PRDI - BF 1 silences CIITA expression in multiple myeloma cells

The Major Histocompatibility Complex (MHC) class II transactivator (CIITA) acts as a master switch to activate expression of the genes required for MHC-II antigenpresentation. During B-cell to plasma cell differentiation, MHC-II expression is actively silenced but the mechanism has been unknown. In plasma cell tumors such as multiple myeloma the repression of MHC-II is associated with the loss of CIITA. We have identified that PRDI-BF1, a transcriptional repressor inhibits CIITA expression in multiple myeloma cell lines. Repression of CIITA depends on the DNA binding activity of PRDI-BF1 and its specific binding site in the CIITA promoter. Deletion of a histone deacetylase recruitment domain in PRDI-BF1 does not inhibit repression of CIITA nor does blocking HDAC activity. This is in contrast to PRDI-BF1 repression of the c-myc promoter. Repression of CIITA requires either the amino-terminal acidic and conserved PR motif or the proline-rich domain. PRDI-BF1 has been shown to be a key regulator of B cell and macrophage differentiation. These findings now indicate that PRDI-BF1 has at least two mechanisms of repression whose function is dependent on the nature of the target promoter. Importantly, PRDI-BF1 is defined as the key molecule in silencing CIITA and thus MHC-II in multiple myeloma cells. INTRODUCTION One of the essential factors required for the transcription of MHC-II genes is the class II transactivator, CIITA (1,2). MHC-II expression correlates with CIITA expression for both constitutive expression on B cells, and cytokine induced expression in other cell types (3-5). With only a few exceptions to the rule CIITA functions as a 'master regulator' for MHC-II expression (2). During B-cell development MHC-II expression is turned on by gest on Sptem er 1, 2017 hp://w w w .jb.org/ D ow nladed from PRDI-BF1 represses CIITA expression 3 at a very early stage and turned off when B-cells terminally differentiate into plasma cells (6,7). The process of extinguishing MHC-II expression in plasma cells is poorly understood. Cell-cell fusion between MHC-II negative myeloma cells and MHC-II positive B cells have indicated the presence of a dominant repressor in the myeloma cell. (8-12). It was later shown that CIITA mRNA transcripts are absent in myeloma cells and introduction of CIITA rescues MHC-II expression (13). Thus, the dominant repressor of MHC-II expression appears to act by repressing CIITA expression. CIITA has four distinct promoters, each transcribing a unique Exon 1, with three of the forms predominating (14). CIITA expression in B-cells is predominantly from promoter 3 (14,15). We have recently characterized the CIITA promoter 3 (CIITAp3) in B-cells (16). In vivo, two elements on the promoter, ARE-1 and ARE-2, are occupied and essential for CIITA expression. However in the myeloma cell line NCI-H929, the CIITAp3 is completely unoccupied suggesting that CIITA transcription in myeloma cells may be mediated through changes in promoter assembly (16). One of the transcription factors induced when B cells differentiate into plasma cells is PRDI-BF1. It has been shown that introduction of the murine homologue of PRDI-BF1 (Blimp-1) into B cell lymphoma cell lines leads to many phenotypic changes associated with differentiation into early plasma cells, (17). Despite its dramatic effect on B-cells, only two targets for PRDI-BF1 have been identified, c-myc (18) and IFN-β (22). In this report we now show that PRDI-BF1 directly suppresses CIITA transcription. This results in the silencing of MHC-II expression long observed in plasma cells and multiple myeloma and as such may be an attractive target for therapeutic regulation of MHC-II. by gest on Sptem er 1, 2017 hp://w w w .jb.org/ D ow nladed from PRDI-BF1 represses CIITA expression 4 EXPERIMENTAL PROCEDURES DNA Constructs and cell lines. The CIITAp3 reporter constructs (16); the c-myc PRF reporter construct (18,19); pc-PRDI-BF1, pcFLAG-PRDI-BF1(331-789) and pcFLAGPRDI-BF1(398-789) constructs (20) have been described previously. The pc-PRDI-BF1 vector was FLAG tagged at the C-terminal end of PRDI-BF1. The three protein deletions, pcPRDI-BF1∆PR (deletes aa 36-166), pcPRDI-BF1∆Pro (deletes aa 308-399) and pcPRDI-BF1∆ZnF (deletes aa 508 to 691) were generated in the pc-PRDI-BF1-FLAG construct. All constructs have been confirmed by sequencing. Raji, IM-9 and CA46 are B-lymphoblastoid cells. U266, 8266, SKO and NCI-H929 cells are multiple myeloma cells. All cells were grown according to ATCC specifications. Transient transfection. Cells were transfected by electroporation as described previously (16). Transfections were normalized by co-transfecting a constant amount of pRL-TK construct (Promega). In order to make direct comparisons between PRDI-BF1's effect on c-myc and CIITA promoters, the control lanes were set at 100, for both TSAand TSA+ samples. Generation of antibodies to PRDI-BF1. Two peptide sequences N-terminal 1-17 residues and C-terminal 720-737 residues were selected to raise polyclonal antibodies in rabbits (Research Genetics). The antibodies were affinity purified from the serum. Nuclear extract preparation and Electrophoretic Mobility Shift Assays (EMSA). Nuclear extracts were prepared according to Dignam et al (21). EMSA was performed as described previously (16) using synthetic oligonucleotides, 2μg of poly(dI:dC) and nuclear extract or in vitro translated proteins. The PRDI-BF1 oligonucleotide spans from by gest on Sptem er 1, 2017 hp://w w w .jb.org/ D ow nladed from PRDI-BF1 represses CIITA expression 5 -190 to -158 base pairs of the promoter. Unlabeled oligonucleotide competitors were used in a 50-100 fold molar excess. RNA analysis. RNA was isolated using Trizol reagent (Gibco BRL). Ribonuclease Protection Analysis (RPA) was performed using a P labeled probe containing the Cterminal (Kpn-I and Xho-I) portion of PRDI-BF1. The RNA probe was hybridized with 15μg of sample RNA and digestion was performed using RNase A/T1 (Ambion) according to manufacturer's specifications. Western Blot and Immunoprecipitation. Whole cell extract was prepared in phosphate buffered saline with 0.1% NP-40, 1mM DTT, 0.1mM PMSF and Complete EDTA-free protease inhibitor cocktail (Boehringer Mannheim). Anti-FLAG M2-Agarose affinity beads (Sigma) were used to immunoprecipitate over-expressed FLAG-tagged proteins. Endogenous PRDI-BF1 was immunoprecipited with Protein A agarose beads and the N or C-terminal PRDI-BF1 antibody at 4C for 1 hour. The immunoprecipitated proteins were resolved on a 10%SDS-PAGE, which was transferred to a PVDF membrane. The membrane was incubated with the indicated primary antibody for 2 h followed by a Antirabbit IgG horse radish peroxidase antibody. The secondary antibody was visualized by ECL plus (Amersham). For direct Western blots or immunoprecipitations 1x10 or 5 x10 cells were used per lane, respectively. In vitro transcription translation reactions. In vitro transcription translation of PRDIBF1 and PRDI-BF∆PR proteins were performed using the TNT quick coupled transcription/translation system (Promega) according to manufacturers specifications. pCDNA blank vector was used as a control. The S labeled proteins were resolved on a by gest on Sptem er 1, 2017 hp://w w w .jb.org/ D ow nladed from PRDI-BF1 represses CIITA expression 6 SDS-PAGE and visualized by autoradiography to estimate the yield of full length product prior to use in the DNA binding assays. RESULTS and DISCUSSION PRDI-BF1 represses the CIITA promoter in myeloma cells. We have previously reported that endogenous CIITA mRNA levels in myeloma cell lines are significantly lower compared to B cells (16). Similarly, we observed that a CIITAp3 promoter construct with 1140 or 545 bp of promoter sequences had low activity in myeloma cells, similar to an SV40 minimal promoter while in B-cell lines the CIITA promoter was 2-4 times more active in comparison (data not shown). This suggested that the repression of CIITA transcription might be contained within the first 545 bp of the promoter. Progressive 5’ deletions of CIITAp3, were transiently transfected into two myeloma cell lines and tested for promoter activity. As shown in Figure 1A deletion of the region from –195 to –151 bp resulted in a seven-fold increase in promoter activity in U266 cells and a 3.2 fold increase in NCI-H929 cells. Further deletion down to position –113 bp removes the previously identified activator element ARE-1 and severely diminished promoter activity. This indicates that the factors necessary for activation of CIITAp3 are present in the myeloma cells, but overall activity is silenced by elements between –195 and –151. Homology searches suggested that the region between -180 and -168 bp could be a putative binding site for the transcriptional repressor, PRDI-BF1. In order to test whether PRDI-BF1 was able to repress CIITAp3, the protein was transiently over-expressed in Bcells. Introduction of PRDI-BF1 repressed transcription from the CIITA promoter (CIITAp3.195) by 60% as compared to vector control (Figure 1B). Deletion of the potential PRDI-BF1 binding site (CIITAp3.151) leads to the inability of PRDI-BF1 to by gest on Sptem er 1, 2017 hp://w w w .jb.org/ D ow nladed from PRDI-BF1 represses CIITA expression 7 significantly repress CIITA transcription. These findings and the observation that the murine homologue of PRDI-BF1 is low or absent in B cells, and increases in plasma cells (17) suggest that endogenous PRDI-BF1 in human myeloma cells could be repressing CIITA transcription by binding to this site on the promoter. PRDI-BF1 mRNA and protein levels are high in human myeloma cells. The relative amount of PRDI-BF1 mRNA in myeloma cells and B cells was determined with both a ribonuclease protection assay (Figure 2A) and also confirmed by Northern blotting (data not shown). Each of the myeloma cell lines contained PRDI-BF1 mRNA while it was undetectable in B-cell lines. We next examined the level of PRDI-BF1 protein present in myeloma cells by immunoprecipitation. PRDI-BF1 specific antibodies were raised against amino and carboxy terminal peptides. Both antibodies specifically immunoprecipited a band of approximately 100 kD from myeloma cells. The band comigrates with PRDI-BF1 protein expressed from the cloned cDNA when transfected into a B cell line (Figure 2B lanes 1 and 2). Importantly, we also tested for the presence of PRDI-BF1 in bone marrow samples from myeloma patients. Although these samples contain only 68% to 92% myeloma cells, PRDI-BF1 was clearly visible in 2 of 3 patients (Figure 2C). Thus PRDI-BF1 is present in myeloma cells concomitant with the loss of CIITA expression. PRDI-BF1 binds to the CIITAp3 in vitro. Identification of direct PRDI-BF1 binding to the region between -195 to -151 bp of the promoter was done with a series of in vitro protein/DNA binding studies utilizing the electrophoretic mobility shift assay (EMSA). Incubation of an oligonucleotide spanning -190 to -158 bp with nuclear extracts from myeloma cells revealed two specific complexes, which were competed by an unlabeled by gest on Sptem er 1, 2017 hp://w w w .jb.org/ D ow nladed from PRDI-BF1 represses CIITA expression 8 self oligonucleotide (Figure 3A). Competition with consensus binding sites such as an IRF-E (lane 3) which have similarity to the PRDI-BF1 site did not alter complexes. To directly identify that PRDI-BF1 is part of the complexes bound to DNA the U266 nuclear extracts were immuno-depleted of PRDI-BF1 by using either the N or C-terminal antibody. Western blots indicate that nearly all of the PRDI-BF1 was removed (Figure 3B). EMSAs performed using these depleted extracts showed that the upper complex completely disappeared, while the lower complex was reduced in intensity (Figure 3C lanes 1-3). Confirmation that only PRDI-BF1 was depleted and that the extract was not degraded was done using either a BASP (lanes 4-6) or NF-1 (not shown) consensus binding site in the EMSA assay. The difference between the two complexes containing PRDI-BF1 is not known as yet but it is possible that PRDI-BF1 is found in multiple isoforms or that it can complex with other proteins. Repressive domains in PRDI-BF1. The PRDI-BF1 protein had been first identified by its ability to bind and repress the PRDI (positive regulatory domain I) site within the IFNβ promoter, following viral induction (22). It is a 789 amino acid protein (depicted in Figure 4A) bearing N and C-terminal acidic domains, a PR domain homologous to RIZ and MDSI-EVI proteins (23), a proline-rich domain between amino acids 331-389, and a C-terminal Zinc finger DNA binding motif. It has been demonstrated that the proline-rich domain of PRDI-BF1 is able to recruit the Groucho family of corepressors (20) which in turn can interact with histone deacetylases (24). This suggested that the proline-rich domain was critical for transcriptional repression. In order to identify the domains responsible for PRDI-BF1 mediated repression of CIITA we constructed three individual domain deletions in the protein and obtained two N-terminal deletions of the protein by gest on Sptem er 1, 2017 hp://w w w .jb.org/ D ow nladed from PRDI-BF1 represses CIITA expression 9 which maintain DNA binding activity (a kind gift from T. Maniatis). We transiently overexpressed these proteins with the CIITAp3.195 reporter in CA46 B-cells (Figure 4A,B). Several observations have been made. First, deletion of Zinc finger DNA binding motif abolishes PRDI-BF1 activity (lanes 7 vs 2). Secondly, deletion of the acidic and PR domains do not affect repressive ability (lanes 3 vs 2) but deletion of the acidic, PR and proline-rich domains completely ablates PRDI-BF1 function (lanes 4 vs 2). Finally, internal deletion of only the proline-rich domain does not affect repression (lanes 6 vs 2), while internal deletion of the PR domain partially (30%) abrogates PRDI-BF1 function (lanes 5 vs 2). These findings indicate that a complex interplay of multiple repression domains exists within the amino terminus of the protein. The proline-rich alone (lanes 3 vs 4) or the acidic and PR domains alone (lanes 6 vs 2) are sufficient for full activity. Because internal deletion of the PR domain maintains only partial function, this suggests that PR domain is a key contributor to repressive activity of PRDI-BF1. However, which domain is most critical for CIITA repression remains unclear. Since deletion of the PR domain partially ablates the ability of the protein to repress CIITA, it is essential to know if this effect is due to deletion of the PR domain or misfolding of the recombinant protein. One way to address the structure of the recombinant protein is to determine whether it maintains DNA binding activity. EMSA was performed using in vitro translated PRDI-BF1 and PRDI-BF1∆PR proteins to make a quantitative comparison between the DNA binding activity of the two. Similar levels of both proteins of the expected molecular weight were expressed by in vitro translation (Figure 4C). As observed in Figure 4D both proteins had comparable DNA binding activities. This finding provides evidence that the loss of activity is not due to catastrophic mis-folding by gest on Sptem er 1, 2017 hp://w w w .jb.org/ D ow nladed from PRDI-BF1 represses CIITA expression 10 although minor alterations are possible. In addition, all the recombinant proteins were abundantly expressed when transfected into CA46 B cells. Detection of the protein expression levels was performed by Western blot (Figure 4B). PRDI-BF1 mediated repression of CIITAp3 does not involve histone deacetylase activity. It has been recently shown that the murine homologue of PRDI-BF1 (Blimp-1) is able to repress c-myc expression by recruitment of histone deacetylase (25). This recruitment was reported to be through the proline-rich domain. Our data in Figure 4A indicates that the deletion of the proline-rich domain did not affect the repression of CIITA. Although the proline-rich domain might have multiple functions, this suggests that histone deacetylases might not be required for CIITA repression. To test the importance of HDAC activity more directly we inhibited HDAC activity and examined the effects on c-myc and CIITA transcription. Addition of a histone deacetylase inhibitor, Trichostatin A (TSA) had no effect on transcriptional repression of CIITA, however, consistent with the previous study it completely abolished repression of c-myc (Table 1). A similar independence from HDAC activity was also observed with endogenous PRDIBF1. U266 myeloma cells were transfected with the CIITA promoter construct only in the presence or absence of TSA. If HDACs were involved in repression of CIITA then a relief of repression would be observed upon addition of TSA. However, no change in transcriptional activity was detected between the TSA treated and untreated cells. This clearly indicates that PRDI-BF1 mediated repression of CIITAp3 does not require HDACs. With respect to PRDI-BF1 mediated repression of CIITA, our findings demonstrate a role of the PR domain in transcription repression. The PR domain of two related family by gest on Sptem er 1, 2017 hp://w w w .jb.org/ D ow nladed from PRDI-BF1 represses CIITA expression 11 members, RIZ and MDSI-EVI1 has been (26) reported to significantly alter the function of the protein and mediate protein-protein interaction (27). The PR domain shares significant sequence identity to the yeast SET domain of proteins that play an important role in determining chromosomal structure and telomeric gene silencing. However the precise mechanism by which PR domains bring about transcriptional repression has not yet been characterized. It has been shown that RIZ1 has tumor suppressor ability, and deletion of the PR domain promotes oncogenesis. A similar observation was noted in the MDS1-EVI1 gene (28). The PR domain of MDS1-EVI1 is a common target of viral insertions and chromosomal translocations in leukemogenesis (29-31) suggesting it may have an important biological function. Other regions within PRDI-BF1 can partially compensate for loss of the PR domain. An N-terminal truncation of the protein eliminating the acidic and PR domain but retaining the proline-rich domain is still able to confer repressive activity. This suggests that the proline-rich domain is a good candidate for such a compensatory effect. In addition, deletion of the N-terminal acidic domain of Blimp-1, significantly reduced its ability to repress c-myc (25). Thus PRDI-BF1 (Blimp1) appears to utilize multiple repression domains whose function is dependent on the target promoter. In summary, we have identified that PRDI-BF1 binds to CIITA promoter and silences CIITA expression in myeloma cells. This now defines the mechanism of MHC-II silencing in plasmacytomas first observed 13 years ago (9,12). By use of histone deacetylase inhibitors and specific deletions of PRDI-BF1 we have clearly demonstrated that the proline-rich domain and histone deacetylase activity are not required for PRDIBF1 mediated repression of CIITA in contrast to their action on c-myc. The findings also by gest on Sptem er 1, 2017 hp://w w w .jb.org/ D ow nladed from PRDI-BF1 represses CIITA expression 12 suggest that that the highly conserved PR domain may have an important role inrepression of CIITA. REFERENCES1. Steimle, V., Otten, L. A., Zufferey, M., and Mach, B. (1993) Cell 75(1), 135-1462. Harton, J. A., and Ting, J. P. (2000) Mol Cell Bio 20(17), 6185-943. Steimle, V., Siegrist, C. A., Mottet, A., Lisowska-Grospierre, B., and Mach, B.(1994) Science 265(5168), 106-1094. Chang, C.-H., Fontes, J. D., Peterlin, M., and Flavell, R. A. (1994) J Exp Med180, 1367-13745. Chin, K. C., Mao, C., Skinner, C., Riley, J. L., Wright, K. L., Moreno, C. S.,Stark, G. R., Boss, J. M., and Ting, J. P. (1994) Immunity. 1(8), 687-6976. Halper, J., Fu, S. M., Wang, C.-Y., Winchester, R., and Kunkel, H. G. (1978) JImmunol 120, 1480-14847. Burrows, P. D., and Cooper, M. D. (1990) Semin Immunol 2(3), 189-958. Sartoris, S., Tosi, G., De Lerma Barbaro, A., Cestari, T., and Accolla, R. S.(1996) Eur J Immunol 26(10), 2456-609. Venkitaraman, A. R., Culbert, E. J., and Feldman, M. (1987) Eur J Immunol 17, 1441-144610. Chang, C. H., Fodor, W. L., and Flavell, R. A. (1992) J Exp Med 176(5), 1465-911. Latron, F., Jotterand-Bellomo, M., Maffei, A., Scarpellino, L., Bernard, M.,Strominger, J. L., and Accolla, R. S. (1988) Proc Natl Acad Sci USA 85, 22292233bygestonSptemer1,2017hp://www.jb.org/Downladedfrom PRDI-BF1 represses CIITA expression13 12. Dellabona, P., Latron, F., Maffei, A., Scarpellino, L., and Accolla, R. S. (1989) JImmunol. 142, 2902-291013. Silacci, P., Mottet, A., Steimle, V., Reith, W., and Mach, B. (1994) J. Exp. Med.180, 1329-133614. Muhlethaler-Mottet, A., Otten, L. A., Steimle, V., and Mach, B. (1997) Embo J16(10), 2851-6015. Piskurich, J. F., Wang, Y., Linhoff, M. W., White, L. C., and Ting, J. P. (1998) JImmunol 160(1), 233-4016. Ghosh, N., Piskurich, J. F., Wright, G., Hassani, K., Ting, J. P., and Wright, K. L.(1999) J Biol Chem 274(45), 32342-5017. Turner, C. A., Jr., Mack, D. H., and Davis, M. M. (1994) Cell 77(2), 297-30618. Lin, Y., Wong, K., and Calame, K. (1997) Science 276(5312), 596-919. Jayachandra, S., Low, K. G., Thlick, A. E., Yu, J., Ling, P. D., Chang, Y., andMoore, P. S. (1999) Proc Natl Acad Sci U S A 96(20), 11566-7120. Ren, B., Chee, K. J., Kim, T. H., and Maniatis, T. (1999) Genes Dev 13(1), 125-3721. Dignam, J. D., Lebovitz, R. M., and Roeder, R. G. (1983) Nucleic Acid Research11, 1475-148822. Keller, A. D., and Maniatis, T. (1991) Genes Dev 5(5), 868-7923. Buyse, I. M., Shao, G., and Huang, S. (1995) Proc Natl Acad Sci U S A 92(10), 4467-7124. Chen, G., Fernandez, J., Mische, S., and Courey, A. J. (1999) Genes Dev 13(17),2218-30bygestonSptemer1,2017hp://www.jb.org/Downladedfrom PRDI-BF1 represses CIITA expression14 25. Yu, J., Angelin-Duclos, C., Greenwood, J., Liao, J., and Calame, K. (2000) Mol.Cell. Bio. 20(7), 2592-60326. Xie, M., Shao, G., Buyse, I. M., and Huang, S. (1997) J Biol Chem 272(42),26360-627. Huang, S., Shao, G., and Liu, L. (1998) J Biol Chem 273(26), 15933-928. Fears, S., Mathieu, C., Zeleznik-Le, N., Huang, S., Rowley, J. D., and Nucifora,G. (1996) Proc Natl Acad Sci U S A 93(4), 1642-729. Morishita, K., Parker, D. S., Mucenski, M. L., Jenkins, N. A., Copeland, N. G.,and Ihle, J. N. (1988) Cell 54(6), 831-4030. Morishita, K., Parganas, E., William, C. L., Whittaker, M. H., Drabkin, H., Oval,J., Taetle, R., Valentine, M. B., and Ihle, J. N. (1992) Proc Natl Acad Sci U S A89(9), 3937-4131. Nucifora, G., Begy, C. R., Kobayashi, H., Roulston, D., Claxton, D., Pedersen-Bjergaard, J., Parganas, E., Ihle, J. N., and Rowley, J. D. (1994) Proc Natl Acad Sci U SA 91(9), 4004-8ACKNOWLEDGMENTSWe thank Dr. Maniatis (Harvard) for the PRDI-BF1 expression vectors and Dr. Dalton (Moffitt) for providing us multiple myeloma patient bone marrow samples. This work issupported by a NIH grant CA80990-R01 and by the Molecular Biology Core Facility atthe H. Lee Moffitt Cancer Center. IGY is supported in part by the Inst. of Biomolecular Science, Univ. of South Florida. FIGURE LEGENDSFigure 1. PRDI-BF1 represses CIITAp3 through the -151 to -195 region.bygestonSptemer1,2017hp://www.jb.org/Downladedfrom PRDI-BF1 represses CIITA expression15 (A) CIITAp3 deletion constructs (15μg) were transiently transfected into the myeloma cells and luciferase activity measured. Deletion of the PRDI-BF1 binding site (-195 to -151) results in a relief of repression. The results are an average of three experiments forU266 cell line and one representative experiment for NCI-H929 cells. (B) CA46 B cellswere transfected with the promoter constructs (15μg) as indicated along with PRDI-BF1expression vector (7μg) or control vector. Deletion of the region between -195 and -151 markedly decreased the ability of PRDI-BF1 to repress CIITA transcription. The resultspresented are an average of five experiments. Figure 2. Expression pattern of PRDI-BF1 in multiple cell types.(A) PRDI-BF1 mRNA expression is detected in each myeloma cell line by RNaseprotection assay. (B) The endogenous PRDI-BF1 protein detected in myeloma cells is ofapproximately the same molecular weight as the cloned cDNA. Lanes 1 and 2 are CA46B cells transfected with blank vector or PRDI-BF1-FLAG expression vector respectively.Lanes 3 and 4 detect endogenous PRDI-BF1 in SKO myeloma cells.IP=Immunoprecipitation, F=FLAG, N and C= N or C-terminal PRDI-BF1 antibodyrespectively. (C) Detection of PRDI-BF1 protein in bone marrow samples of multiple myeloma patients. Peripheral B cells from a healthy donor (lane 1), U266 cells (lane 2),Bone marrow samples from patient 1 with 92% plasma cells, patient 2 (68% plasmacells) and patient 3 (72% plasma cells).Figure 3. PRDI-BF1 binds to CIITAp3 in vitro. (A) EMSA was performed using an oligonucleotide containing the PRDI-BF1 binding site and U266 nuclear extracts. NS =Nonspecific band (B) Depletion of PRDI-BF1 from U266 nuclear extracts byimmunoprecipitation with N and C terminal PRDI-BF1 antibody. Western blot wasbygestonSptemer1,2017hp://www.jb.org/Downladedfrom PRDI-BF1 represses CIITA expression16 performed with the N-terminal PRDI-BF1 antibody. S= the nuclear extract afterimmunoprecipitation. B= immuno-complexed beads. (C) EMSA using immunodepletedU266 nuclear extract indicates the upper complexes contain PRDI-BF1. Lanes 1-3,PRDI-BF1 binding site is used as a probe. Lanes 4-6 a control BSAP transcription factorbinding site is used as the probe to demonstrate that PRDI-BF1 has been specificallydepleted.Figure 4. Repressive domains of PRDI-BF1. (A) CA46 B cells were transientlytransfected with CIITAp3.195 (15μg) and co-transfected with either control, wild type orvarious deletions of PRDI-BF1 (7μg). Results are an average of four experiments. (B)CA46 cells were transfected with 15μg of the PRDI-BF1 deletion constructs and harvested 36 hours later. In the left panel whole cell extracts were subjected toimmunoprecipitation with anti-FLAG antibody followed by Western blot using the sameantibody. In the right panel a direct Western blot was performed using antibody to the N-terminal of PRDI-BF1. Arrows indicate the wild type and mutant PRDI-BF1 proteins.(C) In vitro transcription translation of PRDI-BF1 (lanes 3,4) and PRDI-BF1∆PR (lanes 5,6) proteins. Lanes 1,2 are control (empty expression vector used for in vitro transcription translation reaction). Detection of theS labeled proteins was performed onSDS-PAGE. In the odd number lanes (1,3,5) 2μl of protein has been loaded, in the evennumber lanes 4μl protein has been loaded. (D) PRDI-BF1∆PR retains DNA bindingactivity. EMSA was performed using a P oligonucleotide containing the PRDI-BF1binding site and the in vitro translated proteins. Lanes 1-3 control, Lanes 4-6 PRDI-BF1,lanes 7-9 PRDI-BF1∆PR. SP= cold specific competitor.bygestonSptemer1,2017hp://www.jb.org/Downladedfrom