Y
‘4001
samples showed a n increase in stabiiity and others a decrease but: in all cases, these changes were small. It appears, then, that this immediate increase in stability as a result of irradiation is associated mostly with the free fat of beef in combination with added antioxidants. T h e reason for this effect, however, remains unknown a t this time.
1200
0 (D
5
1000
(n (L
0
I
Literature Cited
600
Ir 400 Q
ILn
200
PROTEIN BOUND LIPIDS FREE FAT
P
P B
B E B
P
P
B
B
P
B
B B
B
B
P
P
B P P
COMPOSITION OF RECONSTITUTED MEAT SAMPLES P i p o r k ; B=Beef Figure 2. Effect of irradiation on stability of fat from reconstituted beefpork combinations
not with ground pork, and onl5- when antioxidants had be-n added to the beef. It \vas first Lhought that this effect might be due to the formation of ne\\- antioxygenic substances or synergists resulting from the effect of irradiation on the proteins of the meat, since many amino acids or peptides show synergistic activity with phenolic antioxidants (,?). Hoiiever. the failure of pork to shoii the same effect is difficult to explain. In an attempt to determine the cause of this effect, the free glyceride fat and bound lipids from samples of freezedried beef and pork liere extracted
separately and the meats reconstituted using various cross-combinations of lipid and nonlipid components. As shown in Figure 2, all the irradiated samples showed an immediate increase in stability. but only the four samples containing free fat from beef showed large increases. the largest being obtained with a sample consisting of pork protein and phospholipids with free fat derived from beef. T h e three samples containing free fat from pork showed only very small stability improvements. Storage of these irradiated samples for 2 weeks a t room temperature showed no significant or consistent effect; some of the stored
(1) Bradshaw, \V. \l’,) Truby: F. K . , Office of Technical Services, U. S. Dept. of Commerce, PB Rept. 154306 (1 962). (2) Chipault, J. R . ? Mizuno, G. R., J. .4GR. FOODCHEM.14, 221 (1966). (3) Chipault, J. R . , Mizuno, G. R., J . .4m. Oil Chemists’ SOC.41, 468-73 (1964). 14) ChiDault. J. R.. Mizuno, G. R., unpublished results. (5) Clausen, D. F.. Lundberg. \V. O., Burr. G. O . , J . Am. Oil Chemists’ Soc. 24, 403-4 (1947). (6) hierritt. C., Jr., Il’alsh, J . T., Bazinet, hl. L., Kramer. R. E., Bresnick. S. R., Ibzd., 42, 57-58
(1965). (7) Rose, D.. Lips, H. J.. Cyr. R., J . FoodSci. 26, 153-155 (1961). (8) SedlBtek. B. A , , .]\ahrung 2 (6), 947-56 (1958). (9) Tonnelat. J. P.. Flanzy. J., ‘4nn. Zootrchnol. 10 (2),97--104 (1961). Receized j a r recierr! October 4, 7965. Accepted January 26, 1966. Presented in part before the Dizsision of Agricultural and Food Chemistry, Symposium on Radiation Presercation of Foods, 150th Meetin,?, ACS, Atlantic City, -T. J . , S$btember 1965. Research sujportrd by the 0. S. Army .Tatick Laboratories, ‘Yatick, Mass., and T h e Hormrl Foundation, Austin, Minn.
F I S H O I L ODORS G. R. MIZUNO, EVELYN McMEANS, and J. R. CHIPAULT
Odor Problems of Fish Oils. Volatile Amines of Menhaden Oil
The Hormel Institute, University of Minnesota, Austin, Minn.
The basic constituents of a highly volatile fraction collected during molecular distillation of menhaden oil have been examined by paper chromatography and thin-layer chromatography. Tentatively identified were ethylenediamine and 1,4-butanediamine as major components with smaller amounts of propyl- and hexylamines. Secondary and tertiary arnines or quaternary ammonium bases were not detected.
T
nature of the compounds responsible for fishy odors remains largely unknown. Some investigators believe that carbonyl compounds resulting from the oxidative deterioration of highly unsaturated fatty acids are responsible for fishy odors ( 9 ) . Others, however, have suggested that these HE
odors are caused by noncarbonyl constituents (4) and that certain nitrogencontaining compounds may be involved in their formation (7, 74). During a study of the nature of fish oil odors samples of volatile compounds collected during molecular distillation of menhaden oil were made available to VOL. 14,
The origin of the samples and the nature of the volatile acidic constituents of the highly volatile fraction have been described (6). This report describes the basic constituents of this fraction.
us.
Experimental Starting Material. Nitrogen determinations and qualitative tests for NO. 3,
MAY-JUNE
1966
229
amines were performed on the forerun streaking and tailing. Much better redistillate and the aqueous and nonsults were obtained by using thin-layer aqueous fractions of the cold trap conchromatography. T h e procedure was similar to that described by Brenner, densate collected during molecular distillation of menhaden oil (70). Niederwieser. and Pataki ( 3 ) . The Silica Amines were isolated only from the Gel G plates were dried at 105' to 110' C. for a t least 4 hours and stored under aqueous portion of the cold trap disatmospheric conditions. Aqueous solutillate. For preliminary experiments the tions of the amine hydrochlorides were condensate recovered after isolation of the spotted on the plates. air-dried, and chroacids as their sodium salts ( 6 ) was used, but for final fractionation and identificamatographed with a solvent consisting of tion, larger amounts of material Lvere 75 grams of phenol, 25 ml of \rater. and obtained directly from the total aqueous 4 ml. of glacial acetic acid. The spots volatile fraction. were revealed by spraying the plates Preliminary Measurements. Total with a 0.1% ninhydrin solution in nitrogen was obtained by Kjeldahl chloroform. analysis, using the digestion mixture and Results and Discussion procedure of Koble (72) in combination with the colorimetric determination Table I shows the results of the qualiof ammonia described by Russell (73). tative tests performed on these menSpot tests as described by Feigl (8) haden oil volatile fractions. The forerun were used to determine the presence or was the first distillate of the molecular absence of the various classes of amines in distillation and represents low molecular these fractions. Tests for tertiary amines weight lipid components such as free were performed Lrith a citric acid-acetic anhydride reagent \vhich gives a red to fatty acids, partial glycerides. unsaponipurple color when heated in the presence fiables, and breakdown products of unof these compounds. The test for saturated fatty acids that could be colsecondary amines was the blue-violet lected on the molecular still condensing color formed when they are treated surface. No amines could be detected with a mixture of sodium nitroprusside in this material or in the organic upper and acetaldehyde, while the iodinelayer of the more volatile compounds azide test for dithiocarbamates was used accumulated in the molecular still cold to detect primary and secondary amines trap. However, both these fractions as a group after they had been treated contained approximately 30 pg. of with carbon disulfide. Isolation of Amines. Either the entire nitrogen per gram as determined by condensate recovered from the isolation Kjeldahl analysis. This amount of of the sodium salts of the acids ( 6 ) , or nitrogen is very close to the limit of 1 to 2 ml. of the aqueous portion of the detection of the qualitative amine tests original cold trap condensate was acidiemployed and it is possible that. even if fied to a pH of 2.0 with 2.Y hydrochloric all the nitrogen present were in the acid. The hydrochlorides of the bases form of amines, it would have been undewere then isolated after distillation of the tected in these samples. water and other volatile compounds, The aqueous layer from the cold trap using the procedure and apparatus employed for preparation of the acid salts contained 15 times this amount of nitro(6). The residual amine hydrochlorides gen (482 pg. per gram) and tests for were dissolved in 100 to 200 pl. ofwater tertiary and secondary amines u e r e and used for qualitative tests and for negative, while the dithiocarbamate test, paper and thin-layer chromatography. Separation of Amines. PAPERCHRO- which detects both primary and secondary amines, was positive. indicating that MATOGRAPHY. For separating the alithe amines in this fraction were all phatic primary monoamines, strips of primary in nature. Whatman No. 1 filter paper, 11.5 X 45 I n preliminary exploratory expericm., were treated by dipping once in 0.15.Y aqueous solution of sodium acements the hydrochlorides of the bases tate and drying in air at room temperaisolated from the aqueous cold trap ture (75). The solvent system was the layer were chromatographed on unupper layer obtained from a n equilitreated Whatman No. 1 filter paper brated mixture of butanol-water-acetic acid (50:49:1, v. 'v.). The chromatograms were developed for 18 hours a t room temperature and the amine spots Table I. Presence of Amines in were revealed by dipping the paper in a Total ninhydrin solution made from 200 mg. Nifroaen of ninhydrin and 15 mg. of ascorbic acid Content,a dissolved in 100 ml. of absolute ethanol. Sample w/G. With this system primary monoamines Standards with one to six carbon atoms were well n-Propylamine Di-n-propylamine separated, while more polar diamines or Tri-n-propylamine amino alcohols remained near the solvent Blank front, with R, values less than that of Menhaden oil volatiles methylamine. distillate THIN LAYER CHROMATOGRAPHY. Forerun Cold trap upper (organic) Chromatography of the volatile bases layer of menhaden oil on paper with the Cold trap lower (aqueous) solvents recommended by Bremner and layer Kenten ( 2 ) for separation of diamines a Procedures described in text. was not successful because of excessive I
-
230
J. A G R . F O O D C H E M .
using several solvent systems and indicator solutions. During these experiments the absence of secondary and tertiary amines and of choline or other quaternary ammonium bases was indicated by the failure of the chromatograms to develop spots when they were treated with sodium nitroprusside-sodium carbonate (7) and phosphomolybdic acid (5), respectively . These amine hydrochlorides were chromatographed on acetate paper using the butanol-water-acetic acid solvent, and when the chromatogram was dipped in the ninhydrin solution, pink to purple spots appeared within a few minutes (Figure 1). This is characteristic of primary amines and again confirms the presence of these compounds in the menhaden oil volatiles. Secondary amines d o not react with ninhydrin until the paper has been heated in an oven a t 65' to 70' for several minutes and, in all cases, the unknown material gave no additional spots upon heating after ninhydrin treatment. As shown also on Figure 1, the primary monoamines with one to four carbon atoms (D-1 to 0 - 4 ) give compact well-separated spots, but the more polar diamine ( A ) or ethanolamine ( B ) migrates only a very short distance from the base line and their R, values are less than that of methylamine. The unknorvn amines from menhaden oil ( C ) gave a long yellowish streak from the base line to the solvent front on which \rere superimposed four light pink to purple spots. The major portion of these unknowns was concentrated in an area consisting of a dark purple spot (C-2) with considerable tailing which moved slightly above the base line with an Rf value less than that of methylamine. There was some slight evidence of separation within the area (C1, C-2). but no conclusion could be drawn regarding the nature of these compounds except that they were not primary monoamines. In addition, two very faint light pink spots (C-3, C-4) could be detected when enough material was used on the chromatogram. One corresponded to propylamine, while
Volatiles from Menhaden Oil Response to Tests for Primary and Secondary Tertiary secondary amines only amines amines only
Q
0
@ Q
8 0 A
B
D
C
E
Figure 2. Thin-layer chromatography of primary amine hydrochlorides A
B
C
D
A and E, ethylenediamine ( l ) , 1,3-proponediamine (2), 1,4-butanediamine (3), 1,5-pentanediamine (4), 1,6-hexanediamine (5). E, methylamine ( l ) , ethylamine (2), propyl amine (3). C, ethanolamine. D, volatile amines from menhaden oil
Figure 1. Separation of primary amine hydrochlorides on Paper A,
-
1,4 butanediamine 6, ethano(putrescine). lamine. C, volatile amines from menhaden oil. D, methylamine (1 ), ‘ethyl(21, propylamine amine (31,butylamine (4)
the other was close to the solvent front. 12:hen log E;!, values, calculated according to the formula proposed by Bremner and Kenten (2) ( R , = 1/R, - 1): were plotted against the carbon number of the primary monoamines, this farther spot corresponded closely to hexylamine. A typical tracing of a thin-layer chromatogram. developed with the phenolwater-acetic acid !solvent designed to separate diamines, is sho\vn in Figure 2. Under the conditions employed here, the unknown material (D)gave five distinct spots. The first two spots had R, values IoIver than those of the monoamines and corresponded closely to ethylenediamine and 1,4-butanediamine (putrescine), respectively. Spot 0 - 3 was yellow with no tinge of pink and, therefore, is not considered to be due to a n amine. Spot 0 - 4 , Ivhich showed a slight pink tinge preceded by a yellow streak, corresponds to propylamine, and 0 - 5 , lvhich \vas also very Lveak, is probably due to the hexylamine detected on the paper chromatogram. The amines present in this volatile fraction from meinhaden oil consist mainly of ethylenediamine and butanediamine with minor amounts of Ca and
C S monoamines. More polar polyamines such as spermine and spermidine have lower R, values than the diamines ( 7 7 ) and, therefore, can be considered absent from this material. If the total nitrogen content of this aqueous fraction, as determined by Kjeldahl analysis, is expressed in terms of butanediamine, it is equivalent to a concentration of 0.15% of this compound in the volatile aqueous phase and of approximately 0.7 p.p.m. in the original menhaden oil. Undoubtedly, only a small fraction of the amines present in the original oil was recovered and detected but. even if we assume that 90y0 of these compounds were lost, their original concentration in the oil must have been extremely low. T h e odor of these compounds is not pleasant, but, in spite of the names given to some of the members of this series (putrescine, cadaverine), it is not of a putrid character, but rather similar to that of ammonia, although considerably weaker. Although these compounds may play a role in the over-all odor of fish oils, the nature of their odor and their low concentrations suggest that their contribution must be rather small. Trimethylamine, which has often been associated with fishy odors ( 7 4 , was not detected in this sample of volatile material. Acknowledgment T h e authors are grateful to Maurice Stansby and Eric Gauglitz, Seattle Technological Laboratory, Bureau of Commercial Fisheries, for supplying the starting materials. VOL.
Literature Cited
(1) Bregoff, M., Roberts, E., Delwiche, C. C., J . Biol. Chem. 205, 565 (1953). (2) Bremner, J. M., Kenten, R. H., Biochem. J . 49, 651 (1951). (3) Brenner, M., Niederwieser, A . , Pataki, G., “Aminosauren und Derivate,” in E. Stahl, “DunschichtChromatographie,” Springer-Verlag, Berlin, 1962. (4) Chang, S.S., Masuda, Y., Mookerjee, B. D., Chem. Ind. (London) 1962, 1023. (5) Chargaff, E., Levine, C., Green, C., J . Biol. Chem. 175, 67 (1948). (6) Chipault, J. R., McMeans, Evelyn, J. AGR.FOODCHEM.13, 15 (1965). (7) Davies, W. L., Gill, E., J . SOG. Chem. Ind. 55, 141-T (1936). (8) Feigl, F., “Spot Tests in Organic Analysis,” 5th ed., Elsevier, New7 York, 1956. (9) Forss, D. A., Dunstone, E. A., Stark, it’.?J . Dairy Res. 27, 211, 373 (1960). (10) Gauglitz, E. J., Gruger, E. H., J . A m . Oil Chemists‘ Soc. 42, 561 (1963). (11) Herbst, E. J., Keister, D. K., il’eaver, R . H., Arch. Biochem. Biuphys. 75, 178 (1958). (12) Noble, E. D., Anal. Chem. 27, 1413 (1955). (13) Russell, J. A, J . Bid. Chem. 156, 457 (1944). (14) Stansbv. M. E.,’ Food Technol. 16, 4, 28 11962). (15) Steiner. M.. Stein von Kamienski. ‘ E., nlaturwissenschaften 40, 483 (1953).’ ~
so.
2
,
Received for reuiew December 8, 1965. Accepted March 14, 1966. Study supported by the U. S. Department of the Interior, Fish and Miildlife Service, and by The Hormel Foundation.
1 4 , NO. 3, M A Y - J U N E
1966
231
