Article pubs.acs.org/est
Reactive Oxygen Species Production Mediated by Humic-like Substances in Atmospheric Aerosols: Enhancement Effects by Pyridine, Imidazole, and Their Derivatives Jing Dou,† Peng Lin,‡,§ Bin-Yu Kuang,‡ and Jian Zhen Yu*,†,‡ †
Division of Environment and ‡Department of Chemistry, Hong Kong University of Science & Technology, Clear Water Bay, Kowloon, Hong Kong, China S Supporting Information *
ABSTRACT: Ambient particulate matter (PM) can cause adverse health effects via their ability to produce reactive oxygen species (ROS). Humic-like substances (HULIS), a complex mixture of amphiphilic organic compounds, have been demonstrated to contain the majority of redox activity in the water-extractable organic fraction of PM. Reduced organic nitrogen compounds, such as alkaloids resulting from biomass burning emissions, are among HULIS constituents. In this study, we examined the redox activities of pyridine, imidazole and their alkyl derivatives using a cell-free dithiothreitol (DTT) assay under simulated physiological conditions (37 °C, pH = 7.40). These compounds were found to have little redox activity on their own as measured by the DTT assay, but they enhanced ROS generation catalyzed by 1,4-naphthoquinone (as a model quinone compound) and HULIS isolated from multiple aerosol samples. The enhancement effect by the individual nitrogen-containing bases was determined to be proportional to their amount in the assay solutions. It is postulated that the underlying mechanism involves the unprotonated N atom acting as a H-bonding acceptor to facilitate hydrogen-atom transfer in the ROS generation cycle. The enhancement capability was found to increase with their basicity (i.e., pKa of their conjugated acids, BH+), consistent with the proposed mechanism for enhancement. Among the imidazole homologues, a linear relationship was observed between the enhancement factors (in log scale) of the unprotonated form of the imidazole compounds (B) and the pKa of their conjugated acids (BH+). This relationship predicts that the range of alkylimidazole homologues (C6−C13) observed in atmospheric HULIS would be 1.5−4.4 times more effective than imidazole in facilitating HULIS-mediated ROS generation. Our work reveals that the ability of atmospheric PM organics to catalyze generation of ROS in cells could be affected by coexisting redox inactive organic constituents and suggests further work deploying multiple assays be conducted to quantify redox capabilities and enhancement effects of the HULIS components.
■
INTRODUCTION Ambient particulate matter (PM) is an important air pollutant known to cause adverse health effects and mortality in humans.1−3 One of the major toxicological mechanisms is through the induction of oxidative stress derived from PMmediated generation of reactive oxygen species (ROS) in cells.4−11 Among the numerous constituents of PM, metals12−16 and quinoid compounds17 have been established as being capable of catalyzing ROS generation in cells. More recently, humic-like substances (HULIS), a complex mixture of water-extractable amphiphilic organic compounds having structural similarities to terrestrial and aquatic humic substances (HS),18 have also been recognized as major redox-active components in ambient PM and can serve as electron carriers to catalyze ROS formation. 19,20 However, the specific components of HULIS that impact ROS formation activity have been largely unexplored. Our previous ultrahigh-resolution mass spectrometric (UHRMS) study identified nitrogen-containing heterocyclic compounds of double-bond equivalency (DBE) of 3 or 4 (i.e., alkaloids) among the most abundant peaks (not necessarily the © XXXX American Chemical Society
most abundant in concentration) detected under positive electrospray ionization mode in the HULIS fraction of ambient and biomass burning (BB) source samples collected in the Pearl River Delta (PRD) region of China.21 Laskin et al.22 also reported alkaloid compounds as abundant constituents in aerosols emitted from various biofuels. Alkaloids are generally basic compounds derived from amino acids in plants and living organisms and can be emitted from smoldering fires with minor pyrolytic and oxidative processing.23 Additionally, some alkaloids, such as imidazole, imidazole-2-carboxaldehyde, and 1N-glyoxal-substituted imidazole, are also reported to be major products of glyoxal reaction with ammonium ions or primary amines on secondary organic aerosol (SOA).24−26 The dithiothreitol (DTT) assay is a chemical method developed for evaluating the redox cycling capacity of catalytically active redox species by measuring how fast DTT Received: December 8, 2014 Revised: May 2, 2015 Accepted: May 11, 2015
A
DOI: 10.1021/es5059378 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology is oxidized.27 In the process of attempting to identify redoxactive organic compounds other than quinones, a few alkaloids (e.g., imidazole, pyrrole, pyridine, pyrimidine, pyrazine, and their derivatives) were tested using the DTT assay, but all showed negligible ROS activity. Prompted by the cyclic voltammetry study of Kipp et al.28 that showed unprotonated imidazole can accept a proton from the reductant, facilitating electron transfer, we examined in this study the effects of some atmospherically relevant alkaloids on modifying the redox activity of both a redox-active model compound (e.g., quinones) and atmospheric HULIS samples. Figure 1 shows
compound) and concomitant consumption of DTT. In the presence of these nitrogen-containing compounds, the oxidation of organic compounds that normally lose hydrogen atoms could be enhanced by the formation of H-bonding. We report here the role of pyridine, imidazole, and their derivatives in ROS production mediated by HULIS.
■
EXPERIMENTAL SECTION Aerosol Samples. Ambient samples of PM2.5 (PM of less than 2.5 μm in aerodynamic diameter) were collected at urban (Guangzhou (GZ)) and suburban (Nansha (NS)) locations in the PRD in 2009. The samples were collected onto prebaked quartz filters using a high volume aerosol sampler (TE-6070 VBL, Tisch Environmental, Inc., USA). The sampling duration for individual samples was 24 h. Fresh BB emissions samples were collected from open burning experiments in a village. Locally harvested rice straw in small bundles and sugar cane leaves in thin piles were burned, simulating the open field burning practiced by farmers. PM2.5 smoke particles were collected about 5 m downwind of the fires. More details about the sampling work can be found in our previous papers.29,30 HULIS Isolation and Determination. HULIS in aerosol filter samples was first isolated from the other constituents in water extracts using solid-phase extraction (SPE) followed by quantification by an evaporative light-scattering detector (ELSD). The water extracts were filtered before the SPE step to avoid any interference redox activity from residual quartz fiber.6 The isolation procedure separates the water-soluble matter into a hydrophilic fraction (the SPE cartridge effluent) and a hydrophobic fraction (i.e., the HULIS fraction or the eluate fraction). In brief, portions of the high-volume filters were extracted in an ultrasonic bath with water, the volume of which used was about 1 mL per 1 cm2 filter area for extraction. The extract was acidified with HCl (pH 2) before it was loaded on an SPE cartridge (Oasis HLB, 30 μm, 60 mg/cartridge, Waters), with the ratio of HCl to the extract set to ∼4 μL per mL. The loaded cartridge was subsequently rinsed with two 1 mL portions of water before elution with 12 mL of methanol. The eluate was collected and evaporated to dryness under a gentle stream of N2 and redissolved in water for HULIS quantification and DTT assay. The majority of the watersoluble metals was separated from the HULIS fraction through the SPE step. The HULIS fraction retains 10% of Fe, 15% of Al,
Figure 1. Conceptual diagram of quinone-catalyzed ROS production and concomitant consumption of DTT (in the box), with imidazole and pyridine compounds facilitating the oxidation of hydroquinone to semiquinone through H-bonding (highlighted in the gray oval).
a perceived conceptual diagram of alkaloids forming H-bonding with a hydroquinone compound (as a model ROS-active
Table 1. Composition Combinations in the DTT Assays That Involve Redox-Inactive N-Bases and Redox-Reactive Single Components or HULIS Mixturesa assay no.
N-base
1.1−1.6 2.1−2.6 3.1−3.6 4.1−4.6 5.1−5.6 6.1−6.6 7.1−7.6 8.1−8.6 9.1−9.6
3-methoxypyridine pyridine imidazole 4-methylimidazole 2-methylimidazole 2,4-dimethylimidazole 2-ethylimidazole 2-ethyl-4-methylimidazole 1,2,4,5-tetramethylimidazole
amt of N-base in individual assays (μmol) 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
9.92 12.4 1.47 1.22 1.22 1.04 1.04 0.91 0.81
49.6 61.8 14.7 6.09 6.09 5.20 5.20 4.54 4.03
99.2 124 29.4 12.2 12.2 10.4 10.4 9.08 8.05
149 185 44.1 18.3 18.3 15.6 15.6 13.6 12.1
ROS-active component (μg) 198 247 58.8 24.4 24.4 20.8 20.8 18.2 16.1
0.3 μg 1,4-NQ or 5−15 μg HULIS
a Each 200 μL sample consisted of 100 μL of an aqueous solution of a certain N-base of different concentrations and 100 μL of a solution of an ROSactive component (either 1,4-NQ or HULIS). The amounts of N-base (in μmol) and the amount of ROS-active component (in μg) in individual assays are listed . The final volume of each assay is 1.2 mL. The incubation time was 30 min for assays involving 1,4-NQ and 90 min for assays involving HULIS.
B
DOI: 10.1021/es5059378 Environ. Sci. Technol. XXXX, XXX, XXX−XXX
Article
Environmental Science & Technology 17% of Cu, and 10 μmol). The linear increase in enhancing DDT consumption with increasing concentrations of pyridine Nbases only held up to the concentration of ∼50 μmol, beyond which the consumption of DTT increased at a much slower rate (Figure S2a,b, Supporting Information). Histidine, an α-amino acid with an imidazole functional group, also shows a positive effect on 1,4-NQ ROS activity. The presence of 6.4 μmol (1 mg) of histidine in the DTT assay yields an enhancement effect between that of pyridine and imidazole at the same molar concentration, which is consistent with the pKa value of histidine (6.10) residing between pyridine (5.14) and imidazole (6.95). Njus et al.40 also found that the imidazole side chain on the histidine residue can facilitate the proton−electron transfer from ascorbic acid to cytochrome b561 by the formation of H-bonding in physiological conditions. However, as histidine is insoluble at higher concentrations, it cannot be tested further in the DTT assay. A second ROS-active model compound, 9,10-phenanthrenequinone (9,10-PQ), was also tested. The DTT consumption rate by 9,10-PQ in the absence of any N-bases was measured to be ∼13000 pmol/min/μg, ∼20 times larger than that of 1,4NQ. Similarly, Li et al.41 reported that on the mass basis 9,10PQ has about 10 times higher ROS activity than 1,4-NQ in the DTT assay; Charrier and Anastasio15 reported it was 21 times on the mass basis (∼28 times on a mole basis). DTT assays were also carried out on the mixtures of 9,10-PQ (0.015 μg) and varying amounts of imidazole (1.5−88 μmol). A linear relationship between ΔDTT9,10‑PQ and the amount of imidazole (R2 = 0.99) was observed, yielding a slope of 413 pmol/min/ (μg 9,10-PQ)/μmol of imidazole, which produces a EFB of 560 pmol/min/(μg 9,10-PQ)/μmol of imidazole. This value is ∼17 times larger than the EFB of imidazole on 1,4-NQ (Table 2). Enhancement of HULIS ROS Activity by N-Bases. We tested the DTT consumption of HULIS in the presence of various amounts of N-bases. As an example, Figure 3b shows the DTT consumption rate of HULIS (from ambient sample NS20091023) as a function of the amount of 2-methylimidazole. A clear linear relationship (R2 = 0.99) is found, indicating
apparent EF = EFB × fB = 10(m × pKa + c) ×
1 1 + [H+]/K a (6)
where m is the slope and c is the intercept of the log (EFB) vs pKa. Figure 4b shows the calculated apparent EF by the imidazole series acting on HULIS (red) and 1,4-NQ (green) in the pKa range of 4−10. With the minimum detectable enhancement factor at ∼0.01 pmol/min/(μg HULIS) per μmol of N-base, the imidazole derivatives with a pKa larger than ∼6 have a measurable enhancement effect on HULIS-mediated ROS generation. At pKa of 9.4 or higher, the unprotonated Nbase fraction would be reduced to
