InteractiveFly: GeneBrief

corto: Biological Overview | References

Polycomb (PcG) and trithorax (trxG) genes encode proteins involved in the maintenance of gene expression patterns, notably Hox genes, throughout development. PcG proteins are required for long-term gene repression whereas TrxG proteins are positive regulators that counteract PcG action. PcG and TrxG proteins form large complexes that bind chromatin at overlapping sites called Polycomb and Trithorax Response Elements (PRE/TRE). A third class of proteins, so-called ìEnhancers of Trithorax and Polycombî (ETP), interacts with either complexes, behaving sometimes as repressors and sometimes as activators. The role of ETP proteins is largely unknown. In a two-hybrid screen, Cyclin G (CycG) was identified as a partner of the Drosophila ETP Corto. Inactivation of CycG by RNA interference highlights its essential role during development. Corto and CycG directly interact and bind to each other in embryos and S2 cells. Moreover, CycG is targeted to polytene chromosomes where it co-localizes at multiple sites with Corto and with the PcG factor Polyhomeotic (PH). corto is involved in maintaining Abd-B repression outside its normal expression domain in embryos. This could be achieved by association between Corto and CycG since both proteins bind the regulatory element iab-7 PRE and the promoter of the Abd-B gene. These results suggest that CycG could regulate the activity of Corto at chromatin and thus be involved in changing Corto from an Enhancer of TrxG into an Enhancer of PcG (Salvaing, 2008a).

In Drosophila, the Bithorax-complex (BX-C) contains the three Hox genes, Ultrabithorax (Ubx), abdominal-A (abd-A) and Abdominal-B (Abd-B), that specify the identities of the third thoracic segment (T3) and the eight abdominal segments (A1 to A8). These genes are expressed in spatially regulated patterns during embryonic development thanks to maternal, gap and pair-rule proteins. Their large cis-regulatory sequences are modular and allow parasegmental regulation. These sequences contain different classes of elements such as initiation elements that respond to early segmentation gene products, insulators and promoter targeting sequences (Salvaing, 2008a).

Hox expression is maintained in the original pattern during later stages of development by the Polycomb-group (PcG) and trithorax-group (trxG) genes. In mutants of PcG or trxG genes, Hox patterns are established correctly but are not maintained. PcG proteins keep Hox genes silenced whereas TrxG proteins keep Hox genes activated thus counteracting PcG action. PcG and TrxG proteins are required for the maintenance of many gene expression patterns. These maintenance proteins form heteromultimeric complexes that bind to chromatin and alter its structure. Current models propose that PcG complexes lead to compact, transcriptionally inactive chromatin, whereas TrxG complexes maintain chromatin in an open conformation that facilitates transcription. In Drosophila, several PcG and TrxG complexes have been purified so far: the Polycomb Repressive Complex 1 (PRC1), the Polycomb Repressive Complex 2 (PRC2), the PhoRC complex, the Pcl-PRC2 complex, the Trithorax Activating Complex 1 (TAC1) and the Brahma Complex (BRM) also called SWI/SNF complex. They are extremely large complexes that contain several proteins including chromatin modifying enzymes such as histone methyl-transferases, acetyl-transferases or deacetylases (Salvaing, 2008a).

Although most PcG mutations suppress trxG mutations and vice versa, a large screen to identify modifiers of the trxG gene ash1 allowed isolation of enhancers that were previously identified as PcG [E(z), E(Pc), Asx, Scm, Psc and Su(z)2]. These genes were then called Enhancers of Trithorax and Polycomb (ETPs). Further molecular data showed that some ETPs encode members of PRC complexes, such as E(Z), PSC or SCM, while some do not. Recently, reclassification of these maintenance proteins has been proposed, the label PcG being kept for members of PRC silencing complexes and the label TrxG for members of complexes that counteract PcG-mediated silencing. A third class of proteins would be represented by PcG/TrxG DNA-binding recruiters or specific co-factors. This study keeps the term ETP for those maintenance proteins that play a dual role in PcG and TrxG functions without belonging to any PcG or TrxG complexes identified so far. The GAGA factor, Gaf, encoded by Trithorax-like (Trl), falls into this category. Indeed, it was first described as an activator of Hox genes, and later shown to play a role in the recruitment of PcG complexes without co-purifying with any PRC silencing complexes. The HMG protein DSP1 also meets the criteria to be an ETP: dsp1 mutants exhibit Hox gene loss-of-function phenotypes but DSP1 is also important for PcG recruitment to chromatin. corto behaves genetically as an ETP. corto mutants present PcG as well as trxG phenotypes and enhance the phenotypes of some PcG, trxG and ETP mutants (Lopez, 2001). Corto directly interacts with Gaf and DSP1 suggesting that ETPs are involved in collaborative processes (Salvaing, 2003; Salvaing, 2006; Salvaing, 2008a).

PcG, TrxG and ETP proteins bind DNA sequences called PRE/TRE that carry the information for the active or silent state of the gene they control. Some PRE/TRE have been shown to maintain this transcriptional state throughout cellular divisions in absence of the initial activator or repressor. Despite massive efforts towards identification of PcG complex targets at genome scale, the mechanism by which the active or inactive state of PRE/TRE is conserved throughout several cell cycles remains still largely unknown. Many PcG and ETP mutants [Asx, corto, E(z), Pc, ph, Psc, Su(z)2, Trl] exhibit proliferation defects as well as chromosome condensation and segregation defects. This suggests that maintenance proteins play a general role in cell cycle control. An attractive hypothesis is that ETPs are critical to maintain the correct association of PcG or TrxG complexes with chromatin during the cell cycle (Salvaing, 2008a).

In a two-hybrid screen using Corto as bait, Cyclin G (CycG), the Drosophila homologue of the mammalian Cyclin G1 and G2 (CycG1, CycG2), was isolated. Vertebrate CycG1 is a transcriptional target of the tumor suppressor p53 (Tamura, 1993; Okamoto, 1994). It is possibly involved in cell proliferation as it is overexpressed in certain cancer cells (Reimer, 1999; Baek, 2003). However, CycG1 induces G2/M arrest and cell death in response to DNA damage (Okamoto, 1999; Kimura, 2001; Seo, 2006). Vertebrate CycG2 acts as a negative regulator of cell cycle, as shown by its high level (Bates, 1996; Bennin, 2002) in cells in which G1/S arrest has been induced by growth inhibitory signals (Salvaing, 2008a).

This study addresses the interactions between Corto and CycG both in vitro and in vivo. CycG is shown to play an essential role during development. Moreover, CycG is targeted to many sites on polytene chromosomes where it co-localizes partially with Corto and with the PcG factor Ph. As an ETP, corto maintains Abd-B repression in embryos. This could be achieved by association between Corto and CycG since both proteins bind to Abd-B regulatory elements, including the iab-7 PRE and the promoter (Salvaing, 2008a).

This study has identified Cyclin G as a new binding partner of the ETP. CycG inactivation leads to lethality showing that this gene is essential in flies. Mammalian genomes encode two G-type cyclins, CycG1 and CycG2, the first one being mainly nuclear whereas the second is mainly cytoplasmic (Horne, 1996). Drosophila has a single homologue, however, it produces at least two different protein isoforms, only the larger being associated with chromatin. These isoforms could combine CycG1 and CycG2 functions. In Drosophila, large scale two-hybrid screens suggested binding of CycG to various Cyclin-Dependent Kinases (CDK) (Cdc2 and Cdk4). Corto and CycG interact in vitro as well as in vivo and form a complex in embryos and presumably also on chromatin. Moreover, Corto interacts with the amino-terminal domain of CycG, which is compatible with the simultaneous binding of CDK and cell-cycle control function of CycG (Salvaing, 2008a).

Requirement of PcG, trxG and ETP genes in cell-cycle control has already been shown in Drosophila. Interestingly, PcG and trxG genes are also involved in self-renewal and proliferation of hematopoietic stem cells in vertebrates. One way they might control cell proliferation is by an epigenetic regulation of genes involved in cell cycle and cell proliferation. Indeed, homologues of Drosophila E(z) and Brm participate in the transcriptional regulation of Cyclin A and E in vertebrates, and in Drosophila, Cyclin A is a PcG target. Alternatively, PcG, TrxG or ETP proteins may interact directly with cell cycle regulatory proteins. Indeed, it has been shown that Brm interacts with Cyclin E, that Mel-18, a human homologue of Posterior Sex Combs, interacts with Cyclin D2 possibly blocking its interaction with Cdks, and this study shows here that the ETP Corto interacts with CycG. These interactions reveal a potential role for these maintenance proteins in regulating the cell cycle independently of transcriptional regulation. This could be a widespread mechanism by which PcG, TrxG and ETP coordinate the chromatin activity status (Salvaing, 2008a).

CycG and Corto co-localize on many sites on polytene chromosomes suggesting that they may have regulated associations. The data show that Corto represses Abd-B in embryos and although it was not possible to test the role of CycG in regulating Abd-B expression in embryos, it was observed that both Corto and CycG bind the iab-7 PRE and the promoter of Abd-B suggesting that they could cooperate in this function. Nevertheless, neither Corto nor CycG were detected on the BX-C locus in salivary glands suggesting that they regulate Abd-B in a tissue-specific manner. The role of the CycG-Corto interaction needs to be further investigated. CycG could regulate Corto activity directly on chromatin by recruiting other factors like kinases or phosphatases thus modifying the phosphorylation status of Corto itself, of histones or other proteins at PRE/TRE and promoters. It has been shown that binding of the PcG protein Bmi1 to chromatin correlates with its phosphorylation status. It will be interesting to investigate whether Corto and CycG bind the iab-7 PRE and promoter of Abd-B simultaneously, to examine their phosphorylation status when bound to chromatin, and to determine if their presence correlates with Abd-B transcriptional activity. One interesting possibility would be that CycG is involved in changing Corto from an Enhancer of TrxG into an Enhancer of PcG (Salvaing, 2008a).

Regulation of Abd-B expression by Cyclin G and Corto in the abdominal epithelium of Drosophila

Polycomb-group (PcG) and trithorax-group (trxG) genes encode important regulators of homeotic genes, repressors and activators, respectively. They act through epigenetic mechanisms that maintain chromatin structure. The corto gene of Drosophila encodes a co-factor of these regulators belonging to the Enhancer of Trithorax and Polycomb class. Corto maintains the silencing of the homeotic gene Abdominal-B in the embryo and it interacts with a cyclin, Cyclin G, suggesting that it could be a major actor in the connection between Polycomb/Trithorax function and the cell cycle. This study shows that inactivation of Cyclin G by RNA interference leads to rotated genitalia and cuticle defects in the posterior abdomen of pupae and that corto genetically interacts with Cyclin G for generating these phenotypes. Examination of these pupae shows that development of the dorsal histoblast nests that will give rise to the adult epithelium is impaired in the posterior segments which identity is specified by Abdominal-B. Using a line that expresses LacZ in the Abdominal-B domain, it was shown that corto maintains Abdominal-B repression in the pupal epithelium whereas Cyclin G maintains its activation. These results prompt a proposal that the interaction between the Enhancer of Trithorax and Polycomb Corto and Cyclin G is involved in regulating the balance between cell proliferation and cell differentiation during abdominal epithelium development (Salvaing, 2008b).

Ubiquitous downregulation of CycG by RNA interference (using da::Gal4 or Act::Gal4 drivers) led to a high percentage of lethality in late third instar larvae or pharates depending on the CycG line and on the sex (Salvaing, 2008a). Lethality was complete in Act::Gal4/+; UAS::dsCycG2/+ males which intriguingly never underwent pupariation and stopped their development as third instar larvae, dying after a few days. In contrast, most females died as late pharates. UAS::dsCycG2/+; da::Gal4/+ as well as Act::Gal4/+; UAS::dsCycG2/+ emerging animals presented defects in the abdominal cuticle restricted to the posterior tergites A4 to A6. Apart from disorientation of abdominal bristles, the tergites of these segments exhibit unsclerotized patches of variable size. Males were more strongly affected than females and also frequently exhibited rotated genitalia (Salvaing, 2008b).

Genetic interactions between CycG and the loss-of-function alleles corto⁴²⁰ and corto⁰⁷¹²⁸ were examined. Their combination with ubiquitous RNAi inactivation of CycG increased lethality, cuticle defects and rotated genitalia. These data suggest that CycG and corto interact genetically and corroborate the existence of a functional relationship between CycG and corto (Salvaing, 2008b).

To understand the underlying defects of the cuticular phenotypes observed in RNAi-inactivated CycG flies, the development of the abdominal epithelium in pupae was addressed. In Drosophila, the abdominal epithelium of adults is derived from a fixed number of diploid histoblast cells, nested within the polyploid larval epithelium. Each abdominal hemisegment contains four histoblast nests, anterior and posterior dorsal, ventral and spiracle nests, that contribute to tergite and sternite of each abdominal segment. Histoblasts start to proliferate at the beginning of metamorphosis, replacing the larval cells, to eventually build up the adult abdominal integument. In wild-type pupae, the anterior and posterior dorsal histoblast nests of each hemisegment begin to fuse between 15 and 18 h APF. Fusion is completed at 24 h APF and the histoblasts have replaced all the polyploid larval cells at 48 h APF. In Act::Gal4/+; UAS::dsCycG2/+ 48 h APF pupae, whereas the dorsal histoblast nests of segment A3 fuse, the histoblast nests of segments A4 to A6 still remained small and unfused. Nevertheless, the development of these flies was not notably delayed, with regard either to puparium formation or to emergence of adult escapers. Therefore, it is concluded that RNAi inactivation of CycG especially impedes abdominal epithelium development of segments A4 to A6 where histoblast proliferation seemed to have stopped completely (Salvaing, 2008b).

It has been shown that corto is involved in the regulation of Abd-B and that Corto and CycG bind to the iab-7 PRE and to the promoter of Abd-B in embryos (Salvaing, 2008a). Since the epithelium defects of RNAi-inactivated CycG individuals affect abdominal segments A4 to A6, and are enhanced in corto mutants, it was hypothesized that they might be associated with misregulation of Abd-B, which specifies posterior abdominal identity. To address the role of corto and CycG in Abd-B regulation in the abdominal epithelium, genetic interactions between Abd-B and corto or CycG mutants was studied. The Fab-7¹ allele was used, in which both the Fab-7 boundary and the iab-7 PRE of the Abd-B cis-regulatory sequences have been deleted. This mutation induces a higher level of Abd-B expression in A6 which leads to a shift of A6 cell identity toward A7. As there is no normal sclerotized A7 segment in wild-type males, Fab7 homeotic A6 to A7 transformation results in loss of cells. As a result, Fab-7¹/+ males thus present a half-reduced A6 segment. It was observed that corto alleles enhance the expressivity of this phenotype leading to complete disappearance of the A6 segment in 100% of the males. Next, the effect of inactivation of CycG was examined in a Fab-7¹/+ genetic context. The expressivity of the Fab-7¹ phenotype was slightly enhanced in most (86%) of the UAS::dsCycG2/+; da::Gal4/Fab-7¹ males but at least a thin A6 segment always persisted. Curiously, no cuticular defects were observed neither in A5 nor in the remaining A6 tergites of these males suggesting that they might partly result from altered Abd-B expression. Overexpression of CycG also led to enhancement of the Fab-7¹ phenotype expressivity but in this case complete disappearance of A6 was observed in 56% of UAS::CycG/+; da::Gal4/Fab-7¹ males and in 100% of Act::Gal4/+; UAS::CycG/+; Fab-7¹/+ males. Lastly, the Fab-7¹ phenotype was investigated in corto, RNAi-inactivated CycG males. Crosses of Act::Gal4; Fab-7¹ females with UAS::dsCycG2/CyO; corto⁰⁷¹²⁸/TM6b males gave only few Act::Gal4/+; UAS::dsCycG2/+; Fab-7¹/corto⁰⁷¹²⁸ male escapers that all exhibited complete disappearance of A6 segment. These results suggest that both corto and CycG participate in maintenance of A6 cell identity by regulating Abd-B expression. corto clearly acted as a repressor of Abd-B since it enhanced the gain-of-function phenotype of Fab-7¹. However, it is not possible to conclude about the precise role of CycG on Abd-B expression since overexpression as well as inactivation led to enhancement of the Fab-7¹ phenotype, although to a lesser extent in case of inactivation. To understand this issue, the expression of Abd-B was addressed in corto mutants and in RNAi-inactivated CycG or overexpressing CycG individuals (Salvaing, 2008b).

Thus, Abd-B expression was examined in the abdominal epithelium of pupae. Since monoclonal anti-Abd-B antibodies show unspecific ubiquitous staining in the pupal epithelium, the HCJ199 strain was used where a P{LacZ} element is inserted in the cis-regulatory sequences of Abd-B. In agreement with published reports, it was observed that LacZ expression mimics Abd-B expression forming a decreasing gradient from A7 (in females) to the posterior part of A4, the expression in this segment being very faint and only detectable at high magnification. This pattern was also observed in Act::Gal4/+; HCJ199/+ control female pupae showing that the Act::Gal4 driver has no effect on Abd-B expression per se. At 24 h APF, LacZ was expressed in polyploid larval cells as well as in proliferating diploid histoblasts. Later on (48 h APF), LacZ was still expressed in the proliferating diploid histoblasts of A7, A6 and A5, and a barely discernible staining could be seen in posterior . In HCJ199/corto⁴²⁰ and HCJ199/corto⁰⁷¹²⁸ 48 h APF pupae, LacZ was also expressed from A7 to the posterior part of A4 but expression in posterior A4 was much stronger than in control pupae. Moreover, some cells in the posterior region of A3 also expressed LacZ. This suggests that, in the abdominal epithelium of pupae as in embryos, corto maintains repression of Abd-B expression. In RNAi-inactivated CycG female pupae (Act::Gal4/+; UAS::dsCycG2/+; HCJ199/+) at 48 h APF, almost complete loss of LacZ expression was observed in A5 and A6, whereas it was still expressed in A7. In contrast, 48 h APF Act::Gal4/+; UAS::CycG/+; HCJ199/+ female pupae that overexpressed ubiquitously CycG showed ectopic expression of LacZ in the whole abdomen. Taken together, these results suggest that CycG has the ability to activate Abd-B expression in the abdominal epithelium and contributes to Abd-B expression maintenance in A6 and A5. Thus, corto and CycG play opposite roles on the control of Abd-B expression in the abdominal epithelium, corto being a repressor and CycG an activator. The expression of Abd-B was addressed in pupae where corto and CycG expressions were simultaneously reduced. In Act::Gal4/+; UAS::dsCycG2/+; HCJ199/corto⁴²⁰ 48 h APF female pupae, although it was not possible to precisely determine segment borders due to impaired histoblast nest development, rescue of LacZ pattern was obseved that extended more anteriorly than in Act::Gal4/+; UAS::dsCycG2/+; HCJ199/+ pupae. Then, Abd-B loss of expression in A5 and A6 induced by CycG inactivation was abrogated when the amount of Corto was simultaneously reduced (Salvaing, 2008b).

This study has show that ubiquitous downregulation of CycG in pupae results in failure of epithelium formation in the posterior abdomen. Abdominal epithelium of adults derives from imaginal histoblasts that are recruited during embryogenic stages and form small group of diploid cells nested in the polyploid larval epithelium. The anterior dorsal nest is composed of about 15 to 18 cells whereas the posterior dorsal nest is composed of about 5 to 6 cells. These cells stay quiescent being arrested in G2 during the larval stages. At the onset of metamorphosis, they first undergo a phase of rapid proliferation triggered by ecdysone signalling and consisting of three synchronous and fast divisions. Also set off by ecdysone signalling are the second phase of histoblast proliferation which is slow and asynchronous and the simultaneous death of the polyploid larval cells. At 24 h APF, the anterior and posterior histoblast nests have fused. Inactivation of CycG impedes the proliferation of histoblasts in the posterior part of the abdomen, the dorsal anterior and posterior nests being still individualized at 48 h APF. This probably results in cuticle defects in the less severely affected individuals that will become adult. Similar cuticle defects have been described in some mutants (Arrowhead, escargot, cdc2, myb, torpedo, EcR) where they extend more often over the entire abdomen. Arrowhead has been shown to be involved in the establishment of abdominal histoblasts during embryogenesis. In RNAi-inactivated CycG larvae, the number of cells in the dorsal anterior and posterior histoblast nests is identical to that of wild-type larvae, suggesting that CycG inactivation does not affect histoblast recruitment during embryogenesis. esgargot and cdc2 are required to maintain diploidy of histoblast cells. In RNAi-inactivated CycG larvae, the size of histoblast nuclei in dorsal nests appears similar to the size of the corresponding wild-type nuclei thus suggesting that CycG inactivation does not affect ploidy. Lastly, the epithelium defects could be related to defects in cell proliferation. This is the case for the myb mutant, which proliferating histoblasts exhibit mitosis defects, or the torpedo mutant, which shows loss of mitotic figures in the histoblast nests at 25 h APF. In RNAi-inactivated CycG pupae at 48 h APF, approximately 100 and 40 cells were observed in the anterior and posterior nests of the A6 segment, respectively, suggesting that they might have undergone the 3 first rounds of division. Then, it could be that cells slow down during the second phase of proliferation. Intriguingly, like in RNAi-inactivated CycG pupae, slowdown of histoblast proliferation in segments A5 and A6 has been observed in the torpedo mutant; torpedo encodes the EGF receptor. This suggests that the role of CycG in the proliferation of the abdominal epithelium could be related to MAP kinase signalling. Furthermore, the use of a dominant-negative form of the Ecdysone-receptor that blocks death of the larval epidermal polyploid cells also induces cuticle defects. In this case, cell-autonomous inhibition of EcR activity leads to abortive delamination and persistence of larval polyploid cells in the pupal epithelium. A similar phenomenon, linked to disruption of ecdysone signal reception, could arise when CycG is inactivated. Interestingly, that Act::Gal4>UAS::dsCycG males never go through pupariation; this could reflect a defect in EcR signalling reception. Although the abdominal cuticle of corto mutants seems to be unaffected, the cuticle defects were enhanced by combining them with CycG inactivation. It suggests that corto and CycG together regulate the formation of the abdominal epithelium during metamorphosis (Salvaing, 2008b).

These data also show that corto and CycG oppositely regulate the expression of the Hox gene Abd-B in the growing pupal epithelium, corto behaving as a repressor whereas CycG behaves as an activator. Since Corto also represses Abd-B in embryos, it can be considered as a global repressor of Abd-B. Nevertheless, neither Corto nor CycG were detected on the BX-C locus in salivary glands suggesting that they regulate Abd-B in a tissue-specific manner (Salvaing, 2008a). In accordance with expression data, reduction of corto or overexpression of CycG leads to enhancement of the gain-of-function phenotype of Fab-7¹ heterozygotes. Surprisingly, whereas loss of Abd-B expression was observed upon inactivation of CycG, a mild enhancement of the gain-of-function phenotype of Fab-7¹ was seen in the same genetic background. This enhancement may result from perturbation of proliferation in the remaining tergite rather than from homeotic transformation of A6 cells to A7 cells. However, it may also reflect the intrinsic mechanism of action of CycG. Indeed, it has been shown that CycG binds both the iab-7 PRE and the promoter of Abd-B. It is well known that PREs have a stronger silencing activity when present in two copies in the genome, a phenomenon called pairing-sensitive repression. Then, if CycG activates Abd-B partly by working at the promoter and partly by limiting pairing-sensitive repression, loss of Abd-B activation at promoter could be overwhelmed by loss of pairing-sensitive repression when a single copy of the iab-7 PRE is present which is the case in the Fab-7¹/+ flies (Salvaing, 2008b).

Finally, in pupae combining RNAi-inactivated CycG and corto mutation, histoblast proliferation is still impeded whereas Abd-B expression seems to be restored. It suggests that the ratio between Corto and CycG activities must be preserved to insure appropriate regulation of Abd-B in the posterior abdomen. Altogether, these results suggest that a tripartite interaction between corto, CycG and Abd-B together regulates the balance between proliferation and differentiation during the formation of the abdominal epithelium at metamorphosis. Further experiments are now required to better understand how these processes are coordinated (Salvaing, 2008b).

Corto and DSP1 interact and bind to a maintenance element of the Scr Hox gene: understanding the role of Enhancers of trithorax and Polycomb

Polycomb-group genes (PcG) encode proteins that maintain homeotic (Hox) gene repression throughout development. Conversely, trithorax-group (trxG) genes encode positive factors required for maintenance of long term Hox gene activation. Both kinds of factors bind chromatin regions called maintenance elements (ME). Previous work has shown that corto, which codes for a chromodomain protein, and dsp1, which codes for an HMGB protein, belong to a class of genes called the Enhancers of trithorax and Polycomb (ETP) that interact with both PcG and trxG. Moreover, dsp1 interacts with the Hox gene Scr, the DSP1 protein is present on a Scr ME in S2 cells but not in embryos. To understand better the role of ETP, genetic and molecular interactions between corto and dsp1 were addressed. This study shows that Corto and DSP1 proteins co-localize at 91 sites on polytene chromosomes and co-immunoprecipitate in embryos. They interact directly through the DSP1 HMG-boxes and the amino-part of Corto, which contains a chromodomain. In order to search for a common target, a genetic interaction analysis was performed. corto mutants were found to suppress dsp1¹ sex comb phenotypes and enhance Antp^Scx phenotypes, suggesting that corto and dsp1 are simultaneously involved in the regulation of Scr. Using chromatin immunoprecipitation of the Scr ME, it was found that Corto was present on this ME both in Drosophila S2 cells and in embryos, whereas DSP1 was present only in S2 cells. These results reveal that the proteins Corto and DSP1 are differently recruited to a Scr ME depending on whether the ME is active, as seen in S2 cells, or inactive, as in most embryonic cells. The presence of a given combination of ETPs on an ME would control the recruitment of either PcG or TrxG complexes, propagating the silenced or active state (Salvaing, 2006).

It is concluded that the two ETPs corto and dsp1 interact genetically and that the proteins they encode (1) directly interact in vitro, (2) co-immunoprecipitate in embryos and (3) co-localize on 91 sites in salivary gland polytene chromosomes. These results suggest that the proteins are simultaneously involved in the regulation of several target genes. DSP1 can bind Corto through one of the two HMG-boxes that also mediate DNA binding. It has been suggested that during nucleoprotein complex formation, the HMG-box B of HMGB bends DNA whereas the HMG-box A mediates interaction with transcription factors, thus promoting their contact with targets. DSP1 seems to follow that scheme to enhance the binding of transcription factors as Dorsal or Bicoid to DNA. What therefore could be the role of the DSP1-Corto interaction in the regulation of common targets? First, DSP1 could bring Corto to the chromatin, where it could further interact with other partners. These partners could be PcG factors or GAGA factor, which have previously been shown to interact with Corto (Salvaing, 2003). Nevertheless, this hypothesis is unlikely since no modification of Corto binding to polytene chromosomes was observed in the dsp1¹ strain. Second, DSP1 could inhibit the interaction between Corto and PcG factors or GAF, thus preventing the silencing of targets that bind both proteins. Third, Corto could modify the DNA bending ability of DSP1 and thus modulate its interaction with other factors, for example TrxG complexes. Indeed, the dsp1 gene has been shown to interact with the TrxG genes trx and brm . The results do not allow discrimination between these last two, non-exclusive possibilities (Salvaing, 2006).

The Hox gene Scr is a common target of Corto and DSP1. Both proteins bind a 10-kb XbaI fragment located 37-kb upstream of the Scr transcription start. Genetic studies have shown that this fragment is required for Scr function in the embryo and in the imaginal disc. In embryos, it restricts the expression of a Scr-lacZ fusion gene to the labial and prothoracic segments, whereas in larvae it is required for Scr expression in the first leg imaginal disc and for Scr silencing in the second and third leg imaginal discs. Interestingly, the function of the 10-kb XbaI fragment is sensitive to a subset of PcG and TrxG mutations and has been genetically characterized as an upstream maintenance element of Scr . At the end of embryogenesis, the Scr expression domain is restricted to the labial and prothoracic segments. In consequence, the mean state of this ME in the whole embryo would be silenced. It can thus be assumed that the global situation in embryos mimics that of the T2 and T3 leg imaginal discs. Conversely, since Scr is expressed in S2 cells, it is proposed that the situation in S2 cells rather mimics that of T1 leg imaginal disc cells. Hence, Corto, which is present on the Scr ME whether active (S2 cells) or silenced (embryos), could be present on this ME in all three leg imaginal discs. In contrast, DSP1, which is present on the ME in S2 cells but not in embryos, could bind the ME only in cells where this element is active, hence in T1 leg imaginal disc cells. It is thus proposed that both Corto and DSP1 proteins localize on this Scr ME in the first leg imaginal disc (Salvaing, 2006).

Some trxG mutants as well as the dsp1 null mutant exhibit a reduced sex comb and previous work has shown that dsp1 interacts with certain trxG genes and regulates Scr expression in T1 discs. HMGB, the vertebrate homologue of DSP1, has been reported to activate and stabilize the TFIID-TFIIA-promoter complex and some TrxG factors have been shown to interact with the RNA polymerase II complex, thus facilitating transcriptional elongation. This leads to a proposal that in the T1 leg imaginal disc, DSP1 facilitates the interaction between a TrxG complex and the transcription machinery, thus maintaining Scr activation. Moreover, Corto has been shown to interact with PcG complexes. The binding of Corto to DSP1 could then impede the interaction between Corto and PcG complexes, thus limiting their recruitment. Therefore, the interaction between the two proteins on the ME would lead to a level of Scr transcription compatible with T1 identity. Conversely, in the T2 and T3 leg imaginal discs, since DSP1 does not bind the ME, Corto would be able to interact with PcG complexes, thus enhancing the silencing of Scr (Salvaing, 2006).

In summary, this study has shown that the two ETPs Corto and DSP1 directly interact and are simultaneously found on a Scr ME when active, whereas Corto alone is found on the same ME when inactive. These data suggest that different combinations of ETP favor the recruitment of either PcG or TrxG complexes, participating in the maintenance of the silenced or active state of ME (Salvaing, 2006).

The Drosophila Corto protein interacts with Polycomb-group proteins and the GAGA factor

In Drosophila, PcG complexes provide heritable transcriptional silencing of target genes. Among them, the ESC/E(Z) complex is thought to play a role in the initiation of silencing whereas other complexes such as the PRC1 complex are thought to maintain it. PcG complexes are thought to be recruited to DNA through interaction with DNA binding proteins such as the GAGA factor, but no direct interactions between the constituents of PcG complexes and the GAGA factor have been reported so far. The Drosophila corto gene interacts with E(z) as well as with genes encoding members of maintenance complexes, suggesting that it could play a role in the transition between the initiation and maintenance of PcG silencing. Moreover, corto also interacts genetically with Trl, which encodes the GAGA factor, suggesting that it may serve as a mediator in recruiting PcG complexes. Corto bears a chromo domain, and evidence is provided for in vivo association of Corto with ESC and with PC in embryos. Moreover, GST pull-down and two-hybrid experiments show that that Corto binds to E(Z), ESC, PH, SCM and GAGA and co-localizes with these proteins on a few sites on polytene chromosomes. These results reinforce the idea that Corto plays a role in PcG silencing, perhaps by confering target specificity (Salvaing, 2003; full text of article).

Search PubMed for articles about Drosophila Corto

Baek, W. K., et al. (2003). Increased expression of cyclin G1 in leiomyoma compared with normal myometrium. Am. J. Obstet. Gynecol. 188: 634-639. PubMed Citation: 12634633

Bates, S., Rowan, S. and Vousden, K. H. (1996). Characterisation of human cyclin G1 and G2: DNA damage inducible genes. Oncogene 13: 1103-1109. PubMed Citation: 8806701

Bennin, D. A., et al. (2002). Cyclin G2 associates with protein phosphatase 2A catalytic and regulatory B' subunits in active complexes and induces nuclear aberrations and a G1/S phase cell cycle arrest. J. Biol. Chem. 277: 27449-27467. PubMed Citation: 11956189

Horne, M. C., et al. (1996). Cyclin G1 and cyclin G2 comprise a new family of cyclins with contrasting tissue-specific and cell cycle-regulated expression. J. Biol. Chem. 271: 6050-6061. PubMed Citation: 8626390

Kimura, S. H., Ikawa, M., Ito, A., Okabe, M. and Nojima, H. (2001). Cyclin G1 is involved in G2/M arrest in response to DNA damage and in growth control after damage recovery. Oncogene 20: 3290-3300. PubMed Citation: 11423978

Lopez, A., Higuet, D., Rosset, R., Deutsch, J. and Peronnet, F. (2001). corto genetically interacts with Pc-G and trx-G genes and maintains the anterior boundary of Ultrabithorax expression in Drosophila larvae. Mol. Genet. Genomics 266: 572-583. PubMed Citation: 11810228

Okamoto, K. and Beach, D. (1994). Cyclin G is a transcriptional target of the p53 tumor suppressor protein. Embo J. 13: 4816-4822. PubMed Citation: 7957050

Okamoto, K. and Prives, C. (1999). A role of cyclin G in the process of apoptosis. Oncogene 18: 4606-4615. PubMed Citation: 10467405

Reimer, C. L., et al. (1999). Altered regulation of cyclin G in human breast cancer and its specific localization at replication foci in response to DNA damage in p53+/+ cells. J. Biol. Chem. 274: 11022-11029. PubMed Citation: 10196184

Salvaing, J., Lopez, A., Boivin, A., Deutsch, J. S. and Peronnet, F. (2003). The Drosophila Corto protein interacts with Polycomb-group proteins and the GAGA factor. Nucleic Acids Res. 31(11): 2873-82. 12771214

Salvaing, J., et al. (2006). Corto and DSP1 interact and bind to a maintenance element of the Scr Hox gene: understanding the role of Enhancers of trithorax and Polycomb. BMC Biol 4: 9. PubMed Citation: 16613610

Salvaing, J., et al. (2008a). The enhancer of trithorax and polycomb corto interacts with cyclin G in Drosophila. PLoS ONE 3(2): e1658. PubMed Citation: 18286205

Salvaing, J., et al. (2008b). Regulation of Abd-B expression by Cyclin G and Corto in the abdominal epithelium of Drosophila. Hereditas 145(3): 138-46. PubMed Citation: 18667003

Seo, H. R., et al. (2006). Cyclin G1 overcomes radiation-induced G2 arrest and increases cell death through transcriptional activation of cyclin B1. Cell Death Differ. 13: 1475-1484. PubMed Citation: 16322753

Tamura, K., et al. (1993). Cyclin G: a new mammalian cyclin with homology to fission yeast Cig1. Oncogene 8: 2113-2118. PubMed Citation: 8336937

date revised: 1 September 2008

Home page: The Interactive Fly © 2008 Thomas Brody, Ph.D.