While numerous studies have demonstrated the association between outdoor exposure to atmospheric particulate matter (PM) and adverse health effects, the actual chemical species responsible for PM toxicological properties remain a subject of investigation. We provide here reactive oxygen species (ROS) activity data for PM samples collected at a rural site in the Po Valley, Italy, during the fog season (i.e., November–March). We show that the intrinsic ROS activity of Po Valley PM, which is mainly composed of biomass burning and secondary aerosols, is comparable to that of traffic-related particles in urban areas. The airborne concentration of PM components responsible for the ROS activity decreases in fog conditions, when water-soluble species are scavenged within the droplets. Due to this partitioning effect of fog, the measured ROS activity of fog water was contributed mainly by water-soluble organic carbon (WSOC) and secondary inorganic ions rather than by transition metals. We found that the intrinsic ROS activity of fog droplets is even greater (> 2.5 times) than that of the PM on which droplets are formed, indicating that redox-active compounds are not only scavenged from the particulate phase, but are also produced within the droplets. Therefore, even if fog formation exerts a scavenging effect on PM mass and redox-active compounds, the aqueous-phase formation of reactive secondary organic compounds can eventually enhance ROS activity of PM when fog evaporates. These findings, based on a case study during a field campaign in November 2015, indicate that a significant portion of airborne toxicity in the Po Valley is largely produced by environmental conditions (fog formation and fog processing) and not simply by the emission and transport of pollutants.
There is a rapidly growing body of epidemiological evidence identifying
major health impacts associated with population exposure to airborne
particulate matter (PM), including, but not limited to, respiratory and
cardiovascular diseases as well as neurodegenerative effects
(Pope et al., 2002,
2004; Dockery and Stone, 2007; Davis et al., 2013; Gauderman et al., 2015).
Even if the international air quality standards for atmospheric PM are based
on mass concentrations (PM
In the present study, aerosol chemical and toxicological measurements were
carried out in the Po Valley, Italy, where radiation fog frequently occurs
during the cold season (up to 25 % of the time in fall–winter months in
rural areas, according to recent studies, Giulianelli et
al., 2014). With approximately 20 million inhabitants (
Fog samples were collected from 30 November to 30 December 2015 at the
meteorological station Giorgio Fea in San Pietro Capofiume
(44
Aerosol samples were collected by a PM
Prior to chemical analysis a quarter of each aerosol quartz-fiber filter was
extracted with 10 mL of 18-M
The magnetic sector inductively coupled plasma mass spectrometry (SF-ICP-MS;
Thermo-Ginnigan Element 2) unit was used for the total elemental analysis of
the aerosol and fog samples. Detailed information about the procedure can be
found elsewhere (Zhang et al., 2008; Okuda et
al., 2014). Briefly, the samples are first digested in an automated
microwave-aided (oven-aided in case of fog samples) digestion system
(Milestone ETHOS
The ROS assay, an in vitro exposure assay of PM extracts to rat alveolar
macrophage cells (cell line NR8383) (Landreman et
al., 2008), was used as a measure of PM toxicity of the samples. In this
method, samples (only filters in this step) were first extracted using an
initial sonication period for 15 min with high-purity 10 mL Milli-Q (18 M
In the in vitro exposure and ROS detection step, the membrane-permeable
2
Spearman rank correlation was used to explore the association of individual
components of the aerosol and fog with the ROS activity. We also applied the
principal component analysis (PCA) to the ambient concentrations of the
chemical components of the aerosols in order to identify the source factors
that contribute to PM levels and ROS activity. In this analysis, the
VARIMAX-normalized rotation approach was employed to identify the
uncorrelated source factors (Henry, 1987). Additionally, we also
applied the multiple linear regression (MLR) analysis to identify the source
factors (as represented by the relevant species) that mostly contribute to
ROS activity of aerosols and fog water. In this analysis, several
combinations of species with high loadings in the PCA were regressed against
the ROS levels, and the ones leading to the highest
When fog forms, aerosol particles are selectively scavenged into fog
droplets, with the smaller and more hydrophobic particles left unscavenged
as interstitial aerosol. This process tends to be selective with respect to
specific chemical components (“partitioning”) and represents a useful way
of studying the toxicological properties of externally mixed PM components
under field conditions. Past studies in the Po Valley (Facchini et al.,
1999; Gilardoni et al., 2014) highlighted that fog scavenging efficiency for
different chemical components of atmospheric particles is related to their
water solubility, with higher scavenging efficiencies for water-soluble (WS)
species (e.g., inorganic ions and water-soluble organic carbon, WSOC) and
lower scavenging efficiencies for hydrophobic compounds (e.g., elemental
carbon and several metals) (Gilardoni et al., 2014).
Therefore, during fog episodes, fog droplets are enriched in WS components,
while water-insoluble (WI) species dominantly partition into interstitial
aerosols (Hallberg et al., 1992;
Fuzzi et al., 1988). Results from the present study are in good
agreement with the published literature on the effect of fog scavenging on
the aerosol properties: Figs. S2 and S3 in the Supplement indicate
that organic as well as inorganic WS components (e.g., methanesulfonate (MSA), oxalate, nitrate,
and sulfate) and metals with high WS fractions (including Ca, Na, and Mg)
have the highest scavenging rates, whereas other combustion-related species
with lower solubility (i.e., Fe, Cr, and Zn) are scavenged much less
efficiently. Metals can be also scavenged by fog
(Mancinelli et al., 2005) but not to the same extent as
secondary organic and inorganic species, resulting in higher metal
enrichment of the interstitial aerosol. The results on partitioning of WSOC
between interstitial aerosols and fog water are presented in Fig. 1c, d.
Contrary to the chemical species discussed above, WSOC is a complex mixture
of chemical species originating from a variety of sources. According to
previous studies in the region (Gilardoni et al., 2014, 2016), oxidized
particulate organic compounds include biomass burning products as well as
secondary organic species, which also account for the products of
transformation of biomass burning compounds upon fog processing. As can be
seen in Fig. 1c, the mass fraction of WSOC in PM
Per-volume (
We present here for the first time results on fog scavenging effects on ROS
activity of the aerosol. The average ROS activity of daytime (i.e, without
fog) and nighttime (in fog as well as interstitial) aerosol bulk extracts
and for fog water is reported in Fig. 1a, b. On a per m
Per-volume (
To further explore the difference in the toxicity of daytime and interstitial aerosols, the ROS activity of daytime and nighttime aerosols was also evaluated based on the filtered ROS analysis protocol, representing the toxicity of the WS components (Fig. 2a, b). As already mentioned, volume-based ROS activity of daytime aerosols was 3–5 times higher than that of interstitial aerosols, but this difference becomes even greater in the filtered extracts (a factor of 7, see Fig. 2a). It should be noted that during the daytime, the per PM-mass filtered ROS activity of aerosols (which relates to the redox activity of only the WS fraction) is almost half of the unfiltered ROS levels (representing the redox activity of both WS and WI fractions). However, during nighttime, the per PM mass ROS activity of the WS fraction contributes approximately only 25 % to the overall per-mass ROS aerosol activity (Fig. 2b). These results imply that during daytime, WS and WI PM fractions contribute roughly equally to the ROS activity of the aerosol, whereas during nighttime, the WI components (comprising mostly elements, metals and insoluble carbonaceous material) contribute as much as 75 % of the aerosols ROS activity. Therefore, even if daytime and interstitial aerosols exhibit the same intrinsic toxicity (Fig. 1b), this is driven by different chemical composition in the two aerosol populations. The most critical difference between daytime and interstitial (i.e., unscavenged) aerosols is that the latter are depleted of WS components, including WSOC (Fig. 1) and inorganic aerosols (Figs. S2 and S3 of the Supplement) which are efficiently scavenged by fog. Therefore, as discussed above, the higher ROS activity of daytime aerosols compared to that of nighttime/interstitial aerosols must be attributed to WS components, while, at nighttime, the aerosol ROS activity is mainly driven by WI components. It is also noteworthy that the highest ROS activity was observed in the fog water, which is enriched in WS components compared to the daytime and interstitial aerosols.
Spearman rank correlation coefficients between per-volume
concentrations of water-soluble species as well as metals/elements in the fog
water samples and the corresponding ROS levels. Correlation coefficients
which were statistically significant (at
The asterisk (
Spearman rank correlation coefficients between concentrations of
the metals/elements, markers of organic aerosol, and water-soluble (WS)
components in the aerosol samples and the corresponding ROS levels.
Statistically significant (
The asterisk (
To investigate the main species driving the ROS activity in fog water, we
performed Spearman rank correlation as well as MLR
analysis between per-volume concentrations of all species and per-volume ROS
activity of fog water, the results of which are presented in Tables 1, 2,
and in Table 3. As can be seen in Table 1, most of the WS species were
strongly correlated with the ROS activity of fog. However, this does not
necessarily mean that all of these species, including inorganic ions, are
responsible for the aerosol redox activity. Previous studies have indicated
that some of these associations with toxicologically innocuous species
(e.g., inorganic ions) are observed because of the colinearity of inorganic
ions with important redox active species, including WSOC, and not because of
the toxicity of these components per se (Ntziachristos et al., 2007;
Cho et al., 2005; Verma et al., 2012a). For
instance, Verma et al. (2012a) found strong positive correlations
between the redox activity of quasi-ultrafine PM (PM
Output of multiple linear regression (MLR) analysis using ROS activity as the dependent variable and ambient concentrations of the measured chemical species as independent variables.
The attribution of ROS activity of fog to dissolved organic substances was also confirmed by the results of the MLR analysis (Table 3), demonstrating that WSOC alone explains 98 % of the variability in the fog water ROS levels. We observed much smaller correlation coefficients between the ROS activity of fog water samples and elemental/metallic components, which have been shown to correlate with particle toxicity in earlier studies (Ntziachristos et al., 2007; Cho et al., 2005; Hu et al., 2010; Verma et al., 2012a). This is due to the lower concentration of metals/elements in the fog water compared to daytime and interstitial aerosols, as shown in Fig. S3.
To investigate the major PM chemical species that contribute to the ROS
activity of daytime and interstitial aerosols, we performed a principal
component analysis (PCA) followed by a linear correlation and a multiple linear
regression (MLR) analysis on the data pertaining to aerosols samples. It
should be noted that, due to the limited number of aerosol samples analyzed
for ROS activity (a total of 6), the data for both daytime and nighttime
(i.e., interstitial) aerosols were combined together, and the following
analyses were performed on the pooled data. Table S2 presents the results of
the PCA conducted on the chemical components of daytime and interstitial
aerosols. As can be seen in the table, two source factors were resolved,
together explaining 95 % of the variance in the data. The first source
factor comprises species that most likely are of combustion origin (e.g.,
traffic, power plants), even though some of them (e.g., Fe, Mn, and Cu) may
also come from nonexhaust traffic sources in other areas
(Sanderson et al., 2014). The metals in this group, including
V, Fe, Cu, Mn, and Ni, are considered toxic and known as redox-active metals
(Argyropoulos et al., 2016; Sowlat et al.,
2016; Wang et al., 2016). The second source factor consists of water-soluble
components, including WSOC as well as the ionic components
(methane-sulfonate, oxalate, sulfate, and nitrate) which are tracers of
secondary aerosols (Hu et al., 2010;
Hasheminassab et al., 2013). We should point out that the
PCA grouping of the chemical species reflects both the possible day-to-day
variations in the contributions of the specific sources to PM mass as well
as the diurnal cycles in concentrations governed by the specific fog
scavenging rates (Figs. S2, S3). Table 2 shows the Spearman rank
correlation coefficients between the WS components and elemental/metallic
species of the aerosols and the corresponding ROS levels. Very strong
positive correlations were observed between the ROS activity of aerosol
samples and many of the WS components, which are also known tracers of
secondary organic aerosols (SOA); for example, the Spearman rank correlation
coefficients between the ROS activity and the concentrations of MSA and
oxalate were 0.83 and 0.83, respectively. Similar to the case of fog
water, the observed association of inorganic ions and low-molecular weight
organic acids with the aerosol ROS activity is probably because of the
colinearity of the ionic species and (unspeciated) redox-active organic
compounds. Interestingly, correlation coefficients are greater for the SOA
tracers (oxalate, MSA) than for the bulk WSOC mixture (
WSOC is the main driver of toxicity in fog water and also a likely
contributor of toxicity of PM
This study reports ROS activity data for PM
A direct comparison of the ROS activity of the fog and aerosol samples collected
in this study with those of ambient aerosol previously reported by other studies
is associated with some uncertainties due mainly to the different size range of
the particles collected and the fact that the ROS activity appears to be strongly
particle-size-dependent (Sioutas et al., 2005; Saffari et al., 2014).
Nonetheless these results point to possible health effects associated with
PM exposure during fog episodes in the Po Valley, the toxicity of which are
comparable or, in many cases, higher than that of the highly toxic
traffic-related PM. Our results show that the toxicity of aerosol particles
accumulating in orographic depressions at the midlatitudes during the cold
season, which is normally peak concentration season for PM in many
continental areas, can be further amplified by the formation of fog, whose intrinsic toxicity is even greater than that of the original aerosol
particles scavenged into the droplets. Moreover, the redox potential of fog
solutes is mainly driven by oxidized organic compounds, which also explains
the excess of ROS activity in fog water with respect to the scavenged
fraction of the aerosol. The effect of secondary organic species on ROS
activity in fog and aerosols in the Po Valley is clearly demonstrated by the
fact that fog water exhibits a higher intrinsic toxicity with respect to
PM1, despite its depletion of redox active metals as a consequence of the
systematically different scavenging rates between WS and WI aerosol species.
The contribution of WS secondary species to ROS activity is also supported
by (a) the results of MLR analysis, (b) the scavenging rate of total
redox-active compounds (71 % of ROS activity) which is higher than that
for most combustion-related metals (typically < 60 %),
corroborating the contribution of hygroscopic particles to ROS activity; and
(c) the results obtained from the filtered extracts indicating that ca.
50 % of ROS activity is attributable to WS species in daytime (i.e.,
out-of-fog) conditions. The origin of secondary organic compounds
responsible for the ROS activity of aerosol WSOC in the Po Valley cannot
fully be elucidated based on this set of data. However, previous
observations at the same site in the fog season showed that SOA are produced
in large amounts by aqueous reactions in fog droplets and deliquesced
aerosols starting from organic compounds emitted from biomass burning
(Gilardoni et al., 2016). Production of biomass
burning SOA is accompanied by the formation of redox active organic
compounds, including hydroquinones. We can therefore speculate that the
enhancement of ROS activity in fog water is at least partly irreversible, as
evaporating fog droplets become enriched on newly formed redox-active
secondary species. The relevance of secondary sources of toxicity in fog and
fog-processed aerosols calls for more stringent controls on possible
precursor emissions, which should be pursued by policy makers including
international authorities (as secondary PM components concentrations are
often triggered by transboundary pollution) (Kiesewetter
and Amann, 2014). Finally, the health effects of the exposure to fog water
toxics depends on fog frequency, and in turn on climate conditions and
climate change. Substantial reductions (
The data presented in this article are available from the authors upon request.
The authors declare that they have no conflict of interest.
This research was funded by Regione Emilia Romagna as part of the Supersito project (Decreto Regionale 428/10) and co-funded by the European Union's Seventh Framework Programme (FP7/2007-797 2013) under grant agreement no 603445 (BACCHUS). The authors would also like to acknowledge the financial support from the United States National Institute of Allergy and Infectious Diseases (award number: 5R01AI065617-15) and the National Institutes of Health (grant number: 5R01ES024936-02). Edited by: Yinon Rudich Reviewed by: Jeffrey Collett and two anonymous referees