Clinica Chimica Acta 337 (2003) 147 152 www.elsevier.com/locate/clinchim Measurement of plasma hydroperoxide concentration by FOX-1 assay in conjunction with triphenylphosphine Dibyajyoti Banerjee*, U.K. Madhusoodanan, M. Sharanabasappa, Sandip Ghosh, Jose Jacob Department of Biochemistry, Kasturba Medical College, Manipal 576119, Karnataka State, India Received 2 June 2003; received in revised form 3 August 2003; accepted 6 August 2003 Abstract Introduction: Hydroperoxides are well-recognized reactive oxygen species which are associated with oxidative stress, a phenomenon of current clinical interest as oxidative stress is associated with a number of disease condition. Ferrous ion oxidation xylenol orange (FOX) methods of hydroperoxide estimation has outdated other methods available for hydroperoxide estimation. Two versions FOX assays are described in the literature, FOX-1 and FOX-2, in which FOX-1 is more sensitive. Methods: we increased the sensitivity of FOX-1 assay by stabilizing the reagent ph 1.7 1.8. Analogous to FOX-2 assay, we have modified FOX-1 assay by using it in conjunction with triphenylphosphine and butylated hydroxytoluene, thus increasing the specificity of FOX-1 assay for hydroperoxide. By modified FOX-1 method, we estimated plasma hydroperoxide concentration of normal human subjects and of diabetic patients and compared with FOX-2 method. Results: The FOX-1 method showed a significant high value of plasma hydroperoxide concentration compared to FOX-2 method both in normal subjects and diabetic patients with a significant correlation. By modified FOX-1 method, the recovery percentage of cumene hydroperoxide was better in biological samples when compared to FOX-2 method. Conclusion: The modified FOX-1 method is equally specific for hydroperoxide determination when compared to FOX-2 but is more sensitive. D 2003 Elsevier B.V. All rights reserved. Keywords: Hydroperoxide; FOX assay; Oxidative stress; Reactive oxygen species 1. Introduction Hydroperoxides (ROOH) are well-established reactive oxygen species (ROS) associated with oxidative stress; a phenomenon of considerable current Abbreviations: ROOH, Hydroperoxide; ROS, Reactive oxygen species; FOX-1, Ferrous ion oxidation xylenol orange version 1; TPP, Triphenylphosphine; XO, Xylenol Orange; CSF, Cerebrospinal fluid. * Corresponding author. Tel.: +91-033-2440-3385; fax: +91-0825-2570061. E-mail address: dibyajyoti5200@yahoo.co.in (D. Banerjee). clinical interest as oxidative stress is related to a number of disease conditions [1]. Ferrous ion oxidation xylenol orange (FOX) assays have been used for the estimation of ROOH concentration and these assays outdated other available methods of hydroperoxide estimation [2,3]. FOX method of ROOH estimation is based upon oxidation of reagent Fe +2 to Fe +3 by sample-oxidizing agents which then binds with reagent xylenol orange (XO) to give a color complex having absorption maximum at 560 nm [2,3]. The Fe +3 XO color complex formation is strictly ph sensitive [4]. Two 0009-8981/$ - see front matter D 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.cccn.2003.08.004
148 D. Banerjee et al. / Clinica Chimica Acta 337 (2003) 147 152 versions of FOX assays are described in the literature, FOX-1 and FOX-2 [2,3]. In the FOX-1 assay, the reagent is prepared in aqueous medium with sorbitol. In the FOX-2 assay, the reagent contains HPLC grade methanol and butylated hydroxytoluene (BHT). In both versions, the reagent contains Fe +2 and H 2 SO 4 [2,3]. FOX assays as such are not specific for hydroperoxide as any other sample-oxidizing agent can oxidize reagent Fe +2 to Fe +3. FOX-2 assay is made specific earlier for ROOH by performing it in presence and absence of triphenylphosphine (TPP) which selectively reduces the hydroperoxides into their corresponding alcohols [2,5]. No such attempt is made with FOX-1 assay to make it specific for plasma hydroperoxide although FOX-1 assay is more sensitive for ROOH estimation when compared with FOX-2 assay [3]. It is postulated that the FOX-1 assay is unsuitable for plasma ROOH estimation. This is because plasma contains unperoxidized lipid, which can react with FOX reagent Fe +2 leading to false positive estimation of ROOH concentration [3]. This problem is overcome in the FOX-2 assay by using BHT, a chain-breaking antioxidant and, thus, this method is recommended for plasma hydroperoxide estimation [2,3]. Fe +3 XO complex formation is strictly ph sensitive and if the reagent is prepared in H 2 SO 4,itis reported to be maximum in the ph range 1.7 to 1.8 [4]. In both versions of FOX assays, ph sensitivity of Fe +3 XO complex is never taken into consideration. In this study, we stabilized the ph of the FOX-1 reagent between 1.7 and 1.8 and used it in conjunction with TPP and BHT to make the method more sensitive and specific for plasma hydroperoxide estimation. 2. Materials and methods 2.1. Materials Xylenol orange, cumene hydroperoxide and butylated hydroxytoluene (BHT) were from Sigma. All the other reagents were at least of reagent grade. Double distilled water was used throughout the study. Glassware were cleaned with warm concentrated nitric acid and thoroughly rinsed with double distilled water before an experiment. A Genesis 10 UV spectrophotometer, Cyberscan 510 ph meter and 1 ml Kontron quartz cuvette were used throughout the study. 2.2. Absorbance of xylenol orange prepared in 25 mmol/l H 2 SO 4 in various ph One hundred micromolar solution of xylenol orange was prepared in 25 mmol/l H 2 SO 4. Its ph was noted. Then it was adjusted to various ph (1.5, 1.55, 1.6,1.65,1.7,1.75,1.8,1.85,1.9 and 1.95) by addition of little amounts of solid Na 2 HPO 4 and kept in different test tubes. Vigorous vortexing using a magnetic stirrer was necessary for complete dissolution of Na 2 HPO 4. The absorbance of prepared xylenol orange solutions at different ph was measured at 560 nm. 2.3. Absorbance of ferric xylenol orange complex at various ph In 10 test tubes, 1 ml of xylenol orange solution prepared as above was taken adjusted at 10 different ph (1.5, 1.55...1.95). In all the tubes, 1 Al of10 mmol/l ammonium ferric sulfate i.e., 10 Amol/l Fe +3 was added using a microlitre (Hamilton) syringe. The tubes were vortexed and kept at room temperature for 45 min. Then, absorbance of the solutions at different ph containing Fe +3 xylenol orange complex was measured at 560 nm. 2.4. Preparation of reagents for FOX assays FOX-2 reagent and 10 mmol/l TPP solution in HPLC grade methanol were prepared as described before [2,5]. Stock FOX-1 reagent contained 100 Amol/l xylenol orange and 100 mmol/l sorbitol in 25 mmol/l H 2 SO 4 adjusted to ph 1.7 1.8 by addition of Na 2 HPO 4. Working FOX-1 reagent was prepared fresh each time before use by addition of 250 Amol/ l ammonium ferrous sulfate. Working TPP reagent for FOX-1 assay of hydroperoxide was prepared by addition of 10 mmol/l TPP and 400 mmol/l BHT in HPLC grade methanol. Before preparation of working FOX-1 reagent, the ph of the stock FOX-1 reagent was checked. The shelf life of TPP reagent was 1 month if stored in dark at 4 jc. The shelf life of stock FOX-1 reagent was 1 week, if stored in a brown bottle in dark at 4 jc.
D. Banerjee et al. / Clinica Chimica Acta 337 (2003) 147 152 149 2.5. Sample collection Blood was collected with heparin as anticoagulant from normal human subjects (n = 20) after obtaining informed consent and centrifuged at 3000 g 10 min immediately to separate clear plasma and plasma hydroperoxide concentration was estimated immediately. Urine, Cerebrospinal fluid (CSF) and ascitic fluid were collected from Clinical Biochemistry Laboratory of Kasturba Medical College Hospital, Manipal. Plasma of diabetic patients (n = 20) was also collected from this lab. The hydroperoxide concentration was estimated immediately after collection. 2.6. Estimation of plasma hydroperoxide Hydroperoxide concentration was estimated from plasma obtained from normal human subjects (n = 20) and also from the plasma of diabetic patients (n = 20). In each plasma sample, hydroperoxide was estimated by FOX-2 method, FOX-1 method in conjunction with BHT and TPP, reagent ph adjusted to 1.7 1.8 (modified FOX-1) and by FOX-1 method in conjunction with BHT and TPP, but without reagent adjustment. The FOX-2 assay was performed as previously described [5] and hydroperoxide concentration was calculated by a freshly prepared standard curve. To perform Modified FOX-1 assay, 90 Al of same plasma sample was taken in two test tubes. In one, 10 Al of methanol containing 400 mmol/l BHT was added. In the other, 10 Al of methanol was added which contained 10 mmol/l TPP and 400 mmol/l BHT. Both the tubes were vortexed and kept at room temperature for 15 min. Then, in both tubes, 900 Al of l of working FOX-1 reagent (ph adjusted to 1.7 1.8) was added. Both tubes were vortexed again and kept at room temperature for 30 min. The absorbance of the solution in the two tubes was measured at 560 nm. Plasma hydroperoxide was calculated from the absorbance difference of the two test tubes (with and without TPP) from a standard curve. The same procedure was repeated with the unadjusted FOX-1 reagent. The percent recovery of hydroperoxide was calculated by FOX-2 assay and above-mentioned FOX-1 assays. 2.8. Statistical methods All values were statistically analyzed and represented as mean F standard error of the mean (including the values of figures). A Student s t-test was used to analyze the values and a p < 0.05 was taken as significant. Correlation coefficient (r) of plasma hydroperoxide concentration values by modified FOX-1 method and by other methods mentioned here were calculated and r significant at least at 5% level was considered significant. 3. Results Freshly prepared 100 Amol/l XO in 25 mmol/ lh 2 SO 4 showed ph around 1.46. The absorbance of the above XO solution was lower in the ph range 1.7 1.8 (Fig. 1). Addition of a strong base followed by development of a purple color in the XO solution. 2.7. Recovery experiments Ten and 0.5 Amol/l of Cumene hydroperoxide was added to plasma, urine, CSF and ascitic fluid. Fig. 1. Absorbance of 100 Amol/l xylenol orange solution in 25 mmol/l H 2 SO 4 at 560 nm adjusted at various ph.
150 D. Banerjee et al. / Clinica Chimica Acta 337 (2003) 147 152 From ph 1.65 to 1.7 Fe +3 XO complex formation showed a sharp rise (Fig. 2). In ph range 1.7 to 1.8, a high absorbance of Fe +3 XO complex was detected (Fig. 2). After ph 2, the Fe +3 X O complex formation was less (data not shown). The standard curve for estimation of ROOH by modified FOX-1 method was more sensitive when ph of the FOX-1 reagent was adjusted to 1.7 1.8 compared to the unadjusted FOX-1 reagent (i.e., ph 1.46) (Fig. 3). The percent recovery of cumene hydroperoxide by FOX-1 assay with reagent ph unadjusted from plasma (n = 25), CSF (n = 25), urine (n = 25) and ascitic fluid (n = 25) were 7.6 F 3.1, 7.8 F 2.9, 8.2 F 2.4, respectively, when 10 Amol/l of cumene hydroperoxide was added to these fluids. An insignificant percent of hydroperoxide was recovered by FOX-2 assay when 0.5 Amol/l cumene hydroperoxide was added to these biological fluids. Less than 55% of hydroperoxide was recovered when 0.5 Amol/l of cumene hydroperoxide was Fig. 2. Absorbance of ferric xylenol orange complex at 560 nm at various ph. In 1 ml of 100 Amol/l XO solution in 25 mmol/l H 2 SO 4 adjusted at various ph 10 Amol/l of Fe +3 was added, vortexed and kept at room temperature for 45 min and the absorbance measured at 560 nm. Fig. 3. Standard curve for estimation of ROOH by modified FOX-1 method with reagent ph adjusted to 1.7 1.8 (a) and reagent ph unadjusted (b). added to the biological fluids by modified FOX-1 assay where reagent ph was not adjusted (i.e., around 1.46). The percent recovery of cumene hy-
D. Banerjee et al. / Clinica Chimica Acta 337 (2003) 147 152 151 Table 1 Plasma hydroperoxide concentration in normal subjects (n = 20) and diabetic patients (n = 20) are shown by modified FOX-1 method (both reagent ph adjusted to 1.7 1.8 and unadjusted) and FOX-2 method Hydroperoxide concentration in normal plasma (Amol/l) Hydroperoxide concentration in diabetic plasma (Amol/l) By modified FOX-1 2.6 F 1.5 6.9 F 2.3 [reagent ph 1.7 1.8] By modified FOX-1 1.7 F 2.5 4.2 F 3.5 r = 0.51, p < 0.01, n =20 r = 0.37, p < 0.05, n =20 reagent ph unadjusted By FOX-2 0.8 F 0.39 2.7 F 1.4 r = 0.46, p < 0.02, n =20 r = 0.48, p < 0.02, n =20 Correlation coefficient of plasma hydroperoxide value by modified FOX-1 method with other methods done in normal and diabetic plasma is shown in columns I and II, respectively. I II droperoxide after adding 10 Amol/l hydroperoxide from the biological fluids was near 100% by modified FOX-1 assay using reagent where ph was adjusted to 1.7 1.8. When 0.5 Amol/l of cumene hydroperoxide was added to the biological fluids, 90 95% of hydroperoxide was recovered by phmodified FOX-1 assay. In normal human subjects, the plasma hydroperoxide concentration using the modified FOX-1 method was significantly higher than either the FOX-2 method ( p < 0.01) or the unmodified FOX-1 method (p < 0.05). The plasma of diabetic patients also showed high hydroperoxide concentration ( p < 0.05) by the ph-adjusted modified FOX-1 method (Table 1). The correlation coefficients of the plasma hydroperoxide concentration by modified FOX-1 and other methods are shown in Table 1. 4. Discussion FOX-1 assay was described earlier as more sensitive than FOX-2 method for estimation of hydroperoxide [3]. The more sensitivity of FOX-1 was due to the use of sorbitol, a radical scavenger, which increases the yield of Fe +3 ion [3]. We increased the sensitivity of FOX-1 assay by ph-stabilizing the FOX-1 reagent in which the reagent showed a minimum absorbance (Fig. 1) with maximum Fe +3 XO complex formation (Fig. 2). Stabilization of FOX reagent ph for estimation of hydroperoxide was advised before [6] but never applied in biological samples. So in this modification, we are using the synergistic effect of sorbitol and ph to get a better sensitivity of hydroperoxide estimation by FOX-1 method. After stabilizing, the ph of the FOX-1 reagent, we obtained good recovery of cumene hydroperoxide from biological samples when compared to FOX-2 method or to FOX-1 method with reagent ph unadjusted. Modified FOX-1 method standard curve has shown a better sensitivity when reagent ph was adjusted compared to when reagent ph was not adjusted (Fig. 3). Plasma hydroperoxide concentration in normal human subjects and in diabetic patients showed a significantly increased value by modified FOX-1 method when compared to FOX-2 method with a significant correlation (Table 1). The chance of false positivity regarding plasma hydroperoxide estimation by modified FOX-1 method is minimized by applying BHT to the sample in the incubation phase, a strategy analogous to the FOX-2 method [2]. Analogous to FOX-2 method, modified FOX-1 method is made specific for plasma hydroperoxide by its performance in presence and absence of TPP, an agent that selectively reduces the hydroperoxide, [2,5]. So in this communication, FOX-1 method is developed as a more sensitive and specific method for estimation of hydroperoxide in biological sample. Sensitivity of assay methods for estimation of plasma hydroperoxide is important because a sensitive HPLC-based method (where TPP was used to increase specificity) reports plasma lipid hydroperoxide concentration around 0.03 umol/l, but the report lacks a statistical assessment [7]. Moreover,
152 D. Banerjee et al. / Clinica Chimica Acta 337 (2003) 147 152 the measurement involves an extraction step which can result in sample hydroperoxide loss [5,7]. Another enzymatic method reports plasma hydroperoxide concentration around 0.5 Amol/l in normal individuals [7]. Therefore, a controversy exists regarding plasma hydroperoxide concentration. If plasma hydroperoxide concentration is really lower than the detection limit of FOX-2 method, modified FOX-1 method can be used as it is equally specific for ROOH and more sensitive for ROOH determination when compared to FOX-2 method. We recommend a detail comparison of modified FOX-1 method with other available ROOH estimation methods. Acknowledgements The authors acknowledge Prakash. P.Y., MSc Student Department of Microbiology, Kasturba Medical College Manipal for meticulously typewriting this manuscript. References [1] McCord JM. The evolution of free radicals and oxidative stress. American Journal of Medicine 2000;108:652 9. [2] Nourooz-Zadeh J. Ferrous ion oxidation in presence of xylenol orange for detection of lipid hydroperoxides in plasma. In: Packer L, editor. Methods in enzymology, vol. 300. San Diego: Academic Press; 1999. p. 58 62. [3] Wolff SP. Ferrous ion oxidation in presence of xylenol orange for measurement of hydroperoxides. In: Packer L, editor. Methods in enzymology, vol. 233. San Diego: Academic Press; 1984. p. 182 9. [4] Gay C, Collins J, Gebicki JM. Determination of Iron in solution with ferric xylenol orange complex. Analytical Biochemistry 1999;273:143 8. [5] Nourooz-Zadeh J, Tajaddini-Sarmadi J, Wolff SP. Measurement of plasma hydroperoxide concentration by the ferrous oxidation-xylenol orange assay in conjunction with triphenylphosphine. Analytical Biochemistry 1994;220:403 9. [6] Gay C, Collins J, Gebicki JM. Hydroperoxide assay with the ferric-xylenol orange complex. Analytical Biochemistry 1999; 273:149 55. [7] Feri B, Yamamato Y, Niclas D, Ames BN. Evaluation of an isoluminal chemiluminescence assay for the detection of hydroperoxides in human blood plasma. Analytical Biochemistry 1988;175:120 30.