Cells Expression in Murine Dendritic Cells and B Differential ...

16 downloads 0 Views 342KB Size Report
The activated T cell-attracting CC chemokine CCL22 is expressed by stimulated B cells and ... mans, but there are exceptions such as the dendritic cell (DC)4-.
The Journal of Immunology

Differential Regulation of CCL22 Gene Expression in Murine Dendritic Cells and B Cells1 Hormas Ghadially,2 Xiao-Lan Ross, Claudia Kerst, Jun Dong, Angelika B. Reske-Kunz, and Ralf Ross3 The activated T cell-attracting CC chemokine CCL22 is expressed by stimulated B cells and mature dendritic cells (DC). We have cloned and sequenced the complete mouse gene, including 4 kb of the 5ⴕ-flanking promoter region, and detected two distinct sites for initiation of transcription by 5ⴕ-RACE. Reporter gene assays indicate that the promoter reflects the specificity of the endogenous gene. Within the proximal promoter region, we identified potential binding sites for NF-␬B, Ikaros, and a putative GC box. All three regions bind proteins. The NF-␬B site was shown to specifically bind NF-␬B subunits p50 and p65 from nuclear extracts of LPS-stimulated B cells, B cell line A20/2J, TNF-␣-stimulated bone marrow-derived DC, and DC line XS106. Furthermore, promoter activity was affected by targeted mutagenesis of the NF-␬B site and transactivation with p50 and p65. The region harboring the putative Ikaros site contributes to promoter activity, but the binding protein does not belong to the Ikaros family. The GC box was shown to specifically bind Sp1 using extracts from LPS-stimulated B cells and A20/2J but not from DC and DC line XS106. Additionally, Sp1 transactivated the promoter in A20/2J but not in XS106 cells, and mutation of the Sp1 site diminished transactivation. Furthermore, binding of the protein complex at the GC box is required for NF-␬B activity, and the spatial alignment of the binding sites is of critical importance for promoter activity. Thus, identical and distinct proteins contribute to expression of CCL22 in DC and B cells. The Journal of Immunology, 2005, 174: 5620 –5629.

C

hemokines are a group of functionally and structurally related small proteins that are critically important in initiating and controlling the homing and migration of leukocytes (reviewed in Refs. 1 and 2). Other functions of individual chemokines or chemokine receptors include, e.g., the regulation of angiogenesis (3) and hematopoiesis (4). Based on the relative position of cysteine residues, four families of chemokines are distinguishable, namely CXC (␣), CC (␤), CX3C (␦), and C (␥). The genes of most of the CXC, as well as the CC chemokines, are clustered and show a similar exon/intron structure, hinting at duplication of common ancestor genes during evolution. For most chemokines, homologues have been identified in mice and humans, but there are exceptions such as the dendritic cell (DC)4chemokine 1 (no mouse homologue) (5) and IL-8 (no clear-cut murine homologue), suggesting relatively recent evolutionary rearrangements and differences in the equipment of the immune system of mice and humans. The CCL22 is present in both mice and humans. In humans, CCL22 is also referred to as macrophage-derived chemokine or stimulated T cell chemotactic protein-1 and in mice as activated B

Clinical Research Unit Allergology, Department of Dermatology, Johannes Gutenberg-University, Mainz, Germany Received for publication July 1, 2004. Accepted for publication January 25, 2005. 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 the Deutsche Forschungsgemeinschaft, SFB548.

2

Current address: Department of Vascular Biology and Thrombosis Research, Medical University of Vienna, Vienna, Austria. 3 Address correspondence and reprint requests to current address: Dr. Ralf Ross, Justus-Liebig-Universita¨t Giessen, Winchester Str. 2, D-35394 Giessen, Germany. E-mail address: [email protected] 4 Abbreviations used in this paper: DC, dendritic cell; LC, Langerhans cell; BM-DC, bone marrow-derived DC; ORF, open reading frame; UTR, untranslated region; EF1␣, elongation factor 1␣1.

Copyright © 2005 by The American Association of Immunologists, Inc.

cell-derived chemokine-1 or DC and B cell-derived chemokine (6 –9). CCL22 has been shown to play an important role in various diseases and mouse disease models. For example, elevated levels of CCL22 have been found in affected tissue of patients suffering from atopic dermatitis (10), atopic asthma (10), and contact hypersensitivity (11). Furthermore, CCL22 has been shown to be critically involved in mouse models of allergen-induced lung inflammation (12), atopic dermatitis (13), and septic peritonitis (14). The major function of CCL22 appears to be the induction of chemotactic migration of activated T cells (7–9, 15), predominantly of the Th2 type (16) and of CD4⫹CD25⫹ regulatory T cells (17), through its specific receptor CCR4. However, another as yet unidentified receptor has also been implicated (18, 19). The murine CCL22 is expressed by activated B lymphocytes and DC (8, 9). Murine epidermal Langerhans cells (LC), an exceptionally homogeneous population of immature DC, produce CCL22 mRNA only upon maturation (9), whereas other myeloid DC transcribe CCL22 constitutively (8). Epidermal LC and other immature DC represent sentinel cells of the immune system. Activation, e.g., by inflammatory stimuli, induces maturation into the most efficient APC known, which is outstanding in activating naive T cells (reviewed in Ref. 20). Applying a differential screening approach, we identified numerous genes differentially expressed during maturation of LC (21, 22). We used regulatory sequences of one of these genes, namely the actin-bundling protein fascin, to transcriptionally target DC (23, 24). Furthermore, the CCL22 gene was identified in this screening as the most abundantly expressed gene induced during maturation of murine epidermal LC (9). The high expression level and the tight regulation of expression, as indicated by the stage specificity, prompted us to isolate the gene and characterize the regulatory elements mediating expression in DC and B cells. Interestingly, DC and B cells use—at least in part—the same promoter elements but different protein complexes are formed. 0022-1767/05/$02.00

The Journal of Immunology

Materials and Methods Animals BALB/cJ mice were bred in the Central Animal Core Facility of the University of Mainz under specified pathogen-free conditions. Mice were used at 2–5 mo of age.

Cells Epidermal cell suspensions were prepared from pelts and used for LC enrichment either directly or following in vitro cultivation for 3 days as described previously (21). LC were enriched from epidermal cell suspensions by immunomagnetic separation with Dynabeads M-450 (Dynal Biotech) loaded with anti-MHC class II mAb 2G9 (anti-mouse I-Ad,b, I-Ed, rat IgG2a as described previously; Ref. 21). MHC class II-positive, bead-coupled LC and cells without attached beads were enumerated using a Neubauer hemocytometer. A purity of ⬃92–95% was obtained as determined by the ratio of bead-rosetted to nonrosetted cells. Cell viability, determined by trypan blue exclusion, was ⬎95%. Bone marrow-derived DC (BM-DC) were cultured as previously described (25) with modifications (26), and maturation was induced by adding either 50 ng/ml TNF-␣ (R&D Systems) or 1 ␮g/ml LPS (SigmaAldrich). MHC class II-positive cells were enriched 24 h later by immunomagnetic separation as described above. B lymphocytes were purified from spleen cells using the Mouse B Cell Recovery kit (Cedarlane Laboratories). Purity as assessed by cytofluorometry was ⬎95%. For stimulation, B cells were cultivated for 2–3 days in the presence of 25 ␮g/ml LPS (Sigma-Aldrich) in IMDM (Invitrogen Life Technologies) supplemented with 5% FCS (Sigma-Aldrich), 2 mM L-glutamine, 5 ⫻ 10⫺5 M 2-ME, 100 IU/ml penicillin, and 100 ␮g/ml streptomycin (complete medium). Naive B cells were isolated by immunomagnetic separation using anti-IgM-coupled Dynabeads. The murine B cell lymphoma A20/2J (27) was cultivated as described previously and stimulated with 25 ␮g/ml LPS for 24 h. The murine fibroblast line NIH 3T3 (28) was cultivated in complete medium. The murine immature and mature DC lines XS52 and XS106, respectively (29), kindly donated by Dr. A. Takashima (University of Texas Southwestern Medical Center, Dallas, TX), were cultivated in complete medium supplemented with 10% supernatant of NS 47 cells (30) and 5 and 10 ng/ml, respectively, of mouse recombinant GM-CSF. GM-CSF was kindly provided by Drs. F. Seiler and D. Krumwieh (Behringwerke AG). Mouse rIL-4 was a kind donation by DNAX.

Isolation of mRNA and RT-PCR mRNA was isolated using the QuickPrep Micro mRNA Purification kit as recommended (Amersham Biosciences). Reverse transcription reaction was performed as described previously (31). RNA concentration and efficiency of reverse transcription reaction were assessed by RT-PCR using primers specific for the housekeeping gene hypoxanthine phosphoribosyltransferase as reported previously (21). Hypoxanthine phosphoribosyltransferase-standardized amounts of cDNA were subjected to PCR with CCL22-specific primers as described previously (9).

5⬘-RACE 5⬘-RACE was performed using the SMART cDNA Library Construction kit (Clontech) for first-strand synthesis. PCR was performed using the 5⬘SMART PCR primer (5⬘-AAGCAGTGGTATCAACGCAGAGT-3⬘; Clontech) and CCL22 gene-specific primer 1 (5⬘-ATGGCTACCCT GCGTCGTGTCCC-3⬘) or primer 2 (5⬘-GACGAGCACCAGGAGTGG3⬘), respectively. Products were separated on an agarose gel, blotted on a nylon membrane (Roche), and hybridized with a digoxigenin-labeled probe derived from CCL22 cDNA. Signals were detected using the PCR DIG Probe Synthesis kit (Roche). Alternatively, products from a PCR using primer 2 and the 5⬘-SMART PCR primer were sequenced using primer 1.

Isolation and characterization of a murine genomic CCL22 clone A murine genomic library derived from a female C57BL/6 mouse (library no. 703) (32) was obtained from the Resource Center/Primary Database. The library was screened by hybridization with a probe containing the mouse CCL22 open reading frame (ORF) (9). Hybridization reactions and signal detection were performed as reported previously (33). P1 clones selected following library screening were used for DNA preparation (NucleoBond Plasmid kit; Clontech). Hybridization with mouse CCL22 cDNA probe was verified by Southern blot analysis. For additional characterization, restriction fragments of P1 clone DNA were subcloned randomly into pZErO 2.1 (Invitrogen Life Technologies), and subclones

5621 hybridizing with mouse CCL22 cDNA probe were identified by colony filter hybridization.

DNA sequencing and sequence analysis Nucleotide sequences were determined by cycle sequencing using the ABI PRISM Dye Termination Cycle Sequencing Ready Reaction kit as recommended and analyzed on a PE 373A DNA sequencer (PerkinElmer). GCrich DNA templates were linearized, heat-denatured (98°C for 5 min) and mixed with 5% (v/v) DMSO before sequencing. Potential transcription factor binding sites were identified by MatInspector V 2.2 (具http://transfac.gbf.de典). Transcription factors mentioned in the text are named according to the first corresponding publication.

Construction of CCL22 promoter reporter gene plasmids To assess activity of the CCL22 promoter, a fragment spanning a sequence ⬃4 kb long containing part of the 5⬘-nontranscribed region and the complete 5⬘-untranslated region (UTR) was subcloned into promoterless firefly luciferase expression vector pGL3-Basic (Promega), resulting in reporter construct pGL-CCL22. Correct orientation was confirmed by sequencing with vector-specific primers (RVprimer3 and GLprimer2; Promega). 5⬘Deletion constructs were obtained by restriction and religation or by nested deletion using the double-stranded Nested Deletion kit (Amersham Biosciences).

Transient cell transfection and luciferase assay Cells were transfected using GenePORTER (Gene Therapy Systems) as recommended by the manufacturer. Briefly, 1 ␮g of reporter DNA and 0.1 ␮g of the coreporter (see next paragraph) were incubated with 4 – 8 ␮l of GenePORTER in 250 ␮l of serum-free culture medium for 1 h before 4 ⫻ 105 cells were added. After 4 h, 2 ml of complete medium were added. Cells were analyzed 24 h posttransfection. To this end, cells were harvested, washed in PBS, and lysed in Passive Lysis Buffer (Promega). Cell lysates were analyzed sequentially for firefly and Renilla luciferase activity using the Dual-Luciferase Reporter Assay System (Promega) in a Turner luminometer TD-20/20 (Turner Design) as recommended by the manufacturer. To normalize absolute firefly luciferase reporter expression values for differences in transfection efficiency and cell recovery, a Renilla luciferase expression construct was cotransfected. To avoid increased coreporter activity due to cytokine- or LPS-mediated stimulation of CMV promoter, pRL-EF1␣ was used as a coreporter construct (24). Renilla expression in pRL-EF1␣ is driven by the strong human elongation factor 1␣1 (EF1␣) gene promoter regulated mainly by GC box-binding transcription factors (34). Reporter activities are calculated as ratio of firefly luciferase test construct activity divided by activity of cotransfected Renilla luciferase reporter construct. For transactivation studies using expression plasmids for Sp1 and Sp3, respectively, luciferase activity was normalized to the amount of protein in the cell lysate. Expression plasmids for NF-␬B p50, p65, and c-Rel were kindly provided by Dr. S. Pettersson (Karolinska Institute, Stockholm, Sweden), and plasmids for the expression of NF-␬B p52 and RelB were kind gifts by Dr. F. Weih (Forschungszentrum Karlsruhe, Karlsruhe, Germany). Expression plasmids for Sp1 and Sp3 (35) were kind gifts of Dr. G. Suske (PhillipsUniversita¨t, Marburg, Germany).

EMSA and site-directed mutagenesis Nuclear proteins were isolated by the method of Schreiber et al. (36), and protein concentration was determined using the Roti-Quant Protein Assay kit (Roth). Complementary oligonucleotides were incubated for 10 min at 95°C and annealed by decreasing the temperature by 1°C/min to 4°C. Following agarose gel electrophoresis, double-stranded oligonucleotides were recovered from the gel by electroelution and end-labeled using T4 polynucleotide kinase (New England Biolabs). Radiolabeled DNA probe (25,000 cpm) was mixed with 1 ␮g of poly(dI:dC) (Amersham Biosciences), 10 ␮g of nuclear proteins, and binding buffer (25 mM HEPES (pH 7.5), 150 mM KCl, 5 mM DTT, and 10% glycerol). Complex formation was allowed to proceed for 30 min at room temperature. For competition experiments, 50-fold molar excess of unlabeled, double-stranded oligonucleotides were added to the binding reaction. For supershift experiments, nuclear proteins were incubated with 2 ␮g of specific Abs for 30 min on ice before the binding reaction. Complexes were separated on 5% nondenaturing polyacrylamide gels. Gels were dried afterward and exposed to Kodak MS films (Eastman Kodak) at ⫺80°C. Site-directed mutagenesis was performed as described previously (37). The sequence of oligonucleotides (top strand) for EMSA and sitedirected mutagenesis were as follows: CCL22I, 5⬘-ACCGCTAAGGG

5622

REGULATION OF THE CCL22 GENE

GACTTTACAATCCTCC-3⬘; CCL22Imut, 5⬘-ACCGCTAAGGATAC TTTACAATCCTCC-3⬘; NF-␬B, 5⬘-GATCGAGGGGACTTTCCCTAGC3⬘; Oct2A, 5⬘-GTACGGAGTATCCAGCTCCGTAGCATGCAAATCCT GG-3⬘; CCL22II, 5⬘-ACCCACAGTTTGGGAGAGTGGCAGTCC-3⬘; CCL22IImut, 5⬘-ACCCACAGTTTGTAAGAGTGGCAGTCC-3⬘; Ikaros, 5⬘GATCATCCATAGGGAAAATTATCCTAG-3⬘; CCL22 III, 5⬘-CGCCTTT GGATCCTCCCGCTGGGGGAAG-3⬘; CCL22IIImut, 5⬘-CGCCTTTGG ATCCTCCCGCTGTGAGAAG-3⬘; Sp1, 5⬘-ATTCGATCGGGGCGGG GCGAG-3⬘; and C/EBP, 5⬘-TGCAGATTGCGCAATCTGCA-3⬘. Abs against Sp4 and NF-␬B subunits were obtained from Santa Cruz Biotechnology. Antisera against Sp1 and Sp3 (35) were kind gifts by Dr. G. Suske (Phillips-Universita¨t, Marburg, Germany).

Statistical analysis of data Statistical analysis of the experimental data was performed by the Student’s t test. A value of p ⬍ 0.05 was considered statistically significant.

Results Isolation of the murine CCL22 gene and identification of transcriptional initiation sites The murine CCL22 gene was isolated from a genomic library with a cDNA probe encompassing the CCL22 ORF. As only the exon/ intron boundaries of the gene had been reported previously (8), we subcloned and sequenced a continuous region of 11,728 bp, covering the murine CCL22 gene, including 4,096 bp of the 5⬘-flanking region (NCBI accession no. AJ315897). Analysis of the sequence confirmed that CCL22 is, as with other CC chemokines, encoded by three exons (Fig. 1A). The exons are separated by two introns of 1114 and 2758 bp, respectively. The sequences of the splice junctions are highly homologous to the corresponding human sequence and identical to the splice junctions in mouse reported by Schaniel et al. (8). To identify the transcriptional initiation site(s), we performed 5⬘-RACE using mRNA of in vitro-cultured LC and LPS-activated B cells, previously shown to express CCL22 (9). Two gene-specific 3⬘-primers were used in combination with a common 5⬘primer, and products were detected by hybridization with a probe containing the first exon of the CCL22 gene. Two lengths of 5⬘UTR were identified in both cell types, indicating the presence of two distinct transcriptional start sites (Fig. 1B). Sequencing of the RACE products revealed that the majority of the transcripts in both cell types are initiated from a site 28 bp upstream of the translation initiation codon (position 4070 of sequence AJ315897), giving rise to a very short 5⬘-UTR. In both cell types, a fraction of the transcripts are initiated from a site 124 bp upstream of the start codon (position 3973 of sequence AJ315897), which corresponds to the start of the cDNA sequenced by Schaniel et al. (NCBI accession no. AF052505; Ref. 8). Interestingly, a nearly perfect TATA box consensus sequence was found ⬃32 bp upstream of the major transcription initiation site, indicating that the majority of the transcripts of both, cultured LC and activated B cells, arise from a promoter containing a TATA box. The CCL22 promoter reflects the activity of the endogenous gene Based on our previous observation that CCL22 is expressed by mature, but not immature, DC and by activated, but not resting, B cells, we searched for an adequate model system to study CCL22 expression. By RT-PCR, we found that the mature DC line XS106 expressed CCL22 but not the immature DC line XS52 and confirmed that LPS stimulation induced expression of CCL22 in the B lymphoma cell line A20/2J (data not shown). In addition, we used the CCL22-negative fibroblast cell line NIH 3T3. To analyze CCL22 promoter activity in these cell lines, a 4069-bp fragment directly upstream of the translational start site was inserted into the vector pGL3-Basic, driving expression of the adjacent firefly

FIGURE 1. A, Schematic depiction of the murine CCL22 gene and CCL22 gene-derived reporter gene constructs. Exons are shown as boxes: f represents translated regions, and 䡺 represents untranslated regions. The two transcriptional start sites are indicated by positions relative to the start of translation. Recognition sites of restriction enzymes are abbreviated as follows: B, BamHI; E, EcoRI; H, HindIII; and S, SpeI. A luciferase reporter gene construct of the 5⬘-flanking promoter region and a series of constructs obtained by successive 5⬘-truncation of the promoter region are represented below. B, Identification of the transcriptional start sites of the CCL22 gene. 5⬘-RACE was performed on mRNA extracted from in vitro-cultured LC and LPS-activated B cells (BC), respectively. Products were separated by agarose gel electrophoresis and detected by Southern hybridization using a probe containing exon I of the CCL22 gene. C, Analysis of the promoter activity of the 5⬘-flanking region of the CCL22 gene. The indicated cell lines were transfected transiently with a reporter gene plasmid containing 4.1 kb of the 5⬘-flanking region of the CCL22 gene and cotransfected with a reporter plasmid coding for renilla luciferase driven by the EF1␣ promoter. Transcriptional activity was calculated as relative light unit relative to pGL3-Basic. Shown are the mean values (⫾ SD) of two independent experiments performed in triplicate. D, Deletion analysis of the CCL22 promoter. Equimolar amounts of luciferase reporter plasmids with 5⬘- to 3⬘-deletions of the CCL22 promoter were transfected transiently into the indicated cell lines and cotransfected with a reporter plasmid coding for renilla luciferase driven by the EF1␣ promoter. A20/2J cells were activated with LPS (25 ␮g/ml) immediately after transfection or were left untreated. The numbers on the abscissae represent the length of the promoter fragment in kb. Transcriptional activity was calculated as relative light unit relative to pGL3-Basic. Shown are the mean values (⫾ SD) of two independent experiments performed in triplicate.

The Journal of Immunology

5623

luciferase ORF (pGL-CCL22). This expression plasmid was transfected transiently into the cell lines, and firefly luciferase activity was measured 24 h posttransfection. A plasmid coding for renilla luciferase driven by the human EF1␣ promoter was cotransfected and used to standardize transfection efficiency. As shown in Fig. 1C, pGL-CCL22 mediated robust reporter activity in the mature DC line XS106, in LPS-stimulated A20/2J cells, and, to a markedly lesser extent (⬃4-fold less, p ⬍ 0.01), in unstimulated A20/2J cells. No activity was detected in XS52 and NIH 3T3. The enhancement of CCL22 expression by LPS stimulation of A20/2J cells was concentration dependent with a maximal induction at 25 ␮g/ml LPS (data not shown). As the activity of the CCL22 promoter fragment reflected the activity of the endogenous gene, we concluded that the major positive or negative regulatory elements necessary for stage-specific CCL22 expression in these cell types were present in pGL-CCL22. Boundaries of the CCL22 promoter To define the extension of the CCL22 promoter and to detect regulatory regions within the promoter, we generated serial 5⬘-deletion constructs of pGL-CCL22. Constructs were transfected into XS106 and A20/2J cells, respectively, and analyzed for activity by luciferase assay. A20/2J cells were either stimulated with LPS directly after transfection or were left untreated. As shown in Fig. 1D, a reporter plasmid containing only the most proximal 250 bp of the CCL22 promoter was able to drive most if not all of the luciferase activity in A20/2J cells. Moreover, full inducibility of the promoter by LPS was retained. Additionally, when 120 bp were removed from the 5⬘-end of the promoter fragment, nearly all activity was abrogated in nonactivated and LPS-activated A20/2J cells, indicating that pivotal regulatory elements are located within this region. In XS106 cells, deletion of the distal 2.5 kb of the promoter had little effect on reporter activity (Fig. 1D). However, additional deletion of the remaining 1.6-kb promoter fragment to 1.4 kb resulted in a ⬃2-fold increase ( p ⬍ 0.001) in luciferase activity, hinting at negatively regulating elements within this 200-bp region. Deletion from 1.4 to 0.55 kb had no significant effect on reporter activity, whereas an additional 5⬘-terminal deletion to 0.25 kb decreased promoter activity back to the level of the full-length promoter. However, removal of additional 120 bp abrogated promoter activity nearly completely ( p ⬍ 0.001). Thus, a small proximal promoter region of 250 bp is sufficient to mediate at least a major part of promoter activity in both cell types. Pivotal regulatory elements are located between positions ⫺250 and ⫺130 bp. Therefore, we searched for putative transcription factor binding sites within the most proximal 250 bp of the promoter using MatInspector software. Three regions harboring conserved transcription factor binding motifs were found, containing a putative binding site for NF-␬B and Ikaros, as well as a G/C-rich element (Fig. 2A). EMSA indicate that all three regions bind nuclear proteins (Fig. 2B). Furthermore, differences in the complexes obtained with nuclear extracts derived from unstimulated vs LPSstimulated A20/2J cells and XS52 vs XS106 cells show that variable complexes are formed at box I and box III (Fig. 2B). To analyze the contribution of the three protein-binding regions to the activity of the CCL22 promoter, reporter plasmids were generated in which these three regions were deleted progressively (Fig. 2C). Deletion of box I, which contains the NF-␬B binding site, resulted in loss of a major part of the inducibility of the promoter by LPS in A20/2J cells ( p ⬍ 0.001). Additional deletion of box II had little effect on the residual promoter activity, whereas removal of box III reduced the activity of the reporter plasmid nearly to background level ( p ⬍ 0.002). In XS106 cells, the deletion of box I resulted in

FIGURE 2. A, Schematic representation of the proximal region of the murine CCL22 promoter. The proximal 250 bp of the CCL22 promoter was analyzed using MatInspector software, and three putative binding sites for transcription factors were identified. The sequences of the putative binding sites are shown in the boxes, and the nucleotides changed by site directed mutagenesis are shown above the wild-type sequence. The position of the binding sites relative to the translation start codon is given in parentheses. The position of the translational start sites are indicated by arrows. B, Binding of proteins to the three sites in the proximal part of the CCL22 promoter. EMSA were performed on oligonucleotides of the three boxes using nuclear extracts isolated from the indicated cell lines. A20/2J cells were activated with LPS (25 ␮g/ml) before the isolation. Data are representative of three independent experiments. C, Serial deletion of the putative binding sites in the proximal part of the CCL22 promoter. Equimolar amounts of the indicated luciferase reporter plasmids were transfected transiently into the indicated cell lines and cotransfected with a reporter plasmid coding for renilla luciferase driven by the EF1␣ promoter. A20/2J cells were activated with LPS (25 ␮g/ml) immediately after transfection. Transcriptional activity was calculated as relative light unit relative to pGL3Basic. Shown are the mean values (⫾ SD) of two independent experiments performed in triplicate.

a slight but statistically not significant decrease of promoter activity. However, the additional deletion of box II resulted in a significant decrease ( p ⬍ 0.001). Most of the residual promoter activity was abrogated by the removal of all three boxes. These results indicate that all three regions contain cis-acting elements that contribute to the promoter activity in A20/2J and XS106 cells. However, the relative contribution is cell type dependent.

5624 Three regulatory regions of the CCL22 promoter bind transcription factors, including NF-␬B p50, p65, and Sp1 EMSA and supershift assays were conducted to evaluate the specificity of the protein complexes formed at box I–III and to identify the respective factors binding to these regions. At the distal box I, harboring a NF-␬B binding site, two DNA/ protein complexes formed, using nuclear extracts from unstimulated A20/2J cells (Fig. 2B). Stimulation of the cells induced the formation of a third complex with a lower mobility. Similarly, three complexes were detected in the DC lines tested with the low mobility complex being prominent in the mature DC line XS106, as compared with the immature DC line XS52. Competition experiments showed that complex formation was sequence specific (Fig. 3A). An excess of unlabeled region I-oligonucleotide or an oligonucleotide with a NF-␬B consensus-binding sequence was able to compete for complex formation, whereas an excess of an unrelated oligonucleotide, harboring an octamer-binding motif, did

FIGURE 3. Identification of proteins binding to the binding sites in the CCL22 promoter. EMSA were performed using oligonucleotides of box I (A), box II (B), and box III (C), and nuclear extracts were isolated from the indicated cell lines in the presence or absence of a molar excess of a cold competitor or in the presence or absence of antisera. Antisera were added to the nuclear proteins 30 min before adding the hot probe. A20/2J cells were activated with LPS (25 ␮g/ml) before the isolation. Competitors: A, wild-type box I (wt), consensus-binding sequence of Oct2a (Oct2a), consensus-binding sequence of NF-␬B (NF-␬B), and mutated box I (mut); and B, wild-type box II (wt), consensus-binding sequence of Ikaros (Ikaros), and mutated box II (mut). Data are representative of two independent experiments.

REGULATION OF THE CCL22 GENE not affect complex formation and neither did an oligonucleotide containing two base substitutions in the NF-␬B core sequence of region I. To analyze whether indeed proteins of the NF-␬B family bind to this region, we conducted supershift assays. An anti-p50 antiserum supershifted a complex formed with nuclear extracts of LPS-activated A20/2J cells and XS106 cells, respectively (Fig. 3A). Addition of anti-p65 antiserum resulted in decreased complex formation in XS106 cells and A20/2J cells. Antisera or mAb directed against p52, RelB, or c-Rel had no substantial effect on complex formation (Fig. 3A). At box II, harboring an Ikaros binding site, three DNA/protein complexes formed, regardless of whether the nuclear extracts were derived from unstimulated/LPS-stimulated A20/2J, XS52, or XS106 cells (Fig. 2B). Excess of unlabeled region II-oligonucleotide blocked complex formation (Fig. 3B). Notably, two base substitutions in the Ikaros core sequence did not affect competition, and an Ikaros consensus binding site did not compete for complex formation, both suggesting that proteins other than Ikaros bind to a distinct sequence of region II. This assumption was verified by supershift assays. Antiserum detecting all known differentially spliced Ikaros isoforms did not induce an additional shift (Fig. 3B). MatInspector detected solely the putative Ikaros site, which is obviously not involved. Thus, the binding site of region II or even the transcription factor binding to this site has not been identified previously. Box III harbors a GC-rich sequence with putative binding sites for transcription factors of the Sp and AP-2 family. Nuclear proteins from unstimulated and LPS-activated A20/2J, as well as XS106, cells formed three distinct complexes, whereas proteins from XS52 cells formed only one complex (Fig. 2B). Competition experiments showed that formation of two of the complexes was specific, whereas the formation of the complex with the highest mobility was nonspecific. An unlabeled oligonucleotide with a two-base substitution in the G-rich sequence did not compete for complex formation, whereas an oligonucleotide with a known Sp1binding sequence did compete for complex formation using nuclear extracts of A20/2J and XS106 cells (data not shown). Binding of factors of the Sp and AP-2 family to box III was analyzed by supershift assays using polyclonal antisera against Sp1, Sp3, Sp4, AP-2␣, AP-2␤, and AP-2␥ (Fig. 3C). The anti-Sp1 antiserum supershifted complex 2 formed with nuclear proteins derived from LPS-stimulated A20/2J cells but not from XS106 cells. The other antisera did not have an effect on complex formation with nuclear proteins of LPS-stimulated A20/2J cells or XS106 cells. As region III also harbors a sequence with some similarity to the C/EBP consensus-binding sequence, we used mAb or antisera against the C/EBP isoforms ␣, ␤, ␥, ␦, ⑀, and ␨ for supershift assays (Fig. 3C). No effect on complex formation with nuclear proteins of LPS-stimulated A20/2J cells or XS106 cells was observed (Fig. 3C). An antiserum cross-reacting with the C/EBP isoforms ␣, ␤, ␦, and ⑀ did not have an effect either (data not shown). In conclusion, Sp1 binds to box III using extracts of LPS-stimulated A20/2J cells but not of XS106 cells. The corresponding factor derived from XS106 cells was not identifiable because antisera against all known factors with specificity for the putative binding sites of box III had no effect in supershift assays. Effect of site-directed mutagenesis of transcription factor binding sites on CCL22 promoter activity To further elucidate the importance of the identified binding sites of the CCL22 promoter and the relative contribution of the respective transcription factors, single nucleotides of the binding sites

The Journal of Immunology

5625

FIGURE 4. A, Effect of mutations in the transcription factor binding sites of boxes I, II, and III on the activity of the CCL22 promoter. Base pairs in the binding sites were changed to the sequence indicated in Fig. 2A using site-directed mutagenesis. Equimolar amounts of the indicated luciferase reporter plasmids were transfected transiently into the indicated cell lines and cotransfected with a reporter plasmid coding for Renilla luciferase driven by the EF1␣ promoter. A20/2J cells were activated with LPS (25 ␮g/ml) immediately after transfection. Transcriptional activity was calculated as relative light unit relative to pGL3-Basic. Shown are the mean values (⫾ SD) of two independent experiments performed in triplicate. B, Effect of mutations in the binding sites on the inducibility of the CCL22 promoter by Sp1 and NF-␬B. Equimolar amounts of the indicated luciferase reporter plasmids were transfected transiently into the indicated cell lines and cotransfected with a plasmid facilitating the expression of the indicated transcription factor. Activity measured after transfection with a plasmid containing no promoter (pGL3-Basic) was set to 1. For experiments using NF-␬B expression plasmids, transfection efficiency was normalized to the activity of the coreporter vector pRL-EF1␣, whereas for experiments using Sp expression plasmids, activity was normalized to the protein content of the cell lysate. Shown are the mean values (⫾ SD) of two independent experiments performed in triplicate. C, Effect of the overexpression of NF-␬B subunits and Sp proteins, respectively, on the activity of the CCL22 promoter. The wild-type luciferase reporter plasmid was transfected transiently into the indicated cell lines and cotransfected with a plasmid facilitating the expression of the indicated transcription factor. Transcriptional activity was calculated as relative light unit relative to pGL3-Basic. For experiments using NF-␬B expression plasmids, transfection efficiency was normalized to the activity of the coreporter vector pRL-EF1␣, whereas for experiments, using Sp expression plasmids activity was normalized to the protein content of the cell lysate. Shown are the mean values (⫾ SD) of two independent experiments performed in triplicate. D, Effect of the substitution of box II and the alteration of the spatial alignment, respectively, on the activity of the CCL22 promoter. Box II was substituted with an unrelated sequence of the same length, or short sequences were inserted into the wild-type reporter plasmid as shown. Equimolar amounts of the indicated luciferase reporter plasmids were transfected transiently into the indicated cell lines and cotransfected with a reporter plasmid coding for Renilla luciferase driven by the EF1␣ promoter. A20/2J cells were activated with LPS (25 ␮g/ml) immediately after transfection. Transcriptional activity was calculated as relative light unit relative to pGL3-Basic. Shown are the mean values (⫾ SD) of two independent experiments performed in triplicate.

were substituted in luciferase reporter constructs (see Fig. 2A), and promoter activity was analyzed by luciferase assays. Substitution of either two nucleotides of the NF-␬B or of the Sp1 binding site drastically reduced promoter activity in LPS-stim-

ulated A20/2J cells ( p ⬍ 0.001 and p ⬍ 0.002, respectively), abrogating inducibility of the promoter (Fig. 4A). In XS106 cells, mutation of the Sp1 binding site had a strong effect on promoter activity as well ( p ⬍ 0.001), indicating that exactly this site is

5626 required, although it is not a known factor of the Sp1 family that binds to this region, as shown. Mutation of the NF-␬B site resulted in a moderate but statistically not significant reduction of promoter activity in XS106 cells as well (Fig. 4A). In accordance with the finding that Ikaros and AP-2 do not bind to the respective sites of the CCL22 promoter, base substitutions in these sites had no effect on promoter activity in both cell lines. Effect of overexpression of transcription factors on CCL22 promoter activity Next, we examined the effect of overexpression of transcription factors on the activity of the CCL22 promoter. CCL22 promoter/ luciferase reporter constructs were cotransfected with expression vectors coding for the NF-␬B proteins p50, p52, p65, RelB, and c-Rel, either separately or as combination of two factors, to allow for formation of functional units of homo- and heterodimers, respectively. In A20/2J cells, overexpression of p65 alone increased promoter activity ⬃2-fold ( p ⬍ 0.001; Fig. 4C). Overexpression of c-Rel had a lesser effect ( p ⬍ 0.05), and overexpression of p50, p52, and RelB had no marked effect. Overexpression of both, p50 and p65, resulted in an ⬃10-fold increase of promoter activity ( p ⬍ 0.001), being by far the most effective combination of NF-␬B proteins in activation of the CCL22 promoter. In contrast to A20/2J cells, XS106 cells require, as reported above, no stimulation to show strong CCL22 promoter activity. Nevertheless, a combination of p50 and p65 further enhanced promoter activity ( p ⬍ 0.001), whereas all other combinations did not augment the already high expression level (Fig. 4C). To investigate the effect of overexpression of NF-␬B subunits on the endogenous CCL22 mRNA production, A20/2J cells were transfected transiently with NF-␬B expression plasmids and analyzed for CCL22 mRNA expression by RT-PCR. To assure comparable transfection efficiency, cells were cotransfected with a renilla luciferase coreporter, and samples displaying similar coreporter activity were chosen for RNA isolation. Although overexpression of p50 alone was not sufficient to induce CCL22 mRNA production, overexpression of p65 or of p50 together with p65 induced endogenous CCL22 mRNA production (data not shown). Taken together, these experiments show that p50 and p65 not only bind to region I of the CCL22 promoter, as indicated by EMSA, but induce expression acting probably as heterodimers. The transcription factor Sp1, shown to bind to region III in EMSA using nuclear extracts from LPS-stimulated A20/2J cells, was analyzed likewise. Cotransfection of expression plasmids coding for Sp1 resulted in an ⬃6-fold increase in luciferase reporter activity of CCL22 reporter constructs in A20/2J cells ( p ⬍ 0.001; Fig. 4C). However, in XS106 cells, cotransfection of expression plasmids coding for Sp1 had no effect on promoter activity, again indicating that Sp1 is involved in induction of expression in A20/2J but not in XS106 cells. Overexpression of Sp3 had no significant effect on promoter activity of reporter constructs in both cell lines (Fig. 4C). Furthermore, overexpression of either Sp1 or Sp3 was not sufficient to induce endogenous CCL22 mRNA production in A20/2J cells (data not shown). Transactivation of the CCL22 promoter by overexpression of Sp1 or NF-␬B p50 ⫹ p65 in A20/2J cells or of NF-␬B p50 ⫹ p65 in XS106 cells was abrogated by introducing a mutation in the respective binding sites (all p ⬍ 0.001; Fig. 4B). Interestingly, mutation of the Sp1 site affected inducibility of the promoter by overexpression of p50 ⫹ p65 in A20/2J and XS106 cells (both p ⬍ 0.001; Fig. 4B), hinting at an interaction between the protein complexes formed at region I and region III. Mutation of the AP-2 site had, as expected, no effect on inducibility by overexpression of Sp1 in A20/2J cells (data not shown).

REGULATION OF THE CCL22 GENE Functional interaction of region I and III depends on the spatial alignment within the CCL22 promoter Interaction between members of the NF-␬B family and Sp1 was reported before, e.g., for the promoter of VCAM-1 or HIV-longterminal repeat (38, 39). Likewise, importance of interaction of the protein complexes formed at box I and III of the CCL22 promoter was suggested by the finding that mutation of the Sp1 binding site markedly affected inducibility by overexpression of p50 ⫹ p65 (Fig. 4B). To demonstrate that the spatial alignment of box I and III is important for the functional interaction of the protein complexes, we inserted small DNA fragments of 18 and 24 bp, respectively. To exclude that effects were caused by disruption of a so-far undetected regulatory element or by disruption of a possible interaction between box II and either boxes I or III, one fragment was inserted between boxes I and II and the other fragment between box II and III. Promoter activity of both modified reporter constructs was markedly reduced in LPS-stimulated A20/2J cells and XS106 cells (all p ⬍ 0.001; Fig. 4D). Moreover, both insertions reduced promoter activity to the same extent compared with the wild-type construct. This reduction was more pronounced in A20/2J cells than in XS106 cells. Keeping in mind the importance of the spatial alignment of box I and III, we replaced box II by an unrelated sequence of exactly the same length to evaluate contribution of region II to promoter activity. The sequence was selected to minimize sequence similarity to avoid that it competes with the wild-type sequence for transcription factor binding. A marked reduction ( p ⬍ 0.001) of promoter activity was observed in XS106 cells (Fig. 4D). In accordance with the data presented above (Fig. 2C), the relative loss of activity by replacement of box II was less pronounced in LPSstimulated A20/2J cells ( p ⬍ 0.02; Fig. 4D). Distinct proteins from primary DC and B cells bind to region I and region III of the CCL22 promoter To investigate whether the same transcription factors are involved in primary cells as in the cell lines tested, EMSA and supershift assays were performed with nuclear extracts from primary B cells and BM-DC. Stimulation of primary B cells with LPS enhanced complex formation (data not shown), an antiserum against NF-␬B p50 induced a supershift, and an antiserum against p65 resulted in decreased complex formation, using nuclear extracts from LPS-activated B cells (Fig. 5). To investigate complex formation in DC, we used nuclear extracts of BM-DC instead of LC because in BM-DC analysis is not hampered by the rapid changes of CCL22 expression observed in LC (Refs. 8 and 9 and our unpublished results). We obtained distinct stages of maturation in BM-DC by stimulation with TNF-␣ and LPS, respectively; the latter representing a stronger maturation stimulus than TNF-␣. Three distinct complexes were formed using nuclear extracts from TNF-␣-stimulated BM-DC. An antiserum against NF-␬B p50 induced a supershift, and an antiserum against p65 resulted in decreased formation of the preexisting complexes, whereas a new low molecular mass complex appears (Fig. 5). In addition, an antiserum against RelB reduced complex formation. However, when extracts from LPS-stimulated BM-DC were used, addition of an antiserum against p50 supershifted both complexes, whereas the addition of an antiserum against p65 had little effect on complex formation. Instead, addition of an antiserum against RelB induced a marked decrease in formation of the lower mobility complex, indicating that p50/RelB rather than p50/p65 heterodimers from nuclear extracts of LPS-stimulated BM-DC bind to region I.

The Journal of Immunology

FIGURE 5. Identification of proteins isolated from LPS-activated B cells and TNF-␣- or LPS-matured BM-DC binding to box I and III, respectively. EMSA were performed on oligonucleotides of box I and box III, respectively, using nuclear extracts isolated from the indicated cells in the presence or absence of a molar excess of a cold competitor or in the presence or absence of antisera. Antisera were added to the nuclear proteins 30 min before adding hot probe. Data are representative of two independent experiments.

We next conducted EMSA with proteins from primary cells using region III as a probe. As had been observed for the B cell line A20/2J, extracts from LPS-activated splenic B cells formed three complexes, of which complex 2 was supershifted using an anti-Sp1 antiserum. When proteins isolated from LPS-matured BM-DC were used, three complexes were detected, of which none was affected by the addition of antisera against Sp1 or Sp3 (Fig. 5).

Discussion The mouse CC chemokine CCL22 is involved critically in regulation of immune responses (8 –12, 14, 15). Of interest are its tight regulation of expression and high expression levels in mature DC and activated B cells. Thus, we characterized the CCL22 gene and its regulation. CCL22 is clustered with CCL17 and CX3CL1 on human chromosome 16q13 (40), apart from the common CC chemokine locus on chromosome 17q11.2, where most CC chemokines are clustered. Likewise, in mice these three chemokines are clustered separately on a short chromosomal fragment spanning ⬍130 kb (19). The separate location and the extensive differences to other CC chemokines on amino acid levels (9) indicate a divergent evolution of this locus. Taking into account that the expression profiles of CCL22 and CCL17 are similar, this may suggest a common regulation of gene expression of these two chemokines. Our data indicate that in case of the CCL22 gene, the regulatory elements required to mediate a strong and highly specific expression are included in a 4.1-kb promoter region, with the major elements active in DC and B cells being located in a small proximal 250-bp region. Interestingly, the same regulatory elements within this proximal region (box I–III) are required in all cells analyzed, and there are similarities but there are also differences in the factors involved. The data obtained from EMSA, supershift analysis, site-directed mutagenesis, and transactivation studies are conclusive in suggesting that the NF-␬B proteins p50 and p65 act together, probably as heterodimers, to activate the CCL22 gene by binding to box I of the promoter in nearly all cells tested, namely the B cell line A20/2J, the DC line XS106, primary LPS-stimulated B cells, and TNF-␣-stimulated DC. Moreover, overexpression of p50 and p65 enhanced the endogenous CCL22 expression in A20/2J cells synergistically, as determined by RT-PCR (data not shown). However, the fact that the mutation as well as the deletion of the NF-␬B site had only a small and statistically not significant

5627 effect indicates that in this cell line, which expresses CCL22 constitutively, this NF-␬B site is of minor importance. In LPS-stimulated DC, p65 is replaced by RelB. A slightly decreased complex formation was already observed with antisera against RelB in TNF-␣-stimulated DC (Fig. 5). Because LPS represents the stronger maturation stimulus compared with TNF-␣, this may suggest that RelB replaces p65 during maturation of DC. However, in contrast to p65 overexpression of RelB does not enhance CCL22 promoter activity in A20/2J and XS106 cells, either alone or in combination with other proteins of the NF-␬B family. Thus, replacement of p65 by RelB may rather reduce than enhance CCL22 expression. Accordingly, RelB had been shown to be the most actively translocated NF-␬B protein in maturing DC (41). Saccani et al. (42) recently demonstrated that p65-containing dimers are replaced completely by RelB-containing dimers at the human CCL22 promoter during maturation of DC. A similar replacement of p65 by RelB at the IL-12p40 promoter correlated with a transcriptional shutdown of the promoter and a drop of polymerase II recruitment. However, they observed that RNA polymerase II levels at the CCL22 promoter peaked when RelB recruitment was maximal and that expression of the I␬B␣ superrepressor, which affects translocation of p65- and c-Rel- but not RelB-containing dimers, rather enhanced than suppressed CCL22 transcription (42). This may suggest that the human and mouse CCL22 promoters differ with respect to RelB activation. Remarkably, the factor Sp1 binds to box III in B cells and the B cell line A20/2J but not in DC and the DC line XS106. Moreover, Sp1 is able to transactivate the promoter in A20/2J but not in XS106 cells, indicating that either the concentration of the factor replacing Sp1 in XS106 is not limiting the expression or Sp1 cannot substitute for this factor in XS106 cells. Nevertheless, exactly the same site is required for activation of CCL22 expression in XS106 cells because substitution of two nucleotides of the Sp1 binding site disrupts activity of the promoter in XS106 cells. Furthermore, the complexes formed at box I and box III interact in both cell lines because a change of the spatial alignment of the boxes affects activity of the promoter in both cases (Fig. 4D). Given these close similarities and the fact that neither Sp1 nor another known member of the Sp1 family binds to the promoter in XS106 cells and DC, a so-far unknown member of the Sp1 family may be active in DC. The fact that LPS did not induce an additional complex to be formed at region III by proteins isolated from A20/2J indicates that Sp1 is already present in the nucleus of these cells before LPS activation. The functional interaction of Sp1 and NF-␬B at the CCL22 promoter, indicated (1) by the importance of the spatial alignment of the binding sites within the promoter and (2) by the fact that a functional Sp1 site is required for NF-␬B transactivation, was demonstrated for other promoters previously (38, 39). The fact that the distance between the NF-␬B and the Sp1 binding sites is considerably greater in the CCL22 promoter may suggest involvement of additional proteins. Such proteins may bind to the promoter themselves, e.g., at box II or even adjacent to the binding sites for Sp1 or NF-␬B. Likely candidates are C/EBP proteins because region III harbors a sequence with similarity to the C/EBP consensus-binding sequence, and interaction of NF-␬B and C/EBP was demonstrated previously (43). The fact that we did not detect binding of C/EBP isoforms ␣, ␤, ␥, ␦, ⑀, and ␨ in supershift assays (Fig. 3C) does not completely rule out an involvement because unstable higher order complexes may remain unnoticed. C/EBP isoforms ␣, ␤, and ␦ were detected in nuclear extracts of A20/2J, and XS106 cells and cotransfection of C/EBP ␣ ⫹ ␦ increased promoter activity in A20/2J cells ⬃5-fold, indicating that heterodimers of C/EBP ␣ and

5628

␦ may activate the promoter (data not shown). However, in XS106 cells, overexpression of C/EBP isoforms ␣, ␤, and ␦, either separately or as combination of two factors, had no effect on reporter activity (data not shown). An additional difference in regulation of the CCL22 promoter in A20/2J and XS106 cells was indicated by luciferase assays following 5⬘-terminal deletion of the promoter region, hinting at an enhancing promoter element located between 550 and 250 bp upstream of the translational start site, which is active in XS106 cells but not in A20/2J cells. Because we detected a putative binding site for GATA proteins in this region, we overexpressed GATA-1, GATA-2, and GATA-3 but did not detect an effect on reporter gene expression in A20/2J and XS106 cells (data not shown). Interestingly, a very recent study by Nakayama et al. (44) investigated the roles of two adjacent NF-␬B sites and a single AP-1 site of the human CCL22 promoter in induction by EBV-encoded LPM-1 in human B cells. Although we identified only one NF-␬B site in the murine CCL22 promoter, a putative AP-1 site is also present in the murine promoter (position 97–100 relative to the translation start site). However, because this site is situated between the two transcriptional start sites, it is unlikely that AP-1 is involved in the regulation of the murine CCL22 promoter. However, it cannot be ruled out that AP-1 is involved in the initiation of transcription of only a fraction of the products. Remarkably, no site with substantial homology to an Ikaros binding site is present in the human CCL22 promoter region, but a very G-rich element, similar to the Sp1 site identified here, is located 175 bp upstream of the translation start codon of the human CCL22 promoter. Although the CCL22 promoter reflects the specificity of the endogenous gene, there may be additional levels of regulation of CCL22 expression not analyzed in detail yet. Saccani and Natoli (45) detected, for example, acetylation of histones H3 and H4 at the human CCL22 locus during activation, a common mechanism of gene activation on chromatin level (46). Enhancer elements are often located far upstream, downstream, or in introns, especially in the first intron. The latter was tested by incorporating a 1.2-kb downstream region, including the first intron of the CCL22 gene into a luciferase reporter construct, but no regulatory sequences were detected (data not shown). The usage of two transcriptional initiation sites 28 and 124 bp upstream of the translational start codon, respectively, offers additional possibilities for regulation of CCL22 expression. Quantitative differences in the usage of these start sites were detected in cultured LC and LPS-activated B cells. The longer 5⬘-UTR contains two small ORF (mini ORF) (position ⫺120 to ⫺96 and ⫺116 to ⫺80, respectively). Such mini ORF in the 5⬘-UTR have been reported to interfere with translation of the corresponding proteins (47), raising the possibility that CCL22 production is controlled not only on the transcriptional but also on the translational level. However, the significance of these mini ORF in the 5⬘-UTR of a proportion of the CCL22 transcripts is not clear. Considering the pivotal role of DC in initiation of novel immune responses and the promising strategies to transcriptionally target DC for therapeutic use (23, 24, 48), thorough knowledge of gene regulation in DC will help to manipulate DC on this level.

Disclosures The authors have no financial conflict of interest.

References 1. Ono, S. J., T. Nakamura, D. Miyazaki, M. Ohbayashi, M. Dawson, and M. Toda. 2003. Chemokines: roles in leukocyte development, trafficking, and effector function. J. Allergy Clin. Immunol. 111:1185.

REGULATION OF THE CCL22 GENE 2. Zlotnik, A., and O. Yoshie. 2000. Chemokines: a new classification system and their role in immunity. Immunity 12:121. 3. Strieter, R. M., P. J. Polverini, S. L. Kunkel, D. A. Arenberg, M. D. Burdick, J. Kasper, J. Dzuiba, J. Van Damme, A. Walz, D. Marriott, et al. 1995. The functional role of the ELR motif in CXC chemokine-mediated angiogenesis. J. Biol. Chem. 270:27348. 4. Broxmeyer, H. E. 2001. Regulation of hematopoiesis by chemokine family members. Int. J. Hematol. 74:9. 5. Adema, G. J., F. Hartgers, R. Verstraten, E. de Vries, G. Marland, S. Menon, J. Foster, Y. Xu, P. Nooyen, T. McClanahan, K. B. Bacon, and C. G. Figdor. 1997. A dendritic-cell-derived C-C chemokine that preferentially attracts naive T cells. Nature 387:713. 6. Godiska, R., D. Chantry, C. J. Raport, S. Sozzani, P. Allavena, D. Leviten, A. Mantovani, and P. W. Gray. 1997. Human macrophage-derived chemokine (MDC), a novel chemoattractant for monocytes, monocyte-derived dendritic cells, and natural killer cells. J. Exp. Med. 185:1595. 7. Chang, M., J. McNinch, C. Elias, III, C. L. Manthey, D. Grosshans, T. Meng, T. Boone, and D. P. Andrew. 1997. Molecular cloning and functional characterization of a novel CC chemokine, stimulated T cell chemotactic protein (STCP-1) that specifically acts on activated T lymphocytes. J. Biol. Chem. 272:25229. 8. Schaniel, C., E. Pardali, F. Sallusto, M. Speletas, C. Ruedl, T. Shimizu, T. Seidl, J. Andersson, F. Melchers, A. G. Rolink, and P. Sideras. 1998. Activated murine B lymphocytes and dendritic cells produce a novel CC chemokine which acts selectively on activated T cells. J. Exp. Med. 188:451. 9. Ross, R., X. L. Ross, H. Ghadially, T. Lahr, J. Schwing, J. Knop, and A. B. Reske-Kunz. 1999. Mouse Langerhans cells differentially express an activated T cell-attracting CC chemokine. J. Invest. Dermatol. 113:991. 10. Schmidt-Weber, C. B., J. G. Wohlfahrt, and K. Blaser. 2001. DNA arrays in allergy and immunology. Int. Arch. Allergy Immunol. 126:1. 11. Greaves, D. R., T. Hakkinen, A. D. Lucas, K. Liddiard, E. Jones, C. M. Quinn, J. Senaratne, F. R. Green, K. Tyson, J. Boyle, et al. 2001. Linked chromosome 16q13 chemokines, macrophage-derived chemokine, fractalkine, and thymusand activation-regulated chemokine, are expressed in human atherosclerotic lesions. Arterioscler. Thromb. Vasc. Biol. 21:923. 12. Gonzalo, J. A., Y. Pan, C. M. Lloyd, G. Q. Jia, G. Yu, B. Dussault, C. A. Powers, A. E. Proudfoot, A. J. Coyle, D. Gearing, and J. C. Gutierrez-Ramos. 1999. Mouse monocyte-derived chemokine is involved in airway hyperreactivity and lung inflammation. J. Immunol. 163:403. 13. Vestergaard, C., H. Yoneyama, M. Murai, K. Nakamura, K. Tamaki, Y. Terashima, T. Imai, O. Yoshie, T. Irimura, H. Mizutani, and K. Matsushima. 1999. Overproduction of Th2-specific chemokines in NC/Nga mice exhibiting atopic dermatitis-like lesions. J. Clin. Invest. 104:1097. 14. Matsukawa, A., C. M. Hogaboam, N. W. Lukacs, P. M. Lincoln, H. L. Evanoff, and S. L. Kunkel. 2000. Pivotal role of the CC chemokine, macrophage-derived chemokine, in the innate immune response. J. Immunol. 164:5362. 15. Tang, H. L., and J. G. Cyster. 1999. Chemokine up-regulation and activated T cell attraction by maturing dendritic cells. Science 284:819. 16. Andrew, D. P., M. S. Chang, J. McNinch, S. T. Wathen, M. Rihanek, J. Tseng, J. P. Spellberg, and C. G. Elias, III. 1998. STCP-1 (MDC) CC chemokine acts specifically on chronically activated Th2 lymphocytes and is produced by monocytes on stimulation with Th2 cytokines IL-4 and IL-13. J. Immunol. 161:5027. 17. Iellem, A., M. Mariani, R. Lang, H. Recalde, P. Panina-Bordignon, F. Sinigaglia, and D. D’Ambrosio. 2001. Unique chemotactic response profile and specific expression of chemokine receptors CCR4 and CCR8 by CD4⫹CD25⫹ regulatory T cells. J. Exp. Med. 194:847. 18. Struyf, S., P. Proost, S. Sozzani, A. Mantovani, A. Wuyts, E. De Clercq, D. Schols, and J. Van Damme. 1998. Enhanced anti-HIV-1 activity and altered chemotactic potency of NH2-terminally processed macrophage-derived chemokine (MDC) imply an additional MDC receptor. J. Immunol. 161:2672. 19. Schaniel, C., F. Sallusto, C. Ruedl, P. Sideras, F. Melchers, and A. G. Rolink. 1999. Three chemokines with potential functions in T lymphocyte-independent and -dependent B lymphocyte stimulation. Eur. J. Immunol. 29:2934. 20. Banchereau, J., and R. M. Steinman. 1998. Dendritic cells and the control of immunity. Nature 392:245. 21. Ross, R., X. L. Ross, J. Schwing, T. Langin, and A. B. Reske-Kunz. 1998. The actin-bundling protein fascin is involved in the formation of dendritic processes in maturing epidermal Langerhans cells. J. Immunol. 160:3776. 22. Ross, R., A. Endlich, K. Kumpf, J. Schwing, and A.B. Reske-Kunz. 1998. Maturation of epidermal Langerhans cells in vitro is accompanied by down-regulation of 4F2 (CD98) as determined by differential display. J. Invest. Dermatol. 110:57. 23. Ross, R., S. Sudowe, J. Beisner, X. L. Ross, I. Ludwig-Portugall, J. Steitz, T. Tu¨ting, J. Knop, and A. B. Reske-Kunz. 2003. Transcriptional targeting of dendritic cells for gene therapy using the promoter of the cytoskeletal protein fascin. Gene Ther. 10:1035. 24. Bros, M., X. L. Ross, A. B. Reske-Kunz, and R. Ross. 2003. The human fascin gene promoter is highly active in mature DC due to a stage-specific enhancer. J. Immunol. 171:1825. 25. Scheicher, C., M. Mehlig, R. Zecher, and K. Reske. 1992. Dendritic cells from mouse bone marrow: in vitro differentiation using low doses of recombinant granulocyte-macrophage colony-stimulating factor. J. Immunol. Methods 154: 253. 26. Lutz, M. B., N. Kukutsch, A. L. Ogilvie, S. Rossner, F. Koch, N. Romani, and G. Schuler. 1999. An advanced culture method for generating large quantities of highly pure dendritic cells from mouse bone marrow. J. Immunol. Methods 223:77.

The Journal of Immunology 27. Kim, K. J., C. Kanellopoulos-Langevin, R. M. Merwin, D. H. Sachs, and R. Asofsky. 1979. Establishment and characterization of BALB/c lymphoma lines with B cell properties. J. Immunol. 122:549. 28. Todaro, G. J., and H. Green. 1966. High frequency of SV40 transformation of mouse cell line 3T3. Virology 28:756. 29. Xu, S., K. Ariizumi, G. Caceres-Dittmar, D. Edelbaum, K. Hashimoto, P. R. Bergstresser, and A. Takashima. 1995. Successive generation of antigenpresenting, dendritic cell lines from murine epidermis. J. Immunol. 154:2697. 30. Schuhmachers, G., S. Xu, P. R. Bergstresser, and A. Takashima. 1995. Identity and functional properties of novel skin-derived fibroblast lines (NS series) that support the growth of epidermal-derived dendritic cell lines. J. Invest. Dermatol. 105:225. 31. Ross, R., R. Kleinz, and A. B. Reske-Kunz. 1995. A method for rapid generation of competitive standard molecules for RT-PCR avoiding the problem of competitor/probe cross-reactions. PCR Methods Appl. 4:371. 32. Ioannou, P. A., C. T. Amemiya, J. Garnes, P. M. Kroisel, H. Shizuya, C. Chen, M. A. Batzer, and P. J. de Jong. 1994. A new bacteriophage P1-derived vector for the propagation of large human DNA fragments. Nat. Genet. 6:84. 33. Ross, R., X. L. Ross, B. Rueger, T. Laengin, and A. B. Reske-Kunz. 1999. Nonradioactive detection of differentially expressed genes using complex RNA or DNA hybridization probes. Biotechniques 26:150. 34. Nielsen, S. J., M. Praestegaard, H. F. Jorgensen, and B. F. Clark. 1998. Different Sp1 family members differentially affect transcription from the human elongation factor 1 A-1 gene promoter. Biochem. J. 333 (Pt. 3):51. 35. Hagen, G., J. Dennig, A. Preiss, M. Beato, and G. Suske. 1995. Functional analyzes of the transcription factor Sp4 reveal properties distinct from Sp1 and Sp3. J. Biol. Chem. 270:24989. 36. Schreiber, E., P. Matthias, M. M. Muller, and 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. Makarova, O., E. Kamberov, and B. Margolis. 2000. Generation of deletion and point mutations with one primer in a single cloning step. Biotechniques 29:970.

5629 38. Neish, A. S., L. M. Khachigian, A. Park, V. R. Baichwal, and T. Collins. 1995. Sp1 is a component of the cytokine-inducible enhancer in the promoter of vascular cell adhesion molecule-1. J. Biol. Chem. 270:28903. 39. Perkins, N. D., A. B. Agranoff, E. Pascal, and G. J. Nabel. 1994. An interaction between the DNA-binding domains of RelA (p65) and Sp1 mediates human immunodeficiency virus gene activation. Mol. Cell. Biol. 14:6570. 40. Nomiyama, H., T. Imai, J. Kusuda, R. Miura, D. F. Callen, and O. Yoshie. 1998. Human chemokines fractalkine (SCYD1), MDC (SCYA22) and TARC (SCYA17) are clustered on chromosome 16q13. Cytogenet. Cell Genet. 81:10. 41. Neumann, M., H. Fries, C. Scheicher, P. Keikavoussi, A. Kolb-Maurer, E. Brocker, E. Serfling, and E. Kampgen. 2000. Differential expression of Rel/ NF-␬B and octamer factors is a hallmark of the generation and maturation of dendritic cells. Blood 95:277. 42. Saccani, S., S. Pantano, and G. Natoli. 2003. Modulation of NF-␬B activity by exchange of dimers. Mol. Cell 11:1563. 43. Ray, A., M. Hannink, and B. K. Ray. 1995. Concerted participation of NF-␬B and C/EBP heteromer in lipopolysaccharide induction of serum amyloid A gene expression in liver. J. Biol. Chem. 270:7365. 44. Nakayama, T., K. Hieshima, D. Nagakubo, E. Sato., M. Nakayama, K. Kawa, and Y. Osamu. 2004. Selective induction of Th2-attracting chemokines CCL17 and CCL22 in human B cells by latent membrane protein 1 of Epstein-Barr virus. J. Virol. 78:1665. 45. Saccani, S., and G. Natoli. 2002. Dynamic changes in histone H3 Lys 9 methylation occurring at tightly regulated inducible inflammatory genes. Genes Dev. 16:2219. 46. Dillon, N., and R. Festenstein. 2002. Unravelling heterochromatin: competition between positive and negative factors regulates accessibility. Trends Genet. 18: 252. 47. Garnier, G., A. Circolo, and H. R. Colten. 1995. Translational regulation of murine complement factor B alternative transcripts by upstream AUG codons. J. Immunol. 154:3275. 48. Takashima, A., and A. Morita. 1999. Dendritic cells in genetic immunization. J. Leukocyte Biol. 66:350.