Receptors of CD4 T Cells Expressing Killer Ig-Like Phenotypic And

The Journal of Immunology

Phenotypic and Functional Characterization of CD4 T Cells Expressing Killer Ig-Like Receptors1

Jeroen van Bergen,2* Allan Thompson,* Arno van der Slik,* Tom H. M. Ottenhoff,* Jacobijn Gussekloo,† and Frits Koning*

Killer Ig-like receptors (KIR) are commonly found on human NK cells, ␥␦ T cells, and CD8 T cells. Although KIR؉ CD4 T cells are found in certain patients, their prevalence in healthy donors is controversial. We now provide definitive proof that such cells are present in most individuals, and report on their frequency, surface phenotype, cytokine profile, and Ag specificity. The number of KIR؉ CD4 T cells detected in peripheral blood increased with age. In contrast with regular KIR؊ CD4 T cells, the majority ,of KIR؉ CD4 T cells lacked surface expression of CD27, CD28, CCR4, and CCR7, but did express CD57 and 2B4. In addition KIR were detected on approximately one-tenth of CD28؊ and CD57؉ memory CD4 T cells. In line with the absence of the Th2 ,marker CCR4, the KIR؉ CD4 cells produced mainly IFN-␥ and little IL-4, IL-10, or IL-17 upon TCR triggering. Furthermore the KIR؉ population contained cells that responded to recall Ags in an HLA class II-restricted fashion. Together, our data indicate that KIR-expressing CD4 T cells are predominantly HLA class II-restricted effector memory Th1 cells, and that a significant, previously unrecognized fraction of effector memory Th1 cells expresses KIR. The Journal of Immunology, 2004, 173: 6719–6726.

atural killer cells use a wide variety of surface receptors KIR expression by CD4 T cells is scarce. However, in patients to recognize infected and malignantly transformed cells. with certain diseases KIR are frequently detected on an expanded The killer Ig-like receptors (KIR)3 constitute the largest CD28Ϫ subset of CD4 T cells (13, 14). In rheumatoid arthritis, N Ϫ family of human NK receptors, with multiple inhibitory and acti- CD28 CD4 T cells are cytolytic and they proliferate in response vating members (1). The inhibitory receptors bind classical HLA to autologous PBMC (15, 16). Moreover, their number correlates class I molecules and mediate “missing self” recognition. The with disease severity (17). When compared with CD28Ϫ CD4 T physiological ligands for the activating KIR are unknown. In ad- cells from healthy control subjects, the cells preferentially express dition to NK cells, T cells also frequently express KIR. However, the disease-associated activating KIR2DS2 molecule (9), which the role these NK receptors play in T cell responses remains costimulates TCR-mediated IFN-␥ production and proliferation largely enigmatic (2). (13). In acute coronary syndromes, CD28Ϫ CD4 T cells can be The KIR family displays a high degree of polymorphism, which activated by KIR cross-linking directly, without the need for TCR is associated with disease. The KIR locus is polymorphic with engagement (14). The Ag specificity of the above-described KIRϩ by guest on September 30, 2021. Copyright 2004 Pageant Media Ltd. respect to both gene content and sequence (3–5). KIR haplotypes CD4 cells is unknown. contain between 5 and 12 highly homologous KIR genes (4), and There are conflicting reports on KIR expression on CD4 T cells up to 9 alleles for individual KIR genes have been described (5). in healthy individuals. According to Anfossi et al. (12), on average KIR3DS1 and its proposed ligand HLA-Bw4 confer resistance to 7 Ϯ 3% (ϮSEM) of KIRϩ T cells carry CD4, indicating that the the development of full-onset AIDS in HIV-infected individuals large majority of healthy donors has KIRϩ CD4 T cells. Indeed, a (6). Also, the KIR2DS2 gene is associated with various autoim- number of KIRϩ CD4 T cell clones have been isolated from mune conditions, such as psoriatic arthritis (7), diabetes (8), and healthy donors (18, 19). In contrast, Warrington et al. (20) detected vascular complications in rheumatoid arthritis (9). KIR expression selectively on CD28Ϫ CD4 T cells, an extremely KIR expression by TCR-␥␦ϩ T cells and a subset of CD8ϩ rare subset in healthy individuals. By consequence, they detected https://www.jimmunol.org TCR-␣␤ϩ cells is firmly established (10–12), but information on KIRϩ CD4 T cells only in a small minority of individuals, containing increased frequencies of CD28Ϫ CD4 cells (9, 14). We now provide definitive proof that KIRϩ CD4 T cells are present in *Department of Immunohematology and Blood Transfusion and †Section of Geron- most individuals, and report on their frequency, surface phenotype, tology and Geriatrics, Department of General Internal Medicine, Leiden University Medical Center, Leiden, The Netherlands cytokine production, and Ag specificity.

Downloaded from Received for publication June 23, 2004. Accepted for publication September 17, 2004. Materials and Methods The costs of publication of this article were defrayed in part by the payment of page Subjects charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. After informed consent was obtained, peripheral blood cells were obtained 1 This work was supported by the Netherlands Organisation for Scientiﬁc Research from healthy adult and 90-year-old donors. Healthy adults were random (NWO) VENI Grant 916.36.025 (to J.v.B.) and the Diabetes Fonds Nederland/Juve- blood bank donors ranging in age from 18 to 70 years (n ϭ 30). Healthy nile Diabetes Research Foundation/Netherland Organisation for Health Research and 90-year-old donors (n ϭ 10) were enrolled in the cohort of the Leiden Development Grant 2001.10.004 (to A.v.d.S.). T.H.M.O. is supported by the NWO, 85-plus study (21, 22), a prospective population-based study of inhabitants the Netherlands Leprosy Foundation, and the European Commission. of Leiden, The Netherlands. Health was deﬁned as no acute illness, no 2 Address correspondence and reprint requests to Dr. Jeroen van Bergen, Department of death within 3 mo after blood sampling, no severe cognitive impairment Immunohematology and Blood Transfusion, Leiden University Medical Center, PO Box and a plasma C-reactive protein level below 10 mg/L. Umbilical cord blood 9600, 2300 RC Leiden, The Netherlands. E-mail address: [email protected] (UCB) cells were derived from the umbilical vein after normal full-term 3 Abbreviations used in this paper: KIR, killer Ig-like receptor; ILT, Ig-like transcript; deliveries (n ϭ 7). Heparinized blood was collected and PBMC or cord LAK, lymphokine-activated killer; UCB, umbilical cord blood. blood mononuclear cells were isolated by density gradient centrifugation

over Ficoll-Hypaque, and subsequently stored at Ϫ80°C in 10% DMSO/ temperature of 61°C, followed by 29 cycles with an annealing temperature 20% FCS until use. of 60°C. All PCRs were conducted in a Peltier Thermal Cycler (PTC-200; MJ Research, Waltham, MA). Antibodies TCR repertoire analysis Directly conjugated Abs to CD3, CD8, CD25, CD27, CD45RA, CD45RO, CD56, CD57, CD94, CCR4, CCR7, TCR-␣␤, KIR3DL1 (CD158e1, clone TCR-␤ chain variable gene segment usage of the clones was established by DX9), and isotype controls were obtained from BD Biosciences (San Jose, FACS using the IOTest ␤ Mark TCR kit (Beckman Coulter), TCR-␣ chain CA) and to CD4 from DAKO (Cytomation, Glostrup, Denmark). The PE- variable gene segment usage was determined by RT-PCR using TCR-␣ conjugated EB6 (CD158a, CD158h, anti-KIR2DL1, KIR2DS1), GL183 chain variable gene segment family specific primers (26). The CDR3 re- (CD158b, CD158j, KIR2DL2, KIR2DL3, KIR2DS2), FES172 (CD158i, gions of the TCRs of selected clones were amplified by PCR, cloned into KIR2DS4), and CD244 (2B4) Abs were from Beckman Coulter (Mijdre- the pCR2.1 vector using TA cloning kit (Invitrogen Life Technologies), cht, The Netherlands). and then sequenced. Unconjugated Abs to NKG2D (BAT221), NKp30 (A76), NKp46 T cell functional assays (BAB281), NKp44 (Z231), and Ig-like transcript (ILT)-2 (F278) were a kind gift from Dr. D. Pende (Instituto Nazionale per la Ricerca sul Cancro, PBMC from HLA-DR3/HLA-DR4 heterozygous individuals were stained Genova, Italy), anti-KIR2D Ab (NKVSF1) from Dr. A. Poggi (Instituto with CD4-FITC and a mix of PE-conjugated Abs (EB6, GL183, DX9, Nazionale per la Ricerca sul Cancro), and anti-KIR2DL4 Ab (cl-33; Ref. FES172), and subsequently plated at 1000 cells/well by a BD Biosciences 23) from Dr. E. O. Long (National Institutes of Health, Bethesda, MD). FACSVantage flow cytometer using CellQuest software. Three wells (3000 Binding of unconjugated Abs was detected using goat anti-mouse IgG cells) of KIRϩ and three wells of KIRϪ CD4 cells were expanded twice, F(ab)2-RPE (DAKO). using mixed irradiated allogeneic PBMC supplemented with 1 ␮g/ml PHA and 20 U/ml IL-2, resulting in 50–100 million cells of each type. This in Flow cytometry vitro expansion did not affect surface KIR expression patterns. ϩ Ϫ Cells were stained with CD3-PerCP, CD4-allophycocyanin, KIR-PE (EB6, The KIR and KIR CD4 bulk cultures were stimulated with plate- ␮ GL183, DX9, FES172 mixed or separately), combined with an Ab to an bound anti-CD3 (1 g/ml in PBS; Janssen-Cilag, Tilburg, The Nether- additional surface marker conjugated to FITC. PE-conjugated Abs to a lands) in 96-well flat-bottom plates (25,000 cells/well). Culture superna- Ϫ limited number of surface markers (CD94, CD244, CCR4, CCR7, CXCR4) tants were harvested 24 h after stimulation and stored at 20°C. Cytokine were combined with the CH-L FITC Ab (directed to KIR2DL2, KIR2DL3, concentrations were determined using the Becton Dickinson Th1/Th2 Cy- KIR2DS2). Staining was performed at 4°C for 1 h, followed by three tometric Bead Array kit according to the manufacturer’s protocols. For the washes with 0.1% BSA in PBS. The cells were then fixed in 0.5% para- detection of IL-17 by sandwich ELISA (BioSource International, Cama- 6 rillo, CA), supernatants were harvested 72 h after stimulation. T cell pro- formaldehyde in PBS and stored at 4°C until acquisition of at least 10 cells 3 per sample on a BD FACSCalibur cytometer and analyzed using the liferation was determined by measuring the incorporation of [ H]thymidine CellQuest (BD Biosciences) analysis software. during the final 16 h of a 96-h culture period. To exclude possible staining artifacts, we performed a number of control For the detection of recall responses, 100,000 irradiated (3000 rad) experiments. First, Ͼ95% of the KIRϩ CD4 T cell population signal dis- PBMC were incubated with one of the following recall Ags: tetanus toxin appeared when CD3, CD4, or KIR Abs were replaced with isotype- (1 Lf/ml; Rijksinstituut voor Volksgezondheid en Milieu (RIVM), matched control Abs. Second, when the small population of KIRϩ CD4 Bilthoven, The Netherlands); Mycobacterium tuberculosis protein purified ϫ cells was isolated by FACS (1, 10, or 1000 cells/well), expanded and sub- derivative (100 diluted) from RIVM; the HLA-DR3-restricted Mycobac- ␮ sequently stained with KIR or control Abs, Ͼ99% of cells were indeed terium tuberculosis HSP65 3-13 peptide (1 g/ml, Ref. 27); CMV lysate ϩ Ϫ ϫ KIR , whereas control KIR CD4 T cells treated in the same manner (100 diluted) from Microbix Biosystems (Toronto, Ontario, Canada); ϩ ␮ always remained negative for KIR. Thus, low numbers of KIR CD4 T CMV pp65 protein (1 g/ml; Austral Biologicals, San Ramon, CA); Can- ϫ cells could be detected reliably. dida albicans lysate (100 diluted); or staphylococcal enterotoxin A (0.1 ␮g/ml; Serva, Heidelberg, NY) in IMDM/10% human serum for4hina by guest on September 30, 2021. Copyright 2004 Pageant Media Ltd. T cell cloning 96-well V-bottom plate. Blocking Abs to HLA class I (W6/32) or class II (PdV5.2), or CD4 (RIV6) were added during the final 30 min of this in- PBMC were stained with CD4-allophycocyanin and KIR-PE (a mix of ␮ ϩ Ϫ cubation, at a final concentration of 10 g/ml. Subsequently, 50,000 au- EB6, GL183, DX9, FES172) Abs as previously described. KIR and KIR tologous responder KIRϩ or KIRϪ CD4 T cells per well were added. T cell CD4 cells were subsequently plated at 1 cell/well by a BD Biosciences proliferation was determined by measuring the incorporation of [3H]thy- FACSVantage flow cytometer using CellQuest software. T cells were cul- midine during the final 16 h of a 96-h culture period. tured in IMDM (Invitrogen Life Technologies, Gaithersburg, MD) containing 20 U/ml recombinant human IL-2 (Cetus, Emeryville, CA) and 6 Results 10% pooled human serum. Mixed irradiated allogeneic PBMC (10 cells/ KIRϩ CD4 T cells are present in most healthy individuals, and well, n ϭ 2–6 donors) supplemented with 1 ␮g/ml PHA were used as feeder cells at the start of the culture and subsequently at 7- to 10-day accumulate with age https://www.jimmunol.org intervals. Cell surface KIR expression levels of these clones were compa- We determined whether KIRϩ CD4 T cells could be detected in rable to NK cells, and remained stable for at least five rounds of restimulation. the peripheral blood of healthy individuals. To this end, PBMC were stained with directly conjugated Abs to CD3, CD4, and KIR. KIR repertoire analysis In healthy adult donors age 18–70, on average 0.2% of CD4 T KIR genotyping was performed as described (8). RT-PCR analysis of KIR cells stained with KIR Abs (Fig. 1). We also determined the fre- ϩ expression on T cell clones was performed as described (24). Briefly, total quencies of KIR CD4 T cells in UCB and in PBMC from healthy Downloaded from RNA was isolated from 5 to 10 ϫ 106 cells, and 8 ␮g of RNA was reverse 90-year-old individuals (Fig. 1). Although KIRϩ CD4 T cells were transcribed using a mix of random and oligo(dT) primers and Moloney virtually undetectable in UCB, KIR were detected on 1.0% of CD4 murine leukemia virus reverse transcriptase (Promega, Leiden, The Neth- ␮ T cells in PBMC from the average healthy nonagenarian. Thus, erlands) in a total volume of 50 l. Negative control reactions were per- ϩ formed in the absence of template or reverse transcriptase. cDNA (0.2 ␮l KIR CD4 T cell numbers appeared to increase with donor age. ␮ per reaction volume of 25 l) was amplified using KIR-specific primers. ϩ We used the primer sets described by Uhrberg et al. (3), but primers were KIR CD4 T cells have an effector memory surface phenotype adjusted when these primers did not match all currently known allele se- To compare the surface phenotype of KIRϩ to that of KIRϪ CD4 quences (5). In addition, primer sets specific for KIR2DL5 and KIR3DL3 T cells, four-color flow cytometry was performed on PBMC from were used (A. Thompson, J. van Bergen, A. van der Slik, and F. Koning, ϩ manuscript in preparation). Primers were used at a concentration of 0.5 healthy adults (Fig. 2). As before, KIR CD4 T cells were distin- ␮ M. Taq polymerase and buffers were obtained from Promega; MgCl2 was guished from other cells using directly conjugated Abs to KIR, used at 2.0 mM. Positive controls for the PCR included cDNA from 3-day- CD3, and CD4. old lymphokine-activated killer (LAK) cultures derived from the donor’s Because KIR are characteristic of NK cells rather than T cells, PBMC or from spleen cells of KIR transgenic or control nontransgenic ϩ mice (25), cloned KIR cDNAs, and a control primer set specific for ␤-actin T cell and NK cell markers were examined. KIR CD4 T cells all (0.05 ␮M). The “touch-down” PCR consisted of 5 cycles with an annealing carried an ␣␤ TCR (Fig. 2A), and were largely negative for CD25, The Journal of Immunology 6721

which is present on most regulatory T cells (28), and also negative for TCR-␣ chain V24, a marker of CD1d-restricted NKT cells (29). Thus, they were regular TCR␣␤ϩ T cells, and not CD4ϩCD25ϩ regulatory cells or NKT cells. CD244 (2B4), a costimulatory molecule that binds CD48 and is expressed on all NK and some T cells (30), was frequently expressed by KIRϩ CD4 T cells, but not by KIRϪ CD4 T cells (Fig. 2B). The large majority of KIRϩ CD4 T cells did not express NK markers CD16, CD56, or CD94. As the observed age-dependent increase in the frequency of KIRϩ CD4 T cells was consistent with a memory phenotype, PBMC were stained with Abs to various memory T cell markers (Fig. 2C). Like memory T cells, KIRϩ CD4 T cells were mostly CD45ROϩ CD45RAϪ, but this was not consistent in all donors (data not shown). The most striking differences between KIRϩ and Ϫ ϩ KIR CD4 T cells were observed for CD27, CD28, and CD57. FIGURE 1. KIR CD4 T cell numbers in healthy donors. KIR expres- Ϫ ϩ ϩ Although KIR cells were almost always CD27 , CD28 , sion frequency on peripheral blood CD4 T cells determined by staining Ϫ ϩ with a mix of KIR Abs (EB6, GL183, DX9, FES172). Together, the KIR CD57 , the majority of KIR CD4 T cells displayed the opposite Abs covered a wide range of KIR: KIR2DL1, KIR2DL2, KIR2DL3, phenotype, characteristic of effector memory cells. KIR2DS1, KIR2DS2, KIR2DS4, and KIR3DL1. Newborn samples were Functionally distinct CD4 subsets are characterized by their che- Ϫ drawn from UCB; adults were random healthy blood bank donors age mokine receptor expression patterns (31). In contrast to KIR CD4 18–70 years. Frequencies were calculated by subtracting background stain- T cells, KIRϩ CD4 T cells did not express CCR7, conﬁrming the ing with isotype control Abs (average 0.03 Ϯ 0.02% (ϮSD) of CD4 T effector memory phenotype of most KIRϩ CD4 T cells indicated cells) from KIR staining. Calculated frequencies were: UCB (n ϭ 7) mean by the absence of CD27 and CD28 (Fig. 2D). Expression of the Ϯ ϭ Ϯ 0.00 0.05%; adults (n 30) mean 0.20 0.15%; 90-year-old donors Th1 markers CCR5, CXCR3, and CXCR6 (31, 32) did not differ n ϭ Ϯ Ϫ ϩ ( 10) mean 0.98 0.89%. consistently between KIR and KIR CD4s (data not shown). KIRϩ CD4 T cells did not express the Th2 marker CCR4, whereas a large subset of KIRϪ CD4 T cells did. Finally, the naive T cell marker CXCR4 was more highly expressed on KIRϪ than on by guest on September 30, 2021. Copyright 2004 Pageant Media Ltd. https://www.jimmunol.org Downloaded from

FIGURE 2. Surface markers of KIRϩ and KIRϪ CD4 T cells. PBMC were stained with CD3, CD4, KIR Abs combined with Abs to markers of T cells (A), NK cells (B), memory cells (C), and Abs to chemokine receptors (D). Histograms were gated on KIRϩ CD3ϩCD4ϩ (KIRϩ)orKIRϪ CD3ϩCD4ϩ (KIRϪ), live lymphocytes were selected using forward light scatter/side light scatter gates. Data are from a single healthy donor, representative of six. Surface marker of interest (thick line) and isotype control (dotted line) are indicated. 6722 KIR EXPRESSING Th CELLS: EFFECTOR MEMORY Th1 CELLS

KIRϩ CD4 T cells. In conclusion, the KIRϩ CD4 T cell population appeared biased toward effector memory Th1 cells and largely devoid of Th2 cells.

Effector memory cells frequently express KIR Overall, the best markers to distinguish KIRϩ from KIRϪ CD4 T cells were CD244, CD27, CD28, and CD57. Therefore, we determined how well these markers predicted KIR expression for the CD4 T cell population. Approximately one-tenth of CD28Ϫ or CD57ϩ cells expressed detectable KIR at their cell surface (Fig. 3A). Compared with the entire CD4 T cell population, of which 0.2% expressed KIR, this was a roughly 50-fold enrichment. Of note, the Abs used detect only a subset of surface expressed KIR and therefore provide an underestimate of the true number of KIRϩ cells. Thus, a sizeable fraction of late-stage memory CD4 cells expressed KIR.

CD4 T cells express a distinct KIR repertoire KIR genes are expressed stochastically and independently from each other. As a result, individual cells express a seemingly random selection of the KIR genes available in the genome, resulting in a diverse repertoire of specificities. The KIR repertoire of CD4 T cells was compared with that of NK cells. To this end, the four KIR Abs thus far used in combi- FIGURE 4. KIR repertoires on CD4 T cells compared with other lym- nation were now used separately in FACS experiments (Fig. 4). phocyte subsets. Cells were gated on live lymphocytes using forward light ϩ ϩ ϩ Ϫ Due to the great diversity in KIR genotypes, the KIR repertoires of scatter/side light scatter and CD3 CD4 (CD4 T cell), CD3 (NK), ϩ ϩ ϩ ϩ ϩ ϩ Ϫ Ϫ NK cells usually differ between individuals (33). For this reason, CD3 CD4 CD28 (CD28 ), or CD3 CD4 CD28 (CD28 ). Numbers indicate percentage of cells within these gates positive for the KIR Abs lymphocyte subsets were only compared within the same individ- EB6 (KIR2DL1, KIR2DS1), GL183 (KIR2DL2, KIR2DL3, KIR2DS2), uals. In all donors examined, the KIR repertoire of CD4 T cells DX9 (KIR3DL1), FES172 (KIR2DS4), or isotype control Abs (mouse differed markedly from the NK cell KIR repertoire (Fig. 4A). For IgG1, IgG2a). To obtain sufficient cell numbers within these gates, 5 mil- example, the EB6 Ab, specific for KIR2DL1 and KIR2DS1, did lion PBMC were acquired per sample. Data are from a single donor, rep- not stain CD4 T cells detectably, whereas the NK cells in the same resentative of 10 (A)or5(B). sample did stain with this Ab. Thus, peripheral blood CD4 T cells display KIR repertoires that are distinct from that of NK cells. CD28Ϫ CD4 T cells have been reported to play a role in rheu- cells frequently expressed KIR (Fig. 3), we compared the KIR

by guest on September 30, 2021. Copyright 2004 Pageant Media Ltd. Ϫ ϩ matoid arthritis and acute coronary syndromes (17, 34). As these repertoire of CD28 to that of CD28 CD4 T cells by FACS (Fig. 4B). Again, the EB6 Ab did not stain either population detectably. In all donors, the GL183 epitope present on KIR2DL2, KIR2DL3, and KIR2DS2 dominated the KIR repertoire of the CD28Ϫ cells, whereas the CD28ϩ cells displayed a broader repertoire of KIR. Therefore, KIR repertoires differed between CD28Ϫ and CD28ϩ CD4 cells.

KIRϩ CD4 T cells need not express KIR2DL4 or an inhibitory https://www.jimmunol.org self-HLA-specific KIR The available KIR Abs have a number of limitations. First, they cover only a subset of KIR. Second, some Abs (EB6, GL183) do not discriminate between activating and inhibitory KIR. Third, at least one Ab (DX9) does not detect every allele of its ligand (KIR3DL1; Ref. 35). To overcome these limitations, KIRϩ CD4 T Downloaded from cell clones were isolated from an adult healthy donor, and KIR expression was determined by RT-PCR (Fig. 5A). We focused our attention on 10 KIRϩ CD4 T cell clones and five KIRϪ control clones from a single donor (Fig. 5A). A number of observations were noteworthy. First, the putative HLA-G receptor KIR2DL4 (36), expressed by all NK (24) and most KIRϩ CD8 T (37–39) cell clones expanded in vitro, was expressed by only two of ten KIRϩ CD4 clones (Fig. 5A). In clones containing detectable KIR2DL4 transcripts (Fig. 5B), KIR2DL4 was found FIGURE 3. KIR frequency on CD4 T cell subsets. PBMC were stained expressed at the cell surface (Fig. 5C). Thus far, surface expression with CD3, CD4, and KIR Abs (EB6, GL183, DX9, FES172) combined bright with Abs to CD27, CD28, or CD57 (A). The CD244 Ab (B) was combined of KIR2DL4 protein has been detected only on CD56 and on with a single KIR Ab (CH-L). Note that, given the inherent limitations of IL-2-activated NK cells (40, 41). Second, a self-specific inhibitory ϩ the KIR Abs, the calculated percentages underestimate the true percentages NK receptor was not obligatory for the KIR CD4 cells. The do- of KIRϩ cells within these CD4 subsets. nor carried two HLA class I alleles that are known ligands for The Journal of Immunology 6723 by guest on September 30, 2021. Copyright 2004 Pageant Media Ltd.

FIGURE 5. KIR and TCR gene expression in KIRϩ CD4 T cell clones. A, Ten KIRϩ and five KIRϪ CD4 T cell clones from a single donor were analyzed for KIR expression by RT-PCR. KIR RT-PCRs were performed for KIR present in the genome of the donor. In all cases, KIR expression detected by Ab (data not shown) matched the KIR genes found expressed by RT-PCR. Control samples included cDNA from LAKs from the same donor, as well as LAKs from a KIR transgenic (KIR tg) and a control mouse (non tg). Cloned KIR cDNAs (2DL1, 3DL3) were also tested. TCR-␤ chain variable gene segment usage was determined by FACS using TCR-␤ chain variable-specific Abs. Of the TCR-␤ chain V22ϩ clones, TCR-␣ chain variable gene segment usage https://www.jimmunol.org was determined by RT-PCR. n.d., Not determined. B, KIR2DL4 RT-PCR on cDNA from selected clones (clones 4, 5, and 7) and LAKs from PBMC of the same donor (ϩ). Control templates were water (Ϫ) and cDNA from LAKs of a nontransgenic control (Ϫ/Ϫ) and a KIR transgenic (ϩ/ϩ) mouse (25). C, FACS staining of clones 4, 5, and 7 with a KIR2DL4-specific (thick line) or isotype control (thin line) Ab. D, The TCR CDR3 regions of two pairs of clones expressing identical TCR-␣ chain and TCR-␤ chain variable gene segments were sequenced and aligned.

inhibitory KIR: HLA-Cw3 and HLA-Cw4, which are ligands for a restricted TCR repertoire (13, 38, 39). The ten KIRϩ clones Downloaded from KIR2DL2 and KIR2DL1, respectively. Yet, none of the clones examined carried at least five different TCR (Fig. 5A). However, expressed KIR2DL1, and only one clone of ten expressed four of ten KIRϩ clones carried TCR-␤ chain V22. Two TCR-␤ KIR2DL2. Furthermore, an Ab to the broadly HLA class I specific chain V22ϩ clones in this panel (clones 7 and 9) expressed distinct inhibitory receptor ILT-2 stained two clones (clones 3 and 5), and TCR and KIR, but the other two (clones 3 and 5) expressed iden- none of the clones stained with Abs to NKp30, NKp44, NKp46, tical TCR and KIR genes (Fig. 5D), demonstrating that the latter CD94, or NKG2D (data not shown). Finally, nine of ten clones two represented a clone that had divided at least once in vivo. expressed KIR2DS2, suggesting that the KIRϩ CD4 T cells of this Thus, the data indicate that KIRϩ CD4 T cells display an oligo- donor frequently express this activating receptor. In conclusion, clonal TCR repertoire. only a minority of KIRϩ CD4 T cell clones expressed KIR2DL4 or an inhibitory NK receptor specific for self-HLA class I. Th1 biased cytokine profile of KIRϩ vs KIRϪ CD4 T cells ϩ TCR diversity of KIR CD4 T cells The absence of Th2 marker CCR4 from KIRϩ CD4s (Fig. 2) sug- KIRϩ CD8 T cells of healthy donors and KIRϩ CD28Ϫ CD4 T gested that these cells were biased toward a Th0/Th1 phenotype. cells from rheumatoid arthritis patients are polyclonal, but do have To test this hypothesis, cytokine production profiles of in vitro 6724 KIR EXPRESSING Th CELLS: EFFECTOR MEMORY Th1 CELLS

expanded KIRϩ and KIRϪ CD4 T cell bulk cultures were compared (Fig. 6A). The in vitro expansion was necessary to obtain large enough numbers of puriﬁed responder cells. Upon stimulation with plate-bound anti-CD3, both CD4 subsets produced sim- ilar amounts of IFN-␥ (Fig. 6A). In contrast, the production of Th2 cytokines IL-4, IL-5, and IL-10 was markedly lower in the KIRϩ CD4 cells. IL-17 is produced by a subset of Th1/Th0 cells (42), but only KIRϪ CD4 cells produced signiﬁcant amounts of this cytokine (Fig. 6A). Thus, the KIRϩ CD4 T cell population included mainly Th1 cells unable to produce IL-17.

KIRϩ CD4 T cells contain HLA class II-restricted recall Ag-specific cells The specificity of KIRϩ TCR-␣␤ϩ CD4 T cells is unknown. Be- cause KIRϩ CD4s expressed memory markers, they might also specifically recognize recall Ags. To test this hypothesis, equal numbers of in vitro expanded KIRϩ and KIRϪ CD4 T cells were stimulated with autologous irradiated PBMC loaded with bacterial, fungal, or viral recall Ags. Both CD4 subsets contained cells that proliferated in response to a number of these recall Ags, CMV lysate in particular (Fig. 6B). The response to CMV lysate could be inhibited by Abs to HLA class II and CD4, but not by an anti-HLA class I Ab (Fig. 6C). Thus, the KIRϩ CD4 T cell population contained HLA class II-restricted memory cells specific for CMV Ags.

Discussion We investigated the frequency, surface phenotype, cytokine profile, and Ag specificity of KIRϩ CD4 T cells in healthy individuals. In contrast with previous reports, low numbers of KIRϩ CD4 T cells were found in the peripheral blood of nearly every healthy adult, and this number increased with age. These cells were strongly enriched for effector memory cells that produced mainly IFN-␥. Furthermore, a subset of KIRϩ CD4 T cells recognized recall Ags in an HLA class II-restricted fashion. by guest on September 30, 2021. Copyright 2004 Pageant Media Ltd. The repertoire of KIRs expressed by KIRϩ CD4 T cells differed markedly from that of NK cells. In contrast with NK cells (24) and in agreement with reports on KIRϩ CD28Ϫ CD4 T cells (9, 14), the majority of KIRϩ CD4 T cells did not express KIR2DL4 or a self-HLA specific inhibitory NK receptor. Most strikingly, the expression of KIR2DL1 and KIR2DS1 was extremely rare on CD4 T cells. KIR2DL1 and KIR2DS1 specifically bind HLA-C group II molecules such as HLA-Cw4 (43). However, the HLA type C of our donors was diverse and could not easily explain the absence of https://www.jimmunol.org KIR2DL1/KIR2DS1 from CD4 T cells. Alternatively, it is conceivable that CD4 T cells lack one or more transcription factors required for the expression of these receptors. Indeed, the promoters of KIR2DL1 and KIR2DS1 lack a binding site for YY1 present in most other KIR promoters, and may therefore be regulated dif- ferently from the other KIR genes (our unpublished observations). Downloaded from Finally, the KIR repertoire may be skewed depending on T cell Ag specificity or affinity (44). Clearly, CD4 T cells and NK cells do not obey the same rules for KIR expression. The KIRϩ CD4 T cells shared a number of features with pre- ϩ viously described KIR CD8 T cells. First, both populations ac- FIGURE 6. Cytokine profile (A), Ag specificity (B), and HLA restric- cumulate with increasing age (12). Second, both are enriched for tion (C)ofKIRϩ CD4 T cells. A, KIRϩ and KIRϪ CD4 T cell bulk cultures effector memory cells (11, 12, 39), characterized by the absence of were stimulated with plate-bound anti-CD3, and cytokines were measured surface expression of CCR7, CD27, and CD28 as well as the pres- from the supernatant 24 h (IFN-␥, TNF-␣, IL-4, IL-5, IL-10) and 72 h ence of CD57 on the cell surface (45). In contrast with KIRϩ CD8 (IL-17) after stimulation. B, These bulk cultures were also used as respond- T cells (12) and NK cells, however, KIRϩ CD4 T cells did not ers in a proliferation assay stimulated with autologous PBMC loaded with show expression of the IL-2/IL-15 receptor ␤-chain (CD122, data recall Ags. C, Abs to HLA class II, but not class I could inhibit the response not shown). This finding fits the observation that IL-15 stimulates to CMV Ags. IL-2 stimulation in the presence of Ab was included as a toxicity control. Data are from a single donor, representative of five. CD8 but not CD4 memory cells (46). Thus, for all TCR-␣␤ T cells, an effector memory surface phenotype is predictive of KIR ex- The Journal of Immunology 6725

pression, suggesting that KIR play an important role in shaping vating KIR can costimulate CD4 T cells (13, 18, 60), and may both CD8 and CD4 effector memory cell responses. thereby recruit low-affinity T cells to the CD4 T cell response. This The KIRϩ CD4 T cells from healthy donors described in this would effectively expand the Ag-specific T cell repertoire. This report share several similarities with the CD28Ϫ CD4 T cell pop- beneficial mechanism may come at a dear price, as activating KIR ulation in rheumatoid arthritis patients (16). CD28Ϫ CD4 T cells in may unleash the autoreactive potential of T cells with a low affinity rheumatoid arthritis patient PBMC also expressed CD57 (16) and for self Ags. KIR (13). Moreover, the KIR repertoire expressed by CD28Ϫ CD4 cells in rheumatoid arthritis patients and healthy donors was com- Acknowledgments parable: the cells rarely expressed the EB6 (KIR2DL1, KIR2DS1) We express gratitude to the blood donors, without whom these studies marker, preferentially stained with the GL183 (KIR2DL2, would not have been possible. We thank Marieke Hoeve for IL-17 ELISA, KIR2DL3, KIR2DS2) Ab (13), and not all KIRϩ cells expressed and Maarten van der Keur and Reinier van der Linden for excellent assis- tance in FACS sorting, Drs. Belkin and Trowsdale for spleen cells from KIR2DL4 (9, 14). Finally, the predominant cytokine produced by ϩ ␥ their KIR transgenic mouse, and Drs. Pende, Poggi, and Rajagopalan for both KIR populations is IFN- (47). However, there were a num- generous sharing of Abs. We are also grateful to Ester M. van Leeuwen for ber of significant differences as well. First, unlike CD28Ϫ cells ϩ advice and reagents concerning CMV. Furthermore, we are greatly in- from rheumatoid arthritis patients (16), KIR CD4 cells from debted to Dr. E. Goulmy for cord blood samples. Finally, we are grateful healthy donors did not proliferate in response to autologous to Drs. Bart O. Roep, Thorbald van Hall, and Rachel Allen for critically Ϫ ϩ PBMC. This was also true for CD28 KIR CD4 T cell clones reading this manuscript. isolated from a healthy donor (data not shown). Furthermore, KIRϩ CD4 T cells did not contain intracellular stores of perforin References (data not shown), contrary to rheumatoid arthritis patient KIRϩ 1. Trowsdale, J. 2001. Genetic and functional relationships between MHC and NK Ϫ receptor genes. Immunity 15:363. CD4 T cells (15). In addition, rheumatoid arthritis patient CD28 2. Vivier, E., and N. Anfossi. 2004. Inhibitory NK-cell receptors on T cells: witness CD4 cells are reportedly more likely to express KIR2DS2 than are of the past, actors of the future. Nat. Rev. Immunol. 4:190. CD28Ϫ CD4 cells from healthy controls (9). Finally, a substantial 3. Uhrberg, M., N. M. Valiante, B. P. Shum, H. G. Shilling, K. Lienert-Weidenbach, ϩ B. Corliss, D. Tyan, L. L. Lanier, and P. Parham. 1997. Human diversity in killer subset (40% on average) of KIR CD4 cells in healthy donors was cell inhibitory receptor genes. Immunity 7:753. CD28ϩ. Thus, KIRϩ CD28Ϫ CD4 T cells in rheumatoid arthritis 4. Hsu, K. C., S. Chida, D. E. Geraghty, and B. DuPont. 2002. The killer cell ϩ immunoglobulin-like receptor (KIR) genomic region: gene-order, haplotypes and patients may represent an expanded or modified subset of KIR allelic polymorphism. Immunol. Rev. 190:40. cells also present in healthy donors. 5. Garcia, C. A., J. Robinson, L. A. Guethlein, P. Parham, J. A. Madrigal, and Tumor- and virus-specific KIRϩ CD8 T cells have been de- S. G. Marsh. 2003. Human KIR sequences 2003. Immunogenetics 55:227. ϩ 6. Martin, M. P., X. Gao, J. H. Lee, G. W. Nelson, R. Detels, J. J. Goedert, scribed (38, 48, 49), but the Ag specificity of KIR CD4 T cells S. Buchbinder, K. Hoots, D. Vlahov, J. Trowsdale, et al. 2002. Epistatic inter- was unknown. We now show that at least some KIRϩ CD4 T cells action between KIR3DS1 and HLA-B delays the progression to AIDS. Nat. Genet. 31:429. proliferate in response to CMV Ags. This finding is in line with the 7. Martin, M. P., G. Nelson, J. H. Lee, F. Pellett, X. Gao, J. Wade, M. J. Wilson, fact that CMV-specific CD4 T cells have high precursor frequen- J. Trowsdale, D. Gladman, and M. Carrington. 2002. Cutting edge: susceptibility cies and tend to lack CD28 (50–52). However, the proliferative to psoriatic arthritis: influence of activating killer Ig-like receptor genes in the ϩ Ϫ absence of specific HLA-C alleles. J. Immunol. 169:2818. response of the KIR CD4 cells did not exceed that of the KIR 8. van der Slik, A. R., B. P. Koeleman, W. Verduijn, G. J. Bruining, B. O. Roep, and population, nor did their IFN-␥ production (data not shown). This M. J. Giphart. 2003. KIR in type 1 diabetes: disparate distribution of activating by guest on September 30, 2021. Copyright 2004 Pageant Media Ltd. was unexpected, as the percentage of memory cells in the KIRϪ and inhibitory natural killer cell receptors in patients versus HLA-matched con- ϩ trol subjects. Diabetes 52:2639. population was much lower than in KIR cells. One way to rec- 9. Yen, J. H., B. E. Moore, T. Nakajima, D. Scholl, D. J. Schaid, C. M. Weyand, and oncile these data would be to postulate that only certain pathogens J. J. Goronzy. 2001. Major histocompatibility complex class I-recognizing re- ϩ J. Exp. Med. 193:1159 induce KIR memory cells. Given the late stage memory pheno- ceptors are disease risk genes in rheumatoid arthritis. . ϩ 10. Battistini, L., G. Borsellino, G. Sawicki, F. Poccia, M. Salvetti, G. Ristori, and type of KIR CD4 T cells and their accumulation with age, it is C. F. Brosnan. 1997. Phenotypic and cytokine analysis of human peripheral blood conceivable that KIRϩ CD4 T cells arise through chronic exposure ␥␦ T cells expressing NK cell receptors. J. Immunol. 159:3723. 11. Mingari, M. C., F. Schiavetti, M. Ponte, C. Vitale, E. Maggi, S. Romagnani, to mostly latent pathogens, such as large DNA viruses like EBV, J. Demarest, G. Pantaleo, A. S. Fauci, and L. Moretta. 1996. Human CD8ϩ T CMV, and HSV, and mycobacteria. Recognition of viral Ags by lymphocyte subsets that express HLA class I-specific inhibitory receptors repre- CD4 T cells classically occurs indirectly via presentation in the sent oligoclonally or monoclonally expanded cell populations. Proc. Natl. Acad. https://www.jimmunol.org Sci. USA 93:12433. context of MHC class II. Alternatively, activating KIR may rec- 12. Anfossi, N., V. Pascal, E. Vivier, and S. Ugolini. 2001. Biology of T memory ognize viral Ags on the surface of infected target cells directly, type 1 cells. Immunol. Rev. 181:269. 13. Namekawa, T., M. R. Snyder, J. H. Yen, B. E. Goehring, P. J. Leibson, analogous to the recognition of a mouse CMV-encoded MHC class C. M. Weyand, and J. J. Goronzy. 2000. Killer cell activating receptors function I-like molecule by the activating NK receptor Ly49H (53, 54). as costimulatory molecules on CD4ϩCD28null T cells clonally expanded in rheu- This would not be detected in our assays because we used nonin- matoid arthritis. J. Immunol. 165:1138. ϩ 14. Nakajima, T., O. Goek, X. Zhang, S. L. Kopecky, R. L. Frye, J. J. Goronzy, and fectious CMV lysate to stimulate KIR CD4 T cells. Another clue C. M. Weyand. 2003. De novo expression of killer immunoglobulin-like recep- Downloaded from ϩ Ϫ to the Ag specificity of KIR CD4 T cells stems from the CD28 tors and signaling proteins regulates the cytotoxic function of CD4 T cells in is the CD57ϩ phenotype of the majority of these cells. Expanded acute coronary syndromes. Circ. Res. 93:106. 15. Namekawa, T., U. G. Wagner, J. J. Goronzy, and C. M. Weyand. 1998. Func- populations of such CD4 T cells have been reported in rheumatoid tional subsets of CD4 T cells in rheumatoid synovitis. Arthritis Rheum. 41:2108. arthritis (16), acute coronary syndromes (34), multiple sclerosis 16. Schmidt, D., J. J. Goronzy, and C. M. Weyand. 1996. CD4ϩ CD7Ϫ CD28Ϫ T cells are expanded in RA and are characterized by autoreactivity. J. Clin. Invest. (55), Crohn’s disease (56), and Wegener’s granulomatosis (57). It 97:2027 Ϫ . is tempting to suggest that KIR2DS2 on CD28 CD4 T cells plays 17. Martens, P. B., J. J. Goronzy, D. Schaid, and C. M. Weyand. 1997. Expansion of ϩ an important role in the recognition of CMV in health and disease. unusual CD4 T cells in severe rheumatoid arthritis. Arthritis Rheum. 40:1106. 18. Mandelboim, O., D. M. Davis, H. T. Reyburn, M. Vales-Gomez, E. G. Sheu, In short, KIR may regulate CD4 T cell responses to chronic patho- L. Pazmany, and J. L. Strominger. 1996. Enhancement of class II-restricted T cell gens as well as to endogenous Ags. responses by costimulatory NK receptors for class I MHC proteins. Science What is the function of KIR on CD4 T cells? Inhibitory KIR on 274:2097. 19. Phillips, J. H., J. E. Gumperz, P. Parham, and L. L. Lanier. 1995. Superantigen- CD4 cells may, in analogy with CD8 cells (39, 58, 59), enhance dependent, cell-mediated cytotoxicity inhibited by MHC class I receptors on T survival of memory T cells, possibly via increased resistance to lymphocytes. Science 268:403. 20. Warrington, K. J., S. Takemura, J. J. Goronzy, and C. M. Weyand. 2001. CD4ϩ, activation-induced cell death, and may also be involved in periph- CD28Ϫ T cells in RA patients combine features of the innate and adaptive im- eral tolerance by inactivating otherwise autoreactive cells. Acti- mune systems. Arthritis Rheum. 44:13. 6726 KIR EXPRESSING Th CELLS: EFFECTOR MEMORY Th1 CELLS

21. der Wiel, A. B., E. van Exel, A. J. De Craen, J. Gussekloo, A. M. Lagaay, 41. Goodridge, J. P., C. S. Witt, F. T. Christiansen, and H. S. Warren. 2003. D. L. Knook, and R. G. Westendorp. 2002. A high response is not essential to KIR2DL4 (CD158d) genotype influences expression and function in NK cells. prevent selection bias: results from the Leiden 85-plus study. J. Clin. Epidemiol. J. Immunol. 171:1768. 55:1119. 42. Aarvak, T., M. Chabaud, P. Miossec, and J. B. Natvig. 1999. IL-17 is produced 22. Weverling-Rijnsburger, A. W., I. J. Jonkers, E. van Exel, J. Gussekloo, and by some proinflammatory Th1/Th0 cells but not by Th2 cells. J. Immunol. R. G. Westendorp. 2003. High-density vs low-density lipoprotein cholesterol as 162:1246. the risk factor for coronary artery disease and stroke in old age. Arch. Intern. 43. Biassoni, R., A. Pessino, A. Malaspina, C. Cantoni, C. Bottino, S. Sivori, Med. 163:1549. L. Moretta, and A. Moretta. 1997. Role of amino acid position 70 in the binding 23. Rajagopalan, S., J. Fu, and E. O. Long. 2001. Cutting edge: induction of IFN-␥ affinity of p50.1 and p58.1 receptors for HLA-Cw4 molecules. Eur. J. Immunol. production but not cytotoxicity by the killer cell Ig-like receptor KIR2DL4 27:3095. (CD158d) in resting NK cells. J. Immunol. 167:1877. 44. Jabri, B., J. M. Selby, H. Negulescu, L. Lee, A. I. Roberts, A. Beavis, 24. Valiante, N. M., M. Uhrberg, H. G. Shilling, K. Lienert-Weidenbach, M. Lopez-Botet, E. C. Ebert, and R. J. Winchester. 2002. TCR specificity dictates K. L. Arnett, A. D’Andrea, J. H. Phillips, L. L. Lanier, and P. Parham. 1997. CD94/NKG2A expression by human CTL. Immunity 17:487. Functionally and structurally distinct NK cell receptor repertoires in the periph- 45. Sallusto, F., D. Lenig, R. Forster, M. Lipp, and A. Lanzavecchia. 1999. Two eral blood of two human donors. Immunity 7:739. subsets of memory T lymphocytes with distinct homing potentials and effector 25. Belkin, D., M. Torkar, C. Chang, R. Barten, M. Tolaini, A. Haude, R. Allen, functions. Nature 401:708. M. J. Wilson, D. Kioussis, and J. Trowsdale. 2003. Killer cell Ig-like receptor and 46. Sprent, J., X. Zhang, S. Sun, and D. Tough. 1999. T-cell turnover in vivo and the leukocyte Ig-like receptor transgenic mice exhibit tissue- and cell-specific trans- role of cytokines. Immunol. Lett. 65:21. gene expression. J. Immunol. 171:3056. 47. Park, W., C. M. Weyand, D. Schmidt, and J. J. Goronzy. 1997. Co-stimulatory 26. Heemskerk, M. H., M. Hoogeboom, R. A. de Paus, M. G. Kester, pathways controlling activation and peripheral tolerance of human CD4ϩCD28Ϫ M. A. van der Hoorn, E. Goulmy, R. Willemze, and J. H. Falkenburg. 2003. cells. Eur. J. Immunol. 27:1082. Redirection of antileukemic reactivity of peripheral T lymphocytes using gene 48. Bakker, A. B., J. H. Phillips, C. G. Figdor, and L. L. Lanier. 1998. Killer cell transfer of minor histocompatibility antigen HA-2-specific T-cell receptor com- inhibitory receptors for MHC class I molecules regulate lysis of melanoma cells ␣ plexes expressing a conserved joining region. Blood 102:3530. mediated by NK cells, ␥␦ T cells, and antigen-specific CTL. J. Immunol. 27. Geluk, A., W. Bloemhoff, R. R. de Vries, and T. H. Ottenhoff. 1992. Binding of 160:5239. a major T cell epitope of Mycobacteria to a specific pocket within HLA- 49. Ikeda, H., B. Lethe, F. Lehmann, N. van Baren, J. F. Baurain, C. de Smet, DRw17(DR3) molecules. Eur. J. Immunol. 22:107. ϩ H. Chambost, M. Vitale, A. Moretta, T. Boon, and P. G. Coulie. 1997. Charac- 28. Sakaguchi, S. 2004. Naturally arising CD4 regulatory t cells for immunologic terization of an antigen that is recognized on a melanoma showing partial HLA self-tolerance and negative control of immune responses. Annu. Rev. Immunol. loss by CTL expressing an NK inhibitory receptor. Immunity 6:199. 22:531. 50. Kern, F., T. Bunde, N. Faulhaber, F. Kiecker, E. Khatamzas, I. M. Rudawski, 29. Kronenberg, M., and L. Gapin. 2002. The unconventional lifestyle of NKT cells. A. Pruss, J. W. Gratama, R. Volkmer-Engert, R. Ewert, et al. 2002. Cytomega- Nat. Rev. Immunol. 2:557. lovirus (CMV) phosphoprotein 65 makes a large contribution to shaping the T 30. Boles, K. S., S. E. Stepp, M. Bennett, V. Kumar, and P. A. Mathew. 2001. 2B4 cell repertoire in CMV-exposed individuals. J. Infect. Dis. 185:1709. (CD244) and CS1: novel members of the CD2 subset of the immunoglobulin 51. Yue, F. Y., C. M. Kovacs, R. C. Dimayuga, P. Parks, and M. A. Ostrowski. 2004. superfamily molecules expressed on natural killer cells and other leukocytes. ϩ HIV-1-specific memory CD4 T cells are phenotypically less mature than cyto- Immunol. Rev. 181:234. ϩ megalovirus-specific memory CD4 T cells. J. Immunol. 172:2476. 31. Sallusto, F., and A. Lanzavecchia. 2000. Understanding dendritic cell and T- lymphocyte traffic through the analysis of chemokine receptor expression. Im- 52. van Leeuwen, E. M., E. B. Remmerswaal, M. T. Vossen, A. T. Rowshani, P. M. Wertheim-van Dillen, R. A. van Lier, and I. J. ten Berge. 2004. Emergence munol. Rev. 177:134. ϩ Ϫ ϩ 32. Kim, C. H., E. J. Kunkel, J. Boisvert, B. Johnston, J. J. Campbell, of a CD4 CD28 granzyme B , cytomegalovirus-specific T cell subset after M. C. Genovese, H. B. Greenberg, and E. C. Butcher. 2001. Bonzo/CXCR6 recovery of primary cytomegalovirus infection. J. Immunol. 173:1834. expression defines type 1-polarized T-cell subsets with extralymphoid tissue 53. Arase, H., E. S. Mocarski, A. E. Campbell, A. B. Hill, and L. L. Lanier. 2002. homing potential. J. Clin. Invest. 107:595. Direct recognition of cytomegalovirus by activating and inhibitory NK cell re- 33. Shilling, H. G., N. Young, L. A. Guethlein, N. W. Cheng, C. M. Gardiner, ceptors. Science 296:1323. D. Tyan, and P. Parham. 2002. Genetic control of human NK cell repertoire. 54. Smith, H. R., J. W. Heusel, I. K. Mehta, S. Kim, B. G. Dorner, O. V. Naidenko, J. Immunol. 169:239. K. Iizuka, H. Furukawa, D. L. Beckman, J. T. Pingel, et al. 2002. Recognition of 34. Liuzzo, G., S. L. Kopecky, R. L. Frye, W. M. O’Fallon, A. Maseri, J. J. Goronzy, a virus-encoded ligand by a natural killer cell activation receptor. Proc. Natl. and C. M. Weyand. 1999. Perturbation of the T-cell repertoire in patients with Acad. Sci. USA 99:8826.

by guest on September 30, 2021. Copyright 2004 Pageant Media Ltd. 55. Markovic-Plese, S., I. Cortese, K. P. Wandinger, H. F. McFarland, and R. Martin. unstable angina. Circulation 100:2135. ϩ Ϫ 35. Gardiner, C. M., L. A. Guethlein, H. G. Shilling, M. Pando, W. H. Carr, 2001. CD4 CD28 costimulation-independent T cells in multiple sclerosis. R. Rajalingam, C. Vilches, and P. Parham. 2001. Different NK cell surface phe- J. Clin. Invest. 108:1185. notypes deﬁned by the DX9 antibody are due to KIR3DL1 gene polymorphism. 56. De Tena, J. G., L. Manzano, J. C. Leal, E. S. Antonio, V. Sualdea, and J. Immunol. 166:2992. M. Alvarez-Mon. 2004. Active Crohn’s disease patients show a distinctive ex- ϩ ϩ null 36. Rajagopalan, S., and E. O. Long. 1999. A human histocompatibility leukocyte pansion of circulating memory CD4 CD45RO CD28 T Cells. J. Clin. Im- antigen (HLA)-G-speciﬁc receptor expressed on all natural killer cells. J. Exp. munol. 24:185. Med. 189:1093. 57. Lamprecht, P., H. Bruhl, A. Erdmann, K. Holl-Ulrich, E. Csernok, U. Seitzer, 37. Uhrberg, M., N. M. Valiante, N. T. Young, L. L. Lanier, J. H. Phillips, and M. Mack, A. C. Feller, E. Reinhold-Keller, W. L. Gross, and A. Muller. 2003. ϩ Ϫ P. Parham. 2001. The repertoire of killer cell Ig-like receptor and CD94:NKG2A Differences in CCR5 expression on peripheral blood CD4 CD28 T-cells and in receptors in T cells: clones sharing identical ␣␤TCR rearrangement express granulomatous lesions between localized and generalized Wegener’s granuloma- highly diverse killer cell Ig-like receptor patterns. J. Immunol. 166:3923. tosis. Clin. Immunol. 108:1. https://www.jimmunol.org 38. Vely, F., M. Peyrat, C. Couedel, J. Morcet, F. Halary, F. Davodeau, F. Romagne, 58. Ugolini, S., C. Arpin, N. Anfossi, T. Walzer, A. Cambiaggi, R. Forster, M. Lipp, E. Scotet, X. Saulquin, E. Houssaint, et al. 2001. Regulation of inhibitory and R. E. Toes, C. J. Melief, J. Marvel, and E. Vivier. 2001. Involvement of inhibitory ϩ activating killer-cell Ig-like receptor expression occurs in T cells after termination NKRs in the survival of a subset of memory-phenotype CD8 T cells. Nat. of TCR rearrangements. J. Immunol. 166:2487. Immunol. 2:430. 39. Young, N. T., M. Uhrberg, J. H. Phillips, L. L. Lanier, and P. Parham. 2001. 59. Mingari, M. C., A. Moretta, and L. Moretta. 1998. Regulation of KIR expression Differential expression of leukocyte receptor complex-encoded Ig-like receptors in human T cells: a safety mechanism that may impair protective T-cell re- correlates with the transition from effector to memory CTL. J. Immunol. sponses. Immunol. Today 19:153. 166:3933. 60. Mandelboim, O., S. Kent, D. M. Davis, S. B. Wilson, T. Okazaki, R. Jackson,

Downloaded from 40. Kikuchi-Maki, A., S. Yusa, T. L. Catina, and K. S. Campbell. 2003. KIR2DL4 is D. Haﬂer, and J. L. Strominger. 1998. Natural killer activating receptors trigger an IL-2-regulated NK cell receptor that exhibits limited expression in humans but interferon ␥ secretion from T cells and natural killer cells. Proc. Natl. Acad. Sci. triggers strong IFN-␥ production. J. Immunol. 171:3415. USA 95:3798.