Bacterial Mutagenicity of Urban Air A Worldwide Phenomenon

a. Background

Tokiwa and co-workers reported in 1976 that organic extracts of particles collected in Ohmuta and Fukuoka, Japan, were active in the Ames Salmonella typhimurium reversion assay with microsomal activation (+ S9); that is, they contained promutagens (Tokiwa et al., 1976). This was reasonable since such extracts were well known to contain many carcinogenic/mutagenic PAHs that are promutagens, e.g., BaP. However, shortly thereafter, a major new discovery was reported: organic extracts of ambient aerosols collected in Duisburg, Germany (Dehnen et al., 1977); Ohmuta and Fukuota, Japan (Tokiwa et al., 1977); Berkeley, California; and Buffalo, New York (Talcott and Wei, 1977); and the Los Angeles Air Basin (Pitts et al., 1977) also showed strong, direct mutagenic activity (- S9). Similar results were reported for ambient particles in Chicago, Illinois (Commoner et al., 1978); Kobe, Japan (Teranishi et al., 1978); Stockholm, Sweden (Löfroth, 1981); and Contra Costa County, California (Wesolowski et al., 1981). Furthermore, unknown "direct" chemical mutagens were associated with fine respirable particles having diameters less than 2.5 /¿m (Commoner et al., f979; Pitts et al., f979; Talcott and Harger, 1980; Tokiwa et al., 1980; Löfroth, 1981).

The Ames assay provided significant microbiological "clues" as to the chemical nature of these airborne mutagens. Extracts of ambient air particles at all locations showed direct (- S9) frameshift activity with strain TA98 (or with earlier related, but less sensitive, strains TA1537 and TA1538), but little or no activity with strain TA1535 (base-pair substitution mutagens). They also displayed mutagenicity on strain TA100.

Concurrently, Salmonella assays of extracts of primary combustion-generated particulate emissions from various sources were shown to be mutagenic in Ames frameshift strains TA98 and TAfOO without microsomal activation (- S9). Therefore, fly ash from coal-fired power plants (Chrisp et al., 1978), wood smoke (Löfroth, f978; Ramdahl et al., 1982a), automobile exhaust (Wang et al., 1978), and diesel soot (Huisingh et al., 1979) must also contain direct mutagens. The question became, what were the specific compounds?

A breakthrough came in 1978, when nitroarenes were reported to be direct-acting, frameshift mutagens present both in ambient air particles collected in Prague (Jäger, 1978; 3-nitrofluoranthene and 6-nitro-BaP) and in extracts of particulate POM emissions in auto exhaust (Wang et al., 1978). These discoveries resulted in major research efforts to identify and quantify nitroarenes and other direct-acting mutagens in emissions and ambient POM; by 1983, some 100 mono- and dinitro-PAHs had been reported to be associated with diesel particles (e.g., Paputa-Peck et al., 1983; Schuet-zle and Perez, 1983; MacCrehan et al., 1988).

In 1983, Tokiwa and co-workers reported dinitropy-renes (which are powerful bacterial mutagens; see Tables 10.16 and 10.19) in ambient particles from Santiago, Chile (Tokiwa et al., 1983), and Nielsen (1983) identified and quantified 1-nitropyrene and several other nitro-PAHs in complex POM mixtures. He and his co-workers also identified several nitro-PAHs in ambient particles collected in a rural area of Denmark (Nielsen et al., 1984). In the United States, nitroarenes were identified and quantified in extracts of ambient particles from St. Louis, Missouri (Ramdahl et al., 1982b); Detroit, Michigan (Gibson, 1982); and southern Ontario, Canada (Sweetman et al., 1982). For more detailed treatments and references, see, for example, articles by Alfheim et al. (1983, 1984b) and Rosenkranz (1996) and reviews by Rosenkranz and Mermelstein (1985a, 1985b), Tokiwa and Ohnishi (1986), and van Houdt (1990).

Reports that certain of the mono- and dinitroarenes found in the exhausts from diesel and gasoline engines, and in ambient air, were animal carcinogens and possible human carcinogens (e.g., see Table 10.14 and IARC, 1989) accelerated research in this new area. In subsequent years, the phenomenon of public exposure to both the direct and activatable bacterial mutagens associated with respirable ambient particles was documented for a variety of air environments throughout the world. These included cities in the San Francisco Bay area (Kado et al., 1986; Flessel et al., 1991), ten localities in the Genoa municipality, Italy (De Flora et al., 1989), four locations in Athens, Greece (Viras et al., 1990), a three-site comparative study in Rio de Janeiro, Brazil, Camden, New Jersey, and the Calde-cott Tunnel, California (Miguel et al., 1990), Rome (Crebelli et al., 1991), and a number of locations in Italy (Nardini and Clonfero, 1992; Scarpato et al., 1993; Barale et al., 1994; Pagano et al., 1996), Hamilton, Ontario, Canada (Legzdins et al., 1994), Barcelona City, Spain (Bayona et al., 1994), five stations in metropolitan Mexico City (Villalobos-Pietrini et al., 1995), Copenhagen, Denmark (Nielsen et al., 1998a, 1999a, 1999b), and Sapporo, Japan (Matsumoto et al., 1998).

b. Particle Size Distribution of PAHs and Mutagenicity

An important aspect of inhalable PAHs is their distribution as a function of particle size in ambient aerosols since size is a key parameter in determining aerosol lung deposition efficiencies (see Chapter 2.A.5).

There have been a number of studies of the size distribution of PAHs and of the associated mutagenicity over the past three decades. For example, Pierce and Katz (1975) measured mass concentrations as a function of particle size for aerodynamic diameters >0.5 /Jim for BaP and several other PAHs and two O-PACs in aerosol samples collected at five sites in, and near, Toronto, Canada. The size distribution of the particles was approximately log-normal, and the majority of the mass of the associated PAHs and PACs was found in particles with diameters below 3 /jlm. Miguel and Friedlander (1978) used an eight-stage, low-pressure impactor with 50% cutoffs down to 0.05-/j,m aerodynamic diameter (Hering et al., 1978, 1979) to determine the mass distributions of BaP and coronene in aerosols collected in Pasadena, California. Approximately 75% of the BaP and 85% of the coronene were associated with particles with diameters less than 0.26 /jum and ~50% of the mass fell in the narrow size range from 0.075 to 0.12 ¡xm. Similar studies have been conducted at various locations around the world, including Belgium, The Netherlands, and the United Kingdom (e.g., see Van Vaeck and Van Cauwenberghe, 1978; and Baek et al., 1991).

To better understand the effects of atmospheric processes (reactions, gas-particle partitioning, etc.) on the size distributions of PAHs in ambient aerosols, Venkataraman and Friedlander (1994b) carried out measurements of gases and particles during winter and summer periods at several sites in the Los Angeles Air Basin. Concurrently, they determined the mass distributions of these PAHs in aerosol particles in vehicular emissions using the same instrumentation and analytical procedures (Venkataraman et al., 1994a). As seen in Fig. 10.15a, the distribution of fluoranthene in ambient aerosols collected during the winter is bimodal with a primary emissions mode at 0.05-0.5 /jlm (mode I) and an accumulation mode at 0.5-4.0 ^m (mode II). A significant fraction of the fluoranthene present in the total sample is present in the accumulation mode, II. Three other 4-ring, semivolatile PAHs sampled (pyrene, chrysene, and benz[a]anthracene) had similar mass distributions.

In contrast, during the winter, the nonvolatile 5- and 6-ring PAHs BaP (Fig. 10.15b), benzo[6]fluoranthene, dibenzanthracene, benzo[g/z/]perylene, benzo[/c]fluo-ranthene, and indeno[ct/]pyrene had ~ 63-82% of their masses in the 0.05- to 0.5-/j,m range, primary emission mode I. However, as seen in Fig. 10.15c, the pattern shifts significantly to larger sizes for aerosols sampled in the summer.

In a similar study, Allen and co-workers (1996) determined the particle size distribution for 15 PAHs with molecular weights ranging from f 78 (e.g., phenan-threne) to 300 (coronene) and associated with urban aerosols in Boston, Massachusetts. As for BaP in the winter (Venkataraman and Friedlander, 1994b), PAHs with MW >228 were primarily present in the fine aerosol fraction (Dp < 2 ftm). A study of 6-ring, MW 302 PAH at the same site showed bimodal distributions, with most of the mass in the 0.3- to 1.0-/j,m particle size size range; a smaller fraction was in the "ultrafine" mode particles (0.09-0.14 /jlm) (Allen et al., 1998). For PAHs with MW 178—202, the compounds were approximately evenly distributed between the "fine" and "coarse" (Dp > 2 jtm) fractions. Polycyclic aromatic hydrocarbons in size-segregated aerosols col lected a month later at a rural site were present to a greater degree in the coarse fraction than those collected in urban Boston, consistent with other observations (e.g., Pierce and Katz, 1975; Van Vaeck and Van Cauwenberghe, 1985; Venkataraman and Friedlander, 1994b). These size distributions are consistent with the condensation of the large, nonvolatile PAHs on small particles during cooling of the exhaust. However, the smaller, semivolatile PAHs become distributed between the smaller and larger particles via continuing vaporization and condensation processes in the atmosphere.

Furthermore, a number of studies have shown that the direct mutagenicity of particles is primarily associated with particles having Dp < 2.5 ^m. For example, Viras and co-workers (1990) report that at two sites in Athens, Greece, 81 and 92% of the total direct activities (TA98 — S9) were associated with particles having Dp < 3.3 yu,m. Furthermore, ~60 and 80%, respectively, were in submicron particles, D < 1.0 yu.m.

Similar results have been reported for the particle size distribution of promutagenic activity (TA98 + S9) of ambient particles collected by Pagano and co-workers (1996) near a busy road in Bologna, Italy. Figure 10.16 shows their data for fractions ranging from <0.4 to <3.3 /am for each of seven, week-long sampling events conducted from November 15, 1994, to March 31, 1995. Except for the first sampling period, the highest activity was in the <0.4-/Am fraction.

c. Variables Influencing Mutagenicity Levels

A number of variables influence the overall levels of mutagenicity to which the general public is exposed through inhalation of ambient air particles. (Note, however, that total exposure also includes gas-phase mutagens; see later.) The factors include, for example, (a) the inherent mutagenic potencies (rev /¿g_l of extract) of the fine particles emitted by each type of combustion a. O

0 0

Post a comment