The chemistry of khat

Sections

ABSTRACT
Introduction
Present status of the chemistry of khat Phenylalkylamines and related compounds

Details

Author: K. SZENDREI
Pages: 5 to 35
Creation Date: 1980/01/01

The chemistry of khat

K. SZENDREI Deparonent of Pharmacognosy, University Medical School of Szeged; Hungary

ABSTRACT

This paper presents a review of literature on the chemical composition of khat (Catha edulis Forsk., Celastraceae). The effect of chewing flesh khat could not be explained satisfactorily by the action of d-norpseudoephedrine which was, for a long time, believed to be the only stimulant in khat. A comprehensive study on the chemical composition of khat was undertaken at the United Nations Narcotics Laboratory with the aim of isolating and characterizing the principles of the flesh plant active on the central nervous system. This work resulted in the detection and isolation of cathinone, a phenylaikylamine characterized as (-)-&alpha-aminopropiophenone. It is the main phenylalkylamine component of fresh khat, and pharmacological studies indicate that it may be the compound responsible for the characteristic stimulant activity and abuse potential of the plant. Some of its "decomposition" or transformation products, such as norpseudoephedrine, norephedrine, 3,6-dimethyl-2,5-diphenyipyrazine, and 1-phenyl-1,2-propanedione, have also been isolated and characterized.

Introduction

The chemistry of khat has been an intriguing puzzle to both plant chemists and pharmacologists for almost a hundred years. The plant is alkaloid-containing, but in contrast to the situation with other alkaloid-containing stimulant plants, our knowledge of khat has progressed slowly and its interpretation has remained fairly controversial until quite recently. Despite the progress achieved since the first attempt to characterize its active principle, the actual active principle either was overlooked [1] or simply escaped isolation and characterization for various reasons, including poor quality of the starting material, inadequate isolation procedures and insufficent purity of the final product. Thus, until not very long ago, the characteristic stimulant activity of the fresh plant material could not be fully explained in terms of the then known khat components.

In his comprehensive review on the medical aspects of khat chewing, Halbach [2] concluded that "for a more detailed assessment of the effects of khat in comparison with other types of commonly used psychostimulants such as coffee, a complete analysis of the genuine active substance in the khat plant would be of great value".

The chemical study of khat goes back to 1887 when Fl?ckiger and Gerock [3] , searching for caffeine as the possible stimulating principle, found no traces of it but discovered instead an alkaloid they named katin. In 1891, Mosso [4] extracted from the plant a basic fraction with stimulant-like properties and called it celastrine. However, the first comprehensive study on khat was carried out by Beitter [[l] , [5] ], who obtained crystalline salts of a substance he concluded was identical to both Fl?ckiger's katin and Mosso's celastrine. The chemical composition of khat was next studied by Stockman [[6] , [7] ], who described three distinct alkaloids, carbine, cathinine and cathidine, without characterizing them structurally. Although the cathine he described was in all probability identical to the substance found by former workers, neither cathinine nor cathidine was later identified as an individual substance [8] . Nevertheless, they have frequently been mentioned in the literature.

An important step forward was the contribution of Wolfes [9] , who, using a technique similar to Beitter's, detected the presence of (+)-norpseudoephe-drine (1) in khat, and concluded that this substance corresponded to katin. He also observed the presence of a water-insoluble base, which, like Stockman's cathidine, can be regarded as an impure representative of the polyester-type khat alkaloids.

(+) - norpseudoephedrine

Full size image: 3 kB, (+) - norpseudoephedrine

In subsequent studies, it was repeatedly stated that (+)-norpseudoephe-drine was the main if not the only phenylalkylamine-type constituent present in the plant. Whereas Alles et al. [10] , and Winterfeld and Bernsmann [11] concluded that carbine was the only extractable base present in substantial amounts in the plant, several authors were able to demonstrate, time and again, the presence of other alkaloidal compounds. Depending upon the extraction and chromatographic procedures, Paris and Moyse [[12] , [13] ] detected three to six alkaloids. They suggested that one of the components might be ephedrine, a suggestion which was repeated by Ristic and Thomas [14] , and Karawya et al. [15, 16], but which could never be substantiated by unequivocal isolation and characterization procedures. Karawya and his co-workers separated three alkaloidal products in addition to cathine and ephedrine: cathinine, cathidine and eduline.[1] However, no structures were proposed for them. Later R?cker et al.[17] , using gas-liquid chromatography/mass spectrometry (GLC-MS), indicated the presence of seven nitrogen-containing substances in the basic fraction, but again, with the exception of cathine, none was adequately characterized.

As mentioned earlier, most authors concluded that the characteristic stimulating effects of khat could be satisfactorily explained by its cathine content. The first to question this conclusion and suggest that the chemistry of khat might be more complex than expected was yon Br?cke [18] . On the basis of simple pharmacological experiments, he felt that the stimulating effect of cathine was too small to be alone responsible for the effect of the fresh plant. He suggested the possible presence of a substance with a more powerful stimulating action. His suggestion was partially supported by Alles et al.[10] and by the fact that consumers show preference and pay ahigher price for fresh khat. Most of the earlier chemical and pharmacological studies were performed on dried plant material of varying quality.

The first serious attempt to resolve this question was undertaken by Brilla [19] and Friebel and Brilla [20] , who searched for a specific substance in the fresh plant that might have a greater activity than cathine. Using a combination of chemical and pharmacological methodology, they compared the effect on locomotor activity of synthetic (+)-norpseudoephedrine oxalate with that of the oxalates prepared from freeze-dried and air- and sun-dried khat samples. The three preparations had qualitatively similar effects, but the oxalate from the freeze-dried plant sample showed a stronger effect on locomotor activity. Differences were also found in the physical and chemical characteristics of the samples, and the authors concluded that the substance isolated from the freeze-dried plant was a cathine-like compound, possibly a labile precursor of cathine, for which no correct structure could be proposed on the basis of the data available.

The earlier indications of the presence in khat of water-insoluble alkaloids such as cathidine were checked by Cais et al.[21] , who carried out extensive extraction and separation experiments on a weakly basic khat alkaloid fraction they described as a mixture of closely related alkaloids. One of these compounds, named cathidine D, was isolated in crystalline form and a partial structure proposed for it. It was suggested that this alkaloid, because of its high molecular weight (611) and complicated nature, might be related to alkaloids like evonine and maytoline originating from the same plant family. Unfortunately, no alkaloid structure of this type was known at that time, mainly because of the lack of adequate techniques for structure determination. A new era in this field began in 1970 with the determination of the chemical structures of maytoline and evonine [22-28]. These structures served as keys for a rapidly increasing number of related structures from various celastraceous plants. On this basis, Cais et al. proposed two structures, ( 2a) and ( 2b), for cathidine D [29] .

Full size image: 13 kB

cathidine D (proposed structures)

Full size image: 5 kB

euonyminol

Using an old "cathidine" sample as starting material, that is, an alkaloid mixture that in the best case might correspond to the original mixture in the plant, Luftmann and Spiteller established the structure of the sesquiterpene core of "cathidine". It was thought to be a mixture of polyesters of a sesquiter-pene polyol, euonyminol (III), and various acids such as acetic acid, benzoic acid, gallic acid trimethyl ether, nicotinic acid and evoninic acid [30] . However, none of these alkaloids was isolated from the mixture.

It has also been reported that khat contains tannins [1, 5, 10, 12, 13, 18]; in one study [31] , some of these tannins were found to be of fiavonoid nature, glycosides of kaempferol ( 4), quercetin ( 5) and myricetin ( 6). In addition, several amino acids [11] , sugar alcohols [32] and ascorbic acid [33, 34] were detected or isolated from the plant.

Full size image: 5 kB

(4): R1= R3 = H (5):R1 = OH;R2 = H (6):R1 = R2 = OH

Q?dan analyzed the volatile components of khat by thin-layer and gas-liquid chromatography (TLC and GLC) and found about 40 components in the essential oil distilled from thc plant [35] . Eleven of these compounds were identified as ocimene ( 7), &beta-phellandrene ( 8), terpinolene ( 9), &alpha-and &beta-pinene ( 10) and ( 11), nerol ( 12), linalool ( 13)&alpha-terpineol ( 14), &alpha- and &beta- thujone ( 15) and ( 16), and fenchone ( 17).

Full size image: 19 kB

Present status of the chemistry of khat Phenylalkylamines and related compounds

Upon the recommendation of the Commission on Narcotic Drugs [36] , a research programme was initiated at the United Nations Narcotics Laboratory on the chemical composition of khat, with special emphasis on the isolation and characterization of the active substances in the fresh leaves. This research was considered to be fundamental for the subsequent evaluation of the effects on health and society caused by the chewing of khat [2] . After careful evaluation of the often confusing chemical and pharmacological literature on khat the following points were selected as guidelines:

  1. In view of the indications of the presence in the fresh leaves of a labile precursor of cathine that would be active on the central nervous system (CNS) and of the consumers' preference for fresh material, it was considered essential that the plant material used in the studies be fresh or that its chemical composition be as close to that of the fresh material as possible;

  2. The sparse data, especiallythe observations of Brilla [19] on the stability of the product isolated from the fresh plant and those of Cais et al. [29] on the chemical nature of cathidine D, suggested that careful separation and purification procedures be used in order to avoid undesirable decompositions and artefact formations;

  3. Most of the early investigations were directed exclusively towards the study of the phenylalkylamine-type compounds, particularly in view of the known stimulant activity of cathine. The possible contribution of other types of compounds to the activity of the plant had been neither carefully studied nor excluded. In fact, Paris and Moyse observed that a tincture obtained from a sample from which most of the alkaloids had been removed was nearly as toxic as the untreated plant material [13] . It was therefore considered worthwhile, if not essential, to elaborate comprehensive fractionation and isolation procedures, with emphasis on the various groups of nitrogen-containing substances.

In order to ensure the desired quality of the starting material, fresh khat was acquired directly in countries where khat is cultivated and consumed (Kenya, Madagascar and the Yemen Arab Republic) and was immediately extracted or freeze-dried on the spot. The crude extracts were concentrated under vacuum and together with the freeze-dried and some fresh plant material were kept in a freezer until further investigation. Aliquots of the concentrated extracts were also freeze-dried and likewise stored in a freezer. The presence of various classes of secondary plant products such as phenylalkylamine-type alkaloids (weak bases), neutral terpenoids, polyphenol glycosides and water-soluble alkaloids [37, 38] in the extracts and the freeze-dried samples were monitored using TLC or GLC.

The monitoring for phenylalkylamine-type compounds gave a complicated picture because of interference by amino acids and the presence of a large numberof nitrogen-free substances .Only (+)-norpseudoephedrine was always identified in the original crude extracts, and, although some additional related compounds were also detected by either GLC or TLC, they could not be satisfactorily separated and identified [37] . Therefore, it became evident that further purification of the material under investigation was necessary.

The simplest qualitative picture was observed with the alkaline chloroform-isopropanol or chloroform extract, which was found to contain mainly phenyl-alkylamine-type compounds. Here again, (+)-norpseudoephedrine was the only identifiable compound. Making use of this observation, a simple solvent-solvent fractionation, combined with pH gradient, was applied to the concentrated ethanolic (methanolic) extract, as detailed in [37] and [38] (figure I).

Full size image: 28 kB

Although this procedure givesonly a rough separation of the substances present into four groups, its principal advantage is that it quasi-quantitatively separates the weakly basic alkaloids from the phenylaikylamine-type compounds. The latter accumulate (in fairly pure state) in the chloroform fraction, thus facilitating subsequent chromatographic examinations by making both TLC and GLC pictures much simpler and easier to interpret. It was found that the most striking feature of the TLC chromatograms was that norpseudo-ephedrine was only a minor constituent of the freshly prepared amine fraction. Instead, this fraction appeared to contain one major and some minor amines (figure II). The major compound had a higher R f-value and gave a slightly different colouration with ninhydrine than norpseudoephedrine; whereas nor-pseudoephedrine immediately gave a stable pink colour upon heating, the colour of the new compound changed quickly from orange to pink when heat was applied.

Figure II

Thin-layer chromatogram of the chloroform fraction

Full size image: 22 kB, Figure II

C = CATHINONE

NE = NOREPHEDRINE

NPE = NORPSEUDOEPHEDRINE

E = EXTRACT

In an attempt to separate and identify these compounds, adsorption column chromatography, preparative GLC and coupled GLC-MS techniques were tried, without much success. Considerable decomposition occurred during these procedures resulting in losses and artefact formation. Thus, the concentrated chloroform fraction containing the amine-type compounds was further purified by the classical alkaloid extraction with 0.5 N sulphuric acid and, after alkalinization to pH 9, quickly re-extracted with chloroform. They were then precipitated in ether as oxalates. Thin-layer chromatography and GLC control of each step indicated that some decomposition occurred during the acid extraction and the oxalate precipitation steps (decomposition was also indicated bythe temporary pink-yellow colour transition during these procedures). However, the bulk of the new major compound remained unchanged after the purification and was isolated from the oxalate mixture by repeated crystallizations in methanol. A structure was proposed for the compound based on comparing its UV, IR, NMR and mass spectra with those of norpseudoephedrine (figures III, IV and V), which were described in [38] .

Full size image: 30 kBFull size image: 247 kBFull size image: 33 kB

The spectral evidence led to the conclusion that the isolated new compound had structure (18), that is (-)-&alpha-am nopropiophenone [38] . Although synthetic racemic &alpha-aminopropiophenoneand the pure isomers had been known for a long time [[39] -[41] ], the compound had not previously been isolated from a natural source. It was assigned the trivial name of cathinone and was synthesized by a modified Gabriel synthesis [42] . Direct comparison of the characteristics of the isolated material with those of the synthetic sample established the correctness of the proposed structure.

Full size image: 4 kB

It is to be noted that the absolute configuration of the asymmetric centre of (-)-cathinone is S, the same as that of the corresponding centre of (+)-norpseudoephedrine [[43] ,[ 44] ]. This suggests that the two substances may be bio-genetically closely related. In fact, while studying the biosynthesis of ephedrine in Ephedra vulgaris, Yamasaki et al. [[45] , [46] ] postulated the existence of a cathinone-type intermediate (figure VI) without, however, detecting or isolating it [47] .

Figure VI

Postulated biosynthesis of ephedrine in Ephedra vulgaris according to [[48] ]

Full size image: 13 kB, Figure VI

The chemistry of khat 17Based on its structural relationship to cathine, and the high instability in the presence of oxygen as well as under alkaline conditions, cathinone may be considered as the suspected and long sought labile carbine precursor in khat [[18] -[20] ]. However, it is not yet clear if it is, or under what conditions it could be, converted to cathine in the plant.

It may also be suspected that if cathinone is the main CNS stimulant principle in khat, its CNS stimulant activity should be similar to but probably stronger than that of norpseudoephedrine. Brilla's pharmacological investigations [19, 20] have already been indicated. A comparison of the physical characteristics of Brilla's "new substance" referred to earlier, which was isolated from the freeze-dried plant, with those of pure cathinone clearly indicates that the material used by Brilla may have been an impure cathinone sample. Thus, the pharmacological data obtained with it are to some extent valid for cathinone, and this is supported by preliminary data from pharmacological research currently being carried out in several laboratories [49] .

In addition to cathinone, the purified chloroform extract was found to contain some minor components and, with one exception, these compounds reacted with ninhydrine, suggesting their amine character. Upon TLC analysis, two compounds were always detected under short wave-length UV light in the upper R f region, one of them giving a typical pink colour with ninhydrine, the other reacting only with Dragendorff's reagent (figure II). Also, both compounds were present in the mother liquor containing cathinone oxalate. In addition, appreciable quantities of the compound giving a positive ninhydrine reaction were detected in the original benzene fraction (figure I), which in principle should not contain strongly basic amines.

The Dragendorff-positive compound was isolated from the combined mother liquors of cathinone oxalate, and its probable chemical structure (20) was deduced from its spectral characteristics [38] . In this connection, the IR spectrum (which greatly resembled the spectrum of diphenyltetrazine ( figure VII)) and the NMR spectrum (figure VIII) were especially informative (for discussion see [38] ).

A direct comparison of these spectral data with those of a synthetic 3,6-dimethyl-2,5-diphenylpyrazine sample [42] supported further the proposed structure.

The isolation of this compound from the mother liquor containing cathinone oxalate suggests that it represents one of the main "decomposition" products of cathinone. In fact: this "decomposition", mentioned repeatedly in the earlier literature, is an oxidative dimerization of cathinone base in the presence of oxygen (figure IX). The reaction had already been studied by Gabriel [40] , who, using synthetic material, was able to isolate the corresponding dihydro derivative (19) as an intermediate in the reaction.

Figure VII

IR spectrum of 3,6-dimethyl-2,5-diphenylpyrazine found in khat

Full size image: 173 kB, Figure VII

Figure VIII

NMR spectrum of 3,6-dimethyl-2,5-diphenylpyarazine found in khat

Full size image: 47 kB, Figure VIII

Figure IX

Possible "decomposition" reactions of cathinone

Full size image: 15 kB, Figure IX

In order to isolate the unknown compound which yielded a positive reaction with ninhydrine, the benzene fraction was further purified using column chromatography. A yellowish, oily material was thus obtained which, according to TLC and GLC analyses, still contained some impurity. Its UV spectrum showed similarities to that of cathinone, but its IR spectrum (figure X) was clearly different from the spectra of phenylalkylamine-type compounds [50, 51]. It did not exhibit OH or NH 2 bands, but showed the presence of two intense > C =O bands.

According to the mass spectrum, the molecular weight of the compound was 148, indicating that there was no nitrogen in the molecule. On the basis of its characteristic fragmentation under electron impact (base peak at m/e 105, strong peaks at m/e 77, 51 and 43 (figure XI) and its NMR spectrum (figure XII)), the compound was characterized as having structure (21) corresponding to 1-phenyl-l,2-propanedione. The NMR spectrum of a sample of the synthesized compound provides support for the proposed structure (figure XIII).

Full size image: 158 kB

Figure XI

Fragmentation of 1-phenyl-l,2-propanedione isolated from khat

Full size image: 10 kB, Figure XI

The structure of this new compound gives rise to a number of questions: is this compound a natural product or an artefact formed during isolation from cathinone (figure IX)? If so, is it biosynthetically linked to cathinone? Investigations now in progress may clarify these points [52] . Recent studies on the decomposition of diethylpropione hydrochloride (the diethyl derivative of cathinone) demonstrated that compound (21) was one of the major decomposition products and it was readily formed from the drug by molecular dissociation under various conditions [53] . The exact mechanism involved, however, has not yet been elucidated.

Although TLC and GLC analyses of the chloroform fraction containing the strongly basic amines gave some evidence that one of the minor compounds was norpseudoephedrine (figure II), no clear-cut separation of all the compounds present could be achieved in any of the solvent systems tried. On TLC, the spot corresponding to norpseudoephedrine was always overlapped by another spot representing one or more compounds. The R f value of this spot was the same as the R f of norephedrine,suggesting that norephedrine might also be present in the mixture. This observation is consistent with the GLC results [38] , which showed that, depending on the conditions employed in the analysis of the extracts, the peak for norpseudoephedrine had a more or less pronounced shoulder. This shoulder was absent in chromatograms of synthetic norpseudoephedrine but could be reproduced by adding 10-20 per cent norephedrine to it.

Full size image: 31 kBFull size image: 32 kB

In order to get further information on the chemical nature of the yet unidentified compounds an indirect approach was considered. The GLC-MS combination seemed to offer the best possibilities, provided optimal conditions for a reliable GLC separation and for obtaining intense molecular ion peaks were established. Gas-liquid chromatography had routinely been used in toxicological analysis and for the identification and quantitation of phenylalkylamine-type compounds in doping control [[50] , [51] , [54] -[57] ]. Difficulties have, however, been frequently encountered due to irreversible absorption and strong tailing tendency, thermal decomposition of some compounds, high volatility and losses of some others, and unsatisfactory separations of the respective isomers [ [51] , [56] , [58] ].

It has been suggested in the literature [[54] -[59] ] that derivatization techniques such as acetylation, silylation, Schiff-base or isothiocyanate formation could be used in order to overcome these difficulties. Some of these reactions, for example trifluoroacetylation, have also been used for the quantitative determination of norpseudoephedrine in biological fluids, with excellent results [[56] , [60] , [61] ]. The advantages of the derivatization techniques consist of increased volatility of the derivatives and, for ephedrine-type molecules, greater stability of the molecular ion. This makes possible the application of the combined GLC-MS technique for the analysis of such compounds or mixtures containing them.

Model experiments on pure reference substances and a mixture containing norephedrine, norpseudoephedrine, cathinone, ephedrine and pseudoephedrine gave, after trifluoroacetylation, excellent peak shapes for each compound and a good GLC separation (figure XIV).

The electron impact mass spectra of the reference compounds yielded only weak molecular ions and in some cases did not show any. However, chemical ionization spectra of the derivatives gave in most cases strong molecular ions. This, together with some characteristicfragments, offered sufficient information for the identification or characterization of each compound using its trifiuoroacetyl derivative (figures XV, XVI and XVII).

The characteristic fragment ions and the corresponding relative intensities indicate that the differences in relative intensities between the molecular ions (M+ + 18 in the case where chemical ionization was carried out using ammonia) and the fragment M- 113 are sufficient for a safe identification of the corresponding enantiomers.

After the reference substances had been analyzed, several khat samples of varying origin (Kenya, Madagascar, Yemen Arab Republic) were subjected to the same analytical procedure. Aliquots of the original chloroform fraction (without further purification; see figure I) were reacted with trifluoroacetic anhydride and the derivatized components of the mixture were separated (figure XVIII), and electron impact and chemical ionization mass spectra of each GLC peak were recorded. The components thus identified were: nor new compound and 3,6-dimethyl-2,5-diphenylpyrazine (in increasing R forder).

Figure XIV Gas chromatogram of the trifluoroacetyl (TFA) derivative of the reference mixture

Full size image: 27 kB, Figure XIV Gas chromatogram of the trifluoroacetyl (TFA) derivative of the reference mixture

Figure XV

Chemical ionization mass spectrum of the trifluoroacetyl derivative of norephedrine

Full size image: 12 kB, Figure XV

Figure XVI

Full size image: 14 kB, Figure XVI

Figure XVII

Chemical ionization mass spectrum of the trifluoroacetyl derivative of cathinone

Full size image: 11 kB, Figure XVII

Thus, the GLC-MS data lend support to our earlier results and also suggest that the "shoulder" on the GLC peak for norpseudoephedrine corresponds to norephedrine. Independently, Schorno and Steinegger [44] identified norephedrine in khat by using NMR spectroscopy. However, ephedrine or pseudoephedrine were not detected in any of the plant samples analyzed by us. This finding is in contrast to some earlier references where ephedrine was suggested to be one of the phenylalkylamine-type components of khat [[12] -[16] ].

As shown in figure XVIII, it was found that a khat sample of Kenyan origin contained a substantial amount of an unknown compound with a higher retention time than norephedrine (24.4 min). Its mass spectrometric fragmentation suggested that it is related to cathinone (figures XIX and XX) in that the ethylamine fragment may be at tached to a cinnamoyl group rather than a benzoyl radical. A careful comparison of its mass spectra with those of cathinone suggest ( 22) as a tentative structure. This compound has not yet been detected in plants and the structure proposed for it need further confirmation.

Figure XVIII

Gas chromatogram of the derivetized components of the chlor form fraction

Full size image: 33 kB, Figure XVIII

Figure XIX Chemical ionization mass spectr um of the trifluoroacetyl derivative of the unknown compound

Full size image: 10 kB, Figure XIX Chemical ionization mass spectr um of the trifluoroacetyl derivative of the unknown compound

Figure XX

Electron impact and chemical ionization fragmentation of the trifluoroacetyl derivative of the unknown compound (22)

Full size image: 19 kB, Figure XXFull size image: 3 kB1

"Eduline" is also the name given to an alkaloid isolated and characterized earlier in Casimiroa edulis, a rutaceous plant species, by Beverman and ??

References

001

A. Beitter, "Pharmakognostisch-chemische Untersuchung der Catha edulis", thesis (Strasbourg, 1900).

002

H. Halbach, "Medical aspects of the chewing of khat leaves", Bulletin of the World Health Organization, vol. 47, 1972, p. 21.

003

F. A. Fl?ckiger and J. E. Gerock, "Contributions to the knowledge of Catha leaves", Pharmaceutical Journal of Transvaal, vol. 18, 1887, p. 221.

004

U. Mosso, "Azione fisiologica del principio attivo de Celastrus edulis", Revista clinics, vol. 30, 1891, p. 65.

005

A. Beitter, "Pharmakognostisch-chemische Untersuchung der Catha edulis", Archiv der Pharmazie, vol. 239, 1901, p. 17.

006

R. Stockman, "The active principles of Catha edulis", The Pharmaceutical Journal and Pharmacist, vol. 89, 1912, p. 676.

007

R. Stockman, "The pharmacological action of Catha edulis and its alkaloids", The Journal of Pharmacy and Experimental Therapeutics, vol. 4, 1912, p. 251.

008

A. D. Krikorian and A. Getahun, "Chat: coffee's rival from Harar, Ethiopia. I. Botany, cultivation and use. II. Chemical composition", Economic Botany, vol. 27, 1973, pp. 353, 378.

009

O. Wolfes, "?ber das Vorkommen von d-Norisoephedrin in Catha edulis", Archiv der Pharmazie, vol. 268, 1930, p. 81.

010

G. A. Alles, M. D. Fairchild and M. Jensen, "Chemical pharmacology of Catha edulis", Journal of Medicinal and Pharmaceutical Chemistry, vol. 3, 1961, p. 323.

011

K. Winterfeld and G. Bernsmann, "Zur Kenntnis der Inhaltsstoffe von Catha edulis Forskal", Archiv der Pharmazie, vol. 63, 1960, p. 991.

012

M. R. Paris and H. Moyse, "Essai de caract?risation du Kat ou th? des Abyssins (Catha edulis Forsk., C?lastrac?es), drogue r?cemment inscrite au tableau B", Annales pharmaceutiques fran?aises, vol. 15, 1957, p. 89.

013

M. R. Paris and H. Moyse, "Abyssinian tea (Catha edulis Forsk., Celastraceae)", Bulletin on Narcotics, vol. 10, No. 2 (1958), p. 29.

014

S. Ristic and A, Thomas, "?ber die Inhaltsstoffe von Catha edulis", Archiv der Pharmazie, vol. 295, 1962, p. 524.

015

M.S. Karawya, M. A. Elkiey and M. G. Ghourab, "A study of the alkaloids of Catha edulis Forsk. growing in Egypt", Journal of the Pharmaceutical Sciences of the United Arab Republic, vol. 9, 1968, p. 147.

016

M. A. Elkiey, M. S. Karawya and M. G. Ghourab, "Estimation of the alkaloids of Catha edulis Forsk. growing in Egypt", Journal of the Pharmaceutical Sciences of the United Arab Republic, vol. 9, 1968, p. 159.

017

G. R?cker and others, "?ber die Alkaloide aus Catha edulis", Planta medica, vol. 24, 1973, p. 61.

018

F. Th. von Br?cke, "?ber die zentral erregende Wirkung des Alkaloides Cathin", Naunyn Schmiedeberg's Archiv f?r experimentelle Pathologie und Pharmakologie, vol. 198, 1941, p. 100.

000

The chemistry of khat 33

019

R. Brilla, "?ber den zentralerregenden Wirkstoff der frischen Bl?ter von Catha edulis Forskal", thesis, Bonn, 1962.

020

H. Friebel and R. Brilla, "?ber den zentralerregenden Wirkstoff der frischen Bl?tter und Zweigspitzen von Catha edulis Forsk.", Naturwissenschaften, vol. 50, 1963,p. 354.

021

M. Cais, D. Ginsburg and A. Mandelbaum, Abstracts of IUPAC Symposium on the Chemistry of Natural Products (Kyoto, 1964), p. 95.

022

S. M. Kupchan, R. M. Smith and R. F. Bryan, "Maytoline, a nicotinoyl sesquiter-pene alkaloid prototype from Maytenus ovatus", Journal of the American Chemical Society , vol. 92, 1970, p. 6667.

023

M. Pailer, E. W. Streicher and J. Leitich, "?ber Evonymus-Alkaloide. 3. Mitt.: Konstitution und Konfiguration von Evonin und Evonolin", Monatshefte f?r Chemie , vol. 102, 1971, p. 1873.

024

A. Kl?sek and others, "Alkaloide aus Evonymus europaea L.", Helvetica Chimica Acta , vol. 54, 1971, p. 2144.

025

H. Wada and others, "Evonine, an alkaloid obtained from Euonymus sieboldiana Blume. I. Relationship of functional groups and partial structure", Tetrahedron Letters , 1971, p. 2655.

026

Y. Shizuri and others, "Evonine, an alkaloid obtained from Euonymus sieboldiana Blume. II. The structure of evonine", Tetrahedron Letters , 1971, p. 2659.

027

H. Wada and others, "Stereochemistry of evonine, neo-evonine, euonymine, and neo-euonymine, alkaloids obtained from Euonymus sieboldiana Blume", Tetrahedron Letters , 1971, p. 3131.

028

K. Sasaki and Y. Hirata, "Structure of new alkaloid, evonine and neoevonine: x-ray analysis of bromoacetyl-neoevonine monohydrate", Journal of the Chemical Society , Perkin II., 1972, p. 1268.

029

M. Cais and others, "Constituents of Catha edulis . Isolation and structure of cathidine D", Tetrahedron, vol. 31, 1975, p. 2727.

030

H. Luftmann and G. Spiteller, "Zur Struktur des Cathidins aus Catha edulis F. Der Polyhydroxygrundk?rper", Tetrahedron, vol. 30, 1974, p. 2577.

031

H. I. El Sissi and M. F. Abd Alla, "Polyphenolics of leaves of Catha edulis", Planta medica , vol. 14, 1966, p. 76.

032

V. Plouvier, "Nouvelles recherches sur le quebrachitol des Sapindac?es et Hippo-castanac?es, le dulcitol des C?lastrac?es et le saccharose de quelques autres families", Comptes Rendues de l&rsquoAcad?mie des Sciences (Paris), vol. 228, 1949, p. 1186.

033

M. J. Mustard, "Ascorbic acid content of some miscellaneous tropical and subtropical plants and plant products", Food Research , vol. 17 1952, p. 31.

034

Nutrition Survey: Ethiopia. Khat, Ethiopian Nutrition Survey (Washington, D.C., 1959), p. 166.

035

S. Q?dan, " Catha edulis , eine wenig bekannte Rausch- und Genussdroge", Planta medica , vol. 21, 1972, p. 410.

036

United Nations, Commission on Narcotic Drugs, Report on Twenty-Fourth Session, para 82 and paras ?

037

"Studies on the chemical composition of khat. I. Extraction, screening investigations and solvent separation of khat components" (MNAR/12/1974).

038

"Studies on the chemical composition of khat. III. Investigations on the phenylalkylamine fraction" (MNAR/11/1975).

039

Chr. Schmidt, "Einwirkung von Phthalimidkalium auf einige sauerstoffhaltige Halogenverbindungen", Berichte der Deutschen Chemischen Gesellschaft, vol. 22, 1889, p. 3249.

040

S. Gabriel, "Wandlungen der Aminoketone", Berichte der Deutschen Chemischen Gesellschaft, vol. 41, 1908, p. 1127.

041

H. Takamatsu, "Studies on optically active phenylpropanolamine derivatives. I. Preparation of optically active &alpha-aminopropiophenones by asymmetric transformation", Journal of the Pharmaceutical Society of Japan, vol. 76, 1956, p. 1219.

042

"Studies on the chemical composition of khat. VIII. Note on the synthesis of cathinone and its &lsquodimer&rsquo (3,6-dimethyl-2,5-diphenylpyrazine)" (MNAR/3/1978).

043

H. Takamatsu, "Studies on optically active phenylpropanolamine derivatives. III. Synthetics of 1-ephedrine and its analogs", Journal of the Pharmaceutical Society of Japan, vol. 76, 1956, p. 1227.

044

X. Schorno and E. Steinegger, "The phenylalkylamines of Catha edulis Forsk.: the absolute configuration of cathinone" (MNAR/7/1978).

045

K. Yamasaki, U. Sankawa and S. Shibata, "Biosynthesis of ephedrine in Ephedra", Tetrahedron Letters, 1969, p. 4099.

046

K. Yamasaki and others, "Participation of C 6-C1 unit in the biosynthesis of ephedrine in Ephedra", Phytochemistry, vol. 12, 1973, p. 2877.

047

S. Shibata, personal communication.

048

K. Nakanishi and others, Natural Products Chemistry, vol. 2 (New York, Academic Press, 1975), p. 301.

049

J. L. Zelger, E. A. Carlini and X. Schorno, "Psychopharmacological investigations in the central stimulant activity of cathinone (&alpha-aminopropiophenone) obtained from Catha edulis", Planta medica, vol. 39, 1980, p. 244.

050

E. G. C. Clarke, Isolation and Identification of Drugs, vols. 1-2 (London, Pharmaceutical Press, 1974-1975).

051

M. Steinigen, "Praktische Hinweise zur Analytik der Sympathomimetica vom Typ der Phenylalkyl- und Phenylakanolamine", Deutsche Apotheker-Zeitung, vol. 113, 1973, p. 81.

052

X. Schorno, personal communication.

053

M. J. Walters, "High-performance liquid chromatographic determination of di-ethylpropion hydrochloride in tablets: isolation and identification of two decomposition products", Journal of Pharmaceutical Sciences, vol. 66, 1977, p. 198.

054

A. H. Beckett and B. Testa, "Gas-liquid chromatographic separation of the optical isomers of some &lsquoephedrines&rsquo and &lsquopseudoephedrines&rsquo", Journal of Pharmacy and Pharmacology, vol. 25, 1972, p. 382.

055

G. Munro and others, "The gas chromatographic analysis of the diastereomeric forms present in some phenylalkanolamines", Journal of Pharmacy and Pharmacology ?

056

M. Donike, "N-trifiuoracetyl-O-trimethylsilyl-phenolalkylamine. Darstellung und massenspezifischer gaschromatographischer Nachweis", Journal of Chromatography, vol. 103, 1975, p. 91.

057

E.T. Lin, D. C. Brater and L. Z. Benet, " Gas-liquid chromatographic determination of pseudoephedrine and norpseudoephedrine in human plasma and urine", Journal of Chromatography, vol. 140, 1977, p. 275.

058

I. L. Martin and (3. B. Baker, "Procedural difficulties in the gas-liquid chromatographic assay of the arylalkylamines", Journal of Chromatography, vol. 123, 1976, . p. 45.

059

H. Brandenberger and D. Schnyder, "Toxikologischer Spurennachweis von biologisch wirksamen Aminen (Amphetamine, Catecholamine, Halluzinogene) durch Gas-Chromatographie und massenspezifischer Detektion (GLC-MD)", Fresenius Zeitschrift f?r analytische Chemie, vol. 261, 1972, p. 297.

060

F. Frosch, "Biopharmazeutische Untersuchungen an D-Nor-pseudoephedrin", Arzneimittel-Forschung, vol. 27, 1977, p. 665.

061

H. Ch. Curtius and others, "Mass fragmentography of dopamine and 6-hydroxy-dopamine. Application to the determination of dopamine in human brain biopsies from the caudate nucleus", Journal of Chromatography, vol. 99, 1974, p. 529.