CD94/NKG2A is an inhibitory receptor expressed by most human NK cells and a subset of T cells that recognizes HLA-E on potential target cells. To study the transcriptional regulation of the human NKG2A gene, we cloned a 3.9-kb genomic fragment that contains a 1.65-kb region upstream of the exon 1, as well as exon 1 (untranslated), intron 1 and exon 2. Using deletion mutants, we identified a region immediately upstream from the most upstream transcriptional initiation site that led to increased transcriptional activity from a luciferase reporter construct in YT-Indy (NKG2A positive) cells relative to Jurkat and K562 (both NKG2A negative) cells. We also localized a DNase I hypersensitivity site to this region. Within this 80-bp segment, we identified two GATA binding sites. Mutation of GATA binding site II (−2302 bp) but not GATA binding site I (−2332 bp) led to decreased transcriptional activity. Pull-down assays revealed that GATA-3 could bind oligonucleotide probes containing the wild type but not a mutated GATA site II. Using chromatin immunoprecipitation assays, we showed that GATA-3 specifically binds to the NKG2A promoter in situ in NKL and primary NK cells, but not in Jurkat T cells. Moreover, coexpression of human GATA-3 with an NKG2A promoter construct in K562 cells led to enhanced promoter activity, and transfection of NKL cells with small interfering RNA specific for GATA-3 reduced NKG2A cell surface expression. Taken together, our data indicate that GATA-3 is an important transcription factor for regulating NKG2A gene expression.
Natural killer cells are a class of lymphocytes that can mediate lysis of many virally infected and tumorigenic cells, without prior sensitization. The consequences of NK cell activation are target cell lysis and/or the production of inflammatory cytokines, such as TNF-α and IFN-γ (1, 2, 3). NK cells express a broad variety of activation receptors that allow them to recognize potential target cells. To prevent damage of normal cells and unwanted inflammation, NK cells also express an array of inhibitory receptors, most of which are specific for MHC class I molecules (4, 5, 6). Class I molecules are expressed by almost all normal cells and thereby play an important role in protecting normal cells from NK cell cytotoxic activity. Upon ligation, inhibitory receptors block activation signals at a very early point in the signaling cascade by recruiting protein tyrosine phosphatases to the proximity of signal initiation (7, 8, 9).
A major NK cell inhibitory receptor in both mice and humans is a heterodimer composed of covalently associated CD94 and NKG2A (or its isoform NKG2B) (2, 4, 7, 10, 11, 12), whose ligand is the nonclassical class I molecule HLA-E (13, 14, 15). CD94/NKG2A expression can also be induced on subpopulations of peripheral blood CD8+ αβ and γδ T cells after activation by TCR triggering combined with cytokine stimulation (16, 17, 18, 19). CD94/NKG2A expression has been shown to be dramatically up-regulated on CD8+ T cells in certain diseases, suggesting that its expression may have pathological consequences. For example, a significant percentage of melanoma infiltrating CD8+ T cells express CD94/NKG2A (20) and, in HIV-infected individuals, the percentage of T cells that express CD94/NKG2A is significantly elevated (21). In addition, the percentage of CD8+ T cells that express CD94/NKG2A in mice is significantly increased after infection with Listeria monocytogenes, polyoma virus, and lymphocytic choriomeningitis virus (22, 23, 24). Thus, CD94/NKG2A inhibitory receptors, already known to be important for regulating cytotoxic and inflammatory responses of NK cells, may also regulate CD8+ T cell responses.
Currently, little data exist on what regulates gene expression of CD94/NKG2A receptors by NK or T cells. The genes encoding the human CD94 and NKG2A molecules are located in the NK complex on chromosome 12p12-p13 that contains 19 genes encoding C-type lectins, including the other members of the NKG2 family (25, 26, 27). The human CD94 gene has seven exons (untranslated exon 1a plus 6 coding exons) and dual promoters leading to two types of transcripts that are different in their 5′ untranslated regions (28, 29). The promoters show differential sensitivity to IL-2 and IL-15 and both contain elements with IFN-γ-activated and Ets binding sites, known as GAS/EBS. In freshly isolated primary NK and CD8+ αβ T cells, only the proximal promoter is active (29). The NKG2A gene has 7 exons (30) that, by alternative splicing of the pre-mRNA, encode both NKG2A and B transcript (31). It has been shown that there are multiple transcription starting sites upstream of both untranslated exon 1 and exon 2 (32, 33).
In this study, we identified a basal promoter 80 bp upstream from the most upstream transcriptional initiation site (32) that contained key promoter regulatory elements for NKG2A gene expression in YT, NKL (both NKG2A positive), but not in Jurkat or K562 (both NKG2A negative) cell lines. Consistent with this data, a DNase I hypersensitive site (DHSI)5 is present within this region in genomic DNA isolated from NKL cells. By mutation analyses, we identified a GATA-3 binding site critical for NKG2A gene expression. Using chromatin immunoprecipitation (ChIP) assays, we showed that the region containing the GATA binding site II (−2302) interacts in situ with the GATA-3 transcription factor in primary NK and NKL cells, but not in Jurkat T (GATA-3 positive) cells. We also demonstrated that GATA-3 is capable of transactivating NKG2A promoter activity and that transfection with small interfering RNA (siRNA) specific for GATA-3 inhibits cell surface expression of NKG2A. Taken together, our data indicate that GATA-3 is an important transcription factor for regulating NKG2A gene expression.
Materials and Methods
Cell isolation and culture conditions
Polyclonal NK cells were isolated by negative selection from peripheral blood using an NK cell isolation kit (Miltenyi Biotec). The purity (>95%) of the isolated cell populations was confirmed by flow cytometry (29). Isolated NK cells were cultured in IMDM (BioWhittaker) supplemented with 500 U/ml rIL-2 (Biological Resources Branch, National Cancer Institute-Frederick Cancer Research and Development Center, Frederick, MD), 10% human AB serum (BioWhittaker), and 2 mM l-glutamine (BioSource International) at 37°C under an atmosphere of 5% CO2.
The human NK leukemia cell line, NKL, was grown in RPMI 1640 (BioSource International) supplemented with 2 mM l-glutamine, 1 mM sodium pyruvate, 200 U/ml rIL-2, 5 μg/ml plasmocin (InvivoGen), and 10% FBS at 37°C under an atmosphere of 5% CO2. The YT-Indy (human NK tumor cell line), Jurkat (human T cell lymphoma), and K562 (human erythroleukemia cell line) cells were grown in RPMI 1640 (BioSource International) supplemented with 2 mM l-glutamine, 10 mM nonessential amino acids, 5 μg/ml plasmocin, and 10% FBS at 37°C under an atmosphere of 5% CO2. NK92 (human NK tumor cell line) cells were cultured in DMEM (BioSource International) supplemented with 2 mM glutamine, 1.5 g/L of NaHCO3, 0.2 mM inositol, 0.1 mM 2-ME, 0.02 mM folic acid, 200 U/ml rIL-2, and 12.5% horse serum (Invitrogen Life Technologies).
DNase I hypersensitivity assay
Human NKL cell nuclei were isolated using the Blood & Cell Culture DNA Mini kit (Qiagen), according to the manufacturer’s instructions. After quantification, nuclei were digested for 5 min at room temperature with varying concentrations of DNase I. The reaction was stopped by the addition of SDS (1% final concentration) and EDTA (10 mM final concentration). The lysates were treated with proteinase K overnight at 55°C, and DNA was isolated using the Blood & Cell Culture DNA Mini kit (Qiagen). Isolated DNA was digested with AhdI and ApalI and analyzed by Southern hybridization with a 508 bp probe corresponding to positions −4466 to −3958 bp upstream of the translational starting site (see Fig. 2 B).
All DNA fragments were derived by PCR using platinum Taq polymerase (Invitrogen Life Technologies) with human genomic DNA (Promega) as template. The NKG2A promoter region was subcloned into the XhoI and SmaI sites of the pGL3-basic reporter plasmid (Promega). The primers used for the promoter region deletion constructs were: P-3905 forward 5′-ggatccagaatttgtttcaagatctg-3′; P-2705 forward 5′-agaactaagagaagttaataatata-3′; P-2505 forward 5′-agtatgcaatgttaagatttagca-3′; P-2335 forward 5′-aagtatcctatattttggaggtgt-3′; P-2295 forward 5′-gatttcacaatagcccatgtgtag-3′; and RV-1 reverse 5′-gcagtgtgtgatgtcagggactgtactc-3′. The primers for the intron 1 deletion construct were ΔInt1-F forward 5′-ggggacagaagagtacagtcccctgacatc-3′ and ΔInt1-R reverse 5′-gactgtactcttctgtccccagaaagtca-3′. Primers were designed for subcloning into the SmaI and XhoI sites of the pGL3-basic reporter plasmid.
Mutant constructs were generated with a QuikChange XL site-directed mutagenesis kit (Stratagene), using a pair of overlapping internal primers that contained a mutant sequence. The primers used for the promoter mutation constructs were: mutGA-1F forward 5′-ggaggtgtgtgatgaggaaaaactgattttcacaatag-3′; mutGA-2F forward 5′-agaccaagttccctatattt-3′; mutGA-1R reverse 5′-ctattgtgaaatcagtttttcctcatcacacacctcc-3′; and mutGA-2R reverse 5′-aaatatagggaacttggtct-3′.
Human GATA-3 cDNA was prepared from total RNA extracted from Jurkat T lymphoma cells by RT-PCR using the following pair of primers: forward 5′-cggaattcgccatggaggtgacggcggacc-3′ and reverse 5′-gctctagactaacccatggcggtgaccatgct-3′. Primers were designed for subcloning into the EcoRI and XbaI sites of the pcDNA3.1 neo+ expression vector (Invitrogen Life Technologies). All plasmid DNA used for transfections were purified with a Qiagen Plasmid Mega kit, according to the manufacturer’s protocol. All restriction enzymes were purchased from New England Biolabs. All custom synthesized oligonucleotides were supplied by Sigma-Genosys. All plasmid inserts were verified by sequence analysis.
Transfection and luciferase assay
A total of 1 × 106 of YT-Indy, Jurkat, or K562 cells were transiently transfected with 5 μg of the reporter construct and 2 μg of pCMV-β-galactosidase plasmid (Invitrogen Life Technologies), an internal control for the normalization of transfection efficiency, using the Bio-Rad GenePulser (300 mV, 0.965 mF). Cells were harvested 40 h after transfection and lysed. Luciferase and β-galactosidase activities were measured with a Monolight 3010 Luminometer (Analytical Luminescence Laboratory) and DU 530 Life Science UV/Vis Spectrophotometer (Beckman Coulter), respectively.
RNA isolation and RT-PCR
Total RNA was isolated from cultured human primary NK, NKL, YT-Indy, Jurkat, or K562 cells using RNeasy mini kits (Qiagen) and synthesis of cDNA was performed using the SuperScript III First-strand Synthesis System for RT-PCR (Invitrogen Life Technologies). To determine the expression levels of NKG2A, GATA-3, and β-actin (internal control), PCR was done with the Platinum PCR SuperMix High Fidelity system (Invitrogen Life Technologies) using specific pairs of primers. The following primers were used: human GATA-3 sense, 5′-gctgtctgcagccaggagagc-3′; antisense, 5′-atgcatcaaacaactgtggcca-3′ (34); human NKG2A sense 5′-ccagagaagctcattgttgg-3′; antisense 5′-ccaatccatgaggatggtg-3′; human β-actin sense 5′-cgcgagaagatgacccagatc-3′; antisense 5′-ttgctgatccacatctgctgg-3′ (35). PCR products were analyzed in 4% or 1% E-gels (Invitrogen Life Technologies).
Cell lysates were prepared and analyzed as previously described (8). Briefly, lysates fractionated by SDS-PAGE were transferred to Immobilon-P membranes (Invitrogen Life Technologies), following by blocking with 4% (w/v) skim milk (Super G) in blocking/washing buffer (10 mM Tris-HCl, pH 7.5, 100 mM NaCl, 0.1% Tween 20) for 1 h. Membranes were probed for 16 h at 4°C with either 1/500 rabbit polyclonal anti-GATA-3 Ab (Active Motif), 1/5000 mouse anti-β-actin mAb (Sigma-Aldrich) or 1/1000 anti NKG2A (8E4) mAb, an Ab derived by Houchins et al. (36). HRP-conjugated anti-IgG (Amersham Biosciences) diluted 1/4000 was used to label the complexed Abs. All Abs were diluted in blocking buffer. Membranes were washed after each step with blocking/washing buffer to remove unbound material. ECL Western blotting kits (Amersham Biosciences) were used to detect the proteins identified by the immunodetection conjugates.
NKL or primary NK cells were harvested, washed, and lysed in a buffer containing 10 mM HEPES, 133 mM KCl, 10% glycerol, 2 mM EDTA, 0.5 mM DTT, 0.5% Nonidet P-40, 1% phosphatase inhibitor mixture 1 (Sigma-Aldrich), 1% phosphatase inhibitor mixture 2 (Sigma-Aldrich), and 1% protease inhibitor mixture (Sigma-Aldrich). Cell lysates were precleared by incubation with streptavidin-agarose beads (Dynabeads M-280; Dynal) for 1 h at 4°C. Precleared cell lysates were incubated with streptavidin-agarose beads (Dynabeads M-280) coupled to biotinylated double-stranded oligonucleotides (Sigma-Genosys) containing a wild type (5′-ggaggtgtgtgatgagataaaactgatttcac-3′) or mutated (5′-ggaggtgtgtgatgagagcaaactgatttcac-3′) GATA binding site II (−2302) from the NKG2A promoter region. The binding reactions were performed overnight at 4°C. Unbound proteins were removed by six washes with lysis buffer. After washing, the oligonucleotide-bound proteins were released in Laemmli sample buffer (Sigma-Aldrich), boiled for 5 min, resolved on 4–12% SDS-PAGE, and detected by immunoblotting as earlier described.
Chromatin preparation and immunoprecipitation were performed with a ChIP Assay kit (Upstate Biotechnology) with minor modifications. Protein-DNA cross-linking was achieved by incubating NKL, YT-Indy, primary NK, or Jurkat cells (1 × 106 cells per condition) with 1% formaldehyde for 10 min at room temperature with gentle agitation. Cross-linking was blocked by adding glycine at a final concentration of 0.125 M and incubating at room temperature for 5 min. Cells were collected by centrifugation and washed in cold PBS containing 1% protease inhibitor mixture (Sigma-Aldrich).
Pelleted cells were resuspended in 200 μl of lysis buffer (Upstate Biotechnology) and incubated on ice for 10 min. Chromatin was then sonicated to an average length of 1–0.2 kb using a Branson Sonifier 450 (Branson Ultrasonics). Sonicated chromatin was then immunoprecipitated with anti-GATA-3 mAb (Santa Cruz Biotechnology) attached to microbeads using a protocol provided by Upstate Biotechnology. The bound protein-DNA complexes were eluted with freshly prepared elution buffer (1% SDS, 0.1 M NaHCO3). After reversing the protein-DNA cross-linking and subsequent DNA recovery, to detect NKG2A promoter region sequences, samples were analyzed by quantitative PCR (MJ Research DNA Engine Opticon 2) with SYBR Green JumpStart Taq ReadyMix (Sigma-Aldrich). The amount of NKG2A promoter related sequence present in each reaction was calculated relative to a standard curve obtained using a serial 10-fold dilution of human genomic DNA (Promega). For each experiment, two independent real-time PCRs were performed. The following primers were used in the ChIP assays: human NKG2A promoter, sense 5′-accaactaagtgacacactttc-3′ and antisense 5′-aggaaattctacacatgggc-3′; negative control primers, sense 5′-atggttgccactggggatct-3′ and antisense 5′-tgccaaagcctaggggaaga-3′. Negative control primers were from a ChIP-IT Chromatin Immunoprecipitaion kit (Active Motif).
Nucleofection with siRNA
Predesigned GATA-3 siRNA (ID no. 43336; Ambion) was used to selectively silence GATA-3 gene expression. The Silencer siRNA (no. 1; Ambion) was used as a negative control. Cells were transfected with 500 nM siRNA by electroporation using the Nucleofection system (Amaxa) as previously described (37). After 18 h of incubation, cells and lysates were analyzed by FACS and immunoblotting, respectively. FACS analysis was performed with PE-conjugated anti-NKG2A Ab (Z-199; Beckman Coulter) to check the cell surface expression of NKG2A. PE-conjugated isotype-matched control Ab was used to monitor background staining levels.
DNase I hypersensitivity mapping
Gene regulatory elements (e.g., promoters, enhancers, and silencers) are often located within regions of open chromatin structure. Such regions can often be identified by their hypersensitivity to DNase I digestion (38, 39, 40). In this regard, to attempt to localize the NKG2A promoter, NKL (NKG2A positive) cells (Fig. 1,A) were used for DNase I hypersensitivity mapping on the human NKG2A gene upstream from exon 6. A prominent DHSI was identified (Fig. 2,A) and localized in the region with multiple transcriptional start sites upstream of the NKG2A untranslated exon 1 (Fig. 2 B). This indicated that this region contains open chromatin structure that would be expected for an NKG2A promoter region.
The basal NKG2A promoter
To determine whether the sequence upstream of the first exon contained a promoter, we cloned a 3.9-kb genomic fragment that begins upstream of the ATG site (Fig. 3,A). First, we tested whether this fragment, which includes the 1.65 kb region upstream from the exon 1, as well as exon 1 (untranslated), intron 1 and exon 2 of the NKG2A gene, possesses promoter activity. This fragment was inserted into a luciferase reporter vector, pGL3-basic. Jurkat (NKG2A negative) and YT-Indy (NKG2A positive) cells (Fig. 1) were transfected with this construct by electroporation. As shown in Fig. 3 B, transfection of YT-Indy cells with the full-length 3.9 kb region upstream of exon 2 led to a 25-fold increase in the relative luciferase activity compared with pGL3-basic. A minimal increase in promoter activity was detected in Jurkat cells. This result indicates that this 3.9-kb region contains a promoter whose function is cell type-dependent.
A series of 5′ deletion mutant constructs of this 3.9-kb genomic fragment were created to locate the region critical for NKG2A promoter activity (Fig. 4,A). These constructs were inserted into the luciferase reporter vector, pGL3-basic. YT-Indy (NKG2A positive) cells, Jurkat (NKG2A negative), and K562 (NKG2A negative) cells were transfected with these deletion mutant constructs. Progressive deletion from −3905 to −2335 bp had a negligible effect on NKG2A promoter activity in YT-Indy cells, but further deletion from −2335 to −2295 bp resulted in a 70–80% reduction in luciferase activity (Fig. 4,B). We did not observe any significant difference in the activity of the deletion constructs in Jurkat (Fig. 4,C) and K562 cells lines (Fig. 4 D). These results indicate that the sequence between −2335 and −2295 bp contains a key cell-specific promoter regulatory region important for NKG2A gene expression.
A deletion mutant construct lacking the NKG2A intron 1 gene segment did not have reduced promoter activity in YT-Indy cells (Fig. 4 E), indicating that intron 1 does not contribute to NKG2A promoter activity in these cells.
Role of GATA-binding motifs in the NKG2A promoter activity
The nucleotide sequence upstream of exon 1, as well as a part of exon 1, is shown in Fig. 5,A. A transcriptional factor database search (TFSearch) of this sequence revealed the presence of a number of potential regulatory elements, including consensus sequences for GATA binding transcription factors (Fig. 5,A). We made constructs with double base pair mutations in both GATA binding sites (Fig. 5,B). The mutations in the GATA binding site I resulted in no change in the promoter activity, but the mutations in GATA binding site II significantly reduced the promoter activity (Fig. 5 B), indicating that this site is involved in the transcriptional regulation of the NKG2A gene.
GATA-3 from NKL and freshly isolated NK cells can bind to GATA binding site II in vitro
All members of the GATA transcription factor family bind to DNA consensus sequences through highly conserved DNA binding domains (41). Mammalian GATA proteins can be divided into two subfamilies based on their structures and their patterns of expression. GATA-1, GATA-2, and GATA-3 are predominantly expressed in hemapoietic cell lineages (42), whereas the GATA-4, GATA-5, and GATA-6 are expressed in the developing heart, gut, and smooth muscle cells (43). GATA-1 and GATA-2 are expressed in erythroid, megakaryocytes, eosinophils, and mast cells (44, 45, 46), whereas GATA-3 is expressed predominantly in T and NK cells (47, 48). Using pull-down assays, we investigated whether NKL and freshly isolated NK cells contain GATA-3 protein capable of binding to GATA binding site II in vitro. Whole cell extracts from NKL and freshly prepared NK cells were incubated with streptavidin-agarose beads coupled to biotinylated double-stranded oligonucleotides containing either a wild type or mutated GATA binding site II designed from the NKG2A promoter (Fig. 5,A). Proteins that bound to the beads were eluted, then separated by SDS-PAGE and the presence of bound GATA-3 protein was determined by immunoblotting with an anti-GATA-3 Ab (Fig. 6 A). As can be seen, the oligonucleotide containing the wild type GATA binding site II was significantly more efficient at binding GATA-3 protein than the oligonucleotide containing the mutated GATA binding site II. These results indicate that GATA-3 protein from NKL and freshly prepared NK cells specifically binds to the GATA binding site II present in the NKG2A promoter in vitro.
Interaction of GATA-3 with the NKG2A promoter in situ
The binding of GATA-3 to the NKG2A promoter in NKL (NKG2A positive, GATA-3 positive), Jurkat (NKG2A negative, GATA-3 positive), or freshly isolated NK cells was analyzed using a ChIP assay. In these analyses, cross-linked chromatin was immunoprecipitated using anti-GATA-3 mAb. As a negative control, we included a reaction containing mAb against STAT5a, a transcription factor that does not have a putative binding site in the defined NKG2A promoter region. After immunoprecipitation and reversal of the cross-links, enrichment of the NKG2A promoter fragment in each sample was monitored by quantitative real-time PCR using primers specific for the human NKG2A promoter region (Fig. 5 A). As an additional control, the same immunoprecipitate was also used to amplify a control DNA genomic region; no enrichment was observed (data not shown).
The results show a 6-fold (NKL) and 4-fold (primary NK) enrichment of the NKG2A promoter sequences in precipitates with GATA-3 Ab compared with control Ab (Fig. 6 B). No significant enrichment was observed in precipitates derived from Jurkat T cells. These results indicate that GATA-3 specifically binds to the NKG2A promoter in NKG2A expressing cells in situ.
Human GATA-3 can transactivate the NKG2A basal promoter
Because GATA-3 mRNA is not expressed in K562 cells (data not shown), we used this cell line to examine whether human GATA-3 can transactivate the basal promoter of the NKG2A gene. This was done by cotransfection of a luciferase reporter vector containing the NKG2A promoter construct with a cDNA expression plasmid encoding human GATA-3. A 1.7- to 2.5-fold increase in luciferase activity was observed in the presence of the GATA-3 expression vector (Fig. 7). Control constructs containing the luciferase reporter gene alone, did not result in any significant increase in luciferase activity. These data provide further support that GATA-3 is involved in positively regulating the expression of the NKG2A gene.
Involvement of GATA-3 in the expression of endogenous NKG2A
To determine whether GATA-3 protein is required for expression of endogenous NKG2A, we used siRNA to inhibit the expression of endogenous GATA-3 in NKL cells. GATA-3 expression was significantly decreased in the GATA-3 siRNA-treated cells (Fig. 8,A). As shown by Fig. 8 B, these GATA-3 siRNA-treated cells had significantly reduced levels of endogenous NKG2A on the cell surface compared with cells treated with control siRNA. These results indicate that GATA-3 is a biologically relevant regulator of endogenous NKG2A expression.
CD94/NKG2A is an inhibitory receptor that regulates NK cell activation (2, 7), and in certain instances, CD8+ T cell function (49). Very little is known about what regulates NKG2A clonal expression on NK cells and induction of expression on CD8+ T cells. For the human NKG2A gene, multiple transcriptional start sites were previously described (Fig. 2 B) (32, 33). Using 5′ RACE and primer extension analyses, it was shown that there are transcription start sites located upstream from exon 2 and multiple transcription start sites upstream from exon 1 of the NKG2A gene (32, 33). All transcripts encode identical proteins, but the larger transcripts have an untranslated exon 1. A putative TATA box and CCAAT box-like element, were found upstream of the ATG codon in exon 2 (32), but neither a canonical TATA box nor a CCAAT box are present upstream of exon 1 (32, 33). The genetic mechanisms regulating expression of these transcripts have not been previously described.
Enhancer and promoter elements are frequently contained within regions of open chromatin structure that are hypersensitive to DNase I digestion (38, 39, 40). Using the NKG2A expressing NKL cell line (Fig. 1), we located a DHSI in the region with multiple transcriptional start sites upstream of untranslated exon 1 of the NKG2A gene (Fig. 2,A). VISTA (Visualization Tools for Alignment) analysis of sequence conservation between the human and mouse NKG2A gene revealed that the hypersensitivity site DHSI resides in a highly conserved region upstream of exon 1 (Fig. 9, red region). The conservation of this sequence across species and its correspondence to a region of open chromatin, along with its proximity to multiple transcription start sites, suggested that this region contained the NKG2A promoter. The fact that a promoter construct containing deletion of the intron 1 gene segment did not decrease promoter activity in YT-Indy cells (Fig. 4) indicates that the region upstream of exon 2 does not play a significant role in the regulation of NKG2A gene expression, which is also supported by the failure to find DHSI (Fig. 2) and significant interspecies homology in this region (Fig. 9).
Using deletion mutant constructs, we localized the basal promoter to a region upstream of untranslated exon 1 within 80 bp of the most upstream transcriptional initiation site (Fig. 4), and within this region, we identified a 40 bp sequence between −2335 and −2295 bp that is important for promoter activity and contains key regulatory elements responsible for cell type-specific expression of the NKG2A gene (Fig. 4). A search of the TFSearch database revealed the presence of two consensus binding sequences for GATA binding transcription factors within this 40 bp sequence (Fig. 5,A). Using site specific mutation analyses (Fig. 5,B), we showed that, GATA binding site II is involved in the transcriptional regulation of the NKG2A gene. This GATA binding site II is conserved in primates and canines (University of California, Santa Cruz, genome browser 〈http://genome.ucsc.edu/Human〉 May 2004 (hg17) assembly), but despite the overall high sequence homology in this region between the human and murine genomes (Fig. 9), this GATA binding site is not conserved in mice, however there are other GATA-like binding sites present within this region. Whether these sites indeed play a role in the activity of mouse NKG2A promoter is a matter for future study.
Transcription factors of the GATA family are zinc finger proteins, that recognize the consensus motif (A/T)GATA(A/G) through a conserved multifunctional DNA-binding domain (41, 50). The functional specificity of a particular GATA is achieved through cell type-specific expression. The GATA-3 transcriptional factor is essential for early T cell development and also for Th2 cell differentiation following T cell activation (47, 48, 51, 52). Several studies have shown that GATA-3 is a critical transcriptional factor for regulating transcription of cytokines like IL-4, IL-5, and IL-13 in Th2 cells (53, 54, 55, 56, 57, 58). In NK cells, GATA-3 is expressed in immature NK precursors, as well as mature splenic CD3− NK1.1+ NK cells (59, 60). GATA-3 appears to play a global role in NK cell differentiation and function e.g., in promoting NK cell maturation, homing to the liver and IFN-γ production (61). Moreover, GATA-3 was also shown to be required for a protective NK cell response against L. monocytogenes (61).
We have demonstrated the capacity of GATA-3 to bind and transactivate the NKG2A promoter. Using pull-down assays, we showed that GATA-3 in NKL and freshly isolated NK cells can specifically bind GATA binding site II (−2302) in vitro (Fig. 6,A). Gene silencing of GATA-3 expression with siRNA reduced surface expression of endogenous NKG2A (Fig. 8,B). Using ChIP assays, we showed that GATA-3 specifically binds the NKG2A promoter in NKL and primary NK cells, but not in Jurkat cells (Fig. 6,B) in situ. This could be explained by the possibility that a multiprotein complex is formed between GATA-3 protein and other transcription factors in the region around −2302 bp in NKG2A expressing cells and that Jurkat cells may lack transcription factor(s) necessary to form this complex. This interpretation agrees with the fact that NKG2A promoter region constructs did not show significant promoter activity in NKG2A negative, GATA-3 positive Jurkat T cells (Figs. 3,B and 4 C). Another possibility is that an inhibitory transcription factor(s) binds to this GATA-3 binding II site in Jurkat T cells preventing the binding of GATA-3 protein to this region. Several such transcription factors that block the ability of GATA-3 to bind to its target sequences have been described (62, 63, 64).
We showed that human GATA-3 has the capacity to transactivate the NKG2A promoter in GATA-3 negative K562 cells. Cotransfection of the reporter construct containing the NKG2A basal promoter with a human GATA-3 expression construct led to a significant, dose-dependent increase in promoter activity. These results support an important role for GATA-3 in NKG2A promoter function and suggest that K562 cell line may contain other transcriptional factors that are necessary for GATA-3 transactivation of the NKG2A promoter. This result also indicates that GATA-3 protein has the potential to stimulate NKG2A promoter activity in cell lines and tissues in which GATA-3 is not normally expressed. For reasons yet to be determined, certain NK cells do not express NKG2A mRNA (65, 66). As GATA-3 appears to be integral to NK cell function (61), a failure to express GATA-3 most likely does not account for NKG2A negative NK cells. More likely other epigenetic events, such as chromatin modification or methylation status, determine whether or not a given NK cell expresses NKG2A.
An open chromatin structure is characterized by increased accessibility to DNase I, restriction enzymes, and transcription factors, as well as by hyperacetylation of histones, particularly H3 and H4. Several recent publications show that GATA-3 can initiate chromatin remodeling. For example, GATA-3 has been shown to be capable of inducing chromatin remodeling of an IL-4/IL-13 intergenic regulatory region that might be involved in the regulation of the IL-4 and IL-13 genes (67). Based on this, it will be interesting to determine whether the decondensation of chromatin required for cell-specific expression of the NKG2A gene is, in part, regulated via the action of GATA-3.
In summary, we have identified the location of the basal promoter of the human NKG2A gene and have shown that it contains a GATA binding site important for its promoter activity. We also have shown that GATA-3 is capable of transactivating NKG2A promoter activity and that it is involved in regulating the expression of endogenous NKG2A. Taken together, our data indicate that GATA-3 is an important transcription factor for regulating NKG2A gene expression in NK cells.
We thank Robert Valas for technical assistance, and Drs. Kerima Maasho, Madhan Masilamani, and Steven Burgess for helpful discussion.
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.
Abbreviations used in this paper: DHSI, DNase I hypersensitive site; ChIP, chromatin immunoprecipitation; siRNA, small interfering RNA.