swmss on me SPECIFIC nwmmumu or 02.2mm m MM FRACTIONS _ Thesis fqr‘the Degree of Pb. EL . MICHIGAN STATE umvensm CATHARINA YUNG - we WANG MG 1970 LIBRARY thSlt‘ Michigan State University This is to certify that the thesis entitled STUDIES ON THE SPECIFIC DISTRIBUTION OF DIELDRIN IN MILK FRACTIONS presented by Catharina Yung-Kang Wang Ang has been accepted towards fulfillment of the requirements for Ph.D. Food Science degree in 7% @M Major profeésor fi ' Date—1.3.4313}! 1970 0-169 ABSTRACT STUDIES ON THE SPECIFIC DISTRIBUTION OF DIELDRIN IN MILK FRACTIONS BY Catharina Yung—Kang Wang Ang Most of the chlorinated hydrocarbon pesticides are lipophilic in nature, thus they are readily accumulated in fatty animal products. The present study was aimed to in- vestigate the specific relationship between dieldrin and different lipid classes in the milk system. Milk was sepa- rated into various fractions by differential solvent extrac- tion methods and by physical methods. The total lipid, phospholipid, and dieldrin contents in these fractions were determined. Several extraction methods for pesticide residue analysis were compared. A method of the Food and Drug Ad— ministration, which provided good recoveries of pesticide as well as total lipids from milk was used in this research. The final determination of residue was carried out by elec- tron capture gas chromatography. Solvent extractions of milk containing added pesti— I cide showed that the fat fraction extracted with non-polar was ar Us We re 78 Se Catharina Yung-Kang Wang Ang hydrocarbon solvents, contained only about one-half of the total amount of dieldrin in milk. This incomplete recovery of pesticide with non-polar solvents might be due to the low recovery of solvents used for the extraction or due to the same factor that depressed the total recovery of solvents from milk. On the other hand, the concentrations of residue in this fraction was much higher than in the fraction which was extracted with ethanol and non-polar solvents after the initial extraction. Therefore, it was shown that dieldrin was more readily extracted with hydrocarbon solvents than were the milk lipids. However, all residue could not be ex— tracted unless all lipid was extracted. The distribution of dieldrin was quite similar in physically separated fractions of milk containing added pes- ticide and milk containing physiologically incorporated pes- ticide. This observation suggests that the results of ex- periments involving distribution of dieldrin in milk from either source might be the same. IThe relative amounts of dieldrin in various fractions closely resembles the relative amounts of total lipids in these fractions and both of them are significantly varied with the separation techniques used. Several independent milk collections and separations were made. Approximate quantities of dieldrin in these fractions were: unwashed cream, 85-87%; skim milk, 10-15%; refined skim milk, l-3%; washed cream, 65-80%; butter, 52- 78%; butteroil, 51-78%; buttermilk, 0.4-2.0%; and butter serum, 0.05-0.08%. Catharina Yung-Kang Wang Ang Comparable concentrations of dieldrin on a fat basis were found in whole milk, skim milk, cream, washed cream, butter and butteroil. Lower levels of residue were observed in buttermilk and butter serum, and lowest values were found in refined buttermilk, refined butter serum and fat globule membrane pellets. Refined skim milk con- tained higher levels of pesticide on a fat basis than other fractions. These data reflect that different factors may influence the distribution of pesticide in milk. The lower levels of residue in membrane materials (buttermilk, butter serum and membrane pellets) were probably related to the higher levels of phospholipid content or high-melting glyceride content, or both, in these fractions. The greater quantities of residue observed in refined skim milk contain- ing extremely low amounts of lipids might be due to the slight solubility of this pesticide in the serum of milk. It is not certain whether dieldrin is less preferen- tially deposited in the lipids of buttermilk and butter serum in the natural original milk system or such a finding is a consequence of re-orientation during separation pro- cesses. Nevertheless, dieldrin apparently has a tendency to be distributed with the neutral lipids or free fats in milk. Analysis of control milk fractions did not lead to any firm conclusions since the residue levels in these sam- ples were too low. STUDIES ON THE SPECIFIC DISTRIBUTION OF DIELDRIN IN MILK FRACTIONS BY Catharina Yung-Kang Wang Ang A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Food Science 1970 ((\ \\ Dr. grad sist Cie ing ACKNOWLEDGMENTS The author sincerely wishes to express her thanks to Dr. L. R. Dugan, Jr. for his guidance throughout the entire graduate program, and for his valuable suggestions and as- sistance in the preparation of this thesis. Appreciation is extended to Dr. J. R. Brunner, Dr. H. A. Lillevik, Dr. P. Markakis and Dr. M. Zabik who con- tributed greatly as members of the author's guidance commit- tee. Special thanks are given to Dr. J. R. Brunner, Dr. L. G. Harmon, Dr. R. V. Lechowich, and Dr. M. Zabik (Ento- mology Department) for the use of their facilities to con- duct research, and to Mr. K. Wilson (Dairy Department) for his very kind assistance in providing milk samples used in this research. Lastly, the author would like to express her appre- ciation to her husband, Peng, for his patience, understand- ing and assistance during this graduate program. ii TABLE ACKNOWLEDGMENTS . . . LIST OF TABLES . . . LIST OF FIGURES . . . INTRODUCTION . . . . REVIEW OF LITERATURE . OF CONTENTS Pesticide Residues in Milk and Other FOOdS O C O 0 Sources of Contamination and Excretion of Residues in Milk Animal Decontamination 0 Metabolic Transformation of Dieldrin Effects of Chlorinated Pesticides in Biological Systems Chlorinated Hydrocarbon Residues in Human Tissues in Relation to Public Health Extraction and Determination of Chlorinated Hydrocarbon Pesticide Residues in Milk . Acid Hydrolysis Alkaline Hydrolysis . Surfactant Reaction Extraction Involving the Use of Polar Solvent One Step Extraction and Distribution of Residues in Milk Processing Techniques and EXPERIMENTAL . . . . Materials . . . Chemicals . . Milk . . . . Cleanup . Residues in iii Page ii vi viii 10 12 15 17 20 25 25 26 27 29 31 34 38 38 38 39 Page Methods . . . . . . . . . . . . . . 39 Extraction of Dieldrin from Milk . . . . 39 Food and Drug Administration Method . . 41 Mojonnier Method . . . . . . . . 42 Surfactant Method . . . . . . . . 43 Alkaline Hydrolysis Method . . . . . 43 Separation of Milk Fractions by Differential Solvent Extraction Methods . . . . . . . . . . . . 45 Cleanup Procedures for the Residue Extract . . . . . . . . . . . . 46 Acetonitrile Partition . . . . . . 48 Florisil-Celite Column Cleanup . . . 49 Determination of Dieldrin by Electron Capture Gas Chromatography . . . . . 49 Separation of Milk Fractions by Physical Methods . . . . . . . . . 51 Preparation of Major Milk Fractions . . . . . . . . . . 51 Preparation of Fat Globul Membrane Pellet Fractions . . . . 55 Analysis of the Physically Separated Milk Fractions . . . . . . . . . 58 Extraction of Lipids and Dieldrin . . . . . . . . . . 58 Determination of Dieldrin . . . . . 6O Phosphorus Determination . . . . . 60 RESULTS AND DISCUSSION . . . . . . . . . . 62 Comparison of the Extraction Effi- ciency of Lipid and Dieldrin in Milk . . . . . . . . . . . . . . 62 Distribution of ipids and Dieldrin in Solvent Separated Milk Fractions . . . . . 69 Distribution of Lipids and Dieldrin in Physically Separated Milk Fractions . . . . 76 Naturally Contaminated Milk . . . . . . 76 Milk Containing Added Dieldrin . . . . . 91 iv BIB] APP] SUMMARY AND CONCLUSIONS PROPOSALS FOR FURTHER RESEARCH BIBLIOGRAPHY APPENDIX Control Milk General Discussion Page 97 99 109 113 115 134 Tab] Table LIST OF TABLES Source of naturally contaminated milk: a record of dieldrin doses, feeding periods of dairy cows, and the col— lection dates of milk samples . . Comparison of extraction efficiency by different methods of lipid and dieldrin in milk . . . . . . . Recoveries of dieldrin from milk at different added levels by the Food and Drug Administration method . . . Distribution of lipids and dieldrin in solvent separated milk fractions . . Distribution of dieldrin in lipid— depleted milk fractions . . . . . Distribution of lipids and dieldrin in major fractions of naturally contamin- ated milk, preparation 1 . . . . . Distribution of lipids and dieldrin in major fractions of naturally contamin- ated milk, preparation 2 . . . . . Distribution of lipids and dieldrin in fat globule membrane pellets of naturally contaminated milk, preparation 1 . . . . . . . . 9a. Distribution of lipids and dieldrin in 9b. major fractions of naturally contamin- ated milk, preparation 3 . . . . . Distribution of lipids and dieldrin in major fractions of naturally contamin— ated milk, preparation 3 . . . . . vi Page 40 63 68 70 75 78 79 81 84 85 Table 10. ll. 12. 13. 14. 15. Table 10. 11. 12. 13. 14. 15. Comparison of the lipid and dieldrin contents in fresh and freeze-dried milk fractions . . . . . . . Distribution of lipids and dieldrin in freeze-dried fat globule membrane pellets of naturally contaminated milk, preparation 3 . . . . . Distribution of lipids and dieldrin in major fractions of raw milk con- taining added dieldrin . . . . Distribution of lipids and dieldrin in freeze-dried fat globule membrane materials of raw milk containing added dieldrin . . . . . . . Distribution of lipids and dieldrin in various fractions of control milk Comparison of the distribution of dieldrin in various milk fractions vii Page 88 89 92 96 98 100 Figure 1. LIST OF FIGURES Page Scheme for the separation of milk fractions by differential solvent extraction . . . . . . . . . . . . 47 Scheme for the separation of major milk fractions by physical methods . . . . 53 Scheme for the separation of fat globule membrane fractions . . . . . . 56 Chemical structuresof dieldrin and its major metabolites . . . . . . . . 136 A typical gas-liquid chromatogram for determination of dieldrin in milk 0 O O O O O O C O O O O O O 137 viii l A ‘3 um! a mo v 8 CO 31’ 1i du th th al INTRODUCTION Pesticide chemicals have been widely used in the past two decades for various purposes and have brought about significant developments in food and agricultural production. However, the use of such chemicals has also generated resi- due problems in our environment. Numerous reports have in- dicated the occurrence of these compounds in a variety of foodstuffs and biological materials, and caused some unin- tentional effects in non-target living organisms. Chlorinated hydrocarbon insecticides account for most of the residue problems. This group of chemicals is very persistent in normal conditions and very hydrophobic as compared to other kinds of pesticides, such as those in the organophosphate and carbamate groups. Consequently, they are more readily found, especially in fatty food products like egg, milk and meat. Detectable amounts of chlorinated hydrocarbon resi- dues in milk and dairy products have been reported frequent- ly. The possible routes of contamination in milk are through the feed ingested by cows, the direct application of these chemicals to animals and their barns, or by accident— al contamination during milking and handling. Y t NUL f: s L pz f M 51 9 d1 ri fr tc f a 5 SL bl Pr du Since milk is one of the major sources of human food, especially for infants and certain invalids, the pres- ence of any persistent toxic chemicals in these products is of great concern. High levels of residues in milk can be prevented ef- fectively through the control of feed contamination, whereas small quantities of these persistent chemicals are almost unavoidable in milk and other food products. Another ap- proach to reduce the pesticide levels is the application of food processing procedures to alter residue concentrations in the finished products. A number of routine dairy proces- sing techniques and other physical treatments have been evaluated for this purpose. The aim of the present research was to study the distribution properties of dieldrin, a commonly used chlo- rinated hydrocarbon insecticide, in milk, which was either from cows which had ingested dieldrin or had dieldrin added to it after milking. Chemical and physical separations of fluid milk into various fractions were made and their lipid and dieldrin contents were determined. This study may as- sist in perceptiveness of the characteristic deposition of such a toxicant in similar biological materials and a possi- ble relation to the toxification mechanism. It may further provide a rational basis for research on the improvement of processing techniques to reduce the amounts of trace resi- dues in milk and milk products. REVIEW OF LITERATURE Pesticide Residues in Milk and Other Foods Shortly after the introduction of DDT for insect control, residues of DDT were found in milk from cows which had ingested DDT (Wilson gt 31., 1946; Carter, 1947). Since that time, many other chlorinated hydrocarbon insecticides, such as dieldrin, BHC, lindane, aldrin, endrin, chlordane, and haptachlor and some of their derivatives have also been found in market milk in various levels as reported by Marth and Ellickson (1959c), Clifford gt a1. (1959), Heineman and Miller (1961), and Duggan (1967). A number of studies on the contamination of residues in food and total diet have been reported in the United States (Mills, 1963; Williams, 1964; Campbell gt $1., 1965; Duggan gt $1., 1966, 1967) and in other countries (Robinson and McGill, 1966; Abbott 23 31,, 1969). Duggan and co-work- ers in the Food and Drug Administration (FDA) have conducted a series of surveys on the residues in food and feed items (Duggan, 1967, 1968; Duggan and Dawson, 1967; Lipscomb, 1968), in the total diet samples (Duggan 33 a1., 1966, 1967; Martin and Duggan, 1968; Corneliussen, 1969), and the die- tary intake of pesticide residues (Duggan and Weatherwax, 1967; Duggan and Lipscomb, 1969). 3 The FDA studies showed that about one-half of the food items and all of the diet samples examined contained pesticide residues. The major residual substances in foods are the chlorinated hydrocarbon insecticides. Animal prod- ucts contained generally higher concentrations of residues as compared to the dietary fruits and crops. More than 50% of the total daily intake of chlorinated residues by the general population were from products of the dairy and meat food classes. Dairy products accounted for about 13.6% of the total daily intake. Since milk is an essential food item for infants and many milk products are important for human diet, the toler- ance levels of all chlorinated hydrocarbon chemicals in milk had been set at "zero" by 1961 (Moffitt, 1963). At that time, the lowest reliable determination level for DDT and DDT-like substances was 2.5 ppm on a fat basis by colori- metric methods. The analytical methods have been improved significantly in the last ten years, especially since the development of electron capture gas chromatographic tech- niques. It has become possible to detect much lower quanti- ties of pesticide chemicals, and many milk samples have been found to contain residues at levels greater than "zero." Although the tolerance levels were maintained at "zero," a series of actionable levels for various pesticide residues were announced by the FDA during late 1963 to 1964; for ex- Ennple, this actionable level for combined DDT and its deriv- éatives was 1.25 ppm on a fat basis or 0.05 ppm in whole milk, and for aldrin and dieldrin was 0.25 ppm on a fat basis or 0.01 ppm in whole milk (Henderson, 1965). In 1967, the tolerance for DDT, DDE and DDD residues combined was established at the previous actionable level (Duggan, 1967; Laben, 1968). A large number of samples of milk and dairy products from different geographical areas in the United States have been examined by the FDA. Duggan (1967) reported that 57% of the samples contained residues. Most frequently found chemicals were DDE, dieldrin, DDT, heptachlor epoxide, DDD, BHC, lindane, aldrin, heptachlor and methoxychlor. They ac- counted for 99.3% of the residues. The organophosphate com- pounds are rarely found in milk, since they are readily hydrolyzed by the rumen fluid of animals (Cook, 1957; Marth and Ellickson, 1959b). The average concentration of each kind of chlorinated residue in milk and dairy products was about one-tenth the tolerance or actionable level. Dieldrin content in dairy products was approximately 0.042 ppm on a fat basis. Sources of Contamination and Excretion of ResiduesTTn Milk The major source of pesticide residues in milk has been found to be the contaminated feed consumed by dairy cows. The feed may become contaminated in several ways, such as the direct application of chemicals to the forage crops, the drift of pesticides from adjoining fields, or the crops may have been grown in contaminated soil (Henderson, 1965; Kiermeier, 1968). Numerous reports have indicated the transfer of chlorinated insecticides from feed into milk of dairy cows. Wilson gt gt. (1946) detected DDT in milk from cows fed DDT— treated pea vine silage. Mann gt gt. (1950) further indi- cated that DDT was found in various products made from con- taminated milk. App gt gt. (1956) observed the presence of dieldrin in the milk from cows fed on dieldrin treated pas- ture. Gannon and associates (1959a, 1959b, 1960) and Marth and Ellickson (1959b) conducted several experiments by feed- ing various chlorinated pesticides including dieldrin, DDT, methoxychlor and heptachlor to cows at different dosage levels for varied periods of time. These workers noted that the rate of appearance, the rate of accumulation from con- tinuous feeding and the concentration of residues in milk varied among pesticide chemicals. They (Gannon and Decker, 1960) found that heptachlor in feed was converted by dairy cows to heptachlor epoxide in milk. The conversion of aldrin to dieldrin in dairy cows was reported by Ivey gt gt. (1961). The transfer of endrin, Kelthane and chlordane from feed to milk were demonstrated by Johnsen gt gt, (1961), Zweig gt gt. (1963) and Westlake gt gt, (1963), respectively. Williams gt gt, (1964) found that DDT, DDD, and DDE were present in milk from cows fed rations containing DDT. Trhe concentrations of DDT and DDD in milk were increased according to the feeding levels of DDT, but DDE concentra- tion apparently was not affected. Dairy cows may become contaminated through the di- rect spray of chemicals to the animals or in their barns. It has been shown that some pesticide compounds can enter into cows through the skin. Carter and Mann (1949) detected DDT in milk from a cow sprayed with DDT. Marth and Ellickson (1959a) indicated that highest levels of DDT in milk generally appeared short- ly after spraying, in some instances DDT appeared in milk up to four months, whereas methoxychlor declined from milk quite rapidly. Radeleff gt gt. (1960) found dieldrin in milk following the spray treatment of Cows with dieldrin. Claborn gt gt. (1960) further reported that DDT and dieldrin were found in higher levels than methoxychlor and toxaphene in milk two days after the cows had been sprayed with these chemicals. Other studies involving the feeding to or the spray- ing of pesticides onto dairy cows have been reviewed by Marth (1962) and Henderson (1965). The chlorinated hydro- carbon pesticides were also found to be deposited in body fat of cows or other animals through a spray treatment or after oral ingestion (Gannon gt gt., 1959a; Radeleff gt gt., 1960; Claborn gt gt., 1960). A close relationship of the residues in the body fat and in the milk fat has been ob- served. Comparatively more studies have been made for DDT than other residues in cows fed contaminated feed. The con- centration of residue in milk fat as well as in body fat was proportional to the residue level in feed, and to the total intake of this compound by the cows. Laben gt gt. (1966) found that an equilibrium of DDT content in milk was achieved after eighteen to twenty-one weeks of continuous feeding to lactating cows, but there was no sufficient data showing an equilibrium between the total intake and the storage of DDT in body fat of cows (Laben, 1968). Both the feeding and spraying studies demonstrated that once the dairy cows were exposed to chlorinated pesti- cides, the residues in the body tissue could be translocated and secreted into milk for a long period of time. When the source of contamination had been stopped, residues in milk as well as in beef tissue were reduced relatively rapidly in an initial stage and this was followed by a slower reduction period. Moubry gt gt. (1968a) indicated that the reduction rate of total DDT and its analogues in the milk fat was much more rapid as compared to the decline of these compounds in the body fat of cows. Chlorinated hydrocarbon pesticides may also be found in milk when the cows have inhalated these chemicals in the barn after spray treatment (Heineman and Miller, 1961; Witt gt gt., 1966b). However, only traces of residues were at- tributed to this route of contamination. m1. edz 196 res the ina low avo: tent 196E sorp prev foun Cont. samp. 19ast applj Condi In recent years, the incidence of high levels of milk contamination have been reduced effectively through educational programs and legislative regulation (Laben, 1968). Today, it is rare to find contaminated milk as a result of the improper use of pesticides directly on cows, the use of treated seeds as feed or from accidental contam- ination during milking and handling (Fries, 1970). However, low levels of residues in milk are relatively difficult to avoid, due to the fact that lots of feed may have been unin- tentionally contaminated with pesticide chemicals (Laben, 1968; Fries, 1970). Forage crops may become contaminated through the ab- sorption of pesticide residues from the soil which has been previously treated with these compounds. Cole gt gt. (1966) found that more than 50% of the forage and grain samples contained DDT at levels from traces to 0.33 ppm, and those samples were from farms where no DDT had been used for at least three years. Some chlorinated hydrocarbon pesticides applied to soils may persist a number of years under normal conditions (Hardee gt gt., 1964). The uptake and translocation of dieldrin and other pesticide residues by plants have been studied by Mumma gt gt. (1966), Wheeler gt gt. (1967), and Nash (1968). Their results suggested that the transfer of these substances from soil to plant could be the source of contamination in food crops and in the milk from cows which had not been fed with any treated feed. 10 The United States Department of Agriculture recently announced that DDT and other persistent chlorinated pesti- cides will be used only in preventing or controlling dis- eases or other "essential uses" by the end of 1970 (Anony- mous, 1969). There should not be much residue problem in the future. However, because of the heavy use worldwide of these chemicals in the past, and the high stability of these compounds, residues have been and will be found in our en- vironment for a long period of time. Feed and forage crops may still be found to contain these residues as a conse- quence of absorption from soil and water, and some traces of residues in milk apparently will not be totally prevented for many years. Animal Decontamination Some of the recent studies were focused on the pos— sibility to decontaminate dairy cows after they had been exposed to pesticide chemicals. Once the contaminated feed is removed, pesticide residues in milk and body fat of cows will decrease gradually as described above. This natural decontamination process is usually slow (Moubry gt gt., 1968b). Since most of the chlorinated hydrocarbons are de- ‘posited in fatty tissues of cows, it was thought that alter- éation of body fat mobilization or storage might be effective :in accelerating the secretion of residues from body fat. IFries gt gt. (1969) found that the rates of elimination of 11 DDT and DDE from milk and dairy cows were not related to the loss of body weight. Only the concentration and total amount of DDD in milk were affected by the body energy loss. However, Miller (1967) reported that when dairy cows were on a low energy diet, the rate of decline for DDT in milk was increased. Feeding thyroprotein did not significantly affect the concentration of DDT or dieldrin in milk fat from con- taminated cows (Braund gt gt,, 1966; Miller, 1967; Stull gt gt., 1968). Later, Braund gt gt, (1969) stated that varying dietary energy did not speed the rate of dieldrin reduction, but feeding thyroprotein decreased the time for dieldrin to reach 1.0 ppm in milk fat. They also showed that the con- centration of dieldrin in whole milk was not affected when the milk fat was decreased by feeding pelleted hay to cows. Another approach, which has been suggested, is the use of drugs to induce liver microsomal enzymes to acceler— ate the detoxification mechanism. Street gt gt. (1966) found that heptobarbital was effective in reducing the stor- age of dieldrin in rats. Oral administration of phenobarbi— tal was also effective in reducing the concentration of DDD in milk of dairy cows, and increased DDD excretion in the urine of young bulls (Alary gt gt., 1968). Fries (1970) thought that only DDD retention was affected by hexobarbital .in rats, and by either hexobarbital or pentobarbital in (:ows, whereas DDT and DDE were essentially unchanged. 12 A report by Wilson (1970) showed that phenobarbital fed to cows, either during the contamination period or in the postcontamination period, did significantly reduce the levels of dieldrin in both the body fat and milk fat of dairy cows, although the fate of possible metabolites was not investigated. Recycling of dieldrin in ruminants was suggested by Cook gt gt. (1967). Wilson gt gt. (1968) and Wilson (1970) demonstrated that the administration of activated charcoal to sheep, goats or cows, which were either ingesting or had ingested dieldrin, would increase the excretion of dieldrin in feces. The efficiency of removal was greater when char— coal and dieldrin were fed at about the same time. Somewhat contrarily, Fries (1970) reported that activated charcoal by itself was not effective in removal of either stored DDT in rats, or the concentrations of dieldrin, DDT and DDD in the milk from previously contaminated cows. Metabolic Transformation of Dieldrin The literature offers abundant evidence of the con- version of DDT to DDE, DDD or DDA in various microorganisms, .rats, ruminants, humans, and plants. It has also been shown ‘that aldrin and heptachlor are readily metabolized to di- eldrin and heptachlor epoxide in biological systems, re— Spectively. Excellent studies on the metabolism of various PeSticides have been reviewed by Hayes (1965), Lykken and 5+ 13 Casida (1969), and Brooks (1969). Only those reports re- garding the biotransformation of dieldrin will be presented. Ludwig and associates (1964) fed rats with labeled aldrin at a 0.2 ppm level in the daily diet, and found that after eight weeks, the total excretion of aldrin and di- eldrin in the urine and feces was equivalent to the daily intake, and almost all of the aldrin was eliminated from the body twelve weeks after the feeding was stopped. However, Quaife gt gt, (1967) questioned the validity of the equil- ibrium since only one feeding level was tested, the rats used had grown rapidly, and the equilibrium was only main- tained for a short time. Datta and co-workers (1965) observed an unusual toxic and alkali—unstable metabolite in the urine of male rats fed dieldrin and aldrin in the diet. The chemical structure and properties of this metabolite have been evalu- ated by Damico gt gt. (1968) and Klein gt gt. (1968). They found that aldrin and dieldrin yielded almost the same metabolite, and which was two to three times more toxic to flies than were aldrin and dieldrin. Structural determina- tions showed that this metabolite was an oxidized, dechlo- rinated compound that no longer has an intact dieldrin sys- ‘tem. The chemical name suggested for this compound is listed in the Appendix, referred to as dieldrin metabolite I. At about the same time, Richardson gt gt. (1968) also identified a metabolite from rat urine, and this metabolite Showed similar properties and structural groups as dieldrin meta diff olit from derf this by ice III as en ro 0X 801 me 14 metabolite I, although Richardson gt gt. proposed a somewhat different chemical name for their compound (dieldrin metab- olite 1'). Richardson gt gt. (1968) also isolated a metabolite from rat feces and identified it as the 6-chloro-monohydroxy derivative of dieldrin. The chemical name suggested for this fecal metabolite is referred to as dieldrin metabolite II in the Appendix. Korte and Arent (1965) found a different metabolite in the urine of rabbits after oral administration of Cl4 labeled dieldrin. They identified this compound as trans-6, 7-hydroxydihydroaldrin, i.e., a 6-chloro, trans-diol formed by hydrolytic cleavage of the epoxide ring. The full chem- ical name is given in the Appendix as dieldrin metabolite III. The toxicity of this compound to mammals was reported as 1/12 to 1/16 that of dieldrin. Matthews and Matsumura (1969) reported that, from tg_zttg and tg_!tttg studies, dieldrin was attacked by three enzyme systems in rat, and the most important degradation route being a glucuronic conjugation system upon one of the caxidation products of dieldrin in the liver. They found .some other metabolites in addition to the three metabolites Inentioned above. In an experiment with Cl4-dieldrin, Hedde gt gt. (1970) determined that about 80% of the radioactive dose was deposited in the carcass of sheep, 9-10% recovered from feces, and 5-9% in the urine over a period of three to four (1511 me tit fc SC W1 15 days. The radioactivity in the urine was contained in six metabolites, four of which were hexane-extractable. Two of these four metabolites were further identified by Feil gt _t. (1970): one appeared to be the trans-6, 7-dihydroxydi— hydroaldrin (dieldrin metabolite III), and another one ap- peared to be identical to dieldrin metabolite II, though a new chemical name was given as dieldrin metabolite II'. Effects of Chlorinated Pesticides in Biological Systems The use of pesticide chemicals in agricultural and food production may affect the yield and the composition of some crops, or the flavor of some fruits and vegetables. The literature in this aspect has been reviewed by Marth (1965). A recent report showed that treatment with pesti- cides had no significant effects on the fatty acid, oil con— tent, and protein content in soybean seeds (Johnson and Jullum, 1969). However, DDT apparently affected the photo- synthesis of some marine phytoplankton (Wurster, 1968) and the nervous system of brook trout (Anderson and Peterson, 1969). Significant reports have been published concerning 'the effects of pesticides, especially DDT and its deriva- ‘tives, on birds and rats. Many studies pointed out that DDT and its metabolites had an effect on the thickness of egg Shell and were responsible for the reduction of population 0f several species of wild birds (Bitman gt gt,, 1969; Wurster and Wingate, 1968). It has been suggested that DDT l6 and dieldrin affect the deposition of calcium to eggshell, by inducing some hepatic microsomal enzymes which affect the metabolism of steroid hormones. Some chlorinated pesticides were found to have an inhibitory effect on enzyme activities, such as that DDT in- hibited the ATP-ase in rat brain (Matsumura gt gt., 1969), and chlordane, aldrin, dieldrin and DDT inhibited ATP—ase functions of rabbit brain (Koch, 1969). Casterline gt gt. (1969) showed that chlordane and some other pesticides had an effect on the esterase activities in serum and tissues of rats fed variable casein diets. Long term administration of DDT to mice has caused liver lesions, hypertrophy and hyperplasia. When mice had been receiving small doses of DDT daily over five genera- tions, the later generations showed more serious tumors and leukemia than the earlier generations (Tarjan and Kemény, 1969). Phillips (1963) reported that DDT affected the utilization of vitamin A and carotene in rat liver. Diel- drin produced a similar effect in rats on a low protein diet (Lee gt gt., 1964). Tinsley (1969) found that rats receiv- ing DDT with a-protein diet had depressed growth and vitamin .A.level in the liver. Rats may become more susceptible to the lethal effect of single toxic oral doses of DDT or other Chlorinated pesticides on a deficient protein diet (Boyd and Stefec, 1969). 17 Tinsley (1966) reported that dieldrin accentuated an essential fatty acid stress and the transformation of lino- leate to arachidonate in rat livers. On the other hand, when rats were raised on a diet deficient in essential fatty acids, a reduced growth rate of female rats, and an in- creased level of polyunsaturated fatty acid together with stearate in the liver lipids of rats were observed (Tinsley and Lowry, 1969). Chlorinated Hydrocarbon Residues in Human Tissues in Relation to Public Health Because of the wide incidence of some persistent pesticide chemicals in food, it is of interest to examine the residue levels in human tissues. There have been a num~ ber of publications providing the results of these surveys carried out in several countries and in various areas of the United States (Hoffman gt gt,, 1964; Dale gt gt,, 1965; Quinby gt gt,, 1965; Brown, 1967; Abbott gt gt., 1968). Generally, adipose tissue or body fat was examined. The most commonly found residues were DDT, DDE, dieldrin, lindane, heptachlor epoxide, DDD and BHC. The concentration of dieldrin in body fat is close to 0.2 ppm throughout the tuorld, while DDT and its metabolites are found at levels from 1 ppm to 30 ppm. The general population in the United States carries approximate 9 to 12 ppm of total DDT and its derivatives (Brown, 1967; Abbott gt gt. 1968). 18 Casarett gt gt. (1968) and Radomski gt gt. (1968) further analyzed the distribution of residues in various human tissues. Adipose fat generally contained the highest levels of residues on a whole tissue basis, whereas bone marrow and liver followed by adipose tissue contained higher concentrations of residues on a fat basis. Brain lipids contained the lowest amounts of residues (Casarett gt gt., 1968). Whether the trace amounts of pesticide residues con- sumed daily over long periods of time will have an adverse effect on human health has developed great interest, espe- cially in recent years. Early in 1956, Hayes and associates performed some experiments with human volunteers who were given daily doses of DDT at levels up to 200 times more than ingested by the general population. It was reported that the storage of DDT in adipose fat was proportional to the total intake, but an equilibrium was achieved in about one year. No illness was observed among the volunteers during the feeding period. Although from this study and other surveys on the general population in different years it appeared that an equilibri- um between the intake and storage of DDT was possible, Quaife gt gt. (1967) questioned the validity of those ex- periments in the methods of sampling the general population and pointed out that the results of analyses by the spectro- Photometric method were not comparable to those by gas- liquid chromatography. 19 A series of studies of dieldrin in relation to human health have been carried out by Hunter and his co-workers in England (Hunter gt gt., 1967; Hunter and Robinson, 1967, 1968; Hunter gt gt., 1969). They reported that the health of occupationally exposed workmen and the volunteers was not affected by body burdens of dieldrin 8 to 60 times greater than the general population. They also showed that there was a correlation between concentration of dieldrin in the adipose tissue and blood, while both were proportional to the daily dosage and total intake, and suggested that the concentration in either tissue might be predicted from the concentration found in the other. This correlation for the concentration of dieldrin in different kinds of samples from the same person was also demonstrated by Hayes and Curley (1968) by the analyses of plasma, fat and urine samples of workers exposed to aldrin, dieldrin and endrin. The storage of dieldrin in human subjects during the surgical stress or in periods of complete fasting was exam- ined by Hunter and Robinson (1968). They found that there was no increase in the concentration of dieldrin in the whole blood in the above cited conditions and since the con- centrations in adipose tissue and blood were related, it was apparent that the total amount of dieldrin in adipose fat would not increase with starvation. As additional information on studies of human vol- lunteers ingesting dieldrin for two years, Hunter and asso- Ciates (1969) pointed out that a steady state of dieldrin 20 concentration in the human body had been achieved. The average biological half-life was about 369 days. When the feeding period had ceased, the concentration of dieldrin in blood decreased gradually. Although several studies pointed out that the pres- ent exposure level for the general population should not have any hazard to health, there is still a lack of investi- gations concerning the effect of pesticide residues on en- zyme activities, hormone functions, and genetic systems in human subjects. It is known that chlorinated hydrocarbons do produce significant histopathological symptoms in liver of animals on a high dosage over a long period of time (Durham, 1963). From limited statistical analyses, Deichmann and Radomski (1968) showed that there was a correlation between the high residues of DDT in the human body and the frequency of human death from various disorders, including liver cancers. Extraction and Determination of Chlorinated Hydrocarbon Pesticide Residues in Milk When DDT was first determined in milk more than twenty years ago, a colorimetric method or a total organochlo- rine estimation method was applied (Wilson gt gt., 1946; Car- ter, 1947). Since 1960, the analytical methods have been im- proved dramatically. Some of the recent reviews of modern methods for pesticide analysis have been prepared by Lisk (1966), Williams and Cook (1967), and Thornburg and Beckman (1969). Official general procedures have been compiled in 21 "Pesticide Analytical Manual," by Food and Drug Administra- tion (FDA, 1968). The analysis of chlorinated hydrocarbon residues in milk usually consists of three steps: the extraction of these substances from samples, removal of co-extracted in- terferring materials, and the determination of residues in the purified extract (Moffitt, 1963; Henderson, 1965; Kroger, 1966; FDA, 1968). The first two steps may also be combined as a single step (Langlois gt gt., 1964a; Stemp and Liska, 1965; Crosby and Archer, 1966). Since these chlorinated pesticides are very hydro- phobic in nature, the extraction is commonly accomplished by employing hydrocarbon solvents. It is also because of the hydrophobic properties that they are readily deposited in the fatty portion of biological materials, such as the adi- pose tissue of animals and the butterfat of milk. In the case of milk, it is known that butterfat is enclosed within the fat globule membrane, which is composed of a lipoprotein complex, together with other constituents (King, 1955). In order to recover the residues quantitatively from milk, the globule membrane should be disrupted to release the butter— fat. Various reagents and a mechanical churning process can be applied to break up the membrane of fat globules in milk. Reagents suitable for these purposes are: (a) strong acid, such as sulfuric acid used in Babcock fat determination, (b) ethyl alcohol, as used in the Mojonnier method to extract fat from milk, (c) surface active agents, such as Triton 22 X-100, and (d) alkaline solution, to hydrolyze the membrane constituents as well as the lipid material and subsequently to make the residues available for solvent extraction. The first three techniques all release the hydro- phobic pesticides with a relatively large portion of butter- fat, which, as well as other edible fat, oil or butter sam- ples, will be “cleaned up" in the second step. Two sub-steps may be involved to accomplish this purpose: solvent parti- tion and column absorption procedures. Usually, the resi- dues are isolated from the fat by partitioning of acetoni- trile (Mills, 1961; modified by Furman and Fehringer, 1967), dimethyl formamide (deFaubert Maunder gt gt., 1964), or dimethyl sulfoxide (Haenni gt_gt., 1962; Eidelman, 1962, 1963, 1967) with either hexane or petroleum ether. Chlorin- ated hydrocarbons are more soluble in acetonitrile (or dimethyl formamide, and dimethyl sulfoxide) than in petro- leum ether (or hexane), whereas butterfat has a reverse solubility. Consequently, the acetonitrile layer contains the pesticide residues with only small amounts of fat mate— rials. This extract is then cleaned up on an absorption column, such as Florisil (Mills, 1961; Wood, 1966), alumina (deFaubert Maunder gt 1., 1964), silica gel (Kadoum, 1968), carbon-celite (Moats, 1964) magnesium oxide (FDA, 1968), Florisil-celite (Shell Chemical Company, 1964), and silicic acid-celite (Samuel, 1966) using various eluting solvents or the combined solvents including petroleum ether, ethyl ether, hexane, methyl chloride, acetonitrile, and dimethyl 23 sulfoxide. Some pigments and traces of lipids will be ab- sorbed on the column, while the chlorinated hydrocarbon in- secticides are eluted in one or more fractions. Other procedures are available to replace the sol- vent partition step: for example, saponification (O'Donnel gt gt,, 1954; Crosby and Archer, 1966), cold-bath precipi- tation (McKinley and Savary, 1962; McCully and McKinley, 1964), sweep co-distillation (Storherr gt gt., 1967; Malone and Burke, 1969), molecular distillation (Bills and Sloan, 1967), deactivated Florisil column absorption (Giuffrida gt gt., 1966), and celite absorption techniques (Singh and Lanthier, 1968). The extracts from these procedures are also subjected to the additional cleanup step using an ab— sorption column as mentioned above. Procedures are also available to clean up butterfat or vegetable oil samples in one step, such as column chroma- tography on carbon-celite or partially deactivated Florisil columns (Moats, 1964; Moats and Kotula, 1966), and channel layer chromatography (Matherne and Bathalter, 1966; Hether- ington and Parouchais, 1970). Even simpler methods are reported to combine the ex- traction and cleanup in a single step (Langlois gt gt., 1964a; Stemp and Liska, 1965; Crosby and Archer, 1966). These techniques will be described later. In all cases, the final step of analysis is the de- termination of residue content in the purified extract, qualitatively or quantitatively or both, by one or more of 24 the following methods: fly bioassay, total organic chloride estimation (Henderson, 1963), paper chromatography (Mills, 1959; Moffitt, gt gt., 1961), colorimetry for dieldrin (Cueto, 1960) or for DDT (Blinn and Gunther, 1963), thin- layer chromatography (TLC) (Kovacs, 1963; Abbott and Thomson, 1965; Moats, 1966), reversed phase paper chromatography (Wang and Chou, 1968), flame-spectrophotometry (Herrmann and Gutsche, 1969), and gas-liquid chromatography (GLC) with either electron capture detector (Burke and Giuffrida, 1964; Burke, 1965) or microcoulometric titrating system (Bosin, 1963; Burchfield and Wheeler, 1966; Giuffrida and Ives, 1969). Residues can be further identified by infrared (Blinn, 1965) or mass spectroscopy (Mumma and Kaniner, 1965). The most widely used methods probably are electron capture GLC, and TLC. Excellent reviews or volumes for the analysis of pesticide residues by GLC and TLC have been offered by Gunther (1964), Westlake and Gunther (1967), Zweig (1967), Gudzinowicz (1967), Wise (1967), and McCully (1969). The literature dealing with methodology and modifi- cations of techniques is massive, only those methods that seem pertinent to the extraction procedure will be described briefly in this review. It should be kept in mind that each extraction step reported may be followed by different clean- up procedures. The final determinations are usually carried out by GLC. The accuracy or the recovery of the residues by 25 each method may be affected by the cleanup step or GLC per- formance in addition to the extraction technique. Acid Hydrolysis Langlois gt gt. (1963) used the Babcock method to extract butterfat from milk, and then analyzed the residue concentrations in this butterfat. Recoveries for added DDT were about 90%, but for dieldrin and lindane were only 70— 75% and 63-67%, respectively. Endrin was lost in this pro- cess. Stanley and LeFavoure (1965) digested animal tissue with a perchloric-acetic acid mixture to destroy the cellu- lar structure and liberate the pesticide-containing fat, which was further cleaned up and analyzed. This method was reported to be suitable for DDT, DDD, lindane, and hepta- chlor, but not for heptachlor epoxide, aldrin, dieldrin and endrin. Alkaline Hydrolysis Schafer gt gt. (1963) described a procedure to de- termine DDT and DDE in milk by alcoholic KOH hydrolysis. Milk fat as well as the fat globule membrane was decomposed and the stable residues were extracted by hexane. In this process, DDT was converted to DDE. Crosby and Archer (1966) and Richardson gt gt. (1967) reported that this method could be applied to the detection of heptachlor, heptachlor epox- ide, dieldrin and endrin as well as DDT+DDE. These workers I ”manus- -..~9'“ vb used of 1C 65-8C Food ceduz a ste dense dielc the c magnt from (alk sodi mixt 80-1 DDD aPhe from Gott One SOdi fere 26 used a rather mild hydrolysis; i.e., heating the mixture of 10 m1 milk and 20 m1 of 25% alcoholic KOH solution at 65-80°C for 15 min. Shell Chemical Company (1964) and Food and Drug Administration (FDA) (1968) recommended a pro- cedure by which the alkaline hydrolysis was carried out on a steam bath and refluxed for 60 or 30 min under an air con- denser for the alkaline stable residues, such as aldrin, dieldrin, and endrin (Shell Chemical Company, 1964), or for the oily materials which were not removed by Florisil and magnesia columns (FDA, 1968). Surfactant Reaction Lampert (1964) suggested the recovery of milk fat from whole milk by using a surface active agent—Triton X-lOO (alkyl-phenoxy polyethoxyethanol) with the aid of urea, sodium tetraphosphate, isopropyl alcohol and water. The mixture of milk and reagents was heated in a water bath at 80-100°C for 10 min. He reported that added DDT, DDE, DDD, aldrin, dieldrin, heptachlor epoxide, lindane, and tox- aphene were recovered at levels comparable to the ether ex- traction techniques. Kroger and Patton (1967) reported that the results from two methods, one was referred to the modified Roese- Gottlieb method involving the use of ethanol and ethers; one was the simplified detergent method using Triton X-100, sodium tetraphosphate and water, were not significantly dif- ferent from each other on a fat basis for the analysis of 27 dieldrin and heptachlor epoxide in naturally contaminated milk. However, Moubry gt gt. (1967) found that the simpli- fied detergent method generally extracted 5-10% less of DDT, DDE, dieldrin and BHC than the extraction by the Mills meth- od (similar to the modified Roese-Gottlieb method) on a whole milk basis. For omental fat samples, Li gt gt. (1968) reported that Triton X-100 was essentially ineffective in aiding the extraction of various chlorinated residues by petroleum ether, while other surfactants, such as Span 80 and crude lecithin, did improve the extraction efficiency for dieldrin and heptachlor epoxide as compared to the extraction with only petroleum ether. Extraction Involving the Use of Polar Solvent Ethyl alcohol has been one of the most widely used reagents in the extraction of organochlorine pesticide resi- dues from fatty foods. The official method prepared by the Food and Drug Administration (FDA, 1968; available since 1962) describes a procedure employing the use of alcohol and potassium oxalate to break up the fat globule membrane and to denature proteins. The released fat, containing pesti- cide residues, is then extracted by ethyl ether and petro- leum ether. This technique is essentially adapted from the Mills method (Mills, 1961), which was a modified Mojonnier fat extraction method. . . .5! I. 5.7-..- ‘ a] m: 28 Shell Chemical Company (1964) recommended the use of a mixture of hexane and ethyl ether for the extraction fol- lowing the alcohol disruption process. Other solvent systems for the extraction of chlorin- ated hydrocarbon pesticide residues from milk have been sug- gested by Giuffrida gt gt, (1966) and deFaubert Maunder gt gt. (1964) using acetone-petroleum ether and acetone-hexane, respectively. Recoveries of various pesticides were report- ed generally satisfactory. Onley (1964) reported that pesticide residues could be extracted directly from the fluid milk instead of butter- fat. According to his procedure, milk was blended with an- hydrous sodium sulfate and a mixture of solvents consisting of acetonitrile, ethyl ether, dioxane and acetone. The ex- tract was chilled in ice, filtered, and diluted with water in the presence of petroleum ether. The petroleum ether solution was further cleaned up on an activated Florisil column, where dioxane was included in the second eluting mixture. The residues in animal tissues are usually extracted by blending the sample with anhydrous sodium sulfate and an organic solvent or mixture of solvents. Taylor gt gt. (1964) used acetone for this purpose. An acetone-acetoni- trile system was used by Saschenbrecker and Ecobichon (1967). Li gt gt. (1968) found that chloroform-methanol extracted more lindane and DDT from animal fat than the extraction 29 with petroleum ether, but this was not true for heptachlor epoxide. The use of Span 80 with petroleum ether provided a better extraction for heptachlor epoxide, a slightly better extraction for DDT, and a similar recovery for lindane as compared to the extraction by chloroform—methanol. Onley and Bertuzzi (1966) developed a method using a mixture of acetone, methyl cellosolve, and formamide to ex- tract the pesticide residues, and to coagulate the fatty materials with calcium stearate. Wood (1969) reported that animal tissues, such as liver and eggs, could be mixed with celite and dimethyl sulphoxide, and then the mixture was packed into a column. Chlorinated residues were eluted from the column with dimethyl sulphoxide. One Step Extraction and Cleanup Langlois and associates (1964a) ground dairy prod- ucts with Florisil to produce a free-flowing powder and which was added to the top of a partially deactivated Florisil column. Organochlorine pesticide residues were eluted with a solvent mixture of 20% methylene chloride in petroleum ether. Recoveries of added DDT, lindane, hepta— chlor, dieldrin, and endrin were 90-96%. Stemp and Liska (1965) modified the above procedure and reported that milk was mixed with methylene chloride- petroleum ether solvent and Florisil, and the liquid was then decanted through a Florisil column. The slurrying technique was repeated several times and the residues were 30 obtained in the elutant from the column. They further re- ported that an aliquot of milk could be directly pipetted onto the top of a Florisil column, and pesticides were eluted with the solvent mixture as above. Good recoveries were observed for chlorinated hydrocarbon residues. The hexane extract from the alkaline hydrolysis pro- cedure was sometimes suitable for electron capture GLC anal- ysis without cleanup (Crosby and Archer, 1966; Richardson gt gt., 1967), especially when the residue levels were suf- ficiently high. This procedure may be considered as an one- step method. The principle of these rapid methods also involves the disruption of protein and fat globule membrane struc- tures in milk by chemical or physical forces, or both forces, and the extraction of pesticide residues was performed at the same time. Recently, Lawrence and Burke (1969) presented a com- parison of various extraction and cleanup procedures used in ten methods for the analysis of chlorinated hydrocarbon re- sidues in milk. All the final determinations were carried out by electron capture GLC with the same conditions. In general, they found that Lampert's method (Lam- pert, 1964, modified by California State Department of Agri- culture) showed the lowest recoveries for the added DDD and the naturally incorporated heptachlor epoxide in milk. Moubry gt gt. (1967) method was better, but was still lower in recoveries. The method of deFaubert Maunder gt gt. . 93:5- "‘. "" (EEL’ 31 (1964) also showed lower values for both residues. It was pointed out by Lawrence and Burke that the weakness of this method appeared to be associated with the dimethyl forma- mide-hexane partitioning technique rather than the extrac- tion procedure. Highest values were obtained by Onley's method (1964). However, Lawrence and Burke questioned the accuracy of the empirical corrections which were used for five variables in Onley's method. Comparable average values, but with varied ranges, were obtained by other methods, which included the method of (a) FDA (1968, using either single extraction or triple extraction techniques), (b) Samuel (1966), (c) Sweep co-distillation (Storherr gt gt., 1967) following the single extraction of fat by FDA method, (d) Langlois gt gt. (1964a), (e) deFaubert Maunder gt gt, (1964), (f) Giuffrida gt gt. (1966). Distribution of Residues in Milk There is no difficulty in realizing that most of the chlorinated hydrocarbon chemicals are found in the lipid portion of milk, since these substances are highly soluble in fat. Mann gt gt. (1950) reported that the concentration of DDT was much higher in the cream and butter than in the Skim milk and whole milk, on a weight basis, providing all fractions were made from the same source of milk obtained from cows receiving DDT in their ration. 32 This pattern of distribution was also proved by Langlois gt gt. (1964b, 1965) for other kinds of chlorinated pesticides. They found that, following separation of pas- teurized milk which had DDT and dieldrin added or pasteur— ized milk from cows fed lindane or endrin, residues were found almost entirely in the cream and butter fractions and the skim milk was essentially free of residues. The amount of residue in each fraction was generally proportional to the fat content in that fraction. Therefore, on a weight basis, butter contained the greatest quantity of residues, followed by cream, whole milk, buttermilk, and skim milk in decreasing order. However, if expressed on a fat basis, butter showed lesser residues of DDT, lindane and dieldrin, but not endrin, than cream and raw milk. Buttermilk con- tained higher levels of the first three residues, but a low— er level of endrin than other fractions. The residue levels in skim milk were reported to be too low to be determined, either on a weight basis or on a fat basis. Beroza and Bowman (1966) demonstrated that both or- ganochlorine (such as aldrin and lindane) and organophos- phate (ethion and imidan) pesticides were essentially re- covered in the cream phase, and only a small percentage remained in the aqueous phase following centrifugation of raw whole milk which was spiked with the pesticides. They suggested that most of the organic pesticides were adsorbed onto the cream phase in milk, for both the non-polar com— pounds (aldrin, ethion, and lindane) and relatively polar 33 compounds (such as imidan). However, the nontpolar com- pounds were poorly recovered from milk by extraction.with hexane and ethyl ether, while the polar compounds were readily extracted by this solvent mixture. Recently, Hugunin and Bradley (1969) reported that skim milk and buttermilk had higher concentrations of DDT and dieldrin per gram fat than the raw whole milk, but in two cases involving Kelthane, the residue level in butter- milk was lower than that in raw whole milk. A decreasing order of the concentration of dieldrin and Kelthane (one case) was observed from a high density buttermilk pellet, to a buttermilk serum phase, to the butteroil. They drew the conclusion that organochlorine pesticide was associated with phospholipid in the milk fat. From the same laboratory, the results of Li §_a_l. (1970) showed that skim milk contained higher concentrations of several chlorinated pesticides, but not of dieldrin, than other fractions which included whole milk, butter, cream, and buttermilk, all expressed on a fat basis. The residue levels of lindane, heptachlor epoxide, and in some cases of DDT and its derivatives were higher in buttermilk than those in raw whole milk. However, buttermilk contained the same concentration of chlordane, as well as in some cases of DDT and its analogs, and a lower value of dicofol (Kelthane) as compared to the content in raw whole milk on a fat basis. Moreover, the concentration of dieldrin was essentially not different among various fractions in all cases. 34 Processing Techniques and Residues in Milk It has been the concern of many researchers as well as food processors whether the presence of organochlorine pesticide chemicals in milk will affect routine dairy pro— cessing, or that some processing techniques will effectively reduce the residue levels in milk and milk products. The effect of residues on dairy processing has been studied by several workers regarding the effects on some microorganisms used in manufacturing of dairy products. Kim and Harmon (1968) and Kim (1969) reported that no effect of methoxychlor, dieldrin, heptachlor and malathion was ob- served on the growth and lactic fermentation ability of lac- tic starter cultures which have been usually used in pro- ducing sour cream, yoghurt, cultured buttermilk and cheese. However, Bradley and Li (1968) stated that dieldrin slightly reduced the acid production by a lactic acid producing or- ganism in making Cheddar cheese. Li gt gt. (1970) again found that most of the pesticides studied, such as dieldrin, heptachlor, and chlordane showed some bacteriostatic or bactericidal action against starter microorganisms in the manufacturing of Cheddar cheese. Comparatively more studies have been made concerning the effects of processing on the residues in milk. Mann gt gt. (1950) demonstrated that pasteurization had no signifi- cant effect on DDT levels in dairy products. Langlois gt gt. (1964b, 1965) and Liska (1968) reported that the concen- trations of various chlorinated residues remained fairly cor Cl”; za1 Prc ca1 of du 0f 35 constant on a fat basis during the manufacture of butter, cheese, and ice cream throughout the processes of pasteuri— zation, homogenization, churning, fermentation and storage. Processing techniques involving high temperature treatments caused some reduction of certain residues. The manufacture of dry whole milk powder either by spray drying or roller drying significantly decreased the concentrations of all residues except methoxychlor, while condensing affected the levels of dieldrin and heptachlor epoxide, and slightly affected telodrin, methoxychlor, chlordane, endosulfan and endosulfan sulfate, but not DDT, lindane and heptachlor. Sterilization was fairly effective, except for lindane, hep- tachlor, and total DDT-like substances. However, a change of DDT to DDE and DDD was observed in some instances. Li gt gt. (1970) manufactured several products in- cluding pasteurized whole milk, 30% cream, butter, cheese, and spray-dried, condensed or sterilized condensed whole milk from the milk of cows fed insecticides. They concluded that, in general, most of the pesticides studied were very stable under various processing conditions. However, re- duced concentrations of dieldrin, lindane, and chlordane were found in spray-dried products,of dicofol in sterilized condensed whole milk, and of DDE in stored products. Di- eldrin and toxaphene showed a slightly higher concentration during storage. Molecular distillation at 200°C and under a pressure 0f 5 x 10-4 torr (1 torr = 1 mm Hg) was found very efficient in remove aldrin, I and Sloal 1 drying 0: have som« ated hyd; that frel ization ' of dield. other ha; ture (Up l95°C anl levels. Unaccept ized pro the loss VaCuum p lindane, had Some 3180 r8n( time (Li MODterey DDD Was 1 36 in removal of added lindane, heptachlor, heptachlor epoxide, aldrin, DDT, DDD and DDE in butteroil as reported by Bills and Sloan (1967). Kiermeier gt gt. (1967) pointed out that freeze- 3 drying of milk at a reduced pressure of 1 x 10- torr might have some reduction effect on the concentration of chlorin- w“?— ated hydrocarbon residues. However, Kroger (1968) found that freeze-drying under 0.1 mm Hg pressure and mild deodor- ization treatments were not effective in reducing the amount 1W7?” ._-.......-- '" of dieldrin and heptachlor epoxide in butteroil. On the other hand, he observed that heating at very high tempera- ture (up to 300°C) as well as steam deodorization at 180- 195°C and 0.01-0.5 mm Hg significantly reduced the residue levels. The high temperature treated butteroils were found unacceptable for further food use, whereas the steam deodor- ized product showed no deleterious side effects, except for the loss of odor and color. Commercial steam distillation- vacuum processing was also fairly effective for reduction of lindane, little effective on dieldrin and DDT and not effec- tive on heptachlor in milk (Ledford gt gt., 1968). Ultraviolet light treatment of milk and butteroil had some effects on reduction of chlorinated residues, but also rendered the milk unacceptable in flavor at the same time (Li and Bradley, 1969). The possible effect of manufacture and storage of Monterey and Cheddar cheese on the content of DDT, DDE and DDD was investigated by Montoure and Muldoon (1968). Their 37 results showed that DDT was not detected in samples of whey at dipping, but higher levels of DDE were found in this phase, and there was a definite change in the ratio of DDE: DDD. Distribution was similar for DDE, DDD and DDT in sam- ples from all other phases, including raw milk, pasteurized milk, whey from pressing, and cheese after pressing. Kim (1969) and Kim and Harmon (1970) reported that lactic cultures did not modify nor degrade the residues of DDT, lindane and aldrin in milk. However, Ledford and Chen (1969) found that some isolated cheese microorganisms de- graded DDT and DDE to DDD in a media of tryptone and yeast extract, and the aerobic growth of geotrichum species caused almost complete disappearance of DDT and DDE. Some other isolated organisms showed no effect. For comparison, it is worthwhile to mention that a significant amount of residues can be removed from fruits and vegetables by washing and peeling, although the follow— ing cooking or canning process has no further effect (Lamb gt gt., 1968; Elkins gt gt., 1968; Farrow gt gt., 1968). Chlorinated residues in crude vegetable oils are removed very efficiently by deodorization, a process usually includ- ed in commercial processing of vegetable oils for human con- sumption (Duggan, 1968; Smith gt gt., 1968). ‘nyuo »' -"‘ ‘ EXPERIMENTAL Materials Chemicals t The major chemicals used in this research and the sources of these chemicals are listed as follows: Kn?” ' ' Dieldrin (99% HEOD, chemical name and structure are shown in the Appendix) was purchased from Applied Science Laboratories, Inc., Ann Arbor, Michigan. Ethyl alcohol was 200 proof and was from Commercial Solvents Corporation, Terre Haute, Indiana. Acetonitrile (CH3CN), hexane (C6H6), petroleum ether, and ethyl ether (CH OCH3) were all of nano- 3 grade and were supplied by Mallinckrodt Chemical Works, St. Louis, Missouri. Acetone was spectrophotometric grade and was also from Mallinckrodt Chemical Works. Distilled water was used throughout the experiments unless otherwise speci- fied. Florisil, 60-100 mesh, was a product of the Floridin Company, packaged by Fisher Scientific Company, Fair Lawn, New Jersey. Celite 545 was manufactured by Johns Manville, and packaged by Fisher Scientific Company. All the minor chemicals used were either of reagent grade or were specified in the text. 38 sources: milk fro Holstein fed wit? and grai Holsteir igan Ste dose of dose da: milk sax determi Consist SamPles eXtract Eldrin tion. ; SamPles 39 Milk The milk used in this research was from three sources: (a) commercial pasteurized homogenized vitamin D milk from a local market, (b) fresh raw whole milk from Holstein cows, Michigan State University Dairy, which was fed with regular rations containing corn silage, alfalfa hay _ .‘3 and grain, (c) naturally contaminated raw whole milk from Holstein cows fed dieldrin in their rations; also from Mich- war” igan State University Dairy. Each cow received an initial dose of 2 g dieldrin (in capsules) followed by a 60 mg dose daily. The feeding periods and the collection dates of milk samples are indicated in Table 1. Methods The methods employed in the present studies for the determination of dieldrin in milk as well as milk fractions consisted of three steps: (a) extraction of dieldrin from samples, (b) cleanup of the extract by removal of co- extracted fat and other substances, (c) determination of di- eldrin by electron capture gas chromatography. Extraction of Dieldrin from Milk Commercial pasteurized homogenized milk was used for the studies on the comparison of extraction methods. Di— eldrin was dissolved in acetone at appropriate concentra- tion. An aliquot of this solution was added to the milk samples and the mixture was allowed to equilibrate for MommdeMm XHHE M0 WQUMD COHUUOHHOU CCU DEM sm300 NHHMD MO WDOHH lwfl DCHDQQW .WUWOD CHHDHUHU N0 UHOUQH m "MHHE UDUHCHEMUCOU %HHQHDUDC M0 WUHSOWII.H @HQWF Ill-4|“ H- Elk-iii] 4O .mzoo soon Eoum xHHE omaoom mo ooumflmcoo mcofiuooaaoo onEMm phase one ocooom oceo .mmo\u:mwoz moon mx\mE H.o no women mm3 ommmoo n .xuflmo mo ucoEuummmo .COmHHS Sax an owofl>onm oHoB coflumEHomcH umcuo one sumo ucmfiommcme mcflpmmw mce .ouon wuflmo >uam~m>flc3 oumum cmmH£0Hz may Eoum oocflmuco mm: xHHZM am\a\e canons» ao\sm\m ms\am\a I- am\em\a ms\oa\ma ow ms\om\o a man ao\m\e ms\ma\mfi cmsoucu ocm as\s~\m ao\am\a ms\HH\NH ms\s\a mm\sm\aa om mo\sm\m m «om m N H sum oumom sme\3oo\ms uumum . omomo nonwooaaoo xfiaz eoHuom n cflueaoflo eoanom Awmmv eaMZmzmo . . mafiomom Canoaoflo mo momoo coauwpomq . u H m 1mm mcflooom m.monEmm xHHE mo mmDMU cofluooaaoo ecu one .msoo mnflco mo mooHH .momoo Canoamflo Mo ouooou o uxHHE ooumcHEmucoo maamusumc mo oousomll.a magma 45 min at miequal milk sam the cont describe Food an Method HSEIFIE a BOO-m oxalate the Cor ethyl a funnel ether 4 to the for at mixtur Phase tracte ether er was 1000-” Chlori and a9 before reWash, 41 45 min at room temperature with intermittent stirring. An equal volume of acetone was added to the control milk samples and these were analyzed in the samemanner as the contaminated samples. Four methods were evaluated and described as below. Food and DrugtAdministration L Method (FDA, 1968; with Slight ModificatiOns) An aliquot of 100 m1 of milk was transferred into a 500—ml separatory funnel. Approximately 1 g'of sodium trwW*’*r“- oxalate (Na oxalate) was added to the separatory funnel and the content was mixed well. One-hundred milliliters of ethyl alcohol (ethanol) were added and again the separatory funnel was shaken thoroughly for 2 min. Fifty m1 of ether and 50 ml of petroleum ether were added successively to the separatory funnel which was shaken thoroughly for at least 1 1/2 min following each addition. The mixture was allowed to separate well and then the aqueous phase was drawn into a second separatory funnel and reex- tracted twice with SO—ml portions of a mixture of ethyl ether and petroleum ether (l:l,v/v). The final aqueous lay- er was discarded and the ether layers were combined in an IOOO-ml separatory funnel containing 600 m1 of 2% sodium chloride (NaCl) solution. The content was mixed cautiously, and again the layers were allowed to separate completely before discarding the aqueous phase. The ether extract was rewashed twice with lOO-ml portions of water which was subseque through cm) usi1 separat ents we steam b of the er was beaker til all contain oven at fat was MO'onni \ method; titY of larger ginal d fat ext adequat fUnnel , Present increase 9iVen as 42 subsequently discarded each time, and was then passed through a column of anhydrous sodium sulfate (Na 804) (2.5 2 cm) using a small amount of petroleum ether to rinse the separatory funnel and the column. All of the ether efflu- ents were collected in a 400-ml beaker and evaporated on a steam bath under a stream of nitrogen. When a large portion of the solvent had been eliminated, the content of the beak- L” er was quantitatively transferred into a tared 30-ml t beaker in which the evaporation of solvent was continued un- r til all of the solvents were driven off. The small beaker E containing fat and pesticide residue was further dried in an oven at 50°C, and then weighed. A portion or all of this fat was used in cleanup procedures for residue analysis. Mojonnier Method (Modified by Walstra and De Graaf, 1962) This method was essentially the same as the FDA method; minor differences involved the use of a small quan- tity of concentrated ammonium hydroxide (NH OH) solution and 4 larger proportions of reagents in the extraction. The ori- ginal design involved the use of a Mojonnier bottle for the fat extraction from 2“) ml of milk. In order to obtain an adequate amount of fat for residue analysis, a separatory funnel and an aliquot of 50 ml of milk were used in the present study. The quantities of other reagents were also increased proportionally. Outlines of these procedures are given as follows: 50 m1 of milk were mixed with 0.75 g of NaCl and 6.25 ml of concentrated NH4OH solution. Fifty 43 ml of ethanol, 125 m1 of ethyl ether and 125 m1 of petroleum ether were used for the first extraction, and 25 ml of ethanol, 75 m1 of ethyl ether and 75 ml petroleum ether were used in the second extraction. The combined ether extract was evaporated on a steam bath and the resulting fat was weighed. Surfactant Method 1' The method reported by Kroger and Patton (1967) was used in this study. The reagent was prepared by dis- ‘Ww‘.u.j~‘ L. .. .h solving 70 g of sodium tetraphosphate and 30 g of Triton X-100 (octyl phenoxy polyethoxyethanol, Sigma Chemical Co., St. Louis, Missouri) in a total of 1,000 ml of water. To 130 m1 of milk in a 250-ml volumetric flask, were added 80 ml of the above reagent. The flask was heated in boiling water bath for 15 min with intermittent swirling every 4-5 min. More reagent was added to bring the liberated fat into the neck of the flask, and the fat was pipetted into a centrifuge tube and spun for l min in a clinical centrifuge. The clear fat obtained was weighed and saved for further uses. Alkaline Hydrolysis Method Two procedures were used: one was the method re- ported by Richardson gt gt. (1967), referred to as the mild alkaline hydrolysis; one was the technique recommended by Shell Chemical Company (1964), referred to as strong alkaline hydrolysis in this study. IL “has; “as: 313;! tion v (KOH) with e aliquc in a v flask nel us flask. which was di evapox trOger hydro] by hee dense; a sepe residt tracts extrac deSer CEGure lipid— 44 Mild alkaline hydrolysis.--Alcoholic alkaline solu— tion was prepared by dissolving 25 g of potassium hydroxide (KOH) in 15 ml of water and the volume was brought to 100 ml with ethanol. Twenty ml of this solution were added to an aliquot of 10 ml of milk in a flask, which was then heated in a water bath at 65°C for 15 min. The content of this flask was quantitatively transferred into a separatory fun— k nel using a total volume of 20 ml of water to rinse the flask. Ten ml of hexane were added to the separatory funnel mw‘qhuam which was thoroughly shaken for 4 min. The aqueous layer was discarded after separation and the solvent extract was evaporated to appropriate volume on a steam bath under ni- trogen. §trong alkaline hydrolysis.--Fifty ml of milk were hydrolyzed with 12.5 g of KOH pellets and 25 m1 of ethanol by heating on a steam bath and refluxing under an air con- denser for one hour. The mixture was then transferred into a separatory funnel using 25 ml water, and the pesticide residues in the aqueous hydrolyzate solution were ex— tracted three times with 40 ml of hexane each time. The extracts were combined and reduced to a suitable volume as described earlier. Both the mild and strong alkaline hydrolysis pro- cedures were also used in the determination of residues in lipid—depleted milk and skim milk fractions. eldrin tracted mm thi fractio tracted resulti the alC. solvent ml port: in a su; transfeJ 2,000 x With a 1 t0 the c with the vent ext the res\ mixed Wi Shaken w ture by extract . and 133581 45 Separation of Milk Fractions by Differential Solvent Extraction Methods Pasteurized homogenized milk containing added di- eldrin was used in this study. The first fraction was ex- tracted by non-polar organic solvents including ethyl ether and this fraction was referred to as the ether-extractable fraction of fluid milk. The residual portion was then ex- tracted with ethanol and non-polar organic solvents and the resulting fraction was designated as the second fraction or the a1coholic-ether-extractable fraction. Three methods employing different non-polar organic solvent systems were compared and described as follows: (a) One hundred m1 of milk were mixed well with 100 m1 portion of a mixture of hexane and ethyl ether (1:1, v/v) in a suitable separatory funnel. Then the mixture was transferred into a centrifuge bottle and centrifuged at 2,000 x G for 15 min. The solvent layer was removed with a pipette and the remaining content was transferred in- to the original separatory funnel followed by extraction with the same solvent mixture for two more times. The sol- vent extracts were then combined, measured, concentrated and the resulting lipid was weighed. The residual portion was mixed with ].g;of Na oxalate and 100 m1 of ethanol and shaken well prior to extraction with the same solvent mix- ture by the same technique as described above. The combined extract was further washed with 100 ml of 5% NaCl solution, and passed through a column of anhydrous Na SO I evaporated 2 4 Ir—I—A and thr and Box the fll same as of peti tion; I tractic of 2% l mainly used 1: ethyl g tracted leum 8t Earlier ations up by a Celite YSis by 10Wing Flori-Si 46 and the fat weighed. This solvent system was used by Beroza and Bowman (1966). (b) Petroleum ether was used for the extraction of the first fraction. The basic technique was essentially the same as in method (a), except that 100 ml, 50 ml and 50 ml of petroleum ether were used successively for the extrac- tion; 1.5 g of Na oxalate was added before the second ex- traction, and that all the extracts were washed with 600 m1 of 2% NaCl solution. The amount of each reagent used was mainly according to the FDA technique (FDA, 1968), which was used in the next method. (c) The first fraction was extracted by the use of ethyl ether-petroleum ether, and the second fraction was ex- tracted by using Na oxalate, ethanol, ethyl ether and petro- leum ether. The techniques involved have been described earlier as the FDA method. A diagram showing the fraction- ations is given in Figure 1. Cleanup Procedures for the Residue Extract All of the extracts containing lipids were cleaned up by acetonitrile-petroleum ether partition and Florisil- celite absorption column techniques prior to the final anal- ysis by electron capture GLC. The hexane extract fol- lowing the alkaline hydrolysis was cleaned up directly on a Florisil-celite column. .61 wry-2""! “" “ 1 [ COMBINE FIRST F 02 ETHER-EL LIPID F: Fi9Ure 1 47 FLUID MILK Extracted 3X with one of the following three non-polar or- ganic solvent systems: ethyl ether-hexane, petroleum ether, and ethyl ether-petro- leum ether Centrifuged at 2,000 x G for 15 min L l I COMBINED UPPER PHASE LOWER PHASE 3 Concentrated Mixed with Na oxalate L and ethanol prior to f FIRST FRACTION extraction with the ' or same solvent system ETHER-EXTRACTABLE as above LIPID FRACTION Centrifuged at 2,000 x G for 15 min COMBINED SOLVENT AQUEOUS LAYER LAYER or LIPID-DEPLETED Washed with PHASE salt solution, dried over an- hydrous NaZSO4 and concentrated SECOND FRACTION or ALCOHOLIC-ETHER- EXTRACTABLE LIPID FRACTION Fi‘gure l.--Scheme for the separation of milk fractions by differential solvent extraction. 48 Acetonitrile Partition This technique was reported by the FDA (1968). Up to $3 g’of extracted fat were dissolved in petroleum ether and transferred into as 125—ml separatory funnel using a total volume of 15 m1 of petroleum ether. Thirty m1 of acetonitrile saturated with petroleum ether were added to the separatory funnel, and the contents thoroughly mixed for one min. When the layers were well separated, the lower phase (acetonitrile) was transferred into a lpOO-ml separa- tory funnel containing 650 ml of 2% NaCl solution and 100 m1 of petroleum ether. The acetonitrile partition step was repeated three more times and all of lower phase was trans- ferred into the same 1,000-m1 separatory funnel, while the final petroleum ether layer with relatively large portions of lipids was discarded. The large funnel was then gently mixed for diluting the acetonitrile with aqueous salt solu- tion and allowing the lipophilic pesticide residue to be re- distributed into the petroleum ether layer. The aqueous acetonitrile portion was drawn into another 1,000-m1 separa- tory funnel and was extracted again with 100 ml of petroleum ether by shaking the funnel vigorously for 15 sec. The aqueous phase was discarded. The petroleum ether extracts were combined in the original separatory funnel, washed twice with 100 ml of 2% NaCl solution, passed through an an- hydrous Na SO4 column (2.5 cm) and concentrated on a steam 2 bath to ca. 20 ml. deactiv action of 20 r cm of a mixtu: thin . about either the co of 2-3 residue effluen- Propria nitr0g« to 0,2 Chror Alto] deteC OPEra Slass atOPo 49 Florisil-Celite Column Cleangp (Shell Chemical Co., 1964) The Florisil-celite absorbent mixture (5:1, w/w) was deactivated with 10% water by mixing the content on a wrist action shaker for 16 hours. A chromatographic column of 20 mm x 400 mm was packed with a layer of glass wool, 2.5 cm of anhydrous NaZSO 10 g (4.2 cm) of the Florisil-celite 4’ mixture, and again a 2.5 cm layer of anhydrous Na SO4 and a 2 thin layer of glass wool. This column was prewashed with about 50 ml of hexane. The concentrated residue extract, either in hexane or in petroleum ether, was transferred onto the column, and eluted into a 400-ml beaker at a flow rate of 2-3 ml per min. Hexane was used to elute pesticide residues from the column until a total of 300 m1 of effluent was received, which was then evaporated to an ap- propriate concentration on a steam bath under a stream of nitrogen gas. Usually the final solutions containing 0.02 to 0.2 ppm of dieldrin were suitable for GLC analysis. Determination of Dieldrin by Electron Capture Gas Chromatography A F & M Hewlett-Packard Model 5750B research gas chromatograph (Hewlett-Parkard Analytical Instruments, Palo Alto, California) equipped with a tritium electron capture detector was used in the later part of this research. The operating conditions were: 4 ft x 1/4 in o. d. coiled glass column prepacked with 3.8% SE-30 on 80/100 mesh Di— atoport S; injection port temperature 240°C; column oven vrvww*““‘”Y-‘ I 200°C: dé flow rat‘ of 90% a size was 0.5 PPm standar< GLC be' tion d. peak h« of resi this st method . 9Taphjc methods height calcula constr Stand; matog the a gas ( the s COluIn I QF-l 50 200°C; detector 210°C; chart speed 0.25 or 0.5 in./min.; flow rate of carrier gas,rmfljxmh 45-50 ml/min; and purge gas of 90% argon-10% methane mixture, 50 ml/min. The sample size was 1-4 ul, usually 2 pl. Standard solutions of dieldrin ranging from 0.005 to 0.5 ppm were prepared in hexane. Usually no less than four standards (different concentrations) were injected into the L GLC before and after the sample injections on each opera- ’ tion day. A standard curve of dieldrin concentration versus L peak height on the chromatogram was made, and the quantity E of residue in the final sample extract was determined from this standard curve. Gaul (1966) has reported that the method of using peak height for calculating gas chromato- graphic peaks showed no significant difference from other methods including disc integration, triangulation, peak height x width at half height, and R x peak height in the t calculated results for dieldrin. A new standard curve was constructed on each operation day because the slope of the standard curve varied from day to day. A typical GLC chro- matogram for determination of dieldrin in milk is given in the Appendix. In the early stage of this research, a Beckman GC-4 gas chromatograph was used. It had an electronic wire as the source of electrons in the detector. A variety of columns have been used on this GC, including: glass or stainless steel column, 4 or 6 ft long, packed with QF-l and OV-l7 or DC 11, or DC 200 on Gas Chromo Q 60-80 mesh ature would speci since 5mm] tions 51 mesh or on Chromosorb w 60/80 mesh. The column oven temper- ature was 150°C to 220°C varied with the column used. It would be very tedious and superfluous in presenting the specific conditions used in each individual determination, since the standard solutions were always injected between samples during the period with the same operation condi- tions. Separation of Milk Fractions by Physical Methods Wm”"”"‘mr Three kinds of milk samples were used: (a) natur- ally contaminated milk from cows fed dieldrin, (b) raw whole milk containing added dieldrin, and (c) control raw whole milk. All milk was from the Michigan State University dairy herd. Milk was collected in stainless steel milk cans at the afternoon milking and brought back to the laboratory immediately for separation. For the preparation of milk containing added dieldrin, an appropriate quantity of di- eldrin solution (1 mg/ml in acetone) was added to the milk and the mixture was allowed to equilibrate by intermittent stirring in a water bath at 37°C for 45 min. Preparation of Major Milk Fractions The method employed was essentially that of Brunner (1965) and Swope (1968). Milk was first separated into skim milk and cream with aDeLaval disc-type separator (Model 9). The resulting cream was diluted with three volumes of warm 52 (39-40°C) 0.25 M sucrose solution (for the naturally con- taminated milk sample 1 and 2) or deionized distilled water (39-40°C) (for other samples) and reseparated as before. This process was repeated three times. A suitable amount of the original skim milk was saved for later analysis while the remainder and the washings were discarded. Washed cream was stored in a large Erlenmeyer flask (3,000 or 4,000 ml) and cooled overnight at 4°C. On the following morning, it was allowed to warm up to 12°C and then churned on an Eberbach rotatory shaker at room temperature for about one hour or until the separation of butterfat and liquid phase was well observed. The buttermilk (liquid phase) was de- canted into another flask for further treatment and the butter was melted in the original flask or other suitable container at 45°C. Butteroil and butter serum fractions were thus obtained. A scheme of the separation procedure is presented in Figure 2. During the entire separation process, each yield of the resulting fractions was weighed, except for skim milk which was estimated by the difference between whole milk and cream. A quantitative portion of each fraction was taken for lipid and dieldrin analysis before that fraction was further separated. The storage conditions and further treatments of various fractions were as follows: Whole milk, skim milk, cream, and washed cream were Stored in Nalgene bottles and kept frozen at -20°C until needed. Butter and butteroil were stored in glass bottles ‘f’rw— w--. - IT 53 FRESH RAW MILK Separated at 35-37°C l 1 CREAM (unwashed) SKIM MILK Washed 3X with 3 volumes of 0.25 M sucrose solu- tion or deionized dis- tilled water at 39-40°C .1 WASHINGS WASHED CREAM Cooled overnight at 4°C; churned at 12°C r‘ l BUTTER BUTTERMILK Melted at 45°C I tn BUTTEROIL BUTTER SERUM (crude) Centrifuged at 1,500 X G for 15 min I ' l j BUTTEROIL INTERFACIAL BUTTER SERUM MIXTURE (refined) Diluted, warmed at 45°C and centrifuged 1 BUTTEROIL INTERFACIAL FL FF (Process epeated until almost all of the free oil was separated) Figure 2.--Scheme for the separation of major milk fractions by physical methods (adapted from Brunner, 1965). .—w- (W? i 54 with 0.01% BHT (butylated hydroxy toluene) added and were kept frozen until needed. Butter serum for the first two preparations of nat- urally contaminated milk fractions, the crude butter serum containing fairly large amounts of butteroil was collected and analyzed; for other preparations, this butteroil was further separated from butter serum by centrifugation at 1,500 X G for 15 min. However, there was still no clear-cut differentiation between the oil and the serum phases. After removal of oil (as much as possible), the pure butter serum phase in the lower portion of the layers was pipetted out. The middle layer which consisted of butteroil, inter- facial fluff and some butter serum was thus left in the cen- trifuge bottle. This mixture was diluted with an equal volume of water, stirred, warmed at 45°C and centrifuged as described previously. This washing and reseparation process (Figure 2) was adapted from Chien and Richardson (1967). The pure oil as well as the pure aqueous phase was removed and the process was repeated until almost all of the butter- oil phase was separated from the interfacial fluff. The butteroil collected in these later steps were all combined to the original butteroil fraction and stored. Only the original pure butter serum, designated as refined butter serum, was saved and stored in Nalgene bottles at -20°C. Buttermilk was preserved by the addition of 0.01% BHT as well as penicillin G (46 mg/l,000 m1) and streptomy- cin sulfate (75 mg/l,000 ml) as suggested by Quirk (1965). ‘fifinT*——“‘r 55 Preliminary tests showed no interference peak on the GLC chromatogram as a result of the use of these chemicals. A small portion of this buttermilk was stored in Nalgene bottle at -20°C for later analysis while the remainder was stored at 4°C until needed for fat globule membrane pellet collections. Prepgtation of Fat Globule Mem- brane Pellet Fractions The procedures involved were primarily those of Swope (1968) with slight modifications as shown in Figure 3. Buttermilk was filtered through four layers of cheesecloth and centrifuged at 2,000 X G for 20 min to eliminate the un- churned fat globules. The liquid was designated as refined buttermilk, and this was subsequently centrifuged at 12,500 X G for 14 min at 4°C to yield a 7,500 8 membrane pellet. The resulting cream-like flotation layer was carefully re- moved from the centrifuge tube with a spatula and a small piece of Kimwipe and discarded. The supernatant was pipet- ted into a second centrifuge tube (care was taken to avoid disturbance of this very loosely—packed 7,500 S pellet) un- til approximately 1. cm depth of the liquid including the pellet was left in the bottom of the centrifuge tube (50 ml capacity). Deionized distilled water was added to this tube, the contents were stirred, and recentrifuged at the same speed. This process was repeated once more to remove the fine particles which might be trapped in the pellet, and the washings were discarded. The original supernatant was ‘15-'- W’ _.._‘__l_ ~5— 56 BUTTERMILK (crude) Filtered Centrifuged at 2,000 X G for 20 min I 7 BUTTERMILK UNCHURNED FAT GLOBULES (refined) Centrifuged at 12,500 X G for 14 min at 4°C PRECIPITATE, FLOTATION LAYER 7,500 S PELLET SUPERNATANT Centrifuged at 30,000 X G for 2.7 hr at 4°C PRECIPITATE, FLOTATION LAYER 230 S PELLET SUPERNATANT Centrifuged at 92,000 X G for 6.3 hr at 4°C PRECIPITATE, CLEAR CENTER FLOTATION TOP 35 S PELLET LAYER LAYER Figure 3.--Scheme for the separation of fat globule membrane fractions (adapted from Swope, 1968). ‘W— mun-o. .-.z ) ‘ 57 further centrifuged at 30,000 X G for 2.7 hr to yield a 230 S pellet. Again the flotation layer was removed, and the pellet was washed twice by the same technique as described for the 7,500 8 fraction. An International Model HR-l high-speed refrigerated centrifuge (International Equipment Company, Needham Heights, Massachusetts) was employed for the collection of these two membrane fractions. The No. 856 rotor used in this centrifuge had dimensions as follows: ra (radius from the center of rotation to air-liquid meniscus) = 5.5 cm; rb (radius from the center of rotation to bottom of liquid) = 10.6 cm. The Svedberg units (S) of the resulting pellets were calculated according to the following equation: 13 S _ 10 1n (rb/ra) WZT where T = separation time in seconds, and . . 2 m (rpm). W = angular veloc1ty in rad/sec = 60 Therefore, the operating speed and time for collection of 7,500 S and 230 S pellets were computed as 10,000 rpm for 13.3 min (ca. 14 min) and 16,500 rpm for 2.7 hr, respective- ly. A Beckman Model L-2 preparative ultracentrifuge (Beck— man Instruments Inc., Spinco Division, Lincolnwood, Illi— nois) was used for collection of the 35 s pellet. The r and rb values for the type-30 rotor used in this cen- a trifuge were 5.0 and 10.5 cm, respectively. The operating ‘w-‘P'._hr.aia i. ~ 58 speed and time for 35 S pellet were 28,000 rpm (92,000 X G) and 6.3 hr, respectively. No washing process was involved for this final precipitate. For the control milk sample, the 35 S pellet was collected directly from the refined but- termilk. The flotation top layer and the final supernatant of this sample and of the milk samples containing added di- eldrin were collected for further analysis, whereas the supernatant and top layer of other samples were discarded. Each pellet was removed from the centrifuge tube with a spatula and a minimal amount of water into a suitable con- tainer and stored at —20°C until used. Analysis of the Physically Separated Milk Fractions Extraction of Lipids and Dieldrin The major fractions were thawed and extracted accord- ing to the procedures described on page 41 (FDA method). Modifications for different fractions were as follows: skim milk, 150-300 ml, reagents used were proportionally in- creased; cream, diluted with approximately six volumes of. water prior to the extraction step; butter, 5 g were dis- solved in 25 ml ether mixture (ethyl ether: petroleum ether, 1:1) and 50 ml of 50% ethanol were used in extraction; butteroil, 5 g were dissolved in ethers, washed with salt solution and water, dried and weighed, no further extraction was involved. For buttermilk and butter serum, since poor separations between ether layer and aqueous layer were (W’s-90 IDA». .- ' A. r- 59 observed, a technique was then adopted, which used 3.5 vol- umes of ethanol, one volume of ethyl ether and 2.5 volumes of petroleum ether in the first extraction and one volume of petroleum ether in each of the second and third extraction. The residual phase was diluted with 2.5 volumes of water (of the sample original volume), and was again extracted with eth- yl ether and petroleum ether as the original FDA procedure. i." _ . 3_--____._ -.. The membrane materiahswere dried under a hood on a warm steam bath (for the first preparation of naturally con- cvwW" taminated milk) or freeze-dried in a Stokes freeze-dryer at 100 u Hg pressure for 24 hr with the heating plate tempera- ture at l3-16°C, finally dried in an oven at 50°C to remove the traces of moisture which might be in the freeze-dried samples. The dried solids were weighed and used for the lipid and dieldrin determinations. The extraction procedures for free lipids and bound lipids were adapted from Cerbulis (1967). Free lipids of the membrane material were extracted with petroleum ether (80 ml per gram of solids) by continuous stirring for 3 hr and the extract was removed by centrifugation. This process was repeated three times, and the extracts were combined, evaporated to dryness on a steam bath with a stream of nitro- gen and the resulting lipids weighed. Bound lipids were next extracted three times from these samples with chloro- form-methanol (2:1, v/v) by continuous stirring for 3 hr us- ing 30 m1 of solvent per gram of dry solids. The extracts were combined, washed twice with 0.2 volume of 0.2% 6O NaCl solution, dried and the resulting lipids weighed. Since chloroform caused some interference on the GLC chromatograms, later extractions of bound lipids used ethanol, ethyl ether, and petroleum ether following the procedure as described earlier for buttermilk. The free- lipid-extracted membrane materials were rehydrated with ten volumes of water prior to the extraction for bound lipids. A preliminary test showed that this revised lipid extraction technique provided a comparable result as the extraction by a solvent mixture of chloroform-methanol (2:1, V/v) re- ported by Folch et a1. (1957). Determination of Dieldrin Dieldrin contents of the extracted lipid fractions were determined by electron capture GLC using the same cleanup techniques and GLC conditions as described previous- ly. Phosphorus Determination Lipid phosphorus was determined essentially according to the procedure of Morrison (1964). The extracted lipids were redissolved in a mixture of chloroform and methanol (2:1, v/v) to a concentration of approximately 30 mg lipids per ml of solution. Aliquots of the solution containing 2-50 pg phosphorus were placed in 40-ml test tubes accurate- ly calibrated at the 12.5 ml and 25 ml level. Solvent was then removed with a stream of nitrogen on a steam bath. One and one-tenth ml of sulfuric acid (95%, sp. gr. 1.84) were 61 added to each tube and this was then heated in a multi-block tube heater (Research Specialty) at 220-230°C until char- ring was completed; usually for 12-15 min. Hydrogen perox- ide (30% H 02, w/v) was added dropwisely and the tube was 2 heated for l min after each addition. This process was re- peated until a clear solution was obtained. The tube was further boiled for 3 min to drive off the excess H202. After the tube was cooled completely, deionized distilled water was added to bring the volume to 12.5 ml. Five-tenths ml of sodium sulfite (16% solution of NaZSO3 anhydrous), 5 ml of ammonium paramolybdate (2% solution of (NH4)6Mo7 024° 4 H20) and 0.5 ml of ascorbic acid (10%) were added succes- sively. Each tube was well shaken and then heated in boil- ing water for l min. After cooling, deionized water was added to bring the volume to 25 ml. The absorbance of the solution was read at 822 nm with a Beckman DU-2 spectro- photometer. Duplicate blanks and standard solutions were carried out at the same time with samples. The phosphorus concentrations in sample solutions were read on the stand- ard curves which were prepared individually for each set of digestion and determination. Phospholipid contents were estimated by multiplying the values of phosphorus by 25 (Swope, 1968). The averages of duplicate determinations are reported. T?" 7. 'F""”“"' RESULTS AND DISCUSSION Comparison of the Extraction Efficiency of Lipid and Dieldrin in Milk Four methods were compared. The results are shown in Table 2. The lipid extraction efficiency by the FDA method was comparable to the Mojonnier method. This was no surprise since the FDA method was essentially a modification of the Mojonnier method. The ether extracts obtained in the FDA method were further washed with a 2% NaCl solution and water to remove some impurities which might affect the base- line on GLC chromatograms for residue analysis. This wash- ing process might also account for the slightly lower lipid recovery by the FDA method. However, this quantity was not significant. The total recoveries of added dieldrin by these two methods showed no significant difference, although the method of Mojonnier seemed to give a slightly higher value. Usually, as described in the "Pesticide Analytical Manual" (FDA, 1968), a portion of the extracted fat was weighed and analyzed for the residue content, and subse- quently the results were expressed on a fat basis. For the above two methods, the concentrations of dieldrin in the ex- tracted lipids were also comparable with each other. 62 63 IIILI \ .. -l lbthLAFW: .Uozqu mHnu an Umuomnuxm uoc mm3 pom x0020 .chHumcHEHmme mum000mso mo mmcmu can mmmum>¢n .c00p0wHU cocoa mo Ema mo.o mchHmucoo x008 cmNchmOEoc omNHusmpmmm HMHoumEEoom I I Amcouumv 00.0 + 00.00 0000.0 + 0000.0 I- on- 000000000: 0:000000 I I 8005 00.0 + 00.00 0000.0 + 0000.0 I- 0.. 000000000: 00000000 00.0 H 00.00 0000.0 H 0000.0 00.0 H 00.0 00.0 H 00.0 0200000000 00.0 H 00.00 0000.0 H 0000.0 00.0 H 00.0 00.0 H 00.0 00000000: 00.0 H 00.00 0000.0 H 0000.0 00.0 H 00.0 00.0 H 00.0 400 c0ue0m0o 00 0002 000:3 000 .wuxm mum>oomm W San .COHumuucmocoo CHHUHwHQ cmuomuuxm pHm00 w noonumz 0.0005 00 cHup0mHo cam UHQHH mo moogume ucmumMMHp an hocmHUHmmm COHuomnuxm mo comHHmmEooun.~ manme 64 The surfactant method showed a lower lipid recovery as compared to the above two methods. Consequently, the total recovery of dieldrin by this method was also much lower than the FDA or Mojonnier method; approximately 10% less lipid as well as dieldrin were extracted by this tech- nique. However, when the isolated fat was analyzed and the results were expressed on a fat basis, only 2-3% less amounts of dieldrin was observed as compared to that found by the other two methods. Lampert (1964), who first devel— oped the surfactant technique for residue analysis in milk, used urea and isopropyl alcohol in addition to Triton X-100 and sodium tetraphosphate. He showed generally good re- coveries on a fat basis for several added insecticides in- cluding dieldrin. However, he indicated that no intention was made to give a quantitative fat separation by this technique. He also stated that good but slower fat separa- tion was feasible if not using urea and isopropanol. The procedure used in the present study was that of Kroger and Patton (1967); they and Moubry 33 31. (1967) applied the simplified technique using Triton X-100, sodium tetraphos- phate and water to separate fat for residue analysis. But neither of them have reported the percentage of fat recov- ered by this technique. Kroger and Patton found that com- parable results for heptachlor epoxide and dieldrin in milk were obtained by this technique and by the Roese-Gottlieb method (similar to the FDA method) on a fat basis. Moubry et a1. did not report the results on a fat basis, but 65 indicated that 5-10% less residues were extracted from milk than by the Mills method (Mills, 1961; also similar to the FDA method). From the results of the present study and the reports of earlier workers, it seems that the lower recovery of the total quantity of residue was due to the incomplete fat recovery; whereas the residue concentration in the iso- lated fat was similar or slightly lower than that in the solvent extracted fat. A report of Lawrence and Burke (1970) suggested that the low recoveries of residues by this surfactant technique probably was due to the low concentra- tions of residues in the fat released by this technique. In examining the alkaline hydrolysis method, it is obvious that the mild hydrolysis, as a consequence of the low heat and a shorter hydrolysis time, showed more advan- tages than the strong hydrolysis with high temperature and longer hydrolysis time. The recovery of added dieldrin as determined by the mild alkaline hydrolysis was comparable to that by the FDA method, whereas the strong alkaline hydroly- sis provided a much lower recovery for dieldrin. This was somewhat surprising because of reports that indicated that dieldrin is thermally stable and not attacked by bases (Porter, 1964). No break-down products were observed on the resulting product GLC chromatograms from either hydrolysis procedure. It appeared that the fragments of dieldrin, if there were any from the strong hydrolysis process, might be hydrophilic and lost in the aqueous phase and washings. Some amounts of dieldrin possibly could be lost with the 66 vapor under the vigorous distillation process of the strong alkaline hydrolysis procedure. Another possibility was that some components in the milk samples might have an effect on the stability of dieldrin with this hydrolysis condition. Neither the FDA (1968) nor the Shell Chemical Company (1964) who recommended the strong alkaline hydrolysis methods has given any information on the recoveries of residues in milk. I The mild alkaline hydrolysis technique had an ad- vantage over the FDA solvent extraction method in eliminat- k ing the acetonitrile partition step. However, the Florisil- v celite column cleanup step was still needed in this tech- nique to provide a suitable extract for GLC analysis. One disadvantage of this method is that an independent fat de- termination should be conducted in order to be able to con— vert the results from a weight basis of whole milk to a fat basis. If a laboratory has personnel to perform fat deter- minations for all samples routinely, or if the results are not required to be reported on a fat basis, then this mild alkaline hydrolysis would be the method of choice in saving analytical time and chemical reagents. The control samples of commercial pasteurized homog— enized milk were found to contain 0.003-0.005 ppm of di- eldrin by all the methods reported as above. No correction for recovery has been made for the purpose of comparing ex— traction efficiency. Since it was desired in the present research to de- termine both the lipid and the dieldrin content in milk 67 samples, the extraction method of FDA was thus chosen for later experiments. Recoveries of dieldrin at different added levels were also examined before the final decision was made. Both the gross and the net recoveries are given in Table 3. Analysis showed that the blank milk samples contained less than 0.005 ppm of dieldrin, where 0.005 ppm was the limit of the analytical sensitivity under those par- ticular experimental conditions. An attempt, by using TLC techniques ("Changes in methods," 1966) to confirm the re- sults of the blank milk samples was not successful because of the low sensitivity and low selectivity of the TLC method as compared to the electron capture GLC method. It was be- cause of this background concentrations of residue in milk samples that the apparent gross recoveries of dieldrin at low levels were above 100%, whereas the net recoveries of dieldrin were not so high. When the contamination level was 0.05 ppm or above, the presence of this low background resi- due did not significantly affect the difference between the gross and net recoveries. Generally, the recovery of added dieldrin at various levels was satisfactory. Statistical analysis showed that an average value of 92.65% with a standard deviation of i 3.29% for the gross recovery and 91.03% i 3.78% for the net recovery were obtained for added dieldrin at levels of 0.050 to 2.000 ppm. Blank samples using water in the place of milk showed 95.0 i 1.0% recovery of dieldrin. Therefore, the less than 100% recovery of di- eldrin from the milk samples was probably due to losses in n. ‘11ng - 68 Table 3.--Recoveries of dieldrin from milka at different added levels by the Food and Drug Administration method. Dieldrin Concentration in Milk, ppm % Recovery_of Dieldrin Added Observed Gross Net 004 None <0.0050b -- -- 1 0.005 0.0087 174.00 74.00 g { 0.010 0.0132 132.00 82.00 g 0.050 0.0460 91.90 86.90 0.100 0.0904 90.42 89.20 0.200 0.1862 93.10 90.60 0.400 0.3794 94.86 93.60 0.800 0.7162 89.52 88.89 1.000 0.9878 98.78 98.28 2.000 1.8000 90.00 89.75 Average and standard deviationc 92.65:3.29 91.03:3.78 Coefficient variationc 3.55% 4.16% __‘ aCommercial pasteurized homogenized milk. bDetection limit. CCalculations were based on the dieldrin concentra- tion levels of 0.050 ppm and above. 69 the multiple steps of transferring, washing, cleaning up and evaporation, and only a small percentage of the loss could be due to the inefficient extraction, if there was any. This FDA method has been subjected to collaborative studies in a number of laboratories, and generally satisfactory re- sults were reported for the analysis of fruits and feed products (Wells, 1967), butterfat (Carr, 1970) and various fatty foods (Johnson, 1965). It was concluded to use this procedure for later experiments in this research. Distribution of Lipids and Dieldrin in Solvent Separated Milk Fractions Three solvent systems were compared in this study. The first fraction was extracted with non-polar hydrocarbon solvents and the second fraction was extracted from the residual portion of the first fraction by using both polar and non-polar organic solvents. Results in Table 4 show that only a small percentage of the total lipid was obtained in the first fraction, although method No. 3 had a higher recovery than the recoveries of the other two methods. How- ever, the concentrations of dieldrin in the first fraction (ether-extractable fraction) were much greater than in the second fraction (alcoholic-ether-extractable fraction). The ratio of dieldrin concentration in fraction 1 and 2 ranged from 3:1 to 27:1, varied with the solvent systems used. In considering the total amount of dieldrin in these two fractions, an almost equal recovery of dieldrin was 70 m. .I . . .lhetb.’ .>0m>00ommmmu .c0000m00 pcoomm Cam 0m000 0:0 000 U0m00 m0nm0omn0xmlum£0mlo00000000 man @0000 m0nm0omu0xmuum£0mo .m0m>00ommmmu .m can .m .0 .oz Uon0mE 000 00000 850000000 Inm£0m 0%:00 0cm .00000 Esm0ou0mm .00300 0>£0mumcmxm£ 0003 mEm0m>m 0sm>0om0 .m0mmn 000003 0 co C0000w00 Ema 0 £003 000000000 0005 0000:000Eon 00000500mmm 00000088000 mm.mm 0m.000 00m.m 00008 000 00.00 0 00.00 mmv.m cam m0 mm.00 m 00.00m m00.0 0m0 m mm.mm 00.000 m0m.m 00009 000 00.00 0 00.00 000.m UCN 0m 00.m0 mm 00.mmm.m 000.0 0m0 m 0m.mm mm.m00 000.m 00009 000 00.00 0 00.00 000.0 0:0 0m 00.00 00 00.000.0 000.0 0m0 0 0cm>0om C00 000mm Ema . . . 00000 00 Um0omu0xm 0 oz oz 00 mum>oomm >00>oomm w ©0Q00 Um0omu0xm :0 ©0m00 w c000omum ©m£0mz m0mE0xoumm4 w 0000mu0cmocoo C0000m0o m.m2000om00 x006 000mummmm 0cm>00m 00 00000000 000 m00m00 mo 0000sn0u0m0onn.0 00009 71 obtained in either fraction for all three solvent systems applied. Beroza and Bowman (1966) reported that about 10% of the total lipids in milk were extracted with the solvent mixture of hexane and ethyl ether, the same solvents as those used for the first fraction of method No. l in the present study, and almost 100% were recovered with solvent system of ethanol and hexane-ethyl ether (same as that used in the second fraction) from fluid milk samples. The pres- ; “psi 6- ent results show a substantially lower recovery of fat with the above solvent system; such as that the first fraction accounted for only 4% of the total extracted lipids, and the sum of the two fractions accounted for only 80% of the total lipids as extracted by method No. 3. The concentration of dieldrin when calculated on the total extracted lipid basis of method No. l was higher than the other two methods, how— ever, the total amount of lipid was low, therefore the ab- solute recovery of dieldrin was not as high as by the other two methods. Lawrence and Burke (1970) also observed that the residue levels tended to be high on a fat basis when only 50-80% of the total fat was extracted from milk by a single solvent extraction technique. Methods No. 2 and No. 3 gave comparable results in total lipid extraction and total dieldrin extraction, where the slight difference could be due to the analytical vari- ance. All of the above studies apparently suggested that the distributions of residue in the lipid fractions 72 extracted differently would be very likely to be different among one another. Therefore, it seems to be definitely im- portant to extract all of the lipids in milk, if the subse- quent residue analysis is to be based on that isolated lipid fraction. Another aspect of the present results concerning the incomplete recovery of dieldrin with non-polar organic solvents will be tentatively elucidated in this discussion. The literature has offered excellent information in the im- portance of milk fat extraction techniques with polar sol- vents or other suitable agents to break the fat globule mem- brane. A question was then raised on whether the added pes- ticides in milk could penetrate the membrane and why the recoveries of these non-polar organochlorine compounds were so low with hydrocarbon organic solvents, if they had not penetrated into the butterfat. Beroza and Bowman (1966) thought that this low recovery was due to the adsorption of these pesticides onto the surface of fat globules rather than penetration into the fat body. A low recovery of these compounds with hexane-ethyl ether was also observed for skim milk in which the pesticides were added, as reported by Beroza and Bowman (1966). They suggested that it was due to the binding forces of non-polar residues to the aqueous phase of milk, although this binding was much weaker than the adsorption on the cream phase in whole milk. However, they did not notice, or did not report the recovery of the solvents used. It was experienced in the present research 73 that most of the non-polar organic solvents, such as hexane, ethyl ether, or petroleum ether, or the combined mixtures of them, were poorly recovered (20-43% as shown in Table 4) once they were mixed with the fluid milk. In some instances, the recoveries of solvents were even lower than that of di- eldrin. Poor separations of the non-polar solvents from the skim milk were also observed. This finding implies that the low recovery of added dieldrin (possibly as well as milk fat) might be due to the low recovery of solvents, or that it may be due to causes similar to those which depressed the complete recovery of added solvents. Unfortunately such causes are not yet thoroughly understood. Milk is a complex system composed of various proteins, lipids and other com- ponents in various forms. The added non-polar organic sol- vents as well as dieldrin could be adsorbed on the hydro- phobic groups of the serum proteins and/or the membrane pro- teins. In connection with this finding, it was then specu- lated that the function of alcohol for pesticide and fat ex- traction appeared to be no less important in rupture of the emulsion between the solvents and milk phases than the rup- ture of fat globule membrane, where the later was indicated as the only essential function of alcohol in the discussion of Moffitt.(l963)and.Kroger and Patton (1967). Additional investigation was made to detect the re- maining residue, if there was any, in the lipid-depleted phase of whole milk and skim milk. The naturally contami- nated milk samples were used in this experiment and the ‘mm— 'v'T—I—. 74 technique of alkaline hydrolysis was the method of choice because the solvent-extractable residue should have been re- moved with the lipid fraction. Table 5 shows the results of such analyses. The concentrations of dieldrin in the lipid- depleted fractions of both the milk and skim milk samples were below the detection sensitivity, under these particular experimental conditions. The strong alkaline hydrolysis procedure again provided a much lower recovery of dieldrin in either whole milk or skim milk as compared to recovery by the mild hydrolysis method. Thus, no attempt was made to use this strong hydrolysis method for the determination of residues in the lipid-depleted fractions. To gain further information concerning the distribu- tion of dieldrin in milk fat, the total extracted lipid was separated into neutral and phospholipid fractions by a batchwise silicic acid technique described by Abramson and Blecher (1965). Isolated milk fat was mixed with silicic acid and packed into a column. The neutral lipids were eluted with ethyl ether and petroleum ether from the column and the phospholipids were subsequently eluted from the col- umn by methanol. The relative quantities of neutral and phospholipids recovered were 99.46 and 0.54%, respectively and the relative quantities of dieldrin recovered with these two lipid fractions were 99.85 and 0.15%, respectively. When a blank solution of petroleum ether containing dieldrin was fractionated in a similar manner, it was observed that 99.65% of the dieldrin was recovered with the effluent of .‘erw _II. ~. . '35 75 Table 5.--Distribution of dieldrin in lipid-depleted milk fractions.a L Dieldrin Concentration on a Fraction Weight Basis, ppm Method 1b Method 2b E I If Whole milk 0.0285 0.0035 Lipid-depleted milkC <0.0003d -_ Skim milk 0.0032 0.0001d Lipid-depleted skim milkC <0.0001d -- aFresh raw whole milk from cows fed dieldrin. bMethod l and 2 refer to the mild and strong alka- line hydrolysis techniques respectively. CLipids were depleted by the FDA extraction proce- dure. dDetection limit. 76 ethers and only 0.35% with methanol. Apparently dieldrin had solubility similar to that of neutral fat in ethers, whereas the trace residues eluted out with the phospholipid or methanol might be due to an incomplete ether extraction, or such a small amount of dieldrin was slightly adsorbed on the silicic acid just as were the phospholipids. However, not much significance could be ascribed to either factor, i‘ since the absolute quantity of dieldrin recovered in the methanol fraction was near the detection limit, and the (#w“*" small amount of residue observed might be an artifact under that particular set of experimental conditions. Addition- ally, once the lipids and residue had been extracted.from the natural milk system with solvents, the original binding forces between the pesticide residues and the milk compon- ents could very well be altered. A second series of experi- ments were, therefore, designed to separate milk fractions by physical forces. The correlation of dieldrin with lipids in these fractions would provide a better insight of the natural deposition of such an organochlorine pesticide re- sidue in milk. Distribution of Lipids and Dieldrin in Physically Separated Milk Fractions Naturally Contaminated Milk Tables 6, 7, 8, 9a and 10 present the results of lipid and dieldrin determination in naturally contaminated milk fractions. Three collections and preparations were 77 made in different periods. It is very obvious that the concentrations of dieldrin residue in milk fat were signifi- cantly decreased in the later collections when the feeding of pesticide had been discontinued. Slightly modified procedures were adapted for each of these three preparations. In preparation 1 (Table 6) fractions of whole milk, skim milk, cream, butteroil, buttermilk and butter serum were analyzed for total lipid, dieldrin and phospholipid content. The concentrations of dieldrin in these fractions were generally proportional to the total lipid contents in these fractions, except that butter serum contained slightly lower levels. The high phospholipid content in this fraction might have an influ- ence on the distribution of residue. However, the same speculation can not be applied to buttermilk, which also contained a high level of phospholipids but had dieldrin level compatible with that in other fractions. In the second preparation (Table 7), both unwashed and washed cream were examined, and the buttermilk was fur- ther refined by centrifugation to remove any unchurned fat. Washing of cream apparently did not affect the content of residue on a fat basis. An interesting finding was made in the comparison of buttermilk and refined buttermilk. When the unchurned floating fat globules were removed, the total lipid content of buttermilk was reduced, but the relative amount of phospholipid in the total lipid was increased. On the other hand, an inverse distribution pattern of _ 0——0- _. 78 Table 6.—-Distribution of lipids and dieldrin in major fractions of naturally contaminated milk, prepara- tion l.a .. I Li id Phospholipid in Dieldrin Concentration, I Fraction (g) Total Lipid ppmb , 0 (%) Extd. Lipid Fraction 1 Whole milk 3.48 0.62 11.05i0.50 0.384:0.017 Skim milk 0.66 1.49 ll.65:0.75 0.077:0.005 Cream 31.53 0.44 11.45:0.40 3.610:0.126 Butteroil 100.00C 0.06 11.90:0.20 11.900:o.200 Buttermilk 0.40 15.79 11.87:0.38 0.047:0.002 Butter serum 12.13 8.38 9.06:1.43 l.098:0.174 aMilk from a cow fed dieldrin (see Table 1. Milk Collection No. l). bAverage and range of duplicate determinations. c Assumed value. 79 Table 7.--Distribution of lipids and dieldrin in major fractions of naturally contaminated milk, prepara- tion 2.a Phospholipid Fraction Lipid in Total Dieldrin Concentration, ppmb (%) Lipid Extd. Lipid Fraction (%) Whole milk 3.67 0.62 4.64:0.05 0.170:0.002 Skim milk 0.47 2.74 4.67:0.58 0.022:0.003- Cream (unwashed) 25.71 0.28 4.55:0.10 l.l70:0.026 Cream (washed) 26.16 0.54 4.39:0.06 1.148:0.016 Butteroil 97.00 0.02 4.80:0.12 4.656:0.116 Butter- milk 4.91 1.56 4.00:0.20 0.196:0.010 Butter- milk (refined) 2.39 3.09 3.20:0.12 0.077:0.003 Butter serum 26.55 3.05 3.65:0.10 0.969:0.027 _ aMilk from cows fed dieldrin (see Table 1. Milk Collection No. 2). bAverage and range of duplicate determinations. dI-‘rlrnn: ‘ (W2 .0 1 - 80 dieldrin was observed, i.e., refined buttermilk contained a lower level of residue than the crude buttermilk on a fat basis. It appears that the association of dieldrin with phospholipids was not so significant as it was with the neutral lipids. The first five major fractions, as shown in Table 7, contained dieldrin at similar levels on a fat basis, whereas L0 lower concentrations of dieldrin were observed in the last three fractions, buttermilk, refined buttermilk and butter serum. These three fractions contained more phospholipids .Qm§,_ .,. ,1-.. 0 .- than were in the other fractions (except skim milk) in this preparation, but did not contain as much phospholipids as that in the corresponding fractions of the previous prepara- tion. It seems possible that some other factors may be relat- ed to the distribution of residue besides the phospholipid contents in these milk fractions. In order to understand the possible binding forces of such a pesticide chemical in the fat globule membrane, the harvested membrane pellets were air-dried and succes- sively extracted by petroleum ether and chloroform-methanol into two fractions. Table 8 indicates that relatively greater amounts of the residue were extracted with the free lipids. Although there were still some residues in the bound lipid fraction, the concentrations were much lower. In addition, considerable interference was encountered in the GLC determination of residues in the bound lipid ex- tracts under these particular experimental conditions, and 81 Table 8.--Distribution of lipids and dieldrin in fat globule membrane pellets of naturally contaminated milk, preparation l.a Free Lipid Bound Lipid Dieldrin Concentration, Pellet in Total in Total ppm Lipidb Lipidb Free Bound Total (%) (%) Lipid Lipid Lipid 7500 S 12.02 87.98 3.94 0.88 1.13 230 S 45.00 55.00 5.80 0.93 3.12 35 S 30.71 69.29 6.43 0.88 2.58 Total 31.70 68.30 5.87 0.90 2.47 aMilk from a cow fed dieldrin (see Table 1. Milk Collection No. 1). Each value is the result of a single de- termination of pooled pellet fractions. bFree lipids were extracted with petroleum ether from the air-dried pellets, bound lipids were successively extracted with chloroform and methanol (2:1,v/v). 82 thus the true concentration of dieldrin probably would be lower than observed. At this point, it still could not be concluded that dieldrin was more associated with the free lipids than the bound lipids, since the higher recoveries of dieldrin with the free lipids might be due to their similar solubilities in petroleum ether, and in the original natural state the dieldrin could be deposited in the bound lipid L“ phase as much as in the free lipid phase. Nevertheless, the data reflect that this organochlorine pesticide was more E readily extracted with non-polar organic solvent than were the bound lipids, which were composed of relatively high levels of phospholipids. Apparently, dieldrin was not so significantly bound to the phospholipids or the protein moiety in the fat globule membrane system. The concentrations of dieldrin were generally lower in the membrane pellets than in the refined buttermilk when they were calculated on total lipid basis. This difference might be attributed to the loss of residues during the air- drying process (but this was very unlikely), or to the lower efficiency of extraction for dried materials (not impossi— ble), or more likely was due to the removal of some flota- tion layers during the multi-centrifugation process, and thus the resulting membrane pellets contained greater amounts of phospholipids than buttermilk. It again appeared that the lower levels of dieldrin in the total lipids were correlated with the high levels of phospholipid content in this fraction. However, the true values of phospholipids 83 in these samples were not determined since all of the ex- tracted lipids had been used for residue analysis. A third series of milk samples were collected in much greater quantities. The cream-washing procedure was modified so as to use deionized distilled water instead of a sucrose solution, thus the membrane pellets would be ob- L tained in a water solution and the subsequently dried mate- J'W rials would be the true dry weight of each pellet without % i the added sucrose. Results of this study are presented in i Table 9a and 9b. First, it is to be observed that these ” milk samples contained much less dieldrin than found in the two previous milk collections, due to the fact that the same dairy cows had received no dieldrin for several months. The decline of residue in milk fat from dairy cows has been well documented in the literature. This observation in the present research only once again confirms those reported by earlier workers. Results in Table 9a and 9b are the average of analy- sis of two independent preparations, where each preparation consisted of six consecutive milk collections and individual fractionations. In this study, the butter fraction was in- cluded in the analysis and the butter serum was further re- fined to eliminate residual amounts of butteroil. Only re- fined buttermilk and butter serum were analyzed because the relatively high contents of phospholipids in these fractions would be of interest in relating to the residue contents. Table 9a, again, shows that dieldrin concentrations were 84 Table 9a.--Distribution of lipids and dieldrin in major fractions of naturally contaminated milk, pre— paration 3.a Phospholipid Fractionb Lipid in Total DieldrinIConcentration,.ppmC (%) Lipid Extd. Lipid Fraction _1 (%) Whole f L milk 4.05 0.71 0.83:0.09 0.034:0.004 E Skim milk 0.58 1.94 0.89:0.10 0.005:0.001 ~“’ Cream (unwashed) 53.63 0.37 0.83:0.02 0.445:0.011 Butter 83.00 0.31 0.86:0.15 0.714:0.125 Butteroil 99.20 0.01 0.87:0.09 0.863:0.089 Butter- milk (refined) 1.67 11.84 0.58:0.04 0.010:0.001 Butter serum (refined) 1.92 23.96 0.50:0.07 0.010:0.001 aMilk from cows fed dieldrin (see Table 1, Milk Col- lection No. 3). bEach value is the average of duplicate prepara- tions. Each preparation consists of six individual milk collections and fractionations. CAverage and range of duplicate preparations. 85 Table 9b.—-Distribution of lipids and dieldrin in major fractions of naturally contaminated milk, pre— paration 3.a Amount of Amount of Amount of Fraction Fraction of Lipid in Dieldrin in Whole Milkb FractionC FractionC (%) (%) (%) Whole milk 100.00 100.00 100.00 Skim milk 93.60d 13.40 14.49 Cream (unwashed) 6.40 84.74 84.78 Cream (washed) 3.74 -- -- Butter 2.43 49.80 51.60 Butteroil 2.00 48.98 50.76 Buttermilk 1.31 0.54 0.40 (refined) Butter serum 0.27 0.13 0.08 (refined) aMilk from cows fed dieldrin (see Table 1, Milk Col- lection No. 3). b Each value is the average of duplicate prepara- tions. Each preparation is the average of six individual milk collections and fractionations. CEach value is the average of duplicate prepara- tions. Each preparation consists of pooled fractions of six individual milk collections and fractionations. dEstimated by the difference between whole milk and unwashed cream. 0 _ 0‘.-_-_—L———.— - WEI 86 comparable within a narrow range in the first five fractions on a fat basis, whereas the concentrations of residue in the refined buttermilk and butter serum were apparently lower than in the other fractions. These two fractions contained significantly less of total lipids than the corresponding fractions from the previous collections and the relative contents of phospholipids were increased. Once again, it shows that the lower dieldrin concentrations in these frac- tions were correlated with the higher phospholipid contents. In this series of milk collections and separations, the yield of each fraction was measured before it was fur- ther processed. Table 9b indicates the percentage distribu- tion of various fractions in milk, and the lipid and di- eldrin accounted for in each fraction. Although skim milk contained very little residue on a weight basis (Table 9a), it accounted for 93.5% of the total milk (by weight); therefore, it still accounted for a significant quantity of) dieldrin of whole milk (14.9%). On the other hand, cream accounted for only 6.4% of the total weight of milk, but it accounted for almost 85% of the total lipids of milk, and contained about 85% of the total residue of milk. From this table, it is again observed that the distribution of di- eldrin was generally parallel with the total lipid distribu- tion in milk. Losses of fraction weight, and losses of the absolute amounts of lipid and dieldrin could be due to the washing process of cream, although the concentration of w y l‘ ‘I"- 1.. . {'M“ ,. 87 dieldrin on a fat basis in the washed cream was not much different from that in the unwashed cream (Table 8). A preliminary test showed that freeze-drying had no significant effect on the dieldrin concentrations of whole milk, skim milk, or buttermilk (Table 10). Kroger (1968) also reported that the levels of dieldrin and heptachlor epoxide in butteroil were not affected by freeze-drying. Thus, the membrane materials obtained in this series of preparations were freeze-dried to render it feasible for the extraction of free lipids by petroleum ether; otherwise this solvent could not be quantitatively separated from the water solution of membrane material. Table 11 illustrates the lipid and dieldrin analy- ses in these dried membrane pellets. The relative amounts of membrane pellets collected in this study were somewhat different from those reported by Swope (1968). In Swope's results, almost 19% of the total amount of pellets were re— covered in the first fraction, 7,500 S pellet; and about 47.5% and 33.5% were recovered in the 230 S and 35 S pellet, respectively; whereas in the present study, the recovery of 7,500 S was only about 2.9% and recoveries of 230 S and 35 S were roughly 43.0 and 54.0%, respectively. This low yield of the 7,500 S pellet in the present research was unfavor— able to residue analysis, because the concentrations of di— eldrin in this series of milk collections were much lower than the earlier collections, and the total lipid content of this 7,500 S pellet was relatively low as compared to the am“ - - : 88 Table 10.--Comparison of the lipid and dieldrin contents in fresh and freeze—dried milk fractions.a % Lipid Extracted on Fraction Fresh Wt. Basis Fresh Freeze—Dried Dieldrin Concentration in Extd. Lipid, ppm Fresh Freeze-Dried Whole milk 3.67 3.89 Skim milk 0.46 0.43 Cream 26.16 28.99 Buttermilk 4.91 4.93 4.62:0.05 4.68:0.29 4.67:0.58 4.45:0.14 4.55:0.20 4.38:0.10 4.00:0.20 4.10:0.15 aMilk from cows fed dieldrin. Freeze-drying condi- tions were at 100 u Hg pressure for 24 hours. bAverage and range of duplicate determinations. 0 d l._:"-.-"C—I-I'.v {Ms 89 .0m+0mum00m00 000oau08 000000 850000000 000 00000 00300 .0000000 >0 000000000 .000000 0c5omn0m “00:00 850000000 >0 0000000X0 .000000 000mu0m0 .00000000800000 000000050 00 00000 000 00000>40 .00000000000 0:0 00000000000 x008 00500>00c0 x00 00 00000000 0008000050 000000 £00m .0008000050 000000 8000 0000000 no 000000000000 000000050 00 00000>0 030 m0 0500> £00mn .00 .02 8000000000 0002 .0 00000 0000 80000000 000 0300 8000 00020 II II II II II mh.hm 00.0 00.0m 00.000 00009 000.0Hmm0.0 m0.lom.o 00.0N 00.nm vm.v0 00.00 mm.m 00.0m m0.vm m mm 000.0Hmmo.o ~0.0Hmm.0 0m.mm mm.hm NN.N0 m0.~m m0.0 00.0w m0.mv m 0mm II II .II II II m0.m0 N0.m 0m.m Nm.N m comb 000000 08 08 0m 0m 09 0m 0.00 va 080m 80 000000 00000000 000000, .8000000800800 80000000 000000800080 0 :0 00000 0 0800080: n 0.m 00000000000 .0008 000000800000 000005000 00 0000000 00000808 0050000 000 000001000000 00 00000000 000 000000 00 00005000000011.00 0000B 90 other two pellets. Consequently, the absolute amount 0 dieldrin in this fraction was too little to be determin accurately. An increasing order of the free lipid contents observed from 7,500 S, 230 S to 35 S pellets , and the lipid contents followed a similar distribution pattern was in agreement with the observation of Swope (1968); though the proportions of total lipid contents in three fractions were slightly different. There were some pho lipids found in the free lipid fraction; it might be du the effect of freeze-drying, by which some of the lipop tein complexes were dissociated and the released phosph lipids were dispersed into the petroleum ether extract. total phospholipid contents in the total lipids of thes pellets were 22-23% approximately, these values were mu greater than refined buttermilk, from which these pelle were collected. This finding confirms the earlier post tion that some neutral fats were removed with the flota layer during the multi-centrifugation processes. Dield analyses were made using the combined lipids, since the separate recovery of residue in different fat fractions might not reflect the original deposition of such a pes f ed were total which al- spho- e to ro- o- The e ch ts ula- tion rin ti- cide chemical in the natural manner in membrane materials. Apparently, the 230 S and 35 S pellet contained compara concentrations of dieldrin on a fat basis, despite the ble dif- ference in their free and bound lipid ratios or the total lipid contents. In comparison of the dieldrin concentrations wwwr“*“' 91 of these pellets with the refined buttermilk fraction, it was again observed that a lower value was found in the pel- lets than in the buttermilk, and a higher phospholipid con- tent was associated with this lower level of dieldrin. Milk Containing Added Dieldrin Determinations were made on two independent prepara- tions of major milk fractions and their individual and aver- age results are shown in Table 12. In general, the distri- ‘awwmt‘m ““ -‘ bution pattern of lipids and added dieldrin were similar to that in the naturally contaminated milk. In this series of experiments, both the crude and refined buttermilk and both unwashed and washed cream were examined. Additionally, a portion of skim milk was also refined by centrifugation at 2,000 X G for 15 min to remove excess cream which was not separated during the first cream separation. The refined skim milk contained much lower lipid contents as compared to the original skim milk, while the phospholipid contents were relatively increased in the refined skim milk; although the direct data of phospholipid contents in the original skim milk were not available for these preparations, they can be assumed not much different from the skim milk with compara- ble total lipid contents in the previous preparations. The refined skim milk fractions from two preparations contained quite different total lipid contents as well as the relative phospholipid contents. This variation was due to the 92 i .Iql.1lul-1. {lum107. | .‘l t. 00.00 mm.0> om.m 0m0.00 mn.m0 0m.0 00.00 0 mm.00 00.00 mn.m m0m.00 mm.m0 mm.o mv.vm m 00.00 00.00 om.m mvm.00 om.m0 00.0 0m.mm 0 000000 m0.0n 00.00 mm.0 mmn.m o>.m0 NN.0 m0.mv 0 00.00 00.00 00.m mmm.m mm.m0 vm.o 00.nv m 00000020 m0.0n 0m.vm 00.0 mm>.m 00.00 00.0 vm.nv 0 80000 om.0m Nv.mm 00.00 m0v.v 00.N0 mm.0 00.0m 0 00.00 00.00 00.0 mvm.v mm.m0 0m.o m0.>m N 0000003050 mm.mm 00.00 00.00 omm.m 0N.m0 om.0 00.0m 0 80000 mm.m mo.m 00.00 m0o.0 00.00 mm.o0 00.0 0 vo.m 00.m 0N0.00 000.0 00.00 0m.0 m0.0 m 000000000 N0.0 mm.0 000.00 000.0 mo.m0 00.00 00.0 0 0008 8000 n0.m0 00.00 00.00 000.0 m0.m0 II 00.0 0 mm.v0 nm.v0 0m0.00 mmo.o ov.m0 II 00.0 m 00.0 00.0 000.00 mmo.o mm.m0 I: 00.0 0 0008 80x0 00.000 00.000 00.000 mom.o no.m0 mm.o nm.m 0 00.000 00.000 00.000 000.0 mm.m0 00.0 mm.m m 0008 00.000 00.000 00.000 000.0 nm.m0 mm.0 mw.m 0 00003 0 000 000 000 00000000 00 00 000 . x002 U0xm OZ 00000000 00000000 00000 000 0 . 00003 00 800 00000 00000000 00 00000000 00 00000 . 00000 00 00000 00 005080 0 005080 00000000 0000000000000 000 0 000 0 I000000 0 00 000000 00000000 0. .0 0 0 h 0.00000000 00000 0000000000 0008 300 00 000000000 0om08 00 00000000 000 000000 00 000050000000||.m0 0000B .80000 00000305 000 0008 00003 0003000 0000000000 000 m0 0000800mm0 .000000000000 030 00 0500> 00000>0 000 m0 0 .0000000000000 000 0000000000 0008 00500>0000 00 m0 00000000000 000m0 .m0m00 000003 00000 0 00 00>00 800 0.0 00 0008 00 00000 003 000000000 93 00.0 N0.o 00.0 Nmo.o 00.0 N0.NN 00.0 0 000000000 00.0 00.0 Nm.o 000.0 N0.0 00.0N 00.0 N 00000 00.0 00.0 00.0 000.0 00.0 00.0N NN.0 0 000000 00.0 00.0 mm.N 000.0 00.0 Nm.0 N0.0 0 00.0 MN.o 00.N m0N.o 0N.0 mm.N 00.N N 000000000 0m.o 00.0 00.0 000.0 00.0 0N.00 00.0 0 0000000000 00.0 mm.N mm.N m0m.o N0.m N0.o 00.0 0 00.0 0m.N 00.N 00N.o 00.0 00.0 00.0 N mo.N mm.m 00.0 0mm.o 00.0 00.0 00.0 0 0000000000 N0.00 00.00 N0.N 000.N0 00.N0 00.0 00.00 0 No.00 00.00 mN.N 00m.m0 0m.m0 00.0 00.000 N 00.00 00.00 00.0 000.N0 00.N0 00.0 00.00 0 000000000 0 000 000 000 00000000 000M0 000 .0 00000000 00000000 00.2 0 m 00000 000 0 z . . . 00003 00 8mm . . 00000 00000000 00 00000000 00 00000 . 00000 00 00000 0 0 005080 00 005080 00000000 0000000000000 000000000000 I00 000 0 00 000000 00000000 . . .00.00000 .N0 00000 94 techniques involved in the first cream separation as well as in the refining separation. Slightly higher levels of dieldrin were measured in the refined skim milk than in the original skim milk on a fat basis. Speculations on this aspect will be included later in this discussion. In comparison of the total recov- ery of lipid or dieldrin in washed and unwashed cream, an obvious loss was observed to be due to the cream washing process. But the losses in this series of preparations were not as great as encountered in the previous series of pre- parations. Comparable to the study of naturally incorporated dieldrin, the added dieldrin was also recovered essentially (BS-87%) in cream phase after separation. Beroza and Bowman (1966) reported a similar observation on recovery of other non—polar chlorinated pesticides. Although the original skim milk accounted for about lO-lS% of the total lipids as well as dieldrin of the total contents in milk, the refined skim milk contributed much less of these substances to the whole milk system (l-3%). Therefore, the absolute quantity Of dieldrin deposited in any fraction of milk varies sig- nificantly with the amount of lipid in that fraction. Com— Parisons can not be made between two corresponding fractions with different lipid contents. Relatively lower concentrations of dieldrin were found in buttermilk and butter serum than in other fractions, and refined buttermilk contained less amounts of residue 95 than the crude buttermilk on a fat basis. These findings were quite similar to those made with the naturally contaminated milk reported earlier. Analyses of the membrane pellets are given in Table 13. The relative amounts of pellets were slightly different from the previous preparation. The total lipid contents of these pellets were comparable to the previous findings, 1 _ while the phospholipid levels were somewhat higher. In con- sidering the levels of dieldrin on a fat basis, no conclu- ‘Q.i~.‘.; Ova. . - ‘ sion can be drawn for any definite distribution pattern for this pesticide among different pellets. Nevertheless, a lower value of dieldrin concentration was again found in these pellets as compared to the refined buttermilk fraction from which they were collected. The final supernatant of the membrane collection was also analyzed. It was observed that the final flotation layer was very loosely formed, and it tended to redisperse into the center clear layer with a slight disturbance, such as when the centrifuge tube caps were removed. Care was taken to avoid this disturbance to the floating materials as much as possible and an aliquot of the top layer was Pipetted out with a syringe. A significant amount of the center layer supernatant was included in this fraction. The clear center layer collected likewise contained some of the floating material, which was almost unavoidable. Analyses (Table 13) show that the top layer contained definitely higher concentration of residue than that in the center .coflumcHEkumU mco Eoum mm3 moam> sommo .mcoflumcfifinmump mumoHHQSU mo mmcmu can mmmuw>¢o .mcoflumummmum pampcmmmpcfi 03» mo xHHEkuusn pmaoom Eonm Umcfimuno mumB muwaammb .mwmmn unmme gmmum m :0 Hm>ma Ema m.o um xHHE ou Umbpm mm3 cHHUHmHom momma hooo.o ma¢.o NH.m mm.va no.m me.o flog Hmcflm momma mooo.o mvm.o hm.m Ah.va NN.h hoa.o kucmo Hmcflm % :oHusHom Hmuumz who Uflmmq .qum pama%wwmuoe Hmuuwwvhuo va ucmvmcummsm § . . cemm coflumuucmocou cflupamflo CH pHmHHosmmonm cw pagan kuumz who mgémfivé om.ommm.m 34m $6». 8.2: 1.58.. Nmo.onmw.H mH.oH¢N.m om.om mm.Hm mo.mm m mm omo.lowv.H 5N.0Hmm.v hv.mm mh.mN Nh.om m omN ovo.o+v~m.o mN.o+Hh.v mh.mm om.hH om.v m oomh i umfimm 3qu .33. vaaamflmpoe 2; 3: 33mm 08mm :oflumuucmocou cfinpamflo ca oflmflaosmmonm oflmflq cofluowum mcmunsz n b||’ 0.:flupawflp Umppm mcwcflmucoo xHHE 3mu mo mamaump ImE mcmuQEmE wasnon umm omfluolmnmmuw CH cfiuoamfio can moflmfla mo aoflusnauumwalt.ma magma 97 layer; although the residue levels in both solutions were substantially low. Control Milk Raw whole milk was fractionated in the same man- ner as the contaminated milk. The purpose of this series of studies was to provide a background information of resi- due levels in various fractions, and if reliable, to examine the distribution of dieldrin at this low level in milk fractions. In addition to the major fractions, pooled mem- QWVV—*“MWW?-—' brane pellets obtained form one process of centrifugation of refined buttermilk at the highest speed (same speed as for 35 S pellet), a clear center layer of supernatant, and the flotation layer were collected and analyzed. Most of these fractions showed practically no peaks in the GLC chromato- gram when their extracts were prepared at concentration levels comparable to those milk samples containing added di- eldrin. Therefore, no adjustment of residue concentrations was made for the fractions from previous preparations of milk containing added dieldrin. When these extracts of the control samples were concentrated by lO-fold, trace amounts of dieldrin were measured. Results are given in Table 14. It appeared that fractions containing little lipids such as the refined fractions of skim milk, buttermilk and butter serum, and the membrane pellets, contained also higher val— ues of residue on a fat basis. However, little reliability can be given to these results due to the fact that more 98 Table 14.--Distribution of lipids and dieldrin in various fractions of control milk. . . Phospholipid Dieldrin Concentration, F . Lipid in Total a raction (%) Li id ppm (g) Extd. Lipid Fraction Whole milk 3.80 0.06 0.063:0.021 0.0024:0.0008 l Skim milk 0.05 14.96 0.195:0.030 0.0001:0.00002 5" (refined) ( 3 Cream 30.78 0.27 0.053:0.012 0.0163:0.0037 2 (unwashed) Cream 46.54 0.24 0.055:0.010 0.0256:0.0047 ” (washed) Butter 83.26 0.21 0.060:0.015 0.0500:0.0125 Butteroil 98.80 0.01 0.059:0.010 0.0583:0.0099 Buttermilk 2.07 5.22 0.230:0.l60 0.0048:0.0033 (refined) Butter 2.03 17.15 0.183:0.082 0.0037:0.0017 serum (refined) Membrane 37.75C 26.79 o.220:o.039 0.083:0.0147C pelletsb Clear 8.07c 12.17 —— -- superna- tantb Flotation 45.74C 2.12 —— -- layer aAverage and range of duplicate determinations (apparent values). bFreeze-dried samples. COn a dry matter basis. 99 interference were observed on the chromatograms of these fractions as compared to the chromatograms of other frac- tions. Furthermore, the concentratiomsof'dieldrin on a fraction weight basis were very low in various fractions, especially in these low-fat fractions; and the concentration levels as shown in Table 14 probably were the limit of de— tection sensitivity and accuracy for dieldrin in each par- ticular fraction under the present experimental conditions. The total lipid and phospholipid contents of the supernatant and flotation layer were also given, although no attempt was made to determine the residue concentrations in these fractions of the control milk. Data showed that the flotation layer contained greater amounts of lipid than the pellets and consequently much lower phospholipid content than the pellets. This observation supports an earlier postulation that a significant amount of neutral fat was re- moved with the flotation layer during centrifugation, and also it justifies the findings that membrane pellet frac- tions contained higher levels of phospholipids than the re- fined buttermilk. General Discussion The relative concentrations of dieldrin in the total lipids of various fractions as reported in Tables 6, 7, 8, 9a, 11, 12, and 13 are compiled in Table 15 for a ready com- parison. Although the observed ratios of dieldrin levels in various fractions were varied from one preparation to Table 15.--Comparison of the distribution of dieldrin in various milk fractions. lOO Approximate Ratio of Dieldrin Concentration (fat basis) Fraction Naturally Contaminated Dieldrin Milk Added Milk (1) (2) (3) (1) (2) Whole milk 1.00 1.00 1.00 1.00 1.00 Skim milk 1.05 1.01 1.07 1.01 1.01 Skim milk -- -— —- 1.17 1.07 (refined) Cream 1.05 0.98 1.00 0.95 1.01 (unwashed) Cream (washed) -- 0.95 -- 0.94 1.01 Butter —- —— 1.01 0.96 1.00 Butteroil 1.08 1.03 1.07 0.98 1.01 Buttermilk 1.07 0.86 -- 0.57 0.71 Buttermilk -- 0.69 0.70 0.53 0.69 (refined) Butter serum 0.82 0.79 -- -- -- Butter serum -- -- 0.60 0.53 0.52 (refined) Membrane pellets 0.22 -- 0.33 0.30a (pooled) Final clear -- -- -- 0.26a supernatant Final top layer -- -- -- 0.39a ‘ aMembrane pellets milk of two preparations. were collected from pooled butter- lOl another, similarities were observed in that buttermilk and butter serum contained generally lower amounts of dieldrin than the fractions of whole milk, skim milk, cream, washed cream, butter and butteroil on a fat basis, while refined buttermilk, refined butter serum and membrane pellets con- tained even lesser amounts of residue. As pointed out earlier, most of these low-residue fractions contained high- er phospholipid contents. Privett gt a1. (1968), and Swope (1968) have provided excellent information of the high phospholipid contents in the plastic skim milk (equivalent to the buttermilk fraction in the present study) and serum (equivalent to butter serum), and membrane pellets, respec— tively. Thus an inverse relationship of the concentrations of dieldrin and phOSpholipid apparently exist. In a few instances in the present research, the phospholipid contents of buttermilk and butter serum were only slightly greater than in the whole milk, but their di— eldrin levels were still low. Therefore, other factors which are associated with the distribution of dieldrin in milk cannot be neglected. The buttermilk fraction was ob- tained by churning the washed cream at 12°C to room tempera- ture (ca. 24°C), and the butter serum was collected from the melting of butter at 45°C. For these reasons, it is very likely that some high-melting fats were separated into these fractions while the low-melting fats were gathered as butter and butteroil. Studies on the high-melting glycer- ides in fat globule membrane have been reviewed by Brunner 102 (1965). In this aspect, the low concentrations of dieldrin in the membrane fractions (buttermilk, butter serum and mem- brane pellets) may also be related with the high levels of high—melting glycerides in these fractions; in other words, this non-polar chlorinated pesticide residue appeared to be oriented into the low-melting glycerides more readily than into the high-melting glycerides. Additionally, the vari- ance of glyceride species, phospholipid compositions, and the fatty acid moieties among these fractions may also in- fluence the distribution pattern of dieldrin. A number of reports have emphasized the importance of rupture of the fat globule membrane in residue analysis (Moffitt, 1963; Kroger and Patton, 1967), based on the postulation that chlorinated pesticide residues are mostly deposited in the fat globules of milk. By physical separa- tion of milk, Langlois gt al. (1964b) and Li gt El, (1970) proved that those physiologically incorporated pesticides were significantly associated with the high-fat fractions, such as cream and butter. The present research confirms these findings, and further indicates that butteroil con- tained higher concentration of dieldrin than the butter on a weight basis. Additionally, the relative amounts of various fractions collected, and the relative amounts of lipid and dieldrin contained in these fractions are pre— sented. If butteroil may account for most of the lipid in milk, it would also account for most of the deposition of residue in milk; although the present separation technique “ftvasa'r‘fi “mu-“r11— _ ._ I. 103 did not provide a quantitative recovery of milk fat in the butteroil fraction. So far there has been little or no doubt that most of the naturally incorporated chlorinated residues as well as butteroil are within the fat globules in milk. On the studies of fortified milk, Beroza and Bowman (1966) reported that above 80% of the total quantity of added pesticides were recovered in the cream phase after separation. They thought that these chemicals were adsorbed onto the membrane of the fat globules rather than having penetrated into the fat body. Langlois 25.2l' (1964b) found that the added DDT and dieldrin were essentially recovered in butter as well as cream. One could still think that these compounds were adsorbed on the serum membrane of the butter phase. However, in the present research, evidence has been given that added dieldrin was mostly found in the butteroil fraction; similar to the distribution of the nat- urally incorporated dieldrin. Questions have again been raised at this point whether the added pesticide can pene- trate through the membrane, or that the recovery of residue in the butter and butteroil phases was only a consequence of re—orientation of this substance during the separation processes; and if the later Speculation is true, then would it be also true for the naturally incorporated pesticides? Although the answers are not yet available, nevertheless all the studies showed that a striking similarity is present between the distribution patterns of dieldrin in the milk 104 system with the different sources of contamination, and that dieldrin if; preferentially deposited into butteroil or neutral fat regardless of whether it is the original dis- tribution or is an altered orientation during separation. Hugunin and Bradley (1969) and Li gt gt. (1970) re- ported that in some instances the concentrations of pesti- cides were higher in the skim milk, buttermilk and high density membrane pellets, in some other instances these high concentrations were observed only for skim milk but not for buttermilk, and in other instances, this difference was not observed at all for the above fractions (such asiknrdieldrin, reported by Li gt gt.) as compared to the raw whole milk on a fat basis. They thought that the higher residue levels observed in skim milk and buttermilk were due to the asso- ciation of pesticides with the phospholipid or lipoprotein portions in these fractions. However, they did not report the fat contents of their milk products or fractions. From the experience in the present research, the fat content of a fraction may vary significantly with the separation tech- niques involved, especially for the lowefat containing frac- tions, such as skim milk and buttermilk. This variation may contribute significant differences in the residue concentra- tions for the low-fat fractions on a fat basis. The varia- tions of the distribution of the same pesticide in corres— ponding fractions among the data of Hugunin and Bradley and Li gt_gt. might be due to the variance in the lipid contents. 105 For the same reason, direct comparison cannot be made of their results with the present study. The skim milk fractions obtained in the present re- search generally contained rather high amounts of total lipids but relatively low amounts of phospholipids as com- pared to the commercial skim milk. Analyses showed that these skim milk contained similar concentrations of dieldrin as the whole milk or cream on a fat basis. When excess cream was further removed from the original skim milk, the total lipid content of the resulting refined skim milk was decreased significantly, whereas the relative phospholipid content was increased. Slightly higher residue concentra- tions on a fat basis were observed in refined skim milk fractions. One possible factor relating to the distribution of dieldrin might be the high amounts of protein in the re- fined skim milk rather than the high phospholipid contents in the total lipids of refined skim milk; because the pro- tein to lipid ratio in the refined skim milk is significant- ly high (for the skim milk containing 0.1% fat, this ratio is about 36:1; Bell and Whittier, 1965). It is possible that some dieldrin is distributed in the serum phase of milk and this quantity does not affect the overall concen- tration of dieldrin in the skim milk which has a relatively high fat content; but this small amount of residue does have a significant effect when the fat content is extremely low as in the refined skim milk. Although some reports showed that the solubility of dieldrin in water is 106 negligible, some other reports have indicated measurable solubilities at 0.25, 0.54, and 1.0 ppm levels at 25, 35 and 1., 1968). Park and Bruce 45°C, respectively (Gunther gt (1968) also reported that the maximum solubility of dieldrin in water was 0.186 ppm. The solubilities of non-polar pes- ticides in protein solutions are even higher. Moss and Hathway (1964) found that dieldrin was soluble at levels as high as 144-210 ppm in the blood serum of rabbits. Simi- larly, it is rational to postulate that the serum of milk may contribute to a non-negligible quantity of dieldrin, and this quantity is significantly important in the refined skim milk. Consequently, when the residue was extracted by sol- vent and reported on a fat basis, the resulted concentration would seem higher in the lipid of refined skim milk than in the original skim milk. Earlier reports indicating that skim milk contained higher levels of residues on a fat basis (Hugunin and Bradley, 1969; Li gt gt., 1970) was probably due to a simi- lar reason: extremely low fat content of their skim milk. Actually, the amounts of residues in skim milk are quite low on a weight basis, so that Langlois gt Q1. (1964b) could not detect these substances directly from the skim milk by the one-step method they used (see Literature Review). Therefore they could not report any level of residues on a fat.basis for their skim milk, while their whole milk and other fractions contained significant amounts of residues on a fat basis. 107 In considering the residue concentrations found in buttermilk, butter serum or their refined fractions and fat globule membrane pellets, tentative explanations postulated here are quite different from those for skim milk and re- fined skim milk. Firstly, the solubility of the protein moiety in buttermilk from thrice-washed cream is different “'30—— i a from those in skim milk (Swope, 1968). The solubilities of pesticides in these two systems of protein solutions or suspensions may be different. Secondly, the protein to ‘7’ Ivan .I-J-‘<.'-‘l~.‘ 0-. . .pa‘Q lipid ratios in these buttermilk and butter serum (approxi- mately 0.72:1.00 to 0.84:1.00, and 0.18:1.00 to 0.20:1.00, respectively, reported by Swope and Brunner, 1968) are much lower than the ratios in skim milk. Thus, the influence of the lipid moiety on the distribution of residue would be greater in a buttermilk or butter serum system than in skim milk; similarly, the protein moiety would be more responsible on the deposition of pesticide in skim milk than in butter- milk. Thirdly, it has been noted that the lipid moiety of buttermilk was very difficult to be completely extracted with the solvent mixture of ethanol, ethyl ether and petro- leum ether as used for whole milk, while the lipid in skim milk was readily extracted with these solvents. A revised procedure using higher proportions of ethanol and petroleum ether was found to be necessary in order to denature the lipoprotein complex of buttermilk or butter serum and to recover a greater quantity of lipid. If only a portion of the lipid is extracted, the subsequent analysis of residue 108 based on that portion of lipid would have a higher concen- tration of pesticide, as discussed earlier. For the above reasons, it is apparently suggested that the factors influencing the distribution of dieldrin in skim milk and buttermilk are different. The present re— sults demonstrated that the lower levels of residue in but— termilk and butter serum were probably related to the high levels of phospholipid content or high-melting glyceride content or both in these fractions. Whereas the protein moiety seemed to have played an important role in contribut- ing the residue concentrations when expressed on a fat basis in the extremely low fat fraction, such as refined skim milk. The variations of the residue levels in buttermilk among the data of earlier workers (Hugunin and Bradley, 1969; Li gt gt., 1970; Langlois gt gt., 1964b) and between their data and the present results might be due to the dif- ferences in the separation techniques applied, the fat con- tents of buttermilk fractions, and the amounts of total lipid extracted from the buttermilk. One difference in the separation techniques used was that they used pasteurized unwashed cream to make buttermilk (Li gt gt., 1970; Langlois gt_gl., 1964b) while a raw and thrice-washed cream was used in the present study. The composition of those buttermilk fractions might be quite differently, and thus influenced the distribution of residues differently. {WT—“W" *-“I.'.—-' l SUMMARY AND CONCLUS I ONS Various methods were compared for the extraction of di- “-1—— eldrin in milk. A method of the Food and Drug Adminis- tration (FDA) and the Mojonnier method provided a com- a.“— .— parable extraction efficiency for lipid as well as for CL...._“ a H1'. ‘4‘? pesticide. The surfactant method recovered less fat from milk, and slightly lower levels of residue in the isolated fat. A mild alkaline hydrolysis technique was comparable with the FDA method for residue analysis, but needed an independent determination of fat content in order to express the residue concentrations to be on a fat basis. A strong alkaline hydrolysis technique pro- vided lowest recovery of dieldrin in milk. Differential solvent extraction of milk showed that the first fraction, that extracted with non-polar hydrocar- bon solvents, contained much higher concentrations of dieldrin than the second fraction, which was subse— quently extracted with ethanol and non-polar solvents. It appeared that dieldrin was more readily extracted with hydrocarbon solvents than were the milk lipids. However, the first fat fraction only accounted for about one-half of the total quantity of dieldrin 109 110 in milk. This incomplete recovery of dieldrin might be due to the low recovery of solvents used for this ex- traction or due to the same factor that depressed the recovery of solvents from milk. The lipid—depleted por- tions essentially contained no residue. Milk containing physiologically incorporated or added dieldrin was separated into various fractions by physi- cal methods. Generally, the distribution pattern of this pesticide was quite similar in the milk system be— tween the two sources of contamination. This observa- tion has led to a proposal that the results of experi- ments involving dieldrin in milk from either source might be the same. The relative amounts of dieldrin in various fractions closely resemble the relative amounts of total lipid in these fractions and both of them are significantly varied with the separation techniques used. Unwashed cream accounted for approximately 85-87% of the total amount of pesticide in milk, and skim milk accounted for about 10-15%, while refined skim milk accounted for only 1-3% of the total residue in milk. Relative recoveries of dieldrin in other fractions were: washed cream, 65— 80%; butter, 52-78%; butteroil, 51—78%; buttermilk, 0.4— 2.0%; and butter serum, 0.05-0.08%. Comparable levels of dieldrin on a fat basis were found in whole milk, skim milk, cream, washed cream, butter and butteroil. Lower levels of residue were observed 111 in buttermilk and butter serum, and even lower values were found in the refined buttermilk and butter serum, and in membrane pellets; although no consistent distri- bution pattern was observed among the three pellet frac- tions of 7,500 S, 230 S and 35 S. An inverse correlation of the residue levels with the levels of phospholipid or high—melting glyceride apparently exists. The slightly [h higher concentrations of dieldrin on a fat basis ob— . served in refined skim milk might be due to the slight ; 4 solubility of this pesticide in the serum of milk. From E another point of view, the residue concentrations on a fat basis as well as on a weight basis for low-fat frac— tions are significantly related to the total fat content of these fractions. The recovery of added dieldrin in butteroil might be due to the re—orientation during separation processes. Similarly, the low concentrations of dieldrin in butter- milk or butter serum may not necessarily prove that such a substance is less preferentially deposited in these fractions in the natural original milk system. Nevertheless, it showed that dieldrin has a tendency to be distributed more favorably with the neutral lipids or free fats in milk. Control milk was also fractionated and analyzed. How- ever, the quantities of residue in these samples were too low to be determined accurately. Although skim milk, buttermilk and butter serum appeared to contain 112 higher levels of residue on a fat basis as compared to other fractions, they contained extremely low concentra- tions of residue on a fraction weight basis. In addi- tion, they exhibited more interference on the GLC chromatogram. Therefore, no firm conclusions on the distribution of dieldrin can be drawn from the analysis of control milk fractions. L1? "I"? ' PROPOSALS FOR FURTHER RESEARCH 1. Determination of dieldrin metabolite in milk. The metabolites of DDT, such as DDE and DDD, have been found in milk, and their relative distribution in vari- ous phases during the manufacture of cheese have been re- ported. However, no study has shown the presence of a metabolite of dieldrin in milk. Reports concerning the identification or isolation of some dieldrin metabolites were from samples of urine, or feces of rats, rabbits or sheep. It would be desirable to determine the possible metabolites in a milk system, and furthermore to determine their distribution in various fractions (which resembles various milk products); because some metabolites are more toxic while others are less toxic than the parent compound. Some metabolites as reported are relatively hydrophilic; the excretion of these substances in the urine is favorable to the health aspect, while the possible excretion of them in the aqueous phase of milk would be unfavorable. There- fore, studies should also include the analysis of any pos- sible hydrophilic metabolites and a different procedure may be necessary for the extraction and determination of these 113 ‘Kgl‘wz'if‘ w-I “—n—c-a n.- 114 substances. The use of radioactive labeled pesticide chem— icals would have advantages in these studies. 2. Effect of enzyme systems on the pesticide residues in milk, and vice versa. Milk is known to contain a number of enzymes, such as lipase, esterase, phosphatase, xanthine oxidase, and lactoperoxidase. Studies have shown that some liver enzymes have some effects on DDT and dieldrin; similarly the milk enzymes may also affect more or less on pesticide residues, and consequently they (enzymes) may influence the final quality and quantity of residues in the finished milk products processed and stored differently. Some pesticides are reported to have an effect on natural enzyme activities in birds, rats and rabbits. The effect could be either an enhancing or inhibitory mechanism. In any case, it would be of value to know whether pesticide chemicals may influence the milk enzyme systems, because these may have a great effect on the quality of milk and milk products. BIBLIOGRAPHY *«r-yw—‘rr't‘er-r—E 1 Ill-ll 'llll.llll 1’! 11.: II'., (II. | BIBLIOGRAPHY Abbott, D. C., R. Goulding, and J. O'G. Tatton. 1968. Organochlorine pesticide residues in human fat in Great Britain. Brit. Med. J. g, 146. Abbott, D. C., D. C. Holmes, and J. O'G. Tatton. 1969. Pesticide residues in the total diet in England and Wales, 1966-1967. II. Organochlorine pesticide residues in the total diet. J. Sci. Food Agr. g9, 245. Abbott, D. C. and J. Thomson. 1965. The application of thin-layer chromatographic techniques to the analy— sis of pesticide residues. Residue Rev. it, 1. Abramson, D. and M. Blecher. 1965. The lipid composition of normal rat thymus. Biochim. Biophys. Acta. fig, 117. Alary, J. G., J. Brodeur, M. G. Cote, J. C. Panisset, P. Lamothe, and P. Guay. 1968. The effect of a pre— treatment with phenobarbital on the metabolism of DDT in dairy cows and young bulls. Rev. Can. Biol. 27, 269. Anderson, J. M. and M. R. Peterson. 1969. DDT: Sublethal effects on brook trout nervous system. Science. 164, 440. Anonymous. 1969. Details on the government curb on uSes of DDT and other hard pesticides have emerged. Chemical and Engineer News, Dec. 1. page 27. App, B. A., R. H. Carter, and R. E. Ely. 1956. Residues on forage, in the soil and in milk following pasture treatment with granulated dieldrin. J. Econ. Entomol. 39, 136. Bell, R. W. and E. O. Whittier. 1965. The composition of milk products. In Fundamentals of Dairy Chemistry, eds. B. H. Webb and A. H. Johnson. Avi Publishing Co., Inc., Westport, Connecticut, page 41. 115 116 Beroza, M. and M. C. Bowman. 1966. Correlation of pesti- cide polarities with efficiencies of milk extraction procedures. J. Assoc. Off. Anal. Chemists. 49, 1007. Bills, D. D. and J. L. Sloan. 1967. Removal of chlorinated insecticide residues from milk fat by molecular dis- tillation. J. Agr. Food Chem. lé' 676. Bitman, J., H. C. Cecil, S. J. Harris, and G. F. Fries. 1969. DDT indUces a decrease in eggshell calcium. Nature. 224, 44. L Blinn, R. C. 1965. Infrared techniques useful in residue "- chemistry. J. Assoc. Off. Agr. Chemists. it, 1009. Blinn, R. C. and F. A. Gunther. 1963. Rapid colorimetric ‘ determination of DDT in milk and butter. J. Assoc. ? Off. Agr. Chemists. 4g, 191. ' Bosin, W. A. 1963. Analysis of pesticide residues using microcoulometric temperature-programmed gas chroma- tography. Anal. Chem. lg, 833. Boyd, E. M., and J. Stefec. 1969. Dietary protein and pes- ticide toxicity: With particular reference to en- drin. Can. Med. Assoc. J. t, 335. Bradley, R. L., Jr. and C. F. Li. 1968. The effect of dieldrin on acid development during manufacture of Cheddar cheese. J. Milk Food Technol. it, 202. Braund, D. G., L. D. Brown, J. T. Huber, N. C. Leeling, and M. J. Zabik. 1969. Excretion and storage of diel- drin in dairy cows fed thyroprotein and different levels of energy. J. Dairy Sci. 23, 172. Braund, D. G., L. D. Brown, N. C. Leeling, and M. J. Zabik. 1966. Effectsof dietary energy level and thyropro- tein on excretion and storage of dieldrin in dairy cows. J. Dairy Sci. 49, 740. Brooks, G. T. 1969. The metabolism of diene-organochlorine (cyclodiene) insecticides. Residue Rev. 31, 81. Brown, J. R. Organo—chlorine pesticide residues in human depot fat. 1967. Can. Med. Assoc. J. 91, 367. Brunner, J. R. 1965. Physical equilibria in milk: The lipid phase. In Fundamentals of Daity Chemistry, eds. B. H. Webb and A. H. Johnson. Avi Publishing Co., Inc., Westport, Connecticut. page 403. 117 Burchfield, H. P. and R. J. Wheeler. 1966. Analysis by microcoulometric gas chromatography. J. Assoc. Off. Anal. Chemists. 49, 651. Burke, J. A. 1965. Gas chromatography for pesticide resi— due analysis: Some practical aspects. J. Assoc. Off. Agr. Chemists. 4t, 1037. Burke, J. A. and Laura Giuffrida. 1964. Investigations of electron capture gas chromatography for the analysis of multiple chlorinated pesticide residues in vege- tables. J. Assoc. Off. Agr. Chemists. 41, 326. L . Campbell, J. E., L. A. Richardson, and M. L. Schafer. 1965. Insecticide residues in the human diet. Arch. En— viron. Health. $2, 831. Carr, R. L. 1970. Collaborative study of the Mills method E for multiple chlorinated pesticide residues in but— ' terfat. J. Assoc. Off. Anal. Chemists. fig, 152. Carter, R. H. 1947. Estimation of DDT in milk by determin- ation of organic chlorine. Anal. Chem. £9, 54. Carter, R. H. and H. D. Mann. 1949. DDT content of milk from a cow sprayed with DDT. J. Econ. Entomol. 43, 708. Casarett, L. J., G. C. Fryer, W. L. Yanger, Jr., and H. W. Klemmer. 1968. Organochlorine pesticide residues in human tissue--Hawaii. Arch. Environ. Health. 11, 306. Casterline, J. L., Jr., C. H. Williams, and J. E. Keys. 1969. Effect of pesticide administration on ester- ase activities in serum and tissues of rats fed variable casein diets. Toxicol. Appl. Pharmacol. 14, 266. Cerbulis, J. 1967. Distributions of lipids in various fractions of cow's milk. J. Agr. Food Chem. ii, 784. "Changes in Method," 1966. J. Assoc. Anal. Chemists. 39, 222. Chien, H. C. and T. Richardson. 1967. Gross structure of the fat globule membrane of cow's milk. J. Dairy Sci. fig, 451. Claborn, H. V., R. C. Bushland, H. D. Mann, M. C. Ivey, and R. D. Radeleff. 1960. Meat and milk residues from livestock sprays. J. Agr. Food Chem. 8, 439. 118 Clifford, P. A., J. L. Bassen, and P. A. Mills. 1959. Chlorinated organic pesticide residues in fluid milk. Public Health Rep. 18, 1109. Cole, H., D. Barry and D. E. H. Frear. 1966. DDT contamin- ation of feed grains and forages in Pennsylvania. Bull. Environ. Contam. Toxicol. t, 212. Cook, J. W. 1957. Action of rumen fluid on pesticides: In vitro destruction of some organophosphate pesti- cides by bovine rumen fluid. J. Agr. Food Chem. 8, 859. Cook, R. M., J. W. Thomas, R. S. Emery, and J. T. Huber. 1967. Dieldrin recycling from the blood to the gut in ruminants. J. Anim. Sci. 88, 1486. Corneliussen, P. E. 1969. Residues in food and feed: Pesticide residues in total diet samples (IV). Pest. Moni. J. 8, 140. Crosby, D. G. and T. E. Archer. 1966. A rapid analytical method for persistent pesticides in proteinaceous samples. Bull. Environ. Contam. Toxicol. t, 16. Cueto, C., Jr. 1960. Colorimetric determination of diel- drin and its application to animal fat. J. Agr. Food Chem. 8, 273. Dale, W. E., F. M. Copeland, and W. J. Hayes, Jr. 1965. Chlorinated insecticides in body fat of people in India. Bull. Wld. Hlth. Org. 88, 471. Damico, J. N., J. Y. T. Chen, Catherine E. Costello, and E. O. Haenni. 1968. Structure of Klein's metabolites of aldrin and of dieldrin. J. Assoc. Off. Anal. Chemists. 8t, 48. Datta, P. R., E. P. Laug, J. 0. Watts, A. K. Klein, and M. J. Nelson. 1965. Metabolites in urine of rats on diets containing aldrin or dieldrin. Nature. 208, 289. deFaubert Maunder, M. J., H. Egan, E. W. Godly, E. W. Ham- mond, J. Roburn, and J. Thomson. 1964. Clean-up of animal fats and dairy products for the analysis of chlorinated pesticide residues. Analyst. 88, 168. Deichmann, W. B. and J. L. Radomski. 1968. Retention of pesticides in human adipose tissue--preliminary re- port. Ind. Med. Surg. 81, 218. 119 Duggan, R. E. 1967. Residues in food and feed: Chlorin- ated pesticide residues in fluid milk and other dairy products in the United States. Pest. Moni. J. l, 2. Duggan, R. E. 1968. Residues in food and feed: Pesticide residues in vegetable oil seeds, oils, and by products. Pest. Moni. J. t, 2. Duggan, R. E., H. C. Barry, and L. Y. Johnson. 1966. Pes— ticide residues in total-diet samples: Ready to consumed food samples (I). Science. 18;, 101. -1A Duggan, R. E., H. C. Barry, and L. Y. Johnson. 1967. Pes- ticide residues in total-diet samples (II). Pest. Moni. J. t, 2. . Emu—”7"“, 3+ ; Duggan, R. E. and K. Dawson. 1967. Pesticide. A report on E I residues in food. FDA Pap. l, 4. Duggan, R. E. and G. Q. Lipscomb. 1969. Dietary intake of pesticide chemicals in the United States (II). Pest. Moni. J. 8, 153. Duggan, R. E. and J. R. Weatherwax. 1967. Dietary intake of pesticide chemicals. Science. 157, 1006. Durham, W. F. 1963. Pesticide residues in foods in rela- tion to human health. Residue Rev. 3, 33. Eidelman, M. 1962. Determination of microquantities of chlorinated organic pesticide residues in butter. J. Assoc. Off. Agr. Chemists. 48, 672. Eidelman, M. 1963. Determination of microquantities of some chlorinated organic pesticide residues in edible fats and oils. J. Assoc. Off. Agr. Chemists. 46, 182. Eidelman, M. 1967. Analysis of chlorinated pesticide resi- dues in fats and oil, utilizing dimethyl sulfoxide: A collaborative study. J. Assoc. Off. Anal. Chem- ists. 88, 591. Elkins, E. R., F. C. Lamb, R. P. Farrow, R. W. Cook, Mar- garet Kawai, and J. R. Kimball. 1968. Removal of DDT, malathion and carbaryl from green beans by com- mercial and home preparative procedures. J. Agr. Food Chem. t8, 962. 120 Farrow, R. P., F. C. Lamb, R. W. Cook, J. R. Kimball, and E. R. Elkins. 1968. Removal of DDT, malathion and carbaryl from tomatoes by commercial and home pre- parative methods. J. Agr. Food Chem. 88, 65. Feil, V. J., R. D. Hedde, R. G. Zaylskie, and C. H. Zachrison. 1970. Dieldrin-Cl4 metabolism in sheep. J. Agr. Food Chem. 88, 120. Folch, J., M. Lees, and G. H. Sloane-Stanley. 1957. A sim- ple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 888, 497. Food and Drug Administration (FDA). 1968. Pesticide Anal- ytical Manual. Vol. I. 2nd edition. Bureau of Regulatory Compliance, Food and Drug Administration, U.S. Dept. of Health, Education, and Welfare, Wash- ington, D.C. Fries, G. F. 1970. Organochlorine pesticides and the dairy industry. J. Dairy Sci. 88, 367. Fries, G. F., W. P. Flatt, and L. A. Moore. 1969. Energy balance and excretion of DDT into milk. J. Dairy Sci. 88, 684. Furman, W. D. and N. V. Fehringer. 1967. Study of modifi- cation of the Mills procedure for chlorinated pes- ticides in dairy products. J. Assoc. Off. Anal. Chemists. 88, 903. Gannon, N. and G. C. Decker. 1960. The excretion of di- eldrin, DDT, and heptachlor epoxide in milk of dairy cows fed on pastures treated with dieldrin, DDT and heptachlor. J. Econ. Entomol. 88, 411. Gannon, N., R. P. Link, and G. C. Decker. 1959a. Storage of dieldrin in tissues and its excretion in milk of dairy cows fed dieldrin in their diets. J. Agr. Food Chem. 8, 824. Gannon, N., R. P. Link, and G. C. Decker. 1959b. Insecti- cide residues in the milk of dairy cows fed insecti- cides in their daily ration. J. Agr. Food Chem. 8, 829. Gaul, Jean A. 1966. Quantitative calculation of gas chroma- tographic peaks in pesticide residue analyses. J. Assoc. Off. Anal. Chemists. 38, 389. 121 Giuffrida, Laura, D. C. Bostwick, and N. F. Ives. 1966. Rapid cleanup techniques for chlorinated pesticide residues in milk, fats and oils. J. Assoc. Off. Anal. Chemists. 88, 634. Giuffrida, Laura, and N. F. Ives. 1969. Microcoulometric gas chromatography: Improvements instrumentation. J. Assoc. Off. Anal. Chemists. 88, 541. Gudzinowicz, B. J. 1967. Gas Chromatggraphic Ana1ysis of Drugs and Pesticides. Edward Arnold (Publishers) Ltd., London, and Marcel Dekker, Inc., New York. Gunther, F. A. 1964. Symposium on instrumentation for the detection and determination of pesticides and their residues in foods. Residue Rev. Vol. 5,Special volume, cal. F. A. Gunther. Academic Press Inc., Publishers, New York. Gunther, F. A., W. E. Westlake, and P. S. Jaglan. 1968. Reported solubilities of 738 pesticide chemicals in water. Residue Rev. 88, 1. Haenni, E. O., J. W. Howard, and F. L. Joe, Jr. 1962. Di- methyl sulfoxide: A superior analytical extraction solvent for polynuclear hydrocarbons and for some highly chlorinated hydrocarbons. J. Assoc. Off. Agr. Chemists. 88, 67. Hardee, D. D., W. H. Gutenmann, D. J. Lisk, G. G. Gyrisco, and C. M. Edmonds. 1964. Zonal accumulation of dieldrin in soil and alfalfa residues. J. Econ. Entomol. 81, 583. Hayes, W. J., Jr. 1965. Review of the metabolism of chlo- rinated hydrocarbon insecticides especially in mam- mals. Ann. Rev. Pharmacol. 8, 27. Hayes, W. J., Jr. and A. Curley. 1968. Storage and excre- tion of dieldrin and related compounds. Arch. En- viron. Health. 18, 155. Hayes, W. J., Jr., W. F. Durham, and D. Cueto, Jr. 1956. The effect of knownrepeated oral doses of chloro- phenothane (DDT) in man. J. Amer. Med. Ass. 162, 890. Hedde, R. D., K. L. Davison, and J. D. Robbins. 1970. Di— eldrin-Cl4 metabolism in sheep: Distribution and isolation of urinary metabolites. J. Agr. Food Chem. 18, 116. 122 Heineman, H. E. O. and C. B. Miller. 1961. Pesticide re— sidues and the dairy industry. J. Dairy Sci. 88, 1775. Henderson, J. L. 1963. Comparison of laboratory techniques for the determination of pesticide residues in milk. J. Assoc. Off. Agr. Chemists. 88, 209. Henderson, J. L. 1965. Insecticide residues in milk and dairy products. Residue Rev. 8, 74. Herrmann, R. and B. Gutsche. 1969. Improvement of detec- tion limits of flame-spectrophotometric determina- ?"‘ tion of chlorinated organic pesticide residues. Analyst. 83, 1033. 3 chromatography, with a rapid spotting technique, for one-stage cleanup in pesticide residue analysis. J. Assoc. Off. Anal. Chemists. 88, 146. Hetherington, R. M. and C. Parouchais. 1970. Channel-layer E Hoffman, w. 5., W. I. Fishbein, and M. B. Andelman. 1964. Pesticide content of human fat tissue. Arch. En- viron. Health. 8, 387. Hugunin, A. G. and R. L. Bradley, Jr. 1969. Location of organochlorine pesticides in milk. J. Dairy Sci. 88, 887 (Abstract), and The 64th Annual Meeting of American Dairy Science Association. June 22-25, 1969. Minneapolis, Minn. Bulletin No. M-41. Hunter, C. G. and J. Robinson. 1967. Pharmacodynamics of dieldrin (HEOD). I. Ingestion of HEOD by human subjects for 18 months. Arch. Environ. Health. 18, 614. Hunter, C. G. and J. Robinson. 1968. Aldrin, dieldrin, and man. Fd. Cosmet. Toxicol. 8, 253. Hunter, C. G., J. Robinson, and K. W. Jager. 1967. Aldrin and dieldrin--The safety of present exposures of the general populations of the United Kingdom and the United States. Fd. Cosmet. Toxicol. 8, 781. Hunter, C. G., J. Robinson, and M. Roberts. 1969. Pharma- codynamics of dieldrin (HEOD). Ingestion by human subjects for 18 to 24 months. Arch. Environ. Health. 18, 12. Ivey; M. C., H. V. Claborn, H. D. Mann, R. D. Radeleff, and G. T. Woodard. 1961. Aldrin and dieldrin content of body tissues of livestock receiving aldrin in their diet. J. Agr. Food Chem. 8, 374. Johnsen, Johnson, Johnson, Kadoum, 123 R. E., P. A. Dahm, M. L. Fairchild, and A. E. Free- man. 1961. Endrin residues on corn stover in rela— tion to dairy cattle feeding. J. Econ. Entomol. 88, 398. B. J. and M. D. Jullum. 1969. Effect of pesti- cides on chemical composition of soybean seed (Glycine max.). Agron. J. 81, 379. L. Y. 1965. Collaborative study of a method for multiple chlorinated pesticide residues in fatty foods. J. Assoc. Off. Agr. Chemists. 88, 668. A. M. 1968. Modifications of the micromethod of sample cleanup for thin-layer and gas-chromatograph- ic separation and determination of common organic pesticide residues. Bull. Environ. Contam. Toxicol. 3, 354. Kiermeier, F. 1968. Pesticies (residues in milk). Annu. Bull., Int. Dairy Fed. Part 5. page 44. Kiermeier, F., M. Rudolphi, and G. Wildbrett. 1967. Be- Kim, S. Kim, S. Kim, S. King, N. havior of chlorinated hydrocarbon insecticide resi- dues during freeze-drying of milk. Z. Lebensm.- Unters, Forsch. 188, 78. In Chemical Abstract. 81, 10443u. C. 1969. Relationship between some insecticides and lactic culture organisms in milk. Ph.D. Thesis. Michigan State University. C. and L. G. Harmon. 1968. Effect of insecticide residues on growth and fermentation ability of lac- tic culture organisms. J. Milk Food Technol. 81, 97. C. and L. G. Harmon. 1970. Relationship between some chlorinated hydrocarbon insecticides and lactic culture organisms in milk. J. Dairy Sci. 88, 155. 1955. The Milk Fat Globule Membrane and Some Associated Phenomena. Commonwealth Agricultural Bureaux, Farnham Royal, Bucks, England. Klein, A. K., J. D. Link, and N. F. Ives. 1968. Isolation KOCh I R- and purification of metabolites found in the urine of male rats fed aldrin and dieldrin. J. Assoc. Off. Anal. Chemists. 81, 895. B. 1969. Chlorinated hydrocarbon insecticides: Inhibition of rabbit brain ATPase activities. J. Neurochem. 18, 269. W'““‘““‘ “I 124 Korte, F. and H. Arent. 1965. Metabolism of Insecticides. IX (I). Isolation and identification of dieldrin metabolites from urine 2f rabbits after oral admin- istration of dieldrin-l C. Life Sci. 8, 2017. Kovacs, M. F., Jr. 1963. Thin-layer chromatography for chlorinated pesticide residue analysis. J. Assoc. Off. Agr. Chemists. 88, 884. Kroger, M. 1966. Observationson the analysis of chlorin- ated hydrocarbon pesticides in milk. Intern. Dairy Congr. Proc. 17th. Munchen. Section A. page 467. Kroger, M. 1968. Effect of various physical treatments on certain organochlorine hydrocarbon insecticides found in milk fat. J. Dairy Sci. 81, 196. Kroger, M. and S. Patton. 1967. Comparison of two milk fat extraction methods for pesticide residue analysis. J. Dairy Sci. 88} 324. Laben, R. C. 1968. DDT contamination of feed and residues in milk. J. Anim. Sci. 81, 1643. Laben, R. C., T. E. Archer, D. G. Crosby, and S. A. People. 1966. Milk contamination from low levels of DDT in dairy rations. J. Dairy Sci. 88, 1488. Lamb, F. C., R. P. Farrow, E. R. Elkins, J. R. Kimball, and R. W. Cook. 1968. Removal of DDT, parathion, and carbaryl from spinach by commercial and home pre- parative methods. J. Agr. Food Chem. 18, 967. Lampert, L. M. 1964. Rapid separation of fat for pesticide residue analysis of milk products. J. Dairy Sci. 81, 1013. Langlois, B. E., B. J. Liska, and D. L. Hill. 1964b. The effects of processing and storage of dairy products on chlorinated insecticide residues. I. DDT and lindane. J. Milk Food Technol. 81, 264. Langlois, B. E., B. J. Liska, and D. L. Hill. 1965. The effects of processing and storage of dairy products on chlorinated insecticide residues. II. Endrin, dieldrin and heptachlor. J. Milk Food Technol. 88, 9. Langlois, B. E., A. R. Stemp, and B. J. Liska. 1963. Milk fat from the Babcock test for insecticide residue analysis in dairy products. J. Dairy Sci. 88, 854. 125 Langlois, B. E., A. R. Stemp, and B. J. Liska. 1964a. Rapid cheanup of dairy products for analysis of Chlorinated insecticide residue by electron capture gas chromatography. J. Agr. Food Chem. 18, 243. Lawrence, J. H. and J. A. Burke. 1969. Ten methods for the analysis of milk for residues of chlorinated pesti- cides. J. Assoc. Off. Anal. Chemists 88, 817. Ledford, R. A. and J. H. Chen. 1969. Degradation of DDT and DDE by cheese microorganisms. J. Food Sci. 88, 386. Ledford, R. A., J. H. Chen, and w. F. Shipe. 1968. Effect r-W of direct steam heating and vacuum treatment on levels of pesticide residues in milk. J. Dairy Sci. 51, 219. Lee, M., K. Harris, and H. Trowbridge. 1964. Effect of the level of dietary protein on the toxicity of dieldrin for the laboratory rat. J. Nutr. 88, 136. Li, C. F. and R. L. Bradley, Jr. 1969. Degradation of chlo- rinated hydrocarbon pesticides in milk and butteroil by ultraviolet energy. J. Dairy Sci. 88, 27. Li, C. F., R. L. Bradley, Jr., and T. Richardson. 1968. Improved extraction of chlorinated hydrocarbon pes- ticides from animal tissues. Bull. Environ. Contam. Toxicol. 8, 317. Li, C. F., R. L. Bradley, Jr., and L. H. Schultz. 1970. Fate of organochlorine pesticides during processing of milk into dairy products. J. Assoc. Off. Anal. Chemists 88, 127. Lipscomb, G. Q. 1968. Residues in food and feed: Pesti- cide residues in prepared baby foods in the United States. Pest. Moni. J. 8, 104. Lisk, D. J. 1966. Detection and measurement of pesticide residues. Science. 154, 93. Liska, B. J. 1968. Effects of processing on pesticide residues in milk. J. Anim. Sci. 81, 827. Ludwig, G., J. Weis, and F. Korte. 1964. Excretion and distribution of aldrin-Cl4 and its metabolites after oral administration for a long period of time. Life Sci. 8, 123. quke Malor Man} Mar Ma] 14a: 14a Ma 126 Lykken, L. and J. E. Casida. 1969. Metabolism of organic insecticide chemicals. Can. Med. Ass. J. 100, 145. Malone, B. and J. A. Burke. 1969. Study of the sweep co- distillation cleanup for chlorinated pesticide resi- dues in edible fats and oils. J. Assoc. Off. Anal. Chemists. 88, 790. Mann, H. D., R. H. Carter, and R. E. Ely. 1950. The DDT content of milk products. J. Milk Food Technol. 18, 340. Marth, E. H. 1962. Chlorinated hydrocarbons deposited in “4 biological materials. II. Animals and animal F' products. J. Milk Food Technol. 88, 72. Marth, E. H. 1965. Residues and some effects of chlorin- ated hydrocarbon insecticides in biological mate- i rial. Residue Rev. 8, l. i Marth, E. H. and B. E. Ellickson. 1959a. Insecticide resi- dues in milk and milk products. I. Insecticide residues in milk from treatment of dairy cows and barns. J. Milk Food Technol. 88, 112. Marth, E. H. and B. E. Ellickson. 1959b. Insecticide resi- dues in milk and milk products. II. Insecticide residues in milk from dairy cattle fed treated crops. J. Milk Food Technol. 188, 9. Marth, E. H. and B. E. Ellickson. 1959c. Insecticide re- sidues in milk and milk products. III. Insecticide residues in dairy products and associated problem. J. Milk Food Technol. 179, 182. Martin, R. J. and R. E. Duggan. 1968. Pesticide residues in total diet samples (III). Pest. Moni. J. 1, ll. Matherne, M. J. Jr., and W. H. Bathalter. 1966. Channel layer chromatography (CLC). A cleanup procedure for pesticide residue analysis. J. Assoc. Off. Anal. Chemists. 88, 1012. Matsumura, F., T. A. Bratkowski, and K. C. Patil. 1969. DDT: Inhibition of an ATP-ase in the rat brain. Bull. Environ. Contam. Toxicol. 8, 262. Matthews, H. B. and F. Matsumura. 1969. Metabolic fate of dieldrin in the rat. J. Agr. Food Chem. 11, 845. Miller, D. D. 1967. Effect of thyroprotein and a low- energy ration on removal of DDT from lactating cows. J. Dairy Sci. 88, 1444. 127 Mills, P. A. 1959. Detection and semiquantitatives estima- tion of chlorinated organic pesticide residues in foods by paper chromatography. J. Assoc. Off. Agr. Chemists. 88, 734. Mills, P. A. 1961. Collaborative study of certain chlorin- ated organic pesticides in dairy products. J. Assoc. Off. Agr. Chemists. 88, 171. Mills, P. A. 1963. Total diet study: Section C. Pesti- cide residue content. J. Assoc. Off. Agr. Chemists. 88, 762. Moats, W. A. 1964. One-step cleanup of chlorinated insec- ticide residues by chromatography on carbon-celite mixtures. J. Assoc. Off. Agr. Chemists. 81, 587. Moats, W. A. 1966. Analysis of dairy products for chlorin- ated insecticide residues by thin-layer chromatog- raphy. J. Assoc. Off. Anal. Chemists. 88, 795. Moats, W. A. and A. W. Kotula. 1966. Single step chromato- graphic cleanup of chlorinated pesticide reSidues, using high elution rates. J. Assoc. Off. Anal. Chemists. 88, 973. Moffitt, R. A. 1963. Residue analysis in the dairy indus— try. In Analytical Methodgtfor Pesticides, Plant 'Growth Regglators, Food Additives. ed. G. Zweig. Academic Press, New York and London. Vol. 1, 545. Moffitt, R. A., B. Burger, and W. Meckler. 1961. Improve- ments in paper-chromatography technique for pesti- cide determinations. J. Assoc. Off. Agr. Chemists. 88, 795. Montoure, J. E. and P. J. Muldoon. 1968. Distribution and stability of DDE, DDD, and DDT in Monterey and Cheddar cheese during manufacture and storage. J. Dairy Sci. 81, 858. Morrison, W. R. 1964. A fast, simple and reliable method for the microdetermination of phosphorus in bio- logical materials. Anal. Biochem. 1, 218. Moss, J. A. and D. E. Hathway. 1964. Transport of organic compounds in the mammal. Biochem. J. 81, 384. Moubry, R. J., G. R. Myrdal, and H. P. Jensen. 1967. Screening method for the detection of chlorinated hydrocarbon pesticide residues in the fat of milk, cheese, and butter. J. Assoc. Off. Anal. Chemists. 88, 885. 128 Moubry, R. J., G. R. Myrdal, and W. E. Lyle. 1968a. Inves- tigation to determine the respective residue amounts of DDT and its analogues in the milk and back fat of selected dairy animals. Pest. Moni. J. 8, 47. Moubry, R. J., G. R. Myrdal, A. Sturges. 1968b. Residues in food and feed. Rate of decline of chlorinated hydrocarbon pesticides in dairy milk. Pest. Moni. J. 2, 72. Mumma, R. O. and T. R. Kaniner. 1965. Identification of halogenated pesticides by mass spectroscopy. J. Econ. Entomol. 88, 491. Mumma, R. O., W. B. Wheeler, D. E. H. Frear, and R. H. Hamilton. 1966. Dieldrin: Extraction of accumula- tion by root uptake. Science. 152, 530. McCully, K. A. 1969. Analysis of multiple pesticide resi- dues: Chemical identity and confirmation of results. World Review of Pest Control. 8, 59. McCully, K. A. and W. P. McKinley. 1964. Determination of chlorinated pesticide residues in fat by electron capture gas chromatography. J. Assoc. Off. Agr. Chemists. 81, 652. McKinley, W. P. and G. Savary. 1962. Procedure for cleanup of butterfat prior to analyses for dieldrin residues. J. Agr. Food Chem. 18, 229. Nash, R. G. 1968. Plant absorption of dieldrin, DDT and endrin from soils. Agron. J. 88, 217. O'Donnel, A. E., M. M. Neal, F. T. Weiss, J. M. Bann, T. J. DeCino, and S. C. Lau. 1954. Chemical determination of aldrin in crop materials. J. Agr. Food Chem. 8, 573. Onley, J. H. 1964. Rapid method for chlorinated pesticide residues in fluid milk. J. Assoc. Off. Agr. Chem- ists. 81, 317. Onley, J. H. and P. F. Bertuzzi. 1966. Rapid extraction procedure for chlorinated pesticide residues in raw animal tissues and fat and meat products. J. Assoc. Off. Anal. Chemists. 88, 370. Park, K. S. and W. N. Bruce. 1968. The determination of the water solubility of aldrin, dieldrin, heptachlor, and heptachlor epoxide. J. Econ. Entomol. 81, 770. _ ___ ..- ‘ — _ . I" 71—. “V ~ 129 Phillips, W. E. J. 1963. DDT and the metabolism of vitamin A and carotene in the rat. Can. J. Biochem. Physiol. 81, 1793. Porter, P. E. 1964. Dieldrin. In Analytical Methods for Pesticides, Plant Growth Regulators and Food Addi- tives, ed. Z. Gunther. Academic Press, New York. Vol. 2, 146. Privett, O. S., L. J. Nutter, and R. A. Gross. 1968. Anal- ysis of milk lipids (1967). 8ymposium: Daity_ Lipids and Lip1d Metabolism, eds. M. F. Brink and D. Kritchevsky. Avi Publ. Co., Inc. Westport, Conn. page 99. Quaife, Mary L., J. S. Winbush, and O. G. Fitzhugh. 1967. Survey of quantitative relationship between inges- tion and storage of aldrin and dieldrin in animals and man. Fd. Cosmet. Toxicol. 8, 39. Quinby, E., W. J. Hayes, Jr., J. F. Armstrong, and W. F. Durham. 1965. DDT storage in the U. S. population. J. Amer. Med. Ass. 191, 175. Quirk, L. W. 1965. Lipids in fractions of the fat globule membrane of cow's milk. M. S. Thesis, Michigan State University. Radeleff, R. D., W. J. Nickerson, and R. W. Wells. 1960. Acute toxic effects upon livestock and meat and milk residues of dieldrin. J. Econ. Entomol. 88, 425. Radomski, J. L., W. B. Deichmann, E. E. Clizer, and A. Rey. 1968. Pesticide concentrations in the liver, brain and adipose tissue of terminal hospital patients. Fd. Cosmet. Toxicol. 8, 209. Richardson, A., M. K. Baldwin, and J. Robinson. 1968. Metabolites of dieldrin (HEOD) in the urine and faeces of rats. Chem. Ind. (London). 18, 588. Richardson, L. A., J. R. Lane, J. T. Peeler, and J. E. Campbell. 1967. Simultaneous determination of selected chlorinated insecticide residues in milk. J. Dairy Sci. 88, 1073. Robinson, J. and A. E. J. McGill. 1966. Organochlorine insecticide residues in complete prepared meals in Great Britain during 1965. Nature. 212, 1037. 130 Samuel, B. L. 1966. An improved screening method for chlo- rinated and thiophosphate organic insecticides in foods and feeds. J. Assoc. Off. Anal. Chemists. 88, 346. Saschenbrecker, P. W. and D. J. Ecobichon. 1967. Extrac- tion and gas chromatographic analysis of chlorinated insecticides from animal tissues. J. Agr. Food Chem. 18, 168. Schafer, M. L., K. A. Busch, and J. E. Campbell. 1963. Rapid screening method for DDT in milk with gas chromatography. J. Dairy Sci. 88, 1025. Shell Chemical Company. 1964. Manual of Methods for the Determination of Residues of Shell Pesticides. Ag- ricultural Chemical Division, Shell Chemical Com- pany, New York, New York. Singh, J. and J. D. Lanthier. 1968. Estimation of chlorin- ated pesticide residues in animal and vegetable oils. J. Assoc. Off. Anal. Chemists. 81, 45. Smith, K. J., P. B. Polen, D. M. DeVries, and F. B. Coon. 1968. Removal of chlorinated pesticides from crude vegetable oils by simulated commercial processing procedures. J. Amer. Oil Chem. Soc. 88, 866. Stanley, R. L. and H. T. LeFavoure. 1965. Rapid digestion and cleanup of animal tissues for pesticide residue analysis. J. Assoc. Off. Agr. Chemists. 88, 666. Stemp, A. R. and B. J. Liska. 1965. Simplified method of sample cleanup for gas chromatographic analysis of chlorinated insecticide residues in milk. J. Dairy Sci. 88, 985. Street, J. C., M. Wang, and A. D. Blau. 1966. Drug effects on dieldrin storage in rat tissue. Bull. Environ. Contam. Toxicol. 1, 6. Storherr, R. W., E. J. Murray, I. Klein, and L. A. Rosenberg. 1967. Sweep co-distillation cleanup of fortified edible oils for determination of organophosphate and chlorinated hydrocarbon pesticides. J. Assoc. Off. Anal. Chemists. 88, 605. Stull,‘3. W., W. H. Brown, F. M. Whiting, L. M. Sullivan, Mary Milbrath, and J. M. Witt. 1968. Secretion of DDT lactating cows fed thyroprotein. J. Dairy Sci. 51, 56. 131 Swope, F. C. 1968. Characteristics of Membrane Lipoprotein Fractions from Cow's Milk. Ph.D. Thesis. Michigan State University. Swope, F. C. and J. R. Brunner. 1968. The fat globule mem- brane of cow's milk: A reassessment of isolation procedures and mineral composition. Michwissenschaft. 23, 470. Tarjén, R. and T. Kemény. 1969. Multigeneration studies on DDT in mice. Fd. Cosmet. Toxicol. 1, 215. Taylor, A., R. E. Rea, and D. R. Kirby. 1964. An improved procedure for the extraction of organochlorine pes— ticides from animal tissue. Analyst. 88, 497. Thornburg, W. and H. F. Beckman. 1969. Pesticide residues. Anal. Chem. 81, 140R. Tinsley, I. J. 1966. Nutritional interaction of dieldrin toxicity. J. Agr. Food Chem. 18, 563. Tinsley, I. J. 1969. DDT effect on rats raised on alpha- protein rations: Growth and storage of liver vita- min A. J. Nutr. 88, 319. Tinsley, I. J. and R. R. Lowry. 1969. Essential fatty acid nutrition: An interaction with dieldrin. J. Amer. Oil Chem. Soc. 88. (Abstract) Walstra, P. and J. J. De Graaf. 1962. Note on the deter- mination of the phospholipid content of milk products. Netherlands Milk Dairy J. 18, 283. Wang, R. T. and S. S. Chou. 1968. Detection of pesticides by reversed phase paper chromatography. Chung Kuo Nung Yeh Hua Hsueh Hui Chih. 8, 63. Watts, J. O. and A. K. Klein. 1962. Determination of chlo- rinated pesticide residues by electronic-capture gas chromatography. J. Assoc. Off. Agr. Chemists 88, 102. Wells, C. E. 1967. Validation of study of a method for pesticide residues in food and animal feeds. J. Assoc. Off. Anal. Chemists. 88, 1205. Westlake, W. E., C. Corley, R. T. Murphy, W. F. Barthel, H. Bryant, and R. L. Schutzmann. 1963. Chemical resi- dues in the milk of cows grazed on chlordane-treated pasture. J. Agr. Food Chem. 11, 244. 132 Westlake, W. E. and F. A. Gunther. 1967. Advances in gas chromatographic detectors illustrated from applica- tions to pesticide residue evaluations. Residue Rev. 18, 175. Wheeler, W. B., D. E. H. Frear, R. O. Mumma, R. H. Hamilton, and R. C. Cotner. 1967. Absorption and transloca- tion of dieldrin by forage crops. J. Agr. Food Chem. 18, 231. Williams, S. 1964. Pesticide residues in total diet sam- ples. J. Assoc. Off. Agr. Chemists 81, 815. Williams, S. and J. W. Cook. 1967. Pesticide residues. Anal. Chem. Annu. Rev. 88, 142R. Williams, S., P. A. Mills, and R. E. McDowell. 1964. Resi- dues in milk of cows fed rations containing low con- centrations of five chlorinated hydrocarbon pesti- cides. J. Assoc. Off. Agr. Chemists. 81, 1124. Wilson, H. F., N. N. Allen, G. Bohstedt, J. Betheoe, and H. A. Lardy. 1946. Feeding experiments with DDT- treated pea-vine silage with special reference. to dairy cows, sheep, and laboratory animals. J. Econ. Entomol. 88, 801. Wilson, K. A. 1970. Personal communication. Wilson, K. A., R. M. Cook, and R. S. Emery. 1968. Effects of charcoal feeding on dieldrin excretion in rumi- nants. Fed. Proc. 81, 558. Wise, J. J. 1967. Thin-layer chromatography. In Analytical Methods for Pesticides, Plant Growth Regulators, and Food Additives, ed. G. Zweig. Academic Press, New York and London. Vol. 5, 47. Witt, J. M., R. C. Angus, J. W. Stull, W. H. Brown, and F. M. Whiting. 1966a. Variations of pesticide resi- dues in Arizona milk. J. Dairy Sci. 88, 1406. Witt, J. M., F. M. Whiting, W. H. Brown, and J. W. Stull. 1966b. Contamination of milk from different routes of animal exposure to DDT. J. Dairy Sci. 88, 370. Wood, B. J. 1966. Elution of dieldrin and endrin from Florisil. J. Assoc. Off. Anal. Chemists. 88, 472. Wood, N. F. 1969. Extraction and clean-up of organochlo- rine pesticide residues by column chromatography. Analyst. 88, 399. 133 Wurster, C. F., Jr. 1968. DDT reduces photosynthesis by marine phytoplankton. Science. 159, 1474. Wurster, C. F., Jr.and D. B. Wingate. 1968. DDT residues and declining reproduction in the Bermuda Petrel. Science. 159, 979. Zweig, G. 1967. Analytical Methods for Pesticides, Plant Growth Regulators, and Food Additives, ed. G. Zweig. Academic Press, New York and London. Vol. 5. Zweig, G., E. L. Pye, and S. A. Peoples. 1963. Residues in butterfat and body fat of dairy cows fed at two levels of Kelthane. J. Agr. Food Chem. 11, 72. who—e; .._ ~14- I APPENDIX CHEMICAL NAMES OF SOME IMPORTANT CHLORINATED PESTICIDES AND THEIR DERIVATIVES Pesticide Aldrin BHC Chlordane DDD(TDE) DDE DDT Dieldrin (HEOD) Dieldrin Metabolite I Dieldrin Metabolite I' Dieldrin Metabolite II Chemical Name 1,2,3,4,10,10-hexachloro-l,4,4a,5,8,8a- hexahydro-l,4-endo-exo-5,8-dimethano- naphthalene 1,2,3,4,5,6-hexachloro-cyclohexane, mixed isomer (Benzene hexachloride) l,2,4,5,6,7,8,8a-octachloro-3a,4,7,7a- tetrahydro—4,7-methanoindan l,l-dichloro-2,2-bis (p-chlorophenyl) ethane l,l-dichloro-2,2-bis (p-chlorophenyl) ethylene l,l,1-trichloro-2,2-bis (p-chloro- phenyl) ethane 1,2,3,4,10,10-hexachloro-6,7-epoxy-1, 4,4a,5,6,7,8,8a-octahydro-l,4—endo-exo- 5,8-dimethanonaphalene (Technical di- eldrin contains not less than 85% of the above ingredient) 1,1,2,3a,7a-pentachloro-5,6-epoxydeca- hydro-2,4,7-metheno-3H-cyclopenta [a] pentalen—3-one l,9,10,10,11-gg9-pentach13r8-435- epoxy-pentacyclo-(7,3,0,0 ' 0 ‘8 0.7: 1)dodecan-l,2-one 1,2,3,4,10,10-hexachloro-6,7-epoxy—l, 4,4a,5,6,7,8,8a-octahydro-4a(or 5)- hydroxy-g§9,l,4-endo-5,8-dimethano- naphthalene 134 135 Dieldrin 9(gyg—epoxy)hydroxy-1,2,3,4,10,10- Metabolite II' hexachloro-6,7-epoxy-1,4,4a,5,6,7,8, 8a-octahydro—l,4-endo-5,8-g8gfdimeth- anonaphthalene Dieldrin 1,2,3,4,10,10-hexachloro-l,4,4a,5,6,7, Metabolite III 8,8a-octahydro—6,7-dihydroxy-g88-l,4- endo-S,8-dimethanonaphthalene Endrin l,2,3,4,10,10-hexachloro-6,7-epoxy-l,4, 4a,5,6,7,8,8a—octahydro-l,4-endo-endo- 5,8-dimethanonaphthalene Heptachlor 1,4,5,6,7,8,8a-heptachloro-2,3-epoxy- Epoxide 3a,4,7,7a-tetrahydro-4,7-methanoindan Lindane 1,2,3,4,5,6-hexachloro-cyclohexane, gamma isomer 136 CI Cl C1 01 Cl Dieldrin (HEOD) O Metabolite I' Metabolite II Metabolite II' Metabolite III has the epoxide ring of dieldrin replaced by two -OH groups. Figure 4.--Chemica1 structures of dieldrin and its major metabolites (see text for references). 137 Solvent Peak Dieldrin Detector Response 2 ,flJ ‘ A ‘F‘ 1‘ 4‘=,--,._ Retention Time (min) Figure 5.--A typical gas-liquid chromatogram for deter- mination of dieldrin in milk. (GLC conditions: Hewlett-Packard model 57503 gas chromatograph; electron capture detector; 3.8% SE-30 on 80/100 mesh Diatoport S column.) “"1111 1111131!!! till/13117111111111“