Exposure to Concentrated Ambient PM_2.5 Shortens Lifespan and Induces Inflammation-Associated Signaling and Oxidative Stress in Drosophila

Exposure to ambient PM_2.5 is associated with human premature mortality. However, it has not yet been toxicologically replicated, likely due to the lack of suitable animal models. Drosophila is frequently used in longevity research due to many incomparable merits. The present study aims to validate Drosophila models for PM_2.5 toxicity study through characterizing their biological responses to exposure to concentrated ambient PM_2.5 (CAP). The survivorship curve demonstrated that exposure to CAP markedly reduced lifespan of Drosophila. This antilongevity effect of CAP exposure was observed in both male and female Drosophila, and by comparison, the male was more sensitive [50% survivals: 20 and 48 days, CAP- and filtered air (FA)-exposed males, respectively; 21 and 40 days, CAP- and FA-exposed females, respectively]. Similar to its putative pathogenesis in humans, CAP exposure-induced premature mortality in Drosophila was also coincided with activation of pro-inflammatory signaling pathways including Jak, Jnk, and Nf-κb and increased systemic oxidative stress. Furthermore, like in humans and mammals, exposure to CAP significantly increased whole-body and circulating glucose levels and increased mRNA expression of Ilp2 and Ilp5, indicating that CAP exposure induces dysregulated insulin signaling in Drosophila. Similar to effects on humans, exposure to CAP leads to premature mortality likely through induction of inflammation-associated signaling, oxidative stress, and metabolic abnormality in Drosophila, strongly supporting that it can be a useful model organism for PM_2.5 toxicity study.

drosophila, PM2.5, lifespan, inflammation-associated signaling, oxidative stress.

A number of studies have demonstrated that exposure to ambient particulate matter with a diameter ≤ 2.5 μm (PM_2.5) is associated with premature mortality. It was estimated that global premature deaths attributed to PM_2.5 exposure in 2010 are ∼3.2 million (Lepeule et al., 2012), ranking as the sixth largest overall risk factor for global premature mortality. Studies on humans and mammalian models have yielded some valuable insight into the mechanisms whereby PM_2.5 exposure leads to premature deaths. It has been demonstrated that PM_2.5 exposure induces diverse detrimental cardiopulmonary and metabolic effects, such as pulmonary and systemic inflammation, systemic oxidative stress, endothelial dysfunction, decreased pulmonary function, hypertension, insulin resistance, and et al. Of them, systemic inflammation and its associated oxidative stress are believed to be mechanistically central in PM_2.5 exposure-associated pathophysiology (Brook et al., 2010).

It is noteworthy that mortality is not only the most important measurements of public health but also the most important index of aging. It is thus not surprising that the majority of those identified adverse effects of PM_2.5 exposure are aging-related. Conceptualizing the essence of aging, Lopez-Otin et al. (2013) categorized the hallmarks of aging as genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion, and altered intercellular communication. Consistent with the notion that PM_2.5 exposure may be pro-aging, several of those hallmarks have been linked to exposure to ambient PM_2.5. For example, one recent study demonstrated that exposure to PM_2.5 is associated with decreases in telomere length and mitochondrial DNA content (Pieters et al., 2015). Mendez et al. (2013) showed that chronic exposure to concentrated ambient PM_2.5 (CAP) induces unfolded protein response, a hallmark of loss of proteostasis, in mouse adipose tissues. Notably, in spite of those pieces of strong evidence that exposure to PM_2.5 promotes aging, there is still a paucity of studies investigating PM_2.5 exposure-associated mortality in animal models. This is likely due to the high cost and long duration associated with studies currently using models such as mice, rats and other higher mammals. Therefore, there is a practical need of new PM_2.5 exposure animal models, in particular those suitable for study on mortality.

Drosophila is a well-studied and highly tractable genetic model organism for understanding molecular mechanisms of human diseases. With its incomparable merits such as powerful genetics, highly conserved disease pathways, and very low costs, Drosophila has been recommended as an alternate animal model for studying reproductive toxicity of environmental chemicals (Tiwari et al., 2011), developmental toxicology (Cummings and Kavlock, 2005), neurotoxicology (Peterson et al., 2008), and nanotoxicology (Chifiriuc et al., 2016; Vecchio, 2015). More importantly, compared to mammalian models, Drosophila has a short lifespan, making them particularly suitable for lifespan assessment. Therefore, in the present study, we evaluated Drosophila models of PM_2.5 exposure through assessing effect of CAP exposure on lifespan, glucose homeostasis, and inflammatory response of Drosophila.

MATERIALS AND METHODS

Drosophila stocks

The white eye strain (w¹¹¹⁸) of Drosophila melanogaster was from Bloomington Stock Center, USA. κB-luc transgenic flies were generously provided by Professor Julie Williams, University of Pennsylvania, USA (Kuo et al., 2010). All flies were reared on standard Drosophila Corn-sucrose, yeast medium, which was prepared per the following recipe: filtered H₂O (15, 094 ml), Agar (181 g), Tormel Yeast (412.5 g), Cornmeal (1008 g), Molasses (1400 ml), Tegosept (37.5 g), 95% Ethanol (137.7 ml) and Propionic Acid (67.5 ml). If not specified, all flies were reared at 22 ± 2 °C, and a 12:12 hour light/dark cycle.

CAP exposure

To perform the whole-body ambient inhalational CAP exposures, flies were housed in standard culturing vials secured with gauze, and exposed to CAP as shown in Figure 1A. The monitoring of exposure atmosphere and ambient aerosol were performed as previously described using a versatile aerosol concentration enrichment system that was modified for long-term exposures (Ying et al., 2009), which was in a mobile trailer and located on the campus of University of Maryland, Baltimore, and thus concentrated the ambient PM_2.5 from this campus area. This concentration system was designed to concentrate fine ambient aerosols only. (Sioutas et al., 1995) Briefly, PM_2.5 was collected by a BioSampler at the inlet and the mass (used for calculation of PM_2.5 concentration in the system) was measured daily. It has been previously demonstrated that PM_2.5 was distributed evenly (within 2% difference) throughout the exposure chamber. (Maciejczyk et al., 2005)

FIG. 1

CAP exposure shortens Drosophila lifespan. A, The diagram of exposure to CAP. B and C, Age-matched adult flies were collected as previously described (Linford et al. 2013). Males and females were sorted on the anesthetic pad, and 50 flies of the same gender were placed into individual vials (4 vials/group). FA/CAP exposure was initiated at the age of 3 days and maintained through the whole lifespan. The death of flies were checked and recorded daily. The survivorship curves of male (B) and female (C) were generated using Graphpad Prizm 5. *P < 0.05 versus FA, curve comparison. During the period of exposure, the average concentrations of PM_2.5 were 4 and 80 µg/m³, FA chamber and CAP chamber, respectively.

All exposures were performed from May 2014 to October 2014. All flies, including both filtered air (FA) and CAP groups, were exposed at exactly the same time. FA-exposed flies received an identical protocol with the addition of a high-efficiency particulate air filter (Pall Life Sciences, East Hills, New York) positioned in the inlet valve to remove PM_2.5 but not other ambient pollutants in the filtered air stream, as previously described (Ying et al., 2009). The exposure protocol comprised exposures for 6 h/day, 5 days/week (no exposure took place during weekends). During the exposure, flies were maintained with temperatures of 18–24 °C, and relative humidity of 40–60%.

Sampling and calculation of PM2.5 concentration in the exposure chambers

To calculate exposure mass concentrations of ambient PM_2.5 and CAP in the exposure chambers, samples were collected weekly on Teflon filters [Teflo, 37-mm, 2-μm pore (Pall Life Sciences, Ann Arbor, Michigan)] and weighed before and after sampling in a temperature- and humidity-controlled weighing room using a Mettler Toledo no. 11106057 microbalance (Mettler Toledo, Columbus, Ohio). The calculated mass concentrations of PM_2.5 in the exposure chambers are presented in the respective figure legends.

Luciferase reporter assay

Flies carrying the κB-luc reporter were exposed to FA/CAP, and 15 flies/vials X 6 vials/group were snap-frozen in liquid nitrogen and stored at −80 °C until performing reporter assay. To measure the activity of luciferase, whole body lysates were prepared with lysis buffer provided by the manufacturer of luciferase assay kit (Promega, Madison, Wisconsin). Luciferase activity in lysates was measured with a luminescence counter (Perkin Elmer, Akron, Ohio). Luminescence data (in arbitrary units) were normalized by protein concentration.

Quantitative PCR

Wildtype flies (w¹¹¹⁸) were exposed to FA/CAP and immediately frozen at −80 °C. Total RNA was extracted from groups of 20 whole flies using Trizol reagent (Invitrogen, Carlsbad, California) according to the manufacturer's instructions. 10 μg total RNA were treated with DNase (TURBO DNA-free Kit; Applied Biosystems, Foster City, California) and 2 μg of DNase-treated samples were used to synthesize cDNA with High Capacity cDNA Transcription kit (Applied Biosystems, Foster City, California). Quantitative polymerase chain reaction (qPCR) was performed in duplicate using the lightcycler 480. "No template", cDNA negative controls were included for each gene set in all PCR reactions to detect contamination. The expression level for each gene was calculated using the Δ^Ct method relative to β-actin. The sequences of all primers were previously described (Abolaji et al., 2015; Chabu and Xu, 2014; Meyer et al., 2014) and are listed in Table 2.

Whole-body and hemolymph glucose and trehalose measurements

To measure whole-body glucose and trehalose levels, flies were exposed to FA or CAP for 5 or 15 days. At the end of exposure, flies were anesthetized with CO₂ and weighed in groups of 10 flies, and transferred to chilled microcentrifuge tubes containing buffer (20 mM HEPESKOH pH7.5, 10 mM KCl, 1.5 mM MgCl₂, 1 mM EDTA, 1 mM EGTA, 1 mM DTT, protease inhibitors, 0.5 mM PMSF) and homogenized using a motorized micropestle. After centrifugation, supernatants were used for determination of whole-body glucose and trehalose levels.

To determine hemolymph glucose and trehalose levels, flies were exposed to FA or CAP for 15 days. At the end of exposure, flies were anesthetized with CO₂ and then remove the head with a syringe, and decapitated fly bodies were quickly transferred to hemolymph collection tube (80 flies/tube, 3 tubes/group), and centrifuge for 6 min at 1500 g at 4ºC. (Bai et al., 2012) The hemolymph at the bottom of the tube was then frozen at –80ºC till glucose and trehalose measurements.

Glucose was determined by using the glucose detection kit (Sigma, St. Louis, Missouri) per manufacturer’s instruction. Trehalose was converted into glucose through addition of trehalase (0.2 U/ml, Sigma, St. Louis, Missouri) and 2-h incubation at 37ºC. At the end of incubation, the total glucose in the product was determined by the glucose detection kit (Sigma, St. Louis, Missouri) per manufacturer’s instruction. Trehalose levels were presented as the difference between those 2 measurements.

Statistical analysis

Data are presented as mean ± SEM from at least 3 independent replicates. Statistical significances were evaluated by t-test and 2-way ANOVA analyses using GraphPad Prism Software.

RESULTS

Exposure Characterization

All exposures in the present study were performed in the summer. The average ambient daily PM_2.5 concentration during the exposure period was 9 μg/m³. The average concentrations of PM_2.5 were between 52 and 91 µg/m³ in the CAP chambers and between 3 and 4 µg/m³ in the FA chambers. Since the exposures were performed for 6 h/day, 5 days/week, the 24-h average CAP concentration was between 17 and 24 μg/m³, which were markedly higher than the annual national ambient air quality standard of 12 μg/m³ set by the US Environmental Protection Agency (U.S. EPA, 2012). The elemental composition of concentrated PM_2.5 was presented in Table 1. Although identifying PM_2.5 emission sources is generally beyond the scope of this paper, the relatively high ratio of Na/Al reflects a more contribution by the marine source than the crustal source, which is consistent with the geographic proximity of the study site to the ocean.(Huang et al., 1999) Whereas the relatively high sulfur suggests that the study site was most strongly affected by secondary aerosols, which are likely to include emissions from coal-fired utility boilers located regionally.(Huang et al., 1999) According to previous study (Jones and Harrison, 2005; Keuken et al., 2012; Reisinger et al., 2008), the concentration of elemental carbon was probably higher or at least around the highest concentration of the elements in Table 1.

Table 1

The Elemental Composition of Concentrated Ambient PM_2.5

Element	Mean (ng/m3)	SD
Na	61.15	30.63
Mg	14.98	7.56
Al	38.46	30.73
Si	123.44	84.62
P	1.56	1.09
S	509.19	126.97
K	35.83	17.31
Ca	62.41	41.56
Ti	6.40	2.84
V	0.30	0.43
Cr	0.48	0.25
Mn	2.24	0.84
Fe	89.12	32.56
Ni	1.19	0.48
Cu	2.74	1.01
Zn	21.19	15.41
As	0.30	0.19
Se	0.17	0.18
Br	15.65	6.14
Sr	0.40	0.21
Ag	0.42	0.82
Sn	1.87	1.31
Ba	2.54	2.43
Ce	0.11	0.59
Pr	0.12	0.33
Er	1.94	0.89
Lu	2.89	1.08
W	0.55	0.63
Ir	0.08	0.17
Pt	0.18	0.32
Au	0.33	0.30
Hg	0.40	0.34
Tl	0.07	0.15
Pb	1.25	0.61

Table 2

The Sequences of Primers

Egr-F	5′-GATGGTCTGGATTCCATTGC-3′
Egr-R	5′-TAGTCT-GCGCCAACATCATC-3′
Reaper-F	5′-GAGCAGAAGGAGCAGCAGAT-3′
Reaper-R	5′-GGACTTTCTTCCGGTCTTCG-3′
Hid-F	5′-TCTACGAGTGGGTCAGGATGT-3′
Hid-R	5′-GCGGATACTGGAAGATTTGC-3′
Pvf1-F	5′-GCGCAGCATCATGAAATCAACCG-3′
Pvf1-R	5′-TGCACGCGGGCATATAGTAGTAG-3′
Upd3-F	5′-ACTGGGAGAACACCTGCAAT-3′
Upd3-R	5′-GCCCGTTTGGTTCTGTAGAT-3′
SOCS36E-F	5′-AAAAAGCCAGCAAACCAAAA-3′
SOCS36E-R	5′-AGGTGATGACCCATTGGAAG-3′
Dro-F	5′-CCATCGAGGATCACCTGACT-3′
Dro-R	5′-CTTTAGGCGGGCAGAATG-3′
Drs-F	5′-GTACTTGTTCGCCCTCTTCG-3′
Drs-R	5′-CTTGCACACACGACGACAG-3′
Dpt-F	5′-ACCGCAGTACCCACTCAATC-3′
Dpt-R	5′-CCCAAGTGCTGTCCATATCC-3′
Thor-F	5′-TACACGTCCAGCGGAAAGTT-3′
Thor-R	5′-CCTCCAGGAGTGGTGGAGTA-3′
Duox-F	5′-ACGTGTCCACCCAATCGCACGAG-3′
Duox-R	5′-AAGCGTGGTGGTCCAGTCAGTCG-3′
Sod1-F	5′-GGAGTCGGTGATGTTGACCT-3′
Sod1-R	5′-GTTCGGTGACAACACCAATG-3′
Jafrac1-F	5′-TGGATCAACACGCCAAGGAA-3′
Jafrac1-R	5′-GGATGCCAGTCTCCTCATCG-3′
Keap1-F	5′-CCAACTTCCTCAAGGAGCAG-3′
Keap1-R	5′-CGGCGACAAATATCATCCTT-3′
Catalase-F	5′-ACCAGGGCATCAAGAATCTG-3′
Catalase-R	5′-AACTTCTTGGCCTGCTCGTA-3′
Inr-F	5′-GCACCATTATAACCGGAACC-3′
Inr-R	5′-TTAATTCATCCATGACGTGAGC-3′
Ilp2-F	5′-ATCCCGTGATTCCACCACAAG-3′
Ilp2-R	5′-GCGGTTCCGATATCGAGTTA-3′
Ilp3-F	5′-CAACGCAATGACCAAGAGAA-3′
Ilp3-R	5′-TGAGCATCTGAACCGAACT-3′
Ilp5-F	5′-GCCTTGATGGACATGCTGA-3′
Ilp5-R	5′-AGCTATCCAAATCCGCCA-3′

Exposure to CAP Shortens Lifespan of Drosophila

PM_2.5 exposure has been well documented to increase premature deaths in humans (Beelen et al., 2008; Kloog et al., 2013; Pope et al., 2002). However, there are few studies investigating the anti-longevity effect of PM_2.5 exposure in animal models. Drosophila is an invaluable model organism for research on longevity. To develop Drosophila models for study on the anti-longevity effect of PM_2.5 exposure, we exposed adult flies to FA/CAP and recorded their lifespans. The resultant survivorship curve revealed that the survival among the flies exposed to FA and CAP was approximately the same during the first 15 days, and after that, exposure to CAP markedly reduced Drosophila survival (Figure 1B and C). Although exposure to CAP reduced lifespan of both male and female Drosophila, the male appeared to be affected more. The 50% survivals of CAP-exposed animals were 20 days for the male Drosophila (versus 48 days of FA-exposed control) and 21 days for the female Drosophila (versus 40 days of FA-exposed control).

Exposure to CAP Activates Pro-Inflammatory Signaling Pathways in Drosophila

Inflammatory response is believed to be central in the pathogenesis of health effects induced by exposure to ambient PM_2.5 (Terzano et al., 2010). Drosophila has many inflammatory components that are also found in mammals. Toll/Nf-κb, Tnf/Jnk, and Upd3/Jak/Stat are 3 signaling pathways central in regulation of inflammation-like response in Drosophila. Their counterparts in mammals, namely Tlrs, Tnf, and Il-6 pathways, are crucial in mammalian inflammatory responses. To determine whether exposure to CAP activates those inflammatory pathways, age-matched adult male flies were exposed to FA/CAP, and subjected to assessments of target gene expression by real-time RT-PCR. Figures 2A–D reveals that CAP exposure time-dependently increased expression of Jnk target genes Egr, hid, reaper, and Pvf. Jak activator upd3, Jak target gene SCOS36E, and NF-κB target genes Drosomycin (Drs) and Diptericin (Dpt) were also transiently increased by CAP exposure. To confirm the activation of pro-inflammatory pathways by CAP exposure, κB-luc transgenic flies were exposed to FA/CAP, and the activity of reporter gene luciferase was assessed. Figure 2J shows that whole-body luciferase activity in CAP-exposed flies was significantly increased, strongly supporting that similar to its effects in humans, CAP exposure may activate pro-inflammatory pathways in Drosophila.

FIG. 2

CAP exposure activates Drosophila pro-inflammatory pathways. Age-matched adult flies (3 days old) were exposed to FA/CAP for 5 (the average concentrations of PM_2.5 were 3 and 52 µg/m³, FA chamber and CAP chamber, respectively), 10 (the average concentrations of PM_2.5 were 3 and 54 µg/m³, FA chamber and CAP chamber, respectively), or 15 (the average concentrations of PM_2.5 were 3 and 73 µg/m³, FA chamber and CAP chamber, respectively) consecutive days, and 30 flies/vial were collected and snap-frozen. The mRNA expression levels of indicated genes were then assessed by real-time RT-PCR. n = 6 (FA) or 7 (CAP). *P < 0.05 versus FA, two-way ANOVA.

Exposure to CAP Induces Systemic Oxidative Stress in Drosophila

Oxidative stress is generally associated with inflammation and also widely believed to be a critical pathophysiological mediator for PM_2.5 exposure-induced health effects (Wang et al., 2015; Weichenthal et al., 2013; Spagnolo et al., 2015). To determine whether exposure to CAP induces oxidative stress in Drosophila, we assessed the expression levels of genes responsive to oxidative stress. Figures 3A–F reveal that CAP exposure significantly increased the expression of Thor, Duox, Sod1 and Catalase, suggesting that CAP exposure may induce oxidative stress. 2′,7′-Dichlorofluorescein (DCFH) oxidation by fly whole-body lysate is a validated index of systemic oxidative stress in Drosophila (Perez-Severiano et al., 2004). To confirm that CAP exposure induces systemic oxidative stress, we prepared fly whole-body lysates from FA- or CAP-exposed drosophila, and assessed DCFH oxidation by those lysates. Figures 3G and H show that CAP exposure significantly increased DCFH oxidation, supporting that exposure to CAP also may induce systemic oxidative stress.

FIG. 3

CAP exposure induces oxidative stress. A-F, age-matched adult flies (3 days old) were exposed to FA/CAP as described in Figure 2. The mRNA expression levels of indicated genes were then assessed by real-time RT-PCR. n = 6 (FA) or 7 (CAP). *P < 0.05 versus FA, 1-way ANOVA. G and H, flies were exposed to FA/CAP for the indicated time and then collected for preparation of tissue lysates. 2′,7′-Dichlorofluorescein (DCFH) oxidation by lysates was then measured as an index of oxidative stress according to the method of Perez-Severiano et al (Abolaji et al. 2015b). The mean fluorescent signals (G) and the calculated oxidation rates (H) are presented. n = 4/group. *P < 0.05 versus FA, 2-way ANOVA.

Exposure to CAP Induces Dysregulated Insulin Signaling in Drosophila

Exposure to ambient PM_2.5 has been linked to adverse cardiometabolic effects in humans, and the latter are believed to contribute to ambient PM_2.5 exposure-induced pre-mature deaths. Drosophila has been established to be a model organism for study on glucose homeostasis. To determine whether CAP exposure induces abnormalities of Drosophila in glucose homeostasis, they were exposed to FA/CAP, and whole-body glucose and trehalose contents were assessed. Figures 4Aand B reveal that 15-day exposure to CAP significantly increased both glucose and trehalose contents. This is paralleled by increased circulating trehalose (Figure 4D) and increased mRNA expression of Ilp2 and Ilp5 (Figs. 4E–H), strongly suggesting that similar to its effects on humans, CAP exposure may induce insulin resistance in Drosophila.

FIG. 4

CAP exposure induces dysregulated insulin signaling. A and B, adult male flies were exposed to FA/CAP for the indicated days, and whole-body glucose (A) and trehalose (B) levels were determined and presented. C and D, adult male flies were exposed to FA/CAP for 15 days (the average concentrations of PM_2.5 were 4 and 91 µg/m³, FA chamber and CAP chamber, respectively), and hemolymph were collected. Glucose (C) and trehalose (D) levels in hemolymph were then determined and presented. E–H, age-matched adult flies (3 days old) were exposed to FA/CAP for the indicated days. The mRNA expression levels of indicated genes were then assessed by real-time RT-PCR. n = 6 (FA) or 7 (CAP). *P < 0.05 versus FA, 2-way ANOVA or student’s t-test.

DISCUSSION

The present study characterized biological responses of Drosophila to CAP exposure. The main findings include that exposure to CAP shortens lifespan, induces inflammation-associated signaling and oxidative stress, and causes dysregulated insulin signaling in Drosophila. To our best knowledge, this is the first study directly demonstrating that exposure to CAP leads to premature deaths in animal models. Mortality is the most important measurement of public health, and a number of epidemiological studies have shown the association between exposure to ambient PM_2.5 and increased premature mortality in humans (Pope et al., 2002; Beelen et al., 2008; Kloog et al., 2013). However, most likely due to the high cost and long duration associated with studies currently using models such as mice, rats and other higher mammals, there is still a paucity of studies investigating the association between PM_2.5 exposure and premature death in animal models. Suitable animal models have historically played an indispensable role in establishment of causal relationships. As such, our demonstration of CAP exposure-induced premature deaths in Drosophila not only replicated PM_2.5 exposure-induced premature deaths in humans, but also may provide a valuable tool for delineation of underlying mechanisms.

Mortality is not only the most important measurement of public health but also the most important index of aging. Therefore, our demonstration of CAP exposure-induced premature death in Drosophila also strongly supports the notion that PM_2.5 exposure is pro-aging. This is consistent with previous studies showing that exposure to PM_2.5 is associated with changes in the core axis of aging in humans including decreases in telomere length and mitochondrial DNA content (Pieters et al., 2015). It is notable that the majority of PM_2.5 exposure-induced adverse health effects such as cardiopulmonary dysfunction, insulin resistance, and pulmonary cancer are clearly age-dependent, suggesting that pro-aging effect of PM_2.5 exposure may be its core toxicity. However, it should be noted that to establish the pro-aging action of exposure to ambient PM_2.5, the demonstration of pre-mature death only is insufficient. A longitudinal evaluation of aging biomarkers should be performed to show that exposure to ambient PM_2.5 promotes their accumulation in flies.

It is noteworthy that although the fly maximum lifespans of the present study are still in the range commonly observed for w¹¹¹⁸, there is considerable early mortality. This may be attributed to the difference between the exposure chambers and the ambient in temperature and/or humidity. To achieve the expected concentration of PM_2.5, temperature and humidity in the exposure chambers have to be maintained within a relatively big range and thus be slightly different from those of the ambient. Since flies are sensitive to both temperature and humidity, the difference in temperature and/or humidity is very likely to cause the early mortality. However, as temperature and humidity in FA and CAP chambers are strictly controlled to be comparable, they are unlikely to confound the assessments of PM_2.5 toxicity. As our concentration system did not remove non-particulate pollutants such as nitrogen dioxide and ozone, the early mortality may also be due to exposure to those non-particulate pollutants.

Inflammation encompassing its associated oxidative stress is believed to be central in the pathogenesis of exposure to ambient PM_2.5 in humans (Terzano et al., 2010). In the present study, we demonstrate that exposure to CAP induced inflammation-associated signaling in Drosophila too, as evidenced by activation of pro-inflammatory signalling pathways including Jnk, Jak, and Nf-κb pathways and induction of oxidative stress. These data strongly support the similarity of responses to PM_2.5 between Drosophila and mammals, and thus validate their application in PM_2.5 toxicology. In addition, our results reveal that activation of those pro-inflammatory signalling pathways in Drosophila is clearly time-dependent. This is consistent with human studies showing temporally and/or spatially different activation of diverse pathways in response to PM_2.5 exposure (Brook et al., 2010). It is particularly interesting that our results demonstrated that activation of Jnk and Jak pathways appears to lead activation of Nf-κb (Figure 2). However, whether this activation precedency equals to a cause–effect relationship has to be determined.

Glucose homeostasis and insulin signaling play a key role in aging (Lopez-Otin et al., 2013). Our present data show that exposure to CAP significantly increased whole-body and circulating glucose levels, paralleled by increased mRNA expression of Ilp2 and Ilp5, indicating that exposure to PM_2.5 induces dysregulated insulin signaling in Drosophila. This is consistent with a rapidly increasing number of studies showing that exposure to PM_2.5 results in metabolic abnormality in humans and mammals (Brook et al., 2013; Chen et al., 2013; Park et al., 2010; Sun et al., 2013; Wagner et al., 2014). Therefore, our results regarding the metabolic effects of CAP exposure not only recall the similarity of responses to PM_2.5 between Drosophila and humans, but also suggest that Drosophila models can be used for study on PM_2.5 exposure-induced metabolic abnormality.

The present study is primarily aimed to validate Drosophila models for study on the toxicity of PM_2.5 exposure. Given that flies are well known to play an important role in ecosystems such as pollinating flowers and pest control, the demonstration of markedly increased mortality in CAP-exposed flies may also be ecologically significant. Currently, the main environmental effects of PM_2.5 are expected to be visibility impairment and environmental damage (primarily from acid deposition) (https://www3.epa.gov/pm/health.html). Our data suggest that ambient PM_2.5 may affect ecosystems directly through its toxicological effects.

This study demonstrates that similar to responses of humans and mammals, Drosophila develop systemic inflammation-like response and dysregulated insulin signaling and have a markedly increased premature mortality in response to exposure to ambient PM_2.5. These similarities, along with their well-known merits in genetics and aging study, validate Drosophila as a valuable model system in which to study PM_2.5 toxicology.

ACKNOWLEDGMENTS

X.W., M.C., Z.H., and L.Q. are thanked for performing the experiments and contributed to acquisition of data. M.Z. and L.C. are thanked for PM composition analysis. Z.Y. is thanked for designing the experiment and interpreting the results. S.R. and N.G.F. are thanked for contributing to CAP exposure of the animals. The manuscript was written by X.W. and Z.Y. All authors read, corrected and approved the manuscript.

FUNDING

This work was supported by the National Institutes of Health (R01ES024516 to Z.Y.), the American Heart Association (13SDG17070131 to Z.Y.) and the National Natural Science Foundation of China (Grant No. 81270342 to Z.Y., 21407082 to X.W., 81500216 to C.M. and 81302452 to L.Q.).