Embryonal Cardiotoxicity of the Helicobacter pylori Lipopolysaccharide

M. A. Panaro*¹, L. Amati², M. Sisto ¹, L. Caradonna², S. Lisi¹, V. Mitolo¹ and D. Fumarola³

¹ Department of Human Anatomy and Histology, University of Bari, Bari, Italy; ² Scientific Institute for Gastroenterological Diseases, Castellana Grotte (Bari), Italy; and ³ Department of Internal Medicine, Immunology and Infectious Diseases, University of Bari, Bari, Italy

*Address correspondence to this author at the Department of Human Anatomy and Histology, University of Bari, Policlinico, piazza Giulio Cesare, I-70124 - Bari, Italy [e-mail: ma.panaro@anatomia.uniba.it]

ABSTRACT:Helicobacter (H.) pylori is the causative agent of the peptic ulcer disease and a co-factor in the development of gastric malignancies. Recently, it has been maintained that chronic H. pylori infections in adults are linked to a higher risk of coronary heart diseases. In this respect, the acute toxic effects of the H. pylori lipopolysaccharide (LPS) on embryonal cardiomyocytes at different developmental stages was evaluated. White Leghorn chick embryos and  smooth (S) - form  NCTC 11637 strain H. pylori organisms were used. Both whole heath- killed H. pylori suspensions (3_10⁶ bacteria/egg) and isolated S-LPS (500 ng/egg) or S-Lipid A (500 ng/egg) were non-lethal to 4-day embryos, becoming moderately lethal (5% to 30%) to 6- and 8-day embryos and highly lethal (> 90%) to 10- to 17-day embryos. The contractile activity of isolated atrial fragments from 10-day embryos was completely inhibited, within 5 min, following treatments with heath-killed H. pylori (3x10⁶/ml), or S-LPS (500 ng/ml), or S-Lipid A (500 ng/ml); the block determined by S-LPS and S-Lipid A was irreversible, while the block by bacterial suspensions was completely reversible upon withdrawal. Following a 24-hour treatment with S-LPS or S-Lipid A of single-cell cultures of cardiomyocytes (isolated from 10-day embryos) a dose-dependent cell loss was observed, as assessed by total protein dosage and direct counting of adherent cells. Propidium Iodide/Annexin V FACS-analysis confirmed the occurrence of cellular necrosis, but did not show any evidence of apoptotic processes. The release of superoxide anion radicals by cultured cardiomyocytes was as follows: S-Lipid A (25 mg/ml) > S-LPS (25 mg/ml) > heath killed H. pylori suspensions (3x10⁶/ml); control cultures did not release detectable amounts of superoxide anion radicals. Furthermore, cultured cardiomyocytes produced increased amounts of NO (N-monomethylarginine-inhibitable) following stimulation with S-LPS (25 mg/ml) or S-Lipid A (25 mg/ml) (but not heath killed H. pylori 3x10⁶/ml suspensions). Under all the above experimental conditions S-polysaccharide proved to be non-toxic. Concluding, H. pylori LPS is relatively non-toxic to the less differentiated cardiomyocytes; cardiomyocytes which are more advanced in their biochemical differentiation become highly sensitive to LPS and produce ROS and NO. ROS are probably responsible for the early toxic actions, while both ROS and NO are likely to be involved in the later degenerative/necrotic effects.

 

INTRODUCTION

     The gram-negative bacterium Helicobacter (H.) pylori is the causative agent of chronic superficial gastritis in humans and the principal microorganism responsible for the development of peptic ulcer disease; furthermore, it is thought to be an important co-factor in the onset of gastric malignancies [1-3]. Recently, different clinical extradigestive disturbances have been linked to chronic H. pylori infections, including cardiovascular disorders (such as ischemic heart disease, primary Raynaud phenomenon, and headache) and autoimmune diseases such as Henoch-Schönlein purpura and Sjögren's syndrome [4, 5]. The relationships between chronic H. pylori infections and coronary heart disease have been extensively investigated and several reports have confirmed the existence of a positive epidemiological correlation [4-9]. Different indirect pathogenetic mechanisms have been advocated to explain the increased cardiovascular risks: release of acute-phase reactants including fibrinogen, reduction of HDL cholesterol or alterations in the serum lipid concentrations, elevation of homocysteine levels, immunological cross-reactivity between H. pylori bacteria and human heat shock proteins, atherogenesis through a low-grade persistent inflammatory stimulation, and platelet activation and aggregation [7-11]. However, other recent reports maintain that H. pylori infection is probably not an important contributor to coronary heart disease events [12-15].

     The biology of H. pylori and the actions of its products on eukaryotic cells have been studied with special reference to the bacterial "virulence factors" (reviewed in [16-18]). Among these, urease and the vacuolating toxin A (vacA) seem to be more directly involved in the pathogenicity of the bacterium. Urease reacts with the gastric urea and generates ammonia which combines with water and yields ammonium hydroxide, a chemical species highly toxic to cells [16]; vacA, after been endocytosed, determines cytoplasmic vacuolation and eventually cell necrosis [19]. Hence, urease and vacA are thought to be the main agents responsible for the lesions at the gastric epithelium level.

     The cellular toxicology of the bacterial lipopolysaccharides (LPSs) and their components has been less extensively investigated, although LPS is held to contribute to the cardiovascular collapse and death observed in patients with sepsis [20]. Direct in vitro studies on myocardial preparations or single cell suspensions derived from Escherichia coli LPS-treated rabbits and guinea pigs demonstrated a reduced shortening of unloaded cells (with decreased action potential duration), as well as lower peak rates of shortening and relengthening [21-23]. Following longer in vitro exposures, E. coli LPS may cause cytoplasmic damage and cell necrosis and, under certain conditions, may trigger an apoptotic program [24-26].

     E. coli LPS is reported to exert seemingly direct effects on some of the ion channels of the cardiomyocytes, resulting in decreased free intracellular Ca²⁺ concentrations ([Ca²⁺]_i) [27]; this would be due mainly to the reduced Ca²⁺ influx through L-type Ca²⁺ sarcolemmal channels, which plays a crucial role in cardiac excitation-contraction coupling [23]. Furthermore, in microglia cells LPS is reported to induce the expression of outward rectifying potassium currents [28, 29].

     Besides the above “direct” effects, LPS is able to start different transduction signaling cascades. According to Comstock et al. [25] LPS ligates the cardiac LPS-receptor, CD14, and the latter triggers an enhanced production of TNF-a (which exerts negative inotropic effects by itself); TNF-a would be eventually responsible for the cellular apoptosis. Furthermore, it appears that LPS is able to modulate the expression of its receptor: the murine heart normally expresses low levels of CD14 mRNA, but after intraperitoneal injection of LPS, induction of CD14 gene expression is considerably enhanced and CD14 antigen colocalizes to the cytoplasm of cells expressing CD14 mRNA [30]. However, according to Cowan et al. [31], LPS may induce the production of proinflammatory cytokines (that cause myocardial dysfunction), by activating specific sets of kinases through a CD14-independent mechanism. IFN-g is thought to be also an important mediator of LPS toxicity [32]. In addition, LPS is reported to stimulate the production of NO (through an inducible nitric oxide synthase (iNOS)): NO, in turn, exerts negative inotropic, negative chronotropic and [Ca²⁺]_i-lowering effects [33] and plays an important role in modulating the L-type calcium currents [34].

     To our knowledge no specific study addressed the cardiotoxicity of the H. pylori LPS during development and in the adult animal.

     Here we review, in an avian model, the acute toxic effects on the embryonal heart, at different developmental periods, of crude suspensions of heat-killed H. pylori organisms and the bacterial endotoxin. These effects have been studied in terms of block of the mechanical activity of the heart in situ or isolated myocardial fragments, and of cytotoxicity to cultured isolated cardiomyo-cytes. The production of Reactive Oxygen Species (ROS) and Nitric Oxide (NO) by cultured embryonal cardiomyocytes challenged with suspensions of heat-killed whole bacteria, LPS and Lipid A have also been evaluated.

     In summary, H. pylori LPS is especially toxic to the more differentiated cardiomyocytes, at intermediate and late developmental stages; H. pylori LPS-challenged cardiomyocytes produce both ROS and NO, the former being probably responsible for most of the observed toxic effects.

EXPERIMENTAL DESIGN

Bacterial Strain and LPS Chemotypes of H. Pylori

     The NCTC 11637 strain of H. pylori was derived from the National Collection of Type Cultures (NCTC), Public Health Laboratory Service, London, England. To obtain smooth-form (S) H. pylori, bacteria were grown in a broth of brain heart infusion (Oxoid, Basingstoke, England) containing 2% (v/v) fetal calf serum (Oxoid), as described by Moran and Walsh [35]. Different LPS chemotypes of the NCTC 11637 strain were a generous gift of Prof. Anthony P. Moran, University College Galway, Ireland.

     Both heat-killed H. pylori S-form suspensions (S-H.p.-SUSP) in PBS and the following products isolated from the bacterium, all dissolved in PBS: the whole LPS (S-LPS), its lipid moiety (S-Lipid A; S-LipA) either phosphorylated or dephos-phorylated, and its polysaccharide moiety (the S-Polysaccharide; S-Poly) were used. These H. pylori products had been isolated as indicated in Pece et al. [36].

Toxicity Tests in Ovo

     Fertilized White Leghorn hens eggs, incubated routinely, were used. At the third day of incubation, 3 ml of albumen were aspirated through a small opening drilled at the acute end of the egg shell and then a larger “window” was opened over the embryo; the window was closed with a coverslip sealed with paraffin. In a set of experiments, S-H.p.-SUSPs in phosphate buffered saline (PBS), with bacterial concentrations ranging from 10⁶/ml to 10⁸/ml, were used to treat chick embryos in the developmental days 4 to 17. One ml of the S-H.p.-SUSP was gently dropped directly over the chorioallantoic membrane. In another set of experiments, 4- to 17-day embryos were treated with different concentrations (ranging from 100 ng/ml to 5 mg/ml) of S-LPS, or S-LipA (phosphorylated or dephosphorylated), or S-Poly, all dissolved in PBS. In all experiments we administered 1 ml of the test solution directly on the chorioallantoic membrane. Control embryos were given 1 ml of PBS. The in ovo toxicity was estimated in terms of embryo lethality within 4 hours.

Toxicity Tests on Chick Embryo Heart Fragments

     Ten-day embryos were transferred to sterile Hanks' Balanced Salt Solution (HBSS) and the heart was isolated by microdissection. The atria were separated from the ventricles and minced with a blade into fragments 0.5 mm ca. in diameter, which were transferred to Minimum Essential Medium (MEM), supplemented with 10% fetal calf serum (FCS), glutamine (2 mM), streptomycin (100 mg/ml) and penicillin (100 U/ml) (complete MEM), and kept at 37°C, with 5% CO₂. The atrial fragments were free-floating and the level of medium was kept low in order to favor gas exchanges. Cardiac fragments were submitted to the toxicity tests 30 to 45 min after isolation. Using a low-power stereomicroscope or an inverted microscope the contraction rate of the fragments was recorded before treatments, after treatments (with either S-H.p.-SUSPs, or S-LPS, or S-LipA (phosphorylated or dephosphorylated), or S-Poly) at different concentrations, and after the withdrawal of the test substance. During the experiment the temperature of the medium was kept at 37°C.

Toxicity Tests on Isolated Cardiomyocytes

(a) Cell adherence tests

     Hearts from 10-day chick embryos were mechanical minced and then disaggregated in a 0.2% collagenase solution for 20 min at 37°C. The cells were gently pelleted and resuspended in complete MEM and plated at a density of 2x10⁵ per 15 mm well (Nunc, Roskilde, Denmark). Cultures were incubated in complete medium supplemented with bromodeoxyuridine (0.1 mM) to prevent proliferation of non-myocytes [37], at 37°C, in a water-saturated atmosphere with 5% CO₂. After 24 hours, the cells were treated by adding to the medium either S-LPS, or S-LipA (phosphorylated or dephosphorylated), or S-Poly, at different concentrations. 24 hours after treatment, the medium was discarded; the plastic-adherent cells were detached by ice-cold PBS and either counted or lysed to measure their total protein content (Bradford method). Control cultures (without treatment with bacterial products) were similarly processed. From the above data we estimated the degree of loss-of-adherence of treated cells (as compared to the controls), cell detachment being assumed as an index of the cytopathic effect.

(b) Annexin V /Propidium Iodide test (FACS)

     Cardiomyocytes, isolated and cultured for 24 hours as indicated before, were treated with 50 mg/ml of S-LPS or S-LipA for 18 hours. Then, an Annexin V-affinity assay was performed, in order to detect the surface exposure of phosphatidyl-serine, combined with Propidium Iodide (PI) staining, to test the membrane integrity [38]. Briefly, in both the experimental and control wells, the adherent cells were detached by ice-cold PBS and all the cells recovered (including the cells that had detached during incubation) were washed and resuspended at 2x10⁵/ml density. An aliquot of 195 ml of the cell suspension was incubated with 5 ml of Annexin V-FITC (Bender MedSystems, Vienna, Austria), for 10 min at room temperature. Then the cells were washed and stained with 1 mg/ml PI (Bender MedSystems) and submitted to bivariate FACS analysis (FACSCalibur, Becton Dickinson, San Jose, CA, USA).

Superoxide Anion Production

     Cardiomyocytes, isolated and cultured for 24 hours as indicated before, were treated for 2 hours with either S-H.p.-SUSPs, or S-LPS, or S-LipA (phosphorylated or dephosphorylated), or S-Poly, at different concentrations. The amount of superoxide anion radicals (O₂^-) released by the cardiomyocytes during the 2-hour treatment was measured by a cytochrome c reduction assay [39]. The concentration of O₂^- in culture supernatants was spectrophotometrically (Cary 50 Scan, Varian, Mulgrave, Australia) measured at 550 nm and expressed as nmol/ml of reduced cytochrome c.

Nitric Oxide Production

     Cardiomyocytes, isolated and cultured for 24 hours as indicated before, were treated for 96 hours with either S-H.p.-SUSPs, or S-LPS, or S-LipA (phosphorylated or dephosphorylated), or S-Poly, at different concentrations, with or without addition of the NO synthase inhibitor, N-monomethylarginine (L-N^GMMA). At the end of treatment the culture supernatants were collected and incubated with the Griess reagent (Carlo Erba, Milan, Italy) (1:1 v/v) for 10 min at room temperature [40]. The absorbance of supernatants was spectrophotometrically measured at 540 nm and the NO₂^- concentration was determined by a standard curve of NaNO₂ and expressed as mmol/ml. To avoid interference by nitrites possibly present in the medium, in each experiment the absorbance of the unconditioned medium was assumed as the “blank”.

IN OVO AND IN VITRO RESULTS

Toxicity in Ovo Experiments

     Four- to 17-day chicken embryos were treated either with PBS alone (controls) or the following H. pylori products: S-H.p.-SUSPs (killed bacteria), S-LPS, S-LipA (phosphorylated or dephosphory-lated), or S-Poly (all in PBS), at different concentrations. PBS alone did not interfere with the embryo development. Doses up to 5 mg/egg of S-Poly proved to be non toxic to the chick embryo in all developmental stages. On the contrary, crude S-H.p.-SUSPs, S-LPS and S-LipA, at appropriate doses, proved to be potentially lethal to the embryo, the percentage of embryo mortality varying, however, very widely according to the developmental stage. Results obtained after treatment of 4- to 17-day chick embryos with S-H.p.-SUSPs (3_10⁶ bacteria/egg), S-LPS (500 ng/egg), or phosphorylated S-LipA (500 ng/egg) are summarized in Table I. Treatment with 3_10⁶ (and up to 10⁸) bacteria had no toxic effect, within 4 hours, on 4-day embryos, but at later stages (6- and 8-day embryos) significant toxic effects were already observed. The embryo mortality suddenly rose to 90% ca. at the 10th day and kept at similar levels up to the 17th day. A similar stage-dependency and comparable percentages of lethality were observed with S-LPS and phosphorylated S-LipA treatment. The toxicity of the dephosphorylated S-LipA did not differ significantly from that of the phosphorylated form (data not shown).

 

    

 At the 10th developmental day, the S-LPS (or S-LipA) log dose-lethality curves (not shown) are very steep in the central section; the median lethal dose (LD50) was estimated to be around 400 ng/egg for both S-LPS and S-LipA.

     Embryos surviving beyond the fourth hour after treatment continued in their apparently unaltered development and did not show specific malformations, when checked at later develop-mental stages until hatching.

Effects on the Mechanical Activity of Atrial Fragments

     Under the above described experimental conditions, most 10-day ventricular fragments spontaneously stopped contracting after isolation, while atrial fragments continued contracting rhythmically, at a frequency of 90-130 beats/min. Following treatment of the atrial fragments with either a S-H.p.-SUSP (3x10⁶/ml), or S-LPS (500 ng/ml), or phosphorylated S-LipA (500 ng/ml), the contraction rate decreased by 60% ca. in two min and eventually the muscle completely stopped contracting within five minutes (Fig. 1). Treatments with dephosphorylated S-LipA (500 ng/ml) yielded similar results (not shown). Treatments with S-Poly (up to 20 mg/ml) did not alter the frequency of contraction of the atrial fragments.

     The toxic effects of the S-H.p.-SUSP were completely reversible upon washing (within less than 5 min); on the contrary, the contraction block after treatment with either S-LPS or S-LipA (either phosphorylated or dephosphorylated) proved to be irreversible (Fig. 1

 

Toxic Effects on Isolated Cardiomyocytes (Apoptosis and Necrosis)

(a) Loss of adherence

     One day after plating, most cells derived from the disaggregation of 10-day embryonal hearts were firmly adherent to the bottom of the culture vessel and over 90% of them exhibited a typical cardiomyocytic shape, roughly resembling elongated rectangles with rounded ends. Following a 24-hour treatment with S-Poly (up to 50 mg/ml) no obvious modification was observed either in cell density or shape. On the contrary, following a 24-hour treatment with either S-LPS or S-LipA (at appropriate concentrations) many cardiomyocytes had turned to a rounded shape (with no evidence of vacuolation) and eventually detached. Thus, the cytopathic effect was quantitated in terms of cell detachment, which was indirectly evaluated by counting cells that still kept adherent, or the amount of total proteins extracted from these cells, as compared to untreated controls.

     Following a 24-hour treatment with either S-LPS or S-LipA (phosphorylated) the number of adherent cells had diminished in a dose-dependent manner. This effect was already statistically significant at a concentration of 5 mg/ml of either S-LPS or S-LipA (the percentage of adherent cells was reduced by 25% ca.); at a concentration of 25 mg/ml the number of adherent cells was reduced by 75% ca., in comparison to the untreated controls (Fig. 2). Dephosphorylated S-LipA yielded similar results.

 

    

The results of the dosages of the total proteins extracted from the adherent cells confirmed the above results. After a treatment with either S-LPS or phosphorylated S-LipA (25 mg/ml; 24 hrs) the proteins extracted from the attached cells were decreased by almost 50% in comparison with the untreated cells. There was no statistical difference between S-LPS treatment and S-LipA treatment (Fig. 3).

 

 

(b) Apoptosis

     The distribution of PI/Annexin V-FITC-stained 10-day myocardial cells was studied by bivariate FACS analysis in cultures treated for 18 hours with S-LPS or phosphorylated S-LipA (4 experiments with S-LPS and 3 with S-LipA, each including a corresponding untreated control culture). One representative example is reproduced in Fig. 4. The results of the different experiments were remarkably consistent; briefly, in both S-LPS and S-LipA challenged cultures: (a) the number of doubly stained elements ("dead" cells) had increased by 11% ca. in the average, while the number of doubly negative elements ("healthy" cells) had decreased by 10% ca. in the average, and (b) the percentage of the Annexin-FITC-positive/PI-negative elements ("true apoptotic" cells) in the treated cultures was very low (around 6%; Fig. 4), and indeed sometimes lower than in controls. Some 13% of the elements were PI+/Annexin-; these were regarded as cells possibly damaged during isolation. On the whole, these results confirm that both S-LPS and S-LipA determine a significant increase in the percentage of cells undergoing necrosis, as compared to controls. On the contrary, the low percentage of the "true apoptotic" cells and the fact that no significant difference was observed in this respect between treated and control cells indicates that S-LPS and S-LipA do not to trigger significant apoptotic processes in cardiomyocytes.

 

 

Quadrant                             Control                     Treated

 

 

Upper Right                            23.2                            34.2

 

Lower Left                              56.9                            47.0

 

Lower Right                              7.5                              6.1

 

Superoxide Anion Production

     Cardiomyocytes treated with S-H.p.-SUSPs, or S-LPS, or S-LipA were found to release significant amounts of superoxide anion within the first two hours. Fig. 5 shows that the O₂^- release was higher with phosphorylated S-LipA (25 mg/ml) (dephosphorylated S-LipA yielded similar results), lower with S-LPS (25 mg/ml) (comparison with S-LipA; p=0,003), and still lower with S-H.p.-SUSP (3x10⁶/ml) (comparison with S-LPS; p=0,012). Untreated cells and S-Poly treated cells did not produce detectable amounts of O₂^-.

 

 

NO Production

     Relatively small amounts of NO were found to be released by untreated cultured cardiomyocytes, under the above described experimental conditions (Fig. 6, A). The amounts of NO released were significantly (p<0.001) increased following either S-LPS (25 mg/ml) or phosphorylated S-LipA (25 mg/ml) treatment, by 150% and 200% respectively (Fig. 6, B and C); however, the difference between these two treatments was statistically not significant. Dephosphorylated S-LipA yielded results similar to those of the phosphorylated form. Production of NO by the treated cells was completely abrogated by the addition of the NO synthase inhibitor, L-N^GMMA (Fig. 6, D and E). Treatments with S-H.p.-SUSP (3x10⁶/ml) or S-Poly did not increase the NO-producing capacities of cardiomyocytes (data not shown).

 

 

CONCLUDING REMARKS

     H. pylori organisms exist in either a smooth (S) or a rough (R) form. The high-molecular-weight S-form LPS consists of an O side chain (formed by polymerized oligosaccharide units) projecting from the cell surface [the O side chain is lacking in the low-molecular-weight R-form LPS], a core oligosaccharide, and Lipid A [41]. The toxicity to chick embryo cardiomyocytes of whole heath-killed S-H. pylori organisms and their S-LPS, S-LipA (phosphorylated and dephosphorylated), and S-Poly was investigated. Dephosphorylation is reported to reduce the toxic potency of the H. pylori S-LipA, as judged from the Limulus amebocyte lysate assay [36], but under the previously illustrated experimental conditions no significant differences were observed.

     The S-Poly proved to be practically devoid of toxicity; and this result is in accordance with its reported lower biological activity. On the contrary, heath-killed bacteria, S-LPS, and S-LipA exerted on embryonal cardiomyocytes two phenomenolo-gically distinct types of toxic actions, differentiated on the basis of their timing. Under the headings A, B, and C we will illustrate the different toxic effects and discuss whether these can be related to a unique or multiple mechanism(s) of action.

A) An early toxic effect is demonstrated by the block of the mechanical activity of the heart. In ovo, the cardiac arrest is observed about three hours after treatment in 10- to 17-day chick embryos, but in isolated atrial fragments from 10-day embryos it takes place within a few minutes. Thus, this toxic action is very fast, the delay observed in ovo being probably related to the time-intervals needed for the absorption at the level of chorioallantoic vasculature and probably the temporary uptake of the chemical species involved by other organs or cellular systems that “filter” the circulating blood.

     Though the experiments on isolated atrial fragments conclusively demonstrate that cardiomyocytes are direct targets for the bacterial products under consideration, it cannot be excluded that in the whole embryo other major targets or sites of biotransformation are also involved.

     The mechanisms of the early toxic actions on cardiomyocytes have not been specifically addressed in the present investigations; however, it is apparent that all the H. pylori products tested primarily exert some form of direct or indirect action on the ion channels involved in the processes of cardiomyocyte contraction. E. coli LPS has been reported to reduce Ca²⁺ influx through the L-type Ca²⁺ sarcolemmal channels, thus determining a decreased concentration of the ([Ca²⁺]_i) [23], while in other cell types (microglia) LPS induces the expression of outward rectifying potassium currents [28, 29].

     The present experiments demonstrated that the H. pylori products are able to activate a metabolic cascade leading to production of highly toxic superoxide anion radicals (O₂^-), which are known to generate other toxic radicals and compounds, such as –OH, _OH (the hydroxyl radical), and H₂O₂ [42]. In granulocytes and monocytes this “respiratory burst” (“oxygen burst”) is a very fast cellular response and is already detectable a couple of minutes after cell stimulation [43]. It has been shown that ROS may be produced also by several non-myeloid cell types, including differentiated cardiomyocytes [44, 45]; in these cells, cytosolic NADH oxidases [45] and the mitochondrial respiratory chain at Site I [44] seem to be primarily involved in ROS production. Furthermore, metabolites produced under hypoxia, such as lactate, were found to increase the levels of ROS production by myocardiocytes [45]; this observation is probably of interest in connection with the above described experiments, since isolated myocardial fragments are likely to be under moderate hypoxic conditions. According to several experimental and clinical reports, oxygen free radicals may mediate an acute impairment of the mechanical myocardial functions [44, 46-50], possibly by determining a condition of intracellular calcium overload [51]. Furthermore, ROS have been shown to modulate directly the inward rectifier K⁺ channels [52] and n-type K⁺ channels [53]. On the other hand, in other conditions a rise in [Ca²⁺]_i (Calcium transient) is an upstream event to the "oxygen burst"; e.g., a stimulation of granulocytes with the polypeptide N-formyl-methionyl-leucyl-phenylalanine (fMLP) activates the phospholipase C (PLC), which catalyzes the breakdown of phosphatidylinositol 4,5-biphosphate (PIP₂) to inositol 1,4,5-triphosphate (IP₃) and diacylglycerol (DAG); the former is responsible for intracellular Ca²⁺ mobilization, which is in turn required for the O₂^- production through the action of a Ca-dependent protein kinase C [42].

     Concluding, the ROS produced by the embryonal cardiomyocytes (possibly together with an imbalance in [Ca²⁺]_i preceding or following ROS production) might be instrumental in the blockade of the ion channels relevant to the contractile activity. Alternatively, S-LPS, S-LipA or other components of the whole heath-killed bacterium might affect -through some other direct or indirect mechanism (see before)- the relevant ion channels, the ROS production being only a concomitant but unrelated phenomenon.

     The block of myocardial contraction by the action of either S-LPS or S-LipA proved to be irreversible upon withdrawal and washing, suggesting a high-affinity binding of the two bacterial products to some receptor structure(s) of cardiomyocytes. On the opposite, the contraction block induced by crude S-H.p.-SUSPs, at the concentrations tested, proved to be fully reversible upon washing. Two hypotheses may be advanced to interpret the latter result: either (a) the toxicity of the heat-killed bacterium is still due to the action of its thermoresistant LPS, but the interaction of the bacterial-bound LPS with the cardiomyocyte receptor structure(s) is weaker than that of the free molecule, thus allowing a reversibility of the toxic effects; or, alternatively, (b) the block by heath-killed H. pylori might be due to the action of other thermostable bacterial products, with low-affinity binding to cardiomyocytes.

     As a comment to the hypothesis (a) it is to be remarked that in the bacterium S-forms the biologically active core oligosaccharide and Lipid A are known to be located deep to the O-polysaccharide [54] and it is possible that this supermolecular surface arrangement prevents LPS from stably binding to the corresponding cell receptor structures.

     In discussing the above hypothesis (b), we have to take into consideration the several "virulence factors" (besides LPS) produced by H. pylori organisms [16-18]. Among these factors, the cytotoxic protein vacA (to which the protein factor cagA is associated), which is activated in a acidic milieu, exerts its vacuolating effects acting on an cytosolic target, after being endocytosed [19]; furthermore, vacA is likely to be denatured by heating. However, vacA has also been reported to form anion-selective, voltage-dependent pores in artificial membranes [55], and it cannot be excluded that vacA could exert other toxic actions directly at the level of the cell membrane. Putative H. pylori effectors for a direct action on the outer cell membrane are porins [56], producing non-specific channels across the membrane and allowing a non-selective Ca²⁺ influx; however, these proteins are possibly denatured by heating. A heat-stable factor, isolated from H. pylori but different from vacA and cagA, was found to act as a local vasoconstrictor [57], but its actions at the cell membrane level are not known. Of course, while the highly toxic urease could not of course be active in the experimental conditions previously reported, phospholipases, proteases, neuramini-dase, fucosidase, or alcohol dehydrogenase, which are reported to be also produced by H. pylori, might play a role in altering the biochemical properties of the myocardiocyte cell membrane. Obviously, the present information is too scanty to evaluate the role, if any, of these "virulence factors" in the early toxic effects on myocardiocytes under the described experimental conditions and further experiments with purified bacterial fractions will be needed to determine the toxicity to the cardiomyocytes of other thermostable non-LPS H. pylori products.

     It is noteworthy that the toxic effects are of relatively limited extent up to the 8th embryonal day, the sensitivity of cardiomyocytes becoming much higher starting from the 10th day. The present observations, parallel -at least partly- our former findings on the toxicity of formylpeptides to the chick embryo [42]. The polypeptide fMLP, which is a chemoattractant and a powerful “activator” of polymorphonuclear and mono-nuclear phagocytes isolated from Escherichia coli cultures, proved to be relatively non-toxic in early embryonal stages, its toxicity becoming much higher during the developmental days 8 to 13 (with a peak around the 10th day), thereafter declining to much lower levels. Furthermore, we were able to demonstrate that the embryonal cardiomyocytes are one of the direct targets for the fMLP toxic actions [58]. Although timings of H. pylori toxicity and fMLP toxicity are not fully superimposable (and indeed the toxic effects differ also in other respects), it is apparent that around the 10th developmental day the chick embryo cardiomyocytes accomplish some further step in their biochemical maturation (perhaps the expression of novel receptors) that makes them more sensitive to some classes of bacterial products.

B) S-LPS and S-LipA also exert a “slower” toxic action, eventually leading –on a time-scale of several hours- to cardiomyocyte detachment and death, as demonstrated in the present experiments by both cellular loss-of-adherence and necrosis/apoptosis tests. These results indicate that S-LPS and S-LipA, besides their immediate action on the specific cellular systems involved in the contractile activity, can affect and eventually disrupt, although to a slower rate, the basic vital functions of the embryonal cardiomyocytes. The present experiments demonstrated that cultured 10-day chicken cardiomyocytes are able to produce sizable levels of ROS following treatments with killed H. pylori organisms, S-LPS, or S-LipA and, in addition, increased (over the controls) amounts of NO following treatments with S-LPS or S-LipA (but not killed H. pylori organisms, at least at the concentrations tested). Other investigators have been able to isolate an inducible nitric oxide synthase (iNOS) cDNA from 10-day cultured chick embryo ventricular myocytes stimulated with LPS [33], and in the rat fetus the NO released by placenta was found to be pharmacologically active as a vasodilator agent on the large vessels [59]. Both ROS and NO are known to be highly toxic to bacteria [e.g., both are reported to be instrumental in the bacterial killing by granulocytes and macrophages [60, 61]] and animal cells [e.g., the ROS released by polymor-phonuclear cells activated under conditions of ischemia/reperfusion can severely damage the organs involved [62, 63]]. Accordingly, it has been shown that both H₂O₂ and O₂^- affect viability of cultured neonatal rat cardiac cells [37]. Furthermore, it has been observed that cardiac myocytes undergo severe degeneration processes in mutant mice deficient in manganese superoxide dismutase, leading to death early in neonatal life [64, 65]; this shows that even the small amounts of O₂^- produced under physiological conditions are able to slowly disrupt cardiomyocytes (and other tissues) unless removed by the specific dismutase. Thus, the O₂^- and NO produced by the heart-derived cell cultures (also including minor aliquots of connective tissue cells and endothelial cells, besides the cardiomyocytes) might by directly or indirectly instrumental in determining the increased cellular losses observed after S-LPS and S-LipA treatments in vitro. However, our present results do not exclude the possibility that biochemical mechanisms other than ROS and NO production may come into play in determining the “late” cytopathic effects of the H. pylori LPS. For instance, the CD14-mediated production of TNF-a [25] and the production of IFN-g or other proinflammatory cytokines through the activation of cascades of specific kinases [31, 32] have been advocated as possible mechanisms accounting for LPS toxicity.

     A further question is whether, following treatment with S-LPS or S-LipA, cardiomyocytes undergo a mere degeneration/necrosis process or they self-eliminate by some sort of, partly active, apoptotic process. Cardiomyocyte degeneration exhibiting the biochemical characteristics of an apoptotic process has been described in different conditions, such as acute myocardial infarction, ischemic cardiomyopathy, dilated cardiomyo-pathy, cardiac allograft rejection, and following reperfusion [66-70]. In in vitro experiments, the loss of neonatal rat cardiac cells following exposure to H₂O₂ or O₂^- is largely determined by the activation of an apoptotic program [37] and E. coli LPS is reported to be able to trigger apoptosis in different tissues [24-26]. In our experiments, however, the low percentage of "true apoptotic" elements in treated cells and the lack of significant differences between treated and control cells (as judged from the surface exposure of phosphatidylserine) seem to rule out the hypothesis that damaged cardiomyocytes play a significant "active" self-destructive role. Obviously, further experimental observations are required in order to define the genetic and biochemical aspects of the degenerative/necrotic processes taking place in S-LPS- and S-LipA-treated embryonal cardiomyocytes.

     The PI/Annexin V FACS analysis demonstrated a relatively high incidence of necrotic and apoptotic phenomena in the control, untreated cardiomyocytes (in the average, 24% of necrotic cells and 6% of apoptotic cells). These results may reflect, on one hand, the fact that cardiomyocytes do not adapt optimally to the in vitro culture conditions [71] and, on the other hand, the relatively high rate of cell necrosis/apoptosis events occurring during the normal morphogenesis of the heart in the chick embryo [72], as well as in other species, including man [73, 74].

C) Summarizing our discussion under the headings A and B, we conclude that H. pylori LPS and its Lipid A moiety (and possibly other, yet undetermined, thermoresistant H. pylori products) are able to affect both the mechanical activity of embryonal cardiomyocytes (the “early” toxic action) and their viability (the “late” toxic effects). However, only those myocardial cells that are well advanced in their differentiation (starting from the mid-incubation period in the chick embryo) are highly sensitive to these toxic agents.

     S-LPS- and S-LipA-challenged cardiomyocytes produce significantly high amounts of ROS and NO. ROS are likely to be directly or indirectly involved in the early toxic effects, while both ROS and NO might be involved in the late degenerative/necrotic effects. However, our present data do not allow to state the exact role(s) of these chemical species in the cascade(s) of biochemical events leading to the final toxic effects. Possibly, the LPS-receptor, CD14, and the production of TNF-a and IFN-g may concur in mediating, at different levels, the LPS toxic actions. Finally, some of the experimental data suggest that both S-LPS and S-LipA, when combined with other molecular components in the intact bacterium, may be comparatively less toxic than the isolated molecules.

     Several clinical evidences, as reported in the Introduction, indicate that chronic H. pylori infections in the adult are possibly associated with a higher incidence of cardiac disorders. In Mammals, the clinical and experimental data on the embryonal/fetal effects of a maternal bacterial infection are relatively scanty. Bacterial LPS has been demonstrated to affect indirectly the fetus through its actions on the placenta. In human placenta, LPS is reported to prime prostaglandin E₂ release [75] and suppress gonadotropin production by a direct action on trophoblast cells and, additively, by promoting ROS release from the neutrophils and monocytes of the organ [76]. In mouse, LPS was found to induce NO release from decidual macrophages [77, 78] and cellular apoptosis in the placental tissue [26]; furthermore, bacterial LPS induces a dramatic increase in maternal serum concentrations of TNF-a, interleukin (IL)-6 and IL-1a and amniotic fluid concentrations of IL-6 and IL-1a [79]. LPS may also affect the embryo directly since it is able to cross the placental barrier: Cai et al. [80] have demonstrated that, in the rat, maternal LPS administration induces an increased expression of IL-1b and TNF-a mRNAs in the fetal brain. As a result of all these direct and indirect toxic actions, maternal bacterial infections or LPS administration are reported to induce premature labor and delivery or fetal death and abortion [81-83]. On the whole, these observations demonstrate that bacterial LPS, or chemical species released by maternal leukocytes and macrophages in response to challenging with bacterial products, may directly or indirectly affect fetal viability. Our present results demonstrate that in the avian embryo H. pylori S-LPS exerts direct toxic actions on cardiomyocytes and suggest that LPS, and possibly other H. pylori organism products able to cross the placental barrier, might be toxic to the heart tissue of the mammalian fetus, especially in the intermediate and late developmental periods, whereas cardiomyocytes seem to be much less LPS-sensitive in earlier developmental periods.

ACKNOWLEDGEMENTS

     The authors thank: Prof. Anthony P. Moran, University College Galway, Ireland, for the generous gift of the LPS chemotypes used in this study; and Dr. Anna Giberna for her skilful secretarial assistance. Paper supported by grants from M.U.R.S.T., C.N.R. and Ministero della Sanità (Rome, Italy).

REFERENCES

[1]        Nomura, A., Stemmermann, G.N., Chyou, P.H., Kato, I., Perez-Perez, G.I., Blaser, M.J. N. Engl. J. Med., 1991, 325, 1132.

[2]        Peterson, W.L. N. Eng. J. Med., 1991, 324, 1043.

[3]        Parsonnet, J., Hansen, S., Rodriguez, L., Gelb, A.B., Warnke, R.A., Jellum, E., Orentreich, N., Vogelman, J.H., Friedman, G.D. N. Engl. J. Med., 1994, 330, 1267.

[4]        Gasbarrini, A., Franceschi, F., Cammarota, G., Pola, P., Gasbarrini, G. Ital. J. Gastroenterol. Hepatol., 1998, 30, 115.

[5]        Gasbarrini, A., Franceschi, F. Biomed. Pharmacother., 1999, 53, 223.

[6]        Mendall, M.A., Goggin, P.M., Molineaux, N., Levy, J., Toosy, T., Strachan, D., Camm, A.J., Northfield, T.C. Br. Heart J., 1994, 71, 437.

[7]        Niemelä, S., Karttunen, T., Korhonen, T., Läärä, E., Karttunen, R., Ikäheimo, M., Kesaniemi, A.Y. Heart, 1996, 75, 573.

[8]        Pasceri, V., Cammarota, G., Patti, G., Cuoco, L., Gasbarrini, A., Grillo, R.L., Fedeli, G., Gasbarrini, G., Maseri, A. Circulation, 1998, 97, 1675.

[9]        Pieniazek, P., Karczewska, E., Duda, A., Tracz, W., Pasowicz, M., Konturek, S.J. J. Physiol. Pharmacol., 1999, 50, 743.

[10]      Sung, J.J.Y., Sanderson, J.E. Heart, 1996, 76, 305.

[11]      Strachan, D.P. Br. Med. Bull., 1998, 54, 87.

[12]      Folsom, A.R., Nieto, F.J., Sorlie, P., Chambless, L.E., Graham, D.Y. Circulation, 1998, 98, 845.

[13]      Koenig, W., Rothenbacher, D., Hoffmeister, A., Miller, M., Bode, G., Adler, G., Hombach, V., Marz, W., Pepys, M.B., Brenner, H. Circulation, 1999, 100, 2326.

[14]      Murray, L.J., Bamford, K.B., Kee, F., McMaster, D., Cambien, F., Dallongeville, J., Evans, A. Atherosclerosis, 2000, 149, 379.

[15]      Whincup, P., Danesh, J., Walker, M., Lennon, L., Thomson, A., Appleby, P., Hawkey, C., Atherton, J. Circulation, 2000, 101, 1647.

[16]      Figura, N. Helicobacter, 1997, 2 (suppl. 1), S3.

[17]      Atherton, J.C. Br. Med. Bull., 1998, 54, 105.

[18]      McGee, D.J., Mobley, H.L. Curr. Top. Microbial Immunol., 1999, 241, 155.

[19]      Molinari, M., Galli, C., de Bernard, M., Norais, N., Ruysschaert, J.-M., Rappuoli, R., Montecucco, C. Biochem. Biophys. Res. Commun., 1998, 248, 334.

[20]      Parker, J.L., Adams, H.R. In Cardiovascular Toxicology; Acosta D., Ed., Raven: New York, 1992, pp. 197-237.

[21]      Hung, J., Lew, W.Y. Circ. Res., 1993, 73, 125.

[22]      Rubin, L.J., Keller, R.S., Parker, J.L., Adams, H.R. Shock, 1994, 2, 113.

[23]      Zhong, J., Hwang, T.C., Adams, H.R., Rubin, L.J. Am. J. Physiol., 1997, 273, H2312.

[24]      Haendele, J., Zeiher, A.M., Dimmeler, S. Eur. J. Pharmacol., 1996, 317, 407.

[25]      Comstock, K.L., Krown, K.A., Page, M.T., Martin, D., Ho, P., Pedraza, M., Castro, E.N., Nakajima, N., Glembotski, C.C., Quintana, P.J., Sabbadini, R.A. J. Mol. Cell. Cardiol., 1998, 30, 2761.

[26]      Ejima, K., Koji, T., Tsuruta, D., Nanri, H., Kashimura, M., Ikeda, M. Biol. Reprod., 2000, 62, 178.

[27]      Wetzel, G.T., Chen, F., Klitzner, T.S. Circ. Res., 1993, 72, 1065.

[28]      Chung, S., Jung, W., Lee, M.Y. Neurosci. Lett., 1999, 262, 121.

[29]      Draheim, H.J., Prinz, M., Weber, J.R., Weiser, T., Kettenmann, H., Hanisch, U.K. Neuroscience, 1999, 89, 1379.

[30]      Fearns, C., Kravchenko, V.V., Ulevitch, R.J., Loskutoff, D.J. J. Exp. Med., 1995, 181, 857.

[31]      Cowan, D.B., Poutias, D.N., Del Nido, P.J., McGowan, F.X. Jr. Am. J. Physiol. Heart Circ. Physiol., 2000, 279, H619.

[32]      Marchant, A., Bruyns, C., Vandenabeele, P., Ducarme, M., Gerard, C., Delvaux, A., De Groote, D., Abramowicz, D., Velu, T., Goldman, M. Eur. J. Immunol., 1994, 24, 1167.

[33]      Shimizu, T., Kinugawa, K., Sugishita, Y., Sugishita, K., Harada, K., Matsui, H., Kohmoto, O., Serizawa, T., Takahashi, T. Cardiovasc. Res., 1998, 38, 405.

[34]      Han, X., Shimoni, Y., Giles, W.R. J. Physiol., 1994, 476, 309.

[35]      Moran, A.P., Walsh, E.J. J. Med. Sci., 1993, 162, 384.

[36]      Pece, S., Fumarola, D., Giuliani, G., Jirillo, E., Moran, A.P. J. Endotoxin Res., 1995, 2, 455.

[37]      von Harsdorf, R., Li, P.-F., Dietz, R. Circulation, 1999, 99, 2934.

[38]      van Engeland, M., Nieland, L.J.W., Ramaekers, F.C.S., Schutte, B., Reutelingsperger, C.P.M. Cytometry, 1998, 31, 1.

[39]      Weening, R.S., Wever, R., Roos, D. J. Lab. Clin. Med., 1975, 85, 245.

[40]      Ding, A.H., Nathan, C.F., Stuehr, D.J. J. Immunol., 1988, 141, 2407.

[41]      Moran, A.P., Helander, I.M., Kosunen, T.U. J. Bacteriol., 1992, 174, 1370.

[42]      Panaro, M.A., Mitolo, V. Immunopharmacol. Immunotoxicol., 1999, 21, 397.

[43]      Lund-Johansen, F., Olweus, J. Cytometry, 1992, 13, 693.

[44]      Ambrosio, G., Zweier, J.L., Duilio, C., Kuppusamy, P., Santoro, G., Elia, P.P., Tritto, I., Cirillo, P., Condorelli, M., Chiariello, M., Flaherty, J.T. J. Biol. Chem., 1993, 268, 18532.

[45]      Mohazzab-H, K.M., Kaminski, P.M., Wolin, M.S. Circulation, 1997, 96, 614.

[46]      Ferrari, R., Ceconi, C., Curello, S., Alfieri, O., Visioli, O. Eur. Heart J., 1993, 14 Suppl., G 25.

[47]      Chan, E.M., Thomas, M.J., Bandy, B., Tibbits, G.F. Can. J. Physiol. Pharmacol., 1996, 74, 904.

[48]      Keith, M., Geranmayegan, A., Sole, M.J., Kurian, R., Robinson, A., Omran, A.S., Jeejeebhoy, K.N. J. Am. Coll. Cardiol., 1998, 31, 1352.

[49]      Mallat, Z., Philip, I., Lebret, M., Chatel, D., Maclouf, J., Tedgui, A. Circulation, 1998, 97, 1536.

[50]      Courtois, M., Maupoil, V., Fantini, E., Durot, I., Javouhey-Donzel, A., Athias, P., Grynberg, A., Rochette, L. Free Radic. Biol. Med., 1998, 24, 121.

[51]      Hess, M.L., Manson, N.H. J. Mol. Cell. Cardiol., 1984, 16, 969.

[52]      Jeglitsch, G., Ramos, P., Encabo, A., Tritthart, H.A., Esterbauer, H., Groschner, K., Schreibmayer, W. Biol. Med., 1999, 26, 253.