Helicobacter pylon Infection and Urinary Excretion of 8-Hydroxy-2- deoxyguanosine, an Oxidative DNA Adduct1

To assess whether Helicobacter pylon-related inflammation increases oxidative DNA damage, we evaluated the association between H. pylon infection and urinary excretion of an adduct of oxidative DNA damage, 8-hydroxy-2-deoxyguanosine (8ohdG). Subjects included 555 healthy persons, ages 20-39, within the Kaiser Permanente Medical Care Program in Northern California. We tested sera for antibodies to H. pylon by ELISA; collected demographic, dietary, smoking, and alcohol data by questionnaire; and assayed 24-h urine samples for SohdG with a newly developed ELISA kit. Two hundred eighty-one subjects provided adequate 24-h urine samples for SohdG and creatinine assays and had detectable levels of SohdG. After adjusting for 24-h urinary creatimne (Ucr) and demographic factors, persons without H. pylon infection had significantly higher amounts of 24-h urinary SohdG than infected persons (geometric mean, 18.04 ,tg 8ohdG/Ucr g versus 14.36 ,.tg SohdG/Ucr g, respectively; P 0.008). Excretion of SohdG was higher in whites and Hispanics (17.44 and 18.09 ,.tgfUcr g) than in blacks (13.21 g.tgfUcr g; P < 0.001). Gender was not significantly associated with 8ohdG excretion (16.18 .tgfUcr g for males versus 16.01 tg/Ucr g for females; P = 0.883). Of the dietary factors evaluated, vitamin C negatively correlated (P < 0.001) and carbohydrate intake positively correlated with 8ohdG excretion (P 0.003). Infection with H. pylon was strongly associated with decreased 8ohdG excretion in the urine. This unexpected finding suggests either that DNA repair is deficient in infected subjects, that inflammation destroys the adduct, or that urinary 8ohdG is not an accurate measure of gastric damage. Introduction Helicobacter pylon causes gastric inflammation and gastric cancer (1-5). ROS3 released by phagocytic cells consequent to inflammation have been implicated in this process (6-8). In support of this, gastric mucosal biopsy specimens from H. pylon-infected individuals contain significantly greater amounts of ROS than biopsy specimens from noninfected persons (9). Moreover, the density of H. pylon observed in the gastric mucosa correlates with levels of ROS in gastric tissue (10). If ROS, released consequent to H. pylon related gastritis, are involved in gastric carcinogenesis, then individuals with active H. pylon infection should have greater indices of oxidative DNA damage in the stomach than uninfected persons. This was recently supported by a small study (n 74), demonstrating significantly higher levels of one marker of oxidative DNA damage, 8ohdG, in mucosal tissue of H. pylon-infected children than in uninfected control tissue (1 1). 8ohdG is one of the most abundant lesions in DNA induced by ROS (12, 13). It results from attack of a singlet hydroxyl or oxygen radical on guanine. 8ohdG lesions can be excised by DNA repair enzymes before mutation occurs, can undergo cell death, or can lead to permanent changes in DNA structure (14-16). During DNA replication, unrepaired 8ohdG lesions induce G-to-T transversions, which may contribute to carcinogenesis (17). Ostensibly, 8ohdG lesions excised from tissues by DNA repair enzymes enter the bloodstream and are filtered and excreted in urine (18). Shigenaga and Ames (19) demonstrated that when 8ohdG is infused i.v., it does not become degraded during its transport from plasma to urine. This suggests that adducts, once they have reached the bloodstream, should be detected in the urine in a quantity reflective of serum levels. Although little is known about individual differences with respect to DNA adduct formation, tolerance for such adduction, and subsequent DNA repair mechanisms, age, metabolic rate, smoking, and dietary caloric and antioxidant intake may influence adduct formation and/or urinary detection (20). However, dietary sources of 8ohdG did not seem to influence the levels of the adduct detected in urine in an animal model (19). Most studies of 8ohdG to date have focused on tissue levels of the adduct. Because urinary assays are noninvasive, their use is far simpler than those involving tissue samples. Moreover, urine levels should reflect total body oxidative stress, a biological measure of great interest. Unfortunately, urinary assays require HPLC with electrochemical detection. This is technically cumbersome and has been inadequately validated as a marker of oxidative damage (19, 21, 22). In the few population-based studies that have been conducted, how3 The abbreviations used are: ROS, reactive oxygen species: 8ohdG. 8-hydroxy2-deoxyguanosine: HPLC. high-performance liquid chromatography: PHA. Personal Health Appraisal; Ucr. urine creatinine. on August 27, 2017. © 1998 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from 92 H. pylon Infection and Urinary Excretion of SohdG 4 i. Parsonnet. unpublished data. ever, results conform to the hypothesis that urinary 8ohdG levels correlate with oxidative stress. For example, Lagorio et a!. (23) showed a positive linear relationship between exposure to benzene and levels of 8ohdG in human urine, whereas Loft et a!. (24) showed greater excretion of 8ohdG in smokers. As yet, however, there has been little consistency in these types of studies. Recently, an ELISA kit was developed to quantify levels of 8ohdG in human urine (Genox Corp., Baltimore, MD). The availability of this simple commercial assay provides the opportunity to evaluate 8ohdG excretion in large population studies. In our study, we hypothesized that, as a result of the active inflammatory process, urinary levels of 8ohdG would be higher in H. pylon-infected individuals than in uninfected persons. To test this hypothesis, we evaluated the association between H.pyion infection and 24-h urinary 8ohdG in healthy young adults, adjusting for diet and a measure of lean body mass. Materials and Methods Study Population. Subjects were drawn from a large pool of PHA participants within the Northern California Kaiser Permanente Medical Care Program. Kaiser Permanente Medical Care Program provides an optional lifestyle and health assessment for both new members and established members who have not seen a doctor in 5 or more years. For our study, we recruited a race/ethnicity stratified sample of healthy young adults (ages 20-39) in the PHA program who were not known to have any underlying inflammatory conditions. After providing informed consent, subjects were asked to complete a questionnaire regarding dietary intake, alcohol and caffeine consumption, exercise, and smoking and then to provide 10 ml of sera and a 24-h urine sample. The Health Habits and History Questionnaire was used to assess dietary intake of macroand micronutrients (25). This semiquantitative food frequency questionnaire, which has been validated in several different populations (26-29), elicits information about the type and quantity of foods eaten over the past month and the daily or weekly consumption of vitamin supplements. Of 3340 members who enrolled in the PHA between June 1 , 1992 and October 3 1 , 1993, 567 were eligible to participate in this study based on age, gender, race, and complete dietary and interview information. Twelve of these subjects reported improbable dietary intake and were excluded from further study. Of the remaining 555 subjects, 281 (50.8%) provided adequate 24-h urine samples and were included in the final analysis. Serum Assays. Sera were frozen at 70#{176}Cbefore testing. They were subsequently assayed by ELISA for the presence of IgG antibodies to H. pylon as described previously (30). One negative control, four low-titer H. pylon-positive controls, and one high-titer H. pylon-positive control were assayed at the same time as the study samples. The cutoff value for seropositivity was set at the mean level of the four low-titer H. pylon controls. This assay has been shown to be greater than 90% sensitive and specific for active H. pyloni infection, as confirmed by bacteriological culturing of pathological specimens from 58 infected and uninfected subjects from a separate study.4 A random selection of 100 sera was also tested for creatinine in a commercial laboratory using a modified Jaffe reaction (Nichols Laboratory, San Luis Obispo, CA). Urine Collection, Storage, and Assays. Urine was collected in plastic containers over a 24-h period by all study participants. Participants were told to void their first morning urine on the first day of collection and to collect every subsequent urine after that time, ending with the following day’s first urine. They were asked to return the urine sample within 24 h of the final collection or to freeze the sample if retained for greater than 24 h. Participants were asked whether the sample they returned represented the complete 24-h period. If the samples were deemed incomplete, the subject and the sample were not included in the analysis. Urine was stored at -70#{176}Cat the Kaiser Permanente Division of Research, where the total urine volumes were recorded. Individual aliquots were transported to our laboratory on dry ice and assayed for 8ohdG using a doubleantibody ELISA kit (Genox Corp.). This assay uses a murine monoclonal primary 8ohdG antibody that is specific to the 8ohdG adduct in urine (N45.l; Ref. 31). Urine samples were thawed and centrifuged to remove particulate matter. Six standards of known concentrations of 8ohdG (0.64, 3.2, 16, 80, 400, and 2000 ng/ml) and six blank wells were run concomitantly with triplicate urine samples. Fifty l of sample urine and 50 p.1 of primary monoclonal antibody reconstituted in 1% BSA/phosphoric acid buffer were added to a microtiter plate that was precoated with 8ohdG-BSA conjugate. Plates were sealed and incubated for 1 h at 37#{176}C, followed by a wash with 200 p3 of 0.05% Tween 20/phosphoric acid buffer. One hundred pA of horseradish peroxidase-conjugated anti-mouse polyclonal secondary antibody solution was then added to each well, incubated, and washed. One hundred p.1 of hydrogen peroxide/phosphoric acid-reconstituted substrate was then added to each well, followed by reaction termination with 100 l of 2 N sulfuric acid. Absorbance readings were taken 3 mm later with a spectrophotometer at 492 nm. The average of the absorbance readings from the triplicate runs of each individual sample was determined, and the value of 8ohdG for each subject was obtained via comparison to a standard reference curve. Creatinine values in urine were determined by a commercial laboratory using a colorimetric modified Jaffe reaction (Corning-Nichols Institute, San Luis Obispo, CA). Validation of the Urinary Assay. Studies using the 8ohdG ELISA assay report excellent specificity of the N45. 1 monoclonal antibody (31-33). An ELISA for the detection of 8ohdG used; the addition of known concentrations of deoxyribonucleosides including 2 ‘-deoxy-adenosine, 2’-deoxycytidine, 2’-deoxyuridine, and 2’-deoxyguanosine to urine samples had little effect on the amount of 8ohdG detected (3 1 ). Similarly, an immunohistochemical study of 8ohdG in renal carcinogenesis showed minimal cross-reactivity between the N45.l antibody and a wide array of purineand pyrimidine-containing compounds or their breakdown products that are ubiquitous in urine. The ELISA assay achieved excellent specificity, a lower limit of detection, and a lower risk of production of 8ohdG during the pretreatment steps associated with HPLC (32). When HPLC with electrochemical detection and the N45.l ELISA assay were used to measure 8ohdG content in identical urine samples, the two methods had a high degree of correlation. The sensitivity was substantially higher, however, with the ELISA assay (limit of detection, 1 ng/ml).5 The amount of 5 H. Ochi, T. Yoshikawa, R. Cutler, M. Takeuchi, and N. Ramarathnam. Development of a monoclonal antibody ELISA for the quantification of 8-hydroxy-2’deoxyguanosine. submitted for publication. on August 27, 2017. © 1998 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from Cancer Epidemiology, Blomarkers & Prevention 93 6 Japan Institute for the Control of Aging, Fukuroi, Shizuoka, unpublished data. Table I Characteristics of study participants (n 279)