Identification of CLEC12B, an Inhibitory Receptor on Myeloid Cells*

Activation of immune cells has to be tightly controlled to prevent detrimental hyperactivation. In this regulatory process molecules of the C-type lectin-like family play a central role. Here we describe a new member of this family, CLEC12B. The extracellular domain of CLEC12B shows considerable homology to the activating natural killer cell receptor NKG2D, but unlike NKG2D, CLEC12B contains an immunoreceptor tyrosine-based inhibition motif in its intracellular domain. Despite the homology, CLEC12B does not appear to bind NKG2D ligands and therefore does not represent the inhibitory counterpart of NKG2D. However, CLEC12B has the ability to counteract NKG2D-mediated signaling, and we show that this function is dependent on the immunoreceptor tyrosine-based inhibition motif and the recruitment of the phosphatases SHP-1 and SHP-2. Using monoclonal anti-CLEC12B antibodies we found de novo expression of this receptor on in vitro generated human macrophages and on the human myelo-monocytic cell line U937 upon phorbol 12-myristate 13-acetate treatment, suggesting that this receptor plays a role in myeloid cell function.

Activation of the immune system has to be tightly regulated to prevent the attack of self-antigens and the development of autoimmunity (1). This is in part accomplished by a sophisticated system of various receptors that create a delicate balance of activating and inhibitory signals (2,3). These receptors are present on different immune cells but play a major role for the innate immune system.
Many receptors share common signaling motifs to transmit their signal into the cell. Activating receptors usually signal via the immunoreceptor tyrosine-based activation motif containing tyrosine residues that can be phosphorylated upon receptor engagement and can recruit Syk family kinases (4). Inhibitory receptors often possess an Immunoreceptor Tyrosine-based Inhibition Motif (ITIM), 4 which can recruit phosphatases upon phosphorylation (5). These phosphatases, like SHP-1 or SHP-2, are able to de-phosphorylate and therefore inhibit intracellular factors that otherwise would promote cellular activation (6).
Interestingly, many activating and inhibitory receptors come in pairs, with a strong homology of the extracellular domains and partially overlapping ligand specificity but opposite signaling capacities (2). One reason for the existence of these receptor pairs might be to avoid extreme immune reactions to minor threats. Examples for these paired receptors are killer cell Iglike receptors on natural killer (NK) cells, which comprise inhibitory and activating receptors, both recognizing the same major histocompatibility complex class I ligand (5,7). Also in the family of lectin-like receptors can be found antithetic pairs that are specific for the same ligand but transmit opposing signals (8). For example, the NKG2A-CD94 heterodimer transmits an inhibitory signal and NKG2C-CD94 signals in an activating fashion while both receptors recognize the same ligand, HLA-E.
NKG2D is a C-type lectin-like receptor and plays an important role for the activation of natural killer cells (9). NKG2D has no intracellular signaling domain but instead pairs with the adaptor molecules DAP10 and DAP12 to mediate an activating signal. So far, no inhibitory counterpart for NKG2D has been described. Therefore we conducted a data base search for molecules with a high homology to the NKG2D extracellular domain. With this approach we identified a novel cell surface receptor encoded within the same chromosomal region as NKG2D.
Here we describe the molecular cloning, biochemical and initial functional characterization of CLEC12B, a new inhibitory receptor of the lectin-like family.

EXPERIMENTAL PROCEDURES
Cells and Antibodies-Cell lines used in this study were human embryonic kidney 293T, Ba/F3 transduced with NKG2D ligands, P815, and NKL. Human monocytes were isolated from peripheral blood mononuclear cells using anti-CD14-positive selection beads (Miltenyi) and used for the in vitro generation of macrophages as described (10). Antibodies used were anti-histidine tag (Serotec, Oxford, UK), anti-SHP-1, * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  anti-SHP-2 (BD Biosciences), anti-actin (Sigma), biotinylated anti-phosphotyrosine (4G10; Upstate, Charlottesville, VA), phycoerythrin-streptavidin, horseradish peroxidase-goat antimouse IgG, biotinylated goat anti-human IgG (Jackson ImmunoResearch, West Grove, PA), horseradish peroxidasestreptavidin (Amersham Biosciences), MOPC-21 (Sigma), and anti-NKG2D (R&D Systems, Minneapolis, MN). Mouse monoclonal anti-CLEC12B antibodies were produced by immunizing mice with Ba/F3 cells transfected with CLEC12B and boosting with purified CLEC12B-Ig fusion protein as described previously (11). Antibodies were tested for their specificity for CLEC12B by enzyme-linked immunosorbent assay and Western blot and fluorescence-activated cell sorter (FACS) analysis. Reverse Transcription PCR, cDNA Constructs, and Retroviral Transduction-PCR analysis of CLEC12B cDNA was performed on Marathon-Ready cDNA libraries (Clontech) using the primer pair cac tca cat ttc agg att ctg ctg gag caa g and cac atc ctt tgg atc cat tga tct ggt caa g followed by a second round of PCR using the primer pair caa ctg ggc aac tcc aac aac ttg tcc atg and gaa cag agc cat ctt ccc aga acc aac ttc tg. Full-length CLEC12B cDNA was amplified using the primer pair atg tct gaa gaa gtg acc tac gcg aca ctc aca ttt c and cta atc caa atc ctc agt ctt cac tgg ggc agc. CLEC12B cDNA linked to a His tag sequence was cloned into the retroviral expression vector pBABE puro CMVϩ containing a puromycin resistance gene. A tyrosine-tophenylalanine mutation within the ITIM motif was generated using site-directed mutagenesis. Retroviral transduction was done using the packaging cell line Phoenix-ampho (American Type Culture Collection, Manassas, VA). Transduced NKL and Ba/F3 were selected in puromycin, and NKL were subsequently subcloned to generate clones stably expressing CLEC12B. Expression was verified by flow cytometry.
Ig Fusion Proteins-The extracellular part of CLEC12B was cloned in-frame with the Fc part of human IgG. Proteins were produced by transient transfection of 293T cells and subsequently purified via protein A-agarose.
Differentiation of U937 Cells-U937 cells were plated in 6-well plates at 0.5 ϫ 10 6 cells/well and stimulated with 10 Ϫ8 M phorbol 12-myristate 13-acetate (PMA). Cells were harvested at different time points. Adherend cells were scraped off the plates and stained for flow cytometry with an anti-CLEC12B antibody.
Cytotoxicity Assay-P815 cells were grown to mid-log phase, and 5 ϫ 10 5 cells were labeled in 100 l of CTL medium (Iscove's modified Dulbecco's medium with 10% fetal calf serum and 1% penicillin/streptomycin) with 100 Ci of (3.7 MBq) 51 Cr for 1 h at 37°C. Cells were washed twice with CTL medium and resuspended at 5 ϫ 10 4 cells/ml in CTL medium. Effector cells were resuspended in CTL medium and preincubated with antibodies (0.5 g/ml final concentration) for 15 min at 25°C. After preincubation, effector cells were mixed with 5,000 labeled target cells/well in a V-bottom 96-well plate. Maximum release was determined by incubation of target cells in 1% Triton X-100 solution. For spontaneous release, targets were incubated without effectors in CTL medium alone. All samples were done in triplicates. After a 1-min centrifugation at 220 ϫ g, plates were incubated for 4 h at 37°C. Supernatant was harvested, and 51 Cr release was measured in a ␥-counter. Per-cent specific release was calculated as ([experimental release Ϫ spontaneous release]/[maximum release Ϫ spontaneous release]) ϫ 100. The ratio between maximum and spontaneous release was at least 4 in all experiments.
Transfection of 293T Cells-293T cells (0.5 ϫ 10 6 /well) were grown for 24 h in a 6-well plate and transfected by calcium phosphate with 3 g/well of total DNA (12). Cells were harvested 48 h after transfection and treated either with pervanadate or phosphate-buffered saline for 10 min at 37°C. Cells were lysed in ice-cold Triton X-100 lysis buffer for 20 min on ice, and lysate was cleared by centrifugation (10,000 ϫ g, 4°C, 15 min).
Immunoprecipitation and Western Blotting-For immunoprecipitation, precleared lysates (13) were first incubated for 1 h at 4°C with 2 g of control IgG1 followed by 2 g of specific antibody, both coupled to recombinant protein G-agarose (Invitrogen). Beads were washed three times in ice-cold lysis buffer and boiled in 2.5ϫ reducing (5% SDS, 25% glycerin, 12.5% 2-mercaptoethanol, 0.156 M Tris-HCl, pH 6.8, 0.01% Bromphenol blue) or nonreducing (no 2-mercaptoethanol) sample buffer. For Western blotting, samples were separated on 10% NuPAGE gels (Invitrogen) and transferred to a polyvinylidene difluoride membrane (Roth, Karlsruhe, Germany). The membrane was blocked in phosphate-buffered saline with 5% milk powder for 1 h at room temperature and incubated with the indicated antibody overnight at 4°C. After washing, the membrane was incubated with the respective horseradish peroxidase-conjugated secondary reagent for 1 h at room temperature and developed using Super Signal West Pico (Pierce).
FACS Staining-1 ϫ 10 5 to 5 ϫ 10 5 cells were resuspended in FACS buffer (phosphate-buffered saline, 2% fetal calf serum) containing the appropriate antibody (10 g/ml). After a 20-min incubation on ice, cells were washed in FACS buffer and stained with phycoerythrin-conjugated secondary antibody (1:200). Cells were subsequently washed and resuspended in FACS buffer containing 2% formaldehyde and analyzed using flow cytometry. Staining with Ig fusion proteins was conducted similarly. Cells were diluted in FACS buffer containing the respective Ig fusion protein (10 g/ml). After incubating on ice, cells were washed and stained with biotinylated goat anti-human IgG antibody. Then cells were incubated with phycoerythrinconjugated streptavidin and analyzed.

RESULTS
In an attempt to identify NKG2D-related molecules, we identified a mouse cDNA clone (AK016908) coding for an unnamed C-type lectin protein product of 275 amino acids (BAB30491). Using this sequence as bait, we identified a related human sequence in the data base designated human macrophage antigen-H (AY358810). Comparing the sequences, we found that human macrophage antigen-H (232 amino acids) was missing exon 6, resulting in a premature stop codon and an incomplete C-type lectin domain. Searching the genomic DNA data base, we identified the missing exon 6, which showed homology to the murine sequence. Using primers spanning exons 1 to 6 we identified the complete cDNA coding for a 276-amino acid type II transmembrane protein (DQ368812) belonging to the C-type lectin family of recep-tors (14,15). The extracellular domain of the putative protein showed highest homology to human NKG2D (36% similarity). However, unlike NKG2D it did not contain a charged amino acid in its transmembrane region but an ITIM sequence (VTYATL) in the cytoplasmic tail (Fig. 1A).
The genetic locus of this receptor was mapped to human chromosome 12p13.2 and mouse chromosome 6qF3 in the vicinity of other C-type lectin-like receptors (Fig.  1B). Overall, the receptor showed a high similarity to CLEC12A (16 -18) (34% similarity), which was located on the same chromosomal locus. We therefore named this new receptor CLEC12B. Using reverse transcription PCR we found expression of CLEC12B in human cDNA libraries prepared from various tissues except brain (Fig. 1C). Interestingly, in mammary gland and ovary we only found a truncated splice variant lacking exon 4. As this exon encodes part of the C-type lectinlike domain, this alternative transcript would yield a non-functional protein.
Because of its high similarity to the NKG2D extracellular domain we speculated that this receptor could be an inhibitory counterpart of NKG2D. Similar to NKG2D, surface expression of this receptor was independent of CD94 expression and the receptor formed disulfidelinked homodimers in transfected cells (data not shown). We produced an Ig fusion protein of the extracellular domain of human CLEC12B to test whether it might recognize the same ligands as NKG2D. Ba/F3 cells stably expressing the human NKG2D ligands MICA, MICB, ULBP1, ULBP2, or ULBP3 were stained at high levels with a NKG2D-Ig fusion protein. However, we detected no binding of the Ig fusion protein containing the extracellular domain of human CLEC12B (Fig. 2), demonstrating no apparent overlap in ligand specificity with NKG2D.
To examine CLEC12B expression and biochemical properties, we produced mouse monoclonal antibodies against human CLEC12B. We selected three subclones that specifically recognized the receptor in transfected cells either in Western blot (clones 12 and 16, Fig. 3A and data not shown) or in flow cytometry (clones 30 and 16, Fig. 3B and data not shown). CLEC12B was detected as a double band in Western blot analysis of transfected cells, possibly due to differential glycosylation of the  (1), colon (2), heart (3), kidney (4), liver (5), lung (6), mammary gland (7), ovary (8), spleen (9), testis (10), or water control (11). The first round of amplification was performed using primers spanning exons 1 to 6. The resulting product was further amplified using primers spanning exons 3 to 5. The expected amplicon size is 384 bp (upper band). The lower band (225 bp) represents a splice variant lacking exon 4. Lower panel, actin control. receptor. Using the antibody clone 30 in flow cytometry, we detected strong expression of CLEC12B on the human promyelocytic cell line U937 after PMA stimulation but not in nonstimulated cells (Fig. 4A). The induction of CLEC12B protein expression was confirmed by Western blot analysis using another anti-CLEC12B-specific mAb (clone 12) (Fig. 4B). Interestingly, the endogenous CLEC12B protein was detected as a single band as compared with the double band of the transfected protein. The reason for this difference is unknown so far. This indicates that CLEC12B can be expressed on monocytic cells after a cell differentiation-inducing stimulus. To further characterize this we purified CD14-positive monocytes from human peripheral blood mononuclear cells. Although we detected no surface expression on freshly isolated monocytes or any other leukocyte population (data not shown), we observed the surface expression of CLEC12B on in vitro differentiated human macrophages (Fig. 4C). This expression was confirmed by Western blot analysis using three independent anti-CLEC12B mAbs (Fig. 4D and data not shown).
To characterize receptor function, we generated a mutant of human CLEC12B by exchanging the ITIM tyrosine with phenylalanine. The His-tagged wild type and the mutant receptors were stably expressed in the NK cell line NKL (Fig. 5, A and B). Triggering the NKG2D receptor in NKL cells in a redirected lysis assay resulted in efficient NK cell cytotoxicity, which was inhibited by co-triggering the CLEC12B WT (Fig. 5C) but not the mutant receptor (Fig. 5D). This demonstrates that CLEC12B can effectively function as an inhibitory receptor and that this activity depends on the ITIM sequence.
ITIM motifs usually couple to phosphatases to promote inhibitory signaling (19). We therefore transfected 293T cells with His-tagged WT or mutant human CLEC12B in combination with SHP-1 or SHP-2. Transfected cells were pervanadatetreated and CLEC12B was immunoprecipitated. Pervanadate treatment induced detectable tyrosine phosphorylation only   in the WT receptor. Interestingly, the lower band of the transfected CLEC12B detected by Western blotting was preferentially phosphorylated (Fig. 6, A and B). Upon phosphorylation the WT, but not the mutant receptor, could recruit SHP-1 and SHP-2 as detected by co-immunoprecipitation (Fig. 6, A and B). This indicates that the inhibitory function of CLEC12B can be mediated by the recruitment of SHP-1 or SHP-2 to its ITIM upon receptor phosphorylation.

DISCUSSION
The cDNA sequence for CLEC12B was found during a data base search for sequences with homology to NKG2D. Because of the high homology of CLEC12B's extracellular domain with NKG2D and its ITIM motif, this new lectin-like receptor appeared to be a potential inhibitory counterpart. Yet binding of CLEC12B Ig fusion proteins to NKG2D ligands could not be detected. Additionally we did not find any expression of CLEC12B on primary T cells, NK cells, or on NK cell lines. This makes it very improbable that CLEC12B is an inhibitory counterpart to NKG2D, at least with respect to expression and competitive binding to common NK cell ligands.
Functional analysis of full-length CLEC12B revealed that this receptor is able to signal in an inhibiting fashion via recruitment of the phosphatases SHP-1 and SHP-2. Although CLEC12B triggering could inhibit NKG2D-mediated NK cell activation (Fig. 4), its inhibitory activity was not limited to NKG2D. Other activating NK cell receptors such as 2B4 (20) were sensitive to the inhibitory effect of CLEC12B (data not shown), demonstrating that CLEC12B can inhibit a broad range of receptormediated activating signals. In preliminary experiments we also investigated the function of endogenous CLEC12B on PMAstimulated U937 cells. Stimulating these cells with lipopolysaccharide induced the production of tumor necrosis factor ␣. However, engagement of CLEC12B by plate-bound anti-CLEC12B antibody did not significantly inhibit lipopolysaccharide-induced tumor necrosis factor ␣ production (data not shown). Plate-bound antibodies only stimulate their respective receptor at the site of contact between the cell and the plastic dish. In contrast, lipopolysaccharide is a soluble mediator and can therefore exert its effect over the entire surface of the cell. It is therefore likely that the local inhibitory effect of CLEC12B engagement was not sufficient to counteract the more global lipopolysaccharide signal. This demonstrates that CLEC12B is only potent in blocking cellular activation when it is co-engaged with an activating receptor in a defined region of the plasma membrane of a cell.
We found CLEC12B mRNA expression in almost all tissues tested. In mammary gland and ovary only a truncated variant of CLEC12B lacking exon 4 could be detected. As this exon encodes a part of the extracellular domain essential for formation of a functional C-type lectin-like domain, the resulting receptor will be nonfunctional. It will be interesting to investigate the regulation of this differential splicing of CLEC12B in different tissues.
We found protein expression of CLEC12B on in vitro differentiated macrophages and on PMA-stimulated U937 cells. We did not detect CLEC12B expression on in vitro generated dendritic cells (generated from CD14 ϩ cells with granulocyte macrophage colony-stimulating factor and interleukin 4 for 6 days) or on any freshly isolated peripheral blood leukocyte population (data not shown). The expression on macrophages could explain why we found CLEC12B expression in various tissues using reverse transcription PCR, as macrophages and related immune cells are spread throughout the body.
CLEC12B may be involved in limiting the activity of monocyte-derived immune cells after cell differentiation and possibly during inflammatory diseases. It will therefore be interesting to investigate CLEC12B expression during various infectious or chronic diseases. Together with the identification of its ligand, this might give more insight into the biological function of this novel receptor.