DNA damage and oxidative stress of tobacco smoke condensate in human bladder epithelial cells

Abstract

Smoking is the major risk factor for bladder cancer (BC), and up to 50% of the cases are attributed to smoking. There are 70 known carcinogens in tobacco smoke; however, the principal chemicals responsible for BC remain uncertain. The aromatic amines 4-aminobiphenyl (4-ABP) and 2-naphthylamine (2-NA) are implicated in BC pathogenesis of smokers based on elevated BC risk in factory workers exposed to these chemicals. However, 4-ABP and 2-NA only occur at several ng per cigarette and may be insufficient to induce BC. In contrast, other genotoxicants, including acrolein, occur at 1000-fold or higher levels in tobacco smoke. There is limited data on the toxicological effects of tobacco smoke in human bladder cells. We have assessed the cytotoxicity, oxidative stress, and DNA damage of tobacco smoke condensate (TSC) in human RT4 bladder cells. TSC was fractionated by liquid-liquid extraction into an acid-neutral fraction (NF) containing polycyclic aromatic hydrocarbons (PAHs), nitro-PAHs, phenols, and aldehydes, and a basic fraction (BF) containing aromatic amines, heterocyclic aromatic amines, and N-nitroso compounds. TSC and NF induced a time- and concentration-dependent cytotoxicity associated with oxidative stress, lipid peroxide formation, glutathione (GSH) depletion, and apurinic/apyrimidinic (AP) site formation, while BF showed weak effects. LC/MS-based metabolomic approaches showed TSC and NF altered GSH biosynthesis pathways, and induced more than 40 GSH-conjugates. GSH-conjugates of several hydroquinones were among the most abundant conjugates. RT4 cell treatment with synthetic hydroquinones and cresol mixtures at levels present in tobacco smoke accounted for most of the TSC-induced cytotoxicity and the AP sites formed. GSH conjugates of acrolein, methyl vinyl ketone, and crotonaldehyde levels also increased due to TSC-induced oxidative stress. Thus, TSC is a potent toxicant and DNA-damaging agent, inducing deleterious effects in human bladder cells at concentrations of <1% of a cigarette in cell culture media.

Graphical Abstract

Introduction

Bladder cancer (BC) is the most common malignancy of the urinary system and the tenth most commonly diagnosed cancer worldwide.¹ Genetic polymorphisms, chronic infection, and environmental and occupational exposures are risk factors for BC.² Tobacco smoking is a major risk factor for BC, accounting for 33% and 50% of the BC cases in women and men, respectively, are smokers.³ Tobacco smoke is comprised of thousands of chemicals, including 70 known carcinogens.⁴ These genotoxic agents include aromatic amines (AA), heterocyclic aromatic amines (HAA), N-nitrosamines (NOC), polycyclic aromatic hydrocarbons (PAH), nitro-PAH, phenolic compounds, aldehydes, volatile hydrocarbons, and metals;⁴ however, the chemicals in combusted tobacco responsible for BC remain uncertain.

The AA 4-aminobiphenyl (4-ABP) and 2-naphthylamine (2-NA) are bladder carcinogens present in tobacco smoke and are implicated in BC pathogenesis of smokers based on the elevated incidence of BC in factory workers in the textile dye and rubber tire industries exposed to high levels of these chemicals.^5,6 The carcinogenicity of AA is driven by their ability to form mutation-prone DNA adducts.⁷ 4-ABP DNA adducts form in human bladder cells in vitro and the bladder genome of some BC patients.^8-15 Moreover, the N-acetyltransferase 2 (NAT2) slow acetylation polymorphism is associated with a significant increase in BC risk in smokers.^16-18 NAT2 is a key enzyme involved in detoxifying AA.⁷ Thus, the mechanistic data and epidemiology support an etiological role for 4-ABP in human BC. However, 4-ABP is only formed at a few ng per cigarette,^19,20 levels that may be insufficient to induce BC in smokers.²¹ p53 is a frequently mutated gene in smoking and non-smoking related BC.²² p53 mutational spectra in smoking-related BC show distinct differences from non-smoking-related BC (http://p53.iarc.fr/). The 280 and 285 codons are mutational hotspots in smokers and nonsmokers, while codon 273 is a mutational hotspot in smoking-related BC. In the human HTB bladder cell line, 4-ABP DNA adducts preferentially form at codons 280 and 285; however, adducts formed at codon 273 are negligible.^23,24 The mutational signatures of tobacco smoke dominate the mutational landscape of bladder tumors but the chemicals responsible for these signatures are unknown.²⁵ Thus, other chemicals in tobacco smoke are likely to contribute to BC development.

Acrolein (Acr), a highly reactive α,β-unsaturated aldehyde, occurs at levels up to 500 μg per cigarette, and it is one of the most abundant genotoxicants in cigarette smoke.²⁶ The International Agency for Research on Cancer (IARC) classifies Acr as probably carcinogenic to humans (Group 2A).²⁷ Acr forms the mutation prone regioisomeric α- and γ-hydroxy-1,N²-propano-2′-deoxyguanosine (α-OH-Acr-dG and γ-OH-Acr-dG) DNA adducts.²⁸ These adducts are detected in the human lung, oral cavity, and other organs.^29-35 Recently, the Tang laboratory detected γ-OH-Acr-dG and α-OH-Acr-dG in the bladder genome of BC patients at 2-fold higher levels than in the bladder genome of patients without BC.¹² The same group also reported that Acr inhibits DNA repair and induces tumorigenic transformation of human bladder cells.³⁶ The Tang group concluded that Acr is a major tobacco-related bladder carcinogen based on these findings.^12,36 Inhalation studies in rodents reported that Acr increased the incidence of nasal cavity adenoma and lymphoma in mice, the nasal cavity cell carcinoma, pituitary gland adenoma, and pituitary gland adenocarcinoma in rats. However, Acr did not induce BC.²⁷ Acr occurs at high levels in cigarette smoke. However, the amount of inhaled Acr that reaches distant organs such as the bladder to cause DNA damage is uncertain because of Acr's high reactivity with macromolecules and efficient detoxication by GSH.^37-39 To the best of our knowledge, there is no published literature reporting Acr levels in smokers' urine. In contrast, the Acr-derived mercapturic acid conjugate N-acetyl-S-(3-hydroxypropyl)-l-cysteine (3HPMA)⁴⁰ is detected in urine and higher in smokers than nonsmokers.⁴¹ Acr also forms endogenously as a byproduct of oxidative stress-induced lipid peroxidation.⁴² Thus, the Acr DNA adducts formed in the bladder genome of BC patients may be the result of oxidative stress, a hallmark feature of BC, rather than by direct exposure to Acr from tobacco smoke.⁴³

There is a strong association between cigarette smoking and BC risk, but surprisingly, there are few mechanistic studies on tobacco smoke condensate (TSC) in human bladder cells.⁴⁴ In this study, we have investigated the cytotoxic effects of TSC in RT4 cells, a well-characterized human bladder cell line.⁴⁵ The TSC from the 1R6F Kentucky reference and commercial Marlboro cigarettes and their subfractions obtained by liquid-liquid extraction were employed for the study. The acid-neutral fraction (NF) is enriched for PAH, nitro-PAH, phenols, including hydroquinone (HQ) and catechol (Cat), and aldehydes, while the basic fraction (BF) is enriched for AA, HAA, and NOC. We anticipated that the BF containing many carcinogens would exert strong cytotoxicity. However, the TSC and NF were far more potent and induced a dose- and time-dependent cytotoxicity, which was associated with the induction of oxidative stress, LPO and AP site formation. In contrast, the BF showed weak effects. We employed targeted and untargeted metabolomics-based approaches to characterize the impact of TSC, NF, and BF on glutathione (GSH) metabolism. TSC and NF significantly altered the GSH biosynthesis pathway. Moreover, 44 GSH conjugates of tobacco toxicants and endogenous electrophiles were detected in TSC- and NF-treated cells. HQ and Cat underwent oxidation to their toxic quinones, which were among the most abundantly formed GSH conjugates and also involved in AP sites formation.

Materials and methods

Chemicals.

Reduced glutathione (GSH), oxidized glutathione GSSG, O-(pyridin-3-ylmethyl)hydroxylamine (PMOA), butyraldehyde, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), calf thymus (CT) DNA, p-Benzoquinone (p-BQ), HQ, 2-methylhydroquinone (2-MeHQ), Cat, p-cresol (p-Cre), m-cresol (m-Cre), and o-cresol (o-Cre), Acr, methylacrolein (MeAcr), crotonaldehyde (Crot), Methyl vinyl ketone (MVK), RNase A (bovine pancreas), RNase T1 (Aspergillus oryzae), proteinase K (Tritirachium album), sodium dodecyl sulfate (SDS), DNase I (type IV, bovine pancreas), alkaline phosphatase (Escherichia coli), and nuclease P1 (Penicillium citrinum) were purchased from Sigma Aldrich (St. Louis, MO, USA). Phosphodiesterase I (Crotalus adamanteus venom) and adenosine deaminase were purchased from Worthington Biochemical Corp. (Newark, NJ). LC/MS grade solvents were purchased from Fisher Chemical Co. (Pittsburgh, PA). Puregene protein precipitation solution was purchased from Qiagen (Germantown, MD). Strata-X 33 μm Polymeric Reversed Phase 30 mg solid-phase extraction (SPE) cartridge was obtained from Phenomenex (Torrance, CA, USA). If not stated otherwise, all other chemicals were purchased from Sigma-Aldrich (St. Louis, MO).

Cell culture.

RT4 cells (ATCC, Manassas, VA) were cultured in McCoy's 5A medium (ATCC, Manassas, VA) containing 10% fetal calf serum (Sigma Aldrich, St. Louis, MO, USA), penicillin (100 IU/mL) (Gibco, Life Technologies, Carlsbad, CA, USA), and streptomycin (100 μg/ml) (Gibco, Life Technologies, Carlsbad, CA, USA) at 37 °C in a humidified atmosphere of 5% CO₂.

Tobacco smoke extract collection and TSC preparation.

Reference cigarettes (1R6F) were obtained from The Center for Tobacco Reference Products (CTRP), University of Kentucky, Lexington. Commercially available Marlboro cigarettes were purchased from a retail store in Taiwan. Cigarettes were conditioned in a 22 °C and 60% relative humidity environment for 48 h before usage. Cigarettes were then smoked employing a Borgwaldt LX1 linear single port smoking machine (Borgwaldt KC, Hamburg, Germany) under a Canadian Intense smoking regimen which consisted of 55 mL puff volume, 2 sec per puff, 30-sec puff interval, and 100% blocked ventilation holes.⁴⁶ Mainstream smoke was collected on Cambridge filter pads (5 cigarettes per pad). Total particulate matter (TPM) was measured by gravimetric analysis of the filter pad. TPM was extracted from the filter pads with ethanol (10 mL per pad) by sonication for 60 min using an ultrasonic water bath (Branson Ultrasonics Corp, Danbury, CT). The ethanol extracts were centrifuged to remove the glass fiber filaments at 3,000 g for 10 min, then concentrated under a gentle stream of nitrogen with DMSO (0.5 mL per pad extract) added as a co-solvent. The concentrated extract (TSC) was stored at −80 °C. Blank pads were extracted following the same procedure and served as negative controls.

NF and BF preparation.

The TSC from reference cigarettes (1R6F) or Marlboro cigarettes (5 cigarette equivalent in 0.5 mL DMSO) was diluted in 0.1 N HCl (10 mL) and fractionated by liquid-liquid extraction using dichloromethane (DCM) (2 volumes, 2 times). The DCM fraction was concentrated under a gentle stream of nitrogen with DMSO added as a co-solvent (0.5 mL per 5 cigarette equivalent). This extract was designated as the neutral fraction (NF). The aqueous fraction was adjusted to pH ~13 with 1N NaOH and extracted with DCM (2 volumes, two times). The retrieved organic fraction was then concentrated under a gentle stream of nitrogen with DMSO added as a co-solvent (0.5 mL per 5 cigarette equivalent) and designated as the basic fraction (BF). Both fractions were stored at −80 °C. Blank pads were extracted following the same procedure and used as negative controls.

Cell viability assays.

Cell viability was performed with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Sigma-Aldrich, St Louis, MO, USA). Cells (1.5 × 10⁴) were seeded in MW96 plates and cultured in complete media. After 48 h, cells were washed with pre-warmed phosphate-buffered saline (PBS) and treated with TSC (75 - 600 μg/ml), NF (75 - 600 μg/mL), BF (75 – 600 μg/mL), hydroquinone mixture (HQ, Res, Cat, 2-MeHQ, p-Cre, m-Cre, and o-Cre) (0.5 – 100 μg/mL), aldehydes (Acr, Crot, MVK, or MeAcr) (0.5 – 200 μg/mL), or blank pads extracts (negative controls) for up to 96 h. Treatments were refreshed with chemicals every 24 h. After treatment, cells were washed with pre-warmed PBS and incubated with media containing MTT (0.5 mg/mL, 100 μL) at 37 °C. After 1 h, the media was removed, and the formazan crystals were dissolved in DMSO (100 μL). The absorbance was measured at 570 nm using a SpectraMax ID3 plate reader (Molecular Devices, San Jose, CA).

ROS measurement.

ROS formation was detected employing 6-carboxy-2',7'-dichlorodihydrofluorescein (Carboxy-H₂DCFDA) (Thermo Fisher Scientific, Waltham, MA)., Cells (1.5 × 10⁴) were seeded in a black clear bottom 96-well microplate and cultured for 48 h. Cells were then washed with pre-warmed Hank's balanced salt solution (HBSS) and stained with 20 μM of Carboxy-H₂DCFDA for 45 min at 37 °C in a humidified atmosphere of 5% CO₂. Cells were then washed with pre-warmed HBSS and treated with TSC (75 - 600 μg/mL), NF (75 – 600 μg/mL), BF (75 - 600 μg/mL), or blank extracts (negative controls) for 6 h. The cell viabilities after 6 h of treatment with the highest doses of TSC, BF, and NF were 75% or greater. The fluorescence of the oxidized Carboxy-H₂DCFDA intermediate was measured over 6 h at Ex/Em = 485/535 nm using a SpectraMax ID3 plate reader (Molecular Devices, San Jose, CA).

Thiobarbituric Acid Reactive Substances (TBARS) assay for malondialdehyde (MDA) measurement.

The MDA levels, an end product of lipid peroxidation, were measured using a TBARS Parameter Assay Kit (R&D Systems, Minneapolis, MN) following the manufacturer's protocol. After 6 h of treatment with blank pad extracts, TSC, NF, or BF (300 - 600 μg/mL), RT4 cells (1 x 10⁶) were resuspended in deionized water (1 mL) and lysed by sonication and freeze and thaw cycles (3 cycles). The lysate was treated with an equal vol of 0.6 N trichloroacetic acid and centrifuged at 12,000 g for 4 min. The supernatants were mixed with the thiobarbituric acid reagent and incubated at 50 °C for up to 3h. The optical density was measured before incubation and every 30 min using a SpectraMax ID3 plate reader (Molecular Devices, San Jose, CA).

GSH, GSSG, and GSH-conjugate measurements.

Cells (3 x 10⁶) were treated with blank pad extracts, TSC, NF, or BF (300 - 600 μg/mL). After 6 h, the cells were washed twice with chilled PBS to remove dead cells, flash-frozen with liquid nitrogen, and scraped from the P60 mm tissues culture dish with 80% CH₃OH/19.5% H₂O/0.5% HCO₂H. The cell lysates were then centrifuged at 4 °C at 21,000 g for 10 min, and the supernatants were vacuum centrifuged to dryness at room temperature. The cell extracts were resuspended at a concentration of 3 million cells in 99.8% H₂O/0.2% HCO₂H (60 μL).

Mass Spectrometric Analysis of GSH, GSSG, and GSH-conjugates.

The relative abundances of GSH, GSSG, and GSH-conjugates were measured by high-resolution accurate mass liquid chromatography-tandem mass spectrometry (HRAMS-LC/MS²). The analyses were performed on an Orbitrap Lumos Tribrid MS equipped with a HESI-II source interfaced to an UltiMate 3000 RSLCnano UHPLC system (Thermo Fisher Scientific, San Jose, CA). Electrospray ionization (ESI) was in the negative ion mode. The HESI-II was set at 2.1 kV, and the sheath and auxiliary gases were set at 35 and 10 arbitrary units, respectively. The ion transfer tube was 275 °C, and the vaporizer temperature was 120 °C. The MS parameters were as follows: RF-lens, 35%, Orbitrap resolution, 120,000 (MS) full width of the peak at its half-maximum (fwhm), at m/z 200, and 30,000 (MS²); activation type, higher-energy C-trap dissociation (HCD) normalized collision energy, 30%; maximum injection time 50 ms; AGC target, 4 x 10⁵ (Full MS) and 5 x 10⁴ (MS²); the quadrupole isolation width was set at 0.5 m/z; and data type, profile. Chromatography was done with an ACQUITY UPLC HSS T3 VanGuard pre-column (5 x 2.1 mm, 1.8 μm particle size, 130 Å pore size) (Waters, Milford, MA), an ACQUITY Waters HSS T3 Column, (150 x 2.1 mm, 1.8 μm particle size, 100 Å pore size, Waters, Milford, MA). The column was heated at 40 °C. The (A) mobile phase consisted of 2 mM NH₄OAc/ 0.05% FA and the (B) mobile phase contained 2 mM NH₄OAc/ 95% CH₃CN/ 0.05% FA (B). The initial composition (0% B) was maintained for 3 min with a flow rate of 65 μL/min, increased linearly to 30% B at 10 min, and then to 50% B at 15 min. After that, solvent B increased to 100% at 20 min. Then the flow rate was increased to 100 μL/min and held for 2 min. Thereafter, the gradient returned to 0% B in 5 min, and the system was re-equilibrated at this mobile phase composition before the next injection. The samples (1 μL) were injected using a 5 μL injection loop. The eluent was diverted to waste for the first 1.9 min of the analysis (0% B). The precursor ions were m/z 306.0765 (GSH), m/z 611.1447 and 305.0687 (GSSG singly and doubly charged, respectively), m/z 414.0977 (HQ-SG and Cat-SG), m/z 428.1133 (MeHQ-SG and MeCat-SG), m/z 442.1291 (DiMeHQ-SG), m/z 364.1184 (Acr-SG), and m/z 378.1441 (Crot-SG and MVK-SG). Data analysis was performed with Xcalibur software (version 4.4). The MS² ion chromatograms (EICs) were generated for GSH, GSSH, and GSH conjugates with a 10 ppm tolerance for mass accuracy with the following ions: m/z 272.0888, 254.0782, 210.0884, 143.0462, and 128.0353.

Kinetics of TSC-induced AP sites.

Cells (3.0 × 10⁶) were treated with TSC, NF, BF (300 μg/mL), or blank pad extracts (negative control) for up to 48 h. The cell media treatments were refreshed with the tobacco extracts after 24 h. At time points 6, 24, and 48 h of treatment, the cells were washed with chilled PBS on ice, scraped with 1 mL ice chilled PBS, and pelleted at 250 g for 10 min at 4 °C. The cell pellets were stored at −80 °C until DNA processing. The pelleted cells were resuspended in 50 mM Tris pH 8.0, 10 mM EDTA, 10 mM BME and the AP sites were derivatized with PMOA (PMOA-dR), extracted, and enzymatically digested following our published protocol.⁴⁷

AP site measurements.

PMOA-dR AP sites were measured with an Orbitrap Lumos Tribrid MS interfaced to an UltiMate 3000 RSLCnano UHPLC using an EASY-Spray Source in the positive ion mode at the MS² scan stage. The EASY-Spray source voltage was set at 2.1 kV and the capillary temperature was 300 °C. The MS parameters were as follows: RF-lens, 40%; Orbitrap resolution; 15,000 (Full MS) and 30,000 (MS²); HCD collision energy, 40%; maximum injection time, 50 ms; AGC target, 4 x 10⁵ (Full MS) and 5 x 10⁴ (MS²); the quadrupole isolation width was set at 1.6 m/z; data type; profile. A Michrom Magic C18 AQ column (150 x 0.1 mm, 3 μm particle size, 100 Å pore size) (Auburn, CA) was employed for chromatography. The mobile phases consisted of 2 mM NH₄OAc (A) and CH₃CN (B). The flowrate was 600 nL/min, and the initial solvent composition (1% B) was maintained for 4 min and then increased to 90% B at 19 min. After holding at 99% B for 1 min, the gradient returned to 1% B in 2 min, and the system was re-equilibrated for 5 min before the next injection. The samples (2 μL) were injected with a 10 μL load loop. The precursor ions for PMOA-dR and its internal standard PMOA-[¹³C₅]dR were m/z 241.1 and 246.1, respectively. Both compounds underwent fragmentation to produce the same product ions. EICs were extracted for m/z 92.0495, 108.0444, 110.0600, and 125.0709, and employed for quantitative measurements.⁴⁷

Metabolomics of TSC in RT4 bladder cells

Cells (3 x 10⁶) were treated TSC, NF, and BF (300 μg/mL) or blank pad extracts serving as the negative control for 6 h, and the cell lysates were prepared as described above. Metabolites were resolved with the same ACQUITY UPLC HSS T3 VanGuard pre-column and ACQUITY HSS T3 column employing the same gradient conditions and ion source parameters as reported for the targeted analyses for GSH, GSSG, and GSH-conjugates. The spray voltage was 2.1 kV in the negative ion mode and 3.0 kV in the positive ion mode. The MS parameters were as follows: RF-lens, 35%, Orbitrap resolution, 120,000; scanning range, m/z 100 – 700 (or m/z 300 – 700); AGC target was 4 x 10⁵ (Full MS). The MS² acquisition was performed by data-dependent acquisition (DDA) (ddMS²) for ions having an intensity above 2 × 10⁴ and a charge state of 1 or undetermined. The data-dependent mode cycle time was 3 s between master scans. A dynamic exclusion of 6 s was employed to eliminate the continuous fragmentation of the same precursor (with a mass tolerance of 10 ppm) after each MS² scan. The Orbitrap resolution at MS² was 30,000; isotope exclusion was assigned; the HCD was stepped at 25, 35, and 65; the maximum injection time was 200 ms; the normalized AGC target was 1.6x10⁶; the quadrupole isolation width was set at 1.6 m/z; data type, profile. Four DDA MS² scans were also obtained with these parameters using the AcquireX data acquisition workflow employing the Deep Scan method. All samples were randomized in the run order to eliminate bias and reduce the variance caused by changes in the column or MS performance.⁴⁸

Data processing

Compound Discoverer^™ (CD3.3, version 3.3.1.111, Thermo Fisher Scientific) was used to process and visualize metabolomics data, search MS² spectra for compound identification, and generate PCA and volcano plots. Feature finding and peak detection mass tolerance settings were set at 7 ppm; a minimum peak intensity of 15,000, and a signal to noise (S/N) threshold of 1.5 required with baseline removal for feature finding and EIC generation. Alignment between samples used the ChromAlign algorithm using a mid- experiment quality control (QC) sample as the reference file. The preferred ions for feature finding were the [M+H]⁺¹ and [M−H]⁻¹ ions, and compound abundances were calculated from the most abundant ion present. Gap filling used the CD3.3 Fill gaps node with real peak detection, 7 ppm mass tolerance, and S/N threshold of 1.5 over baseline signal. The data were normalized using the Apply SERRF QC Correction node with a minimum required 50% minimum present in QC samples, with an allowed relative SD (RSD) of compound intensities < 25%, and a random forest #trees = 200.

The MS² spectra acquired by DDA or AcquireX were searched using the CD3.3 Search mzCloud and Search mzVault workflow nodes and matched against the mzCloud^™ online database (https://www.mzcloud.org/) and the NIST/EPA/NIH 2020 commercial Mass Spectral Library. The mzCloud node used a confidence forward search, and the cosine similarity algorithm. The mzVault node used the NIST search algorithm; both search nodes employed 10 ppm precursor and fragment ion tolerances with best match scores > 70. The assigned compound names used mzCloud IDs as the preferred compound name over those found in mzVault. In some instances, MS² spectra were manually searched against Massbank (https://massbank.eu/MassBank/).⁴⁹ The identification of compounds and confidence levels followed the criteria reported by Schymanski and co-workers.⁵⁰ The chemicals considered "identified" were Level 1 confirmed with a reference standard and Level 2 probable identification by library spectra matches. All MS² spectra identified and discussed in this manuscript were manually interpreted for validity. A detailed summary of the workflow nodes used and their settings are included as a Supplementary text file.

Statistics.

Statistics on targeted analytes and other endpoints were performed with Prism 5.03 (GraphPad Software, La Jolla, CA). The statistical significance effect of treatment within a group was determined by the Student's t-test. The data are reported as the mean ± SD (* P < 0.05; **P < 0.01, ***P < 0.005 versus control). All studies were performed with at least three independent experiments in triplicate.

The effects of TSC, NF, BF, and the negative control filter pad were examined for global changes in the metabolome. Each fraction was assayed in quadruplicate, and three independent experiments were completed. Intra experiment clustering of samples were visualized using PCA, and volcano diagrams were plotted using CD3.3. ANOVA p values were automatically adjusted for the family-wise error rate using Tukey's HSD in CD3.3. Volcano plots with −log10 p-value and log2-fold changes are reported, and significant features were reported with p < 0.05 and log2 FC (treatment/negative control filter) > 1.0 or < −1.0.

Results

TSC cytotoxicity in RT4 cells.

TSC and NF (75 – 600 μg/mL) of the reference cigarettes (1R6F) resulted in a concentration- and time-dependent cytotoxicity in RT4 cells (Figure 1A and 1B). In contrast, no cytotoxicity was observed when RT4 cells were treated with BF at the same concentrations (Figure 1C). A combined NF and BF mixture resulted in comparable cytotoxicity profiles as those observed with NF treatment alone. (Figure 1D). Comparable cytotoxic effects were observed when Marlboro TSC was employed for assay (Figure S1). We note that 600 μg/ml of TSC is equivalent to 2% of a cigarette per mL of cell culture media.

Figure 1.

TSC, NF, and BF cytotoxicity in RT4 cells. Cells were treated with either blank pad extracts (control), TSC, NF, BF, or NF and BF combined (75 – 600 μg/mL) for up to 96 h. Every 24 h the cell treatment was renewed, and cell viability was evaluated by MTT assay. (*P < 0.05; **P < 0.01, ***P < 0.005).

TSC and NF induce oxidative stress in RT4 cells.

It is well established that tobacco smoke induces oxidative stress.^51,52 ROS induction was measured in viable RT4 cells treated with TSC, NF, and BF (75 - 600 μg/mL) after 6 h of treatment where the cell viability was greater than 75%. We employed the cell-permeant indicator carboxy-H₂DCFDA, which is better retained in the cell than its analog, H₂DCFDA, and thus, more efficiently traps the ROS generated over time. TSC and NF induced a concentration-dependent ROS formation, by measurement of the oxidized intracellular fluorescent intermediate of carboxy-H₂DCFDA, while the BF did not induce a significant induction of ROS (Figure 2A). The levels of ROS increased over 6 h of treatment (Figure S2). TSC and NF also significantly decreased the GSH:GSSG ratio in RT4 cells, whereas the BF did not (Figure 2B). Reduced GSH is a vital ROS scavenger, and a decreased GSH:GSSG ratio is a marker of oxidative stress induction.⁵³ Oxidative stress can lead to lipid peroxidation, generating reactive aldehydes, including malondialdehyde (MDA).⁵⁴ A significant concentration-dependent increase in MDA levels occurred in RT4 cells treated with TSC and NF but not in cells treated with the BF (Figure 2C). Oxidative stress induces numerous types of DNA damage, including oxidized DNA lesions, single- and double-strand breaks, and the AP site formation; many of these lesions are toxic and/or mutagenic.⁵⁵ AP site levels increased in RT4 cells treated with TSC and NF but not in cells treated with BF (Figure 2D).

Figure 2.

TSC, NF, and BF oxidative stress induction in RT4 cells. Cells were treated for 6 h with blank pad extracts (control), TSC, NF, and BF (75 – 600 μg/mL), (A) levels of ROS trapped by carboxy-H₂DCFDA, (B) the relative abundance of GSH and GSSG measured by HRAMS LC/MS², (C) MDA levels measured by TBARS assay, and (D) the levels of AP sites levels were measured by targeted HRAMS LC/MS². (*P < 0.05; **P < 0.01, ***P < 0.005).

TSC, NF, and BF impact on the intracellular GSH-metabolome in RT4 cells.

Since TSC treatment resulted in a significant depletion of GSH in RT4 cells, we characterized the impact of TSC, NF, and BF on the GSH metabolome in RT4 cells by full scan and untargeted HRAMS LC/MS² based metabolomics approach scanning the m/z 300 - 700 range in the negative ion mode. Treatment with TSC, NF, and BF led to significant changes in the metabolome of RT4 cells , as shown by the PCA plot (Figure 3A). The volcano plot shows that TSC treatment increased the abundance of 414 compounds and decreased 71 compounds compared to the negative control cells over this mass range (Figure 3B).

Figure 3.

PCA biplots and volcano plots of RT4 cells treated with TSC, NF, and BF. Cells were treated with blank pad extracts (control), TSC, NF, and BF (300 μg/mL) for 6 h, and the RT4 cells metabolome was analyzed by untargeted HRAMS LC/MS² metabolomics scanning from m/z 300 - 700 in the negative ion mode. (A) The PCA clustering biplot shows the separation of the TSC fraction treatments, (B) volcano plot of the log2-fold changes of compounds in cells altered by TSC, and (C) volcano plot of putative GSH-conjugates detected by GSH class coverage. Red (higher in TSC) and green boxes (higher in the negative control) show the cutoff of −1*log10 p value2 (p < 0.01) and log2-fold-change (FC) > 1. Highlighted blue circles correspond to the hydroquinones and catechol GSH conjugates.

GSH and many GSH conjugates undergo HCD fragmentation in the negative ion mode to produce common product ions and a signature fragment ion containing the Cys sulfur atom of the adducted electrophile (Scheme 1).^56,57 We filtered for putative GSH conjugates formed in TSC-treated cells employing class coverage. A minimum of 4 principal fragment ions of GSH within a 10-ppm m/z tolerance were required for tentative identification. This stringent class coverage revealed an increase of 44 putative GSH conjugates (> log2-fold change, p < 0.01) in RT4 cells treated with TSC (Figure 3C). The NF and BF increased the number of GSH conjugates by 39 and 6, respectively, compared to negative control cells. (Figure S3). We hypothesized that the GSH depletion in TSC-treated cells could be attributed to HQ derivatives and Cre (Scheme 2) that undergo redox cycling and form GSH conjugates (Scheme 3).^58-61 The presumed GSH conjugates of HQ, Cat, MeHQ/MeCat, and DiMeHQ were among the most abundant GSH conjugates detected in TSC, and NF treated cells (Figure 3C). RT4 cells were treated with synthetic HQ and Cat at concentrations reported in TSC.⁶² The identities of the HQ-SG (t_R 10.2 min) and Cat-SG (t_R 11.6 min) in TSC-treated cells were confirmed by co-elution with the biosynthetic conjugates using targeted HRAM LC/MS² (Figure 4A). The structures of the isomeric MeHQ-SG and MeCat-SG conjugates were not assigned. 2-MeHQ, 3-MeCat, 4-MeCat, but not 6-MeCat, have been reported in TSC.^62-64 We note that o-cresol (o-Cre), p-Cre, and m-Cre are also present in TSC.^62-64 p-Cre can undergo P450 oxidative metabolism catalyzed by P450 2E1, to form the 4-MeCat, which following further oxidation to the quinone, forms isomeric conjugates with GSH at the aromatic ring (Scheme 3).⁶⁵ P450 2E1 is highly expressed in the liver and to the best of our knowledge, P450 2E1 is not expressed in human bladder.⁶⁶ p-Cresol also undergoes P450 2E1 oxidation to the quinone methide to form a GSH conjugate at the benzylic carbon (Scheme 3).^65,67 The MS² spectra of the GSH conjugates are shown in Figure 4B. All spectra display the major product ions attributed to GSH and the corresponding product ions attributed to the thiolate adducted electrophiles (Figure 4C).

Scheme 1.

GSH fragmentation by HCD in the negative ESI mode

Scheme 2.

Hydroquinones and cresol structures

Scheme 3.

(A) GSH conjugate formation of HQ, Cat, and (B) p-Cresol bioactivation by P450 followed by GSH conjugate formation.

Figure 4.

HRAMS LC/MS² EICs and product ion spectra of GSH conjugates of HQ/Cat isomers in RT4 cells treated for 6 h with blank pad extracts (control), TSC, NF, and BF (300 μg/mL). The GSH conjugates were analyzed in the negative ion mode. (A) EIC of GSH conjugates of HQ derivatives, (B) MS² mass spectra of HQ, Cat, MeHQ/MeCat, and DiMeHQ GSH conjugates and (C) signature product ions (theoretical m/z value) of the electrophiles adducted to the Cys thiol (R-S⁻ ion). The precursor ions monitored were as follows: HQ-SG and Cat-SG m/z 414.0977, MeHQ/MeCat-SG m/z 428.1133, and DiMeHQ-SG m/z 442.1291. The product ions of the GSH moiety were monitored at m/z 128.0348, 143.0457, 179.0457, 210.0880, 254.0779, 272.0886. Peak areas are reported when above the background levels.

Targeted HRAMS LC/MS² was employed to estimate the relative abundances of these GSH-conjugates in TSC-treated RT4 cells. The GSH-conjugates of HQ, Cat, MeHQ/MeCat, and DiMeHQ were highly elevated in TSC and NF while undetected in BF and untreated cells (Figure 5). Thus, the liquid-liquid extraction procedure of the TSC efficiently recovered the HQ derivatives in the NF.

Figure 5.

Relative abundance of HQ, Cat, isomeric MeHQ/MeCat, and DiMeCat GSH conjugates in TSC, NF, and BF treated RT4 cells. Cells were treated with blank pad extracts (control), TSC, NF, and BF (300 μg/mL) for 6 h, and the relative abundances of HQ-SG, Cat-SG, MeHQ-SG, MeCat-SG m/z, DiMeHQ-SG were measured by targeted HRAMS LC/MS² in the negative ion mode. 2-MeHQ (R₁: CH₃, R₂: H, R₃: OH), 3-MeCat (R₁: OH, R₂: CH₃, R₃: H), 4-MeCat (R₁: OH, R₂: H, R₃: CH₃), 2,3-DiMeHQ (R₁: CH₃, R₂: H), 1,6-DiMeHQ (R₁: H, R₂: CH₃) (*P < 0.05; **P < 0.01, ***P < 0.005).

HQ and Cre cytotoxicity in RT4 cells.

Cytotoxic and carcinogenic effects of various HQ and cresol derivatives compounds are well known.^58-60 Thus, we characterized the cytotoxic potency of tobacco-derived HQs and isomeric cresols in RT4 cells. Mixtures of synthetic HQs were prepared at the same concentrations reported in tobacco smoke (Table 1).⁶³ HQ, Cat, Res, 2-MeHQ, p-Cre, m-Cre, and o-Cre mixtures (0.5 – 100 μg/mL) resulted in a concentration and time-dependent cytotoxicity in RT4 cells (Figure 6A). The level of cytotoxicity induced by this mixture in RT4 cells occurred at concentrations close to HQ and Cre levels present in TSC, assuming 100% recovery of these compounds in the TSC preparation.⁶²

Table 1.

Hydroquinone and cresol abundances in cigarette smoke and dose mixtures used for cell treatments^a

Compound	μg per cigarette	Compound	Mixture assayed
			50 μg/mL	12.5 μg/mL	6.25 μg/mL
HQ	122	HQ	11	3	1
Cat	195	Cat	18	5	2
Res	44	Res	4	1	1
p-Cre	22	p-Cre	2	1	0
m-Cre	60	m-Cre	6	1	1
o-Cre	60	o-Cre	6	1	1
2-MeHQ	28	2-MeHQ	3	1	0
Total μg/cig	531	Total μg/mL in cell media	50	13	6.3

^aThe treatment dose of 600 μg/mL is equivalent to 2% cig/mL of a which equals ~ 11 μg/mL of hydroquinone and cresol mixture.

Figure 6.

HQ and Cre derivatives-induced cytotoxicity, AP site formation, and decreased GSH:GSSG ratios in RT4 cells. Cells were treated with a HQ and Cre derivatives mixture containing HQ, 2-MeHQ, Cat, p-Cre, m-Cre, o-Cre, and Res (0.5 – 100 μg/mL) for 6, 24, and 48 h. Cell treatment was renewed with chemicals every 24 h. (A) Cell viability was evaluated by MTT assay. (B) AP site levels were measured by targeted HRAMS LC/ MS². (C) GSH:GSSG ratios as a function of HQ and Cre treatment and incubation time.

Tobacco-associated quinones induce oxidative stress, and AP site formation in RT4 cells.

Several molecular mechanisms of toxicity have been reported for HQ, Cat, and Cre, including ROS formation via redox cycling with their corresponding quinone and semiquinone anion radicals, and quinone methides of Cre (Scheme 3). These intermediates can form DNA adducts (Scheme 4)^58-61 The oxidized quinone intermediate of Cat reacts with DNA by 1,4-Michael addition to form adducts at the N7 atom of guanine and the N3 atom of adenine. These adducts undergo rapid depurination, resulting in AP sites (Scheme 4).^68,69 We examined cell viability (Figure 6A), AP site formation (Figure 6B), and GSH:GSSG ratios (Figure 6C) in RT4 cells treated with a synthetic HQ and Cre derivatives mixture at levels present in tobacco smoke. ⁶³ Treatment with increasing concentrations of this mixture resulted in cytoxicity and induced AP site formation in a concentration- and time-dependent manner (Figure 6B). The HQ and Cre derivatives mixture treatment also decreased the GSH:GSSG ratio (Figure 6C), and increased the Acr-SG levels (M. Bellamri, unpublished observations). The oxidative stress induced by HQ and Cre derivatives is significantly greater than the untreated, control cells but weaker than the NF and TSC fractions (Fig. 1), suggesting that TSC toxicity involves not only HQ and Cre derivatives but also other chemicals.

Scheme 4.

HQ and Cat oxidation and DNA adduct formation. p-BQ forms p-BQ-dAdo and p-BQ-dCyd, and the oxidized benzoquinone of Cat forms the Cat-4-N7-dG and Cat-4-N3-dAdo adducts with concomitant AP-site formation.

p-Benzoquinone (p-BQ) also reacts dC and dA to form 3-hydroxy-3,N⁴-benzetheno-2’-deoxycytidine (p-BQ-dCyd) and 9-hydroxy-1,N⁶-benzetheno-2’-deoxyadenosine (p-BQ-dAdo) (Scheme 4).^70-72 DNA adducts of HQ, which is a metabolite of benzene, have been detected, by ³²P-postlabeling, in HL-60 cells, a human myeloid cell line; however, these DNA adducts did not correspond to any of the principal adducts formed in DNA reacted with p-BQ/HQ or benzene.^70-75 We confirmed that p-BQ reacts with CT DNA to form p-BQ-dCyd and p-BQ-dAdo; however, thus far, we have not successfully detected either DNA adduct in RT4 cells treated with TSC or HQ mixtures by ion trap MSⁿ.⁷⁶ (M. Bellamri and R. Turesky, unpublished observations). The synthesis of stable, isotopically labeled internal standards will be required for quantitative measurements and unequivocal confirmation that BQ does not form stable p-BQ-dCyd or p-BQ-dCyd adducts in genomic DNA. The p-Cre quinone methide also forms DNA adducts; however, their structures’ identities remain uncharacterized.^77,78

Oxidative stress and endogenous GSH-conjugates of lipid peroxidation.

TSC also induced GSH conjugate formation with endogenous electrophiles, including Acr, MVK, and Crot. (Figure 7A). These volatile aldehydes occur in cigarette smoke, but they are not trapped on the Cambridge filter pads and are present in the gas phase.⁴⁶ The occurrence of these GSH conjugates is likely a result of TSC-induced oxidative stress in RT4 cells. HRAMS LC/MSⁿ confirmed the identities of these GSH conjugates in RT4 cells treated with a non-cytotoxic concentration of Acr, Methacrolein (MeAcr), MVK, or Crot (Figure S4). HRAMS LC/MS² EIC of Acr, MVK, and Crot GSH-conjugates and their MS² mass spectra are shown in Figure 7B, 7C and 7D. However, a GSH conjugate of MeAcr was not detected. Characteristic product ions attributed to GSH, and the signature product ions corresponding to the Cys thiol adducted electrophile (R-S⁻ ion) are seen in the mass spectra of each conjugate (Figure 7E). The GSH conjugates of Acr, MVK, and Crot were only detected as the reduced alcohol conjugates and not the aldehydes. The levels of these GSH conjugates increased over 6 h of treatment before being cleared from the cells at 24 h (Figure 8).

Figure 7:

Relative abundance of Acr, Crot, and MVK GSH conjugates in TSC, NF, and BF treated RT4 cells. Cells were treated with blank pad extracts (control), TSC, NF, and BF (300 μg/mL) for 6 h. The GSH conjugates were analyzed in the negative ion mode. (A) The relative abundances Acr-SG, Crot-SG, and MVK-SG were measured by targeted HRAMS LC/MS². (*P < 0.05; **P < 0.01, ***P < 0.005). (B and C) HRAMS LC/MS² chromatograms, (D) MS² scan ion spectra, and (E) signature product ions (theoretical m/z value) of the electrophiles adducted to the Cys thiol (R-S⁻ ion). The precursor ions monitored were: Acr-SG m/z 364.1184, Crot-SG and MVK-SG m/z 378.1441. The extracted GSH ions were m/z 128.0348, 143.0457, 179.0457, 210.0880, 254.0779, 272.0886.

Figure 8.

Kinetics of Acr, Crot, and MVK GSH conjugate formation in TSC, NF, and BF treated RT4 cells. Cells were treated with blank pad extracts (control), TSC, NF, and BF (300 μg/mL) for 3, 6, and 24 h. The relative abundances Acr-SG, Crot-SG, and MVK-SG were measured by targeted HRAMS LC/MS² in the negative ion mode. (*P < 0.05; **P < 0.01, ***P < 0.005). The precursor ions monitored were: Acr-SG m/z 364.1184, Crot-SG and MVK-SG m/z 378.1441. The extracted GSH ions were m/z 128.0348, 143.0457, 179.0457, 210.0880, 254.0779, 272.0886.

TSC and NF toxicants and modulation of GSH biosynthesis pathway in RT4 cells.

We further characterized the RT4 cell metabolome within the m/z 70-700 range in both positive and negative ion modes. Treatment with TSC, NF, and BF significantly altered the RT4 cells' metabolome, as shown by the clustering of the different extracts depicted in the PCA plot (Figure 9A). TSC treatment increased the abundance of 980 compounds and decreased 56 compounds (Figure 9B). Multiple tobacco-associated chemicals were detected, including nicotine, nornicotine, anatabine, cotinine, Cat, pyrogallol, pyrogallol-O-glucuronide, scopoletin, and scopoletin-O-glucuronide (Figure S5).⁷⁹ The identification of the TSC toxicants was supported by matched mass spectra of reference compounds in the mzCloud or NIST high-resolution MS/MS databases. (Figures S6). The identities of these chemicals were confirmed by targeted HRAMS LC/MS². The synthetic chemicals had identical MS² spectra to those toxicants present in TSC-treated RT4 cells (M. Bellamri, unpublished data).

Figure 9.

PCA biplot and volcano plot of RT4 cells treated with TSC, NF, and BF. Cells were treated with blank pad extracts (control), TSC, NF, and BF (300 μg/mL) for 6 h. The RT4 intracellular metabolome was analyzed by untargeted HRAMS LC/MS² over a m/z 70 - 700 range in the positive ion mode. (A) PCA clustering biplot shows the separation of the treatment groups, (B) volcano plot of the log 2-fold changes of compounds in cells altered by TSC. Red (higher in TSC) and green boxes (higher in the negative control) show the cutoff of −1*log10 p value (p < 0.01) and log2-fold-change (FC) > 1 or <−1.

Cat was among the TSC toxicants identified by untargeted MS and MS² scanning employing AcquireX. The retention time and MS² spectrum of Cat in TSC treated RT4 cells matched those of the Cat synthetic standard, whereas the isomeric HQ (t_R = 10.6 min) and Res (t_R = 7.3 min) had different retention times (Figure 10). The signal of Cat response on a per mol basis was >50-fold higher than the signals of HQ or Res. Thus, these other benzene-diol isomers may have escaped detection in the cell lysates. The partitioning of these various chemicals into the NF or BF demonstrates the efficiency of liquid-liquid extraction to separate different classes of tobacco toxicants as a function of their polarities and chemical functional groups.

Figure 10.

HRAMS LC/MS² chromatograms (A) and MS² scan ion spectra (B) of HQ, Cat, and Res synthetic standards and in the lysates of TSC- treated cells. Ion spectra were obtained by targeted HRAMS LC/MS² in negative ion mode.

TSC, NF, and BF also significantly altered GSH biosynthetic pathway compared to the negative control cells. The relative abundances of GSH, GSSG, cysteinylglycine, γ-glutamyl cysteine, glutamate, and S-adenosylmethionine (SAM) were significantly decreased. In contrast, the levels of methionine, γ-glutamylleucine, γ-glutamyl lanine, 5-oxoproline, and S-adenosylhomocysteine (SAH) were increased (Figure 11). These compounds’ identities were based on matching MS² spectra in the mzCloud database (Figure S8). The BF treated cells' GSH metabolome pathway was comparable to those of negative control cells (Figure 11).

Figure 11.

Modulation of intermediates in the GSH biosynthesis pathways in RT4 cells treated with TSC, NF, and BF. After 6 h of treatment with blank pad extracts (control), or TSC, NF, and BF (300 μg/mL), the RT4 cell metabolome was analyzed by untargeted HRAMS LC/MSⁿ over an m/z 70 - 700 window in positive and negative ion modes. The intermediates of GSH biosynthesis pathways were identified by MS² spectra searches using mzCloud or the NIST 2020 library databases. Targeted HRAMS LC/MS² of GSH intermediates was done to confirm the data-dependent data obtained by AcquireX.

Discussion

Despite decades of research on the health risk of tobacco smoking, the chemicals responsible for BC in smokers remain uncertain. Herein we demonstrate that TSC induced a concentration- and time-dependent cytotoxicity, the induction of oxidative stress, lipid peroxidation, and DNA damage in RT4, a well-established human bladder cell line. Only a few studies have investigated TSC toxicity in human bladder cells.^80-85 These earlier studies also reported that a TSC concentration exceeding 1% cigarette per mL induced concentration-dependent cytotoxicity in T24, a human bladder cancer cell line, and SV-HUC-1, an SV-40 immortalized human urothelial cell line. These studies also showed that TSC induces molecular and morphological changes resulting in an epithelial-to-mesenchymal transition and enhanced invasiveness; both are critical features in cancer development. However, to the best of our knowledge, our study is the first to show that chemicals in the NF of TSC are responsible for the induced cytotoxic effects in human bladder cells.

Cigarette smoking is associated with chronic and progressive inflammation, a key factor for the pathophysiologic development of tobacco-related diseases.⁸⁶ Inflammatory microenvironment is characterized by increased ROS and RNS, leading to oxidative stress induction. TSC and NF-induced cytotoxic effects in RT4 cells are associated with oxidative stress induction marked by ROS formation, lipid peroxidation, and GSH depletion. Oxidative stress induction by TSC occurs in several lung cancer cell lines and normal human bronchial epithelial cells.^87-90 TSC-induced oxidative stress is associated with cell transformation, epithelial-to-mesenchymal transition, cell migration, and inflammation, key features in cancer pathology.^87-90 Moreover, in smokers, oxidative stress markers including lipid-peroxidation products, MDA, nitric oxide (NO) production, superoxide dismutase (SOD), and glutathione peroxidase (GSH-Px), and 8-oxo-2'-deoxyguanosine (8-oxo-dG) in exhaled breath condensate and F2-isoprostanes in plasma are significantly higher compared to nonsmokers.^91,92 However, the oxidative stress potency of TSC in human bladder cells has not been reported. Further studies are warranted to understand the role of oxidative stress in BC pathology in smokers.

Our study shows that TSC and NF-induced GSH depletion in RT4 cells are associated with the abundant formation of GSH conjugates of the oxidized benzoquinone derivatives of HQ, Cat, and their methylated derivatives. Moreover, the synthetic hydroquinone and cresol mixtures induced significant cytotoxic effects in RT4 cells at concentrations present in TSC. IARC classifies HQ and Cat as group 3 and group 2A carcinogens, respectively.⁹³ Cat is carcinogenic in rodents and strongly co-carcinogenic on mouse skin when applied together with B[a]P.^94,95 Cat is also a metabolite of benzene, which causes leukemia in humans and animals.^96,97 Hydroquinones and cresols induce cytoxicity marked by their redox cycling and resulting in ROS formation, free radicals, GSH conjugate formation, and GSH depletion.^58-61 HQ undergoes oxidation to p-BQ, which reacts with DNA in vitro; major adducts form with dA and dC (Scheme 4).^70-75 However, these laboratories did not detect either DNA adduct in human cells; thus far, our laboratory also has not detected these lesions in human cells (unpublished observations, M. Bellamri and R. Turesky).

p-Cre also forms DNA adducts following metabolic activation by P450 2E1; however, the structure of the DNA adducts are unknown.^77,78 P450 2E1 is highly expressed in the human liver and plays a crucial role in the metabolism of endogenous molecules and xenobiotics.⁶⁶ The P450 2E1c1/c1 polymorphism is associated with higher BC risk.⁹⁸ However, to the best of our knowledge, the P450 2E1 protein expression and functional activity have not been reported in human bladder tissue or bladder cell lines, including RT4 cells.

Cat and other quinones, including the estrogen quinone, react with the guanine N7 atom, and adenine N3 and N7 atoms, by 1,4-Michael addition, to form depurinating DNA adducts and resultant AP sites (Scheme 4).^60,68,69 The AP site is a ubiquitous DNA lesion induced by various processes, including alkylation and deamination of bases, oxidative stress, or radiation, which damage DNA bases.⁹⁹ DNA glycosylases excise these lesions, and the resulting transitory AP sites are the central intermediate in the DNA base excision repair (BER) pathway.^100-102 The aldehyde reactive probe assay estimates AP site levels at a steady state of up to 30,000 lesions per cell.¹⁰³ BER rapidly repairs individual AP sites, while clustered lesions can persist in cells for more than one day.^104,105 AP sites that escape repair or are misrepaired can lead to G to T and G to C transversion mutations associated with cancer development.¹⁰⁶ The AP site measurements in human organs by quantitative LC/MS methods are limited.^107,108 These studies reported no significant increase in AP site levels in human smoker lung biopsies compared to non-smokers. However, the small number of subjects precludes firm conclusions.¹⁰⁷ Our data demonstrate that tobacco-related hydroquinones and cresol mixtures and TSC treatment significantly increase AP site levels in RT4 cells. To our knowledge, AP site measurements have not been reported in the human bladder. Further mechanistic studies on the role of TSC and its related HQ and Cat derivatives in DNA adduct and AP site formation in smokers' bladder DNA are warranted.

TSC treatment also induces Acr-SG, MVK-SG, and Crot-SG conjugate formation in RT4 cells. Although these volatile enals form in tobacco smoke, they are present in the gas phase and not found in the particulate fraction.⁴⁶ These GSH conjugates are likely formed by endogenous enals formed during the oxidative stress induced by TSC in RT4 cells. The Acr-SG, MVK-SG, and Crot-SG reached their maximum levels at 6 h of treatment with TSC and then were cleared from the cells at 24 h. These GSH enal conjugates were also formed in B16–BL6 mouse melanoma cell line exposed to the tobacco smoke gas phase.^109,110 In contrast to our findings, the GSH enal conjugates rapidly formed within 5 min and then were cleared from B16-BL6 cells by 30 min post-treatment.^109,110 This difference in the kinetics of removal of these GSH enal conjugates is likely due to the cell treatment with the different fractions of tobacco smoke. With the gas phase treatment, the cells are directly exposed to the reactive aldehydes which are rapidly detoxified by GSH and cleared from the cell. However, cells treated with TSC continuously induce ROS, in part through redox cycling of benzene-diols (Schemes 3, 4), resulting in enal formation and prolonged induction of GSH enal conjugates (Figure 2A and Figure S2).

At the outset of this study, we anticipated that the BF containing AA, HAA, and NOC would exert potent cytotoxicity and DNA damage. However, the BF toxicity is negligible compared to the toxic effects of TSC and NF. Extrahepatic P450 1A1, P450 1B1, P450 2A6, P450 2A13, and P450 2E1 bioactivates AA, HAA, and NOC;^111,112 however, these enzymes' functional activities are very low in RT4 cells and likely insufficient to bioactivate these carcinogens in the complex TSC or BF matrix efficiently. In contrast, human liver expresses these key enzymes at high levels allowing for the metabolic activation of a wide range of tobacco toxicants.^111,113 Genotoxic N-hydroxy-AA and N-hydroxy-HAA metabolites formed in the liver can reach the bladder and other extrahepatic organs, where they under further bioactivation by phase II enzymes producing the penultimate intermediates that bind to DNA.^111,114 Therefore, the cytotoxic and genotoxic potency of BF is likely underestimated in this study due to the lack of the hepatic bioactivation system. Studies assessing the deleterious effects of BF on human bladder cells are currently underway employing co-culture systems of human bladder cell lines with primary human hepatocytes.

Conclusion

Investigations employing rodent and human cells have greatly advanced our knowledge of mechanisms of genotoxicity, DNA damage, and DNA repair for many toxicants.¹¹⁵ Our findings highlight the potent toxicity of HQ and other phenols in the acid-neutral fraction of TSC in human bladder cells in vitro. However, isolated cell systems may not capture the entire metabolic fate of TSC toxicants that occurs in vivo. Moreover, the route of exposure and the bioavailability of the toxicant can dramatically impact its biological effects in human target organs. Tobacco smoke induces toxicity in the lung, a primary target organ of tobacco smoke exposure; however, the levels of some chemicals in tobacco smoke that reach distant organs and induce deleterious effects are uncertain. Therefore, extrapolating toxicity mechanisms of tobacco-associated chemicals in isolated cell-based systems for human risk assessment of bladder cancer must be done with caution. Biomarkers of DNA damage in the bladder are required to advance our understanding of the mechanisms of tobacco-associated bladder cancer.