Identification of Distal KIR Promoters and Transcripts

ORIGINAL ARTICLE Identification of distal KIR promoters and transcripts

MJ Stulberg1, PW Wright2, H Dang1, RJ Hanson3, JS Miller3 and SK Anderson1,2 1Laboratory of Experimental Immunology, Center for Cancer Research, National Cancer Institute-Frederick, Frederick, MD, USA; 2Basic Research Program, SAIC-Frederick Inc., NCI-Frederick, Frederick, MD, USA and 3Division of Hematology, Oncology, and Transplantation, University of Minnesota Cancer Center, Minneapolis, MN, USA

A more complete understanding of the transcriptional control of the human and murine class I MHC receptors will help to shed light on the mechanism of selective, stochastic, gene activation that operates in these gene families. Studies of the murine Ly49 class I MHC receptor genes have revealed an important role for distal transcripts originating upstream of the proximal promoter. To date, there have been no reports of distal promoters within the functionally analogous human KIR family of class I MHC receptors. In the current study, reverse transcriptase-polymerase chain reaction (RT-PCR) and RNase protection assays were used to reveal the presence of distal KIR transcripts initiating upstream of the previously characterized proximal KIR promoter. The intergenic promoter elements detected were associated with repetitive elements of the Alu and L1 families. Unlike the proximal KIR promoter, the distal promoter regions were not NK cell-specific. KIR genes expressed in a variegated manner produced a low level of distal transcripts containing a large 5 0 untranslated region. In contrast, the highly expressed KIR2DL4 gene possessed a higher level of spliced distal transcripts that were capable of producing KIR2DL4 protein. The identification of distal KIR promoter elements suggests that intergenic transcripts may influence the expression of KIR genes. Genes and Immunity (2007) 8, 124–130. doi:10.1038/sj.gene.6364363; published online 7 December 2006

Keywords: human; NK cells; KIR; transcription

Introduction present in an individual genotype leads to a wide array of KIR repertoires in the human population.7 Stochastic Natural killer (NK) cells are large granular lymphocytes activation of KIR genes leads to distinct percentages of that have the ability to eliminate tumors and virus- NK cells that express a given KIR. The mechanism infected cells.1 NK cells sense the state of potential target governing selective expression of KIR genes is not cells through the integration of signals received from completely understood. activating receptors recognizing stress-related changes The current hypotheses explaining variegated expres- and the opposing signals generated by inhibitory sion of KIR genes are based on the methylation state of receptors that recognize major histocompatibility com- active versus silent KIR alleles. Three separate studies plex (MHC) class I molecules.2 The killer cell immuno- observed that non-expressed KIR alleles are methylated, globulin-like receptor (KIR) genes constitute a major while treatment with the demethylating agent 5-aza-2- family of MHC receptors expressed by human NK cells.3 deoxycytidine (5-aza) resulted in the de novo expression There are two functionally distinct types of KIR, the of KIR genes.8–10 How methyltransferases are recruited activating and inhibitory receptors. The inhibitory KIR and what governs selective methylation is unknown, and have been shown to recognize specific human leukocyte studies have focused on the KIR promoter elements. antigen (HLA) molecules, while the precise nature of the A promoter identified in the 50 region adjacent to the ligands for the activating KIR is still being explored.4 translational start site has been characterized for many KIR expression is variegated, enabling NK cells to KIR genes, and the possible association of distinct distinguish subtle changes in HLA expression.5–7 Var- promoter properties with the mechanism of variegated iegated expression is beneficial, since virus infection or expression has been discussed.11–15 The KIR2DL4 gene is transformation might only downregulate one HLA allele, unique in that it is expressed by all NK cells, and given and the presence of NK cells that express a subset of this property, the KIR2DL4 promoter has been analyzed inhibitory HLA receptors allows the detection of cells by in vitro promoter assays and DNA footprinting in that have altered expression of a single type of HLA order to compare it with the promoters of the variegated molecule. Variation in the number and type of KIR genes KIR genes.11 The KIR2DL4 gene is also unique since it has a 14 kb intergenic region upstream from the translational start site, in contrast to the approximately 2 kb that Correspondence: Dr SK Anderson, Lab of Experimental Immuno- separates the other KIR genes.16 In addition, KIR2DL4 logy, Center for Cancer Research, NCI-Frederick, Building 560, expression is restricted to NK cells, and in vitro promoter room 31-93, Frederick, MD 21702-1201, USA. E-mail: [email protected] assays did not detect appreciable activity of the KIR2DL4 11 Received 28 September 2006; revised 6 November 2006; accepted 7 promoter in the Jurkat T-cell line. Comparison of November 2006; published online 7 December 2006 KIR2DL4 with KIR3DL1 revealed differences in predicted Distal KIR promoters MJ Stulberg et al 125 transcription factor-binding sites and promoter activity. a 2DL4 3DL1 All of the expressed KIR genes possess a binding site for pA ATG the AML family of transcription factors that is located L1M5 AluSx 1 2 3 approximately 98 bp upstream from the ATG start site, 8 4 and the AML-2 (Runx3) protein has been shown to bind +10 -911 -835 -756 -627 -255 +512 -1637 -2065 -1424 -1130 12,13,15 -1749 this element. The potential importance of the AML (cDNA) 17 site has been revealed by studies of KIR2DL5 alleles. b Two non-expressed KIR2DL5 gene variants, KIR2DL5.2 1347 and –5.4, have a single base pair change that disrupts the 1268 AML site, while the expressed –5.1 and –5.3 variants 1139 have this site intact. Additional in vitro studies have 767 shown that disruption of the AML-binding site in several KIR genes significantly reduced, but did not abolish 502 promoter activity, suggesting that AML might play a role M 13–15 +10 -756 in the opening of the locus in vivo. -255 -627 -835 -911 -1130 -1424 -1749 Several studies have examined the KIR intergenic -1637 -2065 region in order to identify elements regulating promoter Figure 1 RT-PCR detection of KIR intergenic transcripts. (a)A activity.11,12,14 When the full 2 kb intergenic region was schematic of the KIR2DL4-KIR3DL1 genomic region is shown, with exons indicated by numbered boxes. The boundaries of the tested, no additional activity was observed relative to the intergenic region are marked by the KIR2DL4 polyA-addition site smallest active fragment contained within 267 bp of the (pA), and the KIR3DL1 start codon (ATG). The positions of the first ATG start site.11 As an additional upstream bi-directional nucleotide of the PCR primers relative to the KIR3DL1 start codon switch promoter has been associated with the stochastic are indicated by the small numbered arrows. The position of the expression of the murine Ly49 class I MHC receptors,18 exon 4 reverse primer corresponds to its location in a fully spliced we conducted a detailed investigation of KIR RNAs KIR3DL1 mRNA ( þ 512). The position and orientation of repetitive elements are indicated by the large reverse arrows above. (b) PCR of initiated upstream of the known promoter, and the KIR NK cell RNA with intergenic primers. The products resulting from intergenic regions were analyzed for additional promoter PCR with various forward intergenic primers and the exon 4 elements. reverse primer are shown. Lanes are labeled with the position of the 5’ primer used. The molecular weight standard (M) is the 1 kb plus DNA ladder (Invitrogen). The predicted nucleotide length of spliced KIR3DL1 mRNAs containing 50 intergenic sequence are Results indicated by the arrows. Detection of KIR intergenic transcripts The possible presence of intergenic transcripts in the KIR L1M5 AluSx gene cluster was investigated by performing reverse transcriptase-polymerase chain reaction (RT-PCR) on USF AML Myc Elk-1 RNA from purified human NK cells with a series of NF-E2 50 primers contained within the KIR2DL4-KIR3DL1 -1347 -819 intergenic region together with a 30 primer in exon 4 of the KIR3DL1 gene (Figure 1). Primers starting at 255, 627, -1347 -887 756 and 835 bp upstream of the KIR3DL1 start codon were capable of generating spliced cDNAs containing -1258 -819 exons 1–4 preceded by continuous, unspliced 50 intergenic sequence. Primers located further upstream were -1097 -819 unable to produce the predicted products. All of the PCR -911 -819 products were cloned and sequenced, and the predicted 0 spliced KIR3DL1 RNAs containing 5 intergenic sequence -1347 -846 were identified. These results demonstrate the presence of RNA transcripts originating from upstream elements, 01234 and suggest that transcription from an upstream pro- Relative Promoter Activity moter element initiates between 835 and 911 bp upstream of the KIR3DL1 start codon. Figure 2 Localization of elements required for KIR3DL1 distal promoter activity. The À1347 to À819 fragment is shown with the positions of putative transcription factor-binding sites shown as Characterization of the KIR3DL1 distal promoter element open boxes. The relative position and orientation of the L1 and Alu As RT-PCR experiments indicated the possible presence repeats are indicated by the lines above. Values represent the mean of an additional KIR3DL1 promoter in the intergenic and error bars indicate the SEM of at least three independent region containing flanking AluSx and L1M5 repetitive experiments. elements (Figure 1), a series of fragments spanning these repetitive elements was generated by PCR and cloned abolished promoter activity, and the removal of the 50 into pGL3. Constructs were transfected into the YT-Indy Myc/AML and USF transcription factor-binding sites cell line and relative promoter activity was measured. As contained within the L1M5 element decreased activity by shown in Figure 2, weak promoter activity was found in approximately 50% (fragment À1097 to À819). the KIR3DL1 upstream region associated with fragments that contained both of the adjacent L1M5 and AluSx Direct measurement of KIR transcriptional start sites repetitive elements. Removal of the 30 AluSx element In order to determine if a significant fraction of KIR (fragment À1347 to À887) containing Ets and NF-E2 sites transcripts originated upstream of the previously char-

Genes and Immunity Distal KIR promoters MJ Stulberg et al 126 acterized promoters, a ribonuclease protection assay transcripts.11,14,17,19 However, the RPA analysis indicates (RPA) was performed on RNA isolated from peripheral that a preferred transcription start site is located 58 blood NK cells and the NK92 cell line using antisense nucleotides upstream of the KIR3DL1 start codon. KIR3DL1, KIR2DL2 or KIR2DL4 probes (Figure 3). The Interestingly, a TATAAA-like element (TGTAAA) is KIR2DL2 probe was used for the detection of KIR located 25 nucleotides upstream of the major start site. transcripts in ex vivo NK cells, since we found that the A nearly identical pattern of KIR start sites was observed level of transcripts detected with this probe were in NK92 and sorted KIR-positive NK cells with probes significantly higher than KIR3DL1 transcripts, and there- from the KIR2DL1, KIR2DL2 and KIR3DL1 50 regions fore allowed detection of the low level of transcripts (Figure 3a and data not shown). Additional protected present in fresh NK cells. In order to study KIR RNA fragments corresponding to transcripts initiated transcripts in NK92 cells, it was necessary to induce further upstream were detected at a much lower level. demethylation with 5-aza-2-deoxycytidine (5-aza), since The full KIR3DL1 probe sequence was protected as well the KIR2DL4 gene is the only family member expressed as a slightly shorter RNA corresponding to a spliced by this cell line, and treatment with 5-aza is required to RNA lacking the first five nucleotides of intron 1 induce expression of the other KIR genes.8–10 As shown contained in the probe. The full-length protected probe in Figure 3a, KIR3DL1 and KIR2DL2 proximal transcripts potentially represents unspliced KIR3DL1 RNA; how- were only detected in NK92 cell RNA derived from ever, there may be some contribution of contaminating 5-aza-treated cells. Numerous start sites of proximal genomic DNA to this band even though the RNA transcripts were detected, consistent with previous samples were treated with DNase before RPA analysis. studies performed by conducting 5’-RACE on KIR The RPA analysis of KIR2DL4 transcripts is shown in

a Probe: KIR3DL1 KIR2DL2 b KIR2DL4 Probe Probe Yeast Yeast NK92 5-aza NK92 Probe 5-aza Yeast NK92 KIR+ 5-aza

300

Distal Distal 200 Unspliced Spliced (210) 200 200 Distal (192)

100 Proximal 90

100 80

Proximal 100 70 (92) Proximal (62) 60

agtgagtctg -249 agatgtgttttgagcctggccgttg -224 cgcatgatgtgaagtgacaagtcta -199 tgttccgctcttgagcgag cacccactgggcctcatgca -137 gtctgcagTtttcagaaaccctcat -174 aggtagaaagagcctgcgta cgtcaccctcccatgatgtg -97 tcctcccttgactgattcaccactt -149 gtcaacatgtaaactgcatg ggcagggcgccaaatAacAt-57 gaacctcatatgacgtagaagaagc -124 cctGtgcgctgctgagctga gctggggcgcggccgcctgt -17 ctacctatgtccccttcacatgttg -99 ctgcacagacagcaccATGt cgctcatggtcgtcagcatg +24 tggtcaatgtgtcaactgcacgatc -74 gcgtgtgttg GTGAGT splice donor-intron1 +40 cgggcccctcaccAcatcctctgca -49 ccggtcagtcgagccgagtcactgc -24 KIR3DL1 Probe gtcctggcagcagaagctgcaccAT +2 KIR2DL4 Probe Figure 3 Analyses of KIR transcription start sites by RPA. (a) RPA with antisense KIR3DL1 or KIR2DL2 probes. Undigested probe (Probe), 20 mg of total RNA from yeast control (Yeast), NK92, NK92 treated with 1 mM 5-aza for 72 h (5-aza) or KIR-expressing peripheral blood NK cells (KIR þ ) was analyzed and run on an 8% sequencing gel. Arrows indicate the positions of proximal and distal transcripts. The sequence of the KIR promoter region complementary to the KIR3DL1 probe sequence is shown below. Nucleotide numbers relative to the ATG start codon are shown to the right of each line. A TATAAA-related element is boxed. The three major transcription start sites observed are underlined single nucleotides in bold uppercase. The intron 1 splice donor sequence is also underlined. (b) RPA with an antisense KIR2DL4 probe. The positions of protected distal and proximal RNA transcripts are indicated by arrows. The sequence complementary to the probe is indicated below, and nucleotide numbers relative to the ATG start codon are shown to the right of each line. The predicted splice acceptor is underlined, and the nucleotides corresponding to the 50 ends of the protected fragments are shown in bold uppercase.

Genes and Immunity Distal KIR promoters MJ Stulberg et al 127 Figure 3b. A protected RNA initiated at a site similar to however a similar level of activity was seen with a the major KIR3DL1 proximal start site was detected. An smaller fragment (10 713–10 483) that contains predicted additional protected RNA was detected that indicated a Irf-1 and Ets-binding sites. Fragments containing the transcript beginning 190 nucleotides upstream of the region from 11143 to 10 925 and from 10 925 to 10 713 bp ATG start codon. Examination of the sequence adjacent did not demonstrate any promoter activity, indicating to the 50 end of the protected RNA indicated that it might that the core promoter is located in the 10 713–10 483 represent a spliced KIR2DL4 transcript originating from region. As a previous report has demonstrated IL-2- an upstream promoter, since a consensus splice acceptor regulation of KIR2DL4 expression,21 we tested the effect sequence was present immediately upstream. The region of IL-2 on transcription from the KIR2DL4 distal upstream of the previously characterized KIR2DL4 promoter. Treatment of transfected cells with IL-2 promoter was used to search the GenBank human EST (1000 U/ml) had no effect on distal promoter activity database, and a sequence was identified that represented (data not shown). an additional KIR2DL4 transcript with a start site approximately 10 kb upstream of the KIR2DL4 start Promoter activity is associated with repetitive elements in codon and spliced to the site at À190 nucleotides several KIR genes indicated by the RPA assay (GenBank #DA590865). The Due to the high homology of the KIR intergenic regions, presence of this transcript in the NK92 cell line and similar repetitive elements are present in all members of peripheral blood NK cells was confirmed by performing the KIR gene family. Figure 5 shows the promoter PCR with a primer corresponding to the 50 end of the activity detected in fragments containing repetitive Genbank sequence together with an antisense primer elements present in a similar location in the intergenic containing the KIR2DL4 stop codon. region preceding the KIR2DL4, KIR2DL5 and KIR3DL3 genes. Fragments isolated from the homologous region Identification of a distal KIR2DL4 promoter preceding the KIR2DL5 gene possessed very weak The region flanking the 50 end of the novel KIR2DL4 promoter activity, possibly due to the lack of a 30 Ets- transcript start site was tested for promoter activity in binding site present in the KIR3DL1 distal promoter luciferase reporter assays (Figure 4). The 30 ends of the region. Fragments generated from a similar location in promoter fragments tested were located 50 bp down- the KIR2DL4 gene possessed promoter activity that was stream from the predicted transcription start site. The two- to four-fold greater than the KIR3DL1 activity start site of the distal promoter was predicted by the 50 detected in YT-Indy and HEK293T cell lines. In order to end of a human EST clone (GenBank #DA590865) that determine if transcripts from this additional distal was derived using the oligo-capping method which KIR2DL4 promoter (distal-2) could also be spliced to preferentially clones cDNAs containing the 50 end of the proximal region, PCR was performed with a primer mRNAs.20 The largest fragment analyzed (11143 to near the 30 end of the active promoter fragment. A distal 10 483 bp upstream of the KIR2DL4 start codon) contained putative Yy1 and AML-binding sites. This construct exhibited the highest level of promoter activity; 2DL4 distal-1 (11143-10485)

L1MEc 2DL4 distal-2 Yy1 Yy1 Irf-1 AML AML Ets Ets (1373-932) 10,483

2DL5 distal HEK293T 11,143 Jurkat (1354-824) YT-Indy

10,925 3DL1 distal (1347-819)

10,713 3DL3 distal 10,638 (909-1462)

0 5 10 15 20 25 Relative Promoter Activity 0 10203040506070 Relative Promoter Activity Figure 4 Identification of an upstream KIR2DL4 promoter. Reporter constructs containing the indicated fragments upstream Figure 5 Comparison of tissue specificity and promoter strength of the distal KIR2DL4 transcript start site were electroporated into of distal elements. Promoter activity of distal elements from YT-Indy cells, and luciferase activity relative to empty pGL3 was the KIR2DL4, À2DL5, À3DL1, and À3DL3 genes is shown. The measured 48 h post-transfection. The positions of the fragments nucleotide position of each fragment upstream from the start codon relative to the KIR2DL4 start codon are indicated. The schematic is shown in parentheses for each element. Note that the KIR3DL3 shown above indicates the positions of putative transcription factor- element is in reverse orientation relative to the other fragments. The binding sites as open boxes. The reverse arrow indicates the pGL3 constructs were transfected into either HEK293T (open bars), orientation of the L1MEc repetitive element contained in this region. Jurkat (black bars) or YT-Indy (hatched bars). Relative promoter The relative promoter activity represents the average of at least activity shown represents the mean of at least three experiments, three independent experiments. and error bars indicate the SEM.

Genes and Immunity Distal KIR promoters MJ Stulberg et al 128 study revealed differences in transcription factor binding and tissue-specific promoter activity of these two genes.11 The search for specific differences in the KIR2DL4 gene that might explain the lack of variegated expression displayed by other KIR family members has not yet Vector Distal-1 Distal-2 Proximal indicated a candidate element to explain the distinct behavior of these genes. The initial sequence analysis of Figure 6 KIR2DL4 distal transcripts are translatable. The two distal and the proximal KIR2DL4 transcripts were generated by RT-PCR the KIR gene cluster did however reveal one significant and cloned into the pEF6/v5-His-Topo expression vector. Con- difference: the KIR2DL4 gene lacks the large array (23–63 structs were transfected into HEK293T cells and intracellular copies) of 19–20 bp nucleotide repeats found in the first staining was performed with the mAb33 antibody on permeabi- intron of all other KIR genes.16 The functional relevance lized, fixed cells following the protocol of Rajagopalan et al.22 of these repeats has yet to be directly tested, however, one intriguing possibility is their potential involvement in a gene silencing mechanism, since repetitive DNA sequences have been linked to RNA-mediated pathways KIR2DL4 transcript containing the region adjacent to the of gene silencing.23 The current study has revealed an distal-2 promoter spliced to the acceptor site at À190 additional difference between KIR2DL4 and the varie- (Figure 3b) was identified, demonstrating that transcripts gated KIR genes; KIR2DL4 possesses an additional originating from either of the two distal 2DL4 promoters promoter 10 kb upstream that produces spliced mRNAs could be spliced to the proximal promoter region. The capable of expressing KIR2DL4 protein. KIR2DL4 distal-1 promoter activity detected in the YT- The discovery of additional promoter elements up- Indy cell line was slightly higher than the activity stream of the previously characterized proximal KIR observed in HEK293T and Jurkat cells, indicating the promoter indicates a possible role of intergenic tran- possible contribution of NK-specific transcription factors scription in the regulation of KIR gene expression. to the activity of this promoter. The intergenic region Studies of the regulation of cytokine gene clusters have preceding the KIR3DL3 gene contains an L1 repeat in the also revealed the presence of many intergenic transcripts, forward direction and this element possessed even and additional promoter elements.24,25 The additional greater promoter activity when cloned into pGL3 in the upstream promoters identified in the KIR genes are same orientation as the KIR3DL1, KIR2DL5 and KIR2DL4 associated with repetitive elements, and do not behave in elements. In all cases, no promoter activity was detected a tissue-specific manner. Alu repeats within the human when the fragments were cloned into pGL3 such that the b-globin gene cluster have been shown to play a role in repetitive elements were in the opposite orientation (data the silencing of fetal g-globin during adult erythropoi- not shown). With the exception of the KIR2DL4 distal-1 esis, and a recent report has implicated Dicer-mediated promoter, the upstream promoter elements identified silencing as a mechanism associated with the loss of fetal were not NK cell-specific, since higher promoter activity hemoglobin expression during development.26 The pos- was found in HEK293T cells than the YT-Indy NK line. sible role of distal KIR transcripts in generating double- stranded RNA in conjunction with a recently discovered Functional relevance of the upstream KIR2DL4 transcripts antisense KIR transcript is currently under investigation. The detection of a significant level of KIR2DL4 tran- (SK Anderson, unpublished data). scripts originating upstream of the proximal promoter An upstream bi-directional promoter complex (Pro1) (Figure 3b) suggests that these transcripts might con- specifically active in immature NK cells was shown to be tribute to KIR expression. Since two distal KIR2DL4 required for gene activation in the murine Ly49 cluster of 18,27 promoters were identified, primers adjacent to each of class I MHC receptor genes. The mechanism of gene these promoters were used with a 30 primer containing activation in the KIR genes should be distinct however, the KIR2DL4 stop codon to amplify transcripts contain- since the upstream KIR promoter elements are neither ing the full KIR2DL4 coding region. PCR products were NK cell-specific nor bi-directional switch elements. The cloned into the pEF6/v5-His-Topo expression vector and model of variegated Ly49 gene expression proposes that verified by DNA sequencing. Expression constructs the Pro1 element chooses between transcription toward containing proximal and distal cDNAs were transfected or away from the gene in a probabilistic fashion, and into HEK293T cells and intracellular expression was forward transcripts result in the opening of the proximal detected with a KIR2DL4-specific monoclonal antibody.22 promoter region required for expression in mature NK As shown in Figure 6, the alternative transcripts cells. In contrast, the proximal promoter in the KIR genes were capable of producing KIR2DL4 protein at levels is bi-directional. KIR alleles with a high ratio of forward similar to those observed with the proximal promoter to reverse proximal promoter activity are expressed on a transcripts. higher percentage of NK cells than KIR alleles with a low ratio of forward to reverse promoter activity (SK Anderson, manuscript in preparation). Perhaps antisense transcripts produced by the proximal promoter can Discussion associate with distal sense transcripts, leading to the Comparative analysis of the transcriptional control of the production of double-stranded RNA and siRNA- KIR3DL1 and KIR2DL4 genes should yield insights into mediated silencing of the KIR proximal promoter. In the molecular mechanisms underlying NK cell expres- this respect, it is notable that the KIR2DL4 gene produces sion and selective gene activation, since the KIR3DL1 spliced distal transcripts that can express KIR2DL4 gene is expressed in a variegated fashion whereas the protein, perhaps bypassing control mechanisms operat- KIR2DL4 gene is expressed by all NK cells. A previous ing on the proximal promoter region.

Genes and Immunity Distal KIR promoters MJ Stulberg et al 129 The discovery of intergenic KIR promoters and [a-32P]RNA was separated on a 6% denaturing poly- transcripts opens new avenues of investigation into the acrylamide gel and the full-length probe was excised and possible mechanisms that control the allele-specific eluted by overnight incubation at 371C in gel elution methylation of the variegated KIR genes. Future work buffer (Ambion). Approximately 1 Â105 cpm of the gel- on the production of siRNA and potential links to KIR purified probe was added to 20 mg of total RNA in each gene methylation may provide important insights into reaction mixture. The protected RNA products were the regulation of KIR expression. separated on an 8% denaturing polyacrylamide gel. The gel was dried and exposed for B24 h in a Phosphor- Imager cassette (Molecular Dynamics, Sunnyvale, CA, Materials and methods USA). The image was visualized using PhosphorImager Cell lines SI analysis and ImageQuaNT (Molecular Dynamics). HEK293T cells were cultured in Dulbecco’s modified Eagle medium containing 10% fetal bovine serum (FBS), Generation of luciferase reporter constructs 100 U/ml penicillin, 100 U/ml streptomycin (P/S), so- Promoter fragments were generated by PCR using dium pyruvate, and L-glutamine. NK92 cells were oligonucleotide primers and cloned into the pCR2.1- cultured in RPMI 1640 media containing 10% FBS, non- Topo vector (Invitrogen). Inserts were excised with either essential amino acids, b-mercaptoethanol, P/S, sodium SacI/XhoI or XhoI/HindIII enzymes and cloned into the pyruvate, L-glutamine, HEPES, and 1000 U/ml of re- pGL3-basic firefly luciferase reporter vector (Promega, combinant human IL-2. YT-Indy and Jurkat cells were Madison, WI, USA), to create constructs with the insert cultured in RPMI 1640 media containing 10% FBS, P/S, in either the forward or reverse orientation. and L-glutamine. Cell transfection and luciferase assays Preparation of blood NK cells YT-Indy and Jurkat cells were transfected by electro- Mononuclear cells were obtained by Ficoll-Hypaque poration using a BTX ECM 830 (Genetronics, San Diego, (Sigma Diagnostics, St Louis, MO, USA) density gradient CA, USA). YT-Indy transfections included a total of centrifugation of peripheral blood from consenting 5 Â 106 cells in serum free RPMI medium with 10 mgof normal volunteers. The use of cells was approved by pGL3 constructs plus 100 ng of Renilla luciferase pRL- the University of Minnesota Institutional Review Board SV40 vector and electroporated at 250 mV, with three according to the Declaration of Helsinki. NK cells were pulses of 7 ms at an interval of 100 ms. Jurkat transfec- enriched by depletion of T-cells and monocytes using a tions used a total of 1 Â107 cells in serum-free RPMI cocktail of CD3 and CD14 magnetic beads as recom- medium with 30 mg of luciferase construct plus 1 mgof mended by the manufacturer (Miltenyi Biotech, Auburn, pRL-SV40 vector and electroporated at 300 mV, with 1 CA, USA). NK-enriched cells were stained with the pulse of 10 ms. HEK293T cells were transfected using following monoclonal antibodies: allophycocyanin FuGene 6 (Roche Molecular Biochemicals, Indianapolis, (APC)-conjugated CD56, and a cocktail of KIR mAbs, IN, USA) following the manufacturer’s protocol. Luci- including fluorescein isothiocyanate (FITC)-conjugated ferase activity was assayed at 48 h (YT-Indy, HEK293 T) DX9 (anti-CD158e), EB6 (anti-CD158a/h), and GL183 and 24 h (Jurkat) using the Dual-Luciferase Reporter (anti-CD158b/j) (BD Pharmingen, San Diego, CA, USA). Assay System (Promega) according to the manufac- KIR expressing NK cells were sorted using a FACSVan- turer’s instructions. Firefly luciferase activity was nor- tage SE fluorescence-activated cell sorter with the malized relative to the Renilla luciferase activity for each FACSDiva Option (Becton Dickenson, San Diego, CA, transfection. USA).

RT-PCR of KIR transcripts KIR2DL4 Expression vectors Total RNA was isolated from peripheral blood NK cells RNA was isolated from NK92 cells with the RNeasy kit with the RNeasy kit (Qiagen, Valencia, CA, USA) and (Qiagen) and 0.5 mg was used to generate cDNA using the 5.0 mg was used to generate cDNA using the Super- SuperscriptII first-strand cDNA synthesis kit (Invitrogen). Oligonucleotide primers specific for distinct KIR2DL4 50- scriptII first-strand cDNA synthesis kit (Invitrogen, 0 Carlsbad, CA, USA). Approximately 5% of the resulting untranslated regions were used together with a 3 primer cDNA was used per PCR reaction with a series of 24 containing the KIR2DL4 stop codon to generate PCR nucleotide primers at the intergenic positions indicated fragments that were subsequently cloned into the pEF6/ in Figure 1. The cycling conditions were: 941-20 s, v5-His-Topo expression vector (Invitrogen). Expression 581-20 s, and 731-20 s for 35 cycles. vectors were transfected into HEK293T cells using FuGene 6, and KIR2DL4 protein expression was visua- RNase protection assays lized using the mAb33 antibody on permeabilized, fixed The RNase protection assay (RPA) was performed cells following the protocol of Rajagopalan et al.22 according to the manufacturer’s protocol using an RPA kit (RPA II; Ambion, Austin, TX, USA). The probe regions chosen (Figure 1) were generated from genomic Acknowledgements DNA by PCR and cloned into the pCR4-TOPO vector. Insert containing the flanking T7 polymerase-binding This project has been funded in whole or in part with site was obtained by digestion with PmeI and FspI federal funds from the National Cancer Institute and the restriction enzymes. In vitro transcription was performed National Institutes of Health, under contract no. DHHS according to the manufacturer’s protocols using the T7 N01-C0-12400. This research was supported in part by MAXIscript (Ambion) in vitro transcription kit. The the Intramural Research Program of the NIH, National

Genes and Immunity Distal KIR promoters MJ Stulberg et al 130 Cancer Institute, Center for Cancer Research. This work 13 van Bergen J, Stewart CA, van den Elsen PJ, Trowsdale J. was supported in part by NIH Grant R01 HL55417 (JSM). Structural and functional differences between the promoters The contents of this publication do not necessarily of independently expressed killer cell Ig-like receptors. Eur J reflect the views or policies of the Department of Health Immunol 2005; 35: 2191–2199. and Human Services, nor does mention of trade names, 14 Xu J, Vallejo AN, Jiang Y, Weyand CM, Goronzy JJ. Distinct commercial products, or organizations imply endorse- transcriptional control mechanisms of killer immunoglobulin- ment by the US government. like receptors in natural killer (NK) and in T cells. J Biol Chem 2005; 280: 24277–24285. 15 Presnell SR, Zhang L, Ramilo CA, Chan HW, Lutz CT. Functional redundancy of transcription factor-binding sites in the killer cell Ig-like receptor (KIR) gene promoter. Int References Immunol 2006; 18: 1221–1232. 16 Wilson MJ, Torkar A, Haude S, Milne T, Jones D, Sheer S et al. 1 Trinchieri G. Biology of natural killer cells. Adv Immunol 1989; Plasticity in the organization and sequences of human KIR/ 47: 187–376. ILT gene families. Proc Natl Acad Sci USA 2000; 97: 4778–4783. 2 Lanier LL. NK cell recognition. Annu Rev Immunol 2005; 23: 17 Vilches C, Gardiner CM, Parham P. Gene structure and 225–274. promoter variation of expressed and nonexpressed variants 3 Carrington M, Norman P. The KIR Gene Cluster. National of the KIR2DL5 gene. J Immunol 2000; 165: 6416–6421. Library of Medicine (US), NCBI: Bethesda, MD, 2001. 18 Saleh A, Davies GE, Pascal V, Wright PW, Hodge DL, Cho EH 4 Parham P. MHC class I molecules and KIRs in human history, et al. Identification of probabilistic transcriptional switches in health and survival. Nat Rev Immunol 2005; 5: 201–214. the Ly49 gene cluster: a eukaryotic mechanism for selective 5 Ciccone E, Pende D, Viale O, Di Donato C, Tripodi G, Orengo gene activation. Immunity 2004; 21: 55–66. AM et al. Evidence of a natural killer (NK) cell repertoire for 19 Radeloff B, Nagler L, Zirra M, Ziegler A, Volz A. Specific (allo) antigen recognition: definition of five distinct NK- amplification of cDNA ends (SPACE): a new tool for the determined allospecificities in humans. J Exp Med 1992; 175: analysis of rare transcripts and its application for the promoter 709–718. analysis of killer cell receptor genes. DNA Seq 2005; 16: 44–52. 6 Wagtmann N, Biassoni R, Cantoni C, Verdiani S, Malnati MS, 20 Suzuki Y, Sugano S. Construction of a full-length enriched and Vitale M et al. Molecular clones of the p58 NK cell receptor a 5’-end enriched cDNA library using the oligo-capping reveal immunoglobulin-related molecules with diversity in method. Methods Mol Biol 2003; 221: 73–91. both the extra- and intracellular domains. Immunity 1995; 2: 21 Kikuchi-Maki A, Yusa S, Catina TL, Campbell KS. KIR2DL4 is 439–449. an IL-2-regulated NK cell receptor that exhibits limited 7 Valiante NM, Uhrberg M, Shilling HG, Lienert-Weidenbach K, expression in humans but triggers strong IFN-gamma Arnett KL, D’Andrea A et al. Functionally and structurally production. J Immunol 2003; 171: 3415–3425. distinct NK cell receptor repertoires in the peripheral blood of 22 Rajagopalan S, Bryceson YT, Kuppusamy SP, Geraghty DE, two human donors. Immunity 1997; 7: 739–751. van der Meer A, Joosten I et al. Activation of NK Cells by 8 Santourlidis S, Trompeter HI, Weinhold S, Eisermann B, an Endocytosed Receptor for Soluble HLA-G. PLoS Biol 2006; Meyer KL, Wernet P et al. Crucial role of DNA methylation 4: e9. in determination of clonally distributed killer cell Ig-like 23 May BP, Lippman ZB, Fang Y, Spector DL, Martienssen RA. receptor expression patterns in NK cells. J Immunol 2002; 169: Differential regulation of strand-specific transcripts from 4253–4261. Arabidopsis centromeric satellite repeats. PLoS Genet 2005; 9 Chan HW, Kurago ZB, Stewart CA, Wilson MJ, Martin MP, 1: e79. Mace BE et al. DNA methylation maintains allele-specific KIR 24 Jones EA, Flavell RA. Distal enhancer elements transcribe gene expression in human natural killer cells. J Exp Med 2003; intergenic RNA in the IL-10 family gene cluster. J Immunol 197: 245–255. 2005; 175: 7437–7446. 10 Chan HW, Miller JS, Moore MB, Lutz CT. Epigenetic control of 25 Rogan DF, Cousins DJ, Santangelo S, Ioannou PA, Antoniou highly homologous killer immunoglobulin-like receptor gene M, Lee TH et al. Analysis of intergenic transcription in the alleles. J Immunol 2005; 175: 5966–5974. human IL-4/IL-13 gene cluster. Proc Natl Acad Sci USA 2004; 11 Stewart CA, van Bergen J, Trowsdale J. Different and 101: 2446–2451. divergent regulation of the KIR2DL4 and KIR3DL1 promoters. 26 Haussecker D, Proudfoot NJ. Dicer-dependent turnover of J Immunol 2003; 170: 6073–6081. intergenic transcripts from the human beta-globin gene 12 Trompeter HI, Go´mez-Lozano N, Santourlidis S, Eisermann B, cluster. Mol Cell Biol 2005; 25: 9724–9733. Wernet P, Vilches C et al. Three structurally and functionally 27 Tanamachi DM, Moniot DC, Cado D, Liu SD, Hsia JK, Raulet divergent kinds of promoters regulate expression of clonally DH. Genomic Ly49A trans-genes: basis of variegated Ly49A distributed killer cell Ig-like receptors (KIR), of KIR2DL4, and gene expression and identification of a critical regulatory of KIR3DL3. J Immunol 2005; 174: 4135–4143. element. J Immunol 2004; 172: 1074–1082.

Genes and Immunity