zyx zyxw zyxwv Isolation and Structure of Sesangolin, a Constituent of Sesam urn angolense (Welw .) WILLIAM A. J O N E S ~ ,I~O R T REROZA, OX A N D EDWISD. BECKER? Entomology Research Dizizsion, Agmcultural Research Service, U.S. Department of Agriculture, Beltsvzlle, Maryland, and the lTafionallnstztutes of Health, Bethesda, 'Ifarylanrl Receii,ed March 9, 1962 The seed oil of the wild sesame plant, Sesumum angolense is unusually high in synergistic act,ivity with pyrethrum. The oil was shown to contain two synergists, sesamolin and a previously undescribed compound that we have named sesangolin. Analysis of the latter indicated the presence of one methoxyl, two methylenedioxyphenyls, and two benzyl ether groups. Aliphatic double bonds, C-methyl groups, and acetal structures (aside from those on the methylenedioxyphenyl groups) were absent. Permanganate oxidation yielded piperonylic and 6-methoxypiperonylic acids. Nit,ric acid oxidation yielded the dextrorotatory di-?-lactone of a,p-bis( hydroxymethy1)succinic acid. Based on these fragments sesangolin is 2-(3,4meth~lenedioxypheny1)-6-(6-methox~-3,4-methylenedioxyphenyl)-3,7-dioxabicpclo[3.3.0]octane and its bridgehead hydrogens are in a cis configuration. Proton nuclear magnetic resonance spectra were especially useful in elucidating the structure of the compound and a detailed examination of the spectrum supports the assigned formula. Sesangolin is about equal to sesamin in ability to increase the insecticidal potency of pyrethrum. zyxwvu An euamiuation of the seed oil (sesame oil) from thirty-three different strains of Sesamum ?ndicum3 disclosed a maximum amount of ea. 1.5% of sesamin and sesamolin, the only ingredients of sesame oil known to be synergistic with the insecticide pyrethrum. The report of Pearman, ec aZ.,' stating that a seed oil derived from a wild sesame plant of Northern Rhodesia, Sesamum angolense (Welw.), contained 9% sesamin (based on synergistic action with pyrethrins) was therefore of great interest, especially since such an oil may be a valuable source of synergistic materials. The high sesamin value, which was based on bioassay, may have been caused by the presence of sesamolin, a compound known to hare a synergistic activity as much as five times that of sesam1n.j Pearman, et al., did not consider sesamolin as an iligredient because at the time of their work sesamolin was not known to be a synergistic component of sesame oil. They also reported that their oil contained 7.4.% of unsaponifiables, a mlue obviously inconsistent with a 9Oi:, figure for the unsaponifiahle sesamin. These considerations led us to investigate the constituents of the wild sesame oil The petroleum ether extract of S . angolenar seed, upon removal of the solvent, yielded oil and crystals.6 Analysis by the method of Suarea7 of the oil made homogeneous by heating until the crystals dissolved indicated that it contained 3.73% sesamin and 2.69a/, sesamolin. This result confirmed that the percentage of synergistic constit- urnts in the oil was indeed large (totaling 6.42%), but the sesamin content was much below the 9% value reported by Pearman. Ilowever, if the greater synergistic activity of sesamolin as compared with sesamin is taken into account, the high biological activity reported by Pearman, et al., appears to be verified. Because of tlie lack of specificity of Suarez's ultraviolet analysis (most compounds with the 3,4methyleiiedioxyphenyl structure would absorb strongly in approximately the same wave length region) and the fact that the oil came from a new plant species, the isolation of the sesamin and sesamolin from the oil of S. angolense seed was attempted in order to prove their presence. Chromatography of the oil mixture on silicic acid gave two distinct zones. Although the first yielded sesamolin, the second did not give sesamin, but a compound different from either sesamin or sesamolin. X o reference to a previous isolation of the compound could be found and accordingly the name sesangolin is proposed for it. Based on the experimental evidence, detailed below, sesangolin has the structure shown in formula I. zyxwvutsrqpon zyxwvuts zyxwvutsrqpon (1) P a r t of this work n a s submitted b, R'. 4 Jones a s II S thesis t o the American Unirersit\ \? ashineton D C. (2) National Institutes of Health, Bethesda, Maryland (3) M Beroia and I1 L K nman J .4m 021 Chemzsts Soc 32, 7 4 8 (1955) (4) R. W Pearman TI' D Raymond, and J \ Squires Coionzal Plant Anzmal Prod IGr Brit ) 2 , 297 (19.51) (5) II Beroia J .4m Oil Chemisfs' S o c , 31, 302 (1954) (6) The oil was kindly supplied by RIr G D. Glynne Jones, PJ rethrum Board of Kenya, Nakuru Kenqa Colony, Africa (7) C C Suarei R T O'Connor E T Field, and W. G Bickford, A n d Chem,, 24, 668 (1862) 1 The synergistic action of sesangolin was found to equal that of sesamin, increasing the toxicity of pyrethrins to house flies approximately threefold in an equiproportional mixture.* At first sesaiigolin melted at 87-88') but in later experiments it melted a t 101'. Cross-seeding experiments established the higher melting form as the more stable. (8) The tests against house flies were carried out by W. A. Gersdorff and P. G , Piyuett, of the Entomology Researeh Division. SESANGOLIN SEPTEMBER, 1962 The purity of sesangolin was indicated by its sharp melting point, and by the single zone with constant ultraviolet absorption ratios9 obtained for each of the fractions in the chromatography of the oil. The purity of sesangolin was further established by our inability to fractionate it by gas and paper chromatography. Elemental analyses and molecular weight determinations pointed to CzlH& as the formula of sesangolin. The compound contains one methoxyl grouplo but no carbon-linked methy1.l' Since it did not decolorize potassium permang:uiate in acetone e w n wheii heated to 60' for sewral minutes, aliphatic double bonds are absent. h Villavecchia color test,' which indicates the presence of a 3,Pmethylenedioxyphenyl acetal structure such as is present in sesamolin, was negative. S o acetal structure (aside from those on the methylenedioxyphenyl groups) is present because sesangolin survived acid hydrolysis. The hydrogenation of sesaiigolin in acetic acid with a palladium-charcoal catalyst resulted in the rapid uptake of two moles of hydrogen, after which the compound absorbed hydrogen very slowly. The rapid uptake of the two moles of hydrogen is consistent with the presence of the two benzyl ether groups shown in the sesangolin formula. The nitration of sesangolin yielded a diiiitro compound. The ultraviolet spectrum of sesangolin is practically identical with that of sesamin and sesamolin,l? when compared on a molar basis. The close similarity of the chromophoric groups of these compounds was therefore established and the presence of two methylenedioxyphen y1 groups in the sesangolin molecule was strongly indicated. Methylenedioxyl analysesl3 furthrr supported the presence of these groups. The infrared spectrum of sesangolin provides additional evidence for the presence of methvl- 3233 dioxyphenoxy group, does not have a peak in this region. Inspection of the sesangolin formula shows that the methoxy-aryl group is a methylenedioxyphenoxy structure. Sesangolin, when subjected to prolonged permanganate oxidation in refluxing acetone, yielded piperonylic and 6-methoxypiperonylic acid. The isolation of these acids from sesangolin confirms the presence in this molecule of one 3,bmethylenedioxyphenyl group and one 6-methoxy-3,4-methylenedioxyphenyl group, both attached to carbon atoms. There remained oiily the problem of ideii tifyiiig the central nucleus which calculated by difference has a formula of C6H802. This formula is identical with that of the central nucleus of sesamin, sesamolin, asarinin, pinoresinol, and eudesmin. These compounds are known to have the fused tetrahydrofuran structure shown in formula I1 with substituted phenyl groups a t positions 2 and G (or 4 and 8). n'itric acid oxidation of these zyxwvu zyxwvu zyx zyxwv 7 6H1 3 I1 zyx compounds gives an optically active di-7-lactone of a,P-bis(hydroxymethy1)succinic acid (111),the oxo groups fixing the carbon atoms to which the aryl 0 H2C< %=O H \ /c-c \ I \ \ I11 zyxwvutsrqp Proton nuclear magnetic resonance (n.m.r.) (9) M. Boroza. A n d . Chem., 2 2 , 1507 (14.50). >lethod of E. p. Clark. "Seli,imicro Q,lantitati,.e organic studirs were especially useful in the complete struc.\nalysis," Academic Press, Ine., N e w York, N. Y , , 19-13, tural elucidation and actually preceded some of the f l l ) Method of \V. F.Barthel and F. B. Layorre. I n d . i:'ng. Chem., oxidation studies already described. Figure 1 1 6 , 434 (1911). ShOwS the 1i.m.r. spectrum of sesangolin and for (12) P. Budowski, R . T. O ' c o n n o r . E, T. Fipl<i, .I. <tm. o,i Chemists' S o c . . 28, 51 (1951). (13) >Beroza, I. .4nal. Chem., 2 6 , 1970 (1964). (1.5) H. Erdtnisn and J. Gripenbere. .Ictu Chim. Scund., 1 , 7 J (1H47). (14) .\I. Beroza, J . A m . Chem. Sac., 7 1 , 3332 (1955). zyxwvutsrqponmlkjihgfedcbaZYXWV JONES, REROZA,AND BECKER 3234 VOL. 27 The methoxyl group must therefore be in the position indicated in I. Isolation of G-methoxypiperonylic acid from the oxidation of sesangolin (as I/ 695 described above) confirms this conclusion. The remaining features of the sesangolin spectrum are quite similar to those of sesamin, as indicated in Fig. 1. The unresolved multiplet centered a t 175 is due to the two bridgehead protons. The unusual chemical shift for such protons on carbons p to an oxygen is apparently due to the fused tetrahydrofuran ring system. The complex group at ca. 220-270 is due to the two CH2 groups a to the oxygens. The spin coupling in this group has not been analyzed. The t v o doublets a t 274 and 298 are clearly due to the protons CY to both the oxygen and the phenyl groups. The 24 C.P.S. difference in their resonance frequencies indicates an appreciable chemical non-equivalence, whereas the correspoiiding protons in sesamin are equivalent, giving rise to a doublet a t I I 279. The non-equivalence of these protons in 4b2 3$O 2i9 I19 0 sesangolin might result from the presence of the Fig. 1.-N.m.r. spectra of ( A ) sesangolin and (B) sesamin. methoxy on one of the aromatic rings, but the difFrequencies given in c.p.8. from internal tetramethylsilane. ference seems large. Asarinin, the asymmetric stereoisomer of sesamin, has two doublets a t 261 comparison the spectrum of sesamin. Positions and 286, similar to sesangolin, but also shows of lines (at 60 Mc.) are given in C.P.S.from internal resonance lines in the 190-220 region, where sesantetramethylsilane. The lines a t 348 and 350 golin and also sesamin have no lines. We believe are clearly ascribable to the two methylenedioxyl that a more complete analysis of the spectra of groups which are in slightly different chemical sesangolin as well as of sesamin and its stereoisoenvironments. The strong line a t 222 is due to mers is needed before the n.m.r. data can furnish the methoxyl group, the chemical shift suggesting reliable information on the configuration of sesanan aryl rather than alkyl ether (e.g., Jackman16 golin. gives typical values of 224 and 197, respectively). The n.m.r. spectrum thus is seen to be in comThe integral of the spectrum (Fig. 1) shows un- plete accord with formula I and in fact with the questionably that there are only five aromatic other evidence furnishes virtually conclusive proof protons (in the region 386410). Thus one of the of the correctness of this structurc. methylenedioxyphenyl groups must be additionally substituted, presumably by the methoxyl group. Experimental The exact position of the methoxyl is established as follows: Comparison of the aromatic regions Isolation of Sesango1in.-The initial isolation was acof sesangolin and sesamin (as well as other methyl- complished essentially according to the method of Beroza.I8 enedioxyphenyl compounds not shown here) indi- A 2-3-g. sample of S. anydense oil was dissolved in 1:4 chloroform-isooctane and introduced on an 80-g. silicic cates that the three aromatic protons in the un- acid (Merck) column that had been prewashed with 100 ml. substituted methylenedioxyphenyl ring are respon- of 15% ethyl acetate in isooctane and 75 ml. of isooctane. sible for the lines at 402 and 406. The lines a t 386 The column was developed with 3% ethyl acetate in isoooand 410 are of approximately equal intensity, tane, sesamolin being eluted between ’700 and 1200 ml. and between 1200 and 2000 ml. At 50-ml. intervals and the integral shows that the line at 386 is due to sesangolin the absorption a t 280, 288, and 300 mp was determined in an a single proton. The lines at 386 and 410 are thus ultraviolet spectrophotometer The eluate from the first established as arising from the two ring protons of zone gave a 288/300-mp absorbance ratio of approximately the methoxyl-substituted methylenedioxyphenyl 2.0, a value in agreement with that of sesamolin. A white group. These protons must be para to each other crystalline solid, suhsrquently isolated in 3.370 yield from this eluate, was determined to he sesamolin by m.p. (90since the lines are not split. Protons ortho or meta 92”), mixed melting point with an authentic sample, to each other are known to be coupled by approxi- infrared and ultraviolrt spectra The 2881’300-mp absorbmately 8 and 3 c.P.s., respectively, whereas protons ance ratio of the eluate from the second zone was 1.2 (corpara to each other are coupled by less than 1 c.P.s.” responding value of sesamin is 7.0). A white crystalline A I NUMBER OF PROTONS zyxwvutsrq zyxwvutsrqponmlkjihgf zyxwvutsrqponmlkjihgfe zyxwvutsrqponmlkjih zyxwvu zyxwvutsrqponm (16) L. M. Jackman, “Applications of Nuclear IIagnetic Resonance Spectroscopy in Organic Chemistry,” Pergamon Press, New York, iV.Y., 1959, p. 55. (17) J. A. Pople, TV. 0.Schneider, and H. J. Bernstein, “High Resolution Nuclear Magnetic Resonance,” MoGraw-Hill, New York, N. T.,1959, p. 193. solid (sesangolin) melting a t 87-88” after two crystallizati 3119 from methanol was obtained from this eluate; yield 3.3%. Anal. Calcd. for CztH2007: C, 65.62; H, 5.24; mol. wt., 384. Found: C, 65.66, 65.43; H, 5.36, 5.33; mol. wt. (18) M.Berosa, Anal. Chem., 26, 1173 (1954). SEPTEMBER, 1962 SESANQOLIN 3235 zyxw zyxwvutsrq zyxwvutsrqpo zyxwvutsr (Rast), 380; mol. wt. (Signer using methylene chloride),lo 383. In a mixture with sesamolin, sesangolin melted over a wide range, starting to melt a t 74'. To determine whether the solid crystalline residue that separated upon removal of solvent from a petroleum ether extract of S. angolense seed was largely sesangolin, the solid was filtered off, washed with isooctane, and crystallized from absolute ethanol. The resulting crystals melted sharply a t 101'. Tiic identity of this compound was in doubt until its infrared spectrum was compared and found identical with that of sesangolin (m.p. 88'). Polymorphism was confirmed as follows: A small amount of the 101' material was dissolved in a minimum of hot alcohol and seeded with the 88' material. The crystalu, which separated on cooling, melted a t 101". Similarly a crystal melting a t 101' was used to seed the 88" material dissolved in a minimum of hot alcohol. It also produced crystal3 melting a t 101'. Once the 101" material was obtained, the lower melting material could not be produced. The higher melting crystals therefore constitute the more stable form. Sesangolin was also isolated most conveniently by the method of Tracy19 for the extraction of pyrethrum synergists from sesame oil from a sample of S. angolense oil from which t,he solid had been removed: Distilled butyrolactone (1250 ml.) was added to 580 ml. of the oil in a 3-l., three-neck Eask fitted with a thermometer, stirrer, condenser, and heating maiitle. After heating the mixture with stirring a t 150" for 2 hr., the mixture was allowed to come to room temperature while being stirred. The upper layer was feparated and the butyrolactone distilled, leaving a residue that was taken up in petroleum ether (b.p. 60-70"). After rpmoving the petroleum ether-soluble portion, the residue was crystallized three times from absolute alcohol and twice from amyl alcohol; yield was 5 g. of a product melting a t 101'; [a12311+48.5" ( c 0.54 chloroform). TheVillavecchia color test was negative, showing that no sesamolin was present. Gas Chromatography of Sesangolin.*o-The gas chromatography was carried out a t 180' on a 6-ft. column of Gaschrom P impregnated with 0.75% SE-3021 using argon m the carrier gas with a flow rate of 15-17 cc./min. and an ionization unit as detector. A single symmetrical peak (other than that of the solvent) occurring a t approximately 40 nGn. was obtained during an 80-min. run. Paper Chromatography of Sesangoh-This procedure, which identifies 3,4-methylenedioxyphenyl synergists by reversed-phase paper chromatography using 30% aqueous acetic acid as the developing solvent,zz gave a single peak a t RJ 0.54 for sesangolin. Dinitrosesangoh.-A test tube containing 90 mg. of sesangolin in 0.6 ml. of glacial acetic acid was swirled in a 40" bath while 0.75 ml. of 2:3 by volume concentrated nitric-acetic acids wm added dropwise over a 10-min. period. After another 5 min. in the bath the test tube was removed, cooled, and 5 ml. of water added with swirling. In a few minutes crystals began to form. They were filtered off, washed with water, dried, and recrystallized from absolute ethanol; wt. 28 mg., m.p. 153-154'. Anal. Calcd. for C~lHuOllNt: C, 53.17; H, 3.82; N, 5.91. Found: C, 53.28; H, 4.12; N, 5.08, 5.14. Ultraviolet Spectrum of Sesangolin.-A solution containing 0.031 mg. per ml. of isooctane exhibited peaks at 236 and 293 mp. The molar extinction coefficients are: max. 293 mr, 7870; min. 257 mp, 622; max. 236 mp, 8575; min. 222 mp, 5830. Infrared Spectrum of Sesangoh-The infrared spectrum of a 1% solution (w./v.) of sesangolin in carbon disulfide (0.4-mm. light path) was determined with a Model 21 Perkin-Elmer spectrophotometer. Strong (5) and medium (m) peaks were observed a t the following cm.-I values: 2850(m), 1245(s), 119O(s), 1155(m), 104O(s), 940(s), 860(m), 817 and 807 (m). Nuclear Magnetic Resonance Studies.-The spectra were obtained with a Varian HR-60 spectrometer and V-3521 integrator using scan rates of 1-2 c.p.s./sec. for spectra and 7 c.p.s./sec. for integrals. Samples were dissolved in deuterochloroform a t a concentration of about IO%, and tetramethylsilane was used as an internal reference. Line positions were determined by interpolation between audio side bands, and should be accurate to 1-2 C.P.S. Reported values of the integrals are the average of five determinations. Permanganate Oxidation of Sesango1in.-Ten g r a m of potassium permanganate was added gradually over an 8-hr. period to a refluxing solution of 1 g. of sesangolin in 50 ml. of acetone. After standing overnight a t room temperature, the mangnnese dioxide was filtered off, and the acetmr Filtrate held aside. A hot water wash of the manganese dioxide gave an alkaline filtrate that was conccntrated to a small volume and acidified with hydrochloric acid. The solution was extracted with chloroform several times; the chloroform extract was dried over sodium sulfate and evaporated, giving 28 mg. of solid material. The acetone filtrate was evaporated and the residue taken up in 5 7 , potassium hydroxide. The ether extract of this solution yielded 30 mg. of unchanged sesangolin. The aqueous layer was acidified with hydrochloric acid and extracted with ether, which upon evaporation yielded 32 mg. of oily residue, not readily characterizable. The solid residue was chromatographed on silicic acid23 yielding two crystalline compounds. The first (9.4 mg.) which melted a t 226' after recrystallization from absolute ethanol, was shown t o be piperonylic acid by a mixed melting point with an authentic sample, ultraviolet and infrared spectra, and thin-layer chromatography.z3 The second compound (1.7 mg.) was chromatographed again and identified as 6-methoxypiperonylic acid24 by its ultraviolet and infrared spectra, and its R, value and characteristic color with chromotropic acid-sulfuric acid spray on thinlayer chromatography.23 Dilactone from Sesangoh-The procedure followed was that used by Beroza14 for the preparation of the same dilactone from sesamolin; yield from 1 g. of sesangolin w m 39 mg. The crystals from benzene were recrystallized finally from absolute ethanol, m.p. 159-160' [a]2311 +184' ( c 0.23 water). Anal. Calcd. for C6HeO4: C, 50.7; H, 4.3. Found: C, 51.35, 51.10; H, 4.54, 4.70. The melting point of the dilactone was undepressed in admixture with the dilactone obtained from sesamiii by following the same preparative procedure. zyxwvutsrqpo zyxwvutsrq zyxwvutsrqp zyxwvutsrqp zyxwvu (19) R. L. Tracy, U. S. Patent 2,837,534(June 3, 1958). ( 2 0 ) Kindly carried out b y Dr. Henry hl. Fales of the National Heart Institiit?, Bethesda, Maryland. (21) A w l i d Science Lsboratoriea, Inc.. State College, Pennsylvania. (22) M. Beroza, BnaZ. Chem., 28, 1550 (1956). -~ (23) M.Beroza and W. A . Jones, Anal. Chem., 54, 1029 (196'2). (24) R. T. Arnold and N. Bortnick. J , A m . Chem. Soc., 67, 1797 (1945).