Ikaros-CtIP Interactions Do Not Require C-terminal Binding Protein and Participate in a Deacetylase-independent Mode of Repression*

Ikaros and Aiolos are Kruppel zinc finger proteins that play key roles in hemo-lymphoid development and homeostasis. We have previously shown that they can repress transcription through the recruitment of histone deacetylases (HDACs). Here, we provide the first functional evidence that these proteins can also repress gene function in a manner that does not require deacetylase activity. This functionality can be attributed in part to Ikaros interactions with the HDAC-independent corepressor, C-terminal binding protein (CtBP). However, mutations that block Ikaros-CtBP interactions do not abolish Ikaros's repression activity, implicating the involvement of additional corepressors. Consistent with this expectation, we show that Ikaros can interact with a CtBP-interacting protein (CtIP), which has also been linked to a deacetylase-independent strategy of repression. Despite being a CtBP interactor, CtIP's association with Ikaros does not require CtBP but instead relies upon its Rb interaction domain. Significantly, Ikaros can interact with Rb, which itself can repress gene function in a deacetylase-independent manner. A mutation in Ikaros that abrogates CtIP interactions significantly reduces repression, and a double mutation that prevents interaction with both CtIP and CtBP even further alleviates repression. Finally, we show that CtIP and CtBP can interact with the general transcription factors, TATA binding protein and transcription factor IIB, which suggests a possible mechanism for their deacetylase-independent mode of repression.

lished that Ikaros proteins play critical roles during hemolymphopoiesis (9 -11). Ikaros is required from the earliest stages of hemopoiesis, at the level of the hemopoietic stem cell (12), to the later stages of lymphoid cell fate determination; in addition, Ikaros proteins regulate lymphocyte proliferation and homeostasis (13,14). Molecular and biochemical studies aimed at understanding the basis for these complex biological roles have revealed that Ikaros, in addition to functioning as an activator (15), can also potently repress gene expression (16).
Transcriptional repressors can be categorized in several ways. A common approach is to classify them as long range or short range repressors. Members of the former group, such as Groucho and Sir proteins, are capable of making a promoter resistant to all enhancers regardless of their distance from the promoter, whereas short range repressors, such as Kruppel and Giant, act in a less general manner to block the activity of locally bound activators (17). An alternative but non-mutually exclusive approach to repressor classification is based on the utilization, or the lack thereof, of the activity of histone deacetylase enzymes (HDACs) 1 for repressor function. HDACmediated repression is expected to occur through the removal of acetyl groups from the N termini of histones, which presumably creates a compact chromatin configuration that inhibits transcription. Examples of HDAC-recruiting corepressors include the Sin3 and Mi-2␤ proteins (18,19). Histone deacetylase-independent repressors are believed to function through multiple mechanisms, but the strategy that has been most extensively studied is the interaction of such factors with the basal transcriptional machinery; these interactions affect recruitment of the RNA Polymerase II holoenzyme to the promoter or events associated with promoter clearance and re-initiation (20 -22). Examples of HDAC-independent corepressors include the C-terminal binding protein (CtBP) (23) and two CtBP interactors, the CtBP-interacting protein (CtIP) (24) and the histone deacetylase-related protein (HDRP/MITR) (25).
CtBP is a 48-kDa phosphoprotein that was first identified as an interactor of adenovirus E1A and more recently of several Dipteran and mammalian repressors (26,27). Interaction between CtBP and these proteins, in most cases, is mediated through a PXDL(S/R) motif (26,27). Investigation of the mech-* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  anisms behind CtBP-mediated repression has revealed that, although CtBP can interact with histone deacetylases, it can repress transcription even when deacetylase activity is inactivated, suggesting its ability to use alternative repression mechanisms (26,27). In a search for CtBP-interacting proteins, CtIP, a 125-kDA protein with similarity to DNA repair proteins, was identified (24).
In addition to interacting with CtBP, CtIP has also been shown to bind the tumor suppressors, Rb/p130 (28) and BRCA1 (29,34). Rb is a key regulator of the G 1 /S transition of the cell cycle, and CtIP has been implicated in the deacetylase-independent repression pathway of this key regulator (28). BRCA1 is a 1863-amino acid protein composed of an N-terminal RING domain and two C-terminal BRCT domains whose mechanisms of action are poorly understood. Germline mutations of BRCA1 are responsible for many cases of hereditary breast and ovarian cancers (31). CtIP has been shown to interact with BRCA1 through its BRCT domains and to be a component of a BRCA1⅐BARD1 complex (32) as well as a BRCA1⅐LMO4⅐Ldb1 complex (33). Significantly, mutations in BRCA1 found in breast cancer patients prevent interactions with CtIP, suggesting an important role for CtIP in BRCA1's tumor-suppressive function (34,29). CtIP has been implicated in BRCA1's DNA repair function (35). Upon genotoxic stress, such as ␥-irradiation, CtIP becomes phosphorylated by the ATM kinase, which apparently prevents CtIP-BRCA1 interactions, thus allowing BRCA1 to activate genes involved in DNA repair such as p21 and GADD45 (35). These findings have, however, been strongly contested (36). Nevertheless, it is very likely that CtIP plays important roles in regulating the tumor suppressive functions of BRCA1, Rb, and other regulators.
We have previously shown that Ikaros can interact with the HDAC-recruiting factors, Sin3 and Mi-2␤ (16,37). Consistent with such interactions, Ikaros repression of the adenovirus major late (AdML) promoter is relieved by the deacetylaseinhibitor trichostatin A (16). However, Ikaros also interacts with CtBP that can repress transcription in a deacetylaseindependent manner (38). Here, we investigate the HDACindependent repression potential of Ikaros. We show that Ikaros-mediated repression of the thymidine kinase (tk), unlike that of the AdML promoter, is insensitive to deacetylase inhibitors. In addition to CtBP, Ikaros can interact with two corepressors, CtIP and Rb, that can work through a deacetylaseindependent pathway. Mutations that abrogate CtIP interactions reduce repression by Ikaros whereas those that prevent both CtBP and CtIP associations even further alleviate repression. We provide evidence to suggest that Ikaros repression through this pathway may involve interactions with the basal transcriptional machinery.

EXPERIMENTAL PROCEDURES
Plasmids-BXG1, BXG1-Ik1, BXG1-Aio, BXG1-MAD, BXG1-mMAD, the reporters G5tkCAT and G5AdMLPCAT, CDM8-Ik1, CDM8-HA-Ik1, CDM8-FLAG-Aio3, CDM8-FLAG-Helios, CDM8-FLAG-Eos (Daedalus), CDM8-MT-Sin3A, pCMV2-FLAGIk1, pCMV2-FLAGIk1cm, and GST-hCtBP1 have been previously reported (38). Deletion and point mutants of Ik6 were generated by the Stratagene mutagenesis kit using Ik6 in the context of the BXG1 vector, which encodes the Gal4 DBD (amino Transfections-293T and NIH3T3 cell lines were maintained in Dulbecco's modified Eagle's medium with 10% fetal bovine serum (Hy-Clone). Transfections of these cell lines were carried out using the HEPES buffered saline-CaPO 4 method. For repression assays, 1 g of the Gal4 fusion plasmid, 10 g of the Gal4-reporter plasmid, and 0.5 g of the pXGH5 growth hormone transfection efficiency control plasmid were used. Twenty-four hours after transfection cells were fed with fresh media, and 18 -24 h later cells were harvested and processed for CAT assays as described previously (8). In those instances where trichostatin A (Upstate Biotech) was employed, we added the drug to the cells 16 -18 h before harvesting. Growth hormone assays were done as recommended by the manufacturer (Nichols Institute). Transfections were typically performed in duplicate and repeated between three and six times.
Immunoprecipitation and Western Analysis-Whole-cell extracts from 293T cells transfected with the relevant plasmids were prepared as previously described (8) and pre-cleared using Protein G-agarose beads (Roche Molecular Biochemicals). The pre-cleared extracts were incubated with the antibody of interest or the relevant isotype control on ice for 1 h. 30 l of Protein G beads was then added to the extract, and the extracts were rotated overnight. The beads were collected by centrifugation and washed four times with TS buffer. The beads obtained after this procedure were treated with SDS sample buffer, boiled at 95°C for 15 min, and loaded onto a SDS-polyacrylamide gel along with 8 -10% of the cell extract used for the immunoprecipitation. The proteins were transferred to a nitrocellulose membrane, probed with the relevant antibody, and examined by autoradiography with ECL (Amersham Biosciences, Inc.). FLAGM2 purification of Ikaros complexes has been described before (37). Antibodies used were: Myc-tag, MT (Roche Molecular Biochemicals), HA (BAbCO), FLAG M2 (Sigma), Gal4, Sin3B (Santa Cruz Biotechnology), HDAC2 (Zymed Laboratories Inc.), Rb (Amersham Biosciences, Inc.), and anti-Ikaros and Mi-2, which have been previously described (37). CtIP antibodies were generously provided by Dr. R. Baer and Dr. W.-H. Lee. Anti-HDRP was provided by Dr. X. Zhou and Dr. P. Marks.
GST Interaction Assays-GST, GST-TBPN, GST-TBPC, and GST-TFIIB were prepared using previously described protocols (38). 1-2 g of the GST proteins was incubated with proteins programmed in rabbit reticulocyte lysate (Promega) for 1 h at 4°C and washed extensively with MT/phosphate-buffered saline. The beads were then boiled in SDS sample buffer and fractionated on an SDS-polyacrylamide gel. The gels were then dried and visualized by autoradiography.
Histone Deacetylase Assay-Histone deacetylase assays were performed using tritiated chicken reticulocyte histones as described previously (37). Briefly, immunoprecipitates from 293T whole cell extracts were washed 3ϫ in TS buffer and incubated with 100,000 cpm of tritiated acetylated histones for 45 min at 30°C in HD assay buffer. The reaction was stopped by acidification, and the released tritium was extracted with ethyl acetate.

Ikaros Repression of the tk Promoter Does Not Rely on Histone Deacetylase Activity-We have previously shown that
Ikaros repression of the adenovirus major late (AdML) promoter is relieved in the presence of the histone deacetylase inhibitor, trichostatin A. Thus, we suggested that Ikaros mediates repression of this promoter through the action of histone deacetylases (HDACs) (Fig. 1, left panel) (16). Subsequently, we found that Ikaros interacts with the corepressor CtBP, which can repress transcription in a histone deacetylase activityindependent manner (38). Based on this finding we claimed that Ikaros, in addition to repressing transcription through HDAC, can also function using HDAC activity-independent mechanisms (38).
To address the possible ability of Ikaros to repress in a deacetylase activity-independent manner, we set out to identify promoters that Ikaros might repress using this alternate repression mechanism. In a recent report we showed that the Ikaros corepressor, CtBP, can repress the thymidine kinase (tk) promoter in a deacetylase activity-independent manner (38). To determine whether Ikaros's repression of the tk promoter might also utilize a similar strategy, NIH3T3 cells were transfected with G5tkCAT and expression vectors encoding Gal4 DNA binding domain (DBD) fusions of Ikaros and its family member, Aiolos. As controls, we included the empty vector, the vector expressing the Gal4 DBD alone, and Gal4 DBD fusions to the Sin3 interaction domain of MAD (MAD) or a mutant version of this domain that cannot interact with Sin3 (mMAD); the MAD protein serves as a positive control, because it has been shown to repress the tk promoter in a deacetylasedependent manner (39) whereas its mutant variant serves as the negative control. Transfectants were either treated with trichostatin A or left untreated. CAT assays revealed that repression of the tk promoter by Gal4-Ik1 and Gal4-Aiolos, unlike Gal4-MAD, was not relieved over background levels in the presence of the deacetylase inhibitor ( Fig. 1, right panel). Thus, Ikaros-and Aiolos-mediated repression of the tk, unlike the AdML promoter, is not dependent on the activity of histone deacetylases. This observation is highly reminiscent of the corepressors Rb and HDRP, which also repress the tk promoter in a manner that does not rely on HDACs (40,41). This provides the first functional evidence that Ikaros and Aiolos can repress transcription in a manner independent of HDAC activity.
Ikaros Can Interact with the HDAC Activity-independent Corepressor CtIP-Ikaros interactions with CtBP can account, in part, for this alternate repression strategy. Nevertheless, it is highly likely that other corepressors are also involved, because mutations that abolish the Ikaros-CtBP interaction still permit significant levels of repression (38). This supposition is further strengthened by the observation that Aiolos, which cannot interact with CtBP (38), continued to repress the tk promoter in the presence of the deacetylase inhibitor.
To identify other Ikaros-interacting factors that can account for its deacetylase-independent repression strategy, we screened a panel of seven well-established corepressors for binding the most full-length Ikaros isoform, Ik1, through a co-transfection/co-immunoprecipitation approach. These studies highlight the specificity that Ikaros shows in its interactions with corepressors. Of all the proteins tested, only CtIP and our positive control, Sin3A, were capable of binding Ikaros at detectable levels ( Fig. 2A). Significantly, CtIP has been implicated in effecting HDAC-independent repression through the tumor suppressor, Rb (28). Thus, CtIP interactions with Ikaros may play a role in Ikaros's deacetylase-independent repression mechanism.
CtIP Interacts with Ikaros Proteins in Vitro and in Vivo-We next tested whether other Ikaros isoforms, like Ik1, could also interact with CtIP. Co-immunoprecipitation experiments showed that all the tested isoforms, Ik2, -3, and -7, could interact with CtIP (Fig. 3A). To determine whether the CtIP-Ikaros interactions seen in vitro could be recapitulated in vivo, we probed Western blots containing Ikaros complexes immunopurified from resting and cycling T lymphocytes. CtIP was found in complexes from both sources (Fig. 3B). Thus, CtIP appears to be an interactor of Ikaros proteins in lymphocytes.
An Rb-binding Motif on CtIP Is Required for Its Interaction with Ikaros-Because both Ikaros and CtIP bind CtBP (24,38), we tested whether the CtIP-Ikaros interaction was mediated through this common corepressor. Wild type Ikaros (Ik1) and a mutant form that cannot interact with CtBP (Ik1cm) were transfected along with CtIP and tested for binding. Interestingly, Ik1 and Ik1cm were capable of interacting with CtIP, albeit less strongly (Fig. 3C, compare IP lanes 1 and 4). Thus, although CtIP may be recruited to Ikaros through CtBP, it can also be recruited through CtBP-independent means.
In addition to interacting with CtBP, CtIP also interacts with the tumor suppressor and corepressor, Rb (24, 28). To determine whether CtIP might be recruited to Ikaros through an Rb-dependent mechanism, we co-transfected Ikaros with FIG. 2. Ikaros interacts with the HDAC-independent corepressor CtIP. In A, 293T cells were transfected with the Ik1 and either myc-tagged (MT-) Sin3A, Cabin1, BCoR, Groucho, CoREST, or CtIP with Ikaros. Whole cell extracts were immunoprecipitated with myc antibody and immunoblotted with Ikaros antibody to test for interaction. I, input; C, isotype control IP; B, bound fraction from specific IP. In B, interaction between SMRT and FLAG-Ikaros was tested as in A with the indicated antibodies.

FIG. 1. Ikaros represses the tk promoter in an HDAC activity-independent manner.
HDAC activity-independent repression by Ikaros. NIH3T3 cells transfected with 1 g of the indicated Gal4 fusions (BXG1), 10 g of G5AdMLPCAT or G5tkCAT, and 1 g of pXGH5 as a transfection efficiency control plasmid. 16 -18 h before harvest, transfectants were treated (ϩTricho) with trichostatin A (100 ng/ml) or left untreated (ϪTricho). CAT activity was corrected for transfection efficiency using the growth hormone assay. This experiment was done in duplicate three times, and variation between experiments was less than 20%. Fold derepression upon trichostatin treatment is indicated below the graph and was calculated as the increase in corrected CAT activity upon trichostatin treatment divided by the corrected CAT activity in untreated cells. The left panel has been previously reported (16) and is included to allow comparison. wild type CtIP and CtIP mutants that are defective for interactions with Rb (ϪRb) and CtBP (ϪCtBP). Wild type CtIP interacted strongly with Ikaros whereas CtIP defective for interactions with CtBP showed a reduced interaction (Fig. 3C,  compare IP lanes 1 and 2). Interestingly, CtIP that was defective for Rb interactions was significantly impaired in its interactions with Ikaros (Fig. 3C, compare IP lanes 1-3). Taken together, these data show that Ikaros can bind CtIP through a mechanism that relies upon an intact Rb binding domain on CtIP (summarized in Fig. 3D). So, can Ikaros associate with Rb? FLAG-Ik1 and Rb were co-transfected and tested for association by immunoprecipitation. Rb was indeed immunoprecipitated with Ikaros (Fig. 3E), which lends support to our earlier finding that CtIP interactions with Ikaros require a functional Rb interaction motif.
In summary, these data indicate that Ikaros interactions with CtIP do not require CtBP but instead require a functional Rb motif on CtIP.
CtIP Interacts with All Ikaros Family Members-Consistent with the finding that Ikaros does not require CtBP to interact with CtIP, the Ikaros family members, Helios, Aiolos, and Eos, which do not interact with CtBP, can bind CtIP (Fig. 3F). Further support for CtBP-independent recruitment of CtIP to Ikaros and its family comes from the finding that exon 7 of Ikaros, which lacks a CtBP-binding motif, can also interact with CtIP (Fig. 3A, lane 5). These data suggest that a region in exon 7 is most likely the CtBP-independent domain through which CtIP associates with Ikaros.

Mutations That Abolish Ikaros Associations with CtIP
Alleviate Repression-To obtain CtIP interaction-defective Ikaros mutants, we targeted several mutations in exon 7 of the Ikaros isoform, Ik6 ( Fig. 4B and data not shown). Of these mutants, M8, which contains a 20-amino acid deletion spanning residues 416 -435, was found to significantly decrease CtIP binding (Fig. 4, A and B). To determine the role of the CtIP-Ikaros interaction, we tested the effect of this mutation on repression by Ik6, the most potent repressor among the Ikaros isoforms. Like the CtBP interaction mutant (M1), the CtIP mutant (M8) caused a 50% reduction in repression of the tk promoter (Fig.  4B). When both these mutations were incorporated in a single molecule (M9), repression was further reduced to 15% of wild type Ik6 levels (Fig. 4B). These data indicate that CtIP and CtBP are major components of the deacetylase-independent repression strategy of Ikaros on the tk promoter.
CtIP Does Not Interact with HDAC2 and Precipitates Low Levels of HD Activity-What is the mechanism of CtIP-mediated repression? Although CtIP has been implicated in deacetylase-independent repression, little is known about its interactions with deacetylases. To determine whether CtIP can interact with endogenous histone deacetylases, we immunoprecipitated CtIP, and as a positive control Sin3A, from 293T cells. CtIP, unlike Sin3A, was not found associated with any significant amount of HDAC2 (Fig. 5A). This was verified by histone deacetylase assays of these immunoprecipitates, which indicated that HDAC activity associated with CtIP was close to background levels. In contrast, Sin3A, which associates FIG. 3. Ikaros can bind the HDAC activity-independent corepressor CtIP through a CtBP-independent mechanism. A, interaction between CtIP and Ikaros isoforms (Ik1, -2, -3, and -7) or exon 7 (E7) was tested by immunoprecipitation. B, in vivo interaction between CtIP and Ikaros in activated (a) and resting (r) T lymphocytes. Immunopurification of Ikaros-containing complexes was accomplished using a FLAGM2 column as previously described (37). The input (I), final wash (W), and eluate (E) were tested by immunoblot analysis using antibodies to CtIP. C, an Rb but not a CtBP motif on CtIP is critical for interactions with Ikaros. Ik1 wild type (Ik1WT) and Ik1 defective for interactions with CtBP (Ik1cm) were tested for their ability to interact with wild type CtIP, CtIP that cannot interact with CtBP (ϪCtBP), and CtIP that cannot interact with Rb (ϪRb) by IP. The numbers below the immunoblot are included to aid the reader in comparing input and IP lanes. D, a summary of the interaction data obtained in C. E, C33A cells were transfected with Rb and FLAG-Ik1 to determine if the two proteins interact. To this end, whole cell lysates were immunoprecipitated with FLAG and an isotype control antibody. F, association between Ikaros family members and CtIP was tested by IP from whole cell extracts prepared from transfected 293T cells. The asterisks in A, E, and F identify the heavy chain of the immunoprecipitating antibody. strongly with HDACs, supported 18-fold higher activity than background levels (Fig. 5A). These data lend support to the suggestion that CtIP likely utilizes HDAC-independent means to repress gene expression.
Ikaros, CtBP, and CtIP Interact with Components of the Basal Machinery-A well studied HDAC-independent repression strategy involves interactions with the basal transcriptional machinery that negatively affect pre-initation complex assembly and/or promoter clearance (20,22). Using GST fusions of different components of the basal transcriptional machinery, we found that in vitro translated CtIP can bind TFIIB (Fig. 5B) whereas the other deacetylase-independent Ikaros corepressor, CtBP, can associate with both the N (amino acids 1-128) and C termini of TBP as well as with TFIIB (Fig. 6A). Furthermore, Ikaros itself can interact with the C terminus of TBP (amino acids 128 -328) and with TFIIB (Fig. 5B). Taken together these findings raise the possibility that the HDAC activity-independent repression mediated by Ikaros on the tk promoter may occur through interactions with components of the basal transcriptional machinery. DISCUSSION We have previously shown that Ikaros and Aiolos can repress transcription through the recruitment of histone deacetylases (16). In addition, Ikaros interacts with CtBP, which can repress through HDAC activity-independent mechanisms (38). On the basis of this interaction, we proposed that Ikaros might also be capable of this alternate repression strategy. Here, we provide the first functional evidence that Ikaros and Aiolos can effect repression in a manner that is independent of deacetylase activity. However, mutations in Ikaros that abolished interactions with CtBP could still repress transcription, indicating the involvement of other corepressors (38). Consistent with this expectation, we show that Ikaros interacts with the corepressors, CtIP and Rb, which are capable of deacetylase-independent repression. Finally, mutations that abrogate CtIP interactions with Ikaros alleviate repression, and those that prevent both CtBP and CtIP interactions even further reduce repression of the tk promoter.
Because both Ikaros and CtIP contain the CtBP penta-peptide interaction module and because both these factors can bind CtBP (24,38), we considered the possibility that their association was mediated via CtBP. However, Ikaros proteins bearing mutations that prevented interactions with CtBP, were still able to interact with CtIP. Therefore, interaction with CtBP was not essential for CtIP associations with Ikaros. In support of this finding, a domain of Ikaros lacking a CtBP interaction FIG. 4. A mutation in Ikaros that prevents interactions with CtIP alleviates repression. A, 293T cells were co-transfected with CtIP and the indicated Ik6 mutants (diagrammed in B) and tested for binding as described in Fig. 3A. B, effects of CtIP interaction mutations on Ik6 on repression. The indicated Gal4 fusions (1 g) were transfected with the reporter G5tkCAT (10 g) and a transfection control plasmid (0.5 g). Whole cell extracts prepared from the transfectants were assayed for CAT activity. Fold Repression was calculated by dividing the normalized CAT activity of the Gal4 DBD plasmid by that obtained from each Ik6 variant. Transfections were done in duplicate four times. domain (exon 7), as well as the Ikaros family members that cannot associate with CtBP, could bind CtIP. Thus, CtIP is a good candidate for a corepressor that effects repression of the tk promoter by Ikaros and Aiolos.
What is the CtBP-independent mode of CtIP association with Ikaros? CtIP interactions with Ikaros require the former's intact Rb interaction motif, suggesting that CtIP may associate with Ikaros through Rb family proteins. In support of this suggestion, both CtIP and Ikaros can interact with Rb. But why might Ikaros need two independent ways to recruit CtIP? It has recently been shown that the binding of CtBP to its interacting proteins is regulated by the levels of nicotinamide adenine nucleotides, NAD ϩ and NADH; agents capable of increasing NADH levels stimulate interactions between CtBP and its interactors and thereby increase repression (42). Based on these findings, we posit that having a CtBP-independent mode of recruitment would permit Ikaros⅐CtIP complexes to function under conditions that do not favor Ikaros-CtBP interactions.
How does CtIP mediate repression? CtIP cannot interact with HDAC2 and immunoprecipitates very small amounts of deacetylase activity, supporting its classification as a deacetylase-independent repressor. Significantly, CtIP, CtBP, and Ikaros can interact with components of the basal machinery, namely TBP and TFIIB. Several repressors have been shown to effect repression through such interactions. Studies with Rb have shown that it affects the formation of an effective preinitiation complex possibly through its interactions with the TFIID (22) whereas detailed studies with the nuclear receptor corepressor, NCoR, have indicated that it blocks interactions between TAFII32 and TFIIB, which are crucial for transcriptional initiation (20). NCoR has also been hypothesized to lock interactions with basal transcriptional components into a nonfunctional complex or conformation that abrogates transcription. Future studies using in vitro transcription systems will allow us to address the role, if any, of Ikaros and its corepressors' interactions with basal transcriptional factors in repression. Recently, it has also been shown that the CtIP interacting proteins, CtBP and Rb, can repress gene function in a deacetylase-independent manner through the recruitment of Poly-combs (30). This raises the possibility that Ikaros may also utilize this avenue of deacetylase-independent repression.
The importance of the Ikaros-CtIP interaction in deacetylase-independent repression was consolidated by mutational analysis. Mutation of the CtIP interaction site on Ik6 significantly reduced repression of the tk promoter by 50% of wild type levels. Thus, CtIP is a component of the deacetylaseindependent repression by Ikaros of this promoter. Furthermore, mutations that abrogated CtIP and CtBP interactions reduced repression to roughly 15% of levels supported by the wild type protein. Thus, CtIP and CtBP appear to collaborate to repress the tk promoter. The fact that repression is not completely abolished suggests the potential role of still other corepressors. In this context, we have recently found that Ikaros can interact with another corepressor, histone deacetylaserelated protein (HDRP) (data not shown). HDRP was first identified as an interactor of a key muscle regulatory protein, MEF2 (25), and was recently shown to bind CtBP and to repress the tk promoter in an HDAC-independent manner (41). In addition, another Ikaros interactor, Sin3, which usually represses using deacetylases, has also been shown to be capable of HDAC-independent repression; for this function, Sin3 appears to target components of the basal transcriptional machinery (21).
An enigma resulting from these studies is the basis for why repression of the tk versus the AdML promoters requires Ikaros to utilize two different repression strategies. If histone deacetylation is involved in repression through effecting DNA compaction, one would have expected HDAC recruitment to repress all promoters. One possible explanation for a promoter-selective function of HDACs is that the promoter context, defined by the other trans-acting factors bound to it, may only permit HDAC recruitment on a promoter like AdML but not one like tk. Thus, selective recruitment of co-factors by a DNA binding factor, influenced by its binding context, may underlie a transcription factor's promoter-specific transcriptional functions.
Thus far, CtIP has been shown to interact with two tumor suppressors, BRCA1 (29,34) and Rb (28). We have previously shown that dysregulation of Ikaros expression causes rapid development of leukemias and lymphomas (10). CtIP interactions with Ikaros may be involved, in part, in regulating the tumor suppressor function of Ikaros. The availability of the Ikaros-CtIP interaction mutant will allow this hypothesis to be tested. In conclusion, in this report we have presented several lines of evidence to show that Ikaros can function as a deacetylase-independent repressor in addition to its ability to repress through the recruitment of histone deacetylases (Fig. 6). This is a significant step forward in the attempt to molecularly dissect the workings of this key hemo-lymphoid regulator.
FIG. 6. Model summarizing repression strategies of Ikaros. A, in this scenario, Ikaros recruits HDACs through Mi-2␤ and/or Sin3 proteins to a promoter. This recruitment is expected to create a compact chromatin configuration that is not conducive for transcription. B, in this alternate scenario, Ikaros recruits CtBP, CtIP, and Rb to a promoter, all of which can interact with components of the general transcriptional machinery. Such interactions may underlie the HDAC activity-independent mode of Ikaros repression. The corepressors that Ikaros recruits may be dictated by the promoter context.