The C-Type Lectin Receptor Mincle Binds to Streptococcus pneumoniae but Plays a Limited Role in the Anti-Pneumococcal Innate Immune Response

The innate immune system employs C-type lectin receptors (CLRs) to recognize carbohydrate structures on pathogens and self-antigens. The Macrophage-inducible C-type lectin (Mincle) is a FcRγ-coupled CLR that was shown to bind to mycobacterial cord factor as well as certain fungal species. However, since CLR functions during bacterial infections have not yet been investigated thoroughly, we aimed to examine their function in Streptococcus pneumonia infection. Binding studies using a library of recombinantly expressed CLR-Fc fusion proteins indicated a specific, Ca2+-dependent, and serotype-specific binding of Mincle to S. pneumonia. Subsequent experiments with different Mincle-expressing cells as well as Mincle-deficient mice, however, revealed a limited role of this receptor in bacterial phagocytosis, neutrophil-mediated killing, cytokine production, and antibacterial immune response during pneumonia. Collectively, our results indicate that Mincle is able to recognize S. pneumonia but is not required for the anti-pneumococcal innate immune response.


Introduction
Streptococcus pneumoniae frequently colonizes the upper respiratory tract of humans. Depending on the immune status of the host, on preceding viral infections, and on the pneumococcal serotype, this asymptomatic colonization can progress to invasive diseases. These diseases, that include community-acquired pneumonia, sepsis, and meningitis, cause significant mortality especially in children and the elderly [1,2]. Important virulence factors of S. pneumoniae are

Materials and Methods
Bacterial strains S. pneumoniae serotype (ST)3 strain PN36 (NCTC7978), ST2 strains D39 and D39Δcps, ST1 multilocus sequence type (MLST)306 strain and ST9N MLST66 strain were used. Bacteria were grown in THY media at 37°C and 5% CO 2 until they reached a phase of logarithmic growth. For heat-inactivation, bacteria where incubated at 56°C for 1 h while shaking. Trichosporon cutaneum was grown in YEPD medium at 26°C for 2-3 days and was then heatinactivated at 80°C for 20 min.

Production of the CLR-Fc fusion protein library
The library of CLR-Fc fusion proteins was prepared as described previously [37][38][39]. Briefly, murine splenic RNA was reverse transcribed into cDNA using Reverse Transcriptase (New England Biolabs, Ipswich, USA). The cDNA encoding the extracellular part of each CLR was amplified by polymerase chain reaction (PCR) and was then ligated into the pFuse-hIgG1-Fc expression vector (InvivoGen, San Diego, USA). The CLR-Fc vector constructs were either stably transfected into CHO cells or transiently transfected using the FreeStyle Max CHO-S Expression System (Life Technologies, Darmstadt, Germany). Purification of the CLR-Fc fusion proteins from the cell supernatant was performed using HiTrap Protein G HP columns (GE Healthcare, Piscataway, USA). The purity of each CLR-Fc fusion protein was confirmed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and subsequent Coomassie stain, Western Blot using anti-human IgG-HRP antibody (Dianova, Hamburg, Germany) as well as mass spectrometry.

Flow cytometric analysis of Mincle-Fc binding to S. pneumoniae
Heat-inactivated bacteria were labeled with SYTO61 Red Fluorescent Nucleic Acid Stain (2.5 µM in PBS; Life Technologies) at RT for 30 min and were afterwards washed three times in PBS. To analyze the Ca 2+ dependency of the interaction, 50 µL labeled bacteria (3×10 8 cells/mL) were incubated with 20 µg/mL Mincle-Fc diluted in lectin binding buffer or EDTA buffer at 4°C for 1 h. After three washing steps with PBS, Mincle-Fc binding to S. pneumoniae was detected with a PE-conjugated goat anti-hFc antibody (Dianova, Hamburg, Germany). Flow cytometric analysis was performed with a FACSCanto II flow cytometer (BD Pharmingen, Heidelberg, Germany). Data were analyzed using the FlowJo analysis software (Tree Star Inc., Ashland, OR, USA).

Cells and infection
Primary cells were isolated from wild-type (WT), Fcerg1 -/-(encoding FcRγ) [40] or Mincle -/- [29] mice on a C57Bl/6J background. Alveolar macrophages (AMFs) were obtained by bronchoalveolar lavage from mouse lungs. For isolation of alveolar epithelial cells (AECs), lung homogenates were prepared and leukocytes and endothelial cells were depleted from the cell suspension by incubation with biotinylated rat anti-mouse CD45, CD16/32 and CD31 (BD Pharmingen, Heidelberg, Germany) followed by magnetic separation. Microvascular endothelial cells (MVECs) were isolated from lung homogenates by positive magnetic selection using biotinylated rat anti-mouse CD144 (BD Pharmingen). For culturing, plates were coated with fibronectin. Bone marrow-derived macrophages (BMMs) were prepared from the bone marrow and cultured in RPMI 1640 containing 30% L929 cell supernatant and 20% FCS for 10 days. Polymorphonuclear leukocytes (PMNs) were isolated from the bone marrow using the anti-Ly6G MicroBead Kit (Miltenyi Biotech, Bergisch Gladbach, Germany). Cells were infected with S. pneumoniae ST2 (D39) or were stimulated with TDM (Sigma, St. Louis, USA) or LPS (Alexis Biochemicals, San Diego, USA).

Bacterial uptake and killing assays
BMMs were infected with S. pneumoniae ST2 (D39Δcps) for 30 min followed by treatment with 50 mg/mL gentamicin. Intracellular, viable bacteria were determined 30 min or 90 min after treatment. Cells were PBS-washed and lysed with 1% saponin for 10 min. Serial dilutions of the bacterial suspensions were plated on blood agar plates, and CFUs were determined. In PMNs, opsonophagocytic killing assays were conducted as described previously [41]. In brief, S. pneumoniae ST2 (D39Δcps) were pre-opsonized with infant rabbit serum (Pel-Freez Biologicals, Rogers, USA) at 37°C for 30 min before PMN were added. The percentage of viable bacteria was determined relative to control reactions lacking neutrophils after incubation at 37°C for 45 min.

Ethics statement
Animal experiments were performed in strict accordance with the German regulations of the Society for Laboratory Animal Science and the European Health Law of the Federation of Laboratory Animal Science Associations. The protocol was approved by the Landesamt für Gesundheit und Soziales Berlin (Permit No. G 0210/11, G 0304/12, G 0357/12). All efforts were made to minimize suffering.

Determination of bacterial load, cell recruitment and cytokines
Bacterial loads were determined in the bronchoalveolar lavage fluid (BALF) and blood. Serial dilutions of samples were plated on blood agar and CFUs were determined. BAL cells were counted by haemocytometer and differentiated by flow cytometry (FACSCalibur; BD) as described previously [13]. Cytokines were quantified by ELISA in the BALF or by quantitative RT-PCR from total cellular RNA of the lung.

ELISA
Concentrations of IL-6, TNFα and KC in the BAL or in cell-free supernatants were quantified by commercially available sandwich ELISA kits (eBioscience, Frankfurt, Germany; R&D, Minneapolis, USA).

Quantitative RT-PCR analysis
Total cellular RNA was isolated, transcribed to cDNA, and amplified by quantitative RT-PCR using Gene Expression Master Mix (Applied Biosystems, Foster City, USA). TaqMan Gene Expression Assays were purchased from Applied Biosystems.

Data Analysis
Data are expressed as mean ± SEM. For CLR-Fc binding studies, statistical analysis was performed using unpaired Student's t-test. For murine S. pneumoniae infection experiments, analysis was performed using the log-rank test for survival, and the Kruskal-Wallis test followed by Dunn's multiple comparison test for comparison of more than two groups. Data analysis was performed using the Prism software (GraphPad Software, La Jolla, CA). For all statistical analyses, p values < 0.05 were considered significant: Ãp < 0.05, ÃÃp < 0.01, ÃÃÃp < 0.001, ÃÃÃÃp < 0.0001.

Mincle binds to S. pneumoniae
To analyze whether CLRs are involved in S. pneumoniae recognition, we performed an initial ELISA screening for CLR binding to heat-killed S. pneumoniae ST3 using a comprehensive library of CLR-Fc fusion proteins [37] (Fig. 1A). Binding of the murine DC-SIGN homolog Specific intercellular adhesion molecule-grabbing nonintegrin receptor 1 (SIGNR1) was used as a positive control since SIGNR1 on marginal zone macrophages was previously reported to be crucial for S. pneumoniae recognition [42]. Indeed, SIGNR1-Fc exhibited substantial binding to plate-bound S. pneumoniae (Fig. 1A). Besides SIGNR1-Fc, the ELISA-based pre-screening revealed Mincle as a candidate CLR that bound to heat-killed S. pneumoniae (Fig. 1A). In contrast, numerous other CLR-Fc fusion proteins, including MCL, DCAR, DCIR, CLEC-9a, MICL, CLEC-12b, SIGNR3, and MGL1 exhibited no or very weak binding to plate-bound S. pneumoniae. To confirm the specificity of the Mincle/S. pneumoniae ST3 interaction, we performed flow cytometric binding assays ( Fig. 1B-C). In agreement with the ELISA-based binding assay, flow cytometric analysis indicated substantial binding of Mincle-Fc to S. pneumoniae ST3 (Fig. 1B). Furthermore, we performed comparative binding studies with Mincle-Fc and other bacteria and fungi. In a previous study, the fungus T. cutaneum did not induce Mincle activation in a Mincle reporter cell line-based assay [32]. Indeed, whereas a high percentage of S. pneumoniae was recognized by Mincle-Fc, we observed only marginal binding of Mincle-Fc to T. cutaneum. These findings indicate that Mincle specifically recognizes S. pneumoniae.
The Mincle/S. pneumoniae interaction is Ca 2+ -dependent and serotype specific Next, we determined whether the Mincle-Fc binding to S. pneumoniae was mediated in a Ca 2+ -dependent manner. Previously, it was shown that Mincle binds to its known ligands TDM (cord factor) and TDB in a Ca 2+ dependent fashion [24,25]. Pre-incubation of Mincle-Fc in a buffer containing the Ca 2+ chelating agent EDTA resulted in reduced binding of Mincle-Fc to plate-bound TDB as well as S. pneumoniae ST3 suggesting a Ca 2+ -dependent Mincle/S. pneumoniae interaction ( Fig. 2A). This finding was corroborated by flow cytometry since the Mincle-Fc incubation in an EDTA-containing buffer markedly reduced its binding to  S. pneumoniae (Fig. 2B). Next, we performed an ELISA-based competition assay with increasing concentrations of heat-killed S. pneumoniae ST3 to disrupt the Mincle-Fc binding to the Mincle ligand TDB (Fig. 2C). Indeed, Mincle-Fc incubation in the presence of S. pneumoniae led to reduced binding to TDB, thus confirming the specificity of the Mincle/S. pneumoniae interaction. To analyze if Mincle binding to S. pneumoniae is serotype specific, we performed binding studies with different S. pneumoniae serotypes (ST1, ST2, ST3 and ST9N). Indeed, we observed a differential binding of Mincle-Fc to different serotypes. Mincle-Fc exhibited a strong binding to S. pneumoniae ST2 (D39) and ST3 (PN36), weaker binding to ST9N and almost no binding to ST1 (S1 Fig.). Thus, we conclude that Mincle-Fc binding to S. pneumoniae is serotype-specific suggesting a carbohydrate-specific binding of Mincle to S. pneumoniae.
Mincle is not required for production of inflammatory cytokines, phagocytosis or bacterial killing upon S. pneumoniae infection Next, we examined the expression and function of Mincle in different cell types. We found that Mincle expression is low in alveolar epithelial cells and lung microvascular endothelial cells, and higher in alveolar macrophages, bone-marrow macrophages (BMMs) and neutrophils (Fig. 3A). Mincle expression was up-regulated upon pneumococcal infection in lung epithelial cells, endothelial cells and macrophages. Lack of Mincle or FcRγ (Fcerg1-/-) in alveolar macrophages and BMMs did not significantly affect the S. pneumoniae ST2-induced production of TNFα or KC (Fig. 3B-D). In contrast, cytokine production stimulated by TDM, which served as a positive control, was abolished in Mincle -/and Fcer1g -/cells cells (Fig. 3E). Moreover, phagocytosis and killing of S. pneumoniae by macrophages and neutrophils were not significantly affected by Mincle or FcRγ deficiency (Fig. 3F-G). Thus, our data suggest that Mincle is not required for the antibacterial innate immune response during pneumococcal pneumonia Although Mincle signaling did not affect innate immune responses of different cell subsets to S. pneumoniae in vitro, a Mincle-dependent bacterial detection in other cell types, or alternatively Mincle-mediated responses to endogenous danger molecules such as SAP130 [31] could regulate the antibacterial immune response in vivo. We therefore examined the function of Mincle and its adapter molecule FcRγ during pneumococcal pneumonia. Intranasal infection of wild-type mice with S. pneumoniae ST3 led to increased expression of Mincle in the whole lung (Fig. 4A). However, Mincle -/-, Fcer1g -/and wild-type mice did not significantly differ in their survival following S. pneumoniae infection (Fig. 4B). Moreover, mice lacking Mincle or its adapter molecule displayed unaltered bacterial loads in the bronchoalveolar lavage fluid (BALF) as compared to wild-type animals (Fig. 4C). To evaluate systemic dissemination of S. pneumoniae in these animals, bacterial loads in blood and spleen were determined. We detected high bacterial numbers in blood and spleens of all groups of infected mice 48 h p.i. (Fig. 4D and data not shown). According to the unaltered bacterial loads, S. pneumoniaeinfected Mincle -/-, Fcer1g -/and wild-type mice exhibited a similar recruitment of neutrophils and macrophages to the lung (Fig. 4E-F). In addition, S. pneumoniae-induced production of cytokines and chemokines, many of which have been previously associated with Mincle activation in different models [24,31], were not significantly affected by lack of Mincle or FcRγ (Fig. 4G-K). These data collectively indicate that Mincle signaling is not required for antibacterial innate immune responses and resistance to S. pneumoniae ST3 in vivo.

Discussion
Myeloid CLRs are an important class of PRRs expressed by various immune cells and capable of detecting a broad spectrum of microbial and endogenous ligands. However, their function in acute bacterial infection has not been investigated systematically. For S. pneumoniae infection, SIGNR1 is one of the few CLRs shown to be crucial for bacterial recognition, phagocytosis as well as antibacterial defense [43,44]. We therefore set out to identify additional CLRs that are able to bind to S. pneumoniae. To this end, we used a comprehensive library of CLR-Fc fusion proteins covering immunologically relevant members of the Dectin-1 and DCIR family [37]. CLR-Fc fusion proteins are useful tools to identify pathogen/CLR interactions as they display the CRD in a dimeric fashion. Dimeric display allows for multivalent ligand binding and has helped to unravel novel CLR interactions with microbes or endogenous ligands [45,46]. Out of our CLR-Fc library, we identified Mincle as a novel binder of S. pneumoniae, in addition to the known pneumococcal recognition by SIGNR1. We demonstrate that the Mincle/S. pneumoniae interaction occurs in a Ca 2+ -dependent fashion, as has been previously shown for the Mincle ligand TDM and pathogenic fungi [24,30,32].
To determine whether the Mincle/S. pneumoniae interaction impacts the innate immune response, we employed various primary immune cells and the murine infection model of pneumococcal pneumonia. However, we did not observe a functional consequence of this interaction for S. pneumoniae-induced innate responses in vitro and in vivo. Our finding that Mincle and FcRγ-deficiencies did not affect cytokine production and phagocytosis in macrophages suggests that binding of S. pneumoniae to Mincle does not correlate with down-stream signaling. Alternatively, the loss of Mincle signaling in macrophages, neutrophils and mice might have been compensated by other functionally redundant CLRs not coupled to FcRγ, as commonly observed in different infection models using CLR-deficient cell and mouse lines [16]. In addition, PRRs such as the TLRs and NLRs known to be essential for antibacterial immunity against S. pneumoniae [2,47] might have overcome potential Mincle-mediated effects. It appears reasonable that the recognition of prokaryote-specific PAMPs such as peptidoglycan and pore-forming toxins by those receptors might dominate the innate immune response to bacteria. In contrast, the recognition of broadly expressed carbohydrate ligands might play a more important and non-redundant role in infections with eukaryotic pathogens such as fungi and parasites [20,48].
Mincle is known to recognize different glycolipids in M. tuberculosis and fungi [24,27,30]. Given that glycolipids such as lipoteichoic acid are also present in S. pneumoniae [49,50], it is reasonable to speculate that those structures mediate the binding to Mincle. In addition, Mincle also binds to the endogenous protein SAP130, indicating that besides glycolipids nonglycosylated ligands may be recognized by Mincle as well [31]. This renders the recognition of a protein ligand present in S. pneumoniae possible, although it appears unlikely since all binding studies were performed using pneumococci after heat inactivation. The specificity of the S. pneumoniae/Mincle interaction was proven by different observations. First, we detected no or only marginal binding of most other CLR-Fc fusion proteins of the library to the bacterium, second the interaction was Ca 2+ -dependent, and third, it could be competitively inhibited. Considering this specific interaction, it may be interesting to elucidate in future studies whether combined deficiencies in Mincle and other CLRs and/or cross-talk mechanisms between Mincle and other PRRs act in a synergistic manner.