Comparison of ozone-specific (OZAC) and oxygen radical (ORAC) antioxidant capacity assays for use with nasal lavage fluid

Toxicology in Vitro 25 (2011) 1406–1413

Contents lists available at ScienceDirect

Toxicology in Vitro journal homepage: www.elsevier.com/locate/toxinvit

Comparison of ozone-specific (OZAC) and oxygen radical (ORAC) antioxidant capacity assays for use with nasal lavage fluid Joseph M. Rutkowski, Lizzie Y. Santiag, Abdellaziz Ben-Jebria, James S. Ultman ⇑ Department of Chemical Engineering, The Pennsylvania State University, 106 Fenske Laboratory, University Park, PA 16802, USA

a r t i c l e

i n f o

Article history: Received 4 January 2011 Accepted 6 April 2011 Available online 12 April 2011 Keywords: Uric acid Ascorbic acid Glutathione Chelating agents Ozone inhalation Air pollutant

a b s t r a c t Antioxidants in respiratory mucus protect the underlying airway epithelium from damage by ozone (O3), a common outdoor air pollutant. To understand O3–antioxidant interactions and the variation of these interactions among individuals, in vitro assays are needed to measure the total antioxidant capacity of airway lavage fluid, a convenient source of (diluted) mucous samples. Here, we compare the oxygen radical absorbance capacity (ORAC), a general method that uses peroxyl radicals as a reactive substance, to the recently developed ozone specific antioxidant capacity (OZAC), a procedure that directly employs O3. For prepared model mucous antioxidant solutions containing uric acid, ascorbic acid or glutathione, the ORAC and OZAC methods yielded comparable antioxidant capacities. The addition of EDTA or DETAPAC, necessary to prevent auto-oxidation of test solutions during the ORAC assay, unpredictably altered ORAC measurements. EDTA did not have a significant effect on OZAC measurements in either prepared uric acid or ascorbic acid solutions. When assessing antioxidant capacities of nasal lavage samples, the ORAC and OZAC assays were no longer comparable. Because the OZAC of nasal lavage samples was positively related to measured uric acid concentrations whereas the ORAC data were not, the OZAC method appears to provide more realistic mucous antioxidant capacities than the ORAC method. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction Ozone (O3), a ubiquitous component of outdoor air pollution, is an oxidant gas that causes disturbances to the respiratory tract. Even short-term exposures to environmentally-relevant O3 concentrations can cause decrements in lung function in healthy persons (Hazucha, 1987) accompanied by an elevation of inflammatory mediators in the airways (Koren et al., 1989; Ciencewicki et al., 2008). Exercise can exacerbate these effects (McDonnell et al., 1983; Chimenti et al., 2009) and asthmatic patients are particularly susceptible to ozone’s negative effects (Kim et al., 2007). The respiratory mucous layer, which contains an array of chemical substrates that are reactive with O3, is the first line of defense in protecting the respiratory system from oxidative damage. Some of these substrates, such as the low molecular weight antioxidants uric acid (UA), ascorbic acid (AH2), and reduced glutathione (GSH) react with O3 to form benign products (van der Vliet et al., 1999). Reactions with other compounds increases the likelihood of direct and indirect damage to airway epithelial cells (Ballinger et al., 2005). For example, polyunsaturated fatty acids react with O3 to

Abbreviations: UA, uric acid; AH2, ascorbic acid; GSH, glutathione; ITS, indigo trisulfonate. ⇑ Corresponding author. Tel.: +1 814 863 4802; fax: +1 814 865 7846. E-mail address: [email protected] (J.S. Ultman). 0887-2333/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.tiv.2011.04.008

form products such as aldehydes that may ultimately prove harmful to tissue (Frampton et al., 1999). Still other substrates found in mucus, such as surfactants, immunoglobins and cytokines, may have specialized biological functions that are compromised as a result of their reaction with O3 (Muller et al., 1998). A simple test to quantify the overall reaction capacity of biological solutions for O3 can provide information that is impractical to obtain by individually studying all the components of the fluid. One application of such a test would be in the study of ozone-substrate kinetics in in vitro bioreactors (Mudway et al., 1996); Mudway and Kelly, 1998; Kermani et al., 2006). A test of overall reaction capacity would also be useful in evaluating secretions from in vitro airway tissue models (Choe et al., 2006). We previously reported the development of an assay of ozone specific antioxidant capacity (OZAC) that rapidly provides a micromolar equivalent of O3 quenching by antioxidant in model and collected nasal lavage samples (Rutkowski et al., 2003). Briefly, the lavage sample and saline control solutions are each thoroughly mixed with ozonated air and the excess O3 that remains after completion of all ozonation reactions is quenched in a 1:1 color-bleaching reaction with an indicator dye, indigo trisulfonate (ITS). The degree of color loss, as determined by the difference in light absorbance between the control solution and the sample solution, is then used to compute the OZAC. The OZAC thus represents the total reactivity of the sample solution with gaseous O3.

1407

J.M. Rutkowski et al. / Toxicology in Vitro 25 (2011) 1406–1413

The conventional method for determining solution antioxidant potential is a fluorometric assay called oxygen radical absorbance capacity (ORAC) method (Cao et al., 1993). ORAC values for various foodstuffs, e.g., fruit juices, are sometimes seen on grocery store packaging touting the foods’ antioxidant composition (Seeram et al., 2008). The assay is based on the reactivity of a sample with peroxyl radicals generated by the thermal decomposition of an azide compound, 2,20 -azobis dihydrochloride (AAPH) (Cao et al., 1993). To evaluate reactivity, a fluorescent protein indicator (b-phycoerythrin) is added to a sample solution, and the progressive attenuation in fluorescence of this protein is periodically recorded until all of the protein has been oxidized. The sample antioxidant capacity slows the fluorescence decay as compared to a control solution. Normalization by an analogous measurement on an antioxidant standard, 100 lM trolox solution (a water-soluble vitamin E analog), yields the ORAC in trolox units (Cao et al., 1993). In this article, we compared OZAC measurements of prepared antioxidant solutions and nasal lavage samples to ORAC measurements obtained from the same samples. We also examined the influence of two chelating agents, EDTA and DETAPAC, used to limit auto-oxidation (i.e., reaction with environmental oxygen) in such assays.

2.2. ORAC measurement Following the method of Cao et al. (1993), 20 lL of 75 mM sodium phosphate buffer (i.e., the blank solution), 20 lL of 100 lM trolox (i.e., the trolox solution), and 20 ll of either pure antioxidant solution or nasal lavage (i.e., the sample solution) were each placed in one of three round cuvettes (12  75 mm borosilicate glass, Fisher Scientific). The following solutions were then added to these 20 lL aliquots of test solutions to obtain reacting solutions with a total volume of 2 ml: 780 lL of buffer, 1000 lL of 8 lg/ml b-phycoerythrin, and 200 lL of 65 mM AAPH solution. Thus, there was a 100-fold dilution of both the trolox and the sample solutions during the preparation of the reaction solutions. The cuvettes were then immersed in a water bath at 37 °C such that peroxyl radicals produced by the thermal decomposition of AAPH reacted with antioxidants, thereby delaying the oxidation of b-phycoerythrin by excess peroxyl radicals. Measurement of the b-phycoerythrin remaining in each cuvette was determined every 5 min by fluorescence measurement (540 nm excitation; 565 nm emission) on a Quantech fluorometer (Turner) until the b-phycoerythrin fluorescent capacity was extinguished. The area under the fluorescence versus time curve measured for each cuvette (AUC) was used to compute the antioxidant capacity of the sample solution relative to trolox, a water-soluble analog of vitamin E, using Eq. (1):

2. Methods and materials

ORAC ðtrolox unitsÞ ¼ 2.1. Overview of experiments This study consisted of three sets of experiments: (1) the effect of EDTA and DETAPAC on the measurement of ORAC in prepared solutions of UA, AH2 and GSH; (2) a comparison of the ORAC and the OZAC on these prepared solutions; and (3) a comparison of ORAC and the OZAC measurements on nasal lavage samples collected from healthy adults. In the first set of experiments, ORAC values were first measured for 250 to 2000 lM solutions of ethylenediaminetertraacetic acid (EDTA) and diethylenetriaminepentaacetic acid (DETAPAC or DTPA), two chelating agents used in biological solution preparations, to determine their inherent antioxidant capacity. The ORAC of UA, GSH and AH2 solutions at concentrations of 100–500 lM was then compared in the presence and absence of 100 lM EDTA or DETAPAC. In the second set of experiments, ORAC and OZAC values of UA, AH2 and GSH were compared. The OZAC was determined for antioxidant concentrations of 1 to 50 lM. This concentration range was selected because of its relevance to nasal lavage fluid (van der Vliet et al., 1999) and also because it is within the resolution of the current measurement method (Rutkowski et al., 2003). An antioxidant concentration range of 100–500 lM was selected for sample solutions assayed by the ORAC method. While 500 lM is close to the upper limit of the ORAC measurement, the inherent 1:100 dilution of an initial sample limited the antioxidant concentration in the final test solution to only 5 lM, one-tenth of the highest concentration used in the OZAC measurements. To enable a comparison between ORAC and OZAC measurements, the OZAC of 1–50 lM trolox solutions was also measured. In the third series of experiments, nasal lavage samples were obtained from eight healthy nonsmokers on three separate days. In addition to performing OZAC and ORAC measurements on these samples, HPLC analysis of UA concentrations was carried out. It was then possible to compare the OZAC and ORAC of diluted epithelial lining fluid and also to evaluate the contribution of UA, the most abundant low molecular weight antioxidant, to the overall antioxidant capacity.

AUCsample  AUCblank AUCtrolox  AUC blank

ð1Þ

According to this equation, an ORAC measurement of 1 unit indicates that the sample solution has an antioxidant capacity equivalent to a 100 lM trolox solution. 2.3. OZAC measurement The OZAC method was previously described by Rutkowski et al. (2003). Briefly, ozonated air was generated in four closed 3.5-ml quartz cuvettes (R-301-T, Spectrocell, Inc.) positioned equidistant from a pen-ray UV lamp (80-2049-1, Jelight Company, Inc.) while the space between the cuvettes and the UV lamp were externally flushed with nitrogen. After UV irradiation for a period of 5 min, the O3 level in the four curvettes reached 100–200 ppm and two of the cuvettes were injected with 0.5 ml of sample. The other two cuvettes were injected with 0.5 ml of physiologic (0.9% NaCl) saline that acted as positive (i.e., O3-exposed) controls for the measurement. All four cuvettes were immediately removed from the box and vigorously shaken 60 times such that gaseous O3 reacted directly with the undiluted sample. Immediately thereafter, 0.3 mL of a 0.10–0.15 mM solution of indigo trisulfonate (ITS) was injected into each cuvette and the cuvettes were shaken 40 times to quench the remaining O3. The contents of each cuvette were finally transferred into a disposable spectrophotometer microcuvette (220–900 nm, 1.5 ml, 1 cm pathlength, Kartell spa) and the absorbance was measured on a dual-beam spectrophotometer (DU 520, Beckman) at 600 nm using physiological saline as the blanking solution. ITS is a water-soluble derivative of a common blue dye that has a high molar absorptivity of e600 = 23,800 M1 cm1 for 600 nm light (Bader and Hoigné, 1981). Ozone attacks a carbon–carbon double bond of ITS, in a rapid reaction (i.e., rate constant of 1  107 M1 s1) that yields colorless products. Since the reaction of O3 with ITS has a 1:1 stoichiometry, the difference in the absorbance between the sample (AS) and the positive control (APC) is proportional to the number of moles of O3 that reacted with the antioxidants in the sample. The OZAC can be computed in terms of a molar antioxidant concentration in the original sample by applying the Beer–Lambert Law in Eq. (2):

1408

OZAC ¼

J.M. Rutkowski et al. / Toxicology in Vitro 25 (2011) 1406–1413

V ITS þ V S ðAS  APC Þ Be600 V S

ð2Þ

where VITS and VS are the respective volumes of the dye solution and sample solution injected into each cuvette, and B is the pathlength of the cuvette. The upper limit of the OZAC method, approximately 50 lM OZAC, was determined from the AS value measured for a negative control (i.e., a sample solution for which the UV lamp was off). To reduce systematic errors due to differences in UV light path, cuvette construction, etc., OZAC measurements were conducted in pairs, with the positions of the sample cuvettes and the control cuvettes switched between each measurement. Thus, each data point corresponds to the mean of two separate measurements on two sample aliquots. 2.4. Chromatographic analysis Prior to performing an ORAC or OZAC assay, concentrations of UA, AH2, and GSH in the sample solution were determined by reverse phase HPLC (HP1100, Agilent Technologies). An LC18 column (58230-U, Supelco) was used with a 0.2 M KH2PO4 mobile phase at pH 2.15 flowing at 1 ml/min and a sample injection volume of 100 L. A diode array detector (G1315A, Agilent Technologies) was employed to determine the concentrations of UA, AH2 and GSH in pure antioxidant solutions at detection frequencies of 285, 245 and 200 nm, respectively. Electrochemical detection at +0.8 V (LC-4B, Bioanalytical Systems) was necessary to quantify UA concentrations in nasal washings because an unknown compound interfered with the UA absorption spectrum. Standard curves were developed by injecting solutions containing known concentrations of UA, AH2 or GSH into the HPLC, and determining the area under the appropriate output peak of the chromatogram. The output peaks corresponding to UA, AH2 or GSH were identified from the retention times obtained after injecting solutions of the pure antioxidants. All measurements were replicated by injection of three separate samples. 2.5. Solution preparation Chemicals were obtained from the Aldrich/Sigma Chemical Company unless otherwise noted. Solutions with desired concentrations of UA, AH2, GSH, trolox, EDTA or DETAPAC were prepared gravimetrically in physiologic saline (Baxter Healthcare Corp.). Whenever possible, the concentrations of UA, AH2, and GSH in these solutions were verified by reverse-phase HPLC. The pH of all prepared solutions was adjusted to 7.1–7.4 using NaOH and HCl prior to OZAC measurements. For the ORAC assay, a 75 mM sodium phosphate buffer was prepared in distilled deionized water and adjusted to a pH of 7.0 with NaOH and HCl. The b-phycoerythrin solution was prepared in this buffer by diluting the solution purchased from the manufacturer (4 mg/ml in 60% saturated ammonium sulfate, Molecular Probes). The AAPH solution was prepared by dissolving the solid reagent in buffer. For the OZAC assay, fresh solutions of ITS were prepared by dissolving the solid compound in de-ionized, distilled water. The ITS concentration, prepared gravimetrically according to the manufacturer’s specification of purity, was verified by an absorbance measurement at 600 nm. At the O3 concentrations of 100–200 ppm generated in the OZAC assay system, an ITS concentration from 0.10 to 0.15 mM resulted in a reasonable range of absorbance values within the standard curve between the positive control (i.e., O3-exposed saline samples that had the least optical density) and negative control (i.e., unexposed saline that had the greatest optical density) solutions.

For the HPLC analysis, the 0.2 M KH2PO4 mobile phase was prepared gravimetrically by dissolving the solid reagent in distilled deionized water and adjusting the pH to 2.15 using concentrated H3PO4 and KOH. Nasal washings were obtained from eight young adults who were free of any respiratory disease or infection and who did not smoke. On each of the three days in which a subject participated, 5 ml of sterile physiologic saline (Baxter Healthcare Corp.) warmed to 37 °C was instilled into one nostril while the subject kept his or her head tilted back with the velum closed. After 10 s, the subject tilted his or her head forward and the solution was drained into a sterile plastic specimen cup placed in front of the nose. Immediately thereafter, another 5 ml of saline were introduced into the other nostril, and 10 s later, the subject expelled the solution into the same cup. Experimental sessions were repeated when the recovery of instilled fluid was less than 65%. Each of nasal lavage sample was passed through a 0.45 lm filter (Agilent) to remove particulates and debris before further analysis. To ensure that lavage samples did not exceed the 50 lM upper limit of the OZAC measurement, they were diluted 1:4 before measurement in accordance with the assay design (Rutkowski et al., 2003). All human subject procedures were approved by the Penn State Office for Research Protections (Protocol#01M1187) and, prior to donating nasal lavage, all subjects provided their informed consent. 2.6. Data analysis All reported concentrations refer to those initially present in the prepared solutions or present in the nasal lavage sample before it was diluted by 1:4. The ORAC measurement was performed in duplicate on prepared solutions containing either UA or AH2. A single ORAC measurement was taken on all other prepared solutions and on each nasal lavage sample. OZAC measurements were replicated 4 times on all prepared solutions and on each nasal lavage sample. Where replicate measurements are available, the data is generally presented as the mean (±standard deviation). Least squares regressions and analyses of covariance (ANCOVA) were performed with MINITAB software. p-values < 0.05 were considered to be significant. 3. Results 3.1. ORAC method and the effects of chelating agents Fluorescence-time curves were used to calculate the ORAC values for prepared solution (Fig. 1A and B). The area between the fluorescence decay curve of a test solution and the blank relative to that of a 100 lM trolox solution is defined as the antioxidant capacity. In other words, one ORAC unit corresponds to an antioxidant capacity equivalent to a 100 lM trolox solution. At concentrations of 500 lM, each of the small molecular weight antioxidants examined, UA, AH2 and GSH, displayed a decay curve above that of the trolox standard indicating that they had ORAC values greater than one. (Fig. 1A). In the usual protocol for the ORAC method, chelating agents are added to limit auto-oxidation in test solutions. We sought to determine if these agents have an inherent antioxidant capacity that would influence the ORAC measurement. The fluorescent decay curves of 500 lM solutions of either EDTA or DETAPAC were much below a 100 lM trolox solution and were only slightly above the positive saline controls (Fig. 1B). Thus, chelating agent concentrations up to 500 lM have little effect on the ORAC measurement. For chelating agent concentrations greater than 500 lM, however, both EDTA and DETAPAC yielded more substantial antioxidant capacities

J.M. Rutkowski et al. / Toxicology in Vitro 25 (2011) 1406–1413

1409

not affect the ORAC results (p = 0.384 and p = 0.590, respectively) for UA solutions (Fig. 2A). However, the ORAC values of AH2 solutions (Fig. 2B) were significantly increased by either EDTA (p = 0.001) or DETAPAC (p = 0.005). Although the ORAC of GSH solutions was not significantly affected by EDTA (p = 0.113), it was significantly decreased (p = 0.007) by the addition of DETAPAC (Fig. 2C). 3.2. OZAC of prepared antioxidant solutions Whereas the OZAC values for UA and AH2 appeared to be proportional to the antioxidant concentrations, the OZAC of GSH approached a constant value of 21.63 ± 0.04 as its concentration increased (Fig. 3A).

Fig. 1. ORACs for prepared solutions. (A) Fluorescence decay curves of UA, AH2, GSH and trolox solutions. The ordinate is made dimensionless by its initial level and abscissa is made dimensionless by the time required for the control solution to reach an immeasurable fluorescence level. The area between the fluorescence decay curves of the sample and the control (i.e., antioxidant-free solution) is used to calculate the ORAC value for each solution according to Eq. (1): s = 100 lm Trolox, j = 500 lm UA, N = 500 lm AH2,  = 500 lM GSH,  control. (B) Fluorescent decay curves for 500 lm solutions of the chelating agents EDTA and DETAPAC are similar to the control: s = 100 lm Trolox, j = 500 lm EDTA, N = 500 lm DETAPAC,  control. (C) ORAC values for solutions of EDTA and DETAPAC can demonstrate have inherent antioxidant capacities at concentrations above 500 lm: j = EDTA, N = DETAPAC.

(Fig. 1C). At 1000 lM, typical of the concentration used in biological samples, EDTA and DETAPAC had respective ORAC values of 0.49 and 1.16 units. With these substantial antioxidant capacities in mind, we examined the effects of 100 lM concentrations of EDTA or DETAPAC on the ORAC values of UA, AH2 and GSH solutions (Fig. 2). An ANCOVA using antioxidant concentration as a covariate and treatment with or without chelating agent as a fixed variable indicated that the addition of either 100 lM EDTA or 100 lM DETAPAC did

Fig. 2. Effects of chelating agents on ORACs of prepared solutions. (A) Neither the addition of EDTA (p = 0.384) nor DETAPAC (p = 0.590) affect the ORAC of UA solutions. (B) Both EDTA (p = 0.001) and DETAPAC (p = 0.005) cause a statistically significant change in the ORAC values of AH2. (C) The ORAC of GSH solutions is not affected by EDTA (p = 0.113), but significantly changed (p = 0.007) by the addition of DETAPAC: s = neat GSH, j = GSH + 100 lm EDTA, N = GSH + 100 lm DETAPAC.

1410

J.M. Rutkowski et al. / Toxicology in Vitro 25 (2011) 1406–1413

EDTA and DETAPAC solutions reached a plateau value of about 40 lM, close to the upper limit of OZAC measurement. The addition of 1000 lM EDTA appears to reduce the OZAC of both UA and AH2 solutions (Fig. 3C). An ANCOVA of these data indicated that this change was not significant in either the UA solutions (p = 0.051) or the AH2 solution (p = 0.151). 3.3. Comparison of ORAC and OZAC for prepared solutions To compare an ORAC unit to the antioxidant capacities determined by the OZAC method, it was necessary to determine the OZAC value that is equivalent to 100 lM trolox in the sample solution. Thus, OZAC measurements were made at trolox concentrations 1–50 lM (Fig. 4). A linear regression of these data was significant (p < 0.001). Because the intercept of this regression was not significantly different from zero, the regression was repeated with the intercept forced through zero. The resulting slope ± standard error of 1.37 ± 0.05 was significantly different from zero (p < 0.001) and was also different from 1.00 (p < 0.001). That is, 1.37 lmol of trolox is required to remove 1 lmol of O3 from an OZAC sample solution. Since each ORAC unit is equivalent to 100 lmol, the equivalent OZAC is given by

EQUIVALENT OZAC ¼ 137 ORAC UNITS

ð3Þ

In theory, this equation will hold if the peroxidation processes measured in the ORAC assay result in the same stoichiometry as the ozonation processes measured in the OZAC. This equation was applied to the largest set of ORAC values among the ‘‘neat’’ and stabilized solutions in Fig. 2 for comparison to the OZAC values in Fig. 3A. In particular, we graphed LOG(OZAC) and LOG(EQUIVALENT OZAC) as a function of the logarithm of the antioxidant concentration in the sample (Fig. 5). We performed an ANCOVA of these measurements using a fixed classification variable to designate either the OZAC or OZAC–equivalent ORAC method and using the logarithm of the antioxidant concentration as a covariate. The results indicate that there was no difference between the two methods. In addition, the slopes of the LOG–LOG data obtained for UA, AH2 and GSH data were not different from one (p = 0.071,0.082 and 0.184) and their intercepts were no different from zero (p = 0.068,0.0085 and 0.656, respectively). 3.4. ORAC and OZAC of nasal lavage samples

Fig. 3. OZAC of prepared solutions. (A) The slope of the UA data (0.94 ± 0.04) was not significantly different from one (p = 0.229), whereas the slope of the AH2 data (0.85 ± 0.04) was not equal to one (p = 0.015): j = UA, N = AH2,  = GSH,  line of identity. (B) Both EDTA and DETAPAC exhibit inherent antioxidant capacities in the OZAC assay: j = EDTA, N = DETAPAC. (C) Addition of 1000 lM EDTA does not significantly change the OZAC of either UA (p = 0.051) or AH2 solutions (p = 0.151): h = UA, j = UA + 1000 lM EDTA, 4 = AH2, N = AH2 + 1000 lM EDTA.

Linear regressions of the OZAC-concentration data for UA and AH2 had r2 values of 0.98 and 0.97, respectively. Since the intercepts of these regressions were not significantly different from zero for either UA (p = 0.246) or AH2 (p = 0.229), the regressions were repeated with the intercepts forced through zero. In the case of UA, the resulting slope ± standard error of 0.94 ± 0.04 was significantly different from zero (p < 0.001) and not significantly different from 1.00 (p = 0.229). For AH2, the slope ± standard error of 0.85 ± 0.04 was different from zero (p < 0.001) and was also different from 1.00 (p = 0.015). Thus, the overall stoichiometry of the ozonation of UA is 1:1 while it is somewhat less than 1:1 for AH2. We determined the OZAC of EDTA and DETAPAC solutions over a concentration range of 250–2000 lM (Fig. 3B). The OZAC of both

Of the several different substances in nasal mucous that are reactive with O3, UA is perhaps the most abundant (Peden et al., 1993). Thus, UA concentrations in nasal lavage samples from eight

Fig. 4. OZAC of trolox solutions. Regression analysis (solid line) yielded the relationship OZAC = 1.37  (trolox concentration). Thus, the factor to convert ORAC units to an equivalent OZAC concentration is 1.37.

J.M. Rutkowski et al. / Toxicology in Vitro 25 (2011) 1406–1413

1411

a moderate correlation (r2 = 0.207) was found between subjectaveraged ORAC values and UA concentrations (Fig. 6B), this relationship was not significant (p = 0.258). A stronger, significant correlation (r2 = 0.525; p = 0.042) was found between the subjectaveraged OZAC and UA concentrations (Fig. 6D). In other words, a relationship between ORAC and UA concentration among the subjects was not discernable, whereas the data indicates that more than 50% of subject-to-subject differences in OZAC were driven by corresponding differences in UA concentration. An interesting feature of the OZAC-UA linear regression was the existence of a statistically significant (p = 0.001) intercept value of 24.8 ± 4.2 lM (Fig. 6D). This demonstrated that, even in the absence of UA, other antioxidants and O3-reactive substances play an important role in establishing the OZAC value. As was the case for pure antioxidant samples (Fig. 5), each ORAC unit should be equivalent to an OZAC of 137. Whereas the ORAC was measured on full-strength lavage, the OZAC was measured on samples that were prediluted by 1:4. Thus, the OZAC equivalent to an ORAC measurement in nasal lavage is given by

ðEQUIVALENT OZACÞ ¼ 34:2 ðORAC UNITSÞ

ð4Þ

A plot of this line (Fig. 7A and B) indicated that the OZAC and ORAC of nasal lavage did not exhibit equivalent antioxidant capacities, except at the lowest antioxidant capacity. Therefore, while pure solutions of UA, AH2 and GSH had comparable OZAC and OZAC-Equivalent ORAC data, this relationship did not hold for nasal lavage fluid.

4. Discussion

Fig. 5. Comparison of ORAC and OZAC assays for prepared solutions. In the case of the ORAC assay, the ordinate refers to EQUIVALENT OZAC values obtained by using the conversion factor of 137 (Eq. (3)). Statistical analysis indicated that LOG(OZAC) and LOG(EQUIVALENT OZAC) versus LOG (antioxidant concentration) both coincide with the line of identity for (A) UA solutions, (B) AH2 solutions, and C) GSH solutions: s = OZAC (measured), d = 137  ORAC (measured), .line of identity.

healthy nonsmokers were determined by HPLC, and ORAC and OZAC measurements on the samples were compared. Lavage samples for each subject were collected on three separate days as indicated by the connecting lines on Fig. 6. It is apparent from these line segments that within-subject variations in ORAC (Fig. 6A) and in OZAC (Fig. 6C) were not consistently related to the corresponding variations in UA concentration. That is, in some cases a day-to-day increase in UA resulted in an increase in ORAC or OZAC (i.e., line segments of positive slope) whereas in other cases the opposite was true. To determine whether between-subject differences could be explained by UA concentration, the three-day set of ORAC and of OZAC values were averaged for each subject and were then regressed against the subject-averaged UA concentrations. Although

In vitro antioxidant assays utilize various oxygen radical chemistries to quantify the relative capacity of dietary, environmental, and biological substrates and solutions in limiting oxidative damage. Understanding these capacities provides a basis for health and environmental recommendations. Oxygen free radicals formed by environmental pollutants such as O3 can cause respiratory distress, even at low concentrations, (Hazucha, 1987). Oxygen radicals are also produced by naturally-occurring peroxides that immune cells utilize for fighting infection (Koren et al., 1989; Ciencewicki et al., 2008). In either case, extracellular fluids – respiratory mucus or blood – contain a multitude of small molecular weight antioxidants, proteins, lipids, and dietary polyphenols that scavenge oxygen radicals to avoid undesirable cellular damage (Ballinger et al., 2005). Because these protective species react differently with O3 than with peroxides, the use of specific antioxidant assays is desirable as an experimental tool. The long-established ORAC is an indirect method that specifies the antioxidant capacity of aqueous solutions for all peroxyl radicals, no matter what their origin, in terms of the equivalent concentration of water-soluble vitamin E (i.e., trolox) (Cao and Prior, 1998). The OZAC method is a direct method that provides the molar O3-quenching of aqueous solutions that are exposed to gaseous O3. Although O3 may cause damage by formation of free radicals that subsequently participate in free radical chain reactions, O3 is also capable of undergoing direct oxidation of substrates at unsaturated carbon–carbon bounds (Pryor, 1994). Therefore, the use of peroxyl radicals in the ORAC assay as a surrogate for O3 will, in all probability, result in different reaction pathways than those in the OZAC assay that uses O3 itself as an oxidant challenge. In spite of different reaction pathways, we demonstrated that ORAC and OZAC measurements are comparable for simple prepared solutions with known compositions. In those cases, we demonstrated a relationship of the ORAC to the OZAC by using a conversion factor of 137 that accounts for the difference between overall reaction stoichiometries. We also demonstrated that the

1412

J.M. Rutkowski et al. / Toxicology in Vitro 25 (2011) 1406–1413

Fig. 6. ORAC and OZAC for nasal lavage. Nasal lavage samples from eight volunteers collected on three separate days. (A) Sequence of daily ORAC measurements connected by line segments do not show a consistent pattern for an individual subject. (B) Regression analysis of three-day averages of ORAC did not yield a significant correlation (p = 0.258) with three-day averages of UA (p = 0.258). (C) Daily OZAC measurements did not show a consistent pattern for an individual subject. (D) Regression analysis of three-day averages of OZAC indicated a significant correlation with three-day averages of UA (p = 0.042). A significant non-zero intercept (24.8 ± 4.2) of the regression is due to reactions with other molecules in addition to UA in the lavage fluid: s = daily measurements, d = three-day averages of each individual, ——regression.

Fig. 7. Comparison of ORAC and OZAC for nasal lavage. OZAC measurements (ordinate) compared to ORAC measurements (abscissa) on nasal lavage samples from eight volunteers collected on three separate days. (A) Sequence of daily measurements connected by line segments do not show a consistent pattern for an individual subject. (B) Regression analysis of three-day averages of OZAC did not yield a significant correlation (p = 0.123). The data only approximated their expected equivalency, OZAC = 34.3  ORAC, at an ORAC of 1 unit: s = daily measurements, d = 3-day averages of each individual, — predicted line of OZAC–ORAC equivalence.

complex oxidation chemistries of a biologic solution such as nasal lavage fluid, unlike prepared solutions, prohibits a unique relationship between the OZAC and ORAC measurements (Mudway and Kelly, 1998). The ORAC method is based on the reaction of substrates with peroxyl radicals generated by the thermal decomposition of an azide compound (Cao et al., 1993). This requires heating of sample

solutions in an air environment for more than 2 h, and potentially results in an underestimation of ORAC values due to auto-oxidation of substrates in addition to the intended peroxidation. We found that the addition of 1000 lM of EDTA or DETAPAC, chelating agents for preventing auto-oxidation, affected the ORAC values in an inconsistent and unpredictable manner. This undoubtedly occurred because, even at these relatively low concentrations, EDTA

J.M. Rutkowski et al. / Toxicology in Vitro 25 (2011) 1406–1413

and DETAPAC themselves possess relatively high ORAC values and are therefore capable of quenching or otherwise interfering with peroxidation pathways (Ballinger et al., 2005). Auto-oxidation is far less of a problem for the OZAC method because it is performed at room temperature in a matter of minutes. For the sake of completeness, however, we did evaluate the effect of 1000 lM EDTA on the OZAC of UA and AH2 solutions. In this case, the presence of the chelating agent did not affect the results, likely due to the rapid reaction of UA or AH2 with O3 (Kermani et al., 2006). In samples of human nasal lavage, we studied the relationship of ORAC and OZAC measurements to the UA concentration obtained by HPLC. The reason for singling out UA for study is its documented importance in the nasal mucous layer [18] and its insensitivity to autooxidation. We found no relationship between ORAC and UA concentration in nasal lavage. On the contrary, a regression of OZAC measurements with UA had a positive intercept (Fig. 6D) that accounted for about 70% of the overall average antioxidant capacities of the nasal samples. As molecules other than UA must certainly react with reactive oxygen species in nasal lavage, we conclude that the OZAC is physiologically more realistic than the ORAC and that UA contributes about 30% of the antioxidant capacity to the nasal liquid lining layer. In conclusion, we find that both the ORAC and OZAC techniques can be used to rapidly and reproducibly measure antioxidant capacities of aqueous solutions and complex biological samples in vitro. Care must be taken, however, in applying these assays; the researcher must be aware of the specificity of the reaction chemistry in each procedure. The peroxyl radical chemistry of the ORAC method and the necessity to include chelating agents makes it unsuitable for studying the O3 antioxidant capacity of respiratory lavage samples. This does not, however, preclude the use of the ORAC method when peroxyl chemistry is specifically desired or for samples with stable chemistries and/or slower kinetics. The direct use of O3 with the exclusion of chelating agents makes the OZAC technique better suited for measuring the reaction of respiratory mucus with O3. Conflict of interest statement None declared. Acknowledgements This research was supported by Grant No. ES06075 from the National Institute of Environmental Health Sciences at the National

1413

Institutes of Health. A. Ben-Jebria is also affiliated with: Institut National de la Sante’ et de la Recherche Medicale (INSERM), France. References Bader, H., Hoigné, J., 1981. Determination of ozone in water by the indigo method. Water Res. 15, 449–456. Ballinger, C.A., Cueto, R., Squadrito, G., Coffin, J.F., Velsor, L.W., Pryor, W.A., Postlethwait, E.M., 2005. Antioxidant-mediated augmentation of ozoneinduced membrane oxidation. Free Radic. Biol. Med. 38, 515–526. Cao, G., Alessio, H.M., Cutler, R.G., 1993. Oxygen-radical absorbance capacity assay for antioxidants. Free Radic. Biol. Med. 14, 303–311. Cao, G., Prior, R.L., 1998. Comparison of different analytical methods for assessing total antioxidant capacity of human serum. Clin. Chem. 44, 1309–1315. Chimenti, L., Morici, G., Paterno, A., Bonanno, A., Vultaggio, M., Bellia, V., Bonsignore, M.R., 2009. Environmental conditions, air pollutants, and airway cells in runners: a longitudinal field study. J. Sports Sci. 27, 925–935. Choe, M.M., Tomei, A.A., Swartz, M.A., 2006. Physiological 3D tissue model of the airway wall and mucosa. Nat. Protoc. 1, 357–362. Ciencewicki, J., Trivedi, S., Kleeberger, S.R., 2008. Oxidants and the pathogenesis of lung diseases. J. Allerg. Clin. Immunol. 122, 456–468 (quiz 469–470). Frampton, M.W., Pryor, W.A., Cueto, R., Cox, C., Morrow, P.E., Utell, M.J., 1999. Ozone exposure increases aldehydes in epithelial lining fluid in human lung. Am. J. Respir. Crit. Care Med. 159, 1134–1137. Hazucha, M.J., 1987. Relationship between ozone exposure and pulmonary function changes. J. Appl. Physiol. 62, 1671–1680. Kermani, S., Ben-Jebria, A., Ultman, J.S., 2006. Kinetics of ozone reaction with uric acid, ascorbic acid, and glutathione at physiologically relevant conditions. Arch. Biochem. Biophys. 451, 8–16. Kim, D.H., Kim, Y.S., Park, J.S., Kwon, H.J., Lee, K.Y., Lee, S.R., Jee, Y.K., 2007. The effects of on-site measured ozone concentration on pulmonary function and symptoms of asthmatics. J. Korean Med. Sci. 22, 30–36. Koren, H.S., Devlin, R.B., Graham, D.E., Mann, R., McGee, M.P., Horstman, D.H., Kozumbo, W.J., Becker, S., House, D.E., McDonnell, W.F., et al., 1989. Ozoneinduced inflammation in the lower airways of human subjects. Am. Rev. Respir. Dis. 139, 407–415. McDonnell, W.F., Horstman, D.H., Hazucha, M.J., Seal Jr., E., Haak, E.D., Salaam, S.A., House, D.E., 1983. Pulmonary effects of ozone exposure during exercise: doseresponse characteristics. J. Appl. Physiol. 54, 1345–1352. Mudway, I.S., Housley, D., Eccles, R., Richards, R.J., Datta, A.K., Tetley, T.D., Kelly, F.J., 1996. Differential depletion of human respiratory tract antioxidants in response to ozone challenge. Free Radic. Res. 25, 499–513. Mudway, I.S., Kelly, F.J., 1998. Modeling the interactions of ozone with pulmonary epithelial lining fluid antioxidants. Toxicol. Appl. Pharmacol. 148, 91–100. Muller, B., Seifart, C., Barth, P.J., 1998. Effect of air pollutants on the pulmonary surfactant system. Eur. J. Clin. Invest. 28, 762–777. Peden, D.B., Swiersz, M., Ohkubo, K., Hahn, B., Emery, B., Kaliner, M.A., 1993. Nasal secretion of the ozone scavenger uric acid. Am. Rev. Respir. Dis. 148, 455–461. Pryor, W.A., 1994. Mechanisms of radical formation from reactions of ozone with target molecules in the lung. Free Radic. Biol. Med. 17, 451–465. Rutkowski, J.M., Santiago, L.Y., Ben-Jebria, A., Ultman, J.S., 2003. Development of an assay for ozone-specific antioxidant capacity. Inhal. Toxicol. 15, 1369–1385. Seeram, N.P., Aviram, M., Zhang, Y., Henning, S.M., Feng, L., Dreher, M., Heber, D., 2008. Comparison of antioxidant potency of commonly consumed polyphenolrich beverages in the United States. J. Agric. Food. Chem. 56, 1415–1422. van der Vliet, A., O’Neill, C.A., Cross, C.E., Koostra, J.M., Volz, W.G., Halliwell, B., Louie, S., 1999. Determination of low-molecular-mass antioxidant concentrations in human respiratory tract lining fluids. Am. J. Physiol. 276, L289–296.