Mutation Research, 241 (1990) 67-73
67
Elsevier MUTGEN 01541
Chromosome damage in rat pulmonary alveolar macrophages following ozone inhalation K . R i t h i d e c h *, J . A . H o t c h k i s s , W . C . G r i f f i t h , R . F . H e n d e r s o n
and A.L. Brooks * *
Inhalation Toxicology Research Institute, Looelace Biomedical and Environmental Research Institute, P.O. Box 5890, Albuquerque, N M 87185 (U.S.A.)
(Received 30 May 1989) (Revision received 4 December 1989) (Accepted 7 December 1989)
Keywords: Ozone; Inhalation; Chromosomedamage; Pulmonary alveolar macrophages Summary
To determine whether ozone is clastogenic at environmentally relevant exposure levels, rats were exposed for 6 h to 0.0, 0.12, 0.27, or 0.80 ppm ozone. The alveolar macrophages were isolated from animals sacrificed 28 h after the end of the exposure. The mitotic index and frequency of chromosome aberrations were determined. N o change in the mitotic index was detected following 0.12 ppm ozone exposure. A significant decrease in mitotic index was observed after exposure to 0.27 ppm ozone; a significant (4-fold) increase in the frequency of dividing macrophages was detected following exposure to 0.8 ppm ozone. Only chromatid-type aberrations were observed. There was a significant increase in the frequency of cells with chromatid gaps and in the frequency of cells with chromatid deletions. Animals exposed to 0.27 ppm ozone had the highest proportion of cells with chromatid deletions (0.172) relative to background level (0.028). No exchanges or chromosome-type aberrations were detected in any of the animals. These data suggest that ozone, at relatively low levels, is clastogenic in macrophages from exposed rats.
Ozone is a major oxidant gas in photochemical smog and is produced by the action of sunlight on nitrogen oxides and hydrocarbons from a wide
* Associated Western Universities Postdoctoral Fellows at the time of this study. Present address: Medical Department, BrookhavenNational Laboratory, Upton, NY 11973 (U.S.A.). **
Present address: Department of Biology, Battelle Pacific Northwest, P.O. Box 999, Richland, WA 99352 (U.S.A.).
Correspondence: Dr. Kanokporn Rithidech, Medical Department, Brookhaven National Laboratory, Upton, NY 11973 (U.S.A.).
range of environmental sources, especially from automobile exhaust. The levels of ozone present during peak smog events can reach up to 0.6 ppm (NRCC, 1977). This environmental air pollutant, which is generally associated with large cities, can cause damage to the lung when inhaled (Evans et al., 1976; Hasset et al., 1985; Borek et al., 1986; Rabinowitz and Bassett, 1988), and is a potential lung carcinogen (Hasset et al., 1985; Menzel, 1984). Its ability to produce cell transformation, alone and in combination with radiation, has been reported (Borek et al., 1986), and the literature pertaining to its total toxicity has been reviewed by Menzel (1984). Menzel speculated that the
0165-1218/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
68 major site of action was cell membranes, with damage being mediated by oxidation of unsaturated fatty acids and oxidation of thiols or amino acids in either tissue protein or low molecular weight peptides. Evidence that ozone is genotoxic is equivocal, although there is some evidence that ozone may be mutagenic and carcinogenic (Menzel, 1984). It is important to delineate the potential for ozone to cause genetic damage following inhalation at environmentally relevant concentrations, to better understand the role of ozone in lung carcinogenesis. Studies to expand this understanding have been conducted by testing the ability of ozone to produce cytogenetic damage in the blood lymphocytes of man (McKenzie et al., 1977; Merz et al., 1975), Chinese hamsters (Zelac et al., 1971; Tice et al., 1978) and mice (Gooch et al., 1976). The frequency of chromosome aberrations in the bone marrow of mice was also studied (Tice et al., 1978). All these studies were negative for cytogenetic damage induced by ozone, with the exception of those reported by Zelac et al. (1971) and Merz et al. (1975). However, studies conducted in vitro demonstrated that both chromosome aberrations (Gooch et al., 1976) and sister-chromatid exchanges (Guerreo et al., 1979) were produced in cultured cells exposed to ozone. The current study was undertaken to determine if low levels of inhaled ozone could produce cytogenetic changes in chromosomes of lung macrophages. These cells were selected for evaluation, because they presumably receive a much higher dose of ozone than that experienced by cells evaluated in the past, such as the bone marrow and the blood. Furthermore, they can be stimulated to divide by ozone exposure (Hotchkiss et al., 1988). The high dose per unit of exposure and the ability to score cytogenetic damage in these cells make them good candidates for evaluating the clastogenic potency of ozone in vivo, in a relevant cell population. Materials and methods
Animals and exposures
A total of 20, female F 3 4 4 / N rats (12-14 weeks of age; 145-220 g) were obtained from the Inhalation Toxicology Research Institute's colony for
use in this study. Before and after exposures, the rats were housed 2-3 per cage in shoebox-style, polycarbonate cages with sterilized aspen wood chip bedding and filter tops. Food (Wayne Lab Blox, Allied Mills, Chicago, IL) and water were provided ad libitum. Rooms were maintained at 20-22°C, with a relative humidity of 20-50% and a 12-h light/dark cycle, with the light cycle starting at 6:00 a.m. Rats were exposed in whole-body inhalation exposure chambers (HC-1000, Hazleton Systems, Lexana, KS) to nominal ozone concentrations of 0.0 (air control), 0.12, 0.27 and 0.8 ppm for a total of 6 h. Five rats were exposed at each ozone concentration. Because the Institute is located at an elevation of 1728 m, where the average barometric pressure is 625 mm Hg (0.82 sea level atmosphere), the absolute concentrations of ozone in the exposure atmosphere were equivalent to sea level ozone concentrations of 0.0, 0.10, 0.22 and 0.6 ppm (0.0, 0.20, 0.43 and 1.29 m g / m 3, respectively). The rats were randomized by body weight and were conditioned in exposure chambers supplied with filtered air for 24 h prior to ozone exposure. While in the exposure chambers, the animals were housed individually in rack-mounted stainless steel wire cages. Water was available at all times, but food was removed during ozone exposure. Exposures were initiated at 6:30 a.m. and terminated 6 h later. A 6-h exposure approximates the duration of peak environmental ozone concentrations (due to diurnal variations in 03 concentration) during a typical work day. However, the primary reasons for choosing 6 h were: (1) to allow comparison of the results obtained from this single exposure study with other repeated O3-exposure studies conducted at the Institute, and (2) to facilitate the maintenance of the animals within the whole-body inhalation chambers both before and after exposure during the 8-h workday of the technical staff. Ozone was generated with an OREC Model 03VI-00zonizer (Ozone Research and Equipment Corp., Phoenix, AZ). Dilution air was mixed with ozone to bring the total airflow through the exposure chamber to approx. 15 chamber air changes/h. The chamber ozone concentration was controlled by adjusting the intensity of UV radiation within the ozonizer. The concentration of
69 ozone within the c h a m b e r was m o n i t o r e d throughout the exposure with two Dasibi 100 H Ozone Monitors (Dasibi Environmental Corp., Glendale, CA) and a Mast Model 727-3 Ozone Monitor. D a t a were recorded by Linear 0141 strip chart recorders (Linear Instrument Corp., Reno, NV). The exposure-atmosphere sampling probes were positioned within the breathing zone of the rats in the chambers. Chamber temperature and relative humidity were maintained between 21 and 25°C and 40 and 70%, respectively.
Cytogenetic techniques Macrophages were prepared for cytogenetic analysis as previously described (Bice et al., 1977). Briefly, the animals were injected with colchicine (6 m g / k g body weight) 24 h after exposure to arrest metaphase cells. 4 h later, the animals were sacrificed by CO 2 asphyxiation followed by exsanguination, and the trachea, lung and heart were removed intact. The heart-lung block was cooled in saline (4°C) for 30 min, then lavaged with saline to recover macrophages. Slides were coded for analyses. The cytogenetic methods used were reviewed previously (Rithidech et al., 1989). 50 metaphase cells per animal were analyzed for structural and numerical chromosome changes. N u m b e r s of chromatid gaps were recorded for the same cells and were evaluated separately from other structural aberrations. The mitotic index of the macrophages was also determined on 500 cells per animal.
Data analyses The number of cells with chromatid deletions or with chromatid gaps for each animal was treated as a binomial proportion, e.g., the number of cells observed with one or more deletions of the 50 cells examined per animal. A linear logistic model (McCullagh and Nelder, 1983) was used to estimate the proportion of cells with chromatid deletions or gaps in the control group and in each exposure group, by using a m a x i m u m likelihood technique. Our basic reasons for choosing the logit rather than other methods for binomial data were (1) it is widely accepted in the statistical community, (2) it can be easily adapted to a wide variety of experimental designs in generalized linear models, (3) it provides better estimates of standard errors than
several competing methods, and (4) the validity of some of the assumptions of the statistical model can be relatively easily evaluated. Of course the other choice would have been to evaluate the data as aberrations per cell and treat it as Poisson-distributed data. We could again treat the Poisson data in a generalized linear model (i.e., log linear) and the above reasons for using the logit would also be true for this model. Our reasons for treating the data as being binomially distributed were (1) that this seemed to involve fewer assumptions than one has to make for Poisson data (although these assumptions are probably accepted by cytogeneticists), and (2) when the data is analyzed as being Poisson-distributed we arrive at the same conclusions. This technique was used because it involves fewer assumptions than techniques comparing the frequencies of aberrations per cell. The linear logistic model estimates I i = logit(pi) = log[ p i / ( 1 - Pi)] as a linear combination of the 4 exposure levels of ozone, where each exposure level is represented by a variable that indicates whether or not an animal was exposed to that level. The primary advantages of using a linear logistic model in this study were to provide better estimates of standard errors and the ability to evaluate the assumptions of the model. Ri was the number of cells with chromatid deletions of the m i cells examined for a given animal, and Pi was the expected proportion of cells with deletions, so that R i = binomial(m i, Pi)- Using this parameterization, we estimated the logit for the control group and the difference from the control group on the logit scale for the other 3 exposure groups. Likelihood ratio statistics were computed as the reduction in deviance from adding the categorical variables for ozone exposure. This statistic has an asymptotic chi-square distribution and tests whether or not the ozone exposure explains a statistically significant amount of the variability. The same technique was used for the chromatid gap and mitotic index data. All model calculations were performed using G L I M 3.77 (C.D. Payne (Ed.), The G L I M System Release 3.77: Generalized Linear Interactive Modelling Manual, Royal Statistical Society, London, 1987, available from Numerical Algorithm G r o u p Inc., 1101 31st St., Suite 100, Downers Grove, IL 60515-1263).
70
Results The influence of ozone exposure on the mitotic index of the macrophages is illustrated in Fig. 1. The highest level of ozone (0.8 ppm) induced a significant increase in mitotic index relative to the control level ( p < 0.01). The mitotic index in macrophages from animals exposed to 0.27 p p m was significantly decreased from control level ( p < 0.01). The number and type of chromosome aberrations, and the frequencies of cells with chromatid deletions and with chromatid gaps, are listed as a function of ozone concentration in Table 1. The
4 x w
r, z
3
0
9
2
I
I
I
0.2 0.4 0.6 OZONE CONCENTRATION (ppm)
I
I
0.8
Fig. 1. The effects of ozone on mitotic index (MI) in pulmonary alveolar macrophages from exposed rats. Points represent mean MI_SE.
TABLE 1 C H R O M O S O M E ABERRATIONS IN PULMONARY ALVEOLAR MACROPHAGES FROM RATS EXPOSED TO O Z O N E 50 cells/rat, 5 rats/exposure level. Ozone dose (ppm)
Chromatid deletions
Isochromatid deletions
Chromatid gaps
Total aberrations "
Abnormal cells ,.b
Cells with gaps b
0 (0) ¢ 0.12 (0.20) 0.27 (0.22) 0.8 (0.66)
7 12 49 15
0 2 2 1
20 22 33 61
7 14 51 16
1.4+0.4 2.8 + 0.9 8.6 + 1.7 3.2 + 0.3
3.4+0.5 4.4 + 1.7 5.2 ___1.1 9.4 + 1.6
a Chromatid gaps are not included. b Mean -t- SE. c Numbers in parentheses are the ozone concentrations at sea level.
TABLE 2 RESULTS OF LOGISTIC REGRESSION MODEL ANALYSIS F O R CELLS W I T H C H R O M A T I D DELETIONS A N D C H R O M A T I D GAPS P R O D U C E D BY D I F F E R E N T LEVELS OF O Z O N E EXPOSURE Ozone (ppm)
0 0.12 0.27 0.8
Cells with chromatid deletions
Cells with chromatid gaps
Parameter a (S.E.)
Proportion b (95% C.I.)
Parameter a (S.E.)
Proportion b (95% C.I.)
-3.55 0.72 1.98 0.86
0.028(0.013, 0.056(0.034, 0.172(0.130, 0.064(0.040,
-2.62 0.28 0.46 1.16
0.068(0.043, 0.088(0.059, 0.104(0.072, 0.188(0.144,
(0.38) ¢(0.47) ¢(0.42) ¢(0.46)
0.057) 0.092) 0.224) 0.102)
(0.25) c(0.34) ¢(0.33) c(0.30)
0.107) 0.130) 0.148) 0.24)
a Parameters are calculated on a logit scale g = logit(p) = l o g [ p / ( 1 - p)] and p = inverse logit(g) = 1 / [ 1 + e x p ( - g ) ] where p is a proportion of cells with an abnormality and g is the logit of p. b Estimated proportion of cells with abnormality and 95% confidence interval based upon standard errors from logistic model. The proportions are shown as their estimated values for each ozone exposure group and not as the difference from the control group. The 95% confidence limits are inverse logit of g 1.96 S.E. c Estimated difference from control group (0 ppm ozone) on the logit scale. These values are added to the control values to obtain g = logit(p) for the particular exposure level.
71 table illustrates that all of the aberrations were of the chromatid-type and that no exchanges or complex aberrations were observed in this study. Table 2 shows the results of fitting the linear logistic regression model to the data. Likelihood ratio statistics indicated that the categorical variables for ozone exposure levels were statistically significant for chromatid deletions (X 2 = 37.33; d.f. = 3; p < 0.005) and chromatid gaps (X 2 = 19.65; d.f. = 3; p < 0.005). There was no indication of extra binomial variation for either endpoint ( p > 0.05). The exposure to ozone significantly increased the proportion of cells with chromatid deletions and with chromatid gaps. The proportion of cells with chromatid deletions was highest in the 0.27-ppm exposure group. Discussion The results of this study show that ozone exposure induces an increase in the frequency of chromatid-type aberrations in lung macrophages. These results are concordant with those from studies of human lymphocytes exposed in vivo to 0.5 p p m ozone for 6 and 10 h (Merz et al., 1975). Evidence from in vitro studies with human lymphocytes showed the same response. Following a brief exposure (5 or 10 min) to 5 p p m ozone, a significant increase in the number of chromatid deletions was observed in human lymphocytes (Fetner, 1962). Gooch et al. (1976) also reported a similar increase in chromatid deletions in human lymphocytes after an in vitro exposure to 7.23 or 7.95 p p m / h for 30, 60 or 90 min. An increase in the frequency of chromosomal damage was also detected in lymphocytes of Chinese hamsters exposed to 0.2 p p m ozone for 5 h (Zelac et al., 1971). However, the predominant type of damage found in Chinese hamster cells was chromosometype aberrations, especially at longer periods of time after the ozone exposure. Based on the results of these studies, Zelac et al. (1971) postulated that ozone was a radiomimetic agent. Most studies, including our own, failed to detect ozone-related increases in the frequency of chromosome-type aberrations, suggesting that ozone is not a radiomimetic substance, but m a y be S-dependent in its clastogenic action. Tice et al. (1978) failed to detect significant
increases in the levels of chromatid deletions in blood lymphocytes of exposed Chinese hamsters either immediately or 14 days after ozone exposure. However, they did note that there was an increase in chromatid deletions at 7 days post-exposure, as well as an increase in the frequency of chromatid gaps 14 days after ozone inhalation. Other studies using h u m a n blood lymphocytes (McKenzie et al., 1977; G o o c h et al., 1976) or mouse bone-marrow cells (Tice et al., 1978) also failed to detect any clastogenic effects following either in vivo inhalation exposure or in vitro exposure to ozone. The inability to detect cytogenetic effects of ozone in these studies might have been related to both the concentration x time products and the low dose of ozone that reached the target bone-marrow cells and lymphocytes. In the current study, the macrophages were exposed to an effective concentration of ozone that would be expected to be much higher than that received by the bone-marrow cells or by the lymphocytes in previous studies. Our finding of an increase in the level of chromatid gaps in macrophages of animals exposed to the highest concentration of ozone is interesting. A similar increase in the number of chromatid gaps has been reported for lymphocytes of ozoneexposed humans (Merz et al., 1975) and Chinese hamsters (Tice et al., 1978). The biological consequences of chromatid gaps are not well understood, although it has been suggested that they arise from unrepaired, D N A single-strand breakage (Bender et al., 1974). Such damage may, after cell replication, be detected as chromatid deletions in subsequent cell divisions. The observation that there was a decrease in chromosomal damage in animals exposed to the highest concentration of ozone (0.8 ppm) is difficult to understand, because the next lower level (0.27 ppm) had a higher frequency of chromosome aberrations. A possible explanation might be that ozone reacted with the lung tissue in such a way as to induce an influx of macrophages into the bronchoalveolar airspace and to stimulate the lung macrophage population to divide. Thus, both the resident cell population and the mitotic index changed. Since the exposure was terminated after only 6 h, m a n y of the cells recovered by lavage may not have been directly exposed to ozone. The
72
endpoint of the cytogenetic assay used in this study reflects the ratio of cells with chromosomal aberrations to normal cells. Therefore, an ozoneinduced increase in the total number of macrophages recovered by lavage would tend to "dilute out" the macrophage population that was actually exposed to ozone, resulting in an underestimation of the cytogenetic effects of ozone at higher doses. This interpretation is supported by the increase in mitotic index observed in the present study, as well as by the results of a previous study by Hotchkiss et al. (1988). These authors reported a significant increase in the labeling index of alveolar macrophages 18 h after a 6-h exposure to 0.8 ppm ozone, as well as a significant increase in the total number of macrophages recovered by lavage between 18 and 42 h after exposure to 0.8 ppm ozone. It is also possible that macrophages exposed to the highest dose of ozone (0.8 ppm) were too fragile to be recovered by lavage, because of the toxic effects of ozone at this concentration (Dowell et al., 1970; Alpert et al., 1971). In summary, our study demonstrates cytogenetic effects in pulmonary alveolar macrophages after exposure of rats to ozone levels that may be found in many large metropolitan cities. Because macrophages are not the major cells at risk for the induction of cancer or toxicity, chromosome damage in these cells can only be considered indicative of the actual level of ozone exposure or effective dose to lung cells. The cytogenetic damage observed in macrophages might, however, be predictive of cytogenetic damage that would be produced in lung epithelial cells. Additional research needs to be conducted to determine the level of damage in other cell types more relevant to cancer induction or toxicity in the lung resulting from ozone exposure. However, this research did demonstrate that by using the proper cell type, it is possible to detect cytogenetic damage in the lungs of animals that were exposed to environmentally relevant, low concentrations of ozone over a brief time period. Our research suggests that the genetic material in lung cells might be damaged by ozone, thereby increasing the risk of lung cancer and presenting a potential human health problem.
Acknowledgements We thank Drs. D.E. Bice, N.F. Johnson, J.A. Bond and T. Coons for their constructive reviews of the manuscript. The technical help of Mary Jo Waltman is also acknowledged. This research was performed under U.S. Department of Energy Contract No. DE-AC04-76EV01013 and by N I H Grant ES04282, and conducted in facilities fully accredited by the American Association for Accreditation of Laboratory Animal Care.
References Alpert, S.M., D.E. Gardner, D.J. Hurst, T.R. Lewis and D.L. Coffin (1971) Effects of exposure to ozone on defensive mechanisms of the lung, J. Appl. Physiol., 31,247-252. Bender, M.A, H.G. Griggs and J.S. Bedford (1974) Mechanisms of chromosomal aberrations production, III. Chemicals and ionizing radiation, Mutation Res., 23, 197-212. Bice, D.E., R.J. Cox and C.T. Schnizlein (1977) Lavage techniques to obtain maximum numbers of rat alveolar macrophages and their use in migration assays, in: B.B. Boecker, C.H. Hobbs and B.S. Martinez (Eds.), Inhalation Toxicology Research Institute Annual Report, LF-58, pp. 390-394. Borek, C., M. Zalder, A. Ong, H. Mason and G. Witz (1986) Ozone acts alone and synergistically with ionizing radiation to induce in vitro neoplastic transformation, Carcinogenesis, 7, 1611-1613. Dowell, A.R., L.A. Lohrbauer, D. Hurst and S.D. Lee (1970) Rabbit alveolar macrophage damage caused by in vivo ozone inhalation, Arch. Environ. Health, 21, 121-127. Evans, M.J., L.V. Johnson, R.J. Stephens and G. Freeman (1976) Renewal of the terminal bronchiolar epithelium in the rat following exposure to NO 2 or 03, Lab. Invest., 34, 565-578. Fetner, R.H. (1962) Ozone-induced chromosome breakage in human cell cultures, Nature (London), 194, 793-794. Gooch, P.C., D.A. Creasia and J.G. Brewen (1976) The cytogenetic effects of ozone: inhalation and in vitro exposures, Environ. Res., 12, 188-195. Guerrero, R.R., D.E. Rounds, R.S. Olson and J.D. Hackney (1979) Mutagenic effects of ozone on human cells exposed in vivo and in vitro based on sister chromatid exchange analysis, Environ. Res., 18, 336-346. Hasset, C., M.G. Mustafa, W.F. Coulson and R.M. Elashoff (1985) Murine lung carcinogenesis exposure to ambient ozone concentrations, J. Natl. Cancer Inst., 75, 771-777. Hotchkiss, J . A . J . R . Harkema, D.T. Kirkpatrick and R.F. Henderson (1989) Response of rat alveolar macrophages to ozone: quantitative assessment of population size, morphology and proliferation following acute exposure, Exp. Lung Res., 15, 1-16.
73 McCullagh, P., and J.A. Nelder (1983) Generalized Linear Models, Chapman and Hall, London, 261 pp. McKenzie, M.H., J.H. Knelson, N.J. Rummo and D.E. House (1977) Cytogenetic effects of inhaled ozone in man, Mutation Res., 48, 95-102. Menzel, D.B. (1984) Ozone: an overview of its toxicity in man and animals, J. Toxicol. Environ. Health, 13, 183-204. Merz, T., M.A Bender, H.D. Keller and T.J. Kull (1975) Observation of aberrations in chromosomes of lymphocytes from human subjects exposed to ozone at a concentration of 0.5 ppm for 6 and 10 hours, Mutation Res., 31, 299-302. National Research Council Committee (1977) Report on Medical and Environmental Effects of Environmental Pollutants: Ozone and Other Photochemical Oxidants, National Academy of Science (U.S.A.), Washington.
Rabinowitz, J.L., and D.J.P. Bassett (1988) Effects of 2 ppm ozone exposure on rat lung lipid fatty acids, Exp. Lung Res., 14, 477-489. Rithidech, K., B.T. Chen, J.L. Mauderly and A.L. Brooks (1989) Cytogenetic effects of cigarette smoke on alveolar pulmonary macrophages of the rat, Environ. Mol. Mutagen., 14, 27-33. Tice, R.R., M.A Bender, J.L. Ivett and R.T. Drew (1978) Cytogenetic effects of inhaled ozone, Mutation Res., 58, 293-304. Zelac, R.E., H.L. Cromroy, W.E. Bolch Jr., B.G. Dunavant and H.A. Bevis (1971) Inhaled ozone as a mutagen, I. Chromosome aberrations induced in Chinese hamster lymphocytes, Environ. Res., 4, 262-282.
67
Elsevier MUTGEN 01541
Chromosome damage in rat pulmonary alveolar macrophages following ozone inhalation K . R i t h i d e c h *, J . A . H o t c h k i s s , W . C . G r i f f i t h , R . F . H e n d e r s o n
and A.L. Brooks * *
Inhalation Toxicology Research Institute, Looelace Biomedical and Environmental Research Institute, P.O. Box 5890, Albuquerque, N M 87185 (U.S.A.)
(Received 30 May 1989) (Revision received 4 December 1989) (Accepted 7 December 1989)
Keywords: Ozone; Inhalation; Chromosomedamage; Pulmonary alveolar macrophages Summary
To determine whether ozone is clastogenic at environmentally relevant exposure levels, rats were exposed for 6 h to 0.0, 0.12, 0.27, or 0.80 ppm ozone. The alveolar macrophages were isolated from animals sacrificed 28 h after the end of the exposure. The mitotic index and frequency of chromosome aberrations were determined. N o change in the mitotic index was detected following 0.12 ppm ozone exposure. A significant decrease in mitotic index was observed after exposure to 0.27 ppm ozone; a significant (4-fold) increase in the frequency of dividing macrophages was detected following exposure to 0.8 ppm ozone. Only chromatid-type aberrations were observed. There was a significant increase in the frequency of cells with chromatid gaps and in the frequency of cells with chromatid deletions. Animals exposed to 0.27 ppm ozone had the highest proportion of cells with chromatid deletions (0.172) relative to background level (0.028). No exchanges or chromosome-type aberrations were detected in any of the animals. These data suggest that ozone, at relatively low levels, is clastogenic in macrophages from exposed rats.
Ozone is a major oxidant gas in photochemical smog and is produced by the action of sunlight on nitrogen oxides and hydrocarbons from a wide
* Associated Western Universities Postdoctoral Fellows at the time of this study. Present address: Medical Department, BrookhavenNational Laboratory, Upton, NY 11973 (U.S.A.). **
Present address: Department of Biology, Battelle Pacific Northwest, P.O. Box 999, Richland, WA 99352 (U.S.A.).
Correspondence: Dr. Kanokporn Rithidech, Medical Department, Brookhaven National Laboratory, Upton, NY 11973 (U.S.A.).
range of environmental sources, especially from automobile exhaust. The levels of ozone present during peak smog events can reach up to 0.6 ppm (NRCC, 1977). This environmental air pollutant, which is generally associated with large cities, can cause damage to the lung when inhaled (Evans et al., 1976; Hasset et al., 1985; Borek et al., 1986; Rabinowitz and Bassett, 1988), and is a potential lung carcinogen (Hasset et al., 1985; Menzel, 1984). Its ability to produce cell transformation, alone and in combination with radiation, has been reported (Borek et al., 1986), and the literature pertaining to its total toxicity has been reviewed by Menzel (1984). Menzel speculated that the
0165-1218/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
68 major site of action was cell membranes, with damage being mediated by oxidation of unsaturated fatty acids and oxidation of thiols or amino acids in either tissue protein or low molecular weight peptides. Evidence that ozone is genotoxic is equivocal, although there is some evidence that ozone may be mutagenic and carcinogenic (Menzel, 1984). It is important to delineate the potential for ozone to cause genetic damage following inhalation at environmentally relevant concentrations, to better understand the role of ozone in lung carcinogenesis. Studies to expand this understanding have been conducted by testing the ability of ozone to produce cytogenetic damage in the blood lymphocytes of man (McKenzie et al., 1977; Merz et al., 1975), Chinese hamsters (Zelac et al., 1971; Tice et al., 1978) and mice (Gooch et al., 1976). The frequency of chromosome aberrations in the bone marrow of mice was also studied (Tice et al., 1978). All these studies were negative for cytogenetic damage induced by ozone, with the exception of those reported by Zelac et al. (1971) and Merz et al. (1975). However, studies conducted in vitro demonstrated that both chromosome aberrations (Gooch et al., 1976) and sister-chromatid exchanges (Guerreo et al., 1979) were produced in cultured cells exposed to ozone. The current study was undertaken to determine if low levels of inhaled ozone could produce cytogenetic changes in chromosomes of lung macrophages. These cells were selected for evaluation, because they presumably receive a much higher dose of ozone than that experienced by cells evaluated in the past, such as the bone marrow and the blood. Furthermore, they can be stimulated to divide by ozone exposure (Hotchkiss et al., 1988). The high dose per unit of exposure and the ability to score cytogenetic damage in these cells make them good candidates for evaluating the clastogenic potency of ozone in vivo, in a relevant cell population. Materials and methods
Animals and exposures
A total of 20, female F 3 4 4 / N rats (12-14 weeks of age; 145-220 g) were obtained from the Inhalation Toxicology Research Institute's colony for
use in this study. Before and after exposures, the rats were housed 2-3 per cage in shoebox-style, polycarbonate cages with sterilized aspen wood chip bedding and filter tops. Food (Wayne Lab Blox, Allied Mills, Chicago, IL) and water were provided ad libitum. Rooms were maintained at 20-22°C, with a relative humidity of 20-50% and a 12-h light/dark cycle, with the light cycle starting at 6:00 a.m. Rats were exposed in whole-body inhalation exposure chambers (HC-1000, Hazleton Systems, Lexana, KS) to nominal ozone concentrations of 0.0 (air control), 0.12, 0.27 and 0.8 ppm for a total of 6 h. Five rats were exposed at each ozone concentration. Because the Institute is located at an elevation of 1728 m, where the average barometric pressure is 625 mm Hg (0.82 sea level atmosphere), the absolute concentrations of ozone in the exposure atmosphere were equivalent to sea level ozone concentrations of 0.0, 0.10, 0.22 and 0.6 ppm (0.0, 0.20, 0.43 and 1.29 m g / m 3, respectively). The rats were randomized by body weight and were conditioned in exposure chambers supplied with filtered air for 24 h prior to ozone exposure. While in the exposure chambers, the animals were housed individually in rack-mounted stainless steel wire cages. Water was available at all times, but food was removed during ozone exposure. Exposures were initiated at 6:30 a.m. and terminated 6 h later. A 6-h exposure approximates the duration of peak environmental ozone concentrations (due to diurnal variations in 03 concentration) during a typical work day. However, the primary reasons for choosing 6 h were: (1) to allow comparison of the results obtained from this single exposure study with other repeated O3-exposure studies conducted at the Institute, and (2) to facilitate the maintenance of the animals within the whole-body inhalation chambers both before and after exposure during the 8-h workday of the technical staff. Ozone was generated with an OREC Model 03VI-00zonizer (Ozone Research and Equipment Corp., Phoenix, AZ). Dilution air was mixed with ozone to bring the total airflow through the exposure chamber to approx. 15 chamber air changes/h. The chamber ozone concentration was controlled by adjusting the intensity of UV radiation within the ozonizer. The concentration of
69 ozone within the c h a m b e r was m o n i t o r e d throughout the exposure with two Dasibi 100 H Ozone Monitors (Dasibi Environmental Corp., Glendale, CA) and a Mast Model 727-3 Ozone Monitor. D a t a were recorded by Linear 0141 strip chart recorders (Linear Instrument Corp., Reno, NV). The exposure-atmosphere sampling probes were positioned within the breathing zone of the rats in the chambers. Chamber temperature and relative humidity were maintained between 21 and 25°C and 40 and 70%, respectively.
Cytogenetic techniques Macrophages were prepared for cytogenetic analysis as previously described (Bice et al., 1977). Briefly, the animals were injected with colchicine (6 m g / k g body weight) 24 h after exposure to arrest metaphase cells. 4 h later, the animals were sacrificed by CO 2 asphyxiation followed by exsanguination, and the trachea, lung and heart were removed intact. The heart-lung block was cooled in saline (4°C) for 30 min, then lavaged with saline to recover macrophages. Slides were coded for analyses. The cytogenetic methods used were reviewed previously (Rithidech et al., 1989). 50 metaphase cells per animal were analyzed for structural and numerical chromosome changes. N u m b e r s of chromatid gaps were recorded for the same cells and were evaluated separately from other structural aberrations. The mitotic index of the macrophages was also determined on 500 cells per animal.
Data analyses The number of cells with chromatid deletions or with chromatid gaps for each animal was treated as a binomial proportion, e.g., the number of cells observed with one or more deletions of the 50 cells examined per animal. A linear logistic model (McCullagh and Nelder, 1983) was used to estimate the proportion of cells with chromatid deletions or gaps in the control group and in each exposure group, by using a m a x i m u m likelihood technique. Our basic reasons for choosing the logit rather than other methods for binomial data were (1) it is widely accepted in the statistical community, (2) it can be easily adapted to a wide variety of experimental designs in generalized linear models, (3) it provides better estimates of standard errors than
several competing methods, and (4) the validity of some of the assumptions of the statistical model can be relatively easily evaluated. Of course the other choice would have been to evaluate the data as aberrations per cell and treat it as Poisson-distributed data. We could again treat the Poisson data in a generalized linear model (i.e., log linear) and the above reasons for using the logit would also be true for this model. Our reasons for treating the data as being binomially distributed were (1) that this seemed to involve fewer assumptions than one has to make for Poisson data (although these assumptions are probably accepted by cytogeneticists), and (2) when the data is analyzed as being Poisson-distributed we arrive at the same conclusions. This technique was used because it involves fewer assumptions than techniques comparing the frequencies of aberrations per cell. The linear logistic model estimates I i = logit(pi) = log[ p i / ( 1 - Pi)] as a linear combination of the 4 exposure levels of ozone, where each exposure level is represented by a variable that indicates whether or not an animal was exposed to that level. The primary advantages of using a linear logistic model in this study were to provide better estimates of standard errors and the ability to evaluate the assumptions of the model. Ri was the number of cells with chromatid deletions of the m i cells examined for a given animal, and Pi was the expected proportion of cells with deletions, so that R i = binomial(m i, Pi)- Using this parameterization, we estimated the logit for the control group and the difference from the control group on the logit scale for the other 3 exposure groups. Likelihood ratio statistics were computed as the reduction in deviance from adding the categorical variables for ozone exposure. This statistic has an asymptotic chi-square distribution and tests whether or not the ozone exposure explains a statistically significant amount of the variability. The same technique was used for the chromatid gap and mitotic index data. All model calculations were performed using G L I M 3.77 (C.D. Payne (Ed.), The G L I M System Release 3.77: Generalized Linear Interactive Modelling Manual, Royal Statistical Society, London, 1987, available from Numerical Algorithm G r o u p Inc., 1101 31st St., Suite 100, Downers Grove, IL 60515-1263).
70
Results The influence of ozone exposure on the mitotic index of the macrophages is illustrated in Fig. 1. The highest level of ozone (0.8 ppm) induced a significant increase in mitotic index relative to the control level ( p < 0.01). The mitotic index in macrophages from animals exposed to 0.27 p p m was significantly decreased from control level ( p < 0.01). The number and type of chromosome aberrations, and the frequencies of cells with chromatid deletions and with chromatid gaps, are listed as a function of ozone concentration in Table 1. The
4 x w
r, z
3
0
9
2
I
I
I
0.2 0.4 0.6 OZONE CONCENTRATION (ppm)
I
I
0.8
Fig. 1. The effects of ozone on mitotic index (MI) in pulmonary alveolar macrophages from exposed rats. Points represent mean MI_SE.
TABLE 1 C H R O M O S O M E ABERRATIONS IN PULMONARY ALVEOLAR MACROPHAGES FROM RATS EXPOSED TO O Z O N E 50 cells/rat, 5 rats/exposure level. Ozone dose (ppm)
Chromatid deletions
Isochromatid deletions
Chromatid gaps
Total aberrations "
Abnormal cells ,.b
Cells with gaps b
0 (0) ¢ 0.12 (0.20) 0.27 (0.22) 0.8 (0.66)
7 12 49 15
0 2 2 1
20 22 33 61
7 14 51 16
1.4+0.4 2.8 + 0.9 8.6 + 1.7 3.2 + 0.3
3.4+0.5 4.4 + 1.7 5.2 ___1.1 9.4 + 1.6
a Chromatid gaps are not included. b Mean -t- SE. c Numbers in parentheses are the ozone concentrations at sea level.
TABLE 2 RESULTS OF LOGISTIC REGRESSION MODEL ANALYSIS F O R CELLS W I T H C H R O M A T I D DELETIONS A N D C H R O M A T I D GAPS P R O D U C E D BY D I F F E R E N T LEVELS OF O Z O N E EXPOSURE Ozone (ppm)
0 0.12 0.27 0.8
Cells with chromatid deletions
Cells with chromatid gaps
Parameter a (S.E.)
Proportion b (95% C.I.)
Parameter a (S.E.)
Proportion b (95% C.I.)
-3.55 0.72 1.98 0.86
0.028(0.013, 0.056(0.034, 0.172(0.130, 0.064(0.040,
-2.62 0.28 0.46 1.16
0.068(0.043, 0.088(0.059, 0.104(0.072, 0.188(0.144,
(0.38) ¢(0.47) ¢(0.42) ¢(0.46)
0.057) 0.092) 0.224) 0.102)
(0.25) c(0.34) ¢(0.33) c(0.30)
0.107) 0.130) 0.148) 0.24)
a Parameters are calculated on a logit scale g = logit(p) = l o g [ p / ( 1 - p)] and p = inverse logit(g) = 1 / [ 1 + e x p ( - g ) ] where p is a proportion of cells with an abnormality and g is the logit of p. b Estimated proportion of cells with abnormality and 95% confidence interval based upon standard errors from logistic model. The proportions are shown as their estimated values for each ozone exposure group and not as the difference from the control group. The 95% confidence limits are inverse logit of g 1.96 S.E. c Estimated difference from control group (0 ppm ozone) on the logit scale. These values are added to the control values to obtain g = logit(p) for the particular exposure level.
71 table illustrates that all of the aberrations were of the chromatid-type and that no exchanges or complex aberrations were observed in this study. Table 2 shows the results of fitting the linear logistic regression model to the data. Likelihood ratio statistics indicated that the categorical variables for ozone exposure levels were statistically significant for chromatid deletions (X 2 = 37.33; d.f. = 3; p < 0.005) and chromatid gaps (X 2 = 19.65; d.f. = 3; p < 0.005). There was no indication of extra binomial variation for either endpoint ( p > 0.05). The exposure to ozone significantly increased the proportion of cells with chromatid deletions and with chromatid gaps. The proportion of cells with chromatid deletions was highest in the 0.27-ppm exposure group. Discussion The results of this study show that ozone exposure induces an increase in the frequency of chromatid-type aberrations in lung macrophages. These results are concordant with those from studies of human lymphocytes exposed in vivo to 0.5 p p m ozone for 6 and 10 h (Merz et al., 1975). Evidence from in vitro studies with human lymphocytes showed the same response. Following a brief exposure (5 or 10 min) to 5 p p m ozone, a significant increase in the number of chromatid deletions was observed in human lymphocytes (Fetner, 1962). Gooch et al. (1976) also reported a similar increase in chromatid deletions in human lymphocytes after an in vitro exposure to 7.23 or 7.95 p p m / h for 30, 60 or 90 min. An increase in the frequency of chromosomal damage was also detected in lymphocytes of Chinese hamsters exposed to 0.2 p p m ozone for 5 h (Zelac et al., 1971). However, the predominant type of damage found in Chinese hamster cells was chromosometype aberrations, especially at longer periods of time after the ozone exposure. Based on the results of these studies, Zelac et al. (1971) postulated that ozone was a radiomimetic agent. Most studies, including our own, failed to detect ozone-related increases in the frequency of chromosome-type aberrations, suggesting that ozone is not a radiomimetic substance, but m a y be S-dependent in its clastogenic action. Tice et al. (1978) failed to detect significant
increases in the levels of chromatid deletions in blood lymphocytes of exposed Chinese hamsters either immediately or 14 days after ozone exposure. However, they did note that there was an increase in chromatid deletions at 7 days post-exposure, as well as an increase in the frequency of chromatid gaps 14 days after ozone inhalation. Other studies using h u m a n blood lymphocytes (McKenzie et al., 1977; G o o c h et al., 1976) or mouse bone-marrow cells (Tice et al., 1978) also failed to detect any clastogenic effects following either in vivo inhalation exposure or in vitro exposure to ozone. The inability to detect cytogenetic effects of ozone in these studies might have been related to both the concentration x time products and the low dose of ozone that reached the target bone-marrow cells and lymphocytes. In the current study, the macrophages were exposed to an effective concentration of ozone that would be expected to be much higher than that received by the bone-marrow cells or by the lymphocytes in previous studies. Our finding of an increase in the level of chromatid gaps in macrophages of animals exposed to the highest concentration of ozone is interesting. A similar increase in the number of chromatid gaps has been reported for lymphocytes of ozoneexposed humans (Merz et al., 1975) and Chinese hamsters (Tice et al., 1978). The biological consequences of chromatid gaps are not well understood, although it has been suggested that they arise from unrepaired, D N A single-strand breakage (Bender et al., 1974). Such damage may, after cell replication, be detected as chromatid deletions in subsequent cell divisions. The observation that there was a decrease in chromosomal damage in animals exposed to the highest concentration of ozone (0.8 ppm) is difficult to understand, because the next lower level (0.27 ppm) had a higher frequency of chromosome aberrations. A possible explanation might be that ozone reacted with the lung tissue in such a way as to induce an influx of macrophages into the bronchoalveolar airspace and to stimulate the lung macrophage population to divide. Thus, both the resident cell population and the mitotic index changed. Since the exposure was terminated after only 6 h, m a n y of the cells recovered by lavage may not have been directly exposed to ozone. The
72
endpoint of the cytogenetic assay used in this study reflects the ratio of cells with chromosomal aberrations to normal cells. Therefore, an ozoneinduced increase in the total number of macrophages recovered by lavage would tend to "dilute out" the macrophage population that was actually exposed to ozone, resulting in an underestimation of the cytogenetic effects of ozone at higher doses. This interpretation is supported by the increase in mitotic index observed in the present study, as well as by the results of a previous study by Hotchkiss et al. (1988). These authors reported a significant increase in the labeling index of alveolar macrophages 18 h after a 6-h exposure to 0.8 ppm ozone, as well as a significant increase in the total number of macrophages recovered by lavage between 18 and 42 h after exposure to 0.8 ppm ozone. It is also possible that macrophages exposed to the highest dose of ozone (0.8 ppm) were too fragile to be recovered by lavage, because of the toxic effects of ozone at this concentration (Dowell et al., 1970; Alpert et al., 1971). In summary, our study demonstrates cytogenetic effects in pulmonary alveolar macrophages after exposure of rats to ozone levels that may be found in many large metropolitan cities. Because macrophages are not the major cells at risk for the induction of cancer or toxicity, chromosome damage in these cells can only be considered indicative of the actual level of ozone exposure or effective dose to lung cells. The cytogenetic damage observed in macrophages might, however, be predictive of cytogenetic damage that would be produced in lung epithelial cells. Additional research needs to be conducted to determine the level of damage in other cell types more relevant to cancer induction or toxicity in the lung resulting from ozone exposure. However, this research did demonstrate that by using the proper cell type, it is possible to detect cytogenetic damage in the lungs of animals that were exposed to environmentally relevant, low concentrations of ozone over a brief time period. Our research suggests that the genetic material in lung cells might be damaged by ozone, thereby increasing the risk of lung cancer and presenting a potential human health problem.
Acknowledgements We thank Drs. D.E. Bice, N.F. Johnson, J.A. Bond and T. Coons for their constructive reviews of the manuscript. The technical help of Mary Jo Waltman is also acknowledged. This research was performed under U.S. Department of Energy Contract No. DE-AC04-76EV01013 and by N I H Grant ES04282, and conducted in facilities fully accredited by the American Association for Accreditation of Laboratory Animal Care.
References Alpert, S.M., D.E. Gardner, D.J. Hurst, T.R. Lewis and D.L. Coffin (1971) Effects of exposure to ozone on defensive mechanisms of the lung, J. Appl. Physiol., 31,247-252. Bender, M.A, H.G. Griggs and J.S. Bedford (1974) Mechanisms of chromosomal aberrations production, III. Chemicals and ionizing radiation, Mutation Res., 23, 197-212. Bice, D.E., R.J. Cox and C.T. Schnizlein (1977) Lavage techniques to obtain maximum numbers of rat alveolar macrophages and their use in migration assays, in: B.B. Boecker, C.H. Hobbs and B.S. Martinez (Eds.), Inhalation Toxicology Research Institute Annual Report, LF-58, pp. 390-394. Borek, C., M. Zalder, A. Ong, H. Mason and G. Witz (1986) Ozone acts alone and synergistically with ionizing radiation to induce in vitro neoplastic transformation, Carcinogenesis, 7, 1611-1613. Dowell, A.R., L.A. Lohrbauer, D. Hurst and S.D. Lee (1970) Rabbit alveolar macrophage damage caused by in vivo ozone inhalation, Arch. Environ. Health, 21, 121-127. Evans, M.J., L.V. Johnson, R.J. Stephens and G. Freeman (1976) Renewal of the terminal bronchiolar epithelium in the rat following exposure to NO 2 or 03, Lab. Invest., 34, 565-578. Fetner, R.H. (1962) Ozone-induced chromosome breakage in human cell cultures, Nature (London), 194, 793-794. Gooch, P.C., D.A. Creasia and J.G. Brewen (1976) The cytogenetic effects of ozone: inhalation and in vitro exposures, Environ. Res., 12, 188-195. Guerrero, R.R., D.E. Rounds, R.S. Olson and J.D. Hackney (1979) Mutagenic effects of ozone on human cells exposed in vivo and in vitro based on sister chromatid exchange analysis, Environ. Res., 18, 336-346. Hasset, C., M.G. Mustafa, W.F. Coulson and R.M. Elashoff (1985) Murine lung carcinogenesis exposure to ambient ozone concentrations, J. Natl. Cancer Inst., 75, 771-777. Hotchkiss, J . A . J . R . Harkema, D.T. Kirkpatrick and R.F. Henderson (1989) Response of rat alveolar macrophages to ozone: quantitative assessment of population size, morphology and proliferation following acute exposure, Exp. Lung Res., 15, 1-16.
73 McCullagh, P., and J.A. Nelder (1983) Generalized Linear Models, Chapman and Hall, London, 261 pp. McKenzie, M.H., J.H. Knelson, N.J. Rummo and D.E. House (1977) Cytogenetic effects of inhaled ozone in man, Mutation Res., 48, 95-102. Menzel, D.B. (1984) Ozone: an overview of its toxicity in man and animals, J. Toxicol. Environ. Health, 13, 183-204. Merz, T., M.A Bender, H.D. Keller and T.J. Kull (1975) Observation of aberrations in chromosomes of lymphocytes from human subjects exposed to ozone at a concentration of 0.5 ppm for 6 and 10 hours, Mutation Res., 31, 299-302. National Research Council Committee (1977) Report on Medical and Environmental Effects of Environmental Pollutants: Ozone and Other Photochemical Oxidants, National Academy of Science (U.S.A.), Washington.
Rabinowitz, J.L., and D.J.P. Bassett (1988) Effects of 2 ppm ozone exposure on rat lung lipid fatty acids, Exp. Lung Res., 14, 477-489. Rithidech, K., B.T. Chen, J.L. Mauderly and A.L. Brooks (1989) Cytogenetic effects of cigarette smoke on alveolar pulmonary macrophages of the rat, Environ. Mol. Mutagen., 14, 27-33. Tice, R.R., M.A Bender, J.L. Ivett and R.T. Drew (1978) Cytogenetic effects of inhaled ozone, Mutation Res., 58, 293-304. Zelac, R.E., H.L. Cromroy, W.E. Bolch Jr., B.G. Dunavant and H.A. Bevis (1971) Inhaled ozone as a mutagen, I. Chromosome aberrations induced in Chinese hamster lymphocytes, Environ. Res., 4, 262-282.
