Chemical Characteristics and Cytotoxicity to GC-2spd(ts) Cells of PM2.5 in Nanjing Jiangbei New Area from 2015 to 2019

PM2.5 is an air pollutant with complex components. After entering the body through respiration, PM2.5 can not only cause respiratory diseases, but also break through the blood–testis barrier and influence the reproductive system. PM2.5 with different components may result in different toxic effects. In the first five years of Nanjing Jiangbei New Area, industrial transformation would change the concentration and chemical fraction of PM2.5 in the local environment to a certain extent. In this study, PM2.5 collected in Nanjing Jiangbei New Area every autumn and winter from 2015 to 2019 was analyzed. PM2.5 concentration generally decreased year by year. The large proportion of secondary inorganic ions indicated the presence of secondary pollution at the sampling site. PM2.5 was mainly emitted from fossil fuel combustion and vehicle exhaust. The cytotoxicity of PM2.5 samples was evaluated by PM2.5 exposure to mouse spermatocytes (GC-2spd(ts) cells). Cell viability was relatively low in 2016 and 2018, and relatively high in 2017 and 2019. Reactive oxygen species levels and DNA damage levels followed similar trends, with an overall annual decrease. The cytotoxicity of PM2.5 on GC-2spd(ts) cells was significantly correlated with water-soluble ions, water-soluble organic carbon, heavy metals and polycyclic aromatic hydrocarbons (p < 0.01). According to principal component analysis and multiple linear regression, fossil fuel combustion, secondary transformation of pollutants and construction dust were identified as the major contributors to cytotoxic effects, contributing more than 50%.


Introduction
Rapid modernization and energy consumption have led to increased pollutant emissions, and air pollution is gradually becoming a critical environmental problem in China [1]. Air pollution can adversely affect human health and increase the risk of cancer [2]. PM 2.5 , an important air pollutant, was recognized as a class I carcinogen by the International Agency for Research on Cancer (IARC), and its concentration is an important indicator of air pollution [3,4]. PM 2.5 pollution has become a major global environmental problem [5]. PM 2.5 was ranked as the fifth leading cause of death in the 2015 Global Burden of Disease Study, with PM 2.5 exposure resulting in more than 4 million premature deaths [6]. According to National Class I standard of PM 2.5 concentration of 35 µg/m 3 , 81% of Chinese residents lived in areas where PM 2.5 concentration exceeded the standard in 2017 [7]. A growing number of toxicological studies have demonstrated that PM 2.5 can enter and deposit in the alveoli carrying numerous chemicals, bacteria and viruses [8]. PM 2.5 can directly affect the respiratory system by inducing oxidative damage and inflammatory responses, and can also reach other organs through blood circulation, leading to serious multisystem diseases [9]. PM 2.5 exposure can cause vascular endothelial cell damage, lead to cellular dysfunction and increase the risk of cardiovascular diseases such as atherosclerosis [10]. A large number of toxicological studies related to PM 2.5 have focused on the PM 2.5 samples were collected on the roof of the library of Nanjing University of Information Science and Technology, Jiangbei New Area, Nanjing, China ( Figure S1). From 2015 to 2019, 20 samplings were carried out from October to November each year, with a total of 100 samples. PM 2.5 was collected on quartz microfiber filters (Whatman, UK) through a large-flow sampler (Tisch Environmental, Cleves, OH, USA) with a flow rate of 1.13 m 3 /min. The filters were preheated in a muffle furnace at 450 • C for 6 h to remove impurities. Before and after sampling, the filters were weighed after standing in a desiccator at room temperature for 24 h. The sampled filters were cut into pieces according to different years and extracted ultrasonically with ultrapure water for 20 min three times. The suspension was filtered through eight layers of gauze and freeze-dried to obtain PM 2.5 powder. The samples were stored in a refrigerator at −20 • C away from light for later use.

Composition Analysis of PM 2.5
After microwave digestion with 65% HNO 3 , the concentrations of heavy metals (including Pb, Zn, Ba, Cu, Sr, Mn, Cr, Ni and Cd) in the samples were determined by an inductively coupled plasma mass spectrometry (ICP-MS, Thermo Fisher Scientific, Waltham, MA, USA). After sonication with ultrapure water and filtration through the polytetrafluoroethylene (PTFE) membranes with a pore size of 0.22 µm, the concentrations of the water-soluble ions (including Na + , NH 4 + , K + , Mg 2+ , Ca 2+ , F − , Cl − , SO 4 2− and NO 3 − ) and WSOC in the samples were determined by an ion chromatograph (IC, Dionex, Sunnyvale, CA, USA) and a total organic carbon analyzer (TOC-L, Shimadzu, Tokyo, Japan), respectively. After adding the internal standard solution and dichloromethane, the samples were subjected to ultrasonic filtration and rotary evaporation. Then, PAHs concentrations in the samples were measured by a gas chromatography-mass spectrometer (GC-MS, Agilent, Santa Clara, CA, USA). According to the list of class I and class II carcinogens published by IARC, 15 PAHs were detected in this study, including naphthalene The standard curves of all tested substances were linear (r 2 > 0.997). The blank samples with standard reference were analyzed in parallel with tested samples. The recovery rates of all tested substances were within 100 ± 15%. Further method details and instrumental conditions were given in Tables S1-S3.

Cell Culture and Exposure
GC-2spd(ts) cells were used in this study, which were kindly provided by Stem Cell Bank, Chinese Academy of Sciences. GC-2spd(ts) cells were cultured in Dulbecco's Modified Eagle Medium (DMEM, Bio-Channel, China) supplemented with 10% fetal bovine serum (FBS, Gibco, Australia), and maintained in an incubator (Thermo Scientific, Waltham, MA, USA) at a constant temperature of 37 • C and 5% CO 2 . The PM 2.5 powder was dissolved with serum-free DMEM to prepare the PM 2.5 venom with a certain concentration gradient (0, 50, 100, 200, 400 µg/mL) after ultrasonication and vortexing. The concentrations of the PM 2.5 venom were selected based on the results of previous experiments. At this concentration gradient, the cells exhibited variable toxic effects while ensuring a certain viability. For each exposure experiment, GC-2spd(ts) cells in logarithmic growth stages were seeded at a density of 1 × 10 5 cells/mL and cultured for 24 h, and then exposed to PM 2.5 at different concentrations for 24 h. A blank group and a control group were set up in each experiment. The blank group contained only DMEM without cells and PM 2.5 , while the control group contained only DMEM and cells without PM 2.5 . Each sample was repeated three times in parallel.

Cytotoxicity Assay
In this study, cell viability, cellular ROS levels and DNA damage levels of GC-2spd(ts) cells were measured after exposure to PM 2.5 . Cell viability was detected using the CCK-8 kit (Beyotime Biotechnology, Shanghai, China). First, 10 µL of CCK-8 reagent was added to each sample. After incubation at 37 • C in the dark for 2 h, the optical density (OD) of the samples was measured at the wavelength of 450 nm by a microplate reader (Molecular Devices, San Jose, CA, USA). Cell viability is expressed as the ratio of (OD sample -OD blank ) to (OD control -OD blank ). Cellular ROS levels were detected using the DCFH-DA probe. The DCFH-DA powder (Sigma-Aldrich, St. Louis, MO, USA) was dissolved in dimethyl sulfoxide (DMSO, Purity > 99.9%, Beyotime Biotechnology, Shanghai, China) to prepare a 10 mM stock solution. It was then diluted with serum-free DMEM to 10 µM DCFH-DA solution (0.1% DMSO) and added to each sample. After incubation at 37 • C in the dark for 25 min, the cells were washed three times with phosphate-buffered saline (PBS) and collected. The collected cells were measured for fluorescence intensity by flow cytometry (Beckman, Brea, CA, USA) at the excitation wavelength of 488 nm and emission wavelength of 525 nm. The cellular ROS level was expressed as the ratio of fluorescence intensity of the sample group to the control group. For the DNA damage assay, DNA double strand break staining kit by γ-H2AX immunofluorescence (Beyotime Biotechnology, Shanghai, China) was used. When cellular DNA double strand breaks, H2AX will undergo phosphorylation modification to produce γ-H2AX. Therefore, γ-H2AX is often used as a marker of DNA damage, whose content level can reflect the degree of cellular DNA damage. After staining with the kit, the cells were observed and captured under a fluorescence microscope (Jiangnan Yongxin Optics, Nanjing, China). γ-H2AX exhibited green fluorescence at the excitation wavelength of 488 nm. The fluorescence intensity was calculated by the captured images and ImageJ software (v1.52, National Institutes of Health, Bethesda, MD, USA). The DNA damage level was expressed as the ratio of the fluorescence intensity of the sample group to the control group. All experiments were performed in three replicates to ensure the accuracy of the results.

Principal Component Analysis-Multiple Linear Regression (PCA-MLR)
Principal component analysis (PCA) was performed on the chemical components of PM 2.5 in Nanjing Jiangbei New Area using SPSS software (v23.0, International Business Machines Corporation, Armonk, NY, USA). Several principal components that can represent the vast majority of the variation in the samples would be extracted. The sources of PM 2.5 were identified by using the principal factor loadings of the chemical elements after great rotation of the variance. Multiple linear regression analysis (MLR) was then performed by SPSS software (v23.0) to obtain the main sources and their relative contributions.

Statistical Analysis
Statistical analysis was performed using SPSS software (v23.0). The variation of cytotoxicity at different exposure concentrations was analyzed by one-way analysis of variance (ANOVA). The correlation between the composition and cytotoxicity of PM 2.5 was investigated by Pearson correlation analysis, and the results are shown in Figure S2. For all cases, p < 0.05 was considered statistically significant.

Chemical Characteristics of PM 2.5
The average PM 2.5 concentration of Nanjing Jiangbei New Area during the sampling period from 2015 to 2019 was 54.00 ± 18.68 µg/m 3 , 50.08 ± 31.19 µg/m 3 , 38.97 ± 20.87 µg/m 3 , 48.35 ± 19.25 µg/m 3 and 31.28 ± 24.68 µg/m 3 , respectively. As shown in Figure 1, during the sampling period from 2015 to 2019, the number of days that the daily average PM 2.5 concentration reached the national class I standard was 4, 5, 12, 8 and 16 days, respectively. Overall, PM 2.5 concentrations during the sampling period showed a decreasing trend year by year. This may be related to the industrial transformation of Nanjing Jiangbei New Area after its establishment in 2015, where high-tech industries have gradually replaced the traditional chemical industry. In addition, the rebound of PM 2.5 concentrations in 2018 may be due to extreme meteorological conditions and temperature inversion, which were unfavorable to the diffusion of pollutants. After the sampling period of this study in 2018, a heavy haze event lasting for two weeks occurred in Nanjing. As a result, the air pollution control for the autumn and winter seasons of 2019 was advanced to the beginning of October, which contributed to the lowest PM 2.5 concentration during the sampling period in 2019. Since the composition of PM 2.5 varies with the emission sources, seasons, and climate, PM 2.5 concentrations do not adequately explain the risk to human health. Therefore, the contents of water-soluble components, heavy metals and PAHs in sampled PM 2.5 were determined in this study. The mass proportions of major chemical compositions in PM 2.5 are shown in Figure S3. Water-soluble substances were the  [30]. Considered with PM 2.5 concentrations, all measured components had the greatest abundance in the atmosphere in 2015. The unidentified fraction is estimated to be the insoluble carbonaceous component. study in 2018, a heavy haze event lasting for two weeks occurred in Nanjing. As a r the air pollution control for the autumn and winter seasons of 2019 was advanced beginning of October, which contributed to the lowest PM2.5 concentration durin sampling period in 2019. Since the composition of PM2.5 varies with the emission so seasons, and climate, PM2.5 concentrations do not adequately explain the risk to h health. Therefore, the contents of water-soluble components, heavy metals and PA sampled PM2.5 were determined in this study. The mass proportions of major che compositions in PM2.5 are shown in Figure S3. Water-soluble substances were the ured components with the highest proportion of PM2.5, accounting for 45.54-64 which was similar to the previous result [30]. Considered with PM2.5 concentratio measured components had the greatest abundance in the atmosphere in 2015. Th dentified fraction is estimated to be the insoluble carbonaceous component.

Water-Soluble Ions
The proportion of water-soluble ions in PM2.5 samples from 2015 to 2019 ranged 31.63% to 52.94%, which was higher than that in Xiamen (24.4%) and lower than t Zhengzhou (66.1%) [31,32]. As shown in Figure 2, SO4 2− , NO3 − and NH4 + (SNA) we most abundant water-soluble ions, accounting for 53.89-79.71% of the total water-so ions. SNA can affect atmospheric visibility and acidity of precipitation, adversely encing regional climate and the lives of residents [33,34]. SNA are mainly gene through the secondary transformation of SO2, NOx and NH3. The high SNA concentr in PM2.5 samples from 2015 to 2019 indicated relatively severe secondary pollution ar the sampling sites. The ratio of NO3 -concentration to SO4 2-concentration is often u identify whether NOx and SO2 in the atmosphere come from mobile or fixed source If the ratio is greater than 1, the emission sources are dominated by mobile sources vehicle exhaust) and conversely by fixed sources (e.g., coal combustion). The rat PM2.5 samples from 2015 to 2019 were 0.91, 0.58, 0.85, 1.16 and 0.92, respectively. E for that in 2018, the NO3 − /SO4 2− in other years were less than 1, indicating that NO SO2 in Nanjing Jiangbei New Area in autumn and winter were mainly from fixed s emissions. However, the values of NO3 − /SO4 2− generally showed an increasing trend gesting that the contribution of mobile source emissions to PM2.5 pollution was incre NH4 + was significantly correlated with SO4 2− and NO3 − , with correlation coefficients o and 0.95, respectively (p < 0.01). The ratios of the molar concentration of NH4 + to SO sampled PM2.5 for five years were all calculated to be greater than 2, determining that

Water-Soluble Ions
The proportion of water-soluble ions in PM 2.5 samples from 2015 to 2019 ranged from 31.63% to 52.94%, which was higher than that in Xiamen (24.4%) and lower than that in Zhengzhou (66.1%) [31,32]. As shown in Figure 2, SO 4 2− , NO 3 − and NH 4 + (SNA) were the most abundant water-soluble ions, accounting for 53.89-79.71% of the total watersoluble ions. SNA can affect atmospheric visibility and acidity of precipitation, adversely influencing regional climate and the lives of residents [33,34]. SNA are mainly generated through the secondary transformation of SO 2 , NO x and NH 3 . The high SNA concentrations in PM 2.5 samples from 2015 to 2019 indicated relatively severe secondary pollution around the sampling sites. The ratio of NO 3 concentration to SO 4 2concentration is often used to identify whether NO x and SO 2 in the atmosphere come from mobile or fixed sources [35]. If the ratio is greater than 1, the emission sources are dominated by mobile sources (e.g., vehicle exhaust) and conversely by fixed sources (e.g., coal combustion). The ratios of PM 2.5 samples from 2015 to 2019 were 0.91, 0.58, 0.85, 1.16 and 0.92, respectively. Except for that in 2018, the NO 3 − /SO 4 2− in other years were less than 1, indicating that NO x and SO 2 in Nanjing Jiangbei New Area in autumn and winter were mainly from fixed source emissions. However, the values of NO 3 − /SO 4 2− generally showed an increasing trend, suggesting that the contribution of mobile source emissions to PM 2.5 pollution was increasing. NH 4 + was significantly correlated with SO 4 2− and NO 3 − , with correlation coefficients of 0.91 and 0.95, respectively (p < 0.01). The ratios of the molar concentration of NH 4 + to SO 4 2− in sampled PM 2.5 for five years were all calculated to be greater than 2, determining that NH 4 + was present in the atmosphere primarily in the form of ammonium sulfate and ammonium nitrate [36]. Na + and Cl − were also present in relatively high levels, which are generally considered to come from the ocean. However, the correlation between Na + and Cl − in this study was not significant, implying that there may be other sources. The concentration ratio of Cl − to Na + in seawater is 1.8 [37], which in sampled PM 2.5 from 2015 to 2019 were 50.95, 2.11, 1.34, 0.39 and 0.56, respectively. Therefore, it was presumed that Cl − in PM 2.5 in 2015 and 2016 were likely from fossil fuel combustion [38], while Na + were from biomass combustion [39]. In contrast, Na + and Clin PM 2.5 mainly came from the ocean in 2017, 2018 and 2019. In addition, Fwas emitted from fossil fuel combustion [40], K + was mainly from biomass combustion [41], Ca 2+ was mostly influenced by construction dust [42] and both the latter two sources contributed to Mg 2+ [43]. nium nitrate [36]. Na + and Cl − were also present in relatively high levels, which are ge ally considered to come from the ocean. However, the correlation between Na + and C this study was not significant, implying that there may be other sources. The concen tion ratio of Cl − to Na + in seawater is 1.8 [37], which in sampled PM2.5 from 2015 to 2 were 50.95, 2.11, 1.34, 0.39 and 0.56, respectively. Therefore, it was presumed that C PM2.5 in 2015 and 2016 were likely from fossil fuel combustion [38], while Na + were f biomass combustion [39]. In contrast, Na + and Clin PM2.5 mainly came from the ocea 2017, 2018 and 2019. In addition, Fwas emitted from fossil fuel combustion [40], K + mainly from biomass combustion [41], Ca 2+ was mostly influenced by construction d [42] and both the latter two sources contributed to Mg 2+ [43].

Heavy Metals
The fractions of heavy metals in sampled PM2.5 from 2015 to 2019 were 3.99%, 3.3 3.02%, 4.56% and 2.81%, respectively. The highest content of heavy metals in samp PM2.5 was found in 2018, while the content of heavy metals decreased year by year in o years. The concentration of each metal element varied somewhat from year to year, wh may be related to its source. Zn and Cu are mainly from metallurgical industry and combustion [44,45]. Vehicle exhaust and parts wear and are also important sources o and Cu [46,47]. Mn and Ba have high background values in soil and generally come f ground dust [48][49][50]. Pb, Cr and Ni are usually emitted through fossil fuel combus and vehicle exhaust [51]. Sr may come from incineration soot and ground dust [52], w Cd is largely from combustion of coal and fuel oil [53]. In these five years, Zn, Cu, Mn Pb were the most abundant metal elements in the sampled PM2.5, accounting for 93 97.05% of the total measured heavy metals (Figure 3a), which was consistent with the sult of previous research [54]. This may be related to the reserved chemical parks, the n construction sites, and the large number of transportation vehicles during the construc

Heavy Metals
The fractions of heavy metals in sampled PM 2.5 from 2015 to 2019 were 3.99%, 3.38%, 3.02%, 4.56% and 2.81%, respectively. The highest content of heavy metals in sampled PM 2.5 was found in 2018, while the content of heavy metals decreased year by year in other years. The concentration of each metal element varied somewhat from year to year, which may be related to its source. Zn and Cu are mainly from metallurgical industry and fuel combustion [44,45]. Vehicle exhaust and parts wear and are also important sources of Zn and Cu [46,47]. Mn and Ba have high background values in soil and generally come from ground dust [48][49][50]. Pb, Cr and Ni are usually emitted through fossil fuel combustion and vehicle exhaust [51]. Sr may come from incineration soot and ground dust [52], while Cd is largely from combustion of coal and fuel oil [53]. In these five years, Zn, Cu, Mn and Pb were the most abundant metal elements in the sampled PM 2.5 , accounting for 93.43-97.05% of the total measured heavy metals (Figure 3a), which was consistent with the result of previous research [54]. This may be related to the reserved chemical parks, the new construction sites, and the large number of transportation vehicles during the construction phase of Nanjing Jiangbei New Area. To further distinguish the sources of metal elements, enrichment factor analysis was used to determine the contribution of natural and anthropogenic sources to the metal elements in sampled PM 2.5 [55]. In this study, Mn was selected as the reference element to calculate the enrichment factors (EF) of the other eight metal elements, and the results were listed in Table S4. Zn, Cu and Cd were dominated by anthropogenic sources. Ba was mainly from natural sources. Pb, Cr and Ni were contributed by both natural and anthropogenic sources. Sr in 2016 and 2017 was contributed by a combination of natural and anthropogenic sources, while Sr in the remaining years was from natural sources. This indicated that the heavy metals in PM 2.5 in Nanjing Jiangbei New Area from 2015 to 2019 mostly came from emissions of chemical parks and road traffic. During the sampling period from 2015 to 2019, the concentrations of Zn in PM 2.5 showed an overall decreasing trend, while Cu was increasing, indicating that Zn in the atmosphere mainly came from chemical industry emissions, while Cu was more likely to be from vehicle exhaust and parts wear. metal elements, and the results were listed in Table S4. Zn, Cu and Cd were dominated by anthropogenic sources. Ba was mainly from natural sources. Pb, Cr and Ni were contributed by both natural and anthropogenic sources. Sr in 2016 and 2017 was contributed by a combination of natural and anthropogenic sources, while Sr in the remaining years was from natural sources. This indicated that the heavy metals in PM2.5 in Nanjing Jiangbei New Area from 2015 to 2019 mostly came from emissions of chemical parks and road traffic. During the sampling period from 2015 to 2019, the concentrations of Zn in PM2.5 showed an overall decreasing trend, while Cu was increasing, indicating that Zn in the atmosphere mainly came from chemical industry emissions, while Cu was more likely to be from vehicle exhaust and parts wear.

Organic Carbon
Organic carbon is an important component of PM2.5, including primary organic carbon (POC) emitted directly by combustion and secondary organic carbon (SOC) generated by volatile organic compounds (VOCs) through photochemical reactions [56,57]. In general, organic carbon accounts for about 50% of the total mass of PM2.5, with WSOC occupying 10-70% of organic carbon [58][59][60]. The concentrations of WSOC in PM2.5 in Nanjing Jiangbei New Area during the sampling period from 2015 to 2019 ranged from 9.46% to 19.76%, with the highest and lowest concentrations appearing in 2015 and 2018, respectively. In this study, there was a highly significant correlation between WSOC and Cl − (r 2 = 0.94, p < 0.01), suggesting that fossil fuel combustion contributed to WSOC in PM2.5. Meanwhile, WSOC also exhibited relatively strong correlation with SNA, with correlation coefficients of 0.77, 0.82 and 0.83, respectively (p < 0.01). As mentioned above, there was serious secondary pollution at the sampling sites during the sampling period, thus WSOC may be partially generated through secondary reactions. In addition, the concentrations of PAHs in sampled PM2.5 were measured in this study, as shown in Figure 3b. The concentrations of PAHs in sampled PM2.5 from 2015 to 2019 ranged from 161.93 to 450.70 ng/g, accounting for a mere 0.02-0.05%. Despite their low levels in PM2.5, PAHs are carcinogenic and can influence human reproductive function as endocrine disruptors [23,61]. In this study, the total PAHs concentration in general tended to decrease. The 4-6-ring PAHs

Organic Carbon
Organic carbon is an important component of PM 2.5 , including primary organic carbon (POC) emitted directly by combustion and secondary organic carbon (SOC) generated by volatile organic compounds (VOCs) through photochemical reactions [56,57]. In general, organic carbon accounts for about 50% of the total mass of PM 2.5 , with WSOC occupying 10-70% of organic carbon [58][59][60]. The concentrations of WSOC in PM 2.5 in Nanjing Jiangbei New Area during the sampling period from 2015 to 2019 ranged from 9.46% to 19.76%, with the highest and lowest concentrations appearing in 2015 and 2018, respectively. In this study, there was a highly significant correlation between WSOC and Cl − (r 2 = 0.94, p < 0.01), suggesting that fossil fuel combustion contributed to WSOC in PM 2.5 . Meanwhile, WSOC also exhibited relatively strong correlation with SNA, with correlation coefficients of 0.77, 0.82 and 0.83, respectively (p < 0.01). As mentioned above, there was serious secondary pollution at the sampling sites during the sampling period, thus WSOC may be partially generated through secondary reactions. In addition, the concentrations of PAHs in sampled PM 2.5 were measured in this study, as shown in Figure 3b. The concentrations of PAHs in sampled PM 2.5 from 2015 to 2019 ranged from 161.93 to 450.70 ng/g, accounting for a mere 0.02-0.05%. Despite their low levels in PM 2.5 , PAHs are carcinogenic and can influence human reproductive function as endocrine disruptors [23,61]. In this study, the total PAHs concentration in general tended to decrease. The 4-6-ring PAHs were the most abundant, accounting for 72.17-90.97% of the total PAHs, but their concentrations displayed a yearly downward trend as well. For each year, B[ghi]P was the most abundant single PAH, followed by B[b&k]F and IND. These high molecular weight PAHs are generally considered to be derived from fuel and coal combustion [62]. To further identify the sources of PAHs in PM 2.5 , the diagnostic ratio analysis was utilized, and the results were listed in Table 1 [63][64][65]. According to that, PAHs in PM 2.5 during the sampling period from 2015 to 2019 were mainly emitted from combustion sources, including coal, diesel and gasoline. Moreover, PAHs with high concentrations were significantly correlated with Cl − and Zn from industrial sources, but weakly correlated with NO 3 and Cu from traffic sources. Therefore, PAHs in PM 2.5 were majorly contributed by fossil fuel combustion in the surrounding chemical parks in this study. As shown in Figure 4a, the viability of GC-2spd(ts) cells in all five exposure groups decreased with increasing PM 2.5 exposure concentrations. At the PM 2.5 exposure concentration of 50 µg/mL, the cell viability of the exposure group was not significantly different from that of the control group. When the PM 2.5 concentration was increased to 100-200 µg/mL, the cell survival rates of the exposure groups varied significantly from the control group. The highest cytotoxicity of PM 2.5 was presented in 2016, followed by 2018, 2015, 2017 and 2019. At the exposure concentration of 400 µg/mL, the lowest and highest cell survivals were obtained in 2018 and 2019, which were 54.73% and 70.38%, respectively. A linear fit was performed for PM 2.5 exposure concentration and cell viability for each year from 2015 to 2019. It was found that there was a linear negative correlation between PM 2.5 exposure concentration and cell viability for each year (p < 0.05), with r 2 of 0.93, 0.76, 0.88, 0.88 and 0.94, respectively, demonstrating a dose-effect relationship between PM 2.5 exposure and cell viability. Similarly, a study in Beijing revealed that PM 2.5 inhibited the viability of GC-2spd(ts) cells and even induced apoptosis at high exposure concentrations [14]. It has also been suggested that these may be due to the elevated cellular ROS level caused by PM 2.5 [13]. Overall, the cytotoxicity of PM 2.5 in 2017 and 2019 was relatively weak, whereas that in 2016 and 2018 was more able to inhibit the viability of GC-2spd(ts) cells, probably due to the differences in the chemical compositions of PM 2.5 in different years. The highest SNA level in 2016 and the highest heavy metal level in 2018 may contribute to the stronger cytotoxicity of PM 2.5 in these two years.

Cellular ROS
As shown in Figure 4b, the ROS levels of GC-2spd(ts) cells increased with increasing exposure concentrations in the five exposure groups from 2015 to 2019. At the low exposure concentration (50 µg/mL), there was no significant difference in cellular ROS levels between the exposure group and the control group, except for the 2015 exposure group. However, when the exposure concentration was increased to 100 µg/mL, the cellular ROS levels of all exposure groups were significantly different from that of the control group (p < 0.05), indicating that oxidative stress occurred intracellularly in all exposure groups. At concentrations of 50-200 µg/mL, PM 2.5 in 2015 led to the highest ROS levels, which significantly varied from the results of the exposure group in other years (p < 0.05). The highest WSOC concentration, the highest PAHs concentration and relatively high heavy metal concentration were found in PM 2.5 in 2015, and all these components are believed to induce elevated cellular ROS [21][22][23][24][25]. The lowest ROS level at each concentration was observed in the 2019 exposure group, which might be due to the lowest concentrations of water-soluble ions, heavy metals and PAHs in PM 2.5 in 2019. As the exposure concentration increased, the elevation of cellular ROS levels became slower in all exposure groups. This could result from that the cell viability decreased with increased exposure concentrations, and the cells that were apoptotic or dead due to severe oxidative stress failed to be detected. It has also been demonstrated in another study that PM 2.5 exposure not only caused an elevated ROS level in GC-2spd(ts) cells, but also led to cellular DNA damage [14]. By adding antioxidants, DNA damage was improved and apoptosis was inhibited, suggesting that oxidative stress played a key role in the adverse effects of PM 2.5 on GC-2spd(ts) cells.

DNA Damage
When DNA double-strand breaks, H2AX can be phosphorylated to generate γ-H2AX [66], the level of which can reflect the degree of DNA damage. In this study, the green fluorescence of γ-H2AX could be observed under fluorescence microscope after staining, as shown in Figure 5a. Compared with the control group, all exposure groups exhibited more fluorescent spots and higher fluorescence intensity. With the increase of exposure concentration, the number and fluorescence intensity of fluorescent spots in the images increased, and the γ-H2AX content increased. This indicated that more cells underwent DNA damage and that the DNA damage was aggravated. It is evident from Figure 5a that exposure to PM 2.5 in 2015 and 2016 induced a greater degree of DNA damage. To quantify the level of DNA damage in GC-2spd(ts) cells, the fluorescence intensity was calculated for each image, as shown in Figure 5b. With increasing PM 2.5 concentration, the γ-H2AX level in GC-2spd(ts) cells increased significantly (p < 0.01), indicating an elevated level of cellular DNA damage. Zhang et al. also demonstrated that PM 2.5 increased 8-OHdG level in GC-2spd(ts) cells at high exposure doses, inducing DNA damage, which in turn may lead to reduced fertility [13]. Overall, exposure to PM 2.5 in 2015 resulted in the highest γ-H2AX level and the highest degree of DNA damage, followed by PM 2.5 in 2016, and the weakest toxic effect of PM 2.5 was observed in 2019. This trend approximated that of ROS levels, probably because intracellular oxidative stress can induce DNA damage. After DNA damage occurs in germ cells, failure to repair in time or failure to repair will lead to apoptosis or alteration of genetic information, thereby resulting in genotoxicity [67].
lead to reduced fertility [13]. Overall, exposure to PM2.5 in 2015 resulted in the highest γ-H2AX level and the highest degree of DNA damage, followed by PM2.5 in 2016, and the weakest toxic effect of PM2.5 was observed in 2019. This trend approximated that of ROS levels, probably because intracellular oxidative stress can induce DNA damage. After DNA damage occurs in germ cells, failure to repair in time or failure to repair will lead to apoptosis or alteration of genetic information, thereby resulting in genotoxicity [67].

Potential Toxic Components, Sources and Contributions
There are differences in the toxic effects caused by different components of PM2.5. PAHs are intensely carcinogenic and can reduce cell viability, thereby threatening human health [68,69]. Heavy metals are bioaccumulative, which can influence bones and nerves, and even result in genetic damage [70,71]. PAHs and heavy metals in PM2.5 were confirmed by previous studies to induce DNA damage and increased cellular ROS [72,73]. Compared with other components, water-soluble components are relatively less toxic, but they can be distributed throughout the body with the blood, resulting in respiratory and cardiovascular diseases [74,75]. Water-soluble components of PM2.5 can lead to apoptosis by triggering inflammatory responses [76]. In this study, cytotoxicity of GC-2spd(ts) cells was significantly correlated with water-soluble ions, WSOC, heavy metals and PAHs in PM2.5 (p < 0.01). The favorable correlation suggested that the presence of these chemical components may contribute to the reduced cell viability, increased ROS levels, and enhanced DNA damage of GC-2spd(ts) cells. The differences in the chemical fraction of PM2.5 often depend on the source. From 2015 to 2019, the proportion of chemical components of PM2.5 in Nanjing Jiangbei New Area had varied, but these chemical components that probably induce toxic effects were always present. Therefore, tracing the sources of these toxicogenic components, assigning the contribution of each source to cytotoxic effects, and controlling the sources and generation pathways of toxicogenic components can reduce the health risk of PM2.5 to a certain extent. In this study, principal component analysis and

Potential Toxic Components, Sources and Contributions
There are differences in the toxic effects caused by different components of PM 2.5 . PAHs are intensely carcinogenic and can reduce cell viability, thereby threatening human health [68,69]. Heavy metals are bioaccumulative, which can influence bones and nerves, and even result in genetic damage [70,71]. PAHs and heavy metals in PM 2.5 were confirmed by previous studies to induce DNA damage and increased cellular ROS [72,73]. Compared with other components, water-soluble components are relatively less toxic, but they can be distributed throughout the body with the blood, resulting in respiratory and cardiovascular diseases [74,75]. Water-soluble components of PM 2.5 can lead to apoptosis by triggering inflammatory responses [76]. In this study, cytotoxicity of GC-2spd(ts) cells was significantly correlated with water-soluble ions, WSOC, heavy metals and PAHs in PM 2.5 (p < 0.01). The favorable correlation suggested that the presence of these chemical components may contribute to the reduced cell viability, increased ROS levels, and enhanced DNA damage of GC-2spd(ts) cells. The differences in the chemical fraction of PM 2.5 often depend on the source. From 2015 to 2019, the proportion of chemical components of PM 2.5 in Nanjing Jiangbei New Area had varied, but these chemical components that probably induce toxic effects were always present. Therefore, tracing the sources of these toxicogenic components, assigning the contribution of each source to cytotoxic effects, and controlling the sources and generation pathways of toxicogenic components can reduce the health risk of PM 2.5 to a certain extent. In this study, principal component analysis and multiple linear regression were used in combination. Five characteristic factors were extracted by PCA from various chemical components of Jiangbei New Area for five years. The contributions of the five factors to cytotoxic effects were then evaluated by MLR.
The cumulative variance contribution of the five characteristic factors reached 100%. The total variance explained and the rotated factor loading matrix were listed in Tables S5 and S6, respectively. The components with loadings higher than 0.5 in each factor were presented in Figure 6a. The loadings of 4-6-ring PAHs were higher in Factor 1, which were all greater than 0.8. In addition, F − , Cl − , Ni, Zn and WSOC also had large loadings in Factor 1. Combined with the previous analysis, Factor 1 mainly represented fossil fuel combustion source. In Factor 2, Cu, Pb, Cd and Nap exhibited strong loadings. Cu, Pb and Cd are considered to come from vehicle exhaust and parts wear, while Nap can be released from diesel exhaust of transport vehicles and asphalt used for paving [76]. NO 3 − also had a large loading in Factor 2, and its precursor (NO x ) was mainly from vehicle exhaust, indicating that Factor 2 primarily represented traffic source. The loadings of Na + , K + and Mg 2+ were larger in Factor 3, and they were considered to be mainly generated by biomass combustion in this study. Factor 3 was defined as biomass combustion source. The high loadings of Sr and Ba were found in Factor 4, which were initially considered to be from ground dust. However, F − , Cl − , Acp, AcPy and Ant also presented moderate loadings, and these components are usually common species in combustion products. Sr was possibly from the soot of waste incineration [77], and Ba can be detected in large quantities in wood waste incineration products [78]. F − and Cl − were the major components in the combustion products of plastic products. Acp, AcPy and Ant accounted for a significant proportion of the waste incineration products [79,80]. All of these supported Factor 4 as waste incineration source. Factor 5 had the highest loadings of SO 4 2− , NO 3 − and NH 4 + which were generated from the secondary conversion of SO 2 , NO x and NH 3 . Ca 2+ was also of large loading in Factor 5 and is the major component in construction dust. Therefore, Factor 5 represented the mixed source of secondary conversion and construction dust. In this study, these five sources were identified as the sources of major pollutants in Nanjing Jiangbei New Area from 2015 to 2019.
presented in Figure 6a. The loadings of 4-6-ring PAHs were higher in Factor 1, w were all greater than 0.8. In addition, F − , Cl − , Ni, Zn and WSOC also had large loadin Factor 1. Combined with the previous analysis, Factor 1 mainly represented fossil combustion source. In Factor 2, Cu, Pb, Cd and Nap exhibited strong loadings. Cu, Pb Cd are considered to come from vehicle exhaust and parts wear, while Nap can b leased from diesel exhaust of transport vehicles and asphalt used for paving [76]. N also had a large loading in Factor 2, and its precursor (NOx) was mainly from vehicl haust, indicating that Factor 2 primarily represented traffic source. The loadings of K + and Mg 2+ were larger in Factor 3, and they were considered to be mainly generate biomass combustion in this study. Factor 3 was defined as biomass combustion sou The high loadings of Sr and Ba were found in Factor 4, which were initially considere be from ground dust. However, F − , Cl − , Acp, AcPy and Ant also presented moderate l ings, and these components are usually common species in combustion products. Sr possibly from the soot of waste incineration [77], and Ba can be detected in large quant in wood waste incineration products [78]. F − and Cl − were the major components in combustion products of plastic products. Acp, AcPy and Ant accounted for a signifi proportion of the waste incineration products [79,80]. All of these supported Factor waste incineration source. Factor 5 had the highest loadings of SO4 2− , NO3 − and NH4 + w were generated from the secondary conversion of SO2, NOx and NH3. Ca 2+ was als large loading in Factor 5 and is the major component in construction dust. Therefore, tor 5 represented the mixed source of secondary conversion and construction dust. In study, these five sources were identified as the sources of major pollutants in Nan Jiangbei New Area from 2015 to 2019. To further determine the relative contribution of the major sources to cytotoxi fects, multiple linear regressions were performed with standardized five factors standardized cell viability, ROS levels and DNA damage levels, respectively. The re were displayed in Figure 6b. Except for the waste incineration source, the remaining sources all contributed significantly to the cytotoxic effects (p < 0.05). In terms of To further determine the relative contribution of the major sources to cytotoxic effects, multiple linear regressions were performed with standardized five factors and standardized cell viability, ROS levels and DNA damage levels, respectively. The results were displayed in Figure 6b. Except for the waste incineration source, the remaining four sources all contributed significantly to the cytotoxic effects (p < 0.05). In terms of cell viability, the mixed source of secondary conversion and construction dust contributed the most with 30.55%, followed by fossil fuel combustion source with 23.10%, traffic source with 22.27%, biomass combustion source with 14.71% and waste incineration source with 9.37%. For cellular ROS levels, the fossil fuel combustion source was the greatest contributor with 32.54%. Following that, the mixed source of secondary conversion and construction dust contributed 23.23%, traffic source contributed 16.32%, biomass combustion source contributed 14.73% and waste incineration source contributed 13.18%. The ranking of the contributions of the five sources to DNA damage levels was consistent with that to ROS levels. Fossil fuel combustion source contributed the most, followed by mixed source of secondary conversion and construction dust, traffic source, biomass combustion source and waste incineration source, with contributions of 35.12%, 24.88%, 16.06%, 15.07% and 8.87%, respectively. Overall, fossil fuel combustion, secondary conversion of pollutants and construction dust were the dominant sources of cytotoxic effects, contributing more than 50%. The traffic source and biomass combustion source also contributed a large proportion, which were worthy of attention in environmental management. They were the major sources of pollutants in Nanjing Jiangbei New Area from 2015 to 2019. Although many high-tech industries had been introduced, traditional chemical industries still existed and emitted large amounts of pollutants into the atmosphere, posing a potential threat to the health of surrounding residents. During the construction phase of the new area, dust from construction sites and exhaust from transport vehicles were almost daily present, so there might be a high risk of occupational exposure to construction workers, traffic police, drivers, etc. [81]. Straw burning had long been banned, but this phenomenon persisted in the suburbs. In addition, crop residues were often used as kitchen firewood in rural areas, which contributed to the emissions from biomass burning source [82].

Conclusions
In this study, PM 2.5 was collected from October to November in 2015-2019 in Nanjing Jiangbei New Area, and the chemical fractions and toxicity to GC-2spd(ts) cells were determined. From 2015 to 2019, the mass concentration of PM 2.5 showed an overall downward trend. Water-soluble ions were the most abundant component of PM 2.5 , and SNA accounted for the largest proportion of water-soluble ions, indicating the existence of severe secondary pollution at the sampling site. In this study, heavy metals were mainly from fossil fuel combustion and traffic emissions, WSOC from fossil fuel combustion and secondary conversion, and PAHs from combustion emissions from surrounding chemical parks. In terms of cytotoxicity, there was a dose-effect relationship between PM 2.5 exposure and viability of GC-2spd(ts) cells. The strongest inhibition of cell viability was observed in 2016 and 2018, and the cytotoxicity of PM 2.5 was relatively low in 2017 and 2019. The overall downward trends in ROS levels and DNA damage levels during these five years were probably due to similar trends in the concentrations of heavy metals and PAHs. This suggested that heavy metals and PAHs in PM 2.5 were likely to be significant components causing oxidative stress and DNA damage in cells. The trends of ROS levels and DNA damage levels were not exactly consistent with those of cell mortality, implying that there were other mechanisms leading to cell death besides oxidative stress and DNA damage, which deserved further investigation in future studies. The results of this study are based on in vitro exposure experiments. The effects of PM 2.5 on GC-2spd(ts) cells could reflect the reproductive toxicity risk of PM 2.5 to some extent but could not completely restore the toxic effects of PM 2.5 on the organism. Therefore, corresponding in vivo exposure studies are necessary for follow-up. By studying the changes in the reproductive system and germ cells after PM 2.5 exposure in mice, the reproductive toxicity and toxicogenic mechanisms of PM 2.5 can be further investigated. Five characteristic sources were extracted from pollutants at the sampling site for these five years, identified as fossil fuel combustion source, traffic source, biomass combustion source, waste incineration source and mixed source of secondary conversion and construction dust. Their contributions to the toxic effects of GC-2spd(ts) cells were evaluated. Fossil fuel combustion, secondary transformation of pollutants and construction dust were identified as the main contributors to cytotoxic effects. However, from the perspective of regional development, the contribution of an individual pollution source cannot be fixed, so this assessment approach has certain limitations. In addition, elevated levels of cellular ROS and DNA damage often lead to reduced cell viability, which may interfere with the assessment of source contribution. Therefore, studies on PM 2.5 toxicity and source contribution assessment at more time points may be needed. Toxicity studies for different sources of PM 2.5 could also be further conducted to compare their contributions. The reduction in PM 2.5 concentration, cytotoxicity and toxic component concentrations in Nanjing Jiangbei New Area from 2015 to 2019 revealed that environmental improvement might benefit from industrial upgrading, and environmental policy played a considerable role as well.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/toxics11020092/s1. Figure S1. Location of the sampling site. Figure S2. Pearson correlation between cell viability, ROS levels and DNA damage levels of GC-2spd(ts) cells, and individual components of PM 2.5 . More detailed data is listed in the file named "Supplementary data.xlsx". Figure S3 Mass fraction of major chemical compositions in PM 2.5 . Table  S1. Instrumental conditions for microwave digestion and ICP-MS.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.