MLE (maleless) is the other MSL complex member possessing enzymatic activity. mle locus encodes a 1,293-aa protein that contains two dsRNA-bind-ing domains and a helicase/NTPase domain of the DExH subfamily type. Furthermore, the C-terminal portion MLE contains nine glycine-rich heptad repeats (Kuroda et al. 1991). Full-length MLE has helicase activity, and it can resolve RNA:RNA, RNA:DNA and DNA:DNA hybrids with 3' ssDNA overhangs in vitro. It can also utilize all common NTPs in helicase assays with similar Km values, illustrating its broad specificity for substrates (Lee et al. 1997).

The function of MLE in dosage compensation is unclear. Transgenic fly lines carrying mutant versions of MLE helicase domain have revealed some requirements for its function. The mutant form of MLE (MLEGET) that abolishes its ATPase activity but retains ssDNA and ssRNA binding ability cannot rescue mle mutant flies. Richter et al. (1996) mutated the same GKT motif, but their construct (MLEGNT) could restore male viability to approximately 50 %. Point mutation in the conserved DExD box of MLE severely reduced the rescue ability of the transgene. MLEDQID mutant retains ATPase binding ability but abrogates ATPase and helicase activity. No effect was seen when the conserved SAT motif was changed to AAA, even though the mutation is postulated to abolish RNA helicase activity.

When MLEGET construct was expressed in mle null mutant background, the MSL complex could nucleate on entry sites, but it was unable to spread to the surrounding chromatin (Lee et al. 1997; Gu et al. 2000). Helicase activity is thus not strictly required for correct localization. In addition, roX RNAs are not stable in the absence of MLE helicase activity, which suggests that MLE is required for proper complex maintenance but not core complex assembly (Gu et al. 2000). Failure to spread could also be indirect, caused by the lack of roX RNAs.

Maternally supplied MLE stabilizes roX1 transcripts in early embryos (Meller 2003), but this is not essential for dosage compensation, as zygotic MLE is sufficient for male viability. roX2 transcripts associate with MLE in vivo (Meller et al. 2000), and similar to MOF and MSL3, MLE association with the X chromosome is sensitive to RNase treatment. Surprisingly, the MLE C-terminal portion (amino acids 941-1,293) containing only glycine-rich repeats can associate with chromatin in an RNase-sensitive manner (Richter et al. 1996).

Taken together, these results illustrate the interdependence between roX1, roX2 and MLE. First, MLE is required for roX1 and roX2 stability (Gu et al. 2000; Kageyama et al. 2001). Second, an RNA component (either roX RNAs or a yet unidentified RNA) maintains MLE association with the X chromosome (Richter et al. 1996).

MLE is the only MSL complex protein with male-specific lethal phenotype that has been shown to have an additional function not related to dosage compensation. nap (no action potential) is an allele of the mle locus that shows no male-specific lethality (Kernan et al. 1991). Instead, flies are paralyzed in non-permissive (high) temperatures (Wu et al. 1978). Since this phenotype is very similar to Drosophila para mutants, Reenan et al. (2000) reasoned that MLE could be involved in processing of para transcript. Indeed, para mRNA is aberrantly spliced in mlenap mutant flies, such that only 17 % of transcripts are correctly spliced. para mRNA is predicted to form extensive secondary structures, which suggests that MLE is involved in resolving these structures, allowing correct splicing. Strikingly, mle alleles with male-specific lethal phenotype have no effect on para mRNA splicing (Reenan et al. 2000). These results could have several different explanations. First, mlenap is a gain-of-function allele and wild-type MLE has no function in para splicing. Second, it could be a hypomorphic mle allele epistatic to other RNA helicases. In complete absence of mle, other helicases could resolve para secondary structures, but mlenap still retains some function and prevents the access of other helicases to para mRNA. The third explanation is that nap and male-specific lethal phenotypes map to different domains of MLE protein. However, this is unlikely, as mlenap mutation maps in the next amino acid after the GKT motif that was shown to be important for MSL complex spreading (Kernan et al. 1991; Lee et al. 1997).

Biochemical data also support the idea that MLE may have other functions in addition to dosage compensation. First, co-immunoprecipitation experiments have revealed that MLE is only loosely associated with the MSL complex (Copps et al. 1998; Buscaino et al. 2003). Second, the bulk of MLE exists as a monomer in SL2 nuclear extracts, while only a small proportion of it co-fractionates with the 2-MDa MSL complex (Copps et al. 1998).

0 0

Post a comment