21
Mutation Research, 261 (1991) 21-28 © 1991 Elsevier Science Publishers B.V. 0165-1218/91/$03.50 ADONIS 016512189100127Z MUTGEN 01687
Indoor sources of mutagenic aerosol particulate matter: smoking, cooking and incense burning G6ran L6froth, Charlotta Stensman and Margareta Brandhorst-Satzkorn Nordic School of Public Health, S-402 42 Gothenburg (Sweden) (Received 10 October 1990) (Accepted 27 February 1991)
Keywords: Ames Salmonella test; Particulate matter, indoors; Cigarette smoking; Cooking; Incense burning
Summary The emission of aerosol particles and their mutagenic activity as well as the emission of some gaseous pollutants has been studied experimentally in order to compare the emission from some indoor pyrolysis processes. Cigarette (tobacco and herbal) smoking, incense and mosquito-coil burning and frying of experimental lean minced pork emitted particulate matter. Their extracts were mutagenic in the Ames Salmonella test with TA98 and activation as well as, with a higher response, in a microsuspension test with the same strain and activation condition. The response of the particles from the smoking and burning processes varied from 3000 to 50,0000 revertants per gram of smoked or burnt material in the conventional Salmonella test and from 50,000 to 350,000 revertants per gram in the microsuspension assay. The frying of lean minced pork gave an airborne emission of about 53 and 560 revertants per gram Of fried pork, respectively, in the 2 assays. The frying of some common food items following cookbook recipes also emitted mutagenic aerosol particles but the emitted activity was less than that in the pork experiment. Carbon monoxide, isoprene and benzene were present in the emissions from the smoking and burning processes but were not detectable in the frying fumes. The results suggest that incense and mosquito-coil burning can cause indoor air pollution akin to that from cigarette smoking. Indoor air pollution from cooking requires further study.
Indoor air pollution can be caused by several phenomena such as infiltration of outdoor air, temporary indoor processes and slow release of components from construction, household and consumer products. The presence of aerosol particles is, however, to a large extent caused by 2 routes, infiltration of outdoor air and indoor py-
Correspondence: G. L6froth, Nordic School of Public Health, Box 12133, S-402 42 Gothenburg (Sweden).
rolytic processes causing the formation of an aerosol. The present study had the objective of determining the mutagenic activity in 2 Salmonella assays of experimentally produced aerosol particulate matter from some common and uncommon indoor activities. An additional aim was to compare the chemical nature of the mutagenic components by simple chemical fractionation. In addition, it was feasible to determine the emission of carbon monoxide, isoprene and benzene. The
22 common processes studied were ing and frying of proteinaceous the uncommon processes were incense and mosquito coils and herbal cigarettes.
cigarette smokfoods, whereas the burning of the smoking of
Materials and methods
The following processes were investigated: smoking of a common Swedish tobacco cigarette brand, 'Red Prince' (Swedish Tobacco Co., Stockholm, Sweden) and a 'non-tobacco' cigarette brand 'Ginseng herbal cigarettes' (Honeyrose Products Ltd, Stowmarket, Suffolk, U.K.); burning of incense sticks 'Fragrant and Fantastic' (Factory for the Manufacture of Incense, Guangnei Commune, Beijing, China) and incense cones 'Oriental Garden Cone Incense' (Shoyeido & Co., Kyoto, Japan) and insecticidal mosquito coils 'US622' (Kapi Ltd., Nakuru, Kenya). Frying of experimental lean minced pork in a thermostated electric frying pan set at its highest rated temperature (225°C) was also investigated; this experimental pork was of the same type as that used by Berg et al. (1988). Frying of some common Swedish foods in the same pan following cookbook recipes was also studied. Aerosol particles were collected on preweighed glass fiber filters during smoking, burning and frying in a 180-1 plastic box. The rate of collection was about 75 1/min. The determination of carbon monoxide, isoprene and benzene was performed in separate experiments in the box as well as in a 19-m 3 room without collection of particulate matter. Cigarettes were machine-smoked as described by Ling et al. (1987) and the mainstream smoke was vented outside the box. The tobacco cigarettes weighed about 1 g each, including filter, of which about 0.7 g tobacco was burnt. The non-filter herbal cigarettes weighed 0.8-0.9 g, of which 0.5-0.6 was burnt. Pre-weighed amounts of incense materials and insecticidal coils were burnt and any remaining unburnt material was recovered and weighed. The experimental lean pork was minced and then fried, 3-4 patties, weighing 60-80 g, at a time with 6 min cooking on one side and then 4 min on the reverse side in the same manner as
described by Berg et al. (1988). For comparison, commercial pork was obtained, minced and fried in the same manner. The common Swedish food items were: pork chops, Falu sausage (beef, pork, potato flour; 23% fat), Swedish hamburger (minced beef and pork, egg, milk, breadcrumbs, onions and spices), blackpudding (pig and beef blood, beef tallow, rye flour, sugar, spices; 11% fat), Baltic herring (a fish resembling smelt) and pancakes (wheat flour, milk, egg and salt). Some additional information on cooking parameters is listed in Table 3 under Results. With the exception of Swedish hamburgers and blackpudding which were studied in replicate experiments, each food item was analyzed only once with respect to airborne mutagenicity. Temperature measurements during frying were performed with a thermoelement (type T, Pentronic, Sweden) which was placed between the pan bottom and the food. For practical reasons, the measurements were made in experiments without the collection of samples. The aerosol particulate filter samples were extracted by sonication in acetone, the extracts were converted to samples suitable for mutagenicity assays, and the samples were assayed for mutagenicity in the Ames Salmonella plate incorporation assay and a corresponding microsuspension assay as described for environmental tobacco smoke and sidestream cigarette smoke (L6froth et al., 1988). $9 was obtained from livers of Aroclor 1254-induced Sprague-Dawley rats. Its protein content was 30 mg/ml as determined by the method of Lowry et al. (1951). The amount of $9 was 50 /zl/plate in the plate incorporation assay and 10 /xl/tube (plate) in the microsuspension assay. The chemical fractionation into basic and non-basic components was performed by liquidliquid extraction with diethyl ether and sulfuric acid and sodium hydroxide aqueous solutions as described earlier (L6froth, 1981). The extraction of fried food surfaces was done with the acetone method as described by Felton et al. (1981) and the combined acid, neutral and basic fractions were used for mutagenicity testing after dissolution in dimethyl sulfoxide. All samples were tested for mutagenicity at several doses in at least 2 independent tests with
23
graph (Photovac Inc., Canada) as described elsewhere (L6froth, 1989).
at least 2 plates per dose in each test. The response has been calculated from the linear or approximately linear dose-response curve and is here reported either as revertants per gram of burnt or fried material or as revertants per milligram of aerosol particulate matter. For samples not showing a detectable mutagenic response it was assumed that the effect was less than 50 revertants, i.e., about a doubling of the spontaneous rate, at the highest non-toxic dose. Toxicity was evaluated by checking the background lawn microscopically at a magnification of about 200. For 11 sample sets tested in the plate incorporation assay, the average spontaneous and positive control plate counts and their ranges were in the absence of $9, 25 (19-31) for spontaneous and 258 (207-308) for 25 /~g quercetin; in the presence of $9, 43 (37-46) for spontaneous and 313 (252-360) for 2.5 /zg benzo[a]pyrene. The corresponding figures for 19 sample sets tested in the microsuspension assay were: in the absence of $9, 23 (16-29) for spontaneous and 300 (222-458) for 5 /zg quercetin; in the presence of $9, 49 (37-63) for spontaneous and 594 (370-793) for 0.5/zg benzo[a]pyrene. Carbon monoxide was determined with General Electric 15ECS3CO3 Carbon Monoxide Detectors (Wilmington, NH, U.S.A.) calibrated at 0 and 42 + 2 ppm. Isoprene and benzene were determined with a Photovac 10S50 gas chromato-
Results
The microsuspension mutagenicity assay, originally introduced by Kado et al. (1983), gave higher responses than the traditional plate incorporation test (Tables 1 and 2). This has been found in other studies as well (DeMarini et al., 1989; Agurell et al., 1991; Agurell and Stensman, 1991; Bagley et al., 1991). The higher sensitivity is of practical importance for samples which are only available in small amounts, such as particulate matter collected with personal samplers (I~froth et al., 1988).
Smoking and incense burning All the combustion processes resulted in aerosol particles, the extracts of which were mutagenic. The overall results are given in Table 1. All samples gave an enhanced response in the presence of $9 activation and gave a higher response with the microsuspension method than with the regular plate incorporation assay. The tobacco cigarette result was in agreement with our previous results (I_25froth et al., 1988) as was the result with the mosquito coil (Lazaridis and L6froth, 1987). An exception was that the presently obtained aerosol emission is higher. This
TABLE 1 T H E E M I S S I O N O F A E R O S O L PARTICLES AND M U T A G E N I C I T Y F R O M S M O K I N G A N D B U R N I N G PROCESSES Material
Prince cigarettes Herbal cigarettes Incense sticks Incense cones Mosquito coil
Particles ( m g / g smoked or burnt material)
37 " 24 30 51 51 54 51 61
Mutagenic response R e v e r t a n t s / g smoked or burnt material _+S E a Plate incorporation
Microsuspension
TA98 - $9
TA98 + $9
TA98 - $9
TA98 + $9
5 400 _+300 < 1200 b NT c 3 300_+ 300 NT < 2 200 < 1000 3 900 _+300
531005-5800 38300_+2100 NT 13300_+1200 NT 5 200 _+ 600 3200_+ 300 35300_+2600
129000+6000 27600_+1800 27300 _+1300 8000_+2200 14700_+2800 8400 _+1800 10000_+1200 28100_+3500
349000+15000 216000_+17000 245 000 _+18 000 107000_+ 8000 139000_+ 6000 45 600 _+ 4 800 47600_+ 2000 157000_+ 8000
a Standard error of the slope. b < is estimated as < 50 revertants at the highest non-toxic dose. c NT, not tested.
24
may be due to the fact that the earlier sample collection was performed in a room, with the possibility that particles were lost to room surfaces. The mutagenic response per particle mass, 580 rev./mg, which can be calculated for the plate incorporation assay with TA98 + $9 from the data given in Table 1, was very similar to that obtained earlier with the same brand of coils.
250 O
o Od
D
200 150
c~ 100 LxJ
~ LxJ
5O
r-v"
z
~
Mutation Research, 261 (1991) 21-28 © 1991 Elsevier Science Publishers B.V. 0165-1218/91/$03.50 ADONIS 016512189100127Z MUTGEN 01687
Indoor sources of mutagenic aerosol particulate matter: smoking, cooking and incense burning G6ran L6froth, Charlotta Stensman and Margareta Brandhorst-Satzkorn Nordic School of Public Health, S-402 42 Gothenburg (Sweden) (Received 10 October 1990) (Accepted 27 February 1991)
Keywords: Ames Salmonella test; Particulate matter, indoors; Cigarette smoking; Cooking; Incense burning
Summary The emission of aerosol particles and their mutagenic activity as well as the emission of some gaseous pollutants has been studied experimentally in order to compare the emission from some indoor pyrolysis processes. Cigarette (tobacco and herbal) smoking, incense and mosquito-coil burning and frying of experimental lean minced pork emitted particulate matter. Their extracts were mutagenic in the Ames Salmonella test with TA98 and activation as well as, with a higher response, in a microsuspension test with the same strain and activation condition. The response of the particles from the smoking and burning processes varied from 3000 to 50,0000 revertants per gram of smoked or burnt material in the conventional Salmonella test and from 50,000 to 350,000 revertants per gram in the microsuspension assay. The frying of lean minced pork gave an airborne emission of about 53 and 560 revertants per gram Of fried pork, respectively, in the 2 assays. The frying of some common food items following cookbook recipes also emitted mutagenic aerosol particles but the emitted activity was less than that in the pork experiment. Carbon monoxide, isoprene and benzene were present in the emissions from the smoking and burning processes but were not detectable in the frying fumes. The results suggest that incense and mosquito-coil burning can cause indoor air pollution akin to that from cigarette smoking. Indoor air pollution from cooking requires further study.
Indoor air pollution can be caused by several phenomena such as infiltration of outdoor air, temporary indoor processes and slow release of components from construction, household and consumer products. The presence of aerosol particles is, however, to a large extent caused by 2 routes, infiltration of outdoor air and indoor py-
Correspondence: G. L6froth, Nordic School of Public Health, Box 12133, S-402 42 Gothenburg (Sweden).
rolytic processes causing the formation of an aerosol. The present study had the objective of determining the mutagenic activity in 2 Salmonella assays of experimentally produced aerosol particulate matter from some common and uncommon indoor activities. An additional aim was to compare the chemical nature of the mutagenic components by simple chemical fractionation. In addition, it was feasible to determine the emission of carbon monoxide, isoprene and benzene. The
22 common processes studied were ing and frying of proteinaceous the uncommon processes were incense and mosquito coils and herbal cigarettes.
cigarette smokfoods, whereas the burning of the smoking of
Materials and methods
The following processes were investigated: smoking of a common Swedish tobacco cigarette brand, 'Red Prince' (Swedish Tobacco Co., Stockholm, Sweden) and a 'non-tobacco' cigarette brand 'Ginseng herbal cigarettes' (Honeyrose Products Ltd, Stowmarket, Suffolk, U.K.); burning of incense sticks 'Fragrant and Fantastic' (Factory for the Manufacture of Incense, Guangnei Commune, Beijing, China) and incense cones 'Oriental Garden Cone Incense' (Shoyeido & Co., Kyoto, Japan) and insecticidal mosquito coils 'US622' (Kapi Ltd., Nakuru, Kenya). Frying of experimental lean minced pork in a thermostated electric frying pan set at its highest rated temperature (225°C) was also investigated; this experimental pork was of the same type as that used by Berg et al. (1988). Frying of some common Swedish foods in the same pan following cookbook recipes was also studied. Aerosol particles were collected on preweighed glass fiber filters during smoking, burning and frying in a 180-1 plastic box. The rate of collection was about 75 1/min. The determination of carbon monoxide, isoprene and benzene was performed in separate experiments in the box as well as in a 19-m 3 room without collection of particulate matter. Cigarettes were machine-smoked as described by Ling et al. (1987) and the mainstream smoke was vented outside the box. The tobacco cigarettes weighed about 1 g each, including filter, of which about 0.7 g tobacco was burnt. The non-filter herbal cigarettes weighed 0.8-0.9 g, of which 0.5-0.6 was burnt. Pre-weighed amounts of incense materials and insecticidal coils were burnt and any remaining unburnt material was recovered and weighed. The experimental lean pork was minced and then fried, 3-4 patties, weighing 60-80 g, at a time with 6 min cooking on one side and then 4 min on the reverse side in the same manner as
described by Berg et al. (1988). For comparison, commercial pork was obtained, minced and fried in the same manner. The common Swedish food items were: pork chops, Falu sausage (beef, pork, potato flour; 23% fat), Swedish hamburger (minced beef and pork, egg, milk, breadcrumbs, onions and spices), blackpudding (pig and beef blood, beef tallow, rye flour, sugar, spices; 11% fat), Baltic herring (a fish resembling smelt) and pancakes (wheat flour, milk, egg and salt). Some additional information on cooking parameters is listed in Table 3 under Results. With the exception of Swedish hamburgers and blackpudding which were studied in replicate experiments, each food item was analyzed only once with respect to airborne mutagenicity. Temperature measurements during frying were performed with a thermoelement (type T, Pentronic, Sweden) which was placed between the pan bottom and the food. For practical reasons, the measurements were made in experiments without the collection of samples. The aerosol particulate filter samples were extracted by sonication in acetone, the extracts were converted to samples suitable for mutagenicity assays, and the samples were assayed for mutagenicity in the Ames Salmonella plate incorporation assay and a corresponding microsuspension assay as described for environmental tobacco smoke and sidestream cigarette smoke (L6froth et al., 1988). $9 was obtained from livers of Aroclor 1254-induced Sprague-Dawley rats. Its protein content was 30 mg/ml as determined by the method of Lowry et al. (1951). The amount of $9 was 50 /zl/plate in the plate incorporation assay and 10 /xl/tube (plate) in the microsuspension assay. The chemical fractionation into basic and non-basic components was performed by liquidliquid extraction with diethyl ether and sulfuric acid and sodium hydroxide aqueous solutions as described earlier (L6froth, 1981). The extraction of fried food surfaces was done with the acetone method as described by Felton et al. (1981) and the combined acid, neutral and basic fractions were used for mutagenicity testing after dissolution in dimethyl sulfoxide. All samples were tested for mutagenicity at several doses in at least 2 independent tests with
23
graph (Photovac Inc., Canada) as described elsewhere (L6froth, 1989).
at least 2 plates per dose in each test. The response has been calculated from the linear or approximately linear dose-response curve and is here reported either as revertants per gram of burnt or fried material or as revertants per milligram of aerosol particulate matter. For samples not showing a detectable mutagenic response it was assumed that the effect was less than 50 revertants, i.e., about a doubling of the spontaneous rate, at the highest non-toxic dose. Toxicity was evaluated by checking the background lawn microscopically at a magnification of about 200. For 11 sample sets tested in the plate incorporation assay, the average spontaneous and positive control plate counts and their ranges were in the absence of $9, 25 (19-31) for spontaneous and 258 (207-308) for 25 /~g quercetin; in the presence of $9, 43 (37-46) for spontaneous and 313 (252-360) for 2.5 /zg benzo[a]pyrene. The corresponding figures for 19 sample sets tested in the microsuspension assay were: in the absence of $9, 23 (16-29) for spontaneous and 300 (222-458) for 5 /zg quercetin; in the presence of $9, 49 (37-63) for spontaneous and 594 (370-793) for 0.5/zg benzo[a]pyrene. Carbon monoxide was determined with General Electric 15ECS3CO3 Carbon Monoxide Detectors (Wilmington, NH, U.S.A.) calibrated at 0 and 42 + 2 ppm. Isoprene and benzene were determined with a Photovac 10S50 gas chromato-
Results
The microsuspension mutagenicity assay, originally introduced by Kado et al. (1983), gave higher responses than the traditional plate incorporation test (Tables 1 and 2). This has been found in other studies as well (DeMarini et al., 1989; Agurell et al., 1991; Agurell and Stensman, 1991; Bagley et al., 1991). The higher sensitivity is of practical importance for samples which are only available in small amounts, such as particulate matter collected with personal samplers (I~froth et al., 1988).
Smoking and incense burning All the combustion processes resulted in aerosol particles, the extracts of which were mutagenic. The overall results are given in Table 1. All samples gave an enhanced response in the presence of $9 activation and gave a higher response with the microsuspension method than with the regular plate incorporation assay. The tobacco cigarette result was in agreement with our previous results (I_25froth et al., 1988) as was the result with the mosquito coil (Lazaridis and L6froth, 1987). An exception was that the presently obtained aerosol emission is higher. This
TABLE 1 T H E E M I S S I O N O F A E R O S O L PARTICLES AND M U T A G E N I C I T Y F R O M S M O K I N G A N D B U R N I N G PROCESSES Material
Prince cigarettes Herbal cigarettes Incense sticks Incense cones Mosquito coil
Particles ( m g / g smoked or burnt material)
37 " 24 30 51 51 54 51 61
Mutagenic response R e v e r t a n t s / g smoked or burnt material _+S E a Plate incorporation
Microsuspension
TA98 - $9
TA98 + $9
TA98 - $9
TA98 + $9
5 400 _+300 < 1200 b NT c 3 300_+ 300 NT < 2 200 < 1000 3 900 _+300
531005-5800 38300_+2100 NT 13300_+1200 NT 5 200 _+ 600 3200_+ 300 35300_+2600
129000+6000 27600_+1800 27300 _+1300 8000_+2200 14700_+2800 8400 _+1800 10000_+1200 28100_+3500
349000+15000 216000_+17000 245 000 _+18 000 107000_+ 8000 139000_+ 6000 45 600 _+ 4 800 47600_+ 2000 157000_+ 8000
a Standard error of the slope. b < is estimated as < 50 revertants at the highest non-toxic dose. c NT, not tested.
24
may be due to the fact that the earlier sample collection was performed in a room, with the possibility that particles were lost to room surfaces. The mutagenic response per particle mass, 580 rev./mg, which can be calculated for the plate incorporation assay with TA98 + $9 from the data given in Table 1, was very similar to that obtained earlier with the same brand of coils.
250 O
o Od
D
200 150
c~ 100 LxJ
~ LxJ
5O
r-v"
z
~
