Identification and Characterization of the CD226 Gene Promoter*

CD226 is one of the main activating receptors on natural killer cells, and it can induce cytotoxicity to target cells through interaction with its ligands CD155 or CD112. CD226 is also involved in T cell differentiation, activation, and cytotoxicity. The expression of CD226 on natural killer cells and T cells can be regulated by cytokines and chemical stimuli; however, the mechanism of the regulation of the CD226 gene is still unknown. In this study, we have identified two promoters in the human CD226 gene named P1 and P2, which are located at –810 to –287 bp and +33 to +213 bp, respectively, and a negative regulation element between P1 and P2. Both P1 and P2 can be regulated by phorbol ester (12-O-tetradecanoylphorbol-13-acetate) and calcium ionophore (A23187). Bioinformatics analysis shows that, within this CD226 gene region, there are putative binding sites for transcription factors AP-1, Sp1, PEA3, and Ets-1. We have found that transcription factor activating protein-1 (AP-1) can up-regulate CD226 promoters P1 and P2 in human hepatocarcinoma cells, a hepatocarcinoma cell line with low expression of endogenous AP-1 and Ets-1. Interestingly, the transcription factor Ets-1 promotes AP-1-induced P2 activity but inhibits AP-1-induced P1 activity for which a 10-bp AP-1/Ets-1 composite site (CCTTCCTTCC) in P1 may be responsible.

greatly stimulated by treatment of the cells with phorbol ester (TPA) (2). IL-2 and TNF-␣ augment CD226 expression and cytotoxic function of effector cells generated from mixed lymphocyte culture, whereas transforming growth factor-␤ (TGF-␤) could inhibit both of these events. TGF-␤ also prevents the IL-2-induced up-regulation of CD226 expression (12). In 2003, the poliovirus receptor/CD155 and nectin-2/ CD112 were identified as cell surface ligands for human CD226 (13). Soon thereafter, the interaction of CD226 with its ligands was confirmed as one of the major mechanisms that triggers NK cell activation and cytotoxicity against some tumor and leukemic cells (14 -16). Accumulating evidence indicates that CD226 is involved in a variety of immunological functions, including T cell differentiation and cytotoxicity, NK cell cytotoxicity, natural killer T cells apoptosis, megakaryocyte polyploidization, platelet activation and aggregation, monocyte extravasation, dendritic cell maturation, and platelet and megakaryocytic cell adhesion to vascular endothelial cells (1)(2)(3)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15). It has been found that LFA-1 (CD11a/CD18), the ␣L␤2 integrin, physically associates with CD226 in NK cells and anti-CD3 monoclonal antibody-stimulated T cells and that crosslinking of LFA-1 induces tyrosine phosphorylation of CD226, for which the Fyn protein tyrosine kinase is responsible, indicating that CD226 is involved in the LFA-1-mediated intracellular signals (5,16). CD226 on T cells forms a dynamic molecular complex with protein 4.1G and human Large discs, which may serve to cluster and transport LFA-1 and associated molecules (4). Protein kinase C phosphorylates Ser 329 of CD226, which plays a critical role in both CD226 adhesion and signaling as well as for lipid raft recruitment of CD226 and LFA-1-mediated signaling (17,18). LFA-1, CD226, and Fyn are polarized at the immunological synapse upon stimulation with anti-CD3 in T cells (5). Although progress in CD226 molecule distribution, function, and its relationship to some diseases has been made in recent years, the mechanism of CD226 gene regulation is still unknown.
Activating protein-1 (AP-1) is a collective term referring to the dimeric transcription factors of Jun and Fos or activating transcription factor subunits that bind to a common DNA site, the AP-l-binding site. The activity of AP-1 could be modulated by treatment of cells with TPA, known to stimulate protein kinase C (19,20). AP-1 mediates gene regulation response to a variety of extracellular stimuli, including growth factors, cytokines, tumor promoters, and chemical carcinogens (20,21). When regulating gene expression, AP-1 could cooperate with other transcription factors such as Ets-1. Ets is a family of transcription factors that shares a highly conserved Ets domain, a winged helix-turn-helix DNA binding domain that could bind different sites containing a core (A/C)GGA(A/T) motif existing in the promoter region of many genes, including matrix metalloproteinases, uPA, and TIMP-1 (22)(23)(24). Ets-1 is composed of Ets domain, transactivation domain, and pointed domain and can cooperate with the c-Fos⅐c-Jun complex at AP-1 sites to regulate certain promoters (25). In this study, we have identified the promoters of human CD226 gene for the first time and investigated the regulation of CD226 promoters by AP-1 and Ets-1 transcription factors.
Cell Culture and RNA Preparation-Jurkat cells were maintained in RPMI 1640 medium (HyClone) supplemented with 10% fetal calf serum (HyClone), and HHCC cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum at 37°C in 5% CO 2 . Total RNA was extracted from TPA-treated Jurkat cells using TRIzol according to the manufacturer's instructions (Invitrogen).
Plasmid Construction-A 2-kb upstream regulation region of the CD226 gene was amplified by PCR from genomic DNA from human peripheral blood mononuclear cells and cloned into pGL3-basic vector (Promega) by restriction endonucleases HindIII and KpnI. A series of different truncated fragments of the regulation region of the CD226 gene were created by PCR amplification using the primers listed in Table 1 and inserted into pGL3-basic vector by the same restriction endonucleases. Expression vectors of c-Jun, c-Fos and Ets-1 were kindly provided by Prof. D. K. Watson (Medical University of South Carolina).
Transient Cell Transfection-For promoter activity analysis, promoterless pGL3-basic vector was used as a negative control, and plasmid phRL-TK (Promega) was cotransfected as an internal control. Plasmid DNA was prepared using the Maxi-prep kit (Qiagen). Jurkat cells were seeded in a 96-well plate with 1 ϫ 10 5 cells/well, and triple wells were set for each group. 0.1 g of pGL3 vector containing different truncated DNA fragments and 5.0 ng of phRL-TK vector were cotransfected for each well using Lipofectamine 2000 (Invitrogen). For investigation of the regulation of the CD226 promoter by AP-1 and Ets-1, 1 g of total DNA was transfected into HHCC cells in a 24-well plate using Lipofectamine 2000. 48 h after transfection, the cell lysates were prepared for the measurement of luciferase activity. In some groups, Jurkat cells were treated with 30 ng/ml TPA for 4 h, 0.5 g/ml A23187 for 1 h, 100 units/ml IL-2 for 24 h, or 500 units/ml of TNF-␣ for 24 h before luciferase assay.
Luciferase Assays-Cells were rinsed in phosphate-buffered saline at 48 h after transfection and lysed in a passive lysis buffer (Promega). Luciferase activities were measured using the Dual-Luciferase reporter assay system (Promega) with a Turner luminometer and normalized for transfection efficiency to the Renilla luciferase activity. Reported data were represented as the mean from three independent experiments.
Western Blot-Jurkat cells or HHCC cells were pelleted and then lysed by 2ϫ loading buffer. After SDS-PAGE resolution and membrane transfer, the target proteins were probed with rabbit antibody against human c-Jun, Ets-1, or ␤-actin (Santa Cruz Biotechnology). Thereafter, horseradish peroxidaselabeled goat anti-rabbit immunoglobulin was added, and the proteins were detected by horseradish peroxidase substrate (Pierce).
Electrophoretic Mobility Gel Shift Assay-Nuclear extracts were prepared from TPA-treated Jurkat cells according to the method of Dignam (26) with minor modification. Briefly, 3 ϫ 10 7 cells were washed twice with ice-cold phosphatebuffered saline and resuspended in buffer A (20 mM HEPES, pH 7.9, 1.5 mM MgCl 2 , 10 mM KCl, 0.1 mM EDTA, 0.5 mM dithiothreitol, and 0.5 mM phenylmethylsulfonyl fluoride) and left on ice for 15 min. After the lysates were passed several times through a 25-gauge needle, nuclei were recovered by centrifugation at 13,000 rpm for 10 min at 4°C. Nuclei were resuspended in buffer B (20 mM HEPES, pH 7.9, 1.5 mM MgCl 2 , 420 mM NaCl, 0.2 mM EDTA, 20% glycerol, and 1 mM dithiothreitol) and incubated on ice for 30 min. The mixture was centrifuged at 13,000 rpm at 4°C for 15 min, and the supernatant as nuclear extract was collected and stored at Ϫ70°C until use. Protein concentrations of nuclear extract were determined by Bradford assay. For preparation of probes, ␥-32 P was labeled to the end of AP-1 oligo consensus using T4 polynucleotide kinase (Promega), whereas ␣-32 P was labeled to P1 and P2 by PCR amplifica-  amplification of different truncated  DNA fragments T1, T2, T3, T4, T5, T6, and T7, respectively. They are P1 and P8 for  the production of T1, P1 and P2 for T2, P3 and P8 for T3, P3 and P4 for T4, P5 and  P8 for T5, P3 and P6 for T6, and P7 and P8 for T7.

Primers
Sequences Location tion. In some groups, 2.0 g of nuclear extract was pretreated with AP-1 oligo consensus (5Ј-CGCTTGAT-GAGTCAGCCGGAA-3Ј), or Sp1 oligo consensus (5Ј-ATTC-GATCGGGGCGGGGCGAGC-3Ј) for 10 min in binding buffer provided in the electrophoretic mobility gel shift assay kit (Promega). The 32 P-labeled probes were added into nonpretreated or pretreated nuclear extract and reacted for 20 min. The reactions were electrophoresed in 6% non-dena-tured polyacrylamide gel in 0.5ϫ Tris borate-EDTA buffer at 350 V for 15 min. The gel was dried and exposed to a Biomax-MS film (Eastman Kodak) at Ϫ70°C for ϳ10 h. After that, the films were developed, and the binding of transcription factors with promoters or oligo consensus was analyzed according to the density and area of each blot on the developed film.

RESULTS
Defining the Transcription Start Site of Human CD226 Gene-The 5Ј-RACE was carried out to define the transcription start site of human CD226 gene. The mRNA from TPA-treated Jurkat cells was reverse transcribed using gene-specific primers. One DNA band of ϳ500 bp was obtained after nested PCR and then cloned into pMD18-T vector for DNA sequencing. The results from three randomly selected clones revealed that the transcription start site was located at Ϫ229 bp upstream of ATG in mature mRNA of CD226, which was identical to the CD226 genomic sequence (Fig. 1).
Bioinformatics Analysis of the CD226 Promoter Region-A 2-kb DNA sequence of the CD226 upstream regulation region was obtained from GenBank TM , which was used for bioinformatics analysis. There are two TATA-box, located at Ϫ85 to Ϫ80 bp and Ϫ196 to Ϫ191 bp, respectively; one GC-box is located at Ϫ249 to Ϫ243 bp and several putative binding sites of transcription factors, such as AP-1, Sp1, Ets-1, PEA3, and GATA-1 (Fig. 1).
Location and Regulation of the CD226 Promoter-A series of truncated DNA fragments were inserted into pGL3 vectors and transfected into Jurkat cells, and the promoter activity of each truncated fragment was measured by luciferase assay. The results show that truncated DNA fragment 1 (T1) has basic promoter activity, whereas T4 and T7, located at Ϫ810 to Ϫ287 bp and ϩ33 to ϩ213 bp, respectively, have relatively higher promoter activities; these regions are designated as P1 and P2 (Fig. 2). In Jurkat cells, TPA could increase P1 activity by 19.3% but slightly inhibit P2 activity, whereas A23187 could increase both P1 and P2 activities. When treated with TPA and A23187 together, the Jurkat cells showed inhibitory effects for all three fragments (T1, T4/P1, and T7/P2) compared with the groups of TPA-or A23187-treated T4/P1 or A23187-treated T7/P2 (Fig. 3A). This interesting phenomenon is consistent with previous described results (2). The CD226 promoters could also be regulated by IL-2 and TNF-␣. Thus after stimulation for 24h by IL-2 or TNF-␣, the transfected Jurkat cells were lysed for measurement of luciferase activities. The results showed that IL-2 increased the activities of T1,   SEPTEMBER 29, 2006 • VOLUME 281 • NUMBER 39
Tissue-specific Expression of CD226 Promoters-CD226 is mainly expressed in hematopoietic cells, including T cells, NK cells, monocytes, megakaryocytes, and some populations of B cells (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11). Therefore, we examined the expression of CD226 promoters in different cell types to investigate whether the promoters contain tissue-specific elements. The results showed that P1 and P2 have relatively high activities in Jurkat and Daudi cells, which express CD226 and low activities in HL60 and K562 cells but which do not express CD226. Moderate activities of P1 and P2 were found in Dami cells and NK92 cells. Interestingly, P1 and P2 showed relatively different activities among the cell lines described above. P1 had higher activity in Jurkat cells, whereas P2 showed higher activity in Daudi cells, although the difference of activities for P1 and P2 in NK92 and Dami cells was not significant (Fig. 4).
Identification of the NRE in the CD226 Upstream Regulation Region-Based on the analysis of luciferase activities using T4 -T7, we found a NRE, which was located between the two promoters P1 and P2. Both T5-pGL3 and T6-pGL3 containing a DNA fragment Ϫ286 to ϩ32 bp had much lower promoter activities compared with T4-pGL3 and T7-pGL3 (Fig. 2), indicating the fragment Ϫ286 to ϩ32 bp may be a NRE. To further test the inhibitory effect, the NRE was introduced to the downstream of cytomegalovirus (CMV) promoter and transfected into Jurkat cells. The luciferase assay showed that the NRE could inhibit CMV promoter activity with or without TPA treatment, and the inhibition rates reached 85.1 and 73.7%, respectively (Fig. 5).
Regulation of CD226 Promoters by Transcription Factors AP-1 and Ets-1-AP-1 is the main transcription factor activated by TPA, IL-2, or TNF-␣. Ets-1 is a transcription factor involved in many lymphocyte functions and can also associate with AP-1 to regulate target genes. Bioinformatics analysis indicated that there were several putative binding sites of AP-1 and Ets-1. To investigate whether the two transcription factors could regulate CD226 promoters, HHCC cells were employed because of their low expression of endogenous AP-1 and Ets-1 and high efficiency for transient transfection (Fig. 6A). We co-transfected the expression vectors containing these transcription factors with P1-pGL3 or P2-pGL3 into HHCC cells and found that c-Jun, c-Fos, or Ets-1 alone did not regulate P1 and P2. In contrast, the c-Jun/c-Fos heterodimer (AP-1) increased P1 and P2 activities to 56.7-and 12.7fold, respectively. When we co-transfected the AP-1 and Ets-1 expression vectors and P1-pGL3 or P2-pGL3 into the HHCC cells, we found that Ets-1 inhibited AP-1-induced P1 activity but increased AP-1-induced P2 activity (Fig. 6B). Bioinformatics analysis showed that there was an AP-1/Ets-1 composite binding site in P1 (Fig. 1), suggesting that Ets-1 may inhibit AP-1-induced P1  CD226 promoters by stimuli (TPA1, A23187) and cytokines. A, T1-pGL3, T4/P1-pGL3, and T7/P2-pGL3 were transfected into Jurkat cells and cultured for 48 h before cell lysis. In some groups, transfected Jurkat cells were treated with TPA for 4 h or A23187 for 1 h before luciferase assay. B, T1-pGL3, T4/P1-pGL3, and T7/P2-pGL3 were transfected into Jurkat cells and cultured for 24 h. Thereafter, IL-2 (100 units/ml) or TNF-␣ (500 units/ml) was added in the culture system for an additional 24 h before luciferase assay. nil, no treatment.  activity by competing for binding at the site. Deletion of the composite site of AP-1/Ets-1 partially impaired AP-1-induced P1 activity, and Ets-1 could no longer inhibit AP-1-induced P1 activity (Fig. 6C). These results indicate that the composite binding site was responsible for AP-1-induced P1 activity, substantiating the notion that Ets-1 inhibits AP-1 function on P1 by competing for the binding site.
Binding of Transcription Factors with CD226 Promoters-AP-1 functionally regulates CD226 promoters. To investigate whether AP-1 could physically bind with the CD226 promoter, we carried out electrophoretic mobility gel shift assay analysis. In electrophoretic mobility gel shift assay, the specificities of AP-1 bands were confirmed by adding excess amounts of unlabeled specific and nonspecific AP-1 probes to the reaction mixtures. The results indicate that the nuclear extract from TPAstimulated Jurkat cells could bind with CD226 promoters P1 and P2. One main band was found when P1 was incubated with the nuclear extract. This band was partially inhibited by unlabeled AP-1 oligo consensus but almost completely inhibited by the Sp1 oligo consensus. Two strong bands were found when P2 interacted with the nuclear extract, and they were strongly inhibited by unlabeled AP-1 oligo consensus but only partially by unlabeled Sp1 oligo consensus (Fig.  7). These results confirmed that AP-1 and Sp1 could bind CD226 promoters P1 and P2.

DISCUSSION
The phenomenon that a single gene owns alternative promoters is very common and is often related to tissue-specific expression of that gene (27)(28)(29). In this study, we have shown that the human CD226 gene has at least two promoters, termed P1 and P2, and they have distinctive activity in some cell lines derived from different tissues. These results suggest that the two promoters may play a role in tissue-specific expression of human CD226. Previous studies have shown that CD226 mRNA and surface expression in Jurkat cells was up-regulated by treatment of the cells with TPA (2), and it is known that TPA can activate the transcription factors AP-1 and nuclear factor-B through the protein kinase C pathway (24). Here we have also shown that TPA can up-regulate CD226 promoters in Jurkat cells and that AP-1 can increase CD226 promoter activities. These results suggest that the mechanism of up-regulation of CD226 expression by TPA may be through the activation of AP-1. We have also found that ionophore/A23187 up-regulates CD226 expression and promoter activity in Jurkat cells. However, when Jurkat cells were treated with TPA, co-treatment with A23187 resulted in an inhibitory effect on CD226 promoter activities. This finding is consistent with the previous FIGURE 6. Regulation of CD226 promoters by AP-1 and Ets-1. A, the expression of AP-1 and Ets-1 in Jurkat and HHCC cells. Jurkat cells or HHCC cells were pelleted and then lysed by 2ϫ loading buffer. After SDS-PAGE and membrane transfer, the target proteins were probed with rabbit antibodies against human c-Jun, Ets-1, and ␤-actin. Thereafter, horseradish peroxidase-labeled goat anti-rabbit immunoglobulin was added, and the proteins were detected by horseradish peroxidase substrate. B, expression vectors of mock, c-Jun, c-Fos, Ets-1, AP-1, or AP-1 plus Ets-1 were transfected into HHCC cells. Luciferase activities were measured at 48 h after transfection. C, function of composite binding site of AP-1 and Ets-1 in P1. Wild-type P1-pGL3 vector (WT-P1) or AP-1 site-deleted P1-pGL3 vector (AP-1 ⌬ -P1) were transfected into HHCC cells, and luciferase activities were measured at 48 h after transfection. FIGURE 7. Binding of transcription factors to CD226 promoters. The 32 P-labeled probes were added into TPA-treated or non-pretreated nuclear extract from Jurkat cells and reacted for 20 min. The reactions were electrophoresed, and the gel was dried and exposed to a Biomax-MS film. After that, the films were developed, and the binding of transcription factors with promoters or oligo consensus were analyzed according to the density and area of each blot on the developed film. One predominant band could be observed when 32 P-labeled AP-1 probe was incubated with the nuclear extract (lane 2). The unlabeled AP-1 oligo consensus (but not the unlabeled Sp1 oligo consensus) competed the binding of 32 P-labeled AP-1 probe with AP-1 in the nuclear extract (lanes 3 and 4). Both P1 and P2 interacted with nuclear extract from TPA-treated Jurkat cells (lanes 6 and 10), and the binding was inhibited by unlabeled AP-1 and Sp1 oligo consensus at different levels (lanes 7, 8, 11, and 12). report on CD226 mRNA and protein expression (2), but the mechanism of this phenomenon is still unclear. It is well established that AP-1 and nuclear factor of activated T cells are the main downstream transcription factors of TPA and the A23187 signal pathway. It is possible that the interaction of AP-1 and nuclear factor of activated T cell may be involved in the inhibitory effect of A23187 on TPA-induced CD226 expression.
In the present study, we have shown that not only TPA could up-regulate CD226 promoter activities but also AP-1 up-regulated CD226 promoters through an AP-1/Ets-1 binding site. Thus, it is most likely that TPA functions by activating AP-1, and AP-1 in turn promotes CD226 transcription through the binding site. Ets-1 can interact with AP-1, which displayed different regulatory activities on the two promoters of the CD226 gene. In our investigation of the binding of AP-1 to CD226 promoters in vitro, we found that both P1 and P2 could bind with the nuclear extract of TPA-stimulated Jurkat cells, and the bands could be inhibited by unlabeled oligonucleotides of AP-1 and Sp1 to different extents. These results imply that the nuclear proteins that bind with CD226 promoters could be a protein complex, at least composed of AP-1 and Sp1, and the composition of the complex may influence the regulatory outcome.
It is very common for genes to contain negative regulatory elements within the promoter, and this is one of the major mechanisms of gene regulation. Regulations by NRE can occur by two mechanisms. The first is position-dependent or -independent, and the second is gene-specific or -nonspecific. Here we have identified a NRE that is positioned between the two promoters of the CD226 gene. Because the NRE inhibited CD226 promoters, whether downstream or upstream, it is likely that this NRE works in a position-independent manner. Moreover, the NRE of the CD226 gene can also effectively inhibit the CMV promoter, suggesting that this NRE works in a gene-nonspecific manner. The elucidation of the mechanism of the NRE inhibition will be helpful in further understanding CD226 gene regulation.