Molecular mimicry of host sialylated glycans allows a bacterial pathogen to engage neutrophil Siglec-9 and dampen the innate immune response

Although antimicrobial properties of vertebrate innate immune cells are extensively studied, less is understood about mechanisms dampening inflammatory responses. Such negative regulatory systems can be subverted by microbes. Of relevant interest is the “molecular mimicry” of mammalian sialic acid (Sia)–terminated sialoglycans by microbes that are obligate commensals or potential pathogens of humans.¹  Surface Sia expression can blunt alternative pathway complement activation and reduce immunogenicity.¹  However, this may not fully explain convergent bacterial evolution of near-perfect mimicry of vertebrate sialoglycans. For example, the human-specific commensal/pathogen group B Streptococcus (GBS) has a capsular polysaccharide (CPS) that displays the structure Siaα2-3Galβ1-4GlcNAc,²  a sequence identical to one common at termini of human glycoproteins.

Sia-recognizing immunoglobulin superfamily lectins (Siglecs) are type I transmembrane proteins expressed on immune cells.^3,4  The rapidly evolving subgroup of CD33-related Siglecs (CD33rSiglecs) are postulated (but not proven) to negatively regulate inflammatory responses by recognizing host sialoglycans.³  Many CD33rSiglecs have conserved cytoplasmic tyrosine-based motifs, comprising a membrane-proximal immunoreceptor tyrosine-based inhibitory motif (ITIM) and a membrane-distal ITIM-like motif.^3,4  The wide expression of host Sias and the prominence of cognate ITIM-bearing CD33rSiglecs on immune cells suggest that they may function in “self”-recognition, dampening innate immune responses to prevent autoreactivity.³ 

The functional outcome of CD33rSiglec binding to sialylated ligands remains poorly understood. Cross-linking antibodies and/or Siglec transfection into Siglec-deficient cell lines has demonstrated the importance of the ITIM and ITIM-like motifs for inhibiting cellular activation and proliferation,^5-12  and even inducing apoptosis.^13,14  However, interpretation of such data is limited by the use of nonnative transfected cells and/or anti-Siglec antibodies that are unnatural ligands.

Although CD33r Siglecs recognize Sias on the same cell surface (so-called cis interactions),¹⁵  high densities of Sias on adjacent cell surfaces or multimerized polyvalent probes,¹⁶  heavily sialylated glycoproteins,^17-19  or bacterial CPSs²⁰  can engage Siglecs in trans, presumably outcompeting the cis-Sia ligands at contact sites.

Here we show that sialoglycan ligation in trans can transmit negative regulatory signals through Siglec-9, which is prominent on human neutrophils.^3,4  The functional outcome is innate immune subversion by GBS, via molecular mimicry of host sialoglycans.

GBS COH1, a highly encapsulated serotype III strain, was propagated in Todd-Hewitt broth at 37°C to logarithmic growth phase.

Normal human volunteers donated small blood samples for the isolation of neutrophils, with informed consent obtained in accordance with the Declaration of Helsinki, under protocols approved by the University of California, San Diego Human Subjects Institutional Review Board. Isolation was performed using Polymorphprep (Axis-Shield, Oslo, Norway). For prelabeling, neutrophils were resuspended in 500 μL of Hanks balanced salt solution (HBSS), Ca⁻, Mg^−+/−calcein-AM (Invitrogen, Carlsbad, CA) for 30 minutes at 37°C, washed 3 times, and resuspended in 1.5 mL HBSS.

Natural or synthetic glycans in carbonate buffer, pH 9.4, were added (100 μL) to 96-well Immulon 4HBX plates (Thermo Electron, Waltham, MA), incubated overnight at 4°C, washed 3 times, and blocked at room temperature with phosphate-buffered saline (PBS) plus 3% bovine serum albumin (BSA) for 1 hour. Labeled neutrophils were added, and initial fluorescence intensity (FI) was measured. After 30 minutes, nonadherent neutrophils were removed with PBS using a 12-well multichannel pipetter, and final FI was measured.

Neutrophils were incubated for 5 minutes in the presence or absence of the NBSAb (murine IgG1 clone E10-286; BD Biosciences Pharmingen, San Diego, CA) or BSAb (murine IgG2a clone 191240; R&D Systems, Minneapolis, MN) at 1:200 dilution and washed with HBSS plus calcium and magnesium plus 3% BSA before use in various assays.

Neutrophils showed specific binding to polyacrylamide arrays bearing multiple copies of Siaα2-3Galβ1-4GlcNAcβ1, even without prior removal of competing cell surface Sias (Figure 1A). Siglec-9 is the dominant CD33rSiglec on neutrophils that recognizes α2-3–linked Sias and could be involved in binding. To study this, we validated anti–Siglec-9 IgG mouse monoclonal antibodies reacting with the Sia-binding site (BSAb) or not reacting with the Sia-binding site (NBSAb) as functional reagents, using competition assays with Siglec-9-Fc chimeric protein. The BSAb, but not the NBSAb, inhibited Siglec-9-Fc binding to Siaα2-3Galβ1-4GlcNAcβ1–coated wells (Figure 1B). The 2 antibodies are thus validated as functional reagents to dissect differences between blocking or not blocking the Sia-binding site. The BSAb also selectively reduced neutrophil interactions with Sia-containing glycan probes (Figure 1C), confirming that this CD33rSiglec can mediate adhesive interactions with sialoglycans in trans, without removal of competing cis Sias. The high density of sialoglycans and/or their arrangement allows them to defeat competition by cell surface sialoglycans.

The CPS of serotype III GBS presents terminal Siaα2-3Galβ1-4GlcNAc-glycans, a sequence similar to the synthetic glycan–coated wells and to native human cell surface glycans, which can interact with recombinant Siglec-9-Fc.²⁰  To determine whether GBS serotype III CPS can engage human neutrophils via such molecular mimicry, we deposited purified CPS in microtiter wells and treated a subset with sialidase. Neutrophils adhered to native, but not desialylated, CPS (data not shown). Binding was diminished by the BSAb and not by NBSAb (Figure 1D). As with the immobilized synthetic sialoglycans, the dense Siaα2-3Galβ1-4GlcNAc presentation on the GBS serotype III CPS engaged Siglec-9 on neutrophils in trans. Surface extracts of type III GBS that include the GBS CPS in a more native context were isolated by mutanolysin cleavage of peptidoglycan backbones. Again, neutrophil binding to these extracts was Sia-dependent (not shown) and selectively blocked by the BSAb (Figure 1E), confirming Siglec-9 dependency.

Neutrophils are specialized granulocytes that recognize and directly kill microorganisms. Because they interact with the sialylated serotype III GBS CPS in a Siglec-9–dependent manner, we tested the capacity of intact live GBS to alter functional responses. Neutrophils activated by bacterial recognition produce reactive oxidative species and release granule proteins and chromatin that form neutrophil extracellular traps (NETs), which ensnare and kill bacteria.²¹  Neutrophils interacting with live GBS in the presence of the BSAb produced stronger oxidative bursts and released more granule protease, compared with those in NBSAb (Figure 2A,B). Likewise, neutrophils stimulated by GBS generated more NETs in the presence of the BSAb compared with NBSAb (Figure 2C,D). Thus, Siglec-9 engagement by GBS CPS sialoglycans blunts key aspects of neutrophil activation in response to bacterial recognition.

Besides releasing bactericidal factors, neutrophils also synthesize cytokines that amplify or dampen immune responses. Others showed that transfection of Siglec-9 into macrophage cell lines increases production of the anti-inflammatory cytokine interleukin-10 (IL-10).²²  We found that neutrophils that cannot engage sialylated GBS via Siglec-9 resulting from the BSAb produced less IL-10 mRNA (Figure 2E). Thus, bacterial sialoglycans signal through neutrophil Siglec-9 to increase expression of an anti-inflammatory cytokine.

Thus, a novel functional consequence of GBS sialoglycan molecular mimicry is interaction with neutrophil Siglec-9, blocking these cells from responding normally. This could contribute to GBS pathogenicity by promoting evasion of neutrophil-based killing. Indeed, neutrophils kill more GBS on coincubation with BSAb compared with NBSAb (Figure 2F). The blocking effect of the BSAb is dependent on bacterial cell surface sialoglycan expression, as it does not occur with either Sia-negative GBS or another nonsialylated bacterium, group A Streptococcus (Figure 2G).

Taken together, the data demonstrate, for the first time, that microbial evasion of neutrophil responses can occur via engagement of a host inhibitory receptor by precise molecular mimicry of a host sialoglycan. This property of GBS may assist it during invasive blood-borne infections in susceptible persons, such as newborn infants, pregnant women, and the elderly.

This work was supported by National Institutes of Health grants P01HL57345 (A.V.) and HD051796 (V.N.).

National Institutes of Health

Contribution: A.F.C., S.U., and Y.-C.C. performed the research, prepared figures and legends, and read the paper; and A.F.C., A.L.L., V.N., and A.V. designed the research, analyzed the data, and wrote the paper.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Ajit Varki, 9500 Gilman Dr, University of California, San Diego, La Jolla, CA 92093-0687; e-mail: a1varki@ucsd.edu.