Rapid Fluorimetric Method to Detect Total Plasma Malondialdehyde with Mild Derivatization Conditions

Malondialdehyde (MDA), an oxidation product of polyunsaturated fatty acids, is used as an in vivo marker to assess lipid peroxidation in diseases such as atherosclerosis and diabetes (1)(2)(3)(4). In biological matrixes, MDA is measured after derivatization with thiobarbituric acid (TBA) (5). Because TBA reacts with many other aldehydes (6), results are expressed as TBA-reactive substances (TBARS). Several problems are associated with TBARS analysis, in particular, low reproducibility and a lack of specificity that leads to overestimations. To overcome these difficulties, more specific methods have been proposed that require sample pretreatment to precipitate proteins and extract MDA-reactant adducts (6)(7)(8)(9). This additional step is time-consuming and adversely affects precision. The aim of the present study was to develop a rapid and sensitive method to measure MDA in plasma, avoiding sample pretreatment. Tetraethoxypropane (TEP), TBA, and bilirubin were obtained from Fluka. Fatty acid-free bovine serum albumin (BSA), Total Protein Reagent, and Protein Standard were from Sigma, and 2,2′-azobis(2-amidinopropane) (ABAP) was from Wako. We prepared an aqueous stock solution of 1 mmol/L TEP. A 10 μmol/L MDA solution was obtained by diluting TEP in 0.1 mol/L HCl. A 0.025 mol/L TBA solution was prepared daily by dissolving TBA in water. BSA solutions were prepared in 0.1 mol/L HCl. To remove protein-bound MDA, BSA solutions were heated at 80 °C for 1 h and dialyzed for 3 days against 0.1 mol/L HCl in a 3500-Da cutoff dialysis membrane (Spectrapore; Spectrum Medical Industries). The actual protein concentration was verified using the Total Protein Reagent Kit. A 50 mmol/L ABAP solution was prepared in water. Six MDA solutions ranging from 0.05 to 0.5 μmol/L were prepared by diluting the 10 μmol/L stock solution with 0.1 mol/L HCl. Triplicate solutions were used to obtain a dose–response curve. Calibration curves in plasma were obtained by adding MDA (0.2–0.6 μmol/L) in duplicate to six different plasma samples. The MDA-TBA adduct (TMT) was synthesized according to Guzmán-Chozas et al. (10). To investigate the effect of BSA on TMT fluorescence, we transferred 0.2, 0.4, and 0.6 μmol/L acidic TMT solutions (pH 2) to a magnetically stirred cuvette in a FL 55 Perkin-Elmer spectrofluorometer at 30 °C and recorded the fluorescence emission spectra at 549 nm (excitation at 520 nm). BSA (80 g/L) was added to produce a concentration of 10 g/L, and the spectra were recorded again. To study the effect of BSA concentration on TMT fluorescence, we assayed the same TMT solutions in the presence of 1.3–25 g/L BSA. For the assay, 200 μL of MDA or the EDTA-plasma samples was added to 700 μL of 0.1 mol/L HCl (pH 2). After 20 min at room temperature, 900 μL of 0.025 mol/L TBA was added, and the solution was maintained at 37 °C for 65 min. A 1.6-mL aliquot was transferred to a cuvette (kept at 37 °C); 400 μL of 120 g/L BSA was then added and the emission recorded (549 nm). Precision was assessed by analyzing pooled plasma and two calibrator solutions for within- and between-run SD. The detection and quantification limits were calculated according to EURACHEM guidelines (11). MDA recovery experiments were performed by incubating plasma with different added concentrations of MDA (final concentrations, 0.2, 0.4, and 0.6 μmol/L). To ensure that bilirubin did not interfere with the fluorometric detection of TMT, the pigment (12 μmol/L) was added to a plasma sample. We used the method to measure lipid peroxidation in plasma oxidized in vitro with ABAP. Aliquots of pooled EDTA plasma were added to four tubes. ABAP was then added to three of the tubes to give final concentrations of 4.9, 9.5, and 14 μmol/L. The tubes were incubated at 38 °C for 4 h, and MDA was determined. The experiment was repeated three times. Finally, the method was used to measure MDA in the plasma of 12 healthy human volunteers. The presence of protein led to enhancement of the fluorescence of the TMT adduct, as demonstrated by the linear increase in fluorescence intensity at 549 nm (Table 1 ). The fluorescence reached a plateau at a final protein concentration of 20 g/L. On the basis of these results, BSA was added to plasma samples to reach a concentration of 24 g/L before MDA detection. Effect of added BSA on TMT fluorescence at 549 nm (excitation at 520 nm). Mean (SD). Effect of added BSA on TMT fluorescence at 549 nm (excitation at 520 nm). Mean (SD). CD has been often used to study the binding of small molecules to proteins (12). Optically active small molecules show a change in CD on binding to a macromolecule, either because of electronic interactions with the binding site or conformational changes. These changes can be easily detected because most biopolymers have no CD in the visible light region. We used CD to investigate the binding of albumin to TMT (Fig. 1 ). The change in spectra suggests that the fluorescence enhancement of the adduct is attributable to the binding of TMT to BSA. Near-ultraviolet CD signals for BSA (dotted line), TMT (dashed line), and the BSA–TMT complex (solid line). The induced CD spectra were measured for a 1:1 BSA–TMT complex: [BSA] = [TMT] = 12 μmol/L in 10 mmol/L phosphate buffer, pH 2. Cell path length, 1 cm. Below pH 3, human serum albumin acquires a molten globule-like state (13), characterized by a substantial amount of secondary structure and a largely disordered tertiary structure (14). It is reasonable to assume that under our analytical conditions BSA is in the molten globule-like state, with exposed hydrophobic surfaces facilitating binding to TMT. To establish whether the same enhancing effect of BSA could be observed in the presence of plasma, 200 μL of human plasma containing 80 g/L total protein was added to 0.4 μmol/L TMT. An increase in fluorescence intensity was observed similar to that obtained with BSA [215.9 (1.9) vs 217.6 (1.4); n = 3]. On the basis of the enhancing effect of albumin on TMT fluorescence, we have developed a rapid and sensitive method to assess MDA in human plasma without sample pretreatment. The optimal reaction rate between MDA and TBA in the presence of albumin was obtained at 40 °C. Nevertheless, we carried out the assay at 37 °C, slightly sacrificing sensitivity but maintaining plasma in a physiologic condition. A HCl concentration of 0.1 mol/L was adequate to catalyze the reaction between MDA and TBA, as reported previously (6), while maintaining the albumin enhancement effect. The dose–response curve for aqueous solutions has been calculated between 0.05 and 0.5 μmol/L MDA (y = 138.79 ± 3.21x, where y is the fluorescence intensity in arbitrary fluorescence units and x is MDA concentration in μmol/L; R2 = 0.999). In the case of dose–response curves in plasma with added MDA, no matrix effect was observed [mean (SD) sensitivity from six different calibration curves, 140.68 (6.19)]. The detection and quantification limits were 0.015 and 0.025 μmol/L, respectively. Within- (n = 10) and between-run (n = 10) relative SDs were 2% and 3.8%, respectively, for aqueous calibrator solutions and 3.7% and 8.4% for plasma. The mean recovery, calculated on the basis of 20 experiments at three different MDA concentrations, was 97.3% ± 4.3%. The effects of contaminants under conditions similar to those used in the current study have been described (6). The reaction rate for TBA and common aldehydes at 65 °C is markedly lower than the reaction rate for TBA and MDA, allowing analysis at a contaminant:analyte ratio up to 100:1 by weight (6). Because plasma is used in our method, hemoglobin, bilirubin, or high concentrations of triglycerides could affect the results. Bilirubin showed no effect at physiologic concentrations. Regarding hemoglobin, the use of hemolyzed samples should be always avoided when oxidative status is measured because the cellular enzymes released may oxidize plasma components, inducing artificial peroxidation products. Triglycerides could adversely affect the assay because opalescence enhances light scattering. Nevertheless, when we tested samples from hypertriglyceridemic individuals, the assay worked well for triglyceride concentrations up to 4000 mg/L. We encountered problems, however, with a sample containing 11 900 mg/L. To evaluate the effectiveness of the proposed method for detecting MDA as a product of oxidative stress, we incubated plasma with increasing amounts of ABAP, a peroxy radical generator. The MDA values obtained were linear with the ABAP concentration (y = 0.0431x + 0.3109 μmol/L; R2 = 0.989). Finally, applying the method to a group of healthy volunteers, we found MDA values ranging from 0.067 to 0.169 μmol/L [mean (SD), 0.112 (0.034) μmol/L; n = 12], which is in good agreement with published values (5)(15). In conclusion, the proposed method is sensitive, rapid, easy to perform, and able to detect plasma MDA without overestimation as a result of sample handling, other interfering species, and matrix artificial oxidation. This work was supported by a COFIN 2001 grant from the Ministry of Education, University and Research (MIUR)