| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
National Institute of Public Health of Mexico. Av. Universidad No. 655, Col. Santa Ma. Ahuacatitlán, 62508 Cuernavaca, Morelos, Mexico
Correspondence
Lizbeth López-Carrillo
lizbeth{at}insp.mx
Received 2 August 2007
Accepted 20 September 2007
Abbreviations: GC, gastric cancer; OR, odds ratio; SVA, supernatants with vacuolating activity.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
The H. pylori VacA cytotoxin is a virulence factor potentially linked to GC. VacA causes cell vacuolation in cell culture, produces significant epithelial cell damage and increases the level of apoptosis in epithelial cells (Galmiche et al., 2000; Kuck et al., 2001; Cover et al., 2003); this in turn enhances the hyperproliferative response, altering the balance of gastric homeostasis, which changes gastric epithelial cell turnover and permits the persistence of mutated cells. Studies suggest that these processes may promote carcinogenesis (Galmiche et al., 2000; Xia & Talley, 2001).
The clinical and epidemiological evidence regarding the relationship between VacA and GC is unclear (de Figueiredo Soares et al., 1998; Shimoyama et al., 1999; Yamaoka et al., 1999; Enroth et al., 2000; Figueroa et al., 2002). Among patients with GC, some clinical studies have reported positive correlations between antibody host response to VacA and the presence of VacA-positive H. pylori strains (de Figueiredo Soares et al., 1998; Garza-Gonzalez et al., 2004). Some epidemiological studies have shown an increased risk of GC in seropositive VacA subjects (Enroth et al., 2000; Figueroa et al., 2002), but not all studies have reported these findings (Shimoyama et al., 1999; Yamaoka et al., 1999). Two possible explanations for the inconsistent reporting of the VacA/GC association are: (i) the lack of control for key confounding factors such as H. pylori CagA status, and (ii) significant differences in the sensitivity and specificity of the immunoblot kits used to determine the H. pylori VacA status; among these are: Helicoblot (HB2.0 or HB2.1; Genelabs Diagnostics) (de Figueiredo Soares et al., 1998; Enroth et al., 2000; Garza-Gonzalez et al., 2004), the Chiron RIBA Recombinant Immunoblot assay (Chiron) (Shimoyama et al., 1999; Yamaoka et al., 1999) and other homemade immunoblot assays (Hirai et al., 1994; Rudi et al., 1997; Figueroa et al., 2002).
Here, we compared the association between GC and the presence of anti-VacA antibodies using two different detection methods; in addition, we adjusted for CagA status. The study was conducted with serum samples from Mexico, where approximately 80 % of the adult population are seropositive for H. pylori (Torres et al., 1998) and where GC is still a leading cause of death by cancer (Secretaría de Salud, 1993). Unlike the rest of the world (Parkin et al., 1997), the incidence of this type of cancer is not decreasing.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Cases. Inclusion criteria for case subjects were: histologically confirmed adenocarcinoma of the stomach, either intestinal or diffuse type; no previous history of any other cancer; at least 20 years of age; and resident in the study area for at least the previous 6 months. Eligible cases were recruited from 13 tertiary, public, social security hospitals in Mexico, distributed as follows: seven from Mexico City (Federal District), three from the city of Puebla (State of Puebla) and three from the city of Merida (State of Yucatan). A single expert cancer pathologist determined the histological variety of the tumour (intestinal or diffuse), following the criteria established by Lauren (1965). Information about the anatomical subsite location of the tumour was not available for this study. In total, 281 patients who met all of the eligibility criteria were identified (approx. 75 % of the total number of GC patients reported to the Mexican National Cancer Registry during the study period of recruitment for the study areas) and 261 agreed to enrol in the study, giving a participation rate of 92.9 %.
Controls. For each case subject, we recruited a minimum of two age-, sex- and residence-matched control subjects. Controls were randomly identified in selected departments of the participating hospitals. Eligibility criteria included: no previous diagnosis of malignant tumour, diet-related disease (gastritis, peptic ulcer, liver cirrhosis or diabetes mellitus) or immunosuppressive condition; same age (±5 years) as the index case; and resident in the same study area as the index case for at least the previous 6 months. The most common diagnoses among controls were: circulatory system diseases (other than hypertension) (18.8 %); diseases of the nervous system and the sensory organs (except psychiatric disorders) (16.1 %); osteomuscular and connective tissue disorders (14.4 %); injuries and poisonings (10.9 %); and respiratory diseases (10.0 %). Other less common diagnoses were genitourinary diseases and skin problems. We also recruited control subjects who were healthy companions of patients and those attending the hospitals for preventive care purposes, such as vaccination. The participation rate for the control group was 94.6 % (523 out of 553 eligible subjects).
Interviews. After informed consent had been obtained from each subject, they were questioned about their socioeconomic, clinical and dietary characteristics. The interviews were performed in the participating hospitals by trained personnel.
Serum samples. Blood samples (10 ml) were drawn from each subject. After centrifugation, serum aliquots were obtained and stored at –70 °C until the samples were tested. Serum samples from a total of 203 cases and 430 controls were available for this study.
The H. pylori status of each subject was determined. H. pylori IgG antibodies were determined by immunoblot as described below. The H. pylori CagA status was available for this study and the ELISA procedures used to determine it have been described elsewhere (Lopez-Carrillo et al., 2004).
Production of cytotoxin and assessment of vacuolating activity.
A high level toxin-producing H. pylori strain (7494, s1/m1 genotype and cagA+) was used as the source for vacuolating cytotoxin preparation. Colonies obtained from one blood agar plate [tryptic soy agar (Difco) supplemented with 5 % sheep blood plus 10 µg vancomycin ml–1] were inoculated into 100 ml sulfite-free Brucella broth containing 0.2 % cyclodextrin (Sigma) and cultured as described previously (Ayala et al., 2006). The cultures were centrifuged at 15 000 g for 20 min at 4 °C. The supernatants were then pooled and this fraction was termed ‘supernatants with vacuolating activity’ (SVA). Analysis by 10 % SDS-PAGE showed that this fraction was composed of proteins of different molecular mass, among them a band corresponding to VacA (
87 kDa). The SVA fraction was used as the source of VacA for the neutralization assays, as well as antigen for immunoblot analysis. Determination of vacuolating activity for the SVA fraction was conducted essentially as described by Cover et al. (1990). The assay was carried out by the addition of a fixed amount of total SVA protein (6.1 µg) to confluent HeLa cells. To estimate quantitatively the vacuolating activity of the SVA fraction, a neutral red uptake assay was used as reported elsewhere (de Bernard et al., 1998) and is described below. The protein concentration of the SVA was determined using the Lowry method (Lowry et al., 1951).
H. pylori anti-VacA antibody status was determined by two methods: a VacA-neutralizing antibody assay and an in-house immunoblot assay.
Neutralization of VacA cytotoxin activity. The serum from each patient was inactivated as described elsewhere (Cover et al., 1992) and then diluted with Dulbecco's modified Eagle's medium (DMEM) cell-culture medium. Inhibition of VacA activity using patient sera was performed according to a method reported elsewhere (Cover et al., 1992) and modified as follows: equal volumes of H. pylori SVA containing 6.1 µg total protein and serum from each subject were pre-incubated for 1 h at 37 °C. The starting amount of protein in each serum sample was 9.5 µg and ten further twofold dilutions from 1 : 4 to 1 : 2048, were carried out in DMEM culture medium in 96-well plates. Next, 100 µl from each well containing the SVA/serum mixture was transferred to a second 96-well plate containing confluent HeLa cells (7x103 cells per well). The plates were incubated for 18 h at 37 °C with 5 % CO2. For each serum sample, two controls were included: (i) HeLa cells with SVA but without serum (positive control), and (ii) HeLa cells with DMEM culture medium alone (negative control). Experiments were carried out in duplicate. Both vacuolization and inhibition were evaluated by neutral red uptake assay (de Bernard et al., 1998). Cells were stained for 8 min with 0.05 % neutral red. After washing the cells twice with 0.3 % BSA in PBS, intravacuolar neutral red was extracted with 70 % ethanol and 0.37 % HCl. A540 was determined using a Multiskan EX ELISA reader (Thermo Electron). All assays were performed in duplicate. Results were expressed as the mean A540 of experimental wells, minus background A540 (the mean A540 of wells containing HeLa cells incubated with medium only). The net A540 value was used to calculate the cut-off value.
The cut-off point was determined by comparing the net A540 value of wells containing both the SVA fraction and the serum being tested. The net A540 values obtained for each of ten serial dilutions (described above) were compared by analysis of variance. The positive/negative cut-off point was determined using the lowest dilution for which a statistically significant difference was observed in the mean net A540 between two consecutive dilutions. The cut-off point was set as 0.16 absorbance units at a dilution of 1 : 32. Thus, a serum was judged to be positive (i.e. showed neutralizing activity) if the A540 value was lower than 0.16 at a dilution of 1 : 32. Sera with A540
0.16 were considered negative (i.e. no neutralizing activity).
H. pylori and VacA cytotoxin immunoblot assay.
The SVA fraction was concentrated approximately 50-fold by ultrafiltration with a 10 kDa cut-off (Millipore). Proteins were separated by 10 % SDS-PAGE and transferred to nitrocellulose membranes (Millipore). The filter was blocked for 1 h with buffer containing 10 % (v/v) skimmed milk in 20 mM Tris/HCl (pH 7.5) containing 0.5 M NaCl. The membranes were cut into strips and incubated overnight at room temperature with serum samples from each patient at a 1 : 200 dilution. The strips were washed and further incubated for 1 h at room temperature with alkaline phosphatase-conjugated goat anti-human IgG (Sigma). A serum sample was considered to be H. pylori-positive if three or more immunoreactive bands were present on the strip; otherwise, the sample was considered negative. The sample was also judged to be VacA-positive if a band with a molecular mass of 87 kDa was recognized. To confirm that the 87 kDa band corresponded to VacA, one strip was incubated with anti-VacA antibody (bN-20; Santa Cruz Biotechnology). Colour reactions were developed with 5-bromo-4-chloro-3-indolylphosphate and nitro blue tetrazolium.
Statistical analyses.
The general characteristics of cases and controls (age, sex, etc.), as well as H. pylori status (IgG, CagA and anti-VacA antibodies), were compared using Student's t test and 
2 statistics. Crude and adjusted risks (by age and CagA status) for GC according to the presence of H. pylori IgG antibodies and VacA status were estimated using unconditional logistic regression models for all adenocarcinomas and then stratified for histological type of GC (i.e. intestinal and diffuse).
|
RESULTS AND DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
. Age, sex and residence distribution were similar, and no significant difference was found between cases and controls regarding socioeconomic level. Intestinal GC was more frequent than diffuse GC in this population.
|
). A statistically significant difference in CagA-positive status was also found between cases and controls (78.6 vs 66.0 %).
|
). A borderline significant difference was found between cases and controls for the presence of anti-VacA antibodies (68.5 vs 60.5 %) when these were determined by neutralization. In contrast, when immunoblotting was used to detect anti-VacA antibodies, lower and similar percentages were found in both cases and controls (44.3 and 44.2 %, respectively). The discrepancy between the two methods may be due to the fact that most anti-VacA antibodies present in H. pylori-infected patients recognize conformational epitopes of VacA, whilst linear epitopes of VacA are recognized to a lesser degree (Figueiredo et al., 2001). This weak reactivity of VacA in immunoblots using sera from patients with active H. pylori infection has been reported previously (Lepper et al., 2004). The failure to recognize linear VacA epitopes could potentially result in differences in sensitivity and specificity between immunoblotting and neutralization. Therefore, detection of anti-VacA antibodies by inhibition of VacA activity in vitro should produce more reliable data than immunoblot assays. It is important to note that human serum from H. pylori-infected patients, as well as from uninfected persons, contains non-specific VacA-neutralizing activity that does not depend on anti-VacA IgG (Cover et al., 1992). However, such an effect would not have affected our results at the dilution used (1 : 32), as the amount of serum protein at the positive/negative cut-off point was very low (0.30 µg), so any non-specific neutralizing activity would be negligible.
Association of virulence factor with GC
Table 3 shows that although immunoblotting and neutralization both detected anti-VacA antibodies, only when the neutralization assay was used were we able to find a significant odds ratio (OR 1.69, 95 % CI 1.08–2.66). A significant association between H. pylori IgG antigens and GC risk was also detected (OR 2.37, 95 % CI 1.15–4.89) (data not shown).
|
VacA and CagA are involved in different steps of the pathogenic process. VacA plays an essential role in the initial colonization of H. pylori and the subsequent epithelial cell damage of the stomach mucosa (Salama et al., 2001). VacA also suppresses the local immune response, which promotes the presence of a chronic, persistent H. pylori infection (Gebert et al., 2004). CagA-positive H. pylori strains possess an intact Cag pathogenicity island, which is associated with activation of the inflammatory response and mucosal immunity, as well as with altered regulation of multiple genes associated with proliferation and differentiation of gastric epithelial cells. The function of the CagA protein itself is not yet clear, but recent reports have associated it with transformation of the gastric epithelium. This might indicate that the development of GC would only occur if the H. pylori infection is chronic and if the infecting strain is CagA-positive (Amieva et al., 2003; Hatakeyama, 2004). Chronic inflammatory changes are linked to cancer, thus explaining the relative importance of CagA versus VacA in GC. However, we cannot exclude the participation of the co-chaperonin GroES of H. pylori in the induction of inflammation and neoplastic changes (Lin et al., 2006), reflecting the multiple pathogenic possibilities displayed by this bacterium.
VacA is a highly expressed protein in vivo (Boonjakuakul et al., 2004) and some studies have shown that carriers of H. pylori strains that efficiently produce VacA are at higher risk of GC. However, few studies have evaluated the relationship of GC with the presence of anti-VacA antibodies (Rudi et al., 1997; de Figueiredo Soares et al., 1998; Enroth et al., 2000; Figueroa et al., 2002) and results have been inconsistent (Shimoyama et al., 1999; Yamaoka et al., 1999). Thus, the anti-VacA/GC association is questionable, and earlier controversial results may partially be explained by their different methods and the criteria of the assay.
This study showed that the association between anti-VacA antibodies and GC risk depends not only on the type of serological test used to detect anti-VacA antibodies and CagA, but also on the CagA status. This last situation does not support an independent role for VacA in GC risk.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
-induced apoptosis of gastric epithelial cells in Helicobacter pylori infection. Infect Immun 69, 816–821.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
