Helicobacter pylori infection in Africa and Europe: enigma of host genetics

Helicobacter pylori infection is one of the most common bacterial infections. The prevalence varies from 25–50% in developed countries to 70–90% in the third world.¹ Despite improved treatment modalities, H pylori related gastrointestinal pathology, in common with gastritis, peptic ulcers and consecutive bleeding events, gastric MALT lymphoma, or carcinoma, remains a major burden on Western health systems. In the USA, approximately four million people have active peptic ulcers and about 350 000 new cases are diagnosed each year. Four times as many duodenal ulcers as gastric ulcers are diagnosed.² Epidemiological evidence suggests that both infection with H pylori and the consecutive development of clinically relevant pathology are influenced by genetic predisposition as only a fraction of exposed individuals develop infection and likewise a fraction of infected individuals develop ulcers or even gastric cancer.³

Thye et al used H pylori reactive serum immunoglobulin G as a marker of H pylori infection in Senegalese siblings and provided for the first time concrete statistical evidence for a genetic predisposition to H pylori infection. The authors reported an association between IFNGR1 polymorphisms and high antibody concentrations.⁴ Inclusion of the three variants (H318P, L450P, −56 T/C) in the linkage analysis increased the LOD score to 4.2. The two African amino acid exchange variants, H318P and L450P, were not found in 100 unselected Germans.⁴

Immediately, the question arises of whether variation in the interferon γ receptor 1 (IFNGR1) locus is related to H pylori infection or pathology in Caucasian populations We genotyped two polymorphisms at the IFNGR1 locus (rs608914, rs11914) in 344 H pylori infected individuals undergoing upper gastrointestinal endoscopy from northern Germany and 311 healthy blood donors. H pylori infection was tested by rapid urease test from a gastric biopsy or histology. Patients were grouped according to the severity of the mucosal inflammation, ranging from mild inflammation such as gastritis or duodenitis, to erosions and ulcer disease. Polymorphisms were selected from the Applied Biosystems “Assay on Demand” service (https://store.appliedbiosystems.com) and genotyped by Taqman using standard protocols. Because both polymorphisms were non-functional single nucleotide polymorphisms (rs11914: synonymous T/G exchange in exon 1, frequency in blood donors 13.5%; rs608914: C/T exchange about 6.5 kb downstream of the transcriptional start site, frequency in blood donors 31.3%) a haplotype case control analysis was performed using Hapmax⁵ to assess the association of the locus with the respective phenotypes. The markers exhibited a low degree of linkage disequilibrium (LD) (D′ = 0.174) yielding a highly informative haplotype analysis of the locus (frequencies in normal controls: TC 0.586; TT 0.100; GC 0.279; GT 0.035). No significant association with infection status or severity of H pylori associated inflammation was found (table 1 ▶).

Table 1.

Haplotype analysis of infection status and clinical manifestation of Helicobacter pylori infection

Comparison groups	n (groups)	p Value
Infection status (normal controls versus all H pylori positive patients)	311 v 344	0.39
Moderate versus mild pathology in H pylori infected patients (gastric/duodenal erosions versus no pathology or gastritis/duodenitis)	66 v 166	0.33
Severe versus mild pathology in H pylori infected patients (gastric/duodenal ulcers versus no pathology or gastritis/duodenitis)	112 v 166	0.61

The table shows the comparative frequencies of the IFNGR1 haplotype described above.

Susceptibility to H pylori infection was tested by comparison of all H pylori positive patients (n: all subgroups: 66+112+166 = 344) against normal controls (top row). Genetic predisposition for complications of H pylori infection was tested by comparison of patients with moderate pathology (gastric or duodenal erosions, n = 66) and severe pathology (gastric or duodenal ulcers n = 112) against patients with mild or no pathology grouped together (no pathology, gastritis, or duodenitis, n = 166). Significance was assessed by a χ² test of the global likelihood ratio of the case control haplotype estimations.

We conclude that INFGR1 is unlikely to be involved in the aetiology of H pylori infection or the development of clinical sequelae in German Caucasians. This may be due to aetiological differences between African and Caucasian individuals, as suggested pathophysiologically by Mitchell et al, who demonstrated major differences in the IgG subclass response to H pylori infection in the first and third world.⁶ In relation to clinical disease manifestations, the IFNGR1 locus may affect antibody concentrations but not the clinical course of H pylori infection in Caucasians. Alternatively, other immunoregulatory genes in the vicinity of the IFNGR1 locus such as the interleukin 20 receptor α (200 kb distance) or MAP kinases 5 (600 kb distance) could harbour the causative variants. High densitiy LD mapping of the locus is required to unravel the causative genetic variants in both African and Caucasian populations. Our data support the hypothesis that the genetic diversity of the host immune system may contribute to the differences in H pylori prevalence and clinical outcome in African and Caucasian populations.