SELENIUM IN PROTEINS FROM TOXIC FOODSTUFFS*

SELENIUM IN PROTEINS FROM TOXIC FOODSTUFFS* III. THE REMOVAL OF SELENIUM FROM TOXIC PROTEIN WDROLYSATES BY E. PAGE PAINTER AND KURT W. FRANKE (From the Department of Experiment Station Chemistry, South Dakota State College, Brookings) (Received for publication, July 23, 1935) Since Robinson (1933) found selenium in certain samples of toxic foodstuffs reported by Franke (1934), little has appeared to elucidate the chemical r61e of this element in the foodstuffs. Franke and Painter (1935) found that the selenium is linked in the protein of the toxic grains in such a manner that it is not removed by solvents for selenium or inorganic selenium compounds. The selenium in the grains differs in this respect from that reported by Beath and coworkers (1934) who found the selenium compounds removed from certain plants by water extraction. Franke and Painter (1935) also reported that the selenium appeared in solution in organic combination in the hydrolysates when toxic proteins were hydrolyzed with acids. Later they (Painter and Franke, 1935) hydrolyzed toxic proteins with acids and studied the selenium distribution in the insoluble humin and hydrolysates. The proportion of total selenium in the humin increased when humin formation was increased. In another study. (unpublished data) a similarity between labile sulfur and labile selenium was evident when toxic proteins were hydrolyzed with alkaline plumbite. A large proportion of the selenium was in the lead sulfide formed. * Published with the permission of the Director of the South Dakota Agricultural Experiment Station as Communication No. 16 from the Department of Experiment Station Chemistry. These investigations are being carried out under the Purnell Fund and with the cooperation of the Bureau of Chemistry and Soils, Bureau of Plant Industry, Bureau of Animal Industry, and Bureau of Home Economics of the United States Department of Agrioulture. 643

644 Selenium in Proteins In this paper the removal of selenium from toxic protein hydrolysates is reported. Plan of Experiment It has been found (Painter and Franke, 1935) that most of the selenium in toxic proteins is in the hydrolysate when the proteins are hydrolyzed by the usual methods. Since amino acids are the known products of protein hydrolysis, the common methods of separating them from hydrolysates were used to determine whether the selenium-containing compound (or compounds) could be separated in a similar manner. It should be emphasized that orthodox procedures were not rigidly followed, and alterations of methods were freely practised. This appears justifiable, inasmuch as an unidentified compound (or compounds) containing selenium is present in toxic protein hydrolysates. In case a selenium homologue of cystine or methionine were present, it would not have properties exactly like those of the sulfur compounds, but they probably would be similar in some respects. Owing to this probability, particular emphasis is given the two amino acids, cystine and methionine. Furthermore, it was desirable to remove the selenium from a toxic protein hydrolysate to determine whether or not it was still toxic. EXPERIMENTAL The source and method of separation of the protein used in this investigation have been described (Franke and Painter, 1935). The proteins were hydrolyzed with sulfuric acid (Painter and Franke, 1935), the insoluble humin filtered off and washed, and the acid removed with barium hydroxide. The barium sulfate was centrifuged and washed several times with hot water. The washings were added to the neutral (free from sulfuric acid or barium) hydrolysate. This hydrolysate was used as the starting point of all the separation procedures. The solution was concentrated by vacuum distillation or diluted with distilled water when adjustments in volume were necessary. Owing to the amount of humin formed during the hydrolysis, the tryptophane present in the protein was probably decomposed. The work of Gortner and his collaborators (1915, 1917) and Holm

E. P. Painter and K. W. Franke 645 and Greenbank (1923) indicates that this amino acid is practically quantitatively decomposed during acid hydrolysis when humin formation is large. After each separation the approximate quantity of selenium in the fractions was determined. Selenium was determined by the alkaloidal test described by Horn (1934) and the procedure described by Robinson et al. (1934). When a 3 gm. sample gave a negative selenium test, it was considered selenium-free. Extraction Methods Extraction with Organic Solvents-Marx (unpublished data) extracted a toxic protein hydrolysate with chloroform, petroleum ether, alcohol, and benzene. In no case was selenium removed, nor was a crystalline compound obtained. Extraction with Butyl Alcohol-It was found (Potter, unpublished data) that nearly all the selenium passed into the butyl alcohol fraction when a toxic protein hydrolysate was separated by the method described by Dakin (1918). After 72 hours of continuous extraction only a trace of selenium remained in the neutral aqueous solution. All the amino acids except the basic and dicarboxylic acids and glycine (?), hydroxyglutamic acid (?), and tyrosine (?) are extracted by butyl alcohol. Pirie (1932) has shown that methionine is readily extracted by butyl alcohol. That cystine slowly passes into the butyl alcohol fraction has been shown by Zahnd and Clarke (1933) and Hess and Sullivan (1935). They also found that cysteine passes into the butyl alcohol more readily than cystine. Precipitation Methods Precipitation with Phosphotungstic Acid-The well known method for the precipitation of the bases, lysine, arginine, and histidine, together with cystine, precipitated less than half the selenium from toxic protein hydrolysates. Both the methods described by Morrow (1927) and Vickery and Leavenworth (1929) were used. Thimann (1930) has shown that proline is partly precipitated by phosphotungstic acid. The solubility of cystine phosphotungstate depends upon the degree of racemization. Hoffman and Gortner (1922) and Plimmer and Lowndes (1927) found that the phosphotungstate of inactive cystine is much more

646 Selenium in Proteins soluble than the phosphotungstate of I-cystine. Undoubtedly there was appreciable racemization of cystine during hydrolysis by the method used. Precipitation with Copper-Town (1928) and Brazier (1930) use copper carbonate as an amino acid precipitant. Separation of the amino acids is effected by the solubilities of their copper salts in different solvents. The water-soluble copper salts contained from a trace to about half the selenium present in the hydrolysate. Consistent results were not obtained when precipitations were carried out under similar conditions. The amino acid complexes of copper insoluble in water are those of leucine, phenylalanine, and aspartic acid (Town, 1928; Brazier, 1930). Vickery and Leavenworth (1929) have shown that cystine is 95 per cent precipitated by copper hydroxide in a solution containing cystine and histidine. Mueller (1923) prepared the copper salt of methionine with copper carbonate and found that it was almost insoluble in cold water, but moderately soluble in hot water. Vickery and White (193233) use cuprous oxide to precipitate cysteine. Cystine is boiled with sulfuric acid in the presence of tin to reduce it to cysteine. The cuprous mercaptide precipitates quantitatively. By the use of this procedure on toxic protein hydrolysates, approximately half the selenium appeared in the cuprous mercaptide precipitate. Methionine is not precipitated under the conditions employed (Vickery and White, 1932-33). Precipitation with Silver Sulfate and Silver Oxide-The silver precipitation of Kossel and Kutscher (1900) for the determination of arginine, lysine, and histidine is probably the most satisfactory procedure for precipitating these amino acids. By following the modification described by Vickery and Leavenworth (1928) a very small amount of selenium was in the precipitates. A trace of selenium was in the precipitates formed in both neutral and alkaline solutions. However, the amount was so small that it was probably due to a contaminant. These authors have shown (Vickery and Leavenworth, 1929) that a small amount of cystine is precipitated by silver sulfate in acid solution and that none is precipitated by silver oxide in alkaline solutions. Precipitation with Mercury Salts-Three mercury salts, the sulfate, acetate, and chloride, have been studied. Each appeared to be superior to other amino acid precipitants in precipitating

E. P. Painter and K. W. Franke selenium. Precipitation with Hopkins and Cole s reagent (1901-02) in 5 per cent sulfuric acid resulted in only partial precipitation of the selenium present. It was found that more selenium was in the precipitate when precipitation was carried out in a nearly neutral solution. However, appreciable amounts of selenium remained in the solution when mercuric sulfate was used. When mercuric acetate was used, more amino acids were precipitated from solution than when the sulfate or chloride was used, but some selenium remained in solution. Mercuric chloride has proved to be the best precipitant for the selenium compounds yet found. It would completely precipitate the selenium from a nearly neutral hydrolysate. The following procedure was developed, which precipitated the selenium so nearly quantitatively that present methods for the detection of selenium were not sensitive enough to detect it in a reasonable amount of the soluble hydrolysate. To the hydrolysate from 100 gm. of toxic protein in 3 liters of solution solid barium carbonate is added in excess. 2 liters of saturated mercuric chloride are then added while the solution is stirred. A flaky precipitate forms immediately. The solution is allowed to set at room temperature, with frequent stirring, for 1 hour. The mercury precipitate and undissolved barium carbonate are filtered off with a Buchner funnel and the precipitate is washed two or three times with water. The filtrate is practically free from selenium. After the mercury, chlorine, and barium have been removed, and the amino acids dried, the alkaloidal test on a 3 gm. sample is negative. A positive test can be obtained on less than 0.1 gm. of the toxic protein, so less than one-thirtieth of the selenium originally present is in this filtrate. The distillation method (Robinson, Dudley, Williams, and Byers, 1934) indicated less than 3 parts per million of selenium. As in all similar procedures, results are somewhat variable. Over 60 mercuric chloride precipitations have been made. One precaution must be considered. The mercuric chloride must be saturated. It should be dissolved by heating, and the solution used after it has cooled down to room temperature. In general mercury salts precipitate histidine and tryptophane quantitatively, and cystine and tyrosine to a less extent. Mercuric sulfate in 5 per cent sulfuric acid has been used to precipitate

648 Selenium in Proteins cystine (Hopkins and Cole, 1901-02). Under these conditions Vickery and Leavenworth (1929) found 86 per cent of the cystine precipitated. These authors (1928) also use Hopkins reagent to precipitate histidine. Pirie (1932, 1933) and Mueller (1923) and du Vigneaud and Meyer (1931-32) used mercury salts to precipitate methionine. In his later paper, Pirie (1933) describes the combined action of mercuric acetate and phosphotungstic acid to precipitate methionine. In both Pirie s procedures methionine is recovered from a final precipitation in which mercuric chloride is used. He states that both the sulfate and chloride ion have an inhibiting action on the precipitation of methionine in an amino acid solution. In both the older methods for isolating methionine (Mueller, 1923; du Vigneaud and Meyer, 1931-32) the amino acid was precipitated from a solution neutral to Congo red. By the method described for removing the selenium from toxic protein hydrolysates, the selenium in the hydrolysate from 100 gm. of protein is in the mercuric chloride precipitate which contains approximately 5 gm. of amino acids. DISCUSSION The results obtained by precipitating selenium from toxic protein hydrolysates with different amino acid precipitants does not restrict the properties of the selenium compound to those identical with any one amino acid. In fact more than one compound of selenium is indicated. It would be very difficult to conceive a single compound partially precipitated by every one of the precipitants used but only completely precipitated by one. However, it is entirely probable that other precipitants than mercuric chloride would quantitatively precipitate the selenium compounds under very definite conditions. Few conclusions can be reached regarding the chemical properties of the selenium compound (or compounds) when only one property-that of the solubility of several of its metallic complexes-has been studied. The properties in regard to solubility resemble those of cystine, methionine, histidine, and possibly tryptophane. Every amino acid reagent used precipitating these amino acids precipitates some selenium and every hydrolysate fraction containing these amino acids in appreciable quantity

E. P. Painter and K. W. Franke 649 contained some selenium. Since the solubilities of these amino acids resemble each other closely, and since little is known of the solubilities of the compounds of methionine, it is difficult to eliminate the known amino acids. One difficulty has deterred progress toward isolating the selenium-containing compound. When the metallic precipitate of amino acids containing a concentration of selenium was decomposed with hydrogen sulfide, a large proportion of the selenium always was in the metallic sulfide formed. This has made several separations of the selenium-containing compounds of toxic protein hydrolysates by precipitations very difficult. This was pronounced when cystine in a toxic hydrolysate was reduced to cysteine. Upon decomposition of the cuprous mercaptide with hydrogen sulfide, all of the selenium in the precipitate appeared in the cuprous sulfide formed, which suggests that an unstable selenol existed. The selenides of heavy metals are extremely insoluble compounds. If one of the known amino acids contained selenium in one of its usual organic combinations, the selenium would undoubtedly replace a sulfur or oxygen atom (Bradt and Van Valkenburgh, 1929; Bradt, 1930, 1934; Bradt and Green, 1931; Bradt and Crowell, 1931). The probability that it would replace sulfur has been suggested (Painter and Franke, 1935). A compound entirely unlike those of the known protein hydrolysis products is not precluded. There is evidence that most of the selenium is in a compound very similar to cystine. All cystine and cysteine precipitants used precipitate some selenium, and some selenium is removed from toxic hydrolysates under the same conditions that labile sulfur is split off from cystine or cysteine. It has also been shown (Painter and Franke, 1935) that a compound of selenium is present which decomposes more readily in acid solutions than cystine. Mercuric chloride was found to precipitate probably all the selenium from toxic protein hydrolysates. Cystine cannot be completely precipitated in this manner. The quantitative precipitation of cystine by heavy metals is precluded because one-sixth of the sulfur is oxidized when silver (Vickery and Leavenworth, 1930), mercury, or.copper (Simonsen, 1931-32; Preisler and Preisler, 1930, 1932)

650 Selenium in Proteins salts are employed; this may also occur with other metals. Evidently this did not happen with a selenium compound. It is known that sulfur is more easily oxidized than selenium. In an earlier paper (Painter and Franke, 1935) it was stated that the molar selenium-sulfur ratio was approximately 1: 148 in the protein under investigation. Since many of the properties of the selenium-containing compounds closely resemble those of the sulfur-containing amino acids, the difficulty in separating the two is obvious. This difficulty is further augmented by the unstability of the selenium in its naturally occurring linkages. SUMMARY Toxic proteins have been hydrolyzed and the removal of selenium has been attempted by various methods. Most of the selenium passed into the butyl alcohol when extracted from a nearly neutral hydrolysate. Precipitations were carried out with phosphotungstic acid, copper, silver, and mercury salts. In all cases a fraction of the selenium was in the precipitate. A procedure with mercuric chloride was developed whereby all the selenium compounds were precipitated. BIBLIOGRAPHY Beath, 0. A., Draize, J. H., Eppson, H. F., Gilbert, C. S., and McCreary, 0. C., J. Am. Pharm. Assn., 23, 94 (1934). Bradt, W. E., Proc. Zndiana Acad. SC., 40, 141 (1930); 43, 72 (1934). Bradt, W. E., and Crowell, J. H., Proc. Indiana Acad. SC., 41,227 (1931). Bradt, W. E., and Green, J. F., Proc. Indiana Acad. SC., 41,215 (1931). Bradt, W. E., and Van Valkenburgh, M., Proc. Indiana Acad. SC., 39, 165, 171 (1929). Brazier, M. A. B., Biochem. J., 24, 1188 (1930). Dakin, H. D., Biochem. J., 12, 290 (1918). Franke, K. W.,.Z. Nutrition, 8, 597, 609 (1934). Franke, K. W., and Painter, E. P., Cereal Chem., in press (1935). Gortner, R. A., and Blish, M. J., J. Am. Chem. Sot., 37, 1630 (1915). Gortner, R. A., and Holm, G. E.,.Z. Am. Chem. Sot., 39,2477 (1917). Hess, W. C., and Sullivan, M. X., J. Biol. Chem., 108, 195 (1935). Hoffman, W. F., and Gortner, R. A., J. Am. Chem. Sot., 44,341 (1922). Holm, G. E., and Greenbank, G. R., J. Am. Chem. Sot., 46, 1788 (1923). Hopkins, F. G., and Cole, S. W., J. Physiol., 27,418 (190142). Horn, M. J., Znd. and Eng. Chem., Anal. Ed., 6, 34 (1934). Kossel, A., and Kutscher, F., Z. physiol. Chem., 31, 165 (1900). Morrow, C. A., Biochemical laboratory methods for students of the biological sciences, New York, 163 (1927).

E. P. Painter and K. W. Franke 651 Mueller, J. H., J. Biol. Chem., 66, 157 (1923). Painter, E. P., and Franke, K. W., Cereal Chem., in press (1935). Pirie, N. W., Biochem. J., 26, 1270 (1932) ; 27, 202 (1933). Plimmer, R. H. A., and Lowndes, J., Biochem. J., 21, 247 (1927). Preisler, P. W., and Preisler, D. B., J. Biol. Chem., 89, 631 (1930); 96, 181 (1932); J. Am. Chem. Sot., 64,2984 (1932). Robinson, W. O., J. Assn. Of. Agric. Chem., 16, 423 (1933). Robinson, W. O., Dudley, H. C., Williams, K. T., and Byers, H. G., Znd. and Eng. Chem., Anal. Ed., 6, 274 (1934). Simonsen, D. G., J. Biol. Chem., 94, 323 (193132). Thimann, K. V., Biochem. J., 24, 368 (1930). Town, B. W., Biochem. J., 22, 1083 (1928). Vickery, H. B., and Leavenworth, C. S., J. Biol. Chem., 76, 707 (1928); 83, 523 (1929); 86, 129 (1930). Vickery, H. B., and White, A., J. Biol. Chem., 99, 701 (193233). du Vigneaud, V., and Meyer, C. E., J. Biol. Chem., 94, 641 (1931-32). Zahnd, H., and Clarke, H. T., 1. BioZ. Chem., 102, 171 (1933).

SELENIUM IN PROTEINS FROM TOXIC FOODSTUFFS: III. THE REMOVAL OF SELENIUM FROM TOXIC PROTEIN HYDROLYSATES E. Page Painter and Kurt W. Franke J. Biol. Chem. 1935, 111:643-651. Access the most updated version of this article at http://www.jbc.org/content/111/3/643.citation Alerts: When this article is cited When a correction for this article is posted Click here to choose from all of JBC's e-mail alerts This article cites 0 references, 0 of which can be accessed free at http://www.jbc.org/content/111/3/643.citation.full.h tml#ref-list-1