APCs express receptors recognizing microbes and regulating immune responses by binding to corresponding ligands on immune cells. Having discovered a novel inhibitory pathway triggered by ligation of DC-HIL on APC to a heparin/heparan sulfate-like saccharide of syndecan-4 on activated T cells, we posited DC-HIL can recognize microbial pathogens in a similar manner. We showed soluble recombinant DC-HIL to bind the dermatophytes Trichophyton rubrum and Microsporum audouinii, but not several bacteria nor Candida albicans. Dermatophyte binding was inhibited completely by the addition of heparin. Because DC-HIL contains an ITAM-like intracellular sequence, we questioned whether its binding to dermatophytes can induce tyrosine phosphorylation in dendritic cells (DC). Culturing DC with T. rubrum (but not with C. albicans pseudohyphae) induced phosphorylation of DC-HIL, but not when the tyrosine residue of the ITAM-like sequence was mutated to phenylalanine. To examine the functional significance of such signaling on DC, we cross-linked DC-HIL with mAb (surrogate ligand), which not only induced tyrosine phosphorylation but also up-regulated expression of 23 genes among 662 genes analyzed by gene-array, including genes for profilin-1, myristoylated alanine rich protein kinase C substrate like-1, C/EBP, LOX-1, IL-1β, and TNF-α. This cross-linking also up-regulated expression of the activation markers CD80/CD86 and heightened APC capacity of DC to activate syngeneic T cells. Our findings support a dual role for DC-HIL: inhibition of adaptive immunity following ligation of syndecan-4 on activated T cells and induction of innate immunity against dermatophytic fungi.
Dendritic cells (DC)3 are the most potent APCs for initiating and controlling adaptive immune responses through presentation of Ag in the context of costimulatory signals to naive T cells. Subsets of DC include CD11c+/CD4+ lymphoid DC, CD11c+/CD8+ myeloid DC, CD11c+/PDCA-1+ plasmacytoid DC, and I-A/I-E+ epidermal Langerhans cells (LC) (1). In addition, immature DC, which reside in tissues interfacing with the external environment, serve as sentinels that sense and distinguish among different microbes to be internalized and processed, leading to altered DC gene expression profiles required for eliciting pathogen-specific adaptive immunity (2). Thus, DC serve important roles in both innate and adaptive immunity.
DC express several receptors that interact with corresponding surface molecules on microbes, including pattern recognition receptors (PRRs) that bind directly to particular molecular components of a given microbe (pathogen-associated molecular patterns) (3). PRRs include: the leucine-rich repeat protein CD14 that binds to LPS, lipoteichoic acid, and peptideglycan (4); scavenger receptors (SR-A, CD36, and MARCO) that bind low density lipoproteins or lipid A on some bacteria (5); the adhesion molecule integrin CR3 that binds to LPS, lipophosphoglycan, an acylpoly (1, 3) galactoside (APG), and Candida albicans (6); TLRs, consisting of nine members, that bind to zymosan, Staphylococcus aureus, LPS, bacterial flagellin, CpG bacterial DNA (7, 8); and C-type lectin-like receptors, such as dectin-1 (9), dectin-2 (10), and DC-SIGN (11, 12), which bind to many pathogens via polysaccharide moiety on their surface (13). Dectin-1 binds to β-glucan polysaccharide expressed primarily on C. albicans yeast (14); dectin-2 binds to hyphal (but not yeast) forms of C. albicans (15) probably through high-mannose structures (16); and DC-SIGN binds to mannose-type oligosaccharides (11). DC take advantage of distinct ligand specificities of PRRs to distinguish among pathogens, thereby inducing pathogen-dependent up-regulated gene expression. In particular, DC-SIGN may be unique because it uses adhesion properties to control DC migration and T cell activation (17, 18).
Previously, we discovered that DC-HIL is a highly glycosylated type I transmembrane protein of 95 and 125 KDa containing an extracellular Ig-like domain constitutively expressed at high levels by DC and macrophages. We also found DC-HIL to deliver a potent inhibitory signal to T cells, in an Ag-independent manner, by binding a heparin/heparan-like saccharide of syndecan-4 (SD-4) on activated T cells (19, 20). These findings led us to hypothesize that DC-HIL binds microbial pathogens in a similar manner. We found DC-HIL to bind the dermatophytes Trichophyton rubrum and Microsporum audouinii. Coculture of DC with T. rubrum induces phosphorylation of the tyrosine residue in the ITAM-like intracellular sequence of DC-HIL at a level markedly greater than that induced by ligation to the T cell ligand SD-4. This phosphorylation up-regulated expression of genes responsible for DC maturation and augmentation of APC function. Thus, DC-HIL is a PRR for dermatophytic fungi that can heighten APC properties, while also negatively regulating T cell activation.
Materials and Methods
Female BALB/c and C57BL/6 (5- to 8-wk-old) mice were purchased from Harlan Breeders, and OT-I and OT-II (5- to 8-wk-old) mice obtained from Taconic Farms. Following National Institutes of Health guidelines, these animals were housed and cared for in the pathogen-free facility of the Institutional Animal Care Use Center of The University of Texas Southwestern Medical Center.
Fc-fused recombinant protein
Rat mAb generated against CD4 (L3T4), CD8, I-A/I-E (2G9), CD11b (M1/70), CD86 (GL1), CD317 (PDCA-1, eBio927), and hamster mAb against CD11c (N418), CD80 (16A-10A1) were purchased from eBioscience. Secondary Abs were purchased from Jackson ImmunoResearch Laboratories.
UTX-103 rabbit anti-mouse DC-HIL mAb was generated as follows: Rabbits were immunized four times with 0.5 mg of DC-HIL-hFc (fused with human IgG-Fc) 3 wk apart. A week after the third immunization, sera were collected and the titer of anti-mouse DC-HIL evaluated by ELISA using DC-HIL-mFc (fused with mouse IgG-Fc) to eliminate anti-human IgG Ab. Rabbit spleen cells were fused with a rabbit myeloma cell line (Epitomics). One clone (UTX-103 mAb IgG1) was purified from the culture supernatant using protein A-agarose and entoxoin-free buffers. Compared with our previous 1E4 rat anti-DC-HIL mAb (19), this mAb has much higher affinity for native DC-HIL protein. Purified UTX-103 mAb and control IgG preparations used for cross-linking experiments were tested for endotoxin contamination using a LAL Chromagenic Endpoint Assay (HyCult Biotechnology; Cedarlane Laboratories). Endotoxin levels were <0.5 EU/ml.
Preparation of APC
Epidermal cells were isolated from ear skin of BALB/c mice using trypsin (Mediatech); LC were enriched by centrifugation over Histopaque (1.083, Sigma-Aldrich) (21). Bone marrow (BM) DC were harvested from day 6 culture of BM cells with GM-CSF (22). Other DC subpopulations were procured from the spleen cells of naive mice. Macrophages were harvested from the peritoneal cavity of mice stimulated with thioglycolate (21).
Freshly isolated or cultured leukocytes were assayed for surface and intracellular expression by flow cytometry. For surface expression, cells (1 × 105) were incubated with UTX-103 rabbit anti-DC-HIL mAb or the isotypic control IgG1 (each 10 μg/ml). After washing, cells were labeled with fluorescent secondary Ab (PE-anti-rabbit IgG, 1 μg/ml). For intracellular staining, cells (1 × 105) were fixed with 4% paraformaldehyde at room temperature for 30 min. After washing with Dulbecco’s PBS, cells were treated with permeabilization buffer (0.5% Saponin, 0.5% BSA in DPBS) for 10 min, and then incubated with UTX-103 mAb (10 μg/ml) in the permeabilization buffer for 30 min, followed by incubation with the secondary Ab (1 μg/ml). Fluorescence intensity of stained cells was analyzed by FACSCalibur (BD Biosciences).
Binding assays for microbes
Candida albicans (ATCC 10231) and the dermatophytes Trichophyton rubrum (ATCC 14001) and Microsporum audouini (ATCC 10008) were purchased from the American Type Culture Collection and grown in medium recommended by the American Type Culture Collection. C. albicans yeast was transformed into pseudohyphae by culturing freshly prepared yeasts at 37°C for 90 min in HBSS containing 1.25 mM CaCl2, 1 mM MgCl2, 10 mM HEPES (pH 7.7), and 10% heat-inactivated FCS (15). Aliquots of pseudohyphae (4 × 105 cells) were washed with DPBS and incubated with staining buffer (0.1% BSA, 2 mM CaCl2 in DPBS) containing 20 μg/ml Fc proteins on ice for 1 h. After extensive washing with buffer, cells were resuspended in 2.5 μg/ml FITC-anti-human IgG Ab on ice for 30 min. For immunstaining of dermatophyte fungi, single colonies of fungi were grown on Sabouroud’s agar plates, harvested and suspended in DPBS. After washing with DPBS and with water, small aliquots were spotted on slide glass, air dried, and stained with Fc proteins as before. Binding of Fc proteins to microbes was examined by confocal microscopy. In inhibition experiments, 20 μg/ml DC-HIL-Fc was preincubated with indicated concentrations of heparin, chitin, D-galacto-D-mannan, β-glucan, or mannan (all from Sigma-Aldrich) for 30 min on ice before binding to T. rubrum. The fungi also were treated with 1,000 unit of N-glycosidase (PNGase F, from New England Biolabs) at 37°C for 1 h before incubating with DC-HIL-Fc.
Either or both tyrosine residues (amino acids 523 and 529) in the intracellular domain was replaced with phenylalanine using Quick Change Site-Directed Mutagenesis and Pfu polymerase (Stratagene) performed according to the manufacturer’s recommendations for the following oligonucleotide pairs: Y523F 5′-GGTTACCATCTTGCTGTTCAAAAAACACAAGGCG-3′ (5′ primer, where italic letters show mutations) and 5′-CGCCTTGTGTTTTTTGAACAGCAAGATGGTAACC-3′ (3′ primer); Y529F 5′-AAAAACACAAGGCGTTCAAGCCAATAGGAAACTG-3′ and 5′-CAGTTTCCTATTGGCTTGAACGCCTTGTGTTTTT-3′; Y523F/Y529F 5′-CATCTTGCTGTTCAAAAAACACAAGGCGTTCAAGCCAATAG-3′ and 5′-CTATTGGCTT GAACGCCTTGTGTTTTTTGAACAGCAAGATG-3′.
Immunoblotting and tyrosine phosphorylation assay
Whole cell extracts were prepared from untreated or activated BM-DC (cultured for 2 days with 1 μg/ml LPS) and assayed for protein concentration (23). An aliquot (10 μg/lane) was applied to 4–15% SDS-PAGE, followed by immunoblotting using UTX-103 mAb and control IgG (each 1 μg/ml) (23). To examine tyrosine phosphorylation of DC-HIL on DC, BM-DC were treated with three different stimuli: BM-DC (3 × 106 cells/ml complete medium) were cultured with indicated amounts (as dried weight) of T. rubrum or C. albicans hyphae; BM-DC (5 × 106 cells in 500 μl of DPBS) were also incubated with UTX-103 mAb or control rabbit IgG (10 μg/ml) on ice for 30 min, followed by cross-linking with 100 μg/ml goat anti-rabbit IgG; and BM-DC (5 × 106 cells) were cultured in culture dish precoated with SD-4-Fc or control Ig (20 μg/ml). At indicated time periods at 37°C, treated DC were lysed using 500 μl of 2× lysis buffer (15). DC-HIL protein was immunoprecipitated by incubation at 4°C for 3 h with 2–5 μg of UTX-103 mAb and overnight incubation with protein-A agarose (50 μl of 50% slurry). The immune-complexes were dissociated by boiling and then analyzed for expression of phosphotyrosin by immunoblotting using biotinylated anti-phosphotyrosine (0.5 μg/ml) (4G10, Upstate Biotechnology) and HRP-streptavidin (1/10,000 dilution). The blotted membranes were also stripped and reanalyzed using 1E4 rat anti-DC-HIL mAb (1 μg/ml) (23) and HRP-anti-rat IgG (1/10,000 dilution). To examine tyrosine phosphorylation of DC-HIL mutants, COS-1 cells were seeded at a density of 5 × 105 cells/dish and transfected with a plasmid vector encoding wild-type or tyrosine mutants of DC-HIL (2 μg) using Fugene 6 (19). Two days after transfection, cells were cross-linked by UTX-103 mAb plus secondary Ab, and then immunoprecipitated as described above.
Gene expression by cross-linked BM-DC was analyzed using oligonucleotide probe-blotted gene arrays (Dendritic and APC and Autoimmune and Inflammation) according to the manufacturer’s recommendations (SuperArray Bioscience). In brief, BM-DC were harvested and cultured in 96-well plate (1 × 105 cells/well) precoated with UTX-103 mAb or control IgG (20 μg/ml). After 6 h of culture at 37°C, total RNA was isolated from treated cells (1 × 106) using ArrayGrade total RNA isolation kit (SuperArray). Biotin-labeled cDNA probes were prepared using the TrueLabeling-AMP 2.0 kit (SuperArray) and hybridized with gene arrays; hybridization signals were then detected using chemiluminescent detection kit (SuperArray). Image acquisition and analysis of data were performed by ImageQuant 400 (Amersham Biosciences). Expression of some genes up-regulated by the cross-linking was reanalyzed by real-time PCR following the manufacturer’s recommendations (LightCycler FastStart DNA Masterplus SYBR Green I, Roche). Primers for oxidized low-density lipoprotein receptor (Olr-1, GenBank: AY057791) include: 5′-CCCGGAAGCTGGACGAGA-3′ (5′ primer) and 5′-AGAACGGGGAGGTGGTATGG-3′ (3′ primer); myristoylated alanine rich protein kinase C substrate like-1 (Marcksl-1, GenBank: NM_009851): 5′-GCCCCCAGCAGACCCCCATCAT-3′ and 5′-CTCGCCCTGCTCCTGCTCTTCCTC-3′. For production of TNF-α and IL-1β, BM-DC (1 × 105 cells) were incubated similarly with immobilized UTX-103 mAb or control IgG for indicated time periods. BM-DC (2 × 105 cells) were treated with 10 μg/ml dried T. rubrum or Candida hyphae at 37°C for 30 min, immediately after which cells were washed extensively and cultured for 2 days. Culture supernatant was recovered and cytokines measured by ELISA kits (eBionscience).
BM-DC were cultured in 96-well plate precoated with UTX-103 mAb or control IgG (20 μg/ml) for 1 day. Treated DC were harvested and reseeded on 96-well plate at a different cell density and pulsed for 6 h with MHC class II-restricted OVA323–339 and MHC class I-OVA257–264 peptide (each 2 μg/ml) synthesized by the Protein Chemistry Technology Center at The University of Texas Southwestern Medical Center. After pulsing, DC were cocultured with the constant number of CD4+ or CD8+ T cells (1 × 105/well) purified from spleen of unprimed OT-II or OT-I transgenic mice, respectively, using T cell isolation kits (Miltenyi Biotec). Two days after coculture, culture supernatant was recovered and IL-2 and IFN-γ production was assayed by ELISA kits (eBionscience).
Expression of DC-HIL by APC
Using a newly developed UTX-103 rabbit anti-DC-HIL mAb with markedly higher affinity to native DC-HIL than our previously generated 1E4 rat anti-DC-HIL mAb (19), we reexamined surface expression of DC-HIL on different APC subsets by flow-cytomeric analysis (Fig. 1). Epidermal LC were identified as I-A/I-E+ epidermal cells, almost all of which expressed DC-HIL constitutively at high levels on the surface and intracellularly (Fig. 1,A). By contrast, DC-HIL was not expressed by I-A/I-E− epidermal cells. CD11c+ BM-DC also expressed DC-HIL constitutively on their surface. LPS stimulation up-regulated DC-HIL surface expression by some (but not all) CD11c+ DC (Fig. 1,B). This up-regulation was confirmed by immunoblotting of protein extracts from BM-DC using UTX-103 mAb that immunostained two bands (95 and 125 KDa) (Fig. 1,C). In spleen, there are at least three distinct DC subsets (CD11c+/CD4+ lymphoid DC, CD11c+/CD8+ myeloid DC, and CD11c+/PDCA-1+ plasmacytoid DC (1), all of which also expressed DC-HIL constitutively on the surface, albeit at lower levels (Fig. 1,D). DC-HIL expression by these DC subsets appeared invariant of in vivo stimulation (data not shown). Finally, peritoneal macrophages from mice treated with thioglycolliate also expressed surface and intracellular DC-HIL at high levels (Fig. 1 E). These results indicate that DC-HIL is expressed by a wide variety of APC subsets.
DC-HIL binds to dermatophyte cell wall
Many C-type lectin receptors (e.g., mannose receptor, DC-SIGN, dectin-1, and dectin-2) bind to saccharide ligands expressed by microbes (13). Similarly, we showed DC-HIL to bind a heparin/heparan sulfate (HS)-like saccharide of SD-4 on activated T cells (20). We thus posited that DC-HIL may also recognize microbial pathogens through a similar sugar moiety. To address this issue, we performed binding assays using fluorescent-labeled DC-HIL-Fc, dectin-2-Fc (as control), or Fc alone. Neither Staphylococcus aureus, group A streptococci, Pseudomonas aeruginosa, nor Escherichia coli bound to DC-HIL (data not shown). We then examined binding to Candida albicans pseudohyphae consisting of round yeast and filamentous hyphae (Fig. 2,A). Using dectin-2-Fc as a positive control because it is known to bind hyphal (but not yeast) components (15) (Fig. 2,A), we observed that neither DC-HIL-Fc nor Fc alone bind to Candidal pseudohyphae. We then examined binding to dermatophytes (Fig. 2, B and C). DC-HIL-Fc bound to Trichophyton rubrum with high affinity and to Microsporum audouinii at lower affinity. Binding of DC-HIL to T. rubrum was blocked completely by addition of heparin (2 μg/ml) (Fig. 2,D), which we showed previously to inhibit DC-HIL binding to activated T cells (20). To sort the fungal ligands of DC-HIL, T. rubrum fungi were pretreated with N-glycosidase that removes saccharide residues from glycoproteins (Fig. 2,E), or DC-HIL-Fc was preincubated with fungal saccharides, including chitin, galactomannan, β-glucan, and mannan before binding assays (Fig. 2, F and G). N-glycosidase treatment inhibited binding of DC-HIL-Fc to T. rubrum almost completely. None of the saccharide inhibitors showed complete inhibition as was observed with heparin: Chitin and galactomannan were moderate inhibitors and others (β-glucan and mannan) weak inhibitors. These results indicate that DC-HIL can bind dermatophytes, suggesting that the fungal ligands of DC-HIL may be saccharides structurally related to HS, chitin, and/or galactomannan.
Ligation of DC-HIL leads to tyrosine phosphorylation of its ITAM-like motif
Because DC-HIL has an ITAM-like signaling motif (YxxI), we questioned whether binding of DC-HIL to T. rubrum transduces tyrosine phosphorylation of this protein in DC (Fig. 3,A). BM-DC were cocultured with varying doses of C. albicans hyphae or T. rubrum (as dried weight), and tyrosine phosphorylation on DC-HIL was assayed by immunoprecipitation and blotting using anti-p-tyrosine Ab. Tyrosine phosphorylation of DC-HIL was induced in DC following coculture with T. rubrum, but not with C. albicans hyphae even at the highest dose tested, consistent with selective binding by DC-HIL. Such phosphorylation was also detected in BM-DC after cross-linking of DC-HIL with UTX-103 mAb (but not with control IgG) (Fig. 3,B). However, the level of phosphorylation induced by the mAb was considerably less than by T. rubrum. We then questioned whether the T cell ligand SD-4 can induce tyrosine phosphorylation (Fig. 3 C). Treatment of BM-DC with immobilized SD-4-Fc (but not control Ig) induced phosphorylation, albeit at a weaker level compared with the two previous stimulators.
DC-HIL contains two tyrosine residues in its intracellular domain: at aa 523 proximal to the transmembrane domain and aa 529 in the YxxI sequence corresponding to the ITAM-like motif (note that a typical ITAM has two tandem-repeats of YxxI/L) (24). To determine which tyrosine residue is responsible for phosphorylation, point-mutation analysis was performed (Fig. 3,D), in which either or both Tyr 523 or 529 was (or were) mutated to phenylalanine (designated Y523F, Y529F, or Y523F/Y529F). Mutants and wild-type DC-HIL were transfected separately into COS-1 cells and assayed for tyrosine phosphorylation (Fig. 3,E). Surface expression of mutant DC-HIL on COS-1 cells was similar to those of wild-type DC-HIL (data not shown). Wild-type DC-HIL was tyrosine phosphorylated as early as 10 min in COS-1 cells cross-linked with UTX-103 mAb (but not with control IgG). DC-HIL bearing Y523F mutation was phosphorylated at a similar level, whereas Y529F mutant and doubly mutated DC-HIL failed to undergo phosphorylation (Fig. 3 E), thereby identifying the former (tyrosine on aa 529 in the YxxL motif) as the relevant moiety.
Cross-linking of DC-HIL with UTX-103 mAb up-regulates particular genes in DC
Because phosphorylated YxxL (even just one unit) can induce gene expression (25), we examined changes in DC gene expression profile following stimulation of DC-HIL (Fig. 4). For this study, we chose UTX-103 mAb as a surrogate ligand for DC-HIL because cell wall extracts from T. rubrum are toxic to BM-DC (>2 h coculture with extract kills most DC). BM-DC were cross-linked with UTX-103 mAb or control IgG, followed 6 h later by isolation of mRNA, and then gene expression analysis using two different Oligo GEArray Dendritic and APC and Autoimmune and Inflammatory microassays, on which 260 and 440 gene-specific olignucletide probes were blotted, respectively. (These were a total of 662 genes because the two arrays have overlapping gene probes.) Gene expression in DC treated with UTX-103 mAb was expressed relative to that of the house keeping gene GAPDH (Fig. 4, A and C) and compared with that of DC treated with control IgG (evaluated by fold difference). Genes up-regulated more than 2-fold greater than control are listed (Fig. 4, B and D), including profilin-I (an actin-binding protein involved in turnover and restructuring of the actin cytoskeleton) (26), Marcksl-1 (myristoylated alanine-rich C-kinase substrate involved in regulating cell shape, motility, secretion, transmembrane transport, and cycling) (27), CCAAT/enhancer binding protein (C/EBP or Cebpb) involved in LC commitment (28), and a lectin-like receptor for oxidatively modified low-density lipoprotein (LOX-1 or Orl-1) (29, 30). Our analyses showed 23 genes to be up-regulated following cross-linking with UTX-103 mAb, corresponding to 3% of 662 genes tested. Not one gene examined was down-regulated >2-fold. Because endotoxin content in preparations of UTX-103 mAb and control IgG were similar (<0.5 EU/ml), we think it can be excluded as a cause of the changes.
To verify our results, we examined mRNA expression of Marcksl-1 and Orl-1, using real-time PCR. Both genes were chosen because they were highly up-regulated in each microarray (Fig. 5, A and B). Profilin-I and Cebpb (the highest up-regulated genes) were not examined in this manner because we were unable to find relevant primers capable of producing dose-dependent amplification. After BM-DC were cross-linked with UTX-103 mAb or control IgG, mRNA expression was analyzed and expression levels calculated as fold-increases. Consistent with results of microarray-gene expression analysis, Orl-1 and Marcksl-1 genes were up-regulated (12 and 4.5-fold, respectively, greater than controls) 6 h after cross-linking. Their up-regulation was transient because a return to baseline expression levels was noted 2 days after stimulation.
We also examined protein expression of TNF-α and IL-1β by ELISA (Figs. 5,C and D). Without stimulation, BM-DC did not secrete detectable levels of either cytokine, and control IgG treatment induced very low levels of expression. By contrast, treatment with UTX-103 mAb led to markedly elevated levels of TNF-α and IL-1β secretion that lasted at least 2 days. We then questioned whether T. rubrum (or Candida hyphae) stimulates DC to produce these proinflammatory cytokines (Fig. 5, E and F). BM-DC were stimulated with/without dried T. rubrum or Candida hyphae and measured for secretion of these cytokines. DC treated with T. rubrum produced TNF-α and IL-1β cytokines at a level similar or higher than by DC stimulated with cross-linking of DC-HIL. T. rubrum was a stronger stimulator than Candida.
Because IL-1β induces DC maturation and acquisition of strong immunostimulatory capacity (31), we next examined the effect of cross-linking of DC-HIL on DC expression of activation/maturation markers. BM-DC were cross-linked with UTX-103 mAb and then cultured for 2 days. Surface expression of CD80 and CD86 was examined by flow cytometry (Fig. 6, A and B). Treatment with UTX-103 mAb markedly increased expression of CD80 (47 vs 356 mean fluorescent intensity, MFI) and CD86 (18 vs 49). We then examined T cell-stimulatory capacity by UTX-103 mAb-treated DC (Fig. 6, C–E). After cross-linking and pulsing DC with MHC class I- or II-restricted OVA peptide, increasing numbers of these cells were cocultured with a constant number of CD4+ or CD8+ T cells isolated from OT-II or OT-I transgenic mice, respectively. Activation of T cells was measured by IL-2 production for CD4+ T cells (Fig. 6,C) and IL-2 and IFN-γ for CD8+ T cells (Fig. 6, D and E). UTX-103 mAb-treated DC stimulated OT-II CD4+ T cells to produce IL-2, greater than by control IgG-treated DC at each dose tested, with highest response of 8-fold increase. UTX-103 mAb also stimulated DC to exhibit a greater capacity to activate OT-I CD8+ T cells, but to a lesser degree than for OT-II CD4+ T cells: ∼2-fold increase in IL-2 (Fig. 6,D) and 50% increase in IFN-γ (Fig. 6 E). Altogether, cross-linking of DC-HIL up-regulated DC gene expression, resulting in DC maturation and augmented T cell-stimulatory capacity.
While infecting skin, hair, and nails, dermatophytic fungi obtain nutrients from keratinized material (32). Although these fungi usually do not invade living tissue, their metabolic by-products cause inflammation, which is especially severe in immunosuppressed patients (33). An early defense mechanism against dermatophytic infection may be mediated by epidermal LC (skin-resident DC) that reside close to the usual primary site of infection. Having shown that DC-HIL acts as a PRR for dermatophytes and that epidermal LC express DC-HIL constitutively at high levels (human LC also express DC-HIL at high levels; Ref. 34), we posit that DC-HIL exerts antifungal immunity via innate immune recognition and potentiation of LC function. Concurrently, DC-HIL may suppress cutaneous inflammation by attenuating activity of T cells that home to skin. Thus, the net effect of these positive and negative regulations exerted by DC-HIL may determine, at least in part, the outcome of antidermatophyte immunity.
SD-4 is a transmembrane protein heavily laden with HS chains consisting of alternating disaccharide residues (glucuronic acid and iduronic acid with glucosamine). It is expressed constitutively by B cells but not by naive T cells (its expression can be induced by activation; Ref. 20). Despite its expression profile, DC-HIL binds to activated T cells, but not to B cells (our unpublished data). Because the binding is abrogated by heparin or by heparinase treatment of activated T cells, DC-HIL is likely to recognize the structure of HS chain expressed on T cells. Our results suggest that nonself DC-HIL ligands are expressed by T. rubrum and M. audouinii. Unlike TLR ligands, expression of these ligands may be restricted to only some fungi because DC-HIL does not bind bacteria and C. albicans. The cell wall of dermatophytes is made up primarily of chitin, mannan, and galactomannan, none of which structurally resemble HS (35). In fact, chitin and galactomannan did not strongly inhibit DC-HIL binding as heparin did, and mannan was a weak inhibitor. We thus postulate that the putative ligand(s) on dermatophytes are saccharides structurally resembling HS on T cells.
Our findings indicate that the membrane-proximal YxxI sequence of DC-HIL is the functional tyrosine-based signal motif. It is not a typical ITAM because two ITAM units have been shown to be required for signal transduction (36). ITAM was identified originally by mutation analysis of Ag receptors containing multiple activation motifs comprising two YxxL/I sequences with defined spacing between them (24). ITAM is phosphorylated by Src family kinases and the resultant phosphorylated ITAMs are recognized by two Src homology 2 domains of Syk kinases, that enable transduction of signals (37). Thus, the space between two ITAMs is important for recognition by Syk kinases. However, only one ITAM unit was demonstrated to be sufficient to initiate Syk-mediated signal transduction (e.g., dectin-1, which has one YxxL motif; Ref. 9 , 14 , 25). In fact, ligation of dectin-1 by β-glucan (or zymosan) can induce expression of a number of genes necessary for innate immunity against yeast pathogens, cooperatively with TLR (8, 38). Like dectin-1, we speculate that the YxxI motif in DC-HIL is capable of inducing Syk-mediated signaling responsible for potentiating DC function.
Several ligand/receptor pairs controlling T cell activation are involved in reciprocal signaling between APC and T cells. Although the impact of these pairs has been well studied for T cells, relatively less is known about effects on APC. Programmed cell death-1 ligands (PD-L1 and PD-L2/B7-DC) on APC negatively regulate T cell activation by interacting with PD-1 on T cells (39, 40). Because PD-L1 and PD-L2 have short cytoplasmic tails lacking known motifs for signal transduction, these ligands are thought to be incapable of transducing signals following binding to PD-1. PD-L2 possesses a four amino acid-long intracellular domain and recent studied have shown that cross-linking PD-L2 directly potentiates DC function by enhancing DC presentation of Ag-loaded MHC molecules, promoting DC survival and increasing secretion of IL-12 (41, 42). Possibly, PD-L2 associates with a coreceptor (that contains intracellular signaling motifs) through two charged amino acids in its transmembrane domain. To our knowledge, PD-L1-induced signaling has not been formally reported. Herpesvirus entry mediator on APC is a ligand for coregulatory receptors BTLA, CD160, and LIGHT on T cells. Herpesvirus entry mediator is a member of the TNF receptor family that can recruit several members of the TNFR-associated factor family, enabling activation of NF-κB and Jun N-terminal kinase, eventually leading to augmented immune responses (43). These coinhibitory ligands including DC-HIL deliver negative signals to T cells when engaged to their corresponding T cell receptors, and conversely they can transduce positive signals within APC.
Such reciprocal signaling may confer balance in the activation of APC and T cells. Immature DC such as epidermal LC constitutively express coinhibitory ligands at levels higher than costimulatory ligands (44). This is true for DC-HIL, which is expressed highly on epidermal LC. Such DC are less potent activators of naive T cells than mature DC (45), but may down-regulate the status of recently activated T cells because they express high levels of coinhibitory receptors. The DC-HIL ligand SD-4 also is expressed on activated (but not resting) T cells. Shortly after engaging with activated T cells, signals induced by coinhibitory ligands may drive DC to undergo maturation, enabling them to become highly potent APC in activating naive T cells. Infection by T. rubrum may release fungal products containing the DC-HIL ligands that modulate this DC-HIL/SD-4 pathway.
In sum, our results document direct binding of DC-HIL to the cell wall of dermatophytes to transduce a signal potentiating DC function, indicating that DC-HIL is a PRR for these fungi. Thus, DC-HIL can regulate immune responses in a dual manner: previously we showed it to be an inhibitor of adaptive immunity following ligation of SD-4 on activated T cells, and now we present evidence that it can induce innate immunity against dermatophytes.
We thank Irene Dougherty for technical expertise and Susan Milberger for administrative assistance.
The authors have no financial conflict of interest.
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.
This work was supported by grants from the National Institutes of Health (A164927-01) and from Galderma.
Abbreviations used in this paper: DC, dendritic cell; LC, Langerhans cell; PRR, pattern recognition receptor; SD-4, syndecan-4; BM, bone marrow; HS, heparin/heparan sulfate; Marcksl-1, myristoylated alanine rich protein kinase C substrate like-1; PD-1, programmed cell death-1.