
==== Front
Proc Natl Acad Sci U S A
Proc Natl Acad Sci U S A
PNAS
Proceedings of the National Academy of Sciences of the United States of America
0027-8424
1091-6490
National Academy of Sciences

38478694
202312404
10.1073/pnas.2312404121
research-articleResearch ArticleimmunImmunology and Inflammation420
Biological Sciences
Immunology and Inflammation
A type I interferon regulatory network for human plasmacytoid dendritic cells based on heparin, membrane-bound and soluble BDCA-2
Venegas-Solis Francisco a https://orcid.org/0000-0001-5799-8711

Staliunaite Laura a https://orcid.org/0009-0005-5609-7643

Rudolph Elisa a
Münch Carina Chan-Song b
Yu Philipp a
Freibert Sven-A. c d https://orcid.org/0000-0002-8521-2963

Maeda Takahiro e https://orcid.org/0009-0003-0217-0775

Zimmer Christine L. f
Möbs Christian f https://orcid.org/0000-0002-5197-7669

Keller Christian b https://orcid.org/0000-0002-0075-7199

Kaufmann Andreas a https://orcid.org/0000-0002-3936-5379

Bauer Stefan stefan.bauer@staff.uni-marburg.de
a 1
aInstitute for Immunology, Philipps-Universität Marburg, Biomedizinisches Forschungszentrum Marburg, Marburg 35043, Germany
bInstitute of Virology, Philipps-Universität Marburg, Biomedizinisches Forschungszemtrum Marburg, Marburg 35043, Germany
cInstitute for Cytobiology, Center for Synthetic Microbiology, Philipps-Universität Marburg, Marburg 35032, Germany
dCore Facility “Protein Biochemistry and Spectroscopy”, Philipps-Universität Marburg, Marburg 35032, Germany
eDepartment of Island and Community Medicine, Island Medical Research Institute, Nagasaki University Graduate School of Biomedical Science, Nagasaki 852-8523, Japan
fDepartment of Dermatology and Allergology, Philipps-Universität Marburg, Marburg 35043, Germany
1To whom correspondence may be addressed. Email: stefan.bauer@staff.uni-marburg.de.
Edited by Marco Colonna, Washington University in St. Louis School of Medicine, St. Louis, MO; received July 20, 2023; accepted January 10, 2024

13 3 2024
19 3 2024
13 9 2024
121 12 e231240412120 7 2023
10 1 2024
Copyright © 2024 the Author(s). Published by PNAS.
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND).

Significance

Type I IFN is a two-edged sword; it plays an essential role in immune defense against viruses, but on the other hand, inappropriate production of type I IFNs may be harmful to the host by promoting an inflammatory response that finally causes the development of autoimmune or chronic inflammatory diseases. Here, we identified commercial and human plasma–derived heparin as ligand for BDCA-2 resulting in the inhibition of IFN-α production in human plasmacytoid dendritic cells. Upon cellular activation, BDCA-2 is released from the cell surface via ADAM10/17 and functions as heparin antagonist leading to enhanced IFN-α production. This network consisting of heparin, membrane-bound and soluble BDCA-2 regulates type I interferon production, and its dysregulation may promote autoimmune diseases.

Plasmacytoid dendritic cells (pDCs) produce type I interferons (IFNs) after sensing viral/bacterial RNA or DNA by toll-like receptor (TLR) 7 or TLR9, respectively. However, aberrant pDCs activation can cause adverse effects on the host and contributes to the pathogenesis of type I IFN-related autoimmune diseases. Here, we show that heparin interacts with the human pDCs-specific blood dendritic cell antigen 2 (BDCA-2) but not with related lectins such as DCIR or dectin-2. Importantly, BDCA-2–heparin interaction depends on heparin sulfation and receptor glycosylation and results in inhibition of TLR9-driven type I IFN production in primary human pDCs and the pDC-like cell line CAL-1. This inhibition is mediated by unfractionated and low-molecular-weight heparin, as well as endogenous heparin from plasma, suggesting that the local blood environment controls the production of IFN-α in pDCs. Additionally, we identified an activation-dependent soluble form of BDCA-2 (solBDCA-2) in human plasma that functions as heparin antagonist and thereby increases TLR9-driven IFN-α production in pDCs. Of importance, solBDCA-2 levels in the serum were increased in patients with scrub typhus (an acute infectious disease caused by Orientia tsutsugamushi) compared to healthy control subjects and correlated with anti-dsDNA antibodies titers. In contrast, solBDCA-2 levels in plasma from patients with bullous pemphigoid or psoriasis were reduced. In summary, this work identifies a regulatory network consisting of heparin, membrane-bound and solBDCA-2 modulating TLR9-driven IFN-α production in pDCs. This insight into pDCs function and regulation may have implications for the treatment of pDCs-related autoimmune diseases.

plasmacytoid dendritic cells
type I interferons
BDCA-2
glycosaminoglycans
heparin
Deutsche Forschungsgemeinschaft (DFG) 501100001659 369799452 - TRR237 Francisco Venegas-SolisStefan Bauer Deutsche Forschungsgemeinschaft (DFG) 501100001659 114933180-TR84 Francisco Venegas-SolisStefan Bauer
==== Body
pmcPlasmacytoid dendritic cells (pDCs) are a subset of dendritic cells specialized to sense viral/bacterial nucleic acids through endosomal toll-like receptor (TLR) 7 or TLR9. After TLR activation, pDCs produce high amounts of type I interferons (IFNs) that inhibit viral infections and activate the antiviral function of immune cells (1, 2). In addition, type I IFN production by pDCs needs to be under strict control to avoid any adverse effects to the host; this is controlled by IFN regulatory receptors such as blood dendritic cell antigen 2 (BDCA-2) (3).

BDCA-2 is a type II C-type lectin and belongs to the dectin-2 receptor family with function in immunity and homeostasis (4). This receptor is expressed exclusively on the surface of human pDCs (hpDCs) (5), associates with the transmembrane adaptor FcεRIγ (6), and consists of a calcium-dependent C-type carbohydrate recognition domain (CRD), a transmembrane region, and a short cytoplasmic tail (7). Upon ligation with anti-BDCA-2 antibodies, the receptor induces a BCR-like signaling cascade. As a result, it inhibits TLR7- and TLR9-driven type I IFN production (8), prompting the therapeutic use of anti-BDCA-2 antibodies for autoimmune diseases such as systemic or cutaneous lupus erythematosus (SLE and CLE) (9, 10).

Glycosaminoglycans (GAGs) are long, linear polysaccharides with a complex structure due to composition, length, and sulfation pattern (11). Six types of GAGs are known: heparin, heparan sulfate, chondroitin, dermatan, keratan, and hyaluronic acid. GAGs can interact with various proteins as single molecules or as part of a protein (proteoglycans, PGs), and these interactions modulate their numerous biological functions such as extracellular matrix assembly and cell signaling in normal and pathological processes (12).

Heparin is mainly produced and stored by mast cells (13). At sites of tissue injury, it is released and may serve as antithrombotic (14) or it restricts in a complex with proteases invading pathogens (15). Furthermore, unfractionated heparin (UFH) and low-molecular-weight heparin (LMWH) (fragmented UFH, e.g., enoxaparin, dalteparin) are considered essential medicines due to their use as anticoagulants (WHO Model list of essential medicines -23rd list, 2023). Although heparin and other glycosaminoglycans have anti-inflammatory activity (16–18), a role for heparin in the modulation of pDCs activity has not been investigated.

Here, we show that heparin binds to BDCA-2 and inhibits TLR9-driven IFN-α production in pDCs. We found that plasma GAGs inhibit CpG-DNA-induced IFN-α production via BDCA-2 and that this effect is partially abolished by heparinase or chondroitinase treatment. Furthermore, we demonstrate the existence of a soluble BDCA-2 generated by ADAM10/17, which binds to heparin and reverses heparin inhibitory effect on type I IFN production in pDCs. Our study describes heparin as an endogenous ligand for BDCA-2 and identifies soluble BDCA-2 as positive modulator of type I IFN production with implications for the understanding and treatment of type I IFN-mediated autoimmune diseases.

Results

Heparin Inhibits IFN-α Induction in Human Primary Immune Cells.

To study the effect of heparin on the production of type I IFNs in primary immune cells, human peripheral blood mononuclear cells (hPBMCs) or hpDCs from healthy donors were stimulated with CpG-ODN 2216 and different concentrations of heparin (UFH) or enoxaparin. Whereas the TLR9-induced IFN-α production of hPBMCs was reduced only at heparin or enoxaparin concentrations of 20 U/mL (SI Appendix, Fig. S1A), hpDCs demonstrated an almost complete inhibition of CpG-ODN-induced IFN-α production over a wide range of heparin concentrations (Fig. 1). Compared with heparin, enoxaparin only inhibited IFN-α production in hpDCs at the highest concentrations. Since IFN-α production of murine pDCs (mpDCs) was not inhibited by heparin, it suggests that this inhibitory activity is specific to hpDCs (SI Appendix, Fig. S1B). For enoxaparin, a slight reduction in IFN-α production from mpDCs was observed at the highest concentration tested (20 U/mL), suggesting that enoxaparin may weakly act on mpDCs.

Fig. 1. Inhibitory effect of heparin or enoxaparin on TLR9-induced IFN-α production in hpDCs. hpDCs (n = 3) were treated with 100 U/mL heparin, 100 U/mL enoxaparin, or CpG-ODN 2216 at 0.1 µmol/L for 20 h, and IFN-α production was determined by ELISA (Left). Immune cells were treated with indicated concentrations of heparin (Middle) or enoxaparin (Right) for 1 h and then stimulated with CpG-ODN 2216 at 0.1 µmol/L for 20 h, and IFN-α production was determined by ELISA. Relative IFN-α production was defined as the percent ratio of IFN-α produced by cells treated with CpG-ODN 2216 and heparin or enoxaparin and compared to positive control (amount of CpG-ODN 2216-induced IFN-α production observed without heparin or enoxaparin). Each data point represents an individual donor. Statistical significance was analyzed using one-way ANOVA with Dunnett’s post hoc test, *P < 0.05, **P < 0.01, or ***P < 0.001 compared to the positive control.

Heparin Mediates Its Inhibitory Effect via BDCA-2 in hpDCs.

BDCA-2 is exclusively expressed on hpDCs and absent from the mouse genome (5). Because the effect of heparin on IFN-α production was only observed on hpDCs and not on mpDCs, BDCA-2 was considered as possible receptor for heparin.

To investigate whether BDCA-2 functions as heparin receptor, the hpDCs cell line CAL-1 was selected as a model (19). First, the expression of BDCA-2 on CAL-1 was analyzed by flow cytometry (SI Appendix, Fig. S1C and Fig. 2A) and western blotting (Fig. 2B). Since the expression of BDCA-2 on CAL-1 cells was low, a BDCA-2-reconstituted CAL-1 cell line (CAL-1-BDCA-2) with a similar expression of BDCA-2 as hpDCs (Fig. 2C) was created to simulate primary hpDCs. BDCA-2-deficient CAL-1 cells (CAL-1 KO BDCA-2) generated by CRISPR/Cas9 served as negative control (Fig. 2 A and B).

Fig. 2. Inhibitory effect of heparin on type I IFN production depends on BDCA-2. (A) Representative flow cytometry histograms of BDCA-2 expression in CAL-1 cells, BDCA-2-deficient CAL-1 cells (CAL-1 KO BDCA-2), and BDCA-2-reconstituted CAL-1 cells (CAL-1-BDCA-2). Cells were stained with anti-BDCA-2-APC (AC144) labeled antibody and analyzed by flow cytometry. (B) Western blot analysis of BDCA-2 and β-ACTIN expression in CAL-1, CAL-1 KO BDCA-2, and CAL-1-BDCA-2 cells. Cell lysates were analyzed by SDS-PAGE and subsequent western blotting using anti-BDCA-2 (Cat# A58582, EpiGentek). (C) For the analysis of BDCA-2 surface expression, CAL-1-BDCA-2 and hpDCs were stained with anti-BDCA-2-APC (AC144) labeled antibody, and BDCA-2 mean fluorescence intensity (MFI) was determined by flow cytometry. (D) and (E) CAL-1-BDCA-2 cells (D) and CAL-1 KO BDCA-2 cells (E) were treated with 0.25 µg/mL anti-BDCA-2 or isotype control antibody, 4 U/mL heparin, and 0.1 µmol/L CpG-ODN 2216, and IFN-α production was determined after 20 h by ELISA (Left). Immune cells were treated with indicated concentrations of heparin and controls (0.25 µg/mL anti-BDCA-2 or isotype control antibody) for 1 h and then stimulated with 0.1 µmol/L CpG-ODN 2216 for 20 h (Right). IFN-α production was determined by ELISA. (F) CAL-1-BDCA-2 (Left) or CAL-1 KO BDCA-2 (Right) cells were treated with indicated concentrations of heparin and after 1 h, cells were washed and treated with 0.1 µmol/L CpG-ODN 2216 for 20 h, and IFN-α production was determined by ELISA. (G) CAL-1-BDCA-2 (Left) or CAL-1 KO BDCA-2 (Right) cells were treated with 0.1 µmol/L CpG-ODN 2216 and after 1 h, cells were washed and treated with indicated concentrations of heparin for 20 h, and IFN-α production was determined by ELISA. Relative IFN-α production was defined as the percent ratio of IFN-α produced by cells treated with CpG-ODN 2216 and heparin compared to positive control (amount of CpG-ODN 2216-induced IFN-α production observed without heparin). Each data point represents an experiment or donor. Statistical significance was analyzed using Student’s t test (C) or one-way ANOVA with Dunnett’s (D–G) post hoc test, *P < 0.05, **P < 0.01, or ***P < 0.001 compared to the positive control (D–G).

In CAL-1-BDCA-2 cells pretreated with heparin or anti-BDCA-2 antibody (clone AC144), a reduction of IFN-α production upon TLR9 stimulation was observed (Fig. 2D). In contrast, IFN-α production in CAL-1 KO BDCA-2 cells was not affected by anti-BDCA-2 or lower heparin concentrations (Fig. 2E). However, at the highest heparin concentration (Fig. 2E) IFN-α production was affected, possibly due to some uptake interference. Accordingly, confocal microscopy and flow cytometry analysis on CpG-ODN uptake demonstrated that heparin can moderately affect ODN uptake in a BDCA-2-independent manner (SI Appendix, Fig. S2 A–D). To circumvent a possible influence of heparin on CpG-ODN 2216 uptake, two different settings for the same costimulations were established: For the first setting, cells were treated with different concentrations of heparin for 1 h and then washed and incubated with CpG-ODN 2216 for 20 h (Fig. 2F). Alternatively, cells were treated with CpG-ODN 2216 for 1 h and then washed and incubated with different concentrations of heparin for 20 h (Fig. 2G). Under both settings, heparin reduced the production of IFN-α in CAL-1-BDCA-2 cells, but not in CAL-1 KO BDCA-2 cells (Fig. 2 F and G) suggesting a direct heparin-BDCA-2-mediated inhibitory effect on TLR9-induced IFN-α production.

Given the endogenous presence of heparin, the experimental condition where the stimulus follows heparin treatment most accurately mimics physiological conditions and was consequently selected for further experiments.

Antibody-mediated cross-linking of BDCA-2 induces a phosphorylation cascade involving SYK phosphorylation at Y525/526 (8). Of note, heparin incubation dose-dependently induced phosphorylation of SYK (Y525/526) in CAL-1-BDCA-2 cells (SI Appendix, Fig. S2E), although the level of SYK phosphorylation is not as pronounced compared to the phosphorylation induced by anti-BDCA-2 antibodies.

Incubation of pDCs with anti-BDCA-2 antibodies can lead to BDCA-2 internalization (3). Interestingly, heparin reduces surface BDCA-2 expression after an overnight incubation like the treatment with anti-BDCA-2 antibody (clone AC144) (SI Appendix, Fig. S2 F and G). Overall, these results indicate that heparin inhibits the TLR9-induced IFN-α production via BDCA-2.

Specific Interaction of BDCA-2 with Heparin.

To analyze the specificity of BDCA-2–heparin interaction, the extracellular domain (ECD) of BDCA-2 and two closely sequence-related members of dectin-2 family (dectin-2 and DCIR) (SI Appendix, Fig. S3A) were produced in HEK293 cells and purified by Strep-Tactin-mediated affinity chromatography (SI Appendix, Fig. S3 B and C and Table 1). IL-8, a known heparin-binding protein (20) served as positive control. Recombinant IL-8 was detected as two bands in the Coomassie stain and western blot. An unexpected variant at a higher molecular weight was observed, the nature of which is currently unclear and may result from the expression in HEK293 cells. All recombinant proteins were tested in vitro in an affinity protein precipitation assay using heparin-agarose and a modified ELISA with plate-coated heparin (21).

Table 1. Composition of the extracellular domain of soluble C-type lectins and IL-8

Protein	Extracellular domain	Accession number	Expected molecular weight (kDa) + Twin-Streptag	Predicted N-Glycosylation sites (~2,5 kDa per N-glycosylation)	Observed molecular weight (kDa)*	
BDCA-2	N 45–I 213	NP_001358319.1	23.44	3	40	
DCIR	Q 70–L 237	NP_057268.1	23.34	1	36	
Dectin-2	T 42–L 209	NP_001007034.1	22.93	2	30	
IL-8	V 24–S 99	NP_000575.1	12.33	0	15	
*Observed by SDS-PAGE and Coomassie staining or western blotting.

It was shown that BDCA-2, but not dectin-2 or DCIR ECDs, were precipitated by heparin-agarose (Fig. 3A). Similar data were obtained using a modified ELISA with plate-coated heparin (SI Appendix, Fig. S3D and Fig. 3B). Importantly, the interaction of BDCA-2 ECD and IL-8 with heparin was observed in a concentration-dependent manner, whereas dectin-2 and DCIR ECDs failed to interact with heparin (Fig. 3B and SI Appendix, Fig. S3E).

Fig. 3. BDCA-2 interacts with heparin and heparin-derived GAGs. (A) Western blot analysis of the affinity protein precipitation assay of BDCA-2, DCIR, and dectin-2 ECDs using heparin-agarose. Sample input (Upper) and bound proteins (Lower) were analyzed by SDS-PAGE and subsequent western blotting using Strep-Tactin HRP conjugate (IBA). (B) Dose-dependent binding of BDCA-2, DCIR, and dectin-2 ECDs to uncoated or heparin-coated (1 mg/mL) ELISA plates. Bound lectins were detected by Strep-Tactin HRP conjugate (IBA). (C) Analysis of the interaction of BDCA-2 ECD (1 µg/mL) with different GAGs as indicated. ELISA plates were uncoated or precoated with individual GAGs at 1 mg/mL. Bound BDCA-2 ECD was detected by Strep-Tactin HRP conjugate (IBA). (D) Calcium-dependent binding of BDCA-2 ECD (WT) or calcium-binding BDCA-2 mutant ECD (APA) to heparin-coated (1 mg/mL) ELISA plates. Binding studies were performed in the absence or presence of 10 mmol/L Ca2+. Bound proteins were detected by Strep-Tactin HRP conjugate (IBA). (E) Main disaccharide of heparin, composed of glucosamine (GlcN) and iduronic acid (IdoA). Green squares represent the R groups (-H or -SO3H) of the positions 2, 6 or N. (F) Binding of BDCA-2 ECD to ELISA plates coated with heparin or desulfated heparins (2-ODSH, 6-ODSH, and N-DSH) at 1 mg/mL. BDCA-2 ECD binding was detected with Strep-Tactin HRP conjugate (IBA). (G) CAL-1-BDCA-2 cells were treated with different concentrations of 2-ODSH, 6-ODSH, and N-DSH or heparin followed by subsequent stimulation with 0.1 µmol/L CpG-ODN 2216. After 20 h, IFN-α production was determined by ELISA. Relative IFN-α production was defined as the percent ratio of IFN-α produced by cells treated with CpG-ODN 2216 and heparin derivatives compared to positive control (amount of CpG-ODN 2216-induced IFN-α production observed without heparin derivatives). The data represent the mean of four individual experiments. (H) Concentration-dependent binding of BDCA-2 ECD, 3 N-glycosylation mutants (N110Q, N137Q, and N164Q), and PNGase F fully deglycosylated BDCA-2 ECD to plate-bound heparin (1 mg/mL). Protein binding was detected with Strep-Tactin HRP conjugate (IBA). Each data point represents one experiment. Statistical significance was analyzed using one-way or two-way ANOVA with Dunnett’s (G) or Tukey’s (B–D, F, and H) post hoc test, *P < 0.05, **P < 0.01, or ***P < 0.001 compared to the respectively control.

To verify the specificity of the interaction between BDCA-2 and heparin or heparin derivatives, the modified ELISA was performed with different plate-coated GAGs. Herein, BDCA-2 ECD could bind to heparin and the heparin derivatives enoxaparin and dalteparin but not to fondaparinux (synthetic heparin pentasaccharide) or other GAGs (Fig. 3C). In contrast, IL-8 binds only to heparin (SI Appendix, Fig. S3F). The lack of BDCA-2 ECD binding to non-heparin-related GAGs correlated with the inability to modulate TLR9-driven IFN-α production in CAL-1-BDCA-2 cells (SI Appendix, Fig. S3G). Since fondaparinux, a synthetic pentasaccharide analog of heparin (22), did not reduce CpG-ODN-induced IFN-α production, this observation suggests a minimal size requirement for BDCA-2 interaction.

To further confirm BDCA-2–heparin interaction, a microscale thermophoresis (MST) assay was performed. The results showed that heparin (KD = 0.88 µmol/L or 13 µg/mL), enoxaparin (KD = 0.56 µmol/L or 2.52 µg/mL) and dalteparin (KD = 0.99 µmol/L or 4.9 µg/mL) presented a higher affinity to BDCA-2 ECD compared to chondroitin (KD = 6.00 µmol/L or 150 µg/mL) or dermatan (KD = 1.90 µmol/L or 47 µg/mL) (Table 2 and SI Appendix, Fig. S3 H and I). In summary, BDCA-2 specifically interacts with UF and LMW heparin.

Table 2. Dissociation constant (KD) of the interaction between BDCA-2 and different GAGs

Ligand	KD (µmol/L)	SD	
Heparin	0.88	0.19	
Enoxaparin	0.56	0.23	
Dermatan	1.89	0.61	
Chondroitin	6.04	3.11	
Dalteparin	0.99	0.35	
Fondaparinux	2.55	1.64	
The KD values were calculated from the primary data of the microscale thermophoresis measurements.

Characterization of the BDCA-2–Heparin Interaction.

Basic residues such as arginine, lysine, and histidine interact with heparin or negatively charged GAGs via electrostatic interactions. Heparin binding sites contain clusters (1 to 3) of basic residues (XBnX, where n = 1 to 3 and B represent a basic residue and X is a nonbasic residue) (23). Several XBnX heparin binding sites were found in the BDCA-2 sequence and some of these sites are accessible for a possible interaction (PDB: 4zes) (SI Appendix, Fig. S4 A and B), supporting the idea of the BDCA-2–heparin interaction.

BDCA-2 has a C-type CRD with five conserved amino acids (E172, N174, E178, N195, and D195) forming a Ca2+-binding site (7). To evaluate the influence of Ca2+ on the BDCA-2–heparin interaction, a mutated BDCA-2 ECD protein with two mutations (E172A and N174A, termed BDCA-2-APA) in the Ca2+-binding site was engineered. Importantly, BDCA-2–heparin interaction was enhanced by the addition of Ca2+ and impaired by EDTA, while the binding of heparin to the APA mutant was in general weak and not influenced by both treatments (Fig. 3D and SI Appendix, Fig. S4C). Moreover, circular dichroism analysis of the APA mutant demonstrated changes in the secondary structure compared to BDCA-2 ECD (SI Appendix, Fig. S4 D and E), explaining the impaired ability of the APA mutant to interact with heparin in the absence of Ca2+ (Fig. 3D).

To further validate the influence of Ca2+ on the BDCA-2–heparin interaction, the affinity of BDCA-2 ECD and different GAGs was measured in the presence of Ca2+ using MST. Under these conditions, the KD of heparin was reduced from 0.88 to 0.34 µmol/L (Tables 2 and 3 and SI Appendix, Fig. S3 H and I). Interestingly, the KD of enoxaparin and dalteparin were increased with Ca2+ present, showing an unexpected behavior.

Table 3. Dissociation (KD) constant of the interaction between BDCA-2 and different GAGs with calcium (CaCl2 1 mmol/L)

Ligand	KD (µmol/L)	SD	
Heparin	0.34	0.08	
Enoxaparin	4.41	1.17	
Dermatan	4.4	0.85	
Chondroitin	13.55	1.09	
Dalteparin	3.17	1.49	
Fondaparinux	n.d.	n.d.	
The KD values were calculated from the primary data of the microscale thermophoresis measurements.

The main disaccharide unit of heparin contains three main substitutions such as 2-O-sulfation of the iduronic acid and an N- and 6-O-sulfation of the glucosamine residue (Fig. 3E and SI Appendix, Fig. S4F) (18). To assess the importance of heparin sulfation pattern in BDCA-2 binding, the interaction of heparin, 2-O desulfated heparin (2-ODSH), 6-O desulfated heparin (6-ODSH), and N-desulfated heparin (N-DSH) with BDCA-2 ECD was investigated. Of note, desulfated heparins interacted weaker with BDCA-2 ECD compared to fully sulfated heparin (Fig. 3F). Moreover, only 2-ODSH partly inhibited TLR9-induced IFN-α production in CAL-1-BDCA-2 cells (Fig. 3G).

N-glycosylation is a posttranslational modification of proteins involved in protein folding, transport, and function (24). For BDCA-2, NetNGlyc–1.0 (25) detects three possible N-linked glycosylation sites at N110, N137, and N164 (SI Appendix, Fig. S4G) containing the glycosylation motif N-X-S/T. N-glycosylation was confirmed by PNGase F treatment of the BDCA-2 ECD that showed a lower molecular size compared to the untreated protein (SI Appendix, Fig. S4H). To evaluate the importance of N-glycosylation in the BDCA-2–heparin interaction, three single N-glycosylation mutants of the BDCA-2 ECD (N110Q, N137Q, and N164Q) were generated, purified, and used in the modified ELISA. For every mutant, a change in molecular weight was observed (SI Appendix, Fig. S4I), demonstrating N-linked glycans at the predicted positions in BDCA-2. The N-glycosylation mutants of the BDCA-2 ECD showed weaker interactions with heparin compared to fully N-glycosylated BDCA-2 ECD (Fig. 3H). Furthermore, complete N-deglycosylated BDCA-2 ECD by PNGase F treatment lost the ability to interact with heparin (Fig. 3H). Circular dichroism analysis demonstrated that the secondary structure of the mutant ECDs did not change strongly compared to fully glycosylated BDCA-2 ECD. Accordingly, the impaired interaction between the mutants and heparin can not be explained by misfolding (SI Appendix, Fig. S4 J–L).

Overall, these results clearly demonstrate the importance of N-glycosylation, heparin sulfation pattern, and Ca2+ in the BDCA-2–heparin interaction.

Interaction of Plasma GAGs with BDCA-2.

GAGs can be found in plasma, urine, and extracellular matrix (11, 26). To test whether GAG-containing plasma inhibits the TLR9-induced IFN-α production via BDCA-2, stimulation assays with CAL-1-BDCA-2 cells, CpG-ODN 2216, and heparin or plasma were performed. For some experiments BDCA-2 ECD was added to investigate its modulatory role for immune stimulation. Heparin and plasma inhibited the TLR9-induced IFN-α production, and BDCA-2 ECD neutralized the inhibitory effect of both in a concentration-dependent manner (Fig. 4A). These observations suggest that heparin and plasma GAGs interact with BDCA-2 ECD that neutralizes their inhibitory activity. Of note, plasma also inhibited the CpG-ODN 2216-induced IFN-α production of hPBMCs and hpDCs, and BDCA-2 ECD treatment reversed this effect, while dectin-2 ECD was ineffective (Fig. 4 B and C). Like anti-BDCA-2 antibody and heparin treatment, incubation with plasma led to SYK (Y525/526) and BLNK- phosphorylation (Y96) in CAL-1-BDCA-2 cells but not in CAL-1 KO BDCA-2 cells (SI Appendix, Fig. S5 A and B) supporting the view that a plasma component induces the BDCA-2 signaling pathway.

Fig. 4. Heparin and plasma-mediated inhibition of CpG-ODN 2216–induced IFN-α production in hPBMCs, CAL-1-BDCA-2, or hpDCs and its modulation by BDCA-2 ECD. (A) CAL-1-BDCA-2 cells (n = 4) were treated with different concentrations of BDCA-2 ECD in the absence or presence of heparin (4 U/mL) (Left) or human plasma (0.5%) (Right) for 1 h and then treated with 0.1 µmol/L CpG-ODN 2216. After 20 h, the supernatant was harvested, and IFN-α production was determined by ELISA. (B) hPBMCs from four donors were treated with 4 µg/mL BDCA-2 ECD, 2.5% plasma, and 1 µmol/L CpG-ODN 2216 for 20 h, and IFN-α production was determined by ELISA (Left). hPBMCs (n = 4) were treated with indicated concentrations of BDCA-2 ECD with or without 2.5% plasma for 1 h and then stimulated with CpG-ODN 2216 1 µmol/L (Middle). hPBMCs (n = 3) were treated with indicated concentrations of dectin-2 with or without 2.5% plasma for 1 h and then stimulated with CpG-ODN 2216 1 µmol/L for 20 h (Right). After stimulation, IFN-α production was determined by ELISA. (C) hpDCs (n = 3) were treated with indicated concentrations of plasma for 1 h and then stimulated with CpG-ODN 2216 0.1 µmol/L (Left). After 20 h IFN-α production was determined by ELISA. hpDCs (n = 3) were treated with BDCA-2 ECD or dectin-2 (8 μg/mL) with or without 0.25% plasma for 1 h and then stimulated with CpG-ODN 2216 0.1 µmol/L for 20 h (Right). After stimulation, IFN-α production was determined by ELISA. (D) CAL-1-BDCA-2 cells were treated for 1 h with heparin (hep) or 1% plasma (P) untreated or overnight treated with a heparinase mixture consisting of heparinases I, II, and III (Hps) and chondroitinase ABC (Ch) and then stimulated with 0.1 µmol/L CpG-ODN 2216 for 20 h. After stimulation, IFN-α was determined by ELISA (Left). hpDCs (n = 6) were treated for 1 h with plasma 2% or 1%, either untreated or overnight treated with a heparinase mixture consisting of Hps I, II, and III, followed by stimulation with 0.1 µmol/L CpG-ODN 2216 for 20 h. After stimulation, IFN-α was determined by ELISA (Right). Relative IFN-α production was defined as the percent ratio of IFN-α produced by cells treated with CpG-ODN 2216 and heparin and compared to positive control (amount of CpG-ODN 2216-induced IFN-α production observed without heparin or plasma). Each data point represents one individual donor or experiment. Statistical significance was analyzed using one-way or two-way ANOVA with Dunnett’s (C Left panel) or Tukey’s (A–C Right panel and D) post hoc test, *P < 0.05, **P < 0.01, or ***P < 0.001 compared to the respective controls of each experiment.

To confirm that human plasma GAGs inhibit type I IFN production induced by CpG-ODN 2216, highly negatively charged biomolecules were isolated from plasma and ROSA cells (a human mast cell line) (27) using titanium dioxide beads (28) (SI Appendix, Fig. S5 C and D). GAG content and composition were analyzed by polyacrylamide gel electrophoresis and subsequent toluidine staining (29, 30) under conditions with or without GAG lyase treatment by heparinases (Hps) I, II, and III or chondroitinase ABC (ChABC). The main detectable GAG in plasma is chondroitin since ChABC, but not Hps treatment, abolishes the detection of toluidine-stained bands (SI Appendix, Fig. S5E). The amount of heparin was probably too low to be efficiently detected in this assay. However, heparin was detectable in plasma from different donors by Anti IIa Assay Kit (SI Appendix, Fig. S5F) (31–33). Furthermore, extracts from plasma and ROSA cells inhibited the BDCA-2–heparin interaction like exogenous heparin (SI Appendix, Fig. S5G), and both extracts inhibited TLR9-induced IFN-α production from CAL-1-BDCA-2 cells in a concentration-dependent manner (SI Appendix, Fig. S5H). Other plasma components such as fibrinogen or albumin had no modulatory activity (SI Appendix, Fig. S5I).

Of note, treatment of human plasma with Hps (I, II, and III), ChABC or both lyases partially abolished the ability of plasma to inhibit IFN-α production in TLR9-stimulated CAL-1-BDCA-2 cells (Fig. 4D). Similar results were obtained with Hps-treated plasma on CpG-ODN stimulated hpDCs (Fig. 4D). These results confirm that plasma heparin has an inhibitory activity on IFN-α production. In case of chondroitinase treatment, this result suggests that plasma chondroitin may also participate in the modulation of type I IFN production. Since chondroitin did not or only weakly interact with BDCA-2 ECD (Fig. 3C and Table 2) and failed to inhibit IFN-α production (SI Appendix, Fig. S3G), it suggests that the chondroitin-mediated modulatory function may be through a different regulatory receptor on pDCs. However, treatment of heparin with Hps (I, II, and III) but not ChABC completely abolished its inhibitory activity.

In summary, plasma containing heparin inhibits CpG-ODN 2216-induced IFN-α production in hpDCs by interacting with BDCA-2. Importantly, BDCA-2 ECD antagonizes this activity.

Human Plasma Contains solBDCA-2 with Influence on IFN-α Production.

Membrane receptors can exist in soluble form, either through proteolytic cleavage of the extracellular domain or via secretion of a soluble isoform. They can function as decoy receptors, ligand carriers, ligand stabilizers or act as ligands (34). Since the BDCA-2 ECD (a soluble recombinant form of the receptor) neutralized the inhibitory effect of heparin and plasma on type I IFN production, we wondered whether solBDCA-2 may be present in plasma with modulating activity on IFN-α production from pDCs.

SolBDCA-2 in the plasma of 10 healthy donors was detected and quantified by a specific BDCA-2 ELISA (SI Appendix, Fig. S6A), with concentrations ranging from 0.3 to 63 ng/mL (SI Appendix, Table S1; mean of 12 ng/mL). To study the function of solBDCA-2, stimulations were performed on CAL-1-BDCA-2 cells or hpDCs with plasma and CpG-ODN-2216 followed by the quantification of IFN-α and solBDCA-2 in the corresponding plasma. Plasma solBDCA-2 concentration correlated positively with the IFN-α produced by CAL-1-BDCA-2 cells (Fig. 5A, r = 0.62; n = 10; P = 0.05) or hpDCs (Fig. 5B; r = 0.70; n = 8; P = 0.05). Furthermore, plasma heparin concentration correlated negatively with IFN-α produced by CAL-1-BDCA-2 after stimulation with CpG-ODN 2216 and plasma (SI Appendix, Fig. S6B; r = −0.49; n = 10; P = 0.15). Moreover, a slightly positive correlation was observed between the percentage of hpDCs in hPBMCs and solBDCA-2 concentration in plasma (SI Appendix, Fig. S6C; r = 0.25; n = 8; P = 0.5).

Fig. 5. SolBDCA-2 is found in human plasma, and BDCA-2 shedding by ADAM proteases is increased upon cellular activation. (A and B) Correlation between solBDCA-2 concentration in plasma of different donors (n = 10) and IFN-α production after stimulation of CAL-1-BDCA-2 cells (A) or hpDCs (n = 8) (B) with 1% of plasma for 1 h and subsequent 0.1 µmol/L CpG-ODN 2216 stimulation for 20 h. Results were analyzed via ELISA. (C) Western blot analysis of solBDCA-2 in the supernatant of CAL-1, CAL-1-BDCA-2, CAL-1 KO BDCA-2, and CAL-1 KO BDCA-2-reconstituted with BDCA-2 cells (CAL-1 KO BDCA-2-BDCA-2) using heparin-agarose for precipitation. Samples were analyzed by SDS-PAGE and subsequent western blotting analysis using anti-BDCA-2 (Cat# A58582, EpiGentek). (D and E) CAL-1-BDCA-2 cells (D) or hpDCs (E) were treated with indicated concentrations of CpG-ODN 2216, 1 µmol/L ionomycin (IO) and 50 ng/mL phorbol myristate acetate (PMA). After 20 h of incubation, IFN-α production (Left) as well as solBDCA-2 (Right) concentrations were determined by ELISA. (F) CAL-1-BDCA-2 cells were treated with different concentrations of GW280264X (GW) for 1 h and then stimulated with CpG-ODN 2216 0.1 µmol/L or IO 1 µmol/L or PMA 50 ng/mL. After 20 h of incubation, solBDCA-2 production was determined by ELISA. (G) hpDCs (n = 7) were treated with GW (10 µmol/L) or GI (10 µmol/L) for 1 h and then stimulated or not with CpG-ODN 2216 0.1 µmol/L. After 20 h of incubation, solBDCA-2 production was determined by ELISA. (H) PCR analysis of isoforms of BDCA-2 mRNA. RNA from hpDCs (n = 3) was isolated and analyzed for the presence of BDCA-2 isoforms by RT-PCR amplification. The PCR products were analyzed by a 2% agarose gel. Relative solBDCA-2 production was defined as the percent ratio of solBDCA-2 in cells treated with GW or GI and CpG-ODN 2216/IO/PMA compared to positive control (amount of BDCA-2 production observed without the inhibitors). Each data point represents one individual donor or experiment. For correlation analysis, Pearson correlation coefficients were performed (A and B), and a P < 0.05 was considered significant. Statistical significance was analyzed using one-way ANOVA with Dunnett’s (D–F) or Tukey’s (G) post hoc test, *P < 0.05, **P < 0.01, or ***P < 0.001 compared to the negative control (medium) (D and E) and compared to the positive control (F and G).

SolBDCA-2 was further confirmed by heparin-affinity precipitation from the supernatant of BDCA-2-expressing CAL-1 cells (Fig. 5C) followed by western blotting with a BDCA-2 specific antibody (SI Appendix, Fig. S6D). The size of the solBDCA-2 was slightly smaller (approximately 35 kDa) (Fig. 5C) compared to full-length BDCA-2 (approximately 40 kDa, Fig. 2B).

To study whether cell activation increases the amount of solBDCA-2 by proteolytic cleavage via sheddases such as ADAM10 or 17, CAL-1-BDCA-2 cells were stimulated with CpG-ODN, ionomycin (IO; ADAM10/17 activator) (35) or phorbol myristate acetate (PMA; ADAM17 activator) (36). Of note, CpG-ODN, IO and PMA strongly increased the release of solBDCA-2 on CAL-1-BDCA-2 cells (Fig. 5D). Importantly, hpDCs also release solBDCA-2 (Fig. 5E) and CpG-ODN stimulation enhanced the production (Fig. 5E), but compared to CAL-1 cells, IO showed a negative effect and PMA did not influence the production. To confirm the involvement of ADAM proteases, selective inhibitors for ADAM10/17 and ADAM10 such as GW280264X (GW) and GI254023X (GI), were used, respectively (35). The ADAM10/17 specific inhibitor GW significantly reduced the release of solBDCA-2 from cells that were treated with CpG-ODN 2216, IO, and to a lesser extent PMA (Fig. 5F). In contrast, the ADAM10 specific inhibitor GI did not affect the release of solBDCA-2 (SI Appendix, Fig. S6E). In the case of hpDCs, both inhibitors (GW and GI) significantly reduced the release of solBDCA-2 from cells that were treated with CpG-ODN 2216 (Fig. 5G).

The secretion of a soluble isoform of BDCA-2 has been suggested before (3). To confirm the presence of a soluble isoform of BDCA-2, BDCA-2 transcript from hpDCs RNA was analyzed by RT-PCR (Fig. 5H). This result demonstrated the presence of at least two different transcripts and the sequencing of these PCR products confirmed the existence of the full-length mRNA as well as a shorter version without exon 3 coding for the transmembrane region (SI Appendix, Fig. S6 F–I).

Together, these results suggest that solBDCA-2 found in plasma is generated by proteolytic cleavage from surface BDCA-2 by ADAM10/17 and to some extent by secretion of a transmembrane-less soluble isoform. Importantly, cellular activation enhances the release of solBDCA-2.

Dysregulation of solBDCA-2 and Heparin in Inflammatory Diseases.

Since CpG-ODN-activated pDCs increase solBDCA-2 release, we tested whether plasma/serum concentrations of solBDCA-2 and heparin are altered in inflammatory diseases. Accordingly, solBDCA-2 concentrations were quantified in plasma of pDCs-related autoimmune diseases such as bullous pemphigoid (BP) (37, 38) and psoriasis (Pso) (39). Additionally, solBDCA-2 concentration was quantified in serum from scrub typhus (ST) patients. ST is an infectious disease caused by Orientia tsutsugamushi (Ot), which induces type I IFN release (40), antinuclear antibodies (ANA) (41) and is associated with the risk for autoimmune diseases, particularly SLE (42).

SolBDCA-2 in BP (n = 19) and Pso (n = 20) patients was found to be reduced compared to healthy controls (HC; n = 20), whereas heparin concentrations were similar (Fig. 6A). Furthermore, no correlation was observed between heparin and solBDCA-2 levels (SI Appendix, Fig. S7A; r = −0.07; n = 59; P = 0.60). It is noteworthy that plasma from both autoimmune diseases showed a tendency of lower solBDCA-2 levels (P < 0.05 for BP and P = 0.17 for Pso) compared to HC reflecting an influence of certain diseases on solBDCA-2 production.

Fig. 6. Dysregulation of solBDCA-2 in autoimmune diseases and scrub typhus (ST). (A) Determination of solBDCA-2 (Left) and heparin (Right) concentrations in plasma from 20 healthy controls (HC), 19 bullous pemphigoid (BP) patients, and 20 psoriasis (Pso) patients by ELISA and Anti IIa Assay Kit, respectively. (B) Determination of solBDCA-2 concentration in serum from 24 HC and 48 ST patients by ELISA (Left). Determination of heparin concentration in the serum from 15 HC and 15 ST patients by the Anti IIa Assay Kit (Right). (C) Correlation between solBDCA-2 and anti-dsDNA in the serum of HC (n = 19) together with ST patients (n = 24). (D) Correlation between solBDCA-2 and IFN-γ in the serum of healthy controls (n = 19) together with ST patients (n = 24). (E) SolBDCA-2 concentration in the serum of HC (n = 24), 16S-positive (n = 23), and 16S-negative (n = 24) ST patients. (F) Anti-dsDNA antibodies in the serum of HC (n = 47) and ST patients, both 16S-positive and 16S-negative (n = 169). The red dashed line represents the cutoff of 20 IU/mL. Each data point represents an individual donor/patient. *P < 0.05, **P < 0.01, or ***P < 0.001 compared to the healthy control. For correlation analysis, Pearson correlation coefficients were performed (C and D), and a P < 0.05 was considered significant. Statistical significance was analyzed using Student’s t test (B) or one-way ANOVA with Dunnett’s (A, E, and F) post hoc test, *P < 0.05, **P < 0.01, or ***P < 0.001 compared to the HC.

In ST patients, solBDCA-2 was found to be significantly increased and heparin was significantly decreased compared to the HC (Fig. 6B). To assess whether the solBDCA-2 increase in ST correlates with pro-inflammatory cytokines/chemokines during infection or anti-double stranded DNA (anti-dsDNA) antibodies in serum, a total correlation matrix (correlation with samples of ST and HC) was performed. This analysis included data from cytokine and chemokines previously published by Münch et al. (43), as well as recently analyzed data for solBDCA-2 concentrations and anti-dsDNA antibodies (as a serologic marker for SLE) from the same samples of ST and HC. Here, a slight to moderate but significant correlation of solBDCA-2 with IL-6, IFN-γ, and anti-dsDNA antibodies was observed (SI Appendix, Fig. S7B). Of note, solBDCA-2 concentrations correlated positively with anti-dsDNA antibodies (Fig. 6C; r = 0.30; n = 43; P = 0.04), IFN-γ (Fig. 6D; r = 0.33; n = 43; P = 0.02) and IL-6 (SI Appendix, Fig. S7C; r = 0.41; n = 43; P = 0.005). Additionally, a moderate negative correlation was observed between solBDCA-2 and heparin (SI Appendix, Fig. S7D; r = −0.36; n = 30; P = 0.05) as well as a strong negative correlation between anti-dsDNA antibodies and heparin (SI Appendix, Fig. S7E; r = −0.72; n = 30; P < 0.001). Principal component analysis (PCA) of solBDCA-2, heparin and anti-dsDNA antibodies levels (n = 30) demonstrated two different clusters consisting of ST patients and HC (SI Appendix, Fig. S7F).

Münch et al. (43) categorized the samples into patients with bacteremia (16S+), patients without bacteremia (16S−) and HC for analysis of cytokines/chemokines. This assignment was also performed for the analysis of solBDCA-2 and anti-dsDNA antibodies. In this analysis, 16S− presented higher levels of solBDCA-2 and anti-dsDNA antibodies compared to 16S+ and both presented higher and significant levels (16S+ and 16S−) than HC (Fig. 6 E and F). A total of 13 samples from 169 (1 sample from 16S+ group and 12 samples from 16S− group; 7.7%) exceeded the cutoff value of the anti-dsDNA antibody assay (20 IU/mL) as a highly specific biomarker for SLE (Fig. 6F). In a separate analysis of this subgroup of patients with ST (anti-dsDNA positive), a significant increase in solBDCA-2 and anti-dsDNA antibodies was found (SI Appendix, Fig. S7 G and H). Overall, these results show different patterns of solBDCA-2 dysregulation in human autoimmune/inflammatory diseases.

Discussion

pDCs play an essential role in viral infections due to prominent type I IFN production. However, type I IFN production can be negatively controlled by regulatory receptors such as the C-type lectin BDCA-2. This modulatory role is important since prolonged type I IFN production has been implicated in the initiation and/or maintenance of autoimmunity (2). Heparin, a GAG product of mast cells, is widely known for its anticoagulant function and for its use as a medicament (44). Despite that heparin and other GAGs have shown anti-inflammatory activity (16–18), a role of heparin in modulating type I IFN produced by pDCs has not been addressed.

Here, we demonstrate that heparin inhibits type I IFN produced by pDCs through a specific interaction with BDCA-2. This interaction is specific since heparin does not interact with closely related lectins such as DCIR and dectin-2. In addition, other natural GAGs did not interact with BDCA-2 nor inhibit type I IFN production.

We further demonstrated that N-glycosylation of BDCA-2 is necessary for heparin interaction. Similar results were reported for the mannose receptor, where ligand binding and activity also depend on N-glycosylation (45). Presumably, N-glycosylations can ensure a protein conformation suitable for ligand binding. The heparin sulfation pattern is also determining for the interaction with BDCA-2 and the inhibition of IFN since the desulfated heparins (2-ODSH, 6-ODSH, and N-DSH) demonstrated reduced binding affinity and in the case of IFN inhibition, 6-O- and N-sulfation were necessary, whereas the 2-O-sulfation played a minor role. Similarly, the importance of the heparin sulfation pattern in the interaction with antithrombin III and platelet factor 4 has also been observed (46, 47).

Others have identified HCV-derived glycoprotein E2 (48), HIV pg120 (49), as well as macroglobulin and IgG as ligand for BDCA-2 (50). The source of BDCA-2 protein for these binding studies was either the CRD expressed in Escherichia coli (7, 49, 50), or cell surface–expressed BDCA-2 (48, 51). In our work, BDCA-2 ECD was produced in HEK293 cells and since N-glycosylation is involved in heparin binding, recombinant BDCA-2 protein produced in E. coli may not properly define all ligands. Binding studies with cell surface expressed BDCA-2 identified HIV gp120 (49) and HCV-derived glycoprotein E2 (48) as ligand. However, E2 protein binding to BDCA-2 was not specific since it also bound to cell surface DCIR (48) and binding of gp120 was rather weak (51).

Notably, human plasma contains heparin and other GAGs in free form or bound as PG (11, 32, 33). The GAGs isolated from plasma and a mast cell line inhibit type I IFN and interact with BDCA-2. Plasma treated with Hps has a reduced ability to inhibit type I IFN production in pDCs, suggesting that plasma heparin and/or heparin PGs tightly control pDCs activity through BDCA-2. Additionally, in the context of plasma chondroitin and/or chondroitin PGs may also be involved in the control of type I IFN, but presumably via a BDCA-2 independent mechanism.

SolBDCA-2 was detected in human plasma and the activation of pDCs increases the release of solBDCA-2, which is mediated by ADAM10/17. Additionally, a splice isoform lacking a transmembrane domain, may also be involved in the release of solBDCA-2 (3). Regarding function, solBDCA-2 is a natural heparin antagonist that favors the production of type I IFNs. In agreement with our finding, the presence of a solBDCA-2 in plasma has been suggested (52). Overall, pDCs in circulation are in tight control for type I interferon production due to the presence of heparin and this control is lifted once the cells leave the circulation and move to tissues. Once in the tissue, the activation of pDCs produces solBDCA-2, which neutralizes heparin and supports the production of type I IFN (Fig. 7).

Fig. 7. Graphical representation of the IFN regulatory network based on BDCA-2, solBDCA-2, and heparin. pDCs in circulation are in tight control on type I interferon production due to the presence of heparin, and this control is lifted once the cells leave the circulation and move to tissues where the concentration of heparin is lower and the activation of pDCs triggers the release of solBDCA-2. FcRγ: Fc receptor gamma-chain; an adapter protein containing immunoreceptor tyrosine-based activation motif (ITAM). SYK: tyrosine-protein kinase SYK. MYD88: myeloid differentiation primary response protein MyD88; an adapter protein involved in the signaling of TLRs.

Type I IFN produced by pDCs is involved in Pso (39) and autoimmune blistering diseases such as BP (37). Low levels of solBDCA-2 were found in both diseases; this may be explained by pDCs exhaustion, a phenomenon that has been observed in murine models of chronical viral infections/inflammation (53). However, pDCs exhaustion has not been described for both diseases, and therefore, the clinical significance of solBDCA-2 needs further confirmation.

The anti-BDCA-2 antibody litifilimab is beneficial for patients with SLE (9, 10), demonstrating the importance of this receptor in the control of an autoimmune disease. Therefore, understanding the activity of regulatory receptors such as BDCA-2/solBDCA-2 and its ligands is important and may give unique perspectives for possible treatment of autoimmune diseases.

ST is a neglected, emerging infectious disease caused by the obligate intracellular bacterium Ot. Ot infects several types of cells (54), and induces type I IFN (40), as well as ANA antibodies during the acute phase of infection (41), which are associated with an increased risk of developing SLE (42). SolBDCA-2 was up-regulated in ST and correlated positively with anti-dsDNA antibodies, this suggests an increased activation of pDCs during the infection and an essential role of solBDCA-2 in the pathogenesis of infection-related autoimmune diseases. The presence of anti-dsDNA antibodies in ST patients may confirm the relationship of ST with SLE in a subgroup of individuals. Additionally, solBDCA-2 correlated positively with IL-6 and IFN-γ, suggesting a possible synergic/regulatory effect of solBDCA-2 not only for type IFN I. The negative correlation between heparin and solBDCA-2/anti-dsDNA antibodies supports the view that dysregulation in the regulatory network of solBDCA-2 or heparin causes or is a marker for inflammatory disease.

Materials and Methods

All the materials, methods, and statistical analysis used in this study can be found in SI Appendix, Materials and Methods. This section describes all the biochemical, cell culture, stimulations, and imaging methods used in this study. Informed consent was obtained from all blood donors. The local ethics committees of Justus-Liebig-University Gießen and Philipps-University Marburg approved the use of human blood samples.

Supplementary Material

Appendix 01 (PDF)

We thank G. Bein, Institute for Clinical Immunology and Transfusion Medicine, Justus-Liebig-University Giessen, for providing human buffy coats. ROSA cells were kindly provided by Prof. Dr. Michel Arock, Department of Hematological Biology, Pitié-Salpêtrière Hospital, Pierre and Marie Curie University (UPMC). We acknowledge the contribution of the Core Facility “Protein Biochemistry and Spectroscopy” (Oliver Stehling) and Core Facility “FACS” (Hartmann Raifer) of the Philipps-University Marburg. We thank Jörg Bartsch for discussion on ADAM protease-mediated protein shedding. This work was supported by Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) Project-ID 369799452—TRR237—A02 and Project-ID 114933180-TR84—C10 to S.B.

Author contributions

P.Y., S.-A.F., C.K., F.V.-S., A.K., and S.B. designed research; F.V.-S., L.S., E.R., C.C.-S.M., and S.B. performed research; T.M., C.L.Z., and C.M. contributed new reagents/analytic tools; F.V.-S., S.-A.F., and S.B. analyzed data; and F.V.-S., A.K., and S.B. wrote the paper.

Competing interests

The authors declare no competing interest.

Data, Materials, and Software Availability

Microscale thermophoresis analysis data and confocal imaging data are available from Mendeley Data (https://doi.org/10.17632/x7hw356g4d.1) (55). All other data are included in the manuscript and/or SI Appendix. Cell lines and materials used in this study are available from the corresponding author. Previously published data were used for this work (43).

Supporting Information

This article is a PNAS Direct Submission.
==== Refs
1 M. Gilliet, W. Cao, Y. J. Liu, Plasmacytoid dendritic cells: Sensing nucleic acids in viral infection and autoimmune diseases. Nat. Rev. Immunol. 8 , 594–606 (2008).18641647
2 Y. Ye, B. Gaugler, M. Mohty, F. Malard, Plasmacytoid dendritic cell biology and its role in immune-mediated diseases. Clin. Transl. Immunol. 9 , 1–19 (2020).
3 A. Dzionek , BDCA-2, a novel plasmacytoid dendritic cell-specific type II C-type lectin, mediates antigen capture and is a potent inhibitor of interferon α/β induction. J. Exp. Med. 194 , 1823–1834 (2001).11748283
4 B. Kerscher, J. A. Willment, G. D. Brown, The Dectin-2 family of C-type lectin-like receptors: An update. Int. Immunol. 25 , 271–277 (2013).23606632
5 S. Blomberg, M.-L. Eloranta, M. Magnusson, G. V. Alm, L. Rönnblom, Expression of the markers BDCA-2 and BDCA-4 and production of interferon-α by plasmacytoid dendritic cells in systemic lupus erythematosus. Arthritis Rheum 48 , 2524–2532 (2003).13130472
6 W. Cao , BDCA2/FcεRIγ complex signals through a novel BCR-like pathway in human plasmacytoid dendritic cells. PLoS Biol. 5 , 2190–2200 (2007).
7 S. A. F. Jégouzo , A novel mechanism for binding of galactose-terminated glycans by the C-type carbohydrate recognition domain in blood dendritic cell antigen 2. J. Biol. Chem. 290 , 16759–16771 (2015).25995448
8 J. Röck , CD303 (BDCA-2) signals in plasmacytoid dendritic cellsvia a BCR-like signalosome involving Syk, Slp65 and PLCγ2. Eur. J. Immunol. 37 , 3564–3575 (2007).18022864
9 R. Furie , Monoclonal antibody targeting BDCA2 ameliorates skin lesions in systemic lupus erythematosus. J. Clin. Invest. 129 , 1359–1371 (2019).30645203
10 V. P. Werth , Trial of Anti-BDCA2 Antibody Litifilimab for Cutaneous Lupus Erythematosus. N Engl. J. Med. 387 , 321–331 (2022).35939578
11 S. D. Vallet, O. Clerc, S. Ricard-Blum, Glycosaminoglycan-protein interactions: The first draft of the glycosaminoglycan interactome. J. Histochem. Cytochem. 69 , 93–104 (2021).32757871
12 N. K. Karamanos , Proteoglycan chemical diversity drives multifunctional cell regulation and therapeutics. Chem. Rev. 118 , 9152–9232 (2018).30204432
13 M. Krystel-Whittemore, K. N. Dileepan, J. G. Wood, Mast cell: A multi-functional master cell. Front. Immunol. 6 , 1–12 (2016).
14 R. Lassila, A. Jouppila, Mast cell-derived heparin proteoglycans as a model for a local antithrombotic. Semin. Thromb. Hemost. 40 , 837–844 (2014).25393636
15 R. L. Stevens, R. Adachi, Protease-proteoglycan complexes of mouse and human mast cells and importance of their β-tryptase-heparin complexes in inflammation and innate immunity. Immunol. Rev. 217 , 155–167 (2007).17498058
16 I. C. Severin , Glycosaminoglycan analogs as a novel anti-inflammatory strategy. Front. Immunol. 3 , 1–12 (2012).22679445
17 S. J. Fritchley, J. A. Kirby, S. Ali, The antagonism of interferon-gamma (IFN-γ) by heparin: Examination of the blockade of class II MHC antigen and heat shock protein-70 expression. Clin. Exp. Immunol. 120 , 247–252 (2000).10792372
18 M. Lima, T. Rudd, E. Yates, New applications of heparin and other glycosaminoglycans. Molecules 22 , 1–11 (2017).
19 T. Maeda , A novel plasmacytoid dendritic cell line, CAL-1, established from a patient with blastic natural killer cell lymphoma. Int. J. Hematol. 81 , 148–154 (2005).15765784
20 D. P. Witt, A. D. Lander, Differential binding of chemokines to glycosaminoglycan subpopulations. Curr. Biol. 4 , 394–400 (1994).7922353
21 D. J. Hof, E. M. M. Versteeg, C. H. A. van de Lest, W. F. Daamen, T. H. van Kuppevelt, A versatile salt-based method to immobilize glycosaminoglycans and create growth factor gradients. Glycoconj J. 36 , 227–236 (2019).31055697
22 J. C. Reverter, Fondaparinux sodium. Drugs of Today 38 , 185–194 (2002).12532174
23 E. M. Muñoz, R. J. Linhardt, Heparin-binding domains in vascular biology. Arterioscler Thromb Vasc Biol. 24 , 1549–1557 (2004).15231514
24 K. T. Schjoldager, Y. Narimatsu, H. J. Joshi, H. Clausen, Global view of human protein glycosylation pathways and functions. Nat. Rev. Mol. Cell Biol. 21 , 729–749 (2020).33087899
25 R. Gupta, S. Brunak, Prediction of glycosylation across the human proteome and the correlation to protein function. Pac Symp. Biocomput. 7 , 310–322 (2002).
26 S. Bratulic , Analysis of normal levels of free glycosaminoglycans in urine and plasma in adults. J. Biol. Chem. 298 , 101575 (2022).35007531
27 R. Saleh , A new human mast cell line expressing a functional IgE receptor converts to tumorigenic growth by KIT D816V transfection. Blood 124 , 111–120 (2014).24677542
28 M. Wilson, A. Saiardi, Inositol phosphates purification using titanium dioxide beads. Bio. Protoc. 8 , 1–8 (2018).
29 N. Volpi, F. Maccari, Detection of submicrogram quantities of glycosaminoglycans on agarose gels by sequential staining with toluidine blue and Stains-All. Electrophoresis 23 , 4060–4066 (2002).12481260
30 J. A. G. Rodrigues , Mild-acid hydrolysis of a native polysulfated fraction from Acanthophora muscoides generates sulfated oligosaccharides displaying in vitro thrombin generation inhibition. Acta Sci. Biol. Sci. 38 , 7–15 (2016).
31 S. Cavari, L. Stramaccia, S. Vannucchi, Endogenous heparinase-sensitive anticoagulant activity in human plasma. Thromb Res. 67 , 157–165 (1992).1440519
32 H. Engelberg, Plasma heparin levels in normal man. Circulation 23 , 578–581 (1961).13696820
33 N. Volpi, M. Cusmano, T. Venturelli, Qualitative and quantitative studies of heparin and chondroitin sulfates in normal human plasma. BBA - General Subjects 1243 , 49–58 (1995).7827107
34 E. Kefaloyianni, Soluble forms of cytokine and growth factor receptors: Mechanisms of generation and modes of action in the regulation of local and systemic inflammation. FEBS Lett. 596 , 589–606 (2022).35113454
35 S. Kirschke , The metalloprotease ADAM10 generates soluble interleukin-2 receptor alpha (sCD25) in vivo. J. Biol. Chem. 298 , 101910 (2022).35398356
36 F. Bleibaum , ADAM10 sheddase activation is controlled by cell membrane asymmetry. J. Mol. Cell Biol. 11 , 979–993 (2019).30753537
37 Y. Kuriyama , Coordination of retrotransposons and type I interferon with distinct interferon pathways in dermatomyositis, systemic lupus erythematosus and autoimmune blistering disease. Sci. Rep. 11 , 1–13 (2021).33414495
38 N. Ramadan, M. Kurban, O. Abbas, Possible role for plasmacytoid dendritic cells in pemphigus. Ann. Dermatol. 31 , 454–457 (2019).33911627
39 F. O. Nestle , Plasmacytoid predendritic cells initiate psoriasis through interferon-α production. J. Exp. Med. 202 , 135–143 (2005).15998792
40 C. K. Min , A type I interferon and IL-10 induced by Orientia tsutsugamushi infection suppresses antigen-specific T cells and their memory responses. Front. Immunol. 9 , 2022 (2018).30233599
41 M. Koshy , Antinuclear antibodies in scrub typhus: Transient occurrence during acute illness. J. Vector Borne Dis. 55 , 52 (2018).29916449
42 K. A. Chu, W. Chen, C. Y. Hsu, Y. M. Hung, J. C. C. Wei, Association of scrub typhus with the risk of autoimmune diseases: A population-based cohort study. Am. J. Epidemiol. 188 , 1311–1318 (2019).31140560
43 C. C. S. Münch , Multiple Orientia clusters and Th1-skewed chemokine profile: A cross-sectional study in patients with scrub typhus from Nepal. Int. J. Infect. Dis. 128 , 78–87 (2023).36566774
44 A. Onishi, K. St Ange, J. S. Dordick, R. J. Linhardt, Heparin and anticoagulation. Front. Biosci. Landmark 21 , 1372–1392 (2016).
45 Y. Su , Glycosylation influences the lectin activities of the macrophage mannose receptor. J. Biol. Chem. 280 , 32811–32820 (2005).15983039
46 R. Karlsson , Dissecting structure-function of 3-O-sulfated heparin and engineered heparan sulfates. Sci. Adv. 7 , 1–14 (2021).
47 S. E. Stringer, J. T. Gallagher, Specific binding of the chemokine platelet factor 4 to heparan sulfate. J. Biol. Chem. 272 , 20508–20514 (1997).9252363
48 J. Florentin , HCV glycoprotein E2 is a novel BDCA-2 ligand and acts as an inhibitor of IFN production by plasmacytoid dendritic cells. Blood 120 , 4544–4551 (2012).23053572
49 E. Riboldi , Human C-type lectin domain family 4, member C (CLEC4C/BDCA-2/CD303) is a receptor for asialo-galactosyl-oligosaccharides. J. Biol. Chem. 286 , 35329–35333 (2011).21880719
50 J. W. Kim , Identification of serum glycoprotein ligands for the immunomodulatory receptor blood dendritic cell antigen 2. Glycobiology 28 , 592–600 (2018).29796630
51 E. Martinelli , HIV-1 gp120 inhibits TLR9-mediated activation and IFN-α secretion in plasmacytoid dendritic cells. Proc. Natl. Acad. Sci. U.S.A. 104 , 3396–3401 (2007).17360657
52 W. Zhong , Next generation plasma proteome profiling to monitor health and disease. Nat. Commun. 12 , 1–12 (2021).33397941
53 M. Macal , Self-renewal and toll-like receptor signaling sustain exhausted plasmacytoid dendritic cells during chronic viral infection. Immunity 48 , 730–744.e5 (2018).29669251
54 D. H. Paris , Orientia tsutsugamushi in human scrub typhus eschars shows tropism for dendritic cells and monocytes rather than endothelium. PLoS Negl. Trop. Dis. 6 , e1466 (2012).22253938
55 F. Venegas, P. Yu, A. Kaufmann, S.-A. Freibert, S. Bauer, Confocal imaging and microscale thermophoresis dataset from the article: “A novel type I interferon regulatory network for human plasmacytoid dendritic cells based on heparin, membrane-bound and soluble type II C-type lectin BDCA-2”, Mendeley Data (2023).
