Effects of Naphtho[2,1-a]pyrene Exposure on CYP1A1 Expression: An in Vivo and in Vitro Mechanistic Study Exploring the Role of m6A Posttranscriptional Modification

Abstract

Background: Currently, many emerging polycyclic aromatic hydrocarbons (PAHs) have been found to be widely present in the environment. However, little has been reported about their toxicity, particularly in relation to CYP1A1. Objectives: This study aimed to explore the toxicity of naphtho[2,1-a]pyrene (N21aP) and elucidate the mechanism underlying N21aP-induced expression of CYP1A1. Methods: The concentration and sources of N21aP were detected and analyzed by gas chromatography-triple quadrupole mass spectrometry (GC-MS/ MS) and diagnostic ratio analysis. Then the effects of CYP1A1 on the toxicity of N21aP were conducted in male wild-type (WT) and Cyp1a1 knockout mice exposed to N21aP (0.02, 0.2, and 2 mg/kg) through intratracheal instillation. Further, the aryl hydrocarbon receptor (AhR) pathway was examined through luciferase and chromatin immunoprecipitation (ChIP) assays. N6-methyladenosine (m6A) modification levels were measured on global RNA and specifically on CYP1A1 mRNA using dot blotting and methylated RNA immunoprecipitation-quantitative real-time polymerase chain reaction (MeRIP qRT-PCR), with validation by m6A inhibitors, DAA and SAH. m6A sites on CYP1A1 were identified by bioinformatics and luciferase assays, and CYP1A1 mRNA's interaction with IGF2BP3 was confirmed by RNA pull-down, luciferase, and RNA binding protein immunoprecipitation (RIP) assays. Results: N21aP was of the same environmental origin as benzo[a]pyrene (BaP) but was more stably present in the environment. N21aP could be metabolically activated by CYP1A1 to produce epoxides, causing DNA damage and further leading to lung inflammation. Importantly, in addition to the classical AhR pathway (i.e., BaP), N21aP also induced CYP1A1 expression with a posttranscriptional modification of m6A in CYP1A1 mRNA via the METTL14-IGF2BP3-CYP1A1 axis. Specifically, in the two recognition sites of METTL14 on the CYP1A1 mRNA transcript (position at 2700 and 5218), a methylation site (position at 5218) in the 3'-untranslated region (UTR) was recognized by IGF2BP3, enhanced the stability of CYP1A1 mRNA, and finally resulted in an increase in CYP1A1 expression. Discussion: This study systematically demonstrated that in addition to AhR-mediated transcriptional regulation, N21aP, had a new additional mechanism of m6A-mediated posttranscriptional modification, jointly contributing to CYP1A1 expression. Given that PAHs are the metabolic substrates of CYP1A1, this study not only helps to understand the significance of environment-genetic interactions for the toxicity of PAHs but also helps to better understand the health risks of the emerging PAHs at environmental exposure levels.

Headnote

Background: Currently, many emerging polycyclic aromatic hydrocarbons (PAHs) have been found to be widely present in the environment. However, little has been reported about their toxicity, particularly in relation to CYP1A1.

Objectives: This study aimed to explore the toxicity of naphtho[2,1-a]pyrene (N21aP) and elucidate the mechanism underlying N21aP-induced expression of CYP1A1.

Methods: The concentration and sources of N21aP were detected and analyzed by gas chromatography-triple quadrupole mass spectrometry (GC-MS/ MS) and diagnostic ratio analysis. Then the effects of CYP1A1 on the toxicity of N21aP were conducted in male wild-type (WT) and Cyp1a1 knockout mice exposed to N21aP (0.02, 0.2, and 2 mg/kg) through intratracheal instillation. Further, the aryl hydrocarbon receptor (AhR) pathway was examined through luciferase and chromatin immunoprecipitation (ChIP) assays. N6-methyladenosine (m6A) modification levels were measured on global RNA and specifically on CYP1A1 mRNA using dot blotting and methylated RNA immunoprecipitation-quantitative real-time polymerase chain reaction (MeRIP qRT-PCR), with validation by m6A inhibitors, DAA and SAH. m6A sites on CYP1A1 were identified by bioinformatics and luciferase assays, and CYP1A1 mRNA's interaction with IGF2BP3 was confirmed by RNA pull-down, luciferase, and RNA binding protein immunoprecipitation (RIP) assays.

Results: N21aP was of the same environmental origin as benzo[a]pyrene (BaP) but was more stably present in the environment. N21aP could be metabolically activated by CYP1A1 to produce epoxides, causing DNA damage and further leading to lung inflammation. Importantly, in addition to the classical AhR pathway (i.e., BaP), N21aP also induced CYP1A1 expression with a posttranscriptional modification of m6A in CYP1A1 mRNA via the METTL14-IGF2BP3-CYP1A1 axis. Specifically, in the two recognition sites of METTL14 on the CYP1A1 mRNA transcript (position at 2700 and 5218), a methylation site (position at 5218) in the 3'-untranslated region (UTR) was recognized by IGF2BP3, enhanced the stability of CYP1A1 mRNA, and finally resulted in an increase in CYP1A1 expression.

Discussion: This study systematically demonstrated that in addition to AhR-mediated transcriptional regulation, N21aP, had a new additional mechanism of m6A-mediated posttranscriptional modification, jointly contributing to CYP1A1 expression. Given that PAHs are the metabolic substrates of CYP1A1, this study not only helps to understand the significance of environment-genetic interactions for the toxicity of PAHs but also helps to better understand the health risks of the emerging PAHs at environmental exposure levels. https://doi.org/10.1289/EHP14055

Introduction

Polycyclic aromatic hydrocarbons (PAHs) are organic compounds primarily derived from natural processes, such as forest and brush fires, as well as anthropogenic activities, including outdoor indus-trial and vehicular emissions and indoor cigarette smoke.1,2 Cooking methods, such as grilling, roasting, and frying, are also significant sources of PAHs, particularly in indoor settings, such as restaurants and kitchens.3,4 Human exposure to PAHs can occur through various routes, including inhalation, dermal contact, and dietary intake, with respiratory inhalation and dietary intake being the main pathways. Often, these exposure routes overlap, leading to widespread adverse health effects.5,6

Generally, PAHs can be divided into three groups according to the number of aromatic rings: low molecular weight (LMW, 2-3 rings), medium molecular weight (MMW, 4 rings), and high mo-lecular weight (HMW, 5-6 rings).7 A study on PAHs exposure in the Beijing population revealed that ? 85% of LMW-PAHs were acquired through dietary intake, whereas the majority of HMW-PAHs were absorbed through inhalation, constituting 57%.6 LMW-PAHs, being less lipophilic, are easily excreted from the body.8 In contrast, HMW-PAHs exhibit high lipophilicity, increasing the probability of retention in the terminal bronchioles and alveoli, thereby posing significant health risks.8 Of the 16 US Environmental Protection Agency (EPA) priority-controlled PAHs, all HMW-PAHs [benzo[b]fluoranthene (BbF), benzo[k]flu-oranthene (BkF), benzo[a]pyrene (BaP), dibenzo[a,h]anthracene (DBahA), and indeno[1,2,3-cd]pyrene (IcdP)], except benzo[ghi] perylene (BghiP), are considered probable carcinogens.9 Recently, HMW-PAHs with a molecular weight of 302 (MW302) have been studied. They contained chemicals with more mutagenic structures from different sources, such as exhaust gas and gasoline vehicle combustion.10,11 The mutagenic activity of PAHs with MW302 in urban airborne particles accounted for 33% of the total mutagenic activity of PAHs in human lymphoblast cells.12 In fact, some MW302-PAHs were more mutagenic or carcinogenic than priority PAHs, such as dibenzo[a,l]pyrene (DBalP).13 DBalP is a particu-larly strong mutagen with a 30-fold relative potency factor (RPF) of BaP14 and is classified by the International Agency for Research on Cancer (IARC) as probably carcinogenic to humans (Group 2A). Other isomers of DBalP, such as dibenzo[a,i]pyrene (DBaiP) and dibenzo[a,h]pyrene (DBahP), are classified as possibly carci-nogenic to humans (Group 2B) (https://monographs.iarc.who.int/ list-of-classifications). However, to date, toxicity research on MW302-PAH is limited primarily to several dibenzopyrenes. Therefore, a considerable number of PAHs have unknown toxicity, and the 16 priority-controlled PAHs are far from sufficient to assess the health risks of PAHs in the environment.

PAHs are widely recognized for their carcinogenic, teratogenic, and mutagenic properties and have been established as a risk factor for various respiratory diseases, including inflammation, chronic obstructive pulmonary disease, and lung cancer.15-17 Research indi-cates that the majority of PAHs undergo metabolic activation by the cytochrome P450 enzyme (CYP450) to produce epoxide and qui-none metabolites, which are the primary contributors to their toxicity and carcinogenic effects.18 CYP1A1 is typically considered the principal and predominant metabolizing enzyme for PAHs18 and has abundant expression in the lungs.18,19 Noteworthily, PAHs are known inducers of CYP1 A1, which is in turn responsible for the me-tabolism and activation of PAHs.20 Thus, there are fascinating inter-actions between CYP1A1 and PAHs. On one hand, CYP1A1 can metabolically activate PAHs, leading to enhanced toxicity and even carcinogenicity. Animal models revealed that CYP1A1 was essen-tial for the metabolism of BaP to benzo[a]pyrenediolepoxide (BPDE), which binds to DNA to form DNA adducts, leading to DNA damage.18,21 BPDE-DNA adducts have also been detected in the serum of children diagnosed with pulmonary diseases and showing a significant positive correlation with DNA strand breaks.22 DBalP, one of the MW302 PAHs, can be metabolically activated by CYP1A1 to produce DNA adducts with ~ 10-40 times the genotox-icity of BaP.23 Alternatively, PAHs can induce the expression of CYP1 A1, which may be another vital reason for the enhanced toxicity. At low concentrations, BaP induces CYP1 A1 expression, which in turn enhances the metabolic activation of BaP by increasing the formation of B[a]PDE-N2-dG adducts, and often leads to an increase in cell toxicity and DNA damage in BEAS-2B cells.24 Therefore, higher CYP1A1 expression may be a key factor in the toxic effects of PAHs.

In our previous study, we investigated the relationship between nearly 100 PAHs and CYP1A1, including 16 priority-controlled PAHs, and found that the molecular weight and ring arrangement largely influenced the metabolic activity of CYP1A1 and the toxicity of PAHs. Among them, naphtho[2,1-a]pyrene (N21aP), a 6-ring HMW-PAH with MW302, showed strong CYP1A1 metabolic and mutagenic activity.25 N21aP has one more aromatic ring than BaP and is mainly derived from the high-temperature combustion of coal or wood.26 A previous study of particulate matter <2:5 lm in aerodynamic diameter (PM2:5) concentration in Beijing prelimi-narily showed that the N21aP was measured in the urban air samples at a level of one-third that of BaP.27 Simultaneously, a study also showed that N21aP was highly mutagenic, and its mutagenic-ity was comparable to that of BaP using an h1A1v2 cell line, a line of human B-lymphoblastoid.12 The mutagenicity risk of PAHs, depending on the angle-condensation of the "bay region" (phenan-threne structure) and the "fjord region" (pyrene structure), were also identified in our previous study, and PAHs with these struc-tures are more easily metabolized by CYP1A1 to form epoxide products, resulting in mutagenic effects.28 Therefore, the strong mutagenicity of N21aP is most likely caused by CYP1A1 metabolism. Meanwhile, a study found the induction of CYP1A1 by BaP was critical to the toxicity of BaP in BEAS-2B cells.24 Based on the structural similarity of BaP, N21aP probably also plays a role in the induction of CYP1A1; however, to date, no relevant studies have been reported.

Among the various regulatory pathways involved in the expression of CYP1A1 by PAHs, the AhR pathway stands out as the most well established,29 given that the majority of PAHs such as BaP, BkF, benzo[a]anthracene (BaA), BbF, functioned as AhR ligands to activate the AhR in HepG2 cells.30 The CYP1A1 promoter con-tains eight exogenous response elements (XREs) that are used by AhR activation,31 thereby underscoring the crucial role played by AhR in the regulation of the CYP1A gene. A previous study using BEAS-2B cells showed that BaP induces the expression of CYP1A1 through the AhR pathway.24 As a PAH, it needs to be investigated whether N21aP can also regulate CYP1A1 expression through the AhR pathway.

Transcriptional regulation and posttranscriptional regulation are two fundamental mechanisms implicated in the regulation of gene expression. Although transcriptional regulation primarily governs gene regulation, posttranscriptional regulation acts as an additional and consequential layer of regulation.32 Gene expression can be posttranscriptionally regulated through dynamic and reversible RNA modifications.33 Similar to posttranslational modi-fications of proteins, RNA posttranscriptional modifications play a crucial role in regulating significant aspects of RNA function.34 Recently, RNA methylation, particularly the N6-methyladenosine (m6A) modification, has emerged as a prominent area of research within the realm of posttranscriptional modifications. Among the numerous RNA modifications identified, m6A stands out as the most prevalent internal mRNA modification,35 with an average occurrence of 1-2 m6A methylation sites per 2,000 nucleotides in mammalian cells, such as HepG2 cells (a human liver cancer cell line).36,37 m6A-mediated mRNA stabilization is a vital part of the posttranscriptional regulation of gene expression.38 However, de-spite the identification of the involvement of m6A modification in regulating the expression of CYP enzymes, such as CYP1B1,39 the specific role of m6A modification in the regulation of CYP1A1 expression by PAHs remains unexplored. The presence of numer-ous m6A modification sitesonCYP1A1 mRNA40 implies that m6A modification may serve as an additional significant mechanism through which PAHs regulate CYP1A1 expression. Consequently, it is imperative to conduct further investigations to elucidate the role of m6A modification in PAH-induced CYP1A1 expression.

This study aimed to provide a more comprehensive under-standing of CYP1A1 in the metabolic activation of N21aP as a kind of HMW-PAH, which is of particular interest owing to their enhanced stability and greater affinity for adsorption in PM2:5 and subsequent likelihood of higher accumulation in lung tissue. In particular, apart from the well-known AhR pathway, a novel mechanism of m6A modification underlying CYP1A1 expression regulated by N21aP was further explored. This study not only contributes to a novel comprehension of the toxicity of HMW-PAHs but also offers a better understanding of the significant role of CYP1A1 induction in the health risks of PAHs, even at expo-sure levels in realistic environments.

Materials and Methods

Chemicals, Reagents, and the Synthesis of N21aP

The chemicals and reagents used in the present study are listed in Table S1. N21aP was synthesized according to our previous study.25 The specific synthesis diagram of N21aP is detailed in Figure S1. Briefly, to a solution of pyrene (1 g) and bis(pinacolato) diboron (B2pin2; 0:69 g) in hexane (5 mL), a mixture of (1,5-cyclooctadiene)(methoxy)iridium(I) dimer ([{Ir(m-OMe)codg2]; 30 mg), 4,40-di-tert-butyl-2,20-bipyridine (dtbpy; 24 mg), B2pin2 (50 mg), and hexane (2:5 mL) was added and stirred at 80°C for 5 h under a nitrogen atmosphere, which afford 4,4,5,5-tetramethyl-2-pyren-2-yl-[1,3,2]dioxaborolane[2-(Bpin) pyrene] (P1-1). Then tetrahydrofuran (THF; 5 mL) and water (2 mL) containing sodium carbonate (Na2CO3; 383 mg) was added to 2-bromobenzaldehyde (197 mg) and P1-1 (500 mg), and the mixture was purged with nitrogen gas. The resulting solution was combined with palladium

[tetrakis(triphenylphosphine)palladium (PPh3)4; 123 mg], refluxed at 90°C for 11 h, and kept still overnight to afford 2-(pyren-2-yl) benzaldehyde (P1-2). After these steps, to a solution of methoxymethyltriphenyl-phosphonium chloride (3 g) in dry THF (27 mL), a solution of potassium tert-butoxide (t-BuOK; 1 M) in THF, 8:82 mL was added at 0°C. After stirring for 30 min, a solution of P1-2 (600 mg) in THF (27 mL) was added dropwise and stirred at 20°C for 2 h to afford 2-(2-(2-methoxyvinyl) phenyl) py-rene (P1-3). Finally, to a solution of P1-3 (300 mg) in dichlorome-thane (CH2Cl2; 5 mL), 5 drops of methanesulfonic acid (CH3SO3H) were added. Then, the reaction mixture was stirred at 20°C overnight, which was then purified by chromatography on a silica gel column (200-300 mesh silica gel, Synthware) to afford N21aP.

The purity of synthesized N21aP was identified by high-performance liquid chromatography (HPLC, Ultimate 3000) and is presented in Figure S2. Briefly, kromasil 100-5-C18 column (4:6 mm × 250 mm, 5 lm, AkzoNobel) protected by a C18-guard column (4:6 mm× 10 mm, 5 lm, ThermoFisher) was chosen for the separation at 35°C and detected via ultraviolet (UV) detection at 290 nm. The acetonitrile (A) and water containing 0.2% (vol/ vol) of acetic acid (B) were respectively used as the mobile phases. The mobile phase was pumped at a flow rate of 1 mL=min with an isogradient elution of 88% A-12% B for 28 min.

PAH Detection and Source Analysis

Preparation of the atmospheric PM2:5 samples. PM2:5 samples were collected from December 2018 to November 2019 in 13 cities in Jiangsu, a province located in east China, a choice which was derived from one of the annual routine programs of air qual-ity monitoring by Jiangsu Provincial Center for Disease Control and Prevention (Jiangsu CDC), China.41 Atmospheric PM2:5 sampling was conducted for ~ 7 consecutive days at each monitoring site of each month, with a sampling time of not less than 20 h per day. The height the PM2:5 sampler was placed from the ground was set at the height range of the normal activities of the crowd, and the prescribed requirement was 15 m. A glass fiber mem-brane suitable for the detection of organic matter was used in PM2:5 sampling. The samples were stored in a freezer at - 20 °C for subsequent analysis.

Preparation of the particulate matter of fuel oil, biomass, and coal. The samples of fuel oil (93 octane petrol, 97 octane pet-rol, and light diesel oil)42 and biomass (wheat straw, rice straw, and soybean straw)43 combustion were provided by Jianmin Chen (Department of Environmental Science & Engineering, Fudan University, China). Briefly, the particles of fuel oil were collected using a wick burner positioned under a semi-enclosed cylindrical glass cover that provided an ample particle-free air supply. Each combustion experiment of the different fuels was conducted for 10 min and repeated at least three times. For the analysis of organic carbon and PAHs, particles emitted from the combustion of fuel oils were captured using a quartz fiber filter (Whatman).42 For the biomass fuel samples, the combustion chamber consisted of a steel cube and a conical platform. The air inlet incorporated filters to remove particles from incoming ambient air. A 2:25-m2 square combustion platform was set at the center of the chamber's bottom and was slightly elevated above the top of the filter to ensure ample expo-sure of biomass straw to fresh air during combustion. An exhaust tube was situated at the top of the combustion unit, featuring a port near its end for monitoring and sampling purposes. The biomass straws (rice, wheat, and soybean) were collected from suburban regions in Jiangsu Province, totaling 600 g of biomass residue placed horizontally on the square platform to facilitate ignition. During combustion, biomass undergoes 5-10 min of flaming and smolder-ing phases. PM2:5 samples were collected by using 47-mm quartz filters (Tissuquartz, Pall Corp.) at a flow rate of 16:7 L=min.43

PM2:5 samples from coal combustion, including bituminous coal and anthracite coal, were prepared by our laboratory. Briefly, the PM2:5 sampler (TH-150 F, Wuhan Tianhong Environmental Protection Industry Co., Ltd.) was placed at a distance of 1 m from the combustion point at a sampling height of 1:5 m. After ignition of bituminous and anthracite coal (from Taiyuan, China) for 30 min via charcoal stove (XINAISHI), PM2:5 was collected by pre-baked glass fiber film (90-mm diameter, Membrane Solutions) and frozen at -20 °C for subsequent analysis.

PAH Extraction, Detection, and Analysis

The PM2:5 samples were extracted and analyzed according to a pre-vious method.44 One-sixteenth of the sampled PM2:5 membranes were extracted in an ultrasonic bath with dichloromethane (DCM): acetonitrile (ACN) (90:10, vol/vol) three times. The extracts were filtered through syringe filters (0:22 lm, Biosharp) to remove in-soluble particles and were dried by nitrogen sweeping and reconsti-tuted in DCM. Subsequently, the extracts were submitted to commercial solid phase extraction (SPE) cartridges SupelMIP SPE-PAHs (Bellefonte) for clean-up, and finally 50 lL of DCM was added for resolution.

Gas chromatography-triple quadrupole mass spectrometry (GC-MS/MS, TSQ 8000, ThermoFisher) was performed to detect PAHs, including MW302-PAHs, BaP, fluorene (Flu), pyrene (Pyr), and BghiP. The redissolved extracts were separated by a DB-EUPAH capillary column (15 m length×0:25-mm inner diameter ×0:25-mm film thickness, Agilent Technologies) with helium carrier gas at 1:8 mL=min. The temperature was initially held at 135°C for 1 min, raised to 280°C at 20 min, maintained for 30 min, raised to 290°C at 5 min, and maintained for 13 min, raised to 300°C at 15 min, and maintained for 9.5 min. The injector temperature was set at 300°C. A 2-lL sample was injected using the splitless injection mode. The MS was operated in electron ioniza-tion with electron energy of 70 eV. The ion source and transmis-sion line temperatures were all set at 300°C. Argon (99.999%) was used as the collision gas. For multiple reaction monitoring optimi-zation, the precursor ion was dissociated to produce the quantifier ion and qualifier ion. The quantitative ion pairs for LMW-PAHs and HMW-PAHs were 252 m=z to 250 m=z with 27 eV and 302 m=z to 300 m=z with 30 eV, respectively. The qualitative ion pairs for LMW-PAHs and HMW-PAHs were 252 m=z to 248 m=z with 30 eV and 302 m=z to 298 m=z with 35 eV, respectively. The recovery and detection limits of PAHs are listed in Table S2. The solvent blank was analyzed to assess potential carryover and sys-tem contamination. All peaks in the solvent blanks were required to be below the detection limits of all analyses.

Several methodologies, including principal component analysis (PCA), UNMIX, and positive matrix factorization, have been developed to enhance the source attribution of PAHs owing to their unique emission signatures.45 In this study, we employed PCA to explore the potential sources of MW302-PAHs in PM2:5, leveraging the distinct compositions of these compounds. To con-duct the analysis, we used the Xcalibur software and targeted ion pairs at 302 m=z to 300 m=z for the chromatographic profiling of the 34 MW302-PAHs in laboratory-combusted fuel oil, biomass, coal, and PM2:5 samples. To ensure comparability, we normal-ized the peak area of each MW302-PAH by dividing it by the total peak area of all 34 PAHs, reflecting the relative abundance of each compound within the samples. The processed data was then subjected to PCA using Simca software (version 14.1; Umetrics), with a focus on the MW302-PAH content in PM2:5 samples to differentiate between characteristic source profiles.

The diagnostic ratio (DR) of PAHs is a widely recognized and versatile tool for source identification across various environments.46 This method offers numerous advantages, with its ease of interpretation being paramount compared with other available methods. DRs related to PAHs from various sources have been well established and widely applied.47 For instance, the Flu/ Flu+Pyr ratio discriminates between combustion and petroleum sources, with a petroleum limit at 0.40; ratios between 0.40 and 0.50 indicate fossil fuel combustion, whereas those >0.50 sug-gest grass, wood, or coal combustion.48 Similarly, the BaP/BghiP ratio separates traffic from nontraffic sources, with a nontraffic threshold at 0.60, and ratios greater than this value denote traffic-related emissions.49 However, insufficient attention has been given to HMW-PAHs in current environmental studies, resulting in a lack of reference values for DRs based on HMW-PAHs. To address this gap and improve specificity, the present study aimed to establish key DRs for MW302-PAHs by analyzing laboratory-combusted fuel oil, biomass, and coal to trace the origin of HMW-PAHs in PM2.5. The analysis focused on the peak areas of MW302-PAHs (peaks 1-34), selecting a peak area ratio that effectively distinguished between the three combustion sources, which was then adopted as the DR value. This screened DR was subsequently employed for tracing the PM2.5 samples.

In Vivo Mouse Model Assay

For the animal experiments, 8-wk-old male wild-type (WT) C57BL/6 mice (22-25 g) were purchased from Nanjing Medical University. Male mice were selected for the present study to mitigate the impact of the estrous cycle' s stage on female mice and the effect of estrogen on the experimental outcomes.50 Cyp1a1 knockout (KO; Cyp1a1~ ~) mice with a C57BL/6 background were entrusted to Bangyao Bio for construction using CRISPR-Cas9 technology in our previous study.25 Briefly, the mouse Cyp1a1 gene comprises seven exons, with protein translation initiated from exon 2 and concluding at exon 7. Thus, exon 2 was selected as the target region. Two guide RNAs (gRNAs) were designed near exon 2, and the one with superior prediction scores for minimal off-target effects was chosen. The sequences of the selected gRNAs were as follows: Cyp1a1 -single guide RNA1 (sgRNA1): GCTGTCACCGTATTCTGCCTTGG; Cyp1a1^- sgRNA2: CCAACGTTATGACCATGATGACC. Cas9 mRNA and sgRNAs were mixed and injected into the cytoplasm of fertilized mouse eggs to generate specific mouse embryonic cells. Surviving fertilized eggs were subsequently transferred into the fallopian tubes of pseudopregnant mice to produce offspring. All animals were housed in the Animal Core Facility of Nanjing Medical University (Nanjing, China). The animal procedures were approved by the insti-tutional review board of Nanjing Medical University (No. 2003001). The animals were allowed to ingest water and food ad libitum and were housed ina12h:12h light/dark cycle with a relative humidity of 55% ± 10% and a temperature of 22 ± 2°C. The feeding and the experimental operations on animals were in accordance with the animal welfare requirements. At the end of the N21aP exposure, the mice were euthanized with carbon dioxide CO2 to collect serum, bron-choalveolar lavage fluids (BALFs), and tissues. The samples were stored at - 80 °C for subsequent analysis.

This study included three independent animal experiments. In experiment 1, WT mice (n = 3 per group) were administered three concentrations of N21aP (0.02, 0.2, 2 mg/kg) through intratra-cheal instillation to evaluate a dose-dependent induction of lung CYP1A1. Phosphate-buffered saline (PBS) wasused as acontrol at an equivalent volume. The lower exposure of N21aP was consid-ered based on the environmental levels of N21aP from the previous studies in Beijing,27 as well as the reference to the standards of BaP (GB3095-2012). Twenty-four hours after exposure, the mice were euthanized and the lung tissues were collected. In experiment 2, a subacute animal experiment was performed to evaluate the toxicity of N21aP. WT and Cyp1a1-/- mice (n = 5 per group) were treated with 2 mg/kg N21aP through intratracheal instillation twice a week for 4 wk, and the same volume of PBS was used as a control. After 24 h of the final whole-body plethysmography (WBP) mea-surement, the mice were euthanized and the serum, BALFs, and lung tissues were collected. In experiment 3, three concentrations of N21aP (0.02, 0.2, 2 mg/kg) were used to investigate whether inflammatory outcomes were dose dependent. WT mice (n = 3 per group) were exposed to N21aP by continuous 6-d intratracheal instillation, and the same volume of PBS was used as a control. Twenty-four hours after the last treatment of N21aP, the mice were euthanized and the serum, BALFs, and lung tissues were collected.

Assessment of Lung Function in Mice

The respiratory function in awake and freely moving mice was analyzed using WBP (Buxco Electronics Inc.) as described previ-ously.51 One measurement of mice lung function was performed 24 h after the last treatment of N21aP. Briefly, WT or Cyp1a1-/- mice (with or without N21aP treatment, n = 5/group) were accli-mated for 20 min in the cavity. Then, unrestrained mice were monitored for 10 min. Respiratory frequency (fR), enhanced pause (Penh), pause (PAU), and expiratory time (Te) were evaluated.

Enzyme-Linked Immunosorbent Assay of Inflammatory Cytokines

The BALFs and serum of mice were collected for enzyme-linked immunosorbent assay (ELISA). Whole blood taken by eyeball re-moval were collected from mice (n = 5/group) and allowed to stand at room temperature for 2 h before being centrifuged at 3,000 ×g for 30 min. The supernatant was stored at - 80 °C for subsequent analysis. BALF of mice was collected according to our previous study.52 In brief, the left bronchus was ligated, and the right lung was infused with 300 lL of ice-cold PBS three consecutive times (n = 5/group). Subsequently, the collected BALF was centrifuged at 400 × g for 10 min, and the supernatant was stored at - 80 °C for further use.

The levels of the inflammatory cytokine interleukin-1 b (IL-1b) or IL-6 (n = 3/group) were measured using a mouse IL-1b ELISA kit (HS1069-Mu, Hengyuan Biotech Co. Ltd.) or a mouse IL-6 ELISA kit (HS1057-Mu, Hengyuan Biotech Co. Ltd.) according to the manufacturer's instructions. The absorbance was measured at 450 nm using a microplate reader (Infinite M200 PRO, TECAN). The sample concentration was calculated based on the standard curve.

Hematoxylin-Eosin Staining of Lung Tissue from Mice

Consistent with prior intratracheal instillation experiments on mice,53,54 the right caudal lobe was excised for Western blotting and quantitative real-time polymerase chain reaction (qRT-PCR) analysis. The left lung lobe was fixed by perfusion with 4% paraformaldehyde (PFA, BL539A, Biosharp), and 5-lm-thick paraffin-embedded sections of mouse lung tissues were stained with a hematoxylin-eosin (H&E) Staining Kit (C0105S, Beyotime) according to the manufacturer's protocol. H&E sections were exam-ined using slide scans (Pannoramic Scan and Viewer, 3D Histech) to view the images.

RNA Extraction and qRT-PCR

Mouse lung tissues were isolated using RNA Isolater Total RNA Extraction Reagent (R401-01, Vazyme) according to the manufacturer's protocol. The concentration and quality of RNA were assessed using NanoDrop (ThermoFisher), with an optical density (OD)260/OD280 ratio falling within the range of 1.8-2.0 deemed acceptable. Complementary DNA (cDNA) was synthesized using HiScript II Q RT Supermix for qPCR (R222-01, Vazyme). To analyze the relative mRNA expression levels, qRT-PCR was performed using ChamQ Universal SYBR qPCR Master Mix (Q711-02, Vazyme) with a LightCycler 96 (Roche, Switzerland). The PCR program is presented in Table S3. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression was used as the internal control; subsequently, relative mRNA expression levels were calculated using the 2-4 Ct method. The primer sequences in this article were commissioned to be synthesized by Generay and are presented in Table S4.

Western Blotting

Mouse lung tissues were homogenized in lysis buffer (P0013, Beyotime) to extract total protein. The protein concentrations were detected by a Bicinchoninic Acid Protein Quantification Kit (E1 12-01, Vazyme). Proteins (50 lg) were subsequently transferred to a polyvinylidene fluoride (PVDF) membrane (ISEQ00010, 0:2-lm, Millipore) after separation on 10%/12.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Nonspecific binding sites were blocked for 1 h with 5% bovine serum albumin (BSA, ST023, Beyotime) at room temperature. The membranes were then incubated with primary antibodies overnight at 4°C, followed by incubation with peroxidase-conjugated secondary antibodies (Jackson ImmunoResearch) for 1 h at room temperature. After a 2-h wash with tris-buffered saline (ST663, Beyotime) con-taining 0.1% Tween-20 (TBST), the bands were imaged using the Enhanced Chemiluminescence Kit (E411-05, Vazyme) and detected with a Chemiluminescent Imaging System (Tanon). Finally, the relative quantification of the detected protein was ana-lyzed using Image J software (version 2.1.0),55 with GAPDH expression serving as the internal reference. Detailed information on the antibodies used in this study is listed in Table S5.

S9 Preparation and CYP1A1 Activity Assays

WT and Cyp1a1-/- mice (n = 3 per genotype) were employed for the study, with their liver S9 fractions isolated following a 3-d treatment regimen. Mice were administered 100 mg=kg b-naphthoflavone (b-NF, N3633, Sigma-Aldrich), a known in-ducer of CYP1A1 enzyme activity, via continuous gavage. The livers of the mice were used for extraction. Initially, the livers were continually rinsed with a precooled 0:15 mol=L potassium chloride (KCl) solution to eliminate hemoglobin-mediated inhibi-tion of microsomal enzyme activity. Liver homogenates were then prepared by adding 3 mL of a 0:1-mol=L KCl solution per gram of liver using the Tissuelyser (Grinder-48, GallopTech). Subsequently, the homogenates were centrifuged at 9,000 × g for 15 min at 4°C to obtain the S9 fraction in the supernatant. A por-tion of this fraction was analyzed for CYP1A1 activity, and the remainder was partitioned and stored at -80 °C for further use.

CYP1A1 activity was evaluated using the ethoxyresorufin-O-deethylase (EROD) assay kit (GMS18017.1, Genmed Scientifics Inc.) according to the manufacturer's instructions. In a black 96-wellplate, 195 lL of buffer, 25 lL of reaction solution, and 5 lLof substrate solution were mixed. Subsequently, 25 lL of either the standard solution or S9 samples were added to each well, followed by a 10-min incubation at 37°C in the dark. Fluorescence intensity was measured with a microplate reader (Infinite M200 PRO, TECAN) at an excitation wavelength of 530 nm and an emission wavelength of 590 nm. The activity values of the S9 samples were then determined based on the standard curve.

Ames Test

The Ames test was performed according to our previous study.25 Briefly, the frozen Salmonella typhimurium strain TA98 strain

(BeNa Culture Collection) was inoculated into nutrient broth medium and incubated at 37°C on a shaker at 180 revolutions per mi-nute (rpm) for 10 h to reach logarithmic growth. Subsequently, 100 lL of TA98 bacterial solution, 100 lL of N21aP (0.5-, 1-, 2.5-, 5-, 10- lL/plate) and 100 lL of 10% S-9 mixture of WT or Cyp1a1-/- mice were added to the top agar. Then, the top agar was quickly mixed and poured into the bottom petri dish, and was incubated aseptically at 37°C for 48 h. The results of mutagenesis were judged according to the frequency of revertant colonies.

Reduced Glutathione Conjugate of Epoxides ofN21aP

The epoxide was reported to conjugate with reduced glutathione (GSH), form adducts with sulfhydryl groups in cellular proteins or be further converted to reactive quinones, diepoxides, and diol epoxides.56 The conjugation of GSH was performed as previously described.57 Briefly, GSH capture was prepared in a reaction mixture (100 lL) with 1 mM GSH, nicotinamide adenine dinucleotide phosphate (NADPH) regenerating system (011700.02, IPHASE), and 2% liver S9 of WT or Cyp1a1-/- mice (the same S9 used as above) and 20 lM N21aP. Then, the reaction was incubated at 37°C for 1 h and was terminated by adding 100 lL of ice-cold ACN. Subsequently, the incubated mixture was centrifuged at 12,000 rpm for 30 min at 4°C, and the supernatant was analyzed by HPLC-MS/ MS (Q Exactive HF, ThermoFisher). The redissolved extracts were separated by a Hypersil GOLD C18 column (2:1 mm× 100 mm, 1:9 lm, ThermoFisher) at 20°C. Methanol was increased from 5% to 95% in 16 min by using methanol and water as mobile phases. Positive electrospray ionization mode was used for mass spectrome-try, and the spray voltage was 3:5 kV.

In Vitro Cellular Assay

Cell culture and treatment. Human bronchial epithelial cells (HBEs), human embryonic kidney (HEK)-293 cells, and HepG2 cells were purchased from American Type Culture Collection (ATCC). HBE cells were cultured in Dulbecco's Modified Eagle Medium (DMEM, BC-M-005, Biochannel). HEK-293 cells and HepG2 cells were cultured in Eagle's Minimum Essential Medium (EMEM, 320-006, WISENT). All media were supplemented with 10% fetal bovine serum (FBS; BC-SE-FBS01, Biochannel), 100 U=mL penicillin, and 100 lg=mL streptomycin (C0222, Beyotime). Cells were cultured at 37°C and 5% CO2 and no myco-plasma contamination was detected (Figure S3) using the Myco-Lumi Luminescent Mycoplasma Detection Kit for High Sensitivity Instrument (C0298S, Beyotime) according to the manufacturer's instructions. The chemiluminescence meter (BERTHOLD, Centro LB 960) was used for the detection. Notably, the ratio B/A >1:2 indicates the contamination of mycoplasma, and the ratio B/A <0:9 indicates no mycoplasma contamination.

N21aP or BaP was added to the culture medium according to the concentration of treatment. To inhibit methylation, HBE cells were pretreated with 3-deazaadenosine (DAA, a methylation in-hibitor, 100 lM) for 6 h or S-adenosylhomocysteine (SAH, an in-hibitor of the METTL3-METTL14 heterodimer complex, 20 lM) for 24 h. Then, the experiments were conducted after 24 h of cell culture.

Immunofluorescence analysis. Treated cells were fixed with 4% PFA for 30 min. Nonspecific binding sites were blocked for 1 h by 2% BSA at room temperature. Subsequently, primary antibodies were applied to cells overnight at 4°C. Samples were incubated with fluorophore-labeled secondary antibody for 1 h at 37°C. Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI; C1005, Beyotime) for 10 min. The images were detected using a laser confocal microscope (Zeiss, LSM700B). Detailed information on the antibodies is listed in Table S5.

Nuclear and cytoplasmic protein extraction. A nuclear and cytoplasmic protein extraction kit (P0027, Beyotime) was employed to separate the nuclear and cytosolic fractions in accord-ance with the manufacturer's instructions. HBE cells (5× 106 cells) were treated with1 lMN21aP for0, 12, and24h, then rinsed with PBS and harvested by centrifugation. The cytoplasmic frac-tion was isolated by adding Cytoplasmic Extraction Reagent A with phenylmethylsulfonylfluoride (PMSF), followed by a 15-min incubation on ice. Subsequently, Cytoplasmic Extraction Reagent B was added, and the mixture was centrifuged at 16,000 × g for 5 min to obtain the cytoplasmic proteins. For the nuclear fraction, Nuclear Protein Extraction Reagent with PMSF was added to the pellet, incubated on ice for 30 min, and then centrifuged at 16,000× g for 10 min to collect the nuclear proteins. Tubulin and Lamin B1 were used as internal references for the cytoplasmic and nuclear proteins, respectively, as previously reported.58-62

Cell transfection. Small interfering RNA (siRNA) targeting METTL14 (si-METTL14) and IGF2BP3 (si-IGF2BP3) and nega-tive control (NC) siRNA (si-NC) were designed and synthesized by Hanbio. The detailed sequences are presented in Table S6. The METTL14 overexpression plasmid (METTL14-OE) was synthesized by Generay. The detailed sequences are presented in Table S7. Briefly, HBE cells were seeded into 6-well plates, cul-tured to 60% confluence, and then transfected with siRNA (50 nM) or plasmids (2 lg/well) using Lipofectamine 3000 (L3000015, Invitrogen) transfection reagent. The transfected cells were incubated at 37°C for 48 h, and the transfection efficacy was validated using qRT-PCR and Western blotting analysis.

Dual-luciferase reporter assay. The KH domain-truncated IGF2BP3 plasmid (IGF2BP3-DKH) and CYP1A1 WT or mutant plasmids were synthesized by NanJing XinJia Medical Technology Co. Ltd. The detailed DNA sequences used for construction are presented in Table S7. HEK 293 cells were cultured in 96-well plates at 6,000 cells/well for 24 h. Subsequently, the cells were cotransfected with plasmids containing WT or mutant CYP1A1 and si-METTL14 or KH domain-truncated IGF2BP3 plasmids according to the transfection protocol mentioned above. The relative luciferase activity was measured by a dual-luciferase reporter assay according to the manufacturer's protocol (RG027, Beyotime). Briefly, cells were fully lysed and centrifuged, and the supernatant was taken for further testing. Then, 100 lL of the supernatant was added to a black 96-well plate, and the firefly lu-ciferase activity and Renilla luciferase activity were measured by using a chemiluminescence meter (BERTHOLD, Centro LB 960) and expressed in respective relative light units.

AhR-dependent luciferase reporter assay. A luciferase reporter gene plasmid containing the AhR-specific XRE was purchased from Yeasen (No. 11537ES03) to detect human AhR activity in HepG2 cells owing to the presence of a sufficient amount of endoge-nously occurring AhR. Briefly, HepG2 cells were seeded into 6-well plates, cultured to 60% confluence, and then transfected with plasmids (2 lg/well) using Lipofectamine 3000 (L3000015, Invitrogen) trans-fection reagent. Forty-eight hours posttransfection, HepG2 cells were seeded into a 24-well plate at a density of 5 × 105 cells/well and exposed to varying concentrations of N21aP (0, 0.001, 0.01, 0.1, 0.25, 0.5, 1, 5, and 10 lM) or BaP (0, 0.01, 0.1, 1, 5, 10, and 25 lM) for 24 h. Luciferase activity was measured using a Dual-Luciferase Reporter Gene Assay Kit (RG027, Beyotime), and renal luciferase activity was standardized. To investigate the effect of METTL14 on AhR activity, after transfection of the AhR plasmid in HepG2 cells for 24 h, a portion was further transfected with si-METTL14 (50 nM)orMETTL14-OE plasmid(2 lg/well) for 48 h,and another portion was treated with N21aP (1 lM) for 24 h as a positive con-trol, and luciferase activity was measured. The transfection efficacy of METTL14 was validated using Western blotting analysis.

ChIP. The ChIP assay was conducted using the Magna ChIP A/G kit (Magna0017, Millipore), adhering to the protocols pro-vided by the manufacturer. HBE cells, at a density of 1 × 107 cells, were treated with 1 lM N21aP for 24 h and subsequently cross-linked with 1% formaldehyde at room temperature for 10 min. The cross-linking reaction was halted by the addition of 10× glycine for 5 min. Cells were harvested from each dish using a sterile cell scraper, and the collected cells were centrifuged at 800×g for 5 min at 4°C to pellet them. Subsequently, the cell pellet was resus-pended in nuclear lysis buffer. To shear the chromatin into frag-ments of optimal size, the lysate underwent sonication, with a Bioruptor Pico (Diagenode SA) employed for 15 cycles of 30-s on and 30-s off periods. This process resulted in DNA fragments rang-ing from 250 to 1,000 bp. For immunoprecipitation, an anti-AhR antibody (1:50, #83200, Cell Signaling Technology) was used and incubated overnight at 4°C, with immunoglobulin G (IgG) serving as the negative control. The DNA was then purified by using spin columns. Then, the promoter regions of CYP1A1 were quantified by qRT-PCR, as detailed in Table S3. Primer sequences, sourced from a previous study,63 were synthesized by Generay, with detailed information provided in Table S8.

RNA m6A dot blotting assay. Total RNA was isolated by using RNA Isolater Total RNA Extraction Reagent (R401-01, Vazyme), the RNA was diluted by using RNase-free double-distilled water (ddH2O) to 500 ng=lL, 250 ng=lL, and 125 ng=lL. Then, the diluted RNA was subjected to heat shock at 95°C for 3 min and quickly placed on ice. Dot samples were loaded onto a nitrocellu-lose membrane (RPN303B, 0:45-lm, GE Amersham), UV cross-linker (CX-2000, UVP) cross-linked at 120,000 mJ=cm2 for 1 min and dyed with 0.05% methylene blue (MB) solution for 5 min. After being blocked with 5% nonfat milk for 1 h, the membrane was incubated with m6A antibody (1:1,000, Thermofisher) at 4°C overnight and incubated with horseradish peroxidase-conjugated goat anti-rabbit IgG (1:2,000, Jackson ImmunoResearch) at room temperature for 1 h. Finally, the membrane was imaged using the Enhanced Chemiluminescence Kit (E411-05, Vazyme) and visual-ized by a Chemiluminescent Imaging System (Tanon). MB stain-ing was performed as an internal reference control.

Methylated RNA Immunoprecipitation-qPCR. m6A modifi-cation in CYP1A1 was predicted using the sequence-based RNA adenosine methylation site predictor (SRAMP) online tool.40 Subsequently, specific primer targeting the predicted high-confidence m6A sites were designed for methylated RNA immu-noprecipitation (MeRIP)-qPCR. MeRIP-qPCR was performed using the EpiQuik m6A RNA Enrichment (MeRIP) Kit (P-9018-24, Epigentek). Briefly, total RNA was isolated from N21aP-treated HBE cells or mouse lung tissues by using RNA Isolater Total RNA Extraction Reagent (R401-01, Vazyme). Then, RNA (2 lg) was subjected to m6A or IgG beads (5 lg) at 4°C with rotation for 1.5 h. The enriched m6A modifications were then measured by qRT-PCR (Table S3), and the corresponding m6A enrichment in each sample was calculated using normalized input. The primer sequences were commissioned to be synthe-sized by Generay and are presented in Table S8.

Determination of RNA stability. HBE cells transfected with METTL14 or IGF2BP3 siRNA were seeded in a 6-well plate (3 × 105 cells/well). Then, the cells were treated with 5 lg=mL ac-tinomycin D (Act D) to inhibit mRNA transcription. Cellular RNA was extracted at 0 h, 2 h, and 4 h after Act D intervention by using RNA Isolater Total RNA Extraction Reagent (R401-01, Vazyme). The mRNA levelofCYP1A1 was measuredbyqRT-PCR.

RNA-binding protein pull-down assay. Biotin-labeled in vitro transcription CYP1A1 was synthesized by GenePharma. RNA pull-down assays were performed using an RNA-Protein PullDown Kit (P0201, Geneseed Biotech). Briefly, the proteins were extracted from HBE cells using immunoprecipitation (IP) lysis buffer. Then, biotinylated CYP1A1 (50 pmol) was captured with streptavidin magnetic beads at 4°C for 30 min and incubated with the cell lysates at 4°C with rotation overnight. Then, the enriched protein was eluted and obtained, and the enriched proteins were an-alyzed byWestern blotting.

RNA-binding protein immunoprecipitation. The RIP assay was carried out in HBE cells transfected with si-METTL14 or si-NC by using the Magna RIP RNA-Binding Protein Immunoprecipitation (RIP) Kit (17-700, Millipore Corporation) following the manufacturer's instructions. Briefly, transfected cells (1 × 107 cells) were collected and lysed in RIP lysis buffer with protease and RNase inhibitors. IGF2BP3 or IgG antibody (5 lg) was added to the cleared lysates and incubated with mag-netic beads at 4°C with rotation overnight. Then, proteinase K buffer was used to digest and purify the precipitate. Finally, coimmunoprecipitated RNAs were extracted and measured by qRT-PCR (Table S3) and normalized to input. The primer sequences are commissioned to be synthesized by Generay and are presented in Table S8.

Statistical Analysis

All data are presented as the mean ± standard deviation (SD). Student t-test and one-way analysis of variance (ANOVA) followed by Dunnett's or Tukey's multiple comparison post hoc test were used to analyze differences between two or multiple groups by SPSS (version 26.0; IBM Inc.). p < 0:05 indicated that the results were statisti-cally significant. GraphPad Prism software (version 9; GraphPad Software Inc.) was used to visualize the related chart results.

Results

Source and Characteristics of N21aP in PM2:5

Similar to the profile of PM2:5 in China,64,65 the concentration of N21aP showed obvious seasonal variation, with the highest in winter and the lowest in summer in 13 cities in Jiangsu Province (Figure 1A), and the annual average concentration of N21aP was approximately one-third to one-half that of BaP (Figure 1B; Table S9 and Excel Table S1). Notably, as an HMW-PAH,

N21aP exhibited a stronger correlation with PM2:5 in winter during the winter months (December-February) compared with BaP (Figure 1C; Excel Table S2). To explore the source of N21aP in PM2:5, MW302-PAHs from laboratory-combusted fuel oil, bio-mass, and coal were detected, normalized according to the peak area of 34 MW302-PAHs, and processed for PCA, which could basically divide these sources into three major categories (Figure 1D,E; Excel Table S3), suggesting the MW302-PAHs composi-tion of the same pollutant source exhibited high similarity and PCA effectively distinguished between these different sources. Taking winter PM2:5 (December-February) as the representative, peak areas related to MW302-PAHs in winter PM2:5 samples were measured and PCA was used to compare the composition of MW302-PAHs with those generated in-lab combustion of fuel oil, biomass, and coal. These winter PM2:5 samples were dis-tinctly separate from the three combustion sources, forming their own cluster, suggesting that MW302-PAHs in PM2:5 originated from mixed sources (Figure 1F; Excel Table S4).

Furthermore, different in-lab-combusted fuel oil, biomass, and coal were distinguished by DR analysis. The DR method is highly favored for determining PAH sources.47 However, many of the existing DRs are derived from LMW-PAHs. Current environmen-tal studies have not adequately addressed HMW-PAHs, leading to a lack of reference values for DRs based on HMW-PAHs. Thus, considering that the conventional DR method is not commonly employed for MW302-PAH tracing, to address this gap and improve specificity, the present study identified key ratios for the MW302-PAHs by analyzing in-lab-combusted fuel oil, biomass, and coal to identify the origin of N21aP. Two groups of MW302-PAHs (peak 29/peak 28, peak 23/peak 22, corresponding to the peaks of MW302 in Figure 1D and Figure S4, and peak 29 repre-sented the N21aP) were selected to identify the different feature sources by ratio analysis. Figure 1G showed that the two sets of MW302-PAH ratios can basically distinguish three major types of

feature sources, where peak 29/peak 28 and peak 23/peak 22 ratios were assigned into three groups: peak 29/peak 28 > 1:0 or peak 23/peak 22 > 1:5, indicating a biomass combustion source; peak 29/peak 28 between 0.7 and 1.0 or peak 23/peak 22 between 1.2 and 1.5, indicating a fuel oil combustion source; and peak 29/ peak 28 < 0:7 or peak 23/peak 22 < 1:2, implying a coal combus-tion source. Subsequently, given the two sets of DRs (peak 29/ peak 28 and peak 23/peak 22 ratios) effectively distinguished between different combustion sources, they were applied for traceability analysis of HMW-PAHs in the PM2:5 samples. A comparison of peak ratios suggested diverse sources with more samples similar to the profile of fuel oil combustion and coal combustion sources (Figure 1H; Excel Table S5). Meanwhile, the results of the DR method for LMW-PAHs suggested that the sources of LMW-PAHs were a mixture of fossil fuel and coal, which was manifested in the Flu/(Flu+Pyr) values mostly between 0.4 and 0.7, whereas the BaP/BghiP values were in a more scattered distribution (Figure 1I; Excel Table S6).

CYP1A1, Respiratory Function, and Lung Inflammation in Mice Exposed to N21aP

Here, Cyp1a1 KO mice (i.e., Cyp1a1-/-) were first constructed and verified in either CYP1A1 protein (Figure 2A) or mRNA (Figure 2B). Then, the mice were given 2 mg=kg N21aP twice a week for 4 wk through intratracheal instillation, as described in experiment 2. As shown in Figure 2C, WT mice exposed to N21aP exhibited greater respiratory frequency, Penh, PAU, and expiratory time than control mice; by contrast, there were no significant differences in these outcomes between exposed and control Cyp1a1-/- mice. H&E staining revealed damaged alveolar structures, thickened alveolar walls, and numerous infiltrated inflammatory cells in WT mice, which was also mostly alleviated in Cyp1a1-/- mice (Figure 2D). In addition, N21aP-exposed WT mice exhibited lung inflammation, as indicated by the higher levels of IL-1b in BALF and serum and the higher expression of IL-6 and TNF-a, as well as the histone variant c-H2AX (a DNA damage marker), in the lung tissues. These inflammation-related outcomes were not evident in N21aP-exposed

Cyp1a1-/- mice (Figure 2E,F). Furthermore, the inflammation-related outcomes appeared to be N21aP dose dependent, based on the expression levels of IL-1b and IL-6 in BALF and serum (Figure S5A-D) and mRNA and the protein expression of IL-1b, IL-6, and TNF-a inthe lung tissues ofmice (Figure S5E,F).

Effects of CYP1A1 on Activation of N21aP in Vitro

To further explore the critical role of CYP1A1 in metabolic activa-tion of N21aP in vitro, first, EROD activity of the mouse S9 frac-tion was used as a proxy for CYP1A1 activity. We found EROD activity to be much higher in WT mice than in Cyp1a1-/- mice

(Figure S6A). Subsequently, the Ames test revealed a significant mutagenicity of N21aP in the S9 fraction from WT livers, whereas no effect in mutagenic activity was found in the 0.5-10 µM/plate of N21aP in Cyp1a1-/- mice (Figure S6B). Moreover, the formation of N21aP-epoxide was detected by HPLC-MS/MS, and the chro-matogram showed that liver microsomal enzymes of WT mice had a characteristic peak after coincubation with GSH and N21aP, whereas no peak was found in Cyp1a1-/- mice. To further verify the structure, the fragmented structure of minus one molecule of water was examined, and the position of the peak was consistent with the position of the parent ion (Figure S6C).

Effects of AhR on N21aP-Induced CYP1A1 Expression

In addition, N21aP-exposed mice (Figure 3A,B) and HBE cells (Figure 3C,D) exhibited significantly higher expression of CYP1A1 at both the mRNA and protein levels; expression appeared to be dose dependent. Immunofluorescence results also appeared to show higher CYP1A1 expression in HBE cells treated with N21aP (Figure 3E).

Although N21aP-exposed mice and cells did not have signifi-cantly different levels of total AhR protein compared with relevant controls (Figure 3F,G), the AhR protein was lower in the cytoplasm and higher in the nucleus in HBE cells at 12 h and 24 h after 1 lM N21aP exposure (Figure 3H). Furthermore, a ChIP assay was conducted to ascertain the impact of N21aP on the binding affinity of the AhR to the xenobiotic response ele-ments (XREs) within the CYP1A1 promoter regions. The find-ings revealed a notable enrichment of AhR at the CYP1A1 promoter in cells treated with N21aP, as depicted in Figure 3I. Moreover, luciferase reporter gene results revealed higher lucifer-ase activity in N21aP-exposed cells, which appeared to be concentration dependent. These results suggest that N21aP acti-vated AhR signaling in a concentration-dependent manner (Figure 3J), similar to BaP (Figure S7).

Effects of m6A Posttranscriptional Modification on N21aP-Induced CYP1A1 Expression

In additiontoAhR transcriptional regulation, the overall m6A modi-fication level and m6A enrichment of CYP1A1 mRNA in vitro and in vivo were evaluated. We found total m6A levels and the level of m6A on CYP1A1 mRNA were significantly higher inN21aP-treated mice (Figure 4A,B) and HBE cells (Figure 4C,D). In addition, BaP, a classic PAH that induces AhR activation,24 also elevated the expression level of m6A on CYP1A1 (Figure 4E,F). Subsequently, dot blotting with DAA, an inhibitor of m6A methylation that inhibits S-adenosyl methionine (SAM) synthesis,39,66 was performed tover-ify the above effects, and the results showed that DAA-exposed HBE cells had lower m6A levels (Figure 4G) and lower CYPIA1 mRNA expression (Figure 4H). N21aP treatment alone significantly

up-regulated CYP1A1 expression, whereas DAA reversed the highly expression of CYP1A1 caused by N21aP (Figure 4I), which was also observed on BaP (Figure 4J).

Effects of METTL14 on N21aP-Induced CYP1A1 Expression

The expression of proteins related to m6A modification, including methyltransferases (METTL3, METTL14, and WTAP) and deme-thyltransferases (FTO and ALKBH5), was detected in vivo and in vitro. N21aP-exposed mice (Figure 5A) and cells (Figure 5B) demonstrated higher expression of m6A methyltransferase pro-teins, especially METTL14, rather than those of demethylases ALKBH5 and FTO. This finding was also observed in BaP-treated

HBE cells (Figure S8A). Immunofluorescence further confirmed greater METTL14 expression in the cells (Figure 5C). Moreover, knockdownofMETTL14 significantly down-regulated the mRNA and protein expression of both METTL14 and CYP1A1 in HBE cells (Figure 5D,E) and obviously down-regulated N21aP or BaP-induced expression of CYP1A1 (Figure 5F; Figure S8B). Nevertheless, neither HepG2 cells with overexpression of METTL14 nor with knockdown of METTL14 exhibited any significant differences on AhR activity compared with controls (Figure S9A-D). Considering that METTL14 and METTL3 can form methylase complexes and play a major catalytic role in the modification process of m6A,67 SAH, an inhibitor of the METTL3-METTL14 heterodimer complex,68 was selected for intervention. SAH showed similar effects on the down-regulation of CYP1A1 expression in the presence of N21aP or BaP in HBE cells (Figure 5G; Figure S8C).

Identification of m6A Modification Sites in CYP1A1 mRNA in the Presence of N21aP

To identify the m6A modification site of CYP1A1 mRNA targeted by METTL14, the online prediction tool SRAMP40 (Figure 5H) was used to predict the CYP1A1 m6A site [mainly enriched in the coding sequences (CDS) and the 30-UTR]. As shown in Figure 5H, the four top predicted scores were selected: positions at 2700, 2808, and 3796 in the CDS region and position at 5218 in the 30-UTR, namely, CYP1A1-2700, CYP1A1-2808, CYP1A1-3796, and CYP1A1-5218, respectively. Among the four sites, the m6A modification levels of CYP1A1-2700 and CYP1A1-5218 weresignif-icantly greater compared with the corresponding control (Figure 5I). Similar results were obtained for BaP, with significantly higher lev-els of m6A modification at the CYP1A1-2700 and CYP1A1-5218 loci (Figure S8D). Furthermore, luciferase reporters containing ei-ther WT or mutant CYP1A1 (CYP1A1-2700 and CYP1A1-5218) were used to address the effects of m6A modification on CYP1A1 expression. For the mutant construction of CYP1A1, the adenosine bases (A) in m6A consensus sequences (RRACH) were replaced with cytosine (C) to abolish m6A modification (Figure 5J), and the results showed that the WT CYP1A1 (either CYP1A1-2700 or CYP1A1-5218), but not the mutation, had lower luciferase activity in the absence of METTL14, suggesting that METTL14 was actually involved in the methylation of these two sites (Figure 5K). Notably, HBE cells with METTL14 knockdown had lower stability of CYP1A1 mRNA compared with the control (Figure 5L).

Recognition of IGF2BP3 to CYP1A1-5218 loci on the Stability of CYP1A1 mRNA in the Presence of N21aP

Among the m6A readers related to CYP1A1 mRNA stability, IGF2BP3 mRNA levels were significantly higher after N21aP treatment in either mouse lung tissue (Figure 6A) or HBE cells (Figure 6B), which was confirmed by the expression of IGF2BP3 by immunoblotting assay (Figure 6C,D) and the localization of IGF2BP3 in the HBE cellular cytoplasm by immunofluorescence assay (Figure 6E). Silencing IGF2BP3 significantly down-regulated the expression level of CYP1A1 at both the protein and mRNA levels (Figure 6F,G) and obviously reversed the elevation of CYP1A1 expression in HBE cells treated with N21aP (Figure 6H). Furthermore, HBE cells with IGF2BP3 knockdown had lower stabilityof CYP1A1 mRNA compared with the control (Figure 6I). Afterward, a direct interaction between IGF2BP3 and CYP1A1 was observed using a biotin-labeled CYP1A1 probe (Figure 6J). To confirm the identified loci of IGF2BP3 on CYP1A1 mRNA, four possible "UGGAC" sequences on the CYP1A1 gene were chosen (Figure 6K). Following the results in Figure 5, the two m6A modifi-cation sites, CYP1A1-2700 in the CDS region and CYP1A1-5218 in the 30-UTR, were selected for further verification, and IGF2BP3 mutants with truncation of the KH domains (Figure 6L) were con-structed as previously reported (Figure 6L).69 As shown in Figure 6M, overexpression of IGF2BP3 significantly enhanced luciferase activity at CYP1A1-5218 loci rather than CYP1A1-2700 loci, and this enhancement was largely impaired by the mutation in KH1-2, KH3-4, and KH1-4. Moreover, IGF2BP3 was significantly enriched in CYP1A1 mRNA, and the binding between CYP1A1 mRNA and IGF2BP3 was significantly impaired by METTL14 knockdown (Figure 6N).

Discussion

Previous in vivo and in vitro studies have reported that most of the known PAHs are both inducers of CYP1A1 via the classic AhR pathway and substrates of CYP1A1 that result in respiratory damage through metabolic activation.18,24 In addition to the known PAHs, our present study first identified N21aP as a novel substrate of CYP1A1 and suggested a role for N21aP in respiratory inflam-mation outcomes in male mice. In addition to the AhR pathway, our study also suggests that N21aP regulated CYP1A1 expression by enhancing CYP1A1 mRNA stability via METTL14-IGF2BP3- mediated m6A modification. Our research provides novel insights, supporting our hypothesis that N21aP is capable of up-regulating CYP1A1 expression via a dual-regulatory scheme: It not only acti-vates the AhR-mediated transcriptional pathway but also enhances posttranscriptional m6A modifications. We further propose that this synergistic approach facilitated the increased expression of CYP1A1, a phenomenon that may represent a shared regulatory mechanism among various PAHs, as preliminarily suggested by our observations with BaP.

N21aP is one of the HMW-PAHs of 6-ring PAH with MW30226; these PAHs are structurally stable and might be widely present in the environment, especially susceptible to adsorption in PM2:5 and per-sistently harmful to humans.70 Studies have found that HMW-PAHs extracted from environmental and combustion samples exhibit posi-tive mutagenesis reactions71 and significant cancer risks.72 In fact, some dibenzopyrenes with MW302 have been of concern to the European Union and are listed in priority control, such as DBalP, DBaiP, DBahP, and dibenzo[a,e]pyrene (DBaeP).73 N21aP is struc-turally similar to BaP, with one more aromatic ring. Our results dem-onstrated that the concentration of N21aP was approximately one-third to one-half of that of BaP in PM2:5, but it had a better correlation with PM2:5 and more stability than BaP. Therefore, we propose that N21aP was more prone to adsorption on PM2:5 and deposition in the alveoli and bronchi, which we predict would be extremely harmful to respiratory tissues. Furthermore, our study suggested that both N21aP and BaP may have a common source, supporting N21aP as a new important pollutant in the ambient airborne environment.

Given the structural similarity and likely common source to BaP, we argue that the toxicity of N21aP and its association with CYP1A1 should not be ignored. The results in this study revealed that N21aP could be metabolized by CYP1A1 to generate epoxide metabolites that further lead to DNA damage, which should provide a reasonable explanation for the respiratory toxicity of the inhalation of N21aP in WT male mice in which CYP1A1 was originally expressed in lung tissues. Thus, we recognized N21aP as a new toxic PAH given that our study suggested it as a new metabolic substrate of CYP1A1 and could be metabolically activated by CYP1A1 to produce toxic effects. Meanwhile, we found that N21aP could up-regulate the expression of CYP1A1 in vivo and in vitro. Thus, this study suggests N21aP is also both a metabolic substrate and a stabilizing inducer of CYP1A1, simi-lar to BaP.74 The significance of N21aP as an inducer may be greater because the generally lower environmental level of PAH precisely favors its role in inducing expression. CYP1A1 is the dominant meta-bolic enzyme of most PAHs, and its induced expression can greatly affect the intensity of the effect of mixed PAH exposure.75 Thus, the induction of CYP1A1 is vital for the metabolic activation of PAH,29 which must be helpful to fully understand the toxicity and health risk of environmental PAHs.

AhR is the classical and critical pathway for PAHs, such as BaP,24 to regulate CYP1A1 expression.76 Upon binding to the PAH, AhR is activated to translocate into the nucleus and form an AhR: ARNT heterodimer, thereby initiating CYP1A1 transcription through binding to XRE sequences.29 The results of this study suggest that N21aP induced CYP1A1 expression by activating AhR, influencing the AhR binding to the genomic XRE sequence in the CYP1A1 pro-moter. This finding suggests that the AhR pathway may represent a prevalent transcriptional mechanism for PAH-induced CYP1A1 expression. In addition to regulation at the transcriptional level, posttranscriptional regulation plays a crucial role in gene expression. Among the hundreds of posttranscriptional chemical modifi-cations, m6A modification is the most abundant and prevalent internal modification in mammalian mRNAs.77 Interestingly, we first found that N21aP and BaP not only affected AhR activation but also affected the m6A level of CYP1A1 in HBE cells, indicat-ing that m6A modification might also be an additional common regulation in PAH-induced CYP1A1 expression. Although m6A has been reported to regulate gene expression through various mechanisms, including mRNA stability, splicing, translation, and histone modification,78 the specific mechanism by which environ-mental pollutants regulate the expression of metabolic enzymes remains unclear.

The dynamic and reversible m6A modification, regulated by methyltransferases and demethylases, participates in biological processes, suchas cell differentiation, development, and immunity, as well as in the pathological processes of cancers.79 Generally, m6A-methylated mRNA can be synthesized via the methyltrans-ferase complex m6A writer, which mainly comprises METTL3, METTL14, and WTAP.80 METTL3 is the sole catalytic subunit of the complex, whereas METTL14 plays noncatalytic roles in complex stabilization and substrate RNA recruitment.81 METTL14 is involved in various physiological and pathological processes, which are prone to the development of diabetic nephropathy, ather-osclerosis, and various tumors.82 In addition, METTL14 is critical to the stability of mRNA. Cigarette smoke had the higher mRNA stability and expression of DIXDC1 through the up-regulation of METTL14, which in turn accelerated the degeneration and senes-cence of nucleus pulposus cells by activating the Wnt pathway.83

METTL14 is highly expressed in acute myeloid leukemia (AML) and regulates the myeloid differentiation of AML cells by affecting the stability and translation of MYB and MYC mRNA, thereby enhancing the self-renewal and proliferation of AML cells and leading to AML progression.84 Interestingly, we found that METTL14 expression but not METTL3 expression was signifi-cantly up-regulated in N21aP-assoctiated CYP1A1 expression, suggesting that although METTL3 and METTL14 have synergistic effects, they play different regulatory roles in m6A modification. Furthermore, the present study discovered that METTL14 enhanced the m6A level of CYP1A1 mRNA and the m6A enrich-ment region located around the CYP1A1-2700 loci in the CDS region and the CYP1A1-5218 loci in the 30-UTR. In general, m6A modification is a methylated adenosine occurring at the N6 posi-tion, which was enriched in the stop codon, 30-UTR, and inner long exon regions of RNA and occurred in the specific motif "RRACH" (R= G=A=U, H =U=A=C),77 consistent with our results. In sum-mary, METTL14 improves the stability of CYP1A1 mRNA and may be an important mechanism by which N21aP induces CYP1A1 expression. CYP1A1-2700 loci at the CDS region and CYP1A1-5218 loci at the 30-UTR might be the functional sites associated with METTL14-mediated m6A modification of CYP1A1mRNA.

Notably, the effects on targeted mRNAs resulting from m6A predominantly depend on the functions of different m6A reader proteins, including the YT521-B homology (YTH) domain family of proteins (YTHDFs and YTHDCs), heterogeneous nuclear ribo-nucleoproteins (HNRNPs), and insulin-like growth factor 2 mRNA-binding proteins (IGF2BPs).80 Similar to m6A writers, most reader proteins exhibit abnormal expression after exposure to various environmental pollutants.85 For example, studies found that arsenic exposure exhibited greater METTL3 expression, thereby increasing the m6A modification level of PRDM2 and down-regulating P53 expression by promoting PRDM2 mRNA degradation through YTHDF2, ultimately leading to the malignant transformation of human keratinocytes.86 Studies have also shown that m6A modification may be a new mechanism for regulating cytochrome P450 expression, and the m6A reader protein YTHDC2 promotes CYP2C8 mRNA degradation by recognizing m6A modification sites.87 Recently, IGF2BP1, IGF2BP2, and

IGF2BP3 have been confirmed to be able to bind and stabilize m6A-modified mRNA and that IGF2BPs can work with cofactors such as human antigen R (HuR) and Matrin 3 (MATR3) to main-tain the stability of mRNA.69 Here, we found that IGF2BP3 was highly expressed in N21aP-treated cells and that the knockdown of IGF2BP3 was associated with lower CYP1A1 expression in N21aP-treated cells. It has been reported that PM2:5-induced m6A modification regulates the stability of BIRC5 mRNA through IGF2BP3 and promotes the progression of bladder cancer, high-lighting the role of IGF2BP3 in health damage caused by environ-mental pollutants.88 Our results also showed that IGF2BP3 deficiency exhibited less stability of CYP1A1 mRNA. In addition,

direct binding of IGF2BPs to m6A-modified RNAs through its KH domain has been demonstrated in vitro, and the KH domain in IGF2BPs, particularly the KH3-4 dual domain, has been proven to play a crucial role in the recognition and binding of specific mRNAs,69 which should support our results that the KH1-2, KH3-4, and KH1-4 domains in IGF2BP3 were responsible for recognizing the CYP1A1-5218 loci in the 30-UTR of CYP1A1 mRNA. Collectively, these results further revealed that IGF2BP3 could recognize the m6A modification site of CYP1A1 mRNA mediated by METTL14 at the CYP1A1-5218 loci, subsequently enhancing the stability of CYP1A1 mRNA and finally increasing CYP1A1 protein expression.

In summary, the present study elucidated the critical roles of AhR-mediated transcriptional regulation and m6A modification- mediated posttranscriptional regulation in N21aP-induced CYP1A1 expression, which in turn resulted in respiratory inflammation via metabolic activation of N21aP by CYP1A1. Our findings suggest that N21aP induced CYP1A1 expression by promoting CYP1A1 transcription through the AhR activation pathway and by enhancing CYP1A1 mRNA stability in a METTL14-IGF2BP3-dependent manner through m6A posttranscriptional modification (Figure 7), which may be a common regulatory pathway that can be applied to other PAHs, especially HMW-PAHs. This study contributes to a high level of interest and comprehensive understanding of the sig-nificance of environment-genetic interactions for the toxicity of PAHs and provides an important basis for understanding the health risksofHMW-PAHs at environmental exposure levels.

Acknowledgments

S.W. and C.W. conceived, designed, and supervised the study. J.S., L.W. and W.Z. conducted the experiment, analyzed the data, and prepared the draft manuscript. C.C. was responsible for the N21aP synthesis. S.W., J.X. and S.L. were responsible for the PAH extraction, detection, and analysis. X.G., R.X., and X.Z. were responsible for cell experiments and analysis. Y.J., D.Z. and S.X. contributed to animal experiments and data analysis. H.S. provided reliable and qualified PM2:5 samples. All authors read and approved the final manuscript.

This work was supported by the National Natural Science Foundation of China [82173562 (to S.-L.W.), 81973091 (to S.-L.W.), 82273584 (to C.W.), and 81903353 (to C.W.), and 91743205 (to S.-L.W.)], the Jiangsu Provincial College Students' Innovation and Entrepreneurship Training Program [202210312045Z (to C.W.)], the Technology Development Fund of Nanjing Medical University [NMUB2018001 (to C.W.) and NMUB2020007 (to L.W.)], the project funded by the Collegiate Natural Science Foundation of Jiangsu Province [19KJB330003 (to C.W.)], and the research Fund for Large-scale Scientific Instrument Sharing of Nanjing Medical University [ZC2021DY10 (to L.W.)].

References

References

1. Kim K-H, Jahan SA, Kabir E, Brown RJC. 2013. A review of airborne polycyclic aromatic hydrocarbons (PAHs) and their human health effects. Environ Int 60:71-80, PMID: 24013021, https://doi.org/10.1016/j.envint.2013.07.019.

2. Hrdina AIH, Kohale IN, Kaushal S, Kelly J, Selin NE, Engelward BP, et al. 2022. The parallel transformations of polycyclic aromatic hydrocarbons in the body and in the atmosphere. Environ Health Perspect 130(2):025004, PMID: 35225689, https://doi.org/10.1289/EHP9984.

3. Zhang Y, Chen X, Zhang Y. 2021. Analytical chemistry, formation, mitigation, and risk assessment of polycyclic aromatic hydrocarbons: from food processing to in vivo metabolic transformation. Compr Rev Food Sci Food Saf 20(2):1422-1456, PMID: 33506545, https://doi.org/10.1111/1541-4337.12705.

4. Lin C, Huang R-J, Duan J, Zhong H, Xu W. 2022. Polycyclic aromatic hydrocar-bons from cooking emissions. Sci Total Environ 818:151700, PMID: 34798089, https://doi.org/10.1016/j.scitotenv.2021.151700.

5. Huang S, Li Q, Liu H, Ma S, Long C, Li G, et al. 2022. Urinary monohydroxylated polycyclic aromatic hydrocarbons in the general population from 26 provincial capital cities in China: levels, influencing factors, and health risks. Environ Int 160:107074, PMID: 34995968, https://doi.org/10.1016/j.envint.2021.107074.

6. Yu Y, Li Q, Wang H, Wang B, Wang X, Ren A, et al. 2015. Risk of human exposure to polycyclic aromatic hydrocarbons: a case study in Beijing, China. Environ Pollut 205:70-77, PMID: 26017113, https://doi.org/10.1016/j.envpol.2015.05.022.

7. Li Y, Liu M, Hou L, Li X, Yin G, Sun P, et al. 2021. Geographical distribution of polycyclic aromatic hydrocarbons in estuarine sediments over China: human impacts and source apportionment. Sci Total Environ 768:145279, PMID: 33736343, https://doi.org/10.1016/j.scitotenv.2021.145279.

8. Famiyeh L, Xu H, Chen K, Tang YT, Ji D, Xiao H, et al. 2024. Breathing in danger: unveiling the link between human exposure to outdoor PM2.5-bound polycyclic aromatic hydrocarbons and lung cancer risk in an urban residential area of China. Sci Total Environ 907:167762, PMID: 37852504, https://doi.org/10.1016/j. scitotenv.2023.167762.

9. US EPA (US Environmental Protection Agency). 1993. Provisional Guidance for Quantitative Risk Assessment of Polycyclic Aromatic Hydrocarbons. EPA/ 600/R-93/089. https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=30002TUA.txt [accessed 30 June 2023].

10. Zeng Y, Ma H-M, Zhang Q-Y, Tao L, Wang T, Wan C, et al. 2023. Complex poly-cyclic aromatic compound mixtures in PM2.5 in a Chinese megacity: spatio-temporal variations, toxicity, and source apportionment. Environ Int 179:108159, PMID: 37607426, https://doi.org/10.1016/j.envint.2023.108159.

11. Chen S-J, Wang J, Wang T, Wang T, Mai B-X, Simonich SLM. 2016. Seasonal variations and source apportionment of complex polycyclic aromatic hydrocar-bon mixtures in particulate matter in an electronic waste and urban area in South China. Sci Total Environ 573:115-122, PMID: 27552735, https://doi.org/10. 1016/j.scitotenv.2016.08.101.

12. Durant JL, Lafleur AL, Plummer EF, Taghizadeh K, Busby WF, Thilly WG. 1998. Human lymphoblast mutagens in urban airborne particles. Environ Sci Technol 32(13):1894-1906, https://doi.org/10.1021/es9706965.

13. Titaley I, Chlebowski A, Truong L, Tanguay RL, Massey Simonicha SL. 2016. Identification and toxicological evaluation of unsubstituted PAHs and novel PAH derivatives in pavement sealcoat products. Environ Sci Technol Lett 3(6):234-242, PMID: 30079367, https://doi.org/10.1021/acs.estlett.6b00116.

14. US EPA. 2010. Development of a Relative Potency Factor (RPF) Approach for Polycyclic Aromatic Hydrocarbon (PAH) Mixtures (External Review Draft). EPA/635/R-08/012A. https://nepis.epa.gov/Exe/ZyPURL.cgi?Dockey=P100HO25.txt [accessed 30 June 2023].

15. Yu Z, Wang H, Zhang X, Gong S, Liu Z, Zhao N, et al. 2022. Long-term environ-mental surveillance of PM2.5-bound polycyclic aromatic hydrocarbons in Jinan, China (2014-2020): health risk assessment. J Hazard Mater 425:127766, PMID: 34916105, https://doi.org/10.1016/j.jhazmat.2021.127766.

16. Wang T, Song X, Xu H, Zhu Y, Li L, Sun X, et al. 2022. Combustion-derived par-ticulate PAHs associated with small airway dysfunction in elderly patients with COPD. Environ Sci Technol 56(15):10868-10878, PMID: 35834827, https://doi.org/ 10.1021/acs.est.2c00797.

17. Liu KY, Gao Y, Xiao W, Fu J, Huang S, Han X, et al. 2023. Multidimensional anal-ysis of lung lymph nodes in a mouse model of allergic lung inflammation fol-lowing PM2.5 and indeno[1,2,3-cd]pyrene exposure. Environ Health Perspect 131(3):037014, PMID: 36975775, https://doi.org/10.1289/EHP11580.

18. Stading R, Gastelum G, Chu C, Jiang W, Moorthy B. 2021. Molecular mecha-nisms of pulmonary carcinogenesis by polycyclic aromatic hydrocarbons (PAHs): implications for human lung cancer. Semin Cancer Biol 76:3-16, PMID: 34242741, https://doi.org/10.1016/j.semcancer.2021.07.001.

19. Ding X, Kaminsky LS. 2003. Human extrahepatic cytochromes P450: function in xenobiotic metabolism and tissue-selective chemical toxicity in the respiratory and gastrointestinal tracts. Annu Rev Pharmacol Toxicol 43:149-173, PMID: 12171978, https://doi.org/10.1146/annurev.pharmtox.43.100901.140251.

20. Stoddard EG, Nag S, Martin J, Tyrrell KJ, Gibbins T, Anderson KA, et al. 2021. Exposure toanenvironmental mixtureofpolycyclic aromatic hydrocarbons indu-ces hepatic cytochrome P450 enzymes in mice. Chem Res Toxicol 34(9):2145- 2156, PMID: 34472326, https://doi.org/10.1021/acs.chemrestox.1c00235.

21. Puliyappadamba VT, Thulasidasan AKT, Vijayakurup V, Antony J, Bava SV, Anwar S, et al. 2015. Curcumin inhibits B[a]PDE-induced procarcinogenic signals in lung cancer cells, and curbs B[a]P-induced mutagenesis and lung carcinogenesis. Biofactors 41(6):431-442, PMID: 26643788, https://doi.org/10.1002/biof.1244.

22. Kuang H, Dai Y, Ding X, Li Y, Cha C, Jiang W, et al. 2021. Association among blood BPDE-DNA adduct, serum interleukin-8 (IL-8) and DNA strand breaks for children with pulmonary diseases. Int J Environ Health Res 31(7):823-834, PMID: 31722538, https://doi.org/10.1080/09603123.2019.1690638.

23. Madeen EP, Ognibene TJ, Corley RA, McQuistan TJ, Henderson MC, Baird WM, et al. 2016. Human microdosing with carcinogenic polycyclic aromatic hydrocarbons: in vivo pharmacokinetics of dibenzo[def,p]chrysene and metab-olites by UPLC accelerator mass spectrometry. Chem Res Toxicol 29(10):1641- 1650, PMID: 27494294, https://doi.org/10.1021/acs.chemrestox.6b00169.

24. Vázquez-Gómez G, Rocha-Zavaleta L, Rodríguez-Sosa M, Petrosyan P, Rubio-Lightbourn J. 2018. Benzo[a]pyrene activates an AhR/src/ERK axis that contrib-utes to CYP1A1 induction and stable DNA adducts formation in lung cells. Toxicol Lett 289:54-62, PMID: 29545172, https://doi.org/10.1016/j.toxlet.2018.03.012.

25. Chen C, Shen J, Yang L, Zhang W, Xia R, Huan F, et al. 2021. Identification of structural properties influencing the metabolism of polycyclic aromatic hydro-carbons by cytochrome P450 1A1. Sci Total Environ 758:143997, PMID: 33333309, https://doi.org/10.1016/j.scitotenv.2020.143997.

26. Wei S, Liu M, Huang B, Bi X, Sheng G, Fu J. 2011. Polycyclic aromatic hydrocarbons with molecular weight 302 in PM2.5 at two industrial sites in South China. J Environ Monit 13(9):2568-2574, PMID: 21789327, https://doi.org/10.1039/c1em10320b.

27. Wang W, Jariyasopit N, Schrlau J, Jia Y, Tao S, Yu T-W, et al. 2011. Concentration and photochemistry of PAHs, NPAHs, and OPAHs and toxicity of PM2.5 during the Beijing Olympic games. Environ Sci Technol 45(16):6887-6895, PMID: 21766847, https://doi.org/10.1021/es201443z.

28. Chen C, Min Y, Li X, Chen D, Shen J, Zhang D, et al. 2021. Mutagenicity risk pre-diction of PAH and derivative mixtures by in silico simulations oriented from CYP compound I-mediated metabolic activation. Sci Total Environ 787:147596, PMID: 33991922, https://doi.org/10.1016/j.scitotenv.2021.147596.

29. Mescher M, Haarmann-Stemmann T. 2018. Modulation of CYP1A1 metabolism: from adverse health effects to chemoprevention and therapeutic options. Pharmacol Ther 187:71-87, PMID: 29458109, https://doi.org/10.1016/j.pharmthera. 2018.02.012.

30. Goedtke L, Sprenger H, Hofmann U, Schmidt FF, Hammer HS, Zanger UM, et al. 2020. Polycyclic aromatic hydrocarbons activate the aryl hydrocarbon receptor and the constitutive androstane receptor to regulate xenobiotic metabolism in human liver cells. Int J Mol Sci 22(1):372, PMID: 33396476, https://doi.org/10. 3390/ijms22010372.

31. Zanger UM, Schwab M. 2013. Cytochrome P450 enzymes in drug metabolism: regulation of gene expression, enzyme activities, and impact of genetic varia-tion. Pharmacol Ther 138(1):103-141, PMID: 23333322, https://doi.org/10.1016/j. pharmthera.2012.12.007.

32. Corbett AH. 2018. Post-transcriptional regulation of gene expression and human disease. Curr Opin Cell Biol 52:96-104, PMID: 29518673, https://doi.org/ 10.1016/j.ceb.2018.02.011.

33. LiX,Xiong X,ZhangM,WangK,Chen Y, Zhou J, et al. 2017. Base-resolution mapping reveals distinct m1a methylome in nuclear- and mitochondrial-encoded transcripts. Mol Cell 68(5):993-1005.e9, PMID: 29107537, https://doi.org/10.1016/j. molcel.2017.10.019.

34. Roden C, Gladfelter AS. 2021. RNA contributions to the form and function of bio-molecular condensates. Nat Rev Mol Cell Biol 22(3):183-195, PMID: 32632317, https://doi.org/10.1038/s41580-020-0264-6.

35. Roundtree IA, Evans ME, Pan T, He C. 2017. Dynamic RNA modifications in gene expression regulation. Cell 169(7):1187-1200, PMID: 28622506, https://doi.org/10. 1016/j.cell.2017.05.045.

36. Dominissini D, Moshitch-Moshkovitz S, Schwartz S, Salmon-Divon M, Ungar L, Osenberg S, et al. 2012. Topology of the human and mouse m6A RNA methylomes revealed by m6A-seq. Nature 485(7397):201-206, PMID: 22575960, https://doi.org/ 10.1038/nature11112.

37. Liu J, Harada BT, He C. 2019. Regulation of gene expression by N6-methyladeno-sine in cancer. Trends Cell Biol 29(6):487-499, PMID: 30940398, https://doi.org/10. 1016/j.tcb.2019.02.008.

38. Shi H, Chai P, Jia R, Fan X. 2020. Novel insight into the regulatory roles of diverse RNA modifications: re-defining the bridge between transcription and translation. Mol Cancer 19(1):78, PMID: 32303268, https://doi.org/10.1186/s12943-020-01194-6.

39. Isono M, Nakano M, Fukami T, Nakajima M. 2022. Adenosine N6-methylation upregulates the expression of human CYP2B6 by altering the chromatin status. Biochem Pharmacol 205:115247, PMID: 36113565, https://doi.org/10.1016/j.bcp. 2022.115247.

40. Zhou Y, Zeng P, Li Y-H, Zhang Z, Cui Q. 2016. SRAMP: prediction of mammalian N6-methyladenosine (m6A) sites based on sequence-derived features. Nucleic Acids Res 44(10):e91, PMID: 26896799, https://doi.org/10.1093/nar/gkw104.

41. Li Y, Fan Z, Lu W, Xu R, Liu T, Liu L, et al. 2023. Long-term exposure to ambient fine particulate matter-bound polycyclic aromatic hydrocarbons and cancer mortal-ity: a difference-in-differences approach. Chemosphere 340:139800, PMID: 37572709, https://doi.org/10.1016/j.chemosphere.2023.139800.

42. Wu D, Zhang F, Lou W, Li D, Chen J. 2017. Chemical characterization and toxicity assessment of fine particulate matters emitted from the combustion of petrol and diesel fuels. Sci Total Environ 605-606:172-179, PMID: 28666172, https://doi.org/10. 1016/j.scitotenv.2017.06.058.

43. Chen L, Li Q, Wu D, Sun H, Wei Y, Ding X, et al. 2019. Size distribution and chemical composition of primary particles emitted during open biomass burning proc-esses: impacts on cloud condensation nuclei activation. Sci Total Environ 674:179-188, PMID: 31004894, https://doi.org/10.1016/j.scitotenv.2019.03.419.

44. Pietrogrande MC, Demaria G, Russo M. 2023. Determination of particulate poly-cyclic aromatic hydrocarbons in ambient air by gas chromatography-mass spectrometry after molecularly imprinted polymer extraction. J Environ Sci (China) 124:644-654, PMID: 36182170, https://doi.org/10.1016/j.jes.2021.11.021.

45. Larsen RK III, Baker JE. 2003. Source apportionment of polycyclic aromatic hydrocarbons in the urban atmosphere: a comparison of three methods. Environ Sci Technol 37(9):1873-1881, PMID: 12775060, https://doi.org/10.1021/es0206184.

46. Gbeddy G, Egodawatta P, Goonetilleke A, Akortia E, Glover ET. 2021. Influence of photolysis on source characterization and health risk of polycyclic aromatic hydrocarbons (PAHs), and carbonyl-, nitro-, hydroxy- PAHs in urban road dust. Environ Pollut 269:116103, PMID: 33261958, https://doi.org/10.1016/j.envpol.2020. 116103.

47. Khan MF, Latif MT, Lim CH, Amil N, Jaafar SA, Dominick D, et al. 2015. Seasonal effect and source apportionment of polycyclic aromatic hydrocarbons in PM2.5. Atmos Environ 106:178-190, https://doi.org/10.1016/j.atmosenv.2015.01.077.

48. Zhu Z, Li L, Yu Y, Tan L, Wang Z, Suo S, et al. 2023. Distribution, source, risk and phytoremediation of polycyclic aromatic hydrocarbons (PAHs) in typical

urban landscape waters recharged by reclaimed water. J Environ Manage 330:117214, PMID: 36623386, https://doi.org/10.1016/j.jenvman.2023.117214.

49. Li Y, Jia Z, Wijesiri B, Song N, Goonetilleke A. 2018. Influence of traffic on build-up of polycyclic aromatic hydrocarbons on urban road surfaces: a Bayesian network modelling approach. Environ Pollut 237:767-774, PMID: 29217392, https://doi.org/10.1016/j.envpol.2017.10.125.

50. Yoshizaki K, Brito JM, Silva LF, Lino-Dos-Santos-Franco A, Frias DP, Rossi e Silva RC, et al. 2017. The effects of particulate matter on inflammation of respi-ratory system: differences between male and female. Sci Total Environ 586:284-295, PMID: 28174048, https://doi.org/10.1016/j.scitotenv.2017.01.221.

51. Prada-Dacasa P, Urpi A, Sánchez-Benito L, Bianchi P, Quintana A. 2020. Measuring breathing patterns in mice using whole-body plethysmography. Bio Protoc 10(17):e3741, PMID: 33659401, https://doi.org/10.21769/BioProtoc.3741.

52. Ji M, Zhang Z, Li N, Xia R, Wang C, Yu Y, et al. 2018. Identification of 5-hydroxy-methylfurfural in cigarette smoke extract as a new substrate metabolically activated by human cytochrome P450 2A13. Toxicol Appl Pharmacol 359:108- 117, PMID: 30253172, https://doi.org/10.1016/j.taap.2018.09.031.

53. Shannahan JH, Nyska A, Cesta M, Schladweiler MCJ, Vallant BD, Ward WO, et al. 2012. Subchronic pulmonary pathology, iron overload, and transcriptional activity after Libby amphibole exposure in rat models of cardiovascular dis-ease. Environ Health Perspect 120(1):85-91, PMID: 21979745, https://doi.org/10. 1289/ehp.1103990.

54. Wagner JG, Barkauskas CE, Vose A, Lewandowski RP, Harkema JR, Tighe RM. 2020. Repetitive ozone exposures and evaluation of pulmonary inflammation and remodeling in diabetic mouse strains. Environ Health Perspect 128(11):117009, PMID: 33253011, https://doi.org/10.1289/EHP7255.

55. Rueden CT, Schindelin J, Hiner MC, DeZonia BE, Walter AE, Arena ET, et al. 2017. ImageJ2: ImageJ for the next generation of scientific image data. BMC Bioinformatics 18:529, PMID: 29187165, https://doi.org/10.1186/s12859-017-1934-z.

56. Buckpitt A, Boland B, Isbell M, Morin D, Shultz M, Baldwin R, et al. 2002. Naphthalene-induced respiratory tract toxicity: metabolic mechanisms of tox-icity. Drug Metab Rev 34(4):791-820, PMID: 12487150, https://doi.org/10.1081/ dmr-120015694.

57. Chu S, Letcher RJ. 2019. In vitro metabolic activation of triphenyl phosphate lead-ing to the formation of glutathione conjugates by rat liver microsomes. Chemosphere 237:124474, PMID: 31377596, https://doi.org/10.1016/j.chemosphere. 2019.124474.

58. Li X, Yuan B, Lu M, Wang Y, Ding N, Liu C, et al. 2021. The methyltransferase METTL3 negatively regulates nonalcoholic steatohepatitis (NASH) progression. Nat Commun 12(1):7213, PMID: 34893641, https://doi.org/10.1038/s41467-021-27539-3.

59. Zhang J, Li Y, Liu Q, Huang Y, Li R, Wu T, et al. 2021. Sirt6 alleviated liver fibro-sis by deacetylating conserved lysine 54 on Smad2 in hepatic stellate cells. Hepatology 73(3):1140-1157, PMID: 32535965, https://doi.org/10.1002/hep.31418.

60. Qin Y, Chen W, Jiang G, Zhou L, Yang X, Li H, et al. 2020. Interfering MSN-NONO complex-activated CREB signaling serves as a therapeutic strategy for triple-negative breast cancer. Sci Adv 6(8):eaaw9960, PMID: 32128390, https://doi.org/ 10.1126/sciadv.aaw9960.

61. Jin Y, Qi G, Shou Y, Li D, Liu Y, Guan H, et al. 2022. High throughput data-based, toxicity pathway-oriented development of a quantitative adverse outcome pathway network linking AHR activation to lung damages. J Hazard Mater 425:128041, PMID: 34906874, https://doi.org/10.1016/j.jhazmat.2021.128041.

62. Montalbano M, McAllen S, Sengupta U, Puangmalai N, Bhatt N, Ellsworth A, et al. 2019. Tau oligomers mediate aggregation of RNA-binding proteins Musashi1 and Musashi2 inducing Lamin alteration. Aging Cell 18(6):e13035, PMID: 31532069, https://doi.org/10.1111/acel.13035.

63. Vorrink SU, Severson PL, Kulak MV, Futscher BW, Domann FE. 2014. Hypoxia perturbs aryl hydrocarbon receptor signaling and CYP1A1 expression induced by PCB 126 in human skin and liver-derived cell lines. Toxicol Appl Pharmacol 274(3):408-416, PMID: 24355420, https://doi.org/10.1016/j.taap.2013.12.002.

64. Zhang Y, Wu W, Li Y, Li Y. 2023. An investigation of PM2.5 concentration changes in Mid-Eastern China before and after COVID-19 outbreak. Environ Int 175:107941, PMID: 37146469, https://doi.org/10.1016/j.envint.2023.107941.

65. Ma Z, Hu X, Sayer AM, Levy R, Zhang Q, Xue Y, et al. 2016. Satellite-based spa-tiotemporal trends in PM2.5 concentrations: China, 2004-2013. Environ Health Perspect 124(2):184-192, PMID: 26220256, https://doi.org/10.1289/ehp.1409481.

66. Yang WS, Kim JH, Jeong D, Hong YH, Park SH, Yang Y, et al. 2020. 3-Deazaadenosine, an S-adenosylhomocysteine hydrolase inhibitor, attenuates lipopolysaccharide-induced inflammatory responses via inhibition of AP-1 and NF-?B signaling. Biochem Pharmacol 182:114264, PMID: 33035507, https://doi.org/ 10.1016/j.bcp.2020.114264.

67. Liu J, Yue Y, Han D, Wang X, Fu Y, Zhang L, et al. 2014. A METTL3-METTL14 complex mediates mammalian nuclear RNA N6-adenosine methylation. Nat Chem Biol 10(2):93-95, PMID: 24316715, https://doi.org/10.1038/nchembio.1432.

68. Wang P, Doxtader KA, Nam Y. 2016. Structural basis for cooperative function of Mettl3 and Mettl14 methyltransferases. Mol Cell 63(2):306-317, PMID: 27373337, https://doi.org/10.1016/j.molcel.2016.05.041.

69. Huang H, Weng H, Sun W, Qin X, Shi H, Wu H, et al. 2018. Recognition of RNA N6-methyladenosine by IGF2BP proteins enhances mRNA stability and translation. Nat Cell Biol 20(3):285-295, PMID: 29476152, https://doi.org/10.1038/s41556-018-0045-z.

70. Barbosa F Jr, Rocha BA, Souza MCO, Bocato MZ, Azevedo LF, Adeyemi JA, et al. 2023. Polycyclic aromatic hydrocarbons (PAHs): updated aspects of their determination, kinetics in the human body, and toxicity. J Toxicol Environ Health B Crit Rev 26(1):28-65, PMID: 36617662, https://doi.org/10.1080/10937404. 2022.2164390.

71. Wornat MJ, Ledesma EB, Sandrowitz AK, Roth MJ, Dawsey SM, Qiao YL, et al. 2001. Polycyclic aromatic hydrocarbons identified in soot extracts from domes-tic coal-burning stoves of Henan Province, China. Environ Sci Technol 35(10):1943-1952, PMID: 11393972, https://doi.org/10.1021/es001664b.

72. Wang J, Chen S, Tian M, Zheng X, Gonzales L, Ohura T, et al. 2012. Inhalation cancer risk associated with exposure to complex polycyclic aromatic hydrocar-bon mixtures in an electronic waste and urban area in South China. Environ Sci Technol 46(17):9745-9752, PMID: 22913732, https://doi.org/10.1021/es302272a.

73. EU (European Union). 2005. Commission Regulation (EC) No 208/2005 of 4 February 2005 amending Regulation (EC) No 466/2001 as regards polycyclic aromatic hydrocarbons (text with EEA relevance). Off J Eur Union L 34:3-5.

74. Shahid A, Chen M, Lin C, Andresen BT, Parsa C, Orlando R, et al. 2023. The ß-blocker carvedilol prevents benzo(a)pyrene-induced lung toxicity, inflammation and carcinogenesis. Cancers (Basel) 15(3):583, PMID: 36765542, https://doi.org/ 10.3390/cancers15030583.

75. Chen C, Zhang D, Yuan A, Shen J, Wang L, Wang S-L. 2022. A novel approach to pre-dict the comprehensive EROD potency: mechanism-based curve fitting of CYP1A1 activity by PAHs. Sci Total Environ 843:157052, PMID: 35787903, https://doi.org/10. 1016/j.scitotenv.2022.157052.

76. Vogel CFA, Van Winkle LS, Esser C, Haarmann-Stemmann T. 2020. The aryl hydrocarbon receptor as a target of environmental stressors - implications for pollution mediated stress and inflammatory responses. Redox Biol 34:101530, PMID: 32354640, https://doi.org/10.1016/j.redox.2020.101530.

77. He PC, He C. 2021. m6A RNA methylation: from mechanisms to therapeutic potential. EMBO J 40(3):e105977, PMID: 33470439, https://doi.org/10.15252/embj. 2020105977.

78. Wang Y, Li Y, Yue M, Wang J, Kumar S, Wechsler-Reya RJ, et al. 2018. N6-methyladenosine RNA modification regulates embryonic neural stem cell self-

renewal through histone modifications. Nat Neurosci 21(2):195-206, PMID: 29335608, https://doi.org/10.1038/s41593-017-0057-1.

79. Cao X, Geng Q, Fan D, Wang Q, Wang X, Zhang M, et al. 2023. m6A methylation: a process reshaping the tumour immune microenvironment and regulating immune evasion. Mol Cancer 22(1):42, PMID: 36859310, https://doi.org/10.1186/ s12943-022-01704-8.

80. Flamand MN, Tegowski M, Meyer KD. 2023. The proteins of mRNA modifica-tion: writers, readers, and erasers. Annu Rev Biochem 92:145-173, PMID: 37068770, https://doi.org/10.1146/annurev-biochem-052521-035330.

81. Sledz P, Jinek M. 2016. Structural insights into the molecular mechanism of the m6A writer complex. Elife 5:e18434, PMID:27627798, https://doi.org/10.7554/eLife.18434.

82. Guan Q, Lin H, Miao L, Guo H, Chen Y, Zhuo Z, et al. 2022. Functions, mecha-nisms, and therapeutic implications of METTL14 in human cancer. J Hematol Oncol 15(1):13, PMID: 35115038, https://doi.org/10.1186/s13045-022-01231-5.

83. Tu J, Li W, Hansbro PM, Yan Q, Bai X, Donovan C, et al. 2023. Smoking and tet-ramer tryptase accelerate intervertebral disc degeneration by inducing METTL14-mediated DIXDC1 m6 modification. Mol Ther 31(8):2524-2542, PMID: 37340635, https://doi.org/10.1016/j.ymthe.2023.06.010.

84. Weng H, Huang H, Wu H, Qin X, Zhao BS, Dong L, et al. 2018. METTL14 inhibits hematopoietic stem/progenitor differentiation and promotes leukemogenesis via mRNA m6A modification. Cell Stem Cell 22(2):191-205.e9, PMID: 29290617, https://doi.org/10.1016/j.stem.2017.11.016.

85. Zhu X, Fu H, Sun J, Xu Q. 2023. Interaction between N6-methyladenosine (m6A) modification and environmental chemical-induced diseases in various organ systems. Chem Biol Interact 373:110376, PMID: 36736874, https://doi.org/10. 1016/j.cbi.2023.110376.

86. Zhao T, Sun D, Zhao M, Lai Y, Liu Y, Zhang Z. 2020. N6-methyladenosine medi-ates arsenite-induced human keratinocyte transformation by suppressing p53 activation. Environ Pollut 259:113908, PMID: 31931413, https://doi.org/10.1016/j. envpol.2019.113908.

87. Nakano M, Ondo K, Takemoto S, Fukami T, Nakajima M. 2020. Methylation of adenosine at the N6 position post-transcriptionally regulates hepatic P450s expression. Biochem Pharmacol 171:113697, PMID: 31706844, https://doi.org/10. 1016/j.bcp.2019.113697.

88. Liu H, Gu J, Huang Z, Han Z,Xin J, Yuan L, et al. 2022. Fine particulate matter induces METTL3-mediated m6A modification of BIRC5 mRNA in bladder cancer. J Hazard Mater 437:129310, PMID: 35749893, https://doi.org/10.1016/j.jhazmat.2022.129310.