Repressor Critically Dependent on a Cell Type-Specific CR2/CD21

CR2/CD21 Proximal Promoter Activity Is Critically Dependent on a Cell Type-Specific Repressor

This information is current as Daniela Ulgiati and V. Michael Holers of September 25, 2021. J Immunol 2001; 167:6912-6919; ; doi: 10.4049/jimmunol.167.12.6912 http://www.jimmunol.org/content/167/12/6912 Downloaded from

References This article cites 39 articles, 24 of which you can access for free at: http://www.jimmunol.org/content/167/12/6912.full#ref-list-1

Why The JI? Submit online. http://www.jimmunol.org/

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 25, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2001 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. CR2/CD21 Proximal Promoter Activity Is Critically Dependent on a Cell Type-Speciﬁc Repressor1

Daniela Ulgiati† and V. Michael Holers2*†

Transcription of the human complement receptor type 2 (CR2/CD21) gene is controlled by both proximal promoter and intronic elements. CR2 is primarily expressed on B cells from the immature through mature cell stages. We have previously described the presence of an intronic element that is required for both cell- and stage-specific expression of CR2. In this study, we report the identification of a cell type-specific repressor element within the proximal promoter. This repressor sequence is shown by linker scanning mutagenesis to comprise an E box motif. By supershift analysis this element binds members of the basic helix-loop-helix family of proteins, in particular E2A gene products. Mutational analysis demonstrates that binding of E2A proteins is critical for functioning of this repressor. Thus, E2A activity is key not only for early B cell development, but also for controlling CR2 expression, a gene expressed only during later stages of ontogeny. The Journal of Immunology, 2001, 167: 6912Ð6919. Downloaded from

uman complement receptor type 2 (CR2;3 CD21) is a factor 1 (CBF1), a member of the developmentally important 145-kDa protein encoded within the regulators of the Notch signaling pathway. Mutation of this site results in loss of H complement activation gene cluster localized on human function of the silencer and strongly suggests that CBF1 plays a chromosome 1q32 (1). CR2 is the receptor for complement acti- role in controlling human CR2 expression (19). Furthermore, it

vation fragments of C3, specifically, iC3b, C3dg, and C3d (2, 3). was shown that the silencer was unable to repress a heterologous http://www.jimmunol.org/ Additionally, CR2 is the receptor for the EBV and mediates EBV promoter, suggesting specificity for proximal promoter sites. Sim- infection by binding the membrane protein, gp350/220 (4, 5). Hu- ilarly, in the mouse, CR2 expression is regulated by an intronic man CR2 is also the B cell receptor for CD23 (6) and possibly silencer (20) that also requires CR2 proximal promoter sites for IFN-␣ (7). appropriate function (21). Human CR2 is primarily expressed during later stages of B cell In the studies reported herein, we have further analyzed the hu- ontogeny (8); however, it is also expressed on follicular dendritic man CR2 proximal promoter to identify cell type-specific elements cells (9), epithelial cells (10), some thymocytes (11), and a small that could act as putative interaction sites for the CR2 intronic subset of CD4ϩ and CD8ϩ peripheral T cells (12, 13). Within the silencer. We demonstrate the presence of a cell type-specific re-

B cell lineage, CR2 is only found on immature and mature B cells, pressor that shows broad lineage- and stage-specific utilization. by guest on September 25, 2021 and its expression begins at approximately the same stage as IgD and CD23 (8, 14). It has been shown that CR2 is up-regulated after Materials and Methods B cells escape negative selection and migrate to the periphery Cell lines and culture conditions (15Ð17). All human cell lines used in these experiments were obtained from Amer- Previously, we have shown that cell- and stage-specific expres- ican Type Culture Collection (Manassas, VA). Cells lines were maintained sion of human CR2 is controlled by an intronic transcriptional at 37¡C with 5% CO2 in RPMI 1640 with L-glutamine supplemented with silencer, designated the CRS (CR2 silencer). The use of a stable 10% FBS, 100 ␮g/ml streptomycin, and 100 IU/ml penicillin. transfection system and transgenic mice has shown that the CRS Creation and confirmation of mutant CR2 promoter/luciferase element, in conjunction with the CR2 proximal promoter, is able to fusion constructs repress transcription in CR2-negative cell lines and tissues (18). An NheI/XhoI fragment of the CR2 promoter containing nt Ϫ315/ϩ75 was Recent studies have further defined the CRS element and have cloned into the luciferase reporter pGL3-basic vector (CLONTECH Lab- shown a sequence within the silencer crucial to its function. This oratories, Palo Alto, CA). Site-directed mutagenesis was performed using sequence binds the transcriptional repressor C-promoter binding the Quickchange mutagenesis kit (Stratagene, La Jolla, CA), which enabled the incorporation of MluI restriction sites extending 3Ј from positions Ϫ140 (Sp1), Ϫ90 (AP1), Ϫ81 (AP2), Ϫ60 (E box 2) and Ϫ47 (E box 1). *Departments of Immunology and Medicine, and †Division of Rheumatology, Uni- The accuracy of all constructs created was assured by both restriction en- versity of Colorado Health Sciences Center, Denver, CO 80262 zyme digestion and nucleotide sequence analysis. Received for publication August 15, 2001. Accepted for publication October Creation and analysis of mutant CR2 linker scanning constructs 16, 2001. The costs of publication of this article were defrayed in part by the payment of page Linker scanning mutagenesis was performed using the Quickchange mu- charges. This article must therefore be hereby marked advertisement in accordance tagenesis kit (Stratagene). Incorporation of MluI restriction sites was made with 18 U.S.C. Section 1734 solely to indicate this fact. across the E box 2 motif at intervals of ϳ2 bp. An internal deletion con- Ϫ Ϫ 1 This work was supported by the Smyth Professorship (to V.M.H.), National Insti- struct was also made by use of a primer that lacked bp 67 to 61 of the tutes of Health Grant RO1AI31105 (to V.M.H.), and an Arthritis Foundation post- CR2 promoter. All constructs made were confirmed by nucleotide se- doctoral fellowship grant (to D.U.). quence analysis. 2 Address correspondence and reprint requests to Dr. V. Michael Holers, Division of Transfection and measurement of promoter/reporter gene activity Rheumatology, Campus Box B-115, University of Colorado Health Sciences Center, 4200 East Ninth Avenue, Denver, CO 80262. E-mail address: [email protected] Before each transfection, cells were split and grown in log phase to ϳ5 ϫ 5 3 Abbreviations used in this paper: CR2, complement receptor type 2; HLH; helix- 10 cells/ml. Cells were then transfected using the Qiagen Superfect trans- loop-helix; bHLH, basic HLH; CBF1, C-promoter binding factor 1; CRS, CR2 si- fection reagent according to the manufacturer’s specifications with plasmid lencer; LS, linker scanning. DNA prepared using Qiagen Maxiprep-500 columns (Qiagen, Valencia,

CA). Briefly, 10 ␮g of plasmid DNA and 300 ng of pRL-thymidine kinase normalized transcriptional activity (vs Renilla internal control) control vector were complexed in combination with Superfect reagent for was 5.272 Ϯ 1.374 (n ϭ 10) for the Daudi cell line and 5.065 Ϯ 10 min at room temperature. The transfection complexes were then added 0.271 for the Raji (n ϭ 5) cell line. This is in contrast to two non-B dropwise to the cells that had been plated in 5 ml of medium in a six-well tray at a concentration of 5 ϫ 105Ð1 ϫ 106 cells/ml. The cells were then cell lines (that do not express CR2), K562 and U937, which dem- incubated at 37¡C for 48 h following transfection. Cell lysates from the onstrated much lower promoter activity (0.627 Ϯ 0.157 (n ϭ 11) transfected cells were prepared and assayed for both firefly and Renilla and 0.281 Ϯ 0.065 (n ϭ 4), respectively; Fig. 1). luciferase according to the manufacturer’s instructions (Promega, Madison, To determine whether the low level of transcriptional activity WI). All transfection data shown are the mean of 3Ð10 independent trans- fections, with n values shown in each experiment. Additionally, multiple was limited to non-B cells that do not express CR2 or whether it preparations of DNA were used and yielded essentially identical results. was also present in cells of the B cell lineage that do not express Promoter activity is expressed as relative firefly luciferase activity normal- CR2, the Ϫ315/ϩ75 construct was transiently transfected into a ized against Renilla luciferase activity. pre-B cell line (Fig. 1, Reh). These results also demonstrated a low Ϯ EMSA relative level of transcription in this cell line (0.563 0.115). Overall, these results show that even though the CR2-proximal 7 Approximately 8 ϫ 10 cells were used to make nuclear extracts according promoter is active in all cell lines tested compared with the pGL3 to a standard method (22). Extracts were frozen in liquid nitrogen and Ϯ ϭ stored at Ϫ80¡C. Determination of protein concentration was performed vector alone (0.0538 0.005; n 15), the relative CR2 promoter using the Bio-Rad protein assay kit (Hercules, CA). For EMSA, nuclear activity strongly correlates with CR2 expression status. Cell lines extracts were preincubated (10Ð20 ␮g) on ice for 10 min along with 1 ␮g that express CR2 possess a markedly higher basal promoter activ- of poly(dI-dC) in a binding buffer consisting of 4% Ficoll, 20 mM HEPES ity compared with cell lines that do not express CR2.

(pH 7.9), 1 mM EDTA, 1 mM DTT, and 50 mM KCl. When required, Downloaded from competitor oligonucleotides or supershift Abs (Santa Cruz Biotechnology, Functional analysis of the CR2 proximal promoter using site- Santa Cruz, CA) were incubated with the nuclear extract for 30 min on ice. The nuclear extract was then incubated with 80 fmol of 32P-labeled oligo- directed mutagenesis nucleotide for 30 min on ice before loading onto a 6% polyacrylamide gel. Because the Ϫ315/ϩ75 luciferase construct possessed a cell type- ϫ The gel was electrophoresed at 150 V using 0.25 Tris-taurin-EDTA as specific level of activity, mutant constructs were made and tran- the running buffer. EMSA gels were dried under vacuum and exposed to x-ray film. All double-stranded oligonucleotides were end-labeled using siently transfected into K562 (CR2-negative) or Raji (CR2-posi- [32P]ATP and T4 polynucleotide kinase. tive) cell lines. The mutants were constructed by introduction of a http://www.jimmunol.org/ MluI site into previously identified transcription factor binding Results sites (Fig. 2). The results were analyzed to determine the presence Relative CR2 promoter activity strongly correlates with CR2 of either a cell type-specific enhancer element that activated tran- expression scription in the CR2-expressing cell lines or, alternatively, a cell To determine whether the CR2 proximal promoter demonstrates a type-specific repressor element that dampened transcriptional ac- cell type-specific component to its activity, the Ϫ315/ϩ75 lucif- tivity in CR2-negative lines. erase construct was transiently transfected into informative cell Site-specific mutagenesis of both the consensus Sp1 site at Ϫ140 (Fig. 3, construct 2) and the consensus AP1 site at Ϫ90 (Fig. lines that either did or did not express CR2. The Ϫ315/ϩ75 con- by guest on September 25, 2021 struct was used in these studies, as previous experiments revealed 3, construct 3) resulted in no significant change in transcriptional Ϫ ϩ that the transcriptional activity of this construct was comparable to activity compared with the 315/ 75 wild-type construct. This that of a longer construct containing Ϫ1250 to ϩ75 of upstream promoter sequence (data not shown). In mature B cell lines (Fig. 1; Daudi and Raji), which express CR2 at a high level, the mean

FIGURE 2. Mutant constructs used in functional assays. Indicated are several putative transcription factor binding sites as demonstrated by DNase I footprinting and sequence analysis of transcription factor data- FIGURE 1. Transient transfection of the Ϫ315/ϩ75 wild-type construct bases. Also shown are the TATA box and the transcriptional initiation site into CR2-positive and -negative cell lines shows strong correlation of ac- (curved arrow). Site-speciﬁc mutations were made within each motif by tivity with CR2 expression status. Results shown are the mean normalized changing several base pairs to a MluI restriction site in the context of an transcriptional activity Ϯ SEM vs that in the internal Renilla control. otherwise intact proximal promoter. 6914 LINEAGE-SPECIFIC REPRESSION IN THE HUMAN CR2 GENE Downloaded from http://www.jimmunol.org/

FIGURE 3. Normalized transcriptional activity of the CR2 mutant constructs are shown. The site-directed mutant constructs were transiently transfected into CR2-negative K562 cells and CR2-positive Raji cells. The results by guest on September 25, 2021 represent promoter activity and are expressed as normalized transcriptional activity vs the unmodified Ϫ315/ϩ75 construct. result was seen in both cell lines tested. Site-specific mutation of true broadly used repressor element, the E box 2 mutant construct was the consensus AP2 site at position Ϫ81 (Fig. 3, construct 4) transiently transfected into other informative cell lines. When the mu- resulted in a modest 1.7-fold increase in transcriptional activity tant construct was transfected into two mature B cell lines (CR2- in K562 cells and a 1.6-fold increase in activity in Raji cells. positive), Raji and Daudi, only a modest increase in transcriptional These results indicate the possible presence of a repressor activity was seen (1.561 Ϯ 0.139 (n ϭ 5) and 1.128 Ϯ 0.149 (n ϭ 5), element at this site; however, this element does not appear to be respectively; Fig. 4). This was in contrast to the two non-B cell lines cell type specific. Mutation of the E box 1 motif located at tested, neither of which expresses CR2. In the K562 cells, the muta- position Ϫ47 resulted in an 80% reduction of transcriptional tion resulted in a 5.699 Ϯ 0.0745-fold (n ϭ 7) increase in transcrip- activity in Raji cells. This result was also seen in K562 cells, as tional activity compared with the wild-type construct. Similarly, in this mutation resulted in a marked (55%) reduction in promoter U937 cells, a 5.758 Ϯ 0.396-fold (n ϭ 7) increase in promoter activity activity (Fig. 3, construct 6). These results demonstrate the was seen (Fig. 4). To determine whether the repressor element was presence of an activating motif at this site; however, again, its also active in cells of the B cell lineage that did not express CR2, the activity is not cell type specific. E box 2 mutant was transiently transfected into Reh cells, a pre-B cell In marked contrast, when the adjacent E box (Fig. 3, construct line. A similar result was observed in this cell line as in the non-B 5) at position Ϫ60 was mutated, cell type-specific activity was cells. The mutant construct resulted in a 4.636 Ϯ 0.938-fold increase identified. Mutation of this site and transient transfection into Raji in transcriptional activity compared with the Ϫ315/ϩ75 wild-type cells resulted in a modest 1.5-fold increase in promoter activity construct. These results indicate the presence of a cell type-specific (Fig. 3, construct 5). However, in K562 cells the same mutant repressor that is primarily active in cell lines that do not express CR2. demonstrated an almost 6-fold increase in transcriptional activity compared with the Ϫ315/ϩ75 wild-type construct. These results Characterization of factors binding the E box 2 motif reveal the presence of a cell type-specific repressor located at po- The reporter gene analysis of the E box 2 motif demonstrated the sition Ϫ60 of the CR2-proximal promoter. importance of this site as a cell type-specific repressor. To determine the nature of the transcription factors binding to this site, Broad lineage utilization of the cell type-specific repressor EMSA was performed using double-stranded oligonucleotides cor- To determine whether this cell type-specific repression was merely responding to Ϫ73 to Ϫ52 region of the proximal promoter en- limited to a unique effect in either Raji or K562 cells or was due to a compassing the E box 2 motif (Fig. 5, E box 2). The EMSA pattern The Journal of Immunology 6915

Mutant). The mutant oligonucleotide corresponded to the same nucleotides as the wild type (Ϫ73 to Ϫ52); however, positions Ϫ66 to Ϫ61 were changed to nucleotides, ACGCGT, corresponding to the MluI restriction sequence used in the transcriptional assays in Figs. 3 and 4. The wild-type and mutant oligonucleotides were labeled and bound to K562 nuclear extract, and the EMSA proﬁles were compared (Fig. 5B). The mutant oligonucleotide was no longer able to bind protein-DNA complexes A and C and the higher order complexes compared with the wild-type oligonucleotide. These results indicate that the most relevant proteins required for the functional activity of this repressor are in these latter complexes.

Linker-scanning mutagenesis reveals critical nucleotides for repressor function It was apparent from the functional studies that the CR2 promoter possesses a repressor activity and that loss of this activity corre-

lates with loss of specific protein-DNA complexes in EMSA. To Downloaded from FIGURE 4. Transient transfection of the E box 2 mutant construct into determine which specific nucleotides were critical for repressor a number of CR2-positive and -negative cell lines indicate the existence of a cell type-specific repressor at this site. The results are shown as fold function, linker-scanning mutagenesis was performed. Each linker- increase over Ϫ315/ϩ75 wild-type (WT) values (mean Ϯ SEM). scanning mutant was made by introduction of an MluI restriction site across the sequence separated by two nucleotides (Fig. 6A). Additionally, to eliminate the possibility that introduction of the observed when the oligonucleotide was labeled and bound to K562 MluI site artificially introduced a new transcription factor binding http://www.jimmunol.org/ nuclear extracts was complex, with the presence of three protein- site potentially containing activation functions, an internal deletion DNA complexes (Fig. 5A, complexes AÐC) as well as complexes construct was made. This construct had nucleotides Ϫ67 to Ϫ61 de- with slower mobility (Fig. 5A, higher order complex). All protein- leted from an otherwise intact Ϫ315/ϩ75 sequence (Fig. 6A). DNA complexes were shown to be specific, as addition of increas- All mutant constructs along with the internal deletion construct ing amounts of cold self-competitor resulted in abolishment of all were transiently transfected into Raji (CR2-positive) and K562 complexes (Fig. 5A). (CR2-negative) cells, normalized against the 315WT, and com- To determine which protein complexes were relevant for the pared for activity similar to that seen with the original E box 2 cell type-specific repressor effect, an oligonucleotide corre- mutant. Results revealed that all linker-scanning mutants and the sponding to the functional mutant was constructed (Fig. 5, internal deletion were inactive in Raji cells (Fig. 6B; n ϭ 4Ð10), by guest on September 25, 2021

FIGURE 5. EMSA of the putative E box 2 motif using K562 nuclear extract. A, Competition analysis using increasing fold molar excess of unlabeled self-competitor oligonucleotide demonstrates the presence of three speciﬁc protein-DNA complexes (AÐC), along with two higher order complexes. B, An oligonucleotide was designed corresponding to the functional mutant used in the transfection assays. This mutant oligonucleotide was unable to bind the higher order complexes and protein-DNA complexes A and C. 6916 LINEAGE-SPECIFIC REPRESSION IN THE HUMAN CR2 GENE

FIGURE 6. Linker-scanning mutagenesis across E box 2 reveals critical nucleotides for repressor activity. A, Wild-type sequence across the E box 2 consensus sequence is shown. The sequence of mutant constructs is also shown, including the position and sequence of the mutation. The sequence deleted from the internal deletion construct is represented by a gap. Dashed lines represent sequence identity. B, All mutant constructs and the internal deletion were transiently transfected into Raji (CR2-positive) and K562 (CR2-negative) cell lines and normalized against the Ϫ315/ϩ75 wild type (WT). Results are expressed as the mean fold increase over Downloaded from 315WT cells Ϯ SEM. http://www.jimmunol.org/

again indicating that this repressor is not active in CR2-expressing and the higher order complex to the wild-type probe. Similar re- cells. These results were in contrast to those seen in K562 cells. sults were seen using the internal deletion oligonucleotide. These Parallel transfection of linker scanning (LS) mutants (Ϫ62/Ϫ57), results indicate that deletion of positions Ϫ67 to Ϫ61 or mutation LS (Ϫ60/Ϫ55), and the internal deletion possessed similar fold of Ϫ65 to Ϫ61 results in the inability of these complexes to bind increase over the Ϫ315/ϩ75 wild-type construct as the original E (Fig. 7B, LS Int Del, E box 2 Mut). Competition using LS (Ϫ62/ by guest on September 25, 2021 box 2 mutant (Fig. 6B; n ϭ 4Ð10). Thus, the nucleotides that are Ϫ57) resulted in abolishment of the higher order complexes and critical for the function of this cell type repressor are situated from complex C, but not complex A, indicating that nucleotides Ϫ62 to nucleotides Ϫ67 to Ϫ59 of the proximal promoter. The results of Ϫ57 are critical for binding of complex A (Fig. 7B,LS(Ϫ62/ the analysis using the internal deletion construct showed that no Ϫ57)). Interestingly, the opposite was true for LS (Ϫ60/Ϫ55). unanticipated activation sequence was introduced. Competition of wild type with this probe resulted in loss of complexes A and C, but not the higher order complexes. These results Cross-competition EMSA reveals protein-DNA complexes demonstrate that nucleotides Ϫ60 to Ϫ55 are critical for formation important to repressor function of the higher order complexes (Fig. 7B,LS(Ϫ60/Ϫ55)). As all Linker-scanning mutagenesis revealed nucleotides critical for mutations used in the EMSA experiments corresponded to muta- functioning of the CR2 repressor element. EMSA oligonucleotides tions resulting in functional activity, these results suggest that pro- were made that comprised Ϫ73 to Ϫ52 of promoter sequence (Fig. tein-DNA complexes A and C and the higher order complexes are 7A, wild type). Mutant oligonucleotides were designed spanning all necessary for functioning of the repressor. Mutations of nucle- the same region and included MluI restriction sequences substi- otides that result in abolishment of protein-DNA complex A alone tuted at the same positions as those used in the functional assays. (Fig. 7B,LS(Ϫ62/Ϫ57)), the higher order complex alone (Fig. 7B, The MluI sequence (ACGCGT) was substituted in the E box 2 LS (Ϫ60/Ϫ55), or their combination (Fig. 7B, E box 2 mut, LS Int mutant at position Ϫ65 to Ϫ61, the LS (Ϫ62/Ϫ57) at position Ϫ62 Del) all result in loss of repressor function in the transfection assays, to Ϫ57, and the LS (Ϫ60/Ϫ55) at position Ϫ60 to Ϫ55; the LS as observed by an increase in transcriptional activity in K562 cells. Internal deletion (LS Int Del) had nt Ϫ67 to Ϫ61 deleted (Fig. 7A). Cross-competition EMSA was performed using labeled wild-type Supershift assays reveal binding of the basic helix-loop-helix oligonucleotide competed with a 250-fold molar excess of cold (bHLH) protein E2A to the repressor element self, cold E box 2 mutant, cold LS (Ϫ62/Ϫ57), cold LS (Ϫ60/ To further characterize the proteins required for repressor activity, Ϫ55), or cold LS Int Del, competitor. The proteins unaffected by supershift assays were performed. The EMSA oligonucleotide the specific mutations will bind the mutant oligonucleotides, leav- used in these experiments corresponded to the oligonucleotide ing behind only proteins still able to bind the wild-type probe on used in the linker-scanning mutagenesis experiments and com- the gel. This enables identification of each protein-DNA complex prised Ϫ73 to Ϫ52 of promoter sequence (Fig. 7A, wild type). critical for function. Supershift assays were performed using labeled wild-type oligo- Competition of the wild-type oligonucleotide with self resulted nucleotide (Fig. 8, wild type). Various Abs were then added. Ad- in abolishment of all protein-DNA complexes, indicating that all dition of an Ab directed against the bHLH protein E2A resulted in complexes are specific (Fig. 7B, Self). Competition using the E abolishment of protein-DNA complex A (Fig. 8, E2A). Cross- box 2 mutant probe resulted in binding of complex A, complex C, competition experiments using mutant EMSA oligonucleotides The Journal of Immunology 6917 Downloaded from http://www.jimmunol.org/

FIGURE 8. Supershift assay using Abs directed against various bHLH proteins. The oligonucleotide used in these experiments corresponded to Ϫ73 to Ϫ52 of the proximal promoter sequence spanning the E box site 2 motif. Competition using Abs (400 ng) directed against E2A bHLH proteins resulted in abolishment of functionally relevant protein-DNA com- by guest on September 25, 2021 plexes. Abs to other E box binding proteins as shown had no effect on protein-DNA binding.

Discussion Previous work has shown that the human CR2 gene is transcrip- tionally controlled by intronic elements (18). Further analysis in both humans and mice has shown that the intronic silencer is unable to regulate the activity of other heterologous promoters, in- FIGURE 7. EMSA using cold competition oligonucleotides corre- dicating that in both species the intronic silencer must act via in- sponding to functional linker-scanning mutants and internal deletion con- teraction with the CR2-proximal promoter (19, 21). Even though structs. A, Oligonucleotides used in the EMSA show the positions of mu- human and mouse CR2 genes appear to be controlled by a similar tated sequences compared with the wild type. Mutated sequences are in mechanism, the transcription factors involved are likely to be dif- italics and underlined. B, Cross-competition EMSA in K562 cells using labeled wild-type oligonucleotide competed with a 250-fold molar excess ferent due to the lack of significant sequence homology between of cold self, cold E box 2 Mut, cold LS (Ϫ62/Ϫ57), cold LS (Ϫ60/Ϫ55), human and mouse regulatory regions. or cold LS internal deletion oligonucleotides. We have extended the analysis of the transcriptional require- ments for human CR2 and have analyzed the proximal promoter region for candidate sites involved in cell type-specific regulation (Fig. 7) have indicated the importance of protein-DNA complex of CR2. Data obtained from transfection of a Ϫ315/ϩ75 proximal A to the function of the repressor element. Addition of an Ab promoter sequence upstream of a luciferase reporter demonstrated directed against E47 alone resulted in abolishment of one of the several informative results. This construct, although active in all higher order complexes, also shown by cross-competition ex- cell lines tested compared with empty vector control, clearly dem- periments to be involved in repressor function. Protein-DNA onstrated a cell type-specific component to its activity. These re- complex C, a functionally relevant complex, was not effected by sults could indicate the presence of an activator element within the either Ab, indicating that an as yet unidentified protein is also proximal promoter that elevated transcription in CR2-expressing involved in the binding of this repressor motif. Addition of Abs cells or, alternatively, a repressor element within this sequence directed against two other bHLH proteins, namely, USF1 and could be present that dampened transcription in CR2-nonexpress- TFE3 (Fig. 8), did not have an effect on complex formation. ing cells. The use of several site-directed mutants revealed the These results strongly suggest that the E2A proteins E12/E47 presence of a repressor element within CR2-negative cell lines that are functionally relevant proteins in the repressor activity as- inhibited CR2 transcription, demonstrating that the latter mecha- sociated with the E box 2 motif. nism is operating. 6918 LINEAGE-SPECIFIC REPRESSION IN THE HUMAN CR2 GENE

Data collected from linker-scanning mutagenesis revealed that tein-activated protein kinases interact with E47 and are able to the critical nucleotides for repressor function matched an E box phosphorylate this protein, resulting in repression of the transcrip- motif that is known to bind the HLH family of transcription factors tional activity of E47 on an E box-containing promoter (35). Ad- (23). The HLH family of proteins plays a major role in multiple ditionally, E47 phosphorylation inhibits binding of E47 ho- developmental processes. To date, Ͼ240 HLH proteins have been modimers, but allows E47 heterodimer formation, suggesting a identified in many different organisms (24). E box elements and differential regulation of E proteins in B cells and non-B cells HLH proteins have been identified in many promoter and enhancer (36). Class II HLH proteins have also been shown to act as trans- elements that regulate muscle (25), pancreas (26), neuron (27), and criptional repressors. For example, ABF-1 is able to inhibit B cell-specific gene expression (28, 29). Further characterization E47-dependent activation through formation of heterodimers. of the proteins binding the CR2 E box repressor element using Abs E2A-ABF-1 heterodimers may function to actively repress E box- directed against many members of the HLH family of transcription containing genes, or, alternatively, these heterodimers may be tran- factors was undertaken. Supershift analysis revealed competition scriptionally inactive (37). Furthermore, E proteins are able to of functional protein-DNA complexes by Abs directed against form homodimers as well as heterodimers with other members of E2A and E47, indicating that both E12 and E47 are able to bind the the bHLH family and thus differentially regulate transcription. It CR2 E box repressor element. The E12 and E47 proteins arise by has been shown that Id proteins preferentially dimerize with E alternative splicing of the E2A gene (30, 31) and are known as proteins and consequently prevent heterodimers from binding class I HLH proteins. DNA and activating target genes (38). These studies suggest that Due to the large number of HLH proteins within this family, while E2A proteins are, in general, transcriptional activators, several tissue distribution, dimerization capabilities, and DNA binding mechanisms have been discovered that could explain the presence of Downloaded from specificity were used to devise a classification code (32). Class E2A in a repressor complex regulating CR2 promoter activity. I HLH proteins are also known as E proteins and include the In conclusion, we have discovered a cell type-specific repressor following; E12, E47, HEB, E2-2, and Daughterless. All of these element within the human CR2-proximal promoter that displays proteins are widely expressed and are able to form either broad lineage utilization. Furthermore, this repressor appears to homodimers in B cells or heterodimers with tissue-specific class require HLH transcription factors, in particular, gene products of

II HLH proteins in other cell types (23). E proteins, in particular E2A to function. Whether adapter proteins such as the Id family of http://www.jimmunol.org/ E2A gene products, have been shown to be involved in cell proteins are involved and are contributing to the repression mech- differentiation, lineage commitment, and B lineage-speciﬁc anism is currently being examined. gene expression (32). Interestingly, E2A knockout mice lack Previous studies involving the human intronic silencer has pre-B and mature B cells and have reduced numbers of shown critical function for CBF1 (19), a known transcriptional B220ϩCD43ϩ B cell progenitors, indicating a critical role for repressor. CBF1 is a component of the Notch signaling pathway. E2A in B cell development (33, 34). Interestingly, activated Notch 1 and Notch 2 are also able to inhibit As human CR2 is tightly regulated during B cell development, E47 activity (39). Notch signaling has now been tied to two re- E2A may also play a role in the expression of CR2. However, E pressor elements functioning within the human CR2 regulatory proteins have generally been shown to play a role in activating B regions. Data collected from our laboratory have shown that the by guest on September 25, 2021 cell-speciﬁc genes (23). In the case of the CR2-proximal promoter, CR2 silencer must interact with the CR2-proximal promoter for E2A proteins are involved in repression. To date, very little is function. Therefore, it is intriguing to speculate whether the known about the role of E2A proteins in repressing transcriptional E2A-containing promoter repressor element and the CBF1 silenc- activity. However, recent data suggest that mitogen-activated pro- ing element interact with one another or act in concert with Notch

FIGURE 9. Model of the transcriptional regulation of human CR2. The model shows the putative mechanisms involved in either silencing (OFF) or expressing (ON) human CR2 along with possible external signals involved in regulating this process (straight arrows). Indicated are the proteins known to bind the promoter repressor element (E47 and E12) as well as the intronic silencer (CBF1). Also shown is the transcriptional initiation site (curved arrow), exons 1 and 2 (square boxes), and the putative intronic matrix attachment site (hatched line). The Journal of Immunology 6919 signaling. Alternatively, as CR2 expression is tightly regulated, 13. Levy, E., J. Ambrus, L. Kahl, H. Molina, K. Tung, and V. M. Holers. 1992. T several mechanisms may play a role in controlling CR2 gene reg- lymphocyte expression of complement receptor 2 (CR2/CD21): a role in adhesive cell-cell interactions and dysregulation in a patient with systemic lupus erythem- ulation in different cell lineages and stages. These questions are atosus (SLE). Clin. Exp. Immunol. 90:235. currently under investigation. 14. Takahashi, K., Y. Kozono, T. J. Waldschmidt, D. Berthiaume, R. J. Quigg, A. Baron, and V. M. Holers. 1997. Mouse complement receptors type 1 (CR1; Finally, we used the data presented herein together with previ- CD35) and type 2 (CR2; CD21): expression on normal B cell subpopulations and ous results to generate a model of how the human CR2 gene is decreased levels during the development of autoimmunity in MRL/lpr mice. regulated (Fig. 9). Early studies (18) have shown the lack of J. Immunol. 159:1557. 15. Melamed, D., and D. Nemazee. 1997. Self-antigen does not accelerate immature DNase I-hypersensitive sites within cell lines that do not express B cell apoptosis, but stimulates receptor editing as a consequence of development CR2, indicating the possibility of a closed chromatin configuration arrest. Proc. Natl. Acad. Sci USA 94:9267. over the CR2 control elements (Fig. 9, OFF). This is in contrast to 16. Melamed, D., R. J. Benschop, J. Cambier, and D. Nemazee. 1998. Developmental regulation of B lymphocyte immune tolerance compartmentalizes clonal selection CR2-expressing cell lines that possess two hypersensitive sites, from receptor selection. Cell 92:173. one across the proximal promoter and the other within the first 17. Hartley, S. B., M. P. Cooke, D. A. Fulcher, A. W. Harris, S. Cory, A. Basten, and intron, i.e., the silencer site (18). It is interesting to speculate that C. C. Goodnow. 1993. Elimination of self-reactive B lymphocytes proceeds in two stages: arrested development and cell death. Cell 72:325. external signals such as Notch signaling or histone deacetylases 18. Makar, K. W., C. T. N. Pham, M. H. Dehoff, S. M. O’Connor, S. M. Jacobi, and may be relevant in opening the chromatin configuration to allow V. M. Holers. 1998. An intronic silencer regulates B lymphocyte cell- and stage- CR2 to be expressed (Fig. 9, ON). Histone deacetylation appears specific expression of the human complement receptor type 2 (CR2, CD21) gene. J. Immunol. 160:1268. critical for mouse CR2 regulation (21), but to date has not been 19. Makar, K. W., D. Ulgiati, J. Hagman, and V. M. Holers. 2001. A site in the confirmed in the human gene (K. Makar and V. M. Holers, un- complement receptor 2 (CR2/CD21) silencer is necessary for lineage specific published observations). Two repressor elements have been iden- transcriptional regulation. Int. Immunol. 13:657. Downloaded from 20. Hu, H., B. K. Martin, J. J. Weis, and J. H. Weis. 1997. Expression of the murine tified within the human CR2 regulatory regions. The first is CBF1 CD21 gene is regulated by promoter and intronic sequences. J. Immunol. 158:4758. within the intronic silencer (19), and the second are the E2A gene 21. Zabel, M. D., B. L. Byrne, J. J. Weis, and J. H. Weis. 2000. Cell-specific expression of the murine CD21 gene depends on accessibility of promoter and products within the proximal promoter as shown herein. Within the intronic elements. J. Immunol. 165:4437. silenced locus (Fig. 9, OFF), CBF1 may interact with the E2A 22. Li, Y. C., J. Ross, J. A. Scheppler, and B. R. Franza, Jr. 1991. An in vitro transcription proteins to repress transcription. Alternatively, phosphorylation of analysis of early responses of the human immunodeficiency virus type 1 long ter- minal repeat to different transcriptional activators. Mol. Cell. Biol. 11:1883. E47 or another as yet unidentified negative regulator may interact 23. Massari, M. E., and C. Murre. 2000. Helix-loop-helix proteins: regulators of http://www.jimmunol.org/ with the promoter element to repress transcription. Within an ac- transcription in eucaryotic organisms. Mol. Cell. Biol. 20:429. tive locus (Fig. 9, ON), E47/E12 may no longer be repressed by 24. Atchley, W., and W. Fitch. 1997. A natural classification of the basic helix-loop- helix class of transcription factors. Proc. Natl. Acad. Sci. USA 94:5172. external factors. Additionally, Notch may mask the inhibitory ef- 25. Buskin, J. N., and S. D. Hauschka. 1989. Identification of a myocyte nuclear fect of CBF1 and, therefore, allow positive regulatory elements factor which binds to the muscle-specific enhancer of the mouse muscle creatine within the promoter to activate CR2 transcription. kinase gene. Mol. Cell. Biol. 9:2627. 26. Whelan, J., S. R. Cordle, E. Henderson, P. A. Weil, and R. Stein. 1990. Identification of a pancreatic ␤-cell insulin gene transcription factor that binds to and appears to activate cell-specific gene expression: its possible relationship to other cellular factors Acknowledgments that bind to a common insulin gene sequence. Mol. Cell. Biol. 10:1564.

We thank Dr Jim Hagman for his expert advice and helpful comments. 27. Bessis, A., A. Salmon, M. Zoli, N. Le Novere, M. Picciotto, and J. Changeux. by guest on September 25, 2021 1995. Promoter elements conferring neuron-specific expression of the ␤2-subunit of the neuronal nicotinic acetylcholine receptor studied in vitro and in transgenic References mice. Neuroscience 69:807. 28. O’Riordan, M., and R. Grosschedl. 1997. EBF and E47 collaborate to induce 1. Hourcade, D., V. M. Holers, and J. P. Atkinson. 1989. The regulators of com- expression of the endogenous immunoglobulin surrogate light chain genes. Im- plement activation (RCA) gene cluster. Adv. Immunol. 45:381. munity 7:25. 2. Iida, K., L. Nadler, and V. Nussenzweig. 1983. Identification of the membrane 29. Staudt, L., and M. Lenardo. 1991. Immunoglobulin gene transcription. Annu. receptor for the complement fragment C3d by means of a monoclonal antibody. Rev. Immunol. 9:373. J. Exp. Med. 158:1021. 30. Murre, C., P. S. McCaw, and D. Baltimore. 1989. A new DNA binding and 3. Weis, J. J., T. F. Tedder, and D. T. Fearon. 1984. Identification of a 145,000 M r dimerization motif in immunoglobulin enhancer binding, daughterless, MyoD membrane protein as the C3d receptor (CR2) of human B lymphocytes. Proc. and myc proteins. Cell 56:777. Natl. Acad. Sci. USA 81:881. 4. Fingeroth, J. D., J. J. Weis, T. F. Tedder, J. L. Strominger, P. A. Biro, and 31. Sun, X. H., and D. Baltimore. 1991. An inhibitory domain of E12 transcription D. T. Fearon. 1984. Epstein-Barr virus receptor of human B lymphocytes is the factor prevents DNA binding in E12 homodimers but not in E12 heterodimers. C3d receptor CR2. Proc. Natl. Acad. Sci. USA 81:4510. Cell 64:459. 5. Nemerow, G. R., R. Wolfert, M. E. McNaughton, and N. R. Cooper. 1985. Iden- 32. Murre, C., G. Bain, H. Van Dijk, I. Engel, B. A. Furnari, M. E. Massari, tification and characterization of the Epstein-Barr virus receptor on human B J. R. Matthews, M. W. Quong, R. R. Rivera, and M. H. Stuiver. 1994. Structure lymphocytes and its relationship to the C3d complement receptor (CR2). J. Virol. and function of helix-loop-helix proteins. Biochim. Biophys. Acta 1218:129. 55:347. 33. Bain, G., E. Maandag, D. Izon, D. Amsen, A. Kruisbeek, B. Weintraub, I. Krop, 6. Aubry, J.-P., S. Pochon, J.-F. Gauchat, A. Nueda-Marin, V. M. Holers, P. Graber, M. Schlissel, A. Feeney, M. van Roon, et al. 1994. E2A proteins are required for C. Siegfried, and J.-Y. Bonnefoy. 1994. CD23 interacts with a new functional proper B cell development and initiation of immunoglobulin gene rearrange- extracytoplasmic domain involving N-linked oligosaccharides on CD21. J. Im- ments. Cell 79:885. munol. 152:5806. 34. Zhuang, Y., P. Soriano, and H. Weintraub. 1994. The helix-loop-helix gene E2A 7. Delcayre, A. X., F. Salas, S. Mathur, K. Kovats, M. Lotz, and W. Lernhardt. is required for B cell formation. Cell 79:875. 1991. Epstein-Barr virus/complement C3d receptor is an interferon-␣ receptor. 35. Neufeld, B., A. Grosse-Wilde, A. Hoffmeyer, B. W. M. Jordan, P. Chen, EMBO J. 10:919. D. Dinev, S. Ludwig, and U. R. Rapp. 2000. Serine/threonine kinases 3pK and 8. Tedder, T. F., L. T. Clement, and M. D. Cooper. 1984. Expression of C3d re- MAPK-activated protein kinase 2 interact with the basic helix-loop-helix tran- ceptors during human B cell differentiation: immunofluorescence analysis with scription factor E47 and repress its transcriptional activity. J. Biol. Chem. 275:20239. the HB-5 monoclonal antibody. J. Immunol. 133:678. 36. Sloan, S. R., C. P. Shen, R. McCarrick-Walmsley, and T. Kadesch. 1996. Phos- 9. Reynes, M., J. P. Aubert, J. H. M. Cohen, J. Audouin, V. Tricottet, J. Diebold, phorylation of E47 as a potential determinant of B-cell-specific activity. Mol. and M. D. Kazatchkine. 1985. Human follicular dendritic cells express CR1, Cell. Biol. 16:6900. CR2, and CR3 complement receptor antigens. J. Immunol. 135:2687. 37. Massari, M., R. Rivera, J. Voland, M. Quong, T. Breit, J. van Dongen, 10. Birkenbach, M., X. Tong, L. E. Bradbury, T. F. Tedder, and E. Kieff. 1992. O. de Smit, and C. Murre. 1998. Characterization of ABF-1, a novel basic helix- Characterization of an Epstein-Barr virus receptor on human epithelial cells. loop-helix transcription factor expressed in activated B lymphocytes. Mol. Cell. J. Exp. Med. 176:1405. Biol. 18:3130. 11. Tsoukas, C. D., and J. D. Lambris. 1988. Expression of CR2/EBV receptors on 38. Jen, Y. N., H. Weintraub, and R. Benezra. 1992. Overexpression of Id protein human thymocytes detected by monoclonal antibodies. Eur. J. Immunol. inhibits the muscle differentiation program: in vivo association of Id with E2A 18:1299. proteins. Genes Dev. 6:1466. 12. Fischer, E., C. Delibrias, and M. D. Kazatchkine. 1991. Expression of CR2 (the 39. Ordentlich, P., A. Lin, C.-P. Shen, C. Blaumueller, K. Matsuno, C3dg/EBV receptor, CD21) on normal human peripheral blood T lymphocytes. S. Artavanis-Tsakonas, and T. Kadesch. 1998. Notch inhibition of E47 supports J. Immunol. 146:865. the existence of a novel signaling pathway. Mol. Cell. Biol. 18:2230.