The effect of different strains of Helicobacter pylori on platelet aggregation

1Department of Clinical Pharmacology, Royal College of Surgeons in Ireland, Dublin, Ireland; 2Wolfson Digestive Diseases Centre and Institute of Infection, Immunity and Inflammation, University Hospital, Nottingham, United Kingdom; 3Department of Medical Microbiology, Dermatology and Infection, Division of Bacteriology, Lund University, Lund, Sweden; 4Department of Gastroenterology, Beaumont Hospital, Dublin, Ireland; 5Division of Gastroenterology, Department of Medicine, Vanderbilt University School of Medicine, Nashville, Tennessee, USA; 6Division of Gastroenterology, Vancouver General Hospital, University of British Columbia, Vancouver, British Columbia Correspondence: Dr Michael F Byrne, Division of Gastroenterology, Academic Ambulatory Care Centre, Level 5, 5135–2775 Laurel Street, Vancouver General Hospital, Vancouver, British Columbia V5Z 1M9. Telephone 604-875-5640, fax 604-875-5447, e-mail michael.byrne@vch.ca Received for publication May 4, 2006. Accepted August 15, 2006 PA Corcoran, JC Atherton, SW Kerrigan, et al. The effect of different strains of Helicobacter pylori on platelet aggregation. Can J Gastroenterol 2007;21(6):367-370.

H elicobacter pylori plays a significant role in the pathogene- sis of peptic ulcer disease, gastric carcinoma and primary B cell gastric lymphoma.However, only a minority of individuals infected with H pylori develop a clinically significant outcome, such as peptic ulcer disease or gastric cancer.There is ongoing interest in identifying H pylori virulence factors that may predict the risk for symptomatic clinical outcomes.Experience with other bacterial pathogens suggest that H pylori strain-specific factors may influence the pathogenicity of different H pylori isolates.H pylori studies have primarily focused on two groups of putative bacterial virulence factors, the cag pathogenicity island (for which cytotoxic-associated gene A [cagA] is a marker) and the vacuolating cytotoxin (1,2).Some studies (3) have shown the formation of platelet aggregates in H pylori-infected patients, which may explain the association that has long been suggested between H pylori infection and coronary artery disease (CAD) (4-10) and stroke (11), although other studies (12)(13)(14) have failed to show any link.Recent work (15) has suggested that H pylori could function as a triggering factor in thrombotic thrombocytopenic purpura by inducing platelet aggregation through an interaction with von Willebrand factor (vWF).In addition, eradication of H pylori from the gastric mucosa has been associated with improvement of several systemic diseases, including immune thrombocytopenic purpura (ITP).A review of published studies of patients with ITP who underwent H pylori eradication revealed an overall response rate of 52% in 193 patients in whom H pylori was eradicated (16).One of the several suggested mechanisms for involvement of H pylori in ITP is induction of platelet aggregation.
None of these studies have shown any mechanisms by which H pylori might induce platelet aggregation.We have previously described the mechanism by which certain strains of H pylori interact with the main platelet receptor glycoprotein Ib/IX as well as plasma factors such as vWf and immunoglobulin (IgG) to induce platelet aggregation (17).To explore the potential clinical implications of our molecular findings, we investigated 21 different strains of H pylori varying in their pathogenicity factors and their ability to induce platelet aggregation.These results were compared with the disease presentation of the patient from which the bacterium was isolated.

Bacterial growth
H pylori strains were grown and maintained on H pylori agar plates (bioMérieux, France) in a microaerobic environment (Campy pak plus, Becton Dickenson, USA) at 37°C.After 48 h growth, H pylori strains were harvested directly from the plates, with sterile cotton buds, into phosphate buffered saline and resuspended to yield an optical density of 1.6 at 450 nm (approximately 4×10 9 bacteria/mL).Stock cultures were stored at -70°C in broth supplemented with 15% glycerol.

Bacterial strains
Helicobacter strains used in the present study were a kind gift from Dr John Atherton (University of Nottingham, UK) and Dr Torkel Wadstrom (Lund Univeristy, Sweden).Several strains of H pylori were used because the pathogenicity of H pylori varies depending on the presence or absence, among other factors, of the cag pathogenicity island (for which the gene cagA is a marker) and genotype of the vacuolating cytotoxin gene, vacA.Twenty-one strains of H pylori were assessed in the present study.Bacterial characteristics are as follows: A109, J258, J178, J254, J99 (cagA + , vacA s1/m1); J87, 25, 66, 1139, 17874, 17875, 915 (cagA + , vacA + ); J223, J226, J174, J182, J128 (cagA + , vacA s1/m2); 9366, J190, J63, J150 (cagA -, vacA, s2/m2).Bacterial strains were taken from patients who underwent routine upper gastrointestinal endoscopy for a variety of indications.An active peptic ulcer was defined as a circumscribed break in the mucosa, with apparent depth, measuring more than 1 cm in any dimension; an erosion was defined as a definite circumscribed break in the mucosa not fulfilling the criteria for an active ulcer.

Platelet preparation
Whole blood was drawn from healthy human volunteers who had abstained from nonsteroidal anti-inflammatory drugs and acetylsalicylic acid for 10 days.Nine volumes of blood were collected by clean venipuncture using a 19-gauge needle and added to one volume of 3.8% sodium citrate.Platelet-rich plasma (PRP) was prepared through centrifugation of anticoagulated whole blood at room temperature at 150 g for 10 min.The uppermost layer rich in platelets, PRP, was transferred into a 15 mL Greiner tube.The remaining blood was centrifuged at room temperature at 630 g for 10 min to yield platelet-poor plasma.

Platelet aggregation
Platelet aggregation was assayed by light transmission at 37°C using a platelet aggregometer (Bio/Data Corporation, USA).Platelets (suspensions of 500 μL) were warmed to 37°C, and the aggregation response was measured in siliconized cuvettes with continuous stirring at 1200 rpm.All experiments were repeated on at least three occasions using platelets from three different donors and were completed within 4 h of blood sampling.Platelets were tested for normal responses using arachidonic acid (0.5 mg/mL), and/or ADP (2 μM) before each experiment.Fifty microlitres of H pylori bacterial suspension (4×10 9 bacteria/mL of phosphate buffered saline) were added to 450 μL of human PRP.
Platelets respond immediately (within 10 s) to soluble agonists such as ADP.However, with bacteria, there is a lag time to aggregation that is dependent on the type of bacterium.Staphylococcus aureus has a rapid lag time of 2 min to 3 min, while Streptococcus sanguis can have a lag time of approximately 18 min.Not aggregating is usually defined as having a lag time of greater than 30 min.

RESULTS
Of the 21 strains tested, 10 strains induced platelet aggregation.There is no normal value in platelet aggregation; some bacteria induce platelet aggregation, while others do not.For aggregating strains, platelet aggregation took place within 5 min to 6 min.
The ability to induce platelet aggregation is summarized for each of the 21 strains tested in Table 1.There are no normal values in platelet aggregation.Unstimulated PRP generates an aggregation response of 0%; although a small amount of drift can take place especially over a long period.In the present study, we defined aggregation as a 10% change in light transmission.Thus, when bacteria are added to PRP some have no effect and others induce aggregation.In the present case, approximately 50% of the clinical strains induced aggregation.
Data are presented as mean ± SD from experiments performed in triplicate.Figure 1

represents a sample tracing showing induction of platelet aggregation for four of the 21 strains tested.
There was no correlation between the ability to induce aggregation and the type of gastroduodenal disease present in the patient (Table 2).

DISCUSSION
The injury induced by H pylori in the gastroduodenal tract may be due in part to inflammation and thrombosis within the gastric microvasculature.Numerous studies (18,19) have shown platelet activation in animal and human models.In vivo studies in the rat gastric mucosa have characterized the response of microcirculation to extracts of H pylori, including upregulation of neutrophil adhesion molecules, subsequent neutrophil adhesion and extravasation (20,21), and platelet aggregation (18).Platelets can further act on the microcirculation by releasing compounds such as serotonin, plateletactivating factor, thromboxane B 2 and P-selectin, recruiting more neutrophils (19).The occlusion of microvessels by platelet thrombi may lead to mucosal infarction and necrosis, as well as the release of bactericidal peptides from the platelet.These activated platelets would lead to local tissue occlusion, ischemia, generation of cytotoxic products and, eventually, ulceration.Indeed, circulating platelet aggregates and activated platelets have been detected in patients infected with H pylori (18).Whether this is a direct effect of H pylori on platelets or secondary to vascular injury is unknown.
Recent studies have shown both negative and positive associations between H pylori and CAD.Although there are recent studies that suggest that the eradication of H pylori does not have any influence in the treatment of ITP (22), others have shown that eradicating H pylori is associated with improvement of several systemic diseases, including ITP (16).The role of bacterial infection in cardiovascular disease is, at present, an area of much debate.Recent work suggests that pathogen burden is a good predictor of outcome in cardiovascular disease patients.In animal models, exposure to certain pathogens is more effective at inducing atherosclerosis than a high cholesterol diet.The evidence for specific pathogens such as H pylori is weaker, but some studies have shown an association.Equally, the data supporting a role for H pylori in thrombocytopenia are intriguing but far from conclusive.One difficulty is that studies have failed to take into consideration the importance of a prothrombotic phenotype, especially because we have shown that this may only have a 50% prevalence.
Platelet activation is known to contribute to the development of acute phases of myocardial infarction in patients with CAD (23).Some have suggested that H pylori infection may contribute to the development of thrombocytopenia and other cardiovascular diseases by promoting platelet consumption (24).Several theories have attempted to explain how a noninvasive organism can affect platelets, including molecular mimicry and platelet aggregation, but none have been proven.We have previously described a clear molecular mechanism by which certain strains induce platelet aggregation, namely, by binding vWf and inducing glycoprotein Ib-and FcγRIIadependent platelet aggregation in the presence of H pylori antibodies (17).Our present study of 21 strains confirmed that this is potentially a clinically important interaction because almost one-half of the strains studied induced platelet aggregation.We found that 10 of the 21 strains of H pylori induced platelet aggregation.We speculate strongly that the mechanism for platelet-induced aggregation is uniform for all strains of H pylori, because anti-IIb/IIIa and anti-vWf antibodies inhibited a random test of one of the H pylori strains used in the present study (data not shown).We also speculate that this was the uniform mechanism for all five strains in our previous mechanistic study (17), but we accept the valid criticism that   The relative risk is 1.6 (95% CI 0.6 to 4.2) and is nonsignificant (Fisher's exact test) Figure 1) Percentages of platelet aggregation for four different strains of Helicobacter pylori.Numerical lines (1-4) represent strains J179, J226, 6320 and J254, respectively.The percentages of induced platelet aggregations were 1%, 8%, 2% and 29%, respectively.The time was measured in 1 min intervals with the baseline starting at 4 min; overall aggregation took place approximately within 5 min to 6 min.Lag time is defined as the time taken from the addition of bacteria to platelet-rich plasma until the first recognizable signs of aggregation.A range of lag times was observed with the strains used, but the lag time was consistent for each strain not all strains were tested in this way in the present study.The variable results in studies of ITP and other vascular associations of H pylori may be explained, to some degree, by the finding in our present study that not all strains of H pylori activate platelets.For some unknown reason, H pylori tends to induce an aggregation response that is weaker than that seen with certain strains of S aureus or S sanguis, possibly because a different activation method is involved.
Our study findings have some potentially important clinical implications.For example, the response of ITP to H pylori eradication may depend on the infecting strain (12,25).The phenotype of the infecting strain may be important in the outcome of the disease process of ITP.The difference in the clinical response to eradication therapy warrants further investigation into differences in the aggregating phenotype of H pylori.There are similar implications for the other vascular associations of H pylori, such as CAD.Whether H pylori plays a role in CAD, either via platelet aggregation or other methods, may be very strain dependent, further confounding the current confusion.

SUMMARY
Ten of 21 strains of H pylori used in the present study induced platelet aggregation.The clinical relevance of this interaction remains to be clarified, but an increasing number of studies are suggesting a real association with changes in platelet function.The strain dependency in our in vitro model may explain some of the conflicting results with H pylori infection and extraintestinal diseases such as ITP and CAD, disease processes that may be partially dependent on the ability of the infecting strain(s) to induce platelet aggregation.
Effect of Helicobacter pylori strains on platelet aggregation Can J Gastroenterol Vol 21 No 6 June 2007 369

TABLE 1
Twenty-one strains of Helicobacter pylori were tested for their ability to induce platelet aggregationThe strains were derived from patients with benign infection or disease pathologies such as gastric ulcer (GU) and duodenal ulcer (DU).cagA Cytotoxic-associated gene A; vacA Vacuolating toxin gene