The β-Glucan Receptor, Dectin-1, Is Predominantly Expressed on the Surface of Cells of the Monocyte/Macrophage and Neutrophil Lineages

RNA analysis

For RT-PCR analysis, total RNA from various cell lines (RAW264.7, J774, P388D1, and NIH-3T3) and primary cell types (BMDMφ and BMDDC) was prepared using the guanidine isothiocyanate-based RNA isolation kit (Stratagene, La Jolla, CA). First-strand cDNA synthesis was performed using the Advantage RT-for-PCR kit with an oligo(dT) primer (Clontech, Palo Alto, CA), as described by the manufacturer. The βGR transcript was subsequently amplified using primers corresponding to the region of the cDNA encoding residues 66–244 of the primary protein sequence. Dihydrofolate reductase-specific primers (Stratagene) were used as a positive control. Commercially available membranes containing poly(A) mRNA isolated from various mouse tissues were purchased from Origene Technologies (Rockville, MD) and were probed as described by the manufacturer using a full-length βGR cDNA probe.

Tissue and cell preparation

All mice used in this study were C57BL/6J, unless otherwise stated, and were between 8 and 12 wk of age. Animals were kept and handled in accordance with institutional guidelines. Splenocytes were harvested by standard methods using a combination of digestion with Liberase Blendzyme II in RPMI (Roche, Indianapolis, IN) and mechanical dissociation. Femurs were collected, and fresh bone marrow was flushed from within using Liberase Blendzyme II and incubated for 10 min at 37°C to disaggregate cells. Enzymatic activity was quenched with RPMI/20% FCS, erythrocytes were lysed with Gey’s solution, and cell debris was removed by centrifugation through 100% FCS at 300 × g.

Isolation of peripheral blood leukocytes

Mice were sacrificed, and peripheral blood was collected by cardiac puncture into 0.1 vol 100 mM EDTA. Cells were harvested by centrifugation and resuspended in 50 vol Gey’s solution to lyse erythrocytes. Peripheral blood leukocytes were then recovered by centrifugation through FCS as described above.

Isolation of alveolar Mφ

Bronchoalveolar lavage was performed by repeated washes with 1 ml PBS/5 mM EDTA. Resident alveolar Mφ, the major leukocyte population in the lungs, were identified by size and autofluorescence using flow cytometry as previously described (22).

Induction of sterile peritonitis and recovery of peritoneal cells

To induce sterile peritonitis, mice were injected i.p. with 4% thioglycolate (BD Biosciences, Franklin Lakes, NJ) up to 4 days before peritoneal lavage. After humane killing of the animals, inflammatory cells were collected by peritoneal lavage with ice-cold 5 mM EDTA in PBS. Resident peritoneal cells were collected in the same way from untreated animals. Peritoneal Mφ were identified by their expression of F4/80 and CR3 and were distinguished from eosinophils by forward scatter (FSC)/side scatter (SSC) profiles. To confirm the cellular composition of peritoneal exudates, differential counts were performed on cytospin preparations stained with Hema Gurr (VWR International, Poole, U.K.).

FACS analysis

FACS was performed according to conventional protocols at 4°C in the presence of 2 mM NaN₃. Cells were blocked with 5% heat-inactivated rabbit serum, 0.5% BSA, 5 mM EDTA, and 4 μg/ml 2.4G2 (anti-FcγRII and -III) before the addition of primary Abs. Biotinylated Abs were detected using streptavidin-allophycocyanin (BD PharMingen). Cells were fixed with 1% formaldehyde in PBS before analysis.

The following Abs were used in this study: B220-CyChrome (RA3-6B2; BD PharMingen), CD3-CyChrome (17A2; BD PharMingen), F4/80-PE (Serotec), CD11c-PE (HL3; BD PharMingen), Gr-1-PE (anti-Ly6C/G; BD PharMingen), CD49b-PE (DX5-Pan NK-cell; BD PharMingen), 5C6-FITC (anti-CR3/CD11b) (23), 2A11-biotin (rat IgG2b anti-βGR) (20), 5D3-biotin (rat IgG2a anti-Mφ mannose receptor (MR)) (L. Martinez-Pomares, D. M. Reid, G. D. Brown, P. R. Taylor, R. Stillion, S. A. Linehan, S. Gordon, and S. Y. C. Wong, unpublished observations), and irrelevant rat IgG2b-biotin, IgG2a-biotin, and IgG2b-FITC control Abs.

In vitro non-opsonic zymosan binding assay

In vitro zymosan binding assays were performed as previously described (14, 18, 20). In brief, resident or 4-day thioglycolate-elicited peritoneal Mφ were recovered, as described above, and plated at 5 × 10⁵ and 2.5 × 10⁵ cells/well, respectively, in 24-well plates in RPMI/10% FCS overnight. The following day the cells were cooled to 4°C and washed three times with prechilled medium. All experiments were performed at 4°C to prevent receptor internalization, to provide a direct measure of surface receptor involvement, and to prevent local release of opsonins, including complement (24, 25). Zymosan-FITC (Molecular Probes) was added to the Mφ at a ratio of 25 particles/cell for 1 h on ice. For in vitro blocking assays, carbohydrates (laminarin, β-methylglucoside, and mannan; all from Sigma (St. Louis, MO) and used at 100 μg/ml) or Abs (2A11 (20); 5C6 (23), which has been shown to block the CR3-mediated lectin activity (26); or an irrelevant rat IgG2b control; all used at 100 μg/ml) were added to the chilled cells 20 and 60 min, respectively, before the addition of zymosan. After incubation, unbound zymosan was removed by extensive washing with medium, and cells were lysed with 3% Triton X-100. FITC in lysates was quantified using a Titer-Tek Fluoroskan II (Labsystems Group, Basingstoke, U.K.) as previously described (14, 18). For Ab modulation experiments poly-d-lysine-conditioned tissue culture plates were coated with Ab at 100 μg/ml as previously described (27). All experiments were repeated at least three times.

Statistical analysis

Statistics were calculated using GraphPad PRISM (version 2.0; GraphPad Software, Berkeley, CA). One-way ANOVA with Bonferroni multiple comparison test was applied throughout.

Expression of βGR mRNA in macrophages and multiple mouse tissues

We studied the expression of βGR by RT-PCR in several Mφ cell lines and in primary Mφ and DCs. All Mφ cell lines as well as bone marrow-derived Mφ and DC showed evidence of βGR expression, whereas βGR transcript was not detectable in the mouse fibroblast cell line NIH-3T3 (Fig. 1⇓A). Using the full-length coding sequence to screen a multiple tissue Northern blot, we found βGR expression in most murine tissues with the exception of brain, muscle, and skin (Fig. 1⇓B). Notably there was only one discernible transcript detectable in these tissues.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Expression of βGR by primary macrophages and in selected tissues. A, RT-PCR analysis showed the presence of βGR transcript in Mφ cell lines (RAW264.7, J774, and P388D1) and primary Mφ (BMDMφ) and DC (BMDDC), but not in a mouse fibroblast cell line (NIH-3T3). The housekeeping gene dihydrofolate reductase (DHFR) was used as a PCR control. B, A multiple mouse tissue Northern blot probed with βGR cDNA showed widespread expression of βGR in the mouse. Control probing with β-actin confirmed equivalent loading between lanes (data not shown).

Distribution of βGR surface expression in the spleen

We examined the surface expression of βGR using 2A11 on freshly isolated splenocytes (Fig. 2⇓). CD11c^high DC were found to express βGR (population 1) in a similar pattern to that reported previously (15). Notably, however, other CD11c^low/− cells in the spleen, particularly those expressing CR3, exhibited high surface expression of βGR (population 2). To further delineate which cell types were expressing βGR, the cells were subdivided into six populations based on their expression of CR3 and Gr-1 (an mAb recognizing Ly-6G and Ly-6C) and their FSC/SSC profiles (Fig. 2⇓B). Gr-1^highCR3^highSSC^high neutrophils (population 3) exhibited high surface expression of βGR, as did Gr-1^lowCR3⁺SSC^lowMφ (population 4B), which also expressed F4/80 (data not shown). A second, unidentified, population of Gr-1^lowCR3⁺ splenocytes with very high SSC (population 4A) did not show evidence of βGR surface expression. CR3⁺Gr-1⁻ splenocytes, a mixed population containing DC (CD11c^high), NK cells, and other Mφ (both CD11c^int), showed heterogeneity in expression of βGR (population 5). NK cells, which have been shown to recognize β-glucans (8) and are identified by high expression of the DX5 Ag (CD49b) (28), did not show significant labeling with the 2A11 Ab (data not shown). Gr-1^lowCR3⁻ splenocytes (population 6), previously reported to be a T cell subset (29), expressed CD3 and low levels of surface βGR (Fig. 2⇓C). Analysis of all splenic T cells (CD3⁺) and B cells (B220⁺) for βGR surface expression, however, indicated that only a distinct subset of T cells exhibited significant surface expression of βGR (Fig. 2⇓C). βGR⁺CD3⁺T cells were predominantly Gr-1⁺ and CD8⁺, but CD4⁺ cells were also observed (data not shown). Splenic autofluorescent F4/80⁺ Mφ also expressed βGR, but at very low levels (data not shown). Plasmacytoid DC in the spleen, which were identified by their Gr-1⁺B220⁺CD11c^intCR3⁻ phenotype and analyzed in 129/SvEv and BALB/c mice because of the relative scarcity of these cells in C57BL/6 (30), also exhibited low, but detectable, levels of βGR expression (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Distribution of βGR surface expression on freshly isolated splenocytes. A, CD11c^high DC (population 1; purple) were confirmed to express βGR, but highest surface expression was observed on most CR3⁺CD11c^low/neg splenocytes (population 2; brown). B, Gr-1^highCR3^highSSC^high neutrophils (population 3; red) exhibited high surface expression, as did the Gr-1^lowCR3^highSSC^low Mφ (population 4B; light green), which were also F4/80⁺ (data not shown). CR3⁺Gr-1⁻ splenocytes (population 5; yellow), a mixed population that contains DC, Mφ, and NK cells, showed heterogeneous βGR expression. Gr-1^lowCR3⁻ splenocytes (population 6; pink) also exhibited βGR surface expression, albeit at a relatively low level, and these cells were confirmed to express CD3. C, Analysis of all T cells (gated on CD3⁺; orange) showed that only a subset of T cells expressed βGR. B220⁺ B cells (gray) had no obvious βGR surface expression. Unshaded histograms, Control Ab staining; shaded histograms, correspond to the marker indicated.

Expression of βGR on peripheral blood leukocytes

We and others have observed expression of βGR/dectin-1 on both human and mouse PBL by Northern blot (17, 18). We confirmed these observations by FACS by identifying a significant population of PBL that expressed βGR (data not shown). Consistent with the data obtained from the spleen, peripheral blood neutrophils (identified as Gr-1^highSSC^high) and peripheral blood M_o (identified as CR3⁺F4/80⁺SSC^low) exhibited high surface expression of βGR (Fig. 3⇓A).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Expression of βGR by peripheral blood and bone marrow cells. A, Peripheral blood neutrophils (Gr-1^high) and M_o (F4/80⁺CR3⁺SSC^low), gated from total peripheral blood leukocytes as shown in the insets, both showed significant levels of surface staining with 2A11 (shaded histograms) compared with a rat IgG2b control (unshaded histograms). B, Bone marrow was analyzed for the expression of βGR, Gr-1, CR3, and F4/80. Dot plots show βGR expression in relation to Gr-1 staining. Gr-1^high neutrophils are subdivided into two populations by high and low βGR surface expression. βGR^high neutrophils (population 1) had higher SSC and CR3 surface expression than βGR^low neutrophils (population 2), indicating that βGR^high neutrophils may be more mature (data not shown). Gr-1^lowβGR^high cells (population 3) in the bone marrow, most likely of the M_o/Mφ lineage, exhibited the highest βGR surface expression.

Surface expression of the βGR on myeloid cells in the bone marrow

Since we have previously observed the βGR transcript in human bone marrow (18), we examined murine bone marrow for the expression of βGR (Fig. 3⇑B). Gr-1^highCR3⁺ neutrophils were subdivided into two populations. Approximately one-third of the bone marrow Gr-1^high neutrophils had high βGR surface expression; the remaining two-thirds showed intermediate or marginal expression (populations 1 and 2, respectively; Fig. 3⇑B). The βGR^high neutrophils had higher SSC and higher CR3 surface expression than the βGR^low neutrophils (data not shown), suggesting that the βGR^high neutrophils may be in a more advanced state of maturation. This is consistent with the high βGR surface expression detected on circulating peripheral blood neutrophils (Fig. 3⇑A). The Gr-1^low subgroup of bone marrow cells that has been reported to include cells of the M_o/Mφ lineage, myeloid precursors, and hemopoietic stem cells (31) contained cells with the highest βGR surface expression (population 3). The expression of CR3 and F4/80 indicated that these high βGR-expressing cells most likely belonged to the M_o/Mφ lineage (data not shown). Additional unidentified Gr-1^low and Gr-1⁻ bone marrow cells also showed evidence of βGR expression, but these were not characterized further (Fig. 3⇑B).

Expression of βGR by isolated primary Mφ

Freshly isolated resident and thioglycolate-elicited peritoneal Mφ were assayed for surface expression of βGR. We also looked for βGR expression on alveolar Mφ, as we had found a high level of transcript in the lung (Fig. 1⇑). Both freshly isolated alveolar Mφ and thioglycolate-elicited Mφ expressed high surface levels of βGR, whereas resident peritoneal Mφ exhibited lower expression (Fig. 4⇓). Interestingly, we observed an up-regulation of βGR on the surface of resident peritoneal Mφ after 1 day of culture (Fig. 4⇓). Surface expression of βGR on thioglycolate-elicited Mφ was relatively unaffected by 1 day of culture (data not shown). Since the Mφ MR is also a candidate receptor for the nonopsonic recognition of zymosan by resident peritoneal Mφ (see below), we analyzed the surface expression of this receptor on the same cells. Similar to βGR, we found the highest surface expression of the MR on thioglycolate-elicited Mφ, moderate expression on resident alveolar Mφ, but only very limited expression on the surface of resident peritoneal cells (Fig. 4⇓). Unlike βGR, however, the expression of MR on resident peritoneal cells was relatively unaffected by 1 day of culture (Fig. 4⇓). As reported previously, we found that alveolar Mφ expressed negligible CR3 (22) (Fig. 4⇓) and low levels of F4/80 (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Primary Mφ express βGR. Freshly isolated 4-day thioglycolate-elicited peritoneal Mφ (Mφ_thio), resident peritoneal Mφ (Mφ_res) and alveolar Mφ (Mφ_alv), and resident peritoneal Mφ that had been cultured for 1 day (d1 Mφ_res) were assessed for surface expression of βGR, MR, CR3, and F4/80 (not shown). Thioglycolate-elicited and resident alveolar Mφ expressed very high levels of βGR. Resident peritoneal Mφ expressed significantly less surface βGR than the other Mφ studied, but expression was markedly increased after 1 day in culture. Unshaded histograms, Control Ab staining; shaded histograms, receptor-specific staining. Peritoneal Mφ were distinguished from other cells by high expression of F4/80 and CR3, and alveolar Mφ were identified as previously described (22 ).

Surface expression of βGR during peritoneal inflammation

To study the expression of βGR in an inflammatory context, we examined peritoneal exudate cells 18 h after the i.p. administration of thioglycolate, a model of sterile peritonitis. F4/80⁺CR3^highGr-1⁻ Mφ, F4/80⁻CR3⁺Gr-1⁺ neutrophils, and F4/80⁺CR3⁺Gr-1⁻SSC^high eosinophils (32) were then tested for βGR surface expression (Fig. 5⇓). Elicited peritoneal Mφ exhibited the highest βGR expression, and significant amounts were also present on the inflammatory neutrophils. In contrast, recruited eosinophils showed no obvious surface expression of βGR (Fig. 5⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Expression of βGR during inflammation. Sterile peritonitis was induced by i.p. administration of thioglycolate broth 18 h before analysis. Peritoneal exudate cells were examined by FACS with the Mφ/eosinophil marker F4/80 and the neutrophil marker Gr-1 for the expression of βGR. F4/80⁺Gr-1⁻Mφ (Mφ) exhibited very high βGR surface expression. Gr-1⁺F4/80⁻ neutrophils (Neut) also expressed βGR, but F4/80⁺Gr-1⁻SSC^high eosinophils (Eo) did not show evidence of surface expression of this receptor.

Non-opsonic binding of zymosan to resident peritoneal Mφ

We observed that freshly isolated resident peritoneal Mφ had a lower level of surface βGR expression than that on other Mφ studied (Fig. 4⇑). As we had previously shown that βGR was a major receptor for zymosan on thioglycolate-elicited and BMDMφ (20), we wanted to determine whether this was also true for resident peritoneal Mφ. We compared the contribution of βGR on both resident and elicited Mφ and found that the binding of unopsonized zymosan to elicited Mφ was significantly inhibited by β-glucans, as previously reported (Fig. 6⇓) (20). βGR was still a major receptor for zymosan on resident Mφ, but it contributed less to this process than in the thioglycolate-elicited cells. Furthermore, we found that mannan had an inhibitory effect on the binding of zymosan to the resident Mφ, but not to the thioglycolate-elicited cells (Fig. 6⇓A). The combination of β-glucans and mannan did not have an additive effect (data not shown). As with the elicited Mφ, methylglucoside failed to inhibit the initial binding of zymosan to resident peritoneal Mφ, suggesting no involvement of CR3 in this process (data not shown). These results implied that a secondary β-glucan-independent, mannan-inhibitable, nonopsonic binding mechanism was operational on resident Mφ, but was not present on other Mφ examined.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

β-Glucan binding by resident and elicited Mφ. A, Non-opsonic binding of zymosan particles to resident and thioglycolate-elicited Mφ (□ and ▪, respectively). Data are expressed as the mean ± SD as a percentage of the binding to control untreated Mφ measured as relative fluorescence units (RFU). Laminarin significantly inhibited the nonopsonic recognition of zymosan by both cell types. Mannan, however, only inhibited recognition by resident peritoneal Mφ. B, Ab-mediated blocking of nonopsnic zymosan binding by resident peritoneal Mφ. Nonopsonic binding assays were performed, as described in Materials and Methods, after preincubating the Mφ with the mAbs, 2A11 (anti-βGR) and 5C6 (a mAb against CR3 that has been shown to block the lectin activity of CR3 (26 )), or the soluble β-glucans, laminarin and glucan phosphate (not shown). Only 2A11 inhibited nonopsonic zymosan binding to resident peritoneal Mφ. Data are expressed as the mean ± SD percentage of binding to control untreated Mφ, measured as relative fluorescence units (RFU).

To find out which specific receptors were involved, we performed Ab blocking experiments on the resident peritoneal Mφ (Fig. 6⇑B). The anti-βGR mAb, 2A11, blocked the nonopsonic binding of zymosan to resident peritoneal Mφ to the same degree as the soluble β-glucans laminarin (Fig. 6⇑B) and glucan phosphate (data not shown), consistent with βGR being a major β-glucan receptor on Mφ. Anti-CR3 (5C6, which blocks the lectin activity of CR3) had no inhibitory effect (Fig. 6⇑B) consistent with our previous results (20). As surface expression of Mφ MR was low on resident peritoneal Mφ and higher on thioglycolate-elicited Mφ, which do not have a mannan-inhibitable component of zymosan binding, the MR appears not to be involved in the nonopsonic recognition of zymosan by the Mφ (Fig. 4⇑) (20). We also performed Ab modulation experiments using specific mAb to deplete surface receptors from the upper ligand binding surface. Only 2A11, anti-βGR-coated tissue culture wells inhibited the nonopsonic binding of zymosan; anti-Mφ MR and anti-CR3, did not (data not shown).