Research ArticleColitis

IL-22+ CD4+ T Cells Are Associated with Therapeutic Trichuris trichiura Infection in an Ulcerative Colitis Patient

+ See all authors and affiliations

Science Translational Medicine  01 Dec 2010:
Vol. 2, Issue 60, pp. 60ra88
DOI: 10.1126/scitranslmed.3001500

Abstract

Ulcerative colitis, a type of inflammatory bowel disease, is less common in countries endemic for helminth infections, suggesting that helminth colonization may have the potential to regulate intestinal inflammation in inflammatory bowel diseases. Indeed, therapeutic effects of experimental helminth infection have been reported in both animal models and clinical trials. Here, we provide a comprehensive cellular and molecular portrait of dynamic changes in the intestinal mucosa of an individual who infected himself with Trichuris trichiura to treat his symptoms of ulcerative colitis. Tissue with active colitis had a prominent population of mucosal T helper (TH) cells that produced the inflammatory cytokine interleukin-17 (IL-17) but not IL-22, a cytokine involved in mucosal healing. After helminth exposure, the disease went into remission, and IL-22–producing TH cells accumulated in the mucosa. Genes involved in carbohydrate and lipid metabolism were up-regulated in helminth-colonized tissue, whereas tissues with active colitis showed up-regulation of proinflammatory genes such as IL-17, IL-13RA2, and CHI3L1. Therefore, T. trichiura colonization of the intestine may reduce symptomatic colitis by promoting goblet cell hyperplasia and mucus production through TH2 cytokines and IL-22. Improved understanding of the physiological effects of helminth infection may lead to new therapies for inflammatory bowel diseases.

Introduction

Ulcerative colitis (UC), a major form of inflammatory bowel disease (IBD), is characterized by chronic inflammation of the colonic mucosa. Although the etiology of UC is poorly understood, it is believed that an impaired intestinal epithelial barrier, together with defects in mucosal immune regulation, favors the development of pathogenic T helper 17 (TH17) cells (1), probably in response to commensal microbiota (2). Treatment of severe UC is usually through administration of immunosuppressive drugs, whose long-term use is limited by adverse side effects, including increased risk of opportunistic infection. Furthermore, up to 30% of patients develop disease refractory to treatment within 3 years of diagnosis, necessitating colectomy.

IBD is most prevalent in Northern Europe, the United Kingdom, and North America (3) and historically rare in regions of endemic helminth infection such as Asia, Africa, and Latin America (3, 4). This has raised the hypothesis that helminths may protect against the pathologic inflammation underlying IBD as a bystander effect of their ability to modulate the immune system to enable their own survival within mammalian hosts (5, 6). Studies of colitis in mice (6) as well as in clinical trials (7, 8) have suggested that helminth infection can prevent and/or treat IBD. A randomized placebo-controlled trial with Trichuris suis ova as therapy for UC indicated a therapeutic improvement in a disease activity index (7), but the mechanism of action was not characterized. A more complete understanding of this phenomenon could lead to the development of new therapeutics, for example, through infection with live helminths, identification of helminth-derived molecules with immunosuppressive effects, and/or identification of biological pathways activated by helminths that can be targeted through conventional approaches. Because helminth infections can themselves cause inflammation and colitis when immunoregulatory networks are already disrupted (9, 10), it is important to understand the molecular pathways that regulate the balance between inflammation and immunity in the intestinal mucosa if live helminths are used as therapeutic agents.

The immune response to helminth infection is typically characterized by a TH2 response (5), including the induction of eosinophilia and of macrophages activated by TH2 cytokines, termed alternatively activated macrophages. TH2 cytokines such as interleukin-4 (IL-4) and especially IL-13 contribute to wound healing and tissue remodeling (11) and may also promote epithelial integrity and mucosal healing. IL-22, an IL-10 cytokine family member, also promotes wound healing, proliferation, and antiapoptotic pathways in intestinal epithelial cells (12, 13) and furthermore up-regulates antimicrobial peptide and mucus production (1417). IL-22 can be protective in animal models of colitis, partially attributable to the induction of epithelial wound healing and mucus production (13, 17, 18). Although IL-22 was originally described as a cytokine produced by TH17 cells, it is now recognized that IL-22 expression by TH cells can be induced independently of IL-17 expression (19, 20). Therefore, regulation of the production of TH2 cytokines and IL-22 could be a potent avenue for influencing the course of colitis.

Here, we have analyzed the colonic mucosa from a UC patient who obtained symptomatic relief after infecting himself with the nematode parasite Trichuris trichiura. It is estimated that almost a billion people worldwide are infected with T. trichiura, with the highest prevalence in Central Africa, southern India, and Southeast Asia (4, 21). In contrast to T. suis ova, this human parasite establishes chronic, not transient, colonization. Genome-wide transcriptional profiling, flow cytometric evaluation of the immune response, and histological analyses were performed on biopsy samples collected during active colitis and during stable disease remission associated with helminth infection.

Results

Clinical course and histopathology of UC in the setting of T. trichiura infection

We followed the disease course of a 35-year-old man diagnosed with UC in 2003 (Fig. 1 and fig. S1). His initial disease was severe and refractory to mesalamine agents, mercaptopurine, and high-dose steroids. In 2003, his sigmoid colon had extensive ulceration of the mucosal epithelium, with few epithelial-lined mucosal glands remaining (Fig. 2A and fig. S2). Neutrophils heavily infiltrated the glands (Fig. 2B) and the surrounding lamina propria, forming crypt abscesses. In early 2004, the immunosuppressant cyclosporine or colectomy was advised. Instead, the patient chose to infect himself with T. trichiura eggs obtained in Thailand. He ingested 500 in vitro–germinated eggs in late 2004 and an additional 1000 eggs 3 months later. His symptoms improved in the following months, and by mid-2005, he was symptom-free and required no treatment for UC.

Fig. 1

Overview of clinical course and analyses. (A) Time course of UC disease severity (red showing colitis and yellow showing remission) in relation to T. trichiura infection, biopsy collection, and analyses of biopsy tissue by transcriptional profiling and flow cytometry. The subject ingested parasite ova twice in 2004 and again in August 2008. Histopathology slides of biopsies collected in August 2003 and October 2005 were reviewed. In January 2007, biopsies were collected for RNA analysis during a flare of proctitis. When the subject reverted to pan-colitis in July 2008, biopsies were collected for flow cytometry and RNA analysis. In March 2009, a colonoscopy demonstrated mucosal healing and biopsies were again collected for flow cytometry and RNA analysis. (B) Examples of gross pathology seen upon endoscopic examination. In 2007, worms were observed in the cecum, ascending colon, and transverse colon, whereas the sigmoid colon was normal and the rectum exhibited signs of proctitis. In 2008, worms were observed only in the ascending colon and not in the transverse colon; however, the remainder of the colon exhibited signs of severe colitis. In 2009, worms were mainly observed in the ascending colon, and intact mucosa was observed in the remainder of the colon with few signs of inflammation.

Fig. 2

Distinct inflammatory infiltrates characterize colitis-affected and helminth-exposed mucosal tissues. Tissue sections of colon biopsies stained with hematoxylin and eosin. (A) Sigmoid colon in 2003 showed signs of severe colitis including ulcerated epithelium (UE) and crypt abscesses (CA). (B) High-powered view of the sigmoid colon in 2003 showing prominent neutrophils (N), infiltration of glands, and a crypt abscess (CA). (C) Ascending colon in 2005 with cross section of a T. trichiura worm (W), prominent eosinophils (E), and plasma cells (P), as well as goblet cell (G) hyperplasia. (D) Sigmoid colon in 2005 showed restoration of glands (gl) and goblet cells (G). (E) Ascending colon in 2007 with cross section of a T. trichiura worm, with prominent eosinophils and goblet cells. (F) High-power view of the ascending colon in 2007 showing prominent eosinophilia and goblet cells. (G) Proctitis in the rectum in 2007 showed lymphoid aggregates (LA) and neutrophil infiltration into the glands. (H) Severe colitis in the sigmoid colon in 2008 showing loss of glands and ulcerated epithelia and crypt abscess. (I) High-power view of sigmoid colon in 2008 showing pronounced neutrophil infiltration. (J) Sigmoid colon in 2009 showed restoration of glands and goblet cells. Black scale bars, 100 μm; Red scale bars, 50 μm.

He periodically took 5-aminosalicylates either topically or by mouth to control symptomatic flares, although he was generally not under any medical therapy. In October 2005, T. trichiura worms (Fig. 2C) were observed in the cecum, ascending colon, and transverse colon. Despite a prominent infiltration of eosinophils in the lamina propria of these tissues, the mucosa was intact with no significant epithelial cell loss and the glands appeared normal (Fig. 2C and fig. S2). In contrast to 2003, no ulceration or neutrophil infiltration was observed in the sigmoid colon, and glands demonstrated moderate goblet cell hyperplasia and mucus hypersecretion (Fig. 2D).

A brief flare of symptoms in 2007 warranted a colonoscopy. The ascending colon continued to harbor a heavy worm burden (Fig. 1B and fig. S1). This tissue showed no ulceration, with normal glands and many macrophages, eosinophils, and plasma cells (Fig. 2, E and F), similar to the previous colonoscopy (Fig. 2C and fig. S2). The histopathology of the sigmoid colon was essentially unchanged from 2005. Consistent with the flare of proctitis, macroscopic signs of colitis were apparent in the rectum (Fig. 2G). One biopsy contained a predominantly neutrophilic infiltrate in the lamina propria, with cryptitis, a crypt abscess, and several prominent lymphoid aggregates (Fig. 2G). In contrast, other biopsies showed minimal infiltration of the lamina propria, intact epithelium, and goblet cell hyperplasia.

After 3 years of nearly complete disease remission, the patient’s symptoms began to deteriorate in mid-2008, paralleling a decline in stool egg counts from an extremely high number (>15,000 eggs per gram) to more moderate numbers (<7000 eggs per gram). There was active colitis in both the ascending and the sigmoid colons (Fig. 1B and fig. S1). The ascending colon remained colonized by worms (Fig. 1B) and continued to show minimal signs of chronic colitis (fig. S2). In the sigmoid colon (Fig. 2H and fig. S2), evidence for much more severe colitis was observed. Multiple crypt abscesses accompanied prominent neutrophilic infiltrates (Fig. 2I). Changes in tissue architecture were apparent, including a loss of glands and distortion of the epithelial layer (Fig. 2H). Ulceration and granulation tissue indicated that there was chronic inflammation in this tissue.

The patient chose to infect himself again, this time with 2000 T. trichiura eggs. His symptoms improved in the following months, and he required no other medication. He had another colonoscopy in early 2009. Large numbers of worms were observed in the ascending colon (Fig. 1B), and beyond an eosinophilic infiltrate in the lamina propria, no signs of colitis were present. The sigmoid colon still demonstrated mild colitis, with a few scattered neutrophils in the lamina propria and minor crypt distortion. However, goblet cell numbers were restored and no ulceration was noted (Fig. 2J). Thus, renewed T. trichiura colonization was again associated with improvement in histopathologic findings, including normalization of tissue architecture and recovery of epithelial integrity.

Characterization of TH cell responses in the colonic mucosa

Biopsies collected during the 2008 and 2009 colonoscopies were analyzed by flow cytometry for intracellular IL-17, IL-22, IL-4, interferon-γ (IFN-γ), and tumor necrosis factor–α (TNFα) (fig. S3) production. We examined the relative proportions of monocytokine- and polycytokine-producing CD4+ TH cells (Fig. 3A). In 2008, during the relapse of symptomatic colitis, 70% of cytokine-producing CD4+ TH cells from the severely inflamed sigmoid colon expressed only IL-17 (Fig. 3 and fig. S3), whereas less than 1% expressed IL-22. In contrast, biopsies from the ascending and transverse colon, in which epithelial integrity and gland structure remained intact, had more IL-22+ cells (Fig. 3 and fig. S3).

Fig. 3

Induction of IL-22 expression in TH cells of the sigmoid colon is associated with restoration of mucus production. Lymphocytes from colon biopsies collected in 2008 and 2009 were analyzed by flow cytometry for intracellular cytokine expression after a 5-hour PMA-ionomycin stimulation ex vivo. (A) Visualization of the combinations of cytokines expressed by CD4+ T cells from colon biopsies with pie charts in which each slice represents a different cytokine combination. The flow cytometric gating strategy used to generate these charts is shown in fig. S3. Boolean gates were created for cytokine positive cells and used to divide the cytokine-producing cells into distinct populations corresponding to the patterns of cytokines that they are producing. These results were then graphed using the SPICE software to generate the pie charts representing the proportion of cytokine-expressing CD4+ T cells that express combinations of IL-17, IL-22, IFN-γ, and/or IL-4. Cells that are not producing any of these cytokines are not represented in the pie charts. (B) Flow cytometry bivariate contour plots showing the frequencies of CD4+ T cells expressing IL-17, IL-22, IFN-γ, and IL-4 from biopsy samples taken from the sigmoid colon and ascending colon in 2008 and 2009. (C) Mucus production in the sigmoid colon was visualized by the periodic acid–Schiff (PAS) stain. Scale bars, 50 μm.

Analysis of mucosal TH cell responses in 2009, during the period of remission after reinfection with T. trichiura, revealed that IL-17 expression by TH cells in the sigmoid colon was comparable to that of 2008 (Fig. 3A). Furthermore, the IL-17 response was more prominent in the ascending and transverse colon in 2009 than in 2008. However, a significant population of IL-22+ TH cells was now observed in the sigmoid colon (Fig. 3, A and B), such that the proportion of IL-22+ TH cells was similar in the sigmoid and ascending colon. As expected, IL-4+ TH cells were also increased in the helminth-colonized ascending colon (Fig. 3 and fig. S3). Thus, despite the continued presence of CD4+ IL-17+ TH cells, symptomatic remission associated with a reduction in infiltrating neutrophils and restoration of tissue architecture was characterized by the presence of CD4+ IL-22+ TH cells.

Because the protective effect of IL-22 in mouse models of colitis is linked with the stimulation of mucus production by goblet cells (17), we used the periodic acid–Schiff (PAS) stain to determine that although virtually no mucus was detectable in the sigmoid colon in 2008, ample mucus production was present in 2009 (Fig. 3C). A role for TNFα in UC pathogenesis has been suggested by the positive outcomes of several clinical trials with the anti-TNF antibody infliximab (22). TNFα is expressed by several types of immune cells, notably activated macrophages and T cells. Although we found a higher frequency of TNFα+ TH cells in colitis-affected tissue compared to helminth-exposed tissue in 2008 (fig. S3), remission was not associated with the suppression of TNFα responses. The inflammatory response to reinfection with T. trichiura was characterized by a high frequency of TNFα+ TH cells (fig. S3). In summary, active colitis was associated with monocytokine-producing CD4+ IL-17+ TH cells, whereas helminth colonization and disease remission were characterized by the presence of IL-22+ TH cells in the colonic mucosa. High frequencies of mucosal IL-17+ and TNFα+ TH cells were associated with active colitis in 2008, but these cells persisted during symptomatic remission in 2009.

Characterization of TH cell responses in the peripheral blood

T cell responses in the peripheral blood were also analyzed by flow cytometry at three time points: mid-2008, during UC relapse; late 2008, immediately after reinfection with T. trichiura; and early 2009, during remission (Fig. 4). Isolated peripheral blood mononuclear cells (PBMCs) from each time point were cryopreserved at the time of venipuncture and later cultured in parallel in the presence of T. trichiura antigen prepared from an adult worm segment extracted from a biopsy. After 96 hours in culture with medium alone, TH cells from the 2008 time points showed signs of nonspecific activation, expressing many cytokines in the absence of stimulation with T. trichiura antigen (Fig. 4, A and B). This could reflect an acute inflammatory response to helminth infection as well as ongoing colitis-associated inflammation. In contrast, T cells collected during remission (several months after reinfection) did not shown signs of nonspecific activation. Culture with T. trichiura antigen selectively promoted expansion of IL-4+ TH cells in PBMCs that were collected after reinfection (Fig. 4, A and B). Small numbers of IL-22+ TH cells that were specific to T. trichiura antigen could also be detected in 2009 (Fig. 4A). A circulating population of IL-4+ IL-17+ TH cells was recently identified in patients with asthma (23), a condition also associated with chronic TH2-driven inflammation. Similarly, IL-22+ and IL-17+ TH cells that coexpressed IL-4 in response to T. trichiura antigen were present in PBMCs after reinfection (Fig. 4C). This may represent a signature in the peripheral blood of mucosal TH2-associated inflammation.

Fig. 4

IL-4+ TH cells characterize T. trichiura–specific responses in the peripheral blood. Cryopreserved PBMCs collected at three time points (marked with arrows) were thawed and cultured in parallel for 96 hours in the presence or absence of a homogenate prepared from worm fragments collected during colonoscopy. After culture, cells were assayed by flow cytometry for intracellular cytokine expression after a 5-hour PMA-ionomycin stimulation. (A and B) Percentage of CD4+ T cells that produce cytokines in response to T. trichiura antigen stimulation compared to cells cultured in media alone are shown as histograms (A) or bivariate flow cytometry contour plots (B). (C) Flow cytometry gating strategy to illustrate an expansion of T. trichiura antigen-specific IL-4+ cells within both the IL-22+ and the IL-17+ CD4+ T cell subsets from PBMCs that were collected after reinfection (September 2008). Worm antigen was added at 100 μg/ml (high) or 25 μg/ml (low). SSC, side scatter.

Because regulatory T cells (Tregs), which express the transcription factor FoxP3, have been implicated in helminth-mediated immune regulation (5), we quantified FoxP3+ cells in the colonic mucosa by immunohistochemistry (Fig. 5). FoxP3+ cells were more abundant in the colitis-affected tissue than in the helminth-colonized tissue at both time points (Fig. 5). These results suggest that the presence of Tregs in the mucosal tissues is driven predominantly by inflammation rather than helminth colonization, consistent with observations that a high number of Tregs are a marker of immune activation in the intestinal mucosa (24).

Fig. 5

Detection of FoxP3 expression by immunohistochemistry shows that FoxP3+ cells are more abundant in colitis-affected tissue compared to helminth-exposed tissue. (A) Representative low-powered immunohistochemistry images of FoxP3 staining on biopsies collected from the sigmoid colon and the ascending colon in 2008 and 2009. Brown dots are nuclear-stained FoxP3+ cells. Scale bars, 50 μm. (B) Histograms showing the differences in the average number of FoxP3+ cells per field of view between the sigmoid colon and the ascending colon in 2008 and 2009. ***P < 0.0005; *P < 0.05.

Transcriptional profiling analysis of mucosal responses

To identify molecular signatures that are associated with helminth colonization and colitis, we performed gene expression profiling on biopsies collected from colonoscopies in 2007, 2008, and 2009. Each biopsy fragment was treated as an independent sample. In 2007, samples were obtained from the terminal ileum (n = 2), the ascending colon (n = 3), the transverse colon (n = 2), the sigmoid colon (n = 4), and the rectum (n = 3). In 2008, samples were obtained from the ascending colon (n = 4), the transverse colon (n = 4), and the sigmoid colon (n = 4). A sample with a worm and its surrounding tissue was also analyzed (n = 1). Finally, in 2009, samples were obtained from the terminal ileum (n = 3), the ascending colon (n = 5), the transverse colon (n = 3), and the sigmoid colon (n = 3).

Unsupervised hierarchical clustering analysis was performed on the three data sets (Fig. 6 and fig. S4) to determine whether gene expression profiles in different anatomical locations were related. In the first colonoscopy (in 2007), the samples from the ascending colon and transverse colon clustered together, reflecting a similar response to worm colonization in both regions. The samples from the rectum also clustered together strongly because of the common expression of a group of inflammatory genes reflecting the active colitis flare in these tissues at that time (Fig. 6 and fig. S4). The terminal ileum had a distinct profile (probably because it is part of the small intestine), but it also had a profile more similar to that of the normal sigmoid colon than either the colonized or the inflamed regions of the large intestine. In 2008, worms were not found in the transverse colon, and at this time, this region had a profile quite distinct from that of the ascending colon, which remained colonized by worms. The entire sigmoid colon at this time was heavily inflamed, and these samples clustered tightly together. In 2009, when the colon was no longer inflamed and the patient was in remission, the clustering of expression profiles by anatomical location was much less pronounced. Although the terminal ileum remained very distinct, samples from the ascending colon, transverse colon, and sigmoid colon were interspersed in different branches of the tree [Fig. 6: colonoscopy 3 (2009)].

Fig. 6

Transcriptional profiling analysis of biopsy fragments from different regions of the colon in three different colonoscopies. Gene expression patterns of biopsy fragments from the terminal ileum (TI), ascending colon (AC), transverse colon (TC), sigmoid colon (SC), rectum (R), or rectosigmoid colon (SRC) taken during a proctitis flare (2007), severe colitis (2008), and symptomatic remission (2009). Hierarchical clustering analysis was used to organize genes and samples. Each row represents an individual gene and each column an individual biopsy fragment from a specific region. Black indicates median amount of expression; red, greater than median expression; and green, less than median expression. Horizontal bars at the bottom of the figure indicate the clustering or dispersal of the samples affected by colitis (red), helminth colonization (blue), normal (green), and from the uninvolved terminal ileum (purple). Note that in 2009, the horizontal bars of the same colors are not clustered.

Supervised two-way comparisons were then made between biopsies from the helminth-colonized areas and the colitis-affected areas to identify genes that were differentially expressed as a result of colonization (Table 1 and table S1) or colitis (table S2). In 2007 and 2008, the genes that were differentially expressed between areas of helminth colonization and colitis were enriched by gene ontology (GO) analysis for the biological processes of carbohydrate, lipid, fatty acid, and steroid metabolism for helminth colonization (Fig. 7A and fig. S5) and immunity and defense for colitis (Fig. 7B and fig. S5). These results suggest that, whereas colitis is associated with immune-mediated responses, helminth colonization may be associated with increased carbohydrate and steroid metabolism. A large number of genes (Table 1 and table S1) up-regulated in helminth-colonized tissues are uncharacterized, revealing the rudimentary extent of our understanding of the mucosal response to helminth colonization. In 2009, the uninflamed colon had very few (fewer than five) genes that were significantly differentially expressed between the ascending colon (helminth-colonized) and the sigmoid colon (previously inflamed) tissues. When we conducted multiclass significance analysis of microarrays (SAM) analysis to include comparison with biopsy tissues that are unaffected by colitis or helminth colonization, we found that the genes that are most differentially expressed between regions of the colon in 2007 and 2008 were still enriched for the biological processes of carbohydrate and steroid metabolism as well as for granulocyte-mediated immunity (Fig. 7C and fig. S5).

Table 1

Top 50 genes most highly expressed in helminth-colonized tissues relative to colitis-affected tissues from the colonoscopy in 2007.

View this table:
Fig. 7

GO analyses to identify biological processes that are induced in helminth-colonized and colitis-affected regions of the colon. (A) Biological processes induced in helminth-colonized tissues as determined by GO analysis of genes that are significantly up-regulated in these tissues in 2007. X axis indicates the amount of statistical significance [as −log(P)] in enrichment for the indicated biological process. (B) Biological processes induced in tissues with active colitis as determined by GO analysis of genes significantly up-regulated in these tissues in 2007. (C) Biological processes of genes differentially expressed between regions with active colitis, helminth colonization, and normal appearance as determined by multiclass analysis with the significance analysis of microarray (SAM) analysis from samples in 2008.

After completing analysis of the microarray results from the first colonoscopy (2007), we conducted real-time polymerase chain reaction (PCR) analysis to measure expression of genes previously identified to be important in helminth immunity and mucosal responses (Fig. 8A and fig. S6), as well as to verify genes identified by our microarray analysis (Fig. 8B). Helminth-colonized ascending colon expresses significantly (P < 0.05) higher amounts of transcripts for IL-4, IL-25, and RELMβ (resistin-like molecule β) than the rectum. By contrast, IL-17 (Fig. 8A) and TNF (fig. S6) were elevated in the rectum, which had active proctitis. In addition to the well-established inflammatory cytokines (for example, IL-1β, IL-6, and IL-8), this tissue also revealed markedly high expression of the decoy IL-13 receptor α2 (IL-13RA2) and the chitinase 3–like 1 (CHI3L1) (Fig. 8, B and C). The chemokine CCL20 was more highly expressed in the helminth-colonized tissue (in the ascending and traverse colon), whereas CXCL1 and CXCL13 were more expressed in the rectal region with proctitis (Fig. 8B). Helminth-colonized tissues also expressed much higher amounts of metabolic enzymes such as ethanolamine kinase 1 (ERHK1) and hyaluronan synthase 3 (HAS3).

Fig. 8

Real-time PCR validation and analysis of differentially expressed genes in biopsy fragments. Samples from helminth-colonized tissue are shaded in light blue, and samples from tissue with active colitis are shaded in red. TI, terminal ileum; AC, ascending colon; TC, transverse colon; SC, sigmoid colon; SRC, rectosigmoid colon; W, worm granuloma; R, rectum. (A) Box plot showing the minimum, maximum, and median values of real-time PCR measurements of selected proinflammatory mediators from biopsy fragments collected in 2007. *P < 0.05. (B) Real-time PCR verification of selected genes from the microarray profiling in 2007. Each individual spot shows the relative expression of a gene from a particular sample in relative shades of black and red. Darker spots reflect higher expression. Red spots (with borders) reflect 10 times greater amounts of expression than black spots and are used when there are logarithmic-level differences in expression. (C) Real-time PCR measurement of dynamic changes in selected proinflammatory mediators from samples taken in 2007, 2008, and 2009, and compared to samples collected from a normal individual (N) as indicated by an orange vertical bar (Control). Red vertical bars (Col) indicate samples taken from regions with active colitis (also shaded in red), and blue vertical bars (Hel) indicate samples taken from helminth-colonized tissues (also shaded in blue). Tissues with normal appearance (Norm) are indicated in green, and samples from the terminal ileum are indicated in purple. A sample that came from a worm surrounded by granulomatous tissue is shown in light blue.

We conducted additional real-time PCR analysis of biopsies taken from 2008 and 2009 to examine dynamic changes of specific genes (Fig. 8C and fig. S7). We also compared RNA expression in these biopsies to that found in biopsies (n = 6) taken from a healthy individual. The improvement in inflammatory conditions for the colitis-affected regions in 2009 was most markedly associated with decreased expression of IL-8, IL-17A, IL-13RA2, and CHI3L1 (Fig. 8C and fig. S7). In contrast, expression of matrix metalloproteinase 7 (MMP7) remained high in the sigmoid colon tissues for 2009, indicating that the process of tissue repair was ongoing. Also, expression of RELMβ was significantly (P = 0.0002) reduced under conditions of colitis, relative to helminth-colonized tissues and healthy tissues. Helminth-colonized tissues consistently expressed proinflammatory genes also, although in lower amounts than in the colitis-affected regions. These dynamic changes in gene expression provide a molecular picture of inflammation during colitis, tissue recovery, wound healing, and helminth colonization.

Discussion

Helminthic therapy has generated substantial interest clinically, in large part because of the successful outcomes of clinical trials showing a positive effect of T. suis ova treatment on IBD (7, 8) and a growing body of literature demonstrating the ability of helminths to suppress pathologic inflammation in animal models of autoimmunity (5, 6). However, the mechanisms that underlie the benefits of helminth exposure are unknown. We found that intestinal tissue of the colon with active colitis contains monocytokine-producing IL-17+ TH cells in the lamina propria, whereas intestinal tissue that has been colonized by helminths or has undergone mucosal healing contains polycytokine-producing IL-22+ TH cells, supporting a protective role for IL-22 in maintaining mucosal integrity. Although it is difficult to draw firm conclusions from the analysis of a single individual, these findings have generated hypotheses that can be tested through future functional studies in mouse models or with larger clinical cohorts.

We hypothesize that the presence of T. trichiura in the intestinal epithelium activates a TH2 response as well as IL-22+ TH cells to expel the parasites through increased epithelial cell turnover, goblet cell hyperplasia, and increased mucus production in the entire colon (25), which has the bystander consequence of reducing the pathology associated with UC and leading to symptomatic relief. In this subject, the inflammatory response induced by T. trichiura colonization was associated with the repair of the colonic epithelium and glands and a marked restoration of mucus production.

Whereas goblet cell hyperplasia is a characteristic of helminth infection, loss of goblet cells and mucus production is commonly observed in the diseased tissue of UC patients (1). Mucus plays a key role in maintaining the protective mucosal barrier (26). Some mucin gene variants confer a predisposition to UC (27, 28). Furthermore, mucin-deficient mice develop spontaneous colitis (2931). Indeed, the delivery of phosphatidyl-rich phospholipids into the lumen of UC patients has been clinically tested with positive results (26, 32, 33), with the hypothesis that it improves the barrier function of the mucus layer. GO analysis of our microarray data revealed that carbohydrate metabolism pathways were highly up-regulated in helminth-exposed compared to colitis-affected tissue. This may reflect goblet cell hyperplasia in helminth-colonized tissues; indeed, expression of mucins 1 and 4 is up-regulated after helminth colonization (fig. S4).

IL-22 induces intestinal mucus production in mouse goblet cells, and this effect likely partially underlies the ability of IL-22 to suppress colitis. IL-13 and IL-10 (9, 25) also promote mucus production during helminth infection and may contribute to the effect observed in this subject. Hence, the coordinated immune response activated to expel the nematode parasite and to protect the host from the intestinal damage that is caused by the invasion of the epithelium by the parasite may also promote mucosal healing at a distal site affected by colitis. However, it is also possible that the appearance of IL-22+ TH cells in the intestinal mucosa is an indicator of mucosal healing and would occur independently of helminth infection in individuals with UC who spontaneously go into remission. Excess IL-13 production is itself pathogenic in a mouse model of UC (34). It is possible that helminth infection may exacerbate UC symptoms in some patients by elevating the TH2 response in the intestinal mucosa. If helminthic therapy becomes widely used, it will be particularly important to separate patients into groups that may respond, may not respond, or may suffer disease exacerbation from infection.

Transcriptional profiling may distinguish among patients with UC, Crohn’s disease, and irritable bowel syndrome (35), as well as predict responses to TNF antibody treatments (36). Many of the genes identified in these previous studies as associated with UC were also found by us to be up-regulated in colitis-affected tissues (for example, IL-13RA2, CHI3L1, MMP3, MMP7, and PTGS2), as well as established and expected inflammatory mediators (for example, IL-1B, IL-6, IL-8, NOS2A, CXCL1, and CXCL13), validating our approach. In our comparison of mucosal inflammatory responses to helminth colonization with inflammatory responses during UC, we found that many of the genes associated with helminth colonization (Table 1) have unknown functions. IL-13RA2, which we found to be increased with colitis and T. trichiura colonization (Fig. 7), has been associated with fibrosis during chronic colitis, airway inflammation, and schistosomiasis (10, 11, 37, 38). Whether IL-13RA2 plays a protective or a pathogenic role in intestinal inflammation remains to be established. Increased expression of CHI3L1 (also known as YKL-40) has been associated with IBD (39) and airway inflammation (40), where it may be involved in extracellular matrix remodeling. Notably, we found that the expression of RELMβ was inversely correlated with expression of IL-13RA2 and CHI3L1, being suppressed in tissues with active colitis. RELMβ is a protein produced by goblet cells that promotes expulsion of gastrointestinal parasites (41, 42) and reduces the severity of trinitrobenzene sulfonate–induced colitis in mice (43).

Although this study suggests that T. trichiura infection can alleviate symptoms of UC, infection itself can also cause intestinal inflammation that mimics IBD (9, 10). Indeed, proinflammatory cytokines and mediators were up-regulated in helminth-colonized regions of the colon, although at lower amounts than in regions with active colitis. Heavy worm infestation, especially in children, can cause dysentery and rectal prolapse and lead to growth retardation, secondary anemia, and reduced cognitive function. The specific conditions in which T. trichiura infection may lead to any therapeutic benefit are currently unclear, and it is possible that infection can exacerbate existing conditions. Nonetheless, the phenomenon of helminth-mediated immune modulation is now well established (5, 6). Hence, the identification of the mechanisms of these helminth-induced mucosal responses could provide new therapeutic targets for IBD.

Materials and Methods

Biopsy collection and endoscopic analysis

Four to five pinch biopsies were collected from each different anatomical location (terminal ileum, ascending colon, transverse colon, sigmoid colon, or rectum) during ileocolonoscopies in 2007, 2008, and 2009. Two biopsies were sent for histology, and the other biopsies were separated into fragments for immediate flow cytometry analysis or snap-frozen for RNA extraction in Trizol reagent (Invitrogen).

Flow cytometry on colon biopsies and peripheral blood

Colon biopsy specimens were digested with collagenase and dispersed over a filter. PBMCs were isolated from whole blood by density centrifugation. For stimulation of PBMCs with T. trichiura antigen, a homogenate was prepared from an adult worm fragment collected at biopsy. Cryopreserved PBMCs from each time point were thawed and cultured in parallel for 96 hours in the presence or absence of homogenate at 100 μg/ml (high) or 25 μg/ml (low). Biopsy cells and PBMCs (1 × 106) were stimulated with phorbol 12-myristate 13-acetate (PMA) (10 ng/ml) and ionomycin (1 μg/ml) in the presence of brefeldin A (GolgiPlug, BD Pharmingen) for 5 hours at 37°C. Cell surface staining and intracellular cytokine staining were performed with Fix/Perm and Perm/Wash solutions from BD and eBioscience, according to the manufacturer’s instructions. Staining antibodies are listed in table S3. Pie charts were generated with the SPICE software (provided by M. Roederer).

Microarray and real-time PCR analysis of biopsy fragment samples

Total RNA was extracted from biopsy fragments and amplified with an RNA amplification kit (Ambion). Cy5-labeled experimental samples were hybridized against a Cy3-labeled reference consisting of an equal quantity of pooled, amplified RNA from all the experimental samples from that experiment. Two-color hybridizations were performed on Human Exonic Evidence-Based Oligonucleotide (Invitrogen) microarrays (44) printed in-house at the University of California, San Francisco Center for Advanced Technologies. Arrays were scanned with a GenePix 4000B scanner and GenePix PRO version 4.1 (Axon Instruments/Molecular Devices). The Spotreader program (Niles Scientific) was used for array gridding and image analysis. The resulting files were uploaded to Acuity version 4.0 (Molecular Devices), where the raw data were log-transformed and filtered for retention of spots that were of high quality and for removal of nonhuman control spots. Expression profiling data were filtered further (for microarray spots with data in at least 80% of the arrays) before undergoing unsupervised hierarchical clustering analysis and then visualized with Treeview. Filtered data sets were also analyzed for statistically significant genes (either through two comparisons or through multiclass analysis) with the SAM software version 2.23A (45). GO and pathway analyses were performed with Protein Analysis Through Evolutionary Relationships (PANTHER) software (46). For real-time PCR analysis, 100 ng to 1 μg of RNA from each sample was reverse-transcribed, and the resulting complementary DNA (cDNA) was used in quantitative real-time PCRs with SYBR Green labeling. Most of the PCRs used primers designed (in-house) to span introns so that amplification of genomic DNA could be avoided, whereas IL-17 and FoxP3 messenger RNA (mRNA) were measured with TaqMan probes (Applied Biosystems). All values were normalized to β-actin values. Data were analyzed with the t test with GraphPad Prism software.

Footnotes

  • Citation: M. J. Broadhurst, J. M. Leung, V. Kashyap, J. M. McCune, U. Mahadevan, J. H. McKerrow, P. Loke, IL-22+ CD4+ T Cells Are Associated with Therapeutic Trichuris trichiura Infection in an Ulcerative Colitis Patient Sci. Transl. Med. 2, 60ra88 (2010).

Supplementary Material

www.sciencetranslationalmedicine.org/cgi/content/full/2/60/60ra88/DC1

Methods

Fig. S1. Examples of photographs taken during endoscopic examinations in 2007, 2008, and 2009 from different regions of the colon.

Fig. S2. Examples of histopathology images from colitis-affected and helminth-exposed mucosal tissue.

Fig. S3. Flow cytometry analysis of mucosal tissues from 2008 and 2009.

Fig. S4. Microarray analysis of biopsy fragments from different regions of the colon.

Fig. S5. Gene ontology (GO) analysis of genes that are differentially expressed in different regions of the colon.

Fig. S6. Real-time PCR analysis of inflammatory mediators from biopsy fragments in 2007.

Fig. S7. Real-time PCR measurement of dynamic changes in additional proinflammatory mediators in biopsy fragments.

Table S1. Genes that are more highly expressed in the helminth-colonized tissues from the colonoscopy in 2008, relative to the colitis-affected sigmoid colon.

Table S2. Genes that are more highly expressed in the rectum during a flare of proctitis in 2007, relative to helminth-colonized tissues.

Table S3. Antibody panel for multiparameter flow cytometry on colon biopsies.

References and Notes

  1. Acknowledgments: We thank D. Favre for help with fluorescence-activated cell sorting analysis, K. Evason for histopathology assistance, and C. C. Kim and S. Batra for help with microarray analysis. Funding: This work was supported in part by the Sandler Foundation (J.H.M.). J.M.M. is a recipient of the NIH Director’s Pioneer Award Program (DPI OD00329), part of the NIH Roadmap for Medical Research. P.L. was a recipient of a F32 fellowship (AI066470). Author contributions: M.J.B. and J.M.L. performed the experiments, analyzed the data, and wrote the paper. V.K. and U.M. were involved in study design and specimen collection. J.M.M. was involved in study design, provided funding, and wrote the paper. J.H.M. read the histopathology slides and provided funding. P.L. designed the study, performed the experiments, analyzed the data, and wrote the paper. Competing interests: The authors declare that they have no competing interests. Accession numbers: The microarray data are available in GEO under accession numbers GSE25275, GSE25276, and GSE25277.
View Abstract

Related Content

Navigate This Article