InteractiveFly: GeneBrief

female sterile (1) homeotic: Biological Overview | References

BIOLOGICAL OVERVIEW

More than a dozen trithorax group (trxG) proteins are involved in activation of Drosophila HOX genes. How they act coordinately to integrate signals from distantly located enhancers is not fully understood. The female sterile (1) homeotic [fs(1)h] gene is one of the trxG genes that is most critical for Ultrabithorax (Ubx) activation. One of the two double-bromodomain proteins encoded by fs(1)h acts as an essential factor in the Ubx proximal promoter. Three aspects are noted: (1) overexpression of the small isoform FSH-S, but not the larger one, can induce ectopic expression of HOX genes and cause body malformation; (3) FSH-S can stimulate Ubx promoter in cultured cells through a critical proximal region in a bromodomain-dependent manner; (3) purified FSH-S can bind specifically to a motif within this region that was previously known as the ZESTE site. The physiological relevance of FSH-S is ascertained using transgenic embryos containing a modified Ubx proximal promoter and chromatin immunoprecipitation. In addition, FSH-S is involved in phosphorylation of itself and other regulatory factors. It is suggested that FSH-S acts as a critical component of a regulatory circuitry mediating long-range effects of distant enhancers (Chang, 2007).

Drosophila HOX genes control development of body segments via highly restricted expression domains. These domains are first established by transiently expressed segmentation genes in early embryos and then maintained in an epigenetically heritable manner by the Polycomb group (PcG) of repressors, and the trithorax group (trxG) of activators. Like mammalian promoters that are regulated by distant elements, transcriptional regulation of HOX genes also requires coordinated long-range interactions between the basal transcription machinery assembled around the initiation sites and factors recruited at distant regulatory elements. How the epigenetic inheritance imposed by PcG and trxG is integrated into the general framework of such long-range interactions remains unclear. Its elucidation should provide an important model for understanding the regulatory mechanisms of genes under strict developmental control (Chang, 2007).

PcG repressors form at least two types of multimeric complexes that are targeted by sequence-specific binding proteins to a core PcG response element located ~25 kb upstream of the homeotic gene Ultrabithorax (Ubx). These complexes may block the access of the regulatory elements or modify chromatin by associated histone deacetylase and histone methyltransferase activities. In contrast to the highly targeted activities of PcG repressors, trxG activators appear to employ diverse mechanisms for chromatin remodeling and long-range interactions. For example, trithorax (trx) and absent, small or homeotic discs 1 (ash1) encode histone methyltransferases that are targeted to PcG response elements, promoters, and transcribed regions. In addition to these targeted activities, brahma, moira, and osa encode subunits of an ATP-dependent chromatin remodeling complex that can modulate the nucleosome fluidity to provide an open access of regulatory sequences. Moreover, kohtalo and skuld encode subunits of the Mediator coactivator complex that can facilitate interactions between distal factors and basal transcription machinery (Chang, 2007).

How signals provided by distal elements are integrated at the Ubx basal promoter remains unclear. The Ubx proximal region has several unique features. Instead of the consensus TATA box in the −30 region, Ubx contains the initiator around +1 and the downstream promoter element around +30, which are frequently found in genes lacking the TATA box in Drosophila and mammals. The ability of these elements to support Ubx transcription in vitro and in vivo indicates that they represent an authentic basal promoter. However, this basal promoter fails to integrate regulatory signals from distant elements without a proximal region from −200 to −32, revealing a critical requirement for this region in mediating long-range interactions (Chang, 2007).

Interestingly, this critical proximal region (CPR) contains multiple binding sites for Zeste and Trithorax-like (Trl) proteins. The Zeste sites appear to be particularly important, since CPR activity can be substantially replaced by tandem Zeste sites. Consistent with the transactivating role, zeste was initially identified as required for Ubx expression through transvection, a pairing-dependent effect believed to facilitate the transutilization of the regulatory elements on one chromosome by the promoter on homologous chromosome. Zeste protein can stimulate Ubx transcription in vitro and is necessary for the expression of Ubx transgenes containing subsets of regulatory sequences. Paradoxically, zeste is not essential for normal development or for expression of the endogenous Ubx promoter or a Ubx transgene with more complete regulatory sequences. The role of zeste is further complicated by the finding that zeste may be involved in Ubx repression. Clearly, other factors must be required for the activating effect of the Zeste sites in the CPR (Chang, 2007).

The maternal-effect gene female sterile (1) homeotic [fs(1)h] was identified as a transactivator of Ubx by its strong genetic interactions with Ubx, trx, and ash1 mutations (Digan, 1986; Gans, 1980; Shearn, 1989). However, its direct role in homeotic gene activation has been obscured by complex phenotypes in mutant embryos (Huang, 1990). Sequence analysis indicates that fs(1)h encodes two putative proteins of approximately 120 and 210 kDa. The small isoform FSH-S, containing two widely spaced bromodomains (Haynes, 1992; Tamkun, 1992) and the extra terminal (ET) domain at its C terminus (Lygerou, 1994), is identical to the N-terminal half of the large isoform FSH-L (Haynes, 1989). Bromodomains can bind acetylated lysine or histones and are frequently found in transcription or chromatin modification factors, whereas ET domains are found in a small family of double-bromodomain proteins (BET proteins) with no designated function. Several interesting properties have been shown for mammalian BET proteins. For example, human RING3 (or BRD2) is a growth-stimulated nuclear kinase acting on serine and threonine. Mouse BRD2-like protein can be copurified with the Mediator transcriptional coactivator complex. Recently, mouse BRD4 has been shown to be involved in the recruitment of positive transcription elongation factor b (Chang, 2007).

This report provides several lines of evidence to support a direct role of fs(1)h in homeotic gene activation and the idea that FSH-S is primarily responsible for this function. Furthermore, it is shown that FSH-S acts directly on the Zeste site of the CPR. These results support a critical role for FSH-S in integrating signals from distal factors (Chang, 2007).

While many trxG mutations were identified by their suppressing effects on specific homeotic phenotypes caused by PcG mutations, their contributions to regulation of individual HOX genes have not been systematically examined. To address this issue, the effect of mutations of 18 trxG genes was examined on homeotic phenotypes caused by reduced Ubx expression, i.e., transformation of the third to second thoracic segment in adult flies. Interestingly, only fs(1)h [i.e., Df(1)C128], trx, and ash1 mutations showed strong enhancement on Ubx¹³⁰ phenotypes (increased from <1% to ~10%). Other trxG mutations showed weak or no effects on Ubx¹³⁰ mutation, despite that many could suppress PcG phenotypes as strongly as trx mutations. Thus, Ubx activation appeared to be highly sensitive to the dosages of fs(1)h, trx, and ash1. This selective effect was further supported by genetic interactions between fs(1)h and other trxG mutations. Again, fs(1)h showed strong synergistic effects with trx (>20%) and ash1 mutations (~10%) on Ubx phenotypes. By contrast, it showed weaker or no interactions with other trxG mutations. These results strongly suggest that fs(1)h, trx, and ash1 share some common role in certain critical steps of Ubx activation (Chang, 2007).

Loss of fs(1)h function results in complex defects in early embryos, leading to severe body distortion and lethality (Forquignon, 1981; Huang, 1990). These defects hamper the analysis of the role of fs(1)h in Ubx activation. To circumvent these problems, fs(1)h function was inactivated by shifting heat-sensitive fs(1)h¹ mutant embryos from the permissive temperature (21°C) to the restrictive temperature (29°C) during the onset of gastrulation. In wild-type embryos, high levels of Ubx transcripts can be detected in the ventral nerve cord (VNC) in a domain encompassing parasegments (PS) 5 to 12. In mutant embryos, a marked reduction of Ubx transcripts was seen. By contrast, no change was observed for caudal (cad), a HOX gene controlling the development of most posterior segments. Thus, fs(1)h appeared to be required for a subset of HOX genes (Chang, 2007).

fs(1)h encodes two double-bromodomain proteins, FSH-S and FSH-L. To define their roles in HOX activation, the Gal4/UAS binary system was used to induce high levels of FSH-S or FSH-L and examine their effects on HOX expression. UAS transgenes containing epitope-tagged FSH-S or FSH-L were driven by dpp-Gal4 in small subsets of imaginal cells. Targeted expression of FSH-S caused striking defects in the adult. Frequently, adults heads lacked maxillary palpi, and their aristae were transformed into distal legs with claws. Severe defects were also found in thoracic legs, including bifurcation of tibial segments and deletion of tarsal segments. Surprisingly, no discernible defect was seen in adults with targeted FSH-L expression, suggesting that FSH-L and FSH-S act differently (Chang, 2007).

Antenna-to-leg transformations can be induced by ectopic expression of the HOX gene Antennapedia (Antp) in antennal discs. To determine whether extra legs induced by FSH-S might be related to ectopic Antp expression, eye-antennal discs from third instar larvae were stained with an anti-ANTP antibody. Whereas ANTP is normally not expressed in these discs, strong ANTP signals were seen in antennal discs of transgenic animals. Using an anti-Flag antibody to mark tagged FSH-S, extensive overlaps were found between FSH-S and ANTP signals, suggesting that FSH-S is directly involved in ANTP induction. By contrast, no ectopic ANTP was induced by FSH-L, which was consistent with the normal appearance of adult flies. These effects further distinguished the role of FSH-S and FSH-L in HOX activation. Curiously, very little ANTP expression was induced by FSH-S in eye discs, despite its comparable levels in antennal and eye discs. The nature of this tissue-dependent response is unclear. Furthermore, no ectopic Ubx signal was found in eye-antennal and other discs. To avoid problems caused by induction timing or tissue dependence, an en-Gal4 line was used to drive FSH-S expression. Under such conditions, most larvae died before the third instar, while rare adult escapers (less than 1%) showed partial deletion of thoracic segments. In second instar larvae, ectopic Ubx signals could be detected in ventral ganglions. In addition to the transverse rows normally found within PS5 to PS12, Ubx signals appeared in small clusters of cells near the lateral margins of PS4, PS3, and PS2 at anteriorly diminishing frequencies. Occasionally, ectopic Ubx was found in PS4 extending to PS2 on both sides of the ganglion. These results strongly suggested that FSH-S can induce HOX genes (Chang, 2007).

Previous analysis predicted that fs(1)h protein products might be membrane associated, implicating a role in signal transduction. To further characterize FSH-S, antibodies were raised against three regions (S1, S2, and S3) common to both FSH-S and FSH-L. Using affinity-purified antibodies, two common bands were detected in embryonic extracts. The sizes of these two bands were consistent with predicted sizes of FSH proteins (~210 and ~120 kDa). In addition, an antibody specific for FSH-L (i.e., L3) reacted only with the larger protein. The authenticity of these proteins was further confirmed by the analysis of a larval-lethal mutant, fs(1)h¹⁷, which results from an insertion of a copia element in the intron following the FSH-S coding sequences. Unlike many other fs(1)h mutations, this mutation did not cause homeotic effects. Interestingly, the larger protein was severely diminished in mutant larvae at third instar, while the small one was unaffected. These results indicate that these proteins represent the two specific FSH isoforms and, more importantly, that FSH-L is not essential for the homeotic effect (Chang, 2007).

The developmental profiles and subcellular localization of FSH proteins in embryos were analyzed by immunostaining with affinity-purified S1 antibody. Muclear staining was clearly seen in syncytial embryos. Although S1 antibody reacted with both FSH-S and FSH-L, this nuclear staining was attributed to FSH-S, since FSH-L is primarily a centrosomal protein at this stage. In addition, tagged FSH-S was localized in the nuclei in both transgenic lines. The staining intensity appeared to be uniform throughout all developmental stages except in those cells located near invaginating furrows or in VNC. To further confirm the distribution pattern of FSH-S, whole-mount in situ hybridization was performed using a probe from the 3' UTR of FSH-S mRNA, which is absent in FSH-L mRNA. Again, a ubiquitous distribution of FSH-S transcripts was observed (Chang, 2007).

The genetic interactions, induction of HOX gene expression, nuclear localization, and the presence of a double bromodomain raised a strong possibility that FSH-S might directly affect HOX promoters. To test this, the ability of FSH-S to stimulate reporter constructs containing various promoters was tested by cotransfection experiments in a Drosophila haploid cell line which was shown to recapitulate Ubx regulation by trx and Pc. Reporter activities from constructs containing the P1 or P2 promoters of Antp or promoters of Ubx and the Heat shock protein 70 (Hsp70) were assayed following cotransfection of an Act5C-FSH-S effector or an Act5C control vector. The activities of the Antp-P2 and Ubx promoters were stimulated approximately 10-fold and 20-fold, respectively, while the Antp-P1 and Hsp70 promoters were only weakly affected. Thus, the effect of FSH-S appeared to be highly selective (Chang, 2007).

Several trxG genes have been shown to act on regulatory sequences located about 20 kb upstream of the initiation site. Since the UC construct used in this study only contained sequences from −3142 to +360, FSH-S appeared to act via distinct sequences near the basal promoter. To identify the FSH-S response elements (FRE), the effect of FSH-S on a series of Ubx deletion mutants was analyzed. Sequential deletion of 5' sequences from −3142 to −1762 (5Δ1), −628 (5Δ2), or −226 (5Δ3) did not alter the ability of the Ubx promoter to respond to cotransfected FSH-S. Deletion from +360 to +161 (3Δ1) resulted in a general reduction of the promoter activity by about twofold, regardless of the presence or absence of cotransfected FSH-S. Since the stimulatory effect of FSH-S was not affected, this downstream region most likely contains a positive element that is unrelated to FRE. No further effect was observed when sequences from +161 to +36 (3Δ2) were deleted. These results indicated that the FRE is not present in the regions upstream of −226 or downstream of +36. Consistently, a construct containing sequences from −226 to + 36 (3Δ22) was sufficient to respond to FSH-S. Conversely, an internal deletion of sequence from −200 to −32 (InΔ1; In is initiator) almost completely abolished the promoter activity. Since the initiator (ACATTC from −2 to +4) and downstream promoter elements (GGATA from +23 to +27) were intact in InΔ1 construct, the inactivation of the Ubx promoter should reflect the removal of regulatory elements. These results led to the conclusion that the FRE is located between −200 and −32, which corresponds to the CPR determined previously. Further refinement of the boundaries of the FRE was unsuccessful, since deletions from −226 to −127 (3Δ23) or from −127 to −32 inactivated the promoter (Chang, 2007).

Whether any specific domain of FSH-S is required for transactivation was examined. Mutant constructs carrying deletions of the N-terminal half of the first bromodomain (Δ1), the entire second bromodomain (Δ2) and its flanking sequences (Δ3), or both bromodomains (Δ12) or the C-terminal sequences including the ET domain (Δ4-6) were tested in transfection assays. It appeared that deletion of the first bromodomain results in a complete inactivation of FSH-S, suggesting a critical requirement of this domain. However, the full activity of FSH-S was also dependent on the second bromodomain and ET domain, since deletion of these domains resulted in partial inactivation. Interestingly, although FSH-L contains the entire FSH-S sequence, it appeared to be much less active than FSH-S. These results are consistent with the observation that FSH-L could not induce HOX genes in imaginal tissues and support further that FSH-S is primarily, if not exclusively, responsible for the transactivation function of fs(1)h (Chang, 2007).

Next, whether FSH-S could bind any specific sequences in the CPR was examined. An inducible S2 cell line containing the metallothionein promoter-driven Flag-tagged FSH-S was established. Tagged FSH-S was purified by immunoaffinity chromatography from whole-cell extracts after (NH₄)₂SO₄ enrichment. In addition to the major band corresponding to FSH-S, several less abundant proteins were also copurified. Although fs(1)h mutant showed strong genetic interactions with trx or ash1 mutants, FSH-S was not copurified with these proteins or Osa. In addition, FSH-S was not associated with Zeste protein, which was shown to bind the CPR. The ability of purified FSH-S to bind specific sequences of the CPR was demonstrated by EMSAs. Upon addition of increasing amounts of FSH-S to labeled Ubx-5 probe, a slower-migrating band appeared near the top of 3.5% native polyacrylamide gels, indicating the formation of protein-DNA complexes. The exceedingly slow mobility of this band suggested that a multisubunit protein complex is involved. FSH-S is a constituent of this putative complex, since a small but significant supershift was observed when an antibody against FSH-S was briefly incubated with FSH-S protein. A supershift was not observed when an antibody to FSH-L was used instead. Furthermore, this binding was sequence specific, since it could be completely blocked by the addition of excess amounts of unlabeled Ubx-5 or Ubx-6 but not by a random DNA fragment. Similar results were also obtained when Ubx-6 was used as the probe. To further narrow the binding region, four smaller probes from the CPR were used for EMSA. Specific binding was observed with probes Ubx-5b (−167 to ~−94) and Ubx-6a (−104 to ~−35) but not Ubx-5a (−226 to ~−146) or Ubx-6b (−55 to ~+36), indicating that the FRE is located between −167 and −35 (Chang, 2007).

The CPR contains clusters of binding sites for Zeste, Trl (also known as GAGA factor), and NTF-1. To determine whether any of these sites might correspond to FRE, competition assays were performed with DNA fragments containing tandem repeats of Zeste, Trl, or NTF-1 binding sites. Interestingly, only Zeste repeats effectively blocked binding activity. To exclude the possibility that fortuitous binding sites might be generated by multimerizaton of these repeats, an oligonucleotide containing one consensus Zeste site (CGAGTG) was tested with different flanking sequences. This oligonucleotide also blocked the binding activity of FSH-S. Thus, the Zeste site should represent the core FRE (Chang, 2007).

For further analyses of DNA binding properties, FSH-S and recombinant Zeste proteins were compared by an in-gel chemical footprinting technique. The DNA-cleaving ions OP-Cu used in this study gain more access to unprotected sequences than DNase I and are thus capable of revealing detailed differences in binding properties. Similar to studies with DNase I, three sites (Z1 to Z3) were protected by Zeste or FSH-S proteins in the Ubx-6a fragment. Despite an overall similarity, several important differences between these patterns were noticed. For example, the regions unprotected by Zeste produced bands with intensities comparable to those from free probes. However, fainter intervening bands were produced by FSH-S, suggesting weak protection on flanking sequences. Two additional differences were found over the Z1 site. FSH-S appeared to protect more 5′ sequences than Zeste. However, Zeste produced several bands more intense than the control, suggesting DNA distortion in this region. The lack of detectable Zeste protein and the distinct DNA binding properties exhibited by FSH-S clearly support the involvement of a novel binding factor (Chang, 2007).

If the FRE indeed corresponds to the Zeste site, the function of the Zeste site might be inactivated by fs(1)h mutations. Therefore, the effects were examined of fs(1)h mutation on expression of Ubx-lacZ transgenes containing two distal regulatory domains (BXD and ABX) and ~3 kb of immediate upstream sequences in addition to a wild-type CPR (Uβ) or tandem Zeste sites (Uβ-Z). In the wild-type background, strong lacZ signals were observed from PS5 to more posterior parts of the VNC in Uβ embryos. In addition, there was weaker misexpression in anterior parts of the VNC. The misexpression was more pronounced in Uβ-Z embryos. More importantly, lacZ transcripts were severely reduced throughout the entire VNC in both Uβ and Uβ-Z embryos upon inactivation of fs(1)h, indicating a strict requirement of fs(1)h. These results strongly support the physiological relevance of FSH-S to the Zeste site (Chang, 2007).

To further demonstrate that FSH-S is indeed associated with CPR of the endogenous promoter in vivo, chromatin immunoprecipitation assays were performed with formaldehyde-fixed chromatin prepared from male fs(1)h¹⁷ mutant larvae, which contain normal levels of FSH-S but diminishing amounts of FSH-L. Using five pairs of primers to cover sequences of more than 2 kb around Ubx start sites, it was found that FSH-S is preferentially associated with a CPR-containing DNA fragment. Although the antibody used here could cross-react with FSH-L, the contribution of FSH-L to the binding is excluded, because only a minute amount of FSH-L was present in fs(1)h¹⁷ mutant larvae, and, more importantly, no FSH-L signal was detectable in the Ubx promoter. In addition, this association appeared to be promoter specific, since only background signal was detected in the cad promoter (Chang, 2007).

Human RING3 protein, a FSH-S-like protein, has been shown to be a novel nuclear Ser/Thr kinase with scrambled subdomains (Denis, 1996). However, subsequent studies failed to show this activity in the mouse counterpart, FSRG-1, despite more than 90% sequence identity (Rhee, 1998). To determine whether FSH-S could act as a kinase, the kinase activity in FSH-S preparations was examined. Addition of [γ-³²P]ATP resulted in substantial phosphorylation of FSH-S and an additional protein of ~56 kDa. Because this smaller protein was consistently copurified, it will be referred to as FAP56 (FSH-associated protein of 56 kDa). Phosphoamino-acid analysis of in vitro phosphorylated proteins revealed that FSH-S was phosphorylated at the serine residue, while FAP56 was phosphorylated at both serine and threonine residues. Although FSH-S phosphorylation was readily detected by radioactive labeling, no mass increase was found upon incubation with 0.1 mM ATP. However, when treated with calf intestine phosphatase, the mass of FSH-S appeared to decrease slightly, indicating a limited phosphorylation of FSH-S (Chang, 2007).

The kinase activity of RING3 kinase could be restored by renaturation on nitrocellular filter after SDS-PAGE. Using this procedure, no FSH-S phosphorylation was detected in parallel experiments. However, it was reasoned that if FSH-S is a kinase, it must be able to bind ATP. An ATP analog, FSBA, has been used for affinity labeling of ATP binding proteins including kinases. Therefore, the reactivity of FSH-S toward FSBA was examined. Using an FSBA-specific antibody, it was found that FSH-S could indeed be covalently linked to FSBA. More importantly, the degree of cross-linking was substantially reduced by excessive ATP, indicating that FSH-S can bind ATP specifically (Chang, 2007).

Addition of FSH-S to cell extracts in which endogenous kinases were heat inactivated resulted in phosphorylation of many proteins, suggesting the presence of many kinase substrates. The clustering of multiple binding sites for FSH-S (or Zeste) and Trl in the CPR suggests that they might be spatially juxtaposed upon binding to the CPR, raising the possibility that Trl might be a potential kinase substrate. Using in vitro kinase assays, it was found that addition of FSH-S to purified recombinant Trl indeed resulted in its phosphorylation (Chang, 2007).

This report has provided several lines of evidence to support a direct role of FSH-S in HOX gene activation. Unlike other trxG proteins, FSH-S acts directly on the CPR of the Ubx promoter. The revelation of several interesting properties of FSH-S offers important mechanistic insights into the Ubx regulatory circuitry. Lack of functional fs(1)h is known to cause complex developmental defects including homeotic transformation and early embryonic lethality. The contribution of two different fs(1)h products to these effects had not been determined. Based on the following observations, it is suggested that FSH-S is primarily involved in HOX regulation. First, it was shown that FSH-S, but not FSH-L, can effectively activate homeotic promoters in imaginal discs and cultured cells. Second, FSH-S is a nuclear protein, while FSH-L is mainly found in centrosomes and is involved in organization of mitotic spindles in early embryos. Third, FSH-S can bind and function both in vitro and in vivo through a specific motif in the CPR. Lastly, no homeotic phenotype has been observed in an fs(1)h¹⁷ mutant lacking FSH-L. Thus, FSH-S is directly responsible for the homeotic effect of fs(1)h. Although FSH-L contains the entire sequence of FSH-S, these results clearly indicate that it does not play any significant role in the homeotic effect. The complex developmental functions of fs(1)h are very likely to be divided between different isoforms (Chang, 2007).

The abilities of FSH-S to bind a specific motif and to affect promoter activity through the CPR indicate that FSH-S plays an important role at the CPR for activation of the Ubx promoter. Among 18 trxG genes examined, fs(1)h, trx, and ash1 form a small but interesting subgroup that is most critical for Ubx activation and is known to act through specific regulatory sequences. Previous studies have shown that TRX and ASH1 act primarily through distal sequences that are essential for domain-specific Ubx expression. Recently, they have also been implicated in transcriptional elongation by their association with promoter and transcribed sequences. FSH-S is the only factor that functions primarily, if not entirely, on the CPR. Given the critical role of the CPR in promoter activity, FSH-S is very likely to play a key role in integration of activating signals from distal elements and factors. The strong synergistic effects reported in this study for fs(1)h, trx, and ash1 mutations indicate that they are involved in a critical step of Ubx promoter activation and that intimate functional relationships probably exist between these factors. Although they appear to exist in distinct protein complexes, it is highly likely that they interact directly or through associated factors. Such interactions may facilitate the action of TRX and ASH1 in the promoter and more downstream regions. An alternative (but not mutually exclusive) possibility is that FSH-S might be involved in attenuation of the repressing activity of PcG proteins. Since the distal response elements for PcG proteins and TRX/ASH1 are largely overlapping and their histone modification activities are functionally antagonistic, destabilization of PcG complexes could result in more efficient occupancy and/or more potent chromatin modification by TRX and ASH1. In either case, the activities of these distal factors might also be modulated by the kinase activity associated with FSH-S (Chang, 2007).

The stimulatory effects of FSH-S on the Ubx basal promoter also suggest that FSH-S may directly affect the basal transcription machinery. A closely related Saccharomyces cerevisiae protein, BDF1, has been shown to be a TFIID-associated factor, acting potentially as a functional substitute for TAF1 in higher organisms (Matangkasombut, 2000). Mouse BRD4 stimulates transcription by binding to positive transcription elongation factor b (Jang, 2005; Yang, 2005). Although it is unclear whether FSH-S possesses similar activities, the presence of structurally similar domains suggests that it may interact with these basal transcription factors. Therefore, it is speculated that FSH-S provides a dual interface for interactions with distal factors and basal transcriptional machinery for optimal Ubx transcription (Chang, 2007).

The sharing of the same target sequences between FSH-S and Zeste may help clarify a long-standing enigma about the role of Zeste in Ubx regulation. The function of zeste was revealed by a pairing-dependent phenomenon called transvection in which Ubx alleles with defective promoters can partially complement alleles with impaired regulatory sequences. Thus, zeste can facilitate the transutilization of the regulatory sequences on one chromosome by the Ubx promoter on a paired homologous chromosome. However, zeste is not required for expression of an intact endogenous Ubx gene or expression of a Ubx transgene containing more complete regulatory sequences (i.e., 35-kb sequences in 35UZ transgene), despite the fact that Zeste binds to the CPR and is required for expression of Ubx transgenes containing partial regulatory sequences. Moreover, zeste is dispensable for viability. These findings indicate that zeste is not essential for Ubx expression under normal genetic contexts. In contrast, FSH-S is indispensable for Ubx regulation and for development. The ability of FSH-S to bind the same target sequences indicates that FSH-S represents a critical component of a regulatory circuitry that utilizes regulatory signals present on the same chromosome to insure proper transcription of intact Ubx promoter. It is interesting that zeste-independent transvection has also been found that appears to employ the mechanisms that normally operate between the distal elements and the proximal promoter. It is speculated that FSH-S is also very likely to play a role in zeste-independent transvection (Chang, 2007).

The finding of DNA binding activity in FSH-S is surprising, since the double-bromodomain and the C-terminal ET domain, two prominent domains required for the function of FSH-S, are not known for DNA binding activity. It is possible that FSH-S may possess a novel DNA binding domain. Alternatively, the binding activity might be contributed by a factor that is associated with FSH-S, since several proteins were copurified with FSH-S and since recombinant FSH-S did not show the same activity. Further characterization of FSH-S and associated factors is necessary to resolve this question (Chang, 2007).

Another interesting feature of FSH-S is the kinase activity. The structural similarities to the RING3 nuclear kinase, the detection of a similar kinase activity, and the ATP binding activity are consistent with the notion that FSH-S contains a Ser/Thr kinase activity. However, the lack of kinase activity in bacterially expressed FSH-S suggests that posttranslational modification or an additional factor(s) is required for such an activity. It is interesting that, in addition to FSH-S and FAP56, many other proteins including Trl can be phosphorylated by FSH-S in vitro, suggesting a broad substrate specificity. Thus, it seems plausible that FSH-S may modulate the activities of factors that are brought into its proximity. Once it occupies the CPR, it is speculated that FSH-S may affect multiple factors that are in close contact with CPR by either short- or long-range interactions (Chang, 2007).

RNA Polymerase II "Pause" Prepares Promoters for Upcoming Transcription during Drosophila Development.

According to previous studies, during Drosophila embryogenesis, the recruitment of RNA polymerase II precedes active gene transcription. This work is aimed at exploring whether this mechanism is used during Drosophila metamorphosis. In addition, the composition of the RNA polymerase II "paused" complexes associated with promoters at different developmental stages are described in detail. For this purpose, ChIP-Seq analysis was performed using antibodies for various modifications of RNA polymerase II (total, Pol II CTD Ser5P, and Pol II CTD Ser2P) as well as for subunits of the NELF, DSIF, and PAF complexes and Brd4/Fs(1)h that control transcription elongation. It was found that during metamorphosis, similar to mid-embryogenesis, the promoters were bound by RNA polymerase II in the "paused" state, preparing for activation at later stages of development. During mid-embryogenesis, RNA polymerase II in a "pause" state was phosphorylated at Ser5 and Ser2 of Pol II CTD and bound the NELF, DSIF, and PAF complexes, but not Brd4/Fs(1)h. During metamorphosis, the "paused" RNA polymerase II complex included Brd4/Fs(1)h in addition to NELF, DSIF, and PAF. The RNA polymerase II in this complex was phosphorylated at Ser5 of Pol II CTD, but not at Ser2. These results indicate that, during mid-embryogenesis, RNA polymerase II stalls in the "post-pause" state, being phosphorylated at Ser2 of Pol II CTD (after the stage of p-TEFb action). During metamorphosis, the "pause" mechanism is closer to classical promoter-proximal pausing and is characterized by a low level of Pol II CTD Ser2P (Mazina, 2022).

This study aimed to fill a gap in the knowledge regarding how Drosophila uses the RNA polymerase II "pause" to prepare promoters for active transcription at the next stage of development. The main purpose was to determine whether the "pause" is involved in the preparation of genes for transcription at various stages of development. Drosophila development provides a very convenient opportunity for this by allowing the obtaining of material that is highly synchronized in terms of developmental stages, not only during embryogenesis but also during the metamorphosis phase (Mazina, 2022).

Analyzing the pools of "6–8 h genes", "10–12 h genes", and "WL genes" activating during mid-embryogenesis and metamorphosis, RNA polymerase II binding to promoters was observed at the stages preceding the stages of their active transcription. The composition and properties of the "paused" RNA polymerase II complexes were found to differ in mid-embryogenesis and metamorphosis. The "pause" of RNA polymerase II in mid-embryogenesis is characterized by phosphorylation of its Pol II CTD not only by Ser5 but also by Ser2, which corresponds to the "post-pause" state, operating at the transcriptional step after the activity of the p-TEFb complex. In the course of metamorphosis, the genes use the more well-described type of RNA polymerase II "pause", i.e., promoter-proximal pausing, which is characterized by a high level of Pol II CTD Ser5 phosphorylation and a low degree of Pol II CTD Ser2 phosphorylation. The composition of the "paused" RNA polymerase II complexes in embryogenesis and metamorphosis differs in the number of associated elongators; the embryonic "pause" complex lacks Brd4/Fs(1)h due to the low expression level of this protein at this stage of development. The rest of the studied elongation regulators, namely, NELF, DSIF, and PAF, were found to be involved in the RNA polymerase II "pause" both in embryogenesis and metamorphosis (Mazina, 2022).

The performed cluster analysis showed that most of the promoters were not associated with RNA polymerase II before their activation, and the conclusions are valid only for some of the genes preparing for transcription. This is attributed to the limitations of ChIP-Seq. Because the entire embryo and larva were examined, it was not possible to detect tissue-specific binding events. Single-cell techniques may help to overcome this problem, and the implementation of such techniques appears to be a good development for the current work (Mazina, 2022).

It would seem that this is a natural conclusion, since elongation regulators directly interact with RNA polymerase II, but this is not very obvious. The process of elongation regulators recruitment to RNA polymerase II is very unclear and it may well be a multistage process. Additionally, the step of this multistage process may well be the recruitment of elongation regulators onto chromatin through interaction with DNA-binding proteins, and not directly with RNA polymerase II. With this recruitment mechanism, even in the absence of RNA polymerase II, binding of some elongation regulators to promoters would not the detected due to their recruitment by DNA-binding proteins. However, in all the analyzed pools of genes that did not contain RNA polymerase II on the promoters, the binding of elongation regulators with the promoters was not observed. Moreover, the cluster analysis showed that all the analyzed elongation regulators fell into the cluster of genes containing RNA polymerase on the promoters. That is, at least at the stages of Drosophila development that were analyzed, the recruitment of elongation regulators to the promoters occurred together with the recruitment of RNA polymerase II (Mazina, 2022).

It is worth noting that, in previous studies, the binding of elongation regulators with DNA in the absence of RNA polymerase II was observed. In recent works, the NELF-A subunit of the NELF complex was described as being able to bind not only promoters but also enhancers and PRE elements containing a relatively low level of RNA polymerase II. The distribution profile of NELF-A in the genome indicates that this particular NELF subunit can be recruited by DNA-binding proteins separately from other subunits of this complex and, most importantly, separately from RNA polymerase II. Additionally, the recruitment of this subunit may well be an early stage in the assembly of the full NELF complex. The Drosophila Brd4/Fs(1)h protein was previously found to be present not only in promoters and enhancers but also in sites enriched in architectural proteins, mostly not associated with RNA polymerase II. That is, Brd4/Fs(1)h recruitment can also occur not directly to RNA polymerase II, but through an intermediate step of its recruitment to chromatin via DNA-binding (architectural) proteins (Mazina, 2022).

It seems that some elongation regulators can indeed be recruited by DNA-binding proteins as a preliminary step in their binding to RNA polymerase II; however, judging by the data of this article, this does not occur on promoters (Mazina, 2022).

The data suggest that, during Drosophila development, genes prepare in advance for the upcoming transcription by pausing the RNA polymerase II at their promoters. It is assumed that productive transcription of these genes at the appropriate stage is achieved by resolving this "pause". In the case of promoter-proximal pausing, this is the recruitment of the p-TEFb complex to promoters or its activation if it is pre-recruited in an inactive HEXIM-suppressed state. In the case of a "post-pause", the "pause" release can be induced by the recruitment or modification of a certain subunit of the PAF complex, although it is too early to discuss the exact mechanism for this type of pause (Mazina, 2022).

A not entirely clear but interesting question concerns how the increase in the concentration of "pause-releasing" complexes on the targeted promoters is achieved. Is it gene-specific, as in the case with heat shock genes activated by recruitment of HSF1, which stimulates elongation? Or can there be a global change in the intracellular concentration of complexes stimulating elongation at certain stages of development? The change in the expression level of Brd4/Fs(1)h during development that was observed indicates that the second possibility may well be implemented. It is quite probable that for some genes that form partially prepared RNA polymerase II complexes on promoters, an increase in the concentration of Brd4/Fs(1)h in mid-embryogenesis can stimulate their productive transcription (Mazina, 2022).

The most advanced works in this area, namely, the control of gene transcription through the intracellular level of coregulators, refers to genes controlled by poised Pol II and released by TFIIH complex. Some time ago, it was demonstrated that a change in the concentration of TFIIH (a general transcriptional factor stimulating DNA melting and transcription initiation, that is, exit from the poised Pol II state) is controlled by the level of glucose. More recently, the intracellular level of TFIIH has been linked to the transcription of genes responsible for proliferative cell potential using a single-cell approach. It would be extremely interesting to study the level of other regulators that stimulate the release of various types of RNA polymerase "pauses" in cells during development as well as in the case of any external stimuli or the progression of pathologies (Mazina, 2022).

The transcription of developmental genes is under the control of a variety of regulatory systems that control the timing and specificity of transcription in a particular tissue as well as under the influence of master regulator proteins that control transcription in a particular part of the body. It takes time to implement and coordinate all these stimuli. Not surprisingly, developmental genes control their transcription by controlling productive elongation. This approach helps to form a transcriptional hub in the promoter region and ensures the specificity of all the necessary interactions with RNA polymerase II and GTFs. The study of the dynamics of such hubs in development can lead to a better understanding of the mechanisms of transcription regulation in general. Drosophila's rapid development is a convenient experimental model for this goal (Mazina, 2022).

Stepwise modifications of transcriptional hubs link pioneer factor activity to a burst of transcription

Binding of transcription factors (TFs) promotes the subsequent recruitment of coactivators and preinitiation complexes to initiate eukaryotic transcription, but this time course is usually not visualized. It is commonly assumed that recruited factors eventually co-reside in a higher-order structure, allowing distantly bound TFs to activate transcription at core promoters. This study used live imaging of endogenously tagged proteins, including the pioneer TF Zelda, the coactivator dBrd4 (Female sterile (1) homeotic), and RNA polymerase II (RNAPII), to define a cascade of events upstream of transcriptional initiation in early Drosophila embryos. These factors are sequentially and transiently recruited to discrete clusters during activation of non-histone genes. Zelda and the acetyltransferase dCBP (Nejire) nucleate dBrd4 clusters, which then trigger pre-transcriptional clustering of RNAPII. Subsequent transcriptional elongation disperses clusters of dBrd4 and RNAPII. These results suggest that activation of transcription by eukaryotic TFs involves a succession of distinct biomolecular condensates that culminates in a self-limiting burst of transcription (Cho, 2023).

In eukaryotes, the recruitment of RNA polymerase II (RNAPII) to transcription start sites on DNA depends on the assembly of the preinitiation complex (PIC) and is regulated by hundreds of trans-acting factors. In particular, transcription factors (TFs) recruit nucleosome remodelers, histone modifiers, and Mediator to promote the formation of PIC. How these numerous upstream inputs are integrated to give the extraordinary specificity and intricacy of transcriptional regulation remains incompletely understood. A common view suggested by biochemical studies is that these factors are progressively assembled into a single final complex through cooperative interactions. However, other sophisticated processes initiating DNA replication and promoting splicing of mRNAs are governed by a series of distinct and ephemeral complexes in which each complex promotes the next in energy-driven steps. This study examines the possibility that initiation of transcription similarly involves directional transformations of intermediate complexes that would provide additional opportunity for specificity and regulation (Cho, 2023).

Visualizing the composition of transcriptional machinery over time might detect intermediate complexes that integrate the multitude of regulatory inputs of transcriptional control. In recent years, advances in confocal and super-resolution imaging led to the discovery that a wide variety of transcriptional regulators are recruited to form clusters at active genes. These clusters are thought to function as 'transcriptional hubs' by locally enriching transcriptional machinery and enhancing their binding to target DNA sites. Transcriptional hubs are a type of membraneless compartment, whose formation typically involves the multivalent interaction between intrinsically disordered regions (IDRs). Accordingly, IDRs are commonly found in the activation domains of TFs as well as the C-terminal domain (CTD) of Rpb in RNAPII. Similar to the idea that a single final complex is assembled on the DNA to initiate transcription, it has been proposed that the heterotypic interactions between IDRs can give rise to a compartment that simultaneously enriches TFs, coactivators, Mediator, and RNAPII at promoters. Nonetheless, how transcriptional hubs are regulated and whether they undergo compositional changes are still unclear (Cho, 2023).

Studying the dynamics of transcriptional hubs in living cells is complicated by the discontinuous and stochastic nature of eukaryotic transcription, a phenomenon also known as bursting. The Drosophila embryo provides a powerful context to study the timing of events upstream of transcriptional initiation. The early wave of transcription in Drosophila embryos is coupled to the rapid nuclear division cycles such that a few hundred genes initiate a burst of transcription about 3 min after each mitosis. The synchrony of early nuclear cycles and real-time localization of tagged proteins allow one to track activation events prior to the onset of transcription, and tools to knockdown function are available to assess the contribution of events to gene activation. In a recent study, live imaging of endogenously tagged RNAPII revealed the abrupt appearance of RNAPII clusters 2–3 min after mitosis. Brief metabolic labelling revealed foci of nascent transcripts throughout the nuclei in fixed embryos—these foci broadly colocalized with RNAPII clusters, indicating that early-forming RNAPII clusters mark sites of active transcription. Importantly, as nascent transcript levels increased, RNAPII clusters declined and eventually dispersed. These observations are consistent with numerous observations and support a model in which a large excess of RNAPII is recruited prior to initiation, which is then inefficiently converted to elongating RNAPII. What produces this pre-transcriptional RNAPII clustering and how it is coordinated with a burst of transcription are not yet fully understood. In this study, events are followed during the ~2.5 min between mitotic exit and the formation of RNAPII clusters and the fate of these clusters as transcription ensues at about 3 min after mitosis (Cho, 2023).

Zelda (Zld) is a pioneer TF that widely promotes the early wave of zygotic gene expression. Maternally supplied Zld binds to thousands of enhancers and promoters, and its binding sites exhibit increased chromatin accessibility and histone acetylation. Depletion of maternally expressed Zld curtails early zygotic transcription, and the embryos become highly defective at the mid-blastula transition (MBT). The transactivation domain of Zld has been mapped to an intrinsically disordered region. Moreover, fluorescently tagged Zld forms highly dynamic clusters in the nucleus, and previous studies suggest that Zld clusters increase the local concentration of other TFs and facilitate their binding to target DNA. Knockdown of Zld reduces RNAPII 'speckles' in fixed embryos. While these previous studies support a model in which Zld promotes the recruitment of additional components to form transcriptional hubs and facilitates the onset of zygotic transcription, the exact mechanism has not been determined (Cho, 2023).

This study combined genetic perturbation and real-time imaging to delineate a pathway that nucleates and serially transforms transcriptional hubs to trigger initiation of transcription in early Drosophila embryos. Zld is shown to act through transcription coactivators, including the lysine acetyltransferase dCBP and the BET protein dBrd4, to initiate RNAPII clustering at non-histone genes. Importantly, real-time imaging reveals only limited colocalization of these factors at transcriptional hubs, suggesting dynamic and directional changes in the composition such that upstream activators do not stably persist in the hubs with downstream effectors and RNAPII. A model is proposed in which Zld forms numerous largely unstable clusters, some of which trigger a dCBP-dependent step to build more stable dBrd4 clusters; a subset of these dBrd4 clusters then promotes RNAPII clustering near active promoters, and this pool of RNAPII fuels a burst of transcription. Inhibition of transcriptional elongation stabilizes some Zld and dBrd4 clusters, indicating that transcription directly or indirectly promotes their dispersal. Finally, while early inhibition of transcription inhibits RNAPII clustering, abrupt inhibition of transcript elongation after hub formation stabilizes RNAPII clusters. These findings indicate that transcription destabilizes hubs, a feedback that could lead to cycles of RNAPII accumulation and depletion, thereby contributing to the busting feature of transcription. It is suggested that the onset of transcription, like the onset of replication, involves upstream events that directionally modify the machinery to precisely control the process (Cho, 2023).

It has long been recognized that the compartmentalization of transcriptional machinery is a fundamental aspect of eukaryotic gene control. Early cytological studies revealed discrete clusters of RNAPII and nascent transcripts, which were speculated to be stable "transcription factories". Subsequent studies show that rather than genes being recruited to stable factories, numerous factors form hubs or liquid-like condensates transiently at active genes. This leaves open the questions of what governs the dynamics of transcriptional hubs/condensates and how their emergence and dispersal are linked to transcript synthesis. This study used real-time approaches to dissect upstream events in transcriptional initiation whose timing is constrained and synchronized in early Drosophila embryos by coupling to the rapid cell cycles. A cascade of dependencies is documented paralleled by a temporal cascade of cluster formation. The findings indicate that transcriptional hubs directionally pass through a series of intermediate states with different composition, rather than simply enriching all the factors involved in initiating transcription. Specifically, the pioneer TF Zelda acts through coactivators dCBP and dBrd4 to indirectly concentrate pools of RNAPII near promoters. Inhibition of transcription by α-amanitin stabilizes dBdr4 and RNAPII clusters, indicating that transcription directly or indirectly promotes dispersal of transcriptional hub components resulting in negative feedback. It is suggested that the progressive maturation of transcriptional hubs coupled with a negative feedback-loop stimulates a rapid but self-limiting burst of transcription in the early rapid embryonic cycles. These findings have striking parallels to the proposal that non-equilibrium dynamics of transcriptional condensates make direct contributions to sequential transcriptional bursts in the longer cell cycles of more mature cells (Cho, 2023).

The dynamic nature of transcriptional hubs described in this study is distinct from the well characterized transcriptional condensates at nucleoli or histone locus bodies, which are stable compartments and incorporate multiple functionally related components. The dynamic process with its multiple transitions might serve to add precision and sophistication to transcriptional control. First, transitions between discrete steps could provide proofreading steps that test the stability of intermediate complexes to filter out stochastic noise and increase regulatory specificity. Second, additional regulators might promote or prevent passage through the different transitions, thereby allowing the transcriptional hubs to integrate multiple inputs to generate the intricate spatiotemporal expression of developmental genes. In line with these ideas, these data show that the transitions from Zld clusters to dBrd4 and then to RNAPII are each associated with a decline in the number of clusters, suggesting that the maturation of transcription hubs is selective at successive steps. It will be important to learn how this feature contributes to the extraordinary accuracy with which the graded and combinatorial inputs generate transcriptional outputs (Cho, 2023).

The molecular mechanisms that drive the sequential transformation of transcriptional hubs remain to be fully determined. During the first step, Zld and dCBP might directly interact with each other or undergo co-condensation. Alternatively, open chromatin established by Zld could facilitate binding of additional TFs that interact with dCBP. However, it should also be kept in mind that TFs might inhibit deacetylation to indirectly enhance local dCBP-dependent acetylation. In any case, it seems likely, but not yet demonstrated, that dCBP acts by increasing local acetylation to recruit the reader dBrd4. Although dBrd4 might simply bind to histone marks such as H3K27ac, the acetylation of transcriptional machinery could also be involved in recruiting dBrd4. Upon crossing a concentration threshold, dBrd4 clustering might be promoted by multivalent interactions mediated by its own IDR. While the initial clustering of RNAPII appears to spatially coincide with dBrd4 clusters, the subsequent behavior is not consistent with stable partnership, as dBrd4 is lost from temporarily persisting RNAPII clusters. Imaging the period of loss of dBrd4 revealed accompanying features that varied between clusters: abrupt physical rearrangement of foci, simple gradual loss of dBbr4 from complexes, and apparent de-mixing of previously colocalized signals to form largely separate dBrd4 and RNAPII clusters. These behaviors may represent different manifestations of progressive modifications of the biomolecular condensate that reduce the interactions that previously stabilized co-residency of dBdr4 and RNAPII. Finally, both positive and negative effects of transcriptional elongation on the dynamics of transcriptional hubs were observed. The initial requirement of transcription for RNAPII clustering might involve the upstream roles of enhancer RNAs in nucleating RNAPII cluster. In contrast, a later sustained period of transcription of the gene body appears to mediate negative feedback to disperse dBrd4 and RNAPII clusters. This could be explained by a suggested disruption of multivalent interaction between IDRs by the negative charge of nascent RNA but numerous other less direct mechanisms might be responsible. While thw results reveal the timing and coordination of upstream events required for transcription, much more work is needed to provide a mechanistic understanding of the observed processes (Cho, 2023).

Regardless of the molecular details, it is expected that similar regulatory principles are employed by evolutionarily diverse transcription factors to mediate transcriptional activation. For example, in the zebrafish embryo, the pioneer factors Nanog, Pou5f3, and Sox19b similarly recruit CBP/p300 and Brd4 to establish transcriptional competence during early zygotic gene expression. Activation by estrogen receptor α (ERα) also involves histone acetylation and subsequent recruitment of Brd46. Notably, elegant work has shown that dozens of factors are recruited to the ERα target promoter in a cyclical and sequential fashion. It is envision that many of these factors are dynamically recruited to the hubs, and that the enzymatic reactions they carry out contribute to the speed and irreversibility of the transformation of transcriptional hubs. Lastly, it is suggested that the formation of transcriptional hubs in early embryos ensures the rapid initiation of a transcriptional burst within a short interphase window; in other biological contexts, the hubs might serve additional functions such as bridging enhancers and promoters or coordinating expression of multiple loci65. The Drosophila embryos will provide a powerful system to dissect the relationship between transcriptional hubs, chromatin interactions, and transcription dynamics (Cho, 2023).

Different enhancer classes in Drosophila bind distinct architectural proteins and mediate unique chromatin interactions and 3D architecture

Genome-wide studies has identified two enhancer classes in Drosophila that interact with different core promoters: housekeeping enhancers (hkCP) and developmental enhancers (dCP). It is hypothesized that the two enhancer classes are occupied by distinct architectural proteins, affecting their enhancer-promoter contacts. It was determined that both enhancer classes are enriched for RNA Polymerase II, CBP, and architectural proteins but there are also distinctions. hkCP enhancers contain H3K4me3 and exclusively bind Cap-H2, Chromator, DREF and Z4, whereas dCP enhancers contain H3K4me1 and are more enriched for Rad21 and Fs(1)h-L. Additionally, the interactions of each enhancer class were mapped utilizing a Hi-C dataset with <1 kb resolution. Results suggest that hkCP enhancers are more likely to form multi-TSS interaction networks and be associated with topologically associating domain (TAD) borders, while dCP enhancers are more often bound to one or two TSSs and are enriched at chromatin loop anchors. The data support a model suggesting that the unique architectural protein occupancy within enhancers is one contributor to enhancer-promoter interaction specificity (Cubenas-Potts, 2017).

This study characterize the protein occupancy, chromatin interactions and architecture profiles for the two enhancer classes found in Drosophila. Each enhancer class has distinct H3K4 methylation states, is bound by both common and distinct architectural proteins, and is involved in distinct types of chromatin interactions. First, it was established that hkCP enhancers exclusively bind CAP-H2, Chromator, DREF and Z4, while dCP enhancers do not and are preferentially enriched for but not exclusively bound by Fs(1)h-L and Rad21. In addition, hkCP enhancers are more likely than dCP enhancers to associate with multiple TSSs, which promotes a higher transcriptional output. Finally, hkCP enhancers preferentially associate with topologically associating domain (TAD) borders, whereas dCP enhancers are enriched at chromatin loop anchors present inside TADs. Interestingly, enhancers activated by both core promoters exhibit more hkCP enhancer like characteristics, indicating that the both CP enhancers may represent an intermediate among the distinctive hkCP and dCP enhancers. Altogether, these results provide strong correlative evidence, supporting a model suggesting that architectural proteins are critical regulators of enhancer-promoter interaction specificity and that the interactions between enhancers and promoters significantly contribute to the generation of 3D chromatin architecture (Cubenas-Potts, 2017).

The importance of architectural proteins in regulating enhancer-promoter interactions in Drosophila is supported by the observation that the vast majority of architectural protein sites present in the genome correspond to enhancers and promoters. Historically, architectural proteins were identified as insulators, which were functionally demonstrated to block enhancer-promoter interactions. The insulator function of architectural proteins correlates with their enrichment at TAD borders. However, several lines of evidence, including ChIA-PET analysis of CTCF- and cohesin-mediated interactions in mammals, suggest that these architectural proteins help mediate long range contacts among regulatory sequences. In Drosophila this study observed that nearly all of the Group 1 and Group 2 architectural protein sites are associated with enhancers or promoters defined by STARR-seq, TSSs or CBP peaks, suggesting that architectural proteins help mediate enhancer-promoter interactions. Notably, Group 3 architectural proteins include the classic insulator proteins CTCF, CP190, Mod(mdg4) and SuHw, and at least 25% of their peaks cannot be explained by enhancers or promoters. It is interesting to speculate that the non-enhancer-promoter sites may be involved in more classical insulator functions or contributing to the chromatin architecture of inactive regions of the genome (Cubenas-Potts, 2017).

The conclusion that architectural proteins are critical regulators of the specificity between enhancers and promoters is supported by two main lines of evidence. First, the current results demonstrate a strong correlation between each enhancer class and distinct architectural protein subcomplexes. Functional evidence supporting this conclusion comes from mutational analyses of the DRE motif in the distinct enhancer classes, which likely recruits DREF and the other hkCP enhancer associated architectural proteins. Zabidi (2015) demonstrated that the tandem DRE motif alone was sufficient to enhance expression of the housekeeping core promoter and that mutation of DRE motifs within an hkCP enhancer reduced its promoter interactions in a luciferase assay. Furthermore, addition of a DRE motif to a dCP enhancer changed its promoter specificity. Because DREF and potentially BEAF-32 bind to the DRE motif, these results strongly support a model suggesting that the differential occupancy of Cap-H2, Chromator, DREF and Z4 in the two enhancer classes is a critical regulator of their specific interactions with the core promoter types. However, the data cannot discount the notion that unique transcription factor binding at proximal TSSs also contribute to the specificity of enhancer-promoter interactions. Although hkCP enhancer identity is most highly correlated with CAP-H2, Chromator, DREF and Z4 localization, these four architectural proteins are not found in isolation within hkCP enhancers. BEAF-32 and CP190 are also strongly enriched in hkCP enhancers, which are also associated with high occupancy APBSs and TAD borders. Thus, the full architectural protein complement at hkCP enhancers is far more complex than the four hkCP-specific architectural proteins. In addition, architectural proteins that are truly unique to dCP enhancers were not detected. Because dCP enhancers exhibit higher cell type specificity, it cannot be discounted that there are additional dCP enhancers present in the Drosophila genome that were not identified by STARR-seq and thus, excluded from this analysis. From these studies, it is unclear if the enrichment of Fs(1)h-L and Rad21, particularly because Fs(1)h-L and Rad21 are present in hkCP enhancers at lower levels, or the absence of BEAF-32, CAP-H2, Chromator, CP190, DREF and Z4 truly distinguishes the architectural protein complexes found at dCP enhancers. In the future, careful biochemical analyses will be required to gain a comprehensive understanding of the complete organization of architectural protein subcomplexes associated with each enhancer class (Cubenas-Potts, 2017).

hkCP enhancers are associated with multi-TSS chromatin interactions and TAD borders. The promoter-clustering by hkCP enhancers results in a dose-dependent increase in transcriptional output for the interacting genes. Thus, one likely molecular mechanism by which hkCP enhancers promote robust transcriptional activation is by increasing the local concentration of RNA Polymerase II and general transcription factors (GTFs) by bringing multiple TSSs into close proximity. It is interesting that the hkCP enhancers, which form promoter clusters, are associated with TAD borders. It is speculated that the hkCP enhancer interactions involve inter-TAD contacts within the A-type compartment, indicative of the formation of transcription factories (70). From this analysis, it is unclear if the hkCP enhancers alone are sufficient for the formation of the 3D interactions or the neighboring TSSs and their associated transcription factors are also contributing to these contacts. It is hypothesized that the genes recruited to the factories contain the housekeeping promoter motifs (DRE, Ohler 1, Ohler 6 and TCT) and that the hkCP enhancer residents Cap-H2, Chromator, DREF and Z4, are critical to the formation of these 3D contacts (Cubenas-Potts, 2017).

dCP enhancers are more likely to be present within TADs and are enriched on the subTAD-like chromatin loop anchors. dCP enhancers do not form promoter clusters, but are more likely to interact with individual TSSs. One possible explanation for this observation is that the genes interacting with dCP enhancers require the binding of sequence-specific transcription factors, and increasing the concentration of GTFs and RNA polymerase II is not an effective mechanism to promote transcriptional output. The chromatin loop association is consistent with dCP enhancers forming a strong contact with a single TSS. However, it is acknowledged that dCP enhancers are likely one of multiple molecular mechanisms contributing to chromatin loop formation. Surprisingly, the chromatin loops that were observed in Drosophila are distinct from the chromatin loops described in humans. A recent study reported approximately 10,000 chromatin loops in the genome of GM12878 lymphoblastoid cells, but this study detected only 458 chromatin loops in Drosophila utilizing a similar method. The reason why there are so few chromatin loops in Drosophila compared to humans is unclear. It is possible that chromatin loops represent a more precise level of architecture within TADs between specific enhancers and promoters in mammals, but because TADs are significantly smaller in flies (median size 32.5 kb compared to 880 kb in mice, the chromatin loops are not as prominent or easily detected in the Drosophila genome. Notably, it appears that the chromatin loops are generated by different architectural proteins in the two species. The chromatin loops in humans are anchored by convergent CTCF motifs, while the results presented in this study demonstrate that the chromatin loop anchors in Drosophila are depleted of CTCF. Because the chromatin loops in Drosophila show a strong enrichment for Fs(1)h-L, a Brd4 homolog, and the architectural proteins Rad21, Nup98, TFIIIC and Mod(mdg4), it is possible that a combination of transcription and architectural proteins is required for chromatin loop formation in flies, which may be different from mammals . Altogether, it is clear that dCP enhancers are involved in individual contacts with TSSs and are likely one mechanism by which chromatin loops form in Drosophila (Cubenas-Potts, 2017).

Surprisingly, only ~20% and ~12.5% of all hkCP enhancer and ~7.5% and ~8.5% of dCP enhancer interactions involve a TSS or enhancer on the opposite anchor, respectively. The biological significance of the enhancer to non-TSS association is unclear. One possible explanation is that current methods for identifying statistically significant interactions are not sufficiently robust and that many of the enhancer to non-TSS interactions are not representative of biologically significant contacts. However, it cannot be discounted that the non-TSS interactions mediated by enhancers are real and the biological significance of these contacts remains to be determined. Throughout this analysis, the patterns of TSS interactions were compared with each enhancer class instead of drawing conclusions about the absolute number of TSSs bound per enhancer, minimizing the impact of any non-specific interactions within the data. Additional molecular studies for the various type of enhancer interactions (enhancer to promoter, enhancer to non-TSS, etc.) will be required to evaluate the various biological contributions of each (Cubenas-Potts, 2017).

This study found that the functional differences between enhancers that activate housekeeping versus developmental genes are reflected in their chromatin and architectural protein composition, and in the type of interactions they mediate. hkCP enhancers are marked by H3K4me3, associate with TAD borders, and mediate large TSS-clustered interactions to promote robust transcription. This class of enhancers contain the architectural proteins CAP-H2, Chromator, DREF and Z4. In contrast, dCP enhancers are marked by H3K4me1, associate with chromatin loop anchors and are more commonly associated with single TSS-contacts. dCP enhancers are depleted of the hkCP-specific architectural proteins and show an enrichment for Fs(1)h-L and Rad21. The results support a model suggesting that the unique occupancy of architectural proteins in the distinct enhancer classes are key contributors to the types of interactions that enhancers can mediate genome-wide, ultimately affecting enhancer-promoter specificity and 3D chromatin organization. In the future, further characterization of the broadly defined housekeeping and developmental enhancers into smaller subclasses may yield additional levels of regulation and formation of unique architectural protein and transcription factor protein complexes as key mediators of long range chromatin contacts (Cubenas-Potts, 2017).

Double-bromo and extraterminal (BET) domain proteins regulate dendrite morphology and mechanosensory function

A complex array of genetic factors regulates neuronal dendrite morphology. Epigenetic regulation of gene expression represents a plausible mechanism to control pathways responsible for specific dendritic arbor shapes. By studying the Drosophila dendritic arborization (da) neurons, this study discovered a role of the double-bromodomain and extraterminal (BET) family proteins in regulating dendrite arbor complexity. A loss-of-function mutation in the single Drosophila BET protein encoded by female sterile 1 homeotic [fs(1)h] causes loss of fine, terminal dendritic branches. Moreover, fs(1)h is necessary for the induction of branching caused by a previously identified transcription factor, Cut (Ct), which regulates subtype-specific dendrite morphology. Finally, disrupting fs(1)h function impairs the mechanosensory response of class III da sensory neurons without compromising the expression of the ion channel NompC, which mediates the mechanosensitive response. Thus, these results identify a novel role for BET family proteins in regulating dendrite morphology and a possible separation of developmental pathways specifying neural cell morphology and ion channel expression. Since the BET proteins are known to bind acetylated histone tails, these results also suggest a role of epigenetic histone modifications and the 'histone code,' in regulating dendrite morphology (Bagley, 2014).

Dendrites are the primary site of information input to neural circuits, and the shape of dendritic arbors influences the electrophysiological responses of neurons. Due to the existence of highly diverse morphologies among different neuronal subtypes, a question of the relationship between form and function arises: By understanding how the shape of a neuron is specified, it is possible to understand how morphology relates to neural function and how altered morphology relates to dysfunction (Bagley, 2014).

Neurons can be defined by their physiology, morphology, and gene expression. Neuronal diversity is thought to arise from the combinatorial expression of genetic determinants. The dendritic arborization (da) sensory neurons of the Drosophila peripheral nervous system (PNS) constitute a powerful system to study genetic determinants of dendritic arbor morphology. In particular, the use of Drosophila genetic techniques to study the specification of stereotyped, subtype-specific dendritic arbor shapes resulted in the identification of multiple transcription factors, encoded by abrupt (ab), knot/collier (kn/col), spineless (ss), and cut (ct), which regulate dendritic arbor morphology. However, large-scale genomic analyses comparing the transcriptomes of various neural subtypes indicate a daunting amount of varied gene expression and implicate regulation by multiple transcription factors. Thus, a particular neuronal morphology is likely the result of coordination between multiple genomic programs (Bagley, 2014). Epigenetic modifications are one mechanism that could allow coordinated, genome-wide expression profiles. Chromatin is packaged into nucleosomes, where DNA nucleotides wrap an octamer of histone proteins. The chromatin structure can be altered through three main types of modifications, consisting of direct methylation of DNA nucleotides, post-translational histone-tail modifications such as acetylation and methylation, and ATP-dependent chromatin remodeling. ATP-dependent chromatin remodelers were first shown to regulate dendrite morphology when RNAi knockdown of brahma (brm)-associated protein 60kD (Bap60), Bap55, and the ATPase brm altered the dendritic arbors of class I da sensory neurons. In mammalian neurons, the neural-specific Brg/Brm-associated factor (BAF) complex (nBAF), which contains BAF53b and the ATPase Brg, regulates activity-dependent dendrite growth. In addition, the Drosophila BAF53a/b homolog Bap55 regulates dendritic targeting of olfactory projection neurons (PNs) (Bagley, 2014).

The post-translational modification of histone tails involves three types of molecules: The 'writers' add methyl, acetyl, or phospho groups and consist of histone methyl transferase (HMT), histone acetyltransferase (HAT), and kinase enzymes. The 'erasers' remove these modifications and include demethylases (DMTs), histone deacetyltransferases (HDACs), and phosphatases. Finally, the 'readers' are scaffolding proteins that recognize and bind acetyl, methyl, or phosphate modifications to position the 'writer' and 'eraser' enzymes along with transcriptional machinery to the correct genomic position and thereby modify gene expression. The discovery that the Polycomb repressor complex, which binds methylated histone tails, regulates da sensory neuron dendrite morphology indicates a role of histone methylation in dendrite development and a notion supported by the recent finding that the chromodomain Y-like (CDYL) protein negatively regulates dendritic complexity (Qi, 2014). Regarding a role of histone acetylation in dendrite morphogenesis, both HDAC and HAT activities have been implicated in regulating dendrite morphology. Specifically, the Drosophila HDAC1/2 homolog Rpd3 regulates class I da sensory neuron morphology and olfactory PN dendritic targeting. In addition, HDAC2 suppresses dendritic spine density of hippocampal CA1 and dentate granule neurons. The HAT enzyme Pcaf also regulates class I da sensory neuron dendrite morphology. A different HAT enzyme, CREB-binding protein (CBP), regulates the developmental pruning of class IV da sensory neuron dendrites, and mutations in the human homolog CREBBP cause the mental retardation syndrome Rubenstein-Taybi. While these studies indicate a definite role of 'writers' and 'erasers' of histone modifications in regulating dendrite morphogenesis, the role of 'reader' scaffolding proteins associated with histone acetylation has not been thoroughly investigated (Bagley, 2014).

Double-bromo and extraterminal (BET) domain-containing proteins bind acetylated histone tails (Umehara 2010a; Umehara 2010b) and modulate gene expression. In mice, mutations in one BET family member, BRD2, cause neural tube closure defects, behavioral abnormalities, and altered interneuron numbers. In addition, in certain human genomic population studies, mutations in BRD2 have been associated with juvenile myoclonic epilepsy and photosensitivity, which is frequently observed in idiopathic generalized epilepsies. In the current study, evidence is provided for a role of the Drosophila homolog of BRD2, encoded by female sterile 1 homeotic [fs(1)h], in regulating dendrite morphology and sensory function (Bagley, 2014).

This study examined the role of fs(1)h in dendritic development. The effect of a loss-of-function allele [fs(1)h¹¹¹²] was examined on the morphology of class III da sensory neurons in the Drosophila PNS. Overall, fs(1)h¹¹¹² causes a reduction in dendritic arbor complexity, most notably in the finer, higher-order branches. It was possible to partially rescue this reduced morphological complexity by reintroducing Drosophila Fsh-S or the human homolog (huBRD2) proteins. Furthermore, one aspect of the genetic mechanism of action for fs(1)h was found to be regulating the expression of ct (and possibly other genes in the pathway) in multiple da neuron subtypes as well as subtype-specific transcription factors, such as Abrupt for class I and Knot/Collier for class IV da neurons, which in turn affect subtype-specific dendrite development. The data show that fs(1)h regulates genetic pathways controlling dendritic arbor development but does not specify which ion channels are expressed. Finally, the results suggest that the subtype-specific spike morphology is important for an optimal response to relevant sensory stimuli in the mechanosensitive class III da neurons (Bagley, 2014).

The development of a dendritic arbor involves multiple steps beginning with differentiation, where a neuronal precursor acquires a neural fate. Next, neurites begin to extend, and a neuron becomes polarized as neurites are designated as axon or dendrite. The immature axons and dendrites continue to grow as the neuron and the nervous system develop. Initially, the dendritic arbors are simple, with only a few primary dendrites, but as development progresses, the number of branches and overall arbor size increase. The terminal dendrite branches are dynamic throughout development, exhibiting growth, retraction, or stability. In addition, as the animal body size increases, the dendritic field area increases, and therefore a dendritic arbor must scale accordingly. Thus, dendritic development involves a complex plethora of processes, and dendritic morphology could be altered by affecting any of these processes. For instance, if the balance of dendrite dynamics is shifted such that retraction is greater than growth, then dendritic branching will become reduced over time. This appears to be the case in fs(1)h¹¹¹² mutants, since an increase was observed in retracting branches with no change in growth as well as a decrease in the proportion of stable branches in dendritic arbors of fs(1)h¹¹¹² mutant da neuron clones compared with wild-type clones. Alternatively, if scaling of the dendritic arbor is affected, the size of the dendritic arbor will become disproportionately small as the body size of the animal increases throughout development. This does not seem to occur in fs(1)h¹¹¹² mutants because the primary dendrites of fs(1)h¹¹¹² arbors exhibited growth throughout development, although at a delayed rate. Instead, the number of spikes in class III da neuron arbors was reduced early in development and remained reduced throughout development, probably due to the increased amount of dendritic branch retraction and reduced stability. Since the primary dendritic branches were not affected to a large degree by loss of fs(1)h function, it is concluded that the major role of fs(1)h in dendritic development is to regulate dendritic complexity at the level of higher-order dendritic spikes. Moreover, the data suggest that fs(1)h affects dendritic arbor complexity by modulating the dynamics of terminal dendritic branches (Bagley, 2014).

In the da neurons, many molecules are known to regulate dendrite morphology. In particular, Ct, Ss, Ab, and Kn have been shown to regulate subtype-specific morphology of the four classes of da sensory neurons, and these proteins act in parallel genetic pathways. Moreover, the expression of Ct and Ss regulates class III da neuron spike morphology. This study observed a loss of Ct expression in fs(1)h¹¹¹² mutant class III da neuron clones, which suggests that fs(1)h regulates the induction or maintenance of Ct expression throughout class III da neuron development. However, reintroducing Ct expression to class III da neuron fs(1)h¹¹¹² clones did not rescue the nearly absent spike morphology. Therefore, the class III da neuron dendritic phenotype caused by loss of fs(1)h cannot be solely attributed to the loss of Ct protein. Since it is thought that Ct is a component of a genetic pathway responsible for subtype-specific dendritic arbor development, it is possible that fs(1)h regulates Ct expression as well as expression of genes necessary for the Ct pathway to affect dendritic morphology. Therefore, the relationship between ct and fs(1)h does not appear to be a linear pathway, and fs(1)h might regulate both upstream and downstream components of ct. These data indicating that fs(1)h is necessary for the Ct-induced overbranching and spike formation in class I da neuron dendrites support the idea that fs(1)h regulates the expression of downstream components of the Ct pathway, which are necessary for Ct-induced overbranching and spike formation. This hypothesis also explains why reintroducing Ct expression to fs(1)h¹¹¹² clones fails to rescue the dendrite phenotype. It is also known that Ct and Rac1 act synergistically to produce spike morphology. This study examined Rac1 overexpression in a fs(1)h¹¹¹² mutant background and found that Rac1 expression significantly rescued the loss of spikes in class III da neurons. However, Rac1-induced overbranching in class I da neurons was not affected by fs(1)h¹¹¹². Therefore, fs(1)h does not appear to regulate genes downstream from Rac1 but does regulate genes downstream from ct. Since these pathways are known to converge in order to regulate dendritic spike formation, the current data suggest that fs(1)h may be a crucial link between these two pathways. One possible scenario is that ct and Rac1 regulate parallel pathways, but ct may regulate the level of Rac1 expression such that increased Rac1 expression facilitates the formation of spikes. In this model, the results support the hypothesis that fs(1)h is necessary for the ct potentiation of Rac1 expression, which explains why increased expression of Rac1 with UAS-Rac1 causes a rescue of the class III da neuron dendritic phenotype in fs(1)h¹¹¹² mutants. Recent evidence indicates a role for reduced Rac1 expression in social defeat and depressive behavior in mice, possibly through regulating dendritic spine morphology (Golden, 2013). In these behavioral paradigms, reduced Rac1 expression occurred with altered epigenetic marks such that transcriptionally permissive histone H3 acetylation was reduced, while repressive histone H3 methylation was increased. Moreover, administering a class 1 HDAC inhibitor mitigated the reduced Rac1 expression. Thus, these data suggest that Rac1 expression can be regulated by histone acetylation. It is possible that epigenetic reader proteins, such as BET family proteins like fs(1)h, bind acetylated histone marks in the Rac1 promoter to recruit transcriptional machinery and in turn enhance Rac1 expression (Bagley, 2014).

In addition, overexpression of UAS-Fsh-S in class I da neurons did not cause an overbranching phenotype similar to UAS-Ct. In fact, there was no alteration of class I morphology, suggesting that Fsh-S is not sufficient to induce necessary components of the ct pathway to alter dendrite morphology. However, overexpression of Fsh-S in class III and class IV da neurons did cause a decrease in dendritic spike numbers. These data indicate that dendrite morphology may be sensitive to the amount of Fsh-S expression, which was confirmed by modulating the amount of overexpression by reducing GAL4/UAS activity with lower temperature. This may explain why it is possible to achieve only a partial rescue of the fs(1)h¹¹¹² dendritic phenotype with UAS-Fsh-S expression and why overexpression causes a dendritic phenotype similar to the phenotype caused by loss of Fsh-S. In support of this expression level hypothesis, it was observed that Fsh-S overexpression can reduce Ct-induced branching in class I da neurons. Since BRD2 is known to be part of a protein complex (Denis 2006), it is possible overexpression causes a gain-of-function or dominant-negative effect by altering the availability of complex components (Bagley, 2014).

Another possible explanation for the partial rescue of Fsh-S expression concerns the developmental timing of expression. Since these experiments were completed using MARCM, GAL80 is expressed until mitotic recombination occurs to generate the mutant clones. It is likely that GAL80 protein may persist for some time after the clones are formed, and the presence of GAL80 would block GAL4/UAS activity. Therefore, UAS-induced Fsh-S expression may occur at a delayed stage in embryonic development, which could produce a partial rescue. In actuality, a combination of both expression level and developmental timing probably explains the partial rescue of the fs(1)h¹¹¹² phenotype (Bagley, 2014).

While this study focused on the role of fs(1)h in regulating class III da neuron dendrite morphology, phenotypes were observed in other classes of the da neurons as well as expression of Fsh-S in all da neuron classes. In fs(1)h¹¹¹² mutants, a loss of Ct expression was observed in all da neurons that normally express Ct (classes II, III, and IV), suggesting that fs(1)h regulates Ct expression broadly among different neural subtypes. A loss of the class I-specific transcription factor Ab and the class IV-specific transcription factor Kn/Col wer also observed. Thus, it appears that fs(1)h can regulate the expression of subtype-specific gene expression among various neuron subtypes. The loss of Ct or the loss of Kn/Col could explain the reduction in class IV da neuron dendritic arbor complexity, and this further illustrates the pleiotropic nature of the fs(1)h¹¹¹² phenotype. The loss of Ab from class I da neurons should produce an increase in dendritic complexity, but interestingly, this did not occur in fs(1)h¹¹¹² mutants. Thus, these results consistently suggest that fs(1)h is necessary for dendritic arbor complexity, probably by regulating the expression of many different genes. In this manner, fs(1)h could act as a necessary gate for the gene expression responsible for establishing dendritic complexity (Bagley, 2014).

How can fs(1)h regulate gene expression? Histone modifications are a diverse set of post-translational modifications that produce a code whereby epigenetic reader proteins bind these modified histone tails with specificity for particular modifications, such as methylation, acetylation, or phosphorylation. Previous structural studies have shown that the bromodomains of BET family proteins form a hydrophobic pocket enveloping acetylated histone tails (Umehara 2010a; Umehara, 2010b). Moreover, histone acetylation is largely, but not exclusively, regarded as a mark for transcriptional activation. Therefore, fs(1)h may be required for transcriptional activation of gene expression, which has been shown in vitro with respect to Ubx (Chang, 2007). The current data suggest that fs(1)h is required for ct expression and is in agreement with the hypothesis that fs(1)h is a transcriptional activator. It is possible that expression of other genes in the ct pathway also depends on histone acetylation modifications for transcriptional activation, and this activation may require Fsh-S. This would explain the observed nonlinear genetic relationship between ct and fs(1)h. In addition, the results indicate a necessary, but not sufficient, role of fs(1)h in regulating gene expression. This may indicate that BET family proteins require histone acetylation marks to be established but that these scaffold reader proteins do not actively alter histone tail modifications (Bagley, 2014).

Histone modifications, termed the histone code, vary among different cell types and constitute a genome-wide mechanism for coordinating gene expression programs. This is intriguing because fs(1)h contains bromodomains that require histone acetylation to be first established at specific genomic regions in order to influence transcription at these regions. The current results suggest BET family proteins as candidates for reading this histone code to allow the development of dendritic complexity. It is important to note that although many proteins are observed with altered expression in fs(1)h¹¹¹² mutant da neurons, some proteins were unaltered, such as the mechanosensitive ion channel NompC. Furthermore, even though the Ct-induced overbranching in class I da neurons was blocked by fs(1)h¹¹¹², the Ct-induced NompC expression was normal. These data indicate some specificity to the action of fs(1)h in regulating dendritic morphology but not ion channel specification. It is possible that epigenetic 'reader' proteins, such as the BET proteins, coordinate the activity of many genetic pathways but with relevance to a specific outcome, such as regulating dendritic arbor morphology. In this model, the epigenetic 'readers' provide coordination and specificity of genome-wide histone marks to regulate particular aspects of neural cell biology. Moreover, it is conceivable that the specific genes regulated by fs(1)h could vary among different cell types depending on the cell type-specific histone code. This is supported by the different effects of UAS-Fsh-S overexpression in class I versus class III and IV da neurons as well as the loss of expression of cell type-specific transcription factors (Ab and Kn/Col) in fs(1)h¹¹¹² mutants. Currently, there is no atlas of the histone code for individual neural subpopulations. However, as the technology for conducting these types of analyses improves for distinct cell populations, it is conceivable that future studies can provide an answer to how cell type-specific histone modifications affect neural subtype-specific dendritic arbor morphologies (Bagley, 2014).

Finally, the results suggest that the specific morphological shape of the class III da neuron dendrites is important for their ability to appropriately respond to sensory stimuli. The results indicate that pathways regulating dendrite morphology, such as the ct pathway, are reduced in fs(1)h mutants, but other pathways involved in axon morphogenesis or cell type-specific physiology, such as NompC channel expression, remain active. Moreover, the number of spike protrusions correlates with the number of APs produced in response to a mechanosensitive stimulus. This was also observed in another study (Tsubouchi 2012) involving manipulation of the number of spiked protrusions through modulating Rac1 activity. In that study, the gentle touch response increases as spike numbers increase, causing elevated calcium activity detectable with GCaMP fluorescence imaging. Conversely, decreasing the spike numbers results in a decrease of the gentle touch response and calcium activity. One potential caveat to this study is that Rac1 can modulate many aspects of dendritic cell biology through modulating actin cytoskeletal dynamics, and therefore it is unclear whether manipulating Rac1 activity alters the electrophysiological properties or localization of ion channels such as NompC. The finding of a correlation between dendritic spike number and gentle touch/electrophysiological responses in fs(1)h mutant neurons with normal appearance of NompC expression implicates dendritic morphology in regulating touch sensitivity (Bagley, 2014).

Interestingly, NompC is expressed in fs(1)h¹¹¹² mutants, and its distribution throughout the dendritic arbor resembles that of wild-type neurons. While nompC mutants lack a mechanosensory response, neurons lacking fs(1)h still respond to mechanical stimuli, but the magnitude of the response (number of APs) is reduced for a given stimulus intensity. At the behavioral level, this manifests as a reduced response to gentle touch. Therefore, the data suggest that the unique dendritic spike morphology of class III dendrites contributes to their mechanical sensitivity (Bagley, 2014).

While various proteins involved in epigenetic regulation of gene expression have been implicated in dendrite morphogenesis, this study provides evidence that 'readers' of acetylated histone marks regulate dendrite morphology by demonstrating the involvement of BET family proteins in this process. Given the complexity of achieving a comprehensive view of molecularly defined neural subtypes, it is necessary to identify genome-wide mechanisms for molecular diversity that regulate dendritic morphology in order to further understand how morphological diversity is specified. Epigenetic regulators are an intriguing possibility in this endeavor, and future studies comparing gene expression profiles in mutants for regulators of histone modifications among neurons with varied morphologies may be one step forward in answering this fundamental question (Bagley, 2014).

Bivalent complexes of PRC1 with orthologs of BRD4 and MOZ/MORF target developmental genes in Drosophila

Regulatory decisions in Drosophila require Polycomb group (PcG) proteins to maintain the silent state and Trithorax group (TrxG) proteins to oppose silencing. Since PcG and TrxG are ubiquitous and lack apparent sequence specificity, a long-standing model is that targeting occurs via protein interactions; for instance, between repressors and PcG proteins. Instead, this study found that Pc-repressive complex 1 (PRC1) purifies with coactivators Fs(1)h [female sterile (1) homeotic] and Enok/Br140 during embryogenesis. Fs(1)h is a TrxG member and the ortholog of BRD4, a bromodomain protein that binds to acetylated histones and is a key transcriptional coactivator in mammals. Enok and Br140, another bromodomain protein, are orthologous to subunits of a mammalian MOZ/MORF acetyltransferase complex. This study confirmed PRC1-Br140 and PRC1-Fs(1)h interactions and identified their genomic binding sites. PRC1-Br140 bind developmental genes in fly embryos, with analogous co-occupancy of PRC1 and a Br140 ortholog, BRD1, at bivalent loci in human embryonic stem (ES) cells. It is proposed that identification of PRC1-Br140 'bivalent complexes' in fly embryos supports and extends the bivalency model posited in mammalian cells, in which the coexistence of H3K4me3 and H3K27me3 at developmental promoters represents a poised transcriptional state. It is further speculated that local competition between acetylation and deacetylation may play a critical role in the resolution of bivalent protein complexes during developments (Kang, 2017).

Inappropriate activation and/or repression of gene expression underlies many human diseases, yet the mechanisms that execute transitions in developmental gene expression remain poorly defined. How are genes chosen to be initially active or repressed, and how are transitions in gene activity managed with fidelity? Transcription factors clearly regulate these changes, but how can this regulation occur with such specificity when their consensus binding sites and genomic occupancy appear so promiscuous? Together, proteomic and ChIP-seq analyses suggest a model in which PRC1 and MOZ/MORF function to create a poised regulatory state during development (see Model for the role of bivalent complexes in developmental transitions of transcriptional state). As cells differentiate, bivalent protein complexes may eventually be diminished locally, as most loci resolve into either an active or silent state. It is speculated that the choice of activation may occur via increased acetylation, influenced by nearby transcription factors, and subsequent enrichment of Fs(1)h and TrxG proteins such as Ash1, which was specifically recovered in a Br140 pull-down. A transition toward silencing may involve deacetylation and a decrease in TrxG (Kang, 2017).

The retention of some bivalency after initial transcriptional choices are made in embryogenesis is likely to allow critical reversibility for subsequent gene expression programming. However, if transcriptional state is not dictated strictly by the occupancy of bivalent components, how are these states manifested? It is speculated that local post-translational modifications (PTMs) may be critical for the specification of transcriptional state and for reversibility. For example, the Enok subunit of dMOZ/MORF is known to acetylate H3K23, while this mark is incompatible with Pc chromodomain binding to H3K27me3 on the same histone tail. Interestingly, enrichment of H3K23ac from modENCODE data sets on the set of potentially bivalent genes was not observed, but further analysis will be required to investigate the significance of this finding. Competition between the cognate enzymatic activities within bivalent complexes and their interactors may be central to their ability to act as reversible switches of transcriptional state. Future studies to address this hypothesis will require improved approaches to comprehensive PTM detection as well as in vitro reconstitution of key interactions and biochemical activities of bivalent complexes containing the appropriately modified subunits (Kang, 2017).

The results are consistent with recent studies in which PRC1 is found on active genes in many systems, and PRC1 targeting is largely independent of PRC2 (Kahn, 2016). Most exciting is the likely conservation in zebrafish (Laue, 2008) and mice (Sheikh, 2015), based on the opposing genetic activities of PRC1 and MOZ/MORF complexes in regulation of the Hox genes. The reliance on a universal transducer of transcription factor activity in developmental decisions would be an elegant solution to the problem of widespread binding of sequence-specific regulators, as, in the model (see Model for the role of bivalent complexes in developmental transitions of transcriptional state), only local interactions with preset bivalency will result in functional consequences (Kang, 2017).

Key fundamental questions remain. In particular, how are PRC1 and MOZ/MORF targeted in the first place? PREs are cis-acting regulatory elements that can recruit PRC1 and PRC2 to target genes in Drosophila. PREs lack universal consensus sequences but contain combinations of motifs for many DNA-binding proteins. Therefore, diverse protein-protein interactions with the PcG could be critical for initial binding, as postulated from classical genetics. A speculative alternative is that the 5′ TSSs of developmentally regulated genes may remain epigenetically marked throughout the life cycle of the organism to specify the initial association of bivalent complexes. Both BRD4 and BRPF1 have been identified as 'bookmarking proteins' that may retain vital information throughout the cell cycle, based on their ability to remain at their chromosomal binding sites through mitosis (Dey, 2003; Laue, 2008). Furthermore, Fs(1)h and Enok are essential for oogenesis, and genic acetylation is detected very early in embryogenesis (Li, 2014). Finally, the importance of maternal E(z) suggests that H3K27me3 could be at least part of such an inherited mark for developmental genes (Kang, 2017).

In summary, the results provide evidence for bivalent protein complexes that may correspond to a bivalent transcriptional state in Drosophila embryos and mammalian stem cells. Beyond identification of these intriguing protein interactions in flies, it is speculated that their identity reveals a likely role for acetylation in the resolution of bivalency. It is envisioned that the choice toward activation may be triggered and maintained by cell-specific transcription factors that drive the acetylated state, favoring MOZ/MORF and BRD4 bromodomain-dependent association with chromatin. Cell type decisions may be governed by a constant assessment of the amount of acetylation at each TSS, consistent with the enrichment of deacetylases on even very active genes. Deacetylation would favor loss of bromodomain-acetyl interactions and ultimately the loss of coactivators, leading toward the establishment of a stably silenced state. The ability to regulate genes while only partially resolving bivalent complexes is likely to be critical for reversibility in response to changes in cell type-specific transcription factor expression and binding. It is proposed that regulatory elements possess the intrinsic ability to switch fate dependent on this local balance, with de novo targeting rarely required (Kang, 2017).

Coactivator condensation at super-enhancers links phase separation and gene control

Super-enhancers (SEs) are clusters of enhancers that cooperatively assemble a high density of the transcriptional apparatus to drive robust expression of genes with prominent roles in cell identity. This study demonstrates that the SE-enriched transcriptional coactivators BRD4 and MED1 form nuclear puncta at SEs that exhibit properties of liquid-like condensates and are disrupted by chemicals that perturb condensates. The intrinsically disordered regions (IDRs) of BRD4 (see Drosophila female sterile (1) homeotic) and MED1 (see Drosophila Mediator complex subunit 1) can form phase-separated droplets, and MED1-IDR droplets can compartmentalize and concentrate the transcription apparatus from nuclear extracts. These results support the idea that coactivators form phase-separated condensates at SEs that compartmentalize and concentrate the transcription apparatus, suggest a role for coactivator IDRs in this process, and offer insights into mechanisms involved in the control of key cell-identity genes (Sabari, 2018).

SEs regulate genes with prominent roles in healthy and diseased cellular states. SEs and their components have been proposed to form phase-separated condensates, but with no direct evidence. This study demonstrates that two key components of SEs, BRD4 and MED1, form nuclear condensates at sites of SE-driven transcription. Within these condensates, BRD4 and MED1 exhibit apparent diffusion coefficients similar to those previously reported for other proteins in phase-separated condensates in vivo. The IDRs of both BRD4 and MED1 are sufficient to form phase-separated droplets in vitro, and the MED1-IDR facilitates phase separation in living cells. Droplets formed by MED1-IDR are capable of concentrating transcriptional machinery in a transcriptionally competent nuclear extract. These results support a model in which transcriptional coactivators form phase-separated condensates that compartmentalize and concentrate the transcription apparatus at SE-regulated genes and identify SE components that likely play a role in phase separation (Sabari, 2018).

SEs are established by the binding of master TFs to enhancer clusters. These TFs typically consist of a structured DNA-binding domain and an intrinsically disordered transcriptional activation domain. The activation domains of these TFs recruit high densities of many transcription proteins, which, as a class, are enriched for IDRs. Although the exact client-scaffold relationship between these components remains unknown, it is likely that these protein sequences mediate weak multivalent interactions, thereby facilitating condensation. It is proposed that condensation of such high-valency factors at SEs creates a reaction crucible within the separated dense phase, where high local concentrations of the transcriptional machinery ensure robust gene expression (Sabari, 2018).

The nuclear organization of chromosomes is likely influenced by condensates at SEs. DNA interaction technologies indicate that the individual enhancers within the SEs have exceptionally high interaction frequencies with one another, consistent with the idea that condensates draw these elements into close proximity in the dense phase. Several recent studies suggest that SEs can interact with one another and may also contribute in this fashion to chromosome organization. Cohesin, an SMC (structural maintenance of chromosomes) protein complex, has been implicated in constraining SE-SE interactions because its loss causes extensive fusion of SEs within the nucleus. These SE-SE interactions may be due to a tendency of liquid-phase condensates to undergo fusion (Sabari, 2018).

The model whereby phase separation of coactivators compartmentalizes and concentrates the transcription apparatus at SEs and their regulated genes, described in this study and corroborated by (Cho, 2018), raises many questions. How does condensation contribute to regulation of transcriptional output? A study of RNA Pol II clusters, which may be phase-separated condensates, suggests a positive correlation between condensate lifetime and transcriptional output. What components drive formation and dissolution of transcriptional condensates? These studies indicated that BRD4 and MED1 likely participate, but the roles of DNA-binding TFs, RNA Pol II, and regulatory RNAs require further study. Why do some proteins, such as HP1a, contribute to phase-separated heterochromatin condensates and others contribute to euchromatic condensates? The rules that govern partitioning into specific types of condensates have begun to be studied and will need to be defined for proteins involved in transcriptional condensates. Does condensate misregulation contribute to pathological processes in disease, and will new insights into condensate behaviors present new opportunities for therapy? Mutations within IDRs and misregulation of phase separation have already been implicated in a number of neurodegenerative diseases. Tumor cells have exceptionally large SEs at driver oncogenes that are not found in their cell of origin, and some of these are exceptionally sensitive to drugs that target SE components. How can advantage be taken of phase separation principles established in physics and chemistry to more effectively improve understanding of this form of regulatory biology? Addressing these questions at the crossroads of physics, chemistry, and biology will require collaboration across these diverse sciences (Sabari, 2018).

Pumilio Regulates Sleep Homeostasis in Response to Chronic Sleep Deprivation in Drosophila melanogaster

Cells rely on a diverse repertoire of genes for maintaining homeostasis, but the transcriptional networks underlying their expression remain poorly understood. The MOF acetyltransferase-containing Non-Specific Lethal (NSL) complex is a broad transcription regulator. It is essential in Drosophila, and haploinsufficiency of the human KANSL1 subunit results in the Koolen-de Vries syndrome. This study, performed a genome-wide RNAi screen and identify the BET protein BRD4 (see Drosophila Female sterile (1) homeotic) as an evolutionary conserved co-factor of the NSL complex. Using Drosophila and mouse embryonic stem cells, this study characterised a recruitment hierarchy, where NSL-deposited histone acetylation enables BRD4 recruitment for transcription of constitutively active genes. Transcriptome analyses in Koolen-de Vries patient-derived fibroblasts reveals perturbations with a cellular homeostasis signature that are evoked by the NSL complex/BRD4 axis. It is proposed that BRD4 represents a conserved bridge between the NSL complex and transcription activation, and provide a new perspective in the understanding of their functions in healthy and diseased states (De Jesus-Olmo, 2020).

Search PubMed for articles about Drosophila fs(1)h

Bagley, J. A., Yan, Z., Zhang, W., Wildonger, J., Jan, L. Y., Jan, Y. N. (2014) Double-bromo and extraterminal (BET) domain proteins regulate dendrite morphology and mechanosensory function. Genes Dev 28: 1940-1956. PubMed ID: 25184680

Chang, Y. L., King, B., Lin, S. C., Kennison, J. A. and Huang, D. H. (2007). A double-bromodomain protein, FSH-S, activates the homeotic gene ultrabithorax through a critical promoter-proximal region. Mol Cell Biol 27: 5486-5498. PubMed ID: 17526731

Cho, C. Y. and O'Farrell, P. H. (2023). Stepwise modifications of transcriptional hubs link pioneer factor activity to a burst of transcription. Nat Commun 14(1): 4848. PubMed ID: 37563108

Cho, W. K., Spille, J. H., Hecht, M., Lee, C., Li, C., Grube, V. and Cisse, II (2018). Mediator and RNA polymerase II clusters associate in transcription-dependent condensates. Science 361(6400): 412-415. PubMed ID: 29930094

Cubenas-Potts, C., Rowley, M. J., Lyu, X., Li, G., Lei, E. P. and Corces, V. G. (2017). Different enhancer classes in Drosophila bind distinct architectural proteins and mediate unique chromatin interactions and 3D architecture. Nucleic Acids Res 45(4): 1714-1730. PubMed ID: 27899590

De Jesus-Olmo, L. A., Rodriguez, N., Francia, M., Aleman-Rios, J., Pacheco-Agosto, C. J., Ortega-Torres, J., Nieves, R., Fuenzalida-Uribe, N., Ghezzi, A. and Agosto, J. L. (2020). Pumilio Regulates Sleep Homeostasis in Response to Chronic Sleep Deprivation in Drosophila melanogaster. Front Neurosci 14: 319. PubMed ID: 32362810

Denis, G. V., and Green, M. R. (1996). A novel, mitogen-activated nuclear kinase is related to a Drosophila developmental regulator. Genes Dev. 10: 261-271. PubMed ID: 8595877

Denis, G. V., McComb, M. E., Faller, D. V., Sinha, A., Romesser, P. B. and Costello, C. E. (2006). Identification of transcription complexes that contain the double bromodomain protein Brd2 and chromatin remodeling machines. J Proteome Res 5: 502-511. PubMed ID: 16512664

Digan, M. E., et al. (1986). Genetic and molecular analysis of fs(1)h, a maternal effect homeotic gene in Drosophila. Dev. Biol. 114: 161-169. PubMed ID: 3007240

Forquignon, F. (1981). A maternal effect mutation leading to deficiencies of organs and homeotic transformations in the adults of Drosophila. Wilhelm Roux's Arch. 190: 132-138

Gans, M., Forquignon, F. and Masson, M. (1980). The role of dosage of the region 7D1-7D5-6 of the X chromosome in the production of homeotic transformation in Drosophila melanogaster. Genetics 96: 887-902. PubMed ID: 6790337

Golden, S. A., Christoffel, D. J., Heshmati, M., Hodes, G. E., Magida, J., Davis, K., Cahill, M. E., Dias, C., Ribeiro, E., Ables, J. L., Kennedy, P. J., Robison, A. J., Gonzalez-Maeso, J., Neve, R. L., Turecki, G., Ghose, S., Tamminga, C. A. and Russo, S. J. (2013). Epigenetic regulation of RAC1 induces synaptic remodeling in stress disorders and depression. Nat Med 19: 337-344. PubMed ID: 23416703

Haynes, S. R., et al. (1989). The Drosophila fsh locus, a maternal effect homeotic gene, encodes apparent membrane proteins. Dev. Biol. 134: 246-257. PubMed ID: 2567251

Haynes, S. R., et al. (1992). The bromodomain: a conserved sequence found in human, Drosophila and yeast proteins. Nucleic Acids Res. 20: 2603. PubMed ID: 1350857

Huang, D.-H. and Dawid, I. B. (1990). The maternal-effect gene fsh is essential for the specification of the central region of the Drosophila embryo. New Biol. 2: 163-170. PubMed ID: 1982070

Jang, M. K., et al. (2005). The bromodomain protein Brd4 is a positive regulatory component of P-TEFb and stimulates RNA polymerase II-dependent transcription. Mol. Cell 19: 523-534. PubMed ID: 16109376

Kang, H., Jung, Y. L., McElroy, K. A., Zee, B. M., Wallace, H. A., Woolnough, J. L., Park, P. J. and Kuroda, M. I. (2017). Bivalent complexes of PRC1 with orthologs of BRD4 and MOZ/MORF target developmental genes in Drosophila. Genes Dev 31(19): 1988-2002. PubMed ID: 29070704

Lygerou, Z., et al. (1994). The yeast BDF1 gene encodes a transcription factor involved in the expression of a broad class of genes including snRNAs. Nucleic Acids Res. 22: 5332-5340. PubMed ID: 7816623

Matangkasombut, O., et al. (2000). Bromodomain factor 1 corresponds to a missing piece of yeast TFIID. Genes Dev. 14: 951-962. PubMed ID: 10783167

Mazina, M. Y., Kovalenko, E. V., Evdokimova, A. A., Erokhin, M., Chetverina, D. and Vorobyeva, N. E. (2022). RNA Polymerase II "Pause" Prepares Promoters for Upcoming Transcription during Drosophila Development. Int J Mol Sci 23(18). PubMed ID: 36142573

Qi, C., Liu, S., Qin, R., Zhang, Y., Wang, G., Shang, Y., Wang, Y. and Liang, J. (2014). Coordinated regulation of dendrite arborization by epigenetic factors CDYL and EZH2. J Neurosci 34: 4494-4508. PubMed ID: 24671995

Rhee, K., et al. (1998). Expression and potential role of Fsrg1, a murine bromodomain-containing homologue of the Drosophila gene female sterile homeotic. J. Cell Sci. 111: 3541-3550. PubMed ID: 9811568

Sabari, B. R., Dall'Agnese, A., Boija, A., Klein, I. A., Coffey, E. L., Shrinivas, K., Abraham, B. J., Hannett, N. M., Zamudio, A. V., Manteiga, J. C., Li, C. H., Guo, Y. E., Day, D. S., Schuijers, J., Vasile, E., Malik, S., Hnisz, D., Lee, T. I., Cisse, II, Roeder, R. G., Sharp, P. A., Chakraborty, A. K. and Young, R. A. (2018). Coactivator condensation at super-enhancers links phase separation and gene control. Science 361(6400). PubMed ID: 29930091

Shearn, A. (1989). The ash-1, ash-2 and trithorax genes of Drosophila melanogaster are functionally related. Genetics 121: 517-525. PubMed ID: 2497049

Tamkun, J. W., et al. (1992). brahma: a regulator of Drosophila homeotic genes structurally related to the yeast transcriptional activator SNF2/SWI2. Cell 68: 561-572. PubMed ID: 1346755

Tsubouchi, A., Caldwell, J. C. and Tracey, W. D. (2012). Dendritic filopodia, Ripped Pocket, NOMPC, and NMDARs contribute to the sense of touch in Drosophila larvae. Curr Biol 22: 2124-2134. PubMed ID: 23103192

Umehara, T., Nakamura, Y., Jang, M. K., Nakano, K., Tanaka, A., Ozato, K., Padmanabhan, B. and Yokoyama, S. (2010a). Structural basis for acetylated histone H4 recognition by the human BRD2 bromodomain. J Biol Chem 285: 7610-7618. PubMed ID: 20048151

Umehara, T., Nakamura, Y., Wakamori, M., Ozato, K., Yokoyama, S. and Padmanabhan, B. (2010b). Structural implications for K5/K12-di-acetylated histone H4 recognition by the second bromodomain of BRD2. FEBS Lett 584: 3901-3908. PubMed ID: 20709061

Yang, Z., et al. (2005). Recruitment of P-TEFb for stimulation of transcriptional elongation by the bromodomain protein Brd4. Mol. Cell 19: 535-545. PubMed ID: 16109377

Zabidi, M. A., Arnold, C. D., Schernhuber, K., Pagani, M., Rath, M., Frank, O. and Stark, A. (2015). Enhancer-core-promoter specificity separates developmental and housekeeping gene regulation. Nature 518(7540): 556-559. PubMed ID: 25517091

date revised: 1 June 2024

Home page: The Interactive Fly © 2007 Thomas Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.