oxide was condensed with ethyl acetoacetate to yield a compound, 11i.p. 9 1 -Ec,which was latert1 s1iow-n to he a-ncetyl-y-(:~,l-rnethylenediosyf \r). They prepared 1t hy rlehydrogenation oi benzyl )-y-butyrolactone (1 ) . Vi-hen the sodiutrl the ti.7 - rnethylenediosy - 1 - (;3,4,5 - trimethoxy- enolate of this compound was condensed with triphenyl)-~~-hydroxyniethyl-3,1-dihydro-%-naphthoicmethylgalloyl chloride, and the acetyl group reacid lactone (IV),with 1n.p. 24-L8-249", which nioved with dilute sodium hydroxide, a-(3,4,5-tri - methylenedioxybenzyl)- yis an isomer oE a- and i3-apopicropodophyllin, inethosybenzoyl)- yformed' in the dehydration of picropodophyllin. butyrolactone ( I I J , 111.p. 110-111", was obtained. .\nother isomer oi these two compounds, named This compound underwent ring closure when y - a p o p i c r ~ ~ ~ ~ d o p h ~ -has l l i ~been i, obtainedJ by trexted with inethyl alcoholic hydrogen chloride to boiling tr-apoptcroporiophyllin with 2 O q potassium give '1 70f yield of material, m.p. 13S-lt39., hx-druxicie for 24 hours and lactonizing the product. to which they assigned?J the structure of I-hySubsequently, i t was tle~iionstrated~ that y-apo- droxy - (i,T - niethylenedioxy - 1 - (;3,4,.5 - trimethoxypicropodophyllin IS structurally identical with phenyl) - 3 - hydroxymethyl - 1,2,3,4 - tetrahydroH:iworth and Richardson s lactone IY. Their 2-naphthoic acid lactone (111). IThen this conis p t h e i i s was repeated to establish this identity. pound was heated with potassium hydrogen sulIlowe\ er, in the course of that work, certain differ- fate a t 180", it was converted to lactone I\-, ences from the results published by Haworth and 11i.p. 248-219". I n order to account for the Richarclson were found 'These differences form formation of the 1-hydroxy lactone 111, Haworth and Xtkinson6 postulated that a "pinacolinic the subject of the present cointnunication transformation' took place during the ring closure A V P iitimber ( c m - l i . with inethyl alcoholic hydrogen chloride. 'The 4000 3ooo m i 1.m 1200 inoo NKI \ arious steps, as formulated by Haworth ant1 .itkinson, are summarized in Chart I . IVhen Haworth and Atkinson, however, treated the a-acetyl butyrolactone I with methyl alcoholic hydrogen chloride, they did not obtain a hydroxllactone corresponding to 111, but rather a n.eth11 eiter, for which they proposed the structure of uiethvl ~i,7-methylenedioxy-3,1-enclomethyleneoxylI -methyl-1 ,-,,:2.l-tetrahvtIronaphthalene-L'-carbos~-- :. m . p . 142-143' i ate I< = Iwrpiioti y i e c t r n iii chlnrnfnrtn. ranging from ? t i to 5O( a colorless solid, m.1) 140'. This product, after recrystallization iron1 H:iworth and Richardson prepared the lactone ethanol or methanol, formed colorless needles, 1n.p 1\? hy the iollorviiig series of reactions: Safrole 112 -143", apparently identical with Haworth s ( 1 Paper V I ] : A . a;. Schrecker, C.. IT, Greenberg and, J. I. Hartcompound, n1.p. 138-13Q" .inalysis indicated the w e l l , Tars J O I J R N A I , , 74, J l i G D 11932). presence of four methoxyl groups. S o free hy2 ' K D Haworth a n d T. Richardson. J . C h e w . Sof., 348 119.10). ,'? I C . S p S t h . F.lVe.;ielv a n d I-. Kornfeld, B P I . 66, , 1.586 (19x2). droxyl group could lie detected by treatment with .7. C h e w . Soc., 83 (1933). .I.R o b e r t s o n a n d R. R. n ' a t e acetic :liihvdride or pliosphorus trichloride
I n 1936, IIaworth and liichardson" synthesized
J. compound, m.p.%i7-2( , which proved to be {vit h dehytlronn hydropicro pod op h yll i n t d en tical
c,
.\
\V :chr~rkrr nnrl I I
F1
r w I 1 . 'I'FITS J O T J R X A I . , 74, R67G
Nov. 20, 1952
INTERMEDIATES IN SYNTHESIS OF nEHYDROANHYDROPICROPODOPHYI.I.1N
5673
0
Saponification yielded an acid, m.p. 205-208", which did not lactonize when heated with mineral acid. In addition to the compound, m.p. 112-143", there was obtained in several experiments, in yields ranging from 7 to 17%, an isomer crystallizing from dilute ethanol as tiny colorless needles, map. 169170°, which also gave an analysis indicating four methoxyl groups and which also could be saponified to an acid, m.p. 243-244' (foaming). The infrared spectra (Fig. 1) of the compounds, m.p.
142-143" and lGO-170", showed peaks at 1725 and 1i 1 0 em.-', respectively, certainly corresponding to ester groups. The lactone ring present in podophyllotoxin derivatives5j7on the other hand, gives a maximum located in the range of 1745 to 1700 cm.-l All this evidence indicates that both the compounds, m.p. 112-143" and 169-170", respectively, are methyl esters. They may be (7) Paper XII, A . a'.Schrecker and J I,. Hartwell, THISJOURNAL, t o be published.
li VI1
stereoisomers of structure VI (R = 3,4,5-triniethoxyphenyl, R' = CH3) or, less probably, one or both might have the corresponding structure VII. The same applies to the acids obtained on saponification (R' = H). It may be pointed out that while the ester, 111.1). 1G9-170°, was readily hydrolyzed with .iCjO inethyl alcoholic potassium hydroxide, the compound, 1n.p. 142-143", was saponified only in the presence of water. Formation of the lactone IV by treatment of the ester, n1.p. 142-143') with potassium hydrogen sulfate does not involve simple dehydration, as postulated previously,* but rather the removal of the elements of methanol, requiring cleavage of both the ester and the tetrahydrofuran linkages. X compound of structure I11 iriight be expected to be dehydrated readily under mild conditions, such as heating with iodine in benzene; this was found not to be the case. In addition to the esters. 111.1). 1-42-14;;' :tntl 169-
''
HC ::
' -T'
'1
170", a third compound was obtained in the reaction of I1 with methyl alcoholic hydrogen chloride It was isolated in one experiment as colorless ticedles, 1n.p. 185-127" (yield 7%). The same material, melting a t 12'7-128", was formed in high yield when the lactone I1 was treated with a mixture of acetic and hydrochloric acids; it was found to be identical with G,7-methylenedioxy-1-(3,4,5-triinethoxyphenyl) - 3 - chlorornethyl - 3,4 - dihydronaphthalene (\'I111 Ii-hen this compound was rcfluvetl with 1O ( 2 methyl alcoholic potassium hydroxide, it was converted to 6,7-methylenedioxy- 1(:$4,5 - trimethoxyphenyl) - 3 - methylnaphthalene (IX, R = H), 1ii.p. 146-147". The ultraviolet absorption spectrum of I X (K = H) (Fig. 8) greatly resembled that of the derivative in which the 3-methyl group is replaced by a 3-hydroxymethyl group (IX, R = OH)5; that of dehydroanhytlropicropodophyllin (V), in which the chroinoplioric system is conjugated with the lactone carh n y l group, shows the expected shift toward longer w v e lengths arid increased absorption.
OCI13 1'111
OCITa IX
11similar reaction has been described by Haworth and Kelly,* in which the action of a mixture of acetic and hydrochloric acids on veratryl y-hydroxy-6-veratrylbutyl ketoneg gave 6,7-dimethoxy1 - veratryl - 3 - chloromethyl - 3,4 - dihydronaphthalene, which in turn was converted to 6 , i dimethoxy- 1-veratryl-3-methylnaphthalene 8$10 with methyl alcoholic potassium hydroxide. I t may be pointed out that when the a-acetyl lactone I was treated similarly with a mixture of acetic and hydrochloric acids, Haworth and Atkinsone obtained, not a chloromethyldihydronaphthalene, but rather the acid VI (8= CH3, R' = H). Similarly, they isolated G,i-inethylenedioxy-3,l-endomethyleneoxy-1-methyl-I ,2,3,4-tetrahydronaphthalenein an analogous reaction done with methyl y-hydroxy6-piperonylbutyl ketone, and the product did not react with potassium hydrogen sulfate a t 180". This indicates that formation of the 3,4-dihydronaphthalene derivatives seems to require activation of the I-position by an aromatic substituent. Experimental 1, r~-Acetyl-r-(d ,4-methylenedioxybenzyl)-~-butyrolactone
(1) and ZNI
25(l :loo \ Y a w length (nip).
:iXl
4 00
Fig. 2.--Ultraviolet absorption spectra in 95% ethanol of: , 6,7-methylenediosy-(3,4,5-trirnethoxyphenyl)-3inethylii~plithalenr (IS,I< = EI); . . . . . , . . . , , , ., 6,7incthyleiiedioxy - 1 - (:3,4,5 - irimethoxyplieiiyl) - 3 - hytIrosyineth~lnapht11Lllene (IS,li. = 0 1 3 ) ; - ~ -- - , del i ~ t l r o ~ r i l i \ c ~ r o ~ i i ~ r o ~ ~(17. ~~l~~li~lliii
a-(3,4,5-trimethoxybenzoyl)-~-(3,4-methylenedi-
(3) I < . 1). IIaworth and W. Kelly, J . Chem. S O L . ,1646 (1937). (9) Revised structure, cf. ref. 6. (IO) R . D. Haworth and D. Woodcock. J . Chcm. SOL..809 (1938). ( 1 1 ) All our melting points are corrected and were determined with the Hershberg apparatus. Ultraviolet spectra were measured in 95% cthanr,l on a Beckman Model DU spectrophotometer and infrarrd spectra with a Perkin-Elmer model 21 spectrometer. (12) hlicroanalyses were carried out by the hficroanalytical Laboratory. N I A M D , Xational Institutes o f Health, under the dirrction (JF Ur. W.C Alford
Nov. 20, 1052
INTERMEDIATES IN SYNTIIRSTS OF bEIIYDROAN IIYDROPICROPODOPIIYI,I,Ih'
oxybenzyl)-r-butyralactone (11) were prepared according to
5075
The same compound was isolated in 7% yield as colorless needles, m.p. 12.5127' (from ethanol), in addition t o a 26% Haworth and Richardson's procedure.2 Two Isomers of Methyl 6,7-Methyhedioxy-3,I-endo- yield of the ester, m.p. 142-143', when I1 was refluxed for 1 hr. with 10 volumes of half-saturated mcthyl alcoholic m&hyleneoxy- 1- (3,4,5 -trimethoxyphenyl )- 1,2,3,4tetrahydronrlphfhalene-2-cilrboxylate(VI, R = 3,4,5-Trimethoxp hydrogen chloride. 6,7-Methylenedioxy-l-(3,4,5-trimethoxyphenyl)-3-methphenyl, R' = CHI) (Or Equivalent Structure VII).-When I1 was refluxed with 10 volumes of cold saturated methyl ylnaphthalene (IX,R = H).-A solution of 0.50 g. of VIII alcoholic hydrogen chloride for one-half hour,2 there was in 20 cc. of 10% methyl alcoholic potassium hydroxide was obtained, in yields ranging from 26 to 50% (generally refluxed for 19 hours, then diluted gradually with 20 cc. of around 40%), a colorless solid, melting at about 140°, which hot water. Colorless crystals separated; they were colafter recryst,~llization from ethanol or methanol formed lected after chilling, and washed with ice-cold 50% methcolorless needles, m.p. 143-1-13 ', apparently identical with anol, yield 0.43 g. (95%), m.p. 143-146'. Recrystallization the material, m.p. 1J8-139', to which Haworth and Rich- from 50% methanol gave colorless glistening needles, m.p. 14445-146 00 arduon2 had assigned structure 111. The compound appeared to retain tenaciously traces of solvent of crystalliza4nal. Calcd. for C21H2005: C, 71.58; H, 5.72; OCH,, tion; it could, however, be obtained a n a l y t p l l y pure by 26.42. Found: C, 72.34, 72.25;. H,. 5.87, 5.85; OCHI, drying in a high vacuum for two days a t 110 26.81. Anal. Calcd. for G3H11O8: C , 64.48; H, 5.65; 4 o c I & , Because of the appreciable difference between the calcd. 28.98. Found: C, 64.66; H , 5.85; OCHI, 28.67. and found values for carbon, the material was dissolved in The substance was recovered unchanged when refluxed benzene and chromatographed on a1~rnina.l~The apparwith acetic anhydride and sodium acetate, with phosphorus ently homogeneous adsorbate (purplish-blue fluorescence trichloride in benzene, or with iodine in benzene, and also under ultraviolet light) was eluted with benzene in three when an attempt was made to saponify it by refluxing 0.5 g. fractions, each of which was evaporated and recrystallize? from methanol to give material melting at 145.2-146.5 , for one hour with 12.5 cc. of 5% methyl alcoholic potassium 145.8-147.0' and 145.4-147.0', respectively, with the folhydrouide, folloiied by cooling and diluting with water. Saponification to the corresponding acid (R' = H ) was ac- lowing analyses: (a) C, 72._05; H , 6.27; (b) C, 72.17; H, complished by the folloeing procedure: 1.44 g. of the com- 5.86; OCHa, 26.56; (c) C, 12.01; H, 5.85. 6,7-Methylenedioxy-l-( 3,4,5-trimethoxyphenyl)-3-hypound was refluxed with 40 cc. of 10% methyl alcoholic potassium hydroxide for 2 hours, 40 cc. of water was added droxymethyl-3,4-dihydro-2-naphthoic Acid Lactone (IV)._O and refluxing continued for 16 hours. When the clear, color- An intimate mixture of 1.34 g. of the ester, m.p. 142-143 , less solution was acidified with 40 cc. of 2 N hydrochloric obtained in the reaction of I1 with methyl alcoholic hydroacid, an oil separated which crystalliozed on heating. The gen chloride and 2.7 g. of finely-ground potassium hydrogen yield of colorless solid, m.p. 204-211 , was 1.35 g. (97%). sulfate14 was heated at 180' for 10 minutes, and the very dark mass extracted with chloroform. The solution was Recrystallization from dilute ethanol gave felt-like needles, concentrated and chromatographed on a 1 ~ m i n a . l ~Eluting m.p. 205-208'. The product was soluble in aqueous 50with chloroform and concentrating with addition of ethanol dium bicarbonate. Anal. Calcd. for C22H2208: C , 63.76; H , 5.35; 3 OCH3, yielded 0.41 g. (33%) of orange-colored crystals, m.p. 217249". Similar runs gave yields ranging from 24 to 37%, 22.47. Found: C,63.55; H, 5.28; OCH,, 22.59. while the yield dropped t o 15% when the reaction mixture Fractional crystallization of the reaction product obtained was heated for 45 minutes.2 Subliming the crude material with methyl alcoholic hydrogen chloride gave in three e-: in a vacuum (bath 270"), rechromatographing from chloroperiments, in addition to the compound, m.p. 142-143 , form on alumina, eluting the yellow band with chloroform the isomeric methyl ester in yields ranging from 7 t o 17%. and crystallizing twice from chloroform-ethanol furnished This substance crystallized f r o F 50% ethanol as tiny color- analytically and spectrally pure material as apparently less needles, m.p. 169.4-170.0 . colorless needles, m.p. 250.1-251.0" (lit.z 248-249'), Anal. Calcd. for C23H7401: C, 64.48; H , 5.65; 4 OCH3, 350 mp (log E 4.10), 245.5 m p (log E 4.32), =: : :A 280.5 mp 28.98. Found: C, 64.51; H, 5.56; OCHs, 28.84. (log E 3.52), 236.5 mp (log E 4.29). Chloroform solutions While the compound, m.p. 142-143", was not saponified of the compound were pale yellow and did not absorb broby methyl alcoholic potassium hydroxide in the absence of mine rapidly. water, this method of hydrolys$ was successful in the case Anal. Calcd. for C22H~007:C , EB.06; H, 5.09; OCHE, of the isomer, m.p. 169.4-170.0 The yellow solution, obFound: C, 66.78, 60.83; H, 5.19, 5.23; tained by refluxing 0.90 g. of the ester with 15 cc. of 10% 23.49. methyl alcoholic potassium hydroxide for 3 hours, was di- OCH3, 23.71, 23.57. 6,7-Methylenedioxy-1-(3,4,5-trimethoxyphenyl)-3-hyluted with 90 cc. of water and acidified with 15 cc. of 2 N Acid.-The pale yelhydrochloric acid. The colorless gelatinous solid became droxymethyl-3,4-dihydro-2-naphthoic low solution, obtained by refluxing 315 mg. of II' with 325 crystalline after short boiling; yield 0.82 g. (%yo),m.p. 239-240" (effervescence). I n a similar experiment, in mg. of sodium hydroxide, 1.3 cc. of water and 1.3 cc. of which 0.32 g. of the ester was saponified with 8 cc. of 5% ethanol for 10 minu'tes and diluting with hot water, was cooled in ice, acidified with 2 N hydrochloric acid, anti the methyl alcoholic potassium hydroxide for 1 hour, the yield was 0.23 g. (;A%), in addition to 17% of recovered starting flocculent solid extracted with about 100 cc. of chloroform, material. Recrystallization from chloroform-ethanol, then which then was washed with water and dried over sodium from ethanol, gave long thin needles, m.p. 243-244' (foam- sulfate. Evaporating and triturating the solid residue with ing). The material was soluble in aqueous sodium bicnr- benzene yielded 0.28 g.. (85%) of material mrlting a t 251" (when heated slowly from room temperature), and which bonate. melted at 203' with foaming (whei; immersed a t 200°), reAnal. Calcd. for C22Hn08: C, 63.76; H, 5.35; 3 Oc%, Crystallization from 22.47; neut. equiv., 414.4. Found: C, 63.91; H , 5.36; solidified and melted again a t 251 ethanol-benzene, followed by two recrystallizations from OCHa, 22.31; neut. equiv., 416. 30% ethanol gave tiny felt-like colorless needles, m.p. 2496,7-Methylenedioxy- I -(3,4,5trimethoxyphenyl)-3-~hloro- 250' (immersed at room temperature); melting with foaming methyl-3,4-dihydronaphthalene(VIII).-A solution of 5.0 g. at 203', resolidifying and remelting a t 249-250' (immersed of I1 in 75 cc. of glacial acetic acid, to which 50 cc. of concd. a t 200'). The acid was quite soluble in ethanol, moderately hydrochloric acid had been added, was kept at room tem- soluble in chloroforni and sparingly soluble in water and perature for 18 hours, then, after addition of another 25 cc. benzene. of concd. hydrochloric acid, for 5 more hours. The colorAnal. Calcd. for Cn~HtnOs: C, 03.76; €1, 5.36. Found: less needles which had separated from the red solution were C, 63.77; H, 5.58. collected and washed with 30% acetic acid, then with water. Acknowledgments.-The authors wish to thank The yield was 4.26 g. (91%), m.p. 124-127' RecrystalliMrs. Gertrude Y. Greenberg for some of the zation from absolute ethanol gave material melting a t 127preparative work and for measuring the ultra128". Anal. Calcd. for C21H210SC1: C, 64-86; H, 5.44; C1, (13) Alcoa activated alumina, grade F-20. (14) "Potassium bisulfate, crystal"; only starting material was 9.12; OCHs, 23.94. Found: C, 65.31,65.35; H , 5.42,5.74; rerovered when ''fused potassium bisulfate" was used. C1, 9.48; OCHa, 24.20.
'.
.
.
.
violet spectra. AIrs. Iris J. Siewers and Miss .llice 51, Bernardi, Xational Heart Institute,
~~OSIXIBIJ'l'IlJN FKOM T I I E
kindly determified the infrared spectra. B I C T I I E ~ ~13, A ~IARI~I,AND
LABUKATOKY ( J I P CHEMICAL l'ilARMACOLUCY, NAI'IUNAL CANCER INSTITUTE, OF H E A L T H , PUBLIC FIEALTII SERVICE, FEDERAL SECURITY A G E S C Y ]
Components of Podophyllin.
SATIONAL INSTITUTES
IX. The Structure of the Apopicropodophyllins1a2
R t r .ZSTHONY 51'. SCHRECKER AND JONATHAN L. HARTWELL 1iwr;rvrn h l ~ R C I I2 7 , 10.52
The constitution of a-apopicropotiophyllixi has been established 3 5 ill formula I. Structure IIn for P-apopicropodophyllin f,ivored over ITb on spectroscopic evidence A new isomer, y-npop~cropodophvlli~i (1111, 11:~stwen nhtxitied nntl proven f
