Regulation of Co-repressive Activity of and HDAC Recruitment to RIP140 by Site-specific Phosphorylation *

Receptor interacting protein 140 (RIP140) is a versatile transcriptional co-repressor that contains several autonomous repressive domains (RDs). The N-terminal RD acts by recruiting histone deacetylases (HDACs). In a comprehensive proteomic analysis of RIP140 by MS, 11 phosphorylation sites of RIP140 are identified; among them five sites are located in the N-terminal RD including Ser104, Thr202, Thr207, Ser358, and Ser380. The role of phosphorylation of RIP140 in regulating its biological activity and the underlying mechanism are examined using a site-directed mutagenesis approach. Mutations mimicking constitutive phosphorylation or dephosphorylation are introduced. The N-terminal RD phosphorylation, mediated by the mitogen-activated protein kinase (MAPK), enhances its repressive activity through increased recruitment of HDAC. Mutations mimicking constitutive dephosphorylation at Thr202 or Thr207 significantly impair its repressive activity and HDAC recruitment, whereas mutation at Ser358 only slightly affects its HDAC recruitment and the repressive activity. Consistently, mutations mimicking constitutive phosphorylation at either Thr202 or Thr207 convert RIP140 into a more potent repressor, which is less responsive to a disturbance in the MAPK system. Furthermore, constitutive phosphorylation at both Thr202 and Thr207 residues renders RIP140 fully repressive and strongly interacting with HDAC. The activity of this mutant is resistant to the MAPK inhibitor, indicating an essential role for Thr202 and Thr207 in MAPK-mediated modulation of RIP140 function. The study provides insights into the modulation of RIP140 biological activity through a specific cellular signaling pathway that augments phosphorylation at specific residues of RIP140 molecule and alters its cofactor recruitment.

Environmental factors in the extracellular milieu utilize signal transduction pathways to propagate their cues into gene expression (1)(2)(3). Transcriptional factors and their co-regulators are extensively modified at the post-translational level, which usually regulate the critical function and property of the protein (4 -9). Deciphering such modifications and relating them to the biological function is referred to as functional proteomics (10). A variety of post-translational modifications have been found that regulate protein functions, including phosphorylation, acetylation, methylation, glycosylation, ubiquitination, and sumoylation, etc. (11)(12)(13)(14).
Receptor interacting protein 140 (RIP140) 1 is a co-regulator for many transcription factors including nuclear receptors (15)(16)(17)(18)(19). Extensive studies have been conducted to examine the versatile activity of RIP140 in transcriptional regulation. It is known that RIP140 acts, primarily, as a transcriptional corepressor through different mechanisms (20 -22). RIP140 is recruited to nuclear receptors through its nine LXXLL motifs and a modified motif LXXML, where X can be any amino acids (23,24). With respect to its repressive activity, four autonomous repressive domains (RDs) are known. RD1 is located in the aminoterminal region (amino acids 1-495), RD2 and RD3 are located in the central portion (amino acids 336 -1006), and RD4 is located in the carboxyl-terminal region (amino acids 977-1161). These domains function through various mechanisms. The amino-terminal RD acts by recruiting histone deacetylases (HDACs) through an HDAC-interacting domain, which has been mapped to amino acids 78 -303 (22,25). The central region interacts with the carboxyl-terminal binding proteins (CtBP1 and CtBP2) (26). In terms of its physiological action, RIP140-null mice exhibit female reproductive defects (15). Further studies of these animals indicate that RIP140 could play an important role in the regulation of fat accumulation in adipose tissues (27).
Recently, we initiated a functional proteomic study of RIP140 expressed and purified from insect cells. Through extensive mass spectrometric analyses, we found RIP140 are extensively phosphorylated (28) and acetylated (29). To continue the functional proteomic endeavor, we took a systematic mutagenesis approach to uncover the function and the mechanism of actions of specifically modified residues of RIP140. This study reports our systematic studies of the functional role of phosphorylation on RIP140, specifically with respect to its role in the HDAC-mediated repressive activity of RD1.
We previously reported the identification of 10 phosphorylation sites on RIP140 purified from insect cells by LC-ESI-MS/MS analysis (28). These residues included Ser 104 , Thr 207 , Ser 358 , Ser 380 , Ser 488 , Ser 519 , Ser 531 , Ser 543 , Ser 672 , and Ser 1003 . One residue, Thr 202 , appeared ambiguous in our initial analysis, and a reanalysis of the MS/MS data enabled us to verify Thr 202 as an additional phosphorylation site on RIP140. We then determined the effects of phosphorylation on the biological activity of RIP140 and explored the mechanism underlying such an effect in this current study. RIP140 employs multiple repressive pathways including HDAC-and CtBP-triggered events for RD1 and RD2/3, respectively. We focus this initial functional proteomic study on the HDACmediated RD1 activity. We identified mitogen-activated protein kinase (MAPK)-mediated phosphorylation as the critical pathway for RD1-mediated repressive activity, which was caused by enhanced recruitment of HDAC by phosphorylated RIP140. Mutations mimicking either constitutive phosphorylation or dephosphorylation were generated to determine whether the effects were caused by altered charges brought about by phosphorylation on specific residues or the amino acid residues per se. With this systemic approach, we identified two residues, Thr 202 and Thr 207 , crucial for the repressive activity of RD1 and its ability to recruit HDACs. Furthermore, increased negative charge caused by phosphorylation on these two specific residues is essential.
Purification and Mass Spectrometric Analysis of RIP140 -To identify the phosphorylation sites on RIP140, we expressed the Histagged RIP140 in insect cells as a eukaryotic host for mammalian protein expression. The protein was purified by affinity column over Talon resin with 95% homogeneity. The details of the procedure for RIP140 purification was described previously (28). The purified RIP140 was subjected to tryptic digestion, and LC-ESI-MS/MS analysis was conducted as already described to identify the modified residues (28).
Analysis of Biological (Trans-repressive) Activity of RIP140 -The technique for culturing COS-1 cells, transfection experiments, and luciferase and lacZ assay were as described previously (23). Cells were transiently transfected by the calcium phosphate precipitation method with a mixture of pBD-GAL4-RIP140 full length/N-terminal, wild-type (WT)/mutant (Mut) (0.1 g) or pBD-GAL4 (0.1 g), GAL4tk-luciferase (0.5 g) reporter, and a CMV-lacZ internal control (0.05 g). Forty h post-transfection, cells were exchanged with fresh medium containing dextran charcoal (DCC)-treated serum and treated for 8 h with either 1 M chelerythrin chloride (protein kinase C (PKC) inhibitor), 0.5 M KN-93 (calcium-calmodulin-dependent protein kinase II (CaCal) inhibitor), 3 M PD98059 (MAPK inhibitor), or 1 M concentration of anisomycin (MAPK activator) (Calbiochem). Fortyeight h post-transfection, total cell extracts were collected and tested for luciferase and lacZ activity. The fold relative luciferase activity was calculated by normalizing relative luciferase units (RLU) activity found in the experimental samples to the RLU activity found in the pBD-GAL4 alone. Reported values are an average of three experiments with triplicate measurement taken in each experiment.
In Vitro Protein-Protein Interaction Assays (FLAG Pull Down)-3T3 cells were transfected with FLAG-RIP140 or CMV empty vector. Forty h post-transfection, cells were subjected to a 1 M concentration of MAPK activator (anisomycin) or a 3 M concentration of MAPK inhibitor (PD98059) for 8 h in Dulbecco's modified Eagle's medium containing DCC serum. At 48 h post-transfection, cells were harvested, and FLAG-RIP140 was immunoprecipitated by anti-FLAG antibody (Sigma) on protein G-agarose (Sigma) beads (22). The partially purified FLAG-RIP140 fusion protein was incubated with [ 35 S]methioninelabeled HDAC3 protein for pull-down reaction as described previously (23). The specific bound proteins were released by resuspending beads in 20 l of SDS loading buffer, divided in equal amounts, and resolved by SDS-PAGE. One gel was subjected to Western blot analysis probed with anti-FLAG antibody followed by horseradish peroxidase-conjugated secondary antibody and then was detected with ECL (GE Healthcare), which constituted the input. Another gel was fixed, dried, and exposed to a phosphoimager screen (Molecular Dynamics) overnight to detect labeled HDAC protein and visualized by autoradiography. Densitometric analysis was done for labeled HDAC and input. Intensity of labeled HDAC was normalized to input.
His Pull Down-The expression vector for His-RIP140 was expressed and purified from insect cells in presence of phosphatase inhibitor that ensured phosphorylated RIP140 (28). This purified protein was then dephosphorylated by alkaline phosphatase (Roche Diagnostics; 1 unit per 5 g of protein at 37°C for 1 h) treatment. Both phosphorylated and dephosphorylated protein was pulled down by Metal Talon affinity resin (BD Biosciences) pre-equilibrated by extraction buffer (50 mM sodium phosphate buffer, pH 7.0, 300 mM NaCl) and then washed extensively with extraction buffer followed by binding buffer and then tested for interaction with [ 35 S]methionine-labeled HDAC3 protein. For input, Western blot analysis was probed with anti-His antibody (Upstate Biotechnology).
In Vivo Protein-Protein Interaction Assays (Mammalian Two Hybrid Test)-COS-1 cells were maintained as described previously (23). To test the interaction of RIP140 full-length/N-terminal, WT/Mut constructs with HDAC3, cells were co-transfected with pBD-GAL4-RIP full-length/N-terminal, WT/Mut (0.1 g) as bait and pAD-GAL4-HDAC3 (0.1 g) as prey. All co-transfections included GAL4-tk-luciferase (0.5 g) reporter and a CMV-lacZ internal control (0.05 g). Forty h post-transfection, cells were exposed to either 1 M MAPK activator (anisomycin) or 3 M MAPK inhibitor (PD98059) for 8 h in Dulbecco's modified Eagle's medium containing 10% DCC-treated fetal bovine serum medium. Forty-eight h post-transfection, total cell extracts were collected and tested for luciferase and lacZ activity. The fold relative luciferase activity was calculated by normalizing RLU activity found in the experimental samples to the RLU activity found in the pBD-GAL4-RIP140 full-length/N-terminal WT and pAD-GAL4-HDAC3 co-transfection. Reported values are an average of three experiments with triplicate measurement taken in each experiment.

RESULTS
Verification of Phosphorylation at Thr 202 of RIP140 -A MS/MS analysis of tryptic phosphopeptides was carried out to identify the modified residues. Previously, we reported 10 phosphorylation sites including Ser 104 , Thr 207 , Ser 358 , Ser 380 , Ser 488 , Ser 519 , Ser 531 , Ser 543 , Ser 672 , and Ser 1003 (28). Phosphorylation on Thr 202 appeared ambiguous according to the initial analysis of the MS spectrum. To clarify the ambiguity, we reanalyzed the data to verify Thr 202 phosphorylation. In the total ion chromatogram, the tryptic monophosphopeptide spanning residues 199 -212 (SGPTLPDVTPNLIR) eluted as a doubly charged precursor ion at m/z 780.38 (precursor mass 1558.76 Da) at 47.36 min. Product ion analysis of this precursor analysis suggested that some species of the peptide were phosphorylated at Thr 202 , whereas some were phosphorylated at Thr 207 . The MS/MS data clearly showed an 80-amu delta mass shift at y6, y7, y8, and y9 consecutively. The b5 and b6 ions at m/z 456.22 and m/z 533.26 (indicated as b6** in Fig. 1) were identical to those of the unmodified peptide. This clearly indicated Thr 207 phosphorylation. Because the peptide precursor ion was attributed to monophosphorylation and eluted as a single peak in the liquid chromatogram, we thought another phosphorylation site was unlikely to be present. However, upon reexamining the product ion of the precursor ion, we found a relatively high abundant y9 ion (indicated as y9* in Fig. 1) at m/z 1024.6 identical to the unmodified peptide. In addition, the y ions caused by loss of ammonia and H 2 O for this y9 ion were found. This indicated that Thr 207 was not phosphorylated in some species of the precursor peptide. Furthermore, the b2 and b3 ions corresponded to the unmodified peptide. Thus the possibility of Ser 199 phosphorylation was excluded. The doubly charged a8 ion showed a 40-amu shift, signifying Thr 202 phosphorylation. Finally, the protein phosphorylation sites prediction software NetPhos (www.expasy.org) that assisted in the analysis of RIP140 also showed both Thr 202 and Thr 207 to be potential sites for phosphorylation, Therefore, we conclude that Thr 202 can also be phosphorylated in vivo.
Phosphorylation Sites and Corresponding Kinases-All the 11 MS-confirmed phosphorylation sites of RIP140 were compared with the consensus motifs of all known protein kinases (31) as listed in Table I. The N-terminal RD contains five phosphorylation sites. Thr 202 and Thr 207 are potential sites for MAPK phosphorylation, Ser 358 can be phosphorylated either by MAPK or PKC, and the other two at Ser 104 and Ser 380 were unmatched to consensus sequences of known kinases. The remaining six sites in the central and C-terminal RD are potential targets for MAPK, PKC, or CaCal. We then explored the role of these kinases in regulating the repressive activity of RIP140.
Effects of MAPK/PKC/CaCal Inhibitors on Trans-repressive Activity of Full-length RIP140 -To evaluate the intrinsic repressive activity of RIP140, a standard trans-repressive assay was conducted where the full-length RIP140 was fused to a GAL4-BD and analyzed using a GAL4-responsive reporter in a standard cellular system, the COS1 cells. Inhibitors of MAPK, PKC, and CaCal were used to determine the effects of these kinase pathways on the trans-repressive activity of RIP140. The fold relative luciferase activity was evaluated to determine the effect on trans-repression. As shown in Fig. 2, both PKC and CaCal inhibitors have no effect on the trans-repressive activity of RIP140, whereas the specific MAPK inhibitor significantly relieves the repression, suggesting a role for MAPKinduced phosphorylation in mediating the repressive activity of RIP140. Our previous study has identified HDAC recruitment as the primary repressive mechanism of RIP140 through RD1 (22,25). Therefore we further examined the effects of MAPK-induced phosphorylation on the recruitment of HDAC by RIP140.
Effect of Phosphorylation on the Recruitment of HDAC-To examine the direct effect of phosphorylation on HDAC recruitment, we employed an in vitro pull-down assay that directly monitors protein-protein interaction. In vivo phosphorylated or hypophosphorylated RIP140 was prepared by transfecting FLAG-RIP140 into a mammalian cell line that has a full capacity for MAPK signaling, 3T3, followed by an 8-h pulse of 1 M concentration of MAPK activator (anisomycin) or 3 M concentration of MAPK inhibitor (PD 98059) prior to harvesting the cells. The phosphorylated or hypophosphorylated FLAG-RIP140 was partially purified through anti-FLAG-coupled protein G beads (Sigma) and incubated with in vitrotranslated, 35 S-labeled HDAC3. To monitor the constant amounts of input protein, samples were analyzed on a Western blot using anti-FLAG antibody (Fig. 3A, bottom). As shown in Fig. 3A, the MAPK activator marginally enhances interac- Eleven phosphorylation sites of RIP140 were identified by LC-ESI-MS/MS analysis (28). Sequences to these sites were compared to consensus sites of all known protein kinase (34). Correspondence of these sequences to the consensus motifs suggest kinases having a potential role to phosphorylate these residues. The residues have been segregated on the basis of RDs of RIP140.

Residue
Sequences a Sequences adjoining to the residues were compared to consensus motifs of known kinases. tion of RIP140 with HDAC, whereas the MAPK inhibitor drastically reduces the level of HDAC3 recruited to RIP140 (3-fold), suggesting hypophosphorylation of RIP140 impairs its HDACrecruiting ability. The marginal effect of the MAPK activator on RIP140 prepared from this cell line with regards to its HDAC recruitment is consistent with the notion that the MAPK system is constitutively active in the 3T3 cellular background.
To provide further direct evidence, phospho-RIP140 (His-tagged) was purified from insect cells under a condition that stabilizes protein phosphorylation (28). The dephosphorylated state was achieved in vitro by treating this protein preparation with alkaline phosphatase. These two protein preparations were analyzed also in a His pull-down assay by incubating with in vitro-translated, 35 S-labeled HDAC3. As shown in Fig. 3B, ϳ2.5folds more HDAC was recruited to the phospho-RIP140 than to the dephosphorylated protein. The input was also monitored as shown at the bottom of this figure. From these two proteinprotein interaction experiments, it can be concluded that phosphorylated RIP140 more effectively recruits HDAC.
Effects of Phosphorylation on the Trans-repressive Activity of RIP140-N-terminal RD-Because the HDAC-mediated repression involved, primarily, the N-terminal RD1 domain, we thus focused our attention on the RD1 to dissect the roles of specifically phosphorylated residues within this domain. A systematic approach was taken by mutating the five residues independently and in combination within this domain that can be phosphorylated. Furthermore, to test whether the residue itself or the increased site-specific negative charge (30) brought about by phosphorylation on that particular residue triggered the effects on the repressive activity of RD1, both dominant negative and constitutive positive mutations were made in the context of the GAL4 fusion that can be analyzed readily using the GAL4 reporter system. The dominant negative mutations (CN) were made by changing the residue to Ala, whereas the constitutive positive (CP) mutations were made by changing the residue to Glu.
To dissect the effects specifically on the property of RD1, the first series of CN mutants were examined in the context of GAL4-RD1 fusion. As shown in Fig. 4A, in the presence of the MAPK inhibitor (thus the hypophosphorylated state) the repressive activity of WT RD1 is reversed (from 0.2 to 0.7 relative trans-activation). A single mutation at either Thr 202 or Thr 207 renders RD1 much less repressive in the presence of the MAPK activator, whereas the single mutant, at either Ser 104 or Ser 380 , behaves almost like the WT. The repressive activity of the Ser 358 single mutant is also slightly affected, although to a much lesser extent than the Thr 202 or Thr 207 single mutant. Importantly, double mutation at both Thr 202 and Thr 207 completely abrogates the MAPK-mediated transrepression because the same level of activity is detected in the presence of the MAPK activator or inhibitor. The triple mutant at Thr 202 /Thr 207 /Ser 358 behaves essentially like the Thr 202 /Thr 207 mutant. A mock mutation at residue Ser 315 , serving as a negative control, has no effect on the repressive activity as compared with the WT. This result suggests that Thr 202 and Thr 207 are crucial for the MAPK-enhanced transrepressive activity of RD1 and that Ser 358 also modulates the RD1 activity, but to a lesser extent. To monitor the efficiency of the production of these Mut proteins, the expression of transfected expression vector was examined using RT-PCR (to monitor RNA production) and Western blot (to monitor protein production) as shown in Fig. 4B, which verifies the production of a comparable level of mutant proteins.
Effects of Phosphorylation on Interaction of RD1 with HDAC3-To provide a mechanistic insight, the effects of these mutations were evaluated for their effects on the interaction between N-terminal RD and HDAC3 in both phosphorylated and hypophosphorylated conditions using a two-hybrid interaction test as shown in Fig. 5. As predicted, interaction between the WT RD1 and HDAC3 occurs effec-tively in the presence of the MAPK activator, whereas the MAPK inhibitor abrogates almost completely this interaction. Single mutation at either Ser 104 or Ser 380 exerts little effect. Interestingly, single mutation at either Thr 202 or Thr 207 dramatically affects the interaction in the presence of the MAPK activator. Consistent with the effect seen in the repressive activity of the Ser 358 mutant (Fig. 4), interaction of this mutant with HDAC is also less effective than the WT. More dramatically, double mutation at Thr 202 /Thr 207 or the triple mutation at Thr 202 /Thr 207 / Ser 358 almost completely abrogates this interaction. The mock mutation at Ser 315 again has no effect on the interaction. This result provides a mechanistic explanation for the effects of phosphorylation of RD1, specifically at Thr 202 and Thr 207 . Phosphorylation at these residues enhances the recruitment of HDAC by RIP140, thereby making it more repressive. Ser 358 also plays a role in the activity and HDAC recruitment of RD1; however, this residue is less critical than either Thr 202 or Thr 207 .
Role of Site-specific Phosphorylation on the Trans-repressive Activity and HDAC Recruitment of the Full-length RIP140 -The CN mutation studies reveal the essential functional role of two specific residues, Thr 202 and Thr 207 , in regulating the trans-repressive activity of RD1 and its ability to recruit HDAC. However, two relevant questions remain. First, this result does not allow conclusion to be made as to whether the effects are due to phosphorylation or because of changes in amino acid residues. Second, it is unknown if the effects of these RD1 domain mutations remain effective in the context of the full-length RIP140. One primary effect of protein phosphorylation is to impart the site-specific negative charge of the protein, which may subsequently alter its behavior such as interaction with other proteins or its enzyme activity (30). To test whether an increase in the site-specific negative charge of these residues, as predicted from phosphorylation, was responsible for the effects on trans-repressive activity and HDAC recruitment, we generated CP mutations by changing residues Thr 202 , Thr 207 , and Ser 358 singly or in combination into Glu. Furthermore, to evaluate whether these molecular changes can elicit a biological response in the context of the full-length RIP140, these mutations were generated in the backbone of the full-length RIP140. Fig. 6 shows the result of trans-repressive activity of the full-length WT RIP140 and its mutants, where the WT, as predicted, is sensitive to the MAPK inhibitor. Very interestingly, the single CP mutations, at either residue Thr 202 or Thr 207 , dramatically reduce their MAPK sensitivity in the transrepressive assay, indicating that the MAPK-enhanced repressive activity of RIP140 can be mimicked by simply increasing the acidity at Thr 202 or Thr 207 . The single CP mutant at residue Ser 358 exhibits a marginal effect, consistent with the result obtained in the context of the N-terminal RD (Fig. 4). Importantly, the double CP mutant Thr 202 /Thr 207 and the triple CP mutant Thr 202 /Thr 207 /Thr 358 remain fully repressive even in the presence of the MAPK inhibitor, supporting the notion that the increased site-specific negative charge imposed by the MAPK-elicited phosphorylation on Thr 202 /Thr 207 indeed is crucial for the repressive activity of RIP140. A CN triple mutant at Thr 202 /Thr 207 /Thr 358 completely loses its repressive activity irrespective of the status of the MAPK system. To monitor the efficiency of transcription and translation of these constructs, RT-PCR and Western blot analyses were conducted as shown in Fig. 6B where all the mutants are shown to be produced effectively.
The ability of these mutants to interact with HDAC was examined in the two hybrid interaction test as shown in Fig. 7. The WT RIP140 interacts effectively with HDAC, whereas the MAPK inhibitor drastically reduces the strength of this inter- action as predicted. Whereas all the single CP mutants retain their HDAC interaction potency in the presence of the MAPK activator (open bars) or without drug treatment (data not shown), their ability to interact with HDAC is affected by the MAPK inhibitor, but to a lesser extent than the WT RIP140. Importantly, the double CP mutant at Thr 202 /Thr 207 and the triple CP mutant at Thr 202 /Thr 207 /Thr 358 constitutively interact with HDAC even in the presence of the MAPK inhibitor, further supporting the notion that HDAC interaction is also critically regulated by the site-specific negative charge of these residues on RD1 of RIP140, i.e. Thr 202 and Thr 207 . Hence, the MAPK-elicited phosphorylation on these two residues increases the site-specific negative charge of RD1, thereby enhancing its ability to recruit HDAC to RIP140. As predicted, the triple CN mutant at Thr 202 /Thr 207 /Thr 358 can no longer interact effectively with HDAC even in the presence of the MAPK activator. DISCUSSION RIP140 is extensively modified post-translationally. Among 11 phosphorylated residues that are confirmed by the mass spectrometric study (28), 5 are located in the N-terminal RD that encodes an intrinsic repressive activity through recruiting HDACs. This study demonstrates for the first time that RD1 activity of RIP140 is regulated by MAPK-elicited phosphorylation on specific residues of RD1, i.e. Thr 202 and Thr 207 . Furthermore, the phosphorylation-elicited site-specific negative charge on these residues plays a critical role in both recruiting HDACs and conferring a repressive activity. Phosphorylation on Ser 104 or Ser 380 has no effect in these regards, whereas phosphorylation at residue Ser 358 has a marginal effect.
In our assays conducted in the cellular background of 3T3 and COS-1, the repressive potential remains essentially the same in the presence or absence of MAPK activator. Therefore, the MAPK pathway is constitutively active in these cell lines. Hence, the basal (without drug treatment) repressive activity resembles the repressive activity in the presence of the MAPK activator, whereas the repressive activity is dramatically affected by the addition of the MAPK inhibitor.
The HDAC-interacting surface of RIP140 is mapped to its N-terminal domain between amino acids 78 -303 (22,25). In this domain, Thr 202 and Thr 207 are directly involved in the enhanced recruitment of HDAC3 and the repressive activity. Presumably, negative charges brought about by phosphorylation on these residues play a critical role for RIP140 to make a direct contact with HDAC. For instance, charges on these two residues may directly control the binding affinity between RIP140 and HDAC. An interesting finding is with Ser 358 . This residue may not be directly involved in interaction with HDAC; however, its phosphorylation probably facilitates the interaction with HDAC by inducing a conformational change in the tertiary structure of this protein because of the increased negative charge on this site. However, this effect is only apparent in the context of the dissected RD1 (Figs. 4 and 5) because mutation at Ser 358 in the context of the full-length RIP140 exerts no significant effect on either HDAC recruitment or the repressive activity of the full-length protein (Figs. 6 and 7). Conceivably, the conformation of RD1 may be slightly different from its configuration in the context of the intact protein. It remains to be further examined whether Ser 358 phosphorylation has an effect on other aspects of RIP140. Similarly, mutation at either residue Ser 104 or Ser 380 has no effect on the HDAC-mediated repressive potential of RIP140. Whether phosphorylation at these residues plays a role in other aspects of RIP140 also remains to be determined. Alternative, we can not rule out the possibility that these sites can be nonspecifically modified by low affinity kinases in insect cells (28).
The MAPK/PKC/CaCal sites in the RD2 and RD3 domains do not seem to have any effect on HDAC-mediated RD1 activity because mutations at phosphorylated sites within RD1 alone can fully account for the MAPK-sensitive, HDACmediated repression of RIP140. Therefore, the role of phosphorylation in RD2 and RD3 remains to be examined.
Protein phosphorylation is known to be one of the most important intracellular signal transduction events (1)(2)(3)32). It has been estimated that approximately one-third of the mammalian proteins contain covalently bound phosphate groups, and many are subjected to regulation by multiple phosphorylation events (33,34). Recent investigation showed that nuclear co-regulator SRC-3, a known oncogenic factor in FIG. 7. Interaction of HDAC3 with RIP140 full-length (WT/Mut) constructs. A mammalian version of the two-hybrid system was used to examine in vivo interaction between HDAC3 and RIP140 full-length (WT and CP or CN Mut) constructs in COS-1 cells. The GAL4-BD-RIP140 full-length fusion and AD-GAL4-HDAC3 fusions were cotransfected pair-wise in COS-1 cells along with the reporter and a CMV-lacZ internal control. Luciferase units of mutant interaction with HDAC were normalized to WT interaction with HDAC. A site-specific negative charge imparted at Thr 202 and Thr 207 as mimicked by Thr-Glu mutations seems to be crucial for recruitment of HDAC. (Statistics: ❙ versus * (p Ͻ 0.03); † versus * (p Ͼ 0.05);ʈ versus ❙ (p Ͻ 0.03); double curved bars with horizontal bar versus ʈ (p Ͼ 0.05).) breast and prostrate cancer, is regulated by phosphorylation. The phosphorylated SRC-3 is eventually more oncogenic than the unmodified form (30). RIP140 knock-out studies have shown that RIP140, though not involved in the process of adipogenesis, regulates fat accumulation (27,35). RIP140 is also known to have a role in follicular development and is essential for oocyte release during ovulation (35). The functional proteomic studies presented here and the CP and CN mutants of RIP140 generated from these studies would provide powerful tools for further studies of the functional role of RIP140 in these biological systems where the second messenger systems such as the MAPK pathway could play an important role.