Lung Tumor Development in Mice Exposed to Tobacco Smoke and ...

TOXICOLOGICAL SCIENCES 69, 23–29 (2002) Copyright © 2002 by the Society of Toxicology

Lung Tumor Development in Mice Exposed to Tobacco Smoke and Fed ␤-Carotene Diets Ute C. Obermueller-Jevic,* ,1 Imelda Espiritu,† Ana M. Corbacho,* Carroll E. Cross,* and Hanspeter Witschi† ,2 *Center for Comparative Respiratory Biology and Medicine, Department of Medicine, School of Medicine, University of California, Davis, California; and †Center for Health and the Environment and Department of Molecular Biosciences, School of Veterinary Medicine, University of California, Davis, California 95616 Received February 19, 2002; accepted May 7, 2002

cers combined (Szabo, 2001). It also has become obvious that it will be next to impossible to eliminate the cause of most lung cancers worldwide—the smoking of cigarettes and other tobacco products. Although a decrease in smoking trends in the U.S. is encouraging, there remains concern with regard to subpopulations where smoking habits appear to be increasing (Wingo et al., 1999). Tobacco consumption is expanding worldwide, and it is anticipated that death rates will continue to be substantial for the foreseeable future (Lam et al., 2001; Pisani et al., 1999). The emerging field of chemoprevention is now recognized as a new basic and clinical science (Hong and Sporn, 1999). Attempts to find drugs or naturally occurring constituents of the diet that might provide some protection against the development of cancers in populations at high risk go back for many years. There have been some successes, notably with breast and colon cancers. However, the field looks much less promising when it comes to lung cancer, which has the highest frequency of death. As was recently stated, “Hypothesis-driven chemoprevention of lung cancers, when put to the test of randomized large-scale clinical trials, so far has been disappointing, unlike important successes with selective estrogen receptors modulators for breast cancer and non-steroidal antiinflammatory drugs and cyclooxygenase-2 inhibitors for colon cancer” (Omenn, 2000, p. 959). This statement was made after a major disappointment was experienced in three major clinical trials with ␤-carotene (Omenn, 1998). At one time, ␤-carotene had seemed to be an ideal chemopreventive agent, particularly against lung cancer. In epidemiological studies some 30 years ago, inverse associations were discovered between estimated intakes of vitamin A or ␤-carotene (provitamin A) and the risk of developing cancer at various sites (Peto et al., 1981). Experiments with cell and organ cultures, when combined with numerous observational epidemiological studies, seemed to furnish an experimental basis for the anticipated beneficial effect. On the other hand, preclinical studies with animal models had never really provided convincing evidence that ␤-carotene would be an effective chemopreventive agent for respiratory tract tumors, as

In human clinical trials it was found that the putative chemopreventive agent ␤-carotene not only failed to protect active smokers against the carcinogenic action of tobacco smoke, but actually increased their risk of developing lung cancer. In preclinical animal studies, ␤-carotene had been effective against some chemically induced cancers, but not against tumors in the respiratory tract. We exposed male strain A/J mice to tobacco smoke at a concentration of 140 mg/m 3 of total suspended particulate matter, 6 h a day, 5 days a week, for either 4 or 5 months, followed by a recovery period in air for 4 or 5 months, or for 9 months without recovery period. ␤-carotene was added in the form of gelatin beadlets to the AIN-93G diet either during or following tobacco smoke exposure at concentrations of 0.005, 0.05 and 0.5%. In the supplement-fed animals, plasma and lung levels of ␤-carotene were higher than they were in animals fed control diets. Exposure to tobacco smoke increased rather than decreased plasma ␤-carotene levels, but had no significant effect on lung levels. After 9 months, lung tumor multiplicities and incidence were determined. Tobacco smoke increased both lung tumor multiplicities and incidences, but ␤-carotene failed to modulate tumor development under all exposure conditions. Animal studies in a model of tobacco smoke carcinogenesis would thus have predicted the absence of any beneficial effects of ␤-carotene supplementation in current or former smokers, but would have failed to anticipate the increase in lung cancer risk. Key Words: ␤-carotene; lung tumors; cigarette smoke; chemoprevention; strain A mice.

Public health goals place great emphasis on the prevention of cancers. The need for effective prevention strategies is particularly pressing for lung cancer. Deaths from lung cancers account for more deaths than breast, colon, and prostate canThis research was supported by grants from the National Institute of Environmental Health Sciences (NIEHS). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIEHS. 1 Present address: BASF Aktiengesellschaft, 67056 Ludwigshafen, Germany. E-mail: [email protected]. 2 To whom correspondence should be addressed at Center for Health and the Environment, University of California, One Shields Avenue, Davis CA 95616. Fax: (530) 752-5300. E-mail: [email protected]. 23

24

OBERMUELLER-JEVIC ET AL.

opposed to being effective against skin, liver, and gastrointestinal tumors (IARC, 1998). Particularly, no animal studies with tobacco smoke had been done until more recently. Biochemical and morphological changes observed in the lungs of ferrets fed ␤-carotene and exposed for 6 months to tobacco smoke offered some possible mechanisms why in smokers ␤-carotene would actually have increased the risk of developing lung cancer (Liu et al., 2000; Wang et al., 1999). However, no tumor data are available in this model. A mouse lung tumor model (Witschi et al., 1997, 2002) allowed investigators to directly examine whether ␤-carotene would modulate the development of lung tumors induced by tobacco smoke. In this paper we report that ␤-carotene, fed at three different concentrations to mice during and after tobacco smoke exposure, failed to mitigate or to enhance lung tumor development. MATERIALS AND METHODS Materials. Kentucky 1R4F reference cigarettes were obtained from the Tobacco Research Institute, University of Kentucky, Lexington, KY. Diet AIN-93G with added gelatin beadlets containing 10% ␤-carotene was purchased from DYETS, Bethlehem, PA, and stored at 4°C until used; the control diet contained beadlets without any ␤-carotene. All solvents were high-performance liquid chromatography (HPLC) grade except tetrahydrofuran, which was ultraviolet spectroscopy grade, and were purchased from Fluka, Ronkonkoma, NY. Crystalline standards for all-trans-␤-carotene (Type I), all-trans-retinol, and all-trans-retinyl palmitate were obtained from Sigma, St. Louis, MO. Animals. Male strain A/J mice, 5 to 6 weeks old, were purchased from Jackson Laboratories, Bar Harbor, ME. Randomly chosen animals were sent to the Comparative Pathology Laboratory, University of Calfornia, Davis, for a standard rodent health surveillance screen. No evidence for infectious disease (pathogenic agents) or presence of parasites or ova in pelage and cecum were reported. Histopathology was not processed, as no significant macroscopic lesions were noted. Serology was negative for mouse hepatitis virus, Sendai virus, Reovirus type 3, pneumonia virus, parvo, ectromelia, and mycoplasma pulmonis. The animals were housed four to a cage in polypropylene cages with tightly fitting wire screen lids on Teklad bedding material. Animal housing was in accordance with institutional guidelines. Mice were assigned to the different experimental groups after a two-week acclimatization period. At all times during the experiment, including during smoke exposure, water and the test diets were provided ad libitum. The animals were observed daily and weighed weekly. Experimental design. All protocols were approved by the University of California Davis Animal Use and Care Administrative Advisory Committee. The following experiments were conducted: ● Preliminary study (Table 1): in a dose-finding study, four groups of strain A mice were fed AIN-93G diets containing 0.005, 0.05, and 0.5% ␤-carotene or control diet alone. One week later, the animals were killed by pentobarbital overdose. Serum and lung tissue were collected under minimum light exposure and stored at – 80°C until analysis. ● In the first experiment (Table 2), the effects of 0.005 and 0.05% ␤-carotene diets were examined in a split-exposure protocol adopted from our previous studies on chemoprevention against lung tumors induced by tobacco smoke (Witschi, 2000). Animals were either kept in air for 9 months (filtered air controls), or exposed for 5 months to tobacco smoke and then kept another 4 month in filtered air. Animals in treatment group A received control diet during the entire 9 months. Animals in group B were fed a diet containing 0.005% ␤-carotene during months 1–5 only and were then fed control diet, whereas animals in group C were fed 0.005% ␤-carotene diet during the entire

9 months. Groups D and E were similar to groups B and C, except that the diet contained 0.05% ␤-carotene. ● In the second experiment, the split protocol was modified as follows (Table 3). Exposure to tobacco smoke was for 4 months and recovery period in air was for 5 months. Animals in group F received control diet through the entire 9 months. Animals in group G were fed a diet containing 0.5% ␤-carotene during months 1– 4 (smoke exposure) while being fed control diet during months 5–9 (recovery period), whereas animals in group H received control diet while in smoke (months 1– 4) and 0.5% ␤-carotene diet during the recovery period (months 5–9). ● Third experiment (Table 4): this continuous protocol was designed to duplicate exposure conditions similar to the clinical trials where active smokers received ␤-carotene supplementation. To mimic this condition, the animals were exposed to the same tobacco smoke concentrations as used in experiments 1 and 2, but for 6 h a day, 5 days a week, for 9 months. Group I animals were either kept in filtered air for 9 months or exposed to tobacco smoke for 9 months, and received control diet for the entire 9 months. In group J, the diet contained 0.5% ␤-carotene during the entire experiment. At the end of the experiment, plasma and lung samples from a few animals per group were collected for ␤-carotene analysis. Exposure system. The tobacco smoke exposure system was identical to the one used in our previous studies (Witschi et al., 1997, 2002). Briefly, mice were exposed 6 h a day, 5 days a week, to a mixture of 89% sidestream and 11% mainstream smoke generated from burning Kentucky 1R4F reference cigarettes. Chamber atmospheres (relative humidity, 48%, standard deviation [SD] 10%; temperature 21°C, SD 1°C) were monitored for nicotine, carbon monoxide (CO), and total suspended particulates (TSP). Smoke concentrations were gradually increased over the first 5 weeks of each experiment as previously described (Witschi et al., 2000) until they reached the following concentrations (average ⫾ SD): CO, 290 ⫾ 35 ppm; nicotine, 19.2 ⫾ 5.6 mg/m 3; and TSP, 140 ⫾ 24 mg/m 3. Tissue preparation. Animals were killed by pentobarbital overdose. For analysis of tumor incidence and multiplicity, the lungs were manually expanded to inspiratory volume by intratracheal instillation of Tellyesniczky’s fluid and fixed for at least 24 h. The number of tumor nodules visible on the lung surface was counted. The results were expressed as tumor incidence, i.e., percentage of animals with one or several lung tumors, and as tumor multiplicity. Tumor multiplicity was calculated as average number of tumors per lung of all animals, including non–tumor-bearing animals. Histopathology of selected lung tumors was done on H&E stained paraffin sections. Plasma and lung tissue extraction. Extraction of plasma samples was performed according to a modification of the method described by Froescheis et al. (2000). A total of 0.4 ml water, 0.8 ml ethanol, and 1.6 ml n-hexane (0.025% butylated hydroxytoluene, BHT) were added to 0.4 ml plasma. The mixture was vortexed for 2 min, then centrifuged for 10 min (1800 ⫻ g at room temperature). The n-hexane layer was collected in a glass tube, and the sample was extracted again with 1.6 ml n-hexane as described above. The combined n-hexane phases were evaporated to dryness under a stream of nitrogen. The residue was redissolved with 160 ␮l methanol: tetrahydrofuran (9:1 v/v; 0.05% BHT). An aliquot of 50 ␮l was subjected to HPLC analysis. Lung tissue was extracted according to a modification of the method described by Liu et al. (2000). The tissue samples were homogenized on ice with 1 ml chloroform:methanol (2:1 v/v; 0.025% BHT). After homogenization, 100 ␮l saline (8.5 g/l) were added, and the mixture was vortexed for 30 s. The samples were extracted twice with 600 ␮l n-hexane (0.025% BHT), vortexed for 2 min, then centrifuged for 10 min (1800 ⫻ g at room temperature). The two combined extracts were evaporated to dryness under a stream of nitrogen, and the residue was redissolved in 100 ␮l methanol:tetrahydrofuran (1:1 v/v; 0.05% BHT). A 50 ␮l aliquot was subjected to reverse-phase HPLC (RPHPLC) analysis. All procedures were carried out under minimal light exposure. Tetrahydrofuran was further purified on the same day of use and kept in amber glass on ice under nitrogen atmosphere. RP-HPLC analysis. Analysis of ␤-carotene, retinol, and retinyl palmitate in plasma and lung tissue was carried out via RP-HPLC using a Waters system

TOBACCO SMOKE, ␤-CAROTENE, AND LUNG TUMORS (Waters, Milford, MA) consisting of a 1525 binary pump, a 717 plus autosampler, and a 2487 dual wavelength absorbance detector. A C30 YMC Cartenoid威 column was used (5 FM, 250 ⫻ 4.6 mm, Waters, Milford, MA) which was coupled to a YMC Cartenoid威 S guard column (5 FM, 20 ⫻ 4.0 mm). A gradient system was performed with methanol:tert butyl methyl ether:water (81:15:4) as eluent A and methanol:tert butyl methyl ether:water (6:90:4) as eluent B. The gradient was run, with a flow rate of 1 ml/min, as follows: 100% A at 10 min, 50% B at 30 min, 100% B at 40 min, and 100% A at 45 to 50 min. ␤-carotene and retinoids were detected at 450 nm and 340 nm, respectively. Retention times were 4.5 min for retinol, 30 min for retinyl palmitate, and 33 min for ␤-carotene. The temperature in the autosampler was held at 4°C. For ␤-carotene, baseline resolution of all identified peaks was achieved in plasma samples and close to baseline resolution with lung samples. For retinoids, baseline resolution was achieved with all samples. All-trans-␤-carotene, alltrans-retinol, and all-trans-retinyl palmitate were quantified by reference to standard curves constructed using crystalline standards. Statistical analysis. All numerical data were calculated as mean and SD or standard error. Comparisons of body weights and of tumor multiplicity between tobacco smoke-exposed and air-exposed controls were made by Student t-test or, where appropriate, by ANOVA followed by Tukey-Kramer’s posttest. Tumor incidences were compared using Fisher’s exact test. A p value of 0.05 or less was considered to be significant.

RESULTS

In the dose-finding study, male strain A mice were placed on diets containing 0.005, 0.05, or 0.5% ␤-carotene. After one week, the animals were killed, and plasma and lung levels of ␤-carotene, retinol, and retinyl palmitate were determined. The results are presented in Table 1. In the animals fed the control diet and 0.005% ␤-carotene, no ␤-carotene was detectable in the plasma. In mice fed 0.05%, plasma ␤-carotene was detectable at 0.01 nmol/ml, and in the group receiving 0.5% ␤-carotene in diet, levels were six times higher. In the lungs, a trend toward a dose-dependent increase in ␤-carotene over basal levels was found. However, the slope of the curve was very shallow (0.12) and not significantly different from zero (p ⫽ 0.2). Plasma retinol levels were similar in all groups, and no plasma retinyl acetate was detected. In the lungs, retinol levels were considerably lower in the 0.5% group, although the

25

difference from the other groups was statistically not significant, because of large interanimal variations. Retinyl palmitate levels were similar in all animals fed the ␤-carotene diets. In the first tumor experiment, we examined the effects of the two lower levels of dietary ␤-carotene on lung tumor development in tobacco smoke-exposed mice. As documented in previous studies (Witschi et al., 1997, 2002), exposure to tobacco smoke greatly reduced weight gain as long as the animals were in the smoke atmosphere. Once removed, they rapidly gained weight and, at the end of the exposure period, had body weights similar to controls (Table 2). The two dietary concentrations of 0.005 and 0.05% ␤-carotene did not interfere with weight gain in air-exposed or tobacco smoke-exposed animals. Tumor counts showed that in animals fed control diet, tobacco smoke produced a significant increase in both lung tumor multiplicity and incidence compared with animals kept in filtered air. The two ␤-carotene diets did not modulate tumor development in the animals kept in air or in the animals exposed to tobacco smoke. In the second experiment, the mice were exposed for 4 months to tobacco smoke, and a diet containing 0.5% ␤-carotene was used. This concentration of ␤-carotene resulted in a reduction of body weight in the air-exposed animals, which eventually had a 10% lower body weight compared with animals fed control diet (Table 3). Animals exposed to tobacco smoke and ␤-carotene had a 15% reduction in weight gain, and the difference from animals fed the control diet was statistically significant. Tumor counts showed a significant increase in the animals exposed for 4 months to tobacco smoke and fed the control diet over animals kept in air. In the animals fed the 0.5% ␤-carotene diet, there seemed to be a reduction, though statistically not significant, in lung tumor multiplicity; this possible trend was not supported by the tumor incidence data. Given these results, it was decided to evaluate the effects of 0.5% dietary ␤-carotene in a protocol that approximated the clinical trials where active smokers were given dietary supple-

TABLE 1 Concentration of ␤-carotene and Retinoids in Plasma and Lung in One-Week Feeding Study

Plasma ␤-carotene Retinol Retinyl palmitate Lung ␤-carotene Retinol Retinyl palmitate

Control diet

0.005% ␤-carotene

0.05% ␤-carotene

0.5% ␤-carotene

n.d. 1.48 ⫾ 0.28 (4) n.d.

n.d. 1.58 ⫾ 0.24 (4) n.d.

0.01 ⫾ 0.00 (4) a 1.58 ⫾ 0.09 (4) n.d.

0.06 ⫾ 0.01 (4) a 1.41 ⫾ 0.10 (4) n.d.

0.13 ⫾ 0.06 (4) 9.69 ⫾ 4.39 (7) 271 ⫾ 27 (7)

0.17 ⫾ 0.11 (8) 9.76 ⫾ 5.05 (8) 311 ⫾ 57 (8)

0.19 ⫾ 0.13 (8) 9.97 ⫾ 3.19 (8) 319 ⫾ 46 (8)

0.22 ⫾ 0.20 (7) 3.61 ⫾ 2.15 (8) 311 ⫾ 63 (8)

Note. Animals were fed control diet or diets containing 0.005, 0.05, or 0.5% ␤-carotene for 1 week before concentrations of ␤-carotene, retinol, and retinyl palmitate were measured. Plasma concentrations are given in nmol/ml, lung concentrations in nmol/g. All data are given as means ⫾ SEM, with the number of determinations in parentheses. n.d., not detectable. a Pooled plasma samples from four animals each.

26


TABLE 2 Split-Exposure Protocol with Diets Containing 0.005 and 0.05% ␤-Carotene Final body weights (g)

Lung tumor multiplicity

Lung tumor incidence

Treatment

Air*

Smoke*

Air

Smoke

Air

Smoke

Group A Group B Group C Group D Group E

32.4 ⫾ 3.5 (25) 31.9 ⫾ 4.4 (19) 31.9 ⫾ 4.6 (10) 33.2 ⫾ 4.2 (20) 33.4 ⫾ 2.9 (10)

34.0 ⫾ 5.3 (26) 32.6 ⫾ 2.9 (20) 33.4 ⫾ 2.9 (15) 31.7 ⫾ 3.0 (20) 31.7 ⫾ 3.0 (14)

1.2 ⫾ 0.2 1.3 ⫾ 0.3 1.2 ⫾ 0.4 1.5 ⫾ 0.3 1.0 ⫾ 0.3

2.3 ⫾ 0.3** 2.3 ⫾ 0.3** 1.9 ⫾ 0.4 2.2 ⫾ 0.4** 2.3 ⫾ 0.4**

15/25 (60%) 13/19 (68%) 8/10 (80%) 14/20 (70%) 6/10 (60%)

23/26 (88%)** 20/20 (100%)** 12/15 (80%) 17/20 (85%) 13/14 (93%)

Note. Air, filtered air months 1–9. Smoke, tobacco smoke months 1–5 and filtered air months 6 –9. Final body weights are mean ⫾ SD; number of animals in parentheses. Lung tumor multiplicity is given as mean ⫾ SE; lung tumor incidence is nunber of animals with tumors/total number of animals at risk; % tumor-bearing animals in parentheses. Group A, control diet months 1– 9; Group B, ␤-carotene diet (0.005%) months 1–5; control diet months 6 –9; Group C, ␤-carotene diet (0.005%) months 1–9; Group D, ␤-carotene diet (0.05%) months 1–5; control diet months 6 –9; Group E, ␤-carotene diet (0.05%) months 1–9. * No significant differences between treatment groups. ** Significantly different (p ⬍ 0.05) compared with corresponding filtered air group.

ments of the antioxidant. In this experiment, weight gains in animals kept in filtered air and fed the control diet were similar to animals fed the ␤-carotene diet (Table 4). Continuous exposure to tobacco smoke resulted in a decreased weight gain, but no difference between animals fed control diet or ␤-carotene diet was seen. Tumor counts showed that a 9-month exposure to tobacco smoke produced a significant increase in lung tumor multiplicity and lung tumor incidence compared with animals kept in filtered air. The 0.5% ␤-carotene diet did not modulate tumor multiplicities or incidences. Histologically, the tumors in all groups showed the usual adenomatous pattern, with a few tumors beginning to develop small carcinomatous foci, as previously described (Witschi et al., 1997). It should be noted that the animals in this experiment received almost twice the dose of tobacco smoke than the ones in experiments 1 and 2; however, tumor response was practically the same. This adds further evidence to the previous observation that tobacco smoke not only produces lung tumors in mice, but that the accompanying stress and reduced weight gain can inhibit or delay tumor growth (Stinn et al., in press; Witschi et al., 2002).

At the end of this experiment, ␤-carotene levels in plasma and lung were measured. The results are shown in Table 5. In air-exposed animals, plasma ␤-carotene levels were almost twice as high as they had been in the 1-week feeding study. Retinol levels were similar in plasma, although somewhat lower in the lungs. In the plasma of smoke-exposed mice, the ␤-carotene levels were significantly higher than they were in animals fed ␤-carotene and exposed to air, whereas in the lungs of the smoke-exposed animals, they were somewhat lower. The plasma levels of approximately 0.1 to 0.2 ␮M compare to the human physiological concentrations in plasmas of 0.5 to 5 ␮M with or without oral ␤-carotene supplementation, respectively (Biesalski et al., 1996). DISCUSSION

In 1998, IARC concluded that there was sufficient evidence for cancer-preventive activity of ␤-carotene in experimental animals (IARC, 1998). Convincing data were obtained in mouse skin and hamster buccal pouch models of carcinogenesis, with supporting evidence provided by studies on colon,

TABLE 3 Split-Exposure Protocol with Diet Containing 0.5% ␤-Carotene Final body weights (g) Treatment Group F Group G Group H

Air 35.8 ⫾ 4.1 (24) 33.2 ⫾ 4.2 (20) 33.4 ⫾ 2.9 (10)



Smoke

Air

Smoke

Air

Smoke

35.1 ⫾ 4.6 (25) 31.7 ⫾ 3.0 (20)* 31.7 ⫾ 3.0 (14)*

1.0 ⫾ 0.2 1.0 ⫾ 0.2 1.0 ⫾ 0.2

1.9 ⫾ 0.3** 1.4 ⫾ 0.2 1.3 ⫾ 0.2

18/24 (75%) 17/25 (68%) 13/22 (59%)

22/25 (88%) 23/26 (88%) 17/24 (71%)

Note. Final body weights are mean ⫾ SD; number of animals in parentheses. Air, filtered air months 1–9. Smoke, tobacco smoke months 1– 4 and filtered air months 5–9. Lung tumor multiplicity is given as mean ⫾ SE; lung tumor incidence is nunber of animals with tumors/total number of animals at risk; % tumor-bearing animals in parentheses. Group F, control diet months 1– 9; Group G, ␤-carotene diet (0.5%) months 1– 4; Group H, ␤-carotene diet (0.5%) months 5–9. * Significantly lower (p ⬍ 0.05) compared with animals fed control diet ** Significantly different (p ⬍ 0.05) from controls (filtered air).

TOBACCO SMOKE, ␤-CAROTENE, AND LUNG TUMORS

27

TABLE 4 Continuous Exposure Protocol with 0.5% ␤-Carotene Diet Final body weights (g)



Treatment

Air*

Smoke*

Air

Smoke

Air

Smoke

Group I Group J

32.5 ⫾ 4.0 (20) 32.8 ⫾ 3.9 (19)

29.8 ⫾ 2.6 (21)* 28.9 ⫾ 2.9 (21)*

1.2 ⫾ 0.2 0.8 ⫾ 0.2

2.1 ⫾ 0.3** 2.2 ⫾ 0.3**

13/20 (65%) 10/19 (53%)

20/21 (95%)** 19/21 (90%)**

Note. Final body weights are mean ⫾ SD; number of animals in parentheses. Air, filtered air months 1–9. Smoke, tobacco smoke months 1–9. Group I, control diet, months 1–9; Group J, ␤-carotene diet, 0.5% months 1–9. Lung tumor multiplicity is given as mean ⫾ SE; lung tumor incidence is nunber of animals with tumors/total number of animals at risk; % tumor-bearing animals in parentheses. * Significantly lower (p ⬍ 0.05) compared with animals fed control diet. ** Significantly different (p ⬍ 0.05) from controls (filtered air).

liver, and pancreas cancers. On the other hand, the thenavailable studies on tumor development in the respiratory tract failed to show any protective effects of ␤-carotene in the upper airways of hamsters or in the peripheral lung of mice treated with nitrosamines or polycyclic aromatic hydrocarbons (Beems, 1987; Moon et al., 1994; Murakoshi et al., 1992; Wolterbeek et al., 1995; Yun et al., 1995). Two more recent studies were only suggestive of a possible protective effect of the antioxidant in respiratory tract of mice (Conaway et al., 1998) or hamsters (Furukawa et al., 1999). On occasion, enhanced tumorigenesis was seen, but increases were statistically not significant (Beems, 1987; Wolterbeek et al., 1995). The results of the present investigation confirm the general conclusion that ␤-carotene does not effectively prevent or enhance chemical carcinogen-induced lung tumor development in animals. Two recent mechanistic papers, however, make it unreasonable to dismiss a priori the possibility that ␤-carotene might enhance tobacco smoke carcinogenesis in animals, as was so unexpectedly seen in humans in clinical trials (Omenn, 1998). In ferrets fed a dose of ␤-carotene sup-

plement considered to be high (or pharmacological) and close to the dose used in human intervention studies, exposure for 6 months to tobacco smoke led to a proposed possible association of ␤-carotene oxidative metabolites with a subsequent decrease in retinoic acid, leading to a downregulated expression of retinoic acid receptor ␤ and an increase in the expression of c-fos and c-jun (Wang et al., 1999). Such a sequence of events could lead to uncontrolled growth. On the other hand, low (physiological) doses of ␤-carotene failed to set in motion the same events (Liu et al., 2000). Currently, no information is available on whether tobacco smoke and high ␤-carotene supplementation would indeed result in enhanced respiratory tract tumorigenesis in ferrets. Causality between putative mechanisms and tumor development has not yet been established(Lotan, 1999; Woutersen et al., 1999). Ferrets live for about 6 to 10 years. This lifetime and the number of animals needed in a carcinogenesis assay make it questionable whether a rigorous testing of the hypothesis put forward by Wang et al. (1999) is feasible. Because of substantial differences in ␤-carotene absorption

TABLE 5 Concentration of ␤-Carotene and Retinoids in Plasma and Lung in Nine-Month Feeding Study Air exposed

Plasma: nmol/ml ␤-carotene Retinol Retinyl palmitate Lung: nmol/g ␤-carotene Retinol Retinyl palmitate

Tobacco smoke exposed

Control diet

␤-carotene diet

Control diet

␤-carotene diet

n.d. (4) 1.30 ⫾ 0.17 (4) n.d. (4)

0.11 ⫾ 0.01 (3) 1.08 ⫾ 0.09 (4) n.d. (4).

n.d. (4) 1.38 ⫾ 0.22 (4) n.d. (4)

0.22 ⫾ 0.02 (4)* 1.28 ⫾ 0.12 (4) n.d. (4)

n.d. (4) 2.37 ⫾ 0.30 (4) 61 ⫾ 23 (4)

0.26 ⫾ 0.11 (5) 3.48 ⫾ 0.87 (5) 225 ⫾ 46 (5)

n.d. (6) 2.23 ⫾ 1.14 (6) 44 ⫾ 15 (6)

0.19 ⫾ 0.06 (5) 3.80 ⫾ 0.77 (5) 115 ⫾ 27 (5)

Note. Animals were fed control diet or a diet containing 0.5% o␤-carotene for 9 months while being exposed to tobacco smoke (controls kept in filtered air). Concentrations of ␤-carotene, retinol, and retinyl palmitate were measured at time of sacrifice. All data are given as means ⫾ SEM, with the number of determinations given in parentheses (for plasma, n equals number of pooled samples from two animals each, and for lung, n equals number of samples from one animal each); n.d., not detectable. * Significantly different (p ⬍ 0.05) from animals fed ␤-carotene and kept in air.

28


and metabolism between ferrets and mice, the question remains to what extent high and low doses of ␤-carotene are comparable between the two species. In the ferret studies, the animals took in 0.43 or 2.4 mg/kg per day of ␤-carotene, whereas in our mouse studies (assuming 3 g food consumption per day) the low and high doses were orders of magnitude higher (6 to 600 mg/kg per day). This intake is also considerably higher than doses used in clinical trials (about 0.5 mg/kg/day). The highest obtainable lung levels of ␤-carotene in mice fed the 0.5% diet were 0.26 nmol/g lung tissue. Although comparable to levels in normal human lung tissue (0.38 nmoles/g; Schmitz et al., 1991), this was nevertheless 10 times lower than the lung levels in ferrets fed a physiological amount of ␤-carotene (Liu et al., 2000). Clearly, a several hundred-fold higher dose of dietary ␤-carotene is not nearly enough to produce in mice the same tissue levels that can be obtained with much lower supplementation in ferrets. In this context, some interesting differences in ␤-carotene pharmacodynamics between ferrets and other species deserve to be pointed out. In ferrets fed the low and high dose diet, serum ␤-carotene levels were 0.025 and 0.119 nmol/ml plasma and lung levels were 2.83 and 22.3 nmol/g tissue, respectively (Liu et al., 2000). The ratio of lung concentrations (nmol/g) to plasma levels (nmol/ml) is therefore 113 to 183. In mice, the same ratio is much lower (2.36; Table 5). Similar, or even lower, lung to plasma uptake ratios can be calculated for another mouse study: 0.26 to 1.25 (Nakagawa et al., 1996). The lung tissue to plasma level ratio is 1.79 in rats (Shapiro et al., 1984) and 0.2 to 0.5 in preruminant calves (Schmitz et al., 1991). Assuming normal unsupplemented plasma levels of ␤-carotene in humans to be approximately 0.5 nmol/ml and lung levels 0.4 nmol/g, the uptake ratio in man would be approximately 1. Clearly, ferrets seem to have a much more efficient mechanism to accumulate ␤-carotene in their lung tissue than have other species, including man. In the free radical-rich lung atmosphere of ferrets inhaling tobacco smoke, enhanced ␤-carotene oxidation may occur, with formation of eccentric cleavage oxidation metabolites. Accompanied by decreased ␤-carotene levels in plasma and lung, this could be a key element in the postulated mechanism by which ␤-carotene should enhance lung carcinogenesis (Wang et al., 1999). In our experiments, smoke exposure decreased lung ␤-carotene levels to some extent (0.19 nmol/g vs. 0.0.26 nmol/g in controls), but the difference was statistically not significant. Plasma ␤-carotene levels were even higher in the smoke-exposed mice (0.22 nmol/ml) than in the controls (0.11 nmol/ml). This finding contrasts with observations that, in humans, active or passive smoking usually decreases plasma ␤-carotene (Albanes et al., 1997; Farchi et al., 2001; Roidt et al., 1988), although at least a part of this decrease can probably be ascribed to decreased dietary consumption. Even so, the question may be raised whether the mice did not accumulate enough ␤-carotene in their plasma and lung tissue to have an effect as occurs in ferrets, and that this might account for the failure to produce a carcinogenic effect

in the present study. It is questionable whether similarly high tissue levels of ␤-carotene as in ferrets ever could be reached and tolerated in a chronic feeding study with mice. Other attempts to explain enhanced lung carcinogenesis by dietary ␤-carotene supplementation, with or without additional exposure to tobacco smoke ingredients, centered on the hypothesis that induction of cytochrome P450 enzymes and overproduction of reactive oxygen species would place increased oxidative stress on the lung, resulting in enhanced tumor development (Paolini et al., 2001; Perocco et al., 1999). Under conditions used in the present experiment, cigarette smoke is a good inducer of pulmonary mixed function oxidases (Witschi et al., 1997). In our experiments there was never any evidence that even high doses of ␤-carotene, given to control mice kept in air or together with cigarette smoke, would enhance lung tumor formation. Our results showing no effect of ␤-carotene would thus not support the hypothesis that enhanced lung tumor development was a consequence of increased cytochrome P450 activity.. Experiments conducted with human bronchial cells exposed to ␤-carotene and tobacco smoke in vitro also failed to provide evidence for a pro-oxidant effect of ␤-carotene (Arora et al., 2001). After yet another disappointing clinical study (Lee et al., 1999), an editorial called ␤-carotene “a miss for epidemiology” (Marshall, 1999). Because the present animal experiments would have predicted the negative outcome of the clinical trials, they might be labeled a “success” for toxicology. However, the same experiments could also be labeled a “miss,” inasmuch as they did not foretell the unanticipated tumorenhancing effects seen in man. Both epidemiology and animal studies are generally considered to be descriptive only and the question may be asked how important mechanistic considerations are in explaining the “rise and demise of ␤-carotene” (Omenn, 1998). Initially there was a substantial amount of plausible mechanistic data that allowed anticipation of successful chemoprevention (De Flora et al., 1999). When the outcome of the human studies was known, other experiments quickly offered equally plausible mechanisms that explained the adverse outcome (Liu et al., 2000; Paolini et al., 2001). The question thus remains whether these divergent appraisals of data make the ␤-carotene story a miss or a success for mechanistic toxicology. ACKNOWLEDGMENTS We thank Dr. Albert van der Vliet (Center for Comparative Respiratory Biology and Medicine) and Mr. Dale Uyeminami (CHE), University of California, Davis, for their help in different aspects of this project. This publication was made possible by grants ES07908 and ES05707 from the National Institute of Environmental Health Sciences (NIEHS).

REFERENCES Albanes, D., Virtamo, J., Taylor, P. R., Rautalahti, M., Pietinen, P., and Heinonen, O. P. (1997). Effects of supplemental beta-carotene, cigarette smoking, and

TOBACCO SMOKE, ␤-CAROTENE, AND LUNG TUMORS alcohol consumption on serum carotenoids in the Alpha-Tocopherol, BetaCarotene Cancer Prevention Study. Am. J. Clin. Nutr. 66, 366 –372. Arora, A., Willhite, C. A., and Liebler, D. C. (2001). Interactions of ␤-carotene and cigarette smoke in human bronchial epithelial cells. Carcinogenesis 22, 1173–1178. Beems, R. B. (1987). The effect of ␤-carotene on BP-induced respiratory tract tumors in hamsters. Nutr. Cancer 10, 197–204. Biesalski, H. K., Hemmes, C., Hopfenmuller, W., Schmid, C., and Gollnick, H. P. (1996). Effects of controlled exposure of sunlight on plasma and skin levels of ␤-carotene. Free Radic. Res. 24, 215–224. Conaway, C. C., Jiao, D., Kelloff, G. J., Steele, V. E., Rivenson, A., and Chung, F. L. (1998). Chemopreventive potential of fumaric acid, N-acetylcysteine, N-(4- hydroxyphenyl) retinamide and beta-carotene for tobacconitrosamine- induced lung tumors in A/J mice. Cancer Lett. 124, 85–93. De Flora, S., Bagnasco, M., and Vainio, H. (1999). Modulation of genotoxic and related effects by carotenoids and vitamin A in experimental models: Mechanistic issues. Mutagenesis 14, 153–172. Farchi, S., Forastiere, F., Pistelli, R., Baldacci, S., Simoni, M., Perucci, C. A., and Viegi, G. (2001). Exposure to environmental tobacco smoke is associated with lower plasma beta-carotene levels among nonsmoking women married to a smoker. Cancer Epidemiol. Biomarkers Prev. 10, 907–909. Froescheis, O., Moalli, S., Liechti, H., and Bausch, J. (2000). Determination of lycopene in tissues and plasma of rats by normal-phase high-performance liquid chromatography with photometric detection. J. Chromatogr. B Biomed. Sci. Appl. 739, 291–299. Furukawa, F., Nishikawa, A., Kasahara, K., Lee, I. S., Wakabayashi, K., Takahashi, M., and Hirose, M. (1999). Inhibition by ␤-carotene of upper respiratory tumorigenesis in hamsters receiving diethylnitrosamine followed by cigarette smoke exposure. Jpn. J. Cancer Res. 90, 154 –161. Hong, W., and Sporn, M. B. (1999). Prevention of cancer in the next millennium: Report of the chemoprevention working group to the American Association for Cancer Research. Cancer Res. 59, 4743– 4758. IARC (1998). IARC Handbooks of Cancer Prevention Carotenoids. (2). International Agency for Research on Cancer, Lyon, France. Lam, T. H., Ho, S. Y., Hedley, A. J., Mak, K. H., and Peto, R. (2001). Mortality and smoking in Hong Kong: Case-control study of all adult deaths in 1998. BMJ 323, 361. Lee, I. M., Cook, N. R., Manson, J. E., Buring, J. E., and Hennekens, C. H. (1999). ␤-carotene supplementation and incidence of cancer and cardiovascular disease: The Women’s Health Study. J. Natl. Cancer Inst. 91, 2102–2106. Liu, C., Wang, X. D., Bronson, R. T., Smith, D. E., Krinsky, N. I., and Russell, R. M. (2000). Effects of physiological versus pharmacological beta-carotene supplementation on cell proliferation and histopathological changes in the lungs of cigarette smoke-exposed ferrets. Carcinogenesis 21, 2245–2253. Lotan, R. (1999). Lung cancer promotion by beta-carotene and tobacco smoke: Relationship to suppression of retinoic acid receptor-␤ and increased activator protein-1? J. Natl. Cancer Inst. 91, 7–9. Marshall, J. R. (1999). ␤-carotene: A miss for epidemiology. J. Natl. Cancer Inst. 91, 2068 –2069. Moon, R. C., Rao, K. V. N., Detrisac, C. J., Kelloff, G. J., Steele, V. E., and Doody, L. A. (1994). Chemoprevention of respiratory tract neoplasia in the hamster by oltipraz alone and in combination. Int. J. Oncol. 4, 661– 667. Murakoshi, M., Nishino, H., Satomi, Y., Takayasu, J., Hasegawa, T., Tokuda, H., Iwashima, A., Okuzumi, J., Okabe, H., and Kitano, H. (1992). Potent preventive action of alpha-carotene against carcinogenesis: Spontaneous liver carcinogenesis and promoting stage of lung and skin carcinogenesis in mice are suppressed more effectively by ␣-carotene than by ␤-carotene. Cancer Res. 52, 6583– 6587. Nakagawa, K., Fujimoto, K., and Miyazawa, T. (1996). ␤-carotene as a high-

29

potency antioxidant to prevent the formation of phospholipid hydroperoxides in red blood cells of mice. Biochim. Biophys. Acta 1299, 110 –116. Omenn, G. S. (1998). Chemoprevention of lung cancer: The rise and demise of ␤-carotene. Annu. Rev. Public Health 19, 73–99. Omenn, G. S. (2000). Chemoprevention of lung cancer is proving difficult and frustrating, requiring new approaches. J. Natl. Cancer Inst. 92, 959 –960. Paolini, M., Antelli, A., Pozzetti, L., Spetlova, D., Perocco, P., Valgimigli, L., Pedulli, G. F., and Cantelli-Forti, G. (2001). Induction of cytochrome P450 enzymes and over-generation of oxygen radicals in ␤-carotene supplemented rats. Carcinogenesis 22, 1483–1495. Perocco, P., Paolini, M., Mazzullo, M., Biagi, G. L., and Cantelli-Forti, G. (1999). ␤-carotene as enhancer of cell transforming activity of powerful carcinogens and cigarette-smoke condensate on BALB/c 3T3 cells in vitro. Mutat. Res. 440, 83–90. Peto, R., Doll, R., Buckley, J. D., and Sporn, M. B. (1981). Can dietary ␤-carotene materially reduce human cancer rates? Nature 290, 201–208. Pisani, P., Parkin, D. M., Bray, F., and Ferlay, J. (1999). Estimates of the worldwide mortality from 25 cancers in 1990. Int. J. Cancer 83, 18 –29. Roidt, L., White, E., Goodman, G. E., Wahl, P. W., Omenn, G. S., Rollins, B., and Karkeck, J. M. (1988). Association of food frequency questionnaire estimates of vitamin A intake with serum vitamin A levels. Am. J. Epidemiol. 128, 645– 654. Schmitz, H. H., Poor, C. L., Wellman, R. B., and Erdman, J. W., Jr. (1991). Concentrations of selected carotenoids and vitamin A in human liver, kidney and lung tissue. J. Nutr. 121, 1613–1621. Shapiro, S. S., Mott, D. J., and Machlin, L. J. (1984). Kinetic characteristics of ␤-carotene uptake and depletion in rat tissue. J. Nutr. 114, 1924 –1933. Stinn W., Kuhl, H., and Haussmann, H J. (in press). The roles of cigarette smoke exposure-related stress in the A/J mouse lung tumorigenesis model. Eurotox 2002. Toxicol. Lett. [Abstract]. Szabo, E. (2001). Lung epithelial proliferation: A biomarker for chemoprevention trials? J. Natl. Cancer Inst. 93, 1042–1043. Wang, X. D., Liu, C., Bronson, R. T., Smith, D. E., Krinsky, N. I., and Russell, M. (1999). Retinoid signaling and activator protein-1 expression in ferrets given ␤-carotene supplements and exposed to tobacco smoke. J. Natl. Cancer Inst. 91, 60 – 66. Wingo, P. A., Ries, L. A., Giovino, G. A., Miller, D. S., Rosenberg, H. M., Shopland, D. R., Thun, M. J., and Edwards, B. K. (1999). Annual report to the nation on the status of cancer, 1973–1996, with a special section on lung cancer and tobacco smoking. J. Natl. Cancer Inst. 91, 675– 690. Witschi, H. (2000). Successful and not so successful chemoprevention of tobacco smoke-induced lung tumors. Exp. Lung Res. 26, 743–756. Witschi, H., Espiritu, I., Peake, J. L., Wu, K., Maronpot, R. R., and Pinkerton, K. E. (1997). The carcinogenicity of environmental tobacco smoke. Carcinogenesis 18, 575–586. Witschi, H., Uyeminami, D., Moran D., and Espiritu, I. (2000). Chemoprevention of tobacco-smoke lung carcinogenesis in mice after cessation of smoke exposure. Carcinogenesis 21, 977–982. Witschi, H., Espiritu, I., Dance, S. T., and Miller, M. S. (2002). A mouse lung tumor model of tobacco smoke carcinogenesis. Toxicol. Sci. 68, 322–330. Wolterbeek, A. P., Schoevers, E. J., Bruyntjes, J. P., Rutten, A. A., and Feron, V. J. (1995). Benzo[a]pyrene-induced respiratory tract cancer in hamsters fed a diet rich in ␤-carotene. A histomorphological study. J. Environ. Pathol. Toxicol. Oncol. 14, 35– 43. Woutersen, R. A., Wolterbeek, A. P., Appel, M. J., van den Berg, H., Goldbohm, R. A., and Feron, V. J. (1999). Safety evaluation of synthetic ␤-carotene. Crit. Rev. Toxicol. 29, 515–542. Yun, T. K., Kim, S. H., and Lee, Y. S. (1995). Trial of a new medium-term model using benzo(a)pyrene induced lung tumor in newborn mice. Anticancer Res. 15, 839 – 845.