Molecular and Cellular Biochemistry 234/235: 317-326, 2002.

© 2002 Kluwer Academic Publishers.

Soluble metals as well as the insoluble particle

fraction are involved in cellular DNA damage
induced by particulate matter
Ad M. Knaapen, Tingming Shi, Paul J.A. Bonn and Roel P.F. Schins
Particle Research Core, Research Institute for Environmental Health (IUF), Dusseldorf, Germany

Abstract
Exposure to ambient particulate matter has been reported to be associated with increased rates of lung cancer. Previously we
showed that total suspended particulate matter (PM) induces oxidative DNA damage in epithelial lung cells. The aim of the
present study was to further investigate the mechanism of PM-induced DNA damage, in which soluble iron-mediated hydroxyl
radical ("OH) formation is thought to playa crucial role. Using electron spin resonance (ESR) we showed that PM suspensions
as well as their particle-free, water-soluble fractions can generate 'OH in the presence of hydrogen peroxide (HP), an effect
which was abrogated by both deferoxamine and catalase. In addition, PM was also found to induce the 'OH-specific DNA
lesion 8-hydroxydeoxyguanosine (8-0HdG) in the presence ofHP2 as assessed by dot-blot analysis of calfthymus DNA using
an 8-0HdG antibody. In human alveolar epithelial cells (A549), both PM suspensions and the particle-free soluble fraction
elicited formation of DNA strand breaks (comet-assay). Unlike the acellular DNA assays, in epithelial cells the DNA-damag-
ing capacity of the particle suspensions appeared to be stronger than that of their corresponding particle-free filtrates. In con-
clusion, our findings demonstrate that the water-soluble fraction of PM elicits DNA damage via transition metal-dependent
'OH formation, implicating an important role ofHP2' Moreover, our data indicate that direct 'particle' effects contribute to
the genotoxic hazard of ambient particulate matter in lung target cells. (Mol Cell Biochem 234/235: 317-326, 2002)

Key words: DNA strand breaks, epithelial lung cells, Fenton reaction, hydrogen peroxide, hydroxyl radicals, particulate matter

Introduction [e.g. 4-6], whereas others investigated the genotoxic effects

of transition metals because of their specific relation with
Epidemiologic studies have demonstrated a significant asso- respiratory tract carcinogenesis [7].
ciation between exposures to ambient particulate matter (PM) Apart from the genotoxic effects of several individual PM
and the incidence of acute mortality and morbidity due to car- constituents, it is now generally accepted that the ability of
diovascular or pulmonary complications [1,2]. Additionally, PM to generate reactive oxygen species (ROS) plays a piv-
PM has also been associated with increased rates of lung can- otal role in PM-induced adverse health effects [8-11]. For
cer [2, 3]. However, the mechanism of PM-induced carcin- instance, we and others have demonstrated a clear relation
ogenesis is not clear. PM is a complex mixture of aggregates between PM-induced DNA damage and its ability to gener-
of organic and inorganic compounds such as carbonaceous ate ROS [12-14]. Among these ROS, the hydroxyl radical
material, polyaromatic hydrocarbons, metals, salts and ("OH) is of greatest concern, because it is a highly reactive
endotoxins. Studies investigating the possible mechanisms electrophilic species, known for its ability to attack DNA [15].
ofcarcinogenicity induced by PM, have mainly focused on its The most common way for hydroxyl radicals to be generated
genotoxic properties. For instance, several studies concentrated is via the Fenton reaction, which involves the reduction of
on the role of organic extractable mutagens such as PAH's hydrogen peroxide by a transition metal ion [16, 17]. Tran-

Address for offprints: PJ.A. Bonn, Particle Research Core, Research Institute for Environmental Health (IUF), Auf'm Hennekamp 50, 40225 Dusseldorf,
Germany (E-mail: [email protected])
318

sition metals such as Fe and Cu are abundantly present in PM Preparation and characterisation ofPM
(18], and have been shown to be potent inducers of oxida-
tive DNA damage [19-21]. Indeed, previously we showed PM was sampled in Duisburg Germany (1993-1995) on ni-
that induction of the 'OH-specific DNA lesion 8-hydroxy- trocellulose filters (Pore size: 311m, Sartorius; Gottingen,
deoxyguanosine (8-0HdG) by PM or coal fly ash was clearly Germany) using a high volume sampler (HVS 150, Strohlein
related to the ability of these particles to generate 'OH via an Instruments, Germany). PM suspensions were prepared by
iron-dependent Fenton reaction (12, 22]. cutting 45 cm2 ofthe fi Iter, immersed in deionized water (5 ml),
However, studies investigating genotoxic effects of water- shaking (5 min) and subsequently sonication in a waterbath
soluble factors, such as transition metals, fail to evaluate PM (Sonorex TK52; 60 Watt, 35 kHz, 5 min). Because the ob-
as an entirety and thus fail to evaluate possible effects of the tained particle suspension was partly contaminated by small
insoluble particulate fraction. PM is known to be comprised filter fibres (5-10% on mass basis) a blank filter was treated
mostly, in terms of particle numbers, of ultrafine particles the same way and was used as a control in all experiments.
« I00 om), and it is this ultrafine fraction which is reported The final particle concentration of both PM and blank filter
to playa major role in the observed association between PM preparations was estimated by comparative turbidometry at
exposure and adverse health effects [23,24]. Although poorly 405 om using carbon black as a standard (Huber 990, 260 urn).
soluble particles such as quartz and carbon black have been Although part of the PM is water-soluble this method was
shown to cause DNA damage [25-27], the specific role of used to prepare suspensions with equal particle concentra-
the insoluble particle fraction in the genotoxicity ofPM is still tions for all experiments. Previous comparative experiments
poorly described. Therefore, the aim ofthe present study was using gravimetric analysis of PM before and after filter ex-
to expand our earlier work and to investigate the mechanism traction confirmed the correlation between turbidometry and
of PM-induced oxidative DNA damage more closely. More gravimetric analysis. To obtain supernatants, the PM suspen-
specifically, we hypothesize that the insoluble particle core sions were centrifuged for 5 minutes at 13,000 g. Because
contributes to a previously described mechanism of PM-in- our previous work demonstrated that the supernatant still
duced DNA damage in which soluble iron-dependent 'OH contained ultrafine particles [12], in some experiments PM
formation is thought to playa central role (12]. Therefore, suspensions were also filtered through a 0.1 11m filter (Acrodisc
we investigated the ability of both the water-soluble and in- 25 mm syringe filter, Pall Gelman Laboratory, Ann Arbor,
soluble fractions ofPM to induce 'OH and whether these were USA). No ultrafine particles were observed in these filtrates
able to induce the 'OH-specific DNA lesion 8-0HdG in na- (data not shown). Particles were visualised using Scanning!
ked DNA upon direct exposure to PM. To investigate the im- Transmission Electron Microscopy (STEM, Philips CMI2).
plications of these observations for cellular DNA, human Energy dispersive X-ray analysis (EDX) of the elemental
lung epithelial cells were exposed to PM and DNA-strand content of the large (centrifugeable) particle agglomerates
breakage was analysed on a single cell level using the comet present in the suspension showed an elemental profile which
assay. was characterized by S, Ca, Cr, Zn, and Fe. The predominant
element in the sample was Fe. The soluble fraction ofPM was
analysed for the presence of soluble metals using atomic
absorption spectrometry (AAS) and Fe (25311g!g), V (l4.711g!
Materials and methods g) and Ni (76.0 llg/g insoluble PM) were detected. Although
total suspended particulate matter, used in the present study
Chemicals contains non-respirable particles, approximately 70% ofPM
is reported to consist of respirable particles [28]. Further
Albumin, deferoxamine mesylate (DFO), diaminobenzidine- details have been previously described [12].
tetrahydrochloride (DAB), 5,5-dimethyl-l-pyrroline-n-oxide
(DMPO), Hanks' balanced salt solution (HBSS), ethidium
bromide, Dulbecco's modified eagle's medium (DMEM), Measurement ofhydroxyl radicals using electron spin
Hepes buffer, fetal calf serum, horse radish peroxidase, phe- resonance
nol red, phosphate buffered saline (PBS), phorbol myristate
acetate (PMA), and trypsin/EDTA, were obtained from Sigma Generation of 'OH by PM-preparations was investigated in
(St. Louis, MO, USA). Hydrogen peroxide (HP2) was pur- the presence ofhydrogen peroxide and the spin trap 5,5-dime-
chased from Fluka (Seelze, Germany). Catalase was obtained thyl-l-pyrroline-N-oxide (DMPO) [29]. Prior to use, the spin
from Roche (Mannheim, Germany). Calf thymus DNA and trap was purified with active coal. DMPO was dissolved in
RNAse were obtained from Boehringer (Mannheim, Ger- deionized water and active coal was added (30 mg/ml). The
many). For electron spin resonance experiments deionized suspension was shaken continuously, incubated for 20 min.
water was used (Millipore). at 35°C, and subsequently centrifuged (2000 g, 10 min). This