B-ind1, a Novel Mediator of Rac1 Signaling Cloned from Sodium Butyrate-treated Fibroblasts*

Sodium butyrate is a multifunctional agent known to inhibit cell proliferation and to induce differentiation by modulating transcription. We have performed differential display analysis to identify transcriptional targets of sodium butyrate in Balb/c BP-A31 mouse fibroblasts. A novel butyrate-induced transcript B-ind1 has been cloned by this approach. The human homologue of this transcript contains an open reading frame that codes for a protein of 370 amino acids without known functional motifs. In transfected cells, the B-ind1 protein has been found to potentiate different effects of the small GTPase Rac1, such as c-Jun N-terminal kinase activation and transcriptional activity of nuclear factor κB (NF-κB). In addition, we have demonstrated that B-ind1 forms complexes with the constitutively activated Rac1 protein. To investigate the role of B-ind1 in Rac1 signaling, we have constructed several deletion mutants of B-ind1 and tested their ability to affect the activation of NF-κB by Rac1. Interestingly, the fragment encoding the median region of human B-ind1 acted as a dominant-negative variant to block Rac1-mediated NF-κB activity. These data define B-ind1 as a novel component of Rac1-signaling pathways leading to the modulation of gene expression.

Sodium butyrate is a multifunctional agent known as an efficient inhibitor of cell proliferation, an inducer of differentiation in numerous cell lines and apoptosis in certain cell types (1,2). Sodium butyrate exerts most of its biological effects through a reversible inhibition of histone deacetylases (HDAC), 1 resulting in histone hyperacetylation and in selective changes in gene expression. Effects similar to those of sodium butyrate are also induced by structurally unrelated inhibitors of HDAC such as trichostatin A (3). The acetylation state of core histones plays a critical role in the regulation of transcription. In addition, a DNA butyrate-response element conferring the ability of sodium butyrate to induce gene expression has been identified (4,5). However, sodium butyrate also inhibits the expression of several genes including c-myc (6) and cyclin D1 (7). p21 WAF1 , a cellular inhibitor of cyclin-dependent kinase and of the progression in G 1 phase of the cell cycle, is induced by sodium butyrate as well as by trichostatin A (8,9). We have previously reported that in the mouse fibroblast cell line BP-A31 sodium butyrate inhibits the hyperphosphorylation of the tumor suppressor retinoblastoma protein via a transcriptiondependent mechanism (10). However, sodium butyrate induces only a small increase in the level of p21 WAF1 transcript in this cell line, suggesting that other butyrate-induced gene products may be involved in its antiproliferative action.
Van Lint et al. (11) showed that only a small subset of gene promoters is sensitive to inhibition of histone deacetylation by trichostatin A. Thus, cloning of transcripts induced by inhibitors of HDAC may facilitate the identification of effectors responsible for the antiproliferative and/or differentiating actions of this class of drugs. Using the differential display technique (12), we cloned the cDNA of a new gene transcript induced by sodium butyrate in the mouse fibroblast cell line BP-A31. The protein encoded by this transcript (designated B-ind1 for butyrate-induced 1) does not affect cell proliferation and apoptosis but does appear to participate in the control of the transcriptional activity of NF-B that is stimulated by sodium butyrate. In this context, we show that B-ind1 acts downstream of the small GTPase Rac1 in the biochemical pathway leading to NF-B activation. B-ind1 also collaborates with the activated form of Rac1 to induce c-Jun N-terminal kinase (JNK) activity. Taken together, these data implicate B-ind1 as a novel component of the sodium butyrate and Rac1-signaling pathways that influence the regulation of gene expression.

EXPERIMENTAL PROCEDURES
Cell Culture and Transfection-The chemically transformed mouse fibroblast BP-A31 cell line was maintained in minimal essential medium ␣ supplemented with 6% fetal calf serum. For synchronization in early G 1 phase, the cells were cultivated for 1 or 2 days in serum-free minimal essential medium ␣ supplemented with 2.5 M FeSO 4 . Simian COS-7 fibroblast-like cells were cultured in Dulbecco modified Eagle's medium with 10% fetal calf serum.
For transient transfection, COS-7 cells (5 ϫ 10 5 cells/dish) were seeded in 60-mm Petri dishes and transfected the following day with 5 g of DNA using the LipofectAMINE plus kit (Life Technologies, Inc.). The transfection was stopped after 3 h by switching to standard growth medium. Cell extracts were collected 36 h later for kinase assays, Western blot analysis, or immunoprecipitation.
For the luciferase assay, 10 5 cells/dish were seeded in 35-mm Petri dishes and transfected the following day with 1.5 g of DNA. The transfection was stopped after 3 h by switching to standard growth medium. Cell extracts were collected 36 h later. Luciferase activities were measured using the dual luciferase kit (Promega), and firefly luciferase activity in each sample was normalized with Renilla luciferase activity. Light emission was measured for 12-s intervals using a luminometer (Berthold). * This work was supported in part by INSERM and Association pour la Recherche sur le Cancer Grant 9835. 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 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) Z97207 and AJ271091. Plasmids-The (Ig)3-conaluc, conaluc (13), and Rac Q61L (14) constructs have been described previously. Flag-Rac Q61L and Myc-RacV12 were kindly provided by Drs. G. Bokoch and A. Hall, respectively. Flag-JNK1 was a gift from Dr. R. Davis. GST-Jun-(1-79) was expressed in Escherichia coli as described previously (15).
RNA Analysis-Total cellular RNA was extracted by the LiCl/urea method, fractionated by formaldehyde/agarose gel electrophoresis, and analyzed by Northern blotting. The blots were hybridized with the murine B-ind1 cDNA fragment obtained as described below and labeled with the multiprime kit (Amersham Pharmacia Biotech).
Cloning of B-ind1-Total RNA extracted from BP-A31 cells treated with or without 6 mM sodium butyrate in the presence or absence of serum was reverse transcribed in reaction mixtures containing the one-base-anchored oligo(dT) primer (H-T11 A). PCR was performed on the reverse transcription reaction products with the oligo(dT)-anchored primer and the arbitrary primer H-AP3 (5Ј-AAGCTTTGGTCAG-3Ј) (GenHunter Corp.) in the presence of [␣-33 P]dATP. The PCR products were resolved on a 6% acrylamide/urea gel and subjected to autoradiography. The cDNA corresponding to autoradiographic bands overexpressed in butyrate-treated cells but not in serum-treated cells were eluted from the gel, amplified, and subcloned in pCR-TRAP cloning vector (GenHunter Corp.). The murine sequence of B-ind1 was cloned by 5Ј rapid amplification of cDNA ends (RACE) PCR (CLONTECH Marathon) using Taqϩ polymerase (Stratagene). This cDNA (Gen-Bank™/EBI accession number Z97207) was used as a probe for Northern blotting and for screening a human brain cDNA library. The corresponding insert was sequenced and found to encode a 40-kDa protein. cDNA encoding either full-length hB-ind1 or truncated proteins were generated by standard PCR methods using specific primers containing suitable restriction sites on the 5Ј and 3Ј ends. Amplified products were cloned into the expression vector pCDNA3 His, in frame with the N-terminal X-press tag (Invitrogen).
Immunoprecipitation of Rac and Western Blot Analysis-Cells were lysed 36 h after transfection in 500 l of Rac lysis buffer (20 mM Tris, pH 7.5, 2 mM MgCl 2 , 0.5% Nonidet P-40, 150 mM NaCl, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 2 g/ml aprotinin, 2 g/ml leupeptin, and 1 mM sodium vanadate) per dish. After centrifugation at 13,000 rpm for 5 min at 4°C, equal concentrations (300 g) of protein were immunoprecipitated (1 h at 4°C) with 1 g of Flag monoclonal antibody M2, mixed with 30 l of protein G-Sepharose, and incubated for 1 h. Immune complexes were collected by centrifugation and washed 5 times in Rac lysis buffer. Final pellets were suspended in 40 l of Laemmli sample buffer. The samples were separated by SDS-polyacrylamide gel electrophoresis on a 12% (w/v) polyacrylamide gel and electrotransferred onto Hybond-ECL nitrocellulose membranes. Membranes were blocked in Tris-buffered saline (10 mM Tris, pH 8, 150 mM NaCl containing 0.05% Tween 20) and 5% nonfat dry milk for 1 h. The blots were then incubated overnight in blocking solution at 4°C with the Omni-probe antibody (D-8, Santa Cruz Biotechnology). Thereafter, blots were washed in Tris-buffered saline and incubated with horseradish peroxidase-conjugated antibody in blocking solution for 1 h. The blots were visualized by ECL (Amersham Pharmacia Biotech).

RESULTS
Cloning of hB-ind1 Protein-Differential display PCR was used to identify proteins induced by sodium butyrate (12). Serum-starved quiescent BP-A31 murine fibroblasts were incubated for 6 h in the absence or presence of sodium butyrate, and differential display PCR was performed with different primer combinations. With 3 one-base oligo dT-anchored primers and 4 arbitrary primers (16) we detected cDNA in which expression was up-regulated by sodium butyrate in the presence or absence of serum (data not shown). To confirm these results, bands corresponding to differentially expressed cDNA were excised from the gel, and the cDNA was amplified by PCR using the combination primers described above for differential display and then used as probes in Northern blot analysis. This report focuses on the functional characterization of the cDNA designated as B-ind1.
The PCR-derived probe corresponding to B-ind1 hybridized with a single band of 3.5 kilobases. This RNA was weakly expressed in quiescent BP-A31 fibroblasts and was induced by sodium butyrate in the absence or presence of cycloheximide ( Fig. 1A) indicating that the induction of B-ind1 expression does not require de novo protein synthesis. Analysis of the expression pattern of B-ind1 revealed that the mRNA was expressed in all mouse tissues examined (Fig. 1B). A strong signal was detected in the testis, kidney, brain, and liver whereas much weaker signals were observed in the skeletal muscle, spleen, and heart.
To obtain the entire coding sequence, we screened a human brain cDNA library with a partial cDNA clone. Analysis of the complete cDNA clone (hB-ind1) revealed an open reading frame that encoded a predicted protein of 370 amino acids (GenBank/ EBI accession number AJ271091). A search of the data bases using PROSITE revealed no known functional motifs although significant homologies outside of the active site with the putative tyrosine phosphatase PTPLA cloned from mouse and human (17) and with protein coding sequences from organisms as distant as Caenorhabditis elegans and yeast were identified. In addition, hB-ind1 exhibits homology with the human p23 (18) as well as related proteins from other species including chicken, yeast, and C. elegans.
Effect of hB-ind1 on Sodium Butyrate-induced Activation of NF-B-Because sodium butyrate is involved in the control of cell growth and apoptosis, we first examined whether any of these processes might be affected by hB-ind1. In COS-7 cells, using stably expressed hB-ind1 as well as hB-ind1 expressed under transient transfection conditions, we were unable to detect any effect of hB-ind1 on cell proliferation and apoptosis (data not shown).
Previous studies have shown that sodium butyrate induces FIG. 1. Cloning of B-ind1. A, induction of B-ind1 mRNA in BP-A31 cells by sodium butyrate. Serum-deprived BP-A31 cells were incubated for 6 or 9 h with sodium butyrate (NaBut, 6 mM) or cycloheximide (CHX, 1 g/ml), or both. Total RNA (20 g) was analyzed by Northern blotting and hybridized with 32 P-labeled murine B-ind1 cDNA. Equal loading of RNA was confirmed by staining with ethidium bromide. B, B-ind1 mRNA levels in mouse tissues are shown. Total RNA (20 g) from different mouse tissues was analyzed by Northern blotting with 32 Plabeled murine B-ind1 cDNA. Equal loading of RNA was confirmed by staining with ethidium bromide. the activity of NF-B (19), a transcription factor implicated in the control of many cellular processes including inflammatory response, differentiation, cell proliferation, and apoptosis (20). To investigate the potential role of hB-ind1 in sodium butyratedependent activation of NF-B activity, we studied the expression of a luciferase reporter plasmid placed downstream of three tandem NF-B binding sites of the IgG gene and a conalbumin basal promoter, (Ig)3-conaluc. In these experiments, we used COS-7 cells, which are characterized by high expression of transfected genes and a low, butyrate-insensitive content of endogenous B-ind1 mRNA (data not shown). Transient transfection of the promoter (Ig)3-conaluc into COS-7 cells resulted in low basal levels of transcription in untreated cells and an activation of 800-fold with sodium butyrate treatment. In cells transfected with hB-ind1, we observed an additional increase in sodium butyrate induction of (Ig)3-conaluc promoter activity (Fig. 2), raising the possibility that B-ind1 may play a role in sodium butyrate-mediated NF-B activation.
hB-ind1 Potentiates Rac1-induced NF-B Activation-NF-B activity is induced by members of the Rho family of small GTPases (Rho, Rac1, and Cdc42) (21). These proteins function as binary switches by cycling between the active GTPbound state and the inactive GDP-bound state. Constitutive functional activation is achieved by mutations in amino acid positions 12 or 61 analogous to those of p21 ras (22). We investigated whether hB-ind1 may participate in the pathway initiated by small GTPases and leading to NF-B activation. To test this possibility, COS-7 cells were transfected with (Ig)3-conaluc together with an expression vector coding for the activated form of Rac1 (Rac Q61L) in the presence or absence of hB-ind1. As expected, expression of the activated form of Rac1 increased NF-B activity. Interestingly, the effect of Rac1 on the transcriptional activity of NF-B was potentiated in cells cotransfected with hB-ind1 (Fig. 3, A and B). Similar results were obtained with the HeLa cell line (data not shown).
To further investigate the functional importance of B-ind1 in Rac1 signaling and to identify the domains involved, we constructed several deletion mutants of hB-ind1 and tested their ability to affect the activation of NF-B by Rac1. Two truncation mutants, fragments N (amino acids 1-199) and C (amino acids 197-370), failed to potentiate Rac1-induced NF-B acti-vation. Moreover, cotransfection of the median fragment M (amino acids 101-277) inhibited the Rac1-induced NF-B activation (Fig. 3B). When increasing levels of the M mutant expression vector were cotransfected, we observed a dose-dependent inhibition of Rac1-induced activation of NF-B (Fig. 3C). These results strongly suggest that the M fragment acts as a dominant-negative variant to block Rac1-mediated NF-B activity.
hB-ind1 Potentiates Rac1-induced JNK Activation-In addition to NF-B, the Rho family of GTPases controls at least two other transduction pathways, one regulating the mitogen-activated protein kinase/JNK pathway (23) and the second controlling the actin cytoskeleton organization (24). We searched to establish whether expression of hB-ind1 could also potentiate  3. hB-ind1 potentiates Rac1-induced NF-B activation. A, COS-7 cells were transfected with (Ig)3-conaluc (0.5 g) or conaluc together with an empty expression vector or with hB-ind1 (0.5 g) or Rac Q61L (0.5 g) expression vectors alone or in combination. After 36 h, luciferase activities were measured and normalized using the pRL-SV40 reporter as described in Fig. 2. B, COS-7 cells were transfected with (Ig)3-conaluc (0.5 g) and empty expression vector (0.5 g) or expression vectors of Rac Q61L alone or with 0.5 g of hB-ind1 wild-type (WT) or fragments as indicated: N, amino acids 1-199; C, amino acids 197-370; and M, amino acids 101-277. After 36 h, luciferase activities were measured and normalized using the pRL-SV40 reporter as described in Fig. 2. C, COS-7 cells were transfected with (Ig)3-conaluc (0.4 g) together with the empty expression vector or expression vectors of Rac Q61L (0.4 g) alone or with increasing amounts of hB-ind1 M mutant (0.2, 0.4, and 0.8 g). After 36 h, luciferase activities were measured and normalized using pRL-SV40 reporter as described in Fig. 2. Data represent the mean Ϯ S.D. of triplicates (bars). The results are expressed as -fold induction (ratios between the relative luminescence units measured in a given transfection and the empty expression vector). Similar results were obtained in three independent experiments. the stimulation of JNK activity by Rac1. COS-7 cells were cotransfected with Flag-JNK1 and Rac Q61L expression vectors either in the presence or absence of hB-ind1, and JNK activity was determined in Flag immunoprecipitates using GST-c-Jun-(1-79) as substrate. As described previously, coexpression of Rac Q61L led to a significant increase in JNK activity of 3-fold. Expression of hB-ind1 failed to activate JNK but potentiated the Rac Q61L-mediated JNK activation by 5.6-fold. Similar results were obtained with another constitutively active form of Rac, Rac V12 (Fig. 4).
hB-ind1 Forms Complexes with Constitutively Activated Rac1-Our data on the regulation of gene expression by Rac1 are consistent with the possibility that hB-ind1 and the constitutively activated form of Rac1 may interact to produce transactivation of NF-B-dependent promoters. To test this possibility, COS-7 cells were transfected with Flag-Rac Q61L and X-press hB-ind1. Complexes were precipitated with anti-Flag antibody directed toward tagged Rac Q61L followed by immunoblotting with anti-X-press to monitor the presence of hB-ind1. Wild-type hB-ind1 was present in complexes precipitated with the Flag epitope of Rac Q61L (Fig. 5, lane 4). We conclude that hB-ind1 interacts directly or indirectly with the activated form of Rac1. To determine the basis of the inhibitory activity of the median truncation mutant M, we performed coimmunoprecipitation experiments in the same manner as for the wildtype hB-ind1. In these experiments we were unable to detect any interaction between the M mutant and Rac Q61L (Fig. 5,  lane 2). Immunoblotting analysis of whole cell extracts indicates that the mutant form M was expressed at higher levels than the wild-type form (Fig. 5, lane 6 versus 8), so that the observed difference between hB-ind1 wild type or M in terms of binding to the activated form of Rac1 was not due to a difference in cellular content. DISCUSSION There is strong evidence that the biological effects of HDAC inhibitors are mediated through their ability to directly induce or inhibit the expression of several genes. Our purpose was to search for such genes and study the cellular functions of the corresponding products. Analyzing the differential display PCR patterns obtained with 12 primer combinations, we found two transcripts whose induction by sodium butyrate in the mouse fibroblast cell line BP-A31 was confirmed by Northern blotting. One of these transcripts, B-ind1, was present in quiescent, nontreated cells and was induced by butyrate in a protein synthesis-independent manner. Analysis of the complete cDNA obtained by screening of a human brain cDNA library revealed an open reading frame coding for a protein of 370 amino acids.
One important finding of our study is the demonstration that hB-ind1 is involved in the signaling pathway mediated by the small GTPase Rac1. In particular, hB-ind1 potentiated the Rac1-induced transcription and JNK activation. Moreover, the median region comprising amino acids 101-277 acted as a dominant-negative variant to inhibit the induction of the NF-B-dependent promoter by Rac1. The mechanism of action of hB-ind1 may rely on the formation of complexes between Rac1 and its targets. Although the sequence of hB-ind1 reveals no Cdc42/Rac-interactive binding site, a distinctive domain contained in several previously identified Rac1 targets (25), the existence of protein complexes containing Rac1 and hB-ind1 was evidenced by coimmunoprecipitation experiments. Because other proteins without the Cdc42/Rac-interactive binding site associate with Rac1 (26), it is possible that hB-ind1 favors the maintenance of Rac1 complexes with its effector systems. This hypothesis is substantiated by the inhibition of the Rac signaling pathway by the median fragment M of hB-ind1, although this fragment did not form stable complexes with Rac1. The M fragment may act as a dominant-negative variant by competing with endogenous wild-type B-ind1 to sequester downstream targets required in Rac1 signaling.
In addition, the hB-ind1 protein displays homologies with p23 and related proteins from other species including chicken, yeast, and C. elegans. This suggests that hB-ind1 may be a member of a large family of proteins. Human p23 was first identified as a protein associated with the progesterone receptor (18) and was later shown to be in fact a heat shock (Hsp90)binding protein (27,28). Both p23 and its yeast homolog Sba1 (29) appear to act as Hsp90 cochaperones, but each possesses a chaperone activity of its own (30). p23 has been postulated to crystallize as a homodimer. 2 These observations lead us to speculate that hB-ind1 may form homodimers with a chaperone type activity toward Rac1. In that case, the dominantnegative action of the M fragment would result from the formation of M/wild-type B-ind1 dimers, depleting Rac1 of its chaperone.
The role of B-ind1 in the cellular effects of sodium butyrate is not clear. The B-ind1 transcript was present in all tissues examined; its induction by sodium butyrate in cultured cells is not an universal process, and its overexpression in transfected COS-7 cells did not induce cell growth arrest or apoptosis. In the transformed mouse embryo fibroblasts, the cell in which this transcript has been cloned, B-ind1 may participate in the initial differentiation program.
In conclusion, the strategy of cloning transcripts induced by HDAC inhibitors has enabled us to identify a novel protein that 2 D. Toft, personal communication. FIG. 4. hB-ind1 potentiates Rac1-induced JNK activation. COS-7 cells were transfected with Flag-JNK and either the empty expression vector or Rac Q61L, Myc-Rac V12, X-press-hB-ind1 expression vectors alone or in combination as indicated. After 36 h, cells were harvested. Top, lysates were immunoprecipitated with Flag monoclonal antibody M2 and assayed for JNK activity using GST-Jun as substrate. Bottom, anti-M2 immunoblotting of whole extracts confirmed that similar amounts of Flag-JNK were transfected in each sample. The results were quantified by PhosphorImager, and the ratio between 32 P-labeling (GST-Jun) and ECLϩ luminescence (Flag-JNK) is shown (Quantitation).
FIG. 5. hB-ind1 forms complexes with constitutively activated Rac1. COS-7 cells were transfected with expression vectors of X-press-hB-ind1 wild-type (WT) or mutant (M) alone or together with Flag-Rac Q61L as indicated. After 36 h, cells were harvested. Left panel, lysates were immunoprecipitated with Flag monoclonal antibody M2 and analyzed for the presence of hB-ind1 by Western blotting using the X-press antibody. Right panel, expression levels of X-press-hB-ind1 wild-type (WT) or mutant (M) in the cell extracts were verified by Western blotting using the X-press antibody.
is conserved over a wide range of eukaryotic organisms with tissue-specific expression related to specific functions of differentiated cells. The participation of this protein in signaling by the small GTPase Rac1 is an important although perhaps not unique aspect of its activity; B-ind1 and related proteins may have other cellular functions that remain to be discovered.