Low-Dose Exposure to Ganglioside-Mimicking Bacteria Tolerizes Human Macrophages to Guillain-Barré Syndrome-Associated Antigens

Abstract

Early in life, commensal bacteria play a major role in immune development, helping to guide the host response toward harmful stimuli while tolerating harmless antigens to prevent autoimmunity. Guillain-Barré syndrome (GBS) is an autoimmune disease caused by errant immune attack of antibody-bound ganglioside receptors on host nerve cells, resulting in paralysis. Lipooligosaccharides enveloping the prevalent enteric pathogen, Campylobacter jejuni, frequently mimic human gangliosides and can trigger GBS by stimulating the autoimmune response. In low- to middle-income countries, young children are consistently exposed to C. jejuni, and it is not known if this impacts GBS susceptibility later in life. Using a macrophage model, we examined the effect of training these cells with low doses of ganglioside-mimicking bacteria prior to challenge with GBS-associated antigens. This training caused decreased production of proinflammatory cytokines, suggesting tolerance induction. We then screened Campylobacter isolates from 154 infant fecal samples for GM1 ganglioside mimicry, finding that 23.4% of strains from both symptomatic and asymptomatic infants displayed GM1-like structures. Training macrophages with one of these asymptomatic carrier isolates also induced tolerance against GBS-associated antigens, supporting that children can be exposed to the tolerizing antigen early in life. RNA interference of Toll-like receptor 2 (TLR2) and TLR4 suggests that these receptors are not involved in tolerance associated with decreases in tumor necrosis factor (TNF), interleukin-6 (IL-6), or IL-1β levels. The results of this study suggest that exposure to ganglioside-mimicking bacteria early in life occurs naturally and impacts host susceptibility to GBS development. IMPORTANCE In this study, we demonstrated that it is possible to tolerize immune cells to potentially dampen the autoreactive proinflammatory immune response against Guillain-Barré syndrome (GBS)-associated antigens. The innate immune response functions to arm the host against bacterial attack, but it can be tricked into recognizing the host's own cells when infectious bacteria display sugar structures that mimic human glycans. It is this errant response that leads to the autoimmunity and paralysis associated with GBS. By presenting immune cells with small amounts of the bacterial glycan mimic, we were able to suppress the proinflammatory immune response upon subsequent high exposure to glycan-mimicking bacteria. This suggests that individuals who have already been exposed to the glycan mimics in small amounts are less sensitive to autoimmune reactions against these glycans, and this could be a factor in determining susceptibility to GBS.

Keywords: Campylobacter jejuni; Guillain-Barré syndrome; gangliosides; lipooligosaccharides.

FIG 1

Training with E. coli GM1 induces tolerance to C. jejuni HS:19 challenge. The release of TNF (A), IL-6 (B), and IL-1β (C) after training macrophages with C. jejuni HS:19 (HS:19) and E. coli strain CWG308 wild type (WT) and from the WT engineered to express the GM1 mimic (GM1) at the indicated multiplicities of infection (MOI) and subsequently challenging with C. jejuni HS:19 at an MOI of 5. Each data point represents one biological replicate. Error bars indicate standard errors of the means. ****, P < 0.0001; **, P ≤ 0.01, determined by one-way ANOVA.

FIG 2

Training with purified lipooligosaccharides (LOS) from E. coli GM1 is sufficient to induce tolerance to C. jejuni HS:19 challenge. (A) Structures of LOS from E. coli strain CWG308 wild type (WT) and from the WT engineered to express the GM1 mimic (GM1) are shown using the Consortium for Functional Glycomics symbol nomenclature (100). Confirmation of GM1-mimic expression was described in Focareta et al. (57) and shown here by binding with immunogold-labeled cholera toxin B subunit followed by transmission electron microscopy. Scale bars are indicated in each panel. (B) Macrophages were trained with commercial lipopolysaccharide (LPS) or purified LOS from C. jejuni HS:19 (HS:19), E. coli GM1 (GM1), or E. coli WT (WT) before challenging with C. jejuni HS:19 at a multiplicity of infection of 5. Each data point represents one biological replicate. Error bars represent the standard errors of the means. **, P ≤ 0.01 determined by one-way ANOVA.

FIG 3

Neither TLR2 nor TLR4 is responsible for lipooligosaccharide-induced GM1-ganglioside tolerance. Confirmation of TLR2 knockdown (shTLR2) (A) and TLR4 knockdown (shTLR4) (B) macrophages using known agonists for the TLRs including Pam3CSK4 (for TLR2) or commercial lipopolysaccharide (LPS; for TLR4) compared to short hairpin loop RNA control (shScram) macrophages. The release of TNF (C), IL-6 (D), and IL-1β (E) after training shScram, shTLR2, or shTLR4 macrophages with C. jejuni HS:19 (HS:19), E. coli GM1 (GM1), or E. coli WT (WT) at a multiplicity of infection (MOI) of 0.01 and subsequently challenging with C. jejuni HS:19 at an MOI of 5. Each data point represents one biological replicate. Error bars represent standard errors of the means. ****, P < 0.0001; ***, P ≤ 0.001; **, P ≤ 0.01; *, P ≤ 0.05, determined by the Student's t test.

FIG 4

Infants in low- and middle-income countries are exposed to GM1 ganglioside-mimicking C. jejuni. (A) A pie chart of infant Campylobacter isolates separated by display of symptoms (asymptomatic are blue, symptomatic are red) and highlighting the number of GM1 ganglioside mimic-positive samples within each group. (B) Example silver stain of lipooligosaccharides from Campylobacter isolates and (C) the corresponding Western blot using cholera toxin B subunit as a probe. C. jejuni HS:19 was used as a positive control. Unedited silver stain and Western blot images can be found in Fig. S3.

FIG 5

Training with C. jejuni strain 703871 also induces tolerance to C. jejuni HS:19 challenge, which is not mediated by IL-10 release. (A) Fluorescence microscopy done on a GM1-positive infant isolate (703871) using a CTB probe with C. jejuni HS:19 as a positive control and C. jejuni HS:3 as a negative control along with graphs depicting the release of TNF (B) or IL-10 (C) after training macrophages with C. jejuni HS:19 (HS:19), E. coli GM1 (GM1), and E. coli WT (WT) at their respective optimal tolerizing multiplicities of infection (MOI; 0.1 or 0.01) and subsequently challenged with C. jejuni HS:19 at an MOI of 5. Each data point represents one biological replicate. Error bars represent the standard errors of the means. **, P ≤ 0.01; *, P ≤ 0.05 determined by one-way ANOVA test.