Identification of serum glycoprotein ligands for the immunomodulatory receptor blood dendritic cell antigen 2

Abstract

Blood dendritic cell antigen 2 (BDCA-2) is a C-type lectin found on the surface of plasmacytoid dendritic cells. It functions as a glycan-binding receptor that downregulates the production of type I interferons and thus plays a role in oligosaccharide-mediated immunomodulation. The carbohydrate recognition domain in BDCA-2 binds selectively to galactose-terminated bi-antennary glycans. Because the plasmacytoid dendritic cells function in a plasma environment rich in glycoproteins, experiments have been undertaken to identify endogenous ligands for blood dendritic cell antigen 2. A combination of blotting, affinity chromatography and proteomic analysis reveals that serum glycoprotein ligands for BDCA-2 include IgG, IgA and IgM. Compared to binding of IgG, which was previously described, IgA and IgM bind with higher affinity. The association constants for the different subclasses of immunoglobulins are below and roughly proportional to the serum concentrations of these glycoprotein ligands. Binding to the other main serum glycoprotein ligand, α2-macroglobulin, is independent of whether this protease inhibitor is activated. Binding to all of these glycoprotein ligands is mediated predominantly by bi-antennary glycans in which each branch bears a terminal galactose residue. The different affinities of the glycoprotein ligands reflect the different numbers of these galactose-terminated glycans and their degree of exposure on the native glycoproteins. The results suggest that normal serum levels of immunoglobulins could downmodulate interferon stimulation of further antibody production.

Fig. 1.

Identification of serum glycoprotein ligands for BDCA-2. (A) Demonstration of blotting specificity with samples of fetuin and asialofetuin. Aliquots (2 μg) were resolved on a 17.5% SDS polyacrylamide gel and either stained with Coomassie blue or blotted onto nitrocellulose and probed with CRD from BDCA-2 complexed with streptavidin–alkaline phosphatase. (B) Direct probing for glycoprotein ligands in serum. Aliquots corresponding to 0.05 μL of serum and 0.5 μg of human IgG were resolved on a 12.5% SDS polyacrylamide gel. Separate lanes were stained with Coomassie blue or blotted onto nitrocellulose and probed with CRD from BDCA-2 complexed with streptavidin–alkaline phosphatase. (C) Analysis of glycoprotein ligands following affinity chromatography on immobilized CRD. Aliquots of proteins eluted from the affinity column with EDTA were run in parallel and stained and blotted as in (A). (D) Separation of affinity-purified glycoprotein ligands following precipitation with trichloroacetic acid. Gel was stained with Coomassie blue and bands 1–6 were excised for proteomic analysis. Band 1 corresponds to material that did not enter the separating gel following precipitation. Proteins identified in each band are indicated along with their molecular weights (Table I). EDTA, ethylenediaminetetraacetic acid.

Fig. 2.

Binding competition assays for different classes of immunoglobulin binding to BDCA-2. Biotin-tagged CRD from BDCA-2 was immobilized on streptavidin-coated plates that were probed with ¹²⁵I-Man-BSA.in the presence of different concentrations of competing glycoprotein ligands. Experimental values, which are the average of duplicates, are shown as circles. K_I values were determined by nonlinear least squares fitting to a simple binding competition equation, with the fitted curve shown as a continuous line in each graph. Example assays are shown. IgA used in these assays was purified from human serum by affinity chromatography on anti-IgA antibody. The locations of glycosylation sites in each of the three classes of immunoglobulins are shown at the right.

Fig. 3.

Summary of BDCA-2 binding to IgA1 and IgA2 prepared by different methods. (A) Direct purification of IgA1 from serum using the plant lectin jacalin, which binds to the O-linked glycans present in the hinge region of IgA1 but not IgA2 (Figure 2). The IgA1 pool was eluted at low pH and further fractionated by chromatography on the CRD from BDCA-2 immobilized on agarose resin. IgA1 that bound to BDCA-2 was eluted with EDTA. (B) Purification of total IgA from serum by chromatography on immobilized anti-IgA antibodies, followed by fractionation of IgA1 and IgA2 by chromatography on jacalin. The purification schemes are shown schematically, along with the K_I values, determined using the binding competition assay described in Figure 2, for the different preparations. K_I values represent the mean ± standard deviation for three experiments, each performed in duplicate.

Fig. 4.

Binding of BDCA-2 to glycans on heavy chains of immunoglobulins. (A) Approximately 2 μg of each immunoglobulin was separated on 17.5% SDS polyacrylamide gels, which were either stained with Coomassie blue or blotted onto nitrocellulose and probed with BDCA-2–streptavidin–alkaline phosphatase. IgA1 and IgA2 were resolved as shown in Figure 3B. (B) Binding of BDCA-2–strepatividin–alkaline phosphatase complex to fragments of IgG following papain digestion. Following partial digestion of IgG with papain, the remaining fragments were purified on immobilized protein A and separated on 17.5% SDS polyacrylamide gels. Parallel lanes were stained with Coomassie blue and blotted onto nitrocellulose for probing with BDCA-2–strepatividin–alkaline phosphatase complex.

Fig. 5.

Analysis of neoglycolipids derived from IgA glycans. Glycans released from IgA by digestion with protein N-glycosidase were attached to a synthetic analog of phosphatidyl ethanolamine by reductive amination The resulting neoglycolipids were fractionated by reverse phase chromatography in CHCl₃/methanol/water 60:35:8. Aliquots of the resulting fractions were separated on high-performance silica thin-layer chromatography plates in CHCl₃/methanol/water 105:100:28. Left, thin-layer chromatogram stained with orcinol showing the distribution of total neoglycolipids for the full set of elution fractions. Right, chromatograms were soaked with polyisobutylmethacrylate and probed with BDCA-2–strepatividin–alkaline phosphatase complex. Galactose-terminated bi-antennary glycans were processed in parallel for comparison. These standards elute in fraction 3–5.

Fig. 6.

α₂-Macroglobulin native and activated gel blot. Purified α₂-macroglobulin was separated by native polyacrylamide gel electrophoresis before and after treatment with hydroxylamine. The hydroxylamine treatment results in conversion to the electrophoretically fast form. (A) Coomassie blue stained gel. (B) Gel blotted onto nitrocellulose and probed with BDCA-2–streptavidin–alkaline phosphatase complex.