Helicobacter suis : further characterization of the agent and its possible impact on the course of Parkinson's disease

Helicobacter pylori (H. pylori) is a well-known gastric Helicobacter species infecting approximately 50% of the global human population. H. pylori can persist in the human stomach for an entire lifetime if the infection is left untreated. Infections with H. pylori often remain subclinical, yet they may also result in varying degrees of dyspepsia. Approximately 15% of H. pylori-infected individuals will develop severe gastric disease, including peptic ulcer disease, mucosa-associated lymphoid tissue (MALT) lymphoma and adenocarcinoma. Non-Helicobacter pylori Helicobacter (NHPH) species naturally colonize the stomach of animals, but some of them can also infect the human stomach. The most prevalent NHPH species in the human stomach is H. suis, which naturally colonizes the stomach of pigs and non-human primates. Non-human primates probably constitute the original hosts of H. suis, and a host jump from non-human primates to pigs most likely happened between 75,000 and 10,000 years ago. Human H. suis strains are shown to be genetically closely related to porcine H. suis strains, suggesting that pigs may constitute a source for human H. suis infections. Transmission from pigs to humans might occur through direct or indirect contact with H. suis infected pigs, their saliva, and stomach content, but also through the food chain, as it was shown that H. suis can be present and survive in minced pork. NHPHs have been detected in 0.2 to 6% of gastric biopsies from human patients with severe gastric complaints. NHPH infections have been associated with gastritis, gastric and duodenal ulcers, and MALT lymphoma in humans. However, due to their fastidious nature, NHPHs are difficult to isolate and cultivate in vitro, which has hampered the investigation of their pathogenesis in humans. Besides, diagnosis of NHPH infections is difficult and most techniques used for H. pylori diagnosis lack sensitivity and/or are not able to discriminate between different gastric Helicobacter species. It is therefore possible that the prevalence of NHPHs in humans is currently underestimated. Mounting evidence suggests that the effects of a gastric infection with H. pylori or H. suis are not limited to the stomach. H. pylori infections have already been associated with several extra-gastric manifestations, among which Parkinson’s disease (PD). Infections with H. pylori have not only been associated with an increased risk of PD, but also with worse motor function. However, contrasting results on a potential beneficial effect of H. pylori eradication on PD-related motor function have been described. A potential role for H. suis in PD was also suggested. An exceptional high frequency of H. suis DNA was found in gastric biopsies from patients with idiopathic PD (27%) in comparison to a control group (2%). Eradication of the H. suis infection in a PD patient did not only result in a resolution of the gastritis, but also induced an overall improvement of PD-related symptoms. Remarkably, the presence of H. heilmannii (i.e. a zoonotic cat/dog-associated NHPH species) DNA in gastric biopsy samples was not different between idiopathic PD patients and the control group (in both groups approximately 2%). A recent mouse experiment confirmed extra-gastric effects caused by H. suis. After 1 month of infection, H. suis induced gastric inflammation, disruption of the gastric mucosal barrier and low-grade systemic inflammation. Furthermore, disruption of the blood-cerebrospinal fluid barrier, brain inflammation, and associated loss in short-term memory were demonstrated. Adherence to the gastric mucosa is the initial step in gastric colonization and pathogenesis of Helicobacter infections. As mentioned, little is known on the pathogenesis of NHPH infections in humans. Besides, prediction of the function of genes of NHPHs is generally based on extrapolation of what is known for H. pylori. In vitro cell culture experiments may provide initial and valuable information concerning specific genes playing a role in gastric colonization and pathogenesis of NHPHs in humans. In Chapter 1, we tried to gain insights into potential virulence factors of H. suis and H. heilmannii playing a role in the adherence to the human gastric mucosa. Using RNA sequencing, we investigated bacterial genes that are differentially expressed in a H. suis (HS1) and H. heilmannii (ASB1) strain when adhered to the human gastric epithelial cell line MKN7 in comparison to non-adhered H. suis and H. heilmannii bacteria. A list of 134 (83 up-regulated and 51 down-regulated) differentially expressed genes (padj ≤ 0.01; fold change ≥ 2) was identified for the adherent H. suis strain HS1 and a list of 143 (60 up-regulated and 83 down-regulated) differentially expressed genes (padj ≤ 0.01; fold change ≥ 2) was identified for the adherent H. heilmannii strain ASB1. Differentially expressed genes of the H. suis and H. heilmannii strains belonged to multiple functional classes, indicating that adhesion of both strains to human gastric epithelial cells evokes pleiotropic adaptive responses. According to BLASTp analyses, only 2 genes (i.e. the urease accessory protein gene (UreF) and ferrochelatase gene) were commonly up-regulated in H. suis (HS1) and H. heilmannii (ASB1). Four genes (i.e. recombinase A gene (RecA), tellurium resistance gene (TerD) of H. heilmannii with its H. suis homolog general stress protein 16U gene, KH domain RNA binding protein gene (YlqC) of H. heilmannii with its H. suis homolog hypothetical protein gene, and heat shock protein gene (GrpE)) were commonly down-regulated in both pathogens according to BLASTp analyses. These findings suggest that the porcine H. suis strain HS1 and the feline H. heilmannii strain ASB1 use distinct pathways when adhering to human gastric epithelial cells. Indeed, several differences between H. suis (HS1) and H. heilmaniii (ASB1) were found when comparing up- and down-regulated genes implicated in bacterium-host interactions (i.e. Urease genes, NikB, flagella encoding genes, tumor necrosis factor α inducing protein gene (Tipα), gamma-glutamyl transpeptidase (Ggt), outer membrane protein (OMP) genes, and the gene encoding peptidyl-prolyl cis, trans-isomerase (ppi)). Genes that are significantly up-regulated upon adherence to gastric epithelial cells might play a role in gastric pathogenesis and in the ability of Helicobacter to have effects beyond the stomach as well. To further analyze the role of the differentially expressed genes in human gastric colonization and virulence, H. suis and H. heilmannii mutants lacking these genes should be created and tested using in vitro and in vivo models. It should, however, be kept in mind that we only included 1 H. suis strain and 1 H. heilmannii strain in our study, and that the neoplastic background of MKN7 cells (i.e. cell line derived from a well differentiated gastric tubular adenocarcinoma) may have resulted in different receptor phenotypes on H. suis and H. heilmannii in comparison to the in vivo situation. The potential role of H. suis in PD is currently unclear. Based on the finding of an earlier study whereby the eradication of H. suis led to an improvement of both the gastric and neurological symptoms in a PD patient, we hypothesized that a gastric H. suis infection might aggravate PD-related symptoms, such as motor dysfunction, and possibly PD-related pathology. To investigate our hypothesis, we used the 6-hydroxydopamine (6-OHDA) PD mouse model. The 6-OHDA model is often used as an experimental model for PD, whereby 6-OHDA exerts neurotoxic effects on catecholaminergic (i.e. dopaminergic and noradrenergic) neurons. 6-OHDA injection in the brain striatum causes degeneration of dopaminergic neurons, a major pathological hallmark seen in PD resulting in motor dysfunction. Two animal experiments with H. suis infected mice (acute and chronic infection) were performed (Chapter 2). Mice were stereotactically injected in the left brain striatum with either 6-OHDA or the vehicle of 6-OHDA at 21 days (experiment 1) or 500 days after intragastric H. suis inoculation (experiment 2). Control groups receiving only the growth medium of H. suis were included, resulting in 4 groups of mice in both experiments, i.e. 1) control-vehicle, 2) control-6-OHDA, 3) H. suis-vehicle and 4) H. suis-6-OHDA. The mice were subjected to several behavior and motor function tests (i.e. traversal beam, pole, footprint analysis, cylinder) before intragastric inoculation (only in experiment 1), before intrastriatal injection, and before euthanasia. All animals were euthanized at 7 days after intrastriatal injection. Samples of stomach, blood, and brain were taken for further analysis. After 28 days of H. suis infection (experiment 1) and 507 days of H. suis infection (experiment 2), colonization with H. suis was confirmed and inflammation in the corpus and antrum of the stomach was demonstrated. RT-qPCR revealed a significantly decreased expression of the tight junction genes zonula occludens (Zo-) 1 and Zo-3 in the corpus of the stomach after 28 days of H. suis infection (experiment 1). After 507 days of H. suis infection (experiment 2), a significant increased Muc13 expression and a significant decreased expression of the tight junction genes claudin (Cldn) 5 and Zo-1 in the corpus of the stomach was demonstrated. However, no significant changes in gastrointestinal barrier permeability could be demonstrated using the 4 kDa fluorescein isothiocyanate-dextran leakage assay in the H. suis-infected mice in comparison to the control mice in both experiments. Tyrosine hydroxylase staining on brain sections showed a lower % loss of dopaminergic neurons in the H. suis–6-OHDA groups compared to the control–6-OHDA groups. Correspondingly, just prior to euthanasia, motor function of the H. suis–6-OHDA groups was better compared to the control–6-OHDA groups. In both experiments, no significant changes in behavior and motor function between the 4 groups could be demonstrated before intrastriatal injection, suggesting that the gastric H. suis infection alone did not result in behavior and/or motor function alterations. Our study shows that a gastric H. suis infection protects, at least partially, against 6-OHDA-induced dopaminergic cell loss and motor function impairment in a left unilateral intrastriatal 6-OHDA PD mouse model. This potential protective effect of H. suis against 6-OHDA-induced toxicity and motor function impairment needs further elucidation. There are currently not many diagnostic tools available for reliable diagnosis and species identification of NHPH infections in humans. PCR in combination with sequencing of positive PCR amplicons is currently preferred for detection and differentiation of gastric NHPHs. However, this method is labor-intensive, time-consuming, and expensive. Therefore, we investigated whether matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry (MALDI-TOF MS) might constitute a potential tool for future diagnosis of NHPH infections (Chapter 3). This technique is based on the generation of complex fingerprints of specific biomarker molecules by measuring the exact mass/charge ratio of peptides and proteins. Inadequate MALDI-TOF MS identification may occur due to the incompleteness of mass spectrometry databases. The current MALDI Biotyper reference database library only contains 24 Helicobacter entries, most of which belonging to the enterohepatic Helicobacter species and only 7 belonging to gastric species (i.e. only H. pylori strains). Therefore, we identified 93 gastric Helicobacter isolates of 10 different NHPH species using MALDI-TOF MS in order to establish a more elaborate Helicobacter reference database. While the MALDI Biotyper database was not able to correctly identify any of the isolates, the newly created in-house database correctly identified all individual mass spectra and resulted in 82% correct species identification based on the two highest log score matches (with log scores ≥ 2). However, spectra obtained under different growth conditions, for example dry versus biphasic growth conditions, may differ, and agar medium-related peaks may influence reliability of the obtained results, as observed for 4 NHPH species (i.e. H. bizzozeronii, H. cetorum, H. felis, and H. salomonis). In addition, a dendrogram was constructed using all newly created main spectrum profiles. Nine main clusters were formed, with some phylogenetically closely related Helicobacter species clustering closely together and well-defined subclusters being observed in specific species, such as H. suis. Our results suggest that MALDI-TOF MS allows rapid differentiation between gastric Helicobacter species, provided that an extensive database is at hand and variation due to growth conditions and agar medium-related peaks are taken into account. However, further research to enable the use of MALDI-TOF MS for diagnosis of Helicobacter infections in clinical practice and to investigate the sensitivity of this technique on clinical gastric samples is needed. Since H. suis infections have been associated with severe gastric disease and potentially with PD in humans, and because H. suis eradication has been shown to result in significant improvement of both the gastric and PD-related symptoms, effective treatment of H. suis infections in human patients is of great importance. However, not much is known on the antimicrobial susceptibility pattern of gastric NHPHs, including H. suis. Therefore, H. pylori eradication treatment schemes are currently used to eradicate NHPH infections in humans. To improve treatment strategies of H. suis infections in humans, we determined the in vitro susceptibility of 35 H. suis isolates (i.e. 20 porcine isolates and 15 non-human primate isolates) to 15 antibiotics (Chapter 4). A monomodal distribution of minimal inhibitory concentrations (MICs) was seen for β-lactam antibiotics, macrolides, gentamicin, neomycin, doxycycline, metronidazole, and rifampicin. Presence of a bimodal distribution of MICs indicated that 2 porcine isolates (i.e. HS6 and HS10) did not belong to the wild-type population for fluoroquinolones. This was also the case for 1 porcine isolate (i.e. HS4) for tetracycline, 1 porcine (i.e. HS5) and 2 primate isolates (i.e. HSMm R04052c and HSMm R07055b) for lincomycin, and 1 primate isolate (i.e. HSMm R07055b) for spectinomycin. Single nucleotide polymorphisms (SNPs) were present in the gyrA gene of the isolates not belonging to the wild-type population for fluoroquinolones and in ribosomal protein encoding genes of the isolates not belonging to the wild-type population for tetracycline and spectinomycin. MICs of ampicillin, tetracycline, and doxycycline were higher for porcine H. suis isolates compared to primate isolates and in these porcine isolates SNPs were detected in genes encoding penicillin binding and ribosomal proteins. Taken together, our study indicates that acquired resistance occasionally occurs in H. suis isolates and that porcine isolates may be intrinsically less susceptible to aminopenicillins and tetracyclines than primate isolates. Since human H. suis infections have been associated with porcine H. suis strains, the intrinsically lower susceptibility to aminopenicillins and tetracyclines, and the occasional presence of acquired resistance should be taken into account when treating H. suis infections in human patients