Fluorometric Analysis of 2-Thiobarbituric Acid Reactive Substances in Turkey1 C. JO and D. U. AHN2 Department of Animal Science, Iowa State University, Ames, Iowa 50011-3150 ABSTRACT Three experiments were conducted to develop a sensitive and reliable fluorometric thiobarbituric acid (TBA) reactive substances (TBARS) method to determine lipid oxidation products in meat. The first study was conducted to find the optimum sample preparation conditions for meat in the fluorometric method. The second study was to compare the original and the modified methods by using meat and blood samples. The modified fluorometric method was compared with a conventional spectrophotometric method in a third study. Four different extraction solutions (2.5 M acetic acid, 0.5 M hydrochloric acid, 0.8 M perchloric acid, and 1.4 M trichloroacetic acid) and two ratios of extraction solution to TBA (20 mM) solutions (2:1 and 1:1) were examined in the first study. Hydrochloric acid was the optimum among the four extraction solutions tested, and the ratio of extraction solution to TBA solution at 1: 1 was the best for the fluorometric TBARS method in raw ground turkey. The modified fluorometric method had high recovery rates (91%, average), and the regression coefficient of the standard curve prepared with spikes was also high (r2 = 0.99). The analysis of raw meat and plasma samples indicated that the modified fluorometric method had greater sensitivity than the original method. The pH of the reaction mixture played an important role in extraction TBARS from samples, and low pH conditions were preferable to high pH conditions. The amount of lipid oxidation products in raw turkey breast meat indicated that the fluorometric method had much greater sensitivity than the spectrophotometric method. The results from the three studies showed that the modified fluorometric TBARS method was useful for the samples with low lipid oxidation products, such as fresh raw meat. The sensitivity of the modified fluorometric method also facilitates the determination of oxidation products in small quantities of samples. (Key words: fluorometric analysis, spectrophotometric analysis, thiobarbituric acid reactive substances, lipid oxidation, pH) 1998 Poultry Science 77:475–480 However, Kosugi et al. (1989) reported that the absorbance of red pigments in TBA methods may not specifically quantify MDA because other substances may also react with TBA and contribute to absorbance. In addition, heating and acidic condition may cause overestimation of thiobarbituric acid reactive substances (TBARS). Gray and Monahan (1992) reported that MDA interacted with various food constituents such as amino acids, proteins, glycogen, and other food ingredients. The interaction rendered the MDA unavailable to react with TBA reagent and resulted in an underestimation of TBARS. Because of insufficient specificity and sensitivity of conventional TBARS methods, alternative analytical approaches have been developed. Kakuda et al. (1981) developed a HPLC method to assess the amount of MDA in aqueous distillates of chicken meat. They found INTRODUCTION Lipid oxidation is one of the major causes of quality deterioration in meat. Several methods have been developed to assess lipid oxidation products in muscle foods. The thiobarbituric acid (TBA) test (Tarladgis et al., 1960) is among the most widely used to quantify lipid oxidation products in meat and meat products because it is simple and fast. The TBA test determines the amount of malondialdehyde (MDA), a major secondary byproduct of lipid oxidation, in a sample. The determination of MDA in meat or meat products has been studied widely (Hoyland and Taylor, 1991), and significant correlations between the TBA values and sensory scores of poultry meat have been reported (Salih et al., 1987). Received for publication March 28, 1997. Accepted for publication November 1, 1997. 1Journal Paper Number J-17326 of the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa, Project Number 3190, and supported by Food Safety Consortium and Hatch Act. 2To whom correspondence should be addressed: duahn@iastate.edu Abbreviation Key: BHT = butylated hydroxytoluene; DDW = deionized distilled water; MDA = malondialdehyde; TBA = thiobarbituric acid; TBARS = thiobarbituric acid reactive substances; TCA = trichloroacetic acid; TMP = 1,1,3,3-tetramethoxypropane. 475 476 JO AND AHN that HPLC was more sensitive than spectrophotometric methods in assessing lipid oxidation products in aqueous distillates of chicken. Other secondary products of lipid oxidation, such as pentanal (Ang and Young, 1989) or hexanal (Shahidi and Pegg, 1994), also has been used to determine lipid oxidation by a HPLC, but conventional spectrophotometric TBARS methods are still preferred over chromatographic methods because of their simplicity. Yagi (1987) developed a fluorometric method that can determine lipid oxidation products in serum and plasma. Yagi’s fluorometric method had greater sensitivity than the spectrophotometric method, but was not appropriate for meat. The objective of present work was to develop a sensitive, simple, and reliable fluorometric method that can determine lipid oxidation products in meat. To accomplish this objective, the following studies were conducted: 1) the determination of the best sample preparation conditions for meat, 2) the comparison between the Yagi (1987) original and the modified methods using meat and plasma, and 3) the comparison between modified fluorometric and conventional spectrophotometric methods using meat. MATERIALS AND METHODS Chemicals and Reagents Acetic acid, HCl, trichloroacetic acid (TCA), perchloric acid (70%), n-butanol, pyridine, ethanol, and SDS were obtained from Fisher.3 Thiobarbituric acid, butylated hydroxytoluene (BHT), and 1,1,3,3-tetramethoxypropane (TMP) were purchased from Sigma.4 Acetic acid (2.5 M), HCl (0.5 M), perchloric acid (0.8 M), TCA (1.4 M), and SDS (8.1%) were prepared by dissolving an appropriate amount of each chemical in deionized distilled water (DDW). The TBA-TCA solution was prepared by dissolving 20 mM TBA in 15% TCA solution and BHT (7.2%) was dissolved in ethanol (95%). The nbutanol-pyridine solution (15:1, vol/vol) was made by mixing 900 mL of n-butanol with 60 mL of pyridine. All chemicals and reagents were reagent grades and used without further purification. Sample Preparation and Storage Turkey breast meat, ground twice through a 3-mm plate, was used to prepare patties (approximately 100 g). Patties (experimental unit) were stored in a 4 C refrigerator for 7 d, and analyzed by using the spectrophotometric and the fluorometric methods. Meat (3 or 1 g) was weighed into a test tube (50 mL) with 9 mL of 3Fisher Scientific, Pittsburgh, PA 15219-4785. 4Sigma Chemical Co., St. Louis, MO 63178-9916. 5Brinkman Instruments Inc., Westbury, NY 11590-0207. 6Gilford Instrument Laboratories, Inc., Oberlin, OH 44074. 7Barnstead/Thermolyne Corp., Dubuque, IA 52004-0797. DDW and homogenized with a Brinkman polytron5 for 15 s at high speed. BHT (7.2%, 50 mL) was added before homogenization. The homogenate was used for both the spectrophotometric (3 g) and the fluorometric methods (1 g). Blood samples were taken from live turkeys 1 d before slaughter. Blood was collected in a test tube containing anti-coagulant (heparin, 200 unit/mL) and centrifuged for 10 min at 1,000 × g. The plasma was collected and stored in a freezer (–20 C) until used. TBARS Analysis Spectrophotometric Method. The meat homogenate (1 mL) prepared as above was taken into a disposable test tube (10 mL), and 2 mL of the TBA-TCA solution were added. The mixture was vortexed, heated in a 90 C waterbath for 15 min, cooled in a cold water bath for 10 min, and centrifuged at 2,000 × g for 15 min. The absorbance of supernatant was measured at 531 nm with a microsample spectrophotometer.6 Fluorometric Method. Yagi’s original method (1987) was used except that the samples were measured by a fluorometer7 (Model 450, Barnstead/Thermolyne) with 520 nm excitation and 550 nm emission instead of 515 nm excitation and 553 nm emission, due to instrumental limitations. Four extraction solutions, 2.5 M acetic acid, 0.5 M HCl, 0.8 M perchloric acid, and 1.4 M TCA, were prepared. To compare the role of TBA reagent on the yields of TBARS, two extraction solution to TBA solution ratios (2:1, 1:1, vol/vol) were also used. After selection of the best conditions for meat samples, the selected conditions were compared with the original conditions using plasma and meat samples. To examine the effect of pH on the yield of lipid oxidation products, Yagi’s original method (1987) was used with some modification. Hydrochloric acid (2 M) was used to adjust appropriate pH conditions for reaction mixtures before heating, TMP was used as a TBARS standard. For the modified method, 0.5 mL of meat homogenate or 0.5 mL of plasma sample, 200 mL of 8.1% SDS, 1.5 mL of HCl, 1.5 mL of 20 mM TBA, 50 mL of BHT, and 250 mL of DDW were added to a test tube. The sample was vortexed and heated in a 90 C waterbath for 15 min. After cooling for 10 min, 1 mL of DDW and 5 mL of n-butanol-pyridine solution was added. The sample was mixed thoroughly, and centrifuged at 3,000 × g for 15 min. The fluorescence reading of upper layer was measured by a fluorometer with 520-nm excitation and 550-nm emission. Measurements were sequentially made at gain 1 and 5. Statistical Analysis A two-way ANOVA using the General Linear Models procedure of SAS (SAS Institute, 1989), was performed to determine the effect of extraction solution and extraction solution to TBA reagent ratio on the yields of TBARS. Three replications were made and four readings 477 FLUOROMETRIC TBARS ANALYSIS TABLE 1. Fluorescence reading of meat samples with different extraction solutions and extraction solution to thiobarbituric acid (TBA) (20 mM) reagent combinations1 Extraction solution5 Ratio Acetic acid HCl 1:1 2:1 SEM4 Mean SEM5 44.25cx 35.92cy 1.82 40.08d 63.83ax 57.17ay 2.15 60.50a Perchloric acid 51.17bx 46.33by 1.44 48.75c 1.64 Trichloroacetic acid 58.17ax 51.42by 1.09 54.79b SEM2 Mean SEM3 2.35 1.97 54.35x 47.70y 1.161 1.161 a–dMeans within a row with no common superscript differ significantly (P < 0.05); n = 16. within a column with no common superscript differ significantly (P < 0.05); n = 4. 12.5 M acetic acid, 0.5 M hydrochloric acid, 0.8 M perchloric acid, 1.4 M trichloroacetic acid. 2SEM: Among the means of different extraction solutions with same ratios; n = 12. 3SEM: Among the means of two ratios; n = 24. 4SEM: Among the means of the same extraction solution with different ratios; n = 6. 5SEM: AMong the means of the same extraction solutions; n = 24. x,yMeans were made for each replication. Differences among mean values of fluorescence readings were determined by calculating Student-Newman-Keul’s multiple range test. An ANOVA and Student’s t test were performed to determine differences between the original and modified methods. A two-way ANOVA was used to determine the effect of methods and storage time on TBARS values of meat. Each treatment combination had four replications. Mean values and SEM were reported. RESULTS AND DISCUSSION Sample Preparation Conditions The fluorescence readings of samples were significantly higher when the ratio of extraction solutions to TBA solution was 1:1 rather than 2:1 (Table 1). Therefore, the extraction solution to TBA (20 mM) solution ratio of 1:1 was recommended for extraction the TBA reactive substances (TBARS) in samples. Among the four extraction solutions, HCl produced the greatest fluorescence reading in both the ratios of extraction solution to TBA solution at 1:1 and 2:1. When the ratio of extraction solution to TBA reagent was 1:1, however, the yields from HCl and TCA were not significantly different. Tarladgis et al. (1960) originally used HCl as an extraction solution for TBARS. Recently, Rosmini et al. (1995) reported that 10% TCA solution produced the best recovery percentages for “paté” among combinations of phosphoric acid and TCA. The fluorescence yields from different extraction solutions increased in the order of acetic acid < perchloric acid < TCA < HCl when the fluorescence readings at 1:1 and 2:1 were pooled. However, the fluorescence yields of perchloric acid and TCA were not different when the ratio of extraction solutions to TBA solution was 2:1, and those of the TCA and HCl when the ratio was 1:1. Therefore, HCl would be the acid of choice when the ratio of extraction solutions to TBA solution is high (2:1) but both TCA and HCl can be used when the ratio is low (1:1). The top two extraction solutions, HCl and TCA, were selected from the yield study, and sequentially used to test the recovery rates of spikes in meat samples. The HCl had a greater slope than that of the TCA (Figure 1) and the average recovery rates of spikes were 91.0 and 64.6%, respectively (P < 0.01). Therefore, using HCl as an extraction solution produced 1.5 times higher yield than that of the TCA. Although both spikes had high regression coefficients (r2 = 0.99), HCl would be a better extraction solution for fluorometric analysis than TCA. Williams et al. (1983) suggested that TBA methods should report results in terms of their respective standards. Other researchers (Crackel et al., 1988) recom- FIGURE 1. The influence of extraction solutions on the slopes of standard curves prepared by the modified fluorometric method. Either HCl (0.5 M) or trichloroacetic acid (TCA) (1.4 M) was used as an extraction solution (♦: with HCl, y = 97.7x – 21.4, R2 = 0.998; x: with TCA, y = 63.4x – 16.2, R2 = 0.996). 478 JO AND AHN TABLE 2. The amount of lipid oxidation products in meat and plasma samples determined by the original and modified fluorometric methods Method Sample Meat Blood Actual readings from fluorometer Calculated concentration, mg MDA2/kg meat Actual readings from fluorometer Calculated concentration, nmol MDA/mL plasma Original1 Modified SEM 23.75b 0.69a 21.00b 0.88a 57.25a 0.52b 32.00a 0.41b 1.07 0.02 0.87 0.03 a,bMeans (n = 4) within a row with no common superscript differ significantly (P < 0.05). fluorometric method for serum and plasma (Yagi, 1987). 2MDA = malondialdehyde. 1Yagi’s mended determination of the recovery values for MDA and preparation of a standard curve for the calculation of an appropriate TBA conversion. Therefore, using extraction solutions with higher recovery rates would provide more accurate and reliable TBARS values than those with lower rates. Original vs the Modified Fluorometric Methods The modified method had higher readings than those of the original method (Table 2); however, when the amount of MDA was converted to TBARS values by using a standard curve, the modified method had lower values than those of the original method. The explanation of this result was revealed to be related to the pH of the two reaction mixtures prepared from the two methods. In the Yagi (1987) original method, the pH of reaction mixture was around 2.0, and that of the standard was 2.7. Yagi’s original method uses phosphotungstic acid and sulfuric acid to extract TBARS in blood or serum samples. The two acids added are removed from the sample after centrifugation, although a small amount of acids still remains in precipitants and maintains the pH of precipitants at around 2.0. Phosphotungstic acid and sulfuric acid are not added when a standard curve is prepared. Therefore, the pH of the reaction mixture before butanol extraction is higher (pH 2.7) than that of the sample. In the modified method, however, acid was added to both standard and sample preparations and resulted in the pH of the reaction mixtures being around 1.0. Figure 2 shows the effect of pH on the yield of lipid oxidation products determined by the original fluorometric method. The lower the pH and the higher the fluorescence readings were observed, at the pH range of 0.7 to 2.8. These data indicated that the pH difference in the reaction mixtures of the modified and the original methods can cause 1.5- to 2-fold differences in fluorometric readings. The pH effect in standard curves explains the opposite results between the fluorescence readings and the calculated amounts of MDA. The calculated values from the standard curve may produce erroneous values in Yagi’s method unless the pH differences between the standard and samples are considered. Therefore, the calculated amount of MDA from standard curve of modified fluorometric method is reliable because of the pH consistency. Kwon et al. (1965) reported that it is difficult to hydrolyze all MDA bound to meat proteins without using strong acidic conditions and heating. Chen and Waimaleongora-Ek (1981) reported that the pH values influenced the lipid oxidation of ground raw chicken meat as measured by a TBA test. They concluded that the lower the pH values in the sample, the higher were the TBARS values in the pH range of 3.09 to 9.50. Draper et al. (1986), on the other hand, suggested that, if strong acid was used, MDA-TBA complexes would become unstable and cause problems related to the quantitative recovery of MDA during a TBA test. The average recovery rate of our modified fluorometric method, however, was 91% at the pH range of 0.7 to 2.7 and was consistent. The pH of reaction mixture in both standard curve and real samples from modified fluorometric method was 1.0, and were consistent within the pH range (0.7 to 2.7) that we studied for pH effect on yields. FIGURE 2. Effect of pH on the recovery of standards added in meat samples (y = –66.4x + 385.6, R2 = 0.933). 479 FLUOROMETRIC TBARS ANALYSIS TABLE 3. The amount of lipid oxidation products of raw turkey breast meat by the fluorometric and the spectrophotometric methods during the 7-d storage Storage time Method 0 d 1 d 2 d Fluorometric Spectrometric SEM3 0.21ex 0.22abx 0.01 0.28dx 0.14cy 0.01 0.38bcx 0.21cy 0.01 3 d 4 d (mg MDA2/kg meat) 0.33cdx 0.41bx 0.12cy 0.22aby 0.01 0.02 5 d 6 d 7 d SEM1 0.43bc 0.22aby 0.02 0.50ax 0.18by 0.02 0.55ax 0.25ay 0.03 0.01 0.01 a–eMeans within a row with no common superscript differ significantly (P < 0.05). n = 16. within a column with no common superscript differ significantly (P < 0.05). n = 16. 1SEM: Among means of same method with different storage. n = 128. 2MDA = malondialdehyde. 3SEM: Among means of same day with different methods. n = 32. x,yMeans The lipid peroxide levels in normal human serum ranged from 1.86 to 3.94 nmol MDA/mL serum (Yagi, 1987). Our calculated TBARS values of the blood from 17-wk-old turkeys by Yagi’s original method were 0.88 nmol MDA/mL plasma and were 0.41 nmol MDA/mL plasma by the modified method. The modified method required less than half the time of the original method, and the results already discussed indicated that the modified method is superior to the original fluorometric TBARS for meat samples. Modified Fluorometric vs Conventional Spectrophotometric TBARS Methods The comparison of the lipid oxidation products in raw turkey breast meat indicated that the fluorometric method had a greater sensitivity than the spectrophotometric method (Table 3). The TBARS values from the fluorometric method gradually increased from 0.21 to 0.55. The results from the fluorometric method showed significant changes in TBARS values over storage, but TBARS values in the same meat samples obtained from the spectrophotometric method were not consistent. No meaningful differences in TBARS values could be detected during the 7-d storage. Lipid oxidation product in raw meat is generally very small because the conditions in raw meat are unfavorable for the further degradation of the primary oxidation products to the secondary products. Therefore, it is difficult to find differences in lipid oxidation products in raw meat when analyzed by the spectrophotometric method (Raharjo and Sofos, 1993). Siu and Draper (1978) investigated MDA content in various raw or cooked meats. Their TBA values of raw pork and chicken were 0.50 and 0.61 and are in good agreement with our results. There was a significant interaction (P < 0.01) between method and storage time. The differences in TBARS values of turkey raw meat from the two methods increased as the storage time increased (Table 3). This result indicated that even small changes in the status of lipid oxidation in raw meat can be detected when fluorometric method is used. From the results, we concluded that the modified fluorometric method can be successfully applied to determine the lipid oxidation in raw meat. The method is simple, reproducible, and sensitive, and requires small sample volumes. Therefore, the modified fluorometric method would facilitate the determination of oxidation products in small quantities of samples. It is important, however, to prepare both the reaction mixtures and standards at the same pH. REFERENCES Ang, C.Y.W., and L. L. Young, 1989. Rapid headspace gas chromatographic method for assessment of oxidative stability of cooked chicken meat. J. Assoc. Off. Anal. Chem. 72: 277–281. Chen, T. C., and C. Waimaleongora-Ek, 1981. Effect of pH on TBA values of ground raw poultry meat. J. Food Sci. 46: 1946–1947. Crackel, R. L., J. I. Gray, A. M. Pearson, A. M. Booren, and D. J. Buckley, 1988. Some further observation on the TBA test as an index of lipid oxidation in meat. Food Chem. 28:187–196. Draper, H. H., L. G. McGirr, and M. Hadley, 1986. The metabolism of malonaldehyde. Lipids 21:305–307. Gray, J. I., and F. J. Monahan, 1992. Measurement of lipid oxidation in meat and meat products. Trends Food Sci. Technol. 3:315–319. Hoyland, D. V., and A. J. Taylor, 1991. A review of the methodology of the 2-thiobarbituric acid test. Food Chem. 40:271–291. Kakuda, Y., D. W. Stanley, and F. R. Van De Voort, 1981. Determination of TBA number by high performance liquid chromatography. J. Am. Oil Chem. Soc. 58:773–775. Kosugi, H., T. Kojima, and K. Kikugawa, 1989. Thiobarbituric reactive substance from peroxide lipids. Lipids 24:873–881. Kwon, T. W., D. B. Menzel, and H. S. Olcott, 1965. Reactivity of malonaldehyde with food constituents. J. Food Sci. 30: 808–813. Raharjo, S., and J. N. Sofos, 1993. Methodology for measuring malonaldehyde as a product of lipid peroxidation in muscle tissues: A review. Meat Sci. 35:145–169. Rosmini, M. R., F. Perlo, J. A. Perez-Alvarez, M. J. PaganMoreno, A. Gago-Gago, and F. Lopez-Santovena, 1995. TBA test by an extractive method applied to ‘Pate’. Meat Sci. 42: 103–110. 480 JO AND AHN Salih, A. M., D. M. Smith, J. R. Price, and L. E. Dawson, 1987. Modified extraction 2-thiobarbituric acid method for measuring lipid oxidation in poultry. Poultry Sci. 66:1483–1488. SAS Institute, 1989. SAS User’s Guide. SAS Institute Inc., Cary, NC. Shahidi, F., and R. B. Pegg, 1994. Hexanal as an indicator of meat flavor deterioration. J. Food Lipids. 1:177–186. Siu, G. M., and H. H. Draper, 1978. A survey of the malonaldehyde content of retail meats and fish. J. Food Sci. 43:1147–1148. Tarladgis, B. G., B. M. Watts, M. T. Younathan, and L. R. Dugan, Jr., 1960. A distillation method for the quantitative determination of malonaldehyde in rancid flavor. J. Am. Oil Chem. Soc. 37:44–48. Williams, J. C., R. A. Field, G. J. Miller, and R. A. Welke, 1983. Evaluation of TBA methods for determination of lipid oxidation in red meat from four species. J. Food Sci. 48: 1776–1778. Yagi, K., 1987. Lipid peroxides and human disease. Chem. Phys. Lipids. 45:337–351.