Transcriptional Repression and Cofactor Recruitment

     
      The early development of the Drosophila embryo is marked by its progressive subdivision into increasingly precise spatial domains.  This subdivision is achieved through the actions of a hierarchy of maternal and zygotic segmentation genes, many of which encode transcription factors that both positively and negatively regulate the expression of other transcription factors.

      We are particularly interested in the regulation and function of the pair-rule genes, whose correct expression underlies the establishment of metameric pattern in Drosophila.  Our work is focused on the pair-rule segmentation gene, hairy (h), that is needed for proper embryonic segmentation.  Hairy expression in stripes during blastoderm cellularization serves to establish the reiterated pattern of parasegmental units that represent the basic embryonic body plan.  hairy behaves genetically as a negative regulator of a downstream pair-rule gene, fushi tarazu (ftz), during embryonic segmentation.  Consistent with Hairy's role as a primary repressor of ftz expression, ftz stripes are expanded in hairy mutant embryos.

hairy mutant phenotypes.  Hairy is expressed in seven transverse segment-wide stripes in the cellular blastoderm, defining a double segment periodicity.  Hairy is also expressed in an antero-dorsal head patch (right).  This expression is required to repress Ftz expression, leading to seven transverse Ftz stripes in the opposite parasegments.  In hairy mutant embryos, Ftz expression is broadened leading to the loss of alternate segments.  The resulting larval cuticle shows a classical pair-rule phenotype with only half the number of segments remaining (left).

      hairy encodes a nuclear protein with basic and helix-loop-helix (bHLH) domain.  hairy belongs to the Hairy/Enhancer of split (HES) subclass of repressor bHLH proteins including the structurally related Drosophila proteins encoded by deadpan (dpn), and seven members of the Enhancer of split complex [E(spl)-C; E(spl)m3, -m5, -m7, -m8, -mb, -mg, -md], as well as several vertebrate homologs.  These proteins are genetically required throughout development as transcriptional repressors of genes necessary for processes such as sex-determination, segmentation, and neurogenesis.  Members of the HES class share several regions of homology. 

Hairy protein domains and functions.

      HES proteins have a conserved HLH domain, required for protein dimerization, which is preceded by a basic region featuring a signature proline residue, required for DNA binding with specificity for C-box sequences (CACNAG).  HES proteins are also characterized by two other conserved domains: the Orange domain that mediates functional specificity among HES family members, and the C-terminal conserved WRPW tetrapeptide that is necessary and sufficient for the recruitment of Groucho, a WD-repeat containing protein that is not able to bind DNA on its own, but when brought to an endogenous or heterologous promoter serves as a strong repressor of transcription.  

      The HES family of bHLH proteins appears to function as dedicated repressors.  Four major classes of models for transcriptional repression mechanisms have been proposed: 1) repressors prevent activators from binding DNA ("competition"); 2) repressors and activators bind to DNA at independent sites, but the repressors interfere with interaction between the activators and the general transcriptional machinery ("quenching"); 3) repressors and activators bind to DNA at independent sites, with the repressors interacting (directly) with the general transcriptional machinery ("direct repression"); and 4) repressors and activators bind to DNA at independent sites, with repressors recruiting chromatin-modifying enzymes ("chromatin remodeling"). 

      Existing evidence for Hairy action makes the first class of models ("competition") unlikely:  In particular, Hairy-binding C boxes are physically separate from the activator-binding E boxes at target promoters.  Identification of the Groucho co-repressor led to the view that Hairy functions as a promoter-bound repressor: an intact bHLH region is required for Hairy to bind to specific DNA sites where it then recruits the Groucho co-repressor protein.  For the Groucho-like yeast co-repressor protein Tup1, each of three remaining models has been implicated: the Tup1/Ssn6-complex blocks the activation domain of specific DNA binding proteins, interacts with the general transcriptional machinery, and organizes a repressive chromatin structure through direct interaction with the N-terminal regions of histones H3 and H4.  Groucho has recently been shown to interact specifically with the N-terminal region of histone H3 and has been shown genetically to be essential for H3 transcriptional silencing in yeast, suggesting a repression mechanism involving chromatin remodeling.  In flies, Groucho has been reported to recruit Rpd3, a class I histone deacetylase (HDAC), suggesting a mechanism involving chromatin remodeling 

      Recruitment of Groucho, however, does not account for all of Hairy's repressor properties.  We find that Hairy can function genetically as a repressor in the absence of the WRPW motif, and presumably the Groucho co-repressor.

Hairy has been proposed to function as a sequence-specific DNA binding protein that recruits the Groucho co-repressor protein.  Groucho, in turn, has been proposed to recruit Rpd3, a class I histone deacetylase, suggesting a chromatin mechanism (left).  Groucho is required for Hairy-mediated repression: reducing Groucho activity results in de-repression (broadening) of Ftz stripes (left).  Hairy can also repress transcription in the absence of Groucho, presumably through a number of chromatin independent mechanisms (right).

      Our earlier work has highlighted the requirement for multiple Hairy domains for its proper function, suggesting that Hairy is likely to participate in multiple protein-protein interactions/mechanisms. Our focus over the past few years has been on characterizing cofactors required for Hairy-mediated repression, and more recently on the identification of its direct downstream targets.

HAIRY TARGETS.

      We have known for some time that the segmentation gene, ftz, is Hairy's genetic target, but the mechanism by which Hairy represses ftz remains elusive.  While it has long been assumed that Hairy is a DNA binding protein that binds directly to the ftz promoter to regulate Ftz expression, such binding has not been demonstrated. 

We have used DamID, a chromatin-profiling method developed here in Steve Henikoff’s lab, to perform a global and systematic search for direct transcriptional targets of Drosophila Hairy.  Hairy was tethered to E. coli DNA adenine-methyltransferase (Dam) permitting methylation proximal to in vivo binding sites in both Drosophila Kc cells and early embryos.  This approach identified 40 novel genomic targets for Hairy in Kc cells.  We also adapted DamID profiling such that we could use tightly staged collections of embryos (2-6 hours) and found 20 Hairy targets related to early embryogenesis. As expected of direct targets, all of the putative Hairy target genes tested show Hairy- dependent expression and have conserved consensus C-box (Hairy binding site) containing sequences that are directly bound by Hairy in vitro. The distribution of Hairy targets from both the Kc cell and embryo DamID experiments correspond to Hairy binding sites in vivo on polytene chromosomes. In addition to finding putative targets for Hairy in segmentation, we were excited to find groups of targets suggesting roles for Hairy in cell cycle/growth and morphogenesis, processes that must be coordinately regulated with pattern formation. We are currently using this information to develop different classes of targets to use in biochemical assays, as well as to compare to other developmental transcription networks such as the Drosophila Myc/Max/Mad family of independent, yet mechanistically similar, proteins.

HAIRY COFACTORS.

Identification of the Groucho co-repressor solidified the view that Hairy functions as a promoter-bound repressor: an intact bHLH region is required for Hairy to bind to specific DNA sites where it then recruits the Groucho co-repressor protein. Groucho has been proposed to utilize a chromatin remodeling mechanism through its recruitment of histone deacetylase. Recruitment of Groucho, however, does not account for all of Hairy's repressor properties.  Over the past few years, we have identified and characterized a number of additional Hairy-interacting proteins/cofactors using both genetic and protein interaction screens. These factors act in a context-dependent manner and are likely to utilize different mechanisms of repression.

Summary of Hairy-interacting proteins and the region of Hairy required for their interaction.

dCtBP.  dCtBP is required as a cofactor for a number of early developmental repression systems where it functions in a context dependent manner: dCtBP can function as either a co-activator or co-repressor of transcription, with distinct regions of dCtBP being required for activation or repression.  While dCtBP is required for Hairy-mediated repression, reduction of maternal dCtBP activity suppresses the hairy phenotype. dCtBP does not appear to recruit HDACs.  dCtBP interferes with Groucho-mediated repression, and this modulation of Groucho’s repression activity may be a crucial form of regulation in the early embryo, a closed system dependent on maternally provided and ubiquitously distributed proteins.  dCtBP is encoded by a complex locus encompassing at least three distinct genetic complementation groups: mesA, mesB, and 87De.  Alleles from these different complementation groups exhibit distinct as well as overlapping phenotypes.  There are 4 major RNA isoforms of dCtBP that differ in their 3' ends - resulting in proteins that differ by 10 to 110 amino acids.  The different dCtBP complementation groups remove distinct subsets of dCtBP transcripts.  We are currently identifying the molecular lesion(s) associated with each allele in order to correlate these with the phenotypes observed.

dDrap (NC2a).  dDrap is homologous to Dr1-Associated Protein (Drap) or NC2alpha, a factor implicated as a negative regulator of the basal transcription machinery.  In yeast and humans, DRAP (NC2α) has been shown to act as a heterodimer with Dr1 (NC2b) to globally repress transcription.  The transcription of nearly all protein-encoding genes requires the assembly of the general transcription factors (GTFs) and RNA polymerase II on class II promoters.  The Dr1-DRAP heterodimer (NC2) directly inhibits the activity of the GTFs in a chromatin-independent manner by interfering with the interactions between TATA-binding protein (TBP) and TFIIA or TFIIB.  In GST pull down assays, dDRAP and dDr1 interact with one another, dDr1 binds Drosophila TBP, and dDRAP interacts with Hairy.  Additionally, we find that reducing the dose of either dTBP or dTFIIA maternally leads to enhancement of the hairy mutant phenotype.  Our work suggests that one mechanism of Hairy repression likely involves the recruitment of dDRAP, which then acts as a co-repressor with dDr1 to negatively regulate the basal transcription machinery. 

COFACTOR RECRUITMENT

One of the major questions in the field concerns how and when particular cofactors are recruited.  It has been technically challenging to address this question with current methods such as ChIP assays, since cofactor associations may be transient, unstable, or far removed from the DNA binding protein.  Similarly, utilizing expression-based microarray analysis is also not easy, due to the difficulty in sorting direct from indirect interactions with such widely recruited cofactors.  To circumvent these technical issues and as a first step towards understanding the rules governing Hairy cofactor recruitment, we used the DamID approach to determine if the three best characterized Hairy cofactors, Groucho, dCtBP, and dSir2, are recruited to all or a subset of Hairy targets.

Interestingly, we find that Hairy cofactor recruitment is context-dependent.  While Groucho is frequently considered to be the primary Hairy cofactor, we find that it is associated with only a minority of Hairy targets.  The majority of Hairy targets are associated with the presence of a combination of dCtBP and dSir2. Thus, the DamID chromatin profiling technique is providing a systematic means of obtaining a global view of cofactor recruitment requirements during development.

Our present goal is to use developmental, genetic and molecular approaches to examine the properties of Hairy and several of its interacting proteins (dgrn, dCtBP, dSir2, dNC2, dTopors, and Groucho) to help distinguish amongst the possible regulatory mechanisms used by Hairy to mediate transcriptional repression.