Schistosoma mansoni‐specific immune responses and allergy in Uganda

Summary Low allergy‐related disease (ARD) prevalence in low‐income countries may be partly attributed to helminth infections. In the Schistosoma mansoni (Sm)‐endemic Lake Victoria islands (Uganda), we recently observed positive helminth‐allergy associations, despite low ARD prevalence. To understand how Sm‐induced cytokine and antibody profiles might influence allergic response profiles in this population, we assessed Schistosoma worm (SWA)‐ and egg antigen (SEA)‐specific Th1 (IFN‐γ), Th2 (IL‐5, IL‐13) and regulatory (IL‐10) cytokine profiles (n = 407), and total (n = 471), SWA‐, SEA‐ and allergen (house dust mite [HDM] and cockroach)‐specific (as)IgE and IgG4 profiles (n = 2117) by ELISA. Wheeze was inversely associated with SWA‐specific IFN‐γ (P < .001) and IL‐10 (P = .058), and SEA‐specific IL‐5 (P = .004). Conversely, having a detectable asIgE response was positively associated with SWA‐specific IL‐5 (P = .006) and IL‐10 (P < .001). Total, SWA‐, SEA‐ and allergen‐specific IgE and IgG4 responses were higher among Sm Kato‐Katz positive (SmKK+) and skin prick test (SPT)+ individuals compared to SmKK‐ and SPT‐ individuals. However, total and asIgG4/IgE ratios were lower among SPT+ and wheezing individuals. We conclude that, in this population, helminth‐induced antibody and cytokine responses may underlie individual positive helminth‐atopy associations, while the overall IgG4‐IgE balance may contribute to the low overall prevalence of clinical allergies in such settings.

IL-5 and IL-13). 10,11 Helminths, unlike allergens, further induce strong immunoregulation epitomized by IL-10 production. 12 Typically, these cytokines influence the profile of antibodies involved in helminth infection and allergy. Helminth-induced IL-10 may drive immunoglobulin class switching to IgG4 13,14 which, akin to the Th2 cytokine-induced 15 polyclonally stimulated IgE, may inhibit development of allergenspecific effector responses, 5,16 leading to inverse helminth-allergy associations. Conversely, helminth-induced protein-specific IgE may promote strong, cross-reactive helminth-and allergen-specific responses, resulting in positive helminth-allergy associations. 17,18 Emerging epidemiological data on helminth-allergy associations in Uganda reflect the complex interaction between helminths and allergens: while observational analyses in a birth cohort suggested a protective effect of childhood and maternal helminths against childhood eczema 19 that was reversed by maternal anthelminthic treatment, 20 we recently reported positive helminth-allergy associations in a survey conducted in the Schistosoma mansoni (Sm)-endemic Lake Victoria islands, albeit against a backdrop of low ARD prevalence. 21 To establish how Sm-induced cytokine and antibody profiles underpinned helminth-allergy associations in the above survey, we here describe an assessment of Sm-specific cytokine profiles, as well as total, allergenand Sm-specific IgE and IgG4 profiles, and their relationship with Sm infection status, wheeze and atopy.

| Study population
Samples were collected during the baseline household survey preceding a cluster-randomized trial of standard vs intensive anthelminthic intervention (the Lake Victoria Island Intervention Study on Worms and Allergy-related diseases, LaVIISWA; ISRCTN47196031) described elsewhere. 21,22 Briefly, each consenting LaVIISWA participant completed a questionnaire, provided blood, urine and stool and underwent skin prick testing (SPT). Primary allergy-related outcomes were reported wheeze in the previous 12 months and atopy. Wheeze is widely used as a surrogate for asthma in epidemiological studies 23

| Laboratory methods
Two slides from one stool sample per individual were independently examined by different technicians for Sm eggs using the Kato-Katz method. 24 We assessed IFNγ (Th1-type), IL-5, IL-13 (Th2-type) and IL-10 (regulatory) levels by ELISA using supernatants from six-day whole blood cultures stimulated with Schistosoma worm (SWA) and egg antigens (SEA), as previously described. 25 Briefly, heparinized blood was diluted with RPMI 1640 medium (Life technologies, UK) supplemented with penicillin, streptomycin, glutamine and Hepes buffer (all from Life technologies, UK), plated in 96-well culture plates and stimulated (at 37°C, 5% CO 2 ) with 10 μg/mL SWA or SEA (provided by Professor Mike Doenhoff, University of Nottingham) or mitogen (phytohaemagglutinin, PHA, Sigma, UK), or left unstimulated. Supernatants were harvested on day six and stored at −80°C until analysis. Cytokine F I G U R E 1 We hypothesize that the Th2 cytokine-induced Sm-specific IgE promotes potent, Sm-specific, atopic effector responses and Sm elimination from the host, and also cross-reactive responses to some allergens, resulting in positive Sm-allergy associations. By contrast, Smspecific IL-10, IgG4 and/or nonspecific polyclonally stimulated IgE inhibit these allergy-related outcomes. White and shaded arrows denote promotion and inhibition, respectively levels in supernatants were measured by ELISA (Becton Dickinson, USA). The net response to each stimulus was calculated by subtracting the concentration in the unstimulated control well. Response values that were below the dynamic range of the assay and those that were negative after subtraction of the response in the unstimulated well were assigned a value of zero.
HDM and cockroach extract-specific IgE and IgG4 were measured in plasma using an in-house ELISA described previously. 20

| Statistical methods
Our hypothesized mode of action of S. mansoni-induced cytokines and antibodies on allergy-related outcomes is illustrated in Figure 1.
Using STATA 13.1 (College Station, Texas, USA), we performed crosssectional analyses to assess whether Sm Kato-Katz positivity and allergy-related outcomes were associated with antibody and cytokine levels, using the "svy" command to allow for the non-self-weighting cluster survey design. Raw cytokine and antibody responses were skewed, so log 10 (concentration+1)-transformed antibody and cytokine data were used in our regression models; we back-transformed the results to obtain geometric mean ratios and 95% confidence intervals. Crude and age-and sex-adjusted analyses were performed.
Associations between antibody responses were estimated using Spearman's correlation coefficient (r s ). We used a 5% significance level for all analyses. P values quoted in the main text are from adjusted analyses.

| RESULTS
Questionnaire data were obtained from 2316 participants. 22 Their characteristics and those of participants for whom cytokine and antibody responses were assessed are shown in Table 1

| S. mansoni-specific cytokines and allergyrelated outcomes
Individuals who tested positive for Sm by Kato-Katz (SmKK+) had higher geometric mean concentrations of SWA-specific type 2 and regulatory cytokines compared to SmKK-individuals (Table 2), but this was statistically significant only for IL-5 (P = .034). However, there was no dose-response relationship with infection intensity (Table S3A). SEA-specific responses were similar between SmKK+ and SmKK-individuals.
Kato-Katz positivity was also strongly positively associated with total IgE (P < .001) ( Table 3 and Figure S1), which was in turn weakly correlated with SEA-specific IgE but moderately correlated with SWAspecific IgE (r s = .31 and r s = .51, respectively; Table S2). Similarly, total IgG4 (P < .001), total IgG4/total IgE ratios (P = .005) and total IgE/asIgE ratios (P < .05) were positively associated with Kato-Katz positivity. In addition, there was a general dose-response relationship between S.
Cockroach-specific IgE and total IgE were positively associated with cockroach SPT reactivity. HDM-specific IgE and IgG4, SWAand SEA-specific IgE and total IgE, were all positively associated with HDM SPT reactivity (Table 3 and Figure S1). In contrast, cockroach SPT reactivity was inversely associated with total IgG4/total IgE ratios (P = .022), and HDM SPT reactivity with total IgE/HDM-specific IgE ratios (P < .001).

| DISCUSSION
In this highly Sm-endemic setting, associations between wheeze and Sm-specific cytokines and antibodies, when significant, were inverse.
However, SPT reactivity and detectable asIgE were positively associated with the same Sm-specific responses.
In this population, Sm exposure is almost universal, and infection All statistically significant associations between atopy and Smspecific cytokine responses were positive. Associations with whole blood cytokine responses are best interpreted taking into account total cell counts, but these data were unavailable. However, atopy-antibody associations were also positive. Besides, these results mirror our previ-  One limitation of assessing helminth-allergy associations and underlying mechanisms in this population is the almost universal exposure to helminths, and lack of data on duration of infection. We also report a large number of statistical tests, so some apparently "significant" findings could have occurred by chance. As we anticipated that some of our measures might be correlated, we did not formally adjust for multiplicity, instead we focussed on patterns of association and consistency of results, and on biological plausibility with reference to other All antibody concentrations in ng/mL. b All geometric mean ratios and 95% confidence intervals adjusted for survey design, age and sex. c Geometric mean ratios and 95% confidence intervals for associations between antibody levels and SPT reactivity and wheeze were additionally adjusted for SmKK result.
T A B L E 3 (Continued) findings. Another potential limitation is that wheeze was relatively rare in the study population, and hence, some of our comparison groups (such as the age group 1-17 years) had a low prevalence. Besides, reported wheeze could easily be misclassified in this population due to lack of a direct translation of "wheeze" in the native languages. 21 Nonetheless, our results generally agree with our epidemiological observations in the same population, 21 where we found a very low prevalence of clinical allergies, despite positive helminth-atopy associations.