University Microfilms

INFORMATION TO USERS

This dissertation was produced from a microfilm copy of the original document. White the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.

The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.

1. The sign pr "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity.

2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete.

4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.

University Microfilms

300 North Zeeb Road Ann Arbor, Michigan 43106

A Xerox Education Company 73-11,472 CHEN, (Thin-Nan Rolland, 1941- A PHYTOCHEMICAL INVESTIGATION OF THE BARK OF DORYPHQRA SASSAFRAS ENDLICHER. The Ohio State University, Ph.D., 1972 Health Sciences, pharmacy

; • ;/ : University Microfilms, A XEROX Company, Ann Arbor, Michigan

THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED. A PHYTOCHEMICAL INVESTIGATION OP THE BARK OF DORYPHQRA SASSAFRAS ENDLICHER

Dissertation Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By Chin-Nan Rolland Chen, B.Sc., M.Sc*

******

The Ohio State University 1972

Approved by

/ Adviser lege of Pharmacy PLEASE NOTE:

Some pages may have

indistinct print.

Filmed as received.

University Microfilms, A Xerox Education Company ACKNOWLEDGMENTS

I wish to express my sincere appreciation to ,my, . adviser, Professor Jack L. Beal, for his guidance, support, patience and constant encouragement during the course of investigation and throughout my doctoral pro gram. I would like to thank Professor Raymond W. Doskotch for his invaluable advices and determination of n.m.r. spectra and to Professor Lester A. Mitscher for his interest and determination of the circular dichroism spectra. I wish to thank Dr. K. Tomita for providing anonaine hydrochloride and corypalline, Dr. M..- .Shamma for sup plying isocorydine hydrochloride and Dr-. E. Brochmann- Hanssen for furnishing reticuline. Special thanks are due to my fellow graduate stu dents, who generously provided so many stimulating comments and suggestions. And last, but not least, I am especially grateful to my parents for their unfailing support and encourage ment during the years of my education.

ii VITA

October 18, 19*H...... Born - Tainan, Taiwan

I9 6 6...... B.Sc., Pharmacy, Taipei Medical College, Taipei, Taiwan

1967-I96 8...... Teaching Assistant, College of Pharmacy, University of Houston, Houston, Texas

I9 6 8...... M.Sc., Pharmacognosy, College of Pharmacy, University of Houston, Houston, Texas 1968-197 2...... Research Assistant, College of Pharmacy, The Ohio State University, Columbus, Ohio

FIELDS OF STUDY Major Field* Pharmacognosy and Natural Product Chemistry Professor Jack L. 3eal, Adviser

iii TABLE OP CONTENTS Page ACKNOWLEDGMENT, ...... 11 VITA AND FIELDS OP STUDY...... iii LIST OP ILLUSTRATIONS...... viii Chart ..... vlii Table...... ix Figure...... x INTRODUCTION...... 1 Botanical description...... 1 The alkaloids of the family Monimiaceae...... 2 The chemical constituents and pharmacological significance of Doryphora sassafras Endlicher 6 The syntheses of the alkaloids of Doryphora sassafras Endlicher ...... 9 STATEMENT OF PROBLEM...... 12 EXPERIMENTAL...... 13 Materials ...... 13 ‘Methodology...... 13 Chemical analyses...... 13 Valser's reagent...... 13 Mayer' s reagent...... 14 Dragendorff's reagent...... 14 Phosphomolybdic acid reagent...... 15 iv Table of contents (continued) Page Iodoplatinic acid reagent...... 15 Labat's test for methylenedioxy group.. 15 Gibb*s test...... 16 Millon's test...... 16 Evan's test for O-dihydroxyphenol...... 16 Amberlite IRA-*HO anion-exchange resin. 1? Chromatographic analyses...... 17 Physical analyses ...... 18 Melting point determination...... 18 Infrared spectrophotometric analyses... 18 Ultraviolet spectral analyses.• 18 Nuclear magnetic resonance spectrometric analyses...... 19 Mass spectrometric analyses...... 19 Elemental analyses ...... 19 Circular dichroism analyses...... 19 Optical rotation analyses ...... 20 Preliminary investigation...... 20 Quantitative estimation of total alkaloid content...... 20 Pilot study of the extraction procedure.... 23 General extraction and fractionation of the bark of Doryphora sassafras Endlicher...... 26 Thin layer chromatographic examination of the total alkaloids...... 31 v Table of contents (continued) Page Chromatographic separation and isolation of tertiary alkaloids...... 3^ The tertiary non-phenolic precipitate fraction (Fraction B)...... 34

The isolation of alkaloid I (Doryanine) 35 The isolation of alkaloid II (Lirio- denine)...... 38 The isolation of alkaloid III (Iso- corydine)...... 39 The isolation of alkaloid IV (And- naine) ...... 40 The tertiary phenolic precipitate fraction (Fraction D )...... 42 The isolation of alkaloid V (Doryphor- nine)...... **5

Methylation of doryphornine...... 45 Synthesis of doryphornine...... 46 The isolation of alkaloid VI (Reti- culine)...... 54 Methylation of reticuline...... $6

The isolation of alkaloid VII (Base A). 56 Fraction £..... 57 The isolation of alkaloid VIII (Dory- flavine)...... 58 Methylation of doryflavine ...... 60 Acetylation of doryflavine ...... 61 Acetylation of doryflavine dimethylether ...... 62 vi Table of contents (continued) Page The tertiary non-phenolic alkaloid fraction (Fraction F)...... 63 The isolation of doryanine, liriodenine isocorydine and anonaine ...... 63 The tertiary phenolic alkaloid fraction (Fraction G )...... 66 The isolation of alkaloid VI (Reti- culine)...... 66 The isolation of alkaloid IX (Base B).. 66

Attempted methylation of base B...... 69 N-methylation of base B ...... 70 The isolation of alkaloid X (Cory- palline)...... 71 Chromatographic separation and isolation of the total quaternary alkaloid fraction...... 73 The isolation of alkaloid XI (Choline chloride)...... 73 Antimicrobial testing of the isolated compounds ...... 77 DISCUSSION...... 79 SUMMARY ...... 101 REFERENCES...... 103 APPENDIX...... 107

vii LIST OP ILLUSTRATIONS Chart Page I Flow sheet for the extraction of the total alkaloids...... 28 II Flow sheet for the fractionation of the tertiary precipitate alkaloids...... 30 III Flow sheet for the fractionations of the tertiary and quaternary alkaloids...... 32

viii Table Page 1. Results of thin layer chromatographic examination ...... *...... 33 2. Results of the column chromatography of the tertiary non-phenolic precipitate fraction (Fraction B)... • ...... 36 1 3 . Results of the column chromatography of the phenolic precipitate fraction (Fraction D)...... 43 k, Results of the column chromatography of fraction E ...... 58 . 5. Results of the column chromatography of the tertiary non-phenolic alkaloid fraction (Fraction F)...... 6k

6 . Results of the column chromatography of the tertiary phenolic alkaloid fraction (Fraction G)...... 67 7. Results of the column chromatography of the total quaternary alkaloids (Fraction H) 7k

8. Results of the antimicrobial testing...... 78

9 . Comparison of uv, ir and n.m.r. absorption spectra of doryflavine and related com pounds ...... 92

ix Figure Page 1. The synthesis of doryafranine ...... 10 2. The synthesis of doryanine ...... 11 3. Results of thin layer chromatographic analyses of the residues...... 25

4. Elution pattern for fraction B ...... 37 5. Elution pattern for fraction D ...... 44 6. The synthesis of doryphornine...... 4? 7. Elution pattern for fraction F,...... 65 8. Elution pattern for fraction G .... «... 68 9. Elution pattern for fraction H ...... 75 10. Ultraviolet spectra of doryflavine and related compounds...... 94 11. IR spectrum of doryphornine (CHCl-j)...... 108 12. NMR spectrum of doryphornine (CDCl^)...... 109 13. IR spectrum of synthetic thalifoline (KBr)..*. 110 14. IR spectrum of synthetic doryphornine (CHCl^)* 111 15. NMR spectrum of synthetic doryphornine (CDCl^) 112 16. IR spectrum of base A (KBr)...... 113 17* NMR spectrum of base A (CF^COOH)...... 114 18. IR spectrum of doryflavine (KBr)...... 115 19* NMR spectrum of doryflavine (Pyridine-d^) 116 20. IR spectrum of doryflavine dimethylether (chci3)..'..'...... 117 21. NMR spectrum of doryflavine dimethylether (CDC13)...... 118 ■v '

Figure Page 22. Mass spectrum of doryflavine dimethylether.... 119

2 3 . IR spectrum of doryflavine triacetate (KBr)... 120 24. NMR spectrum of doryflavine triacetate (CDCl^) 121 25* IR spectrum of base B (KBr)...... 122 26. NMR spectrum of base B (CF^COOH)...... 123 27. IR spectrum of N-methylated base B (CHCl^).... 124 28. IR spectrum of l-(p-methoxybenzyl)-2-methyl - 6-methoxy-l,2 ,3 ,4-tetrahydro-7-isoquinolinol

Xi INTRODUCTION

Botanical description The plant family Monimiaceae is composed of approx imately *f0 genera and 460 species, which are widely distributed in tropical and subtropical regions. They are trees or shrubs, rarely climbers, and usually possess fragrance (1 , 2 , 3 ), The genus Doryphora is limited to a single species endemic in Australia. Doryphora sassafras Endlicher is the characteristic sassafras tree of New South Wales (^). The tree is indigenous to Eastern Australia and probably received the species name because its bark, leaves, and fruits contain an essential oil having the characteristic odor of sassafras. The plant is described as a tree of considerable size, which grows to an average height of

50 to 80 feet, but in. same places has attained the height of 180 feet. The botanical characteristics of this plant had been given by Bailey (3) "Leaves petiolate, ovate elliptical or oblong-lanceolate, acuminate, coarsely toothed, narrowed at the base, 2 to 4 in, long nearly smooth on the upper side, prominently penniveined and

1 reticulate underneath. Peduncles 2 to 3 lines^long, with a pair of very deciduous bracts of 3 or 4 lines close under the flowers. Perianth-tube about 1 line long when in flower, enlarged and irregularly split when in fruit, segments about 4 lines long, lanceolate, very acute. Anther-appendages nearly as long as the perianth-segments. Carpels slightly hairy, the styles lengthening after fecundation into long plumose awns." It had been reported (4) that the aborigines and also the country people made a tea from the bark, which they drank as a tonic. In addition, the light-yellow wood, which possesses the fragrance of the bark, was found not to be attacked by insects.

The alkaloids of the family Monimiaceae In a recent survey Raffauf reported (5) that approxi mately 50 alkaloids have been isolated and characterized from the Monimiaceae family. Of these alkaloids, the aporphine type is the most abundantly distributed. Boldine (1), isocorydine (2), pukateine Q) and laureline (4) are typical examples.

1 a disused unit of length equal to 0.5 inch. 3

CH O CH.

H C O O

1 R,=R3-H,R2=CH3tR4=OH 3 R =OH, R2=H

2 R1=CH3iR2=R4=H,R3=OCH3 4 Rj-H, R2=0CHa

The bisbenzylisoquinoline alkaloids are also commonly found in this family. On the basis of numbers and posi tion of the ether linkages in the molecule, two distinct structure types of these alkaloids can be classified. The berbamine (j£) and oxyacanthine (6) types involved two ether linkages. o OCH HXO OCH Q O o

OCH HC O

HO CH

OH 6 The micranthine type (£) of alkaloids are those having three ether linkages, of which two are part of a dibenzo-i,^-dioxide system. o o o o

The simple benzylisoquinoline alkaloids sure less frequently distributed in this family. Doryafranine (8,) serves as an example of this type. ,o< o '"CH,

HXO o The liriodenine type of SLlkaloids are rather typical. in this family. Liriodenine (2.) and atheroline (10) are representatives of this group. OCH. HO The aminoethylphenanthrene type of alkaloids, which can be obtained from the Hofmann degradation of a corres ponding alkaloids, were also discovered in this family. Atherosperminine (11) and methoxyatherosperminine (12)

Eire examples.

OCH H.CO H.CO CH H C HXO CH

The proaporphine alkaloid, which is the biogenetic precursor of aporphine alkaloids, was found in one instance. Stepharine (1^) was the example. 6

The isoquinolone alkaloids are present in this family. Doryanine (14) represented the first example of a naturally occurring isoquinolone base. Thalactamine (1£), which was isolated from a Thalictrum minus variety by Mollov, was the other example (6).

OCH HXO

HXO CH

14 15 The above survey indicates that almost all of the alkaloids isolated from this family thus far, are derived from benzylisoquinoline biosynthetically.

The chemical constituents and pharmacological signifi cance of Dorvohora sassafras Endlicher. The chemistry of this plant had been initially studied by Petrie as early as 1921 (4). He studied the oil of the bark and compared it with the essential oil obtain ed from Atherosperma moschatum Labill, Cinnamomum oll- verl Bailey and Sassafras officinale Nees. The oil ob tained from this plant, had a density of 1 .033, and distilled between 60 and 230°. The bark contained 1.35# of the oil, leaves 4.3# and fruits 4#. Penfold (?)

* studied the variation in the yield of the different X 7

constituents of the essential oil of the leaves, according to time of season and origin. Further! Petrie (4) made a preliminary examination of the bark and reported the- following approximate constituents *

(a) (b) Volatile essential oil 1.11795 1.24

Fixed oil 0 .5 6 6 0 .6 3

Resins 1.234. 1.37 Tannins (etc. pptd by Pb(0Ac)2) 1.240 1 .3 8

Reducing sugars (as glucose) 1.000 1 .1 1

Saccharose sugars 0.530 0.59 Calcium oxalate 1 .2 3 0 1.37 Alkaloid (approx.) O .325 0 ,3 6 (a) calculated on air-dry material (b) on material dried at 100°C In addition, Petrie (4) isolated an alkaloid from the bark which was described as an amorphous yellowish-grey powder and was highly electrostatic when brishing it out from one vessel to another, it either strongly adhered, or flew off and scattered. All attempts to obtain it or its salts in crystalline form were unsuccessful. The compound had a melting point of 115° - 11?°C, and possessed

a slightly bitter taste, and the reaction was faintly al kaline to litmus. It dissolved readily in alcohol, chloro form and dilute acidi was very slightly soluble in ether, 8 and wateri and insoluble in petroleum spirits. The equivalent weight of the alkaloid was estimated to be 342 by titration. The elemental analysis suggested the molecular formula ^18^21^4* Tiie which Was assigned the name doryphorine, was found to be present in a concentration of 0 ,30% in the leaves, 0.54# in the bark and 0.1% in the fruits. The pharmacological action of doryphorine on the frog showed the loss of power of movement, and response to touch, the heart beat became slower and, within 8-10 minutes, paralyzed and stopped. The minimum lethal dose for Hvla aurea. a 13 gin. frog, was 7 rag. The alkaloid did not produce convulsion and had no action on nerve, receptive substance or muscle (4). An extensive investigation of the alkaloids of the leaves of this plant was reported by Gharbo et al.,(8) in

1 9 6 5 . Eight alkaloids were isolated as crystalline,pro ducts. Of these, two were identified as liriodenine (£) and choline chloride (16), another two new alkaloids were assigned the names doryanine (14) and doryafranine (8) while the remaining four alkaloids A, fi, C and D remained to be identified.

C H , Cl® V-pCHjCHjO H CH, 16 9

The syntheses of the alkaloids of Doryphora sassafras Endlicher. The syntheses of the two new alkaloids doryafranine and doryanine have subsequently been accomplished by Buck and Cava (9)* Doryafranine had been synthesized on three previous occasions. However, Tomita and Santomi (10) had only obtained this compound as a yellow oil. Sasaki, Ohnlshi and Satoh (11) synthesized this compound as.a yellow gum. On the other hand, this compound was obtained as a byproduct of the ring closure of the diazonium salt to give O-methylmichepressine (12). This product was characterized through its methiodide. The total synthesis leading to doryafranine as a white crystalline product was carried out by Buck and Cava utilizing the following schemet 10

H3C °^ 0 ^ ~ CH2 ^~CI o HCIC. pC»S .e, < ^ X j .coiQr H_CO o

NoBH4 O o CI-?OE«Vpv. . fcOE*

o H,C o o Ll Al H4 CH, o

Figure 1 The synthesis of doryafranine The synthesis of doryanine by Buck and Cava was as presented in the following schemes

Et + H2N-CHjCH(OEt)

K3F e (C N )4

Figure 2 The synthesis of doryanine STATEMENT OF PROBLEM

It has long been recognized that the alkaloids repre sent one of the richest source for medicinal agents, for examplej morphine, codeine, reserpine, ergotamine, ephedrine, quinine, vinblastine and many others. Conse quently, plants which have been reported to contain alkaloids are promising to investigate. Since the previous survey indicated that the bark of Dorynhora sassafras Endlicher had been used as a tonic by the aborigines and the country peoplej and the isolated alkaloid, doryphorine, showed a remarkable pharmacological activity, it was, therefore, considered worthy of a further phytochemical investigation. The purpose of this study is to isolate and separate the alkaloids present in the bark, and determine their chemical and physical properties, so that they can be properly identified and characterized and to provide sufficient material, when possible, to enable pharma cological testing.

12 EXPERIMENTAL

Materials The plant material used in this investigation was the bark of Doryphora sassafras Endlicher (Monimiaceae), which was obtained in 1966 from Australia through Meer

Corporation, 3I8 West 46th Street, New York, New York. It was designated Lot 6-2383** by Meer Corporation and an unground specimen of bark was saved. The remainder of the air-dried bark was ground to a fine particle size by means of a Wiley mill.

Methodology Chemical analyses Valser*s reagent (1*3) - The presence of alkaloids is suggested when 1 ml of an aqueous acidic solution gives a white or yellow precipitate upon treatment with 1 to 2 drops of Valser's reagent. The reagent was prepared1 Potassium iodide, 10 g., in 100 ml. of water was added slowly with stirring to about 15 g. of mercuric iodide. The mixture was stirred until almost all the mercuric iodide was dissolved and the solution was filtered.

13 i4

Waver's reagent (13) - The test solution is similar to that of Valser's reagent. The preparation of this reagent was as followsi Mercuric chloride, l.k g., was

dissolved in 60 ml. of water and poured into a solution containing $ E* potassium iodide in 10 ml. of water.

\ The solution was diluted to 100 ml. with water. Dragendorff1s reagent (1*0 - Alkaloids are detected as orange to red spots on a light orange background when sprayed with Dragendorff's reagent on paper or thin layer chromatograms. The background can be decolorized by spraying with 2 percent aqueous acidic solution. The stock solution was prepared by boiling 2.6 g. of bismuth subcarbonate and 7.0 g. of dry sodium iodide

in 25 ml. of glacial acetic acid for a few minutes, allowing the solution to cool overnight, filtering and mixing 20 ml. of the clear filtrate with 80 ml. of ethyl acetate. The spray reagent was made by mixing 10 ml. of the above stock solution with $0 m l . of glacial acetic acid and 120 ml. of ethyl acetate, and then adding slowly with stirring 10 ml. of water. Phosphomolvbdic acid ~reagent - Alkaloids which have phenolic functions may be detected by spraying phosphomolybdic acid spray reagent on chromatograms and then ex posing to ammonium hydroxide fumes* A deep blue spot indicates a positive test. The reagent was prepared by dissolving 2 g. of

phosphomolybdic acid in 50 ml. of distilled water and

50 ml. of acetone. Iodoplatinic acid reagent (11) - Alkaloids can be detected with this reagent as grey to violet or blue spots on a pink to purple background. The reagent was

prepared by mixing 3 ml. of a 10 percent solution of

chloroplatinic acid with 97 ml. of water and 100 ml.

of 6 percent potassium iodide solution. , Labat*s test for methvlenedioxv group - The alka

loid, 5 mg., was dissolved in 3 ml. of chromotropic acid test solution in a small test tube and warmed in a steam bath for thirty minutes. A positive test for methylenedioxy group was indicated if the color of the solution changed to pink. The test solution was prepared by dissolving 0.5 g. of chromotropic acid in $0 ml. of distilled water and 200 ml. of 12.5 T.T• sulfuric acid. 16

Glbb*s test (16) - The tests were performed on the thin layer plates and were done either before or after spraying with iodoplatinic acid. The plate was immersed • 1 in running water for a few seconds, then wetted with a saturated aqueous solution of sodium bicarbonate, warm ed over a steam bath for a few seconds, and finally treated with a few drops from a pipet of an aqueous suspension of 2,6-dichloroquinonechloroimide. An im mediate bright blue color at the alkaloid spot was taken as a positive test for an unsubstited aromatic position para to a phenolic hydroxy function. Millo^s test (17) - The alkaloid, 1 mg, was sus pended in a few drops of water and to this was added 2 ml. of Millon's reagent. If no change occurred, the mixture was briefly warmed. A red color indicated the presence.of an unsubstituted aromatic position ortho to a phenolic hydroxy function. The reagent was prepared by dissolving one part of mercury in one part of fuming nitric acid and two parts of distilled water. Evan*s test for O-dihydroxvohenol (18) - The tests were performed on the filter paper which was spotted with the compounds. A solution of 10 percent sodium molybdatei0.5 N hydrochloric acid (2il) was sprayed first* 0-dihydric phenols gave a weak yellow color. A

solution of 0 .5 percent sodium nitrite was then sprayed. 17 the spots turning greenish yellow in color. A final spraying with 0,5 N sodium hydroxide solution resulted in the spots immediately becoming a strong pink. Amberlite IRA- frlO anion-exchange resin - The resin was obtained from Mallinckrodt Chemical Works, New York, New York. The resin was purified by washing successively with 10 percent sodium hydroxide, distilled water, 10 percent hydrochloric acid and distilled water. The wash ing process was repeated three times. At the last stage, the resin was in the chloride form. Chromatographic analyses - Thin layer chromatography was performed on silica gel G using either 20 x 20 cm. or 5 x 20 cm. glass plates with a thickness of 0 .2 5 mm of adsorbent. Silica gel G was obtained from E. Merck, Darmstadt, Germany; Distributori Brinkmann Instruments, Inc., Cantiague Rd., Westbury, Long Island, New York. Unless otherwise specified, the solvent system used for the development was butanoliacetic acidtwater (JJ-ilil). The alkaloid spots were revealed by spraying with Drag- endorff's or phosphomolybdic acid reagent or ultravio let light. The adsorbents used for column chromato graphy were 100 mesh silicic acid, which was obtained from Mallinckrodt Chemical Works, St. Louis, Missouri and neutral or acidic alumina, which was obtained from M. Woelm, Germany; Distributor* Watters Associates, Inc., Framingham, Mass. 18

Physical analyses Melting point determination - The melting points of the isolated compounds were measured using a Thomas Hoover capillary melting point apparatus obtained from the Arthur H. Thomas Company, Philadelphia, Pennsylvania. For compounds which melted over 200°C, the melting points were determined- on a Fisher-Johns hot stage melting point apparatus manufactured by Instrument Division, Fisher Scientific Inc., Pittsburgh, Pennsylvania. The temper atures were corrected. Infrared spectronhotometrlc analyses - The infrared absorption spectra were determined in either a potassium bromide pellet or a chloroform solution using a Perkin- Glmer Model 257 Grating Infrared Spectrophotometer, man ufactured by Perkin-Elmer Corporation, Norwalk, Connect icut.

Ultraviolet spectral analyses - The ultraviolet ab sorption spectra were taken in methanol or ethanol us ing a Cary Model 15 Recording Spectrophotometer made by The Allied Physics Corporation of Monrovia, Calif ornia. All the alkaloids were analyzed in ethanol,

methanol or 0 .0 1 percent methanolic potassium hydroxide, or 0 .0 1 percent methanolic hydrochloric acid. Nuclear magnetic resonance spectrometric analyses - The nuclear magnetic resonance spectra were determined in deuterochloroform, deuteropyridine or trifluoroacetic acid using a Varian A-60 Nuclear Magnetic Reson ance Spectrometer manufactured by Varian Associates, Palo Alto, California. Tetramethylsilane was employed as the internal standard. Mass spectrometric analyses - The high resolution mass spectra were determined using an A.E.I. MS-902 Double-Focussing Mass Spectrometer equipped with a. direct inlet system manufactured by Allied Electrical Industries, England. The spectra were determined by Dr. R. L. Foltz of Battelle Memorial Institute, Col umbus, Ohio and by Mr. R. Weisenberger of the Chemis try Department, Ohio State University. Elemental analyses - The elemental analyses were determined by Scandinavian Microanalytical Laboratory, Box 25* Herlev, Denmark. Circular dichroism analyses - Circular dichroism (C.D.) measurements were made on a Durrum-Jasco Model U.V./0RD/CD5 Spectrophotometer with a modernized optical system (Sproul Instruments Corporation). The result of

C.D. expressed in molecular ellipticity (9 ] units were calculated from the following equation1 20

33 x m.w,. xUj I"* c x 1 x 100 M.W, * Molecular Weight of the sample i absorption in degrees c i concentration of the solution in g./ml.

1 i length of the cell in dm. Optical rotation analyses - Optical rotation measure ments were made on a Perkin-Elmer 1^1 Polarimeter manu factured by the Perkin-Elmer Corporation, Norwalk, Conn ecticut. The specific rotation (ajj at temperature t was calculated from the following equation*

f n l * = 1 0 0 l Jd " H T T T Q * corrected observed rotation in degrees

1 = length of the polarimeter tube in dm. c • = concentration of the solution in g./lOO ml. Preliminary investigation ■ Quantitative estimation of total alkaloid content (19) A sample of 20 g. of finely ground bark of Dorvphora sassafras Endlicher was placed in a 250 ml. round bottom

flask. A volume of 150 ml. of 95 percent ethanol was added and the mixture refluxed for two hours. The con tents of the flask were allowed to cool to room temper ature, filtered, and the filtrate was evaporated to dry ness in vacuo at ^0°C. To the residue was added 50 ml. of a 5 percent solution of ammonium hydroxide and 50 ml. of chloroform. The mixture was stirred to dissolve as

* 21 much of the residue as possible in the two solvents. The contents of the flask were then transferred to a separa tory funnel and allowed to separate into two layers. The chloroform layer was collected and the alkaline mother liquor was extracted two additional times with 25 ml. portions of chloroform. The combined chloroform ex tracts were dried over anhydrous sodium sulfate and ev aporated in vacuo at *fO°C to 25 ml. The chloroform so lution was extracted with 5 ml. of 5 percent hydrochlor ic acid. This aqueous acid extractive was designated as the total tertiary alkaloid fraction. The alkaline mother liquor, devoid of tertiary alkaloids, was acid ified with hydrochloric acid and filtered. The fil trate was adjusted to 50 ml* with 2 percent hydrophloric acid and designated as the quaternary alkaloid fraction. Two ml. each of the tertiary and quaternary alkaloid fractions were placed in separate 12 ml. centrifuge tubes, and the volumes made up to 5 ml. each.. Three drops of Valser's reagent were added to each of the tubes, and the resulting precipitate centrifuged for 5 minutes. The heights of the resulting precipitates in these tubes were then compared to the heights pf reference alkaloid standards for quantitative estim ation. Brucine and aspidospermine were employed as the

standards of the tertiary and quaternary fractions, respectively. The concentrations of these reference alkaloids and the interpretation of the test results are as followsi a. Tertiary alkaloids - "brucine, prepared in 2 per cent hydrochloric acid in the following concentrations

0.4 mg./ml. - +1

1 *3 mg./ml. = +2 4.0 mg./ml. = +3 b. Quaternary alkaloids - aspidospermine was pre

pared in 2 percent hydrochloric acid in the following concentrations t

0 .0 5 mg./ml. = +1

0 .1 7 mg./ml. = +2

0 .5 0 mg./ml. = +3 +1 indicates that 10 Kg. of dried plant will

yield approximately 1 g. of alkaloids or 0 .0 1 percent. +2 indicates that 3 Kg. of dried plant will

yield approximately 1 g. of alkaloid or O.033 percent. +3 indicates that 1 Kg. of dried plant will

yield approximately 1 g. of alkaloid or 0 .1 0 percent. +4 indicates an alkaloid content greater than

0 .1 0 percent. It was found that in the Dorvnhora sassfras bark the tertiary alkaloid fraction gave +4 and therefore 23 would have alkaloid contents of greater than 0 .1 0 per cent. The quaternary alkaloid fraction gave a +2 result and therefore would have alkaloid contents of approxi mately 0 .0 3 3 percent. Pilot study of the extraction procedure In order to obtain information that would facili tate the planning of an appropriate approach or best methodology for the isolation of the major alkaloids present, a preliminary examination was carried out in the following manners. . The bark of D. sassafras Endlieher, 100 g., was defatted initially with Skellysolve B. The dried marc was moistened with 50 ml. of 10 percent ammonium hydroxide solution and subsequently extracted exhaustively and successively with Skellysolve B, benzene, ethylene dichloride and ethanol. Evaporation of the extracts _in__vacuo at 40° gave residue I, II, III, and IV respec tively. A similar treatment as above except the dried defatted marc, 100 g., was moistened with 5 ml. of 20 percent sodium carbonate solution rather than 10 percent ammonium hydroxide solution was also carried out. The residues obtained were designated as V, VI, VII, and VIII. Comparison of the various residues with thin layer chromatography utilizing silica gel G as adsorbent and chloroform:methanoltammonium hydroxide solution (9 2:5 *3 ) as developing solvent (see Figure 3) indicated that the differential extraction was not worth the effort. Thus ethanol as a single solvent was used for large scale extraction. Furthermore, the results showed no artifacts had been formed when ammonium hydroxide was employed. The selection of the acids used for the extraction of the residue obtained from ethanol fraction, was based on the quantity of alkaloids being dissolved. In each instance a O.lg sample of the ethanol extract was ex tracted with 7 nil. of solvent. The solvents used were 5 percent hydrochloric acid, 5 percent acetic acid, and I’M.citric acid. In each instance the acidic extract was filtered and adjusted to 10 ml. volume.

To the acidic solution in the test tube was added 3 drops of Valser's reagent, the resulting precipitate was centrifuged. Comparison of the precipitates in dicated 5 percent acetic acid solution was the most suitable solvent. Figure 3 Results of thin layer chromatographic analyses of the residues. NagCC^ NH40H

M 0T 2E z m nr X X

f\s 0 0 V 0 \ y 0 & 0 i b 11 0 0 » » A a < 0 A } 26

General extraction and fractionation of the bark of Doryphora sassafras Endlicher A quantity of 19.4 kg. of the bark of D. sassafras was defatted with Skellysolve B by percolation at room temperature. The percolator used for this extraction was 40 cm. in diameter by 50 cm. in height and was equipped with a pump for circulation of the solvent through the plant material (20). The plant material was allowed to macerate in the solvent for 24 hours and the solvent was then circulated through the plant for \ a period of 12 hours. The percolator was then drained and fresh solvent added to the plant material. This process of maceration and circulation of solvent was repeated until the extract gave negligible amount of residue after evaporation. The combined Skellysolve B percolates (56.9 1») were evaporated in vacuo at 40°C to yield 75 g* of residue, which represented O.38# of the total plant material. The defatted marc was allowed to dry at room temperature and subsequently extracted in a similar manner with 95 percent ethanol. The com bined ethanolic extracts (130 1*) were concentrated in vacuo at 40°C to give 1.6 Kg. of residue, which represented 8.2295 of the total plant material. A 1.1 kg. quantity of the ethanolic extract was extracted with 5 percent acetic acid solution until the aqueous acidic extract no longer gave a positive test 27

for alkaloids with Valser's reagent. The insoluble

residue was 185 g. and was designated as fraction A.

The combined acetic acid extracts, 20 1., were cooled in the refrigerator overnight, filtered, and basified with concentrated ammonium hydroxide to pH 8-9. The voluminous yellow precipitate was filtered, and immediately dissolved in 6.5 !• of 5 percent of sodium » hydroxide solution {The yellow precipitate became dark brown on standing ) (See Chart I.). The fractionation of the yellow precipitate into tertiary phenolic and tertiary non-phenolic fractions was carried out by extracting exhaustively the sodium

hydroxide solution (6 .5 1.) with chloroform. The com

bined chloroform extracts (25 1 .) were dried with sodium sulfate and' evaporated in vacuo at 40°c to yield 62 g. of residue (Tertiary non-phenolic precipitate or Fraction B ). The cooled extracted sodium hydroxide

solution was acidified with glacial acetic acid to pH 6

and then basified with ammonium hydroxide to pH 9 . The

basic solution was filtered to obtain 117 g. of brown precipitate (Fraction C). The filtrate was extracted with chloroform (50 1 .), which was dried and concentrated

to leave a 23 g. of residue (Tertiary phenolic precipi tate or Fraction D ). The remaining alkaline fraction was further extracted with ethyl acetate (2 1 .) to yield 28

Chart I Flow sheet for the extraction of the total alkaloids

Bark of Doryphora sassafras Endlicher (19.4 kg.)

Skellysolve B (5 6 .9 1.) I------1 Skellysolve B soluble Dried marc <75 g.) No alkalod Ethanol (I30 1 .) 1 Ethanol soluble (1.6 kg.) Marc (Discarded) 1.1 kg. of residue 5# Acetic acid (20 1.) Filter 1 Acidic solution Insoluble residue (185 g.) Fraction A

Alkaline solution Yellow precipitate 29

0.391 g« of yellow residue (Fraction E). The mother liquor which did not contain any more alkaloid was discarded (see Chart II). On the other hand, the- alkaline solution which was obtained after removing the yellow precipitate was extracted exhaustively with chloroform. The combined extracts (24 1 .) were dried with anhydrous sodium sulfate and evaporated to yield 58 g. of tertiary alkaloid residue. In a similar fashion as that of yellow precipitate, the residue was fractionated into 28 g. of tertiary non- phenolic alkaloids (or Fraction F) and 20 g. of tertiary phenolic alkaloids (or Fraction G). The remaining alkaline solution, which was obtained after removing the tertiary alkaloids and should contain the quaternary alkaloids, was acidified with acetic acid to pH 2, a solution of Mayer's reagentwas added until the precipitation was complete. The mixture was refrig erated overnight. The precipitate was then filtered and washed with ether to yield a crude brown solid. The dried crude precipitate (30 g.) was dissolved in 2 1 . of a solvent mixture containing acetone (1 part), methanol (3 parts) and water (1 part). The purpose of this modification, instead of using the conventional solvent mixture containing acetone (1 part) and water

(1 part), was intended to decrease the possibility of 30

Chart II Plow sheet for the fractionation of the tertiary precipitate alkaloids

Yellow precipitate 1. 5% NaOH solution (6.5 1*) 2. Chloroform (25 !•)

{------I NaOH solution Chloroform soluble (62 g.) 1. Acetic acid to pH 6 Tertiary non- 2. NHj^OH to pH 9 phenolc precipi tate or Fraction B I----- — I Alkaline Brown solution precipitate (117 g.) Fraction C Chloroform (50 1.) I-- Alkaline solution Chloroform soluble (23 g.) Tertiary phenolic Ethyl acetate precipitate or (2 1 .) Fraction D Ethyl acetate Mother soluble liquor (0.391 g.) (Discarded) Fraction E 31

the ion exchange resin being dissolved in acetone. The solution was filtered, the filtrate (2 1.) was mixed with an equal volume of Amberlite IRA-410 anion- exchange resin in the chloride form and placed in a

6 1. beaker. The mixture was stirred by a mechanical stirrer for 2 days. The resulting pale yellow solution was filtered under suction, and the resin was further washed with the solvent mixture until no more alkaloid could be detected. The combined filtrates (5 1») were evaporated in vacuo at 40°C and freeze-dried to yield 13.6 g. of quaternary chloride (Fraction H) (see Chart III). Thin laver chromatographic examination of the total alkaloids Before the more extensive separation of various frac tions of the total alkaloids was undertaken, an examina tion utilizing thin layer chromatography was carried out. Such examination revealed the number of alkaloids present in individual fractions. The most satisfactory thin layer chromatographic system was found to be using silica gel G as the adsorbent with one of the two following solvent systems:n-butanol1 acetic acid1water (4:1tl) or chloro form 1 methanol: ammonium hydroxide (9 2t5 »3 ) as the develop ing solvent. Dragendorff's spray reagent and ultraviolet light were used to detect the alkaloids. The results of this examination are presented in Table 1. 32

Chart III

Plow sheet for the fractionations of the tertiary and quaternary alkaloids.

Alkaline solution

Chloroform (2*f 1.)

Alkaline solution Chloroform soluble (58 g.) 1« Acetic acid to pH 2 1 * 55^ NaOH solution 2. Mayer's reagent (2 1 .) Filter 2. Chloroform Filtrate Precipitate (1 1.) (Discarded) (30 g.) r I NaOH solution Acetonei Chloroform MeOH*water 1. Acetic soluble acid to (28 g.) (1*3*1) Amberlite pH 6 Tertiary vIRA-410 Cl 2. NH^OH to non-phenolic Quaternary pH 9 alkaloids chloride Chloroform or (1 3 .6 g.) (4 1.) Fraction F Fraction H 1 Mother Chloroform liquor soluble (Discarded) (20 g.) Tertiary phenolic alkaloids or Fraction G 33

Table 1 Results of thin layer chromatographic examination Enactions Weight (g.)' Alkaloid R,

1* . 2*. Fraction A 185 o. o.o4, 0 .0 5 0, 0.44, 0.80 (Insoluble o.li, 0 .1 7 0.91 residue)

Fraction B 62 0, 0 .1 1 , 0 .25. 0, 0.19. 0.26 (Tertiary 0 .3 4 , 0 .4 5 , 0.37, 0.45. non-phenolic 0.54, 0.68 0.53. 0.72 precipitate) 0.84, 0.93 Fraction C 117 0 , 0.04, 0 .0 5 0.81 (tailing) (Brown 0.08, 0.11 precipitate)

Fraction' D 23 0 { 0.04, 0 .0 7 0.25. 0.37, (tertiary 0 .1 3 , 0.16, 0.45, 0.53. phenolic 0.20 O.6 3 . 0.77, precipitate) O .8 3 Fraction E 0.391 Tailing .0.91, 0.4 ( tailing)

Fraction F 28 0 , 0 .2 9, 0.46 0 • 20, 0.311 (Tertiary 0.59 0 .3 8, 0 .4 4 , non-phenolic O.5 5 , 0 .6I, alkaloids) 0.69, 0.93 Fraction G 20 0 , 0 .1 1 , 0 .1 5 , 0 .2 0 , 0 .2 6, (Tertiary 0 .1 8 , 0 .2 5 O.3 6, 0.42, phenolic 0.52, 0.55. alkaloids) 0.59 Fraction H 1 3 *6 Tailing 0 .2 0 , 0.28, (Quaternary 0.33. 0.42, chloride) 0.53 *1 Solvent system 1 chloroform methanol: ammonium hydroxide (92*5*3) *2 Solvent system: n-butanoliacetic acid *water (4:1:1) 3^

Chromatographic separation and isolation of tertiary alkaloids The tertiary non-phenolic precipitate fraction (Fraction B) A column (8 x 80 cm.) was packed in chloroform with a slurry of 100 mesh silicic acid to a height of 69 cm. in the following manner* After mounting on a ring stand and insuring vertical with a level, the column was one-third filled with chloroform. A small cotton pad was moistened with chloroform, dropped into the column and firmly tamped into the neck above the stopcock with the aid of a long glass rod. Sea sand was added to just above the neck of the column. The uniform slurry which was prepared by suspending 1814.4 g. (4 lb.) of silicic acid in 5 1* of chloroform, was poured into the column and allowed to settle at the adjusted flow rate (usually 5 ml/min.). After the adsorbent had settled, a solution of 31 g. of the residue in 100 ml. of chloroform was introduced into the column through a pipette. The stop cock was then opened to allow the solution to flow into the adsorbent. The wall of the column was further rinsed with chloroform. As the last of the solvent ran into the column, the stopcock was closed and a layer of sea sand (ca. 2 cm.) and a filter paper disc (Whatman filter paper, 3 mm.) were placed on the top of the column. 35

Elution was initiated with chloroform, and continued successively with chloroform containing an increasing amount of methanol. The eluent was collected through a fraction collector. Thin layer chromatography was used to monitor combination of the fractions. The results of the column chromatography are presented in Table 2 and Figure if. The isolation of alkaloid I (Dorvanine) - The alka loid (Rf 0 .9 3) was isolated from fraction 101-103. It crystallized as pale yellow needles from chloroform. It was recrystallized twice from a mixture of acetone and petroleum ether (bp 65-110°c) to give colorless needles (200 mg.), m.p. 162-163°G. The ultraviolet absorption spectrum of this com pound had the following A values* 230 mu (log £ H l a A • if.if6 ), 2if5 (^.**9), 258 (if.ifl), 283 (3.81). 29*f (3.85). 326 (3 .6 3), 3*f0 (3 .50 ). There was no shift in A max. under alkaline or acidic condition. The infrared spec trum in a potassium bromide pellet indicated a con jugated carbonyl group at I650 cm”1 and a methylene- — 1 *1 dioxy group at 1030 cm and 9*f0 cm . The latter was confirmed by a positive Labat*s Test. The nuclear magnetic resonance spectrum in deuterochloroform showed the possibility of two aromatic protons at 7.80 & (s, 1H), 36

Table 2 Results of the column chromatography of the tertiary non-phenolic precipitate fraction (Fraction B)

Fraction Eluent Weight Alkaloid number composition (g.) Rf (1 0 0 ml* each)

1 -6 0 CHCl^ 0*398 Negative 6 1 -1 0 0 CHCl^iMeOH (95*5) 0 .1 0 3 Negative

1 0 1 -1 0 3 CHClyMeOH (90*10) I.O38 0.93 10** If 0*375 0.84 105 H 0*359 0 .6 6, 0.84

106 M 0.?4l 0 .6 6 tf 107-115 0 .2 5 2 0.3^. 0.51, 0 .6 2 1 1 6 -1 3 8 CHClyMeOH (70*30) 11.**21 0.18, 0 .2 2 , 0.27 0.39, 0.51, 0 .6 2 ■t 139-150 0.815 0.27, 0.54, O .6 5 0.60

15 1-1 60 CHClyMeOH (50*50) 0.425 0 .2 6 , O.3 6, 0.55 ii 1 6 1 -1 6 5 O .338 0 .1 6 , 0 .2 1 , 0 .3I 0.57 166-215 MeOH 2.199 0 .1 6 , 0 .2 1 , 0.31 0.57 21 6-2 66 2# HC1 in MeOH Zi732 0*32 (tailing) Total 2 6 .1 9 6

The percentage of recovery was '8 5 .36# Fraction Weight in 100 mg. iue^ Elution pattern forFigure fraction^ B isocorydtne Doryanfne Li irteden n • 1

1.421 Fraction Number Anonoine 180

7.73

2 240

270 VJ -o 38

6 ,9 & (s, 1H), two olefintc protons as AB quartet at 7.0 6

(d, 1H,J = 7*5 Hz) and 6 .3 8 b (df 1H,J « 7*5 Hz), two methylenedioxy protons at 6 .0 3 b (s, 2H), one N-methyl group at 3,60 b (s, 3H). The mass spectrum indicated an apparent molecular ion at m/e 203 (100#), and the intense peaks at m/e 175 (7%) and 145 (15%)* The isolation of alkaloid II (Liriodenine) - This alkaloid (Rf 0.8*0 was obtained from fractions 104 and 105. It was recrystallized from chloroform to yield yellow needles (210 mg.), m.p. 282°c. The ultraviolet absorption spectrum of this compound had the follow ing A valuesi 247 mu (log € 4.04), 267 (3 .9 8), TTicljv • 310 (3»46), 413 (3*67)* The infrared spectrum in a potassium bromide pellet showed a conjugated carbonyl group at 1650 cm"* and a methylenedioxy group at 1050 cm”* and 965 cm”*, which was confirmed by a positive Labat's Test. The nuclear magnetic resonance spectrum in deuterochloroform indicated the possibility of one aromatic proton at 7*1 6 (s, 1H), six aromatic protons between 8 .5 b — 7*5 b (m, 6H) and one methylenedioxy group at 6 .3 2 b (s» 2H). 39

The isoiation of alkaloid III (Isocorvdine) - The

residue, 11.421 g, obtained from fraction II6-138 was

rechromatographed over a silicic acid column (5 .5 x 50 cm) which contained 453*6 g (1 lb.) of the adsorbent. The

column was eluted successively with 2 liters of chloro

form, 2 liters of chloroform*methanol (9 0*1 0 ), 2 liters

of chloroform1methanol (7 0 *30 ), 2 liters of chloroform* methanol (50*50) and 2 liters of methanol. The residue obtained from the eluate of chloroform*methanol (7 0*3 0 ) was shown to contain a major component (Rf O.39), but resisted all attempts of crystallization from the base. The alkaloid salt was finally crystallized from 5 per cent ethanolic hydroiodic acid solution. It was re crystallized twice from methanol to yield colorless needles (500 mg.), m.p. 225 °C, [o]*$ + 180.5°(C, 0.04 in methanol). The hydrochloride salt was similarly pre pared and was recrystallized from ethanol as colorless needles, m.p. 275-276°C. In order to obtain the free base, the hydrochloride salt, 50 mg., was partitioned between 50 ml. each of 5 percent ammonium hydroxide solution and chloroform. The alkaline solution was further extracted with an additional 25 ml. of chloro form. The combined chloroform extracts were dried with anhydrous sodium sulfate and evaporated to dryness. The residue was dissolved in methanol and cooled in a refrig erator for 2 weeks, which resulted in deposits of small plates (10 mg.), m.p. 185-186°C. The spectral measurements were made utilizing the hydroiodide salt except the infrared and the nuclear magnetic resonance spectra where the base was used. The ultraviolet spectrum had the following A valuet

21 3 6 3 267 ? mu (log € . ), (3.99). 302 (4.59)* A"*fJhanolicDloX • Na0H 280 mu (log € 4.87), 34-3 (4.?2). The infrared spectrum in a potassium bromide pellet indicated a phenolic hydroxy absorption at 3200 cm“^. The other intensive adsorptions were 1 5 9 0 , 1 5 8 0 , 14-70, 14-50, 1 3 7 0, 1 3 20, 12*1-0, 114-0, 1050 and 820 cm~^. The nuclear magnetic resonance spectrum in deuterochloroform showed the possibility of three aromatic protons at 6.81 & (s, 2H) and 6 .7 & (s, H), three methoxy groups at 3 .91 b (s, 3H),

3.90 8 (s, 3H), 3 .7 b (s, 3H), six methylene protons between 3 .2 b — 2.6 h (m, 6H), and one N-methyl group at 2.52 b (s, 3H). The circular dichroism,spectrum in methanol showed the following absorptions*

M 233 +1 2 1 ,0 0 0 * fel270 “2 2 '7^0 * (®]^iy +2 9 0 3.

The isolation of alkaloid IV (Anonaine) - The crude residue, 2 .1 9 9 g., from fraction 166-215 was dissolved in chloroform and chromatographed over a 100 g. neutral alumina column grade I (2.0 x 30 cm). The chloroform eluates contained a major component (Rf 0.57) and the

* combined fractions were evaporated to dryness. The residue was dissolved in 5 percent ethanolic hy- 41 drochloric acid solution and allowed to stand at room temperature overnight. Some large aggregate needles were deposited and were filtered, washed with ethyl acetate and cold methanol. The alkaloid was further recrystallized from methanol to yield colorless needles (85 mg), m.p.

275-2?5°C. (a) Jp -60°

3 8 6 2?1 the following AUf$?HVllclA # values* 234 mu (log £ . ), (3*»91)» 314 (3 .2 7 ). There was no shift in A under JuQJC • alkaline or acidic condition. The infrared spectrum in potassium bromide pellet indicated one N-H group at

3400 cm"1, one methylenedioxy group at 1060 and 955 cm”1, which was confirmed by a positive Labat*s test. The other intensive absorptions were 1 6 2 0, 1580 , 1 5 0 0 , 1460 and I230 cm"1. The nuclear magnetic resonance spectrum of the base in deuterochloroform showed the possibility of five aromatic protons at 8.10 b (m, 1H), 7 .5 8 to

7.2 b (m, 3H) and 6 .5 8 b (s, 1H), one methylenedioxy group shown as AB quartet at 6,10 b (d, 1H, J = 1Hz), 5*95 & (d, 1H, J = 1Hz), six methylene protons between

3.5 b — 2.7 b (m, 6H) and one N-H proton at 2.18 b (s, 1H). The circular dichroism showed the following absorptions*

H i s +25.3001 (®)24o-35.600i ^ 2 6 5 +89,400) (0 )3 0 5 *6 .700. 42

The tertiary phenolic precipitate fraction (Fraction D)

A chromatographic column (8 cm. x 80 cm.) was pack ed in chloroform with silicic acid (1814.4 g. or 4 lb.) to a height of 69 cm. as previously described. The resi due of fraction D, 23 g., was dissolved in chloroform (100 ml.) and applied to the top of column. Elution was begun using chloroform as an eluent. Thin layer chroma tography was employed utilizing Dragendorff*s reagent and phosphomolybdic acid with ammonia as detecting reagent for examination of the fractions. The results of the column chromatography are presented in Table 3 and Figure 5 . The isolation of alkaloid V (Doryphornine) - This al kaloid (R^p O.8 3) was isolated from fractions 91-1 3 5 . It was recrystallized from chloroform to yield color less needles (40 mg.), mp 215 - 217°C. The ultraviolet absorption spectrum of this com pound had the following IDaXf values: 242 mu (log € 4.64), 271 (3 .W), 282 (3-53). 292 (3-55). 325 <3-3^). 338shoulder (3 .0 5 ). A JJ^hanollc HaOH 30J nu (iog t 4 .83), jkl (l>.4).

This bathochromic shift suggested a possible phenolic function present in the molecule. There was no shift in A max under acidic condition. The infrared spectrum in a potassium bromide pellet indicated a phenolic hydroxy group at 3200 cm"* and a conjugated carbonyl group at *3

Table 3 Results of the column chromatography of the tertiary phenolic precipitate fraction (Fraction D)

Fraction Eluent Weight Alkaloid number composition (e.) (100 ml. Rr each)

1 -3 0 c h c i 3 0.281 Negative 31-90 CHCl^iMeOH (95*5) 0.621 Negative 91-110 CHClyMeOH (90*10) 1.349 O.6 3 , O .8 3 111-135 CHClyMeOH (80*20) 1.348 0 .6 3 , O .8 3 136-145 CHClyMeOH (70*30) 0.653 0, 0.17, 0.59, 0.77 146-165 CHClyMeOH (50*50) 1.492 0 , 0 .4 7 , 0 .5 2 , 0.54, 0.64, O.6 7, O.6 9, 0.77 166-170 CHCl^iMeOH (30*70) 0.323 0.64, 0 .6 9, 0.77 171-180 0.726 0.25, 0.33, 0.46, 0.49, O.5 7 , 0 .6 6, 0 .7 2

181-210 MeOH 6.806 0.25, 0.33, 0.46, 0.49, 0.57, 0.66 O .72 211-220 2% HC1 in MeOH 5.-789 0.23 (tailing) Total 19*054 The percentage of recovery was 82.8# Fraction Weight in 100 mg 4 5 3 2 6 1 iue5 Elution pattern forFigurefraction 5 D 60 90 Fraction Number T20 eHc uI H Re 5 240 150 6.806 i 5.78 it • it

9 180 . 210 ^ 5

1635 cm”1 . The other intensive absorptions were at

1580 , 1 5 2 0 , 1260, 1050 , 850 and 780 cm"1. The nuclear magnetic resonance spectrum in deuterochloroform exhibited the possibility of two aromatic pro

tons at 8 .1 0 6 (s, 1H) and 6.88 8 (s, 1H), two olefinic

protons shown as an AB quartet at 7*10 $ (d, ltf,J = 7 .5 Hz) and 6.4-0 8 (d, 1H, J *= 7 ,5 Hz), one methoxy group at 4.0 & (s, 3H) and one N-methyl group at 3 .6 & (s, 3H). The mass spectrum showed an apparent molecular ion peak at m/e 205 (10055), and the other intensive peaks at 190

(19%) and 162 (315 S), Methylation of dorvnhornine - A solution of 9 mg. of doryphornine in 20 ml. of methanol was treated with

20 ml. of ethereal diazomethane solution which was p generated from 700 mg. of Diazald (N-methyl-N-nitroso- p-toluene-sulfonamide). The solution was allowed to stand at room temperature for 4 days. Evaporation of the solvent and crystallization from chloroform gave 11 mg. of 0-methyl-doryphorninei m.p. 98-100°C. The nuclear magnetic resonance spectrum in deuterochloroform showed one additional methoxy function at 3 .9 5 . Synthesis of doryphornine - The synthetic procedure leading to doryphornine was based on those as reported by Schiff (21) on the synthesis*of thalifoline and Tomita (22) on the synthesis of corypalline with the ex ceptions of last step and modifications. Vanillin (17) was treated with sodium hydroxide to afford O-benzylvanillin (18)t which was converted to 3-methoxy-4- benzyloxy-B-nitrostyrene (1£) by treatment with nitromethane and base. Reduction of the nitrostyrene with lithium aluminum hydride in tetrahydrofuran afford ed 3-methoxy-4-benzyloxy-B-phenethylamine (20), which was converted to its N-formyl analog (21) by refluxing with ethyl formate. Cyclodehydration of the amide (2 1 ) by using the Bischler-Napieralski reaction condition yielded 6-methoxy-?-benzyloxy-3 ,4-dihydroisoquinoline (22). The reagents attempted to effect this cyclization were phosphorous oxychloride, polyphosphoric acid and phosphorous pentachloride. The latter was found to give a good yield. Treatment of the dihydroisoquinoline with methyl iodide produced the methiodide salt (23), which was subsequently oxidized to 2-methyl-6-methoxy- 7-hydroxy-l-oxo-l,2,3,4-tetrahydroisoquinoline (24) with an alkaline potassium ferricyanide solution. The desired product (25 ) was finally obtained by dehydrogenation of the tetrahydroisoquinolone with 10 percent palladium on charcoal. H3CO CHO h 3c HO BzCI NaOH BzO 17 18 19

L iA IH 4

p c i. h 3c o H £ 0 BzO 1. NHJOH n o / » ii * © cfH icffi Bl° 22 £ C H 3 23

1. K3Fe(CN)6 2-HCI

H.CO T P d /c H3C H. HO

25 24 Figure 6 The synthesis of doryphornine 48

(1) O-benzylvanillin (18) To a solution of 28.5 g* of vanillin (12) in ethanol was added a solution of 10 g. of sodium hydroxide in 25 ml. of water. Benzyl chloride, 2 5 .6 ml, was added and the mixture refluxed for 6 hours. The hot solution was decanted from a solid mass of potassium chloride and evaporated to leave an orange oil, which was poured into 50 ml. of 5 percent sodium hydroxide solution. The pro duct separated as a yellow solid after scratching the oil against the flask. It was obtained by filtration and washed the water. The compound was recrystallized from ethanol to give colorless plate, 3 6 .2 g., (80# yield), m.p. 6l-62°C| reported m.p. 60-63°C (23)1 V (chloro- meLX form) (cm"1 ) 1675 (e ® 0)i n.ra.r. (deuterochloroform) b 3*90 (s.- 3H, 0CH3 ) 5.2 (s,. 2H, 0CH2) 7-5 to 7.0 (m, 8H, ArH), 9.90 (s, 1H HC=0). (2) 3-raethoxy-4-benzyloxy-0-nitrostyrene (19)

A solution of 30 g. of O-benzylvanillin (18) in 1 1. of ethanol was cooled to 5 - 10° and 15 ml. of nitromethane was added. A solution of 12.5 g. sodium hydroxide in 250 ml. of ethanol was added dropwise with stirring while the temperature of the reaction was maintained at 5 to 10°C. Ice water was added to dissolve the resulted precipitate and the solution was poured with stirring into a solution of 150 ml. of concentrated hydrochloric acid in 250 ml. of water. The fine yellow precipitate was filtered, washed with water and dissolved ^ 9 in acetone. On cooling the acetone solution deposited crystals which were filtered to give yellow elongated prism, 30 (85# yield), m.p. 120-122°Cj reported m.p. 120-122°C (2 3 ). ^jnax ^cliXorof'or'm) (cm“*) 1630

(olefin), 1510 and 1335 n *m.r. (deuterochloroform), 6 3.9^ (s, 3H, 0 CH^), 5.2 (s, 2H, 0 CH2), 7*20 (m, 8H ArH and CH), 7.95 (d. 1H = 13 Hz, CH). (3) 3-methoxy-4’-benzyloxy-/5-phenethylamine (20) Under an atmosphere of nitrogen gas, 250 ml. of dry tetrahydrofuran and 10 g. of lithium aluminum hydride were placed in a one liter three necked round bottom flask fitted with a reflux condenser (protected from moisture) and a mechanical stirrer. A solution of 30 g. of 3- methoxy-4-benzyloxy-/3-nitrostyrene (1£) in 250 ml. of anhydrous tetrahydrofuran was added dropwise with stirring. After the addition was complete, the mixture was refluxed, with vigorous stirring, on a heating mantle for 5 hours. The mixture was cooled to room temperature and then cooled in an ice bath and to which was added successively 15 ml. of ethyl acetate, 8 ml. of water, 8 ml. of aqueous sodium hydroxide (1570 and 8 ml. of water, over a period of several hours under a continuous stream of nitrogen gas. The precipitated salts were then removed by filtra tion and washed with 500 ml. of tetrahydrofuran. The filtrate and washing were combined and concentrated to 50 leave an orange oil, 21 g., (80% yield). „ (chloro- I ll a X form) cm"'*' 3^00-3200 (NHg), n.m.r. (deuterochloroform),

I 7*3^ On, 5H. ArH), 6.75 (m, 3H, ArH), 5.08 (s, 2H, OCHg), 3-82 (s, 3«, 0CH3), 2.75 (m, CH,,), I.3 6 (s, 2H,

Mil | 2). The hydrochloride salt was prepared from ethanolic hydrochloric acid solution as colorless needle, m.p. 172 - 1740Cj reported m.p. 173“175°C (2^).

(*0 N-formyl-3-methoxy-^-benzyloxy-^J-phenethylamine (2 1 )

A quantity of 20 g. of 3--methoxyl-*f-benzyloxy-

J3 -phenethy1amine (20) and ^0 ml , of ethyl formate were refluxed using a heating mantle. The reaction was moni tored with thin layer chromatography and discontinued after 6 hours when the value had been changed (silica gel G, 2tl benzene»methanol). The mixture was evaporated to leave an oil, which was dissolved in benzene (256 ml.). The benzene solution was extracted three times with

150 ml. of 5 percent aqueous hydrochloric acid solution, and the extractive was discarded. Evaporation of the dried benzene solution yielded a brown oil, 20.5 g. (92% yield), which was sufficiently pure for further reaction. (5) 6-methoxy-7-benzyloxy-3*t’-dihydroisoquinoline hydrochloride (22) A solution of 16.5 g- of N-formyl-3-methoxy-

4-benzyloxy-/J-phenethylamine (2 1 ) in 125 ml. of chloro form was cooled in a dry ice-acetor e bath and 1 6 .5 g« of phosphorous pentachloride was added. The reaction was allowed to stand at room temperature for 24 hours and

then 50 ml. of ethanol was added dropwise with stirring to decompose excess phosphorous pentachloride. The

chloroform was evaporated, and then 50 ml. of water was added cautiously and the resulting solution was extracted twice with 200 ml. of ether to remove unreacted starting material. The aqueous layer was extracted five times with 100 ml. of chloroform, and the combined chloroform extracts were dried with anhydrous sodium sulfate and evaporated, The residual oil was crystallized from acetone to give yellow crystals,12.5 g.» (60% yield), m.p. 191-193°Cj V (chloroform) (cm"*) 1650 (C=N), lu c L X n.m.r. (deuterochloroform), b 7*3k (m, 5H, ArH), 6 .9 (m, 3H, ArH), 5.2 (s, 2H, OCHg), 4.0 (s, 3H, OCH^), 3.20

(m, 4h ). Anal. Calcd. for C ^ H ^ N O g HC1* G,65.79* H,6.21*

N.4.79* Found1 C,65*46j H,6.l6| Nf4.45.

(6) 6-methoxy-7-benzyloxy-3,4-dihydroisoquinoline methiodide (23) A quantity of 6 g. of 6-methoxy-7-benzyloxy-3,4- dihydroisoquinoline hydrochloride (22) was partitioned between 50 ml. of chloroform and 50 ml. of 5 percent sodium hydroxide solution. The alkaline solution was extracted three additional times with 50 ml. of chloro 52

form. The combined chloroform extracts were dried with anhydrous sodium sulfate and evaporated to dryness. The residue was dissolved in 200 ml. of acetone and 16 ml. of methyl iodide was added into the solution. The mixture was refrigerated overnight. The precipitate was filtered and recrystallized with acetone to yield the yellow cry stalline methiodide, 4,2 g. (51# yield), m.p. 193-195°; reported m.p. 194° (22). V (KBr) cm"1 1645 (c=N); max n.m.r. (trifluoroacetic acid), b , 7.48 (m, 5H, ArH), 7.1 (m, 3H, ArH), 5-3 (s, 2H, OCHg^ ^ (s> ^ 0CH^)f

3.8 (s, 3H, NCH3). (7) 2-methyl-6-methoxy-7-hydroxy-l-oxo-l,2,3,4- tetra-hydroisoquinoline (Thalifoline) (24) A quantity of 2 g. of potassium hydroxide and 2 g. of potassium ferricyanide were added to a suspension of 1 g. of 6-methoxy-?-benzyloxy-3,4-dihydroisoquinoline methiodide (2£) in 50 ml. of water and the mixture was refluxed for 30 minutes. The solution was allowed to cool to room temperature and then extracted twice with 100 ml. of chloroform. The extracts were combined and dried with anhydrous sodium sulfate and evaporated to dryness. The residue was dissolved in 100 ml. of 50 percent ethanolic hydrochloric acid solution and heated in the steam bath for 30 minutes. The ethanol was evaporated, 200 ml. of water was added and the aqueous 5 3

acid solution was-basified with ammonium hydroxide to

pH 9 . The alltaline solution was extracted five additional times with 100 ml. of chloroform. The combined chloro form extracts were dried with anhydrous sodium sulfate and evaporated to leave a brown residue, which was puri fied by preparative thin layer chromatography utilizing silica gel G as adsorbent and chloroform*methanol (9*1) as the developing solvent. The band with Rf value 0.84 was scraped off from the thin layer plate and was extracted with 50 ml. of chloroform. The mixture was filtered and washed with additional 50 ml. of chloroform. Evaporation of the dried combined extracts and crystallization of the compound afforded colorless plates, 32-5 mg. (3056 yield), m.p. 208-210°C, reported 209-211°C (21). V (KBr) cm"1 3700-3300 (OH), 1640 (c = 0); n.m.r. (deutero- ehloroform) 6 7.7 (s, 1H, ArH), 6 .6 (s, 1H, ArH), 3 .9

(s, 3H, OCH^), 3.1 (s, 3H, N-CH^). The infrared spectrum of the synthetic product was superimposable to that of natural thalifoline.

(8 ) 2-methyl-6-raethoxy-7-hydroxy-l-oxo-l,2-dihydro isoquinoline (Doryphornine) (25) A quantity of 200 mg. of 2-methyl-6-methoxy-7- hydroxy-l-oxo-l,2,3,4-tetrahydro-isoquinoline (24) was mixed with 100 mg. of 10 percent palladium on carbon in a glass tube and heated in a Wood's metal bath at 260°c for 3 hours. The mixture was cooled to room temperature und extracted with 100 ml. of methanol. The extract was dltered and evaporated to leave a brown residue, which was purified on an adsorption column (0.5 g. .of silicic acid, chloroform used as solvent). Crystallization of the compound from chloroform afforded colorless needles (82 mg., kofo yield), m.p. 215**217°c, The synthetic product was shown to have identical Rf value, infrared - nnd nuclear magnetic resonance spectra with the natural doryphornine. The isolation of alkaloid VI (Reticuline) - The crude residue, 6.8 06 g., from fraction 181-200 was rechromato- rraphed over 300 g. of a silicic acid column (5*5 x 30 cm) The column was eluted successively with 1 1. of chloro form, 1 1. of chloroform* methanol (9*1). 2 1. of chloro form* methanol (8*2), 2 1. of chloroform*methanol (1*1) of methanol. The residue obtained from the eluate.of chloroform*methanol (8*2), which was shown to contain t\ major component (Rf0.^9), was dissolved in a saturated ethanolic oxalic acid solution. Crystals were formed on standing at room temperature overnight. The oxalate Mult was recrystallized from ethanol to afford color less microcrystalline needles (400mg.) m.p. 157-159°C [a )o5 +101.2° (c» 1 .00 in methanol). This alkaloid ttnve a positive Gibb's test. 55

Unless otherwise specified, the spectral measure ments were made using the alkaloid base. The ultraviolet spectrum had the following A values* 200 mu (log e 5.02), 230 (4.56), 282 (4.06) * A JjeltosmoHc NaOH 3Q0 mu XucUw (log € 4.09). The infrared spectrum in chloroform,

indicated a phenolic hydroxy absorption at 3700 cm"’**'. The other intensive absorptions were at 1590, 1510, 1460,

1440, 1375, 1275, 1130 and 1030 cm"'1'. The nuclear magnetic resonance spectrum in deuterochloroform showed the possi bility of five aromatic protons between 6 .8 6 — 6 ,3 5

(to, 5 H), one phenolic hydroxy proton at 5«5 6 (s, broad,

1H) which disappeared after deuterium exchange, two methoxy groups at 3 .8 6 (s, 6H), six methylene protons between 3»° 6 2.5 & (m, 6H) and one N-methyl group at 2.45 5 (s, 3H). When the nuclear magnetic resonance spectrum was measured utilizing deuteropyridine as sol vent, the two methoxy groups appeared at 3 .8 6 (s, 3H) and 3*68 6 (s, 3H). The circular dichroism spectrum in methanol showed the following absorptions* +174,200, +43,600, The mass spectrum showed the molecular ion peak at m/e 329 (0 .5 ?S), and the other intensive peaks at m/e 328 (1#), 286 (0 .3#), 192 (100$),

177 (7055), 176 (1155), 148 {30?S), and 42 (20#). 56

Methylation of reticullne - Reticuline, 25 mg. in 5 ml. of methanol was methylated with diazomethane at room temperature for 2 days. Evaporation of the solu tion gave a brownish residue, which was purified by partitioning between 10 ml. each of chloroform and 5 percent sodium hydroxide solution. The alkaline sol ution was further extracted v/ith two additional 10 ml. of chloroform. Concentration of the combined chloro form extracts and crystallization from ether afforded colorless needles, m.p. 112-114°,(= +112.4° (c *= 0.50 in methanol). The n.m.r. spectrum in deuteo- ehloroform showed two additional methoxy groups at 6.2

(e, 3H) and 6.4 (s, 3H)

The isolation of alkaloid VII (Base A) - This base (R^ 0.44) was isolated from fraction 211-220. It was recrystallized from acetone to yield tan plates {25 mg.) m.p. 169-171°C.

The ultraviolet absorption spectrum of this com pound had the following A values* 234 mu (log E

2.60), 282 (2.92). A “^ hanolio NaCH 297 mu (log E J* 2^2). The infrared spectrum in a potassium bromide pellet showed a phenolic hydroxy group between 3400 to 2800 cm"1 and the other intensive absorptions at 1 5 1 0 , 1450, 1275 , 1225 , and 1125 cm . The nuclear magnetic resonance spectrum in trifluoroacetic acid indicated the possibility of six aromatic protons between 6 .8 5 to 6 .*f & (m, 6h), one methoxy group at 3 .6 $ (s, 3H), six methylene protons between 3*° ~ 2.5 & (m, 6H) and one N-methyl group at

1,9 & (s, 3K). This alkaloid gave a positive Gibb's test. Fraction E The residue of fraction E, 0,3 g., was chromatographed on a 15 g* slurry packed neutral alumina, grade I column (i.5 x 35 ora). Elution was initiated with chloroform, followed by chloroform with increasing amount of methanol. The fractions collected from the column were examined with thin layer chromatography. The compounds were detected with Dragendorff*s reagent or ultraviolet light. The results are presented in Table ' Table 4- Results of the column chromatography of fraction E

Fraction Eluent Weight Alakloid R number Composition (g.) Values (100 ml. each) 1-10 Chloroform 0. 04-5 Negative

11-20 ChloroformiMeOH 0 0.150 .1 5 0 0 .9 1# 0 .0 (1*1) 21-4-0 MeOH 0.035 0 .3 (tailing) Total 0.230

The percentage of recovery, was 7 6 .66#.

The isolation of alkaloid VIII (Dorvflavine) - The compound (R^ 0.91) was obtained from fraction 11 to 20. It was recrystallized from methanol to yield golden yellow plates (80 mg.), m.p. 339°C. A quantity of 30 mg. of this compound was also isolated from the com bined fraction A (I85 g.) and fraction C (117 g*)* The residues of the combined fractions were extracted exhaustively with ethyl acetate in a Soxhlet extractor. The extracts were dried with anhydrous sodium sulfate and evaporated to dryness. The crystalline product was obtained after purification of the residue, O.170 g., over a neutral alumina column (1.5 x 18 cm) and crystalli zation from methanol. This compound gave negative tests with Dragendorff's and Mayer's reagents. The indication of this compound was indeed an alkaloid was due to an elemental analysis of nitrogen (24). This was verified later by a high resolution mass spectral measurement. This alkaloid gave negative Evan's test for 0-dihydroxyphenol; negative Gibh's testj positive Millon's test; and did not react with concentrated sulfuric acid to produce red color. The ultraviolet absorption spectrum of this compound had the following valuest 207 mu (log € 4.42), TTIcLa 230 (4.59), 251 (4.45), 260 (4.4l), 277 (4.32), 290 (4.34), 320 (4.11) and 400 (3 .9 1); A ™®Jhanolic Na0H 244 mu max (log€ 4,61), 263 (4.56), 295 (4.36), 345 (4.23), *00 (4.14).

There was no shift in A max under acidic condition. The infrared spectrum in a potassium bromide pellet indicated a phenolic hydroxy group at 3200 cm-1, a possible N-H function at 3350 cm”1 and a carbonyl group at 1700 cm"1 . The other intensive absorptions are 1500, 1425, 1361, 1300, 1000 and 750 cm”1. The nuclear magnetic resonance spec trum in deuteropyridine showed five aromatic protons at 8.9 b (d, 2H), 8.15 b (3,1 H), 7.9 b (s, 1H), 7*55 6 (s, 1H), and 7.2 & (s, 1H), one methoxy group at 3.98 5 (s, 3»). The mass spectrum of this compound showed an apparent molecular ion at m/e 281 (100$) and the other intensive peaks were at m/e 266 (62$), 238 (18$), 210

(6$), 182 (18$), 140 (12$), 119 (12$), 90 (10$), 63 (6$) and 78 (6$). 60

Anal. Calcd. for C,6 8.3 3 ? H.3.94? N,4.98,

Found* 0,68.09* H.4.16* N,4.91.

Methylation of dorvflavine - A quantity of 30 mg. of doryflavine in 15 ml. of methanol was treated with

30 ml. of ethereal diazomethane which was prepared from 1.2 g. of Diazald. The solution was allowed to stand at room temperature for 3 days and concentrated to dry ness. The residue was purified over 2 g. of silica gel G column chromatography. The chloroform eluate afforded yellow needles, (25 mg.), m.p. 219-221°c. The ultraviolet spectrum in ethanol had the follow ing A ^ x ^ 01 values* 2 34 mu (log € 4.70), 248 (4.53), 261 (4.58), 282 (4.30), 292 (4.53), 322 (4.28), 400 (4.15). There was no shift in A max under alkaline or acidic solution. The infrared spectrum in chloroform showed a possible N-H group at 3460 cm"'1', a carbonyl group at 1700 cm"'1'. The nuclear magnetic resonance spectrum in deuterochloroform indicated five well resolved aromatic protons at 8 .5 & (d, 1H, J =2Hz), 7.78 b (s, 1H), 7.7 b (d, J « 8 Hz), 7.2 b (dd, 1H, J=2Hz and 8 Hz), and 7*05 6 (s, 1H), three methoxy groups at 4.12 S (s, 3H), 4.07 b (s, 3H) and 3-90 b (s, 3H). The high resolu tion mass spectrum showed the molecular ion at m/e 309.0972 (M+ 100$), which corresponds to the empirical formula:

C18Hi5 N°^ (requires m/e 309*308)* The other intensive

peaks were at m/e 3°8.0904 (2j5), 295*0838 (35S), 294.0788 (1255), 279.0530 (2?5), 26 7 .O856 (1 fo), 266.0804

(5#)» 265*0706 (295), 263.0555 W ° ) f 251.0588 (6?5),

■250.0510 (3?o), 238.0879 ( W , 235.0621 (295). 223.0634 (35S). 180.0446 (255), 15^-5522 (35S)t 43.9886 (2#), 18.0106 (4355). Anal. Calcd. for C^H^NO^t 0,69.89; H,4.89.

Found: 0 ,6 9.6 3 * H4.96.

Acetvlation of dorvflavine - A quantity of 20 mg. of doryflavine was dissolved in 1 ml, of pyridine and 1 ml. of acetic anhydride was added. After 2 days at room temperature, the crystals deposited in the flask were removed by filtration and washed with water. The compound was recrystallized twice with methanol and ethanol mixture to afford colorless needle, 21 mg,, m.p. 218-219°C. The ultraviolet spectrum of the acetate had the following .values: 222 mu (log € 4.46), 243 (4.49),

280 (4.31), 320 (3.71), 333 (3*81), and 38O (3 -6 1)* A ^ a n o l i c N*0H 252 mu (log e 5 .00). 277 (*,97). 287

(4.91). 320 (4.61), 400 (4.42). The infrared spectrum in potassium bromide pellet indicated disappearance of all N-H and phenolic OH absorptions, and three additional 62 carbonyl absorptions at 1770, 1?60 and 1730 cm”1 . The nuclear magnetic resonance spectrum in deuterochloroform showed three additional methyl groups at 2.8 b (s, 3H)(

2.48 S (s, 3H) and 2.39 b (s, 3*0. The mass spectrum showed the molecular ion at m/e 407 (M+, 295 $) and the other intensive peaks at m/e 365 (34?S) ( 323 (**9%) t 309 (6%)t 281 (1005$), 266 (205 $) and 43 (53 ^).

Anal. Calcd. for C22H17N07 * 0,64.86* H #4.21, Pound* C,64.721 H.4.20.

Acetvlation of doryflavine dimethylether - A quantity of 20 mg. of doryflavine dimethylether was acetylated with 1 ml. of pyridine and 1 ml. of acetic anhydride and was worked up as previously described. The acetate was crystallized from chloroform to give yellow needles (1 5 mg), m.p. 215-216°C. The infrared spectrum showed no absorption between

3300 to 3000 cm" 1 region, but observed a carbonyl absorp tion at 1710 cm"1 . The n.m.r. spectrum in deuterochloro form showed three methoxy groups at 4.18 b (s, 3H), 4.1 b (s, 3H) and 4.0 b (s, 3H). In addition, a methyl group of the acetate at 2 .8 b £s» 3*0 was also observed. The tertiary non-phenolic alkaloid fraction (Fraction F) The residue of fraction F, 28 g., was dissolved in 120 ml, of chloroform and chromatographed on a chloro form slurry-packed column (7 cm, x 81 xm.) containing 1814.4 g. (or 4 lb) of silicic acid as adsorbent. Elution was initiated with chloroform, followed by chloroform con taining increasing amount of methanol. Consecutive frac tions were collected and combined according to the date of thin layer chromatographic analyses. The chromatographic results are given in Table 5 and Figure 7. The isolation of dor.vanine. liriodenine. isocorvdine and anonaine - A quantity of 120 mg. of doryanine, 250 mg. of liriodenine, 150 mg. of isocorydine hydrochloride and 45 mg. of anonaine hydrochloride were isolated from

fraction 1-40, fraction 41-60, fraction 61 -65 and fraction 71-95 respectively. The compounds were iden tified as previously described in the tertiary non- phenolic precipitate fraction. Table 5 Results of the column chromatography of the tertiary non-phenolic alkaloid fraction (Fraction F)

Fraction Eluent Weight Alkaloid number composition (g.) Rf (100 ml. each)

1-40 chci3 4.715 0.93 41-60 CHClyMeOH (95*5) 0.549 0.69 61-65 CHClyMeOH (9 0*10) 4.641 0.45, 0.56 0.65 6 6 -7 0 CHClyMeOH (80*20) 2.043 0.30, 0.45 0.56, 0.67 71-95 CHClyMeOH (50:50) 0 .9 2 6 0.30, 0.45 O .6 3 M 96-115 7.025 0.42, 0 .6 1 116-125 MeOH O.3O6 0.29, O .38 0.59. 0.69 126-145 2# HC1 in MeOH 4.112 0.31, 0.50 Tatol 23.917 The percentage of recovery was 8 5 .39# Fraction Weight in 100 rag 5 4 3 2 6 1 iue7 Elution patternFigure forfraction 7 F 20 Dorya nlne 40 0.7 a.* 60 Fraction Number 80 203 a. •2.043 100

120 140 160 O' The tertiary phenolic alkaloid fraction (Fraction G) The residue of fraction G, 10 g., was dissolved in

50 ml. of chloroform and chromatographed on a silicic acid column (5*5 x 5° cm) which contained 453.6 g. (1 lb.) of the adsorbent. Elution was initiated with chloroform, followed by chloroform containing an increasing amount of methanol. Consecutive fractions were collected and com bined according to the data of the thin layer chromato graphic analyses. The chromatographic results are pre sented in Table 6 and Figure 8. The isolation of alkaloid VI (reticuline) - This alkaloid was isolated from fraction 51-55 as its oxalate

salt, (550 mg.), m.p. 15?-159°C. The compound was ident ified as previously described. The isolation of alkaloid IX (Base B) - The residue, 4.646 g., obtained from fraction 56-75 was rechromatograph ed over 200 g. of silicic acid column (2.8 x 60 cm). The chloroform eluate was shown to contain one major spot at Rf 0*55* It was crystallized from chloroform to yield grayish needles, (35 mg), m.p. 201-203 -1 5 .6°

(c, 0 .2 5 in methanol). The ultraviolet spectrum of this alkaloid had the following A values* 205 mu (log c 3*^7). 225 (3.18), 281 (2,30)i \ methanolic NaOH mu Table 6 Results of the column chromatography of the tertiary phenolic alkaloid fraction (Fraction G)

Fraction Eluent Weight Alkaloid number composition (g.) Rf (10 0 ml* each)

1-4-5 CHCl^ 0.745 o*55. 0.94 4-6-50 CHClyMeOH (9 0*1 0 ) 0.281 0 .2 0 . 0 .3 1 0.42, 0 .5 2 0*59 51-55 CHClyMeOH (70*30) 0.546 0 .2 0 , 0 .2 6 0*36, 0.52 56-75 CHClyMeOH (50*50) 4.646 0*15. 0.25, 0 .36, 0.55 76-90 CHClyMeOH (30*70) 0.425 0 .1 0 , 0 .2 6 0.41, 0.55 91-100 MeOH O .332 0 .1 0 , 0 .2 6 0.46 10 1-1 20 2% HC1 in MeOH 1.321 Negative Total 8.297 The percentage of recovery was 82,9?# Fraction Weight in 100 mg. 5 6 4 3 2 1 iue8 Elution patternfor fractionFigure8 G 20 Re♦i cvI in 0140 40 Fraction Number Base B 60 80

100 120 o\ a (log € 3*18)i 297 (2 .3 0 ). There was no shift under an acidic condition. The infrared spectrum in potassium bromide pellet indicate the phenolic hydroxy function between 3500 to 3100 cm"*. The other intensive absorp

tions were at 1 6 1 0 , 1 5 8 0 , 1**5 0 , 1 2 50 , 1100 and 1050 cm-1. The compound was not readily soluble in chloroform, therefore the nuclear magnetic resonance spectrum was determined in trifluoroacetic acid. The spectrum showed

the possibility of six aromatic protons between 6 ,9 b to 6.2 b (m, 6H), two methoxy groups at 3.55 & (s, 6H) and

six methylene protons between 3 .5 & — 2.5 b (*n, 6H). The mass spectrum showed a very low abundant molecular ion at m/e 299 (M+ 1%)i the parent peak at m/e 178 (100f) and the other intensive peaks at m/e 298 (2f), 283 (8#),

28^ (99^), 270 (1/a), I63 (20/), 108 (15 ?S), 77 (19^) • 28 (70f) and 18 (70f). Attempted methvlation of Base B - A quantity of 20 mg. of base B in 20 ml. of methanol was treated with 50 ml. of ethereal diazome thane solution which was generated from 1.5 g- of Diazald. The mixture was allowed to stand at 0°C, for 3 days and followed by evaporation of the solvent afforded 25 mg. of brownish residue, which was examined with t.l.c. using chloroform*methanol (9*1) as developing solvent and silica gel G as adsorbent. The chromatogram was sprayed with Dragendorff*s reagent and indicated four spots at Rf 0.06, 0.17, O .7 6 and 0.81. ?0

An attempt to separate these four compounds in a 1 g. silica gel G column was unsuccessful. N-methvlation of Base B - A quantity of 50 mg. of base B was dissolved in a solution containing 1 ml. of 37 percent formaldehyde and 50 ml. of methanol. After stirring the mixture at room temperature for 1 hr., 200 mg. of sodium borohydride was added portionwise. The mixture was further stirred for 1 hr. and evaporated to dryness. The residue was partitioned between 25 ml. of 5 percent ammonium hydroxide solution and 25 ml. of chloroform. The alkaline solution was extracted two additional times with 25 ml. of chloroform. The com bined and dried chloroform extracts were evaporated to afford a yellow oil. The infrared spectrum of this compound in chloroform indicated phenolic hydroxy absorp tion at 35^0 cm"1 . The n.m.r. spectrum in trifluoroacetic acid showed a N-methyl absorption at 2.7 h (s, 3H). The isolation of alkaloid X (corypalline) - This alkaloid (R^, 0.26) was obtained from fraction 91-100 as light tan needles from chloroform. It was recrystall ized twice from chloroform solution to give colorless needle (152 mg.), m.p. 167-168°C. The ultraviolet absorption spectrum of this com

pound had the following A IllcUw valuesi 202 mu (log € 4.43), 225 (3.68), 285 (3.56)* A U ^ hanolic Na0H 245 mu (

logO»97), 293 (3*9 2). There was no shift in A IDcUC_QV under acidic condition. The infrared spectrum in a potassium bromide pellet indicated the phenolic hydroxy absorption between 3200 to 2400 cm"^. The other inten

sive absorptions were 1600, 153°. 1450, 1430, 1370, 1250, 1210 and 1120 cm"'*'. The nuclear magnetic resonance spectrum in deuterochloroform showed the possibility of two aromatic protons at 6.6 b (s, 2H), one methoxy group

at 3 .5 b (s, 3H). six methylene protons between 3 .5-2.5 £

(m, 6H), one N-methyl group at 2.45 b (s, 3H). The mass spectrum indicated an apparent molecular ion at m/e 193 (30$)1 and the intensive peaks at m/e 192 (35 $),

177 (10$), 150 (60$), 135 (5$). 128 (85$) and 18 (100$). Synthesis of corypalline - 6-methoxy-7-benzyloxy-3,

4-dihydroisoquinoline methodide (23). 100 mg., an inter mediate for the synthesis of dorphornine was dissolved % in 50 ml. of methanol and while stirring 200 mg. of sodium borohydride was added portionwise. After two hours, the solution was evaporated to dryness. The

residue was partitioned between 25 ml. of 5 percent

ammonium hydroxide solution and 25 ml. of chloroform. The alkaline solution was extracted two additional times with 25 ml. of chloroform. The combined chloro form extracts were dried with anhydrous sodium sulfate and evaporated to dryness. The compound was crystall

ized from chloroform solution as colorless needles 70 mg., m.p. 167-168°C. The synthetic product was shown to have identical R^. value, infrared and n.m.r. spectra with the natural corypalline. 73

Chromatographic separation and isolation of the total quaternary alkaloid fraction (Fraction H). A quantity of 10 g. of the quaternary alkaloid fraction was dissolved in methanol and adsorbed on 20 g. of acid alumina. The mixture was allowed to air dry in a large evaporating dish. The sample which was mixed with chloro form to form a slurry, was applied to a column (5 x 50 cm) packed with *f00 g. of acid alumina, Grade I, to a height of 25 cm. Elution was initiated with chloroform, followed by chloroform containing increasing amount of methanol. Consecutive fractions were collected and combined according to the data of the thin layer chromatographic analyses. The results are given in Table 7 and Figure 9. The isolation of alkaloid XI fcholine chloride) - The residue of the combined fraction 66-85 was dissolved in absolute ethanol and decolorized with charcoal. On addition of ether and refrigeration overnight, light tan needles were deposited. It was recrystallized two add itional times with ethanol and ether to yield colorless needles (850 mg.), m.p. 288-290°C (Rf 0.17). The compound was very hygroscopic and gave a wine- red color on the silica gel G thin layer plate when sprayed with Dragendorff1 s reagent. The infrared spec trum in a potassium bromide pellet indicated the hydroxy function between 3500 -3IOO cm”\ The other intensive Table 7 Results of the column chromatography of the total i , quaternary alkaloids (Fraction H)

Fraction Elution Weight Alkaloid number composition (g.) Rf (100 ml. each)

1-2 0 Benzene 0.025 Negative 21-40 Benzene*MeOH (90*10) 0.477 0 .20, 0.46

41-65 Benzene*MeOH (70*30) 2.441 0.28, 0 .3 3 0.42, 0 .5 3

66-85 CHClyMeOH (70*30) 2.826 0.17, O .33 0.42, 0.53 86-95 CHClyMeOH (50*50) 0,851 0.42, 0 .5 3 96-125 ' MeOH 1.053. 0.29, 0.50 0.51 Total 9.973 The percentage of recovery was 99-73^ Fraction Weight in 100 mg. 4 6 3 • 5 1 2 Figure 20 9 Elution pattern forfraction H no n f i o h C r o idi l h c Fraction Number

100 140 V a absorptions were 1480, 1350, 1100 and 950 cm"1 . The nuclear magnetic resonance spectrum in trifluoroacetic acid showed the possibility of three methyl groups at 3*35 h (s, 9H), two methylene protons at 3*75 $ (m, 2H) and the other two methylene protons at 4.32 £ (m, 2H). 77

Antimicrobial testing of the isolated compounds An extensive effort was initiated in 1970 to uncover new clinically useful antibiotics from higher plants in this laboratory by Professors Lester A. Mitscher and Jack L. Beal. Consequently the isolated alkaloids from Doryphora sassafras Endlicher were also examined. The method of the testing was agar dilution-streak technique and was performed by Mr. Ruey-ping Leu. The procedure has already been detailed by Mitscher et al (25). Totally six microorganisms were used for screening, namely: 1. Staphylococcus aureus. Smith strain (ATCC no. 13709); 2. Escherichia coli (ATCC no. 9637)1

3 . Salmonella gall inarum (ATCC no. 918*0; **■• Klebsiella pneumoniae AD (ATCC no. 10031)j 5* Mycobacterium smegmatis 607B (ATCC no. 607)I 6 . Candida albicans (ATCC no. 10231). Of the eleven alkaloids isolated, all were tested except doryphornine, which was not examined because of scarcity of sample. Unless otherwise specified, the concentration of the test compounds was carried out at the 100 ug/ml level. The results of this examination are presented in Table 8 . Table 8 Results of the antimicrobial testing

Microorganisms used Compound tested ia 2 3 * 5 6

Doryanine ♦b + + + +

Liriodenine - + - - - Isocorydine HC1 + + + + ♦ + Anonaine HC1 - + «*• - - - Reticuline oxalate + + + + + + _c Doryflavine + + + + + Corypalline + + + + + + Choline chloride + + + + + + Base A + + + + + Base B + + '+ + ♦ ♦ a. The number refers to the microorganism as described in the test. b. - indicates complete inhibition of growthj + indicates no inhibition. c. The minimum antimicrobial concentration was 50 ug/ml. DISCUSSION

Of the eleven alkaloids isolated from the bark of Doryphora sassafras Endlicher three are aporphine bases being the known, compounds liriodenine, anonaine, and isocorydinej three possess the isoquinoline nucleus and are doryanine, corypalline and a new alkaloid assigned the name doryphornine* three proved to be benzylisoquinoline alkaloids and are reticuline, base A and base B. The ubiquitous choline was also identified. The struc ture of an additional weak base assigned the name dory flavine was proposed. The evidences leading to the assignment or postula tion of the structure for an isolated compound will be discussed below. Alkaloid I (Dorvanine) - This compound was found to have superimposable infrared spectrum with an authen tic sample of doryanine, which was previously isolated from the leaves of this plant (8). Thin layer chromato graphic analysis showed the compound had an identical Rf (0.93) with doryanine.

1 80

Moreover, the mixture melting point did not show any

depression. The fragmentation patterns of this compound

in the mass spectrum also fit the assigned structure.

- e -CO H, t o o I CH3 14 m/e 17 5 m/e 2 0 3

-C H 20

C o I c h 3 m/e 14 5 81

Alkaloid II (Liriodenine) - This alkaloid was ident ified as liriodenine (£) on the basis of superimposable infrared spectrum, thin layer chromatographic analysis and mixture melting point observation. An aromatic pro ton which absorbed at ?.l & in n.m.r. spectrum was as

signed the proton at C-3 . This is due to the comparison of the-n.m.r. spectrum of liriodinine and atherospermi- dine (26). The later showed a methoxy absorption at

4.55 5 and disappearance of absorption at 7*1 6 (2 6) R

9 9. R = H 2 6. R = CH3

Alkaloid III (Isocorvdine) - The ultraviolet spectrum KeOH of this compound exhibited A max 2 1 7, 267 and 303 mu (log £ 3 *6 3, 3.99 and 4.59) characteristic of a 1 ,2 ,9 , 10-tetrasubstituted aporphine system. The bathochromic shift and the infrared spectrum indicated a phenolic hydroxy function. The n.m.r. spectrum showed three methoxy groups and one N-methyl group at 3 .91 $,3 .9 0 & , 3.70 & and 2,52 & . The compound was identified as isocorydine hydrochloride on the basis of superimposable infrared spectrum, thin layer chromatography analyses and mixed melting point measurement with an authentic sample. It has been pointed out (27) based on X-ray analy sis that the biphenyl ring of the aporphine alkaloids is appreciably twisted. The alkaloids can exist in the absolute configuration of R (or -) series or its mirror image S (or +) series. Further it was determined (28) that in the 235 to 2^5 niu region, the Cotton effect is independent of the substitution. Consequently, this can be of diagnostic significance. Aporphine alkaloids which belong to the S-configuration exhibit a positive Cotton effect and a negative Cotton effect is observed in the R-configuration in the region 235 to 2^5 mu. This alkaloid showed a positive Cotton-effect in circular dichroism spectrum at 233 mu therefore it belongs to S-series, as structure (2£) shown below. Alkaloid IV (Anonaine) - The ultraviolet spectrum of this compound exhibited A 234, 271 and 31^ mu in o a (log € 3 *8 6, 3.91 and 3 .2 7) characteristic of a 1 ,2-di- substituted aporphine system. The methylenedioxy group of this compound was indicated by a positive Labat's test, absorptions at 1060 and 955 cm-1 in infrared spectrum and 6 .1 6 and 5*95 $ (J =1 Hz) in n.m.r. spec trum. The compound showed a negative cotton effect at mu in the circular dichroism spectrum.

i Prom this data, this alkaloid was identified as anonaine hydrochloride (28.) and was confirmed by compar ison ( ir., tic and mixed m.p.) with an authentic sample.

Alkaloid V (Dorynhornine) - The ultraviolet spectrum of this compound showed a great similarity to that of doryanine. The infrared spectrum indicated the possi bility of a six membered cyclic lactam by the charact eristic absorption at 3002 cm"1 and I635 cm"1. The al kaloid did not show a methylenedioxy group in infrared or n.m.r. spectrum, but indicated the presence of one

methoxy and one phenolic functions. The phenolic func

tion was confirmed by a successful methylation of dory-

phornine with diazomethane. Coordination of.this in

formation led to the postulation of six possible

structures (a to f) for doryphornine.

a. R,=CH3 , n,= H e- R|=CH„B,= H e. R,-CH3,R2-H

b. R|=H,R2=CH3 d. r,= h, r 4=ch3 1. r,-h , r 2 = ch 3

A further consideration of the co-occurrence of

doryanine and corypalline in the bark of the plant sug

gested the structure (a) as the most favorable candidate,

but a choice could not be made on the information obtained

thus far. Accordingly, a synthesis of doryphornine was

carried out. The synthetic compound, as described in

the experimental section, exhibited superimposable in

frared spectrum, identical R^ as that of doryphornine.

Also the mixture melting point did not show depression. The n.m.r* spectrum can also be explained with the proposed structure (2£), The absorption of the N-methyl group at 3*6 5 was shifted downfield by the neighboring carbonyl group and the double bond. The AB type quartet at 7*1 b and 6.4 6 ( J = 7*5 Hz) indicated the presence of the protons at double bond. The Cg aromatic proton was shifted downfield to 8.1 6 because of the deshielding effect of the peri carbonyl group. The fragmentation of doryphornine in the mass spectrum also was in agreement with the established structure.

s 4

25 m/e 20 5

CH3 m/e 19 0 m/e 16 2 86

Alkaloid VI (retlculine) - The compound showed a characteristic benzylisoquinoline ultraviolet absorption

spectrum with A 2 0 0 , 230 and 282 mu (log e 5 .0 2 , ITlELX 4,56 and 4.06). The bathochromic shift and the absorp tion at 3700 cm"'*' in the infrared spectrum indicated a phenolic group. This alkaloid showed two methoxy groups as a singlet, intergrated for six protons at 3 .8 5 in the n.m.r. spectrum when deutrochloroform was used as the solvent. However, two methoxy groups were clearly separated into two singlets at 3*8 & and 3 .6 8 & when deuteropyridine was used as the solvent. The two phen olic hydroxy groups were manifested by a conversion of this alkaloid (2£) to laudanosine (3 0 ) (2 9). From the above data, this alkaloid was identified as reticuline and was confirmed by comparison of the infrared and t.l.c. with an authentic sample. The stereochemistry of the benzylisoquinoline alka loid had been studied (30) and resulted in the follow ing conclusioni the rotatory dispersion curves of the S-series showed three positive Cotton effects in the 200-320 nm region while those of the enantiomeric R- series showed three negative Cotton effects in the same region. 87

This alkaloid showed the positive Cotton effects at 232 mu and 290 mu# Therefore, it belongs to the S- series. The mass spectrum of this compound also agreed with the fragmentation pattern. H-CO

OR 29 R = H 3 0 R = CH.

_ e . H ^ G O ^ C f 29 m /e 3 2 9

-e*/cHj=N-CH, m /e19 2 j-CH3 H.CO

-Stocaf'CH. o m /e 177 I'"' m/ e 2 8 6

X £ k m / e 176 m /e 14 8 CHa 88

Alkaloid VII (Base A) - The ultraviolet spectrum of this compound was characteristic of a benaylisoquinoline alkaloid. The bathochromic shift and infrared spectrum indicated a phenolic hydroxy group. The compound con tains one methoxy and one N-methyl group as shown in the nuclear magnetic resonance spectrum. But the scarcity of the sample prevented a further investigation. Alkaloid VIII (Dorvflavine) - This alkaloid is gol den yellow in color and shows greenish fluorescence under ultraviolet light when it is dissolved in methanol sol ution. It gave a negative test with Dragendorff1s and Mayer's reagents. However, the compound gave a positive qualitative test for nitrogen and showed a molecular ion in the mass spectrum at m/e 281, corresponding to the molecular formula* ^16^11^ 1*• ,J^e elemental an alyses of C, H and N were in good agreement with the calculated values. The ultraviolet spectrum showed rather complex absorptions and the structural skelton could not be readily elucidated. The infrared spectrum showed evidence of a phenolic hydroxy group at 3200 cm”1 and a possible N-H absorption at 3350 cm”'*'. The absorption at 1700 cm”1 indicated a rather unusual conjugated carbonyl group. The n.m.r. spectrum showed one methoxy group, and five unresolved aromatic protons. 89

The mass spectrum showed the intense peak at m/e 266

(62#), which was due to the loss of a methyl group from the* molecular ion* further successive elimination of three 28 mass units of C = 0 group resulted in forma tion of the fragmentation peaks at m/e 238 .(18#), 210 (6#) and 182 (18#)» which were evidenced by observing the metastable peaks at m* 213, 185*7 and 158 (calculat ed m* 212.9, 185.2 and 157.7).

Doryflavine can be acetylated to a triacetate derivative and methylated to a dimethoxy derivative. This difference indicated one 0-H or N-H was relatively Inert to diazomethane compared to the other two groups. A search of the literature was carried out to find a compound which had similar properties as described. The liriodenine type of compound was ruled out on the basis of the molecular formula* liriodenine having

doryflavine Also, liriodenine gave a reddish color with concentrated sulfuric acid while doryphornine did not (31). In consideration of the spectroscopic and analy tical data, the compound has the following functional groups 1 1 OCH^, 2 OH, 5 Ar-H, 1 possible N-H, 1 C = 0. This accounts for all the oxygen and nitrogen in the molecule. The coupling pattern of the aromatic protons in doryflavine dimethylether was very informative and could be assigned; The double bond equivalent of doryflavine is 12 (for C H. (0 V N, D.B.E. * .Ua+2 ) , when a D C e C16Hh N04 d ‘b »E. = ~ ^ - 12). Evidently this compound is a highly conjugated system. Examinations of the ultraviolet spectra of dory flavine and other possible known structures revealed that doryflavine bears a close resemblance in absorption be havior to a group of compounds which have a phenanthrene system. To date, all of these types of compounds were iso lated from the Aristolochiaceae family (32-34). Aristo- lolactam (21) initially derived from the catalytic reduc tion of aristolochlc acid (22), aristo-red (22) and taliscanine (j4) are typical examples. Comparisons of the available ultraviolet, infrared and nuclear magnetic spectra of this type of compounds to doryflavine are presented in Table 9 and Figure 10. €6 *c HO O HOO Q H30 Z

OO H Q

1C ZZ HOO Q eHOO n OV O H /U Z Q Q HOOO

16 Table 9 Comparison of uv, ir and n.m.r. absorption spectra of doryflavine and related compounds.

Spectra Compounds measured Aristolo- Aristo- red Taliscanine Doryflavine Doryflavine 1actam (33) W dimethylether (32,33) uv 222 (22,900) 207 (27,000) A <« > max. 242 (30,840) 244 (38,900) 230 (39.000) 234 (50 ,000)

250 (29.740) 253 (42,400) 255 (35.^90) 251 (28,000) 260 (36,130) 265 (31,500) 260 (26,000) 261 (38,000)

291 (15.050) 294 (19.350) 299 (16,220) 277 (2 1 .000) 282 (30,000)

301 (15.^50) 300 (1 9.100) 290 (2 2 ,000) 292 (3M0 0 )

327 (9,220) 305 (18,800) 320 (1 3 .000) 322 (18,000) 346 (7,190) 335 (5.850)inf,

395 (8,470) 352 (5»000)inf. 400 (21,000) 400 (14,000) 395 (8,200) Table 9 (Continued)

Spectra Compound measured Aristolo- Aristo-red Taliscanine Doryflavine Doryflavine lactam (33) (3*0 dimethylether (32.33) ir (cm-1) N-H 3000-3300a 3000-3300a 3^50 3350a 3^60 c*o 1691 170^ 1690 1700 1700 n.m.r.( & ) b8.62 (Cjk-H, c8.5 (C4-H. d. J=8 Hz) d, J=2Hz) 6.86 (C2-H, 7.7 (C.-H. d, J=8 Hz) d, J=8Hz) 7.53 (C7-H. 7.2 (C--H, d. s) 7 • of d., j«8 & 2 Hz) k 12s) (Cl0' 7.78 (Cp-H.s) 7.33 (Cw-H. 7.0-5* (C10-H.s) t. J»8 Hz) a absent in its acetate derivative, b solvent used was not indicated, c measured in CDCl^ •t- \o o—o-o»

o -o —

d imethyiethe d Doryflavine Aristololaclam A r i s to — rod— i rod— — to s i r A MJl . V . \ \ a

a A V ' - ' and related compounds. related and

\ l ' * \ Figure 10 doryflavine 10 of Figure spectra Ultraviolet I I t i / >v\ \\ \\ / \ >v\ * i Doryflavine — i * U'jL 1 y / \ y \ / A . y. \ '-"i\ 200 200 250 300 350 400 450 ■ ■ i . i . x v\> ' 3 3

- 0 - 1 X Acetylation of doryflavine afforded its triacetate; acetylation of doryflavine dimethylether produced its monoacetate. This behavior also observed in aristolo- lactam and aristo-red, was attributed to the relative inertness of the N-H group of the lactam. Talcing into account the co-occurrence of reticuline, corypalline, doryphornine and negative Evan's test for O-dihydroxy phenol, negative Gibb's test and positive Millon's test, structure of doryflavine therefore, is tentatively proposed as (35). However, a definitive proof for this structure will require more extensive study and the availability of a greater quantity of doryflavine. 96

The fragmentation behavior of doryflavine can be explained accordingly:

H.CO

- e * c h 3.

m /e 2 8 1 m/ e 2 6 6

m =2 1 3 -CO

-CO (C)3H8N0 2f- (c 14h ,n o 3)+ m* = 1 8 5.7 m / e 210 m/e 2 36

-CO m*= 15 8

fci2H8N o) m/ e 1 8 2 Alkaloid IX (Base B) - The ultraviolet spectrum in dicated this alkaloid was of the benzylisoquinoline type. The infrared spectrum and the bathochromic shift demon strated the presence of a phenolic hydroxy group. The nuclear magnetic resonance spectrum showed six aromatic protons* but the coupling system could not be resolved. Two methoxy groups appeared as a singlet at 3.55 $ • The mass spectrum showed a low abundant molecular ion at m/e 299, the parent peak at m/e 1 7 8} the other inten sive peaks at m/e I63 (15#)» 108 (15#)• These evidences, coordinated with the biogenetic consideration led to a speculation of a possible structure (^6 ). Attempts to methylate this base with diazomethane in ether resulted in decomposition of the compounds and the products could not be characterized. N-methylation of base B with formaldehyde and sodium borohydride af forded a yellow oil, which in chloroform showed a rather similar, but not identical spectrum with l-(p-methoxybenzyl)-2-raethyl-6-methoxy-l,2 ,3,4-tetrahydro-7-isoquinolinol (32)» a degraded product from O-methyloxy- acanthine and was kindly furnished by Mr. Wu-nan Wu of this laboratory. T.l.c. analyses showed these two com pounds to have the same Rf 0 .7 5 in benzenet acetone* ammonium hydroxide (30130«0 .5)1 0.45 in n-butanol* acetic acid*water (4*1*1). 98

A further investigation of this compound was dis continued because of the inadequate quantity of the sample available.

HXO Q H„CO e

Q 1 OCH. m / e 17 8 -c h3*

m /e 299

m/e 16 3 99

Alkaloid X (corypalline) - This alkaloid showed ul- MeOH traviolet absorptions at A max 202, 225 and 285 mu (log € ^.^3, 3*68 and 3*58) characteristic of an iso-

* quinoline system. The phenolic hydroxy function was indicated by the bathochromic shift and absorption at 3200 cm“^ in the infrared spectrum. The n.m.r. spectrum showed the presence of one methoxy group and one N-methyl group. The compound was identified as corypalline (38) on the basis of superimposable infrared spectrum, analyses of t.l.c. and mixed m.p. with an authentic sample. A synthesis of this compound from 6-methoxy-7-benzyloxy- 3,4— dihydroisoquinoline methiodide confirmed this assign ment , The fragmentation of this alkaloid in mass spec trum can be explained as follows*

+ •

m/e 177 3 100

Alkaloid XI (choline chloride) - This compound was identified as choline chloride (16,) on the basis of the spectral data (ir. and n.m.r.) and identical with an authentic sample.

CH3 Cl©

H3C *"Y“ CH2CH 2OH CHg 1 6 SUMMARY

A phytochemical investigation of the alkaloids of the hark of Dorvphora sassfras Endlicher (Monimiaceae) was' conducted. A literature survey covering the botanical, chemical and pharmacological significance was presented. Preliminary investigations including a quantitative estimation of the alkaloid content and a pilot study of the extraction procedure were made before a large scale extraction of the bark of this plant was carried out.

A quantity of 1 9A kg of the ground bark was defatted and extracted with ethanol to give 1.6 kg of residue, which represented 8.22% of the total plant material. The alkaloid material was obtained from the residue and divided into tertiary phenolic, tertiary non phenol ic and quaternary fractions by conventional meth ods. The alkaloids doryanine, liriodenine, (+)-isocorydine and (-)anonaine were isolated from the tertiary non-phenolic fractions. The identifications of these four compounds were confirmed by direct comparisons ( ir. spectra, and mixture in.p.) with authentic samples.

101 102 Of the five alkaloids isolated from the tertiary phenolic fraction three were benzyl-tetrahydroisoquinoline alkaloids, namely (t)-reticuline, base A and base B. The latter two alkaloids were not studied further because of scarcity of the samples. The fourth alka loid was a new base assigned the name doryphornine, which was characterized as 2-methyl-6-methoxy-7-hydroxy- l-oxo-l,2-dihydroisoquinoline on the basis of spectral data and a total synthesis. The fifth alkaloidt corypalline was identified on the basis of spectral data and a total synthesis. A new weak base was isolated from fraction E and assigned the name doryflavine, which refers to its genus name and the gold yellow color of the crystal. The structure of this alkaloid is suggested as 9-amino-3,5- dihydroxy-6-methoxy-8-phenanthroic lactam on the basis of spectral data, chemical transformations, elemental analyses, color tests and biogenetic consideration. The ubiquitous choline was isolated from the quaternary alkaloid fraction. Antimicrobial testing was performed and indicated liriodenine and anonaine hydrochloride to be active at the 100 ug/ml level against Staphylococcus aureus. Smith strain (ATCC no. 13704), Klebsiella pneumoniae AD (ATCC no. 10031), Mycobacterium smegmatis 607B (ATCC no. 607) and Candida albicans

(ATCC no. 10231)i doryflavine to be active at the 50 ug/ml level against Mycobacterium smegmatis AD (ATCC no. 607), REFERENCES

1. Hutchinson, J., "The Families Of Flowering Plants" Macmillan and Co., London, p. 89 (1959). 2. Hegnauer, R., "Chemotaxonomie Der Pflanzen" Birkhauser Verlag Basel Und Stuttgart, p. 99 (1969).

3 . Bailey, F. M., "The Queensland Flora" H. J. Dlddams and Co., Brisbane, p. 1287 and p. 1295 (1827-1902). 4. Petrie, J. M., "Chemistry of Doryphora sassafras." Proc. Linn. Soc. N. S. Wales. 37. 139-155 (1912) 5. Raffauf, R. F., "A Handbook Of Alkaloids And Alkaloid- Containing Plants" Wiley-Interscience, New York (1970). 6. Mollov, N. M. and Dutschewska, H. B., "Thalactamine, l-oxo-2-methyl-5,6,7-trimethoxy-l,2-dihydroisoquinoline in a Thalictrum minus Variety," Tetrahedron lrtter. 24,““l951-52 (19^9). 7. Penfold, A. R., "Essential oil of the leaves of Doryphora sassafras." Proc. Linn. Soc. N. S. Wales. 22, 136-157 (1912). 8. Gharbo, S. A., Beal, J. L., Schlessinger, R. H., Cava, M. P. and Svoboda, G. H., "A phytochemical study of Doryphora sassafras I. Isolation of eight crystalline alkaloids from the leaves." Llovdia. 28. 237-244 (1965) “ 9. Buck, K. T., "The syntheses of Doryphora sassafras alkaloids and the preparation and oxidation of N-acylbenzyltetrahydroisoquinolines," Ph.D. Dissertation. The Ohio State University, Columbus, Ohio (1966) 10. Tomita, M, and Santomi,’ M., "Synthesis of hydro- hydrastinine derivatives," J. Pharm. Soc. Japan. 28, 617-624 (1938)

103 104

11. Sasaki, Y., Ohnishi, H., Satoh, N. "Alkaloids of Menispermaceous plants (CXXV). Cleavage of cepharan- thine by metallic sodium in liquid ammonia," Chem. Pharm. Bull. Jan an. 178-18? (1955)* 12. Ito, K., "Alkaloids of Michelia conrpressa 2. Struc ture of michepressine," J. Pharm. Soc. Japan. 81. 703-707 (1 9 6 1). 13. "The United States Pharmacopeia," Fifteenth revision, Mack Printing Co., Easton, Pa., p. IO96 (1955)* 14. Thies, H. and Reuther, F. W.f "Reagent for alkaloids on paper chromatograms," Naturwissenschaften 41. 230-231 (195*0 •‘

15. Randerath, K., "Thin Layer Chromatography," Verlag Chemie Weinheim/Bergstr., Academic Press, p. 56-57 (1963). 16. Stermitz, F. R. and KcMurtney, K. D., "Alkaloids of the Papaveraceae, X. New alkaloids from Argemone gracilenta Greene," J. Org. Chem.. 34. 555-557 (1969). 17. "The United States Pharmacopeia," Fifteenth revision, Mack Printing Co., Easton, Pa., p. 109? (1955)* 18. Coulson, C. B. and Evans, W . C., "Paper chromato graphy and paper electrophoresis of phenol and glycosides,* J. Chromatography. 1., 374-379 (1958).

19. Locock, R. A., "A phytochemical investigation of selected Euoatorium species," Ph.D. Dissertation. The Ohio State University, pp. 40-42 (1965). 2 0 . Fong, H. H. S., "A phytochemical investigation of the non-phenolic tertiary and quaternary alkaloids of Thalictrum rochebrunianum Franc, and sau. (Ranunclaceae), " Ph.p^jssiFtatiQn, The Ohio State University, pp. 6I-6 3 (1 9 65 ) 2 1 . Schiff, P. L., Jr., "The isolation and characteriza tion of the alkaloids of Thalictrum minus L. var. adiantifolium Hort," Ph.D. Dissertation. The Ohio State University, Columbus, Ohio (1 9 67), 22. Tomita, M. and Watanabe, H., "Ullman reaction on the nitrogen-containing heterocyclic compounds," J. Pharm. Sci. Japan. 58. 783-789 (1938) 105

23* Harman, R. E. and Jensen, B. L., "Synthesis and photolysis of Krey’siginone," J. Heterocyclic Chem.. 2. 1077-1081 (1970). 24. Shriner, R. L.f Fuson, R. C. and Curtin, D. Y., "The Systematic Identification Of•Organic Compound," 5th edition, John Wiley & Sons, Inc., N. Y., p. 63 (1967). 2 5 . Mitscher, L. A., Leu, R. P., Bathala, M. S., Wu, W. N.t Beal, J. L. and White, R., "Antimicromial agents from higher plants. I. Introduction, rationale, and methodology," Llovdia. 25, 157-166 ■ (1972). 26. Bick, I. R. C. and Douglas, C. K., "Yellow alkaloids of Atherosoerma moschatum Labill," Tetrahedron Letter 2 5 , 1629-1633 (1964) 27. Shamma, K. and Slusarchyk, W. A., "The aporphine alkaloid alkloids," Chem. Rev.. 6k, 59-79 (1964), 28. Craig, J. C. and Roy, S. K., "Optical rotatory dis persion and absolute configuration-II. Aporphine alkaloids," Tetrahedron. 21. 395-399 (1965). 29. Brochmann- Hanssen, E. and Furuja; T., "A new opium alkaloid, isolation and characterization of (+)-l- (3 1-hydoxy-4■ -methoxybenzyl)-2-mehtyl-6-methoxy- 7-hydroxy-1,2,3,4-tetrahydroisoquinoline (+)-reticuline," Planta Kedica. 1 2 . 328-333 (1964).

3 0 . Craig, J. C. , ?<:artin-Smith, K., Roy, S. K. and Stenlake, J. B., "Optical rotatory dispersion and absolute configuration VIII. Petaline and other benzyltetrahydroisoquinoline alkaloids," Tetrahedron 22, 1335-1339 (1966).

31. Kupchan, S. ?.l., Suffness, M. I. and Gordon, E. M., "The isolation and structure elucidation of oxo- xylopine, a new oxoaporphine alkaloid from Steohania abvssinica Waso." J . Org. Chem.. 35. 1682-1684 (1 9 7 0).

3 2 . Tomita, K. and Sassagawa, S., "Studies on the ingredients of Aristolochiaceous plants II. The ingredients of Chinese drug "Fang-chi," J . Pharm. Soc. Japan. 22. 973-976 (1959). Coutts, R. T., Stenlake, J. B. and Williams, W. D., "The chemistry of the Aristolochia species. Part III Aristolochic acid and related substances from Aristolochia reticulata and A. indica," J. Chem. Soc. 4121-4124 (1967). Maldonado, L. A., Herran, J. and Romo, J., "La taliscanina, un componenta de Aristolochia taliscana Sobretiro de Ciencia. Mex 24, 237-240 (1966). APPENDIX nutfMTTiVJtytU vu*vwuws:&ftm iue1 IRspectrum ofFigure doryphornine11 (CHCl^) 3IM Wt ■ W A V Z W U t l Q J4D9 to (CM** u 00 o h* 2-s-f- iue1 KMHspectrum of doryphornineFigure(CDCl^ 12 ‘ ‘wMi'JV '4 j S‘ iKMCtaiTnufccct} t i l M iue1 IR spectrum of syntheticFigure13 thalifoline (KBr) M fco^i 13* H* o 100 40 50

~3&> «>ei ik e ' lu o w >0 . fl iw n WWMMKR(&f? too

■Figure 14 IR spectrum of synthetic doryphornine (CHC1-) 111 wo 910

7.0 3*0

Figure 15 NMR spectrum of synthetic doryphornine (CDC1~) TWfteMfTTMlClU) 166 0 4 w TOT —*- —— WAVBJWBOUCW-J iue1 IRspectrum of Figurebase(KBr) A 16

TOT •IB r* ■V'«*NUMB8KO<‘% ROD M M looo lit 110 2.0 3.0 4.0 S.0 9JO

<00 m

M ZQ

Figure 17 NMR spectrum of base (CF-COOH) j r ASSORWfcC 13 iue1 IRspectrum of Figuredoryflavine18 (KBr) 4.0 *MQ a5o W*VlKVMOIKU>rt

>■• loo . n OS so 3.0 4.0

M ne

-|C I ■

6.0 slo M-MI*) 3.0 XO 1.0

Figure 19 NMR spectrum of doryflavine (Pyridine-dJ «• N It M ' M H Jimlimlnm MO

75F

Figure 20 IR spectrum of doryflavine dimethylether (CHC1?) 7.0 3.0 4.0 7.0 ■.0 9.0

. > " > m » its o era

fj/> i a.o 3.0 1.0

Figure 21 KMR spectrum of doryflavine dimethylether (CDCl^) . BATTELLE MEMORIAL INSTITUTE MASS SPECTROMETRY LABORATORY SAMPLE CRC-III-108B. DOSKOTCH 09/21/71 RUN NO. 29

' Figure 22 Mass spectrum of doryflavine dimethylether 95 *> 50 • 0 TO 8 0 80 IOC 109

40 xo

Miao ezs WAVffir.-.miB (P4*fi WftVEMUHBOf raft

Figure 23 IR spectrum of doryflavine triacetate (KBr) 3.0 3.0 4.0 T* a'.o rfMir) *.o 1.0 li M IM

^ **1 - » ■ hHf*» »»» j

TO To 3.0 1.0 121 Figure 2k NMR spectrum of doryflavine triacetate (CDC1,) TRMSWrrMenW r n iue2 IR spectrum of Figurebase (KBr)25B . B .(1 MBS WnnMMUtOHL m n*

40 M 122 2.0 3.0 4.0 s.o rpMtrl 6.0 *.0 ’T’ i-i1 ■■■;i I T J09 KB T T

Figure 26 NMR spectrum of base B (CF^COOH) to M *? IM

W US> a®* Toco tto VMttMKKWi

Figure 27 IR spectrum of N-methylated base (CHCl^) TitautMirrMJCXa# 100 ~ v 30 V/MENUMSEKICM 1 iue2 IRspectrum Figure.of28 l-(p-methoxytoenzyl)-2-methyl-6-methoxy y^ ' ■yn^ 3w 1 , 2,3 p4rtetrahydro-7-xsoquinolinol (CHCly Wvi^NIJfJerRlCM'j 00 iyiy> m

i* 125