In Drosophila, the Polycomb group (PcG) of genes is required for the maintenance of homeotic gene repression during development. The Drosophila ortholog of the products of the mammalian Ring1/Ring1A and Rnf2/Ring1B genes has been characterized. Drosophila Ring corresponds to Sex combs extra (Fritsch, 2003: Gorfinkiel, 2004), a previously described PcG gene. Ring/Sce is expressed and required throughout development and the extreme Pc embryonic phenotype due to the lack of maternal and zygotic Sce can be rescued by ectopic expression of Ring/Sce. This phenotypic rescue is also obtained by ectopic expression of the murine Ring1/Ring1A, suggesting a functional conservation of the proteins during evolution. In addition, Ring/Sce binds to about 100 sites on polytene chromosomes, 70% of which overlap those of other PcG products such as Polycomb, Posterior sex combs and Polyhomeotic, and 30% of which are unique. Ring/Sce interacts directly with PcG proteins, because it occurs in mammals (Gorfinkiel, 2004).

Drosophila is part of a protein complex that monoubiquitinates nucleosomal histone H2A. Reducing the expression of mammalian Ring2 results in a dramatic decrease in the level of ubiquitinated H2A in HeLa cells. Chromatin immunoprecipitation analysis has demonstrated colocalization of Drosophila Ring with ubiquitinated H2A at the polycomb response elements and promoter regions of the Drosophila Ubx gene in wing imaginal discs. Removal of Drosophila Ring in SL2 tissue culture cells by RNA interference results in loss of H2A ubiquitination concomitant with derepression of Ubx. These studies identify the H2A ubiquitin ligase, and link H2A ubiquitination to Polycomb silencing (Wang, 2004).

Genetic analysis in Drosophila has unveiled a repression function required for proper regulation of the homeotic genes that determine segmental identities. A large number of genes, collectively known as the Polycomb group of genes (PcG), participate in such a repressive activity. Thus, mutations in the PcG genes lead to homeotic phenotype associated with the indiscriminate expression of genes from the Bithorax complex (BX-C) and/or Anntenapedia complex (ANT-C). PcG related genes have been identified in plants and in vertebrates, and mutations in these genes are, among others, associated with homeotic phenotypes. The PcG are thought to be required for the maintenance of transcriptionally repressed states of the Hox genes, but not for the initiation of their repression. Other transcriptional repressors of the gap and pair rule groups, transiently expressed during development, are responsible for this initiation of repression (Gorfinkiel, 2004).

The molecular mechanism(s) of PcG function is (are) unknown. Several lines of evidence, however, indicate that PcG products work together in multimeric protein complexes in which individual PcG proteins interact with other PcG proteins through conserved domains. Biochemical fractionation of Drosophila nuclear extracts shows two major multimeric complexes. One, termed Polycomb Repressive Complex 1 (PRC1) has a size of about 2 MDa, contains the PcG products Polycomb (Pc), Polyhomeotic (Ph), Posterior sex combs (Psc), Sex combs on midleg (Scm) and Drosophila Ring, some components of the basal transcriptional machinery (TAFs) and other polypeptides (Shao, 1999 and Saurin, 2001). Another complex, of about 600 kDa in size, does not contain any of the above proteins, but instead comprises the products of the extra sex combs (esc), Enhancer of zeste [E(z)] and Suppressor of zeste 12 [Su(Z)12] genes. In contrast to the lack of enzymatic activities associated with the PRC1 complex, the so-called Esc-E(z) complex has histone deacetylase and histone methyltransferase activities. A complex-based function is consistent with the synergistic genetic interactions between any two PcG genes. Additionally, the PcG products are chromosomal proteins that bind specific sites, visualized on salivary gland polytene chromosomes. Many of these binding sites are common for several PcG proteins. The large number of chromosomal sites that bind PcG proteins suggests that the homeotic complexes, BX-C and ANT-C, are only some of many target loci regulated by PcG (Gorfinkiel, 2004).

Repression by PcG proteins occurs through Polycomb response elements (PRE), which are regulatory DNA sequences harbouring functional binding sites for PcG proteins. Until recently, PREs were identified in a few loci, including the homeotic genes of the BX-C and ANT-C complexes. Recently, computational methods have been used in Drosophila to predict PREs on a genome wide scale identifying about 170 candidate PREs, which map to a variety of loci involved in development and cell proliferation. PREs have a modular structure and bind PcG complexes of different composition. How these complexes are targeted to DNA is not known. PREs have DNA binding sites for proteins such as GAGA factor, Zeste and Pleiohomeotic (Pho), which is the only PcG product able to bind DNA. However, Pho is found only in PcG complexes at the earliest stages of Drosophila development. The molecular mechanism(s) by which the PcG repression function uses multimeric complexes is not known (Gorfinkiel, 2004).

In a search for new mammalian PcG genes, Ring1/Ring1A and Rnf2/Ring1B, two mouse genes have been found whose products interact both in vitro and in two hybrid assays with Pc, Psc and Ph homologs (Hemenway, 1998; Satijn, 1999; Satijn, 1997; Schoorlemmer, 1997). Ring1/Ring1A and Rnf2/Ring1B proteins are part of a PRC1 complex isolated from mammalian cells (Levine, 2002). The Drosophila PRC1 complex also contains the ortholog of vertebrate Ring1 proteins, which seems to play an essential role in the in vitro reconstitution of a PRC1 core complex together with Pc, Psc and Ph (Francis, 2001). In mice, null or hypomorphic mutations in the Ring1/Ring1A or Rnf2/Ring1B genes, respectively, show axial skeleton alterations consistent with a PcG function (del Mar Lorente, 2000; Suzuki, 2002; Gorfinkiel, 2004 and references therein).

The product of the Drosophila melanogaster Ring gene (Ring) has been identified as Sex combs extra (Fritsch, 2003; Gorfinkiel, 2004), one of the molecularly uncharacterized PcG mutants in Drosophila. Over-expression of Ring/Sce and also of the murine Ring1/Ring1A can rescue the extreme Pc embryonic phenotype derived from the lack of maternal and zygotic Sce¹, suggesting a functional conservation of the Drosophila and vertebrate proteins during evolution. In addition, Ring/Sce encodes a chromosomal protein that binds to more than 100 specific sites. Direct interactions between Ring/Sce and PcG proteins take place through the same domains as the interactions between their mammalian counterparts (Gorfinkiel, 2004).

The EST databases of the BDGP were searched with either murine Ring/Ring1A or Rnf2/Ring1B cDNAs and two overlapping cDNAs (LD3177 and LD6636) were identified. The complete sequence of cDNA LD3177 was almost identical to a cDNA sequence termed Ring deposited in the databanks (CG5595) (Gorfinkiel, 2004).

By in situ hybridization to polytene chromosomes, Drosophila Ring was located at the end of the long arm of chromosome 3 in section 98A. Interestingly, Sce, a non-molecularly characterized PcG gene, which was defined by a single mutant allele Sce¹, had been mapped by recombination to the 3-92 interval (Breen, 1986). The proximity of such an interval to the cytological localization of the Drosophila Ring gene encouraged the exploration of a possible identity between the Sce and the Drosophila Ring gene (Gorfinkiel, 2004).

The genomic DNA was sequenced from Sce¹ heterozygous embryos corresponding to the Drosophila Ring coding region. Comparing these sequences with the wild type, a deletion of 410 bp was found that removed the codons for the C-terminal 113 amino acids, a small intron and 12 nucleotides of the 3' untranslated region after the termination codon. Therefore, the Sce¹ allele conceptually encodes a truncated Ring protein that is fused in frame to 23 novel amino acids at the C-terminal part of the protein. Fritsch (2003) has made an identical observation. Drosophila Ring will therefore be referred to as Sce (Gorfinkiel, 2004).

In agreement with the presence of Ring in embryonic PcG complexes the data support a PcG function for the Ring protein. Mice bearing null (Ring1/Ring1A) or hypomorphic (Rnf2/Ring1B) mutations show an involvement of the Ring genes in the patterning of the antero-posterior axis (del Mar Lorente, 2000; Suzuki, 2002). However, in contrast to mutations in other vertebrate PcG genes, the alterations of the axial skeleton seen in the Ring mutant mice could not be associated clearly to a deregulation of Hox genes (del Mar Lorente, 2000). Therefore, the role of vertebrate Ring proteins as genuine PcG proteins is strengthened by the data showing a genetic evidence for a PcG function for Sce (Gorfinkiel, 2004).

Cross-species complementation experiments with PcG genes show contrasting results. Thus, M33, the mouse ortholog of Drosophila Pc, was shown to rescue the Pc mutant phenotype in early embryos. However, eed, the mouse ortholog of Drosophila extra sex combs (esc) is not only unable to rescue the embryonic lethality of esc embryos but shows a dominant negative effect on the leg transformation phenotype of esc mutants. It has been suggested that the activity of eed in Drosophila cells is related to its inability to interact with E(z). With respect to Ring, mouse Ring1/Ring1A rescues the cuticle phenotype of Sce embryos, indicating that in early development, at least, the function of Ring is conserved between mice and flies. This might be due to the structural conservation of Ring proteins. The three domains conserved in Ring1/Ring1A and Rnf2/Ring1B are also present in Sce and constitute about 57% of this protein. Whereas the size and degree of conservation of the domains HD2 and HD3 are similar to other protein motifs identified in fly and vertebrate PcG proteins, domain HD1 is somewhat exceptional. This is a 147 amino acids domain in which 78% are identical in fly and vertebrate proteins, particularly in the RING finger motif. An indication of the relevance of the functionality of this region of Ring proteins is the Sce^33M2 allele which shows a phenotype much milder than that of Sce¹ but that is due to a Ring protein with a single amino acid alteration in that region (Fritsch, 2003). The overall structural conservation between Ring proteins seems to dictate a conservation of interaction with other PcG proteins. In addition, Sce interacts with Pc and Psc. In fact, the core of a PRC1 complex isolated from human cells is compositionally similar to that of flies and the biochemical activity of both complexes is similar (Gorfinkiel, 2004).

Despite this conservation, it is possible that Sce serve diverse functions in late development. For example, expression of the mouse M33 protein in flies does not rescue the Pc adult phenotype. Experiments have not addressed the activity of vertebrate Ring proteins at these later developmental stages and, therefore, whether vertebrate Ring proteins can fully substitute for Sce needs to be approached experimentally (Gorfinkiel, 2004).

Previous genetic and biochemical evidence has shown that PcG proteins act as protein complex(es). Sce interacts directly with Pc and Psc, but not with a Ph-fragment, which binds mouse Rnf2/Ring1B. In addition, immunolocalization studies show that Sce binds to approximately 100 sites, which are in part shared by Pc, Psc, Ph, Pcl and Asx binding site, including the ANT-C and BX-C complexes. These results are consistent with the presence of Sce in the PRC1 complex. However, almost a third of the sites that bind Sce do not bind any of the other PcG proteins. This contrasts with the observation that most Sce molecules in cell extracts are found complexed with PcG proteins in the PRC1 complex (Saurin, 2001). The discrepancy, however, may be related to the fact that the characterized PRC1 has been isolated from Drosophila embryos, whereas the Sce chromosomal sites correspond to binding sites in salivary glands from larvae. Psc, another component of the PRC1 complex, is also found in sites, which do not have Pc/Ph/Pcl. It is worth noting that, despite the ability of Sce to interact with Psc, no Sce is found at these unique Psc sites. Nevertheless, some of these sites correspond with predicted PREs. Therefore, the partial overlapping patterns of Sce and other PcG binding sites suggest the existence of different Polycomb complexes in a tissue specific and developmentally controlled manner. An indication of complexes containing subsets of PcG proteins comes from studies in vertebrates where Drosophila Ring proteins are found together with other polypeptides but not Pc or Ph homologs (Gorfinkiel, 2004).

An intriguing result of the studies on the chromosomal binding sites of Sce is that, in contrast to all PcG genes so far studied, the cytological localization of the Sce gene is free of any PcG protein. The absence of PcG proteins at 98A, therefore, suggests that Sce is regulated somehow differently from other PcG loci (Gorfinkiel, 2004).

In summary, the PcG gene Sce encodes the Drosophila ortholog of mammalian Ring proteins. Sce gene binds to Pc and Psc and is a chromosomal protein associated with many sites in polytene chromosomes, which also bind PcG proteins. Finally, Sce is expressed and required throughout development. The extreme Pc phenotype of Sce embryos is rescued by ectopic expression of Drosophila Ring/Sce and Ring1/Ring1A suggesting that the function of these proteins in conserved between flies and mammals, at least in the early stages of fly development (Gorfinkiel, 2004).

GENE STRUCTURE

cDNA clone length - 1457 bp

Exons - 2

Bases in 3' UTR - 137

PROTEIN STRUCTURE

Amino Acids - 435

Structural Domains

Studies on the mammalian RING1 proteins suggest that the C-terminal domain of RING1B interacts with the C-terminal repressor domain of M33, the mammalian homologue of Drosophila Pc (Schoorlemmer, 1997). Similarly, in yeast two-hybrid assays, the RING finger of RING1B has been found to interact with the RING finger of Bmi-1, the orthologue of Psc in the mammalian PRC1 complex (Hemenway, 1998; Satijn, 1999; Levine, 2002). Hence, it appears that RING1 proteins contain two functional domains, an N-terminal RING finger domain that interacts with RING finger proteins such as Bmi-1 and a C-terminal domain that interacts with M33/Pc (Fritsch, 2003).

The comparison between the fly and murine Ring proteins revealed a high degree of conservation. Thus, the three domains (HD1, HD2 and HD3) identified in the murine (and human) proteins are also identified in the fly protein. These domains are separated, as in the mammalian proteins, by non-conserved sequences. Therefore, 78% of the 147 amino acids N-terminal domain (HD1), which contains a Ring finger, are identical between the fly and either of the murine Ring1 proteins. Conservation at the other two domains is lower: 53% and 60% identity with HD2 of Ring1/Ring1A and Rnf2/Ring1B, respectively, and 46% identity between Drosophila Ring HD3 and either HD3 of the murine Ring1 proteins. Curiously, the HD2 of Drosophila Ring is interrupted by a stretch of 11 amino acids (Gorfinkiel, 2004).

date revised: 30 June 2005

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.