Reduction of Particulate Air Pollution Lowers the Risk of Heritable Mutations in Mice

See allHide authors and affiliations

Science  14 May 2004:
Vol. 304, Issue 5673, pp. 1008-1010
DOI: 10.1126/science.1095815


Urban and industrial air pollution can cause elevated heritable mutation rates in birds and rodents. The relative importance of airborne particulate matter versus gas-phase substances in causing these genetic effects under ambient conditions has been unclear. Here we show that high-efficiency particulate-air (HEPA) filtration of ambient air significantly reduced heritable mutation rates at repetitive DNA loci in laboratory mice housed outdoors near a major highway and two integrated steel mills. These findings implicate exposure to airborne particulate matter as a principal factor contributing to elevated mutation rates in sentinel mice and add to accumulating evidence that air pollution may pose genetic risks to humans and wildlife.

Air pollution has the potential to affect millions of humans worldwide and has been associated with an increased risk of lung cancer (1) and of genetic damage in other tissues (25). To investigate whether air pollution induces heritable DNA mutations, we previously exposed sentinel laboratory mice in situ to ambient air for 10 weeks at two field sites: One was located in an urban-industrial area near two integrated steel mills and a major highway on Hamilton Harbour (Ontario, Canada), and the other was in a rural location 30 km away. Comparison of germline mutation rates at expanded-simple-tandem-repeat (ESTR) DNA loci in mouse pedigrees from each site revealed a 1.5- to 2.0-fold increase in mutation rate at the urban-industrial site, providing evidence that air pollution can cause genetic damage in germ cells, inducing transgenerational effects (6). We could not, however, identify causative agents or potential approaches for reducing the mutation risk in urban and industrial areas.

To address these issues, we housed two new groups of sentinel lab mice concurrently for 10 weeks at our earlier urban-industrial site (6). The first group was exposed to ambient air, whereas the second group was housed inside a chamber equipped with a high-efficiency particulate-air (HEPA) filtration system. HEPA filtration removes at least 99.97% of particles 0.3 μm in diameter (7), and the system we used is rated by the manufacturer to remove up to 99.99% of particles down to 0.1 μm in diameter. In addition, HEPA filtration substantially reduces levels of even smaller ambient particles, down to 0.01 μm (8). Mice inside the HEPA filtration chamber were therefore protected from exposure to all airborne particulate matter, with the exception of the smallest ultrafine particles. Simultaneously, we housed third and fourth groups of mice under identical treatment conditions at our rural location, 30 km away, for comparison. Nine weeks after concluding the exposure, we bred the mice and compared germline mutation rates among groups, using pedigree DNA profiling at ESTR loci (912).

Extensive polymorphism at ESTR loci Ms6-hm and Hm-2 allowed us to determine the parental origin of all mutant bands (tables S1 and S2). The offspring of mice exposed to ambient air at the urban-industrial site inherited ESTR mutations of paternal origin 1.9 to 2.1 times as frequently as the offspring in any of the other three treatment groups (Fig. 1A). Mice exposed to HEPA-filtered air at the urban-industrial site had paternal mutation rates that were 52% lower than those of mice exposed to ambient air at the same location (Fig. 1A). Exposure site and HEPA filtration treatment, as well as the interaction between these two variables, explained a significant proportion of the variance in paternal mutation rates (Table 1). In contrast, maternal mutation rates were not significantly affected by either exposure site or HEPA filtration (Table 1). When males exposed to ambient air at the urban-industrial site were mated to unexposed females, their offspring inherited ESTR mutations of paternal origin 2.8 times as frequently as those of rural males mated to unexposed females (Fig. 1B and table S2). In this case, the main effects of site and sex were significant, but there was no significant interaction between the two variables (table S3).

Fig. 1.

(A) The paternal and maternal per-band mutation rates (+95% confidence interval) measured in the offspring of sentinel mice exposed in situ to ambient or HEPA-filtered ambient air at the urban-industrial (Urb-Ind) and rural field sites. Mutation rates are for ESTR single loci Ms6-hm and Hm-2, pooled, and are based on analysis of 94 to 114 offspring (177 to 220 bands) in 17 to 20 pedigrees from each treatment group. (B) Mutation rates that resulted from mating males exposed to ambient air at the rural and Urb-Ind sites to unexposed females. Mutation rates are for Ms6-hm and Hm-2 pooled from 8 pedigrees (45 to 48 offspring, 79 to 91 bands).

Table 1.

Summary of two-way analysis of variance results for the effect of environmental exposure treatment on per-family, paternal and maternal, single-locus ESTR mutation rates.

Source of variance df Paternal Maternal
F value P value F value P value
Exposure site 1,69 7.22 0.0090 3.68 0.0590
HEPA filtration 1,69 8.03 0.0060 0.07 0.7948
Interaction 1,69 13.79 0.0004 1.60 0.2098

HEPA filtration of ambient air therefore reduced ESTR mutation rates at the urban-industrial site, indicating that airborne agents removed by the HEPA filter were necessary for mutation induction. HEPA filtration does not affect gas-phase substances, so the elevation in germline mutation rate in mice exposed at the urban-industrial site must have been caused by exposure to ambient air that contained particulate matter larger than 0.01 to 0.1 μm. Furthermore, this particulate exposure affected ESTR mutation induction primarily in the paternal germ line (Fig. 1). The timing of spermatogenesis in mice and the 9-week delay in breeding after environmental exposure indicate that particulate air pollution affected premeiotic male germ cells (13). This developmental stage was also sensitive to environmental stresses in previous field (6) and laboratory (1416) studies of ESTR mutations. The borderline (near-significant) effect of exposure site observed for maternal mutation rates (Table 1) is similar to our previous findings (6) and suggests that further studies examining ESTR mutation processes in the maternal germ line are necessary.

We measured levels of airborne particulate matter at both field sites on the same 25 days during mouse exposures. Samples were grouped for analysis based on wind direction and the location (upwind or downwind) of sentinel mice at the urban-industrial site relative to the industrial core area of the city of Hamilton. Samples collected on the same dates at the rural site were similarly grouped for comparison. Mean total suspended particulate (TSP), which consists of the respirable fraction as well as larger particles, was 2 to 10 times as high at the urban-industrial site as at the rural site (Wilcoxon signed-rank, P = 0.036) and was associated with the daily number of hours that mice were downwind of the industrial core area (Kendall's Tau = 0.733, P = 0.039) (Table 2).

Table 2.

Mean (± SD) TSP measured at the rural and urban-industrial sites. Samples collected at the urban-industrial site were grouped into six categories based on the daily number of hours that sentinel mice at this location were downwind of the industrial core area. Samples collected on the same dates at the rural site were similarly grouped for comparison. The number of samples collected and the percentage of days during the 10-week mouse exposure with similar downwind times are shown.

Time downwind (hours) Sample size Mean TSP (μg/m3) % exposure days
Rural site Urban-industrial site
0 5 16.2 ± 8.3 38.9 ± 10.5 13
1 to 3 4 23.5 ± 9.3 47.0 ± 17.1 16
4 to 9 3 32.5 ± 9.5 58.5 ± 29.2 14
10 to 18 4 41.9 ± 19.5 127.2 ± 76.6 20
19 to 23 3 10.8 ± 6.3 109.7 ± 35.4 13
24 6 31.7 ± 13.2 115.3 ± 25.3 24

To examine whether this relationship held true for a group of chemical mutagens commonly associated with airborne particulate, we quantified levels of 26 polycyclic aromatic hydrocarbons (PAHs), including the seven carcinogens identified by the U.S. Environmental Protection Agency (table S4). Total PAH concentrations were elevated at the urban-industrial site over the rural site (Wilcoxon signed-rank, P = 0.036) and were associated with the daily number of hours the sampler was downwind of the industrial core area (Fig. 2) (Kendall's Tau = 1.00, P = 0.005). This relationship was much more pronounced for PAHs than for TSP, with PAH levels 4- to 171-fold as high at the urban-industrial site, depending on wind direction (17). The weighted-average daily PAH exposure was 33-fold as high at the urban-industrial site as at the rural site, at 13.4 ng/m3 and 0.4 ng/m3, respectively.

Fig. 2.

Total PAH concentration measured in TSP samples, grouped into categories based on the daily number of hours that sentinel mice at the urban-industrial site were located downwind of the industrial core area. Samples collected on the same dates at the rural site were similarly grouped for comparison. The number of samples comprising these measurements and the percentage of exposure days with similar downwind exposure are given in Table 1. No error bars are displayed, because each measurement presented is from a number of samples pooled within each category.

How exposure to airborne particulate matter induces genetic changes in the male germ line is unknown. Automobile traffic and integrated steel production generate airborne mutagens, including PAHs (18) and heavy metals, and particles may deliver these chemicals and their metabolites to the bloodstream and ultimately the germ cells through the respiratory system. The relationship between PAH concentration and wind direction that we quantified at the urban-industrial site indicates a source of emissions from the industrial area, and total PAH levels were relatively high (∼5 to 30 times as high as in Toronto, Canada's largest city). In addition, benzo[a]pyrene, a potent mutagen and carcinogen often used for assessing risk from PAH exposure, exceeded the recommended lifetime cancer risk guideline (19) of 0.1 ng/m3 by as much as 27-fold on 81% of the mouse exposure days. If PAHs contribute to mutation induction, then the PAH-removal efficiency of our HEPA unit may be an important consideration. Based on chemical analysis of HEPA filters compared to TSP samples from the same location, HEPA filtration blocked 7.1 to 8.0 ng/m3 (55 to 61%) of total PAHs in ambient air at the urban-industrial site. This is likely to be an underestimate (supporting online text), and HEPA filtration probably removed 55 to 100% of particulate-bound PAHs, substantially reducing exposure for the sentinel mice. PAHs may contribute to germline mutation induction, but we cannot yet make specific conclusions regarding their importance.

Human epidemiological studies have associated air pollution exposure with negative health consequences, including cardiovascular (20), respiratory (21), and developmental impairments (22, 23) and lung cancer (1). Identification of the most dangerous air pollutants and their mode of action in producing specific health effects remains uncertain (24). Our study identifies airborne particulate matter as a contributor to heritable mutation induction in mice; however, a direct link between ESTR mutations and health effects has not yet been established. In addition, although elevated germline mutation rates have been documented in both birds (25, 26) and mice (6) near industrial areas, it is not clear whether our results can be extrapolated to humans. Nonetheless, structural changes in DNA have been detected in human sperm after air pollution exposure (27). Data from mouse studies (15, 16) suggest that a relationship may exist between mutation rates at ESTR loci and those in coding regions of the genome that affect phenotype. To reduce the potential risk of harmful heritable mutations for humans and wildlife, along with a suite of other health problems, we suggest that steps be taken to reduce levels of airborne particulate matter in urban environments.

Supporting Online Material

Materials and Methods

SOM Text

Tables S1 to S4

References and Notes

References and Notes

View Abstract

Navigate This Article