As we recently demonstrated, hookworm administration in celiac disease did not result in a clinically significant suppression of pathology, although a trend towards reduced mucosal inflammation (as assessed by histological scoring) was seen with hookworm infection after gluten challenge 
. The infective dose of hookworms used in our trial was low (a total of 15 larvae), which although a safe dose 
, may be insufficient to effectively suppress the immunopathogy of celiac disease.
Surprisingly however, we detected suppression of the basal inflammatory immune response in hookworm infected subjects, reflected in the suppression of duodenal IFN-γ and IL-17A production in trial 1. To extend these results, we developed a biopsy restimulation assay for use in trial 2, using the purified QE65 peptide to stimulate gliadin-specific T cells 
. This allowed us to avoid the use of commercial gliadin preparations containing multiple epitopes, toxic factors and potential endotoxin contamination, which may mask the gliadin-specific T cell response 
. Using this system we detected robust production of IL-2 to QE65 prior to gluten challenge or infection, and strong trends for production of IFN-γ and IL-17A produced in response to QE65. These results indicate that QE65-specific memory T cells reside in the duodenal mucosa, even after >6 months of apparent strict adherence to a gluten-free diet. QE65-specific responses did not appear to be suppressed by hookworm infection, indicating that QE65-specfic T cell responses may remain high during low intensity hookworm infection, while basal inflammatory cytokine production is suppressed.
As candidate mechanisms for suppression of inflammatory immune responses, we proposed IL-10 production, Treg expansion or activation, and cross-regulation or skewing of the inflammatory immune response to a TH2 response. While we did not detect increased basal IL-10 production by duodenal biopsy cells, we observed IL-10 production after gluten challenge in the hookworm-infected participants in trial 2 in response to QE65; there was also a trend for elevated IL-10 production after hookworm infection but prior to gluten challenge. Therefore hookworm infection may be inducing IL-10, which in turn is suppressing the TH1/TH17 response, however a larger study is required to confirm the existence of a pre-gluten challenge IL-10 response. Of note however, the production of IL-10 (along with IFN-γ) in the duodenum during active celiac disease has been reported elsewhere 
, indicating that the post-challenge IL-10 production may be part of a normal celiac response.
Another candidate mechanism for immunomodulation observed in helminth infections is Treg induction. In mouse models, Tregs are induced by helminth infection 
, and are responsible for controlling immunopathology 
. Using Foxp3 as a marker of Tregs, we showed that gluten challenge induces an expansion of mucosal Foxp3+ cells in celiac sufferers irrespective of their hookworm infection status. We also show that in the control, but not the hookworm-infected group, the proportion of CD25+Foxp3+ cells in the circulating CD4+ population increased after gluten challenge, and these cells had increased expression of Foxp3 and CD25. In two recent studies comparing celiac sufferers with active or treated disease, levels of CD4+CD25+Foxp3+ cells in the blood were higher during active disease 
, however this is the first time this effect has been shown during a controlled, short-term gluten challenge. These results, and those from other studies of Tregs in immunopathology 
and surgical inflammation 
, indicate that CD4+Foxp3+ cells often accumulate during inflammation, however they may not be functionally suppressive 
. The suppression of effector T cells by regulatory T cells may be rendered unsuccessful due to upregulation of the negative regulator of TGF-β signalling Smad-7 
. Therefore counter-intuitively CD4+Foxp3+ cell levels in chronic inflammatory diseases correlate with severity of disease, not resolution. Recent work in Crohn's disease also raises the intriguing possibility that the Foxp3+ cells in the mucosa may also be producing IL-17, and so may not be fully committed to either the TH17 or Treg phenotypes 
. Unfortunately functional assays of CD4+Foxp3+ suppression were not within the scope of this study, so we cannot say whether the hookworm infection increased suppressive ability and so suppressed inflammatory cytokine production.
We propose that hookworm infection skews the celiac immune response towards a TH2 phenotype, cross-regulating the inflammatory TH1/TH17 response. When duodenal biopsy cultures from trial 2 were restimulated with QE65 peptide, we found no evidence of a TH2 response prior to infection but we did observe significantly increased production of IL-5 (and a trend for increased production of IL-13) to QE65 once the hookworm infection had become patent. This suggests that the anti-gluten response is being skewed away from a TH1/TH17 towards a TH2 phenotype as a consequence of the immune response to hookworm infection. TH2 responses are often associated with IL-10 production, which in turn suppresses inflammatory responses 
. Evidence for established inflammatory responses being suppressed by TH2 responses comes from coinfection studies, for instance established Bordetella pertussis
-inducedTH1 responses can be suppressed by Fasciola hepatica
TH2 responses in an IL-4-dependent manner 
. Indeed, skewing of the immune response towards TH2 in celiac disease has been suggested previously as a potential immunotherapy 
In summary, we present immunological data from a clinical trial using human hookworm to treat celiac disease. We cultured intestinal biopsies in a gliadin-derived peptide and found inflammatory T cell responses could be measured in the absence of in vivo gluten challenge, and that even short term gluten challenge increased the levels of CD4+Foxp3+ cells in celiac individuals. Hookworm infection suppressed basal production of the inflammatory cytokines IFN-γ and IL-17A, and our data indicate that this suppression may be dependent upon skewing or cross-regulation of the celiac response by a concurrent TH2 response, and/or IL-10 production. Future efforts should explore the potential protective effect of hookworm on a lesser gluten challenge than that used in this study, addressing the impact of infection on a low gluten diet.