ALKALOIDS OF Ochrosia elliptica LABnL.
April 20, 1959 [CONTRIBUTION FROM
TEE
1903
LABORATORY OF CHEMISTRY OF NATURAL PRODUCTS, NATIONAL HEARTINSTITUTE,NATIONAL INSTITUTES OF HEALTH]
Alkaloids of Ochrosia clliptica Labill.' BY SIDNEY GOODWIN, A. F. SMITH^
AND
E. C. HORNING
RECEIWD JULY 28, 1958 Four alkaloids have been isolated from leaves of Ochsosia elliptku Labill. One has been shown to be identical with isoreserpiline. The other three, ellipticine, methoxyellipticine and elliptinine, have been characterized and certain features of their structures have been suggested.
The plant Ochrosia elliptica Labill. (family Apocynaceae) is a small tropical evergreen tree. The genus is a member of the tribe Plumiereae along with the alkaloid producing genera Rauwolfia, Aspidosperma and Holarrhena. Fifteen species of Ochrosia have been described, occurring in Australia, the Andaman Islands, Madagascar, Hawaii and many other Pacific Islands. The species elliptica was first described in New Caledonia. 3 - 4 The only recorded previous chemical examination of Ochrosiu in which alkaloids were isolated was carried out by Greshoff5 a t the Botanical Garden, Buitenzorg, Java. He found the bark of the Javanese species acuminata, ackeringae and coccinea to be rich in alkaloids and distinguished three bases by virtue of their color and solubility; it is not possible to be certain that they correspond to any of the alkaloids found in the present study. The observation that 0. elliptica Labill. (Bleekaria calocarpa Hassk.) was alkaloid-containing was also made by Greshoff. Isolation.-In the present study four crystalline alkaloids were isolated and assigned the names elliptine (now isoreserpiline), ellipticine, methoxyellipticine and elliptinine; ellipticine and methoxyellipticine were distinguished by their bright yellow color and sparing solubility. The isolation studies were carried out on a total of 67 lb. of leaves (wet weight); in all, twelve isolation runs were performed. There appeared to be some variation in the alkaloidal content of the plant material. No single procedure was found to be suitable for all of the alkaloids. Although isoreserpiline was found to belong to the class of alkaloids whose salts are easily extractable from acid solutions with chloroform, this property was not utilized; procedure A, provided the best yield of isoreserpiline (0.28%). The isolation of a mixture of ellipticine and methoxyellipticine presented no difficulty but their separation proved to be extremely tedious. Alumina chromatography was used in this work; other adsorbents were not in(1) Presented in part before the 14th International Congress of Pure and Applied Chemistry, July 1955, Zurich. (2) Ciba Pharmaceutical Company, Summit, N. J. (3) J. J. H. de la Billardiere, Sertum Austro-Caledonicnm, 25 (1824). (4) The genus is not known in the New World except through importation. It is believed that the first introduction of 0. ellipfico was at the Royal Botanic Garden, Port-of-Spain, Trinidad, where it is called 0.moorci. The USDA Plant Introduction Stasion in Florida has a specimen of 0. clliplica which was presumably started from seeds from the Trinidad tree. The plant material used in this study was obtained from clippings of the Florida tree and consisted of relatively new growth of leaves and stems. The fruit was found to be alkaloid-free; neither the trunk bark nor the root hark was examined. (5) M.Greshoff. B n . , 23, a537 (1880).
vestigated and attempts to devise a countercurrent distribution system were not successful. Ellipticine and methoxyellipticine, when crystalline, are sparingly soluble in most solvents used in chromatography, and it was found best to apply a concentrate of the alkaloid extract directly to the chromatographic column (procedure A). After the elution of isoreserpiline, the column wa.s developed with increasing amounts of methanol and the separation was followed by paper chromatography. The yields of the yellow alkaloids were variable; the following are representative : ellipticine, 0.004% ; methoxyellipticine, 0.007%; and a mixture of the two, 0.002%. Elliptinine, the most elusive of the four alkaloids, was isolated in only two instances; the second time in only trace amounts. The isolation providing the better yield (0.0270, procedure B) was characterized by the fact t h a t most of the other alkaloids had been removed by crystallization prior to chromatography. Elliptinine was eluted immediately following methoxyellipticine. A thorough investigation of the mother liquors of the chromatographic fractions which yielded crystalline methoxyellipticine and of the subsequent fractions from other isolations failed to give any crystalline elliptinine, although its presence in small amount was observed by its characteristic ultraviolet spectrum and by paper chromatographic examination. The alkaloids are well resolved in paper chromatography; Rf values of 0.65 for methoxyellipticine, 0.70 for ellipticine and 0.76 for elliptinine were observed in a butanol-acetic acid system. Elliptinine may be differentiated froin the yellow alkaloids by virtue of its easy solubility in methanol. Isoreserpiline (Elliptine).-The analytical results indicated that the alkaloid, m.p. 210-211.5", C23H2806NtI contained three methoxyl groups, no methylimino group, one active hydrogen a t room temperature and a second one a t 100'. Variable Kuhn-Roth values were obtained. The infrared and ultraviolet spectra were characteristic of the /3-alkoxyacrylic ester system, ROC=CCOOCH3.6 The presence of an ester function was confirmed by alkaline hydrolysis. The alkaloid was found to contain no readily acetylated hydroxyl group and no easily hydrogenated double bond. Structural studies were not carried out, since i t was noted (in 1955) t h a t the m.p., rotation, and spectra reported for isoreserpiline,7 an alkaloid from Rauwolfia canescens L., corresponded with those of the Ochrosia alkaloid. The structure proposed for (6) F. E. Bader, Hclu. Chim. Acta. 36, 215 (1953). (7) A. Stoll. A. Hofmann and R. Brunner, ibid., SB, 270 (1955).
1904
SIDNEY GOODWIN,A. F. SMITH AND E. C. HORNING
isoreserpiline by the Swiss group, based on analytical and spectral data, is suported by these observations. The name to be retained in the literature should be isoreserpiline. The active hydrogen data could be interpreted as reaction with the indole S - H a t room temperature and an ether cleavage reaction* of one of the aromatic methoxyl groups at looo. Ellipticine and Methoxyellipticine.-The preparation of homogeneous solvent-free samples of these alkaloids for analysis was a matter of considerable difficulty but it was eventually accomplished with the results that ellipticine was assigned the formula C ,7HI4X2and methoxyellipticine C18H160?;2 with one methoxyl group.g Both alkaloids were analyzed for terminal methyl groups (in two laboratories) with negative results. They are assumed to have the same basic structure since they differ empirically only by a rnethoxyl group and they have similar ultraviolet, visible and infrared spectra. Previously encountered yellow alkaloids from plants of the Apocynaceae are p-carboline anhydronium bases. This system is characterized by the observations t h a t the ultraviolet spectrum in alcohol is unaffected by the addition of acid whereas the addition of alkali produces a bathochromic shift. The unusually complex ultraviolet spectra of ellipticine in alcohol (Fig. 1) and in
Vol. 81
these yellow bases represented a new structural departure for alkaloids of the Apocynaceae. Ellipticine and methoxyellipticine were converted to methiodides whose analyses confirmed the assigned empirical formulas for both alkaloids. When alcoholic solutions of the rnethiodides which were deep yellow and orange, respectively, were made alkaline, reddish-purple solutions were formed, and if sufficient water mere present, amorphous, reddish-purple solids precipitated. This behavior was interpreted as being due to the formation of anhydronium bases of the type found for the purple alkaloid cryptolepine (I1),lo The European workers have shown t h a t cryptolepine (11) is the anhydronium base of quindoline methiodide (I). Although the absorption maxima in the
C",
CIiS
I1
visible are nearly the same (530 nip) for both ellipticine methiodide and quindoline methiodide (in 0.01 .A7 alcoholic potassium hydroxide), the maxima in the ultraviolet are not identical. Sirnilarly the ultraviolet spectra of the parent bases, ellipticine and quindoline, are similar but not identical." It is evident t h a t ellipticine and methoxyellipticine are not quindolinc derivatives; however, the color changes described for the methiodides must be due to anhydronium base forniation and it is therefore suggested that the nitrogen functions in these yellow alkaloids are of the same type as in quindoline; in an abbreviated f ~ r i nthe nitrogen functions may be described as -NH(C=C)C=S-. If it map be assumed that the ring systein of these
_
220
260
300
343
383
420
463
500
540
580
i
623
WAVELEIUC-TH ( m p i .
, clliliticiiie iti
Fig. l.---Ultraviolet spectra o f :
ethaniil; ---------, ellipticine iiictliiutlicic hydroxide solution in cthanol.
iii
0.05 N piitaiziiilu
0.01 N-alcoholic potassium hydroxide were identical in the wave lengths of the maxima and minima; however, the spectrum in acid solution was shifted toward the red. The same results were found for methoxyellipticine. Thus it was evident that (8) lf.S 'Kh.irksch a n d 0 . R e i n m u t h , "Grignard R e a c t i o n s of S o n metallic Substances." Prentice-Hall, I n c . , S e w Y o r k , N . \-,, 1954, p. 1013. (0) l ' h e empirical formulas mentioned earlier1 for these alkaloids !\err I,.iird 01, S,C~I,IIIC\ \I hich m:iy h.ivr l i i c n i n , ~ d c r ~ t > ~ tpiirifir,l clv c)r \ \ e r e sol\ a t e d .
alkaloids is isomeric with quindoline, the11 all of the atonis would be accounted for except for C?HJ which could be satisfied eithcr by an ethyl substituent or by two inethyl groups. Thc nuclear magnetic resonance spectral? of tllipticiiie and xnethox~c1liI)ticixie shoived coilelusively tliat ;ti1 ethyl substituent was not present and th:it tivo c.hcmically shiitctl ~iicthyl groups were present. These conclusions are :ipparent from esaniiixiation of the ii.1n.r. spectrum uf iiiethosyel1il)ticine (Fig. 3 ) ; the signal a t 67 c.p.s. relative t c i benzenc has been assigned to the hydrogen nuclei of the rnethoxl-1 group and those at 00 and 109 c.p.s. to the hydrogen nuclei of inethyl groups which :trc attached to carbon atoms having no attac.liet1 hydrogen atoms. The n.ni.r. spectrum of elliliticine was identical with the exception of the sign:il for the methosyl group. The fact that Kuh11-Roth osidation did not afford acetic acitl int1ic:itcd that the methyl groups were situatetl or1 X I I aro(10) E . Gellert, R a y m o n d - H a m e t a n d E. S c h l i t t l c r I ! ( + ; , ~ h l w A d a , 3 4 , fi42 (1951). ( 1 1) T h e quininduline 5 ) steiii n : t s also eliminated inr ~ l I i ~ ) l i r ii,n ii~ t h e b,isis of i t s ultra\-iulet s p c c t r u r n T h e spectrA c > f iiiiinduiine d n d quinindoline h a v e been pub!ishr.d b y C: .:I Clemo , i r i , i 1) C. I l ~ v l l ~ , c i . J . C h r n . Y r i , O i l (1921 ( 1 2 ) IV,: n r i ind?l)lr,i 1 ) r J S.S h , l ~ , l e r y, i i i i l h I r I. I I C ~ I I I I W I ~ V . i r ~ , i n .\,avci,$teq, I'.ilL . \ I l t , , C,alif., ior t h e x ri.it,l
ALKALOLDS OF Ochrosia elliptica LABILL.
April 20, 1959
1905
matic nucleusIa; ev’dence bearing on this point I I I and the structure o< ellipticine has been provided by Dr. F. A. Hochstein of Charles Pfizer and C0.I4 Ellipticine methiodide was readily reduced by sodium borohydride to form a tetrahydro derivative.I5 Methyl tetrahydroellipticine was found to be identical in melting point and ultraviolet and infrared spectra on direct comparison with an alkaloid from the Peruvian “Quillo B o r d ~ n ” ’ ~ and with the reported melting point and ultraviolet spectrum of alkaloid B from Aspidosperma. ulei Mgf.l6; more detailed comparisons are being carried out in other laboratories. Elliptinine.-This colorless, optically active alkaloid, m.p. 231-233”, was found to have the formula C%H2402N2and to contain one methoxyl, no methylimino, and one terminal methyl group. The infrared spectrum indicated t h a t no carbonyl groups were present; this inciuded the 6 p region with respect to amide groups and conjugated I I I unsaturated esters of the type proposed for akuam67 90 109 micine.” The ultraviolet absorption spectrum of elliptinine was very interesting in that it consisted CHEMICAL SHIFT(C.P. S.), of a shoulder a t 222 mp and a broad maximum a t Fig 2.--Nuclear magnetic resollalice spectrum of inetlioxy311 mp; acidification produced a slight hypso- ellipticine a t 40 mc. in pyridine solution referred to benzene. chromic shift to 308 mp. In a small scale hydrogenation experiment from which a crystalline tion of a small quantity of this Hawaiian species. product could not be obtained, one mole of hydro- Isoreserpiline and elliptinine were not detected, gen was absorbed and the maximum in the ultra- but a new colorless alkaloid was found in small violet spectrum was shifted from 311 to 298 mp, amount. This alkaloid has been partially characI n the infrared spectrum of elliptininium per- terized as its hydriodide, m.p. 215’ dec., [a1589 chlorate very intense bands a t 2.99 (broad) and 27’. The free base has ultraviolet absorption 3.16 p were observed and, in addition, a very weak maxima a t 238 and 290 mp and the spectrum is band was present a t 5.97 p which although having unaffected by acidification. Acknowledgments.-Me are especially grateful the expected wave length for an iminiurn ion was to Dr. F. A. Hochstein of Chas. Pfizer & Co. for a of much too low intensity. The bands near 3 were assigned to indole N H and NH+, respectively. stimulating exchange of information and for On the basis of the infrared interpretation, a samples of “Quillo Bordon” alkaloids. We are grateful to Dr. H . A. Lloyd of this Laboratory for I l l normal enamine system, -C=C-N-, is not samples of quindoline and quinindoline and to Dr. present.lB An interesting speculation based on J. Schrnutz, Dr. A. Wander, A. G. Bern, for the Woodwardlg scheme for the derivation of a samples of some of his Aspidosperma alkaloids. linear carbazoquinuclidine system froin ajmaline We are indebted to the Section of Plant Introducor a biogenetic precursor is structure 111 which is tion, Agricultural Research Service, U. S . Department of Agriculture, for obtaining and identifying compatible with-but of course not required bythe analytical and spectral evidence so far avail- the plant inaterial and to Dr. Karl H. Korte, Division of Forestry, Kahului, Alaui, T. H., for the able for elliptinine. Ochrosiu sanauicensis used in this work. Experimental2o
C ; 7 : . ~ a
(.lI,O
”,
H
’
111
Ochrosicl scindwicensis A.DC.-Ellipticiiie and iiiethoxyellipticine were isolated in the examina(13) N . A. H u g h e s a n d H. R a p o p o r t , THISJ O U R N A L80, , 1604 (1958). (14) F. A . Hochstein, p r i v a t e communication. (15) T h e reduction of u n s a t u r a t e d heterocyclic methiodides h a s been described b y B. W i t k o p , THISJ O U R N A L , 76,3362 (1953) a n d by B. Witk o p a n d 1. B. P a t r i c k , ibid., 7 6 , 4474 (1953). (16) J. S c h m u t z a n d F. H u n r i k e r , H e l u . Chini. Acta, 41, 288 (1958). (17) K . A g h o r a m u r t h y a n d R . R o b i n s o n , Telrahedron, 1, 172 (1957). (18) An e n a m i n e s y s t e m in which t h e corresponding iminium ion is incapable of existence f o r steric reasons ( a s would be t h e case if t h e nitrogen were p a r t of a quinuclidine system) m a y also he discarded on spectral grouds (cf. C. A. G r o b , A. Kaiser a n d E. R e n k , C h e m i s f r y & I ~ i d i ~ s ~598 r s , (1957). (19) R . B. W o o d w a r d , private communication; ( f also R . B. K‘oodward A+igew. Chetri.. 68. 13 (1956).
Isolations. Procedure A.-Two kilograins of ground airdried leaves and stems of Oclzrosia elliplica were stirred at 55’ n i t h 10 1. of 95% ethanol. The alcoholic extract was filtered and the plant material was treated again with ethanol a t 55’. T h e combined alcoholic extracts were concentrated in vacuo t o about 2 1. of black, sirupy residue. The conctntrate was poured slowly into 15 1. of water containing 2 kg. (20) Meltiag points were t a k e n o n a KoEer block. R o t a t i o n s a n d ultraviolet spectra were determined in absolute ethanol, a n d infrared spectra were t a k e n in chloroform, unless otherwise specified. I n s t r u mental measurements were m a d e by Mr. a n d hIrs. H. F. Byers, Misses P. A. W a g n e r a n d C. M o n a g h a n , a n d M r s . K . Warren. Infrared spectra were obtained with a Perkin-Elmer spectrophotometer model21, ultraviolet spectra with a C a r y recording spectrophotometer model 11 a n d rotations were measured in 2-dm. cells using a R u d o l p h photoelectric polarimeter model 80. T h e nuclear magnetic resonance spectra of ellipticide a n d methoayellipticine were measured in pyridine with a Varian H i g h Resolution spectrometer model V4300B a t Varian Associates, Palo Alto, Calif. Extractions mere m a d e by hlr. D. L. Rogerson and h l r . J . D. 1.ink .4nalysrs \\-?re made by h l r ‘!I I l a n ser. Zitrich, and the C l a r k ll!crl,.bnalyticaI 1,aboratory. ljrI>au,L, I l l
1906
SIDNEY
GOODWIN, A. F. SMITH AND E.
of Supercel (with vigorous stirring) and the PH of the mixture was adjusted from 5 to 4 with 75 cc. of glacial acetic acid. The mixture was filtered through a large Buchner funnel and the pale gieen filter cake was resuspended in 7 1. of water containing 50 cc. of glacial acetic acid. (After filtration, the filter cake was washed with a minimum of water, dried a t 32' i n w c u o , and extracted in a soxhlet extractor with ether and alcohol. Although a Mayer test was positive, crystalline alkaloids could not be obtained from these extracts after chromatography or countercurrent distribution attempts.) The combined filtrates were mixed well with a large volume of chloroform and made alkaline with solid sodium carbonate. T h e basic solution was extracted in six portions w-ith 10 X 200 cc. of chloroform for each portion. T h e final basic solutions still gave weak alkaloid tests. T h e combined chloroform extracts were dried over magnesium sulfate and concentrated to 2.51. This chloroform extract was added t o a column (5 X 62 cm.) of 1 kg. of acid-washed alumina. T h e summary of the chromatogram has been arranged in the following order: fraction, solvent, volume of eluant and weight of residue: 1, chloroform (2.5 l.), 0.891 g., non-alkaloidal; 2-4, chloroform (3 l.), 10.91 g., from which 5.50 g. of crystalline isoreserpiline was obtained on crystallization from ether; 5-10, chloroform (4 1.) and 1% methanol-chloroform (3 l.), 2.87 g., non-crystalline, combined with the mother liquors of 2-4 and other suitable fractions from other chromatog r a m for rechromatography; additional amounts of isoreserpiline were obtained but no other crystalline alkaloids were found; 11-14,2,3and 5Y0 methanol-chloroform (6 l.), 1.24 g., non-crystalline, but ellipticine was indicated in 14 by paper chromatogram*'; 15-16, 5% methanol-chloroform (21.) 1.03 g., red-brown tar and yellow solid which on trituration with chloroform yielded 25 mg. of ellipticine (an additional 10 nig. was obtained from the chloroform on concentration); 17-19, 5% methanol-chloroform (1.5 l . ) , 0.79 g. crude residue which afforded 21 mg. of ellipticine which was contaminated with methoxyellipticine (according t o paper chromatogram); 20, 5% methanol-chloroform (0.7l.), 0.19,g. crude.; 8 mg. of crystalline methoxyellipticine contaniinated with ellipticine; 21-24 5% methanolcliloroforrri (6 l.), 1.89 g. of residue from which a yellow solid was obtained i i i the usual way; after three recrystaliization froin methanol, 119 tng. of methoxyeliipticinc was obtaiiied.22 Elution was continued but no crystalline solids could be obtained. Procedure B.-One kilograiri of dried, ground plant ninterial was extracted three times with 1% etlianolic tartaric acid solution at 55". A 2.5-1.concentrate was mixed with l..i 1. of 2 N hydrochloric acid and filtered through Sul)ercel. The filtcr cake RRS washed with dilute acid aiitl with water. The combined filtrates were washed wit11 cliltrroforin (5x); the last wash was still deeply colored. The conibined chloroform extracts were shaken with 10yo sodiurti carbonate solution, dried and the solvent w a s evaporated. The benzene-soluble portion of the residue a f forded no crystalline alkaloids on chromatrigraphy. The acidic solution was made alkaline with solid sodium carbonate and extracted with ethyl acetate. Continuous extraction of the aqueous phase with cliloroforni afforded 1.27 g. of black t a r . The ethyl acetate extracts were extracted with 15Toacetic acid (it was iicceasary to add ether-about cinc-half the volunie of the organic phase-to break u p einulsionsj. The acid extracts were made alkaline and extracted with ethyl acetate. After drying and rcinoving tlie solvent, (21) P a p c r c h r o m a t o g r a m s were o b t a i n e d using W h a t m a n 1 p a p r r and t h e n p p e r p h a s e of 3 m i x t u r e of 1-but:inol-acetic acid--matrr ( 4 1 . 5 ) . T h e alkaloids were detected b y their fluorescence u n d e r ultraviolet l i g h t a n d b y s p r a y i n g w i t h Dragendorff reagent. T h e fluorescent maxirno of ellipticine a n d methnxgeliiplicine, measured in alcohol solution with activation a t 204 m e , were fnund t o be approxim a t e l y 4 4 0 a n d ROO me, respectively: a n experitnent;il model of t h e Rnwnian-Aminco Spectrnphotofluorometer was used t o o b t a i n t h e s e data. ( 2 2 ) I n t h i s ni:mner i t w a s pussihle tn o b t a i n ellipticiue which was comi,letely homojirnenus according t o p a p e r chrom:itojii-ai)hy: howe v e r t h e p a p e r c h r o m a t o g r a m s of methnxyrllipticine even a f t e r repeated crystullizations shuwerl a pcrsi>tent t r a c p ini1,iirity h a v i n g t l i e K, of elliplirine b u t giviug a hluc-green fltiorc.5cenre r a t h e r t h x n t h e b r i l l i a n t yellow-grwn tlus not in good agreement n i t h this structure possibly because of a persistent impurity. A n d . Calcd. for C~6H3006S2: C, 66.06; 11, 8 65; X , 6.16; 3 OCH?,20.48; 2(C)CIi,, 6 62. F O U I I ~ C, : 61.21; H,6.49;N,6.22;OCHS, 19.96;(C)Clia,7.14 Hydrolysis.-A solution of 0.4 g. of isorcserpilinc in 30 ml. of lcA sodium hydroxide in niethanol--water (1:1 ) was refluxcd for 3 hours. The solution was taken t o PH A with hydrochloric acid and evaporated to dryness. Trituration with water afforded a colorless solid having a carboxylat? anion (zwitterion) band in the infrared (3.0 and 6.12 p ) . The coinpound was very unstable and was converted to its (mull) 3.0, 5.98 and kiycirochlix+~e, m . p . 205 208", h,,, 6.12 p . This cornpouiid was rather hygmscopic and unst,ihle. .\lthougli the analysis was not good, the loss of one n r c t I i ( sx)-l group \vas clrarly indicated. (23) Measured using a B e c k m a n 1K-3 sprctruphntorueler.
April 20, 10%
ALKALOIDS O F Ochroyia elliptica LABILL.
Anal. Calcd. for CnHn0,NZCI: C, 60.75; H , 6.26; 2 OCH1, 14.27. Found: C, 61.29; H, 6.17; OCH1, 14.41. Ellipticine.-The most convenient solvent for crystallization was methanol, from which ellipticine crystallized as bright, lemon-yellow needles; however, i t was apparent from initial analytical d a t a t h a t methanol of crystallization was difficult t o remove completely. T h e analytical sample was prepared from yellow crystalline material (48 mg.) obtained from a chromatographic fraction by crystallization froin chloroform; recrystallization from ethyl acetate afforded a mixture of orange rosettes, orange rods and yellow rosettes. The different crystal modifications were indistinguishable by ultraviolet spectra and paper chromatography. The analytical sample melted a t 311-315" dec., Amax (satd. chloroform solution): 2.86 and 6.23 p ; pK. 5.87. The value of the pK. of quindoline was found to 227-234(sh), 238, 245(sh), be 3.6.24 Ultraviolet: Am, 276, 287, 295, 318-322(sh), 333, 343-347(sh), 384 and 401 mp (log c 4.32, 4.36, 4.31, 4.74, 4.90, 4.88, 3.52, 3.71, 3.47. 3.61 and 3.58, resp.); Amin 251, 279, 292, 309, 355 and 393 mp (log e 4.27, 4.73, 4.76, 3.3i, 3.35 and 3.55, resp:). I n 0.01 N ethanolic potassium hydroxide: no change in wave length of maxima and minima but a slight change in extinction coefficients. I n 0.1 N hydrochloric acid in ethanol: Amsx 241, 249, 271(sh), 307, 335 and 426 mp (log e 4.41,4.39,4.29,4.90,3.75and 3.68, resp.); Amin 246,260, 330 and 382 mp (log c 4.38,4.13,3.40 and 3.36, resp.). Anal. Calcd. for CI,H~4N,: C , 82.90; H, 5.73; N, 11.37; (C)CH,, 6.10. Found: C, 82.83; H, 5.77; N, 11.16; (C)CH,, none. Methiodide.-A solution of 44 mg. of ellipticine in 100ml. of acetone was allowed t o stand with a n excess of methyl iodide (10 ml.). T h e bright yellow, crystalline precipitate (59 mg.) was collected, washed with acetone, and crystallized from methanol. T h e methiodide decomposed a t about 360' without melting. Ultraviolet: Amax 223, 242, 251, 27&285(sh), 311, 337-345(sh), 362 and 432 mp (log e 4.47, 4.43, 4.40, 4.24-4.40, 4.90, 3.51-3.57, 3.74 and 3.71, resp.); Ami. 235, 248, 262,335 and 390 mp (log c 4.38: 4,39, 4.13, 3.49 and 3.49, resp.). In 0.05 N alcoholic potassium hydroxide: AUx 246, 272, 336, 376, 403 and 520 mp (log z 4.38, 4.46, 4.75, 3.95, 3.95 and 3.85,resp.); Ami.255,292, 363, 390 and 432 mp (log e 3.97, 4.03, 3.82, 3.85 and 3.59, resp.). Anal. Calcd. for ClaHllNtI: C, 55.68; H , 4.41; N, 7.22; I, 32.69. Found: C, 55.55; H , 4.58; N, 7.37; I, 32.70. Anhydronium Base.-In a very small scale experiment, the anhydronium base was found t o be precipitated as a n amorphous red-purple solid after the addition of 1070 sodium hydroxide solution t o a methanol solution of ellipticine methiodide. It formed a violet-purple solution in chloroform which was completely decolorized on shaking with water; the yellow aqueous phase was not changed on the addition of a small amount of alkali but with excess alkali a red-purple precipitate formed. Cryptolepine was not as easily hydrated; its purple chloroform solution was unchanged on shaking with water. This difference in stability of the anhydronium bases and the differences in pK. values probably result from the same structural features. Methyltetrahydroellipticine.-A solution of t h e methiodide, prepared from 113 mg. of ellipticine, in 100 ml. of methanol and 10 ml. of waterl6 was treated with excess sodium borohydride. The orange color was instantaneously discharged on the addition of the borohydride. After standing for 1 hour, most of the methanol was evaporated i n wcuo. The concentrate was diluted with water and extracted with methylene chloride. Crystallization of the rnethylene chloride residue from ethyl acetate gave 109 mg. of yellowish solid, m.p. 215-220". The product crystallized from methanol as colorless needles, m.p. 21fj-22'; m.m.p. with the "Quillo Bordon" alkaloid 216-220 ; the infrared spectra of the two were identical; ultraviolet: Amsx 227-231( s h ) , 213, 250(sh), 264, 286-288(sh), 296, 318(sh), 328 and 342 nip (log e 4.56, 4.74, 4.63, 4.40, 4.08, 4.35, 3.50, 3.64 and 3.62, resp.); Amin 259, 278,308 and 335 mp (log e 4.34, 3.86, 3.36 and 3.53, resp .). (24) We arc grateful to Dr. Norbert Neuss and Dr. Harold Boaz, Eli Lilly Co., Indianapolis, Ind.. for these data. (25) R. Mirra, J . Cbcm. S O L ,4400 (1957).
1907
Anal. Calcd. for ClaHmN,: C, 81.78; H , 7.63; N, 10.60. Found: C,81.88; H.7.65; N, 10.59. Methoxyel1ipticine.-An analytical sample was prepared from the sample described in the isolation section by sublimation at 180" at about 0.001 mm. Methoxyellipticine could also be crystallized frum ethyl acetate as bright yellow needles, m.p. 280-285' dec., suitable for analysis; equivalent weight: A solution of 29.06 mg. of methoxyellipticine in glacial acetic acid required 2.341 ml. of 0.0448 N perchloric acid in acetic acid; calcd. 276 found 278. UltraA, 246, 277, 294, 307(sh), 337, 353, 403 and 410violet: % 420(sh) mp (log e 4.42,4.66, 4.73, 4.49, 3,81, 3.54, 3.57 and 3.55-3.52, resp.); Amin 223, 254, 281, 327, 348 and 365 m p (log c 4.24, 4.33, 4.64, 3.65, 3.50 and 3.22, resp.). In 0.01 N potassium hydroxide in ethanol: essentially identical with the spectrum in alcohol alone. I n approx. 0.1 N hydrochloric acid in ethanol: Amax 225, 245-261(plateau), 278, 314, 360, 379 and 450 mp (log e 4.26, 4.41-4.40, 4.36, 4.75, 3.79, 3.58 and 3.60, resp.); AmiD 218, 231, 270, 288, 344, 368 and 400 mp (log c 4.22, 4.24, 4.33, 4.28, 3.67, 3.75 and 3.34, resp.). Anal. Calcd. for ClaHlsON't: C, 78.23; H , 5.84; N, 10.14; OCH3, 11.23; (C)CH,, 5.44. Found: C, 78.09, 78.19; H , 5.89, 5.81; X, 10.29; OCHz, 11.06; (C)CHI, none. Methiodide.-A solution of 105 mg. of methoxyellipticine in 150 ml. of acetone was allowed t o stand with a n excess of methyl iodide (10 ml.). The orange, crystalline precipitate (126 mg.) was collected and washed with acetone. The analytical sample crystallized as fine orange needles from methanol. T h e methiodide decomposed .at about 340' without melting. Ultraviolet: Amax 222(sh), 246-254(plateau), 260-265(sh), 281, 316, 364, 384 and 448 mp (log c 4.56, 4.43, 4.41, 4.38, 4.74, 3.75, 3.85 and 3.60, resp.); Ami,, 238, 272, 290, 347, 373 and 407 mp (log e 4.36, 4.36, 4.32, 3.67, 3.74 and 3.43 resp.). I n ethanolic potassium 245, 273, 338, 37&377(sh), 409 hydroxide solution: A?., and 540 mp (log c 4.33, 4.46, 4.69, 394-393, 3.93 and 3.69, resp.); Amin 237, 252,295,393 and 445 mp (log c 4.31,4.32, 4.21, 3.84 and 3.10, resp.). Anal. Calcd. for C19H190X21: C, 54.56; H , 4.58; N, 6.70; I, 30.34. Found: C , 54.37; H , 4.56; N, 6.80; I, 30.47. Elliptinine.-Elliptinine was found t o be very soluble in methanol and sparingly soluble in ethyl acetate, benzene and chloroform. T h e tan solid, m.p. 213-219' (isolated as described), on recrystallization from methanol-benzene afforded a colorless crystalline solid, m.p. 229-233". The analytical sample, m.p. 231-233', [ a ] Z 4 ~-255' o and [alZ441s -593' ( c 0.200), was crystallized from methanol-benzene. Equivalent weight: The titration of elliptinine (15.669 mg.) in glacial acetic acid with 0.0448 iV perchloric acid to the methyl violet end point required 1.100 ml.; calcd. 324, found 318. Ultraviolet: Xmax 222(sh) with end absorption and 311 mp (log c 4.37 and 4.27, resp.); Amin 270 mp (log e 3.39); in 0.1 N bydrochloric acid in ethanol; Amax 217 and 308 mp (log c 4.45 and 4.24, resp.): Ami. 265 mp (log c 3.50). Infrared: Amsx (satd. chloroform solution): 2.80(v.w.), 2.88(m), 6.16(m), and 6.30(m) p ; Amax (Nujol mull): 2.90(v.w.), 3.02(m), 6.16(m) and 6.28(w) p . Anal. Calcd. for CZOHZ~OZK'~: C , 74.04; H, 7.46; N, 8.64; OCHs, 9.57; (C)CH3, 4.63. Found: C , 73.90, H , 7.05; N,8.72; OCHZ,9.83; (C)CH,, 4.02; (N)CH1, 0.73. Perchlorate.-A small sample of elliptinine was converted to the perchlorate in methanol; Am,= (mull): 2.99(s), 3.16 (s), 3.3.5(Nujt~l),5.97(v.w.), 6.15(m), and 6.27(m) p . Hydrogenation.-A solution of 57 mg. of elliptinine in 6 ml. of methanol was added to a pre-treated mixture of 50 mg. of Pd-C and 5 ml. of methanol and stirred in an atmosphere of hydrogen a t atmospheric pressure and room temperature for 2 hours. The change in volume represented the absorption of one equivalent of hydrogen. After removal of the catalyst and solvent, a colorless solid residue, Amax 298 mp, was obtained. This could not be induced to crystallize. Ochrosia sandvicensis A. DC.-One kilogram of ground leaves was treated according to procedure B. After chromatography, a total of 13 mg. of yellow, crystalline solid was obtained. This was shown t o be a mixture of ellipticine aud rnethoxyellipticine by paper chromatography. Isore-
190s
SIDNEY
GOODWIN .4ND E.
serpiline and elliptinine were not detected, although a colorless alkaloid (90 mg.) not found in 0. elliptica was eluted after the yellow alkaloids. Suitable solvents for crystallizaticn could not be found and the alkaloid was converted to its hydriodide in acetone. The salt (which was quite soluble in acetone) was recrystallized from water,
/ C ( ~ S T R I I 1 U T I O NFROM THE
c. HORNING
Yol. 81
m.p. 215' dec., [a]2368 +27" and [ a ] 2 3 4 6 ~ +58" ( c 0.45). The ultraviolet spectrum was examined: Amax 238 and 290 mp and A,:, 258 mp; infrared spectrum: A, (mull) 3.01 and 6.24 ,A.
BETHESDA 14, MD.
LABORATORY O F CHEMISTRY
O F S A T U R A L PRODUCTS, NATIOUAL INSTITUTES OF H E A L T H ]
Alkaloids of Lunasia amar-a Blanco.
HEARTINSTITUTE,
NATIONAL
Structure of Lunacrine
BY SIDNEY GOODWIN AND E. C. HORNING RECEIVED OCTOBER17, 1958 Lunacrine has been assigned the structure I on the basis of the chemical reactions and interpretations of spectra described i n this paper.
Lunacrine was first isolated in 1900 by Boorsmal scured by the presence of other b a ~ i d s . ~The , ~ abfrom the bark of Lzcnasia costzilata Miq. The al- sence of a methylenedioxy group was shown conkaloid was characterized by its melting point clusively by the absence of any signal in the appro(114O), solubility, precipitation reactions and color priate region of the n. m. r. s p e c t r u ~ n . ~ , ~ reactions. 'IVirth? carried out the first studies reThe infrared spectrumg of lunacrine was devoid lated to the structure of lunacrine in 1931, and re- of absorption in the 5 p region while the f i p region ported the empirical formula to be C16H2003N.was very rich in bands; structural assignments were The presence of one methoxyl and one methylimino not possible. The ultraviolet spectrum in alcohol group, and the forniation on zinc dust distillation was unaffected by the addition of alkali, but a of quinoline, skatole and ammonia were also re- marked change occurred on acidification (Fig. 1) ; ported.* -4few years later, Dieterle and Bey13cor- the long wave length bands (313 and 326 mp) exrected the obviously impossible formula and gave hibited a hypsochromic shift (300 mp). Thus the C16H1,OXS which was later confirmed by Steldt nitrogen atom is a part of the chromophore and inand Cheri The alkaloid described by the German volved in a structure which would give this beworkers, although agreeing in melting poitit havior in acid. This irnniediately eliminated from (,llii") with that reported by others,*,*was said to consideration the most prevalent nitrogen-contailibe opticall>-inactive whereas \-alues had been given iiig nucleus in the Rutaceae alkaloids, eiz., quinoline for [ N ] D of -3SO' and, later, -.iS0.3 Lunacririe whose long wave length ultraviolet absorptiori was stated to have a iiiethylenedioxy group by vir- maxima exhibit pronounced bathochromic shifts tue of the formation of the calculated amount of in acid solution. This behavior of lunacrine was i n Imcipitate on heating with phloroglucinol-sulfuric fact reminiscent of I - n i e t h ~ - l - 2 - p h e n y l - ~ - q u i n o l o ~ i ~ ~ , acid.3 Luiiacrim methiodide on treatment with a derirative prepared froni 4-rnethoxy-2-phenq-lsilver osirle followed by 40r; alkali was reported to quinoline, the first alkaloid isolated in the present give a 11roduct, m.p. S.i-8(i0, allegedly isomeric with study.1° In the spectrum of this 4-quinolone the lunacriiie.X The alkaloid was found to be stable typical bifurcated maxima (323 and 337 n ~ p were ) tciwvard perriiangniiate, chroiiiic oxide and other shifted to a single inaxiniuiii a t 304 nip in acid. osidiziiig agents and was riot affected by catalytic This is apparently a characteristic phenomenon of Iiq-tlrogciiaticiri ccin~litions.~ the 4-quinolone system a s it has been found that the long wave length maxima (324 and 337 inp) of T i i tlic present study, lunacrine, m.p.111-118", [a --.i0.1", was isiilated frorn the Leaves of l-methyl-4-quinolo1ieare shifted hypsochroniically (plateau 3OH1.3 nip! in iV hydrochloric acid. Idioi(isL(zunzcirci Blanco of Philippine origin. The ciiipiricd formula and the presetice of one methosyl .liter examination of the litclrature,12it would apa i i t l oiic iiicthyliiniiio group were confirmed by ( 5 ) I.. I1 Briggs. L.D . Cclebruok, I-I. >I. F a l e s a n d W. C. TTildmcin :iiialysis n i i t l , i i i adtlitiiiii, Kuhri-Iloth detcrxni~~atior~ .-Iii,ri Chian , 29, 904 ( I % i J D r . €I 31. FaIes f o r this information. (b) lye are indcl,trd iiit1ic:itcti ciiie tcriiiiiinl iiicth!-l group (or gczt,iG o u d w i n J . N S l i c w l ~ ~ rrind y I - , I:. Johnsun, l'ms J o l . R V . \ r . , diincth>-l). The resistance of lunacririe toward 81,(7)In Spress (1859). catdytic lis-(lrogeiiatitiii was confirmed. S o evi( 8 ) 'i'he first n . m r. s p e c t r i i i n n f lunacriiie was olitnineii :Itirl iritcrt i c ~ i c ewas obtained for the presence of a methylene- p r e t t d b y I l r I1 C o u r o y of n r n n d r i s L'nivrrsity, t o w11-sis ( i f the 9..i--lO.ri p region was obti,
WL'
m.a,
II cdl I n s l , Bo1 Bz~iicnsorp,6 , 15 (1900).
(2) T: H. l y i r r h , P h a i m H.Pikblnii. 68, 1011 (1931); C. A , 26, ,557 (19321 ( 3 ) I I . Dirterle and H. B e y l , A r c h . Pharm , 276, 1 7 4 , 276 (1837). ( 4 ) F. A S t e l d t a n d K. K .Chen, J . A n i . Pharm. A s s o c . , Sci. E d . , 32,
107 (1943!,
s h i f t in t h e s p e c t r u m < i f .I-
