Genome-Wide Association Study Reveals Genetic Link between Diarrhea-Associated Entamoeba histolytica Infection and Inflammatory Bowel Disease

Abstract

Entamoeba histolytica is the etiologic agent of amebic dysentery, though clinical manifestation of infection is highly variable ranging from subclinical colonization to invasive disease. We hypothesize that host genetics contribute to the variable outcomes of E. histolytica infection; thus, we conducted a genome-wide association study (GWAS) in two independent birth cohorts of Bangladeshi infants monitored for susceptibility to E. histolytica disease in the first year of life. Children with at least one diarrheal episode positive for E. histolytica (cases) were compared to children with no detectable E. histolytica infection in the same time frame (controls). Meta-analyses under a fixed-effect inverse variance weighting model identified multiple variants in a region of chromosome 10 containing loci associated with symptomatic E. histolytica infection. An intergenic insertion between CREM and CCNY (rs58000832) achieved genome-wide significance (P value from meta-analysis [P_meta] = 6.05 × 10^-9), and each additional risk allele of rs58000832 conferred 2.42 increased odds of a diarrhea-associated E. histolytica infection. The most strongly associated single nucleotide polymorphism (SNP) within a gene was in an intron of CREM (rs58468612; P_meta = 8.94 × 10^-8), which has been implicated as a susceptibility locus for inflammatory bowel disease (IBD). Gene expression resources suggest associated loci are related to the lower expression of CREM Increased CREM expression is also observed in early E. histolytica infection. Further, CREM^-/- mice were more susceptible to E. histolytica amebic colitis. These genetic associations reinforce the pathological similarities observed in gut inflammation between E. histolytica infection and IBD.IMPORTANCE Diarrhea is the second leading cause of death for children globally, causing 760,000 deaths each year in children less than 5 years old. Amebic dysentery contributes significantly to this burden, especially in developing countries. The identification of host factors that control or enable enteric pathogens has the potential to transform our understanding of disease predisposition, outcomes, and treatments. Our discovery of the transcriptional regulator cAMP-responsive element modulator (CREM) as a genetic modifier of susceptibility to amebic disease has implications for understanding the pathogenesis of other diarrheal infections. Further, emerging evidence for CREM in IBD susceptibility suggests that CREM is a critical regulator of enteric inflammation and may have broad therapeutic potential as a drug target across intestinal inflammatory diseases.

Keywords: diarrhea; epidemiology; genomics; infectious disease; protozoa; public health.

FIG 1

Manhattan plot of diarrhea-associated E. histolytica infection within the first year of life. The genome-wide significance threshold is indicated by a broken line at 5 × 10⁻⁸.

FIG 2

The CUL2-CREM-CCNY region and GTEx results of known expression quantitative trait loci (eQTL). The top graph shows both association results from individual studies (DBC [gray] and PROVIDE [blue]) as well as the meta-analysis of GWAS results (red) in this region. The middle graph shows GTEx eQTLs as vertical lines for GWAS variants with P  < 0.01 within the meta-analysis. The vast majority of overlapping associated SNPs from the meta-analysis are eQTLs for CREM, indicating a role for CREM, instead of CUL2 or CCNY. The bottom graph shows the recombination rate (centimorgans per megabase [cM/Mb]) for this region, illustrating the large block of linkage disequilibrium in this region.

FIG 3

CREM is activated by E. histolytica secreted factors, and CREM expression correlates with E. histolytica clearance in mice. (a) E. histolytica activates CRE-driven gene transcription in intestinal epithelial cells. HCT116 intestinal epithelial cells transfected with a cAMP response element (CRE)-driven reporter exposed to E. histolytica trophozoites (E.HIS) or 10 μm forskolin (FSK) as a positive control for CRE activation. Five biological replicates per condition were assayed. The mean and standard errors of percent CRE induction are shown. (b) Silencing CREM decreases CRE reporter induction by soluble E. histolytica, indicating that CREM is the major transcriptional regulator acting at CREs in response to amebic secreted products. CREM was silenced by 96% compared to GFP controls by quantitative PCR (qPCR). Three biological replicates per condition were assayed. The means plus standard errors (error bars) of percent CRE induction are shown. P values were calculated by unpaired test. Eh SP, E. histolytica secreted product. (c) CREM expression is induced in early infection in mice. Mice with a wild-type humanized leptin receptor gene (WT) or susceptible leptin receptor (Q223R) were infected with E. histolytica trophozoites by intracecal injection and sacrified at 12 h or 72 h postinfection. E. histolytica infection was measured by culture of cecal contents. CREM gene expression data from microarray analysis are shown. (Data from Mackey-Lawrence et al. [12]).

FIG 4

CREM^−/− mice (knockout [KO]) are more susceptible to amebic colitis. (a) Mice were infected with E. histolytica and euthanized on the basis of clinical score or at 72 h postinfection. Infection was assessed by the presence of live amebae upon culturing of cecal contents in growth medium. The numbers above each bar indicate the number of culture-positive mice to the total number of mice in that group. Data shown are from two independent but identical experiments. The superscript a for the KO value indicates that two mice in this group died before the 72-h time point. (b) Fixed cecal tissue was stained using antibody against active caspase-3, a marker for apoptosis (dark brown staining indicated by arrows). Representative images are shown. (c) Apoptotic death of cecal intestinal epithelial cells was scored by investigators blind to treatment groups (red circles indicate culture-positive samples). Error bars represent standard errors of the means (SEM). CREM^−/− mice from Blendy et al. (22) lack DNA binding domain 1a-1b and are CREM and ICER deficient.