The Role of Activating Protein 1 in the Transcriptional Regulation of the Human FCGR2B Promoter Mediated by the -343 G → C Polymorphism Associated with Systemic Lupus Erythematosus*

The inhibitory receptor FcγRIIb is a negative regulator of antibody production and inflammatory responses. The -343 G → C polymorphism in the human FCGR2B promoter is associated with systemic lupus erythematosus. The -343 C mutant promoter has decreased transcriptional activity. In the present study, we show that the transcriptional change correlates with quantitative differences in the interaction of the activating protein 1 complex with the mutant FCGR2B promoter. Promoter pulldown and chromatin immunoprecipitation assays demonstrated binding of c-Jun to the FCGR2B promoter. Phosphorylation of c-Jun was accompanied by transactivation of both FCGR2B promoter variants, whereas dephosphorylation of c-Jun by an inhibitor of c-Jun N-terminal kinase, markedly decreased the promoter activities. The -343 G → C substitution enabled the specific interaction of the transcription factor Yin-Yang 1 with the mutant FCGR2B promoter. Yin-Yang 1 competed with activating protein 1 for binding at the -343 site, and contributed to the repression of the mutant FCGR2B promoter activity. This mechanism could be responsible for the decreased expression of FcγRIIb associated with the -343 C/C homozygous FCGR2B genotype in lupus patients. These findings provide a rationale for the transcriptional defect mediated by the -343 C/C FCGR2B promoter polymorphism associated with systemic lupus erythematosus, and add to our understanding of the complex transcriptional regulation of the human FCGR2B promoter.

Fc ␥-receptors (Fc␥R) 2 bind IgG-containing immune complexes and mediate important immune functions such as phag-ocytosis, degranulation, antibody-dependent cellular cytotoxicity, and production of inflammatory mediators. Human hematopoetic cells express several Fc␥R isoforms encoded by seven separate genes. Unlike activating Fc␥R, Fc␥RIIb is unique in that it contains an immunoreceptor tyrosine-based inhibitory motif in the intracellular domain. A specific amino acid sequence in the immunoreceptor tyrosine-based inhibitory motif domain of Fc␥RIIb allows the recruitment of phosphatases and the initiation of inhibitory signaling. Cross-linking of Fc␥RIIb with activating receptors on B cells and mononuclear phagocytes leads to down-regulation of antibody production, phagocytosis, and cytokine secretion. Several lines of evidence demonstrate that Fc␥RIIb is important in the maintenance of self-tolerance. Fc␥RIIb deficiency is associated with spontaneous development of autoimmune manifestations in several mouse genetic backgrounds (1). Autoimmune prone mouse strains share an Fcgr2 promoter haplotype containing deletions and polymorphisms associated with reduced expression of Fc␥RIIb on the surface of activated B cells and macrophages (2)(3)(4). Whereas the engineered deletion and the natural deficiency in Fc␥RIIb confer susceptibility to development of lupus-like disease, transplantation of bone marrow cells transduced with Fc␥RIIb-expressing retrovirus restored the healthy phenotype (5). This body of data created the impetus for the study of Fc␥RIIb regulation and function in humans.
Genetic studies revealed a significant association of a single nucleotide polymorphism (SNP) encoding the amino acid substitution I232T in the transmembrane of Fc␥RIIb with systemic lupus erythematosus (SLE) in several racial groups (6 -8). The SLE-associated transmembrane polymorphism (Fc␥RIIb-Thr 232 ) induced loss of inhibitory function of Fc␥RIIb through exclusion from sphingolipid rafts (9). In addition, several SNPs present in the FCGR2B promoter correlated with differences in regulatory function and associated with autoimmunity in certain racial groups (10,11). The Ϫ343 G 3 C substitution identified by us in the proximal FCGR2B promoter correlates with decreased transcriptional activity and associates with SLE in European-Americans (11).
In the present study, we investigated the regulation of the human proximal FCGR2B promoter activity, and the mechanism underlying the defect mediated by the Ϫ343 C SNP associated with SLE. The Ϫ343 G 3 C substitution correlated with decreased promoter activity in transiently transfected cells. Electrophoretic mobility shift assays (EMSA) and supershift experiments indicated differences in the binding of activating protein 1 (AP-1) family members to the two FCGR2B promoter sequence variants, with decreased binding of c-Jun to the mutant sequence. The Ϫ343 G 3 C substitution created a new binding site for the nuclear factor Yin-Yang 1 (YY1) that interacted specifically with the mutant sequence competing with c-Jun for DNA binding. The inhibition of YY1 synthesis with specific short hairpin RNA (shRNA) resulted in the up-regulation of the mutant promoter activity. We identify a shift in the binding of YY1 versus c-Jun to the mutant promoter, and this mechanism could be responsible for the repression of FCGR2B promoter activity in SLE patients bearing the Ϫ343 C/C genotype. Our study has implications for the modulation of FCGR2B gene expression in SLE and other diseases having altered expression and function of inhibitory Fc␥RIIb as pathogenic factor.
Plasmids-The Ϫ537 to ϩ43 fragment of the human FCGR2B promoter was amplified with primers containing recognition sites for KpnI in the forward primer (5Ј-GCGCGGTACCGCCATCCTGACATACCTCCTT-3Ј) and XhoI in the reverse primer (5-GCGCCTCGAGCACTC-CCTGGAGCGACGTGGC-3Ј). The products were directionally ligated using a rapid DNA ligation kit (Roche) into the KpnI/XhoI site of the pGL3-Enhancer vectors (Promega, Madison, WI). Constructs containing G at position Ϫ343 were used as template to create Ϫ343 C constructs by sitedirected mutagenesis using the QuikChange kit (Stratagene). The c-Jun trans-Reporting System PathDetect Kit containing the pFC-MEKK plasmid expressing mitogen-activated/extracellular signal-regulated kinase kinase kinase (MEKK) and the pFA2-cJUN plasmid expressing c-Jun was obtained from Stratagene. The pCMV-YY1 vector expressing full-size YY1 was obtained from OriGene.
Cell Transfection and Reporter Assays-Transient transfection of CL-01 cells was performed with Lipofectamine (Invitrogen) according to the manufacturer's instructions. CL-01 cells (1.6 ϫ 10 6 cell/ml) in Opti-MEM I medium (Invitrogen) were incubated with plasmid-Lipofectamine complexes for 5 h, and then supplemented with complete medium and cultured for 48 h. Firefly and Renilla luciferase activities were measured in cell lysates by the dual luciferase assay kit (Promega) using a Wallac MicroBetta Trilux 1450 microplate luminometer. Where indicated, CL-01 cells were co-transfected with shRNA against YY1 1 day prior to the luciferase reporter assay. The TranSilent YY1 shRNA and TranSilent control vector (Panomics) were transfected into CL-01 using AMAXA nucleofector kit. Where indicated, co-transfection with pFC-MEKK, pFA2-cJUN, and pCMV-YY1 was performed.
Promoter Pulldown Assay-Biotinylated WT, MUT, and AP-1 oligonucleotides (Operon, Inc.) were annealed and coupled to streptavidin-agarose CL-4B (Sigma). Nuclear extracts from activated CL-01 cells were incubated overnight with agarose-coupled oligonucletides. Agarose suspensions containing bound nuclear factors were applied to minicolumns, washed, and the bound nuclear proteins were eluted with 400 mM NaCl and separated in 10% SDS-PAGE. Immunoblotting was carried out with anti-YY1, -c-Fos, and -phospho-c-JunP73 antibodies.
Chromatin Immunoprecipitation (ChIP) Assay-ChIP assay was performed according to a published procedure, with several modifications (13). Briefly, CL-01 cells, harvested at baseline or after activation, were treated with 1% formaldehyde for 15 min. The cells were lysed and sonicated with SONICATOR 3000 (MISONIX). Supernatants were incubated overnight at 4°C with anti c-Jun antibodies (sc-45X, Santa Cruz Biotechnology) and anti-XBP (sc-7160X, Santa Cruz Biotechnology), followed by incubation with protein A-agarose in the presence of salmon sperm DNA. The chromatin fraction was eluted, treated with proteinase K, and precipitated. The DNA fragments were amplified by PCR using primers spanning regions Ϫ495 to Ϫ472 and Ϫ278 to Ϫ258 of the FCGR2B promoter.

RESULTS
We investigated the binding of nuclear factors to the nucleotide sequence (Ϫ369 to Ϫ330) containing the Ϫ343 G 3 C substitution (dbSNP number rs3219018) identified by us in the human FCGR2B promoter (11). Two Fc␥RIIb-expressing cell types, the B lymphoma cell line CL-01 and the myelomonocytic leukemia cell line U937, were used as sources for nuclear extracts. EMSA experiments were carried out with doublestranded oligonucleotide probes labeled with [␣-32 P]dCTP. The WT probe corresponded to the sequence of the common FCGR2B promoter that contained Ϫ343 G, whereas the MUT probe contained Ϫ343 C (sequences are shown in Table 1). Analysis of the EMSA pattern revealed that the WT and MUT probes formed distinct DNA-protein complexes (Fig. 1, A and  B). The WT probe formed complex b, and the MUT probe formed complexes b and c with nuclear extract from resting CL-01 cells (Fig. 1A, lanes 2 and 3). Nuclear extracts from U937 incubated with the MUT probe allowed the formation of complex c (Fig. 1B,  lanes 2 and 3). Complex b was not evident with nuclear extracts from U937 cells, suggesting cell-specific differences in the binding of nuclear factors to the FCGR2B promoter probes.
To test the effect of cell activation on the binding of nuclear factors to the WT and MUT probes, we treated CL-01 cells and U937 cells with PMA:ION. Two slow migrating DNA-protein complexes, a1 and a2, were formed when the WT probe was incubated with nuclear extracts from PMA:ION-treated CL-01 and U937 cells (Fig. 1, A and B, lane 5). Complex a1 was predominant in activated CL-01 cells, whereas complex a2 was predominant in activated U937 cells, indicating differences in the formation of the activation-induced complexes in the two cell types (Fig. 1, A and B, lane 5). The formation of complex a1 and a2 with the MUT probe was weak in activated CL-01 and U937 cells (Fig. 1,  A and B, lane 6).  7). C, the sequence of the human FCGR2B promoter between residues Ϫ349 and Ϫ330 containing the Ϫ343 G 3 C SNP. Consensus sequences for transcription factors AP-1 and YY1 aligned with the homologous sequence in FCGR2B promoter. D, Western blots with nuclear extracts from resting and activated CL-01 or U937 cells overlaid with anti-c-Jun-P73, anti-JunB, and anti-c-Fos antibodies.

TABLE 1 Probes and competitors used for EMSA
Computer-based analysis of the FCGR2B sequence containing the Ϫ343 site indicated similarity with the core consensus binding sequence for AP-1 (TGACTCA) (Fig. 1C). We compared the migration pattern of complex a1 and a2 with that of AP-1 by generating a probe (AP-1 probe) containing the canonical AP-1 binding motif (TGACTCA) within the sequence of the FCGR2B promoter (Fig. 1C). The AP-1 probe formed bands with similar mobility with complexes a1 and a2 in activated CL-01 and U937 cells (Fig. 1, A and B, lane 7). These results suggested that AP-1 participated in the formation of activationinduced complex a1 and a2. The expression level of the AP-1 family members, c-Jun, JunB, and c-Fos in nuclear extracts was analyzed in Western blots. The activation of CL-01 and U937 cells with PMA:ION induced the phosphorylation and nuclear translocation of c-Jun, JunB, and c-Fos (Fig. 1D).
To verify the interaction of complexes a1 and a2 with the AP-1 binding sequence, we created an additional probe AP-1 neg bearing a mutated sequence (TAAGAC) of the AP-1 core consensus binding motif ( Table 1). The AP-1 neg probe did not form activation-induced complexes a1 and a2 ( Fig. 2A, lanes 1 and 4). The nuclear factors that participate in the formation of complex b interacted with the AP-1 neg probe, indicating that complex b bound outside the Ϫ343 to Ϫ340 nucleotide sequence ( Fig. 2A, lane 4).
To localize the interaction of nuclear proteins with the WT and MUT promoter sequences, we synthesized a competitor probe corresponding to the FCGR2B promoter left flanking sequence excluding the Ϫ343 site (C L ). In addition, we synthesized competitor probes corresponding to the FCGR2B promoter right flanking sequence containing Ϫ343 G (C WT ) and Ϫ343 C (C MUT ) ( Table 1). We performed EMSA with 200-fold excess of each unlabeled competitor probe. The addition of competitor C L abolished the formation of complex b with the WT and MUT probes ( Fig. 2A, lanes 5 and 6), suggesting that the interaction of complex b with the FCGR2B promoter sequence was upstream and outside of the Ϫ343 G/C polymorphic region. Competitor C L added in excess did not abolish the formation of complexes a1 and c ( Fig. 2A, lanes 5 and 6). Competitors C WT and C MUT did not influence the formation of complex b ( Fig. 2A, lanes 7-10). C WT eliminated complex a1, but did not affect the formation of the MUT-specific complex c ( Fig. 2A, lanes 7 and 8). C MUT eliminated the formation of complex a1 with the WT, and the formation of complex c with the MUT probe ( Fig. 2A, lanes 9 and 10). Complex a1 was eliminated by 200-fold excess of AP-1 oligonucleotide competitor, confirming the interaction of AP-1 factors with the WT promoter sequence at that site ( Fig. 2A, lanes 13 and 15). The excess SP1 competitor did not affect the formation of complex a1 ( Fig. 2A, lanes 14 and 17).
We sought to identify the nuclear factors participating in the formation of the AP-1 transcriptional complex. Antibodies against SP1, used as control, did not form supershifts and did not eliminate the formation of the activation-induced complexes a1 and a2 (Fig. 2B, lane 1). Antibodies specific for c-Fos, c-Jun, and Jun-B enabled the formation of supershift bands with the WT probe (Fig. 2B, lanes 2-4). The incubation of nuclear extract from CL-01 cells with anti-c-Jun antibodies eliminated complex a1 (Fig. 2B, lane 3), whereas anti-c-Fos antibodies eliminated complex a2 (Fig. 2B, lane 2). These results suggested that in CL-01 cells, complex a1 is formed preferentially by c-Jun, and that c-Fos participates mainly in the formation of complex a2.
We investigated the direct interaction of c-Jun with the FCGR2B promoter by ChIP assay. Nuclear extracts were prepared from resting CL-01 cells (0 h) and CL-01 cells activated with PMA:ION for 3, 7, and 30 h (Fig. 2C). Immunoprecipitation with anti-c-Jun antibodies (sc-45X, Santa Cruz Biotechnology) generated a PCR product with specific primers for the FCGR2B promoter (Fig. 2C). We did not detect PCR products by amplification with FCGR2B-specific primers following chromatin immunoprecipitation with control antibody (anti-XBP sc-7160X, Santa Cruz Biotechnology). These results demonstrate the direct interaction of c-Jun with the FCGR2B promoter and suggest that c-Jun could act as transcriptional regulator of the FCGR2B promoter activity at this site.
The formation of complex c was specific for the MUT probe in both CL-01 and U937 cells (Fig. 1, A and B, lanes 3 and 6). The nuclear factor YY1, also known as NF-E1 and UCRBP, has the core consensus binding sequence CCATNTT (14,15). The Ϫ343 G 3 C mutation created a possible YY1 binding site (CCATCAC) to the FCGR2B promoter sequence (Fig. 1C). The MUT-specific complex c was eliminated by an excess of C YY1 competitor in both U937 cells (data not shown) and CL-01 cells (Fig. 2A, lane 16) without affecting other bands. Complex c was not eliminated by addition of the 200-fold excess of C SP1 competitor, used as control (Fig. 2A, lane 17). Supershift formation with YY1 antibodies, but not with SP1 antibodies, established the identity of complex c formed with the MUT probe as being YY1 (Fig. 2B, lanes 5 and 6). Western blot experiments did not show changes in the level of YY1 expression in resting or activated cells (Fig. 2B, right panel). These results suggested that the Ϫ343 G to C nucleotide substitution in the FCGR2B promoter created a new binding site for YY1 that may alter the transcriptional activity of the MUT promoter.
Complex a1 and a2 had the highest intensity with the AP-1 probe (Figs. 1, A and B, lane 7, and 2A, lane 3). The WT probe formed more intense a1 and a2 complexes compared with the MUT probe (Figs. 1, A and B, lanes 5 and 6, and 2A, lanes 1 and  2). These results pointed out a lower ability of the MUT sequence to bind AP-1 compared with the WT sequence.
To confirm that the MUT probe interacted less efficiently with the AP-1 complex, we performed EMSA experiments with nuclear extract from activated CL-01 cells in which 32 P-labeled AP-1-specific probes were preincubated with the competitor oligonucleotides C WT and C MUT (Fig. 3A). The concentration of unlabeled C WT or C MUT ranged from 62.5-to 1000-fold excess. Analysis of the intensity of the AP-1 band indicated that 4 times more C MUT was necessary to attain 50% inhibition of AP-1 complex formation compared with using C WT (Fig. 3A, C WT lanes 2-6, and C MUT , lanes 7-11). The graphic representation of the inhibition of optical density of the AP-1 band by increasing concentrations of C WT and C MUT is depicted in Fig. 3B.
We tested the ability of AP-1 and YY1 nuclear factors to bind to the WT, MUT, and AP-1 probes in promoter pulldown assays. Biotinylated probes coupled to streptavidin-agarose beads were incubated with nuclear extract from activated CL-01 cells. Bound nuclear factors were eluted and analyzed by Western blotting with antibodies against YY1, c-Fos, and phosphorylated c-Jun-P73 (Fig. 3C). YY1 was eluted only from the MUT probe, and was undetectable in pull-down assays using AP-1 and WT probes (Fig. 3C, upper panel). c-Fos was eluted from the AP-1 and WT probes, and was not detected in the eluted fraction obtained from the MUT probe (Fig. 3C, middle  panel). A higher amount of phosphorylated c-Jun-P73 protein was eluted from the WT probe compared with the MUT probe (Fig. 3C, lower panel). The highest amount of phosphorylated c-Jun-P73 was eluted from the AP-1 probe.
We tested whether the Ϫ343 G/C polymorphic region participates in transcriptional regulation of the FCGR2B promoter in luciferase reporter assays. WT and MUT promoter constructs (Ϫ537 to ϩ43) were inserted into the pGL3-Enhancer vector upstream of the luciferase gene, and transfected into CL-01 cells. The Ϫ343 C promoter construct (MUT) had reduced promoter activity (0.6 Ϯ 0.1, p Ͻ 0.01, n ϭ 8) compared with the Ϫ343 G promoter construct (WT) in CL-01 cells (Fig. 4A).
As c-Jun was the major component of the AP-1 complex in activated CL-01 cells, we tested whether c-Jun participates in the transcriptional regulation of the WT and MUT promoter constructs. The inhibitor of c-Jun N-terminal kinase JNKIII is a fusion-penetrating peptide containing a fragment of the human c-Jun. JNKIII disrupts the formation of c-Jun-JNK complexes, and prevents the phosphorylation and activation of c-Jun (16). Treatment of non-activated CL-01 cells with 10 M JNKIII slightly lowered the luciferase activity driven by the WT reporter construct (Fig. 4A). The relative luciferase activity of the MUT promoter construct was higher after treatment of non-activated CL-01 cells with JNKIII compared with mediumtreated cells (Fig. 4A). In activated CL-01 cells, treatment with JNKIII decreased the amount of phosphorylated c-Jun in Western blots (Fig. 4B). Activation of CL-01 cells with PMA:ION for 48 h markedly increased the luciferase activity driven by the WT (3.1 Ϯ 0.4, p Ͻ 0.01, n ϭ 8) and MUT (2.1 Ϯ 0.2, p Ͻ 0.01, n ϭ 8) promoter constructs (Fig. 4C). The difference in relative luciferase activity between WT and MUT remained evident after activation. Interestingly, treatment of PMA:ION-activated CL-01 cells with JNKIII reduced the luciferase activity of the WT (0.9 Ϯ 0.1, p Ͻ 0.01, n ϭ 8) and MUT (0.7 Ϯ 0.2, p Ͻ 0.01, n ϭ 8) reporter constructs to the basal level (Fig. 4C). These results suggested that phosphorylation and activation of c-Jun are critical for regulation of the FCGR2B promoter activity.
The cell permeable compound Bt 2 cAMP is a synthetic analog of cAMP. Bt 2 cAMP triggers intracellular signaling leading to the phosphorylation and activation of c-Jun by a mechanism involving activation of protein kinase A. We evaluated the role of Bt 2 cAMP in the transcriptional regulation of the FCGR2B promoter activity (Fig.  4D). Treatment of CL-01 cells with 1 mM Bt 2 cAMP for 48 h increased the luciferase activity mediated by the WT and MUT promoter constructs (1.8 Ϯ 0.2, p Ͻ 0.01, n ϭ 8 and 1.3 Ϯ 0.2, p Ͻ 0.01, n ϭ 8) The difference in relative luciferase activity between the two promoter constructs was evident before and after activation with Bt 2 cAMP (Fig.  4D). In Bt 2 cAMP-activated CL-01 cells, treatment with JNKIII reduced the luciferase activity of the WT and MUT reporter constructs to the basal level (0.8 Ϯ 0.1, p Ͻ 0.01, n ϭ 8 and 0.5 Ϯ 0.1, p Ͻ 0.01, n ϭ 8) (Fig. 4D), similar to the effect observed with PMA:ION-activated cells.
We tested whether overexpression and activation of c-Jun was associated with changes in the promoter activity of the WT and MUT constructs. CL-01 cells transfected with pFA2-cJUN plasmid expressed increased levels of c-Jun assessed by Western blotting (Fig. 4E). The activation of MEKK leads to phosphorylation of JNK and subsequent phosphorylation and activation of c-Jun. Co-transfection of CL-01 cells with pFA2-cJUN plus pFC-MEKK resulted in up-regulation of the luciferase activity mediated by the WT promoter (1.2 Ϯ 0.2, p Ͻ 0.01, n ϭ 9) (Fig. 4F). Treatment with JNKIII of CL-01 cells co-transfected with pFA2-cJUN plus pFC-MEKK reduced the luciferase activity of the WT reporter construct (0.8 Ϯ 0.1, p Ͻ 0.01, n ϭ 9) (Fig. 4F). Co-transfection of the MUT construct with pFA2-cJUN plus pFC-MEKK was not associated with significant changes in luciferase activity (Fig. 4F).
Two sites in the proximal FCGR2B promoter sequence (Ϫ281 to Ϫ276 and Ϫ181 to Ϫ178) contain the core consensus binding sequence for YY1, thus creating putative binding sites for YY1 in both WT and MUT promoter constructs. We sought to investigate whether the interaction of YY1 and AP-1 with the Ϫ343 FCGR2B promoter sequence was associated with differences in the transcriptional activity mediated by the WT and MUT promoter constructs. We tested whether the co-transfection of YY1 with MEKK and c-Jun induced changes in the promoter activity of the WT and MUT constructs (Fig. 5A). The luciferase activity mediated by the WT promoter was downregulated in CL-01 cells co-transfected with pFA2-cJUN plus pFC-MEKK plus pCMV-YY1 compared with cells transfected with pFA2-cJUN plus pFC-MEKK (0.9 Ϯ 0.2 versus 1.2 Ϯ 0.2, p Ͻ 0.001, n ϭ 9) (Fig. 5A). Co-transfection of CL-01 cells with pFA2-cJUN plus pFC-MEKK plus pCMV-YY1 did not induce major changes in the transcriptional activity of the MUT promoter construct (Fig. 5A).
We investigated whether the inhibition of YY-1 expression was associated with changes in the promoter activity of the WT and MUT constructs. Expression of YY1 was down-regulated with small hairpin RNA against YY1 expressing vector (shYY1). Partial inhibition of YY1 expression by shYY1 was observed in lysates of CL-01 cells by immunoblotting with anti-YY1 antibodies (Fig. 5B). CL-01 cells were transfected with shYY1 or with control shRNA (shContr), followed by transfection with WT and MUT reporter constructs. The WT promoter activity was not different in CL-01 cells transfected with plasmid expressing shYY1 compared with that of shContr, indicating that inhibition of YY1 did not affect the transcriptional activity of the WT promoter (Fig. 5C). The luciferase activity driven by the MUT promoter construct was significantly higher in CL-01 cells transfected with shYY1 compared with shCont (09 Ϯ 0.3 versus 0.6 Ϯ 0.2, p Ͻ 0.01, n ϭ 9) (Fig. 5C). These results suggested that down-regulation of YY1 expression could restore, at least in part, the transcriptional activity of the MUT promoter construct. We believe that this effect is most likely the result of reduced binding of YY1 to the Ϫ343C promoter sequence in cells transfected with shYY1 plasmid.

DISCUSSION
Fc␥RIIb is widely expressed in cells of the immune system, such as B lymphocytes, monocytes, dendritic cells, neutrophils, mast cells, as well as in non-hematopoetic cells. Endowed with the ability to mediate inhibitory signaling upon binding IgGcontaining immune complexes, Fc␥RIIb is an important regulator of antibody-mediated reactions. Generally, Fc␥RIIb deficiency is associated with increased susceptibility and severity to organ-specific and systemic autoimmunity (17). Relatively small reductions in expression of Fc␥RIIb receptors contribute to the breakdown of self-tolerance (4,18). Mice heterozygous for deletions in Fc␥RIIb exhibit only modest reductions in pro-tein expression, but have a predisposition to autoimmunity (19). Recently, the expression of Fc␥RIIb was found to be upregulated on memory B cells of healthy controls, whereas Fc␥RIIb was considerably decreased in memory B cells from SLE patients, reflecting inadequate inhibitory signaling in SLE memory B cells (20). These observations call for a detailed investigation of mechanisms involved in regulation of the FCGR2B gene.
Acquired and genetic factors regulate the expression of Fc␥RIIb and influence its inhibitory potential. Data from our group demonstrated differential modulation of Fc␥RIIb expression in human monocytes and neutrophils by cytokines suggesting that alterations in cytokine levels common during inflammatory and autoimmune reactions could cause changes in Fc␥RIIb expression (21)(22)(23). A role for hormones in the reg-FIGURE 5. Regulation of luciferase activity of WT and MUT FCGR2B promoter constructs by YY1. A, relative luciferase activity of the WT and MUT constructs following co-transfection of CL-01 cells with pFA2-cJUN plus pFC-MEKK, pCMV-YY1, or pFA2-cJUN plus pFC-MEKK plus pCMV-YY1 (**, p Ͻ 0.001, n ϭ 9). B and C, effect of inhibition of YY1 expression with shYY1 on the luciferase activity mediated by the WT and MUT promoter constructs. B, CL-01 cells were transfected with shYY1 or with shContr and YY1 expression was analyzed in Western blotting using CL-01 cell lysates (15 and 7.5 g) and anti-YY1 antibodies. C, relative luciferase activity of CL-01 cells transfected with shYY1 or with shContr followed by co-transfection with WT or MUT constructs. Regulation of the MUT promoter activity in CL-01 cells transfected with shYY1 and shContr (*, p Ͻ 0.01, n ϭ 9). ulation of FCGR2B gene expression was revealed when dihydrotestosterone was found to inhibit Fc␥RIIb expression, suggesting a role for sex hormones in the onset and progression of autoimmune diseases (24).
More recently, inherited defects in Fc␥RIIb expression have been linked to the development of autoimmune diseases. The I232T substitution in the transmembrane domain of Fc␥RIIb is associated with autoimmunity in the Asian population (6 -8, 25). The SLE-associated Thr 232 allele does not alter membrane expression of Fc␥RIIb but is lacking functional inhibitory activity (9). A haplotype containing the Ϫ343 G 3 C SNP in the FCGR2B gene was found to associate with SLE in Caucasians (10). The association of the FCGR2B Ϫ343 C/C allele with SLE correlated with altered nuclear protein binding and with decreased transcriptional activity of the mutant FCGR2B promoter (11).
Primer extension experiments identified two transcription initiation sites in the FCGR2B promoter, and the minimal activity was contained within the first 154 nucleotides of the 5Ј-untranslated region (26). Several potential regulatory elements were identified upstream of the transcription initiation sites in the human FCGR2B promoter sequence (26,27). Two members of the Kruppel family of transcription factors, zinc finger proteins ZNF140 and ZNF91, expressed in T cells possibly bind to the human Fc␥RIIB promoter and could function as transcriptional repressors (26).
In the present study, we detected binding of AP-1 family members to the FCGR2B promoter sequence in activated CL-01 and U937 cells, despite the sequence variation in the AP-1 core consensus motif. The interaction of AP-1 transcription factors with the TGCGTCA sequence contained in the HLA-DR promoter in B-cell lymphoma cells provides another example that AP-1 is capable to interact with non-canonical consensus DNA sequences (28). c-Jun was one of the main components of the AP-1 complex that interacted with the WT promoter in CL-01 cells. In addition, JunB and c-Fos participated in the formation of the AP-1 complex with the WT probe, most likely by forming heterodimers with c-Jun.
Our results suggest a critical role for c-Jun in the transcriptional regulation of the FCGR2B promoter activity. Increased cell activation induced by PMA and Bt 2 cAMP was shown to up-regulate Fc␥RIIb expression and to increase the luciferase activity driven by FCGR2B promoter constructs in U937 cells (10,29,30). We find that the transcriptional activity of the WT and MUT promoter constructs is potently and consistently induced by PMA:ION and Bt 2 cAMP. These activators trigger independent signaling pathways, ultimately leading to the phosphorylation and activation of c-Jun. JNKIII, a specific inhibitor of c-Jun activation, decreased the level of phosphorylated c-Jun and down-regulated the FCGR2B promoter activity. By inducing the synthesis and phosphorylation of c-Jun, JNK may regulate the transactivation of the FCGR2B promoter. Increased activation of JNK in B lymphocytes from SLE patients has been reported (31). Patients with rheumatoid arthritis have constitutive activation of JNK in synovial tissue (32,33). The relationship between the increased activation of JNK during chronic inflammatory and autoimmune responses and the regulation of the FCGR2B gene expression remains to be established.
The homozygous Ϫ343 C/C genotype in the FCGR2B promoter is more frequent in SLE patients (10,11). Analysis of luciferase activity in transiently transfected cells indicated reduced activity of the MUT promoter in both resting and activated cells. Reduced binding of AP-1 to the sequence containing Ϫ343 C correlated with down-regulation of the MUT promoter activity. In another report, a mutant FCGR2B promoter haplotype containing Ϫ343 C/C was associated with increased luciferase activity, presumably mediated by the interaction of YY1 and GATA-4 with the mutant promoter haplotype (34). In agreement with this report, our data show binding of YY1 to the MUT promoter sequence. YY1 has been reported to have a dual effect, acting as a repressor or activator of gene transcription depending on the promoter sequence surrounding the YY1 binding site (14,35,36). In several genes, competition for binding sequences with an activator was identified as the main mechanism for the suppressive activity of YY1 (36). The inhibition of YY1 synthesis with specific shRNA resulted in up-regulation of the transcriptional activity of the MUT reporter construct. Our results suggest that YY1 competes with AP-1 for DNA binding at the Ϫ343 site, a mechanism that could contribute to the transcriptional repression of the mutant FCGR2B promoter activity.
In summary, we propose a mechanism for a non-coding region polymorphism affecting FCGR2B gene activity in SLE. A schematic model showing the specific binding of YY1 to the sequence containing Ϫ343 C and the impaired interaction of the AP-1 transcriptional complex with the MUT promoter leading to down-regulation of the FCGR2B promoter activity is presented in Fig. 6. Our findings may lead to the development of therapeutic approaches for induction of FCGR2B gene expression in autoimmune diseases.