rations. This fraction contained a toxic substance(s) which presumably was present in a byproduct of industrial production of stearte and oleic acids.

The toxic substance was present in a variety of types of fats. These included a sample of triolein, which on the basis of chemical analyses was found to be of “excellent quality" (A. Yartzoff et al., J. Am. vụ Chem. Soc. 38, 60 (1961)). When this trolein was incorporated into a chick ration at a level of 15 per cent, the birds showed severe symptoms of hydropericardium. Simdar symptoms were produced when chicks were fed rations containing distillates or residues secured in commercial production of fatty acids. The toxic substance(s) was iso found in inedible animal tallows, acidulated vegetable oils, and several compecially produced oleic acids and triolein Frestone. W. Horwitz, Friedman, and G. M. Shue. J. Am. Oil Chem. Soc. 38. 418

1. A report from another laboratory confirmed the presence of the chick edema factor in various industrial fats (J. C. Alexander, R. J. Young, C. M. Burnett, and H. D. Hathaway, Poultry Sci. 41, 22 48211.

The presence of the chick edema factor in commercial fats led to the ruling by the Food and Drug Administration that methyl esters of higher fatty acids intended for use food additive must be free of the chick a factor (Federal Register Dec. 9, 11. 26 F.R. 11828; 121.224). The presence of the factor was to be ascertained by a click bioussay based on the volume of periardral fluid in birds fed the fat under investigation.

That salt was necessary in the chicks' Tation for the development of hydropericardium when toxic fats were fed was sugKrsted by Alexander and co-workers. They observed typical symptoms of the disturbance only when the ration contained sodium dloride; extra salt accentuated the condition. Although this syndrome had many of the earmarks of a vitamin E deficiency,

neither antioxidants (including vitamin E) nor selenium had any curative effect.

Sensitivity of different species to the toxic factor in these fats varies considerably. When monkeys were fed a sample of triolein that produced hydropericardium in chicks. the five monkeys died within three months (Yartzoff et al., loc. cit.). The ration contained 25 per cent triolein. The monkeys showed signs of jaundice, pancreatic atrophy and fibrosis, hemosiderosis, fatty liver with necrosis, and gross hemorrhage in the intestinal tract.

Pigs ied a ration composed of equal parts of a highly toxic chick ration and a normal swine ration gained weight and appeared to show no abnormalities (Sanger et al.. loc. cit.). However, when a toxic fat was incorporated into a swine ration at a level of 9 per cent. the five shoats gained 0.72 pounds per day while the controls gained 2 pounds (I. C. Scott, J. Am. Vet. Med. Assn. 137. 258 (1960)). Despite the poor weight gains, the one pig sacrificed about six weeks after the start of the study showed no gross or microscopic lesions attributable to the ration.

The fat from the latter pig was rendered and incorporated into a chick ration at a level of 4 per cent. None of the chicks fed the ration containing the rendered lard developed toxicity signs or symptoms. This finding suggested that the pig did not store appreciable amounts of the toxic substance in its adipose tissue.

The apparent absence of the toxic factor in pork fat is in contrast to chick fat, where storage appears to occur. The unsaponifiable fraction of carcasses of chickens fed the toxic fat was very potent in producing hydropericardium in other birds (Friedman et al., J. Assn. Official Agr. Chem. 42, 129 (1959)).

A more recent study of the response of different species to the chick edema factor utilized the unsaponifiable fraction of a toxic fat. This was fed to young chicks and rats (T. C. Campbell and Friedman, Proc.

Foc. Exp Biol. Med 121, 1283 (1966)). The weandig rats fed the u-aponifiable fraction at a level of 9 g per kilogram body woght per day of the chick odeur factor sho sed a 37 per cent loss in weight over the six day feeding trial. No gross tí--ue alterations were apparent on autopsy. The livers, however, were 21 per cent heavier than those of the controls when expressed on a body weight basis, while the adrenals were 50 per cent heavier. The heart, kidney, and *piech were of normal size.

Chicke fed a comparable amount of concentrate developed hydropericardium in six days The livers of these birds were 15 per cent heavier than those of the controls. The statement was made that "the increase in liver weight in the chicks was not due to moisture or fat." Such an observation should have been documented and checked by histological studies.

A few years ago, R. E. Harman and collaborators (J. Am. Chem. Soc. 82. 2078 (1960)) reported isolation of a crystalline substance from a feed-grade tallow. This compound produced hydropericardium in chicks when incorporated into a commercial ration at a level of 0.1 p.p.m. A report (Yartzofi et al., loc. cit.) indicated that the crystalline substance contained 47 per cont chlorine.

Apparently, there are a number of compounds that behave like the chick edema factor. These compounds move closely with the toxic factor during molecular distillation, on thin layer chromatography, and show similar peaks on gas chromatography Not all these compounds are toxic, and toxicity appears to vary in those that are.

Part of these difficulties may be resolved with characterization and synthesis of our of the toxie substances. By means of single crystal x-ray crystallography, J. S. Cvvel, X. C. Webb, and A. J. Malas reporte (Abstracts, Am. Crystallographic Assn. Meeting. p 27 (1967)) that the compound isolated in their laboratories was 1.2.3.7, 8.9-hexachloro-dibenzo-p-dioxin. This work was supported by infra-red, ultra-violt, and mass spectrometry principally by J. C. Wooton and W. R. Courchene (J. Agr. Food Chem. 12, 94 (1964)). This structure was ́ verified by synthesizing the compound. The latter produced the same lesions, in chicks as the compound isolated from the toxic fat.

1

Significant progress has been made in indentifying one of the toxic substances present in some batches of fats intended for animal foods. There are still a number of unanswered questions, the most important of these being: What is the possible signfcance of these compounds to human health' Since substances with physiological properties similar to the chick edema factor een be stored in the adipose tissue of chicks, can they also accumulate in human tissue when minimal amounts are inadvertently ingested?

Another important question is: What is the source of these compounds? Apparently, the chick edema factor has been associated only with fats and fatty acids subjected to a considerable amount of heat during their processing. If the source of the chlorine in the toxic compound could be identified, it might be possible to devise methods for its removal or elimination before marketing the fat.

Introduction

When the problem of "X" or "Edema" in poultry was brought to our attention in December of 1957, a considerable amount of work had already been done. It was established that drugs added to the feed, as well as contamination by drugs or heavy metals such as lead or arsenic, was not responsible. The disease was not caused by bacterial, viral, or parasitic infection, and evidence was accumulating that the only feed ingredient associated with all the various outbreaks was the fat. The symptoms have been described in detail in the preliminary report of Schnittle, et al. (J. Am. Vet. Med. Assoc., 132, 216 (1958)), which also summarized some of the work leading to the incrimination of the fat added to the ration.

The characteristic symptoms of this disease are the presence of excessive fluid in the pericardial sac, in the abdominal cavity (water belly), and less often subcutaneously, accompanied by high mortality beginning approximately in the third week. The striking resemblance of the "edema disease" symptoms to those recorded for the exudative diathesis syndrome of vitamin E deficiency in chicks made attractive the hypothesis that the disease outbreaks were the result of an induced vitamin E deficiency caused by the use of poor quality fats.

It was convenient, because of other work in progress in our laboratory at the time, to use the A.O.A.C. vitamin D, chick test ration and single comb white Leghorn chicks in an attempt to produce the condition.

Our first feeding trial with a sample of feed collected from one of the "disease areas" was a false start, since after six weeks on the ration the chicks were normal and showed no unusual symptoms upon postmortem examination. The first experiment in which

Presented at the Seventy-second Anand Meeting of the Astacuation of Offrial Agricultural Chemists, Oct. 13 15, 1938, at Washington, D.C.

we successfully produced the characteristic symptoms of the "chick edema discase" involved a sample of fat that had been collected by one of our Food and Drug inspectors at a feed manufacturing plant and which had been known to have produced by-product from the manufacture of oleic the disease, and also a sample of a "tarry" and stearic acid, which presumably had been mixed into this fat sample to the extent of about 40%. In this experiment the vitamin E hypothesis was tested. Also, the suitability of the A.0.A.C. ration for this investigation was checked against a feed resembling more closely a practical commercial ration.

Basal Ration and the Vitamin E Hypothesis-Table 1 shows the composition of the A.O.A.C. basal ration and the special basal ration. The major differences are the substitution of Drackett (an isolated soybean protein) for the casein of the A.0.A.C. diet, and the inclusion of alfalfa leaf meal and linseed oil meals. From the results shown in Table 2, it is clear that: (1) the fatty acid (F.A.) by-product is much more effective in producing the disease than the sample of fat (INV. Fat) that had actually been used by feed manufacturers; (2) the special basal ration is to be preferred to the A.O.A.C. diet, since the severity of the edema symptoms is increased, growth performance is improved, and gizzard crosion is eliminated; (3) the feeding of vitamin E in large

The ure. adduct fraction of the fatty acids produced no unusual effects either in rats or in chicks.

Effect of Unsaponifiable Fraction on Rats. -The significant findings of feeding the unsaponifiable fraction of the fatty acid by

product to rats are seen in Table 4. One ml of the unsaponifiable fraction was fed to cach of four rats every day for 22 days. Control animals received no supplement to the basal ration. The experimental group contained two males from different litters, with litter mates in the control group. The females of both the test and control groups were all from one litter. The growth rate was significantly depressed-females were more drastically affected than males; the size of the thymnus decreased; and the size

of the liver markedly increased, which was accounted for only in part by an increase in liver fat. No other significant differences in the relative weights of organs (kidneys, thyroids, hearts, adrenals, etc.) were noted,

although small differences cannot be demonstrated by such small groups.

Although deleterious effects were produced in rats both by the urea filtrate and by the unsaponifiable portion, symptoms similar to those of the "edema disease" syndrome were not observed in this species.

Effect on Chicks.--When add to the diet of chicks at a level equivalent to 7% of the original fatty by-product, the crea filtrate produced a significant growth de ression but, as Table 5 shows, very few ed ma disease symptoms. The small amount of activity