A parallel study of in vitro sensitivity to benzo[a]pyrene diol epoxide and bleomycin in lung carcinoma cases and controls

1118

A Parallel Study of In Vitro Sensitivity to Benzo[a]pyrene Diol Epoxide and Bleomycin in Lung Carcinoma Cases and Controls Xifeng Wu, M.D., Ph.D.1,3 Jun Gu, M.D., M.S.1 Christopher I. Amos, Ph.D.1 Hong Jiang, M.S.1 Waun Ki Hong, M.D.2 Margaret R. Spitz, M.D.1 1

Department of Epidemiology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas.

2

Department of Thoracic/Head & Neck Medical Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, Texas.

3

School of Public Health, The University of Texas Health Science Center at Houston, Houston, Texas.

Presented in part at the Biomarkers in Risk Assessment Minisymposium, American Association for Cancer Research 88th Annual Meeting, San Diego, California, April 16, 1997. Supported by grants CA 55769 (given to M.R.S.) and CA 68437 (given to W.K.H.) from the National Cancer Institute and the Devereaux Award from the Cancer Research Foundation of America (given to X.W.). The authors thank Dr. T. C. Hsu for his invaluable suggestions regarding the study design and his critical review of the article, Dr. Maureen Goode for editorial assistance, and Ms. Susan Honn for recruiting study subjects. Dr. Waun Ki Hong is an American Cancer Society Clinical Research Professor. Address for reprints: Xifeng Wu, M.D., Ph.D., Department of Epidemiology, Box 189, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Blvd., Houston, TX 77030. Received September 15, 1997; revision received January 23, 1998; accepted March 9, 1998. © 1998 American Cancer Society

BACKGROUND. Because only a fraction of smokers develop neoplastic lesions, host factors may affect their susceptibility to the carcinogenic effects of tobacco smoke. Benzo[a]pyrene diol epoxide (BPDE) is the metabolic product of benzo[a]pyrene (B[a]P), a constituent of tobacco smoke. Therefore, BPDE sensitivity may shed some light on smoking-related carcinogenesis. METHODS. First, differential BPDE sensitivity was tested in five lymphoblastoid cell lines. Then sensitivity to BPDE and bleomycin (an excellent lung carcinoma risk predictor) was tested in parallel in the lymphocytes of 57 lung carcinoma cases and 82 controls. RESULTS. The optimal BPDE treatment duration was 24 hours. The xeroderma pigmentosum cell line was the most sensitive, followed by head and neck cancer, ataxia telangiectasia, and normal cells.The mean breaks per cell for cases and controls were 0.78 and 0.46, respectively (P , 0.0001). BPDE sensitivity was significantly associated with lung carcinoma, with an odds ratio (OR) of 7.26, compared with an OR of 4.56 for bleomycin sensitivity. There was also a dose-response correlation between the quartiles of BPDE-induced breaks and lung carcinoma risk, with ORs of 2.39, 3.12, and 15.03. It is noteworthy that individuals who were sensitive to both BPDE and bleomycin had a significantly increased OR of 38.36. CONCLUSIONS. BPDE sensitivity may be a biologic marker to identify individuals who are susceptible to the carcinogenic effects of tobacco smoke. BPDE and bleomycin sensitivity might represent different repair or sensitivity pathways; however, when these assays are used in parallel, they might refine our ability to identify high risk individuals. Cancer 1998;83:1118 –27. © 1998 American Cancer Society.

KEYWORDS: benzo[a]pyrene diol epoxide, mutagen sensitivity, lung carcinoma, susceptibility.

B

enzo[a]pyrene (B[a]P), one of the most thoroughly studied polycyclic aromatic hydrocarbons, is a constituent of combustion products, such as cigarette smoke. The genotoxicity of B[a]P results from its in vivo metabolic activation by mammalian detoxification enzyme systems to highly reactive electrophilic diol epoxide metabolites, such as benzo[a]pyrene diol epoxide (BPDE), which form covalent adducts in vitro and in vivo on interaction with DNA1– 4 and require excision repair.5– 6 It has been demonstrated that only a fraction of smokers develop neoplastic lesions7 and that host factors may affect susceptibility to carcinogenic action. Therefore, measuring sensitivity to BPDE may shed some light on smoking-related carcinogenesis. A number of genetic, cytogenetic, and biochemical assay systems

BPDE and Bleomycin Sensitivity in Lung Carcinoma/Wu et al.

can be used to measure BPDE sensitivity. For screening large numbers of individuals, the methodology should be as simple and as inexpensive as possible. One convenient test procedure is cytogenetic analysis of cultured lymphocytes. We have shown that bleomycin-induced mutagen sensitivity is an excellent predictor of lung carcinoma risk.8 –15 Wei et al.16 examined in vitro BPDE sensitivity in lung carcinoma cases and controls using the protocol established by Hsu et al.17 for bleomycin with 5-hour BPDE treatment duration. Because the molecular actions of bleomycin and BPDE differ, and because BPDE is a mutagen relevant to the study of lung carcinogenesis, we have extended the research of those studies to identify optimal treatment duration, dose-response correlations, and BPDE sensitivity. In this article, we report the cytogenetic responses of human lymphoblastoid cells to various doses and durations of BPDE. We also evaluate optimal BPDE sensitivity and bleomycin sensitivity in parallel in a subgroup of cases and controls from a molecular epidemiologic study of lung carcinoma to determine whether BPDE sensitivity could be used as a risk predictor for lung carcinoma and whether there are any associations between bleomycin sensitivity and BPDE sensitivity.

1119

ethnicity, and smoking status (current, former, and never).

Data Collection Epidemiologic data were collected by personal interview. After written informed consent was obtained, a structured interview lasting approximately 45 minutes was conducted by trained bilingual interviewers. Institutional guidelines for human subjects were followed. Data were collected on sociodemographic characteristics, recent and prior tobacco use, other life-style habits, and family history of cancer. At the completion of the interview, blood was drawn into sodium-heparinized tubes for cytogenetic and molecular genetic analyses. Laboratory personnel were blinded to case and control status.

BPDE Solution Powder ([6]-anti-BPDE) was purchased from Midwest Research Institute (Kansas City, MO). We used tetrahydrofuran (Sigma Chemical Co., St. Louis, MO) as the solvent. Aliquots of the stock concentration of 1 mM were placed in microcentrifuge tubes (500 mL each) and stored at 220°C in the dark.

Lymphoid Line Experiments

MATERIALS AND METHODS Cell Lines For our preliminary experiments, we used one lymphoid xeroderma pigmentosum (XP) line (GM02485b) and two lymphoid ataxia telangectasia (AT) lines (AT-1 [GM01525c] and AT-2 [GM01526c]) (Coriell Institute of Medical Research, Camden, NJ). The other two lines used (3590P, from a normal blood donor, and 3640P, from a patient with head and neck cancer) were transformed with Epstein–Barr virus in the laboratory of Hsu et al.18 The XP line is sensitive to ultraviolet light and 4-nitroquinoline-1-oxide (4-NQO, an ultraviolet mimetic agent), the two AT lines are sensitive to g-radiation and bleomycin (a radiomimetic agent), and 3640P is sensitive to 4-NQO and to bleomycin, whereas 3590P is resistant to both bleomycin and 4-NQO.19 –20

Study Subjects The cases were 57 newly diagnosed, previously untreated lung carcinoma patients identified at The University of Texas M. D. Anderson Cancer Center. There were no gender, histology, age, or stage restrictions. All patients were white. Eighty-two healthy controls (with no previous cancer history) were identified from the rosters of the largest multispecialty health care group practice in the Houston metropolitan area and were matched to the cases by age (65 years), gender,

Two sets of experiments, each repeated 3 times, were conducted as follows: 1) BPDE treatment duration was assessed by harvesting cells after 1, 3, 5, 6.5, 8, 12, and 24 hours of BPDE treatment to determine the optimal duration of treatment with various doses; and 2) BPDE dose response was measured in cultures treated with BPDE concentrations of 0, 0.1, 0.2, 0.3, and 0.4 mM. Standard RPMI-1640 medium supplemented with 20% fetal bovine serum was used for all cultures. The initial inoculum was 10 mL of growth medium with approximately 5 3 105 cells/mL. These cultures were used for various experiments after 24 hours of incubation.

Primary Blood Cultures For primary blood cultures, the protocol was similar to that for the lymphoid cells, except that each bloodsample culture was initiated with growth medium containing 1% phytohemagglutinin and the cultures were routinely incubated for 72 hours. For the BPDE sensitivity assay, BPDE was added to each culture to a final concentration of 2 mM. The treatment time for the case– control study (24 hours) was selected after analysis of the results of the first set of experiments on lymphoid lines. For the bleomycin sensitivity assay, we used the established protocol.17 Bleomycin was added to each culture to a final concentration of 0.03 U/mL with a treatment duration of 5 hours.

1120

CANCER September 15, 1998 / Volume 83 / Number 6

FIGURE 1. Chromatid breaks induced by various doses of benzo[a]pyrene diol epoxide (BPDE) are shown for various amounts of time in cells from an xeroderma pigmentosum patient (XP) and a normal donor (3590P). The values shown are the mean values of three separate experiments.

Cytogenetic Analysis Cell harvesting followed the standard procedure (1 hour of colcemid arrest, hypotonic KCl pretreatment, fixation in Carnoy’s solution, washing, and air-drying). The slides were subjected to Giemsa staining without banding. The number of breaks in 50 metaphases per sample was counted and expressed as the average number of breaks per cell. Only frank chromatid breaks or exchanges were recorded; chromatid gaps or attenuated regions were disregarded. The slides were coded and read without knowledge of BPDE dose or each subject’s case or control status.

Statistical Analysis BPDE sensitivity and bleomycin sensitivity were analyzed as a continuous variable by using Student’s t test and also as a categoric variable dichotomized at the level of the 75th percentile of breaks per cell in the control groups (using 0.58 breaks per cell [b/c] as a cutoff point for BPDE sensitivity and 0.68 b/c for bleomycin sensitivity). We repeated the analysis with other cutoff points and found that the results did not change substantially. A smoker was defined as a person who had smoked 100 cigarettes in his or her lifetime. Former smokers were defined as smokers who had not smoked for at least 1 year. Age, number of cigarettes smoked, and pack-years were analyzed as continuous variables or categorized by the median value in the control population. Partial correlation coefficients were calculated to evaluate the correlation between variables. To test for significant associations among

mutagen sensitivity (both to BPDE and bleomycin), tobacco use, and cancer risk, univariate odds ratios (ORs) were calculated as estimates of the relative risk. Ninety-five percent confidence intervals (CIs) were computed by the method of Woolf.21 Simple stratified analysis was used to test for interactions between chromosomal breaks and smoking status. Multiple logistic regression was conducted with STATA statistical software.22

RESULTS Effects of Treatment Duration on Lymphoid Lines The most commonly induced chromatid aberrations were simple breaks, isochromatid breaks, and chromatid exchanges. Figure 1 summarizes the mean number of chromatid breaks per cell induced by various BPDE doses and durations of treatment in cells from an XP patient and a normal donor (3590P). Each value was the average of three separate experiments. After treatment with BPDE at final concentrations of 0.1, 0.2, 0.3, and 0.4 mM for 24 hours, the mean breaks per cell value for the XP line versus 3590P were 0.37 versus 0.19, 0.83 versus 0.33, 1.15 versus 0.41, and 1.27 versus 0.54, respectively. It is clear from Figure 1 that the optimal treatment duration for estimating differential BPDE sensitivity in the different cell lines was 24 hours.

Effects of BPDE Dose on Lymphoid Lines Figure 2 shows the dose-response effect of BPDE at the 24-hour time point. When the dose of BPDE was

BPDE and Bleomycin Sensitivity in Lung Carcinoma/Wu et al.

1121

FIGURE 2. The effects of various benzo[a]pyrene diol epoxide (BPDE) doses on induced chromatid breaks in five lymphoid lines are shown. The treatment duration was 24 hours. The values shown are the mean values from three separate experiments. XP: xeroderma pigmentosum; AT: ataxia telangectasia. increased from 0.1 to 0.4 mM, the breaks per cell increased proportionally in the 5 test cell lines. The XP cell line exhibited the greatest sensitivity, followed by the cells of the patient with head and neck cancer and those of the two AT patients. The normal cell line had the lowest sensitivity. Nonlinear dose-response patterns were evident when the BPDE concentration was higher than 0.5 mM (data not shown). Table 1 summarizes the mean values and standard deviations plotted in Figures 1 and 2 to document the variability among the three separate experiments.

Differential BPDE Sensitivity in Healthy Controls and Lung Carcinoma Patients To produce similar genotoxicity, the final concentration of BPDE added to primary blood cultures had to be about 10 times that used for the lymphoblastoid cell lines based on our previous experience. Therefore, we used final BPDE concentrations of 1, 2, 2.5, 3, and 3.5 mM for primary blood cultures to determine the optimal dose. The mean breaks per cell values were 0.29 for 1 mM BPDE, 0.59 for 2.0 mM BPDE, and 0.55 for 2.5 mM BPDE. At 3.0 mM BPDE, there were only a few metaphases. Cells treated with 1 mM BPDE had improved nuclear morphology compared with higher BPDE doses but fewer breaks. The mean breaks per cell value for 2 mM BPDE was similar to that for 2.5 mM BPDE, but the cells had improved nuclear morphology (data not shown). Therefore, we selected 2 mM as the final concentration for the test mutagen in the following human population study. The molecular epidemiologic component is based

on 57 white lung carcinoma patients and 82 controls. There were no significant differences in age, gender, smoking status, pack-years, and number of cigarette smoked between cases and controls (Table 2). The mean age was 64 years for the cases and 65 years for the controls (P 5 0.39). The age range was 38 – 84 years. The mean pack-years of smoking were similar for the cases (45) and controls (46), as were the mean number of cigarettes smoked per day (P 5 0.38). The mean ages at smoking initiation were 17 years for the cases and 18 years for the controls (P 5 0.26). The mean age at smoking cessation was age 54 years for the cases, compared with age 52 years for the controls. BPDE sensitivity data were not available for six of the cases and four of the controls. The bleomycin sensitivity data were not available for one case. Figure 3 summarizes the profile of BPDE sensitivity for lung carcinoma patients and healthy controls. There were significant differences in overall BPDE sensitivity between cases and controls, with mean breaks per cell for cases and controls of 0.78 and 0.46, respectively (P , 0.0001). Nearly 69% of the cases, compared with only 22% of the controls, exhibited BPDE sensitivity (i.e., had $0.58 breaks per cell, Table 3). After adjustment for age and gender, BPDE sensitivity was associated with a statistically significant elevated risk for lung carcinoma (odds ratio [OR] of 8.28, 95% confidence interval [CI] of 3.65–18.78). After adjustment for age, gender, and pack-years, the OR (95% CI) was 8.86 (3.83–20.47). After adjustment for age, gender, pack-years, and bleomycin sensitivity, the OR (95% CI) was 7.26 (3.00 –17.58). Among the same sub-

1122

CANCER September 15, 1998 / Volume 83 / Number 6

TABLE 1 Effects of BPDE Treatment Duration and Dose on Lymphoid Cell Lines Time points (hrs) 1

3

5

6.5

Doses (mM) 3590 P 0.1 0.2 0.3 0.4 XP 0.1 0.2 0.3 0.4

8

12

24

Mean breaks per cell 6 SD 0.15 6 0.04 0.17 6 0.05 0.21 6 0.05 0.23 6 0.04

0.13 6 0.03 0.17 6 0.03 0.21 6 0.05 0.21 6 0.07

0.25 6 0.09 0.40 6 0.12 0.40 6 0.14 0.51 6 0.11

0.24 6 0.09 0.35 6 0.11 0.37 6 0.12 0.43 6 0.12

0.12 6 0.03 0.17 6 0.03 0.24 6 0.07 0.21 6 0.08

0.15 6 0.08 0.15 6 0.07 0.19 6 0.08 0.26 6 0.13

0.19 6 0.02 0.33 6 0.03 0.41 6 0.06 0.54 6 0.07

0.19 6 0.05 0.24 6 0.07 0.32 6 0.08 0.37 6 0.06

0.15 6 0.03 0.19 6 0.06 0.25 6 0.09 0.34 6 0.09

0.45 6 0.14 0.49 6 0.16 0.63 6 0.03 0.58 6 0.14

0.46 6 0.19 0.46 6 0.16 0.62 6 0.03 0.56 6 0.09

0.13 6 0.02 0.20 6 0.04 0.30 6 0.05 0.25 6 0.07

0.17 6 0.10 0.19 6 0.08 0.22 6 0.09 0.35 6 0.10

0.37 6 0.11 0.83 6 0.14 1.15 6 0.21 1.27 6 0.20

Doses (mM) 0.1

0.2

0.3

0.4

0.36 6 0.07 0.61 6 0.25 0.57 6 0.14 0.84 6 0.10 1.13 6 0.17

0.57 6 0.03 0.93 6 0.02 0.67 6 0.06 0.75 6 0.08 1.28 6 0.14

Mean breaks per cell 6 SD

Cell lines 0.15 6 0.03 0.31 6 0.10 0.29 6 0.08 0.46 6 0.10 0.31 6 0.03

3590P AT-1 AT-2 3640P XP

0.21 6 0.08 0.38 6 0.15 0.43 6 0.17 0.51 6 0.04 0.62 6 0.16

BDPE: benzo[a]pyrene diol epoxide; 3590P: sample from a normal blood donor; 3640P: sample from a patient with head and neck cancer; XP: xeroderma pigmentosum; AT: ataxia telangectasia; SD: standard deviation.

TABLE 2 Distribution of Select Host Characteristics, According to Case–Control Status n (%)

Gender Male Female Mean age, yrs (SD) Smoking status Never Former Current Pack-yrs (SD) No. of cigarettes smoked per day

Cases (n 5 57)

Controls (n 5 82)

P value

28 (49.12) 29 (50.88) 63.56 (8.86)

45 (54.88) 37 (45.12) 64.89 (8.96)

0.504 0.389

8 (14.03) 27 (47.37) 22 (38.60) 45.20 (37.28)

6 (7.32) 49 (59.75) 27 (32.93) 46.07 (30.70)

0.252 0.882

23.91 (17.50)

26.33 (14.95)

0.384

SD: standard deviation.

jects, 64% of the cases exhibited bleomycin sensitivity (i.e., had $0.68 breaks per cell), compared with only 26% of the controls. After adjustment for age, gender, pack-years, and BPDE sensitivity, the OR (95% CI) for bleomycin sensitivity was 4.56 (1.91–10.85). Because

we have published much data on bleomycin sensitivity previously, we do not present the bleomycin data here in detail. When subjects were categorized into quartiles of BPDE-induced breaks per cell based on the distribution in the controls (with ,0.30 breaks per cell as the referent category), there was a dose-response correlation between lung carcinoma risk and degree of BPDE sensitivity. The adjusted ORs (95% CIs) for individuals in increasing quartiles of breaks per cell were 3.19 (0.68 –14.94), 4.54 (0.98 –21.09), and 22.72 (5.69 –90.71), respectively. The results of the chi-square test for trend were significant (P , 0.0001). When the data were dichotomized at median age at lung carcinoma diagnosis, the ORs for BPDE sensitivity were markedly higher for younger patients (OR 5 17.44; 95% CI 5 5.11–59.55) than for older patients (OR 5 4.86; 95% CI 5 1.31–18.00). When the data were categorized by median number of cigarettes smoked per day, the ORs for BPDE sensitivity were higher for lighter smokers (OR 5 17.56; 95% CI 5 4.47– 68.99) than for heavy smokers (OR 5 5.09; 95% CI 5 1.53–16.95). The BPDE sensitivity profiles in the controls and

BPDE and Bleomycin Sensitivity in Lung Carcinoma/Wu et al.

1123

FIGURE 3. The percentage distribution of controls and lung carcinoma patients is shown according to response to benzo[a]pyrene diol epoxide (BPDE).

lung carcinoma cases were also assessed by age, gender, and smoking status (Table 4). Although the mean number of breaks per cell for smokers was slightly higher than that for nonsmokers, there were no significant associations between mean breaks per cell and age, gender, pack-years, number of cigarettes smoked per day, duration of smoking cessation, or smoking status for cases or controls. A similar observation was also noted for bleomycin sensitivity. A stratified analysis was performed to examine the interactions between bleomycin sensitivity and BPDE sensitivity (Table 5). Subjects who were not sensitive to either mutagen served as the referent group. The OR (95% CI) for those bleomycin sensitive only was 4.19 (1.28 –13.64); for those BPDE sensitive only, the OR was 7.57 (2.24 –25.63). For individuals who were sensitive to both mutagens, the OR was 38.36 (9.83– 149.67). A synergistic interaction between bleomycin sensitivity and BPDE sensitivity was suggested.

DISCUSSION Although the chromosome breakage syndromes reflect the extreme end of the susceptibility phenotype, within the general population there may be individuals with latent instability that is unmasked by mutagen challenge in vitro. The bleomycin sensitivity assay has been well established by Hsu et al.17 By using a molecular epidemiologic approach, Spitz et al. demonstrated that bleomycin sensitivity was an independent cancer risk predictor for lung carcinoma, head and

neck cancers, and multiple primary cancers after an initial head and neck cancer.8 –15 However, different mutagens act on cells through different molecular mechanisms, and so may activate different repair pathways. Bleomycin, a radiomimetic drug, causes single-strand and double-strand DNA breaks at every stage of the cell cycle. Repair of DNA lesions primarily requires the action of polymerases. However, mutagens that cause genetic damage by different molecular mechanisms may induce different repair pathways and cancer susceptibility. A person who is sensitive to one mutagen can be resistant to others.23 BPDE forms covalent adducts on interaction with DNA in vitro or in vivo1– 4 and requires an excision repair system.5– 6 Because it is a major tobacco carcinogen, BPDE might be a more suitable test mutagen than bleomycin for studying carcinogenesis in the lung and assessing cancer risk. We established that 24 hours was the optimal duration of BPDE treatment for revealing BPDE sensitivity (for the bleomycin assay, we used a 5-hour treatment). Most of the chromosome aberrations we observed at 24 hours were of the chromatid type, supporting our hypothesis that, to observe frequent chromatid breaks, cells after treatment with BPDE must go through the S-phase. Previously, we tested 4NQO at various time points 3, 5, 8, 16, and 24 hours for harvesting and found that 16 and 24 hours were the optimal time points. It appears that BPDE and 4-NQO act in a similar way. It might be that all adduct induc-

1124

CANCER September 15, 1998 / Volume 83 / Number 6

TABLE 3 Risk Estimates of BPDE Sensitivity in Lung Carcinoma Patients and Healthy Controls n (%) BPDE Sensitivity Breaks per cell $0.58 (b/cell) ,0.58 (b/cell) Breaks per cell ,0.30 0.30–0.45 0.46–0.57 $0.58 Age at diagnosis ,68 yrs $0.58 (b/cell) ,0.58 (b/cell) 681 yrs $0.58 (b/cell) ,0.58 (b/cell) Cigarettes ,22 per day $0.58 (b/cell) ,0.58 (b/cell) 221 per day $0.58 (b/cell) ,0.58 (b/cell)

Cases (51)

Controls (78)

Adjusted OR (CI)a

Adjusted OR (CI)b

Adjusted OR (CI)c

35 (68.63) 16 (31.37)

17 (21.79) 61 (78.21)

8.28 (3.65–18.78)

8.86 (3.83–20.47)

7.26 (3.00–17.58)

3 (5.88) 6 (11.76) 7 (13.73) 35 (68.63)

26 (33.33) 19 (24.36) 16 (20.51) 17 (21.80)

1 2.81 (0.61–12.89) 4.48 (0.97–20.61) 20.12 (5.18–78.19)

1 3.19 (0.68–14.94) 4.54 (0.98–21.09) 22.72 (5.69–90.71)

1 2.39 (0.48–11.83) 3.12 (0.64–15.28) 15.03 (3.60–62.76)

10 (62.50) 6 (37.50)

11 (28.95) 27 (71.05)

14.81 (4.73–46.39)

17.44 (5.11–59.55)

14.34 (3.95–52.10)

25 (71.43) 10 (28.57)

6 (15.00) 34 (85.00)

4.26 (1.21–15.03)

4.86 (1.31–18.00)

3.62 (0.92–14.22)

22 (75.86) 7 (24.14)

8 (21.05) 30 (78.95)

12.74 (3.78–42.96)

17.56 (4.47–68.99)

14.47 (3.58–58.59)

13 (59.09) 9 (40.91)

9 (22.50) 31 (77.50)

5.30 (1.67–16.78)

5.09 (1.53–16.95)

4.91 (1.19–20.27)

BPDE: benzo[a]pyrene diol epoxide; OR: odds ratio; CI: confidence interval. a Adjusted by age and gender. b Adjusted by age, gender, and pack-yrs. c Adjusted by age, gender, pack-yrs, and bleomycin sensitivity.

ers require an excision repair system. We mainly use chromatid breaks as predictors of mutagen sensitivity. With chemical treatment duration longer than 24 hours (e.g., 36 – 48 hours), most surviving cells will go through one cell cycle, and chromosome-type aberrations such as translocations, dicentric chromosomes, and ring chromosomes will become dominant. At that time, chromatid breaks alone are not complete reflections of the real damage. Some reciprocal translocations will appear normal under the conventional chromosome staining. Chromosome banding techniques should be used to avoid underestimation. For radiomimetic mutagens such as bleomycin, the genetic action of which is to break single-strand and doublestrand DNA fibers, shorter time treatment might be optimal. Actually, we found that the now-standardized 5-hour bleomycin exposure can easily be changed to 3 hours or even 2 hours without drastically altering the conclusion. Our data also suggest that lung carcinoma cases and controls exhibit different sensitivities to BPDE. BPDE sensitivity was associated with an overall risk for lung carcinoma, with an adjusted OR of 7.26. For the same subjects, we found that bleomycin sen-

sitivity was associated with an overall risk for lung carcinoma, with an adjusted OR of 4.56. As with our data on bleomycin sensitivity, the ORs for BPDE sensitivity were even higher for lighter smokers and younger patients, supporting our hypothesis that mutagen sensitivity constitutes a susceptible phenotype. There was also a dose-response correlation between the quartiles of BPDE-induced breaks and lung carcinoma risk with ORs of 3.19, 4.54, and 22.72. BPDE sensitivity thus appears to be a useful biologic marker for identifying individuals who are sensitive to this tobacco mutagen. Regarding the underlying mechanism, one may speculate whether BPDE sensitivity is related to chromatin structure or DNA repair efficiency. Pandita and Hittelman suggested that chromatin organization and DNA conformation are essential for breakage and recombination at fragile sites.24 By comparing the adduct pattern in BPDE-treated HeLa cells with the pattern in BPDE-treated isolated genomic DNA, Denissenko et al.25 demonstrated that the two patterns were almost identical. This finding suggested that chromatin structure was not a major molecular factor. Wei et al.26 reported that the site specific rates

BPDE and Bleomycin Sensitivity in Lung Carcinoma/Wu et al.

1125

TABLE 4 Mean BPDE-Induced Breaks per Cell, According to Host Characteristics Healthy controls

Characteristic Age (yrs) ,68 $68 Gender Male Female Pack-yrs ,45 $45 Cigarettes smoked/day ,22 $22 Smoking cessation duration (yrs) $13 ,13 Smoking status Never Former Current

Lung carcinoma patients

n

Mean breaks per cell (SD)

P value

n

Mean breaks per cell (SD)

P value

40 38

0.45 (0.21) 0.48 (0.20)

0.56

35 16

0.77 (0.38) 0.81 (0.30)

0.68

44 34

0.46 (0.20) 0.46 (0.21)

0.92

26 25

0.83 (0.35) 0.73 (0.35)

0.30

37 41

0.44 (0.23) 0.48 (0.18)

0.29

19 26

0.87 (0.43) 0.74 (0.31)

0.25

38 40

0.48 (0.22) 0.45 (0.19)

0.56

23 22

0.86 (0.35) 0.73 (0.38)

0.27

24 23

0.42 (0.21) 0.51 (0.23)

0.17

9 14

0.74 (0.35) 0.83 (0.25)

0.47

5 47 26

0.36 (0.22) 0.47 (0.22) 0.47 (0.16)

0.32 0.18

6 23 22

0.66 (0.21) 0.79 (0.29) 0.80 (0.44)

0.29 0.45

BPDE: benzo[a]pyrene diol epoxide; SD: standard deviation.

TABLE 5 Interaction of Bleomycin Sensitivity and BPDE Sensitivity in Lung Carcinoma Risk Bleomycin sensitivity

BPDE sensitivity

Case group n

Control group n

Adjusted OR (CI)a

Adjusted OR (CI)b

Nonsensitive Sensitive Nonsensitive Sensitive

Nonsensitive Nonsensitive Sensitive Sensitive

7 9 11 23

47 14 12 5

1.00 4.08 (1.26–13.20) 6.03 (1.19–19.04) 34.23 (9.25–126.71)

1.00 4.19 (1.28–13.64) 7.57 (2.24–25.63) 38.36 (9.83–149.67)

BPDE: benzo[a]pyrene diol epoxide; OR: odds ratio; CI: confidence interval. a Adjusted by age and gender. b Adjusted by age, gender, and pack-yrs.

of excision repair of BPDE adducts correlated with the overall mutation spectra. Inefficient DNA repair plays an important role in the formation of BPDE adducts and mutation hot spots.26 –27 A correlation between DNA repair capability and mutagen sensitivity has also been reported.28 We found that the lymphoblastoid cells from an XP cell line were the most sensitive, followed by the cells of the patient with head and neck cancer and those of the two AT patients. The cell line of a normal donor had the lowest sensitivity. Although in vitro cytogenetic analyses have been fairly commonly used to study individual sensitivity to genotoxicity and cancer risk in both case– control studies8 –15 and cohort studies,29 –30 one may ask whether BPDE sensitivity in lymphocytes repre-

sents damage in the target tissue (lung tissue). Denissenko et al.25 found a similar adduct pattern in three different cell types: HeLa cells, bronchial cells, and normal human fibroblasts. This suggests that the same adduct pattern is likely to be present in different target cells for lung tissue transformation. We have shown that age, gender, and smoking status do not independently modify the BPDE sensitivity profiles of cases and controls. This result is similar to the previous finding for bleomycin sensitivity.15 We have also shown that tumor stage does not affect the sensitivity profile.15 The specificities of the bleomycin and BPDE assays appear to be similar (75.6% and 78.2%, respectively). The sensitivities of the two assays are also similar (64% for

1126

CANCER September 15, 1998 / Volume 83 / Number 6

the bleomycin assay and 68% for the BPDE assay). A synergistic interaction between BPDE sensitivity and bleomycin sensitivity was suggested in the current study. Only 5 (6.4%) of 78 controls exhibited sensitivity to both bleomycin and BPDE, compared with 23 (46.0%) of 50 cases. Therefore, BPDE and bleomycin sensitivity might represent different repair pathways. However, these assays might facilitate the identification of high risk populations when used together. There were some limitations to this assay. The lowest BPDE concentration tested (1 mM) preserved nuclear morphology best but produced b/c values too low for analysis. In cells exhibiting suitable break frequency, the cell nuclear morphology was not as good as in bleomycin-treated cells. With increasing BPDE concentrations of 1, 2, 2.5, 3.0, and 3.5 mM, the quality of the slides also dramatically decreased: the percentages of well-spread metaphases were 74.6%, 58.8%, 45.3%, 12.7%, 3.6%, and 0%, respectively. The mean b/c value for 2 mM BPDE was similar to that for 2.5 mM BPDE, but the nuclear morphology was better. In summary, several new observations from this study provide a more complete understanding of the BPDE sensitivity assay. BPDE sensitivity based on quantification of BPDE-induced chromatid breaks in cultured lymphocytes may be a useful biologic marker of lung carcinoma development. BPDE and bleomycin sensitivity might represent different repair pathways, but when these assays are used in parallel they might refine our ability to identify high risk populations.

REFERENCES 1. 2.

3.

4.

5.

6.

Phillips DH. Fifty years of benzo[a]pyrene. Nature 1983;303: 468 –72. Arce GT, Allen JW, Doerr CL, Elmore E, Hatch GG, Moore MM, et al. Relationships between benzo[a]pyrene-DNA adduct levels and genotoxic effects in mammalian cells. Cancer Res 1987;47:3388 –95. Wolterbeek AP, Roggeband R, Steenwinkel MJ, Baan RA, Rutten AAJ. Formation and repair of benzo[a]pyrene-DNA adducts in cultured hamster tracheal epithelium determined by 32P-postlabeling analysis and unscheduled DNA synthesis. Carcinogenesis 1993;14:463–7. Shou M, Harvey RG, Penning TM. Reactivity of benzo[a]pyrene-7,8-dione with DNA: evidence for the formation of deoxyguanosine adducts. Carcinogenesis 1993;14: 475– 82. Tang MS, Pierce JR, Doisy RP, Nazimiec ME, Macleod MC. Differences and similarities in the repair of two benzo[a]pyrene diol epoxide isomers induced DNA adducts by uvrA, uvrB, and uvrC gene products. Biochemistry 1992;31: 8429 –36. Van Houten B, Masker WE, Carrier WL, Regan JD. Quantitation of carcinogen-induced DNA damage and repair in

7. 8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21. 22. 23.

human cells with the UVR ABC excision nuclease from Escherichia coli. Carcinogenesis 1986;7:83–7. Law MR. Genetic predisposition to lung cancer. Br J Cancer 1990;61:195–206. Spitz MR, Hsu TC, Wu XF, Fueger JJ, Amos CI, Roth JA. Mutagen sensitivity as a biologic marker of lung cancer risk in African Americans. Cancer Epidemiol Biomarkers Prev 1995;4:99 –103. Spitz MR, Fueger JJ, Beddingfield NA, Annegers JF, Hsu TC, Newell GR, et al. Chromosome sensitivity to bleomycin-induced mutagenesis, an independent risk factor for upper aerodigestive tract cancers. Cancer Res 1989;49: 4626 – 8. Spitz MR, Fueger JJ, Halabi S, Schantz SP, Sample D, Hsu TC. Mutagen sensitivity in upper aerodigestive tract cancer: a case– control analysis. Cancer Epidemiol Biomarkers Prev 1993;2:329 –33. Wu XF, Hsu TC, Annegers JF, Amos CI, Fueger JJ, Spitz MR. A case– control study of nonrandom distribution of bleomycin-induced chromatid breaks in lymphocytes of lung cancer patients. Cancer Res 1995;55:557– 61. Wu XF, Delclos GL, Annegers FJ, Bondy ML, Honn SE, Henry B, et al. A case– control study of wood-dust exposure, mutagen sensitivity, and lung-cancer risk. Cancer Epidemiol Biomarkers Prev 1995;4:583– 8. Strom SS, Wu XF, Sigurdson AJ, Hsu TC, Fueger JJ, Lopez J, et al. Lung cancer, smoking patterns, and mutagen sensitivity in Mexican-Americans. Monogr Natl Cancer Inst 1995;18: 29 –33. Wu XF, Hsu TC, Spitz MR. Mutagen sensitivity exhibits a dose-response relationship in case– control studies. Cancer Epidemiol Biomarkers Prev 1996;5:577– 8. Spitz MR, Hoque A, Trizna Z, Schantz SP, Amos CI, King TM, et al. Mutagen sensitivity as a risk factor for second malignant tumors following upper aerodigestive tract malignancies. J Natl Cancer Inst 1994;86:1681– 4. Wei Q, Gu J, Cheng L, Bondy ML, Jiang H, Hong WK, et al. Benzo[a]pyrene diol epoxide–induced chromosomal aberrations and risk of lung cancer. Cancer Res 1996;56: 3975–9. Hsu TC, Johnston DA, Cherry LM, Ramkisson D, Schantz SP, Jessup JM, et al. Sensitivity to genotoxic effects of bleomycin in humans: possible relationship to environmental carcinogenesis. Int J Cancer 1989;43:403–9. Hsu TC, Shillitoe EJ, Cherry LM, Lin Q, Schantz SP, Furlong C. Cytogenetic characterization of 20 lymphoblastoid lines derived from human individuals differing bleomycin sensitivity. In Vitro Cell Devel Biol 1990;26:80 – 4. Hsu TC, Savaraji N, Trizna Z, Feun L, Furlong C, Schantz S, et al. Cytogenetic studies on the in vitro genotoxicity of 4-nitroquinoline-1-oxide on human lymphocytes. Int J Oncol 1993;3:823– 6. Wang RY, Hsu TC, Liang JC. In situ nick translation, an alternative to cytogenetics in analyzing mutagen effects on human cells. Int J Oncol 1993;2:873– 6. Woolf B. On estimating the relation between blood group and disease. Ann Hum Genet 1995;19:251–3. STATA Release 3.1 Reference Manual, 6th edition. STATA Corporation, College Station, TX, 1993. Hsu TC, Feun L, Trizna Z, Savaraj N, Shirley L, Furlong CL, et al. Differential sensitivity among three human subpopulations in response to 4-nitroquinoline-1-oxide and to bleomycin. Int J Oncol 1993;3:827–30.

BPDE and Bleomycin Sensitivity in Lung Carcinoma/Wu et al. 24. Pandita TK, Hittelman WN. Evidence of a chromatin basis for increased mutagen sensitivity associated multiple primary malignancies of the head and neck. Int J Cancer 1995; 61:738 – 43. 25. Denissenko MF, Pao A, Tang M, Pfeifer GP. Preferential formation of benzo[a]pyrene adducts at lung cancer mutational hotspots in p53. Science 1996;274:430 –2. 26. Wei D, Maher VM, McCormick JJ. Site-specific rates of excision repair of benzo[a]pyrene diol epoxide adducts in the hypoxanthine phosphoribosyltransferase gene of human fibroblasts: correlation with mutation spectra. Proc Natl Acad. Sci U S A 1995;92:2204 – 8. 27. Tornaletti S, Pfeifer GP. Slow repair of pyrimidine dimers at p53 mutation hotspots in skin cancer. Science 1994;263:1436–8.

1127

28. Wei Q, Spitz MR, Gu J, Cheng L, Xu X, Strom SS, et al. DNA repair capacity correlates with mutagen sensitivity in lymphoblastoid cell lines. Cancer Epidemiol Biomarkers Prev 1996;5:199 –204. 29. Hagmar L, Brogger A, Hansteen I, Heim S, Hogstedt B, Knudsen L, et al. Cancer risk in humans predicted by increased levels of chromosome aberrations in lymphocytes: nordic study group on the health risk of chromosome damage. Cancer Res 1994;54:2919 –22. 30. Bonassi S, Abbondandolo A, Camurri L, Dal Pra L, De Ferrari M, Degrassi F, et al. Are chromosome aberrations in circulating lymphocytes predictive of future cancer onset in humans? Preliminary results of an Italian cohort study. Cancer Genet Cytogenet 1995;79:133–5.