Volume 45, Issue 6, Pages 1245-1257 (December 2016) C-Type Lectin Receptor DCAR Recognizes Mycobacterial Phosphatidyl-Inositol Mannosides to Promote a Th1 Response during Infection  Kenji Toyonaga, Shota Torigoe, Yoshitomo Motomura, Takane Kamichi, Jennifer M. Hayashi, Yasu S. Morita, Naoto Noguchi, Yasushi Chuma, Hideyasu Kiyohara, Kazuhiro Matsuo, Hiroshi Tanaka, Yoshiko Nakagawa, Tetsushi Sakuma, Masaki Ohmuraya, Takashi Yamamoto, Masayuki Umemura, Goro Matsuzaki, Yasunobu Yoshikai, Ikuya Yano, Tomofumi Miyamoto, Sho Yamasaki  Immunity  Volume 45, Issue 6, Pages 1245-1257 (December 2016) DOI: 10.1016/j.immuni.2016.10.012 Copyright © 2016 Elsevier Inc. Terms and Conditions

Immunity 2016 45, 1245-1257DOI: (10.1016/j.immuni.2016.10.012) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 1 DCAR Recognizes Mycobacterial Glycolipids (A) Reporter cells expressing FcRγ alone or DCAR + FcRγ were stimulated with the indicated concentrations of M. tuberculosis H37Rv, M. tuberculosis H37Ra, or M. bovis BCG for 24 hr. Induction of NFAT-GFP was analyzed by flow cytometry. (B) Reporter cells expressing Mincle + FcRγ or DCAR + FcRγ were stimulated with M. tuberculosis H37Ra, CHCl3:MeOH:H2O (C:M:W)-treated M. tuberculosis H37Ra, or plate-coated C:M:W extract for 24 hr. Induction of NFAT-GFP was analyzed by flow cytometry. (C) The C:M:W extract was analyzed by HPTLC using C:M:W (65:25:4, v/v/v), stained with copper-acetate reagent, and divided into 20 subfractions (left). Each subfraction was coated onto a plate to stimulate reporter cells (right). (D) The C:M:W extract was analyzed by two-dimensional HPTLC (2D-TLC) using C:M:W (65:30:6, v/v/v) in the first dimension and C:M:AcOH (A):W (80:12:15:4, v/v/v/v) in the second dimension. The HPTLC plates were stained with copper-acetate (left), orcinol (middle), or molybdenum blue (right) reagent. (E) Reporter cells expressing DCAR + FcRγ were stimulated with the indicated concentrations of plate-coated purified lipids for 24 hr. Induction of NFAT-GFP was analyzed by flow cytometry. Data are presented as mean ± SD and are representative of three separate experiments with similar results. See also Figure S1. Immunity 2016 45, 1245-1257DOI: (10.1016/j.immuni.2016.10.012) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 2 Purification of DCAR Ligands from Mycobacteria (A) A crude PIM fraction from M. bovis BCG was subjected to silica gel column chromatography and eluted with C:M:W (80:20:1 to 50:50:1, v/v/v) to yield 104 fractions. Reporter cells expressing DCAR + FcRγ were stimulated with individual plate-coated fractions for 24 hr (top). Each fraction was further analyzed by HPTLC using C:M:W (60:35:8, v/v/v) followed by staining with copper-acetate (middle) or orcinol (bottom) reagent. W, whole fraction. (B) Fraction 56 was further divided into lipids 1 and 2 by HPTLC using C:M:A:W (80:12:15:4, v/v/v/v). The purified lipids 1 and 2 were analyzed by HPTLC using C:M:W (60:35:8, v/v/v) or C:M:A:W (80:12:15:4, v/v/v/v). The HPTLC plates were stained with copper-acetate (left) or orcinol (right) reagent as indicated. (C) Reporter cells expressing DCAR + FcRγ were stimulated with the indicated concentrations of plate-coated purified lipids 1 (fraction 56), 2 (fraction 56), 3 (fraction 53), and 4 (fraction 10) for 24 hr. Induction of NFAT-GFP was analyzed by flow cytometry. Data are presented as mean ± SD and are representative of three separate experiments with similar results. See also Figure S2. Immunity 2016 45, 1245-1257DOI: (10.1016/j.immuni.2016.10.012) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 3 Identification of AcPIM2 and Ac2PIM2 as DCAR Ligands (A–D) Structural analysis of purified lipids 1 (A), 3 (B), 2 (C), and 4 (D). The 1H NMR spectra were measured in CDCl3:CD3OD:D2O (60:35:8, v/v/v) at 304 K (top panels). The ESI-TOF-MS spectra of purified lipids were shown (middle panels). The molecular-related ion peaks were shown in each panel. Chemical structures of AcPIM2 (A), Ac2PIM2 (B), PI (C), and CL (D) from M. bovis BCG (bottom panels). (E) Reporter cells expressing Mincle, Dectin-2, or DCAR were stimulated with the indicated concentrations of plate-coated synthetic AcPIM2 for 24 hr. Induction of NFAT-GFP was analyzed by flow cytometry. (F) Human IgG1-Fc (Ig), Mincle-Ig, Dectin-2-Ig, or DCAR-Ig were incubated with the indicated concentrations of plate-coated synthetic AcPIM2. Bound proteins were detected with anti-hIgG-HRP. Data are presented as mean ± SD and are representative of three separate experiments with similar results. See also Figure S3. Immunity 2016 45, 1245-1257DOI: (10.1016/j.immuni.2016.10.012) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 4 Expression of DCAR on Monocyte-Derived Cells (A) Resident peritoneal exudate cells (PECs) were stained with anti-mouse B220, CD3, CD11b, MHC class II, CD11c (HL3), and CD115. B220−CD3− cells were further divided into R1, R2, R3 (small peritoneal macrophages; SPM), and R4 (large peritoneal macrophages; LPM) populations. (B) PECs from WT (upper panels) or Clec4b1 deficient (Clec4b1−/−) (lower panels) mice were stained with anti-mouse DCAR (2B8; filled histograms) in combination with Abs used in (A). Biotinylated mouse IgG1, κ was used as an isotype-matched control (open histograms). (C) PECs from WT, Clec4b1−/−, or Fcer1g deficient (Fcer1g−/−) mice were stained with anti-mouse DCAR (2B8) in combination with the same Abs used in (A). Surface expression of DCAR on B220−CD3− cells is shown. R4 (B220−CD3−CD11bhiSSChi cells) populations in (A) were excluded. (D) Splenocytes from WT mice were stained with anti-mouse NK1.1, Ly6G, CD11b, CD11c (N418), CD8α, MHC class II, and CD64. NK1.1−Ly6G− cells were further divided into conventional DCs (CD11chi: NK1.1−Ly6G−CD11b+CD11chiCD8α±), monocyte-derived inflammatory cells (CD11cint: NK1.1−Ly6G−CD11b+CD11cint), and monocytes/macrophages (CD11c−: NK1.1−Ly6G−CD11b+CD11c−). (E) Splenocytes from WT (upper panels) or Clec4b1−/− (lower panels) mice were stained with anti-mouse DCAR (2B8; filled histograms) in combination with the same Abs used in (D). Biotinylated mouse IgG1, κ was used as an isotype-matched control (open histograms). (F) Cells were isolated from enzymatically digested lungs of WT mice and stained with anti-mouse NK1.1, Ly6G, CD11b, CD11c (N418), and MHC class II. NK1.1−Ly6G− cells were further divided into CD11bintCD11chi (alveolar macrophages), CD11b+CD11c+MHC class II+, and CD11b+CD11c−MHC class II− cell populations. (G) Total lung cells from WT or Clec4b1−/− mice were stained with anti-mouse DCAR (2B8; filled histograms) in combination with the same Abs used in (F). Biotinylated mouse IgG1, κ was used as an isotype-matched control (open histograms). Data are presented as mean ± SD and are representative of three separate experiments with similar results. See also Figure S4. Immunity 2016 45, 1245-1257DOI: (10.1016/j.immuni.2016.10.012) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 5 DCAR Mediates PIMs-Induced Innate Immune Responses (A–C) PECs from WT, Clec4b1−/− or Fcer1g−/− mice were stimulated with indicated concentrations of plate-coated AcPIM2, Ac2PIM2, or LAM for 48 hr. The concentrations of MCP-1 (A), MIP-2 (B), and IL-6 (C) were determined by ELISA. ∗p < 0.05 compare to WT mice. (D–F) WT and Clec4b1−/− mice were injected with 100 μg of Ac2PIM2 intraperitoneally. At 4 hr after injection, the concentrations of MCP-1 (D), MIP-2 (E), and IL-6 (F) in the peritoneal exudate were determined by ELISA. Each symbol represents an individual mouse. ∗p < 0.05. Data are presented as mean ± SD of triplicate assays from an individual mouse and are representative of three separate experiments with similar results (A, B, and C). See also Figure S5. Immunity 2016 45, 1245-1257DOI: (10.1016/j.immuni.2016.10.012) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 6 Impaired Immune Responses in Clec4b1−/− Mice during Mycobacterial Infection (A) WT and Clec4b1−/− mice were infected i.p. with 1 × 106 CFU of M. bovis BCG. At 2 weeks after infection, the frequency of SPM (CD11b+MHC class IIhi cells) in the peritoneal cavity from uninfected or infected mice was determined by flow cytometry. Each symbol represents an individual mouse. (B) The concentration of MCP-1 in the peritoneal exudate from uninfected or infected mice at 2 weeks after infection were determined by multiplexed bead-based immunoassay. (C) The numbers of CD11b+CD11c+CD64+ cells in the lungs, mediastinal lymphnodes (MLN), and spleen from uninfected or infected mice were determined by flow cytometry. (D) The concentrations of IFNγ and IL-12p40 in the peritoneal exudate from uninfected or infected mice at 2 weeks after infection. (E and F) PECs (E) and splenocytes (F) from infected mice were stimulated with the indicated concentrations of PPD for 4 days. The concentration of IFNγ in the supernatants was determined by ELISA. (G) The bacterial burden in the peritoneal cavities of WT and Clec4b1−/− mice were determined 2 weeks after infection. (H) WT and Clec4b1−/− mice were infected intratracheally with 1 × 106 CFU of M. bovis BCG. At 2 weeks after infection, MLNs were collected from infected mice and stimulated with the indicated concentrations of PPD for 4 days. The concentration of IFNγ in the supernatants was determined by ELISA. (I) The bacterial burden in the lungs of WT and Clec4b1−/− mice were determined 2 weeks after M. bovis BCG infection. (J) WT and Clec4b1−/− mice were infected intratracheally with 1 × 104 CFU of M. tuberculosis H37Rv. At 2 weeks after infection, the concentration of MCP-1 in the lung homogenates from infected mice was determined by ELISA. Data are presented as mean ± SD (E, F, and H) and are representative of two independent experiments with similar results. ∗p < 0.05, ∗∗p < 0.01. See also Figure S6. Immunity 2016 45, 1245-1257DOI: (10.1016/j.immuni.2016.10.012) Copyright © 2016 Elsevier Inc. Terms and Conditions