The retinoblastoma gene product (Rb) induces binding of a conformationally inactive nuclear factor-kappaB.

Nuclear factor-κB (NF-κB) regulates expression of several viral and cellular genes including the human immunodeficiency virus long terminal repeat, major histocompatibility complex class I, and interleukin 2Rα cytokine genes. Here we report that the retinoblastoma gene product (Rb) stimulates binding of the NF-κB p50 homodimer. The addition of Rb protein to an in vitro gel shift binding assay stimulated p50 binding greater than 10-fold. Interestingly, by analyzing NF-κB-dependent transcription activity in vitro, we demonstrate that Rb suppresses transcriptional activity of p50. Chymotrypsin analysis suggests that Rb induces a conformational change in the NF-κB-DNA complex, resulting in binding of a transcriptionally inactive complex. Finally, we demonstrate by coimmunoprecipitation analysis that the Rb-p50 complex is present in Jurkat cell extracts. Our results suggest that Rb may play an important role in regulation of NF-κB transcriptional activity.

. Studies have shown that c-rel and v-rel interact with TBP and TFIIB in vivo and in vitro.
The retinoblastoma susceptibility gene (Rb) is a member of a class of cellular genes referred to as tumor suppressor genes, anti-oncogenes, or recessive oncogenes (17). These proteins are associated with a subset of human cancers, including retinoblastoma, small cell lung cancer, osteosarcoma, and carcinoma of the bladder and breast, which are due to the frequent loss or mutational inactivation of the Rb gene (18,19). Rb is a ubiquitously expressed nuclear protein, the phosphorylation state of which changes during the cell cycle, with a role in regulating cellular proliferation and gene expression (20 -23). Phosphorylation of Rb at the G 1 -S transition allows the progression of cells into S phase and through the cell cycle. Sequences in the Rb protein, commonly referred to as the pocket (approximately 400 amino acids), are important for protein-protein interactions (24 -26) and contain homology with basal transcription factors TBP and TFIIB (27).
Rb interacts with, and negatively regulates, the cellular transcription factor E2F (28,29). Transient cotransfection experiments have demonstrated that an Rb expression plasmid inhibits E2F-dependent transcription, which is correlated with the ability of Rb to interact with E2F (30,31). Rb is physically associated with the E2F DNA-protein complex. It has been proposed that Rb binds to the E2F activation domain and blocks its function as a transcription factor. Independent studies have further suggested that Rb-E2F is an active repressive complex, which inhibits activation of other promoter elements when bound to the E2F sites (28). Constructs containing E2F sites cloned into a promoter with a TATA box and an ATF transcription factor site were tested for activation by ATF in the presence of E2F-Rb. The presence of E2F sites inhibited activation of this promoter regardless of the presence of ATF. This attributes a critical function to Rb since it is described as a regulator that affects promoter activity by switching a E2F site from positive to negative elements. The interaction of E2F and Rb is restricted to the G 0 -G 1 phase of the cell cycle since only underphosphorylated Rb forms complex with E2F. Subsequent phosphorylation of Rb disrupts the Rb-E2F interactions, allowing progression into S phase. Activation of E2F stimulates transcription of S-phase progression genes such as the E2F family, dihydrofolate reductase, thymidine kinase, thymidine synthase, ribonucleotide reductase, and c-myb.
The interaction of Rb with other transcription factors such as elf-1 (32), PU.1 (27), c-myc (33), and ATF (34) have been reported. Interestingly, interaction of Rb with elf-1 suppresses transcription similar to that seen with Rb and E2F. The functional consequence of the interaction of Rb with c-myc or PU.1 remains to be determined but appears to be distinct from that seen with E2F since Rb does not directly inhibit PU.1 or c-myc transcriptional activity in the absence of a binding site for E2F (32). In this report, we present experiments that demonstrate that Rb regulates NF-B p50 transcriptional activity.
Protein Expression and Purification-GST RB fusion or GST proteins were generated in Escherichia coli by growth of the transformed HB101 strain in Luria broth with 100 g/ml ampicillin. The cultures were grown to an absorbance of 0.6, followed by induction with 0.5 mM isopropyl-1-thio-␤-D-galactopyranoside for 3 h. The GST fusion proteins were obtained by sonication of the bacterial pellet in buffer containing 50 mM Tris, pH 7.5, 250 mM NaCl, 0.1% Nonidet P-40, 1 g/ml leupeptin, 1 g/ml aprotinin, and 100 g/ml phenylmethylsulfonyl fluoride at 4°C for 30 min. The extract was purified by centrifugation at 100,000 ϫ g for 30 min. The supernatant was collected and mixed with glutathione-Sepharose beads at 4°C. The GST proteins were eluted by the addition of 20 mM glutathione, pH 7.5, for 10 min, followed by an overnight dialysis in buffer D (20 mM Hepes, pH 7.5, 0.1 M KCl, 20% glycerol, 0.2 mM EDTA, 0.5 mM dithiothreitol, and 0.5 mM phenylmethylsulfonyl fluoride). The purified preparations were analyzed in SDSpolyacrylamide gel electrophoresis gels and then stained with Coomassie Brilliant Blue. The concentrations of the 27-kDa GST, 71-kDa GST RB (379 -792), and 86-kDa GST RB (379 -928) were estimated by comparing the intensities of the bands with those of a known bovine serum albumin standard. One unit is defined as 100 ng of protein. [ 35 S]Methionine-labeled NF-B p65, p50, and c-Rel were in vitro transcribed and translated in rabbit reticulocyte lysates using the TNT system (Promega Corp.). Protein production and yield was confirmed by SDSpolyacrylamide gel electrophoresis and autoradiography. NF-B p50 protein, purified from E. coli, was obtained from Promega. The concentration of this protein is given as gel shift units (gsu), 3.2 gsu/l or 69 ng/gsu.
In Vitro Transcription in Whole-cell Extracts-Templates were prepared by digestion of 100 g of plasmid DNA with 5-10-fold units excess of restriction enzymes for up to 2 h under buffer conditions recommended by New England BioLabs. After the termination of the digest, the DNA was subjected to two phenol:chloroform:isoamyl alcohol (50: 50:1) extractions and a subsequent ethanol precipitation. All incubations were done at 30°C. The in vitro transcription buffer contained 10 mM HEPES, pH 7.9, 50 mM KCl, 0.5 mM EDTA, 1.5 mM dithiothreitol, 6.25 mM MgCl 2 , and 8.5% glycerol. Two to ten g/ml of the HIV-1 LTR chloramphenicol acetyltransferase and AdML DNAs were linearized by digestion with EcoRI and BamHI, respectively. One hundred twentyfive to 500 ng of template were used per reaction. HeLa whole-cell extracts were prepared as described previously (35) and added to a final concentration of approximately 3.75 mg/ml (ϳ55 ng/reaction). The reactions also contained nucleoside triphosphates in water (500 M), [␣-32 P]UTP (400 Ci/mmol) (15 Ci), purified GST fusion protein (1.0 unit), and purified p50 (Promega) at various concentrations. Transcription reactions were terminated by the addition of 20 mM Tris-HCl, pH 7.8, 150 mM NaCl, and 0.2% SDS. 32 P-Labeled RNA was purified and analyzed on a 4% polyacrylamide-urea gel under denaturing conditions as described previously.
Coimmunoprecipitation Assay-Jurkat cells were stimulated with 50 ng/ml 12-O-tetradecanoylphorbol-13-acetate for 1 h. Whole-cell extracts were prepared, and immunoprecipitations were performed as follows. Cells were washed two times with phosphate-buffered saline and resuspended in 400 -600 l of lysis buffer containing 100 mM Tris-HCl, pH 7.5, 10 mM EDTA, 300 mM NaCl, and 0.1% Nonidet P-40. Aprotinin and leupeptin were added fresh at a final concentration of 1 g/ml. After 40 min of incubation, the cells were vortexed briefly, centrifuged at 14,000 rpm for 5 min, and the supernatant was retained for immunoprecipitation. Three hundred g of whole-cell extract was used in each assay. The immunoprecipitation was performed using equivalent amounts of anti-Rb, anti-p50, or rabbit antisera. After 3 h of incubation at 4°C, the antibodies were precipitated using protein A-protein G beads (previously blocked with 4% bovine serum albumin) for 1 h at 4°C. Beads and the complexes bound to them were washed two times with the lysis buffer. The immunoprecipitates were separated on a 4 -20% Tris-glycine gel, transferred onto a nitrocellulose filter, and immunoblotted with anti-Rb antibodies. After 30 min in blocking buffer, as described above, the blot was incubated with the Rb antibodies (1:500 dilution; Pharmingen) for 30 min at room temperature, then washed three times in blocking buffer. Peroxidase-labeled anti-mouse antibodies (1:10000; Amersham Corp.) was added to the primary antibodies. The blot was incubated with the secondary antibody for 30 min at room temperature, washed five times with 1 ϫ TNE (100 mM Tris, pH 7.5, 150 mM NaCl, and 10 mM EDTA), and treated with Amersham ECL Western blotting reagents. The image was developed using Amersham ECL Western blotting reagent and exposed to Kodak XAR autoradiography film.
Gel Shift Assay-The murine Ig-enhancer NF-B oligonucleotide 5Ј-GATCCAGAGGGGACTTTCCGAGAG-3Ј and the TNF-␤ sequence 5Ј-GATCCAGAGGGGCTTCCCCGAGAG-3Ј were Klenow-labeled with [ 32 P]dGTP (Amersham). The labeled oligonucleotides were desalted with G-25 spin columns (Boehringer Mannheim) and precipitated in ethanol. The probe was resuspended in 10 mM Tris, pH 7.5, and 1 mM EDTA. The gel shift reactions were performed in a volume of 20 l in gel shift reaction buffer (10 mM Tris-HCl, pH 7.5, 40 mM NaCl, 1 mM EDTA, and 1 mM dithiothreitol) with purified NF-B p50 (Promega), GST fusion purified protein, 0.5-2 ng (ϳ50,000 cpm) of labeled NF-B oligonucleotide, and 3 g poly(dI⅐dC/dI⅐dC) (Pharmacia) at room temperature for 20 min. The gel shift complexes were analyzed by acrylamide gel electrophoresis as described previously (39). Competition reactions were incubated with a 100-fold excess of unlabeled mutant or wild-type Ig NF-B oligonucleotide to identify a specific NF-B gel shift complex. The specificity of the NF-B gel shift complex was also demonstrated by adding 3 l of anti-p50 polyclonal antibodies (Santa Cruz Biotechnology, Inc.) to the p50 gel shift reactions, producing a supershift complex.

NF-B Protein and GST RB Fusion-Protein Binding Assays-For
GST binding assays, the [ 35 S]methionine-labeled NF-B proteins were incubated with approximately 0.5 g of either GST or GST Rb proteins in buffer D with 0.5% Nonidet P-40 and 0.5 mM dithiothreitol for 2 h at 4°C. Glutathione-Sepharose beads (100 l of 50% suspension) were added for an additional 2 h at 4°C with gentle mixing. The reactions were washed three times with the same buffer, followed by SDS-polyacrylamide gel electrophoresis and autoradiography.

RESULTS
The DNA Binding Activity of the NF-B p50 Homodimer Is Induced by the Rb Pocket-E. coli-purified GST Rb pocket (amino acids 379 -792) fusion protein was tested for its ability to regulate the DNA binding activity of NF-B p50. Approximately 100 ng of purified p50 was incubated with the Ig NF-B oligonucleotide probe. The specificity of the p50 homodimer complex was demonstrated by gel shift competition assay (Fig. 1A, lanes 1Ј, 2Ј, and 3Ј). The p50 gel shift complex was competed by the specific wild-type competitor but not the mutant competitor. Furthermore, the gel shift complex was supershifted with anti-p50 but not a control antibody (data not shown). The concentration of p50 used in the Rb induction assay was titrated to a level where a gel shift complex was minimal (Fig. 1A, lane 1). Increasing concentrations of GST and GST Rb were added to the p50 gel shift reactions. As the concentration of GST Rb increased, the gel shift activity of p50 increased until a saturation level was reached (Fig. 1A, lanes 3 , 5, 7, 9, 11, and 12). In lane 7, the protein concentration of Rb and NF-B is approximately equivalent. In control reactions with GST, the gel shift activity of p50 was not affected. The results presented in Fig. 1B further demonstrate that the p50 induction was due to the Rb domain of the GST-Rb fusion protein. An equal concentration of GST-p53 or bovine serum albumin control proteins did not induce p50 binding (Fig. 1B,  lanes 4 and 5). Several lines of evidence argue against the p50 induction being due to the oxidation-reduction potential of glutathione. (a) The control GST protein went through the same glutathione elution and dialysis, and no induction of p50 binding was observed with this protein. (b) Direct addition of glutathione at concentrations equivalent to the dialyzed protein failed to induce p50 binding (data not shown).
Rb (379 -792) Inhibits the p50-induced Transcription of HIV-1 LTR-In vitro transcription assays with a NFB-inducible template were performed to investigate the functional sig-nificance of Rb on p50 transcription activity. The HIV-1 promoter contains two copies of the nonpalindromic Ig NF-B binding site upstream of the transcription initiation site (Ϫ80 to Ϫ107). The linearized HIV-1 LTR template produces a runoff transcript of 325 bases, which was sensitive to the addition of ␣-amanitin ( Fig. 2A, lanes 1 and 2). Purified NF-B p50 activated HIV transcription in a dose-dependent manner ( Fig.  2A, lanes 3-7). The origin of the higher molecular weight RNA observed at the higher NF-B concentrations is not known. Transcription activity is quantitatively shown in Fig. 2B as a bell-shaped induction curve. In the presence of added p50, a range of 1-8.7-fold induction of basal transcription was observed. Specificity of NF-B p50 activation of HIV-1 LTR transcription was demonstrated by the addition of p50 antibody to the transcription reaction. Antibodies were added to the transcription reaction, where the highest induction of p50-activated transcription was observed. Upon the addition of anti-p50 antibody, but not the control IgG, the p50-activated transcription was inhibited, and only basal levels of transcription were detected (Fig. 2C). These results show that the activated transcription of HIV is due to p50.
The effect of Rb (379 -792) on the p50-induced transcription was analyzed (Fig. 2D). Control reactions with HIV-1 LTR and GST or GST Rb were included as controls (Fig. 2D, lanes 1 and  2). The addition of p50 to the control reactions with GST produced a bell-shaped induction curve (Fig. 2D, lanes 3-7). The difference in p50 titration curves seen in Fig. 2, A and D, are likely due to a difference in specific activity of the p50 used in the two experiments. When Rb (379 -792) was included in the p50-induced transcription assay, a suppression of the p50-induced transcription activity was detected (Fig. 2D, lanes 8 -12). These results are quantitatively demonstrated in Fig. 2E. To demonstrate that Rb does not nonspecifically repress transcription, GST Rb was added to transcription reactions with a control AdML template (Fig. 2F). Transcription from the AdML promoter was compared with reactions containing GST or GST Rb. No differences were observed in the transcription reactions of AdML alone or with GST or GST Rb (379 -792) (Fig. 2F,  lanes 1-3).
One interpretation of the transcription experiments might be that since Rb induces p50 binding, the bell-shaped p50 induction curve is simply offset in the presence of Rb. We, therefore, compared the transcriptional activity of equivalent p50 DNA binding activities in the absence or presence of Rb (Fig. 2G). Gel shift assays were performed with in vitro transcription reaction buffer (Fig. 2G). Under those conditions, equivalent p50-DNA binding activity was observed at 14 ng of p50 in the presence of GST and 3 ng of p50 in the presence of GST Rb. At 56 ng of p50, equivalent p50 binding was observed in the presence of GST or GST Rb (Fig. 2G). By comparing the percentage of p50 bound to the B site and the fold induction of transcriptional activity, we conclude that p50 is not able to activate transcription in the presence of Rb (Fig. 2G).
Rb Induces Conformational Change of NF-B p50 -The transcription analysis suggested that the Rb-induced NF-B complex was transcriptionally inactive. Previous studies by Fujita et al. (12) have demonstrated that NF-B p50 provides strong transcriptional activation only when adopting a chymotrypsin-resistant conformation. Therefore, we analyzed the conformation of p50 bound to the B motif in the presence of Rb using a chymotrypsin sensitivity assay. Two sets of gel shift reactions were studied and compared; the first contained p50 (Fig. 3A, lanes 1-5), and the second contained the Rb-induced p50 gel shift complex (Fig. 3A, lanes 8 -12). The p50 gel shift complexes were subsequently treated with chymotrypsin for various times. The reactions were stopped upon the addition of chymostatin, and the complex containing the labeled B motif was analyzed by electrophoretic mobility shift assay. Quantitative analysis of chymotrypsin digestion demonstrated that the control NF-B and Rb-induced NF-B complex were digested at different rates (Fig. 3A, lanes 1-5 and 7-11; Fig. 3B). For example, at the 10-min time point, the Rb-induced NF-B gel shift complex was reduced by approximately 55%. In contrast, no significant reduction in the amount of control NF-B complex was observed. The difference in chymotrypsin sensitivity is further evident if one compares the NF-B gel shift complex intensity at 0 and 15 min. Although there is 5-10-fold more complex in the Rb-induced control incubation, the signal at 15 min of digestion is equivalent between the two samples (Fig. 3A, lanes 1 and 5, 7 and 11). The chymotrypsin-sensitive nature of the Rb-induced NF-B complex correlates with decreased transcriptional activity.

Coimmunoprecipitation of Rb Protein with NF-B p50 in
Vivo-Rb induction of NF-B binding activity might be mediated through direct protein-protein interaction. Therefore, coimmunoprecipitation assays were performed with Jurkat whole-cell extracts. The whole-cell extract was incubated with anti-p50, anti-Rb, or rabbit IgG. The proteins from these immunoprecipitations were separated on a 4 -20% SDS-polyacrylamide gel, transferred to nitrocellulose, and immunoblotted with anti-Rb antibody. Rb protein was detected in immunoprecipitation reactions that contained anti-Rb or anti-p50 antibodies (Fig. 4A, lanes 1 and 2). In the control reaction, Rb failed to coprecipitate with preimmune rabbit antisera (Fig. 4A, lane 3). Because the IgG heavy chain migrates at ϳ50 kDa, we were unable to perform the reciprocal coimmunoprecipitation of p50 with Rb, i.e. precipitate the Rb protein and immunoblot with anti-p50.
The interaction between Rb and p50 was also demonstrated with in vitro GST binding assays. Glutathione-Sepharose beads containing equal quantities of GST or GST Rb (379 -792) proteins were incubated with in vitro translated p50. NF-B p50 was specifically precipitated in the presence of GST Rb (Fig. 4B, lane 2) but not in the presence of control GST protein. Similar binding assays with other members of the NF-B family, including p65 and c-rel, suggests that the Rb pocket interacts with the Rel homology domain (data not shown).

DISCUSSION
The activity of the NF-B/Rel family of transcription factors is regulated at multiple levels. In this study, we discovered a novel interaction between the cellular transcription factor NF-B p50 and the product of the retinoblastoma susceptibility gene, Rb. This interaction is specific and can be found in both whole-cell extracts of constitutively expressing NF-B cell lines and in vitro using GST binding assays with translated NF-B proteins. Our experiments further demonstrate that Rb stimulates binding of the transcription factor NF-B p50 to DNA. Interestingly, the transcriptional activity of NF-B in the presence of Rb is decreased relative to the control NF-B, suggesting that Rb regulates the transcriptional activity of p50 in  1 and 2). 100% p50 DNA binding activity was observed at 56 ng of p50 with GST or GST Rb (lanes 3 and 4). Lower panel, in vitro transcription reactions were run under identical conditions to compare transcription and DNA binding activity.  1-6) or GST-Rb (lanes 7-11) as described (38). Chymotrypsin (400 g/ml) was added to indicated reactions and incubated for 0, 1, 5, 10, or 15 min. Reactions were terminated by the addition of chymostatin (500 g/ml) and applied to the gel for electrophoretic mobility shift assay as described in Fig. 1. In B, the p50 NF-B gel shift bands from Fig. 4A were quantitated using a phosphorimager. The percentage of NF-B binding was calculated based on the amount of radioactivity in the zero time gel shift band. f, p50 plus GST; }, p50 plus GST-Rb. addition to its DNA binding activity. Studies have shown that Rb binds to several transcription factors in vivo and in vitro and, as a result, regulates their function. Given the physical interaction between Rb and p50 and the function of Rb in repression of E2F activity, one interpretation of these results might be that Rb is a part of the NF-B/DNA complex, blocking the interaction of NF-B with the basal transcription machinery. Using antibody supershift and biotinylated DNA pulldown assays, we have been unable to demonstrate that Rb is a stable component of the DNA-protein complex.
It has been demonstrated previously that the p50 homodimer provides transcriptional activation only when adopting a chymotrypsin-resistant conformation (12). Our present results suggest that the conformation of the p50 homodimer is altered in the presence of Rb, reducing its transcriptional activity and functioning as a transcriptional repressor. This repression, therefore, would be functionally different from that observed with E2F and explains why NF-B was not detected in screening assays to identify promoter elements that are normally targeted by Rb (36). Along these lines, it will be of interest to determine if Rb phosphorylation regulates the interaction of Rb and p50 in a cell cycle-dependent manner.
The results presented in this report suggest that Rb is one of several cellular proteins that function to regulate p50 binding and activity. It has been demonstrated that the cellular protooncogene, Bcl-3, specifically inhibits binding of homodimeric p50 to DNA (37,38). Interestingly, when coexpressed, Bcl-3 and p50 both localize to the nucleus and form a protein-protein complex that is detected in nuclear extracts. In contrast, IB-␣ and IB-␥ inhibit p50 homodimer activity by inhibiting nuclear translocation. In this regard, Rb and Bcl-3 both function as nuclear inhibitors of p50. There are, however, distinct differences in the activities of Bcl-3 and Rb. Although Bcl-3 inhibits binding of the p50 homodimer, it does not inhibit binding of the p50-p65 heterodimer that can bind to the NF-B site and activate transcription. Rb, in contrast, stimulates the binding of an inactive p50 homodimer to the NF-B site, resulting in a decrease in transcription. NF-B plays an important role in the response of lymphocytes to antigens and cytokines. In addition to the cytoplasmic regulation of NF-B transport, Rb induction of transcriptionally inactive p50 homodimer binding may represent another important pathway for the regulation of nuclear NF-B activity.