UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies related to natural products : biosynthesis of coumarins Collier, Peter Lawrence 1971

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
831-UBC_1971_A6_7 C64_5.pdf [ 3.96MB ]
Metadata
JSON: 831-1.0060040.json
JSON-LD: 831-1.0060040-ld.json
RDF/XML (Pretty): 831-1.0060040-rdf.xml
RDF/JSON: 831-1.0060040-rdf.json
Turtle: 831-1.0060040-turtle.txt
N-Triples: 831-1.0060040-rdf-ntriples.txt
Original Record: 831-1.0060040-source.json
Full Text
831-1.0060040-fulltext.txt
Citation
831-1.0060040.ris

Full Text

STUDIES RELATED TO NATURAL PRODUCTS: BIOSYNTHESIS OF COUMARINS BY PETER LAWRENCE COLLIER B.Sc. Honours, University of Alberta, 1967 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in the DEPARTMENT OF CHEMISTRY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA May, 1971 In presenting t h i s thesis in p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library shall make i t f r e e l y available for reference and study. I further agree that permission for extensive copying of t h i s thesis for s c h o l a r l y purposes may be granted by the Head of my Department or by his representatives. It i s understood that copying or p u b l i c a t i o n of t h i s thesis f o r f i n a n c i a l gain shall not be allowed without my written permission. Depa rtment The University of B r i t i s h Columbia Vancouver 8, Canada - i i -ABSTRACT This thesis describes a biosynthetic investigation of coumarins in turpentine-broom, Thamnosma montana Torr. and Frem. In contrast to simple coumarins, the biosynthetic pathways leading to the furanocoumarins were found to be in a state of confusion as indicated by existing published data. None of the results were internally consistent with any general postulate and there was considerable question concerning the actual meaning of a substantial portion of the experimental data. Studies as described here were performed in several different areas. In the i n i t i a l investigations, a detailed study of isolation procedures for the many coumarins present in the plant was necessary. Subsequently, appropriate chemical degradative pathways were developed for umbelliprenin (11), isopimpinellin (2), and alloimperatorin methyl ether (8) to allow isolation of relevant carbon atoms in these coumarins. Finally, 3 incorporation studies with DL-mevalonic-5- H acid were conducted and subsequent degradative reactions were performed on umbelliprenin (11) and alloimperatorin methyl ether (8). The implications of these preliminary experiments in terms of biosynthetic pathways are presented. - i i i -TABLE OF CONTENTS Page TITLE PAGE i ABSTRACT i i TABLE OF CONTENTS i i i LIST OF FIGURES .' iv ACKNOWLEDGEMENTS v i STUDIES RELATED TO THE BIOSYNTHESIS OF COUMARINS INTRODUCTION i 2 DISCUSSION 41 EXPERIMENTAL 60 BIBLIOGRAPHY 85 - iv -LIST OF FIGURES Figure Page 1 Some representative naturally occurring coumarins... 7 2 The shikimic acid pathway to aromatic compounds .... 9 3 Radioactive compounds isolated from H. odorata 13 4 Proposed scheme for coumarin biosynthesis in Melilotus and H_. odorata 16 5 Proposed scheme for biosynthesis of umbelliferone i n Eydrangea 18 6 Proposed scheme for herniarin biosynthesis 21 7 Possible stages of O-methylation in herniarin biosynthesis 22 8 Modified scheme for herniarin biosynthesis 22 9 Involvement of a spirolactone in coumarin biosynthesis 24 10 Suggested biosynthetic scheme for calophyllolide ... 26 11 Degradation of sphondin 28 12 Degradation of pimpinellin 29 13 Proposed scheme for the biosynthesis of furano-coumarins 32 14 Proposed.alternative pathway of furanocoumarin biosynthesis 35 15 Furanocoumarins recovered from Ruta graveolens and Heracleum lanatum 38 16 Proposed scheme for the biosynthesis of marmesin.... 39 17 Typical purification sequence of components from Thamnosma montana, Torr. and Frem 44 18 A typical thin layer chromatoplate of isolated fractions obtained from the chloroform extract 45 19 A thin layer chromatoplate of some authentic samples of compounds in Thamnosma montana Torr. and Frem. .. 46 20 Purification of fractions G, H and I , J and K, and L of the i n i t i a l column chromatography 47 - v -Figure Page 21 Conversion of umbe11iprenin to umbelliferone 48 22 Degradations of isopimpinellin 49 23 Degradations of, alloimperatorin methyl ether ......... 51 24 Recrystallization of various coumarins to constant radioactivity 54 25 Recrystallization of umbelliferone to constant radioactivity 56 26 Recrystallization of alloimperatorin methyl ether d i o l to constant radioactivity 57 27 Recrystallization of the alcohol, compound (48), to constant radioactivity 57 - v i -ACKNOWLEDGEMENTS I wish to express my gratitude to Professor James P. Kutney for his encouragement and guidance throughout the course of this research. I am also extremely grateful to R.N. Young and A.K. Verma for their collaboration and patience with me in this research and for many helpful suggestions. Special thanks i s due also to Miss Diane Johnson for typing the manuscript. - 1 -BIOSYNTHESIS OF COUMARINS - 2 -INTRODUCTION The turpentine-broom, Thamnosma montana Torr. and Frem. (Rutaceae), is found in desert mesas and slopes. These shrubby plants were 1 2 reported to have plant-growth-inhibitor properties. ' In addition, they were reported to have been used by American Indians in folk medicine.^ Bennett and Bonner studied the toxicity of aqueous extracts of leaves of eleven desert plant species and found that Thamnosma montana Torr. and Frem. was the most toxic as judged by the response to tomato.^ The crude material caused the death of young tomato plants at a concen-tration of about 1 mg/ml within seven days. The plants of the Rutaceae family are well known to contain a large number of benzenoid compounds, coumarins, flavones and some quinoline alkaloids. Bennett and Bonner isolated three crystalline compounds from Thamnosma montana and identified two of them as byakangelicin(1) and isopimpinellin (2), respectively.^" - 3 -The structure of the t h i r d compound was elucidated by Dreyer and found to be al l o i m p e r a t o r i n methyl ether d i o l (3) [5-(3'-methyl-2',3'-dihydroxybutanyl)-8-methoxypsoralen]. OCH Dreyer developed a bet t e r e x t r a c t i o n scheme for i s o l a t i n g not only the three compounds obtained by Bennett and Bonner, but was able to separate s i x other compounds. These s i x compounds were g - s i t o s t e r o l (A); three known a l k a l o i d s , N-methylacridone (5), skimmianine (6), and y~fagarine ( 7 ) ; a known furanocoumarin a l l o i m p e r a t o r i n methyl ether (8); and an unknown compound, thamnosin. This was the f i r s t report of N-methylacridone, the parent member of acridone a l k a l o i d s , ~* occurring as a na t u r a l product. However, recent work in our laboratories by T. Inaba has proved this tentative structure to be incorrect. Furthermore, Inaba^ as a result of a complete elucidation of the structure of thamnosin, assigned the following structure to the compound. - 5 -H (10) As a r e s u l t of further recent work i n our l a b o r a t o r i e s on Thamnosma montana, eight other compounds were separated and characterized i n a d d i t i o n to the nine previously i s o l a t e d by the workers mentioned above. These compounds were umbelliprenin (11), isoimperatorin (12), al l o i m p e r a t o r i n methyl ether epoxide (13), thamnosmin (14), bergapten (15), xanthotoxin (16), psoralene (17), and phello.pterin (18). - 6 -Following the i s o l a t i o n and i d e n t i f i c a t i o n of the numerous compounds i n the plant, i t was decided to conduct some bi o s y n t h e t i c work on some of the coumarins. This s e c t i o n of the th e s i s i s concerned with some of the bi o s y n t h e t i c work done on some of the coumarins i n Thamnosma montana. I t i s therefore appropriate to discuss the status of biosynthesis i n t h i s area. Biosynthesis of Coumarins The coumarins form a d i v e r s i f i e d and quite widely d i s t r i b u t e d c l a s s of n a t u r a l l y occurring aromatic compounds. Coumarin i t s e l f i s the simplest member of t h i s c l a s s . Other compounds i n t h i s c l a s s have substituents of varying complexity, ranging from simple hydroxyl or methoxyl groups to isoprenoid side chains and isoprenoid-derived r i n g s . A s t r i k i n g point about the substituted coumarins i s that with only a very few known exceptions, the benzene r i n g of the coumarin nucleus i s oxygenated para to the p o s i t i o n of the side chain attachment, that i s , at p o s i t i o n 7. Figure 1 i l l u s t r a t e s some representative n a t u r a l l y occurring coumarins, bearing a v a r i e t y of d i f f e r e n t functions - 7 -and substitution patterns. Figure 1. Some representative naturally occurring coumarins. Before commencing with the biosynthesis of coumarins, a discussion of the shikimic acid pathway to aromatic compounds (Figure 2) is • - 8 appropriate. - 8 -(Glycolysis-pyruvic acid pathway) D-Glucose (Pentose phosphate pathway) COOH 0 I sT\ / -C — 0 - p — 0 I - ^ 1 1 H2C\OH 0 -HPO, H 0=CH OH H-C-OH A H20P D-Erythrose-4-phosphate COOH C=0 H, COOH 2 HO-C-H H-C-OH H-C-OH DPN. Co CH2OP 3-Deoxy-D-arabino-heptulosonic acid-7-phosphate (DAHP) COOH u H0-I C=0 I C-OH n CH„ DPNH OOOH T° HO-j-H H-C-OH | C-0 1 CH. 3Cyclization^ HQ COOH *''/ Dehydration (5-dehydroquin4se) O OH OH 5-Dehydroquinic acid COOH COOH OH 5-Dehydroshikimic acid COOH OH OH Phosphoenol pyruvate Shikimic acid 5-Phosphoshikimic acid » - 9 -COOH COOH per H CH--0 COOH 0 II 0 H ~\OjT C'^'J^t:;H2 C 0 0 0 1 1 ^COOH OH HO 3-Enolpyruvyl Chorismic acid shikimic acid-5-phosphate |*V Prephenic acid OH COOH \ H CH2 CH COOH NH„ COOH Transamination Cinnamic acid Phenylalanine Phenylpyruvic acid DPN COOH p-Coumaric Tyrosine acid p-Hydroxyphenylpyruvic acid Figure 2. The shikimic acid pathway to aromatic compounds. - 10 -Phosphoenolpyruvic acid and D-erythrose-4-phosphate are independent intermediates in the metabolic pathways of D-glucose. The reaction of the two can be formulated as a concerted reaction between two phosphate esters leading to the formation of orthophosphate and 3-deoxy-D-arabino-heptulosonic acid-7-phosphate (DAHP). The D-arabino configuration i s favored since, in the assumed cyclization of DAHP to 5-dehydroquinic acid, the hydroxyl groups on C-4 and C-5 of DAHP would correspond to the identical configuration of 0-3 and C-4 of 5-dehydroquinic acid. The enzymic condensation of phosphoenolpyruvic acid and D-erythrose-4-phosphate i s formulated as an attack by a nucleophilic group of the enzyme, symbolized here by OH , on the phosphoenolpyruvic acid. This process i s concerted with attack by carbon 3 of the phosphoenolpyruvic acid on the electrophilic carbon atom of the aldehyde and protonation of the carbonyl oxygen atom by an acidic group. The overall result i s the release of orthophosphate or of a transient, phosphorylated enzyme and the open chain form of DAHP. In the limited sense that this reaction may be viewed as an attack by a nucleophilic species on an aldehyde, i t resembles the chemical aldol condensation and the enzymic aldolase condensation. In order to convert DAHP to 5-dehydroquinic acid, the hydroxyl group on C-5 of DAHP i s f i r s t oxidized by diphosphopyridine nucleotide CDPN), to f a c i l i t a t e the elimination of phosphate in the next step. The carbonyl group on C-5 is then reduced by reduced DPN formed in the f i r s t reaction to a hydroxyl group having the original configuration. Finally, the resulting 2,6-diketone i s cyclized to 5-dehydroquinic acid. A l l these reactions are catalyzed by one enzyme or enzyme complex and - 11 -are probably concerted. To convert 5-dehydroquinic acid to shikiraic acid, the former compound is f i r s t dehydrated to 5-dehydroshikimic acid. The enzyme responsible for this was 5-dehydroquinase which was highly specific for this reaction as i t was unable to dehydrate the closely related quinic acid to shikimic acid. The reduction of 5-dehydroshikimic acid to shikimic acid by partially purified 5-dehydroshikimate reductase has also been studied.' A cofactor requirement for the reduction i s triphosphopyridine nucleotide (TPNH) . To convert shikimic acid to prephenic acid, 5-phosphoshikimic acid i s f i r s t formed by phosphate transfer from adenosine-5-triphosphate (ATP) to shikimic acid. This then reacts with enol pyruvate phosphate to give the 3-enol pyruvate ether of phosphoshikimic acid. The prephenic acid is formed by the attack of nucleophilic enolpyruvate on C-l of the ring associated with the phosphate leaving group in a trans manner. With external nucleophilic reagents, such reactions are known to require cis stereochemistry. No exact analogies appear to be known for the type of internal rearrangement postulated for the conversion of enol pyruvate to prephenic acid. Aromatic compounds are formed from prephenic acid by decarboxyla-tion and reductive decarboxylation which yields the immediate precursors of tyrosine and phenylalanine, respectively. The latter two are f i n a l l y formed by transamination and can undergo conversion to cinnamic acid and p-coumaric acid, respectively. Both the acid catalyzed and the enzymic aromatization of prephenic acid may be visualized as initiated by electrophilic attack on the hydroxyl group. Prephenic acid has also been shown to be converted by enzymes to - 12 -p-hydroxyphenylpyruvic acid. Diphosphopyridine nucleotide (DPN) is required for this reaction, suggesting oxidation of C-4 followed by decarboxylation. Biosynthesis of Coumarin: Evidence from feeding experiments showed that coumarin was synthesized from shikimic acid-derived phenylpropane precursors in * . , - J' • 9,10,11 preference to acetate condensation. 9 Brown, Towers, and Wright studied the biosynthesis of coumarin in Hierochloe odorata and Melilotus o f f i c i n a l i s . A number of possible precursors were fed. From this the best precursors of coumarin were found to be o-coumaric acid and cinnamic acid while shikimic acid and L-phenylalanine were found to be slightly less efficient as precursors whereas acetic acid, s a l i c y l i c acid, o-tyrosine» fe r u l i c acid, and melilotic acid were found to be very poor precursors. This work indicated approximately the same results as Kosuge and Conn's feeding experiments in sweet clover, Melilotus alba The results obtained by Brown, Towers, and Wright suggested that the shikimic acid-phenyl-propanoid acid pathway predominates in coumarin biosynthesis in H. odorata. Furthermore, the relative efficient u t i l i z a t i o n of cinnamic acid and o-coumaric acid implied that the former undergoes hydroxylation ortho to the side chain during the process. The effect of varying metabolic periods on the radioactivity of several phenolic constituents of H. odorata including coumarin and structurally related compounds was also studied by Brown, Towers, and 9 14 Wright after the administration of cinnamic acid - 3 - C. The radio-- 13 -activity of the following compounds isolated from the plant was of particular interest (Figure 3). Changes in specific activity of coumarin o-Coumaric acid o-Coumaryl glucoside Coumarin COOH p-Coumaric acid Ferulic a c i d M e l i l o t i c acid Figure 3. Radioactive compounds isolated from H. odorata. 14 with, time after feeding cinnamic acid-3- C to Hierochloe odorata indicated a definite lag phase persisting u n t i l 8 hours from the complete absorption of the precursor. From that time u n t i l about 24 hours, there was a linear increase in specific activity. Changes of specific activity of o-coumaric acid, p-coumaric acid, ferulic acid, melilotic acid, and o-coumaryl glucoside with time from the same experiment indicated that a l l four free acids exhibited a maximum in specific activity at or before 16 hours, and that a l l but melilotic acid attained a relatively high specific activity in 0 to 4 hours. The 14 slow and slight accummulation of C i n melilotic acid again suggested i t to be a secondary metabolic product of cinnamic acid. - 14 -A clearer picture of the distribution of the active carbon was obtained from an examination of the total activity in the various constituents throughout the period of the experiment. From this, i t was clear that the largest reservoir of o-hydroxylated phenyl propanoid material throughout the experiment was o-coumaryl glucoside. It was 14 found that the C content of the glucoside was even higher shortly after feeding than at 4 hours, indicating that cinnamic acid administered to H. odorata undergoes rapid ortho hydroxylation and that glucoside formation also occurs rapidly, with a large accumulation of 14 C i n o-coumaryl glucoside even while absorption of the presursor i s in progress. In order to determine whether o-coumaryl glucoside is an inter-mediate i n coumarin biosynthesis a sample of radioactive o-coumaryl 14 14 glucoside and cinnamic acid-3- C were fed to H. odorata. The C dilution during conversion of cinnamic acid to coumarin was 315 and that of o-coumaryl glucoside was 435, indicating the two compounds to be comparable i n efficiency as precursors of coumarin. 12 Brown has also shown that p-coumaric acid was seventy times less effective than cinnamic acid as a precursor of coumarin i n Hierochloe odorata, while tyrosine was sixty times less effective than phenylalanine. From this i t was clear that a preformed phenolic nucleus posed a decided disadvantage in i t s use as a precursor. In later studies on the biosynthesis of coumarin in H. odorata, 13 14 Brown fed CO^ and showed that coumarin exists in plant cells as a glucoside of o-coumarinic acid (19) and can be recovered by emulsin 14 hydrolysis. A study of the incorporation of C from CC^ into coumarin - 15 -and the aglycone of o-coumaryl glucoside with time was subsequently made. The results indicated that the total activity of coumarin showed (19) a slow, almost uninterrupted increase u n t i l 17 days after which the total declined markedly. The maximum activity of the aglycone, however, occurred at about 4 days and was followed by a marked decrease in activity. In addition to showing the presence of the cis-glucoside of coumarinic acid i n H. odorata, the results of Brown also suggested that o-coumaryl glucoside was a metabolic intermediate rather than an end product, since the peak in i t s total activity is reached earlier than that of coumarin. The demonstration of the presence of coumarinic acid glucoside suggested that glucoside formation was necessary to effect trans-cis inversion which must precede the formation of the coumarin lactone ring. The cis-glucoside would be an intermediate in such a pathway. Also the incorporation of labelled o-coumaryl gluco-side to coumarin, supported the belief that o-hydroxylation was an essential feature of coumarin biosynthesis. Furthermore, coumarin like o-coumaryl glucoside was not an end product, but underwent relatively 14 slow metabolism as revealed by the eventual decline in C in the experiments. - 16 -14 Therefore, Brown proposed the following scheme for coumarin biosynthesis in Melilotus and H. odorata (Figure 4). Figure 4. Proposed scheme for coumarin biosynthesis irt Melilotus and H. odorata. 15 16 Results by Kosuge and Conn , Stoker and B e l l i s , and Gorz and Haskins"^ also independently showed that the conversion of trans-cinnamic acid to coumarin in Melilotus alba occurred by a similar scheme. Furthermore, this plant was shown to contain a trans-cis _ 18 isomerase enzyme system Biosynthesis of 7-0xygenated Coumarins: A. Umbelliferone (7-hydroxycoumarin) (20) The formation of umbelliferone (20) has been studied by Brown, 19 20 21 Towers, and Chen and by Austin and Meyers ' using Hydrangea macrophylla . The latter workers have reported that umbelliferone exists as the free coumarin only to a very small extent i f at a l l , and have identified two bound forms i n Hydrangea (21 and 22). The Glu 7-3-D-Glucosyloxycoumarin or skimmin (21) Glu COOH Glu cis-2,4-di-g-D-Glucosyloxycinnamic acid (22) 14 f i r s t compound was predominant. Experiments with C-labelled compounds, by the two groups have been i n substantial accord and led to the following biosynthetic route being proposed for umbelliferone in Hydrangea (Figure 5). - 18 -Glu (21) (20) Figure 5. Proposed scheme for biosynthesis of umbelliferone in Hydrangea. B. Herniarin (7-methoxycoumarin) (23) Lavender (Lavandula o f f i c i n a l i s Chaix) is one of the few species which elaborate both coumarin and a 7-hydroxylated coumarin, in this 22 case herniarin (23). Therefore, i t was chosen by Brown as a con-venient species in which to compare the biosynthesis of coumarin and a - 19 -7-hydroxylated coumarin. Like coumarin, herniarin'also occurred in the bound state as a glucoside. It seemed quite certain that this glucoside was 2-gluco-syloxy-4-methoxy-cis-cinnamic acid (cis-GMC) (24). The presence of this compound implied the formation, at some stage, of an ortho-hydroxylated precursor and indicated that the lactone ring of 7-hydroxylated coumarins also can be synthesised via ortho-hydroxylation. 22 14 Brown compared in studies on lavender a number of different C labelled compounds as precursors of herniarin, and i n some cases coumarin. L-phenylalanine was incorporated with moderate dilution of 14 C into herniarin as well as coumarin. Glucose was used with lower efficiency as a precursor of both coumarins. However,o-coumaric and p-coumaric acids were ut i l i z e d with a high degree of selectivity. The former was used for the synthesis of coumarin some 150-200 times as e f f i c i e n t l y as i t was used for herniarin synthesis, and the latter was selectively u t i l i z e d for herniarin,synthesis by a slightly lesser factor. Furthermpre}cinnamic acid was found to be a precursor of both coumarins. 1 These findings showed that herniarin, unlike coumarin, was not synthesized by way of the o-coumaric acid - o-coumaryl glucoside pathway. They confirmed the theory that coumarin and herniarin arise via ortho-and para-hydroxylation, respectively, of a common precursor, probably some form of cinnamic acid. 14 The low dilution of C i n trans-GMC suggested that this compound was the intermediate precursor of cis-GMC. The trans-cis inversion necessary for this step would be analogous to that postulated in the - 20 formation of coumarinyl glucoside. In addition, during the conversion of p-coumaric acid to cis-GMC,an ortho-hydroxylation glucoside formation and an 0-methylation of the para-hydroxyl group must occur. Glucoside formation must obviously follow the o-hydroxylation, but the order of the other steps remained in question. In an attempt to elucidate the above problem Brown compared p-14 14 methoxycinnamic acid-a- C and 2,4-dihydroxy-trans-cinnamic acid-1- C (umbellic acid) with p-coumaric acid as precursors of herniarin. The results clearly demonstrated the very high efficiency with which p-methoxycinnamic acid was converted to herniarin. Umbellic acid and umbelliferone, while moderately well utilized,were both poorer precursors than p-coumaric acid and 25-50 times poorer than p-methoxycinnamic acid. Thus., these experiments revealed the following order of precursor efficiencies for herniarin synthesis: trans-GMC > p-methoxycinnamic acid > p-coumaric acid > umbellic acid _ umbelliferone > glucose. Therefore, the scheme for the herniarin biosynthesis proposed by 22 Brown at the time can be represented by the following scheme (Figure 6). The stage at which O-methylation occurred in the biosynthesis of herniarin remained uncertain, however. The two p o s s i b i l i t i e s shown below CFigure 7) seemed the most probable, and as has already been noted a comparison of the two labelled intermediates showed that 22 p-methoxycinnamic acid was much the better herniarin precursor. 23 On the basis of existing evidence, Brown recently proposed the following scheme (Figure 8) for herniarin biosynthesis. Here p-coumaric acid i s converted to an intermediate "X", which may be an enzyme-substrate COOH COOH C6 H11°5 2-Hydroxy-4-methoxy-trans-cinnamic acid 1 trans-GMC CH30 COOH 0 V CH 0 C6 H11°5 (24) (23) ure 6. Proposed scheme for herniarin biosynthesis. - 22 -COOH Umbellic acid Figure 7. Possible stages of O-methylation i n herniarin biosynthesis. i 1 trans-GHC trans-GMC (24) Figure 8. Modified scheme for herniarin biosynthesis. - 23 -complex. To overcome the d i f f i c u l t y posed by the poor efficiency of umbellic acid as an intermediate, i t was suggested that o-hydroxylation and glucosylation occurred without the substrate leaving the enzyme surface, or whatever other complex may be concerned. The product, trans-2-glucosyloxy-4-hydroxy-cinnamic acid (trans-GHC), could then undergo O-methylation to trans-GMC, a known intermediate. However, trans-GHC, an unknown compound, has not yet been tested as a herniarin precursor. C. Novobiocin H2N OCH„ 3 12 A 2 5 Kenner and his co-workers ' have shown that the lactone ring i n the coumarin residue of the antibiotic novobiocin (25), formed by a Streptomyces, originates in a different way. By the use of 14 C they showed that the coumarin residue was formed from tyrosine, 18 and further work with 0 yielded good evidence that the ring oxygen originated from the carboxyl oxygens of tyrosine. They postulated an oxidative cyclization of the amino acid to explain their results. 26 27 Others ' have raised the question whether a similar mechanism may 27 not also operate in higher plants, and Scott, Meyers, and co-workers have suggested the involvement of a spirolactone as shown (Figure 9 ). - 24 -(20) 6-Hydroxycoumarin Figure 9. Involvement of a spirolactone in coumarin biosynthesis. 27 The model reaction shown was demonstrated _in v i t r o , but subsequent ^n vivo work by Austin and Meyers 2^ with a "^C-labelled spirolactone did not bear out the theory. The fact that the coumarins in question actually exist in the c e l l as glucosides of coumarinic acids argued strongly that at least in plants an o-hydroxylation 28 mechanism, rather than oxidative cyclization i s the favored pathway. D, Coumestrol OH (26) - 25 -29 30 The investigations of Grisebach and Barz ' on the coumarano-coumarin, coumestrol (26), have shown that this compound is biosyntheti-cally an isoflavone with the benzenoid ring of the coumarin nucleus originating from acetate, and the remaining nine carbons from phenyl-26 propanoid precursors. As in the formation of isoflavones, an aryl-migration was involved, and a chalcone glucoside was apparently an intermediate. These findings demonstrated that identical structures may have entirely different biosynthetic origins, even in species of the.same family. E. Calophyllolide (27) 31 Recent work by Kunesch and Polbnsky showed that the specific 14 incorporation of (-)-phenylalanine-3- C into a 4-phenylcoumarin (neoflavanoid), calophyllolide (27), supported the biogenetic scheme 32 33 (Figure 10) suggested by Seshadri and O l l i s . Edwards and Stoker presented additional information concerning the isomerization of o-coumaroyl glucoside to coumarinoyl glucoside 34 (19) in the biosynthesis of coumarin and the isomerization of 2-glucosyloxy-4-methoxy-trans-cinnamic acid (trans-GMC) to 2-glucosyloxy-4-methoxy-cis-cinnamic acid (cis-GMC) (24) in the biosynthesis of , . . 35 herniarin. - 26 -Figure 10. Suggested biosynthetic scheme for calophyllolide. The results of Edwards and Stoker indicated that an enzyme is not involved in the isomerization of o-coumaroyl glucoside to coumarinoyl glucoside in M. o f f i c i n a l i s , but that the reaction i s catalyzed by light. Furthermore, their results also indicated that the isomerization - 27 -stage in the biosynthesis of herniarin is a photochemical step, no isomerase enzyme being involved. Consequently, this isomerization stage has now been shown to be photochemical both for coumarin i t s e l f and for a typical 7-oxygenated coumarin. It is probable, therefore, that the isomerization step in the biosynthesis of a l l plant coumarins is entirely photochemical. Biosynthesis of Furanocoumarins: In one of the i n i t i a l studies on the biosynthesis of furano-36 coumarins, Caporale and his co-workers reported the incorporation of 3 14 3 radioactivity from acetate-2- H, tyrosine-2- C and -U- H, mevalonic 14 3 acid-2- C, and succinic acid-2,3- Hinto bergapten (15) and psoralen (17) by leaves of Fiscus carica. OCH (15) (17) 37 Floss and Mothes presented evidence that the coumarin skeleton of furanocoumarins in Pimpinella magna (Umbelliferae) was formed from cinnamic acid and also that para-hydroxylation preceded ortho-hydroxyla-tion, since umbelliferone was a far better precursor than cinnamic acid whereas coumarin gave only poor incorporation, indicating that umbelliferone but not coumarin was an intermediate in the formation of Pimpinella furanocoumarins from cinnamic acid. The co-occurrence of - 28 -simple furanocoumarins, isopropylfuranocoumarinsand isoprenylated furanocoumarins suggested the possibility that the furan ring originated from carbons 1 and 2 of the isoprene residue or from carbons 4 and 5 of 38 39 mevalonic, acid, respectively. Floss and Mothes reported results which confirmed the incorporation of cinnamic acid and apparently c l a r i f i e d the origin of the two extra carbon atom's of the furan ring, and hence allowed some conclusions to be made regarding the pathway of furanocoumarin formation. 14 14 Labelled cinnamic acid-( COOH) and DL-mevalonic acid-4- C were fed to Pimpinella magna and various furanocoumarins present were isolated. In order to establish the direct conversion of cinnamic acid to furanocoumarins, a degradation, using sphondin (28), was performed (Figure 11). The CO^ obtained from C-2 of sphondin had about Cu Powder > C0_(C -2 ) COOH 5' 4' (28) Figure 11. Degradation of sphondin. 90% of the specific activity of the starting material. Thus, most of the activity was located at C-2 of the furanocoumarin, indicating that cinnamic acid was indeed incorporated into the furanocoumarins. 14 Mevalonic acid-4- C was also incorporated into the furanocoumarins. To determine the location of the isotope i n the furanocoumarins isolated, a degradation was attempted. A l l the C would be expected to be in position 5', i f the furan ring was indeed formed from C-4 and C-5 of mevalonic acid. Labelled pimpinellin (29) was degraded by ozonization (Figure 12) and the product was identified as 2,4-dihydroxy-5,6-dimethoxy-m-phthaldehyde (30). (29) (30) Figure 12. Degradation of pimpinellin. The ozonization reaction involved removal of carbon atoms 2, 3, and 5'. Since pimpinellin had a constant specific activity and the dialdehyde had only 11% of the original specific activity, most of the must have been located in C-2, C-3 and/or C-5'.' As shown before, the coumarin portion of the molecule, which includes C-2 and C-3, originated from cinnamic acid. A specific incorporation of C-4 of mevalonic acid was highly unlikely. Thus, i t confirmed that C-4 of mevalonic acid indeed labelled C-5' furanocoumarins. 39 Therefore, the results of Floss and Mothes apparently c l a r i f i e d the origin of the carbon skeleton of the furanocoumarins. Thus, the coumarin portion i s formed from cinnamic acid, while the furan ring originated from an isoprenoid residue, probably via isoprenylation of - 30 -the aromatic ring followed by cyclization and loss of a 3-carbon isopropyl side chain. The general, mechanism of furanocoumarin biosynthesis especially the sequence of the individual steps was also considered by Floss and 39 Mothes. With very few exceptions a l l furanocoumarins have the furan ring attached to position 6 and 7 or 7 and 8 of the coumarin system, the oxygen being in position 7. This would be explained most easily by the assumption that umbelliferone is the compound which i s isoprenylated, and that further hydroxylations and alkylations take place after this step. However, the results of the cinnamic acid and mevalonic acid feeding experiments clearly showed that this was not the case. If isoprenylation of umbelliferone preceded further hydroxylations, i t was expected that the relative specific a c t i v i t i e s of the various furanocoumarins isolated would be the same in the cinnamic acid as in the mevalonic acid experiment, since a l l the reactions which create the difference between the products would occur after the introduction of the isotope in both cases. However, the relative specific a c t i v i t i e s of the various furanocoumarins were not the same in the cinnamic acid and mevalonic acid experiments. The results pointed to the alternative whereby the hydroxylation and methylation pattern was established f i r s t and isoprenylation then occurred as the last step. Since, however, the 7-hydroxy group must be protected from methylation i n order to allow furan ring formation, i t was l i k e l y that umbelliferone was f i r s t converted to the glucoside and this compound was the substrate for further hydroxylations and methylations. The sugar could then be removed at a latter stage to allow furan ring formation to take place. 39 Floss and Mothes also found that, the hydroxylation pattern of the - 31 -furanocoumarins in Pimpinella could be derived by a combination of three types of reactions: (i) a hydroxylation of umbelliferone ortho to the hydroxyl group and para to the lactone oxygen, ( i i ) a hydroxylation in position 5 ("second ortho hydroxylation"), ( i i i ) opening of the lactone ring and recyclization in the opposite direction. An observation in favor of this scheme was that xanthotoxin (16), the only member of a l l the possible mono- and dimethoxylated furano-coumarins of this type which could not be formed by these reactions, was not found in Pimpinella. , (16) The data also suggested that reaction (i) occurred more readily than reaction ( i i ) , since in the cinnamic acid experiment, sphondin (28) had a much higher specific activity than bergapten (15) and isobergapten (31). As expected bergapten and isobergapten had the same specific activity in the cinnamic acid experiment, whereas the value for pimpinellin (29) and isopimpinellin (2 ) were lower. Since the value of pimpinellin was lower than that of i t s isomer i t appeared as i f i t was obligatory for reaction ( i i i ) to occur together with reaction ( i i ) . - 32 -The data from the mevalonic acid experiments also indicated that isoprenylation at position 8 was preferred as a l l angular furano-coumarins had a higher specific activity than the linear ones. This was more pronounced with the pair bergapten and isobergapten, where there was a true competition for the same precursor. 39 On the basis of the results Floss and Mothes proposed the following scheme for the biosynthesis of furanocoumarins (Figure 13). COOH COOH Figure 13. Proposed scheme for the biosynthesis of furanocoumarins. - 33 -40 Recently, Floss and Paikert examined the scheme experimentally. The evidence obtained did not favour the proposed biogenetic scheme and indicated that an alternative pathway, involving isoprenylation at the umbelliferone stage followed by further modification of the benzenbid ring, must be considered. These workers attempted to determine whether the ortho- or the para-hydroxyl group was introduced into the cinnamic acid molecule. The fact that coumarin was not ut i l i z e d i n the biosynthesis did not rule out o-coumaric acid as an intermediate. Both o-coumaric acid 14 and p-coumaric acid with a C label in the carboxyl group were fed to Pimpinella magna. However, the results together with previous work by Brown and his co-workers on the biosynthesis of umbelliferone in Hydrangea macrophylla suggested that o-coumaric acid was not a natural intermediate in the synthesis, but that p-coumaric acid was the l i k e l y precursor. 40 Floss and Paikert also studied the incorporation of umbelli-ferone and i t s 7-glucoside into furanocoumarins. It was found that the glucoside was incorporated slightly less e f f i c i e n t l y than the aglycone 41 Furthermore, earlier experiments had shown very rapid glycosylation of added umbelliferone, and thus, they examined the metabolic fate of umbelliferone in P. magna. Umbelliferone-S-^H was root fed to the plant. After work up of the plants and subsequent enzymatic hydrolysis of the coumarins to yield the aglycones, i t was noted that umbelli-ferone was the only labelled material, that i s , there was no detectable conversion of umbelliferone glucoside to other coumarin glucosides. If the hypothetical pathway of furanocoumarin biosynthesis, as - 34 -postulated by Floss and Mothes, was correct, i t was expected that scopoletin (7-hydroxy-6-methoxycoumarin) (32) would be a precursor of sphondin and pimpinellin, possibly also of isopimpinellin, but not of bergapten and isobergapten. Therefore, scopoletin-methyl-^ was administered to the roots of P. magna and, as a reference, i t s isomer (6-hydroxy-7-methoxyc6umarin-methyl-^H) (33) was fed i n a paral l e l experiment. (32) (33) It was found that scopoletin-me was not incorporated preferentially into any one furanocoumarin as would be expected i f isoprenylation were a late step i n the biogenetic pathway. Furthermore, i t labelled the furanocoumarins in about the same manner of magnitude as i t s 3 isomer 6-hydroxy-7-methoxycoumarin-methyt- H which could not be an intermediate i n the formation of the furanocoumarins of Pimpinella. This indicated that the scopoletin molecule was not transformed into furanocoumarins as a unit, but that the incorporation of radioactivity was possibly only due to demethylation and pa r t i a l r e u t i l i z a t i o n of the labelled fragment. This was also suggested by the rather low rate of incorporation of scopoletin as compared to umbelliferone. Therefore, i t seemed very l i k e l y that scopoletin was not a direct precursor of any of the furanocoumarins of P. magna. 35 40 Therefore, the results presented by Floss and Paikert did not substantiate the assumption that isoprenylation occurred as a late reaction in furanocoumarin biosynthesis and that the hydroxylations and methylations of the coumarins portion occurred at the glucoside stage. An alternative pathway of furanocoumarin biosynthesis had to be considered, involving isoprenylation of umbelliferone (20) followed by further modification of the resulting 6- and 8-dimethylallylumbelliferone (Figure 14). Such a pathway was attractive from a phytochemical point (31) R^H, R2=OCH3, (28) R^OCH^ R2=H (29) RX = R2 = 0CH3 (15) R1=0CH3, R2 = H = 0CH„ Figure 14. Proposed alternative pathway of furanocoumarin biosynthesis. of view since i t could explain the almost exclusive occurrence, in nature, of 6,7- and 7,8-furanocoumarins. 42 Floss, Guenther, and Hadwiger recently presented work on the biosynthesis of furanocoumarins in diseased celery. Upon infection with the fungus Sclerotinia sclerotiorum,celery was known to develop a condition called "pink rot" which has been correlated with the skin-i r r i t a t i n g properties of diseased celery. Two phototoxic psoralen derivatives have been isolated from diseased celery tissue, xanthotoxin (8-methoxypsoralen) (16) and 4,5',8-trimethylpsoralen (34). OCH3 CH3 (16) (34) These workers conducted some feeding experiments to determine whether the two psoralen derivatives originated from the same biosynthetic precursors as other furanocoumarins found i n higher plants, namely cinnamic acid and mevalonic acid with the O-methyl groups presumably coming from the pool (formate or the methyl group of methionine). Xanthotoxin has the same structural pattern as other higher plant furanocoumarins and would thus be expected to be derived from the same biogenetic precursors. On the other hand, the psoralen derivative (34) has some rather unusual features in that i t carries three carbon bound methyl groups. If, as in other furanocoumarins, the ring skeleton of this compound is derived from cinnamate and mevalonate, the three methyl - 37 -groups, or at least the two attached to the coumarin portion, would be expected to originate from the pool by C-methylation. 14 14 14 Cinnamic acid-1- C, methionine-methyl- C, and formate- C were fed to celery plants after infection. As expected, a l l three compounds were incorporated into xanthotoxin, although formate was a considerably less efficient methyl donor than methionine, interestingly, however, none of the three substrates labelled 4,5',8-trimethylpsoralen to any significant extent. It was conceivable that the two products were possibly synthesized at different sites, that i s , xanthotoxin may have been formed by the plant and trimethylpsoralen by the fungal c e l l s . As an alternative explanation of the results of this study, one had, however, to consider the poss i b i l i t y that the two furanocoumarins found i n diseased celery were formed by two entirely different bio-synthetic pathways. 43 Steck, El-Dakhakhny and Brown recently presented material strongly indicating the participation of two a-hydroxyisopropyldi-hydrofuranocoumarins, marmesin (35) and columbianetin (36) as natural intermediates in the elaboration of linear and angular furanocoumarins, respectively. To date, this was the f i r s t time intermediates in the reaction sequence from umbelliferone to furanocoumarins had been identified. - 38 -1 A Trapping experiments were conducted in which umbelliferone-2- C was administered to shoots of Ruta graveolens together with non-14 labelled marmesin, and skimmin-2- C to Heracleum lanatum together with non-labelled columbianetin. In addition to marmesin, the linear furano-coumarins psoralen (17), bergapten (15), and xanthotoxin (16) , were recovered from Ruta. In addition to columbianetin, angelicin (37) , isobergapten (31), sphondin (28), and pimpinellin (29) were recovered from Heracleum (Fig. 15). R2 (17) (Rj_ = R2 = H) (37) (R x = R2 = H) (15) (Rx = OCH3, R2 = H) (31) (R^ = H, R£ = OCH3) (16) (R1 = H, R2 = OCH3) (28) (R^ = OCH3> R2 = H) (29) (Rx = R2 = OCH3) Figure 15. Furanocoumarins recovered from Ruta graveolens. and Heracleum lanatum. From the results, i t was clear that in each species umbelliferone had been converted to the dihydrofuranocoumarin. Furthermore, the lower specific activities of the furanocoumarins were consistent with their formation from marmesin or columbianetin. 43 Steck, El-Dakhakhny and Brown also conducted direct feedings of t r i t i a t e d marmesin and columbianetin to confirm the roles of these compounds in furanocoumarin biosynthesis. In each case there was good conversion to the analogous unsubstituted furanocoumarin and a lower degree of tritium incorporation into the oxygenated furanocoumarins. Umbelliferone was known to occur in Ruta graveolens and in Heracleum lanatum. Therefore, i t can be regarded as a natural inter-mediate in furanocoumarin biosynthes is in these species. The formative route to marmesin proposed by Steck, El-Dakhakhny and Brown can be envisaged as the following (Figure 16). ? (35) i Figure 16. Proposed scheme for the biosynthesis of marmesin. There was an analogous sequence leading to columbianetin, but there was no firm experimental evidence to support this hypothesis. The nature of the reactions leading on to psoralen and angelicin and ultimately to the substituted furanocoumarins, however, remained unclear. As the above discussion reveals, the biosynthetic pathways leading - 40 -to the furanocoumarins, in contrast to those of simple coumarins are indeed in a state of confusion. None of the results are internally consistent with any general postulate and there was question in our minds about the real meaning of a substantial portion of the experi-mental data. For this reason we decided to investigate the biosynthesis of several furanocoumarins isolated from Thamnosma montana. Furthermore, the eventual goal of studying the biosynthesis of the novel dimeric thamnosin system required an i n i t i a l , careful, investigation'of the appropriate monomeric units which would be visualized as the biosynthetic templates for the dimeric natural product. - 41 -DISCUSSION The discussion within the Introduction section of this thesis has revealed that a considerable amount of work has been done on the biosynthesis of some of the simpler monomeric coumarins. Unfortunately, careful analysis of a large portion of the published data leads one to conclude that much s t i l l remains to be done in this area. In many instances the incorporations reported are very low, thereby leading to the isolation of natural products possessing low levels of activity. Furthermore, the eventual characterization'and, i n particular, sufficient degradation to isolate the relevant carbon atoms i s incomplete in a number of instances. It was our feeling that some of the confusions already mentioned in the Introduction are simply due to the incompleteness of the investigations. In order to try and establish more firmly the various postulates i n the literature and also to provide an internally consistent picture of the various coumarins i within the same plant system, we decided to i n i t i a t e investigations i n turpentine-broom, Thamnosma montana Torr. and Frem. As already mentioned, this plant possesses a wonderful array of monomeric and dimeric coumarins. In effect these' aims required investigations in several different areas. These are the following: - 42 -(a) Detailed investigation of suitable isolation procedures to allow isolation and characterization of small quantities of the natural coumarins, (b) Development df appropriate chemical degradation pathways which allow isolation of relevant carbon atoms in the coumarins, 14 3 Cc) Incorporation studies with appropriately labelled ( C and H) precursors. For this purpose, R.N. Young, A.K. Verma, and myself became involved i n separate;but complementary areas of the research. I w i l l now discuss those areas with which I was directly involved but wish to emphasize that a great deal of related work which strengthens some of the conclusions made later i s available from the efforts of my collaborators. For the sake of c l a r i t y , I propose to discuss my work in the order of research areas (a-c). mentioned above. Old l i v i n g plants of' the turpentine-broom, Thamnosma montana Torr. and Frem., containing rather thick woody roots, were obtained in the desert region near Morongo Valley, California with the help of Dreyer and his associates at the Fruit and Vegetable Chemistry Laboratory, Pasadena, California. From the whole aerial part of the plant including the crown as well as the roots, i t was considered desirable to isolate a l l the compounds previously found in the plant as well as any new ones which might be present. In addition to the plant material, Dreyer and his associates also provided us with an acetone extract of Thamnosma montana, Torr. and Frem. Both of these were used to develop a suitable isolation procedure. - 43 -(a) Isolation of Natural Products from Thamnosma montana Torr. and Frem. The plants were air dried thoroughly and ground to a course powder in a high speed Waring blender. Using a Soxhlet extractor, the pulp was extracted with acetone. The resulting solution was f i l t e r e d and the clear acetone solution was evaporated to provide a dark heavy tar. The latter was now extracted with chloroform, f i l t e r e d , and evaporated as before leaving a dark heavy oi l y residue. The crude extract was separated into numerous fractions using column chromatography on deactivated alumina. A typical isolation scheme is portrayed in Figure 17 with a typical thin layer s i l i c a gel chromato-plate of the isolated fractions illustrated in Figure 18. For purposes of comparison the appropriate colour characteristics under ultraviolet light of previously characterized compounds found i n Thamnosma montana Torr. and Frem. are shown on the s i l i c a gel chromatoplate in Figure 19. The compounds of particular interest for our i n i t i a l biosynthetic work were umbelliprenin, alloimperatorin methyl ether, and isopimpinellin. It was desirable, therefore, to i n i t i a l l y accummulate sufficient quantities of the above mentioned compounds along with alloimperatorin methyl ether diol for purposes, of subsequent dilution and degradative work in the biosynthetic investigations. . For this purpose the various fractions from the i n i t i a l column chromatography were combined in an appropriate manner and subjected to additional column chromatography, thin layer chromatography, and f i n a l l y crystallization to yield the desired products as il l u s t r a t e d in Figure 20. - 44 -Ground Plants Crude Extract Soxhlet Extraction (acetone) Extraction (chloroform) Column Chromatography Fraction H K Solvent petroleum ether (30-60) petroleum ether (30-60):benzene (3:1) petroleum ether (30-60):benzene (2:1) petroleum ether (30-60):benzene (1:1) benzene benzene Compounds unidentified non-polar components such as hydrocarbons and waxes benzene:chloroform (1:1) primarily umbelliprenin chloroform chloroform umbelliprenin, isoimpera-torin, alloimperatorin methyl ether thamnosmin, and isopimpinellin chloroform chloroform isopimpinellin, N-methyl-acridone, and alloimpera-torin methyl ether epoxide. chloroform N-methylacridone, and alloimperatorin methyl ether d i o l M chloroform acetone acetone methanol primarily unidentified polar components Figure 17. Typical purification sequence of components from Thamnosma montana Torr. and Frem. - 45 -Fraction A B C D E F G H I J K L M N O 1 1 1 3f 0 0 » Q i 0 ft o ' l o o W 5 0 0 o . o 5 0 0 0 0 0 0 Figure 18. A typical thin layer chromatoplate of isolated fractions obtained from the chloroform extract. Characteristic colours observed on the chromatoplate under ultraviolet light are presented: (1) purple; (2) blue; (3) yellowish-brown; (4) brown; (5) yellowish-green. - 46 -Figure 19. A thin layer chromatoplate of some authentic samples of compounds in Thamnosma montana Torr. and Frem. The legend indicating the names of the compounds and their typical colour under ultraviolet light follows: (1) umbelliprenin - blue; (2) isoimperatorin - yellowish-brown; (3) alloimperatorin methyl ether - yellowish-brown; (4) thamnosmin - purple; (5) isopimpinellin - brown; (6) N-methylacridone - blue; (7) alloimperatorin methyl ether epoxide - yellowish-green; (8) alloimperatorin methyl ether diol - yellowish-brown. - 47 -7 Fraction H and I Fraction J and K Column Chromatography and/or Thin Layer Chromatography and Crystallization 1 umbelliprenin umbelliprenin, isoimperatorin, alloimperatorin methyl ether, thamnosmin, and isopimpinellin isopimpinellin, N-methylacridone and alloimpera-torin methyl ether d i o l N-methylacridone an'd alloimpera-torin methyl ether d i o l Figure 20. Purification of Fractions G, H. and I, J and K, and L of the i n i t i a l column chromatography.. (b) Development of Appropriate Chemical Degradation Pathways. The coumarins of interest primarily for the development of degradative reactions were umbelliprenin (11), isopimpinellin (2), and alloimperatorin methyl ether (8). From these reactions i t was thought possible to isolate relevant carbon atoms in the above mentioned coumarins. Before these degradative reactions could be performed on any active material, isolated from the incorporation experiments, i t was necessary to perfect them on the inactive compounds previously isolated from the plant. R.N. Young, A.K. Verma and myself, consequently, became involved in the rather extensive research on the chemistry of the appropriate coumarins. For the sake of c l a r i t y I w i l l summarize a l l - 48 -of the pertinent reactions which were studied. However, experimental details of only those reactions which were performed by me are recorded in the thesis. ( i ) Degradation of Umbelliprenin The degradative reaction employed for umbelliprenin involved 44 acidic cleavage of the farnesol side chain to yield umbelliferone (20) (Figure 2l). Umbelliprenin (11) was treated with g l a c i a l acetic acid to which a catalytic amount of concentrated sulfuric acid had been added. (11) (20) Figure 21. Conversion of umbelliprenin to umbelliferone. By the above degradative procedure the relevant carbon atoms in the cyclic portion of umbelliprenin could be separated from the iso-prenoid side chain i n the molecule. ( i i ) Degradation of Isopimpinellin There were two degradative sequences developed i n connection with isopimpinellin (2) (Figure 22). i n one sequence isopimpinellin was demethylated by heating i t with anhydrous aluminum chloride to yield 5,8-dihydroxypsoralene (38) which in turn was acetylated by treating i t with acetic anhydride and pyridine to yield 5,8-diacetylpsoralene (39). This latter substance was more easily handled than the former product. In the other sequence isopimpinellin underwent ozonolysis in - 49 -OCH 0 0 C H 3 OCH, (2) " A1C1'3,A (38) (CH3C0)20 0' C 5 H 5 N OCH, (40) \\ .CH, (39) (f ^CH Figure 22. Degradations of isopimpinellin. gla c i a l acetic acid solution. The resulting ozonide was then reduced by using zinc dust to yield 5,8-dimethoxy-6-formylumbelliferone (40). Another degradation attempted with isopimpinellin was to convert i t to 5,8-dimethOxyumbelliferone-6-carboxylic acid (41). It was f e l t 0 OCH (41) -50 -that this compound could in turn be decarboxylated and hence this would provide a technique of isolating the carbon atom attached to position-7 of the coumarin system. In one experiment isopimpinellin was heated with dilute sulfuric acid and potassium dichromate to try and obtain 45,46 1 •, the desired product. In the other case isopimpinellin was subjected to ozonolysis in a glac i a l acetic acid solution, followed by an oxidative work-up using hydrogen peroxide. However, neither procedure produced substantial amounts of the desired carboxylic acid and further attempts to obtain i t were abandoned. ( i i i ) Degradations of Alloimperatorin Methyl Ether There were two degradative sequences developed i n connection with alloimperatorin methyl ether (8) (Figure 23). Alloimperatorin methyl ether i n both cases was converted f i r s t to alloimperatorin methyl ether diol [5-C3'-methyl-2',3'-dihydroxybutanyl)-8-methoxypsoralen] (3) via the epoxide (13). This latter compound was obtained by f i r s t treating the methyl ether with m-chlorcperbenzoic acid i n chloroform at 0°C and then' heating the resulting alloimperatorin methyl ether epoxide with dilute oxalic acid. In one sequence alloimperatorin methyl ether d i o l was f i r s t treated with acetic anhydride in pyridine to yield 5-(3,-methyl-3'-hydroxy-2'-acetoxybutanyl)-8-methoxypsoralene (42) which i n turn was ozonized and reduced, respectively, using ozone in gl a c i a l acetic acid followed by reduction with zinc dust to yield 5-(3'-methyl-3'-hydroxy-2'-acetoxy-i butanyl)-6-al-8-methoxyumbelliferone (43). In the other sequence alloimperatorin methyl ether d i o l was treated with periodic acid in a mixture of acetone free methanol and Figure 23. Degradations of alloimperatorin methyl ether. - 52 -water to yield the aldehyde, compound (44), and acetone (45). The acetone was d i s t i l l e d into p-bromobenzenesulfonhydrazide (46), present in a mixture of glac i a l acetic acid and water, to yield acetone p-bromobenzenesulfonhydrazone (47).^ The aldehyde, compound (44)»was treated with sodium borohydride in a mixture of methanol and chloroform, followed by hydrolysis with water to yield the alcohol, compound (48). (c) Biosynthetic Investigations with Radioactive Mevalonic Acid 3 DL-Mevalonic-5- H acid as i t s dibenzylethylenediamine salt (49) was u t i l i z e d i n the various experiments discussed in this thesis. The OH 3 • I HOC H_-CHo-C-CHoC00H(C£HcCHoNHCHoCH 0NHCH0C>H_), Z Z \ Z o J Z Z Z Z o o 1/ Z CH-3 (49) free acid is unstable and readily lactonizes, therefore, the DBED (dibenzylethylenediamine) salt i s normally employed. In preparation for the plant feedings the methanol solution of the radioactive compound was diluted to a known volume (25 ml) and aliquots of this solution were then used in the experiments described. Similar experiments were conducted by R.N. Young and A.K. Verma 14 using the following possible biosynthetic precursors: glycine-1- C, 14 14 14 glycine-2- C, DL-mevalonic acid-2- C, DL-mevalonic acid-4- C, 3 ' 3 DL-mevalonic acid-2- H, and DL-mevalonic acid-4- H. In the area of research with which I was directly involved, that 3 is with the biosynthetic investigations of DL-mevalonic-5- H acid as i t s dibenzylethylenediamine salt (49), Thamnosma montana plants of about 19 to 22 months of age were ut i l i z e d for the various feedings. - 53 -In a typical biosynthetic feeding an aliquot of the previously mentioned solution of the radioactive compound (49) was fed to a small number of the young plants (6 to 8) for a total exposure time of about ten days. The whole plant was used for the feeding experiments conducted, the radioactive compound being incorporated through the roots with the feeding being hydroponic in nature. After the biosynthetic feeding was completed umbelliprenin, alloimperatorin methyl ether, thamnosmin, isopimpinellin, and a l l o -imperatorin methyl ether epoxide were isolated by the procedure discussed under section (a) of the Discussion and summarized b r i e f l y in Figures 17 and 20. The activity of the combined residual aqueous solutions in which the plants were fed was determined as was the activity of the acetone and chloroform extracts. Before crystallization was ut i l i z e d on each of the isolated coumarins, the amount obtained was determined and then diluted with a weighed amount of cold material previously isolated from the plant. Following this procedure, attempts were made to obtain, by crystallization techniques, constant radioactivity .counts for each coumarin (Figure 24). Subsequent degradation of these compounds provides an indication of the location of the label, and, hence, allows some conclusions to be drawn concerning the biosynthesis of these compounds. (i) Degradation of Radioactive Umbelliprenin After radioactive umbelliprenin (11) had been recrystallized to constant radioactivity (Figure 24), the compound was degraded to umbelliferone (20) as described under Section (b) (i) of the Discussion for inactive umbelliprenin. The umbelliferone obtained was recrystallized to constant radioactivity from ethyl acetate as shown in Figure 25 after partial purification by preparative thin layer chromato-Figure 24. Recrystallization of various coumarins to constant radioactivity. Coumarin Weight Weight Recrystallization Weight after Radioactive Counting Time Isolated Isolated after Number Recrystallization Counts after (minutes) (mg) Dilution (mg.) Recrystallization (mg.) (dpm/mg) umbelliprenin 24.6 40.8 (Rediluted to 64.2) 1 28.6 50,900 40x3(no blank) 2 24.5 19,610 40x3(no blank) 3 22.2 31,750 40x3(no blank) 4 20.6 26,790 40x3 5 18.7 35,400 40x3 .6 .17.7 25,700 40x3 7 14.2 26,650 40x3 8 61.1 3,710 100x3 9 55.8 3,445 100x3 10 53.5 3,335 100x3 alloimperatorin methyl ether 18.1 40.1 (Rediluted to 68.8) 1 2 3 4 29.3 24.7 22.2 19.1 24,000 5,770 1,152 343 40x3(no blank) 40x3(no blank) 40x3(no blank) 40x3 thamnosmin 21.3 31.9 1 2 3 4 5 18.7 16.1 11.5. 10.2 6.7 21,600 13,450 1,542 1,155 368 40x3(no blank) 40x3(no blank) 40x3(no blank) 40x3 40x3 Figure 24.(Continued) Coumarin Weight Weight Recrystallization Weight after Radioactive Counting Time Isolated Isolated after Number Recrystallization Counts after (minutes) (mg) Dilution (mg) (mg) Recrystallization (dpm/mg) isopimpenellin 32.8 49.5 1 36.3 1,130 40x3(no blank) 2 30.2 650 40x3(no blank) 3 28.5 596 40x3(no blank) 4 27.6 349 40x3 5 19.3 186 40x3 6 16.8 175 40x3 7 15.6 171 40x3 (Rediluted 8 45.4 37.3 40x3 to 50.4) 9 43.3 23.3 40x3 10 38.7 13.4 40x3 Ln Ln alloimperatorin methyl ether epoxide 17.8 34.7 1 2 3 4 29.2 22.4 19.4 14.6 1,812 2,035 1,405 266 40x3(no blank) 40x3(no blank) 40x3(no blank) 40x3 - 56 -graphy. Recrystallization Number Radioactive Counts after Recrystallization (dpm/mg) Counting Time (minutes) 1 122 100x3 2 100 100x3 3 113 100x3 Figure 25. Recrystallization of umbelliferone to constant radio-activity. ( i i ) Degradation of Radioactive Isopimpinellin Radioactive isopimpinellin (2) was not subjected to degradative reactions due to the fact that recrystallization to constant radio-3 activity was very d i f f i c u l t and that DL-mevalonic-5- H acid was incorporated apparently only to a very limited extent into this furanocoumarin (Figure 24). ( i i i ) Degradations of Radioactive Alloimperatorin Methyl Ether Since d i f f i c u l t i e s were encountered in attempting to recrystallize alloimperatorin methyl ether (8) to constant radioactivity (Figure 24), this furanocoumarin was f i r s t p a r t i a l l y degraded to alloimperatorin methyl ether d i o l [5-(3'-methyl-2',3'-dihydroxybutanyl)-8-methoxy-psoralen] (3) as described under section (b) ( i i i ) of the Discussion. It was hoped that this latter compound could be recrystallized to constant radioactivity more easily than alloimperatorin methyl ether (8). After isolation the diol was then recrystallized to constant radioactivity from ethyl acetate as shown in Figure 26 after pa r t i a l Recrystal-li z a t i o n Number Weight After Radioactive Counts Recrystallization After Recrystallization (mg ) (dpm/mg) Counting Time (minutes) 1 37.0 32.8 100x3 2 33.0 18.95 100x3 3 25,4 19.85 100x3 4 22.6 17.30 100x3 Figure 26. Recrystallization of alloimperatorin methyl ether diol to constant radioactivity. purification by preparative thin layer chromatoraphy. Alloimperatorin methyl ether diol (3) was subjected to degradation to the aldehyde, compound (44),and acetone (45) which i n turn were converted to the alcohol, compound (48), and acetone p-bromo-benzenesulfonhydrazone (47), respectively, as described under section (b) ( i i i ) of the Discussion. Both compounds were par t i a l l y purified by preparative thin layer chromatography and then recrystallized to constant activity. An insignificant amount of radioactivity was obtained in the acetone derivative. The alcohol, compound (48), however, had significant radioactivity (Figure 27). Recrystal-li z a t i o n Number Weight After Recrystallization (mg ) Radioactive Counts After Recrystallization (dpm/mg) Counting Time (minutes) 1 8.9 15.4 100x3 2 7.3 12.6 100x3 3 6.8 12.9 100x3 Figure 27. Recrystallization of the alcohol, compound (48), to constant radioactivity. - 58 -Conclusions From the degradation of radioactive umbelliprenin (11) to umbelliferone (20) as described under section (c) (i) of the Discussion, i t is worthy to note that most of the radioactivity present in umbelliprenin (> 95%) was lost. This indicated in turn that most of the original radioactivity present in umbelliprenin must have been present in the farnesyl side chain. Therefore this information would suggest that mevalonic acid is not a direct precursor of the coumarin nucleus or aromatic portion of umbelliprenin i n Thamnosma montana Torr. and Frem. However^as expected, the results do indicate that mevalonic acid is the precursor of the farnesyl unit. The well known conversion of mevalonic acid into isopentenyl pyrophosphate and subsequently into If jtf-dimethylallyl pyrophosphate, followed by head-t o - t a i l condensation of three of these 5-carbon units^is obviously the pathway involved. From the degradation of radioactive alloimperatorin methyl ether C8) to alloimperatorin methyl ether diol (3) and subsequent degradation and conversion to the alcohol, compound (48), and acetone p-bromo-benzenesulfonhydrazone (47), respectively, as described under section Cc) C i i i ) of the Discussion, i t is worthy to note that most of the radioactivity present in compound (3) (> 70%) was retained in compound (48). On the other hand, only a trace of radioactivity was found in compound (47). This result was in accordance with other experimental data by other research workers and by Verma and Young in our laboratories. That i s , in alloi.mperatorin methyl ether, i t was believed that the two carbon atoms (4" and 5') in the furan ring were derived from carbon atoms 4 and 5 of mevalonic acid and that the 5-carbon side chain at - 59 -position 5 was derived also from mevalonic acid by condensation with carbon atom 5, of mevalonic acid (Figure 23). Consequently, i f DL-3 mevalonic acid-5- H i s incorporated into alloimperatorin methyl ether as described, no appreciable loss in radioactivity should occur i n the conversion to compound (48) from compound (3) as the original carbon 5 of mevalonic acid is not lost in such a sequence. The data thus described under Section (c) ( i i i ) supports this theory. - 60 -EXPERIMENTAL Throughout this work column chromatography was usually performed on Woelm neutral alumina. This adsorbent was deactivated to activity IV by the addition of water as directed by the manufacturers. On occasion s i l i c a gel for chromatographic adsorption (B.D.H. reagent) was used for column chromatography. In column chromatography of the crude extract where large quantities of adsorbent were used, the adsorbent used was Shawinigan Aluminum oxide, deactivated by the addition of 6% of a 10% acetic acid solution. Except in large scale column chromatography, the solvents were d i s t i l l e d before use. For analytical and preparative thin layer chromatography (TLC), Woelm neutral alumina and s i l i c a gel G (according to Stahl), containing 2% by weight of General Electric Retma p-1, Type 188-2-7 electronic phosphor,were used. The chromatograms, 0.3 mm in thickness, were a i r dried and activated in an oven at 100°C for three hours. For preparative thin layer chromatography a thicker layer (0.5 mm) of adsorbent was used. Detection of bands in both cases was possible with ultraviolet light. Melting points were determined on a Kofler block and are uncorrected. Radioactivity was measured with a Nuclear-Chicago Mark 1 Model 6860 Liquid S c i n t i l l a t i o n counter in counts per minute (cpm). The radio-activity of a sample in disintegrations per minute (dpm) was calculated - 61 -using the counting efficiency which was determined for each sample by the external standard technique*^ u t i l i z i n g the b u i l t — i n barium 133 gamma source. The radioactivity of the sample was determined using an organic s c i n t i l l a t o r solution made up of the following components: toluene (1 l i t e r ) ; 2,5-diphenyloxazole (4 gm ); and 1,4-bis[2-(5-phenyloxazolyl)J benzene (50 mg ) or an aqueous s c i n t i l l a t o r solution made up of the following components: toluene (385 ml); dioxane (385 ml)j methanol (230 ml); naphthalene (80 gm ); 2,5-diphenyloxazole (5 gm ); and l,4-bis[2-(5-phenyloxazolyl)]benzene (62.5 mg ). In practice an organic sample was dissolved i n benzene (1 ml) i n a counting v i a l or, i n the case of a water soluble sample, the sample was dissolved in methanol (1 ml) i n a counting v i a l . Then, i n both cases, the volume was made up to 15 mis with the above organic and aqueous s c i n t i l l a t o r solutions, respectively. For each sample counted the background activity was determined for the counting v i a l to be used by f i l l i n g the v i a l with one of the above s c i n t i l l a t o r solutions and counting to determine the background in cpm. The counting v i a l was emptied, r e f i l l e d with sample to be counted and the s c i n t i l l a t o r solution, and counted again. The difference i n cpm between the background count and the sample count was used for subsequent calculations. (a) Isolation of Natural Products from Thamnosma montana Torr. and Frem. (i) Extraction of Thamnosma montana Torr. and Frem. The plants for this study were obtained i n the desert region near Morongo Valley, California with the help of Dreyer and his associates at the Fruit and Vegetable Chemistry Laboratory, Pasadena California. - 62 -An acetone extract of the plant was also obtained from the same source. The plants were air dried and ground to a powder in a Waring blender. The a i r dr:ied ground powder C393 gm. ) was extracted with acetone for 6 1/2 hours in a large glass Soxhlet extractor. F i l t r a t i o n and evapora-tion of the solvent gave a crude extract (26.3 gm ) as a dark heavy tar. Extraction of the residue with chloroform and subsequent f i l t r a t i o n and evaporation gave a dark heavy o i l y residue (23.8 gm ). ( i i ) Column Chromatography of Chloroform Extract The crude extract (67 gm ) was applied i n chloroform to the top of a column of deactivated Shawinigan .alumina (2,000 gm , deactivated by the addition of 6% by weight (120 ml ) of 10% acetic acid). Elution of the column was performed with the various solvents as illustrated. Fraction Solvent (volume, ml ) wt (gm ) Compounds petroleum ether (2,000) 0.28 A B D E F 25% benzene in petroleum 0.02 ether (1,000) 33 1/3% benzene in petroleum 0.10 ether (1,000) 50% benzene in petroleum 0.37 ether (1,000) benzene (1,000) 0.30 benzene (2,000) 0.86 unidentified non-polar components such as hydrocarbon or waxes 50% chloroform in benzene (1,750) 1.82 primarily umbelliprenin H I chloroform (750) chloroform (1,125) 0.92 umbelliprenin, iso-imperatorin, alloimpera-1.76 torin methyl ether, thamnosmin, and iso-pimpinellin - 63 -Fraction Solvent (volume, ml ) wt (gm ) J chloroform (875) K chloroform (500) L chloroform (500) 6.42 6.93 2.45 Compounds isopimpinellin, N-methylacridone and alloimperatorin methyl ether epoxide N-methylacridone and alloimperatorin methyl ether d i o l primarily unidentified polar components M chloroform (1,000) 1.40 N acetone (1,000) 0.77 0 acetone (1,000) 3.43 P methanol (1,000) 0.24 The numerous fractions obtained from column chromatography were examined under ultraviolet light on a thin layer s i l i c a gel G chromato-plate. Authentic samples of umbelliprenin (11), isbimperatorin (12), alloimperatorin methyl ether (8), thamnosmin (14), isopimpinellin (2), N-methylacridone (5), alloimperatorin methyl ether epoxide (13), and alloimperatorin methyl ether diol (3) were examined in a similar manner for purposes of comparison. The chromatograms were developed i n an ether-hexane mixture (1:1). Fractions A to F: A portion of fractions A to F when examined by TLC showed only the presence of unidentified non-polar components such as hydrocarbons and waxes. Since these fractions had none of the desirable coumarin components, further examination was not conducted. - 64 -Fraction G: A portion of fraction G when examined by TLC showed the presence of at least five compounds with one component being predominant. When compared with authentic samples, this compound proved to be umbelliprenin (Rf 0.88). Fractions H. and I: A portion of fractions H and I when examined by TLC indicated the presence of at least seven compounds. When compared with authentic samples, umbelliprenin (R^ 0.88), isoimperatorin (R^ 0.85), a l l o -imperatorin methyl ether (R^ 0.75), thamnosmin (R^ 0.67), and isopimpinellin (R^ 0.50) were found present. Fractions J and K: A' portion of fractions J and K when examined by TLC indicated the presence of at least eight compounds. When compared with authentic samples, isopimpinellin (R^ 0.50), N-methylacridone (R^ 0.45), and alloimperatorin methyl ether epoxide (R^ 0.44) were present. Fraction L: A portion of fraction L when examined by TLC indicated the presence of at least three compounds. When compared with authentic samples, N-methylacridone (R^ 0.45) and alloimperatorin methyl ether diol (R^ 0.37) were present. Fractions M to P: A portion of fractions M to P when examined by TLC showed primarily unidentified polar components and these were not examined further in this study. The compounds of particular interest, namely,umbelliprenin (11), - 65 -alloimperatorin methyl ether (8), isopimpinellin (2), and alloimperatorin methyl ether diol (3) were isolated from the various fractions of the column chromatography by additional chromatographic purification and/or crystallization. ( i i i ) Isolation of Umbelliprenin from Fraction G Umbelliprenin (11), was obtained from fraction G (1.82 gm. ) primarily by direct crystallization from hexane. Further amounts were obtained by preparative thin layer chromatography on the mother liquors, using thin layer s i l i c a gel G chromatopates developed in an ether-hexane (1:1) mixture. The combined total was then recrystallized several times from hexane to yield a pure white product (0.210 gm ) which was found to be umbelliprenin when compared with an authentic sample previously isolated and characterized in our laboratories by R.N. Young. (iv) Isolation of Coumarins from Fractions H and I Fractions H and I (2.68 gm ) were combined and adsorbed on about 10 grams of alumina. This portion was then added to a column of alumina (activity IV, 100 gm ) and eluted with various solvents as illustrated. Fraction Solvent (volume, ml ) Wt (mg ) Compounds A petroleum ether (150) 36 unidentified non-polar components B 20% benzene in petroleum ether (150) 249 primarily umbelliprenin C 25% benzene in petroleum ether (150) 564 umbelliprenin, and isoimperatorin D 33 1/3% benzene in petroleum ether (80) 339 umbelliprenin, isoimpera-torin, alloimperatorin methyl ether, and thamnosmin - 66 -Fraction Solvent (volume, ml ) Wt (mg ) 50% benzene in petroleum 447 ether (100) Compounds isoimperatorin, a l l o -imperatorin methyl ether, thamnosmin, and iso-pimpinellin F benzene (80) G chloroform (140) 315 isoimperatorin, a l l o -imperatorin methyl ether, thamnosmin, and iso-pimpinellin 315 isoimperatorin, a l l o -imperatorin methyl ether, thamnosmin, and iso-pimpinellin E chloroform (45) 27 primarily unidentified polar components I chloroform (200) 48 The fractions from the column chromatography were examined under ultraviolet light on a thin layer s i l i c a gel G chromatoplate, developed in an ether-hexane (1:1) mixture. Preparative thin layer chromato-graphy, using similar plates and solvent mixture, was used to obtain the desirable coumarins from fractions B to G, inclusive. Further amounts of umbelliprenin (0.323 gm ) were obtained from fractions B, C, and D by preparative thin layer chromatography, followed by crystallization from hexane. Preparative thin layer chromatography on fractions C to G, inclusive, yielded isoimperatorin (0.258 gm ) which was crystallized from a mixture of hexane and ethyl acetate. Preparative thin layer chromatography on fractions D to G, inclusive,yielded alloimperatorin methyl ether (0.238 gm ) and thamnosmin (0.158 gm ) which were both crystallized from a mixture of hexane and ethyl acetate. Preparative - 67 -thin layer chromatography on fractions E, F, and G yielded isopimpinellin (0.203 gm ) which was crystallized from a mixture of hexane and ethyl acetate. The coumarins isolated were identified by comparison with authentic samples previously isolated and characterized i n our laboratories by R.N. Young. (v) Isolation of Coumarins from Fractions J and K Fractions J and K (~»10 gm. ) were combined and adsorbed on about 40 gm of alumina. This portion was then added to a column of alumina (Activity IV, 400 gm ) and eluted with various solvents as ill u s t r a t e d . Fraction Solvent (volume, ml ) wt (mg ) Compounds A 33 1/3% benzene i n petro-leum ether B . 50% benzene i n petroleum ether C 33 1/3% petroleum ether in benzene D E benzene (300) 288 benzene (200) 577 unidentified non-polar components F benzene (125) 519 G benzene (300) 2,035 isopimpinellin and N-methylacridohe H benzene (250) 2,690 primarily N-methylacridone I chloroform (250) 211 primarily N-methylacridone but some alloimperatorin ' methyl ether epoxide J chloroform (1,000) 264 N-methylacridone and unidentified polar K acetone (1,000) 98 components - 68 -The f r a c t i o n s from the column chromatography were examined under u l t r a v i o l e t l i g h t on a t h i n layer s i l i c a gel G chromatoplate, developed s i n an ether-hexane (1:1) mixture. I s o p i m p i n e l l i n was c r y s t a l l i z e d d i r e c t l y out of f r a c t i o n G and a l l o i m p e r a t o r i n methyl ether epoxide was c r y s t a l l i z e d d i r e c t l y out of f r a c t i o n I, using a hexane-ethyl acetate solvent mixture i n both cases. Further amounts of the two compounds were obtained by preparative t h i n l a y e r chromatography on the mother l i q u o r s of the respective c r y s t a l l i z a t i o n s . The combined t o t a l of each compound was then r e c r y s t a l l i z e d from an ethyl acetate-hexane mixture to y i e l d pure i s o p i m p i n e l l i n (0.692 gm ) and pure a l l o i m p e r a t o r i n methyl ether epoxide (0.057 gm ). The coumarins were i d e n t i f i e d by comparison with authentic samples previously i s o l a t e d and characterized i n our l a b o r a t o r i e s by R.N. Young. (vi) I s o l a t i o n of Alloimperatorin Methyl Ether D i o l from F r a c t i o n L Alloimperatorin methyl ether d i o l (3) was obtained from f r a c t i o n L (2.45 gm ) p r i m a r i l y by d i r e c t c r y s t a l l i z a t i o n from e t h y l acetate. Further amounts were obtained by preparative t h i n layer chromatography on the mother l i q u o r s , using s i l i c a g e l G chromatoplates developed i n an ether-hexane (1:1) mixture. The combined t o t a l was then r e c r y s t a l l i z e d from ethyl .acetate to y i e l d a pale yellow product (0.772 gm ) which was found to be a l l o i m p e r a t o r i n methyl ether d i o l when compared with an 4 authentic sample previously i s o l a t e d and characterized by D.L. Dreyer. - 69 -(b) Development of Appropriate Chemical Degradation Pathways (i) Degradation of Umbelliprenin (11) Umbelliprenin (49.8 mg ) was dissolved i n g l a c i a l acetic acid (1.5 ml ) and concentrated sulfuric acid (4 drops). The reaction mixture was stirred at room temperature for 50 hours. The mixture was treated with sodium hydroxide (5%, 50 ml ) and extracted with ether to remove unreacted material. The aqueous solution was acidified with hydrochloric acid (20%, 50 ml ) and extracted with ether. The ether extract was dried over anhydrous sodium sulfate, f i l t e r e d , and evaporated to dryness. Preparative thin layer chromatography, using a s i l i c a gel G chromatoplate developed in an ether-hexane mixture (3:1), was used to obtain the product which appeared as a bright blue band near the baseline. The compound was crystallized from ethyl acetate to yield 18.6 mg of the desired product. Comparison of the latter by mixed melting point and TLC with a commercially available authentic sample of umbelliferone (20), m.p. 230-231°C, indicated the identity of these compounds. ( i i ) Degradations of Isopimpinellin (2) Degradation of Isopimpinellin (2) to 5,8-Dimethoxy-6-formylumbelliferone(40) An ozonolysis of isopimpinellin (2) was conducted, using an ozonolysis medium prepared by passing ozone from a Welsbach . Ozonator through acetic acid (170 ml ) for 11/4 hours. Isopimpinellin (120 mg , 0.5 mmole) was dissolved in a small amount of acetic acid and added to the solution containing ozone (0.91mmole).The reaction mixture was stirred for 3 3/4 hours followed by reduction with zinc dust (492 mg , 7.6 mmole) for several hours. The solution was f i l t e r e d , evaporated to small volume - 70 -(*-25 ml ) and f i n a l l y taken to dryness i n vacuo. The residue (131 mg ) was dissolved in chloroform and subjected to column chromatography, using s i l i c a gel (25 gm ) and the various solvents as ill u s t r a t e d . Fraction Solvent (volume, ml ) benzene (125) A B 33 1/3% chloroform in benzene (125) 50% chloroform i n benzene (125) 33 1/3% benzene in chloro-form (125) wt (mg ) 10 Compounds unidentified non-polar components E F chloroform (125) chloroform (125) 25 primarily 5,8-dimethoxy-6-formyl-34 umbelliferone G chloroform (125) H chloroform (125) I acetone (125) J • acetone (125) K acetone (125) L methanol (250) 10 12 15 17 i 5 30 unidentified polar components and possibly some 5,8-dimethoxyumbelli-ferone-6-carboxylic acid Fractions C to K, inclusive, from the column chromatography were examined under ultraviolet light on a thin layer s i l i c a gel G chromato-plate, developed in an ethyl acetate-chloroform (2:1) mixture. Preparative thin layer chromatography on fractions E and F yielded a light yellow product (24.8 mg ). Crystallization from acetone gave an analytical sample, m.p. 214-216°C, which was identified as 5,8-dimethoxy-- 71 -6-formylumbelliferone (40) when compared with an authentic sample previously isolated and characterized in our laboratories by R.N. Young. IR (KBr) 1758, 1730 (-C0-0-C=C-, aldehyde carbonyl); 1640, MPOH 1593 (aromatic ring). UV X 226 my (sh, log e 4.02), 275 my (log IU3.X e 4.43); xM e 0 H (NaOH added) 238 my (log e 4.26), 269 my (log e 4.20), max MPOH 299 my (log E 4.08), 360 my (log E 4.16); X (HC1 added) 226 my (sh, log E 4.17), 263 my (log e 4.08), 320 my (log E 4.16). NMR (100 MHz) -2.03 (IH, singlet, phenolic OH), -0.23 (IH, singlet, -CHO), 2.15 (IH, doublet, J = 9.5 Hz, H-C4 cbumarin), 3.72 (IH, doublet, J = 9.5 Hz, H-C3 coumarin), 6.00 (3H, singlet, OMe), and 6.02 (3H, singlet, OMe) Anal. Calcd. for C, 0H,.0,: C, 57.60; H, 4.00. Found: C, 57.38; l z 10 b H, 4.07. Degradations of Isopimpinellin (2) to 5,8-Dimethoxy u mbelliferone-6- carboxylic Acid (41) Isopimpinellin (56 mg ) was reacted with sulfuric acid (3%, 10 ml ) and potassium dichromate (169 mg ) dissolved in water (2 ml ) for 6 hours under reflux. The reaction mixture was extracted with a wide variety of solvents, but no stable product was obtained from these or the residual aqueous reaction mixture. An ozonolysis of isopimpinellin was attempted,using an ozonolysis medium prepared by passing ozone through acetic acid (140 ml ) for 1 hour. Isopimpinellin (120 mg , 0.5 mmole) was dissolved in a small amount of acetic acid and added to the solution containing ozone (0.75mmole). The reaction mixture was stirred for 4 3/4 hours - 72 -followed by oxidation with 30% hydrogen peroxide (0.17 gm , 1.5 mmole) for 1 hour. The solvent was removed in vacuo and the residue (139 mg ) dissolved in chloroform and subjected to column chromatography, using s i l i c a gel (12 gm ) and the various solvents as illu s t r a t e d . Fraction A B D Solvent (volume, ml ) benzene (100) 33 1/3% chloroform i n benzene (50) 50% chloroform in benzene (50) 33 1/3% benzene i n chloro-form (50) chloroform (15) wt (mg ) 13 Compounds unidentified non-polar components chloroform (25) 29 primarily 5,8-dimethoxy-6-formyl-umbelliferone G H chloroform (25) chloroform (50) 15 5,8-dimethoxy-6-formylumbelliferone 21 and possibly some 5,8-dimethoxy-umbelliferone-6-carboxylic acid 50% acetone in chloro-form (25) 21 possibly 5,8-dimethoxy-umbellif erone-6-carboxylic acid J acetone (125) K acetone (125) L methanol' (250) 30 13 31 unidentified polar components - 73 -The fractions from the column chromatography were examined under ultraviolet light on a thin layer s i l i c a gel G chromatoplate, developed in an ethyl acetate-chloroform (1:1) mixture. Preparative thin layer chromatography on fractions F, G, and H yielded some of the aldehyde (27.6 mg ), 5,8-dimethoxy-6-formylumbelliferone (40). However, no significant amounts of the carboxylic acid, 5,8-dimethoxyumbelli-ferone-6-carboxylic acid (41) was obtained either by preparative TLC on fractions G and H followed by sublimation or by direct sublimation of fraction I. ( i i i ) Degradations of Alloimperatorin Methyl Ether (8) Conversion of Alloimperatorin Methyl Ether (8) to Alloimperatorin Methyl Ether Diol (3): Alloimperatorin methyl ether (67.8 mg ., 0.239 mmole) and m-chloro-perbenzoic acid (48.8 mg, ,0.284mmole) were dissolved in separate portions (6 1/2 ml ) of chloroform. The solutions were cooled to 0°C, mixed, and stirred for 11 hours at ice bath temperature. The reaction mixture was washed with 5% sodium bicarbonate and water, respectively. Evaporation to dryness and crystallization from an ethyl acetate-hexane mixture yielded a pale yellow solid identified as alloimperatorin methyl ether epoxide (13), m.p. 103-104°C, when compared with an authentic sample previously isolated and characterized in our laboratories by MPOH R.N. Young. IR (KBr) 1717 (C=0), 1588 (C=C). UV x 221 my (log max e 4.40), 244 my (sh, log e 4.30), 251.5 my (log e 4.36), 266 my (log e 4.31), 305 my (log e 4.11). NMR (100 MHz) in CDC13, TMS lock, 1.82 (IH, doublet, J = 10 Hz, H-C^ furanocoumarin), 2.33 (IH, doublet, J = 2Hz, H-C7 furanocoumarin), 3.11 (IH, doublet, J = 2 Hz, H-C, furano-- 74 -coumarin), 3.64 (IH, doublet, J = 10 Hz, H-C^ furanocoumarin), 5.78 (3H, singlet, OMe), 6.5-7.2 (3H; ABC multiplet; JAB = 14.5 Hz, JAQ = 3 Hz, J,,., = 7.5 Hz; benzylic methylene and epoxide protons), 8.51 and BC o 8.70 (6H, two si n g l e t s , ^ C — C ( C H 3 ) 2 ) . Mass spectrum (m/e). 300(M) , 285 (M-15), 271(M-29), 257(M-43), 229(M-71), 201, 199, 186, 171, and 158. Anal. Calcd. for C^H^O^ C, 67.99; H, 5.37. Found: C, 67.91; H, 5.98. Alloimperatorin methyl ether epoxide was treated with oxalic acid (5%, 25 ml ) for 2 hours under reflux. The reaction mixture was extracted with chloroform which was washed with 5% sodium carbonate and water, respectively. The chloroform solution was dried with anhydrous magnesium sulfate and evaporated to dryness to yield a pale yellow solid C75.5 mg ). Preparative thin layer chromatography, using s i l i c a gel G chromatoplates developed in a chloroform-ethyl-acetate (1:1) mixture yielded a pale yellow solid (65.8 mg ). Crystallization from ethyl acetate gave an analytical sample, m.p. 174-176°C, identified as alloimperatorin methyl ether d i o l (40.7 mg ) when compared with characteriza-4 tion data previously acquired for the compound by D.L. Dreyer. [a]^3°-30.6o (95% EtOH). IR (nujol) 3410 (hydroxyl), 1713 (C=0), 1592 F t " OH (C=C). UV x 220 my (log e 4.46),^244 my (log e 4.19), 251 my (log H13.X e 4.23), 266 my (log e 4.19), 309 my (log e 4.06). NMR (60 MHz) in CDC13-deuteriodimethylsulfoxide, TMS lock, 1.75 (IH, doublet, J = 10 Hz, H-C^ furanocoumarin), 2.25 (IH, doublet, J = 2Hz, H-C^ furanocoumarin), 3.00 (IH, doublet, J = 2Hz, H-C, furanocoumarin), 3.73 (IH, doublet, J = 10 Hz, b H-C3 furanocoumarin), 5.83 (3H, singlet, OMe), 6.53 (IH, t r i p l e t , J = 7Hz, H-C,,, furanocoumarin), 6.87 (2H, doublet, J = 7 Hz, benzylic protons), - 75 -8.70 (6H, singlet, C-methyls). Degradation of Alloimperatorin Methyl Ether Diol (3): Acetone free chloroform was prepared by passing chloroform through a celite column impregnated with. 2,4-dinitrophenylhydrazine and d i s t i l l i n g the eluant. Acetone free methanol was prepared by the addition of iodine (25 gm ) to methanol (1,000 ml ), followed by addition of the mixture to sodium hydroxide (500 ml , 1 N) with constant s t i r r i n g . Water (150 ml.) was added to precipitate iodoform. The mixture was f i l t e r e d , refluxed to remove the iodoform odours, and f i n a l l y d i s t i l l e d . Alloimperatorin methyl ether d i o l (21.3 mg } 0.0671 mmole) was dissolved in acetone free methanol (3 ml ) and periodic acid (47.8 mg 0.215mmole) was dissolved in water (3 ml ). The solutions were mixed and stirred for'*-'24 hours. Nitrogen was flushed slowly through the reaction mixture, which was maintained at'*' 50°C, i n order to d i s t i l l over any acetone formed during the course of the reaction. The acetone was collected as a derivative (47) by d i s t i l l a t i o n into p-bromobenzenesulfon-hydrazide (150 mg ),dissolved in a mixture of water (5 ml )and g l a c i a l acetic acid (5 ml ). Prior to' this d i s t i l l a t i o n a blank was d i s t i l l e d into the reagent in a similar manner. Both the blank and the reaction mixture were extracted with acetone free chloroform. The extracts were washed with water, dried with anhydrous sodium sulfate, f i l t e r e d , and evaporated to dryness. Preparative thin layer chromato-i graphy, using Woelm neutral alumina chromatoplates developed i n a - 76 -chloroform-methanol (20:1) mixture was used to isolate the derivative, acetone p-bromobenzenesulfonhydrazone (R^ 0.62), separate from the reagent, p-bromobenzenesulfonhydrazide (R^ 0.76). The blank was discarded as no significant amount of derivative was apparent when the chromatoplate was analyzed under ultraviolet light. Crystallization from chloroform yielded a white analytical sample (10.A mg), m.p. 146-148°C, identified as the derivative, acetone p-bromobenzenesulfon-hydrazone (47) when compared with an authentic sample previously isolated arid characterized in our laboratories by R.N. Young. IR (KBr) 3220 MPOH ( ^N-H)J 1343, 1180 (-S0„). UV X 235 my (log e 4.15). NMR (100 MHz) i n CDC13, TMS lock, 2.1-2.6 (4H, A 2B 2 multiplet, aromatic protons), 3.24 (IH, broad singlet, ^N-H), 8.12 and 8.24 (6H, two singlets, -N=C-(CH3)2). Mass spectrum (m/e) 292 and 290 (M). Anal. Calcd. for C g H ^ ^ S B r : C, 37.18; H, 3.81; N, 9.62. Found: C, 36,86; H, 3.88; N, 9.40. The aqueous non-volatile portion of the reaction mixture was . extracted with chloroform. The extract, was washed with water, dried with anhydrous sodium sulfate, and evaporated to dryness to yield an aldehyde, compound (44), as a crude residue, identified by TLC compari-son to a sample previously isolated and p a r t i a l l y characterized in our laboratories by R.N. Young. NMR (100 MHz) i n CDC13, TMS lock, 0.19 (IH, singlet, -CHO), 1.83 (IH, doublet, J = 10 Hz, H-C^ furanocoumarin), 2.02 (IH, doublet, J = 2 Hz, H-C? furanocoumarin), 2.87 (IH, doublet, J = 2 Hz, H-Cg furanocoumarin), 3.65 (IH, doublet, J = 10 Hz, H-C3 furano-coumarin), 5.62 (2H, singlet, benzylic protons), 5.82 (3H, singlet, OMe). Mass spectrum (m/e) 258(M), 229(M-29), 214(M-44), 201,, 186, and 158. The residue was dissolved in a mixture of chloroform and methanol (1:1) and cooled to 0°C. Sodium borohydride (41.6 mg ) was dissolved in methanol (*»*1 ml) and added to the solution of the aldehyde. The i reduction was conducted f o r i hour with s t i r r i n g followed by hydrolysis with water (~10 ml ) for 1 hour with s t i r r i n g . The reaction mixture was extracted with chloroform. The extract was washed with water, dried with anhydrous sodium sulfate, f i l t e r e d , and evaporated to dryness to yield the crude product. Preparative thin layer chromatography, using s i l i c a gel G chromatoplates developed in an ethyl acetate-chloroform (2:1) mixture yielded the desired product. Crystallization from ethyl acetate gave a pale yellow analytical sample (5.6 mg), m.p. 167-167.5°C, identified as the alcohol, compound (48)9when compared with an authentic sample previously isolated and characterized i n our labora-tories by R.N. Young. IR (KBr) 3450 (OH), 1705-1690 (C=0), 1585 (C=C). Me OH UV x 220 my (log e 4.39), 251 my (log e 4.32), 265 my (log e 4.24), TT13X 306 my (log e 4.11). NMR (100 MHz) in CDC13-D20, TMS lock, 1.92 (IH, doublet, J = 10 Hz, H-C^ furanocoumarin), 2.32 (IH, doublet, J = 2 Hz, H-C^ furanocoumarin), 3.12 (IH, doublet, J = 2 Hz, H-C^ furanocoumarin), 3.66 (IH, doublet, J = 10 Hz, H-C^ furanocoumarin), 5.78 (3H, single^ OMe), 6.08 (2H, t r i p l e t , J = 6 Hz, -CH_20H) , 6.74 (3H, t r i p l e t , J = 6 Hz, benzylic protons). Mass spectrum (m/e) 260(M), 229(M-31), 214(M-46), 201, 186, and 158. Anal. Calcd. for C 1 4 H 1 2 ° 5 : C ' 6 4 , 6 2 5 H> 4.62. Found: C, 64.66; H, 4.61. Mol. Wt. Calcd. for C 1 4H 1 20 5: 260.068. Found (high resolution mass spectrometry): 260.070. - 78 -(c) Biosynthetic Investigations with Radioactive Mevalonic Acid 3 DL-Mevalonic-5- H acid as i t s dibenzylethylenediamine salt (49) was obtained from the New England Nuclear Corporation for the purpose of conducting the biosynthetic investigation. The radioactivity of the 3 9 DL-mevalonic-5- H acid (DBED salt) was 1.0 millicuries(2.2 x 10 dpm) and the specific activity was 221 millicuries per millimole. The radio-active compound (1.21 mg )was obtained dissolved i n methanol (1.0 ml). In preparation for the biosynthetic feeding the methanol solution of the radioactive compound'was diluted with d i s t i l l e d water to a known volume (25 ml ) in a volumetric flask. 3 (i) Feeding of DL-Mevalonic-5- H Acid (DBED Salt) In a typical biosynthetic feeding an aliquot (6.4060 gm ,6.4060/25 x 9 8 2.2 x 10 dpm = 5.64 x 10 dpm) of the solution of the radioactive compound was fed to six young Thamnosma montana plants ('"'19 months old) for a total exposure time of about 10 days. The aliquot was divided up approximately into three equal portions, each portion to be fed to two plants contained in a test tube. The whole plant was used for the feeding which was hydroponic i n nature with the radioactive compound being incorporated through the roots. After the plants had taken up the i n i t i a l radioactive solution, the test tubes were maintained f u l l with d i s t i l l e d water to a level just near the top of the roots. ( i i ) Isolation of Coumarins from the Biosynthetic Feeding The six plants from the biosynthetic feeding were ground to a powder in a Waring blender. The ground powder («11 gm ) was extracted with acetone (750 ml) for 9 3/4 hours in a Soxhlet extractor. F i l t r a t i o n and evaporation of the solvent yielded a dark residue (779.4 mg ). Extraction of the residue with chloroform and subsequent f i l t r a t i o n and evaporation yielded a further dark residue (549.3 mg ). The total radioactivity of the combined residual aqueous solutions in which the plants were fed was determined (7.41 x 10^ dpm) as was the total radioactivity of the acetone extract (23.2 x 10^ dpm) and the chloroform extract (14.6 x 10^ dpm). From the total radioactivity recovered from outside the plants, the corrected radioactivity fed was established (5.64 x 10 8 dpm-0.0741 x 10 8 dpm = 5.57 x 10 8 dpm). The crude extract (549.3 mg ) was adsorbed on about 3 grams of alumina. This portion was then added to a column of alumina (Activity IV, 30 gm )and eluted with the various solvents as il l u s t r a t e d . Fraction Solvent (volume, ml ) Wt (mg ) Compounds A petroleum ether (125) 63 B petroleum ether (125) 11 C 25% benzene in petroleum 24 ether (125) unidentified non-polar compounds such as hydrocarbons and waxes 33% benzene in petroleum ether (125) 50% benzene in petroleum ether (125) 56 26 umbelliprenin, a l l o -imperatorin methyl ether, and thamnosmin benzene (75) 16 umbelliprenin, allo-imperatorin methyl ether, thamnosmin, and isopimpinellin G benzene (50) H benzene (50) 15 9 isopimpinellin and N-methylacridone - 80 -Fraction Solvent (volume, ml ) Wt (mg ) Compounds I benzene (125) 28 chlorophylls, iso-pimpinellin, a l l o -imperatorin methyl ether epoxide, and N-methylacridone J chloroform (125) 16 primarily unidentified polar components K chloroform (125) 19 L , chloroform (125) 11 The fractions from the column chromatography were examined under ultraviolet light on a thin layer s i l i c a gel G chromatoplate, developed in an ether-hexane (1:1) mixture. Authentic samples of umbelliprenin (11), isoimperatorin (12), alloimperatorin methyl ether (8), t thamnosmin (14), isopimpinellin (2), N-methylacridone (5), alloimperatorin methyl ether epoxide (13), and alloimperatorin methyl ether d i o l (3) were examined in a similar manner for purposes of comparison. Preparative thin layer chromatography was used to obtain umbelliprenin (24.6 mg ), alloimperatorin methyl ether (18.1 mg), thamnosmin (21.3 mg), isopimpinellin (32.8 mg)t and alloimperatorin methyl ether epoxide (17.8 mg), identified by comparison to authentic samples previously isolated and characterized in our laboratories. ' In preparation for degradative work to be performed on some of the coumarins isolated, each coumarin was diluted with a weighed amount of cold material previously isolated from the plant. Attempts were made to obtain by crystallization techniques, constant radioactivity counts for each coumarin as ill u s t r a t e d . Coumarin Weight Weight Recrystalliz- Weight After Radioactive Radioactive Isolated Isolated After ing Solvent Final Recryst- counts After status of (mg) Dilution a l l i z a t i o n Final Pecryst- Coumarin (mg) (mg) a l l i z a t i o n (dpm/mg) umbelliprenin 24.6 40.8 (f i r s t dilution) 64.2 (second dilution) hexane 53.5 3,335 constant (incorporation 0.032%) alloimperatorin methyl ether 18.1 40.1 hexane and ethyl, acetate 19.1 343 not constant thamnosmin 21.3 31.9 hexane and ethyl acetate 6.7 368 not constant isopimpinellin 32.8 49.5 ( f i r s t dilution) 50.4 (second dilution) hexane and ethyl acetate 38.7 13.4 not constant alloimperatorin methyl ether epoxide 17.8 34.7 hexane and ethyl acetate 14.6 266 not constant - 82 -( i i i ) Degradation of Radioactive Umbelliprenin Radioactive umbelliprenin (53.5 mg ,3.335 dpm/mg) was dissolved in g l a c i a l acetic acid (2 ml )and concentrated sulfuric acid (5 drops). The reaction mixture was treated with sodium hydroxide (5%, 50 ml )and extracted with ether. The aqueous solution was acidified with hydrochloric acid (20%, 50 ml )and extracted with ether. The ether extract was dried over anhydrous sodium sulfate, f i l t e r e d , and evaporated to dryness. Preparative thin layer chromatography, using a s i l i c a gel G chromatoplate developed in an ether-hexane (3:1) mixture was used to obtain the product (21.5 mg ),identified as umbelliferone (20) by comparison with a commercially available authentic sample. Umbelliferone was recrystallized several times from ethyl acetate to constant radio-act i v i t y (14.0 mg ,122 dpm/mg; 12.2 mg ,100 dpm/mg; 10.4 mg ,113 dpm/mg). (iv) Degradations of Radioactive Alloimperatorin Methyl Ether Since radioactive alloimperatorin methyl ether (19.1 mg , 343 dpm/mg) was not recrystallized to constant radioactivity e f f i c i e n t l y , i t was diluted again (to 68.6 mg ) with cold material and converted to alloimperatorin methyl ether diol which was recrystallized to constant radioactivity. Conversion of Radioactive Alloimperatorin Methyl Ether (8) to Radioactive  Alloimperatorin Methyl Ether Diol (3): Radioactive alloimperatorin methyl ether (68.8 mg ,0.243 mmole) and m-chloroperbenzoic acid (47.6 mg ,0.278mmdle) were dissolved in separate portions (6 1/2 ml )of chloroform. The solutions were cooled to 0°C, - 83 -mixed, and stirred for 12 1/2 hours at ice bath temperature. The reaction mixture was washed with 1% sodium bicarbonate and water, respectively. The chloroform extract was dried with anhydrous sodium sulfate, f i l t e r e d , and evaporated to dryness to yield crude radioactive alloimperatorin methyl ether epoxide (13), identified by TLC comparison to an authentic sample. Radioactive alloimperatorin methyl ether epoxide was treated with oxalic acid (5%, 26 ml )for 3 hours under reflux. The reaction mixture was extracted with chloroform. The chloroform extract was washed with 5% sodium carbonate and water, respectively, followed by drying with anhydrous sodium sulfate and evaporation to dryness to yield the crude product (62.8 mg ). Preparative thin layer chromato-graphy, using s i l i c a gel G chromatoplates developed in a chloroform-ethyl acetate (1:1) mixture yielded a pale yellow solid (46.1 mg ), identified as alloimperatorin methyl ether diol (3) by TLC comparison to an authentic sample. Alloimperatorin methyl ether diol was recrystallized several times from ethyl acetate to constant radioactivity (37.0 mg,32.8 dpm/mg; 33.0 mg ,18.95 dpm/mg; 25.4 mg ,19.85 dpm/mg; 22.6 mg ,17.30 dpm/mg). This represented an incorporation of 0.00007% into the original alloimperatorin methyl ether. Degradation of Radioactive Alloimperatorin Methyl Ether Diol (3): Radioactive alloimperatorin methyl ether diol (22.3 mg ,0.0704 mm) was dissolved in acetone free methanol (3 ml )and periodic acid (47.4 mg ,0.213mmole)was dissolved in water (3 ml ). The solutions were mixed and stirred for/~'24 hours. Acetone, formed during the course of the reaction was isolated as a derivative (47) in a similar manner as i t - 84 -was during the reaction previously conducted on cold material. Crystallization from chloroform yielded the purified derivative (8.0 mg ), identified as acetone p-bromobenzenesulfonhydrazone (47) by TLC comparison to an authentic sample. The derivative when counted, however, displayed insignificant radioactivity. The aqueous non-volatile portion of the reaction mixture was worked up as in the reaction previously conducted on cold material to yield crude aldehyde, compound (44), identified by TLC comparison with an authentic sample. Reduction with sodium borohydride (42.5 mg ) and subsequent hydrolysis and workup as before yielded crude alcohol (9.1 mg), compound (48), identified by TLC comparison with an authentic sample. This degradation product was recrystallized several times from ethylacetate to constant radioactivity (8.9 mg ,15.4 dpm/mg; 7.3 mg t 12.6 dpm/mg; 6.8 mg ,12.9 dpm/mg). - 85 -BIBLIOGRAPHY 1. E.L. Bennett and J. Bonner, Am. J. Botany, 40, 29 (1953). 2. W.H. Muller and H. Muller, Am. J. Botany, 43, 354 (1956). 3. T.H. Kearney and R.H. Peebles, "Arizona Flora", p. 494, University of; California Press, Berkeley and Los Angeles, California, 1960. i 4. D.L. Dreyer, Tetrahedron, 22, 2923 (1966). 5. J.R. Price, "The Alkaloids", Vol. II, p. 353, R.H.F. Manske and H..L. Holmes, Ed., Academic Press, New York, 1952. 6. D.L. Dreyer, private communication. 7. T. Inaba, Ph.D. Thesis, p. 40, The University of British Columbia, 1967. 8. D.B. Sprinson, Advan. in Carbohyd. Chem. , 15_, 235 (1960). 9. S.A. Brown, G.H.N. Towers, and D. Wright, Can. J. Biochem. and Physiol., 38, 143 (1960). 10. T. Kosuge and E.E. Conn, J. Biol. Chem., 234, 2133 (1959). 11.. F. Weygand and H. Wendt, Z. Naturf., 14b, 421 (1959). 12. S.A. Brown, Z. Naturf., 15b, 768 (1960). 13. S.A. Brown, Can. J. Biochem. and Physiol., 40, 607 (1962). 14. S.A. Brown, "Biosynthesis of Aromatic Compounds", p. 15, G. Billek, Ed., Pergamon Press, London, 1966. 15. T. Kosuge and E.E. Conn, J. Biol. Chem., 236, 1617 (1961). 16. J.R. Stoker and D.M. B e l l i s , J. Biol. Chem., 237, 2303 (1962). 17. H.J. Gorz and F.A. Haskins, Crop Sci., 4_, 193 (1964). 18. J.R. Stoker, Biochem. Biophys. Res. Comm.,14, 17 (1964). 19. S.A. Brown, G.H.N. Towers, and D. Chen, Phytochemistry, 3>> 469 (1964), 20. D.J. Austin and M.B. Meyers, Phytochemistry, 4_, 245 (1965). 21. D.J. Austin and M.B. Meyers, Phytochemistry, 4_, 255 (1965). 22. S.A. Brown, Phytochemistry, _2, 137 (1963). - 86 -23. S.A. Brown, Can. J. Biochem. , 43_, 199 (1965). 24. K. Chambers, G.W. Kenner, M.J.T. Robinson, and B.R. Webster, Proc. Chem. Soc, 291 (1960). 25. CA. Bunton, G.W. Kenner, M.J.T. Robinson, and B.R. Webster, Tetrahedron, 19, 1001 (1963). 26. E. Grisebach and W.D. O l l i s , Experientia, 17, 4 (1961). 27. A.I. Scott, P.A. Dodson, F. McCapra, and M.B. Meyers, J. Amer. Chem. Soc. , 85, 3702 (1963). 28. S.A. Brown, Lloydia, 26, 211 (1963). 29. H. Grisebach andW. Barz, Z. Naturf., 18b, 466 (1963). 30. H. Grisebach and W. Barz, Z. Naturf. , 19b, 569 (1964). 31. G. Kunesch and J. Polonsky, Chem. Comm., 317 (1967). 32. T.R. Seshadri, Tetrahedron, 6^, 173 (1959). 33. W.D. O l l i s , Experientia, 22_, 777 (1966). 34. K.G. Edwards and J.R. Stoker, Phytochemistry, 6^, 655 (1967). 35. K.G. Edwards and J.R. Stoker, Phytochemistry, ]_, 73 (1968). 36. G. Caporale, A. Breccia, and G. Rodighiero, "Prepn. Bio-Med. Appl. Labeled Mol., Proc. Symp", p. 103, Venice, 1964. 37. K.G. Floss and U. Mothes, Z. Naturf., 19b, 770 (1963). 38. R.' Aneja, S.K. Mukerjee, and T.R. Seshadri, Tetrahedron, 4_, 256 (1958). 39. H.G. Floss and U. Mothes, Phytochemistry, _5» 161 (1966). 40. H.G. Floss and H. Paikert, Phytochemistry, 8^, 589 (1969). 41. F. Weygand,.H. Simon, H.G. Floss, and U. Mothes, Z. Naturf. 15b, 765 (1960). 42. H.G. Floss, H. Guenther, and L.A. Hadwiger, Phytochemistry, .8, 585 (1969). 43. W; Steck, M. El-Dakhakhny, and S.A. Brown, Tetrahedron Letters, 54, 4805 (1969). - 87 -44. E. Spath, Ber. dtsch. Chem. Ges., 71, 1667 (1938). 45. E.A. Abu-Mustafa and M.B.E. Fayez, J. Org. Chem., 26_, 161 (1961). 46. A. Chatterjee and S.S. Mitra, J. Amer. Chem. Soc, 73^ , 606 (1949). 47. R.J.W. Cremlyn, J. Chem. Soc. (C), 1229 (1966). 48. "Mark 1 Liquid S c i n t i l l a t i o n Systems Manual", Nuclear Chicago Company, p. 18, Section I CL966). 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0060040/manifest

Comment

Related Items