Environmental mutagens in urban air particulates

0 downloads 0 Views 1MB Size Report
Oct 20, 2009 - To cite this article: Barry Commoner , Prema Madyastha , Alice Bronsdon & Antony J. Vithayathil (1978) Environmental mutagens in urban air ...
Journal of Toxicology and Environmental Health

ISSN: 0098-4108 (Print) (Online) Journal homepage: http://www.tandfonline.com/loi/uteh19

Environmental mutagens in urban air particulates Barry Commoner , Prema Madyastha , Alice Bronsdon & Antony J. Vithayathil To cite this article: Barry Commoner , Prema Madyastha , Alice Bronsdon & Antony J. Vithayathil (1978) Environmental mutagens in urban air particulates, Journal of Toxicology and Environmental Health, 4:1, 59-77, DOI: 10.1080/15287397809529645 To link to this article: http://dx.doi.org/10.1080/15287397809529645

Published online: 20 Oct 2009.

Submit your article to this journal

Article views: 3

View related articles

Citing articles: 30 View citing articles

Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalInformation?journalCode=uteh19 Download by: [Aristotle University of Thessaloniki]

Date: 09 June 2016, At: 23:21

ENVIRONMENTAL MUTAGENS IN URBAN AIR PARTICULATES Barry Commoner, Prema Madyastha, Alice Bronsdon, Antony J. Vithayathil

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

Center for the Biology of Natural Systems, Washington University, St. Louis, Missouri

A bioassay capable of detecting carcinogenic substances that are associated with the elevated incidence of cancer in the urban environment would be important for epidemiologic and environmental analyses. The feasibility of using the Salmonella mutagenesis system developed by Ames for this purpose has been tested by analyzing Chicago air paniculate samples. Active material, as evidenced by enhanced rates of mutation, both in the presence of microsomes and in their absence, is readily extractable from samples of air particulates. Dose-response curves have been obtained from such extracts of 15 successive samples, taken at intervals during 1975 from a sampling site in South Chicago. A method for analyzing such data in order to evaluate the relative mutagenic activity of different samples is described. The presence of a number of mutagenic constituents has been demonstrated by means of thin-layer chromatography of particularly active samples, in which the active material is located by mutagenic analysis of successive chromatographic zones. Mass spectrometer analysis of material isolated from an original sample in this way indicates that benzo[a]pyrene and benzo[e]pyrene, which are known to be mutagenic and carcinogenic, are present. It is concluded that within certain constraints, which are described, the methodology can serve the purpose of an environmental bioassay for organic carcinogens.

INTRODUCTION It is now widely recognized that the urban environment in the United States is associated with an elevated incidence of cancer (Hoover and Fraumeni, 1975) and that urban airborne particulates may contain carcinogens (Sawicki, 1967). It is therefore of considerable interest to analyze the occurrence of particulate-associated carcinogens in the urban environment. Such information—in particular, determination of the identities and amounts of different carcinogens, as they are affected by basic Work reported in this paper was supported by the National Science Foundation/RANN grant ENV 75-14127. The authors wish to acknowledge with thanks Commissioner H. W. Poston, city of Chicago Department of Environmental Control, for providing us with samples of air filters used in this study. The mass spectrometer analyses reported in this paper were carried out at the Washington University Mass Spectrometry Resource, established under National Institutes of Health grant RR 00954. The authors wish to thank Drs. Michael K. Hoffman and William F. Holmes for assistance in performing these analyses. Requests for reprints should be sent to Barry Commoner, Center for the Biology of Natural Systems, Washington University, St. Louis, Missouri 63130.

59 Journal of Toxicology and Environmental Health, 4:59-77,1978 Copyright© 1978 by Hemisphere Publishing Corporation 0098-4108/78/0401-0059$2.25

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

60

B. COMMONER ET AL.

environmental variables such as time, place, and weather conditions—is necessary if epidemiologic data on the elevated incidence of cancer in urban areas are to be correlated with specific agents. Such information is also essential in order to determine the origin of carcinogenic agents so that steps can be taken to reduce human exposure to them. Past efforts in this direction have been based on the approach of determining, by means of appropriate chemical analyses, what substances previously identified as carcinogens are present in the particulates. By this means it has been established, for example, that benzo[ff]pyrene, which is known to cause cancer in exposed laboratory animals and humans, often occurs in association with urban air particulates (Hoffman and Wynder, 1968). However, this general approach is difficult to carry out on the scale that is necessary to accomplish epidemiologic correlations or to determine the origins of carcinogens. First, it is impractical to carry out the large numbers of analyses needed for epidemiologic correlations or for tracing the movement of environmental constituents in which dozens of different complex organic compounds need to be identified. Second, this approach would fail to detect substances not yet tested for carcinogenicity, which may readily occur in environmental samples. Biological screening offers an alternative approach that may resolve these difficulties. A screening test capable of rapid detection of carcinogenic activity would obviate the need for numerous total analyses of the compounds present in environmental samples in order to determine whether carcinogenic substances are present. Instead, the screening test, together with conventional physicochemical fractionation procedures, could be used to isolate and identify the relatively few substances, among the many that may occur in environmental samples, that are carcinogenic. A screening test may also detect in environmental samples carcinogenic agents that have not yet been identified as such. Finally, if suitably calibrated, such a screening procedure might be used to estimate the relative concentrations of carcinogenic agents, thus providing the data needed in connection with epidemiologic studies of cancer incidence, and to trace the origins of environmental carcinogens. Among the biological screens that are capable of detecting carcinogens, one of the most promising is that developed by Ames et al. (1973) on the basis of special histidine-negative strains of Salmonella typhimurium. These strains respond to the presence of certain classes of mutagenic substances by reverting to a histidine-positive genome. The frequency of this reverse mutation is readily measured by counting the colonies that appear in an appropriate medium that lacks histidine. While the Salmonella system does not respond to all classes of carcinogens (for example, metallic ions and asbestos are inactive), it is capable of detecting a wide range of organic carcinogens that are likely to occur in environmental samples. The theoretical link between mutagenesis and carcinogenesis on which

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

ENVIRONMENTAL MUTAGENS IN URBAN AIR PARTICULATES

61

this technique was based is the subject of considerable debate. However, the practical applicability of mutagenesis to the detection of carcinogens depends not on the validity of the theory but on the empirical relationship between these two properties, regardless of theoretical considerations. The potential usefulness of the technique has now been established by empirical studies, which show that it is capable of distinguishing between organic compounds that are not carcinogenic toward laboratory animals and those that are, with a reliability of at least 80-90%. This correlation has been demonstrated in our laboratory (Commoner et al., 1976) and by several other groups (McCann et al., 1975). It would appear, therefore, that the Salmonella technique might be a suitable means of developing the foregoing types of data, thereby providing a basis for epidemiologic and environmental analysis of the role of organic carcinogens in the incidence of cancer. In the present study we have evaluated the feasibility of this approach by applying the Salmonella mutagenesis screen to the problem of detecting the carcinogens that occur in urban air particulates in the Chicago area.

METHODS Samples In collaboration with the Department of Environmental Control of the city of Chicago, we have obtained a series of their standard high-volume samples of Chicago air particulates. These are collected on filters by a 25-station network established by the department on the roofs of city high schools. Each sample represents the particulates collected over a 24-h period. The samples were provided to use together with data on the weight of the collected particulates and associated meteorologic information. Extraction Procedures On the basis of preliminary experiments the following general procedures were adopted to extract the air filter samples. Each filter ( 8 X 1 0 in) was shredded and extracted in a Soxhlet apparatus for 4 h in 300 ml of a benzene:hexane:isopropanol (70:10:20) mixture. The extract was then evaporated to dryness and the residue taken up in 10 ml of chloroform. A one-tenth aliquot of the chloroform solution (representing 8 in2 of the filter) was dried, the residue was taken up in 4 ml of dimethyl sulfoxide (DMSO), and aliquots of increasing size were tested for mutagenic activity as described below. A two-tenths aliquot of the chloroform solution (representing 16 in2 of the filter) was dried and taken up in 5 ml of benzene, which was transferred to a separatory funnel together with 5 ml of distilled water. After shaking, the two layers were separated, each was dried, the residues were taken up in 4 ml of DMSO, and aliquots of increasing size were tested for mutagenic activity as described below.

62

B. COMMONER ET AL.

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

Mutagenic Analysis In a preliminary test it was determined that extracts of typical samples yielded the greatest number of mutants with Ames' Salmonella strain TA1538. Accordingly, subsequent tests were carried out with this strain, on plates prepared as described elsewhere (Ames et al., 1975; Commoner et al., 1976). Samples were tested in the presence of a standard microsome preparation prepared from Aroclor-induced rat liver (see Ames et al., 1975) and in its absence. In each test, duplicate plates were prepared at a series of dilutions of the original extract, thus yielding a dose-response curve for each sample. Tests were also carried out on equivalent blanks of the standard filter. Concurrently with each set of analyses, control plates (bacteria alone, bacteria plus microsome preparation, extract alone) were tested, together with plates containing a standard carcinogen (in order to check on the activity of the system). From the numbers of revertant mutant (i.e., histidine-positive) colonies observed on the test and control plates after a 48-h incubation period, the mutagenic activities of different samples, at various dilutions, were determined. All values are based on the averages of duplicate plates. In interpreting the significance of colony counts obtained from experimental samples we have employed an approach developed earlier, based on the comparison of 50 known organic noncarcinogens and 50 organic compounds that had been previously shown to be carcinogenic toward laboratory animals (Commoner et al., 1976). In this comparison we computed a "mutagenic activity ratio" from the quotient (£ — C)/CAv, where E is the number of mutant colonies obtained from the experimental sample; C is the control value (i.e., the number of mutant colonies observed when the experimental material is not included) obtained on the day of analysis; and CAv is the "historical" control value, or the average control value for all runs carried out during the course of the study. We have shown that in the presence of liver microsomes 93% of the noncarcinogens yield mutagenic activity ratios of 2.0 or less and 83% of the carcinogens yield ratios greater than 2.0. About 98% of the noncarcinogens yield ratios of 2.0 or less if microsomes are absent. In the analyses reported below a sample was regarded as possessing statistically significant mutagenic activity if at some concentration it yielded a mutagenic activity ratio greater than 2.0. Isolation and Identification of Mutagenically Active Materials in Air Particulates For this purpose a seven-tenths aliquot of a chloroform solution (representing about 56 in2 of a Washington School filter sample), prepared as described above, was fractionated by means of thin-layer chromatography, using Gelman silica gel-impregnated glass fiber paper, in various solvent systems, as indicated below. The developed chromatogram was

63

ENVIRONMENTAL MUTAGENS IN URBAN AIR PARTICULATES

divided into 1-cm-wide zones, each of which was then extracted with 10% methanol in chloroform. The extracts were dried, and the residues were taken up in DMSO and tested for mutagenic activity against strain TA1538 in the presence of liver microsomes. By means of repeated fractionation in different solvents it was possible to achieve a preparation that contained a single mutagenically active chromatographic peak. This preparation was then subjected to UV analysis (using a Cary model 11 recording spectrophotometer) and mass spectrometry (in a Finnigan model 3300 mass spectrometer, using a probe and an electron impact detector).

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

RESULTS As a preliminary step, analyses were made of benzene:hexane (50:50) extracts of 2-in2 samples of filters collected concurrently from a series of different stations in the Chicago air pollution system. The results, which are shown in Fig. 1, revealed a general proportionality between particulate concentration and the numbers of revertant colonies and identified the Washington School in South Chicago as a site considerably more active than the rest. In an effort to improve the efficiency of extraction it was then found that extracts obtained with benzene:hexane:isopropanol (70:10:20) yielded somewhat higher revertant colony counts than those obtained with benzene:hexane, and the former solvent was used thereafter. Air particulate samples collected at intervals during 1975 from the

240

100

200

300

400

Concentration of Air Particulates (/jg/m3) FIGURE 1. Number of revertant colonies produced per test plate by benzene:hexane extracts of 2-in2 aliquots of- air particulate filters from different Chicago collection sites. Tested on strain TA1538, with rat liver microsomes. The highest value is from the Washington School site.

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

64

B. COMMONER ET AL.

Washington School site have been analyzed. Dose-response curves were obtained for each air filter, with and without the presence of rat liver microsomes, from each of the following: (1) the benzene:hexane:isopropanol extract; (2) the benzene-soluble fraction of the benzene:hexane:isopropanol extract; and (3) the water-soluble fraction of the benzene:hexane:isopropanol extract. The dose-response curves obtained from a typical sample, based on the number of revertant colonies induced by extracts representing increasing amounts of the sample, are shown in Fig. 2. It is evident that in the presence of microsomes this sample exhibits significant mutagenic activity in both the original extract and the benzene fraction derived from it. This is established by the rise in activity with paniculate concentration and by the fact that the highest values obtained represent mutagenic activity ratios that are clearly positive by the criterion described earlier (4.1 and 8.4, respectively). It is possible that there is a slight amount of mutagenic activity in the absence of microsomes, but the effect is borderline, for the mutagenic activity ratios, at the maxima, are only 2.6 and 2.3, respec-

Benzene: Hexane : Isopropanol Extract

§

8

1

Benzene soluble fraction of Benzene: Hexane: Isopropanol Extract

150

100-

50 0 200 Water soluble fraction of Benzene: Hexane: Isopropanol Extract

150100

50 0

50

100

150

200

Equivalent Amount of Air Portlculates/Plate (mg) FIGURE 2. Number of revertant colonies produced per plate by increasing amounts of Chicago air particulate extracts of a filter sample from the Washington School site. Tested on strain TA1538, with (solid line) and without (broken line) rat liver microsomes.

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

ENVIRONMENTAL MUTAGENS IN URBAN AIR PARTICULATES

65

tively. It is also evident from Fig. 2 that the water fraction of the original extract is devoid of mutagenic activity at the concentration tested. The foregoing considerations lead to only qualitative conclusions regarding the absence or presence of mutagenic material in extracts prepared from air particulates. However, for reasons cited earlier, it is of obvious interest to determine how one might compare the levels of mutagenic activity exhibited by the samples obtained at successive intervals from the Washington School site. For this purpose it is necessary to take into account the complex nature in the Ames test. To begin with, such dose-response curves are rarely linear over a wide range of concentrations, frequently falling or rising significantly in slope with concentration. In many instances, because of the toxicity of the substance, the dose-response curve rises to a maximum and then declines sharply with concentration. Another factor that must be taken into account is that some substances are inherently mutagenic, while others require the presence of microsomes (in order to produce active metabolic intermediates). Moreover, some inherently mutagenic substances (for example, /V-methyl nitrosoguanidine) may be inactivated when microsomes are present. Instances that illustrate these effects are evident in the dose-response curves obtained from the Washington School samples. Figure 3 shows 6 of the 15-dose-response curves obtained from these samples for, respectively, the benzene:hexane:isopropanol extracts, the benzene fractions, and the water fractions. In each case the results obtained with microsomes present (solid line) and microsomes absent (broken line) are shown. Approximately linear dose-response curves are exemplified by those obtained from the February 26 and July 31 benzene fractions, with microsomes present. Many of the curves exhibit slopes that decline at higher concentrations. Several curves (e.g., February 8 sample, benzene:hexane:isopropanol extract, microsomes present) show slopes that are lower at the initial concentrations than at the higher ones. Instances of toxicity at higher concentrations can be seen in the July 31 sample, benzene: hexane:isopropanol extract, both with and without microsomes present. This curve also illustrates the inactivating effect of microsomes; at two of the lower sample concentrations the numbers of colonies produced when microsomes are present are lower than those observed when they are absent. In order to devise a procedure capable of comparing the mutagenic activities of different samples that takes into account the foregoing effects, we have adopted the following approach. To begin with, we note, on the basis of our earlier statistical comparison of the mutagenic activities of noncarcinogens and carcinogens, that there is a minimum value of {E — C)/CAv that determines, with the stated reliability figure, that the material is carcinogenic. Hence in the present analyses we may regard a mutagenic activity ratio of 2.0 or greater as indicative of the presence of

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

BENZENE :HEXANE: ISOPRORVNOL EXTRACT

BENZENE FRACTION

100

150

200

50

100

150

200

Equivalent Amount of Air Particulates/Plate (mg) FIGURE 3 . Number of revertant colonies (less control values) produced per plate by increasing amounts of air paniculate extracts collected on six different dates in 1975 at the Washington School site. Tested on strain T A 1 S 3 8 , with (solid line) and without (broken line) rat liver microsomes. The arrows mark the sample concentrations at which experimental minus control values = 2 X C^v (where C^v is the "historical" control value). The reciprocal of the indicated value is the relative mutagenic activity of the sample.

66

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

ENVIRONMENTAL MUTAGENS IN URBAN AIR PARTICULATES

67

active substances in the sample, with a reliability of about 98% if microsomes are absent and of about 93% if microsomes are present. We then determine from the sample's dose-response curve the lowest concentration of the sample at which {E — C)/CAv = 2 . 0 . This value, which is marked by the arrows in Fig. 3 , can be obtained from the dose-response curve by interpolation to determine the sample concentration at which (E — C) = 2.0 X CAv . The value can be determined in this way regardless of the shape of the dose-response curve (specifically, whether a maximum occurs or whether the initial slope is different from that at higher concentrations). 1 Finally, the reciprocal of the sample concentration at which the mutagenic activity ratio is 2.0 may be defined as the "relative mutagenic activity" of the sample. While this procedure does not take into account possible synergistic interactions among separate mutagens present in the sample, it does provide, as a first approximation, relative measures of the mutagenic activities of samples even if they yield dose-response curves that differ in shape. The relative mutagenic activities computed in this way for the benzene:hexane:isopropano! extracts and the. benzene fractions obtained from all 15 Washington School samples are plotted, as a function of sample date, in Fig. 4 . (The corresponding plot for the water fraction is not shown since in every sample the mutagenic activity ratios are zero.) The reported wind direction at each date is also indicated. The data of Fig. 4 support several conclusions. First, it is evident that in the presence of microsomes the level of activity of the benzene fraction generally parallels that of the original extract from which it is derived, provided that there is little or no inherently active material present (i.e., material that is mutagenic in the absence of microsomes). This situation occurs in the latter half of the year, so that in that period both the original extract and the benzene fraction show similar variations in activity with time. However, the activities of the benzene fraction are generally about half of those exhibited by the comparable original extracts, suggesting that active material is lost during the fractionation procedure. Second, it is evident that in the first half of the year several instances occur in which the samples exhibit considerable inherent mutagenic activity, and that at least a good part of this activity is lost when microsomes are present, presumably because of enzymatic inactivation. Thus, for these samples the mutagenic activity in the presence of microsomes is only a gross value. It may include, in addition to the effect of substances that produce mutagenic metabolites only when acted on by 1 It should be noted that, as we have observed earlier (Commoner et al., 1976), the value of CAv in the presence of microsomes is somewhat greater than that when microsomes are absent. This effect, which we have investigated separately, is consistently observed with strain TA1538.

68

B. COMMONER ETAL.

0.2

Benzene-Soluble Fraction

Direction

0.15TA 1538 With Rat Liver Microsomes O--O TA 1538 Without Microsomes

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

O.I

Benzene :Hexane:lsopropanol Extract

Jan

Feb Mar Apr

May June July Aug Sept Oct Nov Dec

FIGURE 4. Relative mutagenic activities (computed as indicated in the examples shown in Fig. 3) of different Chicago air particulate extracts collected on different dates in 1975 at the Washington School site.

microsomes, the effect of these inherently active mutagens that are not inactivated by microsomes.2 Finally, it is evident that the material responsible for mutagenic activity in the absence of microsomes that is reduced when the latter are present is largely lost when the benzene fraction is prepared (i.e., in all cases the values in the, presence of microsomes are greater than those 2 If the inherently mutagenic material were not inactivated by microsomes, then the microsome-activatable material could be estimated simply by subtracting the value obtained in the absence of microsomes from the value obtained when microsomes are present. However, there is no way of knowing from the measurement made in the absence of microsomes what part of that value results from material that is stable in the presence of microsomes and what part from material that

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

ENVIRONMENTAL MUTAGENS IN URBAN AIR PARTICULATES

69

obtained when microsomes are absent). It is possible that part of the inherently active material in the original extract passes into the water fraction in the second step of the procedure, since several samples (e.g., March 22 and July 31) exhibit a consistently rising trend with sample concentration, even though at the highest concentrations the value of (E — C) does not reach the statistical criterion of 2.0 X CAv. The consistent rise in activity with increasing sample concentration suggests that water-soluble active material is in fact present, which would become statistically significant if larger samples were analyzed. Although it is premature to relate these observations to the general data regarding meteorologic conditions, it is perhaps worth noting that most of the high concentrations of inherently mutagenic material observed in the original extract occurred when winds were generally from the northeast quadrant (see Fig. 5). The foregoing observations are indicative of the expected complexity of the mutagenic materials that occur in association with urban airborne particulates. We have further analyzed a particularly active sample, that for December 17, in order to test the feasibility of using the Salmonella technique as a means of isolating and identifying the responsible substances. About 56 in2 of the air filter was extracted in benzene: hexane:isopropanol. The extract was dried and taken up in chloroform, and aliquots were subjected to thin-layer chromatography according to the procedures described earlier. The extracts of successive chromatographic zones were then tested on strain TA1538 in the presence of microsomes. When the original extract was fractionated in a benzene:hexane (1:1) solvent system, as shown in Fig. 6, two mutagenically active components with Rf values of 0 and 0.9 were detected. The zones at Rf = 0.9 and 1.0 were then combined, extracted, dried, and rechromatographed with /7-hexane as the solvent system. As shown in Fig. 7, the procedure yielded a major mutagenically active zone with an Rf value of 0.8 and a minor one at the origin. When the former was further chromatographed, using isooctane as the solvent, as shown in Fig. 8, a single mutagenically active zone with an Rf value of 0.7 was obtained. Under UV light this zone exhibited a strong fluorescence typical of certain polycyclic hydrocarbons. When preparations of pure benzo[c7]pyrene and benzo[e]pyrene were chromatographed in the isooctane solvent system, they yielded the same Rf value as the mutagenically active component, 0.7.

is inactivated by microsomes. Hence, so long as the activity observed in the absence of microsomes is significant relative to that observed in the presence of microsomes, simple subtraction cannot be used to compute what part of the latter value results from microsome-activatable material. Thus, without further analysis a value obtained in the presence of microsomes is capable of quantitative interpretation only if there is little or no activity when microsomes are absent, as is the case in most of the samples obtained in the second half of the year.

B. COMMONER ETAL.

70

Benzene :Hexon«:lsoproponol Extrcct Without Microsomes

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

360*

W

90°

270*

180°

FIGURE 5. Relative mutagenic activities of benzene:hexane:isopropanol extracts of air particulate samples collected from the Washington School site on different dates in 1975, as a function of concurrent wind direction (data of Fig. 4).

It was of obvious interest to identify this constituent. The mutagenically active zone at Rf 0.7 in the n-hexane chromatogram was eluted with 10% methanol in chloroform, dried, and taken up in /7-hexane, and the UV absorption spectrum of the resultant solution was determined. The spectrum is shown in Fig. 9, together with the spectra of pure benzo[o]pyrene, pure benzo[e] pyrene, and a 1:4 mixture of the two isomers. The absorption spectrum of the active material resembles that of a mixture of benzo[ff] pyrene and benzo[e] pyrene. Finally, the same purified preparation obtained from the /7-hexane chromatogram was analyzed by means of mass spectrometry together with a standard sample of benzo[o] pyrene (both isomers yield identical spectra in such an analysis). As shown in Fig. 10, the spectrum of the active component exhibits a strong mass peak at 252, which corresponds to the mass of both

ENVIRONMENTAL MUTAGENS IN URBAN AIR PARTICIPATES

71

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

the. [a] and [e] isomers of benzopyrene, as well as fragmentation peaks which, according to a standard atlas, are characteristic of this substance. The presence of additional peaks—for example, at 266 and 270 and at 238 and 248—suggests that a small amount of some other compound is present as well. These results indicate that the active material isolated by successive thin-layer chromatograms is largely a mixture of benzo[o]pyrene and benzo[e]pyrene. Both isomers are mutagenic toward strain TA1538 in the presence of Aroclor-induced microsomes (McCann et al., 1975). Consequently, benzo[ff]pyrene and benzo[e]pyrene can be identified as two of the substances responsible for the mutagenic activity exhibited by the original extract of the air particulate sample.

no

Senile No. 4 5 0 0 6 Solvent System: Benzene:Hexane(l:t) TAI538, Rat Liver Microsomes

100a? 9 0

1 8

80

£ 50 *S o3 40 X)

5 30 20 10 - 1 0

I

2 3 4 5 6 7 8 9

10 II

Chromatographic Zone (cm. from origin) FIGURE 6. Mutagenic activity of material from successive 1-cm zones from thin-layer chromatographic fractionation of benzene:hexane:isopropanol extract of Chicago air particulate sample (Washington School; December 17, 1975). Tested on strain TA1538 with rat liver microsomes.

Solvent System: n-Hexane TAI538; Rat Liver Microsomes

Monies/'Rate

80-

o

1

1

70-

II

60-

5040-

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

dumber

"B

30-

A

I1

20-

10n- 1 0

1

2

3

4

5

6

7

8

9

10 II

Chromatographic Zone (cm. from origin) FIGURE 7. Thin-layer chromatographic fractionation of mutagenic activity of material from zones at Rf= 0.9 and 1.0 of the chromatogram shown in Fig. 6.

90 Solvent System: Isooctane

80.

T A I 5 3 8 ; Rat Liver Microsomes

in

v

s •5

o

o

XI

- I O I

2

3

4

5

6

7

8

9

10 II

Chromatographic Zone (cm. from origin) FIGURE 8. Thin-layer chromatographic fractionation of mutagenic activity of material from zone at Rf= 0.8 of the chromatogram shown in Fig. 7. 72

ENVIRONMENTAL MUTAGENS IN URBAN AIR PARTICULATES

&

Active Component Irom Chicago Air Porticulate

Benzo(a)pyrene Siandord

Mixtun of Bentotalpyrene ond Benzodlpyrene Standards (1:4]

Beniodipyrtne Standard

oo.

230

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

73

240

250

260

270

280

290

300

JO

320

330 340 230 240

250

260

270

280

290

SOO

3(0

320

330 340

Waveiength-nm

FIGURE 9. Ultraviolet absorption spectra of material from zone at Rf~0.1 of the chromatogram shown in Fig. 8, of pure benzo[fl]pyrene and benzo[e]pyrene, and of a 1:4 mixture of the two isomers.

DISCUSSION It was the purpose of this study to determine how a screening technique, based on the Salmonella mutagenesis test system developed by Ames, might be used to obtain data that are applicable to epidemiologic and environmental analysis of the occurrence of organic carcinogens associated with urban air particulates. The observations reported above show that within certain constraints the technique can be used for this purpose. First, it is evident that the technique can readily determine, qualitatively, whether or not a given sample of air particulates, of a readily accessible size, contains a statistically significant amount of extractable mutagenic material. For this purpose, the appropriate procedure is to determine, from a dose-response curve, whether at any sample concentration the mutagenic activity ratio (£ — C)ICAv exceeds the statistical criterion previously established from tests of standard substances. Such determinations must be made separately with microsomes present and absent. Constraints on this type of determination include the following: (1) the determination relates only to substances that are active on the particular strain of Salmonella that is used; and (2) a false negative result may be obtained if the sample contains sufficient toxic or bacteriostatic material to suppress the growth of mutants. Although the latter situation did not occur in the foregoing results, we have observed it in analyses of effluents from chemical plants (Vithayathil et al., 1977). When this effect does occur it is necessary to separate the bacteriostatic material from the rest of the sample by fractionation before determining whether mutagenic

B. COMMONER ETAL.

74

100.0—

80.0—

Active Componeni from Chicago Air Porticulate

60.0—

40.0—

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

20.0—

'

I

Benzo(a)pyrene major peaks

"

100.0—1

80.0—

Benzo(a)pyrene Standard

60.0—

40.0—

20.0—

0.0—'W

ITTTT1T 250 300

r ,,p,,,|.,. | ...,p. r . r .. 1 ... 1 .. 1 .^,.. r ^.. T ^.... r . [ |....j..1...r..1_j....r

50

100

150

200 m/e

FIGURE 10. Mass spectra of material from zone at Rf=Q.l of the chromatogram shown in Fig. 8 and of benzo[o] pyrene.

material is present. Subject to these constraints and to the previously stated limits of the reliability of the test system, the Salmonella technique can readily be used for the rapid, qualitative detection of organic carcinogens in samples of air particulates. In addition to the data reported here, we have previously reported such analyses for air particulates collected in the St. Louis metropolitan area (Commoner, 1976), and Tokiwa et al. (1977) have reported such results from industrial and residential areas of Japan. It is also evident that, subject to certain constraints, quantitative

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

ENVIRONMENTAL MUTAGENS IN URBAN AIR PARTICULATES

75

estimation of the level of mutagenic activity is possible, based on the analytical procedure described above. In this procedure one determines by interpolation from the dose-response curve the lowest sample concentration at which the mutagenic activity ratio that is representative of statistically significant mutagenic activity occurs. The sample's mutagenic activity is expressed, in relative terms, by the reciprocal of this sample concentration. A major constraint on this procedure is that it is not applicable to data obtained in the presence of microsomes, unless it can be shown that the sample's mutagenic activity in the absence of microsomes is small relative to the value obtained when microsomes are present. Where an initial extract of the sample does not conform to this requirement, it would be necessary to introduce a fractionation procedure that separates inherently active mutagens from those requiring microsomal activation before quantitative estimates of the latter are made. The foregoing results also show that a standard fractionation procedure, such as thin-layer chromatography, in which the separation and purification of mutagenic substances is followed by testing extracts of successive chromatographic zones, can be used to isolate and identify mutagenic compounds that occur in air particulates. While we have reported only one example of this procedure above, we have also used it successfully to isolate the active constituents of samples of urban and rural soil and of aqueous effluents from chemical plants. Finally, it should be noted that wherever in these results a statistically significant level of mutagenic activity is observed, it can be concluded with a high degree of reliability that this activity results from the presence of substances that are, in fact, carcinogenic. The relevant question is the proportion of "false positives" observed in standardized comparisons of carcinogens and noncarcinogens in the Salmonella test system. Thus, we have observed that of 50 noncarcinogens (i.e., substances tested on laboratory animals and found not to induce a significant increase in tumor incidence) tested at a series of concentrations against several of Ames' Salmonella strains, using a variety of different microsome preparations, only 1 substance (/3-aminoazotoluene) was shown to be mutagenic. On these grounds, and from similar comparisons made in other laboratories, it is evident that when statistically significant mutagenic activity is detected in a sample of unknown composition, such as air particulates, one can conclude with a reliability of 95% or more that the sample contains one or more carcinogenic substances. We conclude from these considerations that screening techniques based on the mutagenic responses of Ames' strains of Salmonella are suitable means of analyzing the environmental distribution of organic carcinogens associated with urban air particulates. The requisite data can be obtained, initially, by determining dose-response curves from a suitable extract of the sample. Where constraints such as those discussed above occur, it will

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

76

B. COMMONER ETAL.

be necessary to fractionate the original extract before dose-response curves that are interpretable quantitatively can be obtained. Fractionation, for this purpose and for the purpose of isolating and identifying the active constituents, can be readily carried out by applying the Salmonella screening technique to chromatographic procedures. While the reported analyses of participates from the Washington School site in Chicago were carried out as a means of developing the foregoing procedures, certain substantive aspects of the results are worth noting, at least tentatively. This site, which appears to yield the highest concentrations of carcinogens in air particulates from the Chicago area, is located within a heavily industrialized neighborhood. Steel mills, including coke-oven operations, are present. Since these operations are known to produce high concentrations of benzo[o]pyrene and other carcinogens, the high levels of mutagenic activity that we have observed in air particulates, and direct evidence that benzopyrene isomers occur in them, are not surprising. While the data obtained from this site are insufficient to establish firm correlations with wind direction, they do suggest that with more detailed analyses it will be possible to define the origins of the particulate-associated carcinogens. It would appear therefore that screening procedures based on the Salmonella mutagenesis technique can be used to determine how the environmental distribution of the detectable carcinogens may be associated with the local epidemiology of cancer incidence and with the activities of possible sources of the relevant substances.

REFERENCES Ames, B. N., Durston, W. E., Yamasaki, E., and Lee, F. D. 1973. Carcinogens are mutagens: A simple test system combining liver homogenates for activation and bacteria for detection. Proc. Natl. Acad. Sci. U.S.A. 70:2281-2285. Ames, B. N., McCann, J., and Yamasaki, E. 1975. Methods for detecting carcinogens and mutagens with the Salmonella/mammalian—microsome mutagenicity test. Mutat. Res. 31:347-364. Commoner, B. 1976. Carcinogens in the Environment. In Identification and Analysis of Organic Pollutants in Water, ed. L. H. Keith, pp. 4 9 - 7 1 . Ann Arbor, Mich.: Ann Arbor Science. Commoner, B., Henry, J. I., Gold, J. C., Reding, M. J., and Vithayathil, A. J. 1976. Reliability of bacterial mutagenesis techniques to distinguish carcinogenic and noncarcinogenic chemicals. Final report to the U.S. Environmental Protection Agency, EPA-600/1-76-022. Hoffman, D. and Wynder, E. L. 1968. Chemical analysis and carcinogenic bioassays of organic particulate pollutants. In Air Pollution, ed. A. C. Stern, vol. 2, pp. 187-247. New York: Academic Press. Hoover, R. and Fraumeni, J. F., Jr. 1975. Cancer mortality in U.S. counties with chemical industries. Environ. Res. 9:196-207. McCann, J., Choi, E., Yamasaki, E., and Ames, B. N. 1975. Detection of carcinogens as mutagens in the Salmonella/microsome test: Assay of 300 chemicals. Proc. Natl. Acad. Sci. U.S.A. 72:5135-5139.

ENVIRONMENTAL MUTAGENS IN URBAN AIR PARTICULATES

77

Sawicki, E. 1967. Airborne carcinogens and allied compounds. Arch. Environ. Health 1 4 : 4 6 53. Tokiwa, H., Morita, K., Takeyoshi, H., Takahasi, K., and Ohnishi, Y. 1977. Detection of mutagenic activity in particulate air pollutants. Mutat. Res. 48:237-248.

Downloaded by [Aristotle University of Thessaloniki] at 23:22 09 June 2016

Received September 8, 1977 Accepted October 11, 1977