BIOLOGICAL OVERVIEW

A particular region (or domain) on the Enhancer of zeste protein has been shown to be responsible for gene silencing. To complicate matters, the same domain has been shown to be homologous to Trithorax, a known transcription activator. How can this be? How can a domain presumably involved in gene silencing be homologous to one involved in activation? The sharing of a domain between E(z), a transcription repressor, and Trithorax, a gene activator raises certain questions for which answers are not yet forthcoming: Which functions are located in the homologous region, and is their action competitive at the promoter binding site, or in interaction with the transcription apparatus, or with other proteins?

Two lines of evidence classify E(z) as a member of the trithorax group (trx-g). Firstly, double heterozygous combinations of mutant E(z) and absent, small or homeotic discs 1 (ash1), a trx-g gene, express homeotic transformation phenotypes similar to those found in double heterozygous combinations of mutant trithorax and ash1 alleles. Secondly, within thoracic imaginal discs of larvae hemizygous for mutant alleles of E(z) there is complete loss of accumulation of homeotic proteins Sex Combs Reduced, Antennapedia, and Ultrabithorax and the segmentation protein Engrailed. Similar loss of accumulation of these proteins is observed in trx-g mutations (LaJeunesse, 1996).

E(z) acting as a Polycomb group protein acts in the repression of knirps and giant. Hunchback protein is required to initiate repression of these genes, while E(z) secures and maintains such repression, in conjunction with the other PC-G proteins (Pelegri, 1995). Suppressor of zeste 2 and Posterior sex combs both partner E(z) in its role as a gene repressor. All three of these proteins dissociate from the chromosomes when a temperature sensitive E(z) mutant is inactivated by high temperatures. Gene silencing is lost in this circumstance as polytene chromosomes become decondensed at higher temperatures. Dissociation and decondensation occurs only a few hours after placing flies in non-permissive temperatures. This is graphic evidence that different PC-G proteins co-operate in a multi-protein complex. Inactivation of one protein causes the release of all from their chromosomal association (Rastelli, 1993).

Evidence is presented for direct physical interaction between the Extra sex combs (Esc) and E(z) proteins using yeast two-hybrid and in vitro binding assays. Coimmunoprecipitation from embryo extracts demonstrates association of Esc and E(z) in vivo. The Esc-binding domain of E(z) has been delimited to an N-terminal 33-amino-acid region. Only 6 of 33 amino acids in this region are identical among fly E(z), the murine homologs Ezh1 and Ezh2, and the human homologs EZH1 and EZH2. Site-directed mutations in the Esc protein previously shown to impair Esc function in vivo disrupt Esc-E(z) interactions in vitro. An in vitro interaction also occurs between the human eed (heed) and EZH1 proteins, which are human homologs of Esc and E(z), respectively. The amino acid sequences among these E(z) homologs are much more highly conserved in domains outside the 33 amino acid region. This result suggests that amino acids other than those that are evolutionarily conserved between E(z) and EZH1 participate in the interaction with Esc. It also suggests that Esc residues outside the absolutely conserved portions of the loops connecting blades 3, 4, and 5 of the WD repeat region are involved in binding to E(z) and EZH1. Additional studies will be required to further map the interacting domains of heed and EZH1 and to demonstrate their in vivo association. These results suggest that the Esc-E(z) molecular partnership has been conserved in evolution. Previous studies suggested that Esc is primarily involved in the early stages of Pc-G-mediated silencing during embryogenesis. However, E(z) is continuously required in order to maintain chromosome binding by other Pc-G proteins. In light of these earlier observations and the molecular data presented here, the paper discusses how Esc-E(z) protein complexes may contribute to transcriptional silencing by the Pc-G (Jones, 1998).

At least two lines of evidence indicate that Drosophila E(z) and Esc are not obligate partners at all loci and during all developmental stages. (1) Although both proteins display uniform spatial distributions in nuclei of blastoderm and early gastrulation-stage embryos, the Esc protein is much more limited than E(z) in mid-to-late-stage embryos. (2) There are target genes other than homeotic genes that require E(z), but not Esc, for repression during development. These examples show that the E(z) protein can be recruited to and function at target loci without assistance from the Esc protein. Presumably, E(z) action at these loci involves alternative binding partners. The ability of E(z) to sometimes act independently of Esc is one possible explanation for the differential efficiencies of reciprocal coimmunoprecipitations observed with the two proteins. It has been suggested that in addition to its role in Pc-G repression, E(z) may also be involved in the maintenance of transcriptional activity by trithorax-Group proteins. In contrast, the known role of the Esc protein is limited to repression. A dual role for E(z) could reflect participation in more than one type of protein complex. For example, when E(z) is bound to Esc, it is a component of silencing complexes. However, when E(z) is associated with other, as-yet-unidentified proteins, it may contribute to trx-G-mediated transcriptional activation. In support of this idea, heat inactivation of temperature-sensitive E(z) proteins reduces chromosomal binding by the Trx protein. To assess these possibilities, it will be necessary to define the number and constituents of E(z)-containing complexes isolated from fly embryos (Jones, 1998).

It was reasoned that Enhancer of zeste might reside in a protein complex that possess histone methyltransferase (MTase) activity. Pc contains a chromodomain whose structure and essential residues are homologous to those found in Hp1 and related methyl lysine binding proteins and might therefore recognize a nucleosomal methylation mark. E(z) complex have now been shown to contain an H3 MTase activity. To study the Esc/E(z) complex, Drosophila nuclear extract was fractionated and it was asked if a MTase activity copurifies with E(z) and Esc. A complex containing E(z) and Esc trimethylates lysine 9 and methylates lysine 27 of histone H3 and the trimethylated lysine 9 mark is closely correlated with PcG binding sites on polytene chromosomes. The conjecture that the Esc/E(z) complex contains a histone MTase activity has been confirmed by the finding that the complex immunoprecipitated by anti-Esc or purified biochemically methylates in vitro histone H3 whether assembled in a nucleosome, as a free histone, or in the form of oligopeptides. The purified complex contains several components as predicted from previous studies: in addition to E(z), Su(z)12, Esc, p55, and Rpd3 are found. An additional component of approximately 168 kDa remains to be identified (Czermin, 2002).

The activity of the Esc/E(z) complex leads to trimethylation of lysine 9 and probably also of lysine 27 of H3. In vivo, an antibody directed against dimethylated H3 lysine (9me2K9 antibody) does not detectably stain chromosomal PcG sites while a me3K9 antibody directed against trimethylated H3 lysine decorates all chromosomal PcG sites. Whether three methyl groups are added processively or by independent events and whether lysine 9 and lysine 27 are targeted simultaneously remains to be elucidated. The partial ability of the complex to methylate the peptide acetylated at lysine 9 is accounted for by the presence of the Rpd3 deacetylase (Czermin, 2002).

Polycomb complexes and the propagation of the methylation mark at the Drosophila Ubx gene

Polycomb group proteins are transcriptional repressors that control many developmental genes. The Polycomb group protein Enhancer of Zeste has been shown in vitro to methylate specifically lysine 27 and lysine 9 of histone H3 but the role of this modification in Polycomb silencing is unknown. This study shows that H3 trimethylated at lysine 27 is found on the entire Ubx gene silenced by Polycomb. However, Enhancer of Zeste and other Polycomb group proteins stay primarily localized at their response elements, which appear to be the least methylated parts of the silenced gene. These results suggest that, contrary to the prevailing view, the Polycomb group proteins and methyltransferase complexes are recruited to the Polycomb response elements independently of histone methylation and then loop over to scan the entire region, methylating all accessible nucleosomes. It is proposed that the Polycomb chromodomain is required for the looping mechanism that spreads methylation over a broad domain, which in turn is required for the stability of the Polycomb group protein complex. Both the spread of methylation from the Polycomb response elements, and the silencing effect can be blocked by the gypsy insulator (Kahn, 2006).

The experiments described in this study show clearly that all three PcG proteins tested, Pc, Psc, and E(z), are preferentially located at the PREs. This specificity is clearest and most sharply delineated in the case of Psc and E(z). In the case of PC, the peaks centered at the PREs are much broader, including secondary peaks, and although the binding detected at other Ubx regions decreases to low values, it never reaches the level seen at control sites such as the white gene that possess no PRE. The second basic conclusion from these experiments is that, in contrast to the localization of PcG proteins, the H3 me3K27 profile forms a broad domain that includes the entire Ubx transcription unit and upstream regulatory region. The third important observation is that, contrary to previously published accounts, the PREs themselves appear to contain the lowest levels of me3K27 of the entire domain. This surprising result will be considered first. The lack of apparent methylation at the PREs does not depend on the antibody used or on the level of cross-linking. Comparable results were obtained with two different anti-me3K27 antibodies and with anti-me3K9. Furthermore, the fact that a similar result was obtained with antibody against total histone H3 or histone H2B suggests that nucleosomes are underrepresented at the bxd PRE core. The result is not because of lack of accessibility to the histone or to the epitope: GAGA factor bound to the PRE appears as easily accessible as GAGA factor bound to the Ubx promoter. Salt extraction of the PcG complexes before cross-linking does not qualitatively change the me3K27 binding profile (Kahn, 2006).

In sum, careful quantitative analysis of ChIP indicates that while PcG proteins are principally localized at the PRE, the histone H3 methylation they produce is distributed over the entire Ubx gene. It is evident from this and from the undermethylation of the PRE core that that K27 methylation does not, by itself, recruit PcG complexes. This does not preclude an important role for methylation in PcG binding and silencing but suggests that the relationship between the two requires a more dynamic model (Kahn, 2006).

PREs have been shown to recruit PcG complexes and to produce new binding foci detectable in polytene chromosomes. It is not surprising therefore to find the three PcG proteins tested are associated with the two Ubx PREs. A much smaller peak in microarray profiles for all three proteins can be discerned in the vicinity of the Ubx 3'-exon but its significance is unknown. More surprising was the striking difference between the distributions of Psc and E(z) and that of Pc. E(z) and Psc belong to two different complexes that do not co-precipitate (except in the very early embryo) but are both recruited to the PRE. Pc and Psc are core components of the PRC1-type of PcG complex yet, while Psc is detected almost exclusively at the PREs, Pc has a much broader distribution peaking at the PREs but tailing over considerable distances along the Ubx gene and regulatory regions. The simplest interpretation of this is that a second type of complex containing Pc but not Psc is recruited by a different mechanism to the rest of the Ubx sequences. Alternatively, the same complex, containing both proteins is involved in both cases but the nature of the chromatin contact is different, such that in one case both proteins are well cross-linked to the chromatin but in the second case only Pc is efficiently cross-linked (Kahn, 2006).

Just as striking is the fact that, although the E(z) complex is responsible for the H3 K27 methylation spread over the entire Ubx gene, the E(z) protein is found localized at the PREs. It is concluded that the E(z) complex methylates the Ubx domain by a hit-and-run type of mechanism. Because the methylation is stable, the E(z) complex needs only visit each nucleosome once on the average every cell cycle. It is noted that E(z)-dependent histone H3 K27 dimethylation is highly abundant and widely distributed in the genome but E(z) complexes are not associated with it. Where then does the E(z) that methylates PcG target genes come from? While more complicated scenarios may be imagined, the simplest one involves the E(z) complex bound at the PRE (Kahn, 2006).

It is supposed that the PcG complexes are recruited to PREs by DNA-binding proteins independently of histone methylation. To methylate the entire Ubx domain, the E(z)/ESC histone MTase complex might then detach from the PRE and slide along the chromatin from one nucleosome to the next to survey the entire domain. However, it more likely that both the Pc and the E(z) complexes assembled at the PRE remain associated with the PRE sequences, where they are detected, but that the whole PRE assembly loops over to scan the entire region, methylating all accessible nucleosomes. Such looping models were originally proposed to be mediated by sites of weak PcG complex formation. In a modern version of this type of model, the looping activity would be mediated by the distinct affinity of the Pc protein for histone H3, which is greatly increased by K27 methylation. These affinities would mediate transient interactions of the complexes bound at the PRE with the surrounding chromatin and allow continuous scanning and methylation of unmethylated or hemimethylated nucleosomes (Kahn, 2006).

In such a model, ChIP experiments would always detect a strong PcG presence at the PRE but PcG interactions with the rest of the repressed gene would be distributed over a region, which is very large in the case of the Ubx gene, smaller in the case of the YGPhsW transposon, hence the signal detected at any one site would be weaker in proportion to the extent of the methylated domain. In addition, the contacts between the PcG complex and the rest of the silenced gene would be much more transient than contacts with the PRE. Together, these considerations would explain why ChIP assay gives such low values for PcG proteins over the rest of the methylated domain (Kahn, 2006).

The looping mechanism proposed for the PRE-bound complex strongly resembles that suggested for the interaction between the Locus Control Region and β-globin genes or for enhancer-promoter interactions. Like these interactions, the silencing of a promoter by the PRE is blocked by insulator elements. In transposon constructs, the insertion of a gypsy Su(Hw) insulator between PRE and promoter blocks the spread of methylation. At present, the mechanism of insulator action is not clear and how the block to methylation is achieved is unknown. It is possible that the insulator element produces topological constraints that prevent the PRE-bound complexes from looping beyond the insulator. This would be consistent with the observation that a significant level of Pc presence becomes detectable over the yellow gene when the insulator block is lifted (Kahn, 2006).

Although the data argue against a principal role of histone methylation in the recruitment of Polycomb proteins to their response elements, it seems to be important for both transcriptional repression and stable association of PcG proteins with chromatin. Loss of catalytic E(z) function eventually results in derepression of HOX genes and dissociation of PcG proteins from polytene chromosomes. It is speculated that once the me3K27 domain is established, modified nucleosomes will pave the way for looping interactions of the PRE-bound PcG proteins with the parts of the silenced gene including promoter or enhancer regions. Silencing might then result from hit-and-run interactions with either or both, possibly even resulting in methylation of the associated factors. Alternatively or in addition, trimethylation of K27 and possibly K9 may directly interfere with the signaling cascade of consecutive histone modifications that guide the multistep process of transcription initiation and elongation. Since histone methylation is thought to persist through cell division its immediate presence at the very beginning of the subsequent interphase might win the time necessary for the full assembly of PcG complexes on the PREs before competing transcription has taken over (Kahn, 2006).

A strand-specific switch in noncoding transcription switches the function of a Polycomb/Trithorax response element

Polycomb/Trithorax response elements (PRE/TREs) can switch their function reversibly between silencing and activation by mechanisms that are poorly understood. This study shows that a switch in forward and reverse noncoding transcription from the Drosophila melanogaster vestigial (vg) PRE/TRE switches the status of the element between silencing (induced by the forward strand) and activation (induced by the reverse strand). In vitro, both noncoding RNAs inhibit PRC2 histone methyltransferase activity, but, in vivo, only the reverse strand binds PRC2. Overexpression of the reverse strand evicts PRC2 from chromatin and inhibits its enzymatic activity. It is proposed that the interaction of RNAs with PRC2 is differentially regulated in vivo, allowing regulated inhibition of local PRC2 activity. Genome-wide analysis shows that strand switching of noncoding RNAs occurs at several hundred Polycomb-binding sites in fly and vertebrate genomes. This work identifies a previously unreported and potentially widespread class of PRE/TREs that switch function by switching the direction of noncoding RNA transcription (Herzog, 2014).

By analysis of endogenous transcripts and by using ectopic overexpression strategies in vivo it was demonstrated that transcripts from opposite strands of the vg PRE/TRE have opposite effects on PRE/TRE status. However, in vitro, both noncoding RNAs have equivalent inhibitory effects on the HMTase activity of PRC2. Taking the in vitro and in vivo data together, it is proposed that the specificity of the noncoding RNA interaction with PcG proteins in vivo is not a result of inherently different affinities of PRC2 for different noncoding RNAs but of the availability of a given noncoding RNA (regulated by interactions of that RNA with other molecules) to interact with PRC2 and inhibit its enzymatic activity. It is proposed that the forward-strand noncoding RNA promotes silencing by facilitating pairing between PRE/TREs. PRE/TRE pairing has been shown to be essential for maximum silencing by the vg PRE/TRE, and this silencing is genetically dependent on PcG. Thus, it is proposed that forward strand-induced pairing might facilitate or stabilize PcG-mediated pairing-dependent silencing. E(Z) is detected at the vg PRE/TRE in ChIP analyses but does not interact with the forward strand. Thus, it is proposed that E(Z) binds at the silenced PRE/TRE independently of RNA. The forward-strand noncoding RNA might facilitate or stabilize pairing by binding to additional bridging proteins. These or other proteins might also prevent binding and inhibition of E(Z) by the RNA (Herzog, 2014).

Upon switching to the active state, transcription of the reverse PRE/TRE strand would be incompatible with forward-strand transcription because the reverse transcript runs through the forward-strand promoter. Reverse-strand transcription might thus destabilize pairing, enabling the activation of the PRE/TRE. In addition, the reverse strand binds E(Z) and, upon binding, would inhibit E(Z) HMTase activity, and it might also remove E(Z) from the locus. In this way, multiple self-reinforcing events could contribute to the stable switching of the PRE/TRE into an active state. The vg PRE/TRE responds to TrxG mutations by loss of activation. Whether TrxG-dependent activation acts via the noncoding RNAs will be a key question for future studies (Herzog, 2014).

Genome-wide analysis identifies many sites that might share functional features with the vgPRE/TRE. It is proposed that these elements can exist in a 'neutral' state, in which neither they nor their associated genes are transcribed (this is consistent with the observation that the vg PRE/TRE is not transcribed in tissues or at embryonic stages that do not express vg mRNA). However, in tissues that have the potential to transcribe the gene, the element might be switched either to the forward or reverse mode, thereby boosting either silencing or activation. This might serve to sharpen spatial expression boundaries, to stabilize gene expression states or to accelerate the kinetics of activation or repression (Herzog, 2014).

In conclusion, this work provides a new paradigm linking forward and reverse noncoding transcription to dynamic and developmentally regulated switching of PRE/TRE properties and thus to the maintenance of cell identities during development. Furthermore, the demonstration that any RNA is a potent inhibitor of PRC2 enzymatic activity in vitro but that only specific RNAs are able to bind and inhibit PRC2 in vivo strongly implies that specific RNAs are masked in vivo from interacting with PRC2. This provides an enormous potential for the regulated and reversible RNA-mediated inhibition of local PRC2 activity (Herzog, 2014).

Mbf1 ensures Polycomb silencing by protecting E(z) mRNA from degradation by Pacman

Under stress conditions, the coactivator Multiprotein bridging factor 1 (Mbf1) translocates from the cytoplasm into the nucleus to induce stress-response genes. However, its role in the cytoplasm, where it is mainly located, has remained elusive. This study shows that Drosophila Mbf1 associates with E(z) mRNA and protects it from degradation by the exoribonuclease Pacman (Pcm), thereby ensuring Polycomb silencing. In genetic studies, loss of mbf1 function enhanced a Polycomb phenotype in Polycomb group mutants, and was accompanied by a significant reduction in E(z) mRNA expression. Furthermore, a pcm mutation suppressed the Polycomb phenotype and restored the expression level of E(z) mRNA, while pcm overexpression exhibited the Polycomb phenotype in the mbf1 mutant but not in the wild-type background. In vitro, Mbf1 protected E(z) RNA from Pcm activity. These results suggest that Mbf1 buffers fluctuations in Pcm activity to maintain an E(z) mRNA expression level sufficient for Polycomb silencing (Nishioka, 2018).

Polycomb silencing is essential for the developmental regulation of gene expression. The silencing needs to be robust to tightly repress the expression of developmental genes in undifferentiated cells, such as stem cells, but should also be flexible for rapid release upon differentiation. However, this paradoxical aspect of Polycomb silencing is not well understood (Nishioka, 2018).

Mbf1 was originally identified as an evolutionarily conserved coactivator that connects a transcriptional activator with the TATA element-binding protein (Li, 1994; Takemaru, 1997; Takemaru, 1998). Usually, Mbf1 is present in the cytoplasm; however, under stress conditions, Mbf1 translocates into the nucleus to induce stress-response genes. Previous studies have revealed roles for the coactivator in axon guidance, oxidative stress response, defense against microbial infection, and resistance to drugs such as tamoxifen. However, the cytoplasmic role of Mbf1 has remained elusive, except for mRNA or ribosomal binding (Nishioka, 2018).

Pacman (Pcm/Xrn1) is an evolutionarily conserved 5'-3' exoribonuclease that degrades decapped mRNA (Till, 1998; Jones, 2012). Genetic studies have demonstrated that Drosophila pcm is involved in epithelial closure, male fertility, apoptosis and growth control (Grima, 2008; Lim, 2009; Jones, 2012; Jones 2016; Waldron, 2015). Null mutants of pcm are lethal during early pupal stages, suggesting the enzyme plays an essential role in development (Waldron, 2015; Jones, 2016; Nishioka, 2018 and references therein).

Using a genetic approach in Drosophila, this study shows that cytoplasmic Mbf1 ensures Polycomb silencing by protecting E(z) mRNA from degradation by Pcm. The results thus demonstrate an unexpected component of the regulatory mechanism underlying Polycomb silencing. This mechanism might also allow flexibility in Polycomb silencing, as Mbf1 protein expression declines upon differentiation (Nishioka, 2018).

To address the cytoplasmic role of Mbf1, novel genes were sought that interact with mbf1. Surprisingly, the mbf1 mutation enhanced a classical Polycomb phenotype of Psc and Pc mutants, namely the appearance of an ectopic sex comb tooth or teeth on the male mid-leg. Although mbf1²/+ or mbf1²/mbf1² flies never exhibited the Polycomb phenotype, penetrance of the phenotype in Psc¹/+ increased significantly in Psc¹/+; mbf1²/+, and further increased in Psc¹/+; mbf1²/mbf1². The penetrance was restored to the Psc¹/+ level by expressing wild-type Mbf1 protein from a transgene. Similar effects of the mbf1² allele were observed with the Pc⁶ mutation (Nishioka, 2018).

To gain insight into the mechanism underlying the genetic interaction between Psc and mbf1, the expression of the representative Polycomb group genes Pc, E(z) and pho was analyzed. Results of reverse transcription-quantitative PCR (RT-qPCR) analyses demonstrated a prominent reduction in the expression level of E(z) mRNA in Psc¹/+; mbf1²/+ larvae, whereas Pc and pho mRNA levels remained unchanged. Immunostaining of wing discs demonstrated that E(z) protein expression was severely compromised in Psc¹/+; mbf1²/+ compared with that in wild type, mbf1²/+ or Psc¹/+. By contrast, the expression of Pc and Pho proteins was not significantly affected. Western blot analyses confirmed the marked decrease in the E(z) protein level in both wing and leg discs from Psc¹/+; mbf1²/+. Consistently, Psc¹/+; E(z)⁷³¹/+ exhibited the extra sex comb phenotype, which was comparable to Psc¹/+; mbf1²/+ (Nishioka, 2018).

It is unlikely that Mbf1 affects E(z) transcription because no significant difference was detected in the E(z) mRNA level between wild-type and mbf1²/mbf1² larvae. Consistently, it was not possible to detect any significant difference in the expression of E(z) in the wing disc upon knockdown or overexpression of Mbf1 using a posterior compartment-specific Gal4 driver. When cytoplasmic and nuclear RNA fractions from wing discs were analyzed by RT-qPCR, the nuclear E(z) mRNA level was similar between wild type and Psc¹/+; mbf1²/+. However, the cytoplasmic E(z) mRNA level in Psc¹/+; mbf1²/+ decreased to ~20% of the wild-type level. Collectively, these results suggest that mbf1 regulates the E(z) mRNA level post-transcriptionally in the cytoplasm (Nishioka, 2018).

Considering that Mbf1 binds to mRNA, it was hypothesized that cytoplasmic Mbf1 might bind to E(z) mRNA to protect it from degradation, and thereby regulates the E(z) mRNA level. Results of RNA-immunoprecipitation (RIP) experiments revealed a preferential binding of Mbf1 to E(z) mRNA. A ~10-fold enrichment of E(z) mRNA was found in the anti-Mbf1 antibody pull-down fraction from cytoplasmic extracts of embryos. The pull-down was clearly selective, as enrichment of abundant mRNAs, such as RpL32 and RpL30, was not observed. By contrast, E(z) mRNA was barely detectable in the anti-Mbf1 antibody pull-down fraction from embryonic extracts of the mbf1 mutant, used as a negative control. This is not due to absence of E(z) mRNA in the mbf1 mutant (Nishioka, 2018).

Following the observed preferential binding of Mbf1 to E(z) mRNA, this study focused on the Polycomb phenotype and reduced E(z) mRNA expression level, which were not caused by the mbf1 mutation alone. Enhancement of the Polycomb phenotype and the reduction of E(z) mRNA were only detected in the double mbf1 and Polycomb group gene mutant. To explain the synergistic effect of mbf1 and Polycomb group mutations, it was posited that a component of the mRNA degradation pathway was only activated in the Polycomb group mutant background. Therefore, attempts were made to identify the component of the pathway that was activated in the Psc or Pc mutants. Among the mRNAs tested, only pcm mRNA, which encodes the 5'-exoribonuclease, was upregulated in Psc¹/+ and Pc⁶/+ larvae. Neither the decapping enzyme (Dcp2), components of the exosome [Dis3, Prp6 (CG6841) and Prp40 (CG3542)], nor components in the 3'-deadenylation-mediated pathway (twin and Nab2) appeared to be activated. Western blot analyses revealed a 2-fold increase in the Pcm protein level in wing discs from Psc¹/+ or Pc⁶/+ larvae compared with that from wild type. These results led to an investigation of the effects of the pcm mutation on Polycomb silencing and E(z) mRNA expression (Nishioka, 2018).

Strikingly, the pcm^Δ1 mutation resulted in significant suppression of the Polycomb phenotype in Psc¹/+ and Psc¹/+; mbf1²/+. This suppression was rescued by expressing the wild-type Pcm protein from a transgene. Similar results were obtained using the Pc⁶ mutant. Consistent with this result, the pcm^Δ1 mutation restored the E(z) mRNA levels in Psc¹/+ and Psc¹/+; mbf1²/+ to near wild-type levels (Nishioka, 2018).

In addition to the extra sex comb phenotype, Psc¹/+; mbf1²/+ exhibited misexpression of Ubx in wing discs. The signals appeared as spots consisting of clusters of Ubx-positive cells. The pcm^Δ1 mutation decreased the number of spots per wing disc. The misexpression occurred predominantly around the dorsoventral border in the posterior compartment. Consistently, adult wing defects were observed along the posterior wing margin, which was also suppressed by pcm^Δ1 (Nishioka, 2018).

Importantly, the extra sex comb phenotype was detected under mild overexpression of pcm in mbf1²/hs-pcm double heterozygotes at 25°C, even in the wild-type Polycomb group background. hs-pcm/+ exhibited an ~2.5-fold overexpression of Pcm at 25°C. Nevertheless, hs-pcm heterozygotes in the wild-type mbf1 background did not show any Polycomb phenotype. These results suggest that Mbf1 stabilizes Polycomb silencing against fluctuations in the Pcm protein level in vivo. Enhancement of the Polycomb phenotype was also observed in Psc¹/+; hs-pcm/+ compared with that in Psc¹/+ (Nishioka, 2018).

Biochemical analyses using purified recombinant Mbf1 and Pcm proteins revealed that Mbf1 protects E(z) RNA from degradation by Pcm. RNA protection assays were performed in which in vitro-synthesized E(z) RNA was treated with the RNA pyrophosphatase RppH to convert the 5'-triphosphoryl end into the 5'-monophosphoryl form, which is a Pcm substrate. The RNA was digested with Pcm in the presence or absence of Mbf1. Mbf1 inhibited the digestion of E(z) RNA. In the absence of RppH, RNA degradation was barely detectable, suggesting that the digestion was due to 5'-exoribonuclease activity. Gel filtration of a mixture of Pcm and Mbf1 resulted in the elution of each protein in a clearly separated peak. Furthermore, Mbf1 did not co-immunoprecipitate with Pcm and vice versa. These results suggest that Mbf1 does not inhibit Pcm activity through protein-protein interactions. Collectively, it is concluded that Mbf1 protects E(z) mRNA from degradation by Pcm both in vivo and in vitro (Nishioka, 2018).

It is proposed that cytoplasmic Mbf1 ensures Polycomb silencing by protecting E(z) mRNA from the activity of Pcm. In the mbf1 mutant, E(z) mRNA is free from Mbf1 protein, but pcm expression is downregulated by Polycomb group genes. In the Polycomb group mutant, Pcm expression is upregulated, but E(z) mRNA is partly protected by Mbf1. In the mbf1 Polycomb group double mutant, E(z) mRNA is free from Mbf1 protein and is subject to Pcm attack. Whereas Mbf1 is highly expressed in undifferentiated cells, such as those of embryos, larval testis, ovary, imaginal discs and neuroblasts, its expression is reduced in differentiated tissues, similar to the situation in the mbf1 mutant. This would facilitate the rapid release of developmental genes from Polycomb silencing upon differentiation. Interestingly, expression of mammalian Mbf1 [also termed endothelial differentiation-related factor 1 (Edf1)] and Ezh2 declines immediately after the onset of differentiation (Nishioka, 2018).

A recent study demonstrated that Pcm prevents apoptosis in imaginal discs and downregulates specific transcripts such as hid and reaper (Waldron, 2015). However, suppression of apoptosis did not rescue the lethality of a pcm null mutation at the early pupal stage. Therefore, there might be other targets of Pcm that are essential for early pupal development. The present study indicates that E(z) mRNA could be one such target (Nishioka, 2018).

The mRNA-binding activity of Mbf1 was selective, but might not be strictly specific to E(z) mRNA. Although Polycomb silencing is central to the developmental regulation of gene expression, there could be other mRNAs that bind to Mbf1 in a similar manner, thereby modulating another biological function. Therefore, RIP-seq analysis was conducted to identify Mbf1-bound mRNAs. To ensure robustness of the RIP-seq data, the results were compared independently with two publically available datasets and identified 804 commonly enriched mRNAs. Among these, the enrichment of four representative mRNAs (GstD5, Ide, Tep2 and Pebp1) was confirmed by RIP RT-qPCR analyses. Interestingly, the expression levels of these four mRNAs decreased in Psc¹/+; mbf1²/+ and increased in pcm^Δ1/Y compared with those in wild type, suggesting that the model can be applied to a wider range of mRNAs than just E(z). However, dependency on the Mbf1/Pcm antagonism appears to differ among the mRNAs (Nishioka, 2018).

Gene ontology and pathway analyses of the 804 genes revealed some interesting properties of the Mbf1-associated mRNAs. The gene ontology terms 'glutathione metabolic process', 'oxidation-reduction process' and 'neurogenesis' which includes E(z), are consistent with the fact that previous studies found defects in oxidative stress defense and axon guidance in the mbf1 mutant (Liu, 2003; Jindra, 2004). Also of interest are the groups 'positive regulation of innate immune response' and 'defense response to Gram-negative bacterium', as Arabidopsis MBF1 is involved in host defense against microbial infection. Moreover, pathway analysis of the enriched genes implicated Mbf1 in 'drug metabolism', as previously suggested for tamoxifen resistance. This raises an intriguing possibility that Mbf1 contributes to various types of stress defense, metabolic processes and neurogenesis as both a nuclear coactivator and as a cytoplasmic mRNA-stabilizing protein. Although mbf1 null mutants are viable under laboratory conditions, evolutionary conservation of mbf1 suggests that it has essential role(s) under real-world stress conditions (Nishioka, 2018).

RNA-DNA strand exchange by the Drosophila Polycomb complex PRC2

Polycomb Group (PcG) proteins form memory of transient transcriptional repression that is necessary for development. In Drosophila, DNA elements termed Polycomb Response Elements (PREs) recruit PcG proteins. How PcG activities are targeted to PREs to maintain repressed states only in appropriate developmental contexts has been difficult to elucidate. PcG complexes modify chromatin, but also interact with both RNA and DNA, and RNA is implicated in PcG targeting and function. This study shows that R-loops, three-stranded nucleic acid structures formed when an RNA hybridizes to a complementary DNA strand, thereby displacing the second DNA strand, form at many PREs in Drosophila embryos, and correlate with repressive states. In vitro, both PRC1 and PRC2 can recognize R-loops and open DNA bubbles. Unexpectedly, this study found that PRC2 [E(z), Esc and Su(z)12] drives formation of RNA-DNA hybrids, the key component of R-loops, from RNA and dsDNA. These results identify R-loop formation as a feature of Drosophila PREs that can be recognized by PcG complexes, and RNA-DNA strand exchange as a PRC2 activity that could contribute to R-loop formation (Alecki, 2020).

During Drosophila embryogenesis, transiently expressed transcription factors activate homeotic (Hox) genes in certain regions of the embryo and repress them in others to dictate the future body plan. Polycomb Group (PcG) proteins form a memory of these early cues by maintaining patterns of Hox gene repression for the rest of development. This paradigm for transcriptional memory is believed to be used by the PcG at many genes in Drosophila, and to underlie the conserved and essential functions of PcG proteins in cell differentiation and development from plants to mammals. Polycomb response elements (PREs) are DNA elements that can recruit PcG proteins, but they also recapitulate the memory function of the PcG-when combined with early acting, region-specific enhancers in transgenes, they maintain transgene repression in a PcG-dependent manner only in regions where the early enhancer was not active. PREs contain a high density of binding sites for transcription factors that can recruit PcG proteins through physical interactions. However, the widespread expression, binding pattern, and properties of factors that bind PREs cannot explain how PREs can exist in alternate, transcription-history dependent states to maintain restricted patterns of gene expression, or how they can switch between states. Furthermore, DNA sequences with PRE-like properties have been difficult to identify in other species despite the conservation of PcG complexes, their biochemical activities, and their critical roles in development (Alecki, 2020).

RNAs may provide context specificity to PcG protein recruitment and function. Some PREs, and some PcG-binding sites in mammalian and plant cells, are transcribed into ncRNA, while others reside in gene bodies, and thus are transcribed when the gene is expressed (Herzog, 2014). Both the direction and level of transcription have been correlated with the functional state of PREs. The PcG complex Polycomb Repressive Complex 2 (PRC2) has a well-described high affinity for RNA. RNA is suggested to recruit PRC2 to specific chromatin sites13, but RNA binding can also compete for chromatin binding and inhibit PRC2 activity. One way for RNA to interact with the genome is by the formation of R-loops, three-stranded nucleic acid structures formed when an RNA hybridizes to a complementary DNA strand, thereby displacing the second DNA strand. R-loops have been linked to regulation of transcription and chromatin previously, through a variety of mechanisms. This includes links to PcG regulation in mammalian cells. The formation of R-loops over genes with low to moderate expression is associated with increased PcG binding and H3K27 trimethylation (H3K27me3) in human cells and R-loops have recently been implicated in promoting PRC1 and PRC2 recruitment in mammalian cells, although other evidence suggests they antagonize recruitment of PRC2. It is hypothesized that R-loop formation could biochemically link RNA to PcG-mediated silencing through PREs and tested this idea in the Drosophila system (Alecki, 2020).

This study identified R-loop forming sequencing in Drosophila embryos and S2 cells and observe that ~25% of PREs form R-loops. Interestingly, PREs that form R-loops are more likely to be bound by PcG proteins compared with PREs that do not form R-loops, suggesting that R-loops may be involved in PcG targeting. In vitro, PRC1 and PRC2 recognize R-loops and open DNA-bubbles. Further, when provided dsDNA and RNA, PRC2 induces the formation of RNA-DNA hybrids, the key components of R-loops. These data suggest a mechanism for RNA to contribute to targeting of PcG proteins via R-loop formation induced by the RNA-DNA strand exchange activity of PRC2 (Alecki, 2020).

The demonstration that PRC2 induces the formation of RNA-DNA hybrids in vitro, that PRC2 and PRC1 recognize R-loops in vitro, and that R-loops are present at PREs in vivo suggest a mechanistic model for how RNAs could induce or maintain the OFF state of PREs. If PREs (or the gene they control and in many cases are embedded in) are highly transcribed, the RNA could compete for PRC2 binding to chromatin, as has been demonstrated in vitro and in vivo. However, a lower level of transcription through a PRE (or transcription in an orientation that is favourable for R-loop formation) could allow R-loops to form, possibly via the RNA-DNA hybrid forming activity of PRC2. R-loop formation will repress additional RNA production by preventing RNA polymerase passage allowing recruitment of additional PRC2 (by PRE-binding transcription factors or interactions with other PcG proteins) and its retention on chromatin. PRC2 could then modify histones to maintain a repressive chromatin state. The R-loop, in conjunction with H3K27me3 and PRE-binding transcription factors, would also promote binding of PRC2 and PRC1. R-loops may also interfere with binding or function of proteins that promote the active state of PREs, although this remains to be tested. The data indicate that both coding and ncRNAs form R-loops. The regulation of these RNAs and therefore of R-loops could provide transcriptional memory and developmental context specificity to PcG recruitment by transcription factors that constitutively recognize PREs. A conceptually similar model for how high levels of RNA production at PREs could promote the ON state and low levels the OFF state has been proposed; R-loop formation provides one mechanism by which it can occur. Although this model is highly speculative at this time, it integrates many observations, and provides testable hypotheses (Alecki, 2020).

Observations in Drosophila are also consistent with a possible connection between R-loops and PcG function. The helicase Rm62 interacts genetically with both PcG and TrxG genes, and colocalizes with the PRE-binding protein Dsp1 on polytene chromosomes. Rm62 is the Drosophila homologue of the DDX5 helicase, which can unwind RNA-DNA hybrids in vitro and is implicated in R-loop resolution in vivo. A recent genome-wide RNAi screen for TrxG interacting genes (which should antagonize PcG function) identified the gene for RNaseH140. RNA has been suggested to be important in switching PREs between OFF and ON states, although this has been contested by experiments aiming to test whether transcription through a PRE can switch it to the active state. Resolution of R-loops by cellular RNases or RNA-DNA helicases could contribute to switching PRE states, which will be intriguing to test. It is also likely that even in the simple model suggested in the paper (see Model for the role of R-loop formation driven by PRC2 in PcG gene silencing), the levels of RNA corresponding to 'low' and 'high', and the strength of the effect will depend both on the genomic context and the sequences of the RNAs that are produced (Alecki, 2020).

R-loop formation is observed at ~30% of PREs; these may represent a specific class of PREs. Most R-loops are believed to form co-transcriptionally, so that R-loops would be predicted to depend on PRE transcription. Indeed, >70% of R-loops formed at PREs overlap an annotated coding or non-coding RNA, and PREs with R-loops are more likely to have RNA Pol II signal in ChIP-seq experiments. However, ~67% of PREs where R-loops were not observed also overlap an annotated transcript. Further, a fraction of PREs with R-loops (and a fraction of total R-loops) either do not overlap any annotated transcripts, or overlap a transcript in the opposite orientation as the R-loop. While some of these discrepancies likely reflect incomplete annotation of rare transcripts, they raise the intriguing possibility that the RNA used to form the R-loops could be supplied in trans. Careful analysis of the RNA component of R-loops at PREs will be needed to resolve this. Although speculative at this time, the ability of PRC2 to induce RNA-DNA hybrids could contribute to non-co-transcriptional R-loop formation (Alecki, 2020).

This study finds that PRC2 can induce RNA-DNA strand exchange from RNA and linear dsDNA in vitro. A small number of other proteins have been shown to have similar activity, using various types of substrates. These include the repair proteins Rad52/RecA and PALB2, the human capping enzyme (CE)50, the viral protein ICP8 and the telomere-inding protein TRF2. Like the activity of PRC2, none of these reactions require ATP hydrolysis (although R-loop formation by RecA is stimulated by ATPγS), and most use linear DNA substrates or an unpaired or ssDNA region. The exceptions are TRF2 and ICP8. ICP8 can mediate R-loop formation from an RNA and a supercoiled plasmid. TRF2 stimulates invasion of RNA oligos into a supercoiled plasmid encoding a telomeric DNA array, but the mechanism is believed to be induction of positive supercoiling by TRF2 that facilitates DNA unwinding and RNA invasion. RNA-DNA strand exchange has been investigated most closely for Rad52, and its homologue RecA. Rad52 has been shown both to carry out 'inverse strand exchange' where Rad52 first binds the dsDNA, allowing RNA strand exchange, and to use an RNA-bridging mechanism, in which Rad52 first binds the RNA, and can bridge two dsDNA fragments by forming RNA-DNA hybrids with segments of each of them. Both of these mechanisms are candidates to mediate RNA-mediated repair of DSBs. PRC2 requires a DNA end for RNA-DNA strand exchange in vitro; for this activity to occur in vivo, either a DNA break would be required, or PRC2 would need to be able to use DNA opened by (an)other factors, or by transcription. These requirements may limit PRC2 strand exchange activity at PREs. In order to fully understand the impact of this activity in vivo and to what extent PRC2 contributes to R-loop formation at PREs, additional experiments will be necessary. Interestingly, Topoisomerase II interacts with a subunit of PRC1, colocalizes with PcG proteins in the BX-C, and is implicated in PRE-mediated silencing; transient Topo II induced breaks have been implicated in regulation of transcription and chromatin compaction, and could also be used by PRC2. It is also possible that the activity of PRC2 contributes to RNA-DNA strand exchange at DNA breaks where RNA-DNA hybrids have been shown to form and where PRC2 is recruited (Alecki, 2020).

The connection between RNA and PRC2 has been recognized for some time, in species from plants to humans, but mechanisms beyond RNA binding by PRC2 have not previously been described. This discovery of PRC2-mediated RNA-DNA strand exchange, suggests one mechanism to connect RNA to PcG targeting and function (Alecki, 2020).

GENE STRUCTURE

Exons - seven

Bases in 3' UTR - 66

PROTEIN STRUCTURE

Amino Acids - 760

Structural Domains and Evolutionary Homologies

Molecular analysis of the Enhancer of zeste gene predicts a 760-amino-acid protein product. A region of 116 amino acids near the E(z) carboxy terminus is 41.2% identical (68.4% similar) to a carboxy-terminal region of the Trithorax protein. This portion of the Trithorax protein is part of a larger region previously shown to share extensive homology with a human protein (ALL-1/Hrx) implicated in acute leukemias. Over this same 116 amino acids, E(z) and ALL-1/Hrx are 43.9% identical (68.4% similar). Otherwise, E(z) is not significantly similar to any previously described proteins. Since this region of sequence similarity is shared by two proteins with antagonistic functions, it may comprise a domain that interacts with a common target, either nucleic acid or protein. Opposite effects on transcription might then be determined by other portions of the two proteins (Jones, 1993).

The C-terminal domain of Trx has been termed a tromodomain; the more generally accepted term for this is a SET domain (for Su(var)39, E(z) and trx, the three proteins in which the domain was first identified). With regard to the antagonistic functions between Trx and E(z), some E(z) mutants enhance the trx phenotypes rather than repressing them (Manuel Diaz, personal communication to the editor of The Interactive Fly).

date revised: 5 August 2023

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

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