Urinary Pyrene Carboxylic Acid as a Novel Exposure Biomarker of Woodsmoke

Abstract

Quantifying people’s exposure to wildfires is essential for assessing related health risks. While hydroxyl metabolites of polycyclic aromatic hydrocarbons (PAHs) are commonly used exposure biomarkers of combustion-originated air pollutants, methylated PAHs are more abundant in woodsmoke than other sources. Thus, urinary PAH carboxylic acids, which are metabolites of methylated PAHs, may serve as more sensitive biomarkers of wildfire exposure. In this exploratory study, we developed an LC-MS/MS method to simultaneously quantify hydroxylated and carboxylic metabolites of PAHs and methyl-PAHs in urine. This method was then applied to 56 urine samples collected from 8 campers before, during, and after a 4-hour exposure to campfire. Campers also wore silicone wristbands to monitor ambient PAHs. We found that 1-pyrenecarboxylic acid (1-PYRCA) levels increased significantly at 4 h (96.9%, 95% CI: 2.60–101%), 6 h (96.8%, 95% CI: 5.85–107%), and 8 h (92.5%, 95% CI: 3.59–99.2%), and returned to baseline levels at 24 h. In contrast, the campfire exposure did not significantly increase other urinary PAH metabolites. Wristband PAHs also significantly increased during the 4-hour exposure. These results suggest the use of urinary 1-PYRCA as a sensitive exposure biomarker for woodsmoke and potentially for assessing exposure to wildfires.

Graphical Abstract

Introduction

Climate change is driving an increase in the frequency, intensity, and duration of wildfires.^1,2 Rising global temperatures and drier conditions have made wildfires near-annual events in North America, severely impacting air quality and posing significant health risks.³ Epidemiological studies have linked wildfire smoke exposure to respiratory and cardiovascular illnesses, findings that are supported by controlled human and animal studies.^4,5 Despite these known risks, the lack of feasible biomarkers at the dose-response level limits timely, accurate evaluation of wildfire smoke exposure and its health consequences. Developing sensitive exposure biomarkers is crucial for evaluating and understanding the wildfire health impacts, particularly in a rapidly changing climate.

Polycyclic aromatic hydrocarbons (PAHs) are products of incomplete combustion of fuels (e.g., fossil fuels, biomass), waste, and tobacco. Urinary hydroxylated PAHs (OH-PAHs), particularly 1-hydroxypyrene (1-OHPYR), are widely recognized as biomarkers of PAH exposure.^6,7 As a metabolite of pyrene, 1-OHPYR is a sensitive and reliable marker for both occupational and environmental PAH exposure, with studies linking it to oxidative stress, inflammation, and DNA damage.^8–11 However, woodsmoke, a significant source of PAH exposure, also contains relatively higher levels of methyl-PAHs (MPAHs), a subset of alkylated PAHs, compared to fossil fuel emissions.¹² Despite their abundance, there has been limited research on metabolites of methyl-PAHs as potential biomarkers for woodsmoke exposure.¹³ Given the distinct composition of woodsmoke, metabolites of MPAHs may serve as more sensitive and specific biomarkers for assessing wildfire smoke exposure, compared to OH-PAHs such as 1-OHPYR which is a generic biomarker for all combustion sources.

Previous studies have identified hydroxylated-MPAHs and carboxylic acids (PAH-CAs) as the primary metabolites of MPAHs, detectable using gas chromatography-mass spectrometry (GC-MS).^14–17 For example, a method was developed to measure 11 methyl-naphthol (Me-OHNs), metabolites of methylnaphthalene, in urine via GC-MS.¹³ Similarly, GC-MS has been used to quantify 2-naphtholic acid (2-NAPCA) and 2-phenanthrene carboxylic acid (2-PHECA) in urine samples.¹⁸ Furthermore, urinary levels of 2-PHECA have been shown to correlate with personal exposure to phenanthrene in ambient air.¹⁷ While GC-MS methods provide robust detection capabilities, their application can be limited by lengthy sample preparation and instrument run times. High-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS), on the other hand, offers greater sensitivity, higher throughput, and versatility, making it well-suited for analyzing complex biospecimen such as urine.

This study aims to (1) develop a HPLC-MS/MS method for the simultaneous quantification of five pairs of OH-PAHs and PAH-CAs in human urine samples, with their chemical structures illustrated in Figure S1; and (2) validate the method using 56 urine samples collected from eight volunteers before, during, and after 4-hour campfire smoke exposure. By assessing which of these target metabolites are associated with campfire exposures, we aim to identify biomarkers specific to woodsmoke (and hence wildfire) exposure.

Materials and Methods

Study design and sample collection.

We used biomass-fueled campfires as a proxy for wildfire and exposed non-smoking volunteer campers to controlled campfires managed by study staff. Participants were ≥18 years, free of cardiopulmonary disease in the past six months, and willing to remain near a campfire for 4 hours. They were instructed to avoid barbecued or fried foods for 2 days before and during exposure and completed 24-hour dietary recalls (ASA24) on the exposure day. Written informed consent was obtained, and the protocol was approved by Duke University’s Campus Institutional Review Board. Campfires were fueled by hardwood, dead leaves, and branches, simulating wildfire conditions as described previously.¹⁹ During exposure, a personal micro-aethalometer measured black carbon in real time. Each participant wore a pre-cleaned silicone wristband 4 hours before, during, and after exposure to assess ambient PAHs. Wristbands were stored at −20°C until analysis. Urine samples were collected in 50 mL tubes at 0, 1, 2, 4-, 8-, 12-, and 24-hours from the start of the exposure, stored on ice and transferred to a −20°C freezer.

Sample Processing and Analytical Methods.

Urine samples were processed to quantify OH-PAHs and PAH-CAs using a validated LC-MS/MS method. The procedure involved enzymatic hydrolysis, liquid-liquid extraction (LLE), and solid-phase extraction (SPE), followed by quantification using electrospray ionization tandem mass spectrometry. Method performance (including LOD, recovery, linearity, and matrix effects) was evaluated using spiked pooled urine and standard calibration curves.

Silicone wristbands were analyzed for a suite of 23 PAHs using gas chromatography–high-resolution mass spectrometry (GC-HRMS). The analytical methos were previously published²⁰. Deuterated internal standards were used to estimate recovery and quantify PAHs. Among the 23 target compounds, 12 PAHs were consistently detected above LOD across samples and were included in statistical analyses.

Full analytical details are in the Supplementary Materials (Sections S1, Tables S1–S3, Figure S1).

Statistical analysis.

The exposure start time was defined as t = 0 h. External exposure was assessed using wristband total PAHs and black carbon. Wristband PAHs averaged from −4 to 0 h for pre-exposure, 0 to 4 h during exposure, and 4 to 8 h post-exposure. Internal exposure was measured through urinary PAH metabolites. Statistical comparisons were performed using the Wilcoxon signed-rank test, a non-parametric test for paired data. Post-hoc pairwise comparisons (during vs. before, during vs. after) followed a significant Friedman test, with *p*-values adjusted via Bonferroni correction (α = 0.0167).

To investigate the temporal changes in PAH metabolites, we used mixed-effects models with sample collection time as a factor (before, 2h, 4h, …, 24h corresponding to t = 0, 2, 4, …, 24 hours) and a random intercept for participants. Significance was determined at p < 0.05 (*), and highly significant results were indicated by p < 0.01 (**). All statistical analyses were conducted using R with the lme4 and lmerTest packages (www.r-project.org).

Results

Method Performance.

The LC-MS/MS chromatograms of PAH-CAs and OH-PAHs showed clear separation in both water and urine matrices (Figure 1). The LOD and calibration standard curve results are provided in Table S3. The mean recoveries for all analytes ranged from 61.0% to 118% for the low spike (2 ng/mL), 53.6% to 101% for the medium spike (10 ng/mL), and 58.0% to 120% for the high spike (40 ng/mL) (Table S4). Relative standard deviations across the three concentrations remained below 25%, indicating good reproducibility and reliability of the method. Among the 56 urine samples analyzed, five OH-PAHs and five PAH-CAs were quantified, with detection rates ranging from 69.6% to 100%, except for 2-OHDF, which was not detected in any of the samples. The median concentrations and ranges of these metabolites, adjusted for creatinine, are summarized in Table S5.

Figure 1. LC-MS/MS Chromatogram of 5 ng/mL Standards in Water and Urine Samples: (a) PAH-CA (Carboxylic Acid Group) and (b) OH-PAH (Hydroxylated-PAH Group).

2-NAPCA: naphthalene-2-carboxylic Acid; 4-FLUCA: fluorene-4-carboxylic acid; 4-DFCA: dibenzofuran-4-carboxylic acid; 2-PHECA: phenanthrene-2-carboxylic acid; 1-PYRCA: 1-pyrenecarboxylic acid; 2-OHNAP: 2-naphthol; 2-OHFLU: 2-hydroxy-fluorene, 2-OHDF: 2-hydroxyl-dibenzofuran; OHPHE: hydroxy phenanthrene; 1-OHPYR: 1-hydroxy pyrene

Participant characteristics.

The study participants consisted of eight healthy adults (four men, four women) with a mean age of 34.8 ± 7.2 years and BMI of 23.3 ± 4.8 kg/m².

External Exposure of Woodsmoke.

External exposure of woodsmoke was evaluated by measuring black carbon in the air and PAHs in the wristband. The average concentrations of summed PAHs in the wristbands were 14.7 ± 5.1 ng/g before exposure, 76.9 ± 56.4 ng/g during exposure, and 19.7 ± 10.5 ng/g after exposure. The PAHs levels during the exposure were significantly higher as compared to pre- (p < 0.001) or post-exposure (p < 0.001) periods. There was no significant difference in PAHs levels between pre- and post-exposure values (p = 0.12). Friedman tests confirmed significant exposure-related changes in 9 out of 12 individual PAHs (p ≤ 0.002); concentrations of chrysene, benzo(a)pyrene, and benzo(c)phenanthrene did not vary significantly (Figure S3). The average concentrations of black carbon were 13.8 ± 9.6 μg/m³ during the exposure. The black carbon concentration was positively correlated with wristband PAH levels during the exposure (Rₛ = 0.95, p < 0.001).

Temporal Patterns of PAH Metabolites.

Urinary PAH metabolite concentrations adjusted by specific gravity (SG) and creatinine are highly correlated with each other (Rₛ = 0.89, p < 0.001). Unadjusted, SG-adjusted, and creatinine-adjusted levels of 1-OHPYR and 1-PYRCA exhibited comparable trends (Figure S5). Creatinine-adjusted values were selected for reporting to align with standardized methodologies. Concentrations of hydroxy-PAHs and PAH-CAs increased following woodsmoke exposure, peaking between 4- and 8-hours post-exposure and returning to baseline approximately 24 hours later. Concentrations of 1-PYRCA increased significantly after 4 hours (96.9%, 95% CI: 2.60–101%), 6 hours (96.8%, 95% CI: 5.85–107%) and 8 hours (92.5%, 95% CI: 3.59–99.2%), respectively, displaying a clear rise-and-fall excretion pattern across participants (Figure 2). Other metabolites showed elevated levels following exposure but lacked statistical significance. OH-PAHs exhibited a more modest increase, with levels rising by approximately 10% at 12 hours post-exposure.

Figure 2. Temporal changes in PAH metabolites in urine before, during, and after 4-hour woodsmoke exposures at intervals of 0, 1, 2, 4, 8, 12, and 24 hours.

Results are based on mixed-effects models, with time as a fixed effect and a random intercept for participants. p < 0.05 indicates significance (*), and p < 0.01 indicates highly significant results (**).

Discussion

To identify sensitive biomarkers for quantifying individual wildfire exposure, we developed a method to measure urinary PAH-CAs and OH-PAHs. In our controlled campfire study, 1-PYRCA emerged as a more sensitive biomarker of wildfire exposure than other PAH metabolites, showing significant increases during the exposure period at 4-, 6-, and 8-hours post-exposure. This supports its potential utility for short-term exposure assessment woodsmoke in real-world environments.

This study presents the first characterization of five urinary PAH-CAs using LC-MS/MS. While the analytical parameters for OH-PAHs in HPLC and ESI-MS/MS are well-established, this method expands the scope of PAH exposure assessment by simultaneously measuring both OH-PAHs and PAH-CAs. In our study, all five PAH-CAs exhibited a characteristic loss of 44 Da, [COO] (Table S2). The fragmentation of PAH-CAs is consistent with other carbolic acids, such as fatty acids, eicosanoids, and bile acids.²¹ Previous studies have shown that OH-PAHs exhibit characteristic ion fragmentation, with a loss of 28 Da ([CO]) in the negative mode of ESI.^22,23 Notably, the collision energies required for OH-PAHs were higher than those for PAH-CAs, reflecting the greater dissociation energy needed to cleave hydroxyl ([OH]) groups compared to carboxyl ([COOH]) groups. These findings highlight structural differences in fragmentation patterns between PAH-CAs and OH-PAHs, contributing to the development of more robust analytical methods for PAH exposure assessment.

The biological half-life of 1-PYRCA could not be reliably estimated due to the small sample size and follow-up time limited to 24 hours. Nevertheless, our results indicated that campfire-induced 1-PYRCA elevations were fully resolved within 24 hours, supporting its use for capturing short-term exposures. In our previous study, urinary cyclooxygenase metabolites of arachidonic acid reached their peak at 12 hours following exposure,¹⁹ suggesting that exposure biomarkers like 1-PYRCA respond earlier than health effect biomarkers. This temporal pattern supports its applicability for detecting acute PAH exposures, such as from such as from woodsmoke and likely wildfire smoke.

The baseline and post-exposure concentrations of 1-OHPYR (average across 2–8 h post-exposure) were 0.02 ± 0.01 and 0.02 ± 0.03 μg/g creatinine, similar to a prior study of 2 h woodsmoke exposure (0.12 ± 0.06 and 0.11 ± 0.06 μg/g creatinine).²⁴ Other hydroxy-PAHs showed comparable patterns (Figure S2). The earlier study took place in a closed yurt-like structure, likely causing higher exposure than our open-air setting. To our knowledge, this is the first report of 1-PYRCA baseline and post-exposure levels (0.05 ± 0.05 and 0.64 ± 0.57 μg/g creatinine, respectively). The median baseline 2-PHECA was 0.551 μg/g creatinine (IQR: 0.16–0.85), similar to our Beijing study (0.22; IQR: 0.12–0.48).¹⁷ In Beijing, urinary 2-PHECA correlated weakly with 2-methylated-PHE in paired personal PM_2.5 samples (Pearson r = 0.15), suggesting substantial contributions from non–air pollution sources. Despite low baseline wristband PAHs in our study, 2-PHECA levels matched those in Beijing, again indicating likely influence from non-air sources that may mask campfire-related exposure to methylphenanthrene.

PAH exposure originates from various sources, including industrial emissions, traffic exhaust, residential heating, personal care products, smoking, secondhand smoke, diet, wildfires, and volcanic eruptions.^25–27 Different PAH metabolites are linked to specific sources.²⁸ In this study, the high baseline levels of 2-OHNAP (Figure S4) may be attributed to routine daily exposures, such as mothballs, insecticides, air fresheners, and tobacco smoke. One participant exhibited elevated baseline levels and was the only individual whose pre-exposure wristband showed detectable levels of BaP, Benzo(e)pyrene, and chrysene (Figure S3), likely due to exposure to barbecue smoke the night before the field study. Correspondingly, increased baseline levels of 2-PHECA, 2,3-OHPHE, and 2-OHFLU in this participant suggest potential contributions from dietary intake or cooking-related emissions. In contrast, 1-PYRCA consistently exhibited low baseline levels across these scenarios, underscoring its specificity as a biomarker for woodsmoke exposure.

Inhalation, ingestion, and dermal absorption are the main pathways through which PAHs enter the human body. Participants were advised to avoid PAHs-rich food before, during, and after the exposure session. ASA24 data (Figure S6) indicated that diet was unlikely to explain the increases in urinary 1-PYRCA levels. Although elevated wristband PAHs confirmed dermal contact, previous studies reported that short-term dermal exposure (e.g., 6 hours) did not significantly increase urinary PAH metabolites, likely due to skin’s low uptake rates.²⁹ These results, along with high black carbon levels during the session, suggest inhalation was the dominant exposure route.

In our previous study, urinary 1-OHPYR concentrations in local residents were strongly associated with traffic emissions in Rochester, NY³⁰. This association is consistent with the combustion characteristics of different PAH sources. High-temperature combustion in vehicle engines leads to more complete oxidation, producing primarily unsubstituted (parent) PAHs with minimal branched structures such as methyl groups. In contrast, biomass combustion takes place at lower temperature, generating more branched PAHs, including 1-methylpyrene.^31,32 These findings suggest that 1-PYRCA can be a more sensitive exposure biomarker of woodsmoke than hydroxylated-PAHs. However, an ideal wildfire biomarker should also be specific because wildfire usually co-exists with other sources (e.g., traffic). The specificity of 1-PYRCA to woodsmoke needs to be examined in future studies.

The study’s limitations include a small sample size (n=8), which limits its statistical power and generalizability of the findings. The sample size of our study was determined a priori based on previous studies that successfully quantified the half-time of hydroxylated-PAHs following cigarette smoking (with 8 subjects),³³ barbecued chicken consumption (with 9 subjects),³⁴ and smoked salmon consumption (with 8 subjects).³⁵ We used a sample size of 8 to test the hypothesis that campfire exposure would lead to greater increases in PAH-CAs concentrations than hydroxylated-PAHs. However, our results indicate that the exposure was not high enough to induce observable increases in most urinary PAHs metabolites, limiting our ability to estimate the biological half-lives of PAH-CAs. Moreover, previous studies have reported between-individual variations in xenobiotic metabolism influenced by external exposure and characteristics such as sex, BMI, and age. In our study, the small sample size and variable exposures exposure limited our ability to assess these contributions, though individual responses by sex are shown in Figure S7 to illustrate potential heterogeneity. Additionally, urine samples were collected only up to 24 hours post-baseline, limiting the ability to accurately characterize PAH metabolites and preventing reliable estimation of biological half-lives. Future studies should address these limitations by increasing sample sizes to improve statistical power and extending the collection period to capture the full kinetics of methylated PAH metabolism and urinary excretion. It is also important to assess biomarker specificity in real-world settings, where smoking, alcohol, and other environmental exposures may affect metabolite levels. Finally, while the controlled campfire exposure enabled reproducible exposure assessment, it may not fully replicate the complexity of real-world wildfire smoke, including differences in particulate composition, chemical mixtures, and environmental conditions. To enhance the generalizability of our findings, future studies should validate these biomarkers during actual wildfire events.

This study underscores the potential of urinary 1-PYRCA as a sensitive biomarker for short-term woodsmoke exposure, with significant applications in health studies related to wildfire smoke. The LC-MS/MS method developed here, capable of simultaneously quantifying hydroxylated and carboxylic metabolites of PAHs, offers a robust and comprehensive tool for assessing exposures to both PAHs and methylated PAHs. This versatile approach can be widely applied in research investigating the health impacts of PAH exposure and air pollution, supporting more effective evaluation and mitigation of associated risks.

Supplementary Material

Supporting Information.

Method details; OH-PAHs levels before, during, and after woodsmoke exposure in previous studies; PAH concentrations in wristbands; urinary OH-PAHs and PAH-Cas levels over time; concentrations of 1-OHPYR and 1-PYRCA (unadjusted and adjusted, by sex); PAH reference standards and chemical details; method performance characteristics; recovery rates in spiked urine; urinary PAH levels from exposed participants, and diet taken.

Synopsis:

Urinary 1-pyrenecarboxylic acid can be an exposure biomarker of woodsmoke and needed to be validated for wildfire exposure.