J. Plant Res. 111: 599-604, 1998 Journal of Plant Research (~) by The Botanical Society of Japan 1998 Short Communication Biosynthesis and Metabolism of Purine Alkaloids in Leaves of Cocoa Tea (Camellia ptilophylla) Hiroshi A s h i h a r a ' * , M i s a k o Kato' a n d Y e C h u a n g xing 2 Department of Molecular Biology and Biochemistry, Graduate School of Life Sciences, Ochanomizu University, Otsuka, Bunkyo-ku, Tokyo, 112 8610 Japan 2 School of Life Sciences, Zhongshan University, Guangzhou 510275, China The biosynthesis and metabolism of purine alkaloids in leaves of Carne/I/a pt#ophylla (cocoa tea), a new tea resource in China, have been investigated. The major purine alkaloid was theobromine, with theophylline also being present as a minor component. Caffeine was not accumulated in detectable quantities. Theobromine was synthesized from E8-~=C]adenine and the rate of its biosynthesis in the segments from young and mature leaves from flush shoots was approximately 10 times higher than that from aged leaves from I year old shoots. Neither cellfree extracts nor segments from C. pt#ophylla leaves could convert theobromine to caffeine. A large quantity of F2~C] xanthine taken up by the leaf segments was degraded to ~ via the conventional purine catabolic pathway that includes allantoin as an intermediate. However, small amounts of [2-~C-l xanthine were also converted to theobromine. Considerable amounts of [8 ~4C] caffeine exogenously supplied to the leaf segments of C. ptilophylla was changed to theobromine. These results indicate that leaves of C. ptilophylla exhibit unusual purine alkaloid metabolism as i) they have the capacity to synthesize theobromine from adenine nucleotides, but they lack adequate methyltransferase activity to convert of theobromine to caffeine in detectable quantities, ii) the leaves have a capacity to convert xanthine to theobromine, probably via 3methylxanthine. Key wordr Can~lia ptilophylla~Cocoa tea---Purine alkaloid metabolism ~ Theobromine Purine alkaloids, such as theobromine (3, 7-dimethylxanthine) and caffeine (1, 3, 7-trimethylxanthine), have been found in more than 60plant species (Kihlman 1977). Recently, extensive studies of purine alkaloid metabolism in leaves of tea (Camellia sinensis) and coffee (Coffea arabica) have elucidated the biosynthetic and catabolic pathways of caffeine in some detail (Ashihara et al. 1996a, b, 1997, Crozier et al. 1998). It has been shown that (i) caffeine is synthesized in young leaves via xanthosine--,7 methylxanthosine-, 7-methylxanthine ,theobromine ,caffeine pathway in which the three methylation steps are catalyzed by S adenosylmethionine (SAM)-dependent N-methyltransferases (NMTs) (ii) degradation of caffeine occurs gradually in the leaves via caffeine-, theophylline--* 3-methylxanthine--, xanthine- ~uric acid ,allantoin ,allantoic acid-*~CO2+NH3 degradation pathway. In contrast to tea and coffee plants, some purine alkaloid producing species accumulate theobromine rather than caffeine. This is the case in seeds of Theobroma cacao (cocoa) (Senanayake and Wijesekera 1971) and leaves and flowers of Camellia irrawadiences (Nagata and Sakai 1985, Ashihara and Kubota 1987, Fujimori and Ashihara 1990). In Paullina cupana (guarana), caffeine is the major purine alkaloid in cotyledons, testa, aril and septa, while theobromine predominates in the pericarp and leaves (Baumann et al. 1995). Little is known about the metabolism of purine alkaloids in theobromine rich plants. Although other possibilities exist, one explanation is that they lack SAM: theobromine NMT activity and are therefore unable to convert theobromine to caffeine. It has however yet unclear as to how many NMTs are involved in the three methylation steps in the caffeine biosynthetic pathway. In tea leaves, the available evidence suggests that one SAM-dependent NMT catalyses the last two steps of methylation in the caffeine biosynthetic pathway (Kato eta/. 1996~ Kato and Ashihara 1997). Similar bifunctional NMT has been partially purified from coffee endosperm (Mazzafera et al. 1994). In contrast, Baumann and his colleagues have been claimed that two different NMTs participate the last two steps in caffeine biosynthesis in Coffea arabica (Baumann eta/. 1983, M6sli Waldhauser et al. 1997). In 1988, Chang eta/. discovered C. ptilophylla as a new tea resources, which was named cocoa tea. Subsequently, Ye et al. (1997) observed that theobromine accumulated in young shoots of this species in concentrations as high as 7% of the dry weight, which is well above the levels observed in guarana, cocoa and C. irrawadiences. Leaves of C. ptilophylla were obtained, in January, from plants growing in a natural field at the Zhongshan University, Guangzhou, China. The developmental stages of the leaves were categorized as (i) young, (ii) mature, and (iii) aged. Young leaves were the most recently emerged, weighed ca. 120 mg (fresh weight) and were ca. 35 mm long and 15 mm in width. Mature leaves comprised the fully expanded second leaf below the apex (ca. 90 mm long and 45 mm in width, H. Ashihara et al. 600 weight ca. 680 mg) from newly emerged shoots; while similarly sized aged leaves (weight ca. 700 rag) from more than 1 year old shoots near the base of trees were dark green. Radiochemicals were purchased from Moravek Biochemicals Inc. (Brea, CA, USA). All other chemicals were purchased from Sigma (St. Louis, MO, USA). Purine alkaloids were extracted from dried leaves (100 mg) with methanol. After complete evaporation of the methanol, purine alkaloids were dissolved in distilled water (5 ml) and aliquots (5 10/zL) were analysed by HPLC. Analysis was carried out with a Shimadzu HPLC system, Type LC 10A on a ferruleless column (250 mmX4.6 mm i.d.) packed with a 5 /zm ODS Hypersil support (Shandon, Runcorn, Cheshire, UK) eluted a flow rate of 1 mL min 1with a 25 min gradient 0 40% gradient of methanol in 50 mM sodium acetate, pH 5.0. This gradient elution reversed phase HPLC system separates 11 different purine derivatives and is also able to resolve ~4Clabelled uric acid, allantoin, and allantoic acid (Ashihara et al. 1996b). The absorbance at 270 nm and radioactivity (in the case of tracer experiments) were monitored using a Shimadzu UV VlS monitor, type SPD-10A, and a Raytest radioanalyzer, model Ramona 2,000 (Raytest isotopenmessgerate Gmbh, Straubenhardt, Germany), respectively. In the case of tracer experiments with 14C-labelled compounds, C. ptilophylla leaf segments (ca. 4 mmX4 mm, 100 mg fresh weight) were incubated in 2 mL medium, comprised of 30 mM potassium phosphate buffer, pH 5.6, 10 mM sucrose and a radiolabelled substrate. The incubation was carried out in a 30 mL Erlenmeyer flask, in a shaking water bath at 27 C for 21 hr. The 30 mL Erlenmeyer flask had a centre well containing a small glass tube into which was inserted a piece of filter paper wetted with 0.1 mL of a 20% aqueous potassium hydroxide solution. At the end of the incubation period, the glass tube and filter paper from the centre well were transferred to a 50 mL-flask containing 10 mL of distilled water, after thorough shaking, radioactivity in a 0.5 mL aliquot was determined by liquid scintillation counting in order to estimate the amount of ~4CO2 released. The leaf segments were separated from the incubation medium by filtering through a tea strainer. They were then washed with 50 mL distilled water and transferred to 1 mL of extraction medium, comprised of 20 mM sodium diethyldithiocarbamate in 80% methanol and stored at --80 C. The leaf segments were subsequently ground with a mortar and pestle with the same extraction medium (ca. 5 mL), the resultant tissue homogenate was centrifuged at 12,000Xg for 5-min and the supernatant and pellet separated. The pellet was resuspended in extraction medium (ca. 5 mL) and recentrifuged. The supernatant fractions, containing the methanol-soluble metabolites were combined and reduced to dryness in vacuo, aliquots were analysed by TLC using 200 x 200 mm sheets of micro crystalline cellulose (Spotfilm, Tokyo Kasei Kogyo Co., Tokyo, Japan). The solvent system used was n-butanol/acetic acid/water (4 : 1 : 2, v/v) which separated the radioactive metabolites, which had been identified with the HPLC radiocounting. The radioactivity of 14C on the TLC sheets was analysed using a Bio Imaging Analyzer (Type FLA-2000, Fuji Photo Film Co., Ltd., Tokyo, Japan). Extraction, desalting, and assay of NMT activity from young C. ptilophylla leaves were carried out according to the methods described by Kato et al. (1996). Briefly, C. ptilophylla leaves (ca. 1.3 g fresh weight) were homogenized in 100 mM sodium phosphate buffer (pH 7.2), containing 5 mM 2 mercaptoethanol, 5 mM Na2EDTA, 0.5% (w/v) sodium ascorbate, and 2.5% insoluble polyvinylpolypyrrolidone. The homogenate was centrifuged at 20,000Xg for 15 min and the supernatant was treated with solid ammonium sulphate. The protein fraction precipitated between 30 and 80% saturation was collected and dissolved in 50 mM TrisHCl buffer (pH 8.5) containing 2 mM Na2EDTA, 2 mM 2mercaptoethanol, and 20% glycerol. After desalting with Sephadex G-25, the protein fraction was used immediately for the NMT assay. The reaction mixture contained 100 mM Tris-HCl buffer (pH 8.5), 0.2 mM MgCI2, 0.2 mM substrate, 4.3 ,uM [methyl 14C]SAM (specific activity 2.15GBq mmol 1), and enzyme preparation in a total volume of 100 ,uL. The reaction was initiated by the addition of the enzyme preparation and the mixture was incubated in a shaking water bath at 27 C for 15 min. The enzymatic reaction was terminated by the addition of 1 mL of chloroform. After shaking, a 0.5 mL aliquot of the lower organic layer was dried, scintillation cocktail added and the level of 14C-labelled product(s) determined by liquid scintillation counting. Table1 indicates clearly that the major purine alkaloid accumulated in C. ptilophylla leaves was theobromine together with a small amount of theophylline. Neither caffeine nor paraxanthine (1,7 dimethylxanthine) were detected. These results confirm the previous observation reported by Ye et al. (1997) using the young leaves and shoot of C. ptilophylla. Table1. Levelsof endogenous purinealkaloids in young, matureand aged leavesof cocoa tea (Camellia ptilophylla) plants Common name Theobromine Theophylline Paraxanthine Caffeine Chemical name 3, 7-dimethylxanthine 1,3-dimethylxanthine 1,7-dimethylxanthine 1,3, 7-trimethylxanthine Young 145.4___7.2 9.44-0.4 nd~ nd Data expressedas/zmoles g-1 dry weight_sB (n----6)are shown. nd, not detected. Mature 55.5+--3.3 1.1 +--0.1 nd nd Aged 48.8+2.2 3.9___0.2 nd nd Purine Alkaloid Metabolism in Cocoa Tea Leaves 601 Table 2. Overall metabolism of 9.1 ,uM [8-~4C]adenine (specific activity 2.04 GBq mmol-~) in young, mature, and aged leaves of cocoa tea (Camellia ptilophylla) plants Metabolites a) Methanol-soluble compounds ATP Young Mature Aged 22.7+-2.7 1.64+-0.12 38.24-1.3 3.38+-0.15 17.1 4-1.1 6.46+-0.49 SAM + ADP AM P 6.364-0.21 1.074-0.20 9.784-0.91 2.834-0.03 3.07 4-0.40 1.644-0.22 Allantoic acid Allantoin 0.144-0.14 0.334-0.03 0.38+_0.02 0.83+-0.02 0.654-0.15 0.854-0.08 Xanthine Urea Adenine Theobromine Theophylline 0.254-0.94 2.12+-0.22 1.364-0.18 9.12+- 1.94 nd* 1.624-0.05 6.204-0.17 1.95+-0.12 9.974-0.34 nd 0.304-0.01 1.87+-0.08 1.14__.0.05 0.96+-0.24 nd Caffeine Unidentified metabolite nd 0.26+-0.02 nd 1.284-0.07 nd 0.194-0.02 16.2+-2.8 23.24-0.8 28.54-1.4 61.2_+5.5 38.6+-2.1 54.4+-2.5 27.1 4-0.9 26.9+_0.8 24.6___0.4 b) Methanol-insoluble compounds (mainly nucleic acids) c) CO2 Total uptake (kBq) Duration of incubation was 21 hr. The rates of incorporation of radioactivity into individual metabolites are expressed as percentage of total radioactivity taken up by the leaf segments. Total uptake is expressed as kBq 100 mg-1 fresh weight. Average values and___sg(n=3) are shown. * nd, not detected. Table3. Metabolism of 4.5/zM [2-~4C]theobromine (specific activity 2.07GBq mmol-~) and 4.8 ~M [2-~4C]xanthine (specific activity 1.94GBq mmo1-1)in young leaves of cocoa tea (Camallia ptilophylla) plants Metabotite a) Methanol-soluble compounds Allantoin Xanthine Theobromine Theophylline Caffeine b) Methanol-insoluble compounds c) CO2 Total uptake (kBq) [2-14C]theobromine [2-14C]xanthine 92.6+-0.4 nd* nd 3.47+-1.24 0.64+0.35 1.55+-0.43 92.6+_0.4 nd nd 0.474-0.09 nd nd 3.794-0.29 3.64+_0.14 3.564-2.26 93.04-3.5 3.72+_0.26 7.22+--0.29 Duration of incubation was 21 hr. Values are expressed as percentage of total radioactivity taken up by the leaf segments. Total uptake is expressed as kBq 100 mg-1 fresh weight. Average values and+-sg (n=3) are shown. nd, not detected. Since adenine is the most effective precursor for the synthesis of caffeine in tea leaves (Suzuki and Takahashi 1976), [8 ~4C]adenine was selected as a precursor to examine the biosynthesis of purine alkaloid in C. ptilophylla leaves. Adenine is readily salvaged for adenine nucleotide synthesis by adenine phosphoribosyltransferase in plants, appearing to be converted to AMP and entering the AMP route of purine alkaloid biosynthesis (for review, see Suzuki et al. 1992). The metabolic fate of [8-14C]adenine in young, mature, and aged C. ptilophylla leaves is shown in Table 2. After a 21- hr incubation, more than 98% of radioactivity taken up by the segments was metabolized. The most heavily labeled compounds were catabolites of purine nucleotides such as CO2, xanthine, allantoin, allantoic acid, and urea, together with methanol-insoluble compounds, comprising mainly nucleic acids. Less than 10% of radioactivity w a s recovered in theobromine. Theobromine biosynthesis in aged leaves w a s reduced to one tenth of that in young and mature leaves. Despite being detected as an endogenous purine alkaloid along with 602 H. Ashihara et aL XMF ~.~ IMP (4) (3) ~ AMP ~, (2) Adenine (1) 0 O ~ N ~ ,N H Ribose Xa~thosine SAM l (5) SAH 4-~ ~.__L "> N H N Ribose 7-Methylxanthosine H20 - ~ (13) Uricacid ~ eurine Catabolism H (11) O Allantoin ~ (14) (15) H Glyoxylate+ Urea O~.NAN': ~ (1~) CO2+ NH3 &H 3 3-Methylxanthine ;;Meth~lxanthine // O (12) ~ Xanthine (19) 0 /CH 3 HN ~ , , ~ N + O O H HN~,% O~.NJ~N/2 N "CH3 1,3.D~n;ctphhYl~a~n~ine O /CH3 O~NALZ 3,7-Dimethylxanthine SAM SAH (Theobromine) (8) 1,3,7-Trimethylxanthine (Caffeine) Fig. 1. Possiblepurine alkaloid metabolic pathways operating in leaves of cocoa tea (Camellia ptilophylla). Arrow with a vertical bar represents a blocked conversion. Enzymes: (1) adenine phosphoribosyltransferase; (2) AMP deaminase; (3) IMP dehydrogenase; (4) 5'-nucleotidase; (5)SAM: xanthosine 7N methyltransferase; (6) 7-methylxanthosine nucleosidase; (7) SAM: 7-methylxanthosine 3N-methyltransferase; (8) SAM: theobromine 1N methyltransferase; (9) 7N-demethylase; (10) 1N-demethylase; (11) 3N-demethylase; (12) xanthine dehydrogenase; (13) uricase; (14) allantoinase; (15) allantoicase; (16) urease; (17) 1Ndemethylase; (18) 7N-demethylase; (19) SAM: xanthine 3N-methyltransferase; (20) SAM: 3 methylxanthine 7N-methyltransferase; (21) purine nucleosidase; (21) SAM: 3-methylxanthine 1N-methyltransferase. Several enzymes shown above have not yet been demonstrated in higher plants. IMP, inosine 5'-monophosphate; SAH, S adenosylhomocysteine; XMP, xanthosine 5'-monophosphate. theobromine, incorporation of label into theophylline was not observed. Radioactivity was recovered as AMP, ADP, ATP and SAM. It should be borne in mind that the turnover of adenine nucleotides and SAM is likely to be much faster than that of the purine alkaloids (Yabuki and Ashihara 1991, Y. Yama and H. Ashihara, unpublished observation). Thus, most of the [8-~4C]adenine applied to the leaves may be converted to adenine nucleotides during the early stages of incubation, with a limiting level of substrate lowering incorpo- ration in the later stages. The high incorporation of radioactivity into ATP in aged leaves, shown in Fig. 1, does not necessarily reflect an elevated rate of ATP synthesis. It is more likely to be a consequence of a slow rate of further metabolism of ATP in the aged leaves compared to young leaves where there is a rapid turnover of ATP. The data shown in Table 2 indicate a higher rate of theobromine production in new young and mature leaves from the flush shoots of C. ptilophylla than in aged leaves from old shoots. However, theobromine biosynthesis in young leaves of C. ptilophylla appears to be slower than the rate of caffeine biosynthesis in young leaves of C. sinensis (Ashihara et al. 1995). Under similar experimental conditions, more than 50% of the [8 14C]adenine taken up by young C. sinensis leaves was incorporated into caffeine compared with less than 10% for theobromine in the present study (Ashihara et al. 1995). The available evidence suggests that a single NMT participates in the conversion of 7-methylxanthine to caffeine in tea leaves (Kato et al. 1996, Kato and Ashihara 1997). Substrate specificity studies with crude and highly purified NMT preparations from young tea leaves demonstrated that while paraxanthine was the most active methyl acceptor, 7methylxanthine and theobromine could also be utilised (Fujimori et al. 1991, Ashihara et al. 1995, Kato et al. 1996, Kato and Ashihara 1997) The substrate specificity of NMT activity in gel filtrated extracts from young leaves of C. ptilophylla was investigated. With 7-methylxanthine, NMT activity, 0.16 pkat mg 1 protein, was detected with the substrate being converted to theobromine. In contrast, the preparations failed to methylate the dimethylxanthines, theobromine, theophylline, and paraxanthine (1,7-dimethylxanthine). Thus, the NMT preparations from C. ptilophylla and C. sinensis have different substrate specificities and, as a consequence, the main caffeine biosynthetic pathway in tea via theobromine and the minor pathway via paraxanthine (Kato et al. 1996), are both blocked in C. ptilophylla. Experiments with young leaf segments of C. ptilophylla which demonstrated that [2 14C]theobromine was not converted to caffeine in detectable quantities (Table 3), provided further evidence of the absence of SAM: theobromine NMT activity in C. ptilophylla. This is the major fate of theobromine in leaves of both C. sinensis (Ashihara et al. 1997) and Coffea arabica (Ashihara et al. 1996a, b). The data presented in Table 3 suggest that [2 ~4C] theobromine is catabolized very slowly by C. ptilophylla leaves, with over 90% of the recovered radioactivity being unmetabolized substrate and only 3.6% being released as ~4CO2. In marked contrast, [2-1~]xanthine was catabolized readily by the young leaf segments. More than 90% of the recovered radioactivity was 14CO2while xanthine represented less than 2%. There was also trace incorporation of label into theobromine indicating the existence of a minor salvage pathway. A similar pathway, with 3 methylxanthine acting as the intermediate, occurs in C. sinensis leaves where the salvaged theobromine is further converted to caffeine (Ashihara et al. 1997, Ito et al. 1997). Although no label was associated with theophylline in this 21 h incubation, it is Purine Alkaloid Metabolism in Cocoa Tea Leaves 603 Table 4. Metabolismof 9.6,uM [8-14C]caffeine(specific activity 1.92GBq mmo1-1) by leaf discs from young, mature and aged leaves of cocoa tea (Camellia ptilophylla) plants Sample Young Mature Aged Total uptake (kBq) 9.51 + 1 . 4 1 10.35___1.83 11.02-+0.92 Caffeine (%) 86.88--+0.99 91.18-+0.67 68.39_+0.47 Theobromine (%) 8.25-+0.68 5.61 ___0.58 8.26-+0.47 Theophylline (%) nd* nd nd CO2 (%) 2.02-+1.20 0.19___0.04 0.28-+0.12 Duration of incubationwas 21 hr. Valuesare expressedas percentageof total radioactivitytaken up by the leaf segments. Total uptake is expressed as kBq 100 mg-1 fresh weight. Averagevalues and--+ SD(n----3)are shown. * nd, not detected. feasible that endogenous theophylline could originate from 3-methylxanthine. Data obtained following the incubation of young, mature and aged leaf segments of C. ptilophylla with [8-~4C] caffeine are presented in Table 4. There was relatively little metabolism of the substrate and only small amounts of 14CO2were released. However, a significant amount of the E8 14C] caffeine, 5 8% of the recovered radioactivity, was metabolized to [~4C~theobromine, a conversion, which involves the action of a 1N demethylase. Recent work by Ashihara et al. (1996b, 1997) has demonstrated that the conversion of caffeine to theobromine is not present in leaves of either C. sinensis or Coffea arabica, although this reaction has been observed in studies with coffee fruits (Suzuki and Waller 1984, Mazzafera et al. 1991). The C. ptilophylla 1N-demethylase that converts [8 14C]caffeine to E14C]theobromine is, however, unlikely to play a significant role in the production of endogenous theobromine as the leaves of C. ptilophylla do not contain detectable quantities of caffeine (Table 1). Figure 1 summarises the likely pathways involved in purine alkaloid biosynthesis in C. ptilophylla. Leaves of C. ptilophylla, like those of C. sinensis and Coffea arabica, possesses enzymes which convert purine nucleotides to theobromine (steps 2 7). However, the NMT which catalyses the methylation of theobromine to caffeine (step 8) is either absent or its activity is extremely low in young leaves of C. ptilophylla and, as a consequence, theobromine rather than caffeine is the predominant purine alkaloid. In addition, C. ptilophylla appears to possess an alternative minor route to theobromine. This is the salvage of xanthine to theobromine, which in C. sinensis proceeds via a xanthine~3 methylxanthine~theobromine pathway (steps19 and 20). Arguably, 3-methylxanthine may be the immediate precursor of theophylline (step 21) which, while an endogenous purine alkaloid in C. ptilophylla (Table1), paradoxically, did not accumulate in any of the metabolism experiments. Studies with E2-14C~theobromine indicate that it is relatively stable with only small amounts undergoing demethylation at the 3N and 7N positions (steps 18 and 11), to xanthine which enters the conventional purine catabolic pathway to be degraded to CO2 and NH3 (steps 12 16). The authors thank Dr. Alan Crozier, the University of Glasgow, for his kind offer of the specially packed HPLC column and for his critical reading of the manuscript. This work was supported in part by a Grant-in-Aid for Scientific Research (N0.10640627) from the Ministry of Education, Science, Sports and Culture of Japan to H.A. References Ashihara, H., Gillies, F.M. and Crozier, A. 1997. Metabolism of caffeine and related purine alkaloids in leaves of tea (Camellia sinensis L.). Plant Cell Physiol. 38:413 419. Ashihara, H. and Kubota, H. 1987. Biosynthesis of purine alkaloids in Camellia plants. Plant Cell Physiol. 28: 535 539. Ashihara, H., Monteiro, A.M., Gillies, F.M. and Crozier, A. 1996a. Biosynthesis of caffeine in leaves of coffee (Coffea arabica L.). Plant Physiol. 111: 747-753. Ashihara, H., Monteiro, A.M., Moritz, T., Gillies, F.M. and Crozier, A. 1996b. Catabolism of caffeine and related purine alkaloids in leaves of Coffea arabica L. Planta 198:334 339. Ashihara, H., Shimizu, H., Takeda, Y., Suzuki, T., Gillies, F.M. and Crozier, A. 1995. Caffeine metabolism in high and low caffeine containing cultivars of Camellia sinensis. Z. Naturforsch. 50c: 602 607. Baumann, T.W., Koetz, R. and Morath, P. 1983. N Methyltransferase activities in suspension cultures of Coffea arabica L. Plant Cell Rep. 2 : 3 3 35. Baurnann, T.W., Schulthess, B.H. and Hanni, K. 1995. Guarana (Paullinia cupana) rewards seed dispersers without intoxicating them by caffeine. Phytochemistry 39:1063 1070. Chang, H., Ye, C., Zhang, R., Ma, Y. and Zhang, P. 1988. A discovery of new tea resource Cocoa tea tree containing theobromine from China. Acta Sci. Natur. Univ. Sunyatseni 27: 131-133. Crozier, A., Baumann, T.W., Ashihara, H., Suzuki, T. and Waller, G.R. 1998. Pathways involved in the biosynthesis and catabolism of caffeine in Coffea and Camellia. 16th Coll. Sci. Int. Cafe Nairobi (in press). Fujimori, N. and Ashihara, H. 1990. Adenine metabolism and the synthesis of purine alkaloids in flowers of Camellia. Phytochemistry 29: 3513-3516. Fujimori, N., Suzuki, T. and Ashihara, H. 1991. Seasonal variations in biosynthetic capacity for the synthesis of caffeine in tea leaves. Phytochemistry 30: 22452248. Ito, E., Crozier, A. and Ashihara, H. 1997. Theophylline metabolism in plants. Biochim. Biophys. Acta 1336: 604 H. Ashihara e t al. 323-330. Kihlman, B. 1977. Caffeine and Chromosomes. Elsevier, Amsterdam. Kato, M. and Ashihara, H. 1997. Purification of caffeine synthase from tea leaves. Proc. 61st Annu. Meet. Bot. Soc. Jpn, Narashino, Chiba, p. 208. Kato, M., Kanehara, T., Shimizu, H., Suzuki, T., Gillies, F.M., Crozier, A. and Ashihara, H. 1996. Caffeine biosynthesis in young leaves of Camellia sinensis: in vitro studies on N=methyltransferase activity involved in the conversion of xanthosine to caffeine. Physiol. Plant. 9 8 : 6 2 9 636. Mazzafera, P., Crozier, A. and Magalhaes, C. 1991. Caffeine metabolism in Coffea arabica and other species of coffee. Phytochemistry 3 0 : 3 9 1 3 3916. Mazzafera, P., Wingsle, G., Olsson, O. and Sandberg, G. 1994. S AdenosyI-L-methionine: theobromine 1 N methyltransferase, An enzyme catalysing the synthesis of caffeine in coffee. Phytochemistry 47:1577 1584. Mosli Waldhauser, S.S., Kretschmar, J.A. and Baumann, T.W. 1997. N methyltransferase activities in caffeine biosynthesis: biochemical characterization and time course during leaf devel- opment of Coffea arabica. Phytochemistry 44:853 859. Nagata, T. and Sakai, S. 1985. Purine base pattern of Camellia irrawadiensis. Phytochemist~ 24:2271 2272. Senanayake, U.M. and Wijesekera, O.B. 1971. Theobromine and caffeine content of the cocoa bean during its growth. J. Sci. Food Agric. 22:262 263. Suzuki, T., Ashihara, H. and Waller, G.R. 1992. Purine and purine alkaloid metabolism in Camellia and Coffea plants. Phytochemistry 31:2575 2584. Suzuki, T. and Takahashi, E. 1976. Caffeine biosynthesis in Camellia sinensis. Phytochemistry 15:1235 1239. Suzuki, T. and Waller, G.R. 1984. Biodegradation of caffeine: Formation of theophylline and caffeine in mature Coffea arabica fruits. J. Sci. Food Agric. 3 5 : 6 6 70. Yabuki, N. and Ashihara, H. 1991. Catabolism of adenine nucleotides in suspension cultured plant cells. Biochim. Biophys. Acta 1073: 474-480. Ye, C., Un, Y., Zhou, H., Cheng, F. and Li, X. 1997. Isolation and analysis of purine alkaloids from Camellia ptilophylla Chang. Acta Sci. Natur. Univ. Sunyatseni 36: 30-33. (Received June 26, 1998: Accepted October 13, 1998)