This paper is the first report on an extensive

CHEMISTRY I. FATTY BY OF SLASH-PINE MORELET) CONSTITUENTS HALL (PINUS CARIBIEA, OF THE PHLOEM AND J. ALFRED Products Department OLE GISVOLD (From the Forest Laboratory, * Forest Service, United Slates of Agriculture, Madison) February 25, 1935) Downloaded from www.jbc.org by guest, on November 10, 2009 (Received for publication, This paper is the first report on an extensive investigation under way on the chemical mechanisms involved in the formation and possible translocation of oleoresin within the tree. The necessarily voluminous discussion of the theories of phytosynthesis and metabolism involved has been partly covered already (1). The determination of the nature of the constituents in tissues conceived to be actively associated with the formation of oleoresin is required as a preliminary to further work on seasonal variations. It is hoped that the investigation will not only aid in clarifying the processes of oleoresin synthesis, but also the larger problems of metabolism and translocation within the tree. The slash-pine (Pinus caribaa, Morelet) of the southeastern United States is not only an important and typical oleoresinproducing tree, but is available in extreme profusion in pure stands of young saplings that are well suited for investigations in which a differentiation of tissues is desired. This species has therefore been selected for intensive examination. Microscopical investigations by Gerry (2) have shown that the phloem of slash-pine is extremely rich in labile materials as compared with woody tissue. Devaux and Bargues (3) demonstrated the presence of either oleoresin or oleoresin-yielding material in the phloem and rays of Pinus pinaster (maritime pine). Since the rays are closely associated with the resin passages of the xylem, and form, in effect, a physiological link between these and the * Maintained consin. at Madison, in cooperation 585 with the University of Wis- 586 Chemistry of Slash-Pine. I phloem, the latter was selected as the starting point for the investigation. The question of the existence of oleoresin as such or in combination prior to its appearance in the resin passages of the xylem cannot be considered as settled, since no entirely reliable staining technique exists for distinguishing between fats and oleoresins in situ. Furthermore, no oleoresinous constituents have been detected in the petroleum extract of the phloem. This extract consists entirely of fatty and unsaponifiable material. The ether extract following the petroleum extract apparently contains no free oleoresinous constituents, but does yield a very small amount of a complex glycosidic substance which contains combined oleoresinous bodies. This substance will be discussedin a subsequent article. The present paper deals exclusively with the constituents found in the petroleum (60-70”) extract of the phloem. EXPERIMENTAL Downloaded from www.jbc.org by guest, on November 10, 2009 The material examined was collected in April, 1934, in the forests near Cogdell, Georgia. At the time of collection the top growth was in the stage of bud elongation and oleoresin production from working trees was entering the period of vigorous activity. Saplings of 2 to 4 inches diameter were felled and topped well below the preceding season’s growth in order to avoid chlorophyll-bearing phloem. The cortex was carefully removed and the phloem peeled from the wood in long, narrow strips. These were dried in a shelf drier in a current of air at 45” for about 9 hours to a brittle condition. They were then broken up and packed in tight containers for shipment to Madison. From tree to container, the longest elapsed time was about 14 hours. At the laboratory the dried material was ground in a Wiley mill to approximately 40 to 60 mesh, packed in air-tight containers, and kept at 3” until ready for extraction. The ground material contained 5.4 per cent water and 5.5 per cent of material extractable in petroleum of boiling range 60-70”. 38 kilos were completely extracted with such petroleum in a continuous percolator and from the extract 2 kilos of crude fat were recovered by evaporation of the solvent under a partial vacuum. The product was slightly greenish brown in color. It had a saponification value of 137. J. A. Hall and 0. Gisvold The crude fat was saponified with alcoholic potassium hydroxide in the usual manner, the alcohol removed under a vacuum, and the soaps dissolved in water (2.5 liters per 500 gm. of fat). These solutions were exhausted with ether in continuous extractors. The ethereal extracts, upon evaporation, yielded yellow, semicrystalline masses (Fraction A) which were examined. The soap solutions were decomposed with a slight excess of sulfuric acid. A layer of ethereal solution of fatty acids floated above a flocculent precipitate on the aqueous layer. This precipitate was removed by centrifuging and examined (Fraction B). The fatty acids (Fraction C) were collected by extra&ion of the aqueous liquids with ether and evaporation of the solvent. All crystalline preparations were dried for analysis at 56”, 0.1 mm., for 24 hours over phosphorus pentoxide. Rotations were observed with a Franz Schmidt and Haensch (Berlin) quartz wedge saccharimeter equipped with a Ventzke scale and an electric sodium lamp as light source. The melting points were made with a Thiele apparatus with a thermometer calibrated against a Bureau of Standards thermometer and are not corrected for stem emergence. Carbon and hydrogen were determined by a semimicromethod developed from the original Pregl procedure. Unsaponijiable Matter (Fraction A) Downloaded from www.jbc.org by guest, on November 10, 2009 This fraction amounted to approximately 200 gm. of a yellow, oily mass containing some crystalline material. By extensive fractional crystallization from alcohol, six fractions of crystals were obtained (total 3 gm.), all of which gave the LiebermannBurchard sterol reaction and melted at 137-138”. One of these fractions was purified further by the modified (4) digitonin procedure of Windaus (5). The purified sterol was unchanged in melting point after crystallization from alcohol. Analysis--[a]: = -23.4” (0.04 Cz9Hso0. Calculated. Found. gm. in 5 cc. of CHCl, 2 dm. tube). C 84.06, H 12.08 “ 83.53, 83.56, “ 12.33, 12.33 The acetate was prepared by heating with acetic anhydride, and crystallizing from alcohol until it showed a constant melting point of 126”. Chemistry Analysis-[a]: CJMh. of Slash-Pine. I = -29.7” (0.05 gm. in 5 cc. of CHCL, 2 dm. tube) Calculated. C 81.22, H 11.35 Found. “ 80.98, 81.10, “ 11.62, 11.64 For purposes of comparison, a sterol of the same melting point from the seeds of Pinus sabiniana was purified in the same manner and analyzed. Analysis-[a]: = -22.7”; C 83.64, 83.74, H 12.44, 12.22. All the available sterol (4 gm.) was purified over the digitonide, acetylated, and the acetate crystallized from alcohol. The last fraction (1 gm.) was rejected, and saponification equivalents determined on the first two by the procedure of Sandqvist and Bengtsson (6). CJbOa. Calculated. Found. Fraction “ 1. 2. Saponification “ I( equivalent 1‘ “ 123 122.48, 123.04, average 122.76 121.4 , 123.4, average 122.4 Downloaded from www.jbc.org by guest, on November 10, 2009 Since it has been shown that this procedure is far more reliable in determination of empirical form&e. for sterols than combustion analysis, the formula CkH600 is proposed for this sterol. The melting points are those of sitosterol and the rotations are within the range of possible mixtures of its various stereoisomers. Therefore, although the combustion figures actually agree much better with other formuke, the substance is probably a sitosterol. Alcohol, C16N340-From the mother liquors of the sterol fractions about 1 gm. of a waxy body was obtained, which separated in poorly formed leaflets. It was soluble in all ordinary organic solvents. After several crystallizations from a mixture of alcohol and ether the melting point remained constant at 74-75”. The substance did not absorb bromine and gave no sterol reactions. Analysis--[a]! = +9.8” (0.447 gm. in 5 cc. of CHCl,, 2 dm. tube). C16H8d0. Calculated. C 79.34, H 14.05, mol. wt. 242 Found. “ 78.84, 79.13, “ 14.19, 14.29, “ ‘( 241 (Rast) The acetate, prepared by boiling with acetic anhydride, at 64-65”. melted J . A. Hall and 0. Gisvold 589 This alcohol reacted with Beckmann’s chromic acid mixture, but when warmed with the latter in the amount calculated for the oxidation of a primary or secondary hydroxyl to aldehyde or ketone, only the starting material was recovered in diminished quantity. No carbonyl compounds could be isolated. The substance gave a negative reaction to Deniges’ reagent (7) and is therefore probably not a tertiary alcohol. The known alcohol of the above formula, cetyl alcohol, melts at 49” and is optically inactive. This substance is either a secondary alcohol or a primary alcohol with a branched chain containing an asymmetric center. Probability favors the former. Since it is apparently a hitherto unknown substance, the name isocetyl alcohol is suggested. TABLE I Oily Unsaponifiable Downloaded from www.jbc.org by guest, on November 10, 2009 Distillation Amount cc. of Material Remarks B.p. (0.005 mm. approximate) “C. 5 50 45 25 15 3 80-100 169-170 170-172 172-173 190-200 200-210 1.4591 1.4525 1.4533 1.4557 1.4695 0.8749 0.8731 0.8750 0.8795 Temperature rose very rapidly On liquid, part was solid Solid Tarry residue The remainder of the unsaponifiable material consisted of 225 cc. of a yellow oil. This was distilled under a pressure of approximately 0.005 mm. as shown in Table I. The solid portion of Fraction 5 was separated by pressing upon porous tile. After several crystallizations from a mixture of alcohol and ether, the melting point was constant at 74-75” and no depression was observed when mixed with the alcohol, C16H340, previously described. Fraction 6 consisted of a mixture of this alcohol with the sterol previously described. From a small portion of Fraction 7, the sterol was again isolated over the digitonide. Hence, it appears that Fractions 5 to 7 consisted largely of these two substances. Aside from a small quantity of a lower Chemistry of Slash-Pine. I boiling substance in Fraction 1, Fractions 2 to 4 evidently consisted almost entirely of a single body or a mixture of closely similar compounds. These fractions did not give the Liebermann-Burchard sterol reaction, but developed a reddish orange color with the reagent. From Fraction 3, a crystalline bromide was prepared in glacial acetic acid at 0”. After repeated crystallization from dilute methyl alcohol, it appeared as platelets, melting point 123-125”. It was not pure and was unstable, so could not be prepared for analysis. The original oil of Fraction 3, however, was analyzed with the following results. la], = G0HS603. -0"; iodine No. 104.5. Calculated. C 77.59, Found. “ 77.72, H 12.07, “ 12.13, mol. “ wt. I‘ 464 466 (Rast) Downloaded from www.jbc.org by guest, on November 10, 2009 77.63, 12.14, The above formula and iodine value are satisfied by a cycle and two double bonds. The substance could not be acetylated or benzoylated. A determination of hydroxyl by the method of No colorations were given Zerewetinoff gave no active hydrogen. with fuchsin bisulfite or sodium nitroprusside, nor could any addition reactions be obtained with a number of carbonyl reagents. Methoxyl determinations by the procedure of Clark (8), with the temperature of the condenser maintained at 4’, gave 6.6 per cent (theory, (OCHZ) 6.65 per cent). Somewhat higher values were obtained with higher condenser temperatures. The aeberTollens reaction (9) for methylene oxide groups gave a distinctly positive result although not nearly so pronounced as in the case of This substance is evidently a complex aromatic methylene ethers. ether containing one methoxyl group and other ether linkages of unknown character. Fraction B. Isolation of a Phytosterolin Fraction B, separated by centrifuging from the et,hereal acid solutions, was repeatedly washed with hot water and alcohol. It was practically insoluble in alcohol or ether, but was repeatedly crystallized from hot dioxane as rather poorly defined crystals. Two fractions were obtained: Fraction 1, 0.5 gm., m. p. 290”; Fraction 2, 0.4 gm., m. p. 225’, with not very well defined melting points. Neither fraction was entirely free from ash. Both gave the Liebermann-Burchard sterol reaction. J. A. Hall and 0. Gisvold Analysis of Fraction C,,HsuOa. 1 Calculated. Found. (Corrected C 72.9, H 10.4 “ 72.1, 72.3, “ 10.5, for ash 0.32, 0.41) 591 10.7 The analysis is in fair agreement with the formula for a hexoside of the sterol oc&rring free, C&HsoO. The substance was hydrolyzed with amyl alcoholic HCl according to Power and Salway (lo), and a sterol isolated melting at 136-137”. Analysis-[[cu]f: = -24”. Cz9Hso0. Calculated. Found. C 84.06, “ 83.40, H 12.08 “ 12.22, Downloaded from www.jbc.org by guest, on November 10, 2009 83.35, 12.21 The aqueous portion of the hydrolysate reduced Fehling’s solution and gave a positive Molisch reaction, but no osazone could be prepared. Fraction 2, according to analysis, was obviously impure. It was hydrolyzed as before but was refluxed for 1 hour instead of 2. The sterol recovered melted at 136-137”. Analysis--[a]: = -25”; C 83.53, 83.44, H 12.62, 12.61. In this case, the aqueous hydrolysate yielded a phenylosazone melting at 204”. Evidently both these fractions consisted of a phytosterolin. Fraction C. The Fatty Acids Identijication of n-Caproic Acid-The ethereal solution of fatty acids after removal of the sterolin fraction was evaporated and the residue distilled with steam. The acid distillate was extracted with ether and the ethereal solution shaken out with sodium carbonate solution. From this, by acidification and extraction with ether, 0.5 cc. of an oily acid was obtained, which had the odor of one of the lower fatty acids. A Duclaux distillation indicated that the preparation was probably impure. It was converted over the sodium salt to the p-bromophenacyl ester (11). After repeated crystallizations from dilute alcohol the ester melted constantly at 69-70”. The corresponding n-caproate (from n-caproic acid, Eastman) melted at 70-71”, and the mixture of the two at 69-70”. Mixed melting points with the n-valerate (m. p., 63”) and isovalerate (m. p., 68”) gave large depressions. The acid was therefore identified as n-caproic acid. 592 Chemistry of Slash-Pine. I The ethereal solution after extraction with sodium carbonate yielded 0.1 cc. of a neutral fragrant oil with the odor of a high aliphatic alcohol. The non-volatile fatty acids were separated into solid and liquid fractions according to the procedure of Twitchell (12). When the lead salts of the solid acids were precipitated from 3430 cc. of 95 per cent alcohol per 200 gm. of mixed acids at +4’, the precipitated salts carried down a considerable quantity of semisolid salts which could be easily separated from the solid salts by means of ether. This fraction was later found to consist of almost pure oleic acid, while the fraction obtained from the lead salts soluble in alcohol at 4’ was fairly rich in linoleic. Solid Acids-The solid acids were recovered from the lead salts in the usual manner. Identification of Palmitic Acid-25 gm. of the crude mixture were distilled at 0.01 mm. A few gm. only were distilled over at 170”. This product, recrystallized from alcohol, melted at 62”. No depression was observed when mixed with palmitic acid (Eastman), melting point 62.5’. Analysis GsHssO~. Downloaded from www.jbc.org by guest, on November 10, 2009 Calculated. Found. C 75.00, H 12.50 “ 75.11, “ 12.83 The p-nitrobenzyl ester melted at 4243” and showed no depression with the corresponding palmitate. The methyl ester gave a saponification number of 208 (methyl palmitate, 208). Identification of DodecosanicAcid (Behenic)-The main portion of the crude acids was extensively fractionally crystallized from mixtures of alcohol and ethyl acetate without obtaining a pure fraction as determined by melting points and neutralization values. Fractional distillation of the methyl esters gave some pure methyl palmitate, but the higher fractions were mixtures. However, clean separations were obtained by the following procedure. 5 gm. of mixed acids dissolved in 200 cc. of alcohol were neutralized while hot with 0.1 N alcoholic potassium hydroxide. Upon cooling, a potassium salt separated as a gel which was easily broken up by stirring, filtered off by suction, and washed with cold alcohol. Repetition of this process on the acids obtained from the insoluble potassium salts gave clean separations of palmitic acid from the acid contained in the insoluble salt. J. A. Hall and 0. Gisvold The acid recovered from the insoluble potassium salt, and recrystallized three times from a mixture of alcohol and ether, melted constantly at 78-79”. Analysis CzzHJa02. Calculated. C 77.65, H 12.94, N. E.* 165 “ 77.53, 77.74, ” 13.31, 13.13, “ “ 164.5 Found. * Neutralization equivalent. The p-bromophenacyl ester melted at 91-92’ and thus cannot be distinguished from the stearate by this means. However, when mixed with the stearate in approximately equal amounts, the melting point dropped to 85”. The methyl ester, prepared by refluxing with absolute methyl alcoholic HCl and recrystallized several times from a mixture of 60: 40 isopropyl ether and ethyl acetate, melted constantly at 61-61.5”. However, consistent carbon and hydrogen values could not be obtained upon combustion. l A careful examination of all fractions failed to reveal the presence of stearic acid. The solid acids, therefore, consisted entirely of palmitic and dodecosanic acids, with the latter in slightly larger proportion. Liquid Acids. Identijkation of Oleic and Linoleic Acids-From the fraction of the lead salts which was only slightly soluble in alcohol at +4” and soluble in ether, about 90 gm. of liquid fatty acids were obtained. These were distilled at 0.01 to 0.015 mm. at 190-194” with very little residue. The distilled acids gave iodine values indicating a mixture and were converted again to lead salts. These were fractionally recovered from alcoholic solution by evaporation. The first two fractions of recovered acids (26 gm., 13 gm.) were mixed and analyzed. Analysis Downloaded from www.jbc.org by guest, on November 10, 2009 Ct8H3402. Calculated. Found. C 76.60, H 12.06, N. E. 198.6, I No. 90.07 “ 76.62, “ 11.45, “ “ 195, “ “ 93 (Wijs) 1 Some confusion exists as to the melting point of methyl behenate. Values reported are: Klein (Handbuch der Pflanzenanalyse, Vienna, 2, 392 (1932)), 52”; Sudborough, Watson, and Ayyar (J. Indian Inst. SC., 9A, 25 (1926)), 52”; Tutin and Clewer (J. Chem. Sot., 106, 1845 (1914)), 51”; Power and Salway (J. Chem. Sot., 106, 201 (1914)), 58-59”. Although the preparation reported here is admittedly impure, it melts at 61-61.5”. It is possible that various workers have dealt with isomers. 594 46-47”. Chemistry of Slash-Pine. I The p-bromophenacyl ester was prepared and found to melt at No depression occurred when it was mixed with the corresponding ester of oleic acid (from Kahlbaum oleic, m. p. of ester, 47”). The main portion of the liquid acids was therefore oleic acid. That portion of the lead salts which was completely soluble in alcohol at +4” was fractionated by evaporation of the alcoholic solution, and the small fraction most soluble in alcohol separated. The acid recovered from this lead salt had an iodine value of 141. Bromination of this fraction according to the procedure of FarnSteiner (13) yielded a bromide, insoluble in petroleum ether, and melting at 112” (linoleic tetrabromide, m. p. 115”). This bromide was completely soluble in ether, indicating the absence of linolenic acid. Analysis CIS&#&~. Calculated. Found. C 34.67, “ 34.65, H 5.33, “ 5.58, Br 53.3 “ 53.1 Downloaded from www.jbc.org by guest, on November 10, 2009 34.71, 5.57, The liquid acids therefore tion of linoleic acids. consisted of oleic and a small propor- SUMMARY A petroleum ether (60-70”) extract of the phloem Morelet, yielded the following substances: a C29H500; a sterolin, C&H6006; a wax alcohol, C&H340; ether, &,Hs603; and n-caproic, palmitic, dodecosanic, linoleic acids. caribsa, of Pinus sitosterol, a complex oleic, and All combustions and rotations and several other determinations on small quantities of material were performed in the Biochemical Research Laboratory of the University of Wisconsin by Dr. Eugene Schoeffel. The services of this laboratory were placed at our disposal by Professor Karl Paul Link. BIBLIOGRAPHY 1. 2. 3. 4. Hall, J. A., Chem. Rev., 13, 479 (1933). Gerry, E., U. S. Dept. Agric., Misc. Pub. 209, 33 (1935). Devaux, H., and Bargues, A., Bull. Inst. Pin, No. $7, 131 (1927) Gisvold, O., Dissertation, University of Wisconsin (1934). J. A. Hall and 0. Gisvold 5. 6. 7. 8. 9. 10. 11. 12. 13. 595 Windaus, A., Z. physiol. Chem., 66, 110 (1910). Sandqvist, H., and Bengtsson, E., Ber. them. Ges., 64,2167 (1931). DenigBs, G., Compt. rend. Acad., 126,1277 (1898). Clark, E. P., J. Assn. Of. Agric. Chem., 16, 136 (1932). Weber, K., and Tollens, B., Ann. Chem., 299, 316 (1898). Power, F. B., and Salway, A. H., J. Chem. Sot., 103, 399 (1913). Reid, E. E., and Judefind, W. L., J. Am. Chem. Sot., 42,1043 (1920). Twitchell, E., J. Znd. and Eng. Chem., 13,806 (1921). Farnsteiner, K., Z. Untersuch. Nahrungsu. Genussmittel, 2, 17 (1899). Downloaded from www.jbc.org by guest, on November 10, 2009