A Novel LPS-Inducible C-Type Lectin Is a Transcriptional Target of NF-IL6 in Macrophages¹

Protein-carbohydrate interactions serve multiple functions in the immune system. A number of animal lectins (sugar-binding proteins) mediate both pathogen recognition and cell-cell interactions using structurally related Ca²⁺-dependent carbohydrate-recognition domains (C-type CRDs)³ (1). Macrophages express several types of C-type lectins, such as macrophage mannnose receptor and macrophage asialoglycoprotein binding protein (M-ASGP-BP). Macrophage mannose receptor mediates first-line defense against pathogens by direct phagocytosis (2). The mannose receptor binds sugars such as mannose, N-acetylglucosamine, and fucose that are not common in terminal positions on mammalian oligosaccharides but are frequently found on the surfaces of microorganisms. M-ASGP-BP is expressed on the surface of activated macrophages, recognizes terminal galactose/N-acetylgalactosamine units, and may participate in the interaction between tumoricidal macrophages and tumor cells (3, 4, 5).

Here, we report a novel C-type lectin, macrophage-inducible C-type lectin (Mincle), which is a transcriptional target of NF-IL6 in peritoneal macrophages (PMφ). NF-IL6 was initially identified as a nuclear factor that binds to the IL-1 responsive element in the IL-6 gene (6). Cloned NF-IL6 exhibits homology with CCAAT/enhancer binding protein (C/EBP), a member of the basic leucine zipper family of transcription factors (7). NF-IL6 has been reported by other groups under the names AGP/EBP, LAP, IL-6DBP, rNFIL-6, C/EBPβ, and CRP2 (8, 9, 10, 11, 12, 13). NF-IL6 exhibits a low transcriptional activity unless activated by inflammatory stimuli, which induce phosphorylation of NF-IL6 and augment its transcriptional activity (14, 15, 16). NF-IL6 appears to play an important role in activated macrophages (17). NF-IL6-deficient (−/−) mice are extremely susceptible to infections with microorganisms such as Listeria monocytogenes and Candida albicans (18, 19). PMφ from NF-IL6 (−/−) mice were defective in intracellular killing of L. monocytogenes and displayed impaired tumoricidal and tumoristatic activity. The macrophages used for these studies were able to produce normal amounts of NO, which is thought to play an important role in the elimination of intracellular bacteria and parasites, thus suggesting that a NO-independent, NF-IL6-dependent pathway may be involved in Listeria killing and tumoricidal activity (18).

Overexpression of NF-IL6 protein have established that IL-6, macrophage chemoattractant protein-1 (MCP-1), macrophage inflammatory protein-1α (MIP-1α), MIP-1β, osteopontin, CD14, and lysozyme are downstream genes of NF-IL6 in several hemopoietic cell lines (20, 21, 22). However, expressions of these genes by NF-IL6 (−/−) macrophages are unexpectedly comparable to those observed by wild-type (WT) macrophages probably because other C/EBP families, such as C/EBPδ, could in part compensate for the lack of NF-IL6 in vivo (23). So far, the transcriptional targets of NF-IL6 have been unsuccessfully defined except G-CSF in macrophages (18). Identification of other targets is essential to clarify the NF-IL6 function in more detail.

LPS (Escherichia coli O55: B5) and thioglycolate broth (Brewer’s formula) were purchased by Difco (Detroit MI). IFN-γ, GM-CSF, IL-6, and TNF-α were obtained from Genzyme (Cambridge, MA). Restriction and DNA modification enzymes were products of Toyobo (Otsu, Japan). [α-³²P]dCTP (3000Ci/mmol) and [γ-³²P]ATP (3000Ci/mmol) were obtained from Amersham Pharmacia Biotech (Little Chalfont, U.K.).

The NF-IL6 (−/−) mice generated by homologous recombination have been described previously (18). NF-IL6 (−/−) and WT mice were littermates derived from intercrossing hemizygous females and males.

PMφ were collected by peritoneal lavage with PBS at 4 days after i.p. injecting 2 ml of sterile thioglycolate into 8- to 12 wk-old mice. PMφ were plated on 10-cm plastic dishes at 2.5 × 10⁶ cells/dish in macrophage culture medium (MCM). MCM consists of DMEM (Nissui, Tokyo, Japan) supplemented with 10% FBS, 2 mM l-glutamine, 50 U/ml of GM-CSF, 100 μg/ml streptomycin, and 10 U/ml penicillin G. After 2 h incubation to allow for adherence of macrophages, the dishes were washed vigorously to remove nonadherent cells. Fresh MCM was added on day 2, and fresh MCM without GM-CSF was added on day 4 of culture. PMφ were treated with appropriate reagents on day 5 for harvesting RNA.

Murine B cell leukemia (BCL-1), myeloma cells (MOPC 315), thymoma cells (EL-4), and human monocytic leukemia cells (THP-1) were cultured in RPMI 1640 medium (Life Technologies, Gaithersburg, MD) containing 10% FBS, 100 μg/ml streptomycin, and 10 U/ml penicillin G. Murine NK cells (5E3) were cultured in RPMI 1640 supplemented with 10% FBS and 500 U/ml of IL-2 (Genzyme). NF-IL6 M1 cells were cultured in MEM (Life Technologies) supplemented with twice the normal concentration of amino acids and vitamins, 10% FBS, 400 μg/ml of geneticin (Life Technologies), and 50 μg/ml of hygromycin B (Boehringer Mannheim, Mannheim, Germany) (21). Murine macrophage cells (RAW264.7), embryonic fibroblast cells (NIH3T3), and human embryonic kidney cells (293T) were cultured in DMEM supplemented with 10% FBS, 100 μg/ml streptomycin, and 10 U/ml penicillin G.

WT and NF-IL6 (−/−) PMφ at about 80% confluence were treated with 100 ng/ml LPS and 100 U/ml IFN-γ for 16 h. Total RNA was prepared by RNAeasy kit (Qiagen, Hilden, Germany) following poly(A)⁺ RNA selection using Oligotex-dT30 Latex beads (TaKaRa, Otsu, Japan). Then all procedures were performed according to PCR-select cDNA subtraction kit (Clontech, Palo Alto, CA).

RACE (5′ and 3′) were performed using a Marathon cDNA amplification kit (Clontech). Using 1 μg of poly(A)⁺ RNA from WT PMφ stimulated with 100 ng/ml LPS and 100 U/ml IFN-γ, a library of adaptor-ligated double-stranded cDNA was constructed as described by the manufacturer’s instruction. To obtain a full-length cDNA of mouse Mincle (mMincle), an antisense primer, 270F (5′-GAGAAAATGGGGCTCCAGGAAGAGTG-3′) and a sense primer, 92R (5′-CCCTAAAGGAACCTTCAGCAGCAGTC-3′) were designed from the sequence of the cDNA fragment obtained by subtraction cloning for 5′-RACE and 3′-RACE, respectively.

To obtain human Mincle (hMincle), cDNA synthesized from THP-1 cells mRNA extracted after LPS stimulation (5 μg/ml) was subjected to PCR using degenerate primers. The 5′ primer (5′-GTGAGGCATCAGGTbTCAG-3′) and 3′ primer (5′-DATRTTGTTGGGYTCNCC-3′) were designed to cover a portion of mMincle CRD. PCR conditions were 94°C for 30 s, followed by 35 cycles of 94°C for 5 s and 50°C for 30 s. The resulting 311-bp PCR product was subcloned into pT7Blue T vector (Novagen, Madison, WI) and sequenced. An antisense primer (5′-CCCAGTTCAATGGACAATTCTTG-3′) and a sense primer (5′-ACGGCACACCTTTGACAAAGTCTCTG-3′) were designed from the sequence and 5′- and 3′- RACE were performed using an adaptor-ligated double-stranded cDNA prepared from LPS-stimulated THP-1 cells.

Total RNA (5 μg/lane) was separated on 1.0% agarose gels containing 6.0% formaldehyde. After transfer to a HybondN⁺ membrane (Amersham Pharmacia Biotech), hybridization and wash were performed as described previously (21). RsaI-RsaI cDNA fragment of mMincle (nucleotide 1188–1404), SphI-PstI cDNA fragment of mouse NF-IL6 (X62600; nucleotide 135–897), cDNA fragment of mouse MIP-2 (X53798; nucleotide 149–840), and cDNA fragment of mouse G3PDH (M32599; nucleotide 566-1017) were used as probes. cDNA fragments were radiolabeled with [α-³²P]dCTP (3000 Ci/mmol) by use of the Megaprime DNA labeling system (Amersham Pharmacia Biotech).

Interspecific back-cross progeny were generated by mating (C57BL/6J × Mus spretus)F₁ females and C57BL/6J males as described (24). A total of 205 N₂ mice were used to map the Mincle locus. DNA isolation, restriction enzyme digestion, agarose gel electrophoresis, Southern blot transfer, and hybridization were performed essentially as described (25). All blots were prepared with HybondN⁺ membrane. The mMincle cDNA fragment (nucleotides 1549–2517) was labeled with [α³²P]dCTP using a nick translation labeling kit (Boehringer Mannheim); washing was done to a final stringency of 1.0× SSCP, 0.1% SDS, 65°C. A fragment of 3.7 kb was detected in BamHI-digested C57BL/6J DNA and a fragment of 6.5 kb was detected in BamHI-digested M. spretus DNA. The presence or absence of the 6.5-kb BamHI M. spretus-specific fragment was followed in back-cross mice.

A description of the probes and RFLPs for the loci linked to Mincle including Atp6e, Slc2a3, and Cd4 has been reported previously (26). Recombination distances were calculated using Map Manager, version 2.6.5 (http://mcbio.med.buffalo.edu/mapmgr.html). Gene order was determined by minimizing the number of recombination events required to explain the allele distribution patterns.

The mMincle expression vector was constructed in pcDNA3.1(+) (Invitrogen, Carlsbad, CA). To construct pcDNA3.1(+)-mMincleFlag, forward and reverse primers were devised to introduce an optimal Kozak consensus sequence and Flag epitope (DYKDDDDK), respectively. The forward primer sequence was 5′-GGTCGACCACCATGAATTCAACCAAATCG-3′ and the reverse primer sequence was 5′-CTCACTTGTCATCGTCGTCCTTGTAGTCCAGAGGACTTAT-3′. PCR was performed using double-stranded cDNA generated in the course of RACE as template. Amplified product was verified by sequencing after subcloned into pT7Blue T vector and excised by Sal I digestion. Then mMincleFlag cDNA fragment was ligated to the XhoI site of pcDNA3.1(+) with correct orientation. The day before transfection, 293T cells were seeded on a 6-well plate at 2.0 × 10⁵ cells/well. Four micrograms of pcDNA3.1(+)-mMincleFlag or pcDNA3.1 empty vector was transiently transfected by calcium-phosphate precipitation method. Cells were freed from culture plates using 0.02% EDTA in PBS at 48 h following transfection and washed in flow cytometry buffer (PBS with 2% FBS and 0.1% NaN₃). Cells were incubated for 20 min on ice with 15 μg/ml biotin-conjugated anti-Flag M2 mAb (BioM2; Sigma-Aldrich, St. Louis, MO), washed in flow cytometry buffer, and labeled with 5 μg/ml FITC-streptavidin (PharMingen, San Diego, CA). Control consisted of cells treated with FITC-streptavidin alone. After a final wash in flow cytometry buffer, mMincle-Flag expression was analyzed using FACScalibur using CellQuest software (Becton Dickinson, Lincoln Park, NJ).

The day before transfection, 293T cells were seeded on a 100-mm plate at 2.0 × 10⁶ cells. Twenty micrograms of pcDNA3.1(+)-mMincleFlag or pcDNA3.1 empty vector was transiently transfected by calcium-phosphate precipitation method. Cells were lysed with lysis buffer containing 0.5% Nonidet P-40, 150 mM NaCl, 1 mM EDTA, and 10 mM Tris-HCl, pH 7.4. Cell lysates were immunoprecipitated with protein-G Sepharose (Amersham Pharmacia Biotech) together with 10 μg/ml anti-Flag M2 mAb (Sigma-Aldrich). Immunoprecipitates were washed four times with lysis buffer and suspended with Laemmli sample buffer. After boiling for 5 min, samples were separated on a gradient (10–20%) SDS-polyacrylamide gel and electrically transferred to a nitrocellulose membrane. The membrane was incubated with BioM2 mAb, subsequently treated with HRP-conjugated streptavidin (Genzyme), and then analyzed for immunoreactivity by the enhanced chemiluminescence detection system (DuPont, Boston, MA).

A 129/Sv mouse liver genomic DNA library in the λFix II vector was purchased from Stratagene (La Jolla, CA). Approximately 1 × 10⁶ plaques were screened with the ³²P-labeled EcoRI- EcoRI cDNA fragment of mMincle (nucleotides 126–496). Positive plaques from the genomic library were enriched after two further rounds of screening. DNA from two independent clones was purified using the Wizard prep kit (Promega, Madison, WI). Digestion with several restriction enzymes revealed that these two clones contain the identical sequence. Following digestion with BamHI and SalI, five resulting fragments (5.4 kb, 5.1 kb, 3.2 kb, 0.8 kb, and 0.6 kb) were subcloned into pBluescript KS (+) and sequenced. The 5.1kb BamHI-BamHI fragment containing exon I and the 5′-flanking region was designated pBS/Bam1-8 and employed for promoter-luciferase construction. Appropriate 5′ primers and a common 3′ primer were synthesized to amplify the promoter regions of mMincle. The following primers were used to generate pGL3-1783/+69, pGL3-1190/+69, pGL3-240/+69, and pGL3-61/+69; −1783 (5′-CGACGCGTGGTTTGCAGCCCCATAGGAG-3′), −1190 (5′-CGACGCGTATGATGGCACACCATGATAG-3′), −240 (5′-CGACGCGTAAATCGGGACCAAGTTAGAC-3′), −61 (5′-CGACGCGTCAAGAGAGGAAATTCTGAC-3′), and +69 (5′-GAAGATCTCCCCTGGAAAGTGAGTCTTG-3′). Amplified products were subcloned into pT7Blue T vector and sequenced for confirmation. The inserted fragments were cut out with Mlu I and BglII digestion and ligated to pGL3 basic vector (Promega) at the same restriction sites. To create pGL3-166/+69, pGL3-240/+69 was digested with Mlu I and Aat I, blunt-ended, and re-ligated. NF-IL6 binding site mutation was generated by Quick Change site-directed mutagenesis kit (Stratagene). The mutagenic primer (with altered nucleotides underlined) was 5′-CCTTGTCCTTGTGCCCCAGAGGAAATTCTG-3′. The mutated construct was confirmed by sequencing. Transient transfection into NF-IL6 M1 cells and luciferease assay were done as described previously (21).

Primer extension was performed as described (27). Briefly, an oligonucleotide primer complementary to nucleotides 81–112 of Mincle cDNA was synthesized and end-labeled with [γ-³²P]ATP and T4 polynucleotide kinase. Ten micrograms of total RNA from LPS-stimulated PMφ was hybridized to 10⁴ cpm of the labeled oligonucleotides in 10 mM PIPES, pH 6.4, 1 mM EDTA, pH 8.0, and 0.4 M NaCl at 30°C for 16 h. Following ethanol precipitation, the samples were dissolved in 20 μl reaction buffer containing 50 mM Tris/HCl, pH 8.3, 10 mM MgCl₂, 1 mM DTT, 75 mM KCl, 1 mM dNTPs, and 20 U RNase inhibitor. Reverse transcription was performed at 42°C for 30 min by adding 200 U Superscript II (Life Technologies). The extension products were ethanol precipitated and analyzed on 6% polyacrylamide 7 M urea sequencing gels. A sequence reaction was set up separately with the same nonradiolabeled primer, using the template of genomic clone pBS/Bam1-8, which covered the mMincle 5′-flanking region, and was run in parallel with the extension products on the same sequencing gel.

The following single-stranded oligonucleotides were synthesized: P1 (position −76 to −45 of the mMincle promoter), 5′-CCTTGTCCTTGTGCAAGAGAGGAAATTCTG-3′ and 5′-GTCAGAATTTCCTCTCTTGCACAAGGACAAGG-3′; mP1, 5′-CCTTGTCCTTGTGCCCCAGAGGAAATTCTG-3′ and 5′-GTCAGAATTTCCTCTGGGGCACAAGGACAAGG-3′; IL-6 (position −165 to −138 of the human IL-6 promoter), 5′-GGACGTCACATTGCACAATCTTAATAAT-3′ and 5′-ATTATTAAGATTGTGCAATGTGACGTCC-3′. Underlined nucleotides represent the mutant sequences. Complementary DNA oligonucleotides were annealed by heating in a buffer containing 20 mM Tris-HCl, pH 7.5, 10 mM MgCl₂, 50 mM NaCl at 75°C for 5 min and cooling at room temperature. Probes were then labeled by filling in with [α-³²P]dCTP using a Klenow fragment. NF-IL6 M1 nuclear extracts were prepared as described previously (21). pGEX-NF-IL6 plasmid inserting human NF-IL6 (amino acids 24–345; Ref. 7) downstream of the GST gene in pGEX-4T-2 (Amersham Pharmacia Biotech) is a generous gift from S. Hashimoto. GST protein and GST-NF-IL6 fusion protein were expressed in E. coli BL21 and purified by incubation with glutathione-coupled Sepharose beads (Amersham Pharmacia Biotech). Protein concentration was determined by BCA protein assay reagent (Pierce, Rockford, IL). Nuclear extracts (6 μg) or GST fusion proteins (100 ng) were incubated for 20 min at room temperature with 1 × 10⁴ cpm of the labeled DNA probe in 25 μl of binding buffer containing 10 mM HEPES-KOH, pH 7.8, 50 mM KCl, 1 mM EDTA, pH 8.0, 5 mM MgCl₂, 10% glycerol, and 3 μg of poly(dI-dC) (Amersham Pharmacia Biotech). Competition assays were conducted in the same manner, except that the above reaction mixture was preincubated with a 100-fold molar excess of unlabeled competitor oligonucleotides for 30 min at 4°C before the addition of the labeled probe. Supershift assays were performed using 200 ng of anti-C/EBPβ Ab (C-19; Santa Cruz Biotechnology, Santa Cruz, CA) and preincubated with the above reaction mixture at 4°C for 30 min before the addition of the labeled probe. Samples were loaded on native 5% polyacrylamide gels, and electrophoresis was conducted at 30 mA in 25 mM Tris, pH 8.5, 190 mM glycine, and 1 mM EDTA. Gels were subsequently dried for autoradiography.

We used the subtraction cloning method to compare the differences of mRNA expression profiles between WT and NF-IL6 (−/−) PMφ. WT and NF-IL6 (−/−) PMφ were stimulated for 16 h with 100 ng/ml LPS and 100 U/ml IFN-γ. Tester and driver cDNAs were synthesized using poly(A)⁺ RNA extracted from WT and NF-IL6 (−/−) PMφ, respectively. Subtractive hybridization was performed to concentrate cDNA species preferentially existing in tester cDNA. Two independent subtractive libraries were constructed to reduce false positives owing to individual differences. Differential hybridization was performed against 1000 colonies for each library, and positive clones were sequenced. Among these positive clones, commonly isolated ones from both libraries were subjected to further study. Comparing with WT PMφ, we identified four cDNAs whose expressions were dramatically reduced in NF-IL6 (−/−) PMφ (M.M. and S.A., unpublished observations). One of these cDNAs was novel and designated macrophage-inducible C-type lectin (Mincle) based on its expression characteristics and primary structure of the encoded protein as described below.

Mincle expression was assessed by Northern blotting in PMφ following stimulation with LPS. Mincle mRNA expression was hardly detected in untreated PMφ, but appeared within 2 h following LPS treatment (Fig. 1,A). Three mRNA species hybridized with this probe. The shortest mRNA (1.7 kb) is found to be most abundant. Mincle mRNA expression sustained for at least 16 h after LPS stimulation in WT PMφ. The levels of NF-IL6 mRNA were elevated within 30 min following LPS treatment before the induction of Mincle mRNA in WT PMφ (Fig. 1,A). Although NF-IL6 (−/−) PMφ induced expression of Mincle mRNA by 2 h following LPS stimulation, the expression levels were much lower than WT ones. Comparable levels of MIP-2 mRNA in both cell types indicate that NF-IL6 (−/−) cells respond efficiently to LPS stimulation. Next, we examined Mincle mRNA expression in WT and NF-IL6 (−/−) PMφ after various inflammatory stimuli (Fig. 1,B). Following 4 h stimulation with IFN-γ, IL-6, TNF-α, and LPS, total RNAs were extracted and Northern blotting was performed using Mincle-specific probe. A strong Mincle mRNA induction was observed in WT PMφ following treatment with IFN-γ, IL-6, TNF-α, or LPS (Fig. 1,B). In contrast, much lower levels of expression were detected in NF-IL6 (−/−) PMφ in response to these stimuli. IL-1β, phorbol myristate acetate, or ionomycin treatment could not induce Mincle mRNA effectively both in WT and NF-IL6 (−/−) PMφ (data not shown). Although we used thioglycolate-elicited PMφ that were subsequently expanded with GM-CSF in this experiment, the similar results were obtained from resident PMφ and thioglycolate-elicited PMφ cultured without GM-CSF (data not shown). These results indicate that several inflammatory stimuli strongly induce Mincle mRNA expression in an NF-IL6-dependent manner. Because LPS stimulation induced the strongest Mincle expression in PMφ, we tested Mincle expression in various cell lines after LPS treatment (Fig. 1 C). Mincle expression was observed in macrophage RAW264.7 cells and dramatically augmented after LPS stimulation. A low level of expression was also detected in M1 myeloblastic leukemia cells after stimulation with LPS. However, other cell lines BCL-1 (mature B cell), MOPC (myeloma), 5E3 (NK cell), EL4 (thymoma), and NIH3T3 cells (embryonic fibroblast) could not express any detectable levels of Mincle mRNA. Furthermore, Mincle mRNA was undetectable in brain, heart, lung, spleen, kidney, skeletal muscle, and testis (data not shown), suggesting that it may display macrophage-restricted expression.

We have cloned 2.5- and 1.7-kb mMincle cDNAs by the RACE technique. Sequence analysis revealed that the difference is due to the alternative usage of polyadenylation sites in the 3′ untranslated region, but not due to the alternative splicing (data not shown). Both the 2.5- and 1.7-kb cDNA contain the same open reading frame of 642 bp encoding a polypeptide of 214 amino acids, with a calculated molecular mass of ∼24.4 kDa. This polypeptide was devoid of NH₂-terminal hydrophobic signal peptide but contained a transmembrane domain of 24 amino acids predicted by hydrophobicity analysis (data not shown) (28). These features on the primary structure of Mincle indicated that it is a type II integral membrane protein and is therefore composed of a NH₂-terminal cytoplasmic domain of 21 residues and extracellular COOH-terminal domain of 169 aa (Fig. 2). We also identified a human homologue by a combination of degenerative PCR and RACE techniques using LPS-stimulated THP-1 cell cDNA and aligned it with mMincle (Fig. 2). We have cloned 2.2- and 1.0-kb hMincle cDNAs (data not shown), both of which include a single open reading frame of 657 bp encoding a polypeptide of 219 aa. Sequence comparison of hMincle with mMincle revealed overall identity of 67% and similarity of 85% (Fig. 2). mMincle has a putative N-linked glycosylation site at Asp¹⁰⁷, while hMincle has two sites at Asp⁶² and Asp¹⁰⁷. Two potential phosphorylation sites for protein kinase C (Ser³ and Thr¹²) were found within the cytoplasmic tail of both mouse and human Mincle (Fig. 2). These sites are compatible with protein kinase C phosphorylation motif, Ser/Thr-X-Arg/Lys.

A homology search of the available databases did not reveal any amino acid sequence identical with Mincle. But the extracellular domain of Mincle showed similarity to a wide range of C-type lectins. Mouse and human Mincle have 136- and 141-aa lengths of CRD, respectively. This domain exhibits the greatest degree of homology to macrophage C-type lectin (MCL), which is related to group II C-type lectins (29). The carboxyl terminus of mMincle protein was aligned with CRDs of mMCL and three members of group II C-type lectins: mouse CD23 (mCD23), mouse asialoglycoprotein receptor 1 (mASGR-1), and mouse macrophage asialoglycoprotein binding protein (mM-ASGP-BP) (Fig. 3). The sequence of rat mannose binding lectin A (rMBL-A) was included because it has been analyzed in detail (30, 31, 32, 33). The alignment showed that the putative sequence is 44, 40, 38, 36, and 31% identical with the CRDs of mMCL, mCD23, mASGR-1, mM-ASGP-BP, and rMBL-A, respectively. In addition, it indicated that apart from Asp¹⁵³ and Gln¹⁸⁹, mMincle protein shares 12 of the 14 invariant and 18 of the 18 highly conserved amino acids used to define C-type lectins (Fig. 3) (34). mMincle protein also conserved all amino acids required for calcium cation association within the rMBL-A CRD. Taken together, these observations suggested that Mincle is indeed a C-type lectin.

A genomic DNA fragment obtained in the course of promoter analysis (see Materials and Methods) contained the first five exons of mMincle (nucleotides 1–611) in addition to the 5′-flanking region. The first three exons correspond to separate functional domains of the receptor polypeptide (cytoplasmic tail, transmembrane sequence, and extracellular neck region), as is characteristic to group II C-type lectins (data not shown) (35). The position of the fourth and fifth exons precisely match the position of those found in the CRD of chicken hepatic lectin, the prototype of group II C-type lecin (data not shown) (35).

The mouse chromosomal location of Mincle was determined by interspecific back-cross analysis using progeny derived from matings of [(C57BL/6J × M. spretus)F₁ × C57BL/6J] mice. This interspecific back-cross mapping panel has been typed for over 2700 loci that are well distributed among all the autosomes as well as the X chromosome (24). C57BL/6J and M. spretus DNAs were digested with several enzymes and analyzed by Southern blot hybridization for informative RFLPs using a mouse Mincle cDNA probe. The 6.5-kb BamHI M. spretus RFLP (see Materials and Methods) was used to follow the segregation of the Mincle locus in back-cross mice. The mapping results indicated that Mincle is located in the distal region of mouse chromosome 6 linked to Atp6e, Slc2a3, and Cd4. Although 95 mice were analyzed for every marker and are shown in the segregation analysis (Fig. 4), up to 167 mice were typed for some pairs of markers. Each locus was analyzed in pairwise combinations for recombination frequencies using the additional data. The ratios of the total number of mice exhibiting recombinant chromosomes to the total number of mice analyzed for each pair of loci and the most likely gene order are: centromere Atp6e (2/110), Slc2a3 (1/136), Mincle (1/167), Cd4. The recombination frequencies (expressed as genetic distances in centiMorgans (cM) ± SE) are Atp6e (1.8 ± 1.3 cM), Slc2a3 (0.7 ± 0.7 cM), Mincle (0.6 ± 0.6 cM), Cd4.

We have compared our interspecific map of chromosome 6 with a composite mouse linkage map that reports the map location of many uncloned mouse mutations (provided from Mouse Genome Database, a computerized database maintained at The Jackson Laboratory, Bar Harbor, ME). Mincle mapped in a region of the composite map that lacks mouse mutations with a phenotype that might be expected for an alteration in this locus (data now shown).

The distal region of mouse chromosome 6 shares regions of homology with human chromosomes 22 and 12 (summarized in Fig. 4), suggesting that the human homologue of Mincle will map to one of these two chromosomes as well.

The amino acid sequence suggests that Mincle is a type II integral membrane protein as described above. To confirm the surface expression of Mincle protein, a carboxyl-terminal Flag-tagged mMincle cDNA was transiently transfected into 293T cells and mMincleFlag protein was detected by flow cytometric analysis using anti-Flag M2 mAb without permeabilizing cells. Staining of mMincleFlag-transfected cells with the M2 mAb revealed strong surface expression of mMincleFlag protein (Fig. 5,A, left). Mock-transfected cells showed no actual staining with the M2 mAb (Fig. 5,A, right). Simultaneously, mMincleFlag protein expression was detected by Western blot analysis (Fig. 5 B). mMincleFlag-transfected or mock-transfected 293T cells were immunoprecipitated and blotted with anti-Flag M2 mAb. The observed mMincleFlag protein corresponded to a molecular mass of ∼35 kDa, although the calculated molecular mass of mMincleFlag is 25.4 kDa. This result suggests that mMincle protein extensively acquired posttranslational modifications.

To determine the transcription initiation site of the mMincle gene, we performed a primer extension analysis using a primer corresponding to nucleotides 81–112 of the cDNA (Fig. 6). After hybridization with total RNA from LPS-stimulated PMφ, the primer was extended with reverse transcriptase. Identical reactions were conducted side by side with yeast tRNA as negative control. The extension products are displayed alongside the mMincle antisense sequence generated with the same oligonucleotide primer (Fig. 6). We identified a single predominant transcription initiation site, corresponding to 125 bases upstream of the translation start codon. The nucleotide at this side is C, which we tentatively assigned +1 as the transcriptional initiation site.

A mouse genomic DNA library was screened with the ³²P-labeled EcoRI- EcoRI cDNA fragment of mMincle (nucleotides 126–496). Sequencing of an isolated clone revealed that it included segments of the 5′-flanking region and the first five exons. The 1.8-kb promoter region of the mMincle gene was completely sequenced from both strands (Fig. 7). Computer-assisted search using the TFSEARCH program (36) revealed a number of potential binding sites for various transcription factors. The location of a select few of these thought to be potentially relevant to the regulation of this gene by LPS are shown. These include motifs that may bind NF-IL6, NF-κB, AP-1, and c-Ets. We found putative binding motifs for NF-IL6 at positions −1222 to −1210, −1108 to −1095, −929 to −917, −632 to −619, and −69 to −56. A canonical TATA box is located 29–24 bp upstream of the transcription initiation site.

To examine whether NF-IL6 could trans-activate the mMincle promoter, we constructed a series of the mMincle promoter-luciferase plasmids with various 5′ deletions. A 1852-bp fragment, including 1783 bp of the 5′-flanking region and 69 bp of the 5′-noncoding region, was fused to a promoterless luciferase reporter vector pGL3 and luciferase expression constructs with various 5′ deletions were prepared (Fig. 8). Luciferase reporter constructs were transiently transfected into NF-IL6 M1 cells by an electroporation method. The cells were split into two equal parts posttransfection and cultured in the presence (1 mM) or absence of isopropyl-β-d-thiogalactoside (IPTG). NF-IL6 M1 cells inducibly expressed human NF-IL6 protein within 4 h after IPTG addition as described previously (data not shown) (21). After 18 h, the cultures were harvested and cellular extracts were prepared. The luciferase activity levels were normalized to the cotransfected Renilla luciferase activities and presented as values relative to that of the promoterless pGL3 basic vector. Transfection of the pGL3-1783/+69, -1190/+69, -240/+69, and -166/+69 constructs resulted in about 3- to 5-fold induction by NF-IL6 expression. Deletion of the sequence from position −166 to −61 impaired the responsiveness to IPTG exposure completely (Fig. 8). This region contains one NF-IL6 binding motif (TKNNGNAAK) at position −66 to −58. Mutations were then introduced in the NF-IL6 consensus sequence within pGL3-240/+69 vector to form pGL3-240m. Two adenine and one guanine residues that were thought to be essential for NF-IL6 binding were all changed to cytosine residues. pGL3–240m activity was assayed and found not to respond to NF-IL6 expression (Fig. 8). This result indicates that the binding of NF-IL6 to position −66 to −58 is required to activate the mMincle promoter.

We next tested whether NF-IL6 could actually interact with position −66 to −58 of the mMincle promoter. A double-stranded oligonucleotide P1 spanning position −76 to −45 was radiolabeled and used in EMSA. As shown in Fig. 9,A, when P1 probe was incubated with nuclear extracts prepared from IPTG-treated NF-IL6 M1 cells, broad binding activities appeared (Fig. 9,A, lane 2). Competition assays were done with an authentic NF-IL6 binding site from the human IL-6 promoter (IL-6), P1, and mutated P1 (mP1) harboring the same mutation as the reporter gene assay. A 100-fold molar excess of unlabeled IL-6 and P1 oligonucleotides competed away the nuclear protein-DNA complexes (Fig. 9,A, lanes 3 and 4), but mP1 could not abolish them at all (Fig. 9,A, lane 5). Polyclonal Ab to NF-IL6 inhibited the formation of these complexes and resulted in supershifted complexes (Fig. 9,A, lane 6). To confirm thoroughly that NF-IL6 protein binds to P1 oligonucleotide, a purified GST-NF-IL6 fusion protein was examined instead of NF-IL6 M1 cells nuclear extracts and found to form a specific complex with the DNA probe (Fig. 9,B, lane 2). The DNA-protein complex could be abolished by competition with unlabeled IL-6 and P1 oligonucleotides but not by mP1 oligonucleotide (Fig. 9 B, lane 3–5). These results clearly demonstrate that NF-IL6 binds to position −66 to −58 of the mMincle promoter, which accords with the NF-IL6 binding consensus sequence.

The goal of our study is to clarify the function of NF-IL6 in macrophages through identification of its downstream targets. By subtraction cloning, we have isolated a novel C-type lectin, Mincle, whose expression in response to inflammatory stimuli was severely impaired in NF-IL6 (−/−) macrophages. Mincle gene expression is strongly induced by LPS and several proinflammatory cytokines, including IFN-γ, IL-6, and TNF-α, in WT PMφ. In the present study, the Mincle mRNA expressions were observed only in PMφ, macrophage cell line RAW264.7, and myeloblastic leukemia cell line M1 following inflammatory stimuli. Mincle gene expression may be restricted to myelomonocytic lineage cells stimulated with inflammatory mediators.

Macrophages infiltrate into various inflammatory regions such as rheumatoid arthritis joints (37), many types of tumors (38), and wounds (39), where concentrations of proinflammatory cytokines are elevated. Mincle may be up-regulated in those conditions and play some role on infiltrating macrophages. NF-IL6 (−/−) macrophages lack activities of bacteria killing and tumor cytotoxity even though they are fully activated by treatment with LPS and IFN-γ (18). Low levels of Mincle induction in NF-IL6 (−/−) PMφ could raise a possibility that Mincle may play a role in bacteria killing and tumor cytotoxity. An expected function of Mincle is the recognition of the surface carbohydrates of microorganisms or tumor cells, followed by macrophage activation. Activated macrophages could kill ingested bacteria or lyse tumor cells as NK cells kill tumors and virally infected cells following recognizing target cells through NKR-P1 (40, 41). Mincle may be involved in such immune surveillance processes by activated macrophages under transcriptional control of NF-IL6. However, extensive studies are required to discuss the biological function of Mincle exactly.

Although Mincle is a target gene of NF-IL6 in addition to G-CSF, the promoter sequence of Mincle is quite different from that of G-CSF. G-CSF promoter contains a segment between −165 and −196 bp, G-CSF gene promoter element 1 (GPE1), which plays an important role in the LPS-inducible expression of the G-CSF gene in macrophages (42). Subsequent studies on the GPE1 revealed that both an NF-κB binding element and an adjacent NF-IL6 binding element within GPE1 are critical for induction of the G-CSF promoter by TNF-α and IL-1β (43). It has also been demonstrated that both NF-κB p65 and NF-IL6 can bind to the GPE1 and form a ternary complex with the DNA. Although the Mincle gene has three putative NF-κB binding motifs in the 1.7-kb promoter region, these sites exist apart from the indispensable NF-IL6 binding element. It is possible that NF-κB is involved in the inflammatory signal-induced Mincle gene transcription, but NF-κB may not directly interact with NF-IL6 on the Mincle promoter. The Mincle gene promoter harbors potential binding elements for other inflammation-activated transcription factors, such as AP-1 and c-Ets. These transcription factors may cooperatively act with NF-IL6 for the Mincle gene induction in macrophages.

Recent reports showed that the activities of C/EBPα, -β, -δ, and -ε are redundant in regard to the expression of IL-6 and MCP-1 (22, 23). The ectopic expression of C/EBPα, -β, -δ, or -ε is sufficient to confer the LPS-inducible expression of IL-6 and MCP-1 to P388 lymphoblasts, which normally lack C/EBP factors and do not display LPS induction of proinflammatory cytokines. In fact, C/EBPδ exhibits an expression pattern similar to that of NF-IL6 and has been shown to be involved in the regulation of several genes induced during inflammation (44). It is likely that the lack of NF-IL6 is partly compensated for by the induction of C/EBPδ upon LPS treatment in vivo. The precise mechanism that other NF-IL6 family proteins could compensate for IL-6 and MCP-1 expression but not G-CSF and Mincle expression in LPS-stimulated macrophages remains to be elucidated.

Mincle protein sequence exhibits the highest similarity to the members of group II C-type lectins, most of which mediate glycoprotein endocytosis. The prototype for this group is the asialoglycoprotein receptor, a hepatic cell-surface protein that binds terminal galactose residues exposed upon partial desialylation of circulating glycoproteins (45). This receptor directs turnover of serum glycoproteins, leading to their internalization and delivery to lysosomes via an endocytic pathway. Human asialoglycoprotein receptor H1 subunit contains a single cytoplasmic tyrosine at position 5 that is located within a critical internalization motif for the receptor (46). Rapid internalization of cell-surface receptors generally requires a short stretch of amino acids containing a tyrosine residue (47). Other members of group II C-type lectins, such as M-ASGP-BP (4), CD23 (48), gp120-binding C-type lectin (49), and Kupffer cell fucose receptor (50) carry tyrosine residues in the cytoplasmic region. However, the cytoplasmic region of Mincle protein contains no tyrosine residues, indicating that Mincle protein could not mediate efficient endocytosis. Another C-type lectin that possesses no intracytoplasmic tyrosine residue is MCL, whose protein sequence exhibits the highest homology to Mincle.

Mincle and MCL present several common features; the protein sequences display the highest similarity to the members of group II C-type lectins, transcripts are abundantly expressed in macrophages, cytoplasmic regions contain no tyrosine residue, and the murine genes are mapped on chromosome 6. Thus, Mincle and MCL could be classified together into a derivative of group II C-type lectins.

In conclusion, we have isolated a novel C-type lectin, Mincle, whose expression is strongly induced in response to inflammatory stimuli under the regulation of NF-IL6 in macrophages. Identification of Mincle ligand and targeted disruption of Mincle gene should help elucidate its physiological role in vivo.

We thank E. Nakatani and A. Reuss for their excellent technical assistance and T. Shimada, K. Nakashima and K. Hoshino for their helpful suggestions.