MHC class II is expressed in restricted lineages and is modulated in response to pathogens and inflammatory stimuli. This expression is controlled by MHC CIITA, which is transcribed from multiple promoters. Although factors required for induction of CIITA are well characterized, less is known about the mechanisms leading to repression of this gene. During plasma cell differentiation, B lymphocyte-induced maturation protein-1 (PRDM1/Blimp-1) represses promoter (p)III of CIITA, responsible for constitutive expression in B cells. pIV is inducible by IFN-γ in epithelia, macrophages and B cells. An IFN regulatory factor-element (IRF-E) in CIITA-pIV, which is bound by IRF-1 and IRF-2, is necessary for this response. This site matches the PRDM1/Blimp-1 consensus binding site, and PRDM1/Blimp-1 is expressed in cell lineages in which this promoter is operative. We, therefore, investigated whether PRDM1 regulates CIITA-pIV and found that PRDM1 bound to CIITA-pIV in vivo and the IRF-E in vitro. PRDM1 repressed IFN-γ-mediated induction of a CIITA-pIV luciferase reporter in a fashion dependent on an intact consensus sequence and competes with IRF-1/IRF-2 for binding to the IRF-E and promoter activation. In human myeloma cell lines that express IRFs, PRDM1 occupancy of CIITA-pIV was associated with resistance to IFN-γ stimulation, while short interfering RNA knockdown of PRDM1 led to up-regulation of CIITA. Our data indicate that PRDM1 is a repressor of CIITA-pIV, identifying a target of particular relevance to macrophages and epithelia. These findings support a model in which PRDM1/Blimp-1 can modulate the cellular response to IFN-γ by competing with IRF-1/IRF-2 dependent activation of target promoters.

MHC class II (MHC-II)3 is the major determinant of Ag presentation to CD4+ T cells, and the cognate interaction of a CD4+ T cell with an Ag-presenting cell has distinct functional consequences dependent on the context of costimulatory molecules, cytokine milieu, and microenvironment. The expression of MHC-II is, therefore, tightly controlled (1). Constitutive expression is primarily restricted to bone marrow-derived APCs, including dendritic cells and B cells (2). On other cell types, MHC-II is either absent or inducible only in the context of inflammatory cytokine stimulation, in particular IFN-γ (1).

The expression of MHC-II is largely controlled at the level of transcription. MHC-II promoters are regulated by the assembly of an enhanceosome and the subsequent recruitment of MHC CIITA (1, 3). The basal transcription factors assembled at MHC-II promoters are widely expressed, while CIITA is tightly regulated. CIITA is, therefore, viewed as the master regulator of MHC-II expression (4, 5).

CIITA is transcribed from three major promoters in a lineage and stimulus specific fashion. In dendritic cells and B cells, expression of CIITA is constitutive and mediated by transcripts originating from promoter (p)I and pIII, respectively (2). In macrophages, epithelia, endothelium, and B cells, CIITA can be induced in response to IFN-γ, and this response is primarily mediated through pIV (2, 6, 7, 8).

Although the factors involved in the inducible and constitutive expression of CIITA are well defined, less is known about the molecular mechanisms mediating repression of this gene. During terminal B cell differentiation, expression of MHC-II is repressed in the majority of normal plasma cells and this phenotype holds true for the malignant plasma cells of myeloma (9, 10). The repression of MHC-II correlates with loss of CIITA expression from pIII. This is attributable to B lymphocyte-induced maturation protein-1 (PRDM1/Blimp-1), which binds to and inhibits this promoter (11, 12).

PRDM1 was originally identified as a postinduction repressor of the IFN-β response to viral infection (13). The murine homolog, Blimp-1, was then characterized as a regulator of terminal B cell differentiation (14), and subsequently this process has been shown to be absolutely dependent on Blimp-1 (15). The number of known PRDM1/Blimp-1 target genes is limited and in addition to IFN-β and CIITA-pIII, includes c-Myc (16), Pax-5 (17), SPI-B, and ID3 (18). PRDM1/Blimp-1 acts as a transcriptional repressor through competitive promoter occupancy (13, 19), or by recruiting histone deacetylase activity (20, 21) and the histone H3 lysine-9 methyl-transferase, G9a (22). In mature tissues, PRDM1/Blimp-1 is additionally expressed in squamous epithelia (23), macrophages (24), and T cells (25, 26, 27). However, targets of PRDM1/Blimp-1 specifically relevant to these lineages are currently undefined. The preferred DNA binding sequence for PRDM1/Blimp-1 overlaps with a subset of IFN regulatory factor (IRF)-1 and IRF-2 binding sites (19, 28). This finding was predicted from the original description of PRDM1 as a transcriptional repressor, which binds to the IRF site in positive regulatory domain I (PRDI) of the IFN-β promoter (13).

Three sequence elements determine the IFN-γ responsive induction of CIITA at pIV: these are an IFN-γ activation sequence (GAS) bound by STAT-1, an E-box bound constitutively by USF-1, and an IFN regulatory factor element (IRF-E) (6, 7, 29). This IRF-E site is required for IFN-γ-mediated CIITA induction (6). It is bound by IRF-1/IRF-2 (30, 31), is conserved between mouse and human, and is a close match for the PRDM1/Blimp-1 consensus sequence. CIITA-pIV is regulated in an IFN-γ-dependent fashion in lineages in which PRDM1/Blimp-1 is expressed in mature tissues, including B cells, macrophages, and squamous epithelia (2, 32, 33). Therefore, we sought to determine whether PRDM1 is a regulator of CIITA-pIV. Our data establish CIITA-pIV as a direct target of PRDM1-mediated repression and demonstrate that PRDM1 acts in a competitive fashion with IRF-1/IRF-2 to modulate CIITA-pIV induction in response to IFN-γ. This supports a model in which PRDM1/Blimp-1 may determine the threshold of activation of a subset of IFN-γ responsive promoters.

The human myeloma cell lines U266 and NCI-H929, the Ramos B cell line, the Jurkat T cell line, and HeLa cells were maintained in RPMI 1640 containing 10% heat-inactivated FCS. COS cells were maintained in DMEM containing 10% heat-inactivated FCS. For transfections, HeLa or COS cells were seeded to reach 60% confluency on the day of transfection.

The bicistronic expression vector pIRES2-EGFP (Clontech) was modified by insertion of an N-terminal Myc-epitope tag. Full-length (FL) PRDM1 and sequence encoding the C terminus of PRDM1 (Δ527, encoding 528–789 aa) were amplified using the forward primers (5′-CTTCGAATTCATGAAAATGGACATGGAGGA-3′) and (5′-CTGGAATTCTACCCGCTGAAGAAGCAGAACGGC-3′), respectively, and the reverse primer (5′-TACCGTCGACATCTTAAGGATCCATTGGTTCAACTG-3′) and cloned into the modified pIRES2-EGFP vector.

To generate expression vectors for human IRF-1 and IRF-2, cDNAs were amplified using primers (5′-TTAAGAATTCACCATGCCCATCACTCGGAT-3′) and (5′-ATATGGATCCCTACGGTGCACAGGGAAT3′) for IRF-1 and primers (5′-TTAAGTCGACAAACCATGCCGGTGGAAAGGATG-3′) and (5′-ATATGGATCCTTAACAGCTCTTGACGCG-3′) for IRF-2 and were cloned into pIRES2-EGFP.

To generate expression vectors under the control of the ubiquitin promoter, the Gateway System (Invitrogen Life Technologies) was used. Vector pENTR-D-TOPO (Invitrogen Life Technologies) was first modified by insertion of the pBluescript KS multiple cloning site, generating pENTR-KS. The inserts of FL PRDM1 and Δ527 PRDM1, including the IRES2-EGFP sequences, were excised from the pIRES2-EGFP vectors using NheI and XbaI and cloned into the XbaI site of pENTR-KS. The IRES2-EGFP insert of the parent vector pIRES2-EGFP was cloned into pENTR-KS between BamHI and XbaI. Expression vectors were generated using the destination vector pLenti6/Ubc/V5-Dest and the Gateway recombinase system, according to manufacturer’s instructions. pcDNA3-β-actin-EGFP was provided by E. Morrison (University of Leeds, U.K.).

Expression vectors pGEX-FL PRDM1 and pGEX-Δ527 PRDM1 for GST fusion proteins of FL and C-terminal 261 aa of PRDM1 were generated by subcloning coding sequences from pIRES2-EGFP FL and Δ527 PRDM1 into pGEX-6P-1 (Amersham Biosciences) between EcoRI and SalI.

All plasmids generated were sequenced.

For immunoprecipitation of PRDM1, 1 × 107 H929 myeloma cells were lysed in RIPA buffer containing protease inhibitors. Protein lysates were incubated with non-immune rabbit IgG or rabbit anti-PRDM1 followed by protein G-Sepharose. For whole cell lysates, 1 × 106 cell equivalents were lysed in RIPA buffer. Samples were quantified using the DC protein assay system (Bio-Rad), separated on SDS-PAGE, transferred to nitrocellulose, and detected using either conventional ECL (Perbio) or the Odyssey Infrared Imaging system (LI-COR).

The CIITA-pIV sequence extending from −496 to +91 relative to the transcriptional start site was amplified with the forward primer (5′-AATTCTCGAGTCATGGTAACACAGGT-3′) and reverse primer (5′-ATATAAGCTTGACTTTGGTCACCTAC-3′) and cloned into the vector pXPG (34). Mutations were introduced into the IRF-E site using the Gene-Tailor site-directed mutagenesis kit (Invitrogen Life Technologies) and the forward oligonucleotides (5′-GCTGCAGAAAGAAAGTGAAACGGAAAAAGAA-3′) and (5′-CTCAGCGCTGCAGAAAGAAACTGAAACGGAA-3′) with reverse oligonucleotides (5′-TTTCACTTTCTTTCTGCAGCGCTGAGCTCG-3′) and (5′-TTTCTTTCTGCAGCGCTGAGCTCGGGCCAG-3′).

For luciferase assays, two to four replicate transfections were performed for each condition. All transfections were performed with GeneJuice reagent (Novagen) according to the manufacturer’s instructions. Experiments were done using the Promega luciferase assay system and analyzed on a Berthold Lumat LB Luminometer. Each experiment was repeated three times with similar results. Results were confirmed with separate plasmid preparations. Results are displayed as the fold change compared with luciferase activity in unstimulated cells transfected with the equivalent reporter construct.

The rabbit Ab generated against the PRDM1 N terminus, used in Western blotting, has been described previously (35). The rabbit Ab to PRDM1, used in ChIP assays, was raised against a mixture of GST-fusion proteins generated from the expression vectors pGEX-FL PRDM1 and pGEX-Δ527 PRDM1. Rabbit polyclonal Abs to IRF-1 (H-205) and IRF-2 (C-19) were purchased from Santa Cruz Biotechnology. Mouse monoclonal anti-β-actin clone AC15 was purchased from Sigma-Aldrich. Secondary Abs used were HRP-conjugated goat anti-rabbit IgG or rabbit anti-mouse IgG (Jackson ImmunoResearch Laboratories), donkey anti-mouse IgG coupled to IRDye800, and goat anti-rabbit IgG coupled to AlexaFluor 680 (Molecular Probes). Non-immune purified control rabbit IgG was purchased from Upstate Biotechnology.

Nuclear extracts were prepared as described previously (36) from either untransfected COS cells or COS cells transfected with FL PRDM1, IRF-1, or IRF-2 using GeneJuice reagent (Novagen). For EMSA, the double-stranded probes used contained the following sequences:

CIITA-pIV WT-F (5′-GCAGAAAGAAAGTGAAAGGGAAAAAGAACTG-3′), CIITA-pIV WT-R (5′-CAGTTCTTTTTCCCTTTCACTTTCTTTCTGC-3′), CIITA-pIV MUT-F (5′-GCAGAAAGAAACTGAAACGGAAAAAGAACTG-3′), CIITA-pIV MUT-R (5′-CAGTTCTTTTTCCGTTTCAGTTTCTTTCTGC-3′).

DNA probes end-labeled with [γ-32P]ATP using T4 polynucleotide kinase were incubated with nuclear extract in the presence of poly(dI:dC) (Amersham Biosciences) for 30 min at room temperature. Supershift was performed by the addition of Ab to the extract before mixing with radioactive probe, and competition assays included the addition of unlabeled probe to the reaction mixture. Samples were separated on a 4% polyacrylamide gel containing 5% glycerol.

RNA was isolated by the TRIzol method (Invitrogen Life Technologies) according to manufacturer’s instructions and subjected to DNase I treatment (DNAFree, Ambion). cDNA was generated using random hexamers and Superscript II reverse transcriptase (Invitrogen Life Technologies). Quantitative real-time PCR was performed using SYBR Green MasterMix (Applied Biosystems) on an ABI Prism 7700 system (PerkinElmer) and evaluated with SDS1.9 software (Applied Biosystems). Specificity of PCR was monitored with melting curve analysis, and amplification of a single product of the expected size was verified on agarose gels in initial experiments. Relative expression was normalized to GAPDH. For quantitative RT-PCR, the following primers were used: GAPDH-F (5′-AACAGCGACACCCACTCCTC-3′), GAPDH-R (5′-CATACCAGGAAATGAGCTTGACAA-3′), CIITA pIII-F (5′-GGCTGGGATTCCTACACAATG-3′), CIITA pIII-R (5′-CAACTCCATGGTGGCACACT-3′), CIITA pIV-F (5′-GAGCTGGCGGGAGGGAG-3′), CIITA pIV-R (5′-TGGTGGCACACTGTGAGCTG-3′), TAP1-F (5′-CGATACCTTCACTCGAAACTTAACTCT-3′), and TAP1-R (5′- GGCCCATGGTGTTGTTATAGATC-3′).

ChIP was performed as described (37). Input DNA was prepared from an equal volume of chromatin as used for immunopreciptation, resuspended in the same final volume as the ChIP samples, and a standard curve of input DNA was generated. Target sequences were analyzed by real-time PCR, the fraction of input chromatin precipitated (percentage input) was calculated from the standard curve with subtraction of background present in control immunoprecipitates generated with non-immune purified rabbit IgG. The data are representative of a minimum of two independent chromatin preparations for H929 and U266 and represent average and SD of at least two independent immunoprecipitations. For control promoter sequences, amplification of input chromatin as well as ChIP with Abs to acetylated or trimethylated histone H3-lysine 9 (Abcam) was used to confirm detection. Amplification of input chromatin and ChIP with Ab to acetyl H3-lysine 9 was used to confirm detection of CIITA promoter sequences in chromatin from the Ramos B cell line.

The following primers were used: CIITA pIII PR-F (5′-TCAGTCCACAGTAAGGAAGTGAAATT-3′), CIITA pIII PR-R (5′-GAAACAAGTGAGGGATCATCAAAAA-3′), CIITA pIV PR-F (5′-GGCCACAGTAGGTGCTTGGT-3′), CIITA pIV PR-R (5′-CTCGTCCGCTGGTCATCCT3′), TAP1 PR-F (5′-GGCGAGAAGCTCAGCCATT-3′), TAP1 PR-R (5′-TAGTCTGGGCAGGCCACTTT-3′), DNTT PR-F (5′-CACTTTGGCAGGAAGCTGTTG-3′), DNTT PR-R (5′-CCCTCCTACACGAAGGGTTTT-3′), KRT-10 PR-F (5′-TGGACACACCCTCTCAGTATATAAAGG-3′), KRT-10 PR-R (5′-AGAGTAGTGCTTGCTTGAGCTGTATC-3′), β-actin PR-F (5′-CGGCCAACGCCAAAAC-3′), β-actin PR-R (5′-GCACAGTGCAGCATTTTTTTACC-3′), β-globin F (5′-TTTTGTTCCCCCAGACACTCTT-3′), and β-globin R (5′-GGGTAATCAGTGGTGTCAAATAGGA-3′).

siRNA oligos PRDM1 (siRNA ID 106411, sense sequence 5′ GGAAAGGACCUCUACCGUUtt 3′) and silencer negative control no.1 (catalog no. 4611) were purchased from Ambion. These oligos were labeled with the FAM silencer siRNA labeling kit (catalog no. 1634; Invitrogen Life Technologies) according to the manufacturer’s recommendations. A total of 1 × 106 cells was transfected using Lipofectamine 2000 (Invitrogen Life Technologies) with 100 nM of labeled siRNA. The transfection was conducted according to the manufacturer’s recommendations (Invitrogen Life Technologies). Control samples of U266 were treated with Lipofectamine 2000 only. Following transfection, the U266 were incubated for 48 h before analysis.

To directly address whether PRDM1 binds CIITA-pIV, we used ChIP to examine the interaction of PRDM1 with CIITA-pIV in vivo. We generated a rabbit polyclonal Ab that effectively immunoprecipitates PRDM1 from lysates of myeloma cell lines, which are representative of the terminal stage of B cell differentiation and express high constitutive levels of this protein (Fig. 1 A).

FIGURE 1.

PRDM1 binds to CIITA-pIV. A, Lysates from H929 myeloma cells were immunoprecipitated (IP) with control nonimmune rabbit IgG (NI) or rabbit Ab to FL PRDM1, separated by SDS-PAGE, transferred to nitrocellulose, and probed with PRDM1 Ab specific for the N terminus. Whole cell lysates (WCL) from H929 myeloma cells and Jurkat T cells were included in the analysis. B, Chromatin prepared from H929 myeloma cells was immunoprecipitated with control rabbit IgG or anti-PRDM1 and the amount of CIITA-pIII or CIITA-pIVsequence present in each was quantified by real-time PCR. The fraction of precipitated chromatin relative to control IgG is displayed as percentage input. The percentage of control promoter sequences, β-actin, TDT, KRT-10, and β-globin, in the PRDM1 ChIP is displayed as an average (H929 Controls). B.D., Below the limits of detection. These data are derived from four separate immunoprecipitations from two independent chromatin preparations for H929 and duplicate immunoprecipitations for Ramos and are displayed as the mean ± SD. C, Nuclear extracts were prepared from untransfected COS cells (−) or COS cells that had been transfected with PRDM1 (+) and incubated with 32P-labeled oligonucleotide probe encompassing the CIITA-pIV IRF-E, and subjected to gel mobility shift assay. A unique complex was obtained from PRDM1-containing extracts. The specificity of interaction was assessed by the inclusion of 10- or 100-fold unlabeled probe. To confirm the presence of PRDM1 in the complex, extracts were preincubated with either control nonimmune rabbit IgG (NI) or rabbit anti-PRDM1 prior to mixing with labeled probe.

FIGURE 1.

PRDM1 binds to CIITA-pIV. A, Lysates from H929 myeloma cells were immunoprecipitated (IP) with control nonimmune rabbit IgG (NI) or rabbit Ab to FL PRDM1, separated by SDS-PAGE, transferred to nitrocellulose, and probed with PRDM1 Ab specific for the N terminus. Whole cell lysates (WCL) from H929 myeloma cells and Jurkat T cells were included in the analysis. B, Chromatin prepared from H929 myeloma cells was immunoprecipitated with control rabbit IgG or anti-PRDM1 and the amount of CIITA-pIII or CIITA-pIVsequence present in each was quantified by real-time PCR. The fraction of precipitated chromatin relative to control IgG is displayed as percentage input. The percentage of control promoter sequences, β-actin, TDT, KRT-10, and β-globin, in the PRDM1 ChIP is displayed as an average (H929 Controls). B.D., Below the limits of detection. These data are derived from four separate immunoprecipitations from two independent chromatin preparations for H929 and duplicate immunoprecipitations for Ramos and are displayed as the mean ± SD. C, Nuclear extracts were prepared from untransfected COS cells (−) or COS cells that had been transfected with PRDM1 (+) and incubated with 32P-labeled oligonucleotide probe encompassing the CIITA-pIV IRF-E, and subjected to gel mobility shift assay. A unique complex was obtained from PRDM1-containing extracts. The specificity of interaction was assessed by the inclusion of 10- or 100-fold unlabeled probe. To confirm the presence of PRDM1 in the complex, extracts were preincubated with either control nonimmune rabbit IgG (NI) or rabbit anti-PRDM1 prior to mixing with labeled probe.

Close modal

To validate the ChIP technique with this Ab, we precipitated PRDM1 from chromatin prepared from the human myeloma cell line H929 and assayed CIITA-pIII, a known PRDM1/Blimp-1 target promoter, as a positive control (11, 12). The specificity of interaction was assessed by analyzing the promoter sequences of a range of genes (β-actin, TDT, KRT-10, and β-globin), which are not known PRDM1/Blimp-1 targets and lack the PRDM1/Blimp-1 consensus sequence (displayed as H929 controls). In chromatin from H929 cells, PRDM1 Ab precipitated a fraction of input CIITA-pIII sequence (0.73% ± 0.16) eight-fold greater than the fraction of control promoter sequences (0.08% ± 0.04) (Fig. 1,B). As an additional negative control, PRDM1 ChIP was performed on chromatin prepared from the Ramos B cell line, which does not express PRDM1. The fraction of precipitated CIITA promoter sequence was analyzed and no difference between control and PRDM1 precipitations was detected (Fig. 1 B).

Having validated the technique, we addressed the association of PRDM1 with CIITA-pIV using primers that span the CIITA-pIV IRF-E site. As for CIITA-pIII, a substantial fraction of input CIITA-pIV was detected in PRDM1 ChIP from the H929 myeloma cell line (1.41% ± 0.1), while there was none detected in samples prepared with chromatin from the Ramos B cell line (Fig. 1 B). Thus, PRDM1 associates with CIITA-pIV in vivo.

To establish whether PRDM1 directly interacts with the CIITA-pIV IRF-E site, binding of PRDM1 to this region was analyzed by EMSA (Fig. 1,C). COS cells, which do not express PRDM1, were transfected with an expression vector encoding PRDM1. Nuclear extracts were incubated with radiolabeled oligonucleotide spanning the CIITA-pIV IRF-E alone or in the presence of excess cold competitor. A major complex was present in the nuclear extracts of PRDM1 transfected but not control COS cells (Fig. 1,C). To confirm the identity of this complex, PRDM1 antiserum was added to the nuclear extract, which led to its ablation and supershift (Fig. 1 C). Together, the ChIP and EMSA data demonstrate that PRDM1 associates with CIITA-pIV in vivo and binds to the IRF-E site in vitro.

At its known target genes, PRDM1/Blimp-1 acts as a transcriptional repressor. To address the ability of PRDM1 to repress CIITA-pIV, a luciferase reporter was generated spanning bases −496 to +91 relative to the CIITA-pIV transcriptional start site, which includes the IRF-E at position −66 to −55 (2). This construct gave a low level of basal promoter activity in HeLa cells and gave the expected substantial enhancement following IFN-γ stimulation for 6 h. Cotransfection of the CIITA-pIV construct with PRDM1 expression vector had a negligible effect on basal promoter activity. However, promoter induction by IFN-γ stimulation was markedly repressed (Fig. 2,A). Cotransfection with an equivalent amount of empty expression vector or a vector encoding β-actin gave no inhibition of promoter activity in response to IFN-γ (Fig. 2 A). Transfection of PRDM1 had no effect on a control thymidine kinase promoter construct (data not shown).

FIGURE 2.

PRDM1 represses IFN-γ stimulation of CIITA-pIV. A, HeLa cells transfected with a luciferase reporter construct containing pIV of CIITA and either 150 ng empty expression vector or 150 ng of expression vector encoding β-actin, or expression vector for FL PRDM1, or the C-terminal zinc finger domains of PRDM1 (Δ527 PRDM1) under the control of the ubiquitin promoter, were evaluated for luciferase activity after treatment with medium alone or 200 IU/ml IFN-γ for 6 h. These data are derived from four replicates and are displayed as the mean ± SD. Data are representative of three independent experiments. B, HeLa cells were transfected with the CIITA-pIV promoter luciferase construct and incremental amounts of the FL PRDM1 expression vector under control of the CMV promoter and assayed after IFN-γ stimulation (200 IU/ml) for 6 h. These data are derived from duplicate samples and are displayed as the mean ± SD. Data are representative of three independent experiments.

FIGURE 2.

PRDM1 represses IFN-γ stimulation of CIITA-pIV. A, HeLa cells transfected with a luciferase reporter construct containing pIV of CIITA and either 150 ng empty expression vector or 150 ng of expression vector encoding β-actin, or expression vector for FL PRDM1, or the C-terminal zinc finger domains of PRDM1 (Δ527 PRDM1) under the control of the ubiquitin promoter, were evaluated for luciferase activity after treatment with medium alone or 200 IU/ml IFN-γ for 6 h. These data are derived from four replicates and are displayed as the mean ± SD. Data are representative of three independent experiments. B, HeLa cells were transfected with the CIITA-pIV promoter luciferase construct and incremental amounts of the FL PRDM1 expression vector under control of the CMV promoter and assayed after IFN-γ stimulation (200 IU/ml) for 6 h. These data are derived from duplicate samples and are displayed as the mean ± SD. Data are representative of three independent experiments.

Close modal

PRDM1 binds to DNA via its C-terminal zinc-finger domains, while recruitment of corepressors is dependent on its central proline-rich region (13, 20, 21, 22). To determine whether the DNA binding region was sufficient to mediate transcriptional repression an expression vector encoding the C-terminal zinc-finger domains but lacking the proline-rich region was generated (Δ527 PRDM1, 528–789 aa of PRDM1). Cotransfection with Δ527 PRDM1 was sufficient to mediate the observed repression of the CIITA-pIV luciferase vector (Fig. 2 A).

To investigate the sensitivity of the CIITA-pIV to PRDM1, cotransfection was performed with incremental amounts of FL PRDM1 vector. A dose response was evident over the range of cotransfected PRDM1 vector doses (Fig. 2,B). This response was reproducible using either CMV or ubiquitin promoter-driven expression of PRDM1 (data not shown). Cotransfection with empty expression vector at the highest dose did not mediate repression of IFN-γ induction of pIV (data not shown and Fig. 2 A).

Thus, PRDM1 acts as a potent transcriptional repressor of CIITA-pIV, and in this assay system, the C terminus of PRDM1 encompassing the DNA-binding domain is sufficient to mediate this effect. These results are consistent with a significant role for competitive promoter occupancy in the repression of CIITA-pIV by PRDM1.

The overlap of PRDM1 and IRF-1/IRF-2 consensus sequences predicts either competitive occupancy of mutual sites or the formation of hetero-complexes with IRF proteins. Previous data examining PRDM1 and IRF binding at PRDI of IFN-β favor the former (13, 19). To directly address this question at the IRF-E of CIITA-pIV, we prepared nuclear extracts from COS cells transfected with vectors encoding IRF-1, IRF-2, or PRDM1 and examined the formation of complexes in the presence of the CIITA-pIV IRF-E oligonucleotide. Extracts containing PRDM1, IRF-1, or IRF-2 alone generated complexes distinct from each other (Fig. 3,A). The identity of each of these complexes was confirmed by incubation with the appropriate Ab (Fig. 3, A and B).

FIGURE 3.

PRDM1 competes with IRF-1 and IRF-2 for binding to the CIITA-pIV IRF-E site. A, Occupancy of the CIITA-pIV IRF-E was assessed by mobility shift assay with nuclear extracts from COS cells transfected with IRF-1, IRF-2, or PRDM1. Extracts containing IRF-1 or IRF-2 were mixed with incremental amounts of extract containing PRDM1 alone or in the presence of specific antibodies. The band corresponding to IRF-1 (•); IRF-2 (○); PRDM1 (arrowhead). B, Binding to the CIITA-pIV IRF-E was further analyzed by mobility shift assay for the co-occupancy of IRF-1 and IRF-2 vs PRDM1. Labeled oligonucleotide was incubated with nuclear extracts containing IRF-1, IRF-2, or PRDM1, alone or in combination. Ab to IRF-1, IRF-2 or PRDM1 was included as indicated. The bands are designated as in A, but additionally, co-occupancy of the CIITA-pIV IRF-E is marked by (••), representing two molecules of IRF-1 (○○), representing two molecules of IRF-2, or a combination of IRF-1 and IRF-2 (○•). C, The effect of PRDM1 on the ability of IRFs to activate the CIITA-pIV promoter luciferase construct was evaluated. HeLa cells were cotransfected with the luciferase construct and empty vector, two doses of FL PRDM1 (6.25 or 62.5 ng), IRF-1 (5 ng), or IRF-2 (5 ng) and combinations of these vectors. Luciferase activity was assayed 24 h post transfection. These data are derived from triplicate samples and are displayed as the mean ± SD. Data are representative of three independent experiments.

FIGURE 3.

PRDM1 competes with IRF-1 and IRF-2 for binding to the CIITA-pIV IRF-E site. A, Occupancy of the CIITA-pIV IRF-E was assessed by mobility shift assay with nuclear extracts from COS cells transfected with IRF-1, IRF-2, or PRDM1. Extracts containing IRF-1 or IRF-2 were mixed with incremental amounts of extract containing PRDM1 alone or in the presence of specific antibodies. The band corresponding to IRF-1 (•); IRF-2 (○); PRDM1 (arrowhead). B, Binding to the CIITA-pIV IRF-E was further analyzed by mobility shift assay for the co-occupancy of IRF-1 and IRF-2 vs PRDM1. Labeled oligonucleotide was incubated with nuclear extracts containing IRF-1, IRF-2, or PRDM1, alone or in combination. Ab to IRF-1, IRF-2 or PRDM1 was included as indicated. The bands are designated as in A, but additionally, co-occupancy of the CIITA-pIV IRF-E is marked by (••), representing two molecules of IRF-1 (○○), representing two molecules of IRF-2, or a combination of IRF-1 and IRF-2 (○•). C, The effect of PRDM1 on the ability of IRFs to activate the CIITA-pIV promoter luciferase construct was evaluated. HeLa cells were cotransfected with the luciferase construct and empty vector, two doses of FL PRDM1 (6.25 or 62.5 ng), IRF-1 (5 ng), or IRF-2 (5 ng) and combinations of these vectors. Luciferase activity was assayed 24 h post transfection. These data are derived from triplicate samples and are displayed as the mean ± SD. Data are representative of three independent experiments.

Close modal

To examine the formation of hetero-complexes, incremental amounts of PRDM1 containing extract were added to samples containing IRF-1 or IRF-2. The inclusion of PRDM1 did not result in any new complex, suggesting that PRDM1 does not co-occupy the IRF-E with either IRF-1 or IRF-2. Furthermore, there was a decrease in the binding of IRF-1 and IRF-2 in samples containing the highest amount of PRDM1 (Fig. 3,A). Previous work has demonstrated the simultaneous occupancy of IRF-1 and IRF-2 at the IRF-E site within CIITA-pIV (30). EMSAs performed using either IRF-1- or IRF-2-containing extracts alone resulted in the detection of complexes that are likely to represent the binding of two molecules, referred to as homo-occupancy (Fig. 3 B) (30). Mixing equal amounts of IRF-1- and IRF-2-containing extracts generated a new complex that could be ablated with the inclusion of Ab to either IRF-1 or IRF-2, suggestive of hetero-occupancy of the IRF-E (30). The inclusion of PRDM1-containing extract with samples containing both IRF-1 and IRF-2 resulted in a complex of the expected size for PRDM1 occupancy alone, which is similar to the size of the IRF-1/IRF-2 hetero-complex. In the presence of the highest amount of PRDM1-containing extract, formation of both IRF-1- and IRF-2-containing complexes was diminished. Together, these data are consistent with the previous published report of IRF-1/IRF-2 hetero-occupancy and provide evidence for mutually exclusive binding of PRDM1 and these factors to the CIITA-pIV IRF-E.

Expression of either IRF-1 or IRF-2 with a CIITA-pIV reporter construct is sufficient to activate transcription without the requirement for IFN-γ stimulation (7, 30). If PRDM1-dependent repression of CIITA-pIV is mediated through competition at the IRF-E, then promoter induction by IRF-1 or IRF-2 alone should be repressed by PRDM1. To ascertain whether expression of PRDM1 can prevent the direct induction of CIITA-pIV by IRF proteins, HeLa cells were cotransfected with CIITA-pIV reporter construct and combinations of vectors encoding IRF-1, IRF-2, and PRDM1. Transfection of either IRF-1 or IRF-2 alone increased the transcriptional activity of the reporter construct (Fig. 3 C). Inclusion of PRDM1 expression vector in the transfection resulted in the inhibition of IRF-1- or IRF-2-mediated activation of CIITA-pIV, consistent with competition at the IRF-E site.

If repression of CIITA-pIV by PRDM1 requires binding to the IRF-E site and competition with IRF-1 and IRF-2, then altering the IRF-E sequence to favor IRF rather than PRDM1 binding should eliminate or reduce repression. The IRF-E site of CIITA-pIV is a dimeric IRF binding site (38). Alignment of the CIITA-pIV IRF-E with the PRDM1/Blimp-1 consensus (19, 28), the IRF-E consensus (39), and the IRF binding specificity determined by crystal structure (38, 40), suggested that mutation of the third and ninth residues of the predicted PRDM1/Blimp-1 consensus binding site would preserve binding of IRF, to provide IFN-γ responsive induction, while disrupting binding of PRDM1/Blimp-1 (Fig. 4,A). The relative importance of these residues to PRDM1 binding also has been suggested by methylation interference studies at the PRDI site within the IFN-β promoter (13). The ability of PRDM1 to associate with this mutated IRF-E site (GAAACTGAAACGG) was assessed by EMSA (Fig. 4,B). PRDM1 binding to the mutant site was markedly reduced in comparison with wild type. This change in affinity was confirmed in reciprocal competition experiments with the wild-type and mutant oligos (Fig. 4,B). To verify that these mutations introduce a specific defect in PRDM1 binding, EMSAs were additionally performed using extracts containing either IRF-1 or IRF-2. Both IRF-1 and IRF-2 maintain binding to the mutant oligo, confirming a more profound effect of these two residues on the ability of PRDM1 to interact with the IRF-E of CIITA-pIV (Fig. 4 C).

FIGURE 4.

PRDM1 represses CIITA-pIV via binding to the IRF-E site. A, Alignment of sequences representing consensus binding sites for IRFs and PRDM1/Blimp-1 and mutant IRF-E sequence. Evaluation of PRDM1 (B), IRF-1 or IRF-2 (C) binding to an oligonucleotide containing CIITA-pIV sequence with a mutated IRF-E site (mut pIV) corresponding to that shown in A. Nuclear extracts from untransfected (−) or PRDM1- (+), IRF-1- (+), or IRF-2- (+) transfected COS cells were evaluated for binding to the mutant probe by gel shift assay. Specificity of the interaction was determined by the inclusion of 10- or 100-fold cold competitor. The mutant probe was similarly assayed for its ability to inhibit the binding of PRDM1 to wild-type sequence (wt pIV). The migration of IRF-1 (•); IRF-2 (○). D, HeLa cells were cotransfected with FL PRDM1 expression vector (150 or 9 ng) and a mutated CIITA-pIV luciferase reporter and evaluated for luciferase activity following 6 h of IFN-γ (200 IU/ml) stimulation. These data are derived from duplicate samples and are displayed as the mean ± SD. Data are representative of three independent experiments.

FIGURE 4.

PRDM1 represses CIITA-pIV via binding to the IRF-E site. A, Alignment of sequences representing consensus binding sites for IRFs and PRDM1/Blimp-1 and mutant IRF-E sequence. Evaluation of PRDM1 (B), IRF-1 or IRF-2 (C) binding to an oligonucleotide containing CIITA-pIV sequence with a mutated IRF-E site (mut pIV) corresponding to that shown in A. Nuclear extracts from untransfected (−) or PRDM1- (+), IRF-1- (+), or IRF-2- (+) transfected COS cells were evaluated for binding to the mutant probe by gel shift assay. Specificity of the interaction was determined by the inclusion of 10- or 100-fold cold competitor. The mutant probe was similarly assayed for its ability to inhibit the binding of PRDM1 to wild-type sequence (wt pIV). The migration of IRF-1 (•); IRF-2 (○). D, HeLa cells were cotransfected with FL PRDM1 expression vector (150 or 9 ng) and a mutated CIITA-pIV luciferase reporter and evaluated for luciferase activity following 6 h of IFN-γ (200 IU/ml) stimulation. These data are derived from duplicate samples and are displayed as the mean ± SD. Data are representative of three independent experiments.

Close modal

To determine whether the reduced PRDM1 binding affinity to the mutated IRF-E site is mirrored in a reduced efficiency of PRDM1-dependent repression, we mutated the corresponding bases in the CIITA-pIV luciferase reporter. Consistent with maintenance of IRF-1/IRF-2 binding, IFN-γ-mediated induction of the mutated luciferase construct was not affected. However, the ability of PRDM1 to repress the mutated vector was severely compromised (Fig. 4 D). Together, these data establish that the IRF-E sequence is necessary for PRDM1-mediated repression of CIITA-pIV.

The competitive binding of PRDM1 and IRFs to the IRF-E, and the results of our luciferase assays predict that IFN-γ stimulation of CIITA-pIV should be repressed in the H929 myeloma cell line. We, therefore, examined the effects of IFN-γ stimulation on expression of CIITA from pIV and on PRDM1 occupancy of the promoter in these cells.

RNA and chromatin were prepared from H929 and HeLa cells left untreated or treated for 24 h with IFN-γ. Both H929 and HeLa cells express minimal levels of CIITA-pIV transcripts in the unstimulated state, as assessed by quantitative RT-PCR (Fig. 5,A). Following IFN-γ treatment for 24 h, minimal induction of CIITA-pIV transcripts was detectable in H929 cells, compared with the substantial increase seen in HeLa cells. To demonstrate that H929 cells did not possess a global defect in the induction of IFN-γ target genes, expression of TAP1 and LMP2, which are regulated from a shared IFN-γ responsive bidirectional promoter, was examined (41). In contrast with CIITA-pIV, H929 constitutively express TAP1 and LMP2, and both genes are further induced in response to IFN-γ treatment (Fig. 5,B and data not shown). HeLa cells express barely detectable TAP1 in the unstimulated state but reach levels similar to H929 upon IFN-γ stimulation (Fig. 5,B). Furthermore, to demonstrate that H929 cells did not possess an intrinsic defect in the expression of IRFs we evaluated the protein levels of IRF-1 and IRF-2 in cells before and after stimulation with IFN-γ. H929 cells expressed these factors constitutively, while neither was detectable in unstimulated HeLa cells (Fig. 5 C). Results similar to H929 also were evident in an additional myeloma cell line U266. Upon stimulation, the levels of IRF-1 reached in each cell line were similar, while IRF-2 remained largely unchanged in the myeloma cell lines but was induced in HeLa. These data exclude a global defect in IFN-γ signaling in the H929 myeloma cell line.

FIGURE 5.

CIITA-pIV is nonresponsive to IFN-γ stimulation in myeloma H929 cells. A, The level of CIITA transcripts originating from pIV in HeLa or H929 cells after treatment with 200 IU/ml IFN-γ for 24 h was assessed by quantitative real-time PCR and normalized to GAPDH. Data are displayed relative to the untreated HeLa samples. B, The IFN-γ-inducible target gene TAP1 also was amplified from the same samples and quantified in a similar fashion. These data were derived from four separate stimulations and are displayed as the mean ± SD. C, Equal amounts of cellular lysates from HeLa, H929, or U266 cells left untreated or stimulated with 200 IU/ml IFN-γ for 24 h were evaluated by immunoblotting for levels of IRF-1, IRF-2, or control β-actin.

FIGURE 5.

CIITA-pIV is nonresponsive to IFN-γ stimulation in myeloma H929 cells. A, The level of CIITA transcripts originating from pIV in HeLa or H929 cells after treatment with 200 IU/ml IFN-γ for 24 h was assessed by quantitative real-time PCR and normalized to GAPDH. Data are displayed relative to the untreated HeLa samples. B, The IFN-γ-inducible target gene TAP1 also was amplified from the same samples and quantified in a similar fashion. These data were derived from four separate stimulations and are displayed as the mean ± SD. C, Equal amounts of cellular lysates from HeLa, H929, or U266 cells left untreated or stimulated with 200 IU/ml IFN-γ for 24 h were evaluated by immunoblotting for levels of IRF-1, IRF-2, or control β-actin.

Close modal

The high basal levels of TAP1 and LMP2 mRNA in H929 is consistent with constitutive expression of IRF-1 and IRF-2, as both IRFs have been shown to contribute to regulation of the bidirectional promoter (41, 42). This contrasts with the low levels of CIITA-pIV mRNA. To address whether the difference in TAP1/LMP2 and CIITA-pIV regulation was reflected in differential promoter occupancy, we performed ChIP analysis on chromatin prepared from untreated and IFN-γ-treated H929 cells. Samples were immunoprecipitated with control Ab, Ab to PRDM1, IRF-1, or IRF-2, and the fraction of precipitated input chromatin from the CIITA-pIV and the bidirectional TAP1/LMP2 promoter was assessed (41, 43). The promoter of the cytokeratin-10 gene (KRT-10), which lacks an IRF-E sequence and is not expressed in lymphoid lineage cells was amplified as a negative control to assess nonspecific binding. The association of PRDM1 with CIITA-pIV was preserved despite IFN-γ stimulation, consistent with maintenance of PRDM1-mediated repression of the promoter (Fig. 6,A). Although the IRF-E site of the bidirectional TAP1/LMP2 promoter can be aligned with the PRDM1/Blimp-1 consensus sequence (19, 28, 41, 43), there was poor association of PRDM1 with the TAP1/LMP2 promoter (Fig. 6, A and D). In contrast, both IRF-1 and IRF-2 showed high levels of association with TAP1/LMP2 both before and after stimulation but showed only low-level association with CIITA-pIV (Fig. 6, B and C). The promoter of the negative control KRT-10 showed minimal association with PRDM1, IRF-1, or IRF-2 (Fig. 6, A–C). Together, these data demonstrate that the weak response of the CIITA-pIV to IFN-γ stimulation in H929 occurs despite expression of IRF-1 and IRF-2 and is reflected in binding of PRDM1 but not IRF-1 or IRF-2 to CIITA-pIV.

FIGURE 6.

PRDM1 and IRF-1/IRF-2 show reciprocal patterns of binding at CIITA-pIV and the TAP1/LMP2 promoter. A, Chromatin prepared from H929 cells after treatment with medium alone or 200 IU/ml IFN-γ for 24 h was immunoprecipitated with control rabbit IgG or anti-PRDM1, and the percentage input of CIITA-pIV, TAP1/LMP2 promoter, or KRT-10 promoter sequence was determined by real-time PCR. Percentage input of promoter sequences after precipitation with Ab to IRF-1 (B) or IRF-2 (C) from the same chromatin. These data were derived from four separate immunoprecipitations performed with each Ab and are displayed as the mean ± SD. D, Alignment of TAP1/LMP2 promoter sequence with PRDM1/Blimp-1 consensus sequence.

FIGURE 6.

PRDM1 and IRF-1/IRF-2 show reciprocal patterns of binding at CIITA-pIV and the TAP1/LMP2 promoter. A, Chromatin prepared from H929 cells after treatment with medium alone or 200 IU/ml IFN-γ for 24 h was immunoprecipitated with control rabbit IgG or anti-PRDM1, and the percentage input of CIITA-pIV, TAP1/LMP2 promoter, or KRT-10 promoter sequence was determined by real-time PCR. Percentage input of promoter sequences after precipitation with Ab to IRF-1 (B) or IRF-2 (C) from the same chromatin. These data were derived from four separate immunoprecipitations performed with each Ab and are displayed as the mean ± SD. D, Alignment of TAP1/LMP2 promoter sequence with PRDM1/Blimp-1 consensus sequence.

Close modal

Expression of IRF-1 and IRF-2 is sufficient to drive CIITA-pIV activation in the absence of PRDM1 (Fig. 3,C) (7, 30). In a cell line expressing PRDM1 as well as IRF-1 and IRF-2, reducing the level of PRDM1 would be predicted to lead to CIITA-pIV activation in the absence of extracellular stimuli. We, therefore, sought to knockdown PRDM1 expression in H929 using siRNA but were unable to achieve significant reduction. In contrast, we were able to achieve substantial reductions in PRDM1 mRNA and protein in the myeloma cell line U266 (Fig. 7 A and data not shown).

FIGURE 7.

CIITA-pIV activity is regulated by the level of PRDM1 in the myeloma cell line U266. A, U266 cells were treated with lipofectamine alone (none) or transfected with either an oligonucleotide designed to mediate siRNA knockdown of PRDM1 (PRDM1) or a control oligonucleotide (control). The resulting level of PRDM1 protein was assessed by Western blotting. B, The expression level of PRDM1 protein in the myeloma cell lines U266 and H929 and the B cell line Ramos was determined by Western blotting with anti-PRDM1 specific for the N terminus. C, The level of CIITA transcripts originating from pIV relative to GAPDH was determined by real-time PCR in HeLa or U266 cells treated with medium alone or IFN-γ (200 IU/ml) for 24 h. Data are displayed relative to untreated HeLa cells. D, Chromatin prepared from H929 or U266 cells was immunoprecipitated with control rabbit IgG or rabbit anti-PRDM1 and the quantity of KRT-10 promoter, CIITA-pIII or -pIV, and TAP1/LMP2 sequence present was assessed by real-time PCR. Chromatin prepared from U266 cells was immunoprecipitated with anti-IRF-1 (E) or anti-IRF-2 (F) and the percentage input of CIITA-pIV, TAP1/LMP2 promoter, or KRT-10 promoter was assayed by real-time PCR. These data were derived from two separate immunoprecipitations performed with each Ab and are displayed as the mean ± SD. G, Expression of CIITA from pIII and pIV was evaluated by real-time PCR in U266 cells left untreated or transfected with control oligonucleotide or siRNA to PRDM1. Data are expressed as the mean ± SD derived from duplicate samples. Data are representative of four independent experiments.

FIGURE 7.

CIITA-pIV activity is regulated by the level of PRDM1 in the myeloma cell line U266. A, U266 cells were treated with lipofectamine alone (none) or transfected with either an oligonucleotide designed to mediate siRNA knockdown of PRDM1 (PRDM1) or a control oligonucleotide (control). The resulting level of PRDM1 protein was assessed by Western blotting. B, The expression level of PRDM1 protein in the myeloma cell lines U266 and H929 and the B cell line Ramos was determined by Western blotting with anti-PRDM1 specific for the N terminus. C, The level of CIITA transcripts originating from pIV relative to GAPDH was determined by real-time PCR in HeLa or U266 cells treated with medium alone or IFN-γ (200 IU/ml) for 24 h. Data are displayed relative to untreated HeLa cells. D, Chromatin prepared from H929 or U266 cells was immunoprecipitated with control rabbit IgG or rabbit anti-PRDM1 and the quantity of KRT-10 promoter, CIITA-pIII or -pIV, and TAP1/LMP2 sequence present was assessed by real-time PCR. Chromatin prepared from U266 cells was immunoprecipitated with anti-IRF-1 (E) or anti-IRF-2 (F) and the percentage input of CIITA-pIV, TAP1/LMP2 promoter, or KRT-10 promoter was assayed by real-time PCR. These data were derived from two separate immunoprecipitations performed with each Ab and are displayed as the mean ± SD. G, Expression of CIITA from pIII and pIV was evaluated by real-time PCR in U266 cells left untreated or transfected with control oligonucleotide or siRNA to PRDM1. Data are expressed as the mean ± SD derived from duplicate samples. Data are representative of four independent experiments.

Close modal

To determine whether U266 cells represent a comparable model to H929, we examined PRDM1, IRF-1, and IRF-2 protein expression, as well as CIITA-pIV mRNA and promoter occupancy by ChIP. U266 cells expressed similar levels of PRDM1, IRF-1, and IRF-2 protein to those seen in H929 (Figs. 5,C and 7B). In U266 cells, CIITA-pIV mRNA was present both in untreated and IFN-γ treated cells, but at a low level, compared with that reached in IFN-γ-treated HeLa (Fig. 7,C). PRDM1 was associated with CIITA-pIII and CIITA-pIV to a similar degree in U266 and H929 (Fig. 7,D). IRF-1 and IRF-2 showed only marginally greater association with CIITA-pIV than the control KRT-10 promoter but were strongly associated with the TAP1/LMP2 promoter (Fig. 7, E and F). PRDM1 was not associated with TAP1/LMP2 in U266, and both TAP1 and LMP2 mRNA were expressed at high levels (Fig. 7 D and data not shown). The pattern of CIITA-pIV occupancy by PRDM1 and IRF-1/IRF-2 is, therefore, similar in U266 and H929. The constitutive low-level CIITA-pIV expression observed in U266 is, therefore, not attributable to a shift from PRDM1 to IRF-1/IRF-2 promoter binding, and other elements, such as the STAT-responsive GAS site most likely account for the basal promoter activity.

If PRDM1 is competing with IRF-1/IRF-2 to repress promoter activity in U266, then decreasing PRDM1 protein expression should lead to an increase in CIITA-pIV mRNA in the absence of IFN-γ stimulation. As predicted, knockdown of PRDM1 led to a five-fold increase in CIITA-pIV expression (Fig. 7 G). Moreover, a similar increase was observed from CIITA-pIII. These data indicate that the activity of CIITA-pIV is modulated by the level of PRDM1 in these cells and are consistent with competition between PRDM1 and IRF-1/IRF-2 for promoter regulation.

We conclude that PRDM1/Blimp-1 is a transcriptional repressor of CIITA-pIV, binding at the IRF-E site of this promoter to inhibit IFN-γ and IRF-1/IRF-2-dependent induction. The original identification of PRDM1 as a transcriptional repressor binding to the PRDI of the IFN-β promoter was the first indication that it competes with IRFs for promoter occupancy (13, 28). These data have been confirmed and extended in a recent paper, which defined the Blimp-1 consensus sequence and the general overlap of this sequence with IRF-1/IRF-2 binding sites (19). At the IFN-β promoter, the induction of PRDM1/Blimp-1 following cellular viral infection leads to competitive binding at the PRDI site (13, 19). Our data demonstrate that this model also is applicable to an IFN-γ responsive promoter, and that PRDM1/Blimp-1 expression may alter the cellular response to this cytokine. We find a close correlation between PRDM1 expression levels and activity of the CIITA-pIV, both in the luciferase assay system and in a myeloma cell line. Our results indicate that, with high levels of PRDM1, the response of CIITA-pIV to IFN-γ is largely eliminated. This is consistent with the requirement for IRF binding demonstrated previously for CIITA-pIV induction (6, 7, 30). At low levels of PRDM1 expression, the response of CIITA-pIV is reduced but not eliminated, thus PRDM1/Blimp-1 may act to dampen, as well as silence, the IFN-γ response.

PRDM1/Blimp-1 can mediate transcriptional repression by a number of mechanisms, including the recruitment of epigenetic modifiers (20, 21, 22). The fact that the Δ527 PRDM1 construct was sufficient to mediate CIITA-pIV repression is consistent with a primary role for promoter occupancy, as this truncated protein includes the DNA binding domain but lacks the regions necessary for recruitment of HDAC and Groucho (20, 21). Nevertheless, our data do not exclude a role for epigenetic modification in PRDM1/Blimp-1-mediated repression of CIITA-pIV.

Because not all IRF-1/IRF-2 sites match the PRDM1/Blimp-1 consensus, only a subset of IFN-γ responsive target genes will be modulated in this fashion. This is demonstrated by the inability of PRDM1 to repress a CIITA-pIV IRF-E mutant that has been altered to preserve IRF-1/IRF-2 binding while eliminating the PRDM1 consensus. Moreover the bidirectional promoter regulating TAP1/LMP2 includes an IRF-E consensus sequence, which diverges from the PRDM1/Blimp-1 consensus at only one position. Nevertheless, PRDM1 does not substantially associate with the TAP1/LMP2 promoter, which is constitutively active, IFN-γ inducible, and bound by IRF-1 and IRF-2 in the myeloma cell line H929.

In B cells, transcription of CIITA initiates at both pIII and pIV (32). Along with published data (11, 12), our results demonstrate that PRDM1/Blimp-1 functions as a repressor of both of these promoters. These findings suggest that PRDM1/Blimp-1 acts in a comprehensive manner to extinguish MHC-II expression in this lineage. The biological relevance for this may be provided by the migration of plasma cells to sites of inflammation. Following immunization, the majority of newly generated plasma cells migrate to the bone marrow (44). In this case, PRDM1/Blimp-1 can mediate repression of CIITA at pIII to turn off the expression of MHC-II. However, a portion of plasma cells migrate to sites of inflammation (45, 46, 47). The expression of PRDM1/Blimp-1 in these plasma cells is likely to prevent the up-regulation of CIITA from pIV in environments of ongoing IFN-γ production. PRDM1/Blimp-1 also is expressed in macrophages and epithelia in which CIITA-pIII is not active. In these cell types, CIITA-pIV is IFN-γ inducible. Repression of CIITA-pIV provides a novel target for PRDM1/Blimp-1-dependent regulation in such cells and suggests that PRDM1/Blimp-1 expression in these lineages would oppose IFN-γ-mediated MHC-II induction and Ag presentation. We have shown that cellular stress, including the unfolded protein response of the endoplasmic reticulum, induces PRDM1/Blimp-1 in monocyte/macrophages (35). Therefore, PRDM1/Blimp-1 may act in this context to regulate MHC-II expression.

It has been suggested previously that pIV of CIITA may provide a common target for developmental, pathological, and pharmacological inhibition of MHC-II expression (1). Our data provide the first evidence of a factor capable of blocking transcription from CIITA-pIV by direct promoter binding and suggest an important role for PRDM1/Blimp-1 in repressing MHC-II expression in inducible cell lineages.

We thank Graham Cook and Peter Cockerill for their critical reading of this manuscript and helpful suggestions. We are grateful to the members of the Bonifer laboratory for valuable discussions and technical advice.

The authors have no financial conflict of interest.

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.

1

This work was supported by a Medical Research Council Clinician Scientist Fellowship (R.M.T.) and a Leeds University Research Fellowship (G.M.D.).

3

Abbreviations used in this paper: MHC-II, MHC class II; Blimp-1, B lymphocyte-induced maturation protein-1; FL, full length; PRDI, positive regulatory domain I; PRDM1, PR-domain containing protein-1; IRF, IFN regulatory factor; IRF-E, IRF-element; ChIP, chromatin immunoprecipitation; p, promoter; GAS, IFN-γ activation sequence; siRNA, short interfering RNA.

1
Reith, W., B. Mach.
2001
. The bare lymphocyte syndrome and the regulation of MHC expression.
Annu. Rev. Immunol.
19
:
331
-373.
2
Muhlethaler-Mottet, A., L. A. Otten, V. Steimle, B. Mach.
1997
. Expression of MHC class II molecules in different cellular and functional compartments is controlled by differential usage of multiple promoters of the transactivator CIITA.
EMBO J.
16
:
2851
-2860.
3
Masternak, K., A. Muhlethaler-Mottet, J. Villard, M. Zufferey, V. Steimle, W. Reith.
2000
. CIITA is a transcriptional coactivator that is recruited to MHC class II promoters by multiple synergistic interactions with an enhanceosome complex.
Genes Dev.
14
:
1156
-1166.
4
Ting, J. P., J. Trowsdale.
2002
. Genetic control of MHC class II expression.
Cell
109
:
S21
-S33.
5
Reith, W., S. LeibundGut-Landmann, J. M. Waldburger.
2005
. Regulation of MHC class II gene expression by the class II transactivator.
Nat. Rev. Immunol
5
:
793
-806.
6
Muhlethaler-Mottet, A., W. Di Berardino, L. A. Otten, B. Mach.
1998
. Activation of the MHC class II transactivator CIITA by interferon-γ requires cooperative interaction between Stat1 and USF-1.
Immunity
8
:
157
-166.
7
Piskurich, J. F., M. W. Linhoff, Y. Wang, J. P. Ting.
1999
. Two distinct γ interferon-inducible promoters of the major histocompatibility complex class II transactivator gene are differentially regulated by STAT1, interferon regulatory factor 1, and transforming growth factor β.
Mol. Cell Biol.
19
:
431
-440.
8
Piskurich, J. F., Y. Wang, M. W. Linhoff, L. C. White, J. P. Ting.
1998
. Identification of distinct regions of 5′ flanking DNA that mediate constitutive, IFN-γ, STAT1, and TGF-β-regulated expression of the class II transactivator gene.
J. Immunol.
160
:
233
-240.
9
Tominaga, N., S. Katagiri, M. Ohnishi, H. Nakao, K. Oritani, H. Yagura, T. Tamaki, Y. Kanayama, T. Yonezawa, S. Tarui.
1989
. Analysis of surface antigen expression of human immunoglobulin-secreting cells: phenotypic heterogeneity in normal counterparts of myeloma cells.
Br. J. Haematol.
73
:
302
-308.
10
Burrows, P. D., M. D. Cooper.
1990
. Regulated expression of cell surface antigens during B cell development.
Semin. Immunol.
2
:
189
-195.
11
Piskurich, J. F., K. I. Lin, Y. Lin, Y. Wang, J. P. Ting, K. Calame.
2000
. BLIMP-I mediates extinction of major histocompatibility class II transactivator expression in plasma cells.
Nat. Immunol.
1
:
526
-532.
12
Ghosh, N., I. Gyory, G. Wright, J. Wood, K. L. Wright.
2001
. Positive regulatory domain I binding factor 1 silences class II transactivator expression in multiple myeloma cells.
J. Biol. Chem.
276
:
15264
-15268.
13
Keller, A. D., T. Maniatis.
1991
. Identification and characterization of a novel repressor of β-interferon gene expression.
Genes Dev.
5
:
868
-879.
14
Turner, C. A., Jr, D. H. Mack, M. M. Davis.
1994
. Blimp-1, a novel zinc finger-containing protein that can drive the maturation of B lymphocytes into immunoglobulin-secreting cells.
Cell
77
:
297
-306.
15
Shapiro-Shelef, M., K. I. Lin, L. J. McHeyzer-Williams, J. Liao, M. G. McHeyzer-Williams, K. Calame.
2003
. Blimp-1 is required for the formation of immunoglobulin secreting plasma cells and pre-plasma memory B cells.
Immunity
19
:
607
-620.
16
Lin, Y., K. Wong, K. Calame.
1997
. Repression of c-myc transcription by Blimp-1, an inducer of terminal B cell differentiation.
Science
276
:
596
-599.
17
Lin, K. I., C. Angelin-Duclos, T. C. Kuo, K. Calame.
2002
. Blimp-1-dependent repression of Pax-5 is required for differentiation of B cells to immunoglobulin M-secreting plasma cells.
Mol. Cell Biol.
22
:
4771
-4780.
18
Shaffer, A. L., K. I. Lin, T. C. Kuo, X. Yu, E. M. Hurt, A. Rosenwald, J. M. Giltnane, L. Yang, H. Zhao, K. Calame, L. M. Staudt.
2002
. Blimp-1 orchestrates plasma cell differentiation by extinguishing the mature B cell gene expression program.
Immunity
17
:
51
-62.
19
Kuo, T. C., K. L. Calame.
2004
. B lymphocyte-induced maturation protein (Blimp)-1, IFN regulatory factor (IRF)-1, and IRF-2 can bind to the same regulatory sites.
J. Immunol.
173
:
5556
-5563.
20
Ren, B., K. J. Chee, T. H. Kim, T. Maniatis.
1999
. PRDI-BF1/Blimp-1 repression is mediated by corepressors of the Groucho family of proteins.
Genes Dev.
13
:
125
-137.
21
Yu, J., C. Angelin-Duclos, J. Greenwood, J. Liao, K. Calame.
2000
. Transcriptional repression by blimp-1 (PRDI-BF1) involves recruitment of histone deacetylase.
Mol. Cell Biol.
20
:
2592
-2603.
22
Gyory, I., J. Wu, G. Fejer, E. Seto, K. L. Wright.
2004
. PRDI-BF1 recruits the histone H3 methyltransferase G9a in transcriptional silencing.
Nat. Immunol.
5
:
299
-308.
23
Chang, D. H., G. Cattoretti, K. L. Calame.
2002
. The dynamic expression pattern of B lymphocyte induced maturation protein-1 (Blimp-1) during mouse embryonic development.
Mech. Dev.
117
:
305
-309.
24
Chang, D. H., C. Angelin-Duclos, K. Calame.
2000
. BLIMP-1: trigger for differentiation of myeloid lineage.
Nat. Immunol.
1
:
169
-176.
25
Cattoretti, G., C. Angelin-Duclos, R. Shaknovich, H. Zhou, D. Wang, B. Alobeid.
2005
. PRDM1/Blimp-1 is expressed in human B-lymphocytes committed to the plasma cell lineage.
J. Pathol.
206
:
76
-86.
26
Kallies, A., E. D. Hawkins, G. T. Belz, D. Metcalf, M. Hommel, L. M. Corcoran, P. D. Hodgkin, S. L. Nutt.
2006
. Transcriptional repressor Blimp-1 is essential for T cell homeostasis and self-tolerance.
Nat. Immunol.
7
:
466
-474.
27
Martins, G. A., L. Cimmino, M. Shapiro-Shelef, M. Szabolcs, A. Herron, E. Magnusdottir, K. Calame.
2006
. Transcriptional repressor Blimp-1 regulates T cell homeostasis and function.
Nat. Immunol.
7
:
457
-465.
28
Keller, A. D., T. Maniatis.
1991
. Selection of sequences recognized by a DNA binding protein using a preparative southwestern blot.
Nucleic Acids Res.
19
:
4675
-4680.
29
Morris, A. C., G. W. Beresford, M. R. Mooney, J. M. Boss.
2002
. Kinetics of a γ interferon response: expression and assembly of CIITA promoter IV and inhibition by methylation.
Mol. Cell Biol.
22
:
4781
-4791.
30
Xi, H., D. D. Eason, D. Ghosh, S. Dovhey, K. L. Wright, G. Blanck.
1999
. Co-occupancy of the interferon regulatory element of the class II transactivator (CIITA) type IV promoter by interferon regulatory factors 1 and 2.
Oncogene
18
:
5889
-5903.
31
Xi, H., G. Blanck.
2003
. The IRF-2 DNA binding domain facilitates the activation of the class II transactivator (CIITA) type IV promoter by IRF-1.
Mol. Immunol.
39
:
677
-684.
32
Piskurich, J. F., C. A. Gilbert, B. D. Ashley, M. Zhao, H. Chen, J. Wu, S. C. Bolick, K. L. Wright.
2005
. Expression of the MHC class II transactivator (CIITA) type IV promoter in B lymphocytes and regulation by IFN-γ.
Mol. Immunol.
:
33
Waldburger, J. M., T. Suter, A. Fontana, H. Acha-Orbea, W. Reith.
2001
. Selective abrogation of major histocompatibility complex class II expression on extrahematopoietic cells in mice lacking promoter IV of the class II transactivator gene.
J. Exp. Med.
194
:
393
-406.
34
Bert, A. G., J. Burrows, C. S. Osborne, P. N. Cockerill.
2000
. Generation of an improved luciferase reporter gene plasmid that employs a novel mechanism for high-copy replication.
Plasmid
44
:
173
-182.
35
Doody, G. M., S. Stephenson, R. M. Tooze.
2006
. BLIMP-1 is a target of cellular stress and downstream of the unfolded protein response.
Eur. J. Immunol.
36
:
1572
-1582.
36
Schreiber, E., P. Matthias, M. M. Muller, W. Schaffner.
1989
. Rapid detection of octamer binding proteins with ‘mini-extracts’, prepared from a small number of cells.
Nucleic Acids Res.
17
:
6419
37
Follows, G. A., H. Tagoh, P. Lefevre, D. Hodge, G. J. Morgan, C. Bonifer.
2003
. Epigenetic consequences of AML1-ETO action at the human c-FMS locus.
EMBO J.
22
:
2798
-2809.
38
Fujii, Y., T. Shimizu, M. Kusumoto, Y. Kyogoku, T. Taniguchi, T. Hakoshima.
1999
. Crystal structure of an IRF-DNA complex reveals novel DNA recognition and cooperative binding to a tandem repeat of core sequences.
EMBO J.
18
:
5028
-5041.
39
Tanaka, N., T. Kawakami, T. Taniguchi.
1993
. Recognition DNA sequences of interferon regulatory factor 1 (IRF-1) and IRF-2, regulators of cell growth and the interferon system.
Mol. Cell Biol.
13
:
4531
-4538.
40
Escalante, C. R., J. Yie, D. Thanos, A. K. Aggarwal.
1998
. Structure of IRF-1 with bound DNA reveals determinants of interferon regulation.
Nature
391
:
103
-106.
41
White, L. C., K. L. Wright, N. J. Felix, H. Ruffner, L. F. Reis, R. Pine, J. P. Ting.
1996
. Regulation of LMP2 and TAP1 genes by IRF-1 explains the paucity of CD8+ T cells in IRF-1−/− mice.
Immunity
5
:
365
-376.
42
Rouyez, M. C., M. Lestingi, M. Charon, S. Fichelson, A. Buzyn, I. Dusanter-Fourt.
2005
. IFN regulatory factor-2 cooperates with STAT1 to regulate transporter associated with antigen processing-1 promoter activity.
J. Immunol.
174
:
3948
-3958.
43
Wright, K. L., L. C. White, A. Kelly, S. Beck, J. Trowsdale, J. P. Ting.
1995
. Coordinate regulation of the human TAP1 and LMP2 genes from a shared bidirectional promoter.
J. Exp. Med.
181
:
1459
-1471.
44
Manz, R. A., A. Thiel, A. Radbruch.
1997
. Lifetime of plasma cells in the bone marrow.
Nature
388
:
133
-134.
45
Hauser, A. E., G. F. Debes, S. Arce, G. Cassese, A. Hamann, A. Radbruch, R. A. Manz.
2002
. Chemotactic responsiveness toward ligands for CXCR3 and CXCR4 is regulated on plasma blasts during the time course of a memory immune response.
J. Immunol.
169
:
1277
-1282.
46
Muehlinghaus, G., L. Cigliano, S. Huehn, A. Peddinghaus, H. Leyendeckers, A. E. Hauser, F. Hiepe, A. Radbruch, S. Arce, R. A. Manz.
2005
. Regulation of CXCR3 and CXCR4 expression during terminal differentiation of memory B cells into plasma cells.
Blood
105
:
3965
-3971.
47
Odendahl, M., H. Mei, B. F. Hoyer, A. M. Jacobi, A. Hansen, G. Muehlinghaus, C. Berek, F. Hiepe, R. Manz, A. Radbruch, T. Dorner.
2005
. Generation of migratory antigen-specific plasma blasts and mobilization of resident plasma cells in a secondary immune response.
Blood
105
:
1614
-1621.