Modulation of MSL1 Abundance in Female Drosophila Contributes to the Sex Specificity of Dosage Compensation

Corresponding author: Mitzi I. Kuroda, Howard Hughes Medical Institute, Baylor College of Medicine, 1 Baylor Plaza, Houston, TX 77030., mkuroda{at}bcm.tmc.edu (E-mail).

Communicating editor: V. G. FINNERTY

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Dosage compensation in Drosophila is the mechanism by which X-linked gene expression is made equal in males and females. Proper regulation of this process is critical to the survival of both sexes. Males must turn the male-specific lethal (msl)-mediated pathway of dosage compensation on and females must keep it off. The msl2 gene is the primary target of negative regulation in females. Preventing production of MSL2 protein is sufficient to prevent dosage compensation; however, ectopic expression of MSL2 protein in females is not sufficient to induce an insurmountable level of dosage compensation, suggesting that an additional component is limiting in females. A candidate for this limiting factor is MSL1, because the amount of MSL1 protein in females is reduced compared to males. We have identified two levels of negative regulation of msl1 in females. The predominant regulation is at the level of protein stability, while a second regulatory mechanism functions at the level of protein synthesis. Overcoming these control mechanisms by overexpressing both MSL1 and MSL2 in females results in 100% female-specific lethality.

FOR the majority of animal species, males and females have the same complement of chromosomes, except for the sex chromosomes. Females in many species have two X chromosomes, whereas males only have one. This inequality would result in disproportionate sex-linked gene expression if not prevented by dosage compensation. Dosage compensation is achieved by a variety of mechanisms in the organisms studied to date. In Drosophila, the amount of transcription of most genes from the single male X chromosome is increased to equal that of the two female X chromosomes. Four autosomal genes are known to be required for this process: maleless (mle) and male-specific lethal 1, 2, and 3 (msl1, 2, 3), and they are collectively referred to as the msls (FUKUNAGA et al. 1975 ; BELOTE and LUCCHESI 1980 ; UCHIDA et al. 1981 ). The MSL proteins are postulated to form a multiprotein complex, which can be detected at hundreds of sites along the length of the male X chromosome (KURODA et al. 1991 ; GORMAN et al. 1993 ; PALMER et al. 1993 ; BASHAW and BAKER 1995 ; GORMAN et al. 1995 ; KELLEY et al. 1995 ; ZHOU et al. 1995 ). An additional gene involved in dosage compensation, males absent on first (mof), encodes a protein with an acetyl transferase domain (HILFIKER et al. 1997 ). The MOF protein may provide the link between the X chromosome association of the MSLs and the detection of histone H4 acetylated at lysine 16 (H4Ac16) in a pattern that is highly coincident with the MSLs (TURNER et al. 1992 ; BONE et al. 1994 ; RASTELLI et al. 1995 ; HILFIKER et al. 1997 ). The current model postulates that the MSL proteins, along with MOF and H4Ac16, alter chromatin architecture to facilitate transcription.

The effectiveness of MSL-mediated dosage compensation relies upon sex-specific regulation that turns the mechanism on in males and keeps it off in females. Setting this regulatory switch is critical to the survival of both males and females. While all four msls and mof are transcribed in both sexes, the lack of MSL2 protein and the reduced levels of MSL1 and MSL3 proteins in females indicate that the msls are regulated post-transcriptionally (KURODA et al. 1991 ; PALMER et al. 1993 ; BASHAW and BAKER 1995 ; GORMAN et al. 1995 ; KELLEY et al. 1995 ; ZHOU et al. 1995 ; HILFIKER et al. 1997 ).

Analyses of msl2 have determined that its transcript is the primary target of negative regulation in females (BASHAW and BAKER 1995 ; KELLEY et al. 1995 ; ZHOU et al. 1995 ). Ectopic expression of msl2 in females from the H83M2 transgene is sufficient to assemble the MSLs into complexes on both X chromosomes (KELLEY et al. 1995 ). Additional studies of msl2 regulation revealed that the female-specific Sex lethal (Sxl) protein has the ability to bind the polyuridine tracts within the 5' and 3' untranslated regions (UTRs) of the transcript, and these Sxl-binding sites are required for proper regulation of msl2 (BASHAW and BAKER 1997 ; KELLEY et al. 1997 ). The association of SXL protein with the msl2 transcript in females interferes with the translation of msl2, but the precise mechanism is not known.

Although ectopic expression of MSL2 protein in females results in the assembly of MSL complexes on the X chromosomes, up to 20% of these females survive to adulthood (KELLEY et al. 1995 ), suggesting that their X chromosomes may not be fully dosage compensated. Thus, one or more of the other MSLs may be limiting in females under these conditions. msl1 is a candidate for negative regulation in females because: (1) the level of MSL1 protein is dramatically lower in females compared to males (PALMER et al. 1994 ; GORMAN et al. 1995 ), (2) MSL1 protein accumulates to much higher levels in females ectopically expressing MSL2 compared to wild-type females (KELLEY et al. 1995 ), (3) the females that ectopically express MSL2 are sensitive to the endogenous level of msl1 (KELLEY et al. 1995 ), and (4) the 3' UTR of msl1 contains putative SXL-binding sites (PALMER et al. 1993 ).

In this study, we examined the mechanism of msl1 sex-specific regulation and found evidence for two levels of post-transcriptional control. Our results support the hypothesis that females negatively regulate the amount of MSL1 protein present in their cells by controlling MSL1 protein stability. We also found evidence for down regulation of MSL1 protein synthesis in females. Due to these regulatory mechanisms, MSL1 fails to accumulate in females in the absence of MSL2. When transiently overexpressed in females, MSL1 has an affinity for chromatin, which may provide insight into its function within the MSL complex.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Fly stocks:
The msl1²⁶⁹ and msl1¹⁹⁰ alleles have been characterized (PALMER et al. 1993 ). The msl1^L60 allele is a 2-kb deletion generated by irradiation (R. KELLEY, unpublished data). This deletion begins in the 5' UTR and removes most of the coding region. Females with the H83M2 msl2 transgene (KELLEY et al. 1995 ) or the msl2-NOPU transgene (KELLEY et al. 1997 ) were used for assaying the effects of msl1 transgenes in females. The characteristics of yw flies, and the CyO, TM3, and TM6 balancer chromosomes are documented in LINDSLEY and ZIMM 1992 .

Transgenic constructs:
The M1-FL transgene consists of 8.2 kb of genomic DNA from the msl1 locus in the pCaSpeR 4 (pCSR4) vector (PIRROTTA 1988 ). The M1-MPU transgene was identical to the M1-FL transgene, except the four clustered polyuridine tracts were mutated. PCR mutagenesis (HIGUCHI et al. 1988 ) was performed to introduce mutations in each of the four polyuridine tracts. The primer sequences used for the mutagenesis are as follows: primer A: 5'AAA GTT TTG TCG TTT CTT AT3'; primer B: 5'CTA CAA TGG TGG CAA TTA A3'; primer 1: 5'CAT AAG TTA TAT TGG GCA TCT CTC TGG AGA AAA ACA ATG3'; primer 2: 5'GGT TAT AAG GGC TAA AGA GAG AGA ATA TTT AAA AAT ATA AG3'; primer 3: 5'TTT AGC CCT TAT AAC CAT TCT CTC TGT CTT TTA TAC TG3'; primer 4: 5'CAC AAT CAA ATC CTA AGA GAG AGA CAA ATT ATC AGA G3'. Primers A and B are proximal and distal to the polyuridine tracts, respectively, and the numbered primers contained the mutated bases required to alter each of the four tracts. The M1-T3UTR transgene consists of 6.4 kb of msl1 genomic DNA. The msl1 sequence in this transgene ends 99 bp after the first polyadenylation site. The M1-ECTOPIC transgene was constructed by cloning the coding sequence of msl1 into the pCaSpeR HS 83 vector (HORABIN and SCHEDL 1993 ).

Transgenic lines:
Plasmid DNA of the msl1 constructs was purified using a Qiagen column (Qiagen, Inc., Chatsworth, CA) and injected into yw; +; Ki p^p [Js 2-3]/+ embryos (ROBERTSON et al. 1988 ), according to published protocols (RUBIN and SPRADLING 1982 ; SPRADLING and RUBIN 1982 ). The G₀ progeny were mated to w; msl1²⁶⁹/CyO flies. Transgenic male F₁ progeny that also contained the CyO balancer chromosome were mated to homozygous w; msl1²⁶⁹ virgin females to determine the chromosome of insertion and to determine the ability of the transgene to rescue homozygous msl1²⁶⁹ male lethality. All transgenic progeny were identified through the expression of the white gene in a white mutant background.

Crosses:
The crosses used to study the phenotypic effect of overexpressing MSL1 and MSL2 in females were as follows. w; msl1²⁶⁹/CyO; [w⁺, H83M2] females were crossed to w; [w⁺, M1-ECTOPIC]/TM3 males to produce the H83M2 and H83M2/M1-ECTOPIC progeny. w; msl1¹⁹⁰/CyO females were crossed to w; msl1²⁶⁹; [w⁺, M1-ECTOPIC]/+ males to produce msl1/+ M1-ECTOPIC progeny. yw females were mated to w; [w⁺, M1-ECTOPIC]/TM6 males to produce M1-ECTOPIC progeny with two endogenous copies of msl1⁺. Progeny from which salivary gland chromosomes were immunostained were generated by crossing w; msl1²⁶⁹; [w⁺, H83M2] females to w; [w⁺, M1-ECTOPIC]/TM6 males. To assay the phenotypic effects of the genomic msl1 transgenes, w; msl1²⁶⁹/CyO; [w⁺, H83M2] females were crossed to w; [w⁺, msl1 transgene]/TM3 males and the progeny was analyzed. To study the MSL X chromosomal-banding pattern associated with each of the genomic msl1 transgenes, w; msl1²⁶⁹; [w⁺, msl1 transgene] females were crossed to w, [w⁺, msl2-NOPU]/Y; msl1^L60/CyO males, and chromosomes from the salivary glands of the resulting progeny were immunostained. In the case of lines M1-T3UTR 47A and M1-FL 48D, females that were heterozygous for msl1 over the CyO balancer were used for the cross. Unless otherwise indicated, all crosses were performed at 25°.

Western analysis:
Western analysis was performed on protein extracts from whole adult flies or whole third instar larvae. Whole adult and larval lysates were made according to the protocol of PALMER et al. 1994 . The extracts were electrophoresed through a polyacrylamide SDS gel and transferred to nitrocellulose membrane (PALMER et al. 1994 ). Affinity-purified rabbit anti-MSL1 antibody (PALMER et al. 1994 ), anti-MSL2 antibody (KELLEY et al. 1995 ), anti-MLE antibody (RICHTER et al. 1996 ), or rabbit anti-LEONARDO antibody (SKOULAKIS and DAVIS 1996 ) was used at the appropriate dilution. The MSL1, MSL2, MLE, and LEO proteins were detected on the blot using an alkaline phosphatase-conjugated goat anti-rabbit IgG detection kit (Promega, Madison, WI).

Northern analysis:
Northern analysis was performed on 5 µg of poly(A)⁺ RNA from adult females. The RNA was electrophoresed through a formamide gel and transferred to Hybond N⁺ membrane (Amersham, Arlington Heights, IL). A 3.0-kb fragment of msl1, which contains the open reading frame, was used as a probe to identify the msl1 transcripts. A probe for the alcohol dehydrogenase (Adh) gene was used to indicate the amount of RNA in each lane.

Polytene chromosome squashes:
Polytene chromosome squashes were performed according to standard protocols (KURODA et al. 1991 ; BONE et al. 1994 ). For the heat shock experiment, the larvae were placed at 37° for 1 hr and allowed to recover for 1 hr at 22° before their salivary glands were removed. The tissue was immunostained according to PIRROTTA et al. 1988 with a few modifications. Affinity-purified rabbit anti-MSL2 antibody (KELLEY et al. 1995 ) and goat anti-MSL1 antibody (PALMER et al. 1994 ) were diluted appropriately in PBT (137 mM NaCl, 2.6 mM KCl, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, 0.2% Triton X-100) with 2% BSA and added to the tissue for 16 hr at 4°. The secondary antibodies, FITC- or Texas Red-conjugated donkey anti-rabbit or anti-goat (Jackson Immuno Research Laboratories, Inc., West Grove, PA), were diluted 1:200 in PBT and 2% BSA and incubated with the tissue for 4 hr at 22°. After incubating with the secondary antibodies, the tissue was stained with 2 µg/ml bisbenzimide (Hoechst 33258, Sigma Chemical, St. Louis, MO) for 5 sec. The tissue was mounted by placing 2% n-propyl galate/80% glycerol between the tissue and a coverslip. The chromosomes were observed using a Zeiss Axioskop (Zeiss, Thornwood, NY) and epifluorescence optics. Photographs were taken with Ektachrome 400 film.

In the case of the msl2-NOPU transgene squashes, there was a mixed population of msl1²⁶⁹/msl1^L60, msl1²⁶⁹/CyO, and in two crosses msl1^L60/CyO females. After the salivary glands were removed, the larval bodies were individually homogenized in 50 µl of 10 x PCR buffer (12.5 mM MgCl₂, 500 mM KCl, 100 mM Tris, pH 8.3, 0.1% gelatin), plus 0.2 mg/ml Proteinase K; then the samples were held on ice for 1–2 hr. The homogenates were incubated at 37° for 30 min and extracted with phenol:chloroform:isoamyl alcohol (25:24:1) and ethanol precipitated. After resuspending the pellets in 50 µl of deionized water, 2 µl was used in 25-µl PCR reactions to identify the presence of the L60 allele and when necessary, also the 269 allele of msl1. The primers used for the amplification were as follows: 5'ATC TCT ATG CCC CCA ATC AA3' and 5'TGT GGT AAT CGT TAC TGT TAA CTC TGG3' for the L60 allele and 5'GCA TGA ATC AGG ACT TCG AGC ACC3' and 5'CTA CAA TGC TGG CAA TTA A3' for the 269 allele. Both sets of primers are 3.6 kb apart in the msl1 locus but amplify 1.1-kb fragments from their respective alleles. The amplification consisted of 30 cycles of 94° for 1 min, 50° for 1 min, and 72° for 2 min. The PCR reactions were electrophoresed through 1% agarose gels, and the products were visualized through the use of ethidium bromide. After the genotype of each larva was determined in this manner, the appropriate slides were incubated with antibodies and the immunofluorescence protocol was followed.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

MSL1 protein stability is regulated by MSL2:
Several results have led to the hypothesis that MSL2 affects the level of MSL1: the level of MSL1 protein in msl2 mutant males is significantly lower than in wild-type males (PALMER et al. 1994 ), while MSL1 levels are increased in females that ectopically express MSL2 from the H83M2 transgene (KELLEY et al. 1995 ; Figure 1A). MSL2 is not required for msl1 RNA synthesis (PALMER et al. 1994 ); therefore, there are two potential mechanisms by which MSL2 could affect MSL1 protein levels: MSL2 may facilitate the synthesis of MSL1 protein or MSL2 may physically interact with MSL1, thereby stabilizing it.

View larger version (38K): In this window In a new window Download PPT slide   Figure 1. Western analysis of male and female third-instar larvae using an antibody against the MSL1 protein. A: (1) wild-type male, (2) wild-type female, (3) w; msl1L60/+; [w+, H83M2]/+ male, and (4) w; msl1L60/+; [w+, H83M2]/+ female. B: (1) wild-type male, (2) wild-type female, (3) w; msl1269/ msl1L60; [w+, M1-ECTOPIC]/+ males, (4) w; msl1269/msl-1L60; [w+, M1-ECTOPIC]/+ females, (5) w; msl-1269/msl-1L60; [w+, M1-ECTOPIC]/[w+, H83M2] males, and (6) w; msl-1269/msl-1L60; [w+, M1-ECTOPIC]/[w+, H83M2] females. As a control for loading, the blots were also analyzed using an antibody against the LEO protein.

To test how MSL2 regulates MSL1 accumulation, we constructed a transgene, M1-ECTOPIC, in which the 5' and 3' UTRs and the intron of msl1 were removed. The M1-ECTOPIC transgene contained the open reading frame of the MSL1 protein, the promoter of the hsp83 gene, and the 3' UTR from the transformer 2 gene. Although any regulatory elements located outside the msl1 open reading frame were eliminated, we found that the M1-ECTOPIC transgene retained sex-specific regulation. This transgene encodes sufficient MSL1 protein to rescue msl1 mutant males and to produce a strong 170-kD band on Westerns (Figure 1B). Females carrying the same transgene have much lower levels of MSL1 but dramatically increase MSL1 if ectopic MSL2 is supplied by an H83M2 transgene (Figure 1B). Because an endogenous msl1⁺ gene and our M1-ECTOPIC transgene display similar MSL1 protein profiles (low in the absence of MSL2 and high in the presence of MSL2), we conclude that protein stabilization is the primary mode of MSL1 regulation.

Constitutive expression of MSL1 is lethal to females ectopically expressing MSL2:
In the simplest model, msl1 would be regulated solely by protein stability. If this were true, males and females would synthesize the same amount of MSL1 protein. The MSL1 protein would turn over rapidly in females, but it would be shielded from destruction by binding to MSL2 in males. From this model, we would predict that lethal amounts of MSL1 would accumulate in H83M2 females which ectopically express high levels of MSL2. Although most H83M2 females die, ~20% survive to adulthood, indicating that full dosage compensation involves a second limiting component in addition to MSL2 (KELLEY et al. 1995 ).

The phenotype of the H83M2 females can be suppressed by reducing the level of endogenous msl1 by half (KELLEY et al. 1995 ). This indicates that msl1 may be a limiting factor in normal females and may be subjected to a second level of negative regulation. In addition to modulation of MSL1 protein stability, there may be a negative regulatory mechanism to reduce the synthesis of MSL1 protein in females.

To investigate the hypothesis that H83M2 female survival is due to the limiting abundance of MSL1 protein, we generated females with both the H83M2 and M1-ECTOPIC transgenes. The M1-ECTOPIC transgene alone had no phenotypic effect on females, presumably due to the absence of MSL2. However, overexpression of both MSL1 and MSL2 resulted in >99% female-specific lethality (Table 1). These transgenic females die during late larval development, which coincides with the time of death of msl mutant males (BELOTE and LUCCHESI 1980 ; OKUNO et al. 1984 ). Males carrying these two transgenes were fully viable.

Females that are msl1/+; [H83M2]/[M1-ECTOPIC] survive to the third instar larval stage, allowing examination of MSL complex formation on polytene chromosomes. Chromosomes were immunostained with antibodies to MSL1 (Figure 2) and MSL2 (data not shown). No MSL1/MSL2 complexes were detected in females with the M1-ECTOPIC transgene alone (Figure 2A). However, MSL1/MSL2 complexes were present along the length of the X chromosomes in dying females with both M1-ECTOPIC and H83M2 transgenes (Figure 2C). The X chromosomes of females that carry only the H83M2 transgene display a more puffed morphology compared to wild-type females (KELLEY et al. 1995 ). This feature is exacerbated in females carrying both the H83M2 and M1-ECTOPIC transgenes, suggesting elevated transcription (Figure 2C; LUCCHESI and SKRIPSKY 1981 ). In addition, several autosomal sites of MSL association were observed in flies carrying both transgenes regardless of their sex.

View larger version (66K): In this window In a new window Download PPT slide   Figure 2. Polytene salivary gland chromosomes from third-instar female larvae. (A and B) A single nucleus from a w; msl1269/+; [w+, M1-ECTOPIC]/TM6 female. (C and D) A single nucleus from a w; msl1269/+; [w+, M1-ECTOPIC]/[w+, H83M2] female. (E and F) A single nucleus from a w; msl1269/+; [w+, H83M2]/TM6 female. The chromosomes were stained with anti-MSL1 antibodies shown in A, C, and E and with bisbenzimide shown in B, D, and F. Arrows in C, areas of diffuse chromatin.

Taken together these data indicate that full msl-mediated dosage compensation is 100% lethal to females. This level of dosage compensation is not achieved when solely MSL2 is expressed in females, suggesting that MSL1 levels are controlled by both MSL2-dependent and MSL2-independent mechanisms.

Modulation of MSL1 levels through polyuridine tracts:
SXL protein directly represses translation of msl2 mRNA by binding to multiple polyuridine tracts in the 5' and 3' UTRs (BASHAW and BAKER 1997 ; KELLEY et al. 1997 ). The finding that two out of three msl1 transcripts contain polyuridine tracts identical to those in msl2 lead to the suggestion that SXL protein might also directly repress translation of some msl1 mRNAs (PALMER et al. 1993 ). Female cells lacking SXL derepress msl1 expression (PALMER et al. 1994 ), but this result is difficult to interpret because msl2 is also derepressed. We tested whether SXL could partially repress msl1 translation in females by assaying a set of transgenes in vivo, which differed with respect to their polyuridine tracts.

The msl1 gene produces three transcripts in both males and females (Figure 3B; PALMER et al. 1993 ). Each transcript contains the same open reading frame, but they differ in the length of their 3' UTRs. The two longest msl1 transcripts possess a cluster of three U₈ sites and one AU₇ site for which SXL has a high affinity in vitro (PALMER et al. 1993 ; SAMUELS et al. 1994 ).

View larger version (12K): In this window In a new window Download PPT slide   Figure 3. A schematic diagram of the msl1 locus (A), transcripts (B), transgenes (C), and sequence of the mutated polyuridine tracts in the M1-MPU transgene (D). Solid black lines, msl1 noncoding sequences. Dashed lines, noncoding sequences that are from loci other than msl1. Solid rectangles, msl1-coding sequence. Vertical lines in the 3' UTR, position of the polyadenylation sites. Circles, placements of the polyuridine tracts. Open circles in the M1-MPU transgene, polyuridine tracts altered by PCR mutagenesis. The sequence of the mutant polyuridine tracts is shown in D, where the boldface letters identify the bases that were changed from uridines to cytosines. The msl1 locus, transcripts, and transgenes are drawn to scale; however, the circles representing the polyuridine tracts are not.

To access the potential involvement of the polyuridine tracts in msl1 regulation, we assayed three P-element transgenes: full-length [M1-FL], truncated 3' UTR [M1-T3UTR], and mutant polyuridine tracts [M1-MPU] (Figure 3C). All of the transgenes contained 2.8 kb of msl1 sequence upstream of the translation start site and the msl1 open reading frame, including the intron. The M1-FL transgene and the M1-MPU transgene contained 2.0 kb of 3' UTR, which included all three polyadenylation sites and the cluster of polyuridine tracts. The polyuridine tracts within the M1-MPU transgene were mutated by site-directed mutagenesis using a PCR method to modify the tracts by changing selected uridines to cytosines (Figure 3D; HIGUCHI et al. 1988 ). Similar disruptions of a stretch of uridine residues with cytosines has been demonstrated to abolish Sxl binding (SAMUELS et al. 1994 ; WANG and BELL 1994 ; KELLEY et al. 1997 ). The M1-T3UTR transgene was different from the other two transgenes in that it was truncated 0.3 kb after the stop codon and included only the first polyadenylation signal, thus eliminating the polyuridine tracts. Each transgene produced the appropriate msl1 transcripts (Figure 4A). The M1-T3UTR transgene produced an increased amount of the shortest transcript compared to the M1-FL and M1-MPU transgenes, suggesting that the total level of transcription from each of the transgenes was equivalent. In addition, each transgene produced full-length MSL1 protein (Figure 4B) in sufficient amounts to rescue msl1 mutant males.

View larger version (62K): In this window In a new window Download PPT slide   Figure 4. (A) Northern analysis of adult female flies carrying the msl1 transgenes, using probes to the msl1 and the Adh genes. All of the transgenic females were w; msl1L60; [w+, msl1 transgene] with the exception of line M1-FL 22B, which was heterozygous for the transgene. The lines M1-MPU 17E and M1-MPU 31B were also analyzed and were determined to make all three transcripts at a level slightly lower than that of M1-MPU 7C (data not shown). The msl1L60 allele contains a 2-kb deletion within the locus and produces a set of three small mRNAs, as seen in all lanes except the wild-type lane. (B) Western analyses of adult flies carrying the msl1 transgenes using antibodies to the MSL1 and MLE proteins. All of the transgenic flies were w; msl1L60; [w+, msl1 transgene]/[w+, H83M2], except for M1-FL-22B, which was w; msl1269/msl1L60; [w+, M1-FL-22B]/[w+, H83M2].

Measuring MSL1 synthesis is impractical in wild-type females, because MSL1 is efficiently degraded in the absence of MSL2. Therefore, our assays were performed in females that make saturating (H83M2) or intermediate (msl2-NOPU) levels of MSL2, so that the MSL1 protein synthesis would be reflected in the accumulation of stable protein. We measured accumulation of MSL1 by Westerns, the formation of MSL complexes on the X chromosomes, and the toxicity to females. Surprisingly, none of the three msl1 constructs tested was as toxic to females as the endogenous msl1⁺ locus. The reason for this lowered expression is not known, but it does not hinder a comparison between the three constructs.

We assayed the female toxicity of the msl1 transgenes by measuring the increased severity of the developmental delay observed in females expressing abundant MSL2 from the H83M2 transgene. Developmental delay was measured as the number of days between the eclosion of the first H83M2 and the first [msl1 transgene]/H83M2 female progeny of a cross. The presence of the M1-FL transgene did not prolong the developmental delay of H83M2 females, but two of the three M1-MPU and both M1-T3UTR transgenic lines showed an increased developmental delay of an additional 1–2 days (Table 2). This suggests that the wild-type 3' UTR lowers msl1 activity in vivo, and this repression requires polyuridine tracts.

The MSL2 protein made by the H83M2 females is not toxic when the msl1⁺ dose is cut by half in msl1/+ heterozygotes (KELLEY et al. 1995 ). Introduction of M1-FL into msl1/+; H83M2 females fails to restore female toxicity, demonstrating again that, for unknown reasons, our wild-type transgene is less active than an endogenous copy of msl1⁺. Nevertheless, when the putative SXL-binding sites are destroyed as in the M1-MPU transgene, female development is again slowed, indicating that inappropriate dosage compensation is restored. This demonstrates that in vivo the polyuridine tracts are able to reduce msl1 activity, but the effect is modest in that female development is only delayed; all females eventually reach adulthood.

In the adult females, we noticed a scalloped wing phenotype. This phenotype resembled that of Beadex, an X-linked locus postulated to be dosage sensitive (GREEN 1953 ; LIFSCHYTZ and GREEN 1979 ). Some females from each of the msl1 transgenic lines exhibited this phenotype (Table 2), and the percentage of affected females correlated with the developmental delay. While this result is intriguing, it is unclear that the wing scalloping is due to dosage compensation.

We next tested if the phenotypic differences observed with the various transgenes reflected in vivo levels of MSL1. In our hands, Western analysis of adult flies did not reflect the subtle differences in MSL1 abundance predicted by our genetic data (Figure 4B). However, when we directly visualized the amount of MSL1 in complexes on larval X chromosomes, we detected differences in MSL1 abundance that correlated with our developmental delay data. Females carrying the msl2-NOPU transgene produce an intermediate level of MSL2 protein (KELLEY et al. 1997 ). The number of MSL bands observed on the X chromosome in msl2-NOPU females is highly dependent on the dose of msl1⁺ (R. KELLEY, unpublished data). The msl1 transgenes were crossed into an msl1^- [msl2-NOPU] background, so that the number of MSL bands present on the females' X chromosomes was dependent on the amount of MSL1 produced solely from the msl1 transgene. Females that contained an M1-FL transgene had 20–50 sites. Females with the M1-T3UTR transgene had 30–70 sites. With the exception of line 31B, females carrying the M1-MPU transgene had 40–80 sites. An example of typical MSL1 X chromosome staining for each msl1 transgene is shown in Figure 5A–C. MSL1 and MSL2 were colocalized on the X chromosomes in all lines examined. This result shows that within the unavoidable variation due to the different insertion sites, slightly less MSL1 protein is on the X chromosomes in females which carry 3' polyuridine tracts compared to flies without them.

View larger version (20K): In this window In a new window Download PPT slide   Figure 5. (A–C) Immunofluorescence of chromosomes from third-instar female larvae using antibody to the MSL1 protein. The chromosomes are from w [w+, msl2-NOPU]/w; msl1269/msl1L60; [w+, msl1 transgene]/+ females. Each micrograph is labeled with the name of the msl1 transgenic line from which the chromosomes were taken. (D) A summary of the correlation between the developmental delay phenotype, the wing phenotype, and the number of MSL bands on the X chromosomes for each of the msl1 transgenes tested.

Previous work indicated that SXL represses msl1 expression post-transcriptionally, but it was not known if this was direct or indirect (PALMER et al. 1994 ). The genetic and immunofluorescence data presented here (summarized in Figure 5D) indicate that synthesis of MSL1 in females is weakly repressed by the four high-affinity Sxl-binding sites that are present in the two longest msl1 transcripts. Similar to the regulation of msl2, the two longest msl1 transcripts may be bound by SXL, so that MSL1 protein is only translated efficiently from the shortest transcript (Figure 6).

View larger version (24K): In this window In a new window Download PPT slide   Figure 6. A model for the sex-specific regulation of msl1. In males (top), all three msl1 transcripts are translated into protein, resulting in an abundance of MSL1. MSL2 interacts with MSL1 and stabilizes it, and this partial complex is able to associate specifically with a small number of sites on the X chromosome. MSL3 and MLE, and possibly MOF and rox RNA, are added to the complex, and full dosage compensation is achieved, resulting in viable males. In females (bottom), only the shortest msl1 transcripts are efficiently translated, which reduces the level of MSL1 protein present in females compared to males. This MSL1 protein does not have MSL2 to stabilize it, so it is degraded. The lack of MSL2 and low levels of unstable MSL1 prevent any MSL complexes from forming, resulting in viable females.

In summary, females prevent MSL1 accumulation by two distinct SXL-dependent mechanisms. The cluster of polyuridine tracts in the 3' UTR is necessary for a small reduction in MSL1 synthesis, presumably by direct SXL binding. However, this is obscured by the more significant contribution of MSL1 degradation when separated from MSL2. This destruction of MSL1 indirectly depends upon SXL through its role in msl2 regulation.

The role of MSL1 in dosage compensation:
It has been difficult to ascertain the functions of individual MSL proteins because they are only detectable as complexes. MSL1 and MSL2 are always observed together even in the absence of MSL3 or MLE (LYMAN et al. 1997 ). While the MSL1 protein has no significant similarity to other known proteins, it contains an acidic amino terminus characteristic of proteins that are involved in transcription and chromatin remodeling (PALMER et al. 1993 ). MSL1 does not normally accumulate in the absence of MSL2; therefore, to assay MSL1 in the absence of MSL2, we transiently expressed MSL1 from the M1-ECTOPIC transgene in wild-type females.

M1-ECTOPIC larvae were heat shocked at 37° for 1 hr, allowed to recover at room temperature for 1 hr, and then the polytene chromosomes from their salivary glands were studied by immunofluorescence. In females, the MSL1 protein was associated with sites on all chromosome arms (Figure 7). The MSL1 protein in males was associated with the X chromosome and also with multiple individual sites on the autosomes (data not shown). The MSL1 association with chromatin was detectable for only 2 hr post-heat shock. The transient expression of MSL1 was not sufficient to recruit MSL2 to the autosomal sites in males or to any of the sites in females, but it indicated that MSL1 has affinity for chromatin, consistent with its postulated role (PALMER et al. 1993 ). In contrast, overexpressed MSL2 caused by heat shock induction of a transgene results in exclusive association with MSL1 to sites only on the X chromosome (R. KELLEY, unpublished data). Based on these results, it is possible that MSL1 provides the MSL complex with its affinity for chromatin, but that MSL2 is required to specifically target the complex to the X chromosome.

View larger version (30K): In this window In a new window Download PPT slide   Figure 7. Polytene salivary gland chromosomes from heat-shocked female third-instar larvae. (A and B) A single nucleus from a yw female, (C and D) a single nucleus from a w; [w+, M1-ECTOPIC]/+ female. Each was isolated 1 hr after a 1 hr 37° heat shock. The chromosomes were stained using antibodies against MSL1 (A and C), or bisbenzimide (B and D) to visualize the DNA.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Previous studies have shown that MSL1 levels depend upon the presence of MSL2 (PALMER et al. 1994 ; KELLEY et al. 1995 ). Here we show that MSL2 regulates MSL1 levels through protein stabilization, not induced protein synthesis. The M1-ECTOPIC transgene lacks all flanking regulatory regions and is driven by a promoter that is equally active in males and females. The MSL1 protein produced by the M1-ECTOPIC transgene still depends upon MSL2 to accumulate, much like protein made by the wild-type msl1 gene.

Our results are consistent with data that demonstrate a physical interaction between MSL2 and MSL1 (KELLEY et al. 1995 ; K. COPPS, personal communication). The inherent instability of MSL1 protein could be due to PEST sequences in MSL1, first postulated by PALMER et al. 1993 and confirmed here by the PESTFIND program (ROGERS et al. 1985 ) (Table 3). While the significance of these sequences in MSL1 has not yet been determined, PEST sequences occur in a family of nuclear proteins, many of which are known to have short half-lives (CHEVAILLIER 1993 ).

It has been difficult to determine the individual functions of MSL1 and MSL2 because they are always associated with each other. However, the observation that MSL1 and MSL2 can associate with the male X chromosome without MSL3 or MLE (PALMER et al. 1994 ; GORMAN et al. 1995 ; LYMAN et al. 1997 ) suggests that either one or both of these proteins is able to bind chromatin. In this study, we have separated MSL1 and MSL2 by transiently overexpressing MSL1 in a wild-type female, in the absence of MSL2 protein, and found that MSL1 has the ability to bind chromatin, but lacks specificity for the X chromosome. We propose that MSL1 binds chromatin but is inherently unstable, while MSL2 stabilizes MSL1 and provides X chromosome specificity to the complex.

Although translational repression of MSL2 should be sufficient to repress dosage compensation in females, MSL1 may be regulated in tandem to prevent nonspecific association with chromatin. One can speculate that if MSL1 were abundant and stable in females, it might bind randomly to chromatin and recruit MOF. In turn, MOF could induce changes in the composition of acetylated histones and in the transcription levels of autosomal genes. Thus, it may be beneficial for females to have two mechanisms to reduce the amount of MSL1 protein.

A second level of regulation was first suggested by inspection of the msl1 3' UTR sequence (PALMER et al. 1993 ). The presence of multiple SXL-binding sites leads to the idea that females might down regulate synthesis of the MSL1 protein despite the fact that they contain abundant msl1 RNA. Moreover, recent work has shown that SXL represses translation of msl2 RNA by binding multiple polyuridine tracts (BASHAW and BAKER 1997 ; KELLEY et al. 1997 ). We found that destruction of the SXL-binding sites in the 3' UTR leads to a small increase in MSL1 activity in vivo. Due to the subtlety of the effect, Western blotting and viability analyses were inconclusive. Instead, systems previously determined to be highly sensitive to MSL complex levels (KELLEY et al. 1995 , KELLEY et al. 1997 ) were used to determine that the lack of polyuridine tracts results in retarded female development and an increase in the amount of MSL complexes on the X chromosomes. Our data suggest that SXL represses the translation of msl1 and msl2 mRNA by the same mechanism. This repression is totally effective for msl2. However, msl1 lacks SXL-binding sites at the 5' UTR, and the shortest msl1 transcript lacks all SXL-binding sites; thus, translational repression of msl1 is only weakly effective.

The absence of a more drastic phenotype in flies carrying H83M2 and a genomic msl1 transgene is likely due to the structure of the msl1 transgenes. The failure of any of the msl1 constructs to exhibit the potency of an endogenous locus suggests that we inadvertently omitted an important element such as an enhancer. This possibility was confirmed by putting a 15-kb genomic msl1 transgene (P[ry⁺]msl1; PALMER et al. 1993 ) into the H83M2 background. Females with these two transgenes and one endogenous copy of msl1 were developmentally delayed (data not shown). However, even the 15-kb msl1 construct did not affect female viability like the endogenous msl1 locus. Although our msl1 transgenes were not as potent as the endogenous msl1 locus, we were able to compare their effects by using sensitized genetic backgrounds. In this way, we were able to identify a role for the polyuridine tracts in msl1 regulation.

In our model (Figure 6), msl1 is regulated at two levels in females. Our data indicate that MSL1 synthesis is negatively regulated in females through a mechanism that requires the polyuridine tracts and SXL. In addition, MSL2 and MSL1 must assemble into complexes or MSL1 monomers are destroyed. Of the two levels of regulation, protein degradation is far more critical than translational control. The fact that substantial levels of MSL1 accumulate in H83M2 females shows that SXL only weakly represses MSL1 translation. Furthermore, when synthesis of MSL1 is artificially equalized in males and females using the M1-ECTOPIC transgene, males still accumulate high levels of MSL1 and females destroy most of the MSL1 protein made, just as in wild-type flies.

Previous data supported the hypothesis that regulation of msl2 alone controlled the sex-specificity of dosage compensation (BASHAW and BAKER 1995 ; KELLEY et al. 1995 ; ZHOU et al. 1995 ; BASHAW and BAKER 1997 ; KELLEY et al. 1997 ). We now show that msl1 is regulated in a sex-specific manner and contributes to the sex-specificity of dosage compensation. MSL1 and MSL2 act together to enable complex formation or the stable association of the MSL complex with the X chromosome. Because regulation is targeted in females to these two critical genes, dosage compensation is prevented.

	ACKNOWLEDGMENTS

We thank E. Skoulakis and R. Davis for kindly providing us with the LEONARDO antibody and R. Kelley for providing msl1^L60, H83M2, and msl2-NOPU flies. We thank V. Meller, R. Kelley, and M. Palmer for critical reading of the manuscript. This work was supported by National Research Service Award HG00177-01 (K.A.C.), National Institutes of Health grant GM45744 (M.I.K.), and by the Howard Hughes Medical Institute (M.I.K. and K.A.C.).

Manuscript received March 23, 1998; Accepted for publication June 19, 1998.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BASHAW, G. J. and B. S. BAKER, 1995  The msl-2 dosage compensation gene of Drosophila encodes a putative DNA binding protein whose expression is sex specifically regulated by Sex-lethal. Dev. 121:3245-3258[Abstract].

BASHAW, G. J. and B. S. BAKER, 1997  The regulation of the Drosophila msl-2 gene reveals a function for Sex-lethal in translational control. Cell 89:789-798[Medline].

BELOTE, J. M. and J. C. LUCCHESI, 1980  Male-specific lethal mutations of Drosophila melanogaster. Genetics 96:165-186[Abstract/Free Full Text].

BONE, J. R., J. LAVENDER, R. RICHMAN, M. J. PALMER, and B. M. TURNER et al., 1994  Acetylated histone H4 on the male X chromosome is associated with dosage compensation in Drosophila. Genes Dev. 8:96-104[Abstract/Free Full Text].

CHEVAILLIER, P., 1993  PEST sequences in nuclear proteins. Int. J. Biochem. 25:479-482[Medline].

FUKUNAGA, A., A. TANAKA, and K. OISHI, 1975  Maleless, a recessive autosomal mutant of Drosophila melanogaster that specifically kills male zygotes. Genetics 81:135-141[Abstract/Free Full Text].

GORMAN, M., M. I. KURODA, and B. S. BAKER, 1993  Regulation of sex-specific binding of the maleless dosage compensation protein to the male X chromosome. Cell 72:39-49[Medline].

GORMAN, M., A. FRANKE, and B. S. BAKER, 1995  Molecular characterization of the male-specific lethal-3 gene and investigations of the regulation of dosage compensation in Drosophila. Dev. 121:463-475[Abstract].

GREEN, M. M., 1953  The Beadex locus in Drosophila melanogaster: genetic analysis of the mutant Bx^r49k. Z. indukt. Abstamm.-u. Vererbungslehre 85:435-449.

HIGUCHI, R., B. KRUMMEL, and R. SAIKI, 1988  A general method of in vitro preparation and specific mutagenesis of DNA fragments: study of protein and DNA interactions. Nucleic Acids Res. 16:7351-7367[Abstract/Free Full Text].

HILFIKER, A., D. HILFIKER-KLEINER, A. PANNUTI, and J. C. LUCCHESI, 1997  mof, a putative acetyl transferase gene related to the Tip60 and MOZ human genes and to the SAS genes of yeast, is required for dosage compensation in Drosophila. EMBO J. 16:2054-2060[Medline].

HORABIN, J. I. and P. SCHEDL, 1993  Sex-lethal autoregulation requires multiple cis-acting elements upstream and downstream of the male exon and appears to depend largely on controlling the use of the male exon 5' splice site. Mol. Cell Biol. 13:7734-7746[Abstract/Free Full Text].

KELLEY, R. L., I. SOLOVYEVA, L. LYMAN, R. RICHMAN, and V. SOLOVYEV et al., 1995  Expression of MSL-2 causes assembly of dosage compensation regulators on the X chromosome and female lethality in Drosophila. Cell 81:867-877[Medline].

KELLEY, R. L., J. WANG, L. BELL, and M. I. KURODA, 1997  Sex lethal controls dosage compensation in Drosophila by a nonsplicing mechanism. Nature 387:195-199[Medline].

KURODA, M. I., M. J. KERMAN, R. KREBER, B. GANETZKY, and B. S. BAKER, 1991  The maleless protein associates with the X chromosome to regulate dosage compensation in Drosophila. Cell 66:935-947[Medline].

LIFSCHYTZ, E. and M. M. GREEN, 1979  Genetic identification of dominant overproducing mutations: the Beadex gene. Mol. Gen. Genet. 171:153-159[Medline].

LINDSLEY, D. L., and G. G. ZIMM, 1992 The Genome of Drosophila melanogaster. Academic Press, Inc., San Diego.

LUCCHESI, J. C. and T. SKRIPSKY, 1981  The link between dosage compensation and sex differentiation in Drosophila melanogaster. Chromosoma 82:217-227[Medline].

LYMAN, L. M., K. COPPS, L. RASTELLI, R. L. KELLEY, and M. I. KURODA, 1997  Drosophila Male-Specific Lethal-2 protein: structure/function analysis and dependence on MSL-1 for chromosome association. Genetics 147:1743-1753[Abstract].

OKUNO, T., T. SATOU, and K. OISHI, 1984  Studies on the sex-specific lethals of Drosophila melanogaster. Jpn. J. Genet. 59:237-247.

PALMER, M. J., V. A. MERGNER, R. RICHMAN, J. E. MANNING, and M. I. KURODA et al., 1993  The male-specific lethal-one (msl-1) gene of Drosophila melanogaster encodes a novel protein that associates with the X chromosome in males. Genetics 134:545-557[Abstract].

PALMER, M. J., R. RICHMAN, L. RICHTER, and M. I. KURODA, 1994  Sex-specific regulation of the male-specific lethal-1 dosage compensation gene in Drosophila. Genes Dev. 8:698-706[Abstract/Free Full Text].

PIRROTTA, V., 1988 Vectors for P-element transformation in Drosophila, pp. 437–456 in Vectors: A Survey of Molecular Cloning Vectors and Their Uses, edited by R. L. RODRUGYEZ and D. T. DENHARDT. Butterworths, Boston.

PIRROTTA, V., S. BICKEL, and C. MARIANI, 1988  Developmental expression of the Drosophila zeste gene and localization of zeste protein on polytene chromosomes. Genes Dev. 2:1839-1850[Abstract/Free Full Text].

RASTELLI, L., R. RICHMAN, and M. I. KURODA, 1995  The dosage compensation regulators MLE, MSL-1 and MSL-2 are interdependent since early embryogenesis in Drosophila. Mech. Dev. 53:223-233[Medline].

RICHTER, L., J. R. BONE, and M. I. KURODA, 1996  RNA-dependent association of the Drosophila maleless protein with the male X chromosome. Genes to Cells 1:325-336[Abstract].

ROBERTSON, A. M., C. R. PRESTON, R. W. PHILLIS, D. JOHNSON-SCHLITZ, and W. K. BENZ et al., 1988  A stable genomic source of P element transposase in Drosophila melanogaster. Genetics 118:461-470[Abstract/Free Full Text].

ROGERS, S., R. WELLS, and M. RECHSTEINER, 1985  Amino acid sequences common to rapidly degraded proteins: the PEST hypothesis. Science 234:364-368.

RUBIN, G. M. and A. C. SPRADLING, 1982  Genetic transformation of Drosophila with transposable element vectors. Science 218:348-353[Abstract/Free Full Text].

SAMUELS, M. E., D. BOPP, R. A. COLVIN, R. F. ROSCIGNO, and M. A. GARCIA-BLANCO et al., 1994  RNA binding by Sxl proteins in vitro and in vivo. Mol. Cell. Biol. 14:4975-4990[Abstract/Free Full Text].

SKOULAKIS, E. M. C. and R. L. DAVIS, 1996  Olfactory learning deficits in mutants for leonardo, a Drosophila gene encoding a 14-3-3 protein. Neuron 17:931-944[Medline].

SPRADLING, A. C. and G. M. RUBIN, 1982  Transposition of cloned P elements into Drosophila germ line chromosomes. Science 218:341-347[Abstract/Free Full Text].

TURNER, B. M., A. J. BIRLEY, and J. LAVENDER, 1992  Histone H4 isoforms acetylated at specific lysine residues define individual chromosomes and chromatin domains in Drosophila polytene nuclei. Cell 69:375-384[Medline].

UCHIDA, S., T. UENOYANA, and K. OISHI, 1981  Studies on the sex-specific lethals of Drosophila melanogaster. Jpn. J. Genet. 56:523-527.

WANG, J. and L. R. BELL, 1994  The Sex-lethal amino terminus mediates cooperative interactions in RNA binding and is essential for splicing regulation. Genes Dev. 8:2072-2085[Abstract/Free Full Text].

ZHOU, S., Y. YANG, M. J. SCOTT, A. PANNUTI, and K. C. FEHR et al., 1995  Male-specific lethal 2, a dosage compensation gene of Drosophila, undergoes sex-specific regulation and encodes a protein with a RING finger and a metallothionein-like cysteine cluster. EMBO J. 14:2884-2895[Medline].

This article has been cited by other articles:

				I. V. Kotlikova, O. V. Demakova, V. F. Semeshin, V. V. Shloma, L. V. Boldyreva, M. I. Kuroda, and I. F. Zhimulev The Drosophila Dosage Compensation Complex Binds to Polytene Chromosomes Independently of Developmental Changes in Transcription Genetics, February 1, 2006; 172(2): 963 - 974. [Abstract] [Full Text] [PDF]

				V. Morales, C. Regnard, A. Izzo, I. Vetter, and P. B. Becker The MRG Domain Mediates the Functional Integration of MSL3 into the Dosage Compensation Complex Mol. Cell. Biol., July 15, 2005; 25(14): 5947 - 5954. [Abstract] [Full Text] [PDF]

				C.-G. Lee, T. W. Reichman, T. Baik, and M. B. Mathews MLE Functions as a Transcriptional Regulator of the roX2 Gene J. Biol. Chem., November 12, 2004; 279(46): 47740 - 47745. [Abstract] [Full Text] [PDF]

				B. P. Rattner and V. H. Meller Drosophila Male-Specific Lethal 2 Protein Controls Sex-Specific Expression of the roX Genes Genetics, April 1, 2004; 166(4): 1825 - 1832. [Abstract] [Full Text] [PDF]

				A. J. Greenberg, J. L. Yanowitz, and P. Schedl The Drosophila GAGA Factor Is Required for Dosage Compensation in Males and for the Formation of the Male-Specific-Lethal Complex Chromatin Entry Site at 12DE Genetics, January 1, 2004; 166(1): 279 - 289. [Abstract] [Full Text] [PDF]

				R. L. Kelley and M. I. Kuroda The Drosophila roX1 RNA Gene Can Overcome Silent Chromatin by Recruiting the Male-Specific Lethal Dosage Compensation Complex Genetics, June 1, 2003; 164(2): 565 - 574. [Abstract] [Full Text] [PDF]

				H. Oh, Y. Park, and M. I. Kuroda Local spreading of MSL complexes from roX genes on the Drosophila X chromosome Genes & Dev., June 1, 2003; 17(11): 1334 - 1339. [Abstract] [Full Text] [PDF]

				Y. Park, R. L. Kelley, H. Oh, M. I. Kuroda, and V. H. Meller Extent of Chromatin Spreading Determined by roX RNA Recruitment of MSL Proteins Science, November 22, 2002; 298(5598): 1620 - 1623. [Abstract] [Full Text] [PDF]