 Ozone Selectively Inhibits Growth of Human Cancer Cells

 Science Vol. 209, 22 Aug 1980, pp. 931-933

 Abstract:

 The growth of human cancer cells from lung, breast, and uterine 
 tumors was selectively inhibited in a dose-dependent manner by 
 ozone at 0.3 to 0.8 part per million of ozone in ambient air 
 during 8 days of culture.  Human lung diploid fibroblasts served 
 as noncancerous control cells.  The presence of ozone at 0.3 to 
 0.5 part per million inhibited cancer cell growth 40 and 60 
 percent, respectively.  The non-cancerous lung cells were 
 unaffected at these levels.  Exposure to ozone at 0.8 part per 
 million inhibited cancer cell growth more than 90 percent and 
 control cell growth less than 50 percent.  Evidently, the 
 mechanisms for defense against ozone damage are impaired in 
 human cancer cells.

 The effects of ozone on human health have been a focus of public 
 concern and scientific investigation for more than two decades 
 (I-4).  Considerable attention has been devoted to assessing its 
 cellular effects (5) because it is the major constituent of the 
 ground-level oxidants in polluted air.  Much has been learned 
 about the effects of ozone on normal tissue, but little is known 
 about its action on cancer cells.  We have conducted experiments 
 in which continuous exposure to ozone at 0.3 ppm (6) selectively 
 inhibited the growth of human cancer cells 40 percent in 8 days.

 Controlled levels of ozone (0.3 to 0.8 ppm) were continuously 
 generated by ultraviolet irradiation of filtered ambient air.  
 The ozonated air, containing 5 percent carbon dioxide, was 
 introduced at a constant flow rate of 4.0 liter/min into an 
 environmental chamber in an incubator maintained at 37 degrees 
 Celsius (Fig. 1).  The ozone levels were assayed daily with a 
 spectrophotometric ozone analyzer.  For comparison, noncancerous 
 human lung diploid fibroblasts (7) were cultured in the chamber 
 along with the cancer cells.  The cancer cells were from 
 alveolar (lung) adenocarcinomas (8), breast adenocarcinomas (9), 
 uterine carcinosarcomas, and endometrial carcinomas (10).  All 
 the cells were grown in 60-mm petri dishes in 10 ml of medium 
 and were placed in the chamber at the same time.  Control cells 
 were incubated in an adjoining compartment receiving filtered 
 ambient air containing 5 percent carbon dioxide (4.0 liter/min).  
 Three petri dishes for each cell type were removed from each of 
 the two compartments every 48 hours for 8 days, and the number 
 of cells per plate were counted.  All of the cancer cells showed 
 marked dose-dependent growth inhibition in ozone at 0.3 and 0.5 
 ppm (Fig 2.).  There was no growth inhibition of the 
 noncancerous lung cells at these ozone levels, and they were 
 morphologically identical to the corresponding control cells.  
 At 0.8 ppm, the growth of the noncancerous cells was inhibited 
 50 percent, but all four types of cancer cells were inhibited 
 more than 90 percent.

 After being cultured through 14 passages, the noncancerous cells 
 exhibited measurable growth inhibition and morphological changes 
 (vacuolation) in ozone at 0.5 ppm, suggesting that aging 
 increases the sensitivity of normal lung cells to ozone (Fig 3).  
 In cultured human diploid fibroblasts, morphological changes and 
 a gradual decrease in rate of growth have been attributed to a
 buildup of cellular damage with each successive division 
 (11,12).  Ozone may accelerate processes similar to those 
 naturally causing cellular damage and may decrease the growth 
 rate of the aging fibroblast colony.  However, in ozone at 0.5 
 ppm, all of the human cancer cells (which do not age) had growth 
 rates several times lower than that of the aged, noncancerous 
 cells (Fig 2.).

 Evidently, cancer cells are less able to compensate for the 
 oxidative burden of ozone than normal cells.  The marked 
 sensitivity of cancer cells to ozone raises questions about the 
 possible mechanisms of oxidative inhibition of their growth.  
 Virtually every major component of normal cells has been found 
 to be affected by elevated ozone levels (5).  However, 
 glutathione in its reduced form (GSH) has been credited with 
 providing the first line of defense against the peroxides and 
 free radicals generated in all cells by ozone and oxygen (1, 13-
 15).  It deactivates peroxides and radicals by donating one 
 hydrogen atom to the reactive species.  Loss of a GSH hydrogen 
 (oxidation) results in formation of oxidized glutathione (GS-
 SG).  The cellular respiratory system is responsible for 
 reducing GS-SG to GSH.  The GSH-linked respiratory system in 
 normal and cancer cells, before and after exposure to ozone, 
 must be examined to learn whether a functional impairment of 
 this system is associated with the marked sensitivity of cancer 
 cells to the oxidant.

 These findings lead us to believe that ozone--alone, in 
 combination with radiation therapy (16), or in chemotherapy 
 utilizing electrophilic compounds (17)--may have therapeutic 
 value for patients with certain forms of lung cancer.


 Frederick Sweet
 Ming-Shian Kao
 Song-Chiau D. Lee

 Department of Obstetrics and Gynecology, Washington University 
 School of Medicine, St. Louis, Missouri, 633110.

 Will L. Hagar

 City of St. Louis Air Pollution Control, St. Louis, 63103

 Wileen E. Sweet

 Air Quality Section, East-West Gateway Coordinating Council, St. 
 Louis, 63102.



 Figure 1.  Schematic diagram (not shown) of the system used for 
 culturing human cells in ozonated ambient air.  Filtered ambient 
 air was mixed with carbon dioxide (5 percent) and introduced 
 into a dual chamber incubator (National 331).  Half was 
 conducted through a calibrated ozone generator consisting of a 
 quartz glass tube irradiated with ultraviolet light and then 
 into a hermetically sealed (20 by 20 by 20 cm) glass and 
 stainless steel environmental chamber containing a gasketed 
 access door.  Output of ozone from the generator varied less 
 than 1 percent per day.  The ozone content of the vented air 
 from the chamber was measured daily with a spectrophotometric 
 ozone analyzer (Dasibi 1003-AH).  Malignant and normal human 
 cells were incubated in chamber E saturated with water vapor.  
 Corresponding cells serving as controls were incubated in the 
 adjoining compartment, also saturated with water vapor.


 Figure 2.  Inhibition by ozone of growth of malignant and non-
 malignant cells in culture on day 8.  Each of the cell types 
 were grown in 10 ml of Dulbecco's modified Eagle's minimum 
 essential medium containing 10 percent calf serum.  In a typical 
 experiment, 12 dishes per cell line (usually three or four cell 
 lines were tested per experiment) were loaded into the 
 environmental chamber with an equal number of control dishes in 
 the adjoining compartment (Fig. 1).  The initial population was 
 3 x 10(5) cells per dish.  Every 48 hours three dishes for each 
 cell type were removed from both compartments and the cells were 
 tested for viability with 0.4 percent trypan blue and counted 
 with a hemocytometer.  Each data point represents the number of 
 experimental cells divided by the number of corresponding 
 control cells per dish multiplied by 100 (the percentage of 
 control growth) and is plotted against the measured level of 
 ozone in the environmental chamber.  The percentage of growth 
 inhibition is calculated by subtracting the percentage of growth 
 from 100.  The data are from cell counting on day 8 of 
 incubation.  There is a nearly linear relation between 
 inhibition of the growth of the cancer cells and increasing 
 ozone levels.  The noncancerous cell line IMR-90 began to 
 display measurable growth inhibition only when ozone levels 
 exceeded 0.5 ppm, a level that produced approximately 60 percent 
 inhibition in all of the cancer cells lines tested.  There was 
 some growth inhibition in noncancerous cells aged through 14 
 passages.  The mean populations of the cells serving as controls 
 were as follows (per dish on day 8): IMR-90, 34.8 x 10(5); A-
 549, 36.5 x 10(5); MCF-7, 57.0 x 10(5); endometrial 
 adenocarcinoma, 64.2 x 10(5); myometrial carcinosarcoma, 121.1 x 
 10(5).


 References:

 1.  D.L. Dunsworth, C.E. Cross, J.R. Gillespie, C.G. Plopper, in 
     Ozone Chemistry and Technology.  J.S. Murphy and J.R. Orr, 
     Eds. (Franklin Institute, Philadelphia, 1975), chap. 2.

 2.  H.E. Stokinger and D. Coffin, in Air Pollution, A.C. Stern, 
     Ed. (Academic Press, New York, 1968), vol. 1, pp. 446-546.

 3.  H.D. Kerr et al., Am. Rev. Respir. Dis. 111, 763 (1975).

 4.  J.D. Hackney, W.S. Linn, C.D. Law, S.K. Karuza, Greenberg, 
     R.D. Buckley, E.E. Pedersen, Arch. Environ. Health 30, 385 
     (1975).

 5.  B.D. Goldstein, Rev. Environ. Health 2, 177 (1977).

 6.  Normal human subjects tolerated breathing 0.5 ppm ozone in 
     air 2 hours per day for 1 week or 0.25 ppm ozone 2 hours per 
     day for 3 weeks (4).  The two groups engaged in light 
     exercise during exposure.  Although both groups developed 
     chest discomfort and moderately decreased respiratory 
     function during exposure, their removal from the oxidative 
     environment resulted in rapid disappearance of the symptoms.  
     The mean dose-response curves from this study show a no-
     detectable-effect level at 0.25 to 0.30 ppm.  A similar 
     study (3) found that human subjects tolerated exposure to 
     0.5 ppm ozone for up to 6 hours.  Pulmonary function was 
     affected and chest discomfort developed at this level, with 
     no significant differences observed between smokers and 
     nonsmokers.

 7.  These cells (IMR-90) were obtained from the Human Aging Cell 
     Repository and plated 48 hours after shipping.  This cell 
     type was characterized by W.W. Nichols, D.G. Murphy, V.I. 
     Cristofalo, L.H. Toji, A.E. Greene, and S.A. Dwight [Science 
     196, 60 (1977)].

 8.  This cell line (A-549) was described by D.J. Glard, S.A. 
     Aaronson, G.J. Todard, P. Arnstein, J.H. Kersey, H. Dorsik, 
     and W.P. Parks [J. Natl. Cancer Inst. 51, 1417 (1973)]; M. 
     Lieber, B. Smith, A. Szakal, W. Nelson-Rees, and G.A. 
     Todardo [Int. J. Cancer 17, 62 (1976)]; and K.L. Jones, N.S. 
     Anderson III, and J. Addison [Cancer Res. 38, 1688 (1978)]

 9.  This cell line (MCF-7, estrogen-sensitive) was described by 
     H.D. Soule, J. Vazquez, A. Long, S. Albert, and M. Brennam 
     [J. Natl. Cancer Inst. 51, 1409 (1973)] and by K.B. Horwitz, 
     M.E. Kostlow, and W.I. McGuire [Steroids 26, 785(1975)].

 10. Human uterine carcinosarcoma cells and endometrial 
     adenocarcinoma cells were obtained from pathologically 
     confirmed gynecologic tumors and developed as new cell 
     lines.  The endometrial adenocarcinoma cell line is 
     estrogen-sensitive.  Both were described by M.S. Kao and 
     S.C.D. Lee (27th Annual Meeting of the Society for 
     Gynecologic Investigation, Denver, 20 to 23 March 1980), 
     abstr. 7.

 11. J.R. Smith and R.G. Whitney, Science 207, 82 (1980); S.C.D. 
     Lee, P.M. Bemiller, J.N. Bemiller, A.J. Papelis, Mech. 
     Ageing Dev. 7, 417 (1978).

 12. P.M. Bemiller and L.H. Lee, ibid. 8, 417 (1978)

 13. C.K. Chow and A.L. Tappel, Lipids 1, 518 (1972).

 14. C.K. Chow, Nature (London) 260, 721 (1976).

 15. R.E. Kimball et al., Am. J. Physiol. 230, 1425 (1976).

 16. R.E. Lee, Semin. Oncol. 1, 254 (1974)

 17. O.S. Selawry, ibid., p. 259.

 18. Parts of this report were presented at the 27th Annual 
     Meeting of the Society for Gynecologic Investigation, 
     Denver, 20 to 23 March 1980 (abstracts 7 and 150), and at 
     the 73rd Annual Meeting of the Air Pollution Control 
     Association, Montreal, 24 June 1980 (poster session 27).  We 
     thank W. Nelson-Rees for his gift of A-549 cells; the MCF-7 
     cells were obtained from E.M. Jensen.  We also thank C.M. 
     Copley, Jr., H.M. Camel, and T. Morgan for their 
     constructive criticism of the manuscript.  Correspondence 
     should be addressed to F.S.

 24 April 1980; revised 11 June 1980.
