EFFECT OF TOTAL LIPIDS AND . PHOSPHOUPIDS 0N WARMED-OVER ~ FLAVOR MEASURED BY TBA ANALYSIS? w . -- ~ -- m MUSCLE FROM SEVERAL SPECIES . 1 -_-- - w MLcchN STATE UNWERSITY - BLAENE ROBERT waLseN _1974V Thesis for the aegree AHA. S. . . ————— ..... 1115.513. IIIII III III IIII II III III III IIII II IIIII III III II III 102134 7, ——-—, A A ~ * S smomo av I IIDAG & SflNS' 800K RIIIDERY INC. I LIICBFI {Y BINDERS ISPRHE ..':T MICIIIII .1 I" ( ABSTRACT EFFECT OF TOTAL LIPIDS AND PHOSPHOLIPIDS ON WARMED-OVER FLAVOR MEASURED BY TBA ANALYSIS IN MUSCLE FROM SEVERAL SPECIES BY Blayne Robert Wilson In the present study five species of meat animals were investi- gated to determine the relationship between the lipid components and the development of warmed-over flavor as measured by the 2—thio- barbituric acid (TBA) test. Samples of red muscle were obtained from the longissimus dorsi muscles of mature beef and mutton. Thigh muscles (red muscle) and breast muscles (white muscle) were sampled from both young hen turkeys and mature laying chickens. Market age pork semitendinosis muscle was divided into its predominantly red and white portions, which were sampled separately. TBA values were determined on raw samples, on cooked samples immediately following heating in a boiling water bath to an internal temperature of 70°C and on samples refrigerated at 4°C for 48 hours. after cooking. it was noted that mean TBA values for raw red muscle samples showed relatively low values for mutton (0.14), pork (0.24), beef (0.95), and chicken (1.26), while raw turkey samples had an aver— age TBA value of 5.85. Raw white samples showed similar results Blayne Robert Wilson with mean TBA values of 0.69 for pork, 1.61 for chicken and 2.42 for turkey. dean TBA values immediately after cooking increased consider- ably for all samples with the exception of mutton (red muscle), which showed no increase. After storage at 4°C for 48 hours, mean TBA values for red muscle samples were 2.96 for mutton, 3.71 for beef, 6.03 for pork, 9.20 for chicken and 11.47 for turkey. White muscle samples showed similar, although smaller, increases in mean TBA values for pork (5.83), chicken (8.60) and turkey (8.63). These results show that beef, chicken and turkey TBA values for red and white samples approximately doubled during storage at 4°C for 48 hours, while pork white muscle had an approximately 4-fold increase, pork red muscle had an approximately 6-fold increase and mutton red muscle increased about 20-fold. Total lipid and phospholipid measurements were made on raw muscle samples. Total lipid values for red muscles averaged 1.86% for turkey, 4.74% for chicken, 5.47% for pork, 5.58% for mutton and 14.79% for beef, while mean total lipid values for white muscle samples were 0.79, 1.52 and 8.88% for turkey, chicken and pork, respectively. Phospholipids as a percentage of total lipid for red muscle samples averaged 3.56, 16.73, 17.25, 35.43 and 42.25% for beef, pork, mutton, turkey and chicken, respectively. Mean values for white muscle samples for phospholipid as a percentage of lipid were 11.97% for pork, 42.40% for chicken and 64.42% for turkey. Average levels of phospholipid as a percentage of tissue for all samples ranged from 0.50 to 1.00% with the exception of chicken red muscle, which had a value of 1.60%. TBA values of lipid levels were correlated to determine their Blayne Robert Wilson relationships. Over-all correlation coefficients indicated a signi~ ficant (P < .05) negative relationship between TBA values and total lipid. However, significant (P < .10) positive correlations between TBA values and total lipid for pork red and white muscle samples suggest that pork muscle with higher lipid levels is more subject to warmed-over flavor development. The over-all association between TBA values and phospholipid as a percentage of muscle tissue was essentially zero. A significant (P < .05) positive correlation was established between TBA values and phospholipid as a percentage of lipid. These positive relationships indicated that as the percentage of phospholipid in the total lipid increases, there is an increase in warmed-over flavor. Thus, with the exception of pork, a close relationship was found to exist between TBA values and phospholipid as a percentage of lipid. These results indicate that phospholipids are an important factor in warmed- over flavor development as measured by the TBA test. EFFECT OF TOTAL LIPIDS AND PHOSPHOLIPIDS ON WARMED-OVER FLAVOR MEASURED BY TBA ANALYSIS IN MUSCLE FROM SEVERAL SPECIES BY Blayne Robert Wilson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1974 ACKNOWLEDGEMENTS The author expresses his sincere appreciation to Dr. A. M. Pearson for his guidance and encouragement throughout the course of graduate study and during the preparation of this thesis. Acknowledgement is given to Dr. R. A. Merkel, Professor of Food Science, and Dr. R. W. Luecke, Professor of Biochemistry, for serving as members of the author's guidance committee. Thanks is given to Tom Obremski for his assistance in statistical analysis. The author wishes to express special appreciation to his parents for their constant support and interest throughout his college career. Finally, the author is grateful to his wife, Lani, and children, Daniel and Lisa, for their support, understanding and sacrifice. ii TABLE OF CONTENTS Page INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . 1 REVIEW OF LITERATURE . . . . . . . . . . . . . . . . . . . . . 3 Oxidation of Meat Lipids . . . . . . . . . . . . . . . . 3 Muscle lipid composition . . . . . . . . . . . . . . 3 Levels of muscle lipids . . . . . . . . . . . . . . . 4 Oxidation of phospholipids in cooked meats . . . . . 5 warmed-over flavor o o o o o o o o o o o o o o o o o 7 Mechanisms of Lipid Oxidation . . . . . . . . . . . . . 9 Autoxidation.................... 9 Catalysts of lipid oxidatio . . . . . . . . . . . . 11 2-Thiobarbituric Acid (TBA) Test . . . . . . . . , . 12 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . . 17 mterials O I I I O O O O O O O O O O O O O O O O O 0 I O 17 Solvents and Chemicals . . . . . . . . . .'. . . . . 17 Samples . . . . . . . . . . . . . . . . . . . . . . 17 2-thiobarbituric ac d test for lipid oxidation , . Total lipid analysis . . . . . . . . . . . , , , , 20 Isolation of phospholipid fraction . . . . , , , , , 20 Statistical treatment . . . . . . . . . . . . . . . . 21 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . . . . 23 TBA Values of Red Muscles of Different Species . . . . , , 23 TBA Values of White Muscles of Different Species , , , , , 25 Total Lipids and Phospholipid Levels in Red and White Muscles from Different Species . . . . . . , . . . 27 iii Page Relationship of Total Lipids and Phospholipid Levels to Increases in TBA Values of Cooked Muscle During Storage at 4°C . . . . . . . . . . . . . . . . . . . . . 30 Relationship of TBA values and total lipids . . . . . 30 Relationship of TBA values and phospholipid as percentage of tissue . . . . . . . . . . . . . . 32 Relationship between TBA values and phospholipid as a percentage of lipid . . . . . . . . . . . . . 33 Over-all relationships between TBA values and lipids O O O O O O O O O O O O O O O O O O O O O 0 3S SWY O O O O O O O O O O I O O O O O O O O O O O O O O O O O 38 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . 41 APPENDIX 0 0 O O O O O O O O O I O O O I I O O O O I O O O O O 47 iv LIST OF TABLES Table Page 1 TBA levels and standard deviations for the red muscles from mutton, beef, pork, chicken and turkey O O O O O O O O O O O O O O O O O O O O O O O O 2 4 2 TBA levels and standard deviations for the white muscles from pork, chicken and turkey. . . . . . . . . 26 3 Lipid levels and standard deviations for red muscles from mutton, beef, pork, chicken and turkey O O O O O O O O O O O O O O O O I O O 0 O O I O 28 4 Lipid levels and standard deviations for white muscles from pork, chicken and turkey . . . . . . . . 29 5 The correlation coefficients of TBA values and total lipid levels . . . . . . . . . . . . . . . . . . 31 6 The correlation coefficients between TBA values and phospholipid as a percentage of tissue . . . . . . 32 7 The correlation coefficients of TBA values and phospholipid as a percentage of total lipids . . . . . 34 8 A summation of correlation coefficients for the three lipid parameters evaluated against TBA values 0 O O O O O O O O O O O O O O O O O O O O O O O 36 LIST OF APPENDIX TABLES Table Page I TBA values for all muscles for several species . . . . . 47 II Total lipid values and phospholipids as a percentage of tissue and as a percentage of lipid . . . . . . . . 48 vi INTRODUCTION The oxidation of lipids is a significant factor in deterioration of quality in meat and meat products. A gradual deterioration occurs in fresh and frozen meat due to lipid oxidation. When freshly cooked meat is stored for short periods of time, however, it rapidly develops a stale flavor, commonly referred to as warmed-over flavor. The increased use of precooked meat products in institutions and in the home signi- ficantly increases the importance of understanding lipid oxidation and its adverse effects upon the acceptability of meat products. The phospholipid fraction of meat lipids has been implicated as the cause of oxidative deterioration in cooked meat and the resulting rancid flavor, which deve10ps rapidly during storage. The relatively high content of unsaturated fatty acids in the phospholipid fraction of meat lipids appears to be related to the lability of meat to devel- opment of warmed-over flavor (Love and Pearson, 1971). Earlier studies have shown that cooked meat from various animal species appears to undergo lipid oxidation at different rates under the same conditions. However, little attention has been paid to phos- pholipid concentrations and their effect upon the susceptibility of muscle lipids to undergo autoxidation. Although significant research has been carried out on the problem of warmed-over flavor, more studies are needed to elucidate the role that phospholipids play in autoxidation of cooked meat. This study was undertaken to investigate the relationship between development of warmed-over flavor in cooked meat and the levels of phos- pholipids in various muscles from different species of animals. REVIEW OF LITERATURE Oxidation of Meat Lipids Muscle Lipid Composition The two general classifications of meat lipids are depot or intermuscular lipids and tissue or intramuscular lipids (Love, 1972). Depot lipids, as the name suggests, are stored in large deposits as adipose tissue. Tissue lipids are interspersed throughout the muscle tissue as an integral part of the cellular structure. The tissue lipids are closely associated with tissue proteins and contain a significant portion of the total phospholipids (Watts, 1962). Lea (1962) reported that high lipid concentrations are not neces- sary to bring about serious autoxidation problems. He further indicated that lipid oxidation is often due to catalysts of lipid oxidation in foods. The more metabolically active tissues and organs usually contain a higher proportion of their fatty acids in the form of complex lipids, particularly as phospholipids (Sato and Herring, 1973). Even though the phospholipid portion of meat is relatively low, the suscep— tibility of phospholipids to oxidation makes them an important factor in determining meat quality (Love and Pearson, 1971). The relative lability of the phospholipid fraction results from its high content of unsaturated fatty acids (Watts, 1954; Younathan and Watts, 1959; Lea, 1962; Love and Pearson, 1971). Hornstein g£_§l, (1961) indicated 3 that 19% of the fatty acids in beef muscle phospholipids have four or more double bonds, while only 0.1% of the triglyceride fatty acids in beef show this level of unsaturation. In pork muscle phospholipids, 16% of the fatty acids have four or more double bonds compared to about 0.1% for triglyceride fatty acids (Hornstein g£_§l,, 1961). Phospholipid levels have been reported to be relatively constant in muscles from different animals or carcass locations, while total lipid and neutral lipid levels are more variable (Bernstein £5 21., 1967; O'Keefe g£_gl,, 1968). The phospholipids make up less than 1% of total muscle weight, while the triglyceride fraction is about five times as large (Hornstein g£_gl,, 1961). Luddy gt 3;. (1970) indicated that the phospholipid fatty acids from light or white muscles contain a predominance of monoenes, while polyunsaturated fatty acids are present in higher amounts in the phospholipids from dark or red muscle. Levels of Muscle Lipids Fat is a major component of the carcasses from meat animals and is exceeded in quantity only by the water content. Fat comprises 18-30% of the carcass weight of market steers and 12-20% of the live weight of an average market hog (Dugan, 1971). Watts (1962) reported lipid levels in lean beef muscle to be from 2-4%, while pork muscle lipid levels ranged from 5-7%. Although Kramlich gt_§l, (1973) indicated that phospholipids are found in only small concentrations in animal fats, they indicated. phospholipids are important as functional and structural components of cells and membranes. Phospholipids usually occur as phosphogly- cerides in muscle tissue and comprise about 0.5 to 1.0% of lean muscle (Kramlich £5 31,, 1973). As total muscle lipid decreased from 5% to 1%, the percentage of phospholipid to total lipid showed an increase from less than 10% to nearly 70% (Dugan, 1971). Oxidation of Phospholipids in Cooked Meat Initial studies by Watts (1954, 1961) were concerned with adipose tissue lipid oxidation as a cause of meat rancidity. It was then noted that flavor deterioration of cooked meat did not correlate with values for oxidation of neutral lipids (Timms and Watts, 1958). Later work indicated more oxidation of the phospholipid fraction of rancid, cooked pork than for the neutral lipids (Younathan and Watts, 1960). After observing that the phospholipid fraction and the total lipids from beef and pork quickly became rancid upon exposure to atmospheric oxygen, Hornstein g£_§l, (1961) concluded that neutral fat developed off-flavors less readily than the phospholipids. It quickly became apparent that even though phospholipids were present in relatively small quantities in muscle tissue, they had a significant effect on meat rancidity because of their rapid oxidation (Watts, 1962). Rapid oxidation results from the relatively high level of unsaturated fatty acids in the phospholipids (Lea, 1957). According to Corliss (1968), the phosphorylated bases may also affect the oxidation of unsaturated fatty acids in the phospholipid molecule. He reported that the induction time interval for oxidation of phospholipids appears to be a function of the nitrogen containing moiety. He found that the choline portion of phosphatidyl choline exerts a smaller prooxidant effect than does the ethanolamine moiety of phosphatidyl ethanolamine. Thus, Corliss (1968) concluded that the rate of phospholipid oxidation during the steady state is a function of the unsaturation of the fatty acid components of phospholipids. Campbell and Turkki (1967) reported that neutral lipids are lost more readily than phospholipids during cooking of meat. Thus, the rela- tive concentration of the meat phospholipids to total lipids increases during cooking. At the same time, cooking pork or beef by a dry heat method failed to make any significant changes in the fatty acid com- position of the phospholipids (Campbell and Turkki, 1967). In contrast, other workers (Chang and Watts, 1952) have reported that cooking does not alter the fatty acid composition of ether-extractable lipids from meat and poultry. Unsaturated fatty acids are widely distributed in animal tissue (Sato and Herring,l973). Hornstein.g£ El. (1961) have indicated that the most plentiful polyunsaturated fatty acids in animal tissue are the C18, C20, and C22 acids with two to six double bonds. These fatty acids are subject to rapid oxidation, making them mainly respon- sible for the development of rancidity in meat tissues, including that of warmed-over flavor (Lea, 1962). Although it is possible for oxidation to take place in fatty acids containing only one double bond, such as oleic, the methylene group, which is located between the two double bonded carbon atoms, is significantly more susceptible to oxidative attack than are carbon atoms adjacent to a single double bond (Gunstone and Hilditch, 1945; Watts, 1954; Lundberg, 1962). Thus, Watts (1954) pointed out that linoleic acid, which has one active methylene group, will oxidize ten to twelve times as rapidly as oleic acid. She also stated that lino- 1enic acid with two of these labile carbon atoms, oxidizes twice as fast as linoleic acid. Lundberg (1962) concluded that fatty acid derivatives of oleic or other polyunsaturated fatty acids with methy- 1ene-interrupted unsaturation are subject to increasingly rapid oxida- tion rates, increasing in an exponential manner. The complexity of the oxidation process increases when the hydroperoxides, which are the primary autoxidation products, decompose to form more free radicals, which, in turn, accelerate the chain reaction (Dugan, 1961). Watts (1954) reported that hydroperoxides do not contribute to the off-flavors and odors developed by rancid lipids. The secondary products formed by hydroperoxide degradation, such as aldehydes, ketones, and acids, are susceptible to continued oxidation and are responsible for the rancid odors and flavors in lipids (Lea, 1962; Lundberg, 1962; Scherwin, 1972). Warmed-Over Flavor " "stale" and "rancid" have Various terms such as "warmed—over, been used to describe the flavors which develop in stored, cooked meat (Timms and Watts, 1958). Younathan and Watts (1960) proposed the idea that undesirable flavor changes are due to degradation products of the unsaturated phospholipids, since these lipids are rapidly oxidized in cooked meats. Watts (1962) presented data showing that cooked meats develop rancid flavors, which occur as a result of oxidative changes in muscle lipids during subsequent storage. She reported that heating of fresh lean beef and pork cause protein denaturation, making the meat susceptible to lipid oxidation. Oxidation is accompanied by large increases in thiobarbituric acid (TBA) values and development of rancid odors. Cooked meat from other species of animals shows similar increases in oxidative flavor changes as evidenced by increased TBA values in cooked, refrigerated chicken and turkey (Jacobsen and Koehler, 1970). Satolggngl. (1973) also reported similar results with turkey muscle. They showed significant warmed-over flavor development in light and dark turkey muscle after two days storage at 4°C. Sato and Hegarty (1971) showed that warmed-over flavor also develops in uncooked ground meat in about the same degree and rapidity as in cooked meat. They found large increases in TBA values of raw ground beef within one hour after grinding, if the meat was exposed to air at room temperature. Thus, they postulated that any process which disrupts the muscle membrane system, such as cooking or grinding, will result in exposure of the lipid component to oxygen and other oxidative catalysts, thus, accelerating development of rancidity. Investigation of compounds contributing to meat flavor has re- sulted in some studies dealing with the contribution of lipids to meat flavor. Hornstein and Crowe (1960) and Hornstein g; 31. (1961) have shown the importance of the depot fat in the development of species-specific aromas of cooked beef, pork and lamb. However, they did not attempt to investigate any possible effects of intramuscular fat on meat flavors. Herz and Chang (1970) indicated that the carbonyls are the most numerous class of compounds in meat flavor concentrates, with hexanal being identified as a major component. Some evidence indi- cates that hexanal is a product of lipid oxidation in cooked, fresh meat (Cross and Ziegler, 1965). After isolating the volatile fractions from cured and uncured pork, Cross and Ziegler (1965) observed that hexanal was present in the uncured pork, but absent from the cured product. The failure of cured, cooked meat to undergo the extensive lipid oxidation observed in cooked, uncured meat has been noted by Younathan and Watts (1959). Thus, some carbonyl compounds involved in meat flavor may be a result of lipid oxidation and may effect cooked meat flavor in an undesirable manner. Mechanisms of Lipid Oxidation The mechanism of lipid autoxidation, which occurs as a spontan- eous reaction between lipids and atmospheric oxygen, has been reported in numerous studies (Dugan, 1961; Lundberg, 1962; Schultz g£_gl,, 1962; Scherwin, 1972; Sato and Herring, 1973). Autoxidation Recent evidence indicates that lipid oxidation occurs as a sequen- tial reaction based on free radical formation. Lipid oxidation is believed to occur in three stages: initiation, propagation and term- ination (Dugan, 1961; Ingold g£_al,, 1962; Scherwin, 1972). Dugan (1961) stated that initiation corresponds to the oxidative induction period of the lipid. He indicated that unstable free radi- cals are formed, which act as catalysts for additional free radical formation in the substrate. He described the reaction as follows: ZRH + 02 ———9 2R- + 20H Dugan (1961) stated that the induction period is the period of time required for development of rancidity before it becomes detectable 10 by organoleptic means. Principal initiators of autoxidation are agents such as light, heat and heavy metals, particularly copper and iron (Watts, 1962). Dugan (1961) stated that in the prepagation stage, previously formed fatty free radicals (Ro) combine with molecular oxygen to form peroxide-free radicals (ROO°). He indicated that further reactions with the substrate by these peroxide-free radicals form more free radicals (R'-) and hydrOperoxides (ROOH) as indicated in the following summation: R- + 02 ——-—‘: R00- R00' + R'H ————) ROOH + R'- Kaunitz (1962) and Lundberg (1962) presented evidence to indicate that decomposition of the hydroperoxides during the propagation stage leads to compounds, which are responsible for the rancid odor and flavor of oxidized lipids. According to Ingold (1962), termination of the chain reaction occurs upon destruction or deactivation of the free radicals. He proposed the following reactions to demonstrate some of the possibilities: R-+R-——-—) RR R'+R00°——-> ROOR ROO° + ROO“—————9 ‘ROOR + 02 Dugan (1961) indicated that antioxidants may function as inhibitors by reacting with free radicals during propagation stage to form inert 11 products, which result in termination of the chain reaction. Catalysts of Lipid Oxidation Several compounds have been implicated as catalysts affecting oxidation in animal tissue lipids (Tappel, 1953; Lin. 1970; Sato and Hegarty, 1971). It is generally accepted that iron porphyrins, which are present at significant levels in muscle tissue as heme compounds, act as prooxidants (Tappel, 1962). These include some hemoglobin from the blood and appreciable quantities of the muscle pigment, myoglobin (Craig g£_§l,, 1966). In a review of meat pigment chemistry, Fox (1966) described myoglobin as existing in three forms in the fresh state, oxymyoglobin, reduced myoglobin and metmyoglobin. He further stated that oxymyo- globin contributes to the bright, red color of meat, while reduced myoglobin is purplish-red in color, and metmyoglobin causes the un- desirable dark brown color in fresh meat. The balance that exists between these three pigment forms is controlled by atmospheric oxygen levels and the activity of enzyme reducing systems in the tissues (Watts £5 31., 1966). Fox (1966) also indicated that the pigments are irreversibly converted to denatured ferric hemichromagen during heating. Hetmyoglobin, in which the iron is in the ferric state, has also been reported to be a catalyst of lipid oxidation in fresh meat (Greene 35 al., 1971). Recent evidence has implicated non-heme iron as playing a major role in acceleration of oxidation in muscle lipids (Maclean and Castell, 1964; Liu and Wafés, 1970; Sato and Hegarty, 1971). Liu and Watts (1970) detected significant lipid oxidation of meat in the 12 absence of metmyoglobin. Maclean and Castell (1964) added trace amounts of iron to cod muscle and produced a rancid odor. Sato and Hegarty (1971) demonstrated that non-heme iron accelerated lipid oxidation in water extracted cooked meat. They also reported that myoglobin and hemoglobin do not act as prooxidants in cooked meat. Love (1972) verified the effectiveness of non-heme iron as a prooxidant in cooked meat, and concluded that myoglobin was not a major prooxidant. 2-thiobarbituric Acid (TBA) Test Watts (1954) and Dugan (1955) have reviewed the objective methods that have been developed to measure oxidative deterioration in lipids. They indicated that the common tests are the determination of carbonyl compounds, the active oxygen method and peroxide values, all of which measure degree of unsaturation, and the 2-thiobarbituric acid test (TBA), which measures the level of malonaldehyde. All of these methods have some limitations and are more successful in some systems than in others. Most tests for oxidative rancidity involve extraction of the fat (mainly triglycerides) from meat tissues. The phospholipids and proteolipids, which are primarily involved with oxidation and rancidity (Younathan and Watts, 1960), are not extracted by normal hydrocarbon solvents, but require more polar solvents, such as methanol and ethanol, (Lea, 1957). An advantage of the TBA test is that the fat does not need to be extracted from the rest of the muscle tissue (Tarladgis £5 31., 1960). It would, therefore, be expected to measure malonaldehyde 13 produced from autoxidation occurring in all of the lipid fractions. Malonaldehyde is an end product of lipid oxidation, while the peroxides are intermediate compounds and do not accumulate because of their rapid decomposition (Watts, 1954). Although_malonaldehyde.itself does not contribute to off-odors and flavors, a highly significant correlation between the TBA test and sensory evaluation of rancidity has been established (Tarladgis gt_al,, 1960). Kwon and Watts (1963) ~~..._ .. 4.. noted fhat malonaldehyde production is a useful indicator of flavor deterioration. The TBA test has been employed to determine the degree of rancidity in a wide range of products, including dairy products (Dunkley, 1951; Biggs and Bryand, 1953), meats (Tarladgis gt_gl., 1960; Zipser and Watts, 1962), fish (Schwartz and Watts, 1956; Sinn- huber and Yu, 1958), baked and cereal products (Caldwell and Grogg, 1955) and fats and oils (Sidwell gt al., 1954; Sinnhuber st 31., 1958). The 2-thiobarbituric acid (TBA) test has been used extensively to study lipid oxidation (Lea, 1962). Kohn and Liversedge (1944) found that animal tissues produced a red color when reacted with 2-thiobarbituric acid. They proposed that.an unidentified carbonyl compound was responsible for the development of the red color. Bern- heim gt 31. (1947) postulated that the color was a product of unsat- urated fatty acid oxidation. Wilbur gt_al, (1949) observed a similar color reaction between various aldehydes and sugars and concluded that it was due to production of a three carbon compound. Early work by Powick (1923) suggested that the compounds respon- sible for the Kreis color reaction test for rancid fats were epihydrin aldehyde or its acetals. However, Patton and cadworkers (1951) found 14 evidence that malonic dialdehyde is responsible for the Kreis color reaction. Patton and Kurtz (1951) also were able to show that malonic dialdehyde produced a red color upon heating in the presence of 2- thiobarbituric acid. Sinnhuber st 31. (1958) proposed that the red pigment was a condensation product of two molecules of 2-thiobarbituric acid and one molecule of malonaldehyde with the elimination of two water molecules, which is illustrated as follows: N S\/N N\ as / \011 o\ /o \ on H0 // SH 2 + \c-caz-c/ -———>HC1 I \ 1120 N H H N =CH-CHPCH- N + 2H 0 S/ \ v 2 H 0H OK In 1962, Dahle st 31, proposed a mechanism.of malonaldehyde forma- tion from conjugated fatty acids containing three or more double bonds. Lillard and Day (1964) reported malonaldehyde formation from the oxi- dation of various dienals, which are produced as a result of unsaturated fatty acid oxidation. This indicates the possibility that malonalde- hyde can be produced from autoxidation of polyunsaturated fatty acids and the continued oxidation of secondary autoxidation products. A number of methods have been reported in the literature for performing the TBA test (Sidwell gt gl., 1954; Turner gt El}: 1954; Sinnhuber and Yu, 1958; Sinnhuber gt 31,, 1958; Tarladgis st 31., 1960; Tarladgis gt 31., 1964; Marcuse and Johansson, 1973). Sidwell st 31. (1954) described a steam distillation method for dried milk, with a direct color reading after reaction with TBA. Turner gt El: (1954) reacted pork with TBA in combination with trichloroacetic and phosphoric 15 acid. An isoamyl alcohol-pyridine mixture was used to extract color. Sinnhuber and Yu (1958) used a similar method for fishery products, but extracted the color with petroleum ether. Tarladgis gt_al, (1960) re- ported a technique using steam distillation rather than passing live steam through the sample as was done by Sidwell gt_§1, (1954). Tarladgis §t_§l, (1964) later observed that TBA structure was altered upon acid- heat treatment and proposed a new extraction technique, which allowed the TBA to react with a sample of filtrate in the dark for 15 hours at room temperature. However, it was not shown whether or not acid and heat are needed to release malonaldehyde from the oxidized product. The ability to quantify the "TBA Number" as mg of malonaldehyde per 1,000 g of sample came about as a result of the discovery that l,1,3,3, tetraethoxypropane (TEP) yields malonaldehyde upon acid hydrolysis (Sinnhuber and Yu, 1958). Kwon and Watts (1963) postulated that pre-formed malonaldehyde reacted with other food components and was not distillable. Thus, they proposed the term "distillable malonaldehyde" for use when describing the TBA test. After observing malonaldehyde in aqueous solution, Kwon and Watts (1964) concluded that malonaldehyde has the capacity to enolize from its diketo form (I) to its enolate anion (II), which is not volatile. A third volatile chelated form (111) is also possible (Kwon and Watts, 1964). These three forms are shown below: *k ° A E E“ H\ / 'H H\ / \o 2 .2 )2 .._——-> E . o/ \u o \H \o/ (I) (II) (III) 16 Kwon and watts (1964) also indicated that in aqueous solution, almost all (96%) of the malonaldehyde is in the enolic form (11), and that the various forms are pH dependant. They also presented evidence that the enolic form (II) occurs at pH > 7, while the chelated form (111) dominates at pH < 3. Therefore, as indicated by Kwon and Watts (1964), maximum volatilization of free-preformed malonaldehyde would occur at pH < 3. This points out that the acid is added in order to free the malonaldehyde from other possible combinations in food constituents. The TBA test can be directly applied to lipid-containing material without previous fat extraction (Lea, 1962). However, Kwon gt 3;. (1965) indicated that substances other than lipid oxidation products can react with the TBA reagent to give the red color measured in the TBA test. As an alternative method, the product can be acidified and steam distilled with the TBA reaction being carried out on the distillate (Tarladgis gt 31., 1960). Trace amounts of Fe2+ or Fe3+ have been reported to cause increased TBA values (Wills, 1964). Ascorbic acid has also been alluded to as a cause of high TBA values (Wills, 1966). EXPERIMENTAL Materials Solvents and Chemicals All solvents used were of the highest purity grade. Only reagent grade chemicals were used. Water was distilled and deionized before use 0 Samples Three 200 g samples from each of five different species were obtained from various sources at Michigan State University and were analyzed for development of warmed-over flavor. Those species with significant amounts of white muscle (i.e., pork, chicken, and turkey) were sampled for both red and white muscle, while beef and mutton were sampled only for red muscle. Breast (white) and thigh (red) muscles were obtained 24 hours post-mortem for chicken and turkey. Mutton longissimus dorsi (red) muscle was sampled 24 hours after slaughter. Pork semitendinosis muscle was divided into the predominantly red and white sections and a sample from each section was removed 48 hours after slaughter. Beef longissimus dorsi (red) muscle was obtained 72 hours post-marten. All samples were trimmed of excess subcutaneous adipose tissue and ground twice through a 1/8th inch plate. Ten gram portions were 17 18 weighed and placed in heavy glass tubes.. Raw samples were then sub- jected to TBA analysis. Samples to be cooked were placed in a boiling water bath until they reached an internal temperature of 70°C. Zero day samples were analyzed immediately for TBA values, and the remaining cooked samples were refrigerated for 48 hours and then analyzed for development of warmed-over flavor by TBA analysis. Portions of the raw samples were frozen and stored at —10°C for later lipid analysis. The frozen raw tissue samples were removed from freezer storage and thawed at 4°C for 24 hours. The thawed samples were then analyzed for total lipids and phospholipid content. The lipid data were then evaluated against the TBA numbers obtained in order to determine the relationship between phospholipid levels and development of warmed- over flavor in the cooked, stored meat. Z-Thiobarbituric Acid Test for Lipid Oxidation A TBA method similar to the distillation method of Tarladgis gt a1. (1960) was used. The distillation apparatus consisted of a 250 ml round bottom flask, which was attached to a Friedrich conden- sor with a three way connecting tube. Electric heating mantles were used to facilitate rapid distillation. A 10 g sample was homogenized with 50 m1 of distilled, deionized water in a Virtis homogenizer for 2 minutes at low speed. The mixture was transferred quantitatively into a 250 m1 round bottom flask along with 47.5 ml of distilled, deionized water. The pH of the mixture. was lowered to 1.5 by the addition of 2.5 m1 4N HCl. Boiling chips were added and a small amount of Dow antifoam was sprayed into the flask to control the foam level. The slurry was then steam distilled at high heat until 50 m1 of distillate was collected. The distillate 19 was mixed and 5 ml were transferred to a 16 mm x 125 mm screw cap test tube.- Then 5 m1 of 0.02 M TBA reagent (0.02 M 2-thiobarbituric acid in 90% glacial acetic acid) were added. The tubes were capped and mixed, then placed in a boiling water bath for 35 minutes. After cooling in cold water for 10 minutes, absorbance was read on a Beck- man DU spectrophotometer at 538 am against a blank containing dis- tilled, deionized water and TBA reagent. A standard curve and recoveries were obtained in order to deter- mine the distillation constant needed to convert absorbance to mg of malonaldehyde per 1000 g of meat. Dilutions of 1, 3, 5, 7 and 9 x 10.8 moles of l, 1, 3, 3-tetraethoxypropane (TEP) per 5 ml were used in order to obtain a standard curve. The dilutions replaced the 5 ml of sample distillate and the procedure was continued as previously outlined. The percentage of recovery was determined by addition of 5 x 10.8 moles of TEP to the sample before distillation and deter- mining the amount recovered. Recovery was determined to be 78%. Tar- ladgis gt a1. (1960) reported a percentage of recovery of 68%. The distillation constant (k) was calculated according to the method described by Witte gt_gl, (1970) as follows: S 6 100 k - A x MW x 10 x P where: S - conc. in moles/5 ml distillate (5 x 10-8) A - absorbance MW - molecular weight of malonaldehyde (72) "d I % recovery TBA numbers were calculated using the distillation constant (k) ob- tained and reported as mg of malonaldehyde per 1000 g of meat. 20 Total Lipid Analysis Extraction of total lipid was performed using a modification of the technique described by Folch §t_§1, (1957). Frozen raw meat samples were thawed and weighed to the nearest 0.0001 g. To approxi- , mate natural moisture levels, 17 ml distilled water were added to 33 g of raw meat in a Waring blender. After adding 50 m1 of chloro- form, the mixture was blended at low speed for 1 minute. After adding 100 m1 of methanol, the slurry was blended for an additional 1 minute at low speed. Two more 30 second blendings were made after the addi- tion of 50 m1 chloroform and 50 ml water, respectively. This mixture was placed in a Coors No. 3 Bfichner funnel with Whatman No. 1 filter paper under slight suction. The blender was rinsed with 10-15 ml aliquots of chloroform and methanol, which were added to the contents of the Bfichner funnel. After placing the filtrate in a 500 m1 separa- tory funnel and rinsing the sidearm flask with 10-15 ml aliquots of methanol and chloroform, the lipid-containing chloroform layer was quantita- tively decanted into a 250 ml volumetric flask. The contents of the flask were then brought to volume with additional chloroform. Measured 10 ml aliquots of the extract were pipetted into tared beakers and the chloroform was evaporated under a stream of nitrogen. Percent lipid, based on the weight of the lipid in the 10 ml aliquot, was calculated as follows: wt. of lipid (3)710 ml x 25 wt. of sample (g) lipid % - x 100 Isolation of Phospholipid Fraction Phospholipid separation from the total lipid was accomplished using the method of Choudhury gt_al, (1960). This method involves 21 separation on activated silicic acid, in which neutral lipids are preferentially removed by washing with chloroform, followed by solu- bilization of the phospholipids with methanol. Silicic acid was activated by drying in a 100°C oven for 16 hours. A slurry was formed by adding 10 m1 of chloroform and 10 ml of lipid extract to 5 g of activated silicic acid in a 250 ml beaker. This slurry was then transferred to a 150 m1 60-M Bfichner funnel fitted with a fritted disk and washed with 400 ml of chloroform using slight vacuum. After placing a clean side-arm flask under the Bfichner funnel, the phospholipid containing silicic acid was washed with 300 m1 meth- anol. Each solvent was then decanted into a 500 ml round bottom flask. A Rotovapor was used to reduce the solvent volume to 5-10 ml. The remaining solvent was then poured into a dried, tared 15 m1 ampule and evaporated to dryness with nitrogen. Percentages of phospholipid in the original tissue samples were calculated using the following formula: wt. of phospholipid (g)/10 ml x 25 x sample weight 100 % phospholipid in tissue - Percent of phospholipid in total lipid was calculated according to the following formula: wt. of_phospholipid Z phospholipid 1“ lipid ' wt. of total lipid x 100 Statistical Treatment Statistical analysis was calculated using a Michigan State Univ- ersity computer package program identified as PLOTXY and run on a Control Data Corporation (CDC) 6500 computer. Factors analyzed for 22 correlation coefficients were TBA values, total lipid, phospholipid as a % of tissue and phospholipid as a % of total lipid. Significance of the computed correlation coefficients was determined by using Fisher's Distribution of "t" table (Mason, 1970). RESULTS AND DISCUSSION TBA Values for Red Muscles of Different Species Samples were prepared by removing all subcutaneous fat and then grinding to achieve homogeneity. Portions of each sample were analyzed for TBA levels in either the raw state or after cooking. The cooked samples were then divided into two portions, which were either tested immediately by TBA analysis or refrigerated for 48 hours prior to measurement for warmed-over flavor development by TBA analysis. Table 1 presents the means and standard deviations of TBA values for red muscles. Appendix Table I contains the raw data. The data show the susceptibility of the red muscles to lipid oxidation upon cooking and storage. Watts (1962) has stated that "the threshold value, i.e., the TBA number at which off-odor becomes detectable, is approximately one." Thus, the values recorded for the raw samples show that mutton, beef and pork have initial TBA numbers below the threshold with values of 0.14, 0.95 and 0.24, respectively. Chicken was slightly above the threshold level with a TBA value of 1.26, while the raw turkey samples were well above the threshold level with a TBA number of 5.85. Beef, pork and chicken samples showed considerable increase in TBA values upon cooking. The beef and pork samples had TBA numbers of 1.92 and 1.07, respectively, while the chicken samples increased about 23 24 three-fold in TBA numbers with a value of 3.13. Only the mutton samples with a mean value of 0.15 remained below the threshold level, showing no increase in average TBA value. Cooking resulted in only a slight increase in the TBA numbers for turkey muscle with a mean of 6.00. Table 1. TBA levels and standard deviations for(the red muscles from mutton, beef, pork, chicken and turkey a 0 day 48 hour Sample Raw cooked cooked Mutton