Sialic acid-binding immunoglobulin-like lectin (Siglec) 8 in patients with eosinophilic disorders: Receptor expression and targeting using chimeric antibodies

Abstract

Background: Sialic acid-binding immunoglobulin-like lectin (Siglec) 8 is selectively expressed on eosinophils, mast cells, and basophils and, when engaged on eosinophils, can cause cell death.

Objective: We sought to characterize surface and soluble Siglec-8 (sSiglec-8) levels in normal donors (NDs) and eosinophilic donors (EOs) and assess the efficacy of anti-Siglec-8 antibodies in inducing eosinophil cell death in vitro.

Methods: Eosinophil expression of Siglec-8 was assessed by using flow cytometry and quantitative PCR. Serum sSiglec-8 levels were measured by means of ELISA. Induction of eosinophil death by IgG₄ (chimeric 2E2 IgG₄) and afucosylated IgG₁ (chimeric 2E2 IgG₁ [c2E2 IgG₁]) anti-Siglec-8 antibodies was evaluated in vitro by using flow cytometry and in vivo in humanized mice.

Results: Siglec-8 was consistently expressed on eosinophils from NDs and EOs and did not correlate with absolute eosinophil count or disease activity. sSiglec-8 levels were measurable in sera from most donors unrelated to absolute eosinophil counts or Siglec-8 surface expression. c2E2 IgG₁ and chimeric 2E2 IgG₄ were equally effective at inducing cell death (Annexin-V positivity) of purified eosinophils from NDs and EOs after overnight IL-5 priming. In contrast, killing of purified eosinophils without IL-5 was only seen in EOs, and natural killer cell-mediated eosinophil killing was seen only with c2E2 IgG₁. Finally, treatment of humanized mice with anti-Siglec antibody led to robust depletion of IL-5-induced eosinophilia in vivo.

Conclusions: Siglec-8 is highly expressed on blood eosinophils from EOs and NDs and represents a potential therapeutic target for eosinophilic disorders. Enhanced killing of eosinophils in the presence of IL-5 might lead to increased efficacy in patients with IL-5-driven eosinophilia.

Keywords: Siglec-8; apoptosis; eosinophil; eosinophilic gastrointestinal disease; hypereosinophilic syndrome; inhibitory receptor; mast cell; mastocytosis; monoclonal antibody; soluble receptor.

FIG 1.

Eosinophil count and surface expression of Siglec-8. A, Absolute eosinophil count (AEC) and B, eosinophil surface expression of Siglec-8 expressed as mean fluorescence intensity (mfi) of normal (ND; white circles) and eosinophilic subjects (EO) on treatment (blue circles) or untreated (red circles). EO subjects are classified as Active (AEC≥1500/μL or marked tissue eosinophilia and AEC>500/μL), Responded (clinically improved with AEC≤1000/μL on therapy) or Resolved (clinically improved with AEC≤1000/μL on no therapy). The horizontal bars represent geometric means with 95% CI. Gray shading indicates AEC<500/μL. C, Changes in AEC and Siglec-8 mfi pre-and post-treatment in EO subjects. *P<0.05, Wilcoxon matched pairs, signed rank test, D, Correlation between AEC and Siglec-8 mfi on eosinophils from ND (white circles) and EO subjects on treatment (blue circles) or untreated (red circles). Spearman correlation P=NS.

FIG 2.

SIGLEC-8 mRNA expression in blood eosinophils. A, Eosinophil levels of Siglec-8 mRNA in ND (white circles), and EO, untreated (blue circles) and on treatment (red circles). Geometric means with 95% CI are indicated by horizontal lines. B, Correlation between AEC and SIGLEC-8 mRNA. The vertical dotted line indicates AEC=500/μL. C, Correlation between mRNA and eosinophil surface expression of Siglec-8 (Δmfi); P<0.01, Spearman correlation.

FIG 3.

Serum soluble Siglec-8 levels from various donors. A, AEC and B, soluble serum levels of Siglec-8 in ND (white circles), untreated EO (blue circles), EO on treatment (red circles), systemic mastocytosis (SM) patients (brown circles) and SM-Eo (orange circles). Geometric means with 95% CI are indicated by horizontal lines. Changes in AEC (C), and changes in serum sSiglec-8 (D) at two time points (T1 and T2) in ND (white circles) and EO subjects (blue circles) without a change in treatment and pre-and post-treatment in F/P⁺ and F/P^neg EO subjects treated for 1–3 months with imatinib 400 mg daily (F/P⁺; n=7), prednisone 5–35 mg daily (F/P^neg; n=5) or hydroxyurea 500 mg daily (F/P^neg; n=1) (red circles) (Wilcoxon matched pairs, signed rank test, ** P<0.01 and *P<0.05). E, Correlation between soluble Siglec-8 levels and surface expression of Siglec-8. F, Correlation between soluble Siglec-8 levels SIGLEC-8 mRNA levels. Gray shading in panels A and E indicate AEC<500/μL. Dotted horizontal lines in panels B-D and F-H represent the upper and lower limits of detection for soluble Siglec-8.

FIG 4.

Soluble Siglec-8 levels in subjects with eosinophilic esophagitis (EoE). A, Serum levels of sSiglec8 in Control (white circles), GERD (blue circles), Inactive and Active EoE (red circles). P=ns, one-way ANOVA with post hoc Bonferroni, B, Serum sSiglec8 levels in paired samples from EoE subjects during active and inactive disease. P=ns, Wilcoxon matched pairs, signed rank test. C-F, Correlation between serum Siglec-8 levels and peak intra-epithelial eosinophils per high powered field (eos/hpf) in esophageal biopsies (C), AEC (D), intraepithelial mast cells/hpf (E) and degranulated mast cells/hpf (F) in esophageal biopsies. Significance was assessed by Spearman correlation. Dotted horizontal lines in panels A and B represent the upper and lower limits of detection for soluble Siglec-8.

FIG 5.

Eosinophil apoptosis induced by c2E2 IgG1 and c2E2 IgG4 antibodies to Siglec-8. A, Annexin-V+ eosinophils from ND (white circles; n=8) and EO (red circles; n=29) after overnight incubation in the presence of c2E2 IgG1 or c2E2 IgG4 (10 μg/mL) following pre-incubation with IL-5 (30 ng/mL for 18h). B and C, Dose-dependent effect of anti-Siglec-8 antibodies (10⁻³ to 10 μg/mL) on eosinophil cell death in presence of IL-5. Mean±SEM % Annexin-V⁺ or % 7-AAD⁺ (n=36) in the presence of c2E2 IgG4 or c2E2 IgG1 were compared to their respective isotype controls. Wilcoxon matched pairs, signed rank test (* P<0.05, ** P<0.01, *** P<0.001; ****, P<0.0001).

FIG 6.

Eosinophil killing in the presence of NK cells and c2E2 IgG1 and c2E2 IgG4. A and B, In vitro NK killing assay performed on eosinophils from ND (white circles; n=8) and EO (red circles, n=14) after 4 hrs incubation in presence of anti-Siglec-8 antibodies (10 μg/mL). Annexin-V (A) and 7-AAD (B) staining is represented for each subject. (*, P<0.05, **, P<0.01; ***, P<0.001, Wilcoxon test).

FIG 7.

Eosinophil depletion from peripheral blood leucocyte preparations induced by c2E2 IgG4 in vitro and in vivo. A,B, Dose-dependent effect of c2E2 IgG4 (10⁻⁵ to 10μg/mL) on eosinophil depletion after overnight incubation in the presence (closed circles) and absence of IL-5 (open circles) in ND (A) and EO (B). Data are expressed as the % of eosinophils remaining normalized to the % of eosinophils in culture medium alone (n=5 per group, graphs show mean+/−SEM). Eosinophil depletion in the presence of c2E2 IgG4 (blue) was compared to the respective isotype control (orange). *P<0.05, **P<0.001. C, Percentage of blood eosinophils detected by flow cytometry in engrafted humanized NSGS mice prior to and following administration of hIL-5 (***, P<0.001, Mann Whitney test). D, Percentage of human (hEPX⁺) eosinophils 18 hours post-treatment with c2E2 IgG1, m2E2 IgG1 (equivalent of human IgG4) or isotype control antibody (mIgG1). (*, P<0.05, Wilcoxon test, n=5 mice). Engraftment of mice with CD34⁺ cells from donor 1 are indicated by white circles and from donor 2 by green circles.