The carboxy-terminal amino acid sequence of dMBD2/3 resembles that of mammalian MBD2 and MBD3, both of which belong to histone-deacetylase-containing corepressor complexes. MBD3 is within the Mi-2/NuRD complex, whereas MBD2 is within the MeCP1 complex. Does dMBD2/3 also associate with Drosophila histone deacetylases (dHDACs)? It was found that in-vitro-translated dHDAC1 and endogenous dHDAC1 from a nuclear extract of Drosophila cells were bound by immobilized dMBD2/3–GST fusion protein. An in vivo interaction was demonstrated by transiently co-expressing haemaglutinin (HA) epitope-tagged dMBD2/3 and dHDAC1 in mammalian cells; immunoprecipitates of HA-dMBD2/3 also contained dHDAC1. In addition, dMBD2/3 associated with the corepressor protein dMi-2 dMBD2/3 in vitro and in vivo. Therefore, like mammalian MBD3, dMBD2/3 can interact with components of the Mi-2/NuRD histone deacetylase complex (Tweedie, 1999).
The methyl-CpG-binding domain of dMBD2/3 differs at sites that are highly conserved among vertebrate family members, including a nine amino acid deletion and an opa-likerepeat insertion. Is this because the methyl-CpG-binding domain is superfluous in an organism without CpG methylation? Or does it reflect structural or functional divergence in the evolutionary interval that separates insects and mammals? To distinguish between these alternatives, an insect was sought that has methylated DNA to ask whether its MBD2/3 orthologue resembles mammalian or Drosophila MBD proteins. The presence of 5-methylcytosine in the genome of the cricket Acheta domesticus was established by probing blots of HpaII-digested single-stranded DNA with an antibody against the modified base. An A. domesticus cDNA encoding an MBD2/3-like protein (aMBD2/3) was isolated by degenerate RT-PCR. The methyl-CpG-binding domain of aMBD2/3 closely resembled the mammalian MBD2 and MBD3 domains, but differed from the dMBD2/3 domain, suggesting that the common insect ancestor of Drosophila and A. domesticus encoded an aMBD2/3-like protein. RT-PCR from A. domesticus RNA also identified a second transcript (aMBD2/3Delta) that is alternately spliced at the same positions as dMBD2/3. Both insect MBD proteins, together with their major splice-variant forms, were tested for DNA binding. Only aMBD2/3 formed a strong complex with the methylated probe, and its binding was specifically competed by unlabelled methylated DNA. None of the other proteins, including dMBD2/3, bound to methylated DNA. Therefore, methyl-CpG binding by insect MBD2/3 proteins correlates with genomic DNA methylation (Tweedie, 1999).
Vertebrates produce proteins of the MBD2/3 family that either do or do not bind methylated DNA. Mouse Mbd2 and Xenopus laevis MBD3 bind efficiently to methylated DNA, and mammalian MBD3 shows weak specific binding in vitro. Both mammals and X. laevis produce an abundant splice variant of MBD3 that can no longer bind DNA. It is proposed that the gene encoding aMBD2/3 resembles an ancestral member of the MBD2/3 family that encodes two corepressor functions. This is achieved by alternative splicing of a single transcript to yield DNA-binding and non-DNA-binding forms of the protein. Since Drosophila appears to have lost CpG methylation, mutations that inactivate the methyl-CpG-binding domain of dMBD2/3 may have been allowed to accumulate. Whether divergence of the Drosophila domain has also given rise to novel functions is not known. The C-terminal domain of dMBD2/3, however, still associates with HDACs, presumably in the Drosophila Mi-2/NuRD complex. Persistence of the gene encoding dMBD2/3 without a functional methyl-CpG-binding domain is likely to reflect a requirement for its conserved C-terminal domain in corepressors that do not directly bind DNA (Tweedie, 1999).
Methyl-DNA binding proteins help to translate epigenetic information encoded by DNA methylation into covalent histone modifications. MBD2/3 is the only candidate gene in the Drosophila genome with extended homologies to mammalian MBD2 and MBD3 proteins, which represent a co-repressor and an integral component of the Nucleosome Remodelling and Deacetylase (NuRD) complex, respectively. An association of Drosophila MBD2/3 with the Drosophila NuRD complex has been suggested previously. The molecular interactions between MBD2/3 and the NuRD complex show the two MBD2/3 isoforms precisely cofractionated with NuRD proteins during gel filtration of extracts derived from early and late embryos. In addition, MBD2/3 forms multimers, and engages in specific interactions with the p55 and MI-2 subunits of the Drosophila NuRD complex. These data provide novel insights into the association between Drosophila MBD2/3 and NuRD proteins. Additionally, this work provides a first analysis of the architecture of the Drosophila NuRD complex (Marhold, 2004a).
Attempts were made to confirm the association between MBD2/3 and the MI-2 complex at the functional level. It has been shown previously that MBD2/3 and MI-2 interact in vitro (Tweedie, 1999). Similarly, both proteins have been co-fractionated in protein extracts from Drosophila SL-2 cells (Ballestar, 2001). In order to look for a genetic interaction between MBD2/3 and Mi-2, homozygous MBD1 flies were crossed with flies carrying a heterozygous mutant allele for Mi-2 (Mi-24). Compound heterozygotes for both mutations had significantly rougher and smaller eyes in about 25% of the progeny, while both homozygous MBD1 flies and heterozygous Mi-24 flies had completely normal eyes. This result strongly suggests a functional interaction between MBD2/3 and MI-2. The interaction between MBD2/3 and MI-2 was examined by determining the subcellular distribution of MI-2 protein in MBD1 mutants. Wild-type and mutant embryos were immunostained with a specific antiserum against MI-2 and the subnuclear distribution of the protein was examined by confocal microscopy. This revealed a homogeneous distribution of MI-2 in wild type embryos. However, the protein appeared to be absent from about 10-15 nuclear foci in the mutant. These results are consistent with a functional interaction between MBD2/3 and MI-2, and suggest that the MI-2 complex might be absent from a subset of target loci in MBD2/3 mutants (Marhold, 2004b).
In order to analyze the relationship between MBD2/3 and MI-2 in greater detail, double immunostaining was performed. MBD2/3 has been shown to form nuclear foci at the cellular blastoderm stage that remain detectable until after gastrulation (Marhold, 2002). However, the precise nature of these foci could not be determined further because of the lack of suitable antibodies. As a prerequisite to double immunostaining experiments a monoclonal MBD2/3-specific antibody (MBD 8E7) was raised by immunizing rats with an MBD2/3 peptide. The peptide was selected from the exon 2 region of MBD2/3 that is not present in the short MBD2/3 isoform, and the antibody recognized a single band in Western blots from embryonic nuclear extracts that corresponds to the long isoform of MBD2/3. The specificity of the antibody was confirmed by the absence of detectable signals in Western blots of protein extracts from homozygous MBD1 embryos. Similarly, immunostaining of homozygous MBD1 embryos with 8E7 antibody failed to detect any signals above the background level. Double immunostaining of wild-type embryos with the 8E7 antibody and an MI-2-specific antiserum revealed a speckled nuclear pattern for MBD2/3, which is in agreement with previous observations (Marhold, 2002). By contrast, MI-2 is found in a rather ubiquitous distribution in embryonic nuclei. This result argues against MBD2/3 being an integral component of all MI-2 complexes and suggested a more peripheral association with only a subset of MI-2 complexes (Marhold, 2004b).
To look for a potential interaction between MBD2/3 and methylated DNA attempts were made to double immunostain embryos with antibodies against MBD2/3 and 5-methylcytosine. However, 5-methylcytosine staining requires an extensive sample denaturation with 2 M hydrochloric acid, which affects the distribution of MBD2/3 epitopes. Therefore band shift assays were used to determine the affinity of MBD2/3 for methylated oligonucleotides. Two previous studies have failed to detect an interaction between MBD2/3 and CpG-methylated probes (Ballestar, 2001; Tweedie, 1999). A third study described a preferential binding of the short MBD2/3 isoform to a human DNA probe containing a single methylated CpG (Roder, 2000). However, the short isoform lacks parts of the methyl-DNA binding domain, which rendered the significance of this finding unclear. In addition, CpG methylation is virtually absent from Drosophila genomic DNA and most of the 5-methylcytosine is present in the context of CpT and CpA dinucleotides (Lyko, 2000). This prompted a test of a CpT- and CpA-methylated oligonucleotide probe in band shift assays. In a first set of experiments the affinity was tested of both MBD2/3 isoforms to an oligonucleotide probe that was densely methylated at CpG dinucleotides on both strands. Consistent with the results obtained by others (Ballestar, 2001; Tweedie, 1999), no interactions between MBD2/3 or MBD2/3Delta and CpG-methylated DNA were detected. Under the same conditions, human MBD2 revealed a readily detectable interaction with the probe. To analyze the interaction between MBD2/3 and non-CpG methylated DNA, a double-stranded oligonucleotide was used that contained eight 5-methylcytosine residues in an asymmetrical CpT and CpA context. This revealed a protein-DNA complex for the long isoform of Drosophila MBD2/3, but not for the short isoform, that lacks part of the methyl-DNA binding domain. Most of the shifted signal was detected in the high-molecular weight range, which may indicate that MBD2/3 had formed dimers or oligomers during the binding reaction. The specificity of the interaction between MBD2/3 and the CpT/A-methylated oligonucleotide was confirmed by the addition of unlabelled competitor probe, which strongly reduced the band shift signal. Furthermore, human MBD2 showed no detectable affinity for the CpT/A-methylated probe. From these results, it is concluded that the long isoform of MBD2/3 interacts directly and specifically with CpT/A-methylated DNA (Marhold, 2004b).
The mouse MBD2 protein loses its defined localization pattern in DNA methyltransferase-mutant cell lines (Hendrich, 1998). In order to confirm the methyl-DNA binding activity of MBD2/3 in vivo, the localization of MBD2/3 was examined in embryos with decreased levels of DNA methylation. Dechorionated pre-blastoderm embryos were incubated with the DNA methyltransferase inhibitor 5-azacytidine under conditions that cause efficient and specific demethylation of genomic DNA. Drug- and mock-treated embryos were then double-immunostained with MBD2/3- and DNA-specific antibodies. This revealed a clear mislocalization of MBD2/3 in demethylated embryos. The protein lost its defined focal pattern and showed a homogeneous cellular distribution. Similar results were also obtained after an RNAi-mediated knockdown of the Dnmt2 DNA methyltransferase protein. These data suggest that the localization of MBD2/3 requires wild-type levels of DNA methylation and thus confirms the interaction between MBD2/3 and methylated DNA in vivo (Marhold, 2004b).
DNA methylation in Drosophila is restricted temporally during development and occurs at a significantly lower frequency than in mammals. Thus, the regulatory functions, if any, of this form of DNA modification in Drosophila are unclear. However, the presence of homologs of vertebrate methyl-CpG-binding proteins implies functional consequences for DNA methylation in flies. This work describes the properties of MBD-like, a Drosophila homolog of vertebrate MBD2 and MBD3. MBD-like and MBD-likedelta (a splice variant) failed to bind model methylated DNA probes, inconsistent with their function as mediators of methyl CpG-directed transcriptional repression. However, the MBD-like proteins exhibit transcriptional and biochemical properties consistent with roles as components of a histone deacetylase-dependent corepressor complex similar to the vertebrate Mi-2 complex. The two proteins are differentially expressed during development, suggesting functional specialization. MBD-like and/or MBD-likedelta is present at the chromocenter on larval polytene chromosomes as well as at discrete bands interspersed along the euchromatic chromosome arms, many of which are coincident with known ecdysone-induced loci. This banding pattern suggests gene-specific regulatory functions for MBD-like and the Drosophila Mi-2 complex (Ballestar, 2001).
The two MBD-like proteins were expressed in bacteria, purified, and their binding properties were compared to Xenopus MBD3, a protein previously demonstrated to bind selectively to methylated DNA. A repetitive sequence probe consisting of 12 repeats of the trinucleotide GAC synthesized in either the unmodified or in the fully methylated form was used. Thus, this probe examines only the capacity of proteins to bind methyl CpG in a single sequence context. In South-Western assays using immobilized protein, neither Drosophila MBD-like isoform interacted with the probes, regardless of methylation status, where Xenopus MBD3 selectively bound the methylated probe. This implies that either the proteins fail to bind or that they are unable to refold on the membrane surface. The next step was to examine solution interactions with DNA, using an electrophoretic mobility shift assay. Neither Drosophila protein bound DNA under the same conditions where Xenopus MBD3 bound methylated DNA selectively. It is concluded that, in keeping with the predictions based on amino-acid sequence, neither MBD-like nor MBD-likedelta bound methyl-CpG-containing DNA in the context of the assays used (Ballestar, 2001).
The sequence similarity between the region encoded by exon 3 of MBD-like and vertebrate MBD2 and MBD3 implies conservation of function. One potential role for this region is the interaction with other proteins. Since Drosophila contains homologs of many proteins known to be components of HDAC-containing corepressor complexes in other systems, it was asked whether MBD-like might be a component of such a complex (or complexes) in Drosophila (Ballestar, 2001).
Anti-MBD-like polyclonal antiserum was used to investigate potential interactions between MBD-like and known components of corepressor complexes. Immunoblot analysis confirmed that the antiserum recognized both isoforms of MBD-like. Interestingly, only the shorter isoform, MBD-likedelta, was detected in nuclear and whole extracts from S2 cells. The presence of a minor high molecular mass band was also observed. This weak band may be due to immuno-cross-reactivity to any of the other MBD family members in Drosophila. Immunoprecipitations were performed from S2 nuclear extracts and the precipitated proteins were assayed for enzymatic activities associated with corepressor complexes. Immune serum, but not preimmune serum, efficiently precipitated histone deacetylase activity and ATPase activity. Like vertebrate and Drosophila Mi-2, the precipitated ATPase activity was stimulated by nucleosomes. These results indicate that MBD-likedelta is associated with an undefined histone deacetylase and a nucleosome-stimulated ATPase in S2 nuclei, suggesting inclusion of MBD-likedelta in a Drosophila Mi-2 complex. Also, MBD-likedelta copurifies with dMi-2 and MTA1-like, the Drosophila homolog of the 80-kDa subunit of vertebrate Mi-2/NuRD complexes, consistent with its inclusion in a Drosophila Mi-2 complex similar to that observed in vertebrates (Ballestar, 2001).
A well-characterized transcription assay was used to assess the consequences of recruiting the MBD-like isoforms to a promoter. Both MBD-like and MBD-likedelta were fused to the Gal4 DNA-binding domain (DBD). A Gal4-Groucho fusion and the Gal4 DBD were used as positive and negative controls for transcriptional repression, respectively. Gal4 fusions of MBD-like, MBD-likedelta, and Groucho mediated dose-dependent transcriptional repression following transfection into S2 cells. Repression required a Gal4 site in the reporter plasmid; transfection of MBD-like or MBD-likedelta lacking the Gal4 DBD failed to repress transcription. The expression levels of the transfected Gal4 fusion proteins were equivalent as determined by immunoblot. Furthermore, repression by the two bound MBD-like isoforms was similar to that observed with the well-characterized repressor Groucho. If MBD-like represses transcription through recruitment of a Drosophila Mi-2 complex, transcriptional repression should be sensitive to inhibitors of histone deacetylases, such as TSA. Repression mediated by binding MBD-like or MBD-likedelta to a promoter was largely relieved by TSA; this effect was qualitatively very similar to that of TSA on Groucho-mediated repression. It is concluded that both isoforms of MBD-like function as transcriptional corepressors through recruitment of histone deacetylase activity, consistent with the proposed function of the Drosophila Mi-2 complex (Ballestar, 2001).
To investigate target gene specificity of MBD-likedelta, its distribution on Drosophila salivary gland polytene chromosomes from third instar larvae was examined. In this stage, exclusively the shorter mRNA, corresponding to MBD-likedelta, was detected. Polytene chromosomes are ideal for this analysis since they are thought to reflect the biochemical and structural properties of chromatin of diploid interphase cells, and the pairing of the large number of DNA strands allows for the identification of individual chromosomal sites by light microscopy. To visualize the banding pattern of the polytene chromosomes, the chromosomes were counterstained with DAPI, which stains brightest in the condensed, banded regions of euchromatin and the constitutively condensed heterochromatin at the chromocenter. Immunofluorescence staining with the antibody to MBD-likedelta revealed preferential association with 29 euchromatic sites as well as weaker association with ~100 euchromatic sites and centric heterochromatin. Interestingly, 69% of the predominant sites correspond to developmentally regulated loci that are transcriptionally induced by pulses of the steroid hormone 20-hydroxyecdysone (ecdysone) during the late larval and prepupal periods. Developmental puffing patterns correspond to transcriptional activity. Prior to the third instar larval ecdysone peak, 74EF and 75B are unpuffed and transcriptionally inactive. Following the ecdysone pulse, these loci decondense and become transcriptionally active. Binding of MBD-likedelta to the ecdysone-responsive early puff loci 74EF and 75B was observed at all developmental puff stages examined including prior to ecdysone exposure, shortly after ecdysone exposure, when the genes are strongly activated, and at later stages during puff regression (Ballestar, 2001).
The polytene chromosome pattern of MBD-likedelta does not significantly overlap the sites recently defined for Drosophila SIN3, supporting the idea that MBD-likedelta is not a component of a SIN3 containing complex. This analysis suggests that MBD-likedelta functions in a gene-specific manner and provides a foundation for the identification of potential MBD-likedelta target genes. Furthermore, it provides a basis for understanding the genomic distribution of the Mi-2 complex in Drosophila (Ballestar, 2001).
Home page: The Interactive Fly © 1997 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.