Sialylation of <scp>IgE</scp> reduces <scp>FcεRIα</scp> interaction and mast cell and basophil activation in vitro and increases <scp>IgE</scp> half‐life in vivo

Abstract

Immunoglobulin E (IgE) plays a crucial role in allergic reactions, including systemic anaphylaxis, by binding to the high-affinity IgE receptor FcεRI (FcεRIα + FcεRIβ + FcεRIγ) on mast cells and basophils and, upon allergen-mediated aggregation, inducing the release of inflammatory mediators.1-3 The allergic potential of IgE antibodies (Abs) is further affected by IgE glycosylation-dependent interactions with membrane-bound and soluble carbohydrate-binding proteins (e.g., galectins, siglecs, and C-type lectin receptors) as well as different structural features (E1-4).1, 3 Many healthy individuals harbor allergen-specific IgE Abs in the absence of an allergic disease (E5, E6), suggesting qualitative differences in IgE-triggered responses. Similarly, not all allergic individuals develop severe disease forms (e.g., asthma), and it is still unclear why allergen immunotherapy (AIT) is effective in some individuals but not in all (E6). Different natural or AIT-induced inflammatory states of the specific T- and B-cell response may lead, as described for IgG (E7),4 to differently glycosylated IgE Abs that contribute distinctly to the induction of allergic responses. IgE Abs have several conserved N-glycosylation sites in the constant region of their heavy chains. Most of them are of the complex type and potentially galactosylated and terminal sialylated2, 5 (Figure 1A,B). By comparing the effects of sialylated versus enzymatically de-sialylated IgE monoclonal (m)Abs, a recent study suggested that highly sialylated IgE Abs have a stronger potential to induce allergic reactions than lower sialylated IgE Abs.6 Notably, this observation is opposite to the findings that sialylation dampens the inflammatory potential of IgG, IgA, and IgM Abs (E8-E10).4, 7 Here, we further investigated the role of IgE sialylation on FcεRIα interaction, mast cell and basophil loading and activation, and IgE-mediated passive systemic anaphylaxis (PSA) in mice. Therefore, we generated an anti-TNP (2,4,6-trinitrophenol) murine IgE mAb with three different sialylation levels: (i) originally sialylated (originally sial; glycosylation form generated by the hybridoma cell line), (ii) enzymatically de-sialylated (de-sial), and (iii) enzymatically additional sialylated (high sial) (Figure 1B–D). Both enzymatic IgE glycan modifications did not influence antigen binding (Figure 1E). De-sialylated anti-TNP IgE mAbs showed a stronger interaction with human and murine FcεRIα than originally sialylated and additionally sialylated IgE mAbs, whereas the latter two sialylated IgE glycoforms showed a comparable low or no interaction with FcεRIα in these assays, respectively (Figure 1F). This observation was unexpected due to the mentioned findings about the lower allergic potential of de-sialylated compared to sialylated IgE Abs6 but is in line with a stronger interaction of de-sialylated IgG Abs with Fcγ receptors than sialylated IgG Abs (E8). A comparable trend was observed when human mast cells differentiated from CD34+ blood cells were activated with the three anti-TNP IgE glycoforms and TNP-OVA as antigen (Figure S1). Moreover, less sialylated anti-TNP IgE mAbs showed an enhanced loading on the human mast cell line LAD2 as well as on human blood basophils, which correlated positively with their activation (Figure 1G–J and Figure S1). Low sensitivity of the FcεRIα interaction assay and of the differentiated human mast cell assay and/or additional factors to FcεRIα binding might explain the higher loading and activation potentials of originally sialylated (orig. sial) compared to additional sialylated (high sial) IgE Abs in the LAD2 mast cell and blood basophil assays. Together, these data suggest that the higher interaction of de−/less sialylated IgE with mast cells and basophils correlate with their stronger activation potential. Next, we compared the potential of the three anti-TNP IgE mAb glycoforms in an IgE-mediated passive systemic anaphylaxis (PSA) mouse model (Figure 2A and Figure S1). In accordance with the data described by Shade et al., 20206 but in contrast to the expectation based on our in vitro data, intravenous (i.v.) injection of de-sialylated anti-TNP IgE mAbs, followed by i.v. injection of TNP-coupled ovalbumin (TNP-OVA) as “allergen” 6 or 24 h later, hardly induced any anaphylaxis, as measured by the drop in body core/rectal temperature, compared to the sialylated IgE mAbs (Figure 2B and Figure S1). However, we realized that de-sialylated anti-TNP IgE mAbs could neither be detected in serum nor on blood basophils already 6 h after injection (Figure 2C,D and Figure S1). These data suggested a highly reduced serum half-life of de-sialylated IgE mAbs, which has also been described for de-sialylated IgA Abs in an asialo-glycoprotein receptor-dependent manner (E11). To counteract the potential quick clearance of de-sialylated IgE mAbs, we blocked asialo-glycoprotein receptors with an excess of de-sialylated (asialo)-α1-acid glycoprotein (AGP; orosomucoid) (E11) before sensitizing mice with the anti-TNP IgE glycoforms (Figure 2E and Figure S1). As anticipated, de-sialylated IgE mAbs remained in the circulation and showed serum levels similar to additionally sialylated IgE mAbs 6 h after injection (Figure 2F). Accordingly, anti-TNP IgE-induced PSA could also be observed with de-sialylated IgE mAbs (Figure 2G). These data showed that it is difficult to investigate the potential of de-sialylated IgE Abs in serum half-life-subjected mouse models without blocking asialo-glycoprotein receptors. However, unexpectedly from our in vitro human and mouse FcεRIα interaction and human cell loading and activation data, we observed no differences in anti-TNP IgE-blood basophil-loading and IgE-induced PSA in vivo in mice, neither between originally sialylated (orig. sial) and additionally sialylated (high sial) IgE mAbs without blocking asialo-glycoprotein receptors nor between de-sialylated (de-sial) and high sial IgE mAbs after blocking asialo-glycoprotein receptors (Figure 2B,D,G,H and Figure S1). Model-dependent saturation with IgE mAbs or any stabilizing effects in the steady state status of the mice may explain comparable anti-TNP IgE loading on and activation of effector cells in our in vivo experiments. Alternative models that consider more strongly the differential interaction of distinctly glycosylated IgE (m)Abs with target cells and/or different genetic and environmental factors that influence the inflammatory immune status and therefore the expression of potentially important carbohydrate-binding proteins (E1-E3)1, 3 might elucidate the different allergic potentials of distinctly sialylated IgE Abs. Moreover, further studies need to investigate whether certain carbohydrate-binding proteins can stabilize the half-life of low sialylated IgE Abs in vivo, and which site-specific IgE sialylation/de-sialylation profiles affect IgE binding to FcεRIα and carbohydrate-binding proteins and therewith cell loading and activation. Together, our in vitro data suggest that less sialylated IgE Abs have a higher binding and correlating activation potential on mast cells and basophils than higher sialylated IgE Abs. In humans, different natural or AIT-mediated inflammatory conditions might induce distinctly sialylated IgE Abs. Less sialylated IgE Abs might have a higher allergic potential under certain circumstances than higher sialylated IgE Abs. This might explain why some individuals with specific IgE Abs develop allergies and others do not, why some patients develop mild and others severe disease conditions, or why AIT works in some but not in all individuals (E6). Immune conditions and cell phenotypes of various individuals/patients/cohorts should be determined to further analyze the allergic potential of differently glycosylated IgE Abs. LD, JP, G-ML, YCB, MK, CP, SL, JSB, HBL, and JR involved in IgE generation, ELISA, LAD2 assays and PSA. LD, JP, CK, LK, and VB involved in IgE analysis. LG, CS, and SD involved in assay with human mast cells differentiated from CD34+ blood cells. JB, TW, and UJ involved in human blood basophil assays. LD, JP, JB, TW, JR, and ME involved in statistical analysis. JP, JB, VB, SD, UJ, JR, and ME involved in supervision. JP, LD, JR, and ME involved in writing—original draft. All authors approved the final version. Open Access funding enabled and organized by Projekt DEAL. This study was supported by the Volkswagen-Foundation (97301) (ME), the German Center for Lung Research (DZL) (UJ, ME), and the Deutsche Forschungsgemeinschaft ((DFG, German Research Foundation) 398859914 (EH 221/10–1); 400912066 (EH 221/11–1); 429175970 (RTG 2633); and 390884018 (Germany's Excellence Strategies - EXC 2167, Precision Medicine in Chronic Inflammation (PMI)) (ME). JSB was a PhD student of the RTG 2633. All authors declare to have no conflict of interest. Figure S1. Appendix S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts