes. Our findings suggest that Groucho and associated maternal and zygotic bHLH repressors define the threshold XSE concentrations needed to activate SxlPe and that they participate directly in sex signal amplification. We present a model in which the XSE signal is amplified by a feedback mechanism that interferes with Gro-mediated repression in XX, but not XY embryos.

Figure 1. Binding of Dpn to canonical and non-canonical DNA sequences at SxlPe

A) DNase I footprinting with the indicated units of full-length MBP-Dpn fusion protein. One unit equaled 0.3 pmole (15 nM) MBP-Dpn. Left panel, protection of Dpn-binding site 4, right panel, protection of Dpn-binding sites 1, 2, and 3. Six bp core sequences are indicated. Probes extended from −204 to −373 (left) and −229 to +72 (right). Protection of Dpn-binding site 3 is also visible in Fig. 6 of Hoshijima et al., (1995). B) Electrophoretic mobility-shift assays (EMSA). Indicated units of GST-Dpn bHLH fusion protein were incubated with ³²P-labeled oligonucleotides and the complexes resolved on polyacrylamide gels. One unit equaled 0.3 pmole (15 nM) GST-Dpn bHLH protein. Core sequences for the Dpn-sites are shown. Sequences of probes (1+2), 4a, 3, 3Cm, 4Cm and (1+2)m are in Table 1. C) Binding site competition in EMSA. Complexes were formed between GST-Dpn bHLH protein (0.02 units) and a ³²P-labeled site (1+2) probe and challenged with 10- to 160-fold molar excesses of oligonucleotides (1+2), 3, or 4 as competitors.

Figure 2. Location of protein-binding sites at SxlPe

Diagram represents sequences from −1.1 kb to +1 relative to start of transcription. Triange apices denote positions of identified protein-binding sites. Ten binding sites for the activator Sc/Da (Yang et al. 2001) and two binding sites for the activator STAT (Avila and Erickson 2007) are shown above the line. The five HES-class repressor-binding sites are numbered and shown below the line. Core HES-binding sequences are capitalized. Sequences from + 42 to −392 are sufficient for sex-specific expression of SxlPe but sequences to −1.4 kb are needed for near wild-type expression (Estes et al. 1995).

Figure 3. Canonical and non-canonical DNA sequences mediate HES protein-binding at SxlPe

(A) Dpn-VP16 activates transcription in SL-2 cells via predicted Dpn-binding site 1, 2, 3, and 4 sequences. Four copies of Dpn-binding sequences were joined to a −95bp SxlPe-luciferase reporter and co-transfected with an actin5C promoter-Dpn-VP16 expression vector. Data are expressed as luciferase activity with actin-Dpn-VP16 relative to the actin5C promoter control (+/− one standard deviation). (B) Repression of SxlPe-lacZ by anteriorly-expressed hairy-engrailed (hb-h-en). In situ hybridizations detect SxlPe-lacZ mRNA in embryos carrying wild-type (wt) or mutant (1⁻2⁻, 3⁻4⁻, 1⁻2⁻3⁻4⁻) Dpn-binding sites. Female embryos shown, Dpn-site mutant transgenes responded similarly to hb-h-en in males. (C) Ectopic expression of SxlPe-lacZ transgenes carrying Dpn-binding site mutations. In situ hybridizations to detect lacZ mRNA. All embryos carry two copies of the indicated SxlPe-lacZ transgenes inserted on an autosome.

Figure 4

Sxl protein in gro^mat- embryos. Embryos from mothers bearing gro^E48 germline clones were immunostained stained for Sxl. Embryonic stages are mid-cellularization (left) and gastrulation (right). (Top panels) XX and XY embryos bearing normal doses of the X-linked Sxl gene. (Bottom panels) XX and XY embryos each with one functional copy of Sxl⁺ were the progeny of females with FRT82B gro^E48 germ cells and y w cm Sxl^f1 ct/Y males.

Figure 5. Time course of SxlPe activation in wild-type, Δdpn², and maternal gro^E48 mutant embryos

Wild-type and mutant embryos were stained following in situ hybridization. Black and white panels show surface views of embryonic nuclei at indicated nuclear cycles. Dots represent nascent transcripts from the X-linked Sxl locus. Cycle 12 nuclei were illuminated with UV and visible light to enhance DAPI-stained nuclei. Color panels show peak accumulation of SxlPe-derived mRNA in early cycle 14. Embryos were progeny of wild-type (wt, w¹¹¹⁸) females and males, w¹¹¹⁸; Δdpn²/CyO females and males, or females with FRT82B gro^E48/FRT82B gro^E48 germ lines and w¹¹¹⁸/Y males. Cycle 12 embryos from Δdpn²/ crosses could not be distinguished from Δdpn² heterozygotes or wildtype of the same sex. Time courses are representative of repeated stainings of embryos from four separate inductions of gro^E48 germline clones and five series of embryo collections from crosses between Δdpn² heterozygotes.

Figure 6. Maternal Hey negatively regulates SxlPe

Surface views of embryos at indicated nuclear cycles stained after in situ hybridization to detect nascent transcripts from the X-linked SxlPe. Top row: wild-type XX embryos. Middle and bottom rows: XX and XY progeny of mothers carrying Hey^f06656 germline clones (Hey^mat-). Wild-type XY embryos do not activate SxlPe.

Figure 7. Model for dose-sensitive regulation of SxlPe

(Top) In XX embryos Gro and other products of maternally supplied mRNAs establish initial threshold XSE concentrations by binding to SxlPe. XX embryos exceed threshold [XSE] in cycle 12. Increased histone acetylation, arising from the activation of SxlPe, inhibits Gro-mediated repression allowing the XSE proteins to more effectively stimulate transcription from SxlPe during cycles 13 and 14. (Bottom) In XY embryos continued translation of gro and other maternal mRNAs maintains repression potential above XY XSE concentrations until cycle 13. Zygotic expression of Dpn combined with maternal gro thereafter maintains repression potential above XY [XSE]. The amounts of XSEs differ by two-fold in XX and XY embryos. The amounts of Gro and other maternal or autosomal regulators are equal in both sexes, but their repressive potentials differ because of the proposed feedback mechanism. Y axis represents XSE concentrations and repressor/corepressor function. XSE mRNAs are degraded early in cycle 14. Time scale; cycle 13 is 18 min long and begins 112 min after fertilization (Foe et al., 1993).