A PHYTOCHEMICAL INVESTIGATION OF THE ROOTS AND TUBERS OF RUMEX HYMENOSEPALUS FAMILY POLYGONACEAE
Item Type text; Dissertation-Reproduction (electronic)
Authors Buchalter, Leonard, 1922-
Publisher The University of Arizona.
Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.
Download date 29/09/2021 09:03:58
Link to Item http://hdl.handle.net/10150/284768 This dissertation has been microfilmed exactly as received 67-381
BUCHALTER, Leonard, 1922- A PHYTOCHEMICAL INVESTIGATION OF THE ROOTS AND TUBERS OF RUMEX HYMENOSEPALUS FAMILY POLYGONACEAE.
University of Arizona, Ph.D., 1966 Chemistry, pharmaceutical
University Microfilms, Inc., Ann Arbor, Michigan A PHYTOCHEMICAL INVESTIGATION OF THE ROOTS AND TUBERS OF RUMEX HYMENOSEPALUS FAMILY POLYGONACEAE
by Leonard Buchalter
A Dissertation Submitted to the Faculty of the COLLEGE OF PHARMACY In Partial Fulfillment of the Requirements For the Degree of DOCTOR OF PHILOSOPHY In The Graduate College THE UNIVERSITY OF ARIZONA
1966 THE UNIVERSITY OF ARIZONA
GRADUATE COLLEGE
I hereby recommend that this dissertation prepared under my
direction by T.pnnar^ Bnp.Vialt.pr
entitled A Phyhnr.hprr.i p.al Tnvpsti gflH nn nf thf. Rnohs and
Tn-h^-rs nf Piimp-ir HvwiftrinsftT)ali]Br Family Pol vgonaceae,
be accepted as fulfilling the dissertation requirement of the
degree of nnr.t.rvr of Philosophy
ertatioji Director Date
After inspection of the dissertation, the following members
of the Final Examination Committee concur in its approval and
recommend its acceptance:*
7-/3-U
/ /f
*This approval and acceptance is contingent on the candidate's adequate performance and defense of this dissertation at the final oral examination. The inclusion of this sheet bound into the library copy of the dissertation is evidence of satisfactory performance at the final examination. STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the Uni versity Library to be made available to borrowers under rules of the Library,, Brief quotations from this dissertation are allowable without special permission, provided that ac*> curate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.
SIGNED; ACKNOWLEDGMENTS
I wish to thank my research director, Dr. Jack Cole, for his advice, guidance and continual encouragement; and the other members of the College of Pharmacy faculty for their encouragement throughout this investigation, I also wish to thank members of the faculty of the Department of Chemistry, as well as my undergraduate instructors in Pharmacy at the Philadelphia College of Pharmacy and Science. I wish to thank the Cancer Chemotherapy National Service Center and the National Cancer Institute, U.S. Public Health Service, Bethesda. Maryland, for financial support in part by contract PH-43-63-1136 and research grant CY-5076-MC, and also the American Cancer Society for support from an Institutional Grant at the University of Arizona,, I am especially thankful to my wife, Barbara, and my mother for their inspiration, assistance, and encourage* ment in this endeavor.
iii TABLE OF CONTENTS
Page
LIST OF ILLUSTRATIONS ...... vii LIST OF TABLES...... ix ABSTRACT. . x INTRODUCTION 1 TANNIN EXTRACTION 3 GENERAL CHEMICAL CHARACTERIZATION OF TANNIN EXTRACT .. 6 Preliminary Chemical Analysis...... 6 Infrared Analysis...... i ••••••••• 8 Ultraviolet Analysis ••••••••••••• 9 General Chemical Characterization of the Tannin Fraction. • •••••• 9 Paper Chromatographic Analysis 10 General Classification Based on Preliminary Paper Chromatographic Investigation , 14 FRACTIONATION OF TANNIN EXTRACT ...... 16
t General Fractionation Procedure . 16 Paper Chromatographic Analysis of Fractions A and B...... 16 Infrared Analysis of Fractions A and B...... 18 Classification of Fractions A and B... 21 Conversion of Fractions A and B from Leuco- anthocyanidins to Anthocyanidins. . . 21 Purification of Anthocyanidins Developed in Fractions A and B Prior to Separation 23 General Chemical Analysis of Fraction B...... 25 Separation of Anthocyanidins by Paper Chromatography. ••••••••••• 25 Preparative Paper Chromatography for Isolation of Anthocyanidins .. . . , 29 Visible Spectral Analysis of Anthocyanidins. • 30 Infrared Spectrum of Cyanidin from Preparative Paper Chromatography. . . 32 iv V
TABLE OF CONTENTS—Continued
Page
Degradation by Potassium Fusion of Anthocyanidins • 35 Paper Chromatography Identification of Fragments ...... 36 Summary of Phytochemical Investigation of Fraction B 36 General Chemical Analysis of Fraction A, . , • • . 39 Conversion of Fraction A to Anthocyanidins . . 41 Paper Chromatography of Anthocyanidins of Fraction A. .. . . 41 Structure Proof of Anthocyanidins of Fraction A ...... 42 Potassium Fusion of Fraction A...... 42 Isolation of Leucocyanidin from Fraction A • • 43 Summary of Chemical Analysis of Fraction A • . 46 Theoretical Implications of Phytochemical Analysis of Fraction A • 47 Condensation by Acid 49 Enzymic Condensation Through a Quinone Polymerization Mechanism ...... 49 Polymers with Ether Linkages • •.«••••• 50 ANTI-TUMOR TESTING. 55 History of Testing ...•• ••••• 55 Examination of Five Different Varieties for Anti-Tumor Activity. • •••••• 56 Preliminary Extracts .••••...••••••. 57 Summary of Anti-Tumor Testing. 57 NON-POLAR EXTRACT ...... 65 Preliminary Evaluation 65 Paper Chromatography of Non-Polar Fraction . . . . 66 Extraction Procedure ••••••••••••••• 66 Paper Chromatography of Chrysophanic Acid, Physcion and Emodin. • ••••«.•» 67 Preliminary Thin-Layer Chromatography. ...••• 70 Preparative Thin-Layer Chromatography 72 Isolation of Chrysophanic Acid and Physcion. ... 76 Identification of Chrysophanic Acid. . • . . • 77 Identification of Physcion . 81 Acetate Derivative. 82 vi TABLE OF CONTENTS--Continued
Page Isolation and Identification of Emodin ...... 84 Summary of Materials Identified in7 Non-Polar Extract ...... 89 SUMMARY AND CONCLUSIONS » . , . . . 91 REFERENCES. 95
f LIST OF ILLUSTRATIONS
Figure Page 1. Extraction Procedure for Tannin Fraction • • • • • 5 2. Infrared Spectrum of Tannin Extract. II 3. Ultraviolet Spectrum of Tannin Fraction 12 4. Paper Chromatography of Tannin Extract 15 5. Solvent Fractionation of Tannin Fraction • • • . • 17 6. Infrared Spectrum of Fraction A. ••••••••• 19 7. Infrared Spectrum of Fraction B. ••••••••• 20 8. Conversion of Fractions A and B to Anthocyanidins . . 22 9. Anthocyanidin Purification Procedure 26 10. Infrared Spectrum of Cyanidin Chloride, Commercial Sample, 33 11. Infrared Spectrum of Cyanidin Chloride, Synthesized from Fraction B..,..«•••••. 34 12. Degradation of Anthocyanidins by Potassium Fusion, 37 13. Basic Structures of Leucoanthocyanidins and Anthocyanidins from Fraction B ...... 40 14. Infrared Spectrum of Monomeric Leucocyanidin ... 45 15. Structures Illustrating Types of Condensations . . 53 16. Extraction Procedure for the Non-Polar Extract . . 68 17. Thin-Layer Chromatography of Non-Polar Extract , , 73 18. Infrared Spectrum of Chrysophanic Acid from Non-Polar Extract . . 79
vii viii / LIST OF ILLUSTRATIONS (Continued) Figure Page 19. Infrared Spectrum of Chrysophanic Acid, v Commercial Sample 80 20. Infrared Spectrum of Physcion from Non-Polar Extract ...... 83 21. Infrared Spectrum of Qnodin from Non-Polar Extract 87 22. Infrared Spectrum of Emodin, Commercial Sample...... 88 23. Anthraquinone Pigments of Non-Polar Extract 90 LIST OF TABLES Table Page
1. Qualitative Tests for Anthocyanidins. 24 2. Rf Values of Anthocyanidins Developed from Fraction B ...... 28 3. Visible Spectral Analysis of Anthocyanidins from Fraction B, • 31 4. Rf Values of Commercial Samples and Fusion Fragments* 38 5. Anti-Tumor Test Results ••••..•••••••• 60 6. Paper Chromatography of Non-Polar Extract . . . . . 71 7. Thin-Layer Chromatography of Non-Polafc Extract 74 8. Rf Values Thin-Layer Chromatography on Silica Gel Layers of Material (A; and Commercial Sample of Chrysophanic Acid. • ••••• 78 9. Rf Values Thin-Layer Chromatography on Silica Gel Layers of Natural Product from Spot No. 1 and Commercial Sample of Emodin 85
ix ABSTRACT
The roots and tubers of Rumex Hvmenosepalus yielded a tannin fraction which demonstrated anti-tumor activity. A phytochemical analysis was undertaken to attempt to iden tify the material responsible for the anti-tumor activity,, The tannin fraction was identified as one containing poly- phenolic flavanoidal units. The tannin fraction was separated into a polymeric condensed fraction, and a fraction consisting of monomeric units, by means of solvent fractionation with water and ethyl acetate* The polymeric condensed fraction of the tannin extract was chemically identified as consisting of polymeric leucoanthocyanidin (flavan-3:4-diol) units. The individual leucoanthocyanidins present in the condensed portion consisted of leucocyanidin (5:7:3':4*-tetrahydroxyflavan-3:4-diol)9 leucodelphinidin (5:7:3*:4*:5*-pentahydroxyflavan-3:4-diol), and leucopelargonidin (5:7:3*-trihydroxyflavan-3:4-diol). The presence of these units in the condensed polymer of the tannin extract was demonstrated by conversion of the leucoanthocyanidins to their corresponding monomeric antho- cyanidins, cyanidin, pelargonidin, and delphinidin. The identification of the corresponding anthocyanidins was x xi accomplished by paper chromatography, visible and infrared spectral analysis, and fragmentation studies through potassium fusion. The non-condensed portion of the tannin extract was identified as monomeric leucoanthocyanidin units of the same type contained in the condensed fraction. The monomeric leucoanthocyanidins (flavan-3:4-diols) were identified as leucocyanidin, leucodelphinidin and leucopelargonidin. The identification of the monomers was accomplished by conversion of the monomers to their corresponding anthocyanidins, and the subsequent identification of these monomers as cyanidin, delphinidin and pelargonidin. Proof of structure of the anthocyanidins was carried out by paper chromatography, visible and infrared spectral methods and identification of fragments from potassium fusion. The anthocyanidins were polyhydroxy derivatives-.of 2-phenylbenzopyrylium or flavylium cation. Cyanidin is (3:5:7:3*:4,-pentahydroxyflavylium), delphinidin (3:5:7:3*:4':51-hexahydroxyflavylium) and pelar- gonidin (3:5:7:4*-tetrahydroxyflavylium). All of the antho- cyanidins yielded phloroglucinol (2,4,6 tirihydroxybenzene) from potassium fusion; cyanidin yielded protocatechuic acid (3,4 dihydroxybenzoic acid); delphinidin yielded gallic acid (3,4,5 trihydroxybenzoic acid), and pelargonidin yielded (p-hydroxybenzoic acid) on potassium fusion. These fragments xii resulted from fragmentation of the basic flavylium nucleus, and were a means of identifying the corresponding anthocyanidins. The monomeric leucoanthocyanidin, leucocyanidin, was isolated from a polycaprolactam column chromatography, and was identified by conversion to cyanidin, which was identified by the same methods as previously reported. In anti-tumor tests carried out by the National Insti tute of Health Cancer Chemotherapy National Service Center, it was verified that the primary anti»tumor activity resided in the condensed polymeric leucoanthocyanidin fraction of the tannin extract. Fractional purification of the tannin extract by separation of the monomers present was found not to increase anti-tumor activity. Subsequent phytochemical investigation of the non- polar (ether) extract of the roots and tubers resulted in the identification of the polyhydroxy anthraquinone pigments, chrysophanic acid (1,8 dihydroxy-3-methyl-anthraquincne), physcion (1,8 dihydroxy-3-methoxy-6-methyl anthraquinone) and emodin (1,3,8 trihydroxy-6-methyl anthraquinone)6 The identi fication of the anthraquinone pigments was accomplished by thin-layer preparative chromatography on silica gel layers, infrared and ultraviolet spectral analysis, paper chromatog raphy, and other qualitative organic analytical procedures. Testing of the non-pol^r extract showed that this fraction was devoid of anti-tumor activity. INTRODUCTION
Rumex Hvmenosepalus. a dicotyledon of the family Polygonaceae, also known by the common name Canaigre, is a plant native to the southwestern United States, In tests performed on the various extracts of the plant by the Cancer Chemotherapy National Service Center, National Cancer Institute, Bethesda, Maryland, it was discovered that the tannin-containing fraction of the plant exhibited anti-tumor activity against Sarcoma 180 and Walker 256 Test Systems, An investigation of this fraction was begun in order to attempt to identify the material responsible for the anti-tumor activity, A literature search revealed the existence of several methods of extracting tannins from the plant as employed by leather chemists. Chemical compounds already discovered in the plant include substances such as chryso- phanic acid, physcion (1), and possibly emodin (2), Poly phenols are said to predominate in Canaigre tannins (3). Members of this classification of compounds vary little in phenolic reactivity and therefore are very difficult to separate chemically (3). Polyphenols are capable of a great deal of mutual solubilization, resulting in solid solutions which tend to behave as if they were single substances (4). The roots, the major tannin-containing part of the plant, 1 also contain sugars and starches, which make the usual methods of tannin removal more complicated (4). Therefore, a special method of tannin extraction (Figure 1) was developed to fill the requirements of this investigation. In addition to the tannin extraction procedure, another extraction procedure was developed to obtain a non-polar extract of the ground roots and tubers, (See Figure 16). The phytochemical investigation was thus broadened to both the tannin and non-tannin-containing fractions of the dried roots and tubers. The specific purpose of the investigation of these two fractions was to determine the chemical species responsible for the anti-tumor activity. Prior to this investigation, no phytochemical analysis of the tannin- fraction had been undertaken. TANNIN EXTRACTION
A 600 C5n. sample of the roots and tubers of the plant was ground into a damp reddish-brown meal-like material in a Wiley mill equipped with a 4 millimeter size screen. The dry ground material was then washed with petroleum ether and ethyl ether. The residues obtained from these extractions were set aside for possible future investigation. The plant marc was air dried for a 24-hour period. It was then extracted with 4 liters of a 95% ethanol-methanol (1:1) mixture for 120 hours. The resulting reddish-brown solution was separated from the marc by filtration and allowed to evaporate to dryness. The plant marc was discarded. The amorphous brown residue of the methanol-ethanol extraction was dissolved in approximately 1 liter of distilled water. The solution was then washed repeatedly with a total of 2 liters of chloroform. (Figure 1) The washed water extract was then frozen and lyo- philized at -50° C.to dryness. Approximately 120 Gnu of a light orange-brown, semi-crystalline powder was obtained. The lyophilized extract was submitted to the Cancer Chemo therapy National Service Center, National Cancer Institute, Bethesda, Maryland, for antitumor testing. Testing in Walker 256 Test System (5) gave a value of 35% T/C(Test/Control)
3 4 at 60 mg«/Kg. Also against the Sarcoma 180 Test System (5) a value of 14% T/C at a dose of 90 mg./Kg. was obtained. In the Sarcoma 180 and Walker 256 Solid-Tumor Test Systems a material can be considered confirmed if there is an active, response (T/C = 42%) on each dose-response experi ment regardless of dose. The average survival on the active responses must be = 83%. (6) The results of tests carried out on the tannin fraction indicated activity. 5
Roots and Tubers 1) Petroleum ether wash 2) Ethyl ether wash 3) 95% ethanol-methanol
>iarc Solution (red-brown) (Discard) 1) Evaporate to dryness 2) Dissolve in distilled water
Water Solution 1) Chloroform wash 2) Lyophilize
Semi-crystalline Powder
Figure 1. Extraction Procedure for Tannin Fraction GENERAL CHEMICAL CHARACTERIZATION OF TANNIN EXTRACT
Preliminary Chemical Analysis The amorphous orange-brown powder obtained by the tannin fraction extraction procedure was then subjected to a series of chemical tests in order to obtain some indication of its chemical nature. The following tests were employed: 1. Sodium Fusion (7). Elemental analysis by sodium fusion showed the absence of nitrogen, sulfur, or halogens. 2. jdH Analysis. Aqueous solutions of the extract demonstrated indicator-like properties by color changes with change of pH. Aqueous solutions were orange-brown at pH of 2.1, orange at pH of 8.4 and at pH of 13.4 were a red-violet color. 3* Ferric Chloride Test Solution (8). One drop of ferric chloride solution (9%) was added to 0.5 Gm» of the tannin extract in 5 ml. of methanol. A blue-black color produced indicated the presence of phenols. 4. Gelatin Precipitation lest (9). To 1 ml. of an aqueous gelatin solution (1%) was added 1 ml. of 0.5 Gm. of the tannin extract dissolved in 5 ml. of water. Precipitation of the gelatin indicated a positive test for tannins. 5. 2»4-Dinitrophenvlhvdrazine (10,11). This test for carbonyl groups was negative.
6 7 6. Mayers Test (Mercuric-Potassium Iodide Test Solution) (12), Reagents: (a) Mercuric Chloride (1.353 On.) in 60 ml. of water. (b) Potassium Iodide (3.000 Gm.) in 10 ml. of water. The two solutions were mixed and water was added to make 100 ml. of solution. The test solution ( 5 ml.) was added to 0.5 Gm. of the tannin extract in 5 ml. of water to which one. drop of dilute HC1 had been added. No precipitate formed, indicating the absence 6f alkaloids. 7. Liebermann-Burchard Test for Steroids (13). A 0.5 Gm. sample of the extract dissolved in alcohol was treated with 5 drops of acetic anhydride and then 2 drops of sulfuric acid. Color changes characteristic of steroids did not occur. 8. Solubility Test. The extract was soluble in water and the resultant solution gave an acid reaction to litmus and did not liberate carbon dioxide from a freshly prepared saturated solution of sodium bicarbonate, indicating the absence of free carboxyl groups. The extract was soluble in alcohol and insoluble in all non-polar solvents. 9. Melting Point. No sharp melting point was obtainable up to 300° C. The material reddened over 200° C. 10. Hydrolysis. Aqueous solutions of the tarm in extract were subjected to hydrolytic procedures, using both dilute and concentrated solutions of hydrochloric acid and sodium hydroxide solution, and refluxing. The material was non-hydrolyzable under these conditions, but color changes were noted. 8
11. Leucoanthocvanidin Test (14). After boiling an alco holic solution (0.5 Gm. in 5 ml.) of the tannin extract with 10% hydrochloric acid for 15-30 minutes, a red precipitate appeared. This is a positive test for leucoanthocyanidins.
Infrared Analysis A Perkin-Elmer Infracord spectrophotometer was used to obtain an infrared spectrum of the tannin fraction. (Figure 2) The spectrum was obtained utilizing a potassium bromide pellet. The spectrum showed broad absorption at 2.90 to 3.00 microns, a short band at 5.95 microns, a long band at 6.25 microns, a medium band at 6.60 microns, and a long band at 6.95 microns. In the region from 8.00 to 15.00 microns a short broad absorption occurred at 8.30, 9.15, 9.80, 11.50, 12.20 and 13.10 microns. The spectrum obtained showed corre lation in all areas with a spectrogram for tannins and cate~ chins of purified Quebracho extract tannins (15). The weak plateau in the carbonyl region has been reported by Jones (16) in both Quebracho and Redwood tannins. The infrared spectrum demonstrated the presence of phenolic hydroxyl groups, probably hydrogen bonded, aromatic absorption, and the lack of strong carbonyl-functional absorption. The absence of any aliphatic stretching vibrations was also noted. Comparison with spectrograms of other tannin material, as reported above (16), verified the tannin-like nature of the extract. 9
Ultraviolet Analysis An ultraviolet spectrum was made of the tannin extract by a Beckman DB Spectrophotometer, A 2 x 10"^ molar aqueous solution of the extract showed a maximum absorption at 285 millimicrons with a minimum at 265 millimicrons and a levelling off at 350 millimicrons. This ultraviolet absorption is the characteristic absorption of leucoanthocyanidins (flavan-3I4-diols) and catechins (17). (Figure 3)
General Chemical Characterization of the Tannin Fraction Most of the tannins can be classified as hydrolyzable or condensed. Current investigations (18) show that hydro lyzable tannins consist of glucose or other saccharides poly- es,terified by gallic acid or by phenolic acids clearly derivable from gallic acid. The tannin fraction obtained from Rumex Hvmenosepalus was clearly of the non-hydrolyzable type, as all attempts to hydrolyze the extract failed. One of the well known tests for condensed tannins is the formation of an insoluble red precipitate upon boiling in dilute mineral acid (19). Bate-Smith (14) applied this test to the extracts of leaves; a positive reaction being obtained in the majority of the woody, dicotyledonous plants tested. The red color was found to be due to the generation of anthocyanidins from the leucoanthocyanidins present in the 10 plant tissues. A flavan-3J4-diol structure was suggested for the leucoanthocyanidins, and it was concluded that the leucoanthocyanidins were the substances most commonly respon sible for the reactions in plant tissues attributable to tannin (19)0 Since the tannin fraction from Rumex Hvmeno° sepalus gave a positive result when boiled in dilute mineral acid, it indicated the presence of leucoanthocyanidins in the extract. The results of qualitative organic tests, infrared and ultraviolet analysis, and hydrolytic procedures indicated that the lyophilized extract obtained from the dried roots and tubers of Rumex Hvmenosepalus could be generally classified as a polyphenolic, flavanoidal condensed tannin.
Paper Chromatographic Analysis
The amorphous tannin fraction was then subjected to paper chromatography, Whatman No„ 1 filter paper was used in chromatography, and a descending solvent saturated tank method was employed. The solvent system found most suitable for the paper chromatographic investigation was 1-butanol^glacial acetic acid, and water (4:ls5); upper phase only. An aqueous solution containing 0.5 Gm. of the tannin in 5 ml.of water was spotted on the Whatman No. 1 filter paper, and the spots were developed by descending chromatography for six hours. Development of the chromatogram (18 inches) yielded results shown on Figure 4. A light yellow streak extended from the 4000 3000 2000 1500 CM" 1000 900 800 700
7 8 9 10 11 WAVELENGTH (MICRONS)
Figure 2. Infrared Spectrum of Tannin Extract 0 240 260 280 300 320 Millimicrons
Figure 3. Ultraviolet Spectrum of Tannin Fraction 13 origin to a slightly visible spot at Rf value 0.72. No other visible streaks or spots appeared on the chromatogram. The chromatogram was then examined under an ultraviolet lamp. Two additional spots appeared. At Rf 0.54 an ultraviolet lavender spot appeared over the streak extending from the origin, and at Rf 0.64 a light blue ultraviolet spot appeared. The spot at Rf 0.72 appeared lavender in ultraviolet light. After drying, the chromatogram was cut into strips and tested with the following chromagenic sprays: (1) 3% p-toluene sulphonic acid (20) (2) 5% ethanolic vanillin hydrochloric acid (21) (3) 5% ferric chloride solution The first two test reagents are useful for detecting leuco- anthocyanidins on paper. Five percent ferric chloride solution is used to detect phenol's. One strip cut from the chromatogram was sprayed with 3% p=»toluene sulphonic acid and heated at 103° C. for 5-10 minutes in an oven. The streaking material from the origin turned pink scarlet red. This is a positive test for leuco- anthocyanidins based on the conversion of leucoanthocyanidins to the corresponding anthocyanidins. The color of the streaking material, although pink scarlet, was not uniformly colored but varied in shades. A light green color appeared at Rf 0.28 in the streak. One strip cut from the chromatogram was sprayed with a 5% ethanolic vanillin hydrochloric acid solution. The 14 streak from the origin turned various shades of pink, red and scarlet after heating in the oven. The individual spots at higher Rf values became pink red with this spray. Leucoanthocyanidins turn red in this reagent. A second strip cut from the chromatogram was sprayed with a 5% solution of ferric chloride. With this reagent, the streak turned blue-black on heating in the oven, and the individual spots also turned blue and green on heating in the oven. This test is indicative of phenolic materials.
General Classification Based on Preliminary Paper Chromatography Investigation
The results obtained from this paper chromatographic investigation indicated that the tannin was of phenolic character, and the condensed portion contained polymeric leucoanthocyanidin units, which are accompanied in the tannin fraction with monomeric leucoanthocyanidins (Flavan-3J4~diols)„ The absence of other flavanoids such as flavanols, catechins, flavones, isoflavones, aurones and chalcones was determined by exposing the material on paper to examination with ammonia fumes in visible and ultraviolet light. In no instance did these tests indicate the presence of flavanoidal types other than monomeric and polymeric leucoanthocyanidins (22,23). The presence of cinnamic acid and hydroxylated derivatives was ruled out by chromatography on paper of the tannin frac» tion with commercial samples of caffeic acid, ferulic acid, chlorogenic acid, p-coumaric acid, and cinnamic acid in
1-butanol-glacial acetic acid-water (4:1:5). 15
Origin
. Solvent Front
Figure 4. Paper Chromatography of Tannin Extract FRACTIONATION OF TANNIN EXTRACT
General Fractionation Procedure
When it became apparent that the tannin fraction consisted of polymeric and monomeric materials, an effort was made to separate these two entities into single fractions. After experimentation with various solvent extraction procedures, it was determiiied that a separation of the two fractions could be obtained by the use of ethyl acetate. The extraction procedure is shown in Figure 5. Two fractions were obtained from the solvent extraction procedure. The water soluble fraction extract- able with ethyl acetate will be referred to as Fraction A. The water soluble non-ethyl acetate extractable fraction will be referred to as Fraction B. Both fractions were submitted for anti-tumor testing and the results appear in Table 5.
Paper Chromatographic Analysis of Fractions A and B
The solid materials obtained as Fraction A and
Fraction B were applied to Whatman No. 1 filter paper, and subjected to paper chromatography, using 1-butanol-
glacial acetic acid-Water(4:1i5) as the solvent. system. The chromatogram was developed over a period of six hours by descending method,.. The solvent front extended ovejr a distance of 18 inches, from the origin. The 16 17
Tannin Fraction 1) Distilled water 2) Ethyl Acetate
Aqueous Fraction Ethyl Acetate Fraction (Brown Solution) (Straw-Yellow) 1) Filter 1) Anhydrous Magnesium Sulfate 2) Evaporate 2) Filter 3) Evaporate
Golden-Brown Scaly Material Yellow-Brown Amorphous Powder
Fraction ]B Fraction A
Figure 5, Solvent Fractionation of Tannin Fraction 18 result of this paper chromatography experiment indicated that Fraction A consisted only of the three spots detected at Rf values (0.54, 0.64, and 0.72) in the original paper chromatography of the tannin extract (Figure 4). Fraction B consisted only of the streaking that appeared in the original paper chromatography of the tannin extract, extending from the origin to Rf value 0.72 (Figure 4).
Infrared Analysis of Fractions A and B
The solid materials obtained as Fraction A and Fraction B were then subjected to infrared analysis# Spectra were obtained by using potassium bromide pellets and a Perkin Elmer recording spectrophotometer. The spectrum obtained from Fraction A (Figure 6) was similar to the spectrum obtained of the original tannin extract. (Figure 2) The resolution of the spectrum obtained from Fraction A was sharper than that obtained from the original tannin extract. There was a slight deviation in absorbance between 14 and 15 microns in the spectrum of the tannin extract and that of Fraction A.
The infrared spectrum obtained from Fraction B (Figure 7) showed the same basic absorption as that of Fraction A, but the peak at 2,9 to 3.2 microns was much broader (hydrogen bonding) in Fraction B, and the resolution in the fingerprint area of the spectrum was not nearly as sharp as that in the Fraction A spectrum. 4000 3000 1000 900
7 8 9 10 11 WAVELENGTH (MICRONS)
Figure 6. Infrared Spectrum of Fraction A 4000 3000 2000
7 8 9 10 11 WAVELENGTH (MICRONS)
Figure 7. Infrared Spectrum of Fraction B 21
From these spectra, it was concluded that both
Fractions A and B consisted of molecules essentially the same as those in the original tannin extract. Since Fraction B was condensed material, this may account for the lack of sharp absorption.
Classification of Fractions A and B
The natural chromogens (24) may be classified into the following three groups: (a) those that are insoluble in water and the usual organic solvents, or give only colloidal solutions, (b) those readily soluble in water and not extracted by ethyl acetate, and (c) those which may be extracted from aqueous solutions by means of ethyl acetate. This is a use ful classification of substances isolated from natural sources, and the divisions correspond broadly to (a) condensed polymers, (b) glycosides or diglycosides, and (c) monomers, including flavan-3i4-diols. Fraction A, extracted from the tannin extract, corresponded to monomers, including flavan-3«4~ diols and Fraction B corresponded to condensed polymers.
Conversion of Fractions A and B from Leucoanthocyanidins to Anthocvanidins
Both Fractions A and B when boiled for 15-30 minutes in ethanolic 10% hydrochloric acid gave red solutions, indi cating that both fractions consisted of leucoanthocyanidins which were convertible to anthocyanidins.(See Figure 8)» 22
Aqueous Solution of Tannin Extract 1) Ethyl Acetate
Ethvl Acetate Fraction A Water Soluble Fraction B
1) Evaporate 1) Evaporate
Monomeric^ Polymeric Leucoantho cvanid in s Leucoanthocvanidins 1) Ethanolic 1) Ethanolic 10% HC1 10% HC1 2) Boil 30 2) Boil 30 minutes minutes Red-Violet Solution Red-Orange Solution
1) Evaporate 1) Evaporate
lAnt hoc vanid insl {Anthocyanidinsl
Figure 8. Conversion of Fractions A and B to Anthocyanidins 23
A series of qualitative tests for the presence of anthocyanidins was carried out and the results are listed in Table 1. All the results were positive.
Purification of Anthocyanidins Developed in Fractions A and B Prior to Separation
It was necessary to develop a method of separating and identifying the anthocyanidin pigments which were produced from Fractions A and B, and in so doing characterize the leucoanthocyanidins from which they were derived. To identify the anthocyanidins produced from Fractions A and B, a pro cedure for purifying anthocyanidins was carried out. This procedure, outlined below, takes advantage of the fact that anthocyanidins are soluble in water, but insoluble in non-hydroxylie solvents such as ether, acetone, chloroform and benzene. Also they are completely precipitated from aqueous or alcoholic solutions in the form of blue lead salts, which are soluble in glacial acetic acid giving a dark red color. (25) To an aqueous solution of the crude anthocyanidins lead acetate tfas. added until all the blue lead salts were precipitated. The precipitate was separated from the super natant liquid by centrifuging. The blue lead salts were then converted into chlorides by the addition of 5% methanolic hydrochloric acid. A small amount of 207o aqueous hydrochloric acid was added. A large excess of ether was added, which transferred the pigment into the aqueous layer. This layer 24
Table 1 QUALITATIVE TESTS FOR ANTHOCYANIDINS
REAGENT FRACTION. A FRACTION B ANTHOCYANIDINS ANTHOCYANIDINS
5% FeCl3 Blue-black when Blue-black when neutral neutral-
Lead Acetate Blue Lead Salts Blue Lead Salts Test Solution Precipitate Precipitate
NHg Fumes on Red spot turns Red-orange spot Paper blue turns blue
HC1 Fumes on Blue spot turns Blue spot turns Paper red red-orange
isoamyl Extractable from Extractable from Alcohol acid solution acid solution
Visible Spectra max. 500-540 m/u max. 500-540 nyJ. Ethanol .1% HC1
10% Aqueous Turns blue Turns blue NaOH
Precipitated Red in Glacial Red in Glacial Lead Salts * Acetic Acid Acetic Acid
*This is not a reagentp but is included here since the pre cipitated blue lead salts of a&thocyanidins are soluble in glacial acetic acid giving a dark red color. 25 was separated from the supernatant liquid and heated on a boiling water bath for about three minutes. The solution was cooled quickly and extracted with isoamyl alcohol. The isoamyl alcoholic liquid extracted the pure anthocyanidins, and when this liquid was evaporated to a small volume, the purified anthocyanidins were applied to paper for chromato graphic analysis and separation. In all of the following procedures in which paper chromatography was used to identify anthocyanidins from Fractions A and B, the material was first purified by the method described above and outlined in Figure 9*
General Chemical Analysis of Fraction B Separation of Anthocyanidins by Paper Chromatography
Since Fraction B showed the highest level of anti-tumor activity, it was decided to investigate it initially. One Gm. of the golden brown scaly material obtained from evaporation of the water soluble non-ethyl acetate extractable Fraction B, was boiled for one-half hour in 30 ml# of ethanolic 10% hydrochloric acid. The red-orange solution obtained was evaporated to dryness, and 0.5 Gm. of the anthocyanidins obtained were then purified by the procedure of lead salt precipitation outlined previously in Figure 9. The isoamyl
i alcoholic solution of the red-orange material was concentrated to a small volume and applied to Whatman No. 1 filter paper for chromatographic analysis. Descendihg chromatography was used and the chromatogram,developed for six hours, ran 18 inches. 26
1) Water
Aqueous Solutic n Anthocyanidins 1) Lead Acetate Solution 2) Centrifuge
Blue Lead sJlts Precipitate Aqueous Layer
1) 5% Methanolic HC1 (Discard) 2) 20% Aqueous HC1 3) Excess Ether
EtherL Layer Aqueous Layer (Discard) 1) Boil 3 minutes 2) Isoamyl alcohol
Aqueous>1 Layer Isoamvl Alcohol Layer (Discard) 1) Evaporate
Purified Arithocvanidins
Figure 9. Anthocyanidin Purification Procedure 27 The following three solvent systems were utilized as they had proved effective for identifying anthocyanidins on paper.(26,27) Solvent System No. 1: Formic Acid-3N HC1 (1:1) Solvent System No. 2: Water-Acetic Acid-Conc.HCl (10:30:3) Solvent System No. 3: Acetic Acid-Conc.HCl-Water (5:1:5) All three solvent systems showed the presence of three anthocyanidin pigments. The Rf values in all solvent systems matched with those described in references for delphinidin, cyanidin, and pelargonidin.(26,27) The pigments retained their colors best in Solvent System No. 1. The following Rf values were obtained with this system. At Rf value of 0.11 a light violet-brown spot appeared. At Rf value of 0.22 a red spot appeared, and at Rf value of 0.33 a light orange spot appeared. These Rf values in this solvent system corresponded with reference (26,27) Rf values to: Rf value 0.11: Delphinidin Rf value 0.22: Cyanidin Rf value 0.33: Pelargonidin The Rf values obtained in Solvent Systems 2 and 3,described above, verified the presence of these three anthocyanidins. See Table 2. The next step in the investigation of Fraction B was to isolate and prove the structure of the three pigments that were tentatively identified in this fashion. 28
Table 2 Rf VALUES OF ANTHOCYANIDINS DEVELOPED FROM FRACTION B
SOLVENT SYSTEM Rf VALUE COLOR REFERENCE Rf (26927)
#1 Formic Acid- 0.11 Violet Brown Delphinidin 0.11 3N HC1 (1:1) 0.22 Red Cyanidin 0.22 0.33 Light Orange PelargonidinO.33
n Water-Acetic 0.29 Violet Delphinidin 0.29 Acid-Conc.HCl (10:30:3)* 0.50 Pink Red Cyanidin 0.50 0.71 Orange PelargonidinO.71
- -
n Acetic Acid- 0.20 Violet Delphinidin 0.22 Conc.HCl- Water (5:1:5) * 0.32 Red Cyanidin 0.34 0.55 Light Orange PelargonidinO.55
All spots on paper responded to ammonia fumes by becoming blue or blue-grey, and upon exposure to concentrated HC1 fumes reverted to their original colors* A commercial sample of cyanidin was used to verify Rf values with those of the literature. *The acetic acid used in solvent systems #2 and #3 was glacial acetic acid. 29
Preparative Paper Chromatography for Isolation of Anthocvanidins
It was necessary to obtain sufficient quantity of the pure pigments for visible and infrared analysis and degradation procedures. The method used for isolation of the pigments was the method of preparative paper chromatography,, This method was used successfully by previous investi gators (28), and consisted of streaking the material to be investigated on a large number of sheets of Whatman No. 3 filter paper. For this investigation, 25 sheets were streaked with purified anthocyanidin mixture obtained by conversion of leucoanthocyanidins from water soluble non-ethyl acetate extractable Fraction B. After streaking the 25 sheets, they were developed in tanks which were saturated with solvent, and descending chromatography was used. The solvent system employed for this purpose was Solvent System No. 1 (Page 27)9 which consisted of formic acid 88% and 3N.HC1 (1:1), as this solvent system gave the best resolution, and the pigments maintained their colors on the paper. The developed chromato® grams were air dried for 24 hours, and examined for antho- cyanidins. Results obtained were the same as those previously reported with the yield of cyanidin being the highest, delphinidin next, and the yield of pelargonidin the least. Strips corresponding to the individual anthocyanidins were then cut from each paper and separately extracted with methanolic 1% HC1 from the paper strips. The paper strips were cut 30 into small pieces prior to extraction, and were extracted for 24 hours to assure complete extraction from the paper.
Visible Spectral Analysis of Anthocvanidins
Although anthocyanidins characteristically exhibit intense absorption in the 500-550 millimicron region, addition of a few drops of aluminum chloride solution produces a bathochromic shift (15-50 millimicrons) of the principal A maximum of those anthocyanidin derivatives which contain adjacent hydroxy1 groups (29,30). This reagent is useful in distinguishing anthocyanidins such as cyanidin and peonidin, which have almost identical spectra in alcohol alone. The three pigments obtained from the preparative papers were subjected to visible light analysis in a Beckman DB Spectrophotometer. The results of this visible analysis were compared with the values obtained by Harborne (31) and each material was also analyzed after addition of three drops of aluminum chloride to the methanolic .1% HC1 in which they were originally taken. The results of this analysis (Table 3) verified the pigments that had been previously identified on paper. Delphinidin showed a maximum absorption at 545 millimicrons and a bathochromic shift of 20 millimicrons when aluminum chloride was added. Pelargonidin showed a maximum absorption in the visible region at 520 millimicrons and no bathochromic 31
Table 3
VISIBLE SPECTRAL ANALYSIS OF ANTHOCYANIDINS FROM FRACTION B
PIGMENT A MAXIMUM A MAXIMUM A MAXIMUM A MAXIMUM METHANOLIC ALUMINUM REFERENCE REFERENCE .1% HC1 CHLORIDE (31) (31) METHANOLIC ALUMINUM .1% HC1 CHLORIDE
DELPHINIDIN 546 nyuu 569 vyn. 546 nyix. 569 n\JLL
CYANIDIN 535 vyco 555 nyU. 535 m/LL. 553 nytX-
P ELARGONIDIN 520 mJUL 520 nyxr 520 vyUL 520 nytX.
All data reported was obtained from a Beckman DB Spect rophotometer• 32 shift occurred on the addition of aluminum chloride, Cyanidin showed a maximum absorption of 535 millimicrons and a bathochromic shift of 20 millimicrons when aluminum chloride was added.
Infrared Spectrum of Cvanidin From Preparative Paper Chromatography
After identifying the anthocyanidins produced from the leucoanthocyanidins of tannin fraction B by paper chromatography and visible spectral analysis, an attempt was made to obtain an infrared spectrum of cyanidin since a commercial sample was available for comparison. The commercial sample showed traces of impurities on paper chromatography, so it was purified by precipitation with lead acetate and extraction with isoamyl alcohol. The material thus obtained was chromatographed on paper in formic acid- 3N jHCl (lsl) and showed a higher degree of purity. The material obtained on evaporation of the isoamyl alcohol was - used to make a KBr pellet of the commercial sample.(Figure 10) A similar pellet was made from the cyanidin extracted from the preparative paper chromatography of the anthocyanidins obtained from Fraction B, (Figure 11) These spectra showed excellent correlation in all regions of the infrared from 2 to 15 microns. It was noted in making infrared spectra of cyanidin, both commercial sample and 4000 3000 ••• •1• • • •1
7 8 9 10 11 12 13 14 15 WAVELENGTH (MICRONS)
Figure 10, Infrared Spectrum of Cyanidin Chloride, Commercial Sample
03 4000 3000 1000 900 800
7 8 9 10 11 WAVELENGTH (MICRONS)
Figure 11. Infrared Spectrum of Cyanidin Chloride, Synthesized from Fraction B * 35 natural product material showed short absorption in the 5,9-6.0 micron regions.
Degradation by Potassium Fusion of Anthocvanidins
In order to substantiate further chemically the structure of the anthocyanidins, the anthocyanidins were degradated by fusion with potassium in the following manner (32 ). The anthocyanidins, obtained from Fraction B by boil ing Fraction B in ethanolic 10% HC1 for one-haIf hour, were purified by the method of lead salt precipitation pre viously outlined, and extracted with isoamyl alcohol. The isoamyl alcohol was evaporated, yielding a dark red semi- crystalline powder. One part of this powder was thoroughly mixed in a platinum vessel with 20 parts of potassium hydroxide pellets in the presence of a small amount of water. The mixture was quickly heated and agitated, whereupon the decomposition of the salts occurred at 250° C. After heating for a few minutes, the reaction mixture was rapidly cooled and dissolved in water. After acidification with a slight excess of hydrochloric acid, the solution was shaken with ether. The ether extract was shaken with saturated solution of sodium bicarbonate to remove acidic fragments, whereby phenolic fragments remained in the sodium bicarbonate solution. The sodium bicarbonate solution was acidified and extracted with ether which extracted the acidic fragments. Both the phenolic 36 and acidic portions on evaporation yielded yellow-brown semi-solid residues on evaporation of ether. From the three fundamental anthocyanidins contained in the mixture the corresponding degradation products were obtained as shown in Figure 12.
Paper Chromatography Identification of Fragments
Commercial samples of gallic acid, protocatechuic acid, p-hydroxybenzoic acid and phloroglucinol were obtained, and chromatography of degradation products of potassium fusion and commercial samples was carried out in the following three solvent systems: 1) lwButanol, Glacial Acetic Acid,Water(4:1:5)(Upper phage) 2) Phenol, Water (9:1) 3) 5J/o Glacial Acetic Acid 5% Ferric Chloride solution was used as the chroma® genie spray to identify the fragments produced on Whatman No. filter paper. The results of the paper chromatography of the products of potassium fusion are shown in Table 4.
Summary of Phvtochemical Investigation of Fraction B
Fraction B of the tannin extract has been identified as consisting of condensed, polymeric leucoanthocyanidin units. The presence of the leucoanthocyanidin units was demonstrated chemically by the conversion of the leucoantho°> cyanidins to the corresponding anthocyanidin pigments. 37
PHENOLS ANTHOCYANIDIN ACIDIC FRAGMENTS
HOOCo
PELARGONIDIN p-HYDROXY BENZOIC ACID
*Cr HOOG
PHLOROGLUCINOL. CYANIDIN PROTOCATECHUIC ACID
HOOC-^ VoH N=SDH
DELPHINIDIN GALLIC ACID Figure 12. Degradation of Anthocyanldins by Potassium Fusion * Dotted lines indicate cleavage points. 38
Table 4
Rf VALUES OF COMMERCIAL SAMPLES AND FUSION FRAGMENTS
SOLVENT SYSTEM SOLVENT SYSTEM SOLVENT SYSTEM No* 1 No. 2 No. 3 BAW (4:1:5)* Phenol-Water 5% Acetic Upper Phase (9:1) Acid
SAMPLE Rf SPOT COLOR Rf SPOT COLOR Rf SPOT COLOR
Gallic Acid 0.73 Blue Grey 0.13 Blue Grey 0.48 Blue Grey Fusion Fragment 0.73 Blue Grey 0.13 Blue Grey 0.48 Blue Grey
Protocatechuic 0.89 Green 0.45 Green 0.58 Green Acid Fusion Fragment 0.89 Green 0.45 Green 0.58 Green
p-Hydroxy Benzoic Acid 0.95 Yellow 0.77 Yellow 0.68 Yellow Fusion Fragment 0.95 Yellow 0.77 Yellow 0.68. Yellow
Phloroglucinol 0.77 Light Grey 0.28 Light Grey 0.65 Light Grey Fusion Fragment 0.77 Light Grey 0.28 Light Grey 0.65 Light Grey
Spray Reagent: 5% FeCig *l-Butanol-Glacial Acetic Acid»Water 39
The chemical identification of cyanidin, delphinidin, and pelargonidin was demonstrated by visible spectral analysis, infrared analysis, paper chromatographic identification, and fragment analysis by potassium fusion degradation. The condensed portion of the tannin fraction thus consists of leucocyanidin (5:7:3':4'-tetrahydroxyflavan-3:4- diol), leucodelphinidin (5:7:3*:4•: 5 1-pentahydroxyflavan- 3:4-diol), leucopelargonidin (5:7:4•-trihydroxyflavan-3:4- diol), in polymeric form. The structures of the leucoanthocyanidins differ only in the hydroxylation patterns of Ring B. The basic structure is shown in Figure 13, along with the structures of the anthocyanidins. The anthocyanidins are all polyhydroxy derivatives of 2-phenylbenzopyrilium or flavylium cation. Cyanidin is (3:5:7:31:4*-pentahydroxyflavylium), delphinidin is (3:4:7:3*:4*^'-hexahydroxyflavylium) and pelargonidin is (3:5:7:4*-tetrahydroxyflavylium).
General Chemical Analysis of Fraction A
Fraction A has been previously generally classified a$ consisting of monomers, including flavan-3:4-diol. (See page 21.) Three monomers were apparent at Rf values 0.54, 0.64 and 0,72 in paper chromatography of the original tannin extract in 1-butanol-glacial acetic acid-water (4:1:5) solvent system. 40
BASIC STRUCTURE OF LEUCOANTHOCYANIDINS
Leucocyariidin R-ft'= OH R" = H Leucodelphinidin R-R*-R" = OH
Leucopelargonidin R1 = OH R-R" = H
ANTHOCYANIDIN STRUCTURES
Delphinidin (3:5:7:3*:41:5hexahydroxyflavylium) Chloride •OH _
Cyanidin (3:5:7:3*:4*-pentahydroxyflavylium) Chloride CI"
Pelargonidin (j :5:7:4*-tetrahydroxyflavylium) Chloride
Figure 13. Basic Structures of Leucoanthocyanidins and Anthocyanidins From Fraction B 41
Since Fraction A gave general qualitative tests on paper for leucoanthocyanidins, it was decided to convert the monomeric leucoanthocyanidins to their corresponding antho cyanidins, and identify them chemically as had been done previously with the condensed fraction.
Conversion of Fraction A to Anthocvanidins
One gram of the golden brown powder obtained from the evaporation of Fraction A was dissolved in 300 ml# of ethanol to which had been added 3 ml* of concentrated HC1, The solu tion was boiled for one-half hour, and the color changed from an orange to a deep red color. Upon evaporation of the ethanol, 0.5 Gn, of a deep red semi-crystalline material was obtained. This material responded to the qualitative tests for anthocyanidins as shown in Table 1, and was subse quently purified by the method of lead salt precipitation as outlined in Figure 9,
Paper Chromatography of Anthocvanidins of Fraction A
The anthocyanidins produced from Fraction A were then chromatographed on Whatman No, 1 filter paper in three different solvent systems. Solvent System No, 1 consisted of formic acid-3N HC1 (1:1); Solvent System No, 2 consisted of water-acetic acid-concentrated HC1 (10:30:3); Solvent System No, 3 consisted of acetic acid-concentrated HCl-water (5:1:5). Each solvent system showed the presence of three anthocyanidin pigments and the Rf values were identical to 42 those obtained with the anthocyanidins developed from the condensed tannin fraction, (See Table 2). The Rf values in the three solvent systems used, corresponded to the values of those listed (26,27) for cyanidin, delphinidin, and pelargonidin.
Structure Proof of Anthocyanidins of Fraction A
In order to further substantiate the structures of the anthocyanidins produced from Fraction A, it was necessary to repeat the preparative procedure of paper chromatography outlined on page 29. After extracting the three different anthocyanidins from the preparative papers, visible spectral analysis was identical to that shown in Table 3, The visible spectral analysis showed that the pigments corresponded to cyanidin, delphinidin, and pelargonidin.
Potassium Fusion of Fraction A
One-half gram of a mixture of the three anthocyanidins from Fraction A was fused with potassium hydroxide pellets as outlined on page 35. The products of the degradation were then chromatographed in three different solvent systems. The solvent systems used to identify the fragments by paper
chromatography were: 1-butanol-glacial acetic acid«»water (4:1:5) upper phase, phenol-iwater (9:1),5% glacial-acetic acid.. The fragments obtained from the fusion were chromatographed with commercial samples of gallic acid, protocatechuic acid, p-hydroxy benzoic acid, and phloroglucinol. In all three 43 solvent systems, the Rf values of the fragments obtained from fusion of anthocyanidins corresponded to those of the commercial samples used* The Rf values obtained were identical with those shown in Table 4. These fragments were those that would be obtained from cyanidin, delphinidin, and pelargonidin.
Isolation of Leucocvanidin from Fraction A
A small glass column, one inch in diameter and suit able for chromatography, was packed to twelve inches with polycaprolactam powder, (S) Farbwerke Hoechst. To the top of the column was added 1 Gm. of the golden brown powder representing Fraction A, Absorbent cotton was placed above the added golden brown powder. Ovqr a period of six houps,
100 ml«of 20% acetic acid were added to the column. Three distinct zones appeared visibly after the addition of the acetic acid. A half-inch long chocolate brown zone appeared near the top of the column. One inch below, a very small orange band appeared, and about one-eighth inch below this band appeared another light orange band. The column was carefully extruded from the glass and the individual bands carefully cut from the polycaprolactam and extracted with methanolic 1% HC1, Upon evaporation the chocolate brown spot near the origin of the column yielded 0.75 Gm. of a deep golden brown powder. The second zone yielded 0,1 Qn, of golden brown scaly material, and the third zone yielded 0,1 Gm, of light orange scales. 44
An infrared spectrum was then made from the golden brown crystalline powder extracted from near the origin of the polycaprolactam column. The infrared spectrum showed evidence of hydrogen bonding in the phenolic OH area from 2.8 to 3,2 microns. There was a lack of strong carbonyl absorption and the other bands appeared at 6,2, 6.7, 9.7, 11.0, 13.1 and 14,3 microns. This spectrum indicated that the material was phenolic, aromatic, and with hydrogen bonding of the OH groups. See Figure 14. After examination of the infrared spectrum, it was decided to attempt to convert the golden brown material to an anthocyanidin by the method previously used for this purpose. A 200 milligram sample of the material was boiled
in ethanolic 10% HC1 for 30 minutes. The solution turned from a golden orange to a deep red color. The ethanol was evaporated, leaving a red colored semi-crystalline residue. An infrared spectrum was made of the material, using a potas« sium bromide pellet. The infrared spectrum obtained was identical with the one previously made from a commercial sample of cyanidin. (Figure 10) Since the material was convertible to cyanidin, it was believed that the original material extracted from the column was leucocyanidin. In order to prove that the material from the column was leucocyanidin, 0.2 Gm. of the material was fused with potassium hydroxide pellets as described on page 35. Since the only fragments obtained from the fusion 4000 3000 1000 900
7 8 9 10 11 12 13 14 15 WAVELENGTH (MICRONS)
Figure 14. Infrared Spectrtun of Monomeric Leucocyanidin
•P* Ul 46 were protocatechuic acid and phloroglucinol, this evidence verified that the material extracted from the colvunn was leucocyanidin.
Summary of Chemical Analysis of Fraction A
The presence of monomeric leucoanthocyanidins, leucocyanidin, leucodelphinidin, and leucopelargonidin in Fraction A was demonstrated by conversion of the leuco anthocyanidins to their corresponding anthocyanidins and then isolating these anthocyanidins by preparative paper chroma tography, Visible spectral analysis, paper chromatography in three solvent systems, and degradation by potassium hydroxide J:usion were used to identify the anthocyanidins. The data obtained corresponded to that previously obtained for cyanidin, delphinidin, and pelargonidin, whose structures were proved in the same manner as for Fraction B. The monomeric leucoanthoCyanidin, leucocyanidin was isolated from a polycaprolactam powder (S) column chromatography... The structure of leucocyanidin was proved by infrared analysis and degradation by potassium hydroxide fusion. Further proof of leucocyanidin was accomplished by conversion to cyanidin and identification of this pigment in three solvent systems, by paper chromatography. 47
Theoretical Implications of Phvtochemical Analysis of Tannin Fraction
Examination of the logs of Quebracho wood (Schinop- sis sp.), which is the source of one of the most widely- used tannin extiracts, revealed the presence of monomeric leucofisetinidin. (33,34,35) Roux (35) considers this compound to be the most important precursor and prototype of Quebracho tannins which were estimated (35) to contain from 20 to 40 percent of polymeric leucofisetinidins. With respect to coniferous tannins, Bate-Smith and Swain (36) reported that maritime pine wood (Pinus maritimus) leuco- anthocyanidins yielded cyanidin upon treatment with mineral acid, Hillis (37) subsequently reported that tannins from Pinus radiata bark yielded cyanidin and traces of delphinidin upon similar acidic treatment. Other than noting the forma tion of anthocyanidin, these investigators did not further characterize the leucoanthocyanidins. Monomeric and poly meric leucoanthocyanidins have been carefully differentiated by Roux and co-workers. (38) Based on the work of previous investigators outlined above, the characterization of monomeric leucoanthocyanidins in the water soluble ethyl acetate extractable fraction of Rumex Hymenosepalus seems logical. The chemical characteri zation of these monomers has verified the presence of leucocyanidin, leucodelphinidin, and leucopelargonidin as monomers in the same tannin that contains polymers consisting of the same units. 48
It has been presumed by many workers that the term leucoanthocyanidin (sugar free) and leucoanthocyanin (sugar containing) refer to the monomeric C-15 molecule (39). ((I),Figure 15, page 53) This presumption followed from the fact that compounds of proven structure akin to (I) have been isolated from many sources and shown to yield flavylium salts on boiling in mineral acid solution.(40) It is understood, however, that the substances present in plants which give this reaction may be not only monomers, but polymers whose structure at present is unknown. These polymers may be of any size and could be formed by condensa tion of leucoanthocyanidins alone, or by co-polymerization with other phenolic nuclei (44).' Whatever the structure of the polymers, they would be defined as leucoanthocyanidins, provided they yielded an anthocyanidin on heating in the presence of mineral acid. Although various investigators (42) agree that a large majority of condensed tannins are primarily derived through polymerization of flavan-3-ols and flavan-3;4-diols, the structure of these polymers, i.e., the mode of linkage between the flavanoid units, has not yet been ascertained. In general, there are three structural theories: 1) Condensation by acid. 2) Enzymic condensation through a quinone polymeri zation mechanism.
3) An unspecified mechanism to give polymers with ether linkages. 49
Condensation by Acid
The Heidelberg school represented by Freudenberg (43), Mayer (44) and others have maintained that the catechin tannins are formed in nature, just as in the laboratory, by the post mortem action of acids on catechins. Two mechanisms have been proposed: (Roman numerals refer to structures shown in Figure 15, pages 53 and 54,)
1) In one of these, the basic structure (II) may react as a pair of tautomeric diphenylpropenes (III) and (IV) to form a polymer (V) similar to that involved in the first stage of styrene polymerization. 2) In the second of these, the molecule reacts bifunctionally in the presence of hydrochloric acid, i.e., electrophilic at carbon atom ntamber two of a molecule in which ring fission has given a secondary benzyl alcohol (VI) and nucleophilic at either carbon atom six or eight (VII). These two molecules
condense to form a dimer (VIll) capable of further polymeri zation.
Enzvmic Condensation Through a Quinone Polymerization Mechanism
Auto-oxidative and enzymic polymerization of flavan- 3-ols and flavan-3:4-diols, has been studied by Hathway (45). As a result an alternative hypothesis to the Freudenberg theory has been suggested. Catechin was auto-oxidized in phosphate buffer (pH 8) to give a polymeric product charac terized by elemental analysis and ultraviolet absorption 50 spectra. From a study of the oxygen balance and the obser vation that hydrogen peroxide was accumulated, it was assumed that quinone formation was involved. Since the spectrum differed from that of the auto-oxidation product from pyrocatechol, it was assumed that a head-to-tail polymer (IX) had been formed. Gallocatechin, leucodelphinidin and catechin were aerobically oxidized (46) in phosphate buffer (pH 7) and oak cambium polyphenoxidase to give a polymer with an ultraviolet spectrum similar to an isolated polymeric tannin from oak bark. The spectra were also similar to the oxidation products of catechol and 5-methylpyrogallol, so a tail-to-tail quinone polymerization mechanism was invoked (X). Since the gallo- catechin metabolite was oxidized more rapidly than the other metabolites, it was concluded that oak bark tannin is principally formed by aerobic oxidation in the cambium of the gallocatechin.'
Polymers with Ether Linkages
In the third general structural theory, the linking of flavanols or other phenols through ether linkages has been proposed. Kirby and White (47) examined the ultraviolet spectra of acetate derivatives of Quebracho tannin fractions and have concluded that approximately one-half of the oxygen atoms were occupied in ether linkages. In his most recent work9 Roux (48) has suggested that the mode of condensation is such that the flavan-354-diol structure remains intact. 51
Ether links between flavanoid units through the four position of leucoanthocyanidins to the six or eight position of some other or similar flavanoid nuclei were considered to be a likely mode of ether linkage* Several lines of investigation have suggested linkages such as (XI1) and (XI) as the most likely mode of combination of the flavan—354-diols in pine bark tannin (49)e Whether the structures (XI) and (XII) are generally applicable to all condensed tannins derived from flavan-3-ols and 3«!4-diols remains to be established. The condensation by acid theory has been criticized by the fact that the acid catalyzed reaction requires a low pH (^2) and a high temperature 0 50° C). Neither of these conditions are present in the plants. Regarding the postu- lation of enzymic condensation through a quinone polymeri zation mechanism, the presence of quinone groups in condensed tannins has not yet been demonstrated. Quinones (especially o-diquinones) show a strong carbonyl stretching band at 1650-1685 cm-1 in their infrared spectra; a band that is lacking in the infrared spectra of naturally occurring con densed tannins to date. Since the flavan-3;4-diol structure remains intact when the condensed polymer is converted to monomeric anthocyanidins, it seems probable that the actual mode of structural linkage of the condensed polymeric tannins favors the ether linkage hypothesis„ As reported in this work, the infrared spectra of the polymer is very similar to 52 that of the monomer, and both show an absence of any strong carbonyl absorption or quinone absorption. Continued infrared and ultraviolet spectral analysis should yield a further insight into this problem. 53
CONDENSATION BY ACID MECHANISM
(I)
•HO, a.
(II) (Ill)
—'X (V) (IV)
OH
(VIII) Figure 15. Structures Illustrating Types of Condensation 54
QUINONE POLYMERIZATION MECHANISM
(IX)
POLYMERS WITH ETHER LINKAGES
(XI) (XII)
Figure 15 (Coi^td.) ANTI-TUMOR TESTING
History of Testing
The investigation of Rumex Hvmenosepalus emanates from the anti-tumor properties located in the aqueous- ethanol-methanol tannin fraction of the roots and tubers.
The anti-tumor activity of the initial and subsequent extracts was determined by the Cancer Chemotherapy National
Service Center, National Cancer Institute, Bethesda,
Maryland.
The preliminary extracts of Rumex Hvmenosepalus were prepared and submitted to the screening center by
Dr. Mary Caldwell, et al.. of the University of Arizona,
College of Pharmacy. Botanical identification was confirmed by Robert Barr, Research Associate at the University of
Arizona. This was done as part of a project to screen plants indigenous to Arizona and the Southwest for their possible anti-tumor properties.
Due to the positive confirmation received from the testing center on Rumex Hvmenosepalus. further fraction studies were pursued both from a purely phytochemical viewpoint, and if possible, to isolate the anti-tumor constituent present in the plant.
55 56
Examination of Five Different Varieties for Anti-Tumor Activity
The roots and tubers of five different varieties* of Rumex Hvmenosepalus were extracted with methanol and the extracts obtained were submitted for anti-tumor testing. The varieties were labelled I, II, III, IV, and V„ The results of the testing are listed in Table 5. In order to determine if the varieties affected the anti-tumor activity, a chemical investigation of the contents of the methanol extracts was undertaken. All five extracts were spotted on Whatman No. 1 filter paper and developed in 1-butanol-glacial acetic acid-water (4:1:5), descending chromatography, tank saturation. All of the extracts showed the presence-of monomeric and polymeric leucoanthocyanidins, previously determined by paper chromatography (Figure 4). In addition, all the methanol fractions showed the presence of yellow pigments near the solvent front, Rf 0.90. The presence of non-polar materials in the methanol extracts was demonstrated by extracting each methanol extract with chloroform, resulting in a yellow solution, and then spotting these yellow solutions on silica gel plates and developing in chloroform. Each methanol extract from the
*Refers to selections of Rumex Hvmenosepalus based on tannin content as collected by Norris Gilbert, University of Arizona Agricultural Experimental Station, Mesa, Arizona. 57 different varieties showed the presence of the same non-polar materials as shown on thin-layer plate in Figure 17, page 73 •
Preliminary Extracts
In an attempt to locate the anti-tumor agent, the following extracts were submitted for anti-tumor testing, 1) Methanol-ethanol-water extract (Tannin Fraction) 2) Fraction A (Water soluble ethyl acetate extractable material obtained from original tannin extract.) 3) Fraction B (Water soluble, non-ethyl acetate extrac table material obtained from original tannin extract.) 4) Ether extract (Corresponds to non-polar constituents of roots and tubers.) 5) Five different varieties of methanol extracts. 6) Defatted water extract (Corresponds essentially to tannin fraction, but emphasis placed on completely removing any non-polar materials.) The results of testing the extracts listed above are shown in Table 5.
Summary of Anti-Tumor Testing
An analysis of the active fractions submitted reveal that the anti-tumor activity lies in the tannin fraction. It is also apparent that non-polar constituents 58 which may be in the tannin extract due to incomplete defatting are not responsible for the activity. Fractionation of the original tannin fraction shows t that the activity is maintained in the part of the tannin fraction corresponding to the polymeric or condensed portion. Removal of the monomers by ethyl acetate extraction showed that this fraction was not responsible for anti-tumor activity, and did not enhance the activity of the extract. The anti-tumor activity is retained in the condensed portion of the tannin extract. The ethyl acetate extractable fraction showed an absence of anti-tumor activity, indicating that the monomeric leucoanthocyanidins in the tannin extract are inactive, Methanolic extracts of five different varieties of Rumex all showed anti-tumor activity, with one fraction showing a better activity than the others, (Methanol Extract V), Subsequent chemical investigation of the methanol extracts showed the presence of orange-yellow non-polar constituents in all of them, which may account for the variability of activity, and in only one case, Methanol Extract V, was the anti-tumor activity greater than that shown by the original tannin extract. The methanolic extracts all showed the presence of monomeric and polymeric leuco anthocyanidins. 59
An analysis of these results indicates that the anti-tumor activity is due to the fraction of the tannin extract previously chemically characterized as consisting of polymeric condensed materials, of which the units present have been identified as leucocyanidin, leuco- delphinidin, and leucopelargonidin. A study of the exact method of condensation of the polymeric units would probably give a deeper insight into the anti-tumor activity. 60 Table 5 ANTI-TUMOR TEST RESULTS
EXTRACT TUMOR WT. DOSAGE SURVIVORS RESULTS & DATE LOSS RANGE % T/C
Fraction A 400 0/4 wcacs 1/4/'65 --- 200 0/4 MDDO
www 100 0/4 www
Fraction A «a was 100 0/4 www 3/25/'65 « 65 0/4 — 45 0/4 -1.9 30 4/4 79 Fraction B 600 0/4 1/4/'65 400 0/4 —
200 0/4 ———
WWW 100 3/3 26
Fraction B www 100 0/4 rnaca l/ll/»65 -0.7 65 2/4 -1.7 45 4/4 93 -1.0 30 4/4 54
Defatted 400 0/4 — Water Extract ... 200 0/4 OBsaai 12/23/*65 --- 100 0/4 www •2,5 50 4/4 45 61
Table 5-Continued
ANTI-TUMOR TEST RESULTS
EXTRACT TUMOR WT. DOSAGE SURVIVORS RESULTS & DATE LOSS RANGE % T/C
Ether 400 0/4 Extract 3/24/»66 200 0/4
100 0/4 MM*
• mmm 50 0/4
Ether 50 0/4 83 Extract 4/12/'66 35 0/4 23 0/4 -0.2 15 4/4
Methanol 400 0/4 ••N I 200 0/4 100 0/4
-4.2 50 2/4 ataan
-3.5 50 4/4 77 -3.7 35 4/4 59 -2.1 23 4/4 109
-1.8 15 4/4 84 Methanol 400 0/4 II 200 0/4
100 0/4 50 0/4
-4.3 50 3/4 40 62
< Table 5-Continued ANTITUMOR TEST RESULTS
EXTRACT TUMOR WT DOSAGE SURVIVORS RESULTS & DATE LOSS RANGE % T/C
Methanol -3.4 35 4/4 89 II (Continued) -2.5 23 4/4 92 -2.9 15 4/4 108 Methanol —2. 8 400 4/4 73 III 200 0/4
100 0/4 sacscB -3.8 50 2/4 ... 400 0/4 ... 200 4/4 ... 100 4/4 -3.0 50 4/4 ...
-2.3 50 0/4 sau -2.5 35 4/4 71 -1.2 25 4/4 • 47 15 4/4 89 Methanol ... 400 0/4 IV 200 0/4 100 0/4 -0.4 50 1/4 50 0/4 -3.5 35 4/4 33 63
Table 5-Continued
ANTI-TUMOR TEST RESULTS
EXTRACT TUMOR WT. DOSAGE SURVIVORS RESULTS & DATE LOSS RANGE % T/C
Methanol -3.0 23 4/4 88 IV (Continued) -1.4 15 4/4 104
Methanol MMM 400 0/4 V -6.8 200 2/4 — 100 0/4 o»oo«o -1.3 50 3/4 38 -1.6 50 3/4 18 -2.5 35 2/4 -2.8 23 4/4 75 -2.1 15 4/4 82
-2.6 50 2/4 — -2.9 35 4/4 61 -2.8 23 4/4 60 -2.5 15 4/4 77
*Tannin -11.0 30 6/6 76 Fraction - 3.0 15 6/6 84 120 0/6 -15.0 60 . 6/6 35 64 Table 5-Continued
ANTI-TUMOR TEST RESULTS
Code to Tables of Anti-Tumor Activity
Dose: In mg,/Kg.
Survivors: Number of animals surviving out of number started on test.
Tumor Weight: The mean tumor weight of the test animals, T The mean tumor weight of control animals, C
Per cent T/C: Ratio of tumor weight or survival time of test animals to that of control animals; expressed as %•
Test System: Those marked with asterick * --WM Walker 256 (Intramuscular) All others -- SA Sarcoma 180
Vehicle: Water, unless otherwise specified.
Results of Testing Furnished by: Cancer Chemotherapy National Service Center National Cancer Institute Bethesda, Maryland NON-POLAR EXTRACT
Preliminary Evaluation The isolation of chrysophanic acid and physcion from Rumex Hvmenosepalus was reported previously (1), It was reported that sliced vacuum dried tubers of at least two different strains of Rumex Hvmenosepalus were extracted with benzene to give 1.2% and 0,3% of chryso phanic acid, and physcion respectively. Although the possible presence of emodin has been reported previously, no trace of it could be found (2). It was noted that the tubers containing the greatest amounts of yellow deposits had the less vigorous growth (50). In the extraction procedure developed for extracting the tannin fraction, it will be noted (Figure 1), that at various stages in the extraction procedure, petroleum ether, ethyl ether and chloroform were used to free the tannin fraction of non-polar material. The possibility occurred that residual amounts of non-polar material which had not been removed from the tannin fraction during the defatting process, were responsible for the variability in anti-tumor activity. It was then decided to broaden the scope of the phytochemical investigation to the non-polar fraction, to determine if this fraction was responsible for the anti-tumor activity.
65 66
Paper Chromatography of Non-Polar Fraction
The petroleum ether extract and the ether extract which had been collected from the defatting process of the tannin fraction were first exposed to paper chromatography on Whatman No. 1 filter paper, using 1-butanol-glacial acetic acid-water (4:1:5) as the solvent system, upper phase only, using descending chromatography. The results of the paper chromatographic investi gation showed only a large visible yellow spot at Rf 0.90 in the petroleum ether extract, and in the ether extract, a large yellow spot at Rf 0.90 visible and three ultra violet visible spots at Rf 0.54, 0.73 and 0.84. The large yellow spots in both cases turned pink-red when exposed to ammonia fumes.
Extraction Procedure
Since paper chromatography of the petroleum ether and ether extracts saved from the defatting process showed residual amounts of polar materials, it was decided to develop an extraction procedure which would be completely free of any polar or tannin materials. An extraction scheme utilizing solvent fractionation was developed for this purpose (Figure 16). A new batch (600 Qm.) of the ground roots and tubers was air dried for twenty-four hours, and then extracted with 67 one liter of ether. The resultant ether extract was yellow- green. To this yellow-green solution was then added an equal quantity of water in a separatory funnel, and it was shaken continuously for one-half hour. The ether phase was separated from the aqueous phase. The resulting ether solution was evaporated to dryness. The result was 4 Gm. of an amorphous fragrant yellowish extract. To this extract was added one liter of chloroform. Once again an equal quantity of water was added to the chloroform solution and shaken in a separatory funnel for one-half hour. The chloroform phase was separated from the aqueous phase and then evaporated to dryness. The residue (5 Gm.) was an orange-yellow highly fragrant amorphous solid, A portion of this amorphous residue was dried and submitted for anti-tumor testing. The result of the anti-tumor testing of this fraction is in Table 5 (page 64), The amorphous yellow-orange residue obtained by the extraction procedure outlined on page 68 will be referred to as the non-polar extract.
Paper Chromatography of Chrvsophanic Acid. Phvscion and Bnodin
Since anthraquinone pigments had been previously reported to be present in non-polar extracts of the roots and tubers, it was decided to try to identify them by paper chromatography. Successful paper chromatography separation of anthraquinone pigments has been reported by Shibata, et al. (51). 68
DRIED ROOTS AND TUBERS
ETHER EXTRACT (Yellowish-Brown) 1) Add water
WATER SOLUTION ETHER SOLUTION (Yellowish-Orange) 1) Evaporate
DRY RESIDUE 1) Add chloroform 2) Add water
WATER SOLUTION CH LOROFORM (Light Brown) 1) Evaporate
AMORPHOUS RESIDUE (Yellow-Orange) Fragrant
Figure 16. Extraction Procedure for the Non-Polar Extract 69
The solvent system used for the paper chromatography on Whatman No. 1 filter paper was (Petroleum ether, b.p. 30° - 60° C.) saturated with 977» methanol. A chloroform solution of the non-polar extract was spotted at the top of the paper and descending chromatography was used in tanks 60 cm. high and 30 cm. in diameter. The develop ment was carried out in a well-sealed tank which was cooled by ice. The chromatogram was developed for eight hours. After air drying, the developed paper was sprayed with 0.5% methanolic magnesium acetate solution and heated in an oven at 90° C. for five minutes.
After spraying, the following spots appeared: at
Rf 0.52 a large spot that was yellow turned pink; at Rf 0.89
a small spot that was yellow turned orange; at Rf 0.92 a spot that was yellow turned orange. The Rf values and the
colors developed by the chromagenic spray corresponded to
those obtained by Shibata, et al. (51), Rf 0.52 (Emodin), Rf 0.89 (Physcion) and Rf 0.92 (Chrysophanic Acid).
After identification of these three pigments, the paper chromatography was repeated, using commercial samples
of chrysophanic acid and emodin, which were chromatographed
with the non-polar extract and using quinizarin as an Rf
reference. The commercial samples used were obtained from
K & K Labs, Plainview, New York. Upon repetition of the
paper chromatography with commercial samples and development
of the paper after spraying with 0.5% methanolic magnesium 70 acetate, it became apparent that the colors developed and the Rf values of the spots of the non-polar material corresponded exactly to those of chrysophanic acid, emodin and physcion. Since quinazirin has the same Rf value (0.89) as physcion, in the system used, it was a verification of the presence of physcion, as they both had the same Rf value, but developed different colors on spraying. Physcion turned from yellow to orange and quinizarin turned from pink to purple, (Table 6) This experiment verified the presence of chrysophanic acid, physcion and emodin in the non-polar extract, but since only one solvent system was used, it was decided to utilize thin-layer chromatography on silica gel for further investigation of the non-polar extract and additional evidence for the presence of the three pigments identified on paper.
Preliminary Thin-Layer Chromatography
The yellow-orange residue from the non-polar extraction procedure was exposed to exploratory thin-layer chromatography on silica gel. The solvent used was chloroform, employing ascending chromatography in saturated tanks. Ceric sulfate 2% was used as a chromagenic spray, and the developed plates were also examined under ultraviolet light. Thin-layer chromatography showed the presence of seven separate materials in the non-polar extract. These - ~ 71 Table 6 PAPER CHROMATOGRAPHY OF NON-POLAR EXTRACT
Material Rf Rf in Color with Visible Color literature .5% ethanolic (50) magnesium acetate
Unknown 0.52 Pink Yellow
Emodin 0.52 0.52 Pink Yellow (Commercial)
Quinazirin 0.89 0.89 Purple Pink
Unknown 0.89 Orange Yellow
Unknown 0.92 Orange Yellow
Chrysophanic Acid (Commercial) 0.92 0.92 Orange Yellow
Paper: Whatman No. 1 Filter Paper Solvent: Petroleum Ether (b.p, 30° - 60° C.) saturated with 97% methanol Temperature: 25° - 30° C. Spray: 0.5% Methanolic Magnesium Acetate 72 seven materials showed the presence of orange-yellow pigments both at low Rf and high Rf values (Figure 17). Examination of the plates under ultraviolet light revealed the presence of three fluorescent materials in addition to the visible orange materials, and spraying with 2% eerie sulfate and heating the plate with an open flame after drying, showed the presence of tw
Preparative Thin-Layer Chromatography
In order to isolate the material which was separated by thin layer, twenty preparative thin-layer plates were made. A mixture of 350 Gm. of Silica Gel G according to 73
SOLVENT FRONT
C^> 3 2 1 ORIGIN All numbers above are spot: numbers listed in Table 7, page 74. Figure 17• Thin-Layer Chromatography of Non-Polar Extract 74 Table 7 THIN-LAYER CHROMATOGRAPHY OF NON-POLAR EXTRACT 2% Spot No. Visible Color Ultraviolet Color Ceric Sulfate Reagent Color 1 Orange-Yellow Bright Orange Grey-Green 2 Light Yellow Bright Blue-Green Blue 3 Colorless Light Grey 4 Orange-Yellow Bright Orange Grey-Green 5 Violet 6 Light Grey Light Green Grey 7 Light Yellow Blue Violet SOLVENT SYSTEM: CHLOROFORM CHROMATOGRAPHIC MEDIA: SILICA GEL LAYERS TANK SATURATION, ASCENDING CHROMATOGRAPHY METHOD: Material spotted across bottom of 20 x 20 cm. preparative plate. 75 Stahl, per 750 ml. of water was spread one millimeter thick by a Desaga applicator by Brinkman, on plates 20 x 20 cm. These plates were activated by heating in an oven at 130° C. for one hour. At one inch from the base of the plates, they were spotted with the non-polar extract from an ether solution. The plates were developed utilizing chloroform as the solvent system and ascending chromatography in glass tanks. The plates were developed one inch from the top and allowed to dry. A strip, one inch from the side of each plate, was sprayed with 2% eerie sulfate and heated lightly with an open flame. The seven spots shown in Figure 17 were apparent. In isolation procedures the areas which fluoresced under ultraviolet light and gave a positive eerie sulfate reaction were carefully marked and scraped from the prepara- tive plates. To assure that materials were single sub stances, the extracted spots were re-chromatographed and resprayed. If contaminated with other than one material, they were re-chromatographed until purity was assured. All materials from thin-layer preparative chromatography used in chemical identification procedures were purified in this manner. 76 Isolation of Chrysophanic Acid and Phvscion The orange material representing spot no. 4 on the preparative thin-layer plates was carefully removed from the preparative plates under ultraviolet light. This spot showed a bright orange fluorescence in the ultraviolet light. The material was placed in a small flask and extracted with ether, and then chloroform. The yellow solution was then filtered and evaporated to dryness. Evaporation of the solvents resulted in a small amount (30 mg.) of a bright yellow-orange semi-crystalline material. This material was re-extracted with ether and re-chromatographed on a preparative thin-layer plate. The resultant spot appeared to be a single component, but on spraying with eerie sulfate the presence of two materials became evident. An ether solution of the re-chromatographed material was then spotted once again on another preparative thin-layer plate. After testing various solvent systems, it was determined that benzene separated what appeared to be one single material into two distinct spots--a rather thick orange spot at Rf value 0.56 and another lighter orange spot behind the first spot at Rf 0,50. These materials were carefully removed from the preparative plate, and extracted separately with chloroform for 24 hours. The material from a Rf 0.56 yielded a bright 77 yellow-orange crystalline solid. The material from the Rf 0,50 yielded a yellowish orange crystalline solid. These two solids will be referred to as Compound A and B, respectively. Identification of Chrvsophanic Acid The yellow-orange crystalline material (A) was simultaneously chromatographed on thin-layer plates using silica gel with a commercial sample of chrysophanic acid, utilizing three different solvent systems. The commercial sample of chrysophanic acid used was obtained from K & K Labs, Plainview, New York, The following solvent systems were used: Solvent System No, 1: Chloroform Solvent System No. 2: Benzene Solvent System No, 3: Benzene-Glacial Acetic Acid(60s30) The Rf values of the material (A) and the commercial sample of chrysophanic acid were the same in all three solvent systems on the thin-layer plates, (Table 8) An infrared spectrum of material (A) was then made, using a potassium bromide pellet. The infrared spectrum of material (A) and that of a commercial sample of chryso phanic acid were identical over all wave lengths,(Figures 18, The material (A) from spot no. 4 of the preparative plate was then examined by visible spectral analysis in chloroform. This material showed a maximum absorption at 78 Table 8 Rf VALUES THIN-LAYER CHROMATOGRAPHY ON SILICA GEL LAYERS OF MATERIAL (A) AND COMMERCIAL SAMPLE OF CHRYSOPHANIC ACID SOLVENT SYSTEM Rf VALUE Rf VALUE MATERIAL (A) CHRYSOPHANIC ACID Chloroform 0.75 0.75 Benzene 0.56 0.56 Benzene-Glacial 0.90 0.90 Acetic Acid (60:30) 4000 3000 m li 7 8 9 10 11 WAVELENGTH (MICRONS) Figure 18. Infrared Spectrum of Chrysophanic Acid from Non-Polar Extract vO 4000 3000 900 800 700 ill .'jV.I'M II 4 5 6 7 8 9 10 11 12 13 14 15 WAVELENGTH (MICRONS) Figure 19. Infrared Spectrum of Chrysophanic Acid, Commercial Sample oo o 81 435 millimicrons which was identical with the absorption maximum of a commercial sample of chrysophanic acid in chloroform. Based on the identical Rf values in three different solvent systems on thin-layer chromatography, using silica gel; identical infrared spectra; identical absorption of both the natural product and commercial sample in the visible spectra; and previous identification by paper chromatography, it was concluded that the material (A) was chrysophanic acid (1,8-dihydroxy-6»methyl anthraquinone)• Identification of Phvscion The orange-yellow material (Spot B) which had an Rf value of 0.50 on thin-layer chromatography on silica gel, using benzene as a solvent system, was carefully cut from the preparative thin-layer plate. The material was extracted with ether and petroleum ether for 24 hours and then filtered, After several crystallizations from chloro form, the material was obtained as short, orange colored needles. The material was insoluble in water, soluble in alcohol, in petroleum ether, in chloroform and in benzene. It melted sharply at 207° C. It was insoluble in 5% hydrochloric acid, but soluble in 57o sodium hydroxide solution, forming a dark, red color in the latter. This color reaction is characteristic of anthraquinone pigments related to emodin. 82 An infrared spectrum was then made of the material, utilizing a potassium bromide pellet0 The material showed very close correlation with a commercial sample of emodin, except there is no carbony1 absorption in emodin, but the unknown material showed a weak carbonyl absorption at 5.7 microns. Since physcion is the 3-methyl ether of emodin it is felt that the carbonyl shows up in physcion, since hydrogen bonding is not as pronounced as with the 3-hydroxyl group, which is present in emodin. (Figure 20) Due to the fact that physcion was identified pre viously by paper chromatography and the melting point was exactly correct for physcion, it was decided to attempt to make a derivative to further verify the identity of the material. Acetate Derivative Ten milligrams of the material from Spot B were dissolved in 2,5 ml. of acetic anhydride and two drops of concentrated sulfuric acid were added and the mixture heated for a few minutes. It was then cooled and poured into 75 ml of water and allowed to stand overnight. The yellow solid which separated was re-crystallized from glacial acetic acid, and the yellow crystals which resulted had a sharp melting point of 187° C. The melting point previously reported for physcion diacetate was 186°<-187° C. (52) 4000 3000 2000 1500 CM*' 1000 900 800 700 O.u 0.0 .10 .20 £.30 O tfj.40 ,40 <50 .50 .60 .70 1.0, oo oo 3 4 5 6 7 9 12 13 14 15 WAVELENGTH (MICRONS) Figure 20. Infrared Spectrum of Physcion from Non-Polar Extract oo Co 84 As a result of previous identification by paper chromatography, identical melting point listed in literature, infrared analysis and correlation of diacetate derivative, the material from Spot B on the preparative chromatography plate was assumed to be physcion (3-methyl ether of emodin). Isolation and Identification of Emodin The preparative thin-layer chromatography on silica gel layers of the non-polar extract showed the presence of a visible orange pigment at low Rf value (Spot#l, Figure 17). The orange pigment gave a bright orange fluorescence in ultraviolet light. This orange pigment was carefully cut from twenty thin-layer preparative plates and extracted with ether and chloroform continuously for a period of 24 hours. The yellow solution was filtered and the solvents evaporated® A residue of orange-yellow semi-crystalline material resulted. Repeated crystallization from chloroform resulted in the appearance of orange-like crystals which gave a sharp melting point at 255° G. This is the melting point listed for emodin (1,3,8 trihydroxy-6-methyl anthraquinone). Since a commercial sample (K & K Labs, Plainview, New York) was available, the natural product was first chromatographed on thin-layer of silica gel, using three different solvent systems. The Rf values of the natural product and those of the commercial sample were identical in all three solvent systems used. (Table 9) 85 Table 9 Rf VALUES THIN-LAYER CHROMATOGRAPHY ON SILICA GEL LAYERS OF NATURAL PRODUCT FROM SPOT NO* 1> AND COMMERCIAL SAMPLE1'OF EMODIN SOLVENT SYSTEM Rf VALUE Rf VALUE NATURAL PRODUCT EMODIN SPOT NO. 1 (COMMERCIAL SAMPLE) Chloroform 0.20 0.20 Benzene-Methyl Formate Formic Acid(75-24-1) 0.57 0.57 Benzene-Glacial Acetic Acid (60:30) 0.82 0.82 * Refers to Spot Noa 1, Figure 170 86 An infrared spectrum, using potassium bromide pellet, was then made of the natural product (Figure 21), and also the commercial sample of emodin (Figure 22). The spectra were superimposable over all wave lengths examined in the infrared from 2.5 to 15 microns. Bands in both spectra were noticed at 2.90, 3.40, 6.15, 6.90, 7.30, 7.50, 8.25, 8„60, 9.10, 9.70, 11,00, 11.40, 12.15 and 13.20 microns. The natural product and the commercial sample were then exposed to visible spectral analysis in chloroform, using a Beckman DB Spectrophotometer. The curves obtained were once again superimposable with maximum absorption * occurring at 440 millimicrons in both samples. Based on the fact that emodin had been previously identified by paper chromatography (Table 6), accurate melting point, identical infrared and visible spectra of the natural product and a commercial sample of emodin, and identical Rf values of the natural product and a commercial sample of emodin on thin-layer silica gel plates using three different solvent systems, it was concluded that Spot No. 19 appearing as an orange pigment on preparative thin-layer chromatography on silica gel layers, was emodin (1,3,8-tri- hydroxy-6-methyl anthraquinone). 4000 3000 1000 900 mill • 7 8 9 10 11 WAVELENGTH (MICRONS) Figure 21. Infrared Spectrum of Bmodin from Non-Polar Extract oo VJ 4000 3000 2000 1500 CM 1000 900 800 700 Q-0_ ..I'.'i'. + 1111II11.1.1.1 IF 0.0 .20 £.30 .30 O £.40 .40 <.50 .50 .60 .60 .70 .70 1.0 oo oo 3 4 5 6 7 8 9 10 11 12 13 14 15 WAVELENGTH (MICRONS) Figure 22. Infrared Spectrum of Emodin, Commercial Sample oo oo 89 Summary of Materials Identified in Non-Polar Extract The following hydroxyanthraquinones were identified in the non-polar extract: (Figure 23) 1) Emodin (1,3,8 trihydroxy-6-methyl anthraquinone) 2) Physcion (1.8 dihydroxy-3-methoxy-6-methy1 anthra quinone; 3) Chrysophanic Acid (1,8 dihydroxy-3-methyl anthra quinone) These materials were identified by isolation from preparative thin-layer chromatography on silica gel layers. The structures were proved by infrared and visible spectral analysis and other qualitative organic procedures. EMODIN (1,3,8 trihydroxy-6-methyl anthraquinone) -^NdCH3 PHYSCION (1,8 dihydroxy-3-methoxy-6-methyl anthraquinone) CHRYSOPHANIC ACID (1,8 dihydroxy-3-methyl anthraquinone) Figure 23. Anthraquinone Pigments of Non-Polar Extract SUMMARY AND CONCLUSIONS The ground, dried roots and tubers of Rumex Hymenosepalus have been examined phytochemically in an effort to isolate the material responsible for anti-tumor activity. A special method of extraction was developed to obtain a purified tannin extract. It was determined that the condensed polymeric portion of the tannin frac tion consisting of leucoanthocyanidin units was responsible for the anti-tumor activity. A method of isolating the active fraction was developed by solvent extraction from the primary tannin extract. This solvent fractionation yielded another fraction from the tannin extract which consisted of monomeric leucoanthocyanidin units of the same type present in the condensed fraction. It was determined by submitting these fractions for anti-tumor tests that fractionation of the main tannin extract did not result in enhanced anti-tumor activity. Since the tannin extract was determined to be of the non-hydrolyzable type, a special method was used to determine the chemical composition. This method consisted of converting the condensed fraction consisting of leucoanthocyanidins to their corresponding anthocyanidins and the identification of the 91 92 same by preparative paper chromatography, visible spectral analysis and identification of fusion products from potassium^hydroxide fusion. The chemical composition of the monomeric portion of the tannin extract was identified in the same manner. The monomeric leucoanthocyanidin, leucocyanidin, was isolated from a polycaprolactam column, and was identified by conversion to cyanidin and subse quent identification of this material. In an effort to identify the anti-tumor material, methanol extracts of a selection of five varieties were made. The five extracts were submitted for anti-tumor activity. The methanol extracts all showed a degree of activity, but the activity did not exceed that shown by the condensed portion of the original extract. Subsequent chemical investigation of the methanol extracts demonstrated the presence of non-tannin constituents, as well as material which had previously been identified in the original tannin extract. A special method of extraction was then developed to obtaj^i a non-polar extract completely free of tannin material. This method was developed so that the non-polar extract could be tested for anti-tumor activity, and to attempt to verify the presence of chrysophanic acid and physcion previously reported, and to isolate emodin, the 93 presence of which had not been completely verified by previous investigators. The non-polar extract was obtained and submitted for anti-tumor testing and was found to be devoid of activity, A new solvent system was developed to identify the anthraquinone pigments present in the non-polar extract. The method consisted of preparative thin-layer chromatography on silica gel and the use of chloroform as a solvent system to separate the material on the thin-layer plates, Chrysophanic acid, physcion and emodin were separated and identified in the non-polar extract. The method of thin-layer chromatography on silica gel layers definitely established the presence of emodin as this pigment was detected at low Rf value on the preparative plates utilizing chloroform as the solvent system. In addition to the materials identified in the non-polar extract, other materials were detected and their identification is being attempted. Other methods used to definitely identify the materials were paper chromatography, visible spectral analysis and infrared spectral analysis with the aid of commercial samples that were available. 94 The result of the phytochemical analysis showed that the anti-tumor activity resided in the condensed portion of the tannin extract. Subsequent investigation of the non-polar extract showed that it was inactive. Fractionation of the tannin extract by solvent extraction from water with ethyl acetate did not result in increased anti-tumor activity. REFERENCES 1. Hillis, W. E., Australian J. Chem., 8, 290 (1955). 2. U, S, Comm. Agr. Report, 119 (1878). 3. Roux, D, G., J. Society Leather Trade Assoc., 36. 148 (1952)*: ; 4. Gordon, T. C., Beebe, C. W., and Rogers, J. S., J. Amer. Leather Chemists Assoc.« 42, 118 (1947). 5. U. S. Dept. Health, Education, and Welfare Protocols for Screening Chemical Agents and Natural Products Against Animal Ttamors and Other Biological Systems, Cancer Chemotherapy Reports, Dec. (1962). Ibid. No. 25, p. 53 (1962). 7. Cheronis, N. D., and Entrikin, J. B., "Semi Qualitative Organic Analysis, 2nd. Ed., Interscience Publishers, Inc., 1958, p. 1/5. 8. "United States Pharmacopeia," 16th Edition, Mack Printing Co., Easton, Pa., 1960, p. 1072. 9. Bryant, C. F.« J. Amer. Pharm. Assoc.. 39. 400 (1950). 10. Cheronis, N. D., and Entrikin, J. B., "Semi Qualitative Organic Analysis," 2nd. Ed., Interscience Publishers, Inc., 1958, p. 264. 11. Shriner, Ralph L., Fuson, Reynold C., and Curtin, David Y., "The Systematic Identification of Organic Compounds," 5th Ed., John Wiley and Sons, Inc., New York, 1964, p. 126. 12. Wilson, C. W., Amer. Chem. Soc., 61. 2303 (1939). 13. Feiser, L. F., and Feiser, M., "Natural Products Related to Phenahthrene," 3rd. Ed., Reinhold Publishing Co., New York, 1949, p. 100. 14. Bate-Smith, E. C., and Lerner, N. H.» Biochem. J., 58, 122 (1954). 95 96 15. Putnam, R. G., Weiederhorn, N. M., et al.. J. American Leather Chemists Assoc.. (Chemistry of Vegetable Tannins). 49, 425 (1954). 16. Jones, J.,American Leather Chemists Assoc.. 32. 167 (1949;. 17. Jurd, L., in "Chemistry of Flavanoid Compounds," T. A. Geissman, ed., Macmillan, 1962, p. 154. 18. Schmidt, 0. T., Prog, in Chem. Ore. Nat. Prod.. 13, 70 (1956). 19. Herbert, H. L., in "Chemistry of Flavanoid Compounds," T. A. Geissman, ed., Macmillan, 1962, p. 563. 20. Roux, D. G., Nature. 180. 973 (1957). 21. Venkataraman, K., in "Chemistry of Flavanoid Compounds," X. A. Geissman, ed., Macmillan, 1962, p. 76. 22. Harborne, J. B., J. Chromatography. 1, 473 (1958), 23. Harborne, J, B,, and Geissman, T. A., Ibid. 2_, 581 (1959)® 24. Robinson, G.. M., and Robinson, R., J. Chem. Soc.. 744-752 (1935). 25. Hayashi, Kozo, in "Chemistry of Flavanoid Compounds," T. A. Geissman, ed., Macmillan, 1962, p. 252. 26. Hayashi, KozoL and Abe, Y., Botan. Mag. Tokyo. 69, 577 (1956). 27. Roux, D, G., Nature. 180, 973 (1957). 28. Roux, D, G,, Maihs, E. A., and Paulus, E., Biochem. J.. 28, 834 (1961). 29. Geissman, T. A., Jorgensen, E. C., and Harborne, J. B., Chem. and Ind. (London), 1389 (1953). 30. Geissman, T. A., Jurd, L., Arch. Biochem. Biophysics. 56, 259 (19553. 31. Harborne, J. B.j. Biochem. J.. 70, 22 (1958). 32. Wilstatter, R., and Mallison, H., Anna1en. 408. 40 (1915). 33. Freudenberg, K.', and Weinges, K., Ibid.. 613. 61(1958). 34. Roux, D. G., Nature,. 181. 1454 (1958). 97 35. Roux, D. G., and Evelyn, S. R., Biochem. J., 69, 530 (1958). i, 36. Bate-Smith, E, G,, and Swain, T,, Chem. and Ind. (London), 377 (1953). 37. Hillis, W. E., Soc. Leather Trade Chemists. 38, 91 (1954). 38. Roux, D. G., jL*. M*. Leather Chem. Assoc.. 54, 614 (1959). 39. Swain, T., in "Chemistry of Flavanoid Compounds," T. A. Geissman, ed., Macmillan, 1962, p. 536. 40. Roux, D. G., and Maihs, A. E., Biochem. J.. 74, 44(1960). 41. White, T,, in "The Chemistry and Technology of Leather," Flaherty, et al., eds., vol. 2, Reinhold, New York, 1958, p. l3£. 42.. Bate-Smith, E. C., and Swain, T., Chem. and Ind.(London). 377 (1953). 43. Freudenberg, K., Scj,t Prpc. Roval Put*, Soc.. 27, 153 (1956). 44. Mayer, W., and Baumi, G., Anna1en. 611. 264 (1958), 45. Hathway, D. E., Chem. Soc.T 520 (1958). 46. Hathway, D. E., Biochem. J.. £0, 34 (1958). 47. Kirby, K. S., and White. T., gps, Leather Trade Chem.T 3£, 215 (1954). 48. Roux, D. G., jla. Leather Chem. Assoc. T 54, 614(1959). 49. Hegert, H. L., in paper presented at the 131st Mtg. Am. Chem. Spc., cf. Abstract of Papers, 6E (1957). 50. Chem. AbstractsT 11782G (1955). 51. Shibata, S., Tahito, M., and Tanahu, 0., J.A.C.S.. 72, 2789-90 (1950). 52. Neelahanton, S., Xfld. Acad. Sgj., 33A, 142-147 (1951).