Human Siglec-10 can bind to vascular adhesion protein-1 and serves as its substrate

Abstract

Leukocytes migrate from the blood into areas of inflammation by interacting with various adhesion molecules on endothelial cells. Vascular adhesion protein-1 (VAP-1) is a glycoprotein expressed on inflamed endothelium where it plays a dual role: it is both an enzyme that oxidizes primary amines and an adhesin that is involved in leukocyte trafficking to sites of inflammation. Although VAP-1 was identified more than 15 years ago, the counterreceptor(s) for VAP-1 on leukocytes has remained unknown. Here we have identified Siglec-10 as a leukocyte ligand for VAP-1 using phage display screenings. The binding between Siglec-10 and VAP-1 was verified by different adhesion assays, and this interaction was also consistent with molecular modeling. Moreover, the interaction between Siglec-10 and VAP-1 led to increased hydrogen peroxide production, indicating that Siglec-10 serves as a substrate for VAP-1. Thus, the Siglec-10-VAP-1 interaction seems to mediate lymphocyte adhesion to endothelium and has the potential to modify the inflammatory microenvironment via the enzymatic end products.

Figure 1

Siglec-10 interacts with VAP-1. (A) The amino acid sequence obtained from randomly picked clones after 4 rounds of selection and its match to Siglec-10 and the binding of the corresponding synthetic peptide to recombinant VAP-1 (100 ng/well) in ELISA. The results are mean ± SEM from 3 separate experiments and triplicate wells in each experiment. (B) Binding of recombinant VAP-1 to Siglec-10–Ig chimera. Siglec-10–Ig chimera was immobilized onto ELISA microtiter wells via an anti-hIgG antibody. hIg was used as a negative control. The results are presented as relative binding ratios and are mean ± SEM from 3 separate experiments, each having triplicate wells. (C) Binding of recombinant VAP-1 to CHO cells expressing Siglec-10 and to mock-transfected control cells as detected by anti–VAP-1 antibody. Negative controls are incubations with antibodies without the addition of recombinant VAP-1. The results are mean ± SEM of mean fluorescence intensities measured by flow cytometer from 2 separate experiments. (D) Binding of CFSE-labeled Siglec-10 transfectants to CHO cells expressing VAP-1 and mock controls. The results are mean ± SEM of fluorescent intensities measured by fluorometer from 7 separate experiments, each having duplicate wells. (E) The contribution of sialic acids to the interaction was tested by treating either the cells expressing VAP-1 and or the cells expressing Siglec-10 with neuraminidase to remove the sialic acids. The results are means of fluorescent intensities ± SEM from 4 separate experiments, each having duplicate wells. *P < .05. **P < .01. ***P < .001.

Figure 2

Siglec-10⁺ lymphocytes bind to vessels using VAP-1. (A) Expression of human VAP-1 on mesenteric lymph node vasculature of KOTG mice (left) detected by FITC-Jg-2.10 antibody. A high endothelial venule is pointed out by a white arrow. Lack of expression shown in KO mice (right). Staining with a negative control antibody is shown in the inset. Scale bar represents 50 μm. (B) Purity of the B cells and their Siglec-10 expression used for ex vivo binding assays. Fluorescence-activated cell sorter histograms of CD19 and Siglec-10 expression are shown. Negative control (neg co) antibody was polyclonal anti–P-selectin antibody. (C) Ex vivo frozen section binding assays were used to analyze lymphocyte binding to vessels in mesenteric lymph nodes obtained from VAP-1 KO and VAP-1 KOTG mice. The function of Siglec-10 was blocked by incubating the cells before the assay with anti–Siglec-10 antibody. The results are shown as percentage of control binding (number of KOTG vessel-bound lymphocytes incubated with a nonblocking control mAb is defined as 100%; mean ± SEM). ***P < .001.

Figure 3

Siglec-10 binds to the enzymatic groove of VAP-1 and acts as a substrate. (A) Examples of competitive stainings with Siglec-10–Ig chimera and anti–VAP-1 antibody Jg-2.10. CHO-VAP-1 transfectants were stained with the anti–VAP-1 antibody either in the presence of Siglec-10–Ig chimera or CD44-Ig (negative control) and analyzed by fluorescence-activated cell sorter. Percentages of cells positively stained with anti–VAP-1 mAb or a negative class-matched control antibody are shown. (B) Heart sections of KOTG mice were first incubated either with a control (CD44-Ig) or Siglec-10–Ig chimera and stained thereafter with anti–VAP-1 mAb (Jg-2.10). Some brightly positive vessels in the control section and fewer and less bright ones in the section pretreated with Siglec-10–Ig chimera are pointed out by arrows. Scale bar represents 50 μm. (C) Enzymatic activity of VAP-1 was measured as H₂O₂ (pmol) produced in 1 hour in the presence of Siglec-10 transfectants or mock-transfected control cells. The results are presented as relative SSAO activity ± SEM from 3 separate experiments, each having duplicate wells. A representative experiment showing the amount of H₂O₂ produced at different time points during a 1-hour measurement is presented in the upper left corner. (D) Binding of Siglec-10 transfectants to CHO cells expressing enzymatically active VAP-1, enzymatically inactive VAP-1, or mock controls. The results are mean fluorescence intensities ± SEM from 7 separate experiments, each having duplicate wells. (E) Binding of Siglec-10 peptide (CATLSWVLQNRVLSSCK-biotin) and mutated Siglec-10 peptide (CATLSWVLQNAVLSSCK-biotin) to recombinant VAP-1 (400 ng/well). The results are mean of relative binding ± SEM from 3 separate experiments with triplicate wells *P < .05. **P < .01. ***P < .001.

Figure 4

Interaction of VAP-1 with Siglec-10 C2-type domain 2. (A) The 3-dimensional structure of VAP-1 (monomer A, yellow; and monomer B, green) with a peptide derived from Siglec-10 CE loop (purple ball-and-stick) docked into the active site of monomer B. Arg293 in the docked peptide is labeled and covalently bound to topaquinone (TPQ), which is in an active conformation. Arm1 and the RGD site in domain A are close to the opening of the active site in VAP-1. At the right, there is a 3-dimensional model of the second Siglec-10 C2 domain shown in an orientation fitting into VAP-1. The amino acids corresponding to the docked ligand are shown as purple ball-and-stick in the Siglec-10 model, and the topaquinone-binding arginine is colored yellow and labeled. (B) Stereo view of Siglec-10 (purple) binding to topaquinone (green) in VAP-1. Grid maps for the amide (blue wires) and carboxylate (red wires) probes are shown at a level of −6 kcal/mol. The asparagine (Asn) in Siglec-10 that corresponds to aspartate in Siglec-11 is labeled. (C) Binding of CHO cells expressing Siglec-10 to CHO cells expressing wild-type VAP-1, the RGD mutant of VAP-1, or mock controls was determined with cell-cell adhesion assay as explained in “Assays with transfectants.” The results are presented as a relative binding ratio ± SEM from 5 separate experiments, each having duplicate wells. *P < .05. (D) Binding of CHO cells expressing Siglec-11 to VAP-1 transfectants. (E) Binding of CHO cells expressing Siglec-G to VAP-1 transfectants. The results of panels D and E are relative binding ratio ± SEM from 2 separate experiments, each having quadruple wells.