Bruton's Tyrosine Kinase (BTK) and Vav1 Contribute to Dectin1-Dependent Phagocytosis of Candida albicans in Macrophages

β-glucan exposure on Candida albicans yeasts and hyphae

To facilitate imaging of phagocytosis, we applied two new imaging tools. First, we used a Candida strain that expresses a variant of blue fluorescent protein (Candida-BFP). Second, to study β-glucan exposure on C. albicans yeasts and hyphae, we site-specifically fluorescently labeled the extracellular carbohydrate recognition domain of Dectin-1 (Dectin1-CRD-Alexa647) using the bacterial enzyme sortase [7]. Candida-BFP was incubated in DMEM with 10% serum for 15, 30, 90 and 180 min followed by staining with Dectin1-CRD-Alexa647 ( Figure 1A ). At 15 and 30 min we observed moderately stained C. albicans yeasts with increased staining of bud scars (white arrows), congruent with increased β-glucan exposure at these sites [13]. At 90 and 180 min, formation of hyphae was extensive with strong, homogeneous β-glucan exposure. Under these growth conditions, when compared to C. albicans yeast, C. albicans hyphae thus expose higher levels of β-glucan, which is expected to affect signaling through Dectin-1.

Identification of Dectin1-interacting proteins

To identify proteins that interact with Dectin-1 during phagocytosis of live C. albicans, we performed immuno-isolation experiments using the RAW-Dectin1 macrophage cell line (Esteban et al., 2011). The macrophages were co-incubated with live C. albicans for one hour, a time point that marks early hyphal formation and increased β-glucan exposure. Cells were lysed and Dectin-1, together with its interacting partners, was immunoprecipitated using anti-HA antibodies. Proteins present in these samples were analyzed and identified by SDS-PAGE, followed by LC-MS-MS. From the list of proteins, we selected two that were retrieved in complex with Dectin-1 in the samples with C. albicans, but that were absent from the control sample. These proteins were Bruton's Tyrosine Kinase (BTK) and the guanine nucleotide exchange factor Vav1.

To investigate the expression of BTK, Vav1 and other proteins already known to be involved in Dectin1-mediated phagocytosis, we incubated RAW-Dectin1 macrophages with live C. albicans. Phagocytosis continues throughout the period of coincubation, but C. albicans morphology changes from yeast form to hyphal form over time. Total cell lysates from the different time points were analyzed by immunoblot. Spleen tyrosine kinase (Syk), a known interactor of Dectin-1, was present at constant levels at all-time points ( Figure 1B ). Phosphorylation of Syk increased at 15 min and then waned. BTK and Vav1 were likewise present at constant levels. PLCγ2, a key enzyme in phagocytosis, was present at constant levels throughout the time course; its phosphorylation (at Y1217) was most pronounced around 90 and 120 min.

Next we confirmed the interaction between Dectin-1 and BTK, and between Dectin-1 and Vav1. Immunoprecipitation with polyclonal anti-BTK antibody, followed by immunoblotting with anti-HA, showed an increased interaction between BTK and Dectin-1 starting at 15 min, which gradually increased ( Figure 1C ). The interaction between Vav1 and Dectin-1 is strongest at the later time points ( Figure 1C ). Quantification of the Dectin-1/BTK and Dectin-1/Vav1 interactions using ImageJ software showed that the level of the Dectin-1/BTK complex peaks at 60–90 min, while the strongest Dectin-1/Vav1 interaction was observed at 180 min ( Figure 1D ). We conclude that BTK and Vav1 are expressed at constant levels, and that their interactions with Dectin-1 are strongest at the later time points when C. albicans has formed hyphae.

Confirmation of BTK and Vav1 localization to the phagocytic cup

To study the subcellular localization of BTK, Vav1 and Syk, we constructed stable RAW macrophage cell lines that express these proteins as N- or C-terminal mCherry fusions in the RAW-Dectin1 background. In unstimulated cells, BTK-mCherry and mCherry-Vav1 were cytosolic, whereas mCherry-Syk localized to both the cytosol and nucleus ( Figure 2 , top). Next the distribution of the mCherry-tagged proteins was investigated after coincubation with Candida-BFP for 30, 90 or 180 min. At 30 min, when C. albicans yeast-form cells are present, BTK-mCherry, mCherry-Vav1 and mCherry-Syk showed a clear localization to the Candida-BFP-containing phagocytic cup ( Figure 2A , arrows). After 90 and 180 min, when C. albicans had formed extensive hyphae, BTK-mCherry and mCherry-Vav1 showed enrichment in a cuff region of the phagocytic cup when engulfing Candida-BFP hyphae ( Figure 2A , arrows). mCherry-Syk was more evenly distributed along the phagosomal membrane, with additional foci of enrichment outside the cuff region. There was no enrichment of BTK-mCherry or mCherry-Vav1 in membranes of closed, more mature, phagosomes. N- or C-terminal mCherry fusions with vimentin, expressed as controls in the same RAW macrophage cell line, did not show recruitment to the phagocytic cup (data not shown). 3D reconstructions of RAW macrophages in the process of ingesting C. albicans hyphae showed that the cuff regions of BTK/Vav1/Syk recruitment are cylindrical sleeves surrounding the phagosomal membrane ( Figure 2B ). Recruitment of BTK-mCherry, mCherry-Vav1 and mCherry-Syk to the phagocytic cup was quantified by comparing the fluorescence intensity in the cup/phagosomal membrane to that of the cytosol. Recruitment of mCherry-Syk was strongest, followed by mCherry-Vav1 and BTK-mCherry ( Figure 2C ).

Is recruitment of BTK and Vav1 to the phagocytic cup dependent on Dectin-1? We incubated the RAW mCherry cell lines with β-glucan-coated beads or with zymosan, the phagocytosis of which is Dectin1-dependent [3]. All three fusion proteins localized to phagosomes containing β-glucan-coated beads or zymosan. Recruitment of BTK, Vav1 and Syk therefore indeed relies on Dectin-1 (Figure S1A and B). To address the difference in geometry of yeast versus hyphal particles, we incubated the RAW mCherry cell lines with UV-killed C. albicans yeasts or hyphae. While the shape of the phagocytic cup differs, recruitment of BTK, Vav1 and Syk occurs in all cases (Figure S1C and D). Regardless of the geometry of the ingested particle, BTK and Vav1 are recruited to the phagocytic cup but not to the mature phagosome during Dectin1-mediated phagocytosis ( Figure 2D ). Immunofluorescence was performed to investigate the localization of Dectin1 during phagocytosis of C. albicans hyphae. Dectin1 was enriched in the cuff region of the phagocytic cup to which BTK/Vav1/Syk were recruited, but also showed areas of enrichment outside the cuff region ( Figure 3 ). Phagosomes that contain completely internalized C. albicans yeasts or hyphae showed very little Dectin1 staining.

Phosphoinositide metabolism during C. albicans phagocytosis

Different PIs are present in the phagosomal membrane when the phagocytic cup forms, as well as during phagosomal maturation. To visualize membrane PI composition, PI-binding protein domains fused with fluorescent proteins have been used as imaging tools (biosensors or probes) [14]. We transfected the RAW-Dectin1 cell line with biosensors for PI(4,5)P₂ (PH-PLCδ-GFP), PI(3,4,5)P₃ (PH-BTK-GFP), PI(3,4,5)P₃/PI(3,4)P₂ (PH-Akt-RFP) and DAG (C1-PKCδ-GFP) to study their formation during C. albicans phagocytosis.

PI(4,5)P₂ was present in the plasma membrane at rest, but multiple regions of enrichment were observed at 30 and 90 min at sites where macrophages contacted Candida-BFP yeast or hyphae (Figure S2). Membranes of sealed phagosomes no longer showed PI(4,5)P₂ enrichment, consistent with the reported localization of PI(4,5)P₂ during FcγR-medicated phagocytosis [10]. PI(3,4,5)P₃ as visualized by the BTK-PH domain showed enrichment in some, but not all, of the PI(4,5)P₂-rich regions (Figure S2, arrows).

Next we investigated the localization of PI(3,4,5)P₃/PI(3,4)P₂ and DAG. High levels of both PIs were present in cups or phagosomes containing C. albicans. PI(3,4,5)P₃/PI(3,4)P₂ and DAG colocalized in some early phagosomes, but phagosomes containing only PI(3,4,5)P₃/PI(3,4)P₂ or only DAG were also present ( Figure 4A , white and green arrows). Recruitment of PH-Akt-RFP (PI(3,4,5)P₃/PI(3,4)P₂) and C1-PKCδ-GFP (DAG) or both to Candida-containing cups and phagosomes was quantified at the 30-min time point ( Figure 4B ). More than 20% of cups/phagosomes were PI(3,4,5)P₃/PI(3,4)P₂-positive and a comparable percentage was positive for both PI(3,4,5)P₃/PI(3,4)P₂ and DAG. At this stage 4% of cups/phagosomes were positive for DAG only. PI(3,4,5)P₃/PI(3,4)P₂ were more prominent in phagocytic cups, while DAG was more abundant in sealed phagosomes. These results indicate that during Dectin1-mediated phagocytosis of β-glucan-exposing C. albicans yeast or hyphae, PI(3,4,5)P₃/PI(3,4)P₂ are formed early during initiation of phagocytosis and that DAG-rich membranes/phagosomes appear at a more advanced stage ( Figure 4C ).

The presence of different PIs allows the sequential docking of a specialized set of effector proteins to the membrane during phagosomal maturation. DAG is a potent second messenger, activating members of the Protein Kinase C (PKC) family. We transfected the RAW-Dectin1 cell line with GFP-tagged PKCα, PKCβ1, PKCδ, PKCε or PKCζ to determine their localization during C. albicans phagocytosis. They all localized to the C. albicans-containing phagocytic cup ( Figure 4D ), albeit to varying degrees. The recruitment of PKCδ and PKCε was most pronounced, while the recruitment of PKCα, PKCβ and PKCζ was moderate ( Figure 4E ).

Colocalization of BTK and Vav1 with PI(3,4,5)P but not with DAG

BTK and Vav1 both contain a Pleckstrin Homology (PH) domain that can bind to PI(3,4,5)P₃. Vav1 also contains a putative DAG-binding C1 domain, incapable of binding DAG owing to the presence of hydrophilic and non-charged residues in key binding positions [15]. The Syk polypeptide does not contain predicted PH or C1 domains and presumably localizes to the phagocytic cup through interaction of its tandem SH2 domains with the phosphorylated ITAM-like motif of Dectin-1. We investigated the possible colocalization of BTK, Vav1 and Syk with PI(3,4,5)P₃ and/or DAG during phagocytosis of C. albicans. The BTK-mCherry, mCherry-Vav1 and mCherry-Syk RAW cell lines were transfected with the biosensor construct PH-BTK-GFP to visualize formation of PI(3,4,5)P₃ and then incubated with Candida-BFP. The BTK-mCherry and mCherry-Vav1 fusion proteins colocalized with PI(3,4,5)P₃ at the phagocytic cup, suggesting binding ( Figure 5A,B ), while recruitment of mCherry-Syk was more diffuse, confirming PI(3,4,5)P₃-independent recruitment of Syk ( Figure 5C , arrows). Next, we examined colocalization of BTK, Vav1 and Syk with DAG using the C1-PLCδ-GFP biosensor. BTK-mCherry showed some colocalization with DAG at 30 min. After 90 min, BTK-mCherry and DAG clearly localized to different regions of the phagocytic cup ( Figure 5D , arrows). mCherry-Vav1 and mCherry-Syk showed a similar distribution, and neither colocalized with DAG ( Figure 5E,F ). BTK and Vav1 thus bind to PI(3,4,5)P₃ and not to DAG during phagocytosis of C. albicans, consistent with the presence of PH domains in both proteins and a non-DAG binding C1 domain in Vav1. The colocalization of BTK and Vav1 with PI(3,4,5)P₃ suggests a role for these proteins at an early stage of phagocytosis, as PI(3,4,5)P₃ marks cups and immature phagosomes (see above).

BTK and Vav1 localize to areas of F-actin formation at the phagocytic cup

Rearrangement of the actin cytoskeleton drives phagocytosis, enabling engulfment of the fungal particle by the macrophage. We investigated cytoskeletal changes during Candida phagocytosis by electron microscopy using fixation with tannic acid, a method that preserves actin structures [16]. RAW-Dectin1 macrophages were incubated with C. albicans for 30 min. Areas of decreased staining intensity were observed surrounding the C. albicans-containing phagocytic cups, indicative of actin polymerization ( Figure 6A ). These cuffs had a smooth appearance, were of uniform thickness across the sections examined and were distinct from the cytosol, which was more granular in appearance. We also examined actin polymerization with the biosensor LifeAct, which reports on the distribution of filamentous (F-) actin, in combination with biosensors that bind to PI(3,4,5)P₃ and DAG. F-actin formation was detectable at the phagocytic cup of Candida-BFP yeast and hyphae at 30 and 90 min. There is a clear separation of DAG-rich regions of the phagosome from the F-actin rich membranes ( Figure 6B ). This suggests regional membrane specializations, with different functionalities and different peripheral proteins associated with them. F-actin showed perfect colocalization with PI(3,4,5)P₃ at the Candida-BFP containing phagocytic cup ( Figure 6C ). Also, BTK-mCherry, mCherry-Vav1 and mCherry-Syk colocalize with F-actin after 30 and 90 min of C. albicans phagocytosis ( Figure 6D ). We conclude that PI(3,4,5)P₃-rich areas are formed in the course of Dectin1-mediated phagocytosis of C. albicans and that BTK and Vav1 are recruited to these areas, with ensuing formation of F-actin.

Inhibition of BTK affects phagocytosis and production of DAG

Having established interactions of BTK and Vav1 with Dectin-1 in the course of phagocytosis and their recruitment to PI(3,4,5)P₃-rich membranes, we next investigated the functional importance of these proteins in phagocytosis of C. albicans. We synthesized the highly selective irreversible BTK inhibitor PCI-32765 [17]. The IC₅₀ of this BTK inhibitor for BTK, Tec kinase and Syk is 0.5, 78 and >10,000 nM, respectively, corresponding to a BTK selectivity of 156 fold (Tec) and >10,000 fold (Syk) [17]. In B cells, PCI-32765 irreversibly inhibited autophosphorylation of BTK (IC₅₀: 11 nM), phosphorylation of BTK's physiological substrate PLCγ (IC₅₀: 29 nM), and phosphorylation of downstream kinase ERK (IC₅₀: 13 nM) [17]. RAW-Dectin1 macrophages were pre-incubated with different concentrations of BTK inhibitor (starting from 50 nM), followed by coincubation with Candida-BFP for 1 hour and staining with fluorescently labeled Concanavalin A to distinguish accessible (extracellular) particles from internalized Candida-BPF particles. Preincubation of RAW-Dectin1 macrophages with 50 nM PCI-32765 reduced uptake of C. albicans by 30% and increasing inhibitor concentrations progressively blocked phagocytosis ( Figure 7A ). These results indicate an important role for BTK during C. albicans phagocytosis.

PLCγ, which converts PI(4,5)P₂ into DAG, is a key enzyme during phagocytosis by innate immune cells and can be regulated by BTK [18], [19]. We hypothesized that during phagocytosis of C. albicans, early localization of BTK to the phagocytic cup activates PLCγ to increase synthesis of DAG and so enables recruitment of PKC members to the phagosomal membrane. To address this possibility, RAW-Dectin1 macrophages were incubated with BTK inhibitor or the PLC inhibitor U73112, followed by coincubation with C. albicans for 1 hour. Total DAG levels in these macrophages were measured by lipid extraction, DAG kinase assays and quantification of the product, phosphatidic acid, by thin layer chromatography. Addition of the PLC inhibitor U73112 reduced total DAG levels by 50% while the BTK inhibitor did not significantly affect total DAG levels ( Figure 7B and C ). However, in samples without inhibitor we observed a small increase in DAG levels in the presence of C. albicans, possibly due to increased production of DAG at the phagocytic cup during phagocytosis. To examine the effect of the BTK inhibitor on local production of DAG at the phagocytic cup, we performed confocal microscopy with the C1-PLCδ-GFP biosensor and the PKCε-GFP construct in the absence and presence of the BTK inhibitor. Addition of the inhibitor strongly reduced recruitment of the C1-PLCδ-GFP biosensor and the PKCε-GFP construct to the phagocytic cup ( Figure 7D and E ). These results underscore the importance of investigating local changes in lipid composition as opposed to changes in total levels. BTK is thus involved in the production of DAG and the subsequent recruitment of PKC family proteins at the phagocytic cup ( Figure 7F ). The interactions of BTK and Vav1 with Dectin-1, their recruitment to PI(3,4,5)P3-rich membranes, the possible contribution of Vav1/BTK to actin rearrangements and the role of BTK in the production of DAG are summarized in Figure 7G .

btk−/− and vav1−/− macrophages display reduced phagocytosis

To further investigate the contributions of BTK and Vav1 to phagocytosis we determined the phagocytic capacity of peritoneal macrophages and bone marrow-derived macrophages (BMDMs) from wild type mice and from dectin-1, btk and vav1 knockout mice. Phagocytic indices were determined after incubation of peritoneal macrophages with zymosan for 30 min or with C. albicans for 30 min or 1 hour. While Dectin1-deficient peritoneal macrophages did not phagocytose zymosan, the btk and vav1 deficient cells showed an intermediate phenotype ( Figure 8A ). Uptake of C. albicans by btk and vav1 knockout peritoneal macrophages at the 30-min time point was also significantly reduced compared to wild type cells (p<0.05). In addition, the btk−/− and vav1−/− BMDMs displayed a reduction in phagocytosis similar to that seen for the dectin1−/− BMDMs (p<0.05) ( Figure 8A ). BTK and Vav1 are thus important contributors to Dectin1-mediated uptake of C. albicans by BMDMs.

btk−/− and vav1−/− mice show increased sensitivity to C. albicans infections

In vivo immune responses of wild type and dectin1−/−, btk−/− and vav1−/− mice to C. albicans were tested in a systemic candidiasis model. Tail vein injection of the four groups of mice with C. albicans showed that dectin1−/− mice were most susceptible to C. albicans infections, while the majority of the wild type animals survived the systemic infection. Btk−/− and vav1−/− animals displayed an intermediate phenotype, both during systemic infection with 5×10⁴ colony forming units (CFU) ( Figure 8B ) and with 1×10⁵ colony forming units (CFU) of C. albicans (data not shown). We found no statistical difference in C. albicans loads in kidneys harvested from mice about to succumb to infection ( Figure 8C ). In addition, histological analysis of the kidneys showed similar fungal loads ( Figure 8D ) and comparable immune cell invasion of the tissues in the different groups ( Figure 8E ). Elevated chemokine and cytokine levels in the kidney represent early responses to C. albicans infection and correlate with virulence [20]. We determined secretion of the proinflammatory cytokines TNFα and IL-6. Wild type, dectin1−/−, btk−/− and vav1−/− peritoneal macrophages secreted both TNFα and IL-6 ( Figure 8F,G ) in response to exposure to C. albicans. Btk−/− and vav1−/− macrophages generally secreted more TNFα and IL-6 than wild type and dectin1−/− macrophages (the difference between wild type and btk−/− TNFα secretion reached statistical significance). Next we investigated TNFα and IL-6 levels in mouse kidney during systemic C. albicans infection. Kidneys were harvested 11 days after tail vein injection with 5×10⁴ CFU. Although levels of TNFα and IL-6 were slightly higher in the dectin1−/−, btk−/− and vav1−/− mice than in wild type, these differences did not reach statistical significance ( Figure 8H,I ). We conclude that disease progression in response to C. albicans systemic infection is accelerated in dectin1−/−, btk−/− and vav1−/− animals compared to wild type animals.

Ethics statement

Animals used in this study were housed at the Whitehead Institute for Biomedical Research, which is certified by the United States Office of Laboratory Animal Welfare (OLAW) under the guidance of the Public Health Service (PHS) Policy on Humane Care and Use of Laboratory Animals. Whitehead Institute's Animal Welfare Assurance was approved 11/3/2009 (IACUC, A3125-01) All studies were carried out in accordance with procedures approved by the Massachusetts Institute of Technology Committee on Animal Care (Ploegh lab, CAC# 1011-123-14).

Cells and culture conditions

C. albicans strain SC5314 was cultured in YPD + Uri (2% bactopeptone, 1% yeast extract, 2% glucose and 80 µg/ml uridine) at 30°C. To generate a blue fluorescent protein (BFP)-expressing C. albicans strain, the GFP sequence of the pENO1-yEGFP3-NAT plasmid [28] was replaced with the TagBFP sequence (Evrogen) with codon usage adapted for C. albicans. C. albicans SC5314 was transformed with the pENO1-TagBFP-NAT plasmid and selected with 200 µg/ml nourseothricin (Werner Bioagents, Jena, Germany) resulting in the Candida-BFP strain. The RAW-Dectin1-LPETG-3×HA cell line (RAW-Dectin1) [7] was used for most phagocytosis experiments. Cells were grown in DMEM medium with 10% inactivated Fetal Bovine Serum (IFS) at 37°C and 5% CO₂. For the production of retrovirus, HEK293T cells (ATCC) were transfected using TransIT transfection reagent (Mirus) and virus-containing supernatant was harvested after 24 hours.

Plasmids, cloning, protein expression and construction of stable cell lines

The imaging constructs used in this study are listed in Table 1. A Dectin1-CRD-LPETG-His bacterial expression vector was cloned and expressed as described for Dectin1-CRD [13]. Mammalian expression vectors were constructed for N- or C-terminal mCherry fusions in vectors based on the retroviral plasmid pMSCVpuro (Clontech) (pMSCVpuro-mCherry-N and pMSCVpuro-mCherry-C). The BTK, Vav1 and Syk open reading frames were cloned from mouse spleen cDNA into the tagging vectors, resulting in N- or C-terminal fusion of all three genes (mCherry-BTK, BTK-mCherry, mCherry-Vav1, Vav1-mCherry, mCherry-Syk and Syk-mCherry). The vectors were used to create stable cell lines in the RAW-Dectin1 background by retroviral transduction and selection with puromycine. The resulting six cell lines were tested for expression of the mCherry-fused proteins by immunoblotting and microscopy and the BTK-mCherry, mCherry-Vav1 and mCherry-Syk cell lines were selected for further experiments. For expression of the biosensor constructs in the RAW macrophage cell lines, cells were transfected using Fugene (Roche).

Immunoprecipitation, mass spectrometry, blotting and antibodies

To identify proteins that interact with Dectin-1 during phagocytosis, RAW-Dectin1 macrophages were incubated with live C. albicans at MOI 5 for 1 hour or left unchallenged. Cells were harvested by scraping into ice-cold PBS and lysed in NP40 buffer (25 mM Tris pH 7.4, 150 mM NaCl, 5 mM MgCl₂ with 0.5% NP40 and protease inhibitors). Epitope-tagged Dectin-1 was immunoprecipitated from the total lysates with anti-HA beads (Roche). Eluates were run on a SDS-PAGE gradient gel and silver stained to visualize proteins. Each lane was cut into regions according to molecular weight, which were then reduced, alkylated and subjected to trypsin digestion. The resulting peptides were extracted, concentrated in vacuo, and analyzed by reverse-phase chromatography and tandem mass spectrometry. The resulting CID spectra were searched against a species-specific database generated from NCBI's non-redundant database using SEQUEST. For the generation of anti-BTK and anti-Vav1 antisera, BTK and Vav1 were cloned from mouse spleen cDNA into bacterial expression vector pET28a with an N-terminal His tag (pET28a-BTK and pET28a-Vav1). Vectors pET28a-BTK and pET28a-Vav1 were used to transform Rosetta cells and transformants were induced with IPTG for protein expression. His-Vav1 was isolated from the soluble fraction and His-BTK was isolated from the insoluble fraction in 8 M urea. Both proteins were purified using NiNTA beads and BTK was refolded by stepwise dialysis to eliminate urea, followed by FPLC purification of the peak containing the monovalent BTK. Purified Vav1 and BTK were injected in rabbits and serum was harvested and used for immunoblot and immunoprecipitation experiments (anti-BTK and anti-Vav1). Other antibodies used were: anti-Syk (Cell Signaling), anti-phospho-Syk (Tyr525/526; Cell Signaling), anti-PLCγ2 (Cell Signaling), anti-phospho-PLCγ2 (Tyr1217; Cell Signaling), anti-p97 [44], anti-HA-HRP (Roche). For immunoblotting, protein extracts were separated on 8% or 12% SDS-polyacrylamide gels and transferred to a nitrocellulose membrane using a semi-dry system. For BTK and Vav1 immunoprecipitation experiments, cells were lysed in NP40 buffer followed by immunoprecipitation with 2 µl of anti-BTK or anti-Vav1 antisera and 30 µl of Protein-A beads (Repligen). Beads were washed and eluates were analyzed using SDS-PAGE gels.

Spinning disk confocal and electron microscopy

Confocal images were collected in the W.M. Keck Facility for Biological Imaging using a PerkinElmer Live Cell Imaging spinning disk confocal system and Volocity software. The PerkinElmer Live Cell Imaging spinning disk confocal system was mounted on a Zeiss Axiovert 200M with a 100× 1.4NA Plan-Apochromat objective. Excitation light was generated by gas and solid state lasers (Argon laser for 488 nm, Krypton laser for 568 nm, solid state laser for 405 nm and 647 nm) and passed through an AOTF for wavelength selection and laser power control. A quadruple bandpass filter separated the excitation and emission light inside the CSU-22 confocal scanhead (Yokogawa) and a filter wheel (Prior Scientific) provided selection of emission filters (TagBFP & RFP: dual-band 445/60 and 615/70 nm; GFP: 527/55 nm). Volocity image acquisition software was used to capture images from a Hamamatsu Orca-ER cooled-CCD camera and to control all the equipment. For 3D reconstructions of phagocytic cells, Z planes were acquired at 0.15 µM distance and Volocity image acquisition software was used to create the XYZ views. Electron microscopy sections were examined using a FEI Tecnai Spirit at 80 KV. Routine morphology was performed by trimming and fixing the tissue in 2.5% gluteraldehyde, 3% paraformaldehyde with 5% sucrose in 0.1 M sodium cacodylate buffer (pH 7.4) and 0.2% tannic acid. Samples were post fixed in 1% osmium in veronal-acetate buffer. The tissue was stained in block overnight with 0.5% uranyl acetate in veronal-acetate buffer (pH 6.0), then dehydrated and embedded in em812 resin. Sections were cut on a Leica Ultracut UCT microtome with a Diatome diamond knife at a thickness setting of 50 nm, stained with uranyl acetate, and lead citrate. The sections were examined using a FEI Tecnai Spirit at 80 KV.

Phagocytosis assays

For general confocal microscopy, Candida-BFP was added to RAW-Dectin1 macrophages at an MOI of 10, fixed with 4% PFA in PBS and mounted on slides in 50% glycerol. β-1,3-glucan conjugated beads were a kind gift of Jatin Vyas and prepared as described [45]. Zymosan A (Sigma) was labeled with Alexa647 carboxylic acid, succinimidyl ester (Invitrogen) by incubation in 0.1 M Na₂CO₃ at room temperature. Candida-BFP was UV-killed by exposure to 100.000 µJ/cm² in a UV-crosslinker for four rounds. Recruitment of fluorescent proteins to the phagocytic cup or phagosome was quantified using ImageJ software according to the method of Flannagan and Grinstein [46]. Phagocytic indices of RAW-Dectin1 cells, BMDMs or peritoneal macrophages were determined by incubation with Candida-BFP or zymosan-Alexa647 at MOI 10 for 30 minutes or 1 hour. Cells were fixed in 4% PFA and stained with Concanavalin A-FITC (Sigma) to distinguish unengulfed yeasts. Inhibitors were added to the media at the indicated concentrations for 1 hour prior to incubation with Candida-BFP. Images were aquired by confocal microscopy and the number of intracellular Candida-BFP per macrophage was determined by counting 75-200 cells per experiment. Dectin1-CRD-LPETG was incubated with Staphylococcus aureus sortase A enzyme and GGG-Alexa647 probe resulting in Dectin1-CRD-LPETGGG-Alexa647 that was used for staining of Candida-BFP yeasts and hyphae. For immunofluorescence, cells were grown on coverslips and fixed in 4% PFA in PBS, washed with PBS and incubated in 50 mM NH₄CL in PBS for 10 min. Next, cells were incubated in Binding Buffer (0.1% Saponin, 0.2% BSA in PBS) for 30 min followed by incubation in Binding Buffer with anti-HA-Alexa488 (Invitrogen) antibody for 60 min, several washes with PBS and mounting for spinning disk confocal microscopy.

Synthesis of BTK inhibitor PCI-32765

4-Amino-3-(4-phenoxyphenyl)-1H-pyrazolo[3,4-d]pyrimidine (1) was prepared from 4-phenoxybenzoic acid and malonitrile as described (International Patent Publication No. WO 01/019829 and [17]. Alkylation of pyrazole (1) with 3-methanesulfonyl N-Boc hydroxypiperidine (2) followed by removal of the Boc-protecting group and acylation with acryloyl chloride gave the racemate of PCI-32756 in three steps (Figure S3). All chemicals were of commercial sources and were used as received. DriSolv anhydrous CH2Cl2, DriSolv anhydrous MeOH, DriSolv anhydrous DMF were purchased from EMD Chemicals. Redistilled, anhydrous N,N′- diisopropylethylamine (DiPEA), trifluoroacetic acid (TFA), triisopropylsilane (TIS) N-methylpyrrolidone (NMP) was obtained from Sigma-Aldrich. LC-ESI-MS analysis was performed using a Micromass LCT mass spectrometer (Micromass MS Technologies, USA) and a Paradigm MG4 HPLC system equipped with a HTC PAL autosampler (Michrom BioResources, USA) and a Waters Symmetry 5 µm C8 column (2.1×50 mm, MeCN∶H₂O (0.1% formic acid) gradient mobile phase, 150 µL/min). HPLC purifications were achieved using an Agilent 1100 Series HPLC system equipped with a Waters Delta Pak 15 µm, 100 Å C18 column (7.8×300 mm) using A: H₂O, B: MeCN and C: 1% aqueous trifluoroacetic acid as mobile phase (3 mL/min). (R/S)-1-Boc-3-Hydroxypiperidine (1.05 g, 5 mmol) was dissolved in CH₂Cl₂ (25 mL) and subsequently triethylamine (1.39 mL, 10 mmol) and methanesulfonyl chloride (0.394 mL, 5.1 mmol) were added. After stirring overnight, the reaction was concentrated under reduced pressure, redissolved in ethyl acetate, washed with water and brine, dried over MgSO₄ and concentrated in vacuo. The crude mesylate was dissolved in anhydrous DMF (20 mL). Pyrazole 1 (1.02 g, 3.33 mmol) and potassium carbonate (0.92 g, 6.66 mmol) were added and the reaction was stirred until TLC analysis showed complete conversion. The reaction was diluted with water and extracted with CH₂Cl₂. The organic layer was dried over MgSO₄, concentrated in vacuo. Purification over silica gel chromatography (CH₂Cl₂→MeOH/CH₂Cl₂) gave intermediate 3. Intermediate 3 was dissolved in dioxane (20 mL) and freshly prepared hydrochloric acid (35 mmol) was added. The solution was stirred for 1 h, concentrated in vauo. The crude amine (1 g, 2.58 mmol) was redissolved in CH₂Cl₂ (10 mL). To this was added Et₃N (1.8 mL, 12.9 mmol) and acryloyl chloride (0.22 mL, 2.71 mmol). After 5 h, the reaction was quenched by the addition of water. The solution was extracted and the organic layer was dried over MgSO₄ and concentrated in vacuo. The crude product was purified by reverse phase HPLC (28→43%B in 20 min, 3 mL/min) affording the title compound (46.3 mg, 0.105 mmol) as a white solid. LC/MS: R_t 8.52 min; linear gradient 5→80% B in 10 min; ESI/MS: m/z = 441.2 [M+H]⁺. ¹H NMR (400 MHz, CD₃OD) δ ppm 8.39 (s, 1H), 7.69-7.66 (m, 2H), 7.44-7.39 (m, 2H), 7.21-7.08 (m, 5H), 6.82 (dd, J = 16.8, 10.8 Hz, 0.6H), 6.66 (dd, J = 16.8, 10.8 Hz, 0.4H), 6.17 (d, J = 16.8 Hz, 0.6H), 6.15 (d, J = 16.8 Hz, 0.4H) 5.77 (d, J = 10.8 Hz, 0.6H), 5.65 (d, J = 10.8 Hz, 0.4H), 4.95-4.93 (m, 1H), 4.56 (d, J = 12.4 Hz, 0.6H), 4.24 (d, J = 11.2 Hz, 1H), 4.06 (d, J = 14.0 Hz, 0.6H), 3.89 (dd, J = 12.8, 8.8 Hz, 0.4H), 3.58 (dd, J = 12.4, 9.6 Hz, 0.6H), 3.42-3.36 (m, 1H), 2.45-2.34 (m, 1H), 2.28-2.23 (m, 1H), 2.15-2.05 (m, 1H), 1.80-1.68 (m, 1H). 2.72 (t, J = 7.6 Hz, 2H), 2.15 (dt, J = 7.2, 2.8 Hz, 2H), 1.96 (d, J = 2.8 Hz, 1H), 1.82 (quin., J = 7.6 Hz, 2H), 1.52 (quin., J = 7.2 Hz, 2H).

Quantitative analysis of cellular diacylglycerol content

RAW-Dectin1 macrophages were coincubated with Candida albicans in 6-well dishes for 1 hour, washed with phosphate-buffered saline and subjected to lipid extraction [47]. The chloroform/methanol phase was dried under N₂ and DAG kinase assays were performed as described [48]. See supplementary materials for further details. The dried lipids were dissolved in 40 µl of solubilizing buffer (7.5% octyl-ß-D-glucoside and 5 mM cardiolipin in 1 mM diethylenetriaminepenta acetic acid (DETAPAC, pH 7.0) by vigorously vortexing for 20 sec and incubating at RT for 10 min. Then, 100 µl of 2× reaction buffer (100 mM imidazole HC1, pH 6.6, 100 mM NaCl, 25 mM MgCl₂, and 2 mM EGTA), 4 µl of 100 mM freshly prepared DTT and 20 µl of E. coli DAG kinase (Sigma-Aldrich)) were added while keeping the samples on ice. The reaction was initiated by addition of 3 µCi [γ³³P]-ATP prepared by dilution in 20 µl of 1 mM DETAPAC, pH 6.6. After vortexing briefly, the reaction was incubated at 25°C for 30 min. Lipids were extracted as described above and the reaction products were analyzed by TLC, which was developed in acetone followed by CHCl₃/MeOH/acetic acid (65/15/5 [vol/vol/vol]) solution. Radiolabelled lipids were detected by exposure to imaging screens (BAS-MS; FujiFilm), scanned on a BAS-2500 (FujiFilm)), and quantified with Quantity One software.

Animal care, primary cells and mouse model for systemic candidiasis

Animals were housed at the Whitehead Institute for Biomedical Research and maintained according to protocols approved by the Massachusetts Institute of Technology Committee on Animal Care. C57BL/6 mice were purchased from Jackson Labs, dectin1−/− [49], btk−/− [36] and vav1−/− mice [39] were kind gifts from Stu Levitz, Whasif Khan and Victor Tybulewicz, respectively. Bone marrow-derived macrophages (BMDMs) were differentiated from mouse bone marrow by growth in DMEM (high glucose; Gibco) with 10% HI-FBS (Hyclone) and 5% M-CSF-containing culture supernatant from L929 cells. Experiments were performed after 7 days of differentiation. Peritoneal macrophages were harvested by peritoneal lavage with upto 10 ml PBS. Cells were seeded for experiments in DMEM (high glucose; Gibco) with 10% HI-FBS (Hyclone) and used for experiments the next day. For cytokine analysis, macrophages were incubated with C. albicans and supernatants were harvested after 16 hours. Cytokine concentrations were determined by Discovery assay cytokine array by Eve Technologies. For the mouse model of systemic candidiasis, 5×10⁴ or 1×10⁵ CFU of C. albicans SC5314-derived strain Candida-BFP was administered intraveneously to age-matched C57BL/6 wild type, dectin-1, btk and vav1 −/− mice in a final volume of 200 µl in PBS. Mice were weighed and monitored daily and euthanized when >20% of initial body weight was lost. Kidneys were harvested for enumeration of fungal burden, histology or cytokine analysis. To determine fungal burden, kidneys were homogenized and lysates were plated on YPD plates with antibiotics. For histology, tissues were fixed in 4% formalin in PBS, embedded and sectioned in paraffin and slides were stained with Hematoxylin and Eosin (H&E) or Gomori Methenamine Silver (GMS). For cytokine analysis, kidneys were homogenized in 1 ml PBS followed by cytokine analysis using a Th1/Th2/Th17 mouse cytometric bead array (BD Biosciences) and LSR Flow Cytometer according to the manufacturer's directions.

Statistical analysis

For statistical analysis unpaired t tests were used with 95% confidence interval. All graphs show means and standard deviations of three independent experiments.

Supplemental information

Supplemental information includes three figures.