Chlorogenic Acid Biosynthesis: Characterization of a Light-Induced

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 258, No. 1, October, pp. 226-232, 1987

Chlorogenic Acid Biosynthesis: Characterization of a Light-Induced Microsomal 5O-(4-Coumaroyl)-D-quinate/shikimate 3’-Hydroxylase from Carrot (Daucus carota L.) Cell Suspension Cultures’ THOMAS KijHNL, ULRICH KOCH, WERNER HELLER, AND ECKARD WELLMANN

Biologisches Institut II der Universitti Fm&wg, Schiinzlestrasse I, D-7800 Frea’burg, Federal Republic of Germany Received March 19,1987, and in revised form June 16,1987

Microsomal preparations from carrot (Daucus carota L.) cell suspension cultures cat- alyze the formation of tram-50-caffeoyl-D-quinate (chlorogenate) from trans-5-O-(4- coumaroyl)-D-quinate. trans-5-0-(4-Coumaroyl)shikimate is converted to about the same extent to trans-5Gcaffeoylshikimate. trans-4-0-(4-Coumaroyl)-D-quinate, tram-3-0-(4- coumaroyl)-D-quinate, trane-lcoumarate, and c&Z-5-0-(4-coumaroyl)-D-quinate do not act as substrates. The reaction is strictly dependent on molecular oxygen and on NADPH as reducing cofactor. NADH and ascorbic acid cannot substitute for NADPH. Cytochrome c, Tetcyclacis, and carbon monoxide inhibit the reaction suggesting a cytochrome P- 450-dependent mixed-function monooxygenase. Competition experiments as well as induction and inhibition phenomena indicate that there is only one enzyme species which is responsible for the hydroxylation of the 5-0-(4-coumaric) esters of both D-quinate and shikimate. The activity of this enzyme is greatly increased by in V&JOirradiation of the cells with blue/uv light. We conclude that the biosynthesis of the predominant caffeic acid conjugates in carrot cells occurs via the corresponding I-coumaric acid esters. Thus, in this system, 5-O-(4-coumaroyl)-D-quinate can be seen as the final intermediate in the chlorogenic acid pathway. 8 1987Academic press. I~C.

Chlorogenic acid, 5-0-caffeoyl-D-quinic zymatic studies have shown that the es- acid, is one of the most widespread hy- terification of D-quinate requires hydrox- droxycinnamic acid derivatives known in ycinnamic acids activated by conjugation the plant kingdom (1). Various physiolog- with /3-D-glucose or coenzyme A. Kojima ical roles have been suggested for this and Villegas recently reported on the con- compound. These include growth regula- version of 4-~COUmarOyl-D-glUCOSe (2) and tion and disease resistance ((1) and liter- caffeoyl-D-glucose (3) to the corresponding ature cited therein). 5-O-esters of D-quinate by an enzyme Numerous reports on the biosynthesis of preparation from sweet potato. These re- chlorogenic acid have been published. En- actions have not yet been described for other plant systems. Transesterification ’ This work was supported by Deutsche For- via hydroxycinnamoyl-CoA esters me- schungsgemeinschaft (We 567/5-l). diated by hydroxycinnamoyl-CoA:D-quin- * Present address: Gesellschaft ftir Strahlen und ate hydroxycinnamoyl transferase (CQT),4 Umweltforschung Mtinchen, Ingolst&dter Landstrasse 1, D-8042 Neuherberg, FRG. ’ Abbreviations used: CQT, hydroxycinnamoyi- 3 To whom correspondence should he addressed at CoA:D-quinate hydroxycinnamoyl transferase; CST, Lehrstuhl fur Botanik, Biologisches Institut II der hydroxycinnamoyl-CoAshikimate hydroxycinnamoyl IJniversit%t, Schlnzlestraase 1, D-7800 Freihurg, FRG. transferase.

0003-9861/87 $3.00 226 Copyright 0 1987 by Academic Press, Inc. All rights of reproduction in any form reserved. Daucus carom 5-@(4-COUMAROYL)-D-QUINATE/SHIKIMATE 3’-HYDROXYLASE 227 on the other hand, has been demonstrated HO-..? OOH in a wide range of plants (4). In some cases, CQT activity is increased in response to HCf$OhoH external stimuli, such as light (5-7) or low temperature (8). This increase in enzyme activity is accompanied by accumulation of HO... COOH chlorogenic acid in the tissue. NADPH CQT exhibits high specificity for the ac- ceptor D-quinate. The relative specificities - 02 wQoJ-Q(,~ for the CoA esters differ from one plant system to the other, 4-coumaroyl-CoA and FIG. 1. Hydroxylation of trons-5-0-(4-coumaroyl)- caffeoyl-CoA being almost always the only D-quinate to trans-5-0-caffeoyl-n-quinate (chloro- substrates (4). The question remains as to genate). at which level the 3-hydroxylation of the 4-coumaroyl moiety occurs in vivo. MATERIALS AND METHODS The hydroxylation of both free 4-coumarate (9, 10) and 4-coumaroyl-D-quinate Materials. Chlorogenic acid was purchased from Roth (Karlsruhe). Methanol-d, and trimethylsilane (11) has been reported to be catalyzed by were from EGA-Chemie (Weinheim). Cytochrome P- phenolases. These enzymes have, however, 450 specific inhibitors were kindly provided by Pro- not been proved to be specifically involved fessor H. Grisebach, Freiburg. All other chemicals in chlorogenic acid biosynthesis. and enzymes were obtained or prepared as described Recent investigations on elicitor-induced previously (12). parsley (Petroselinum crispurn Miller) cell Cell cultures. Carrot (Daucus curotu) cell suspension suspension cultures have shown that hy- cultures were a kind gift from Dr. U. Matern, Freiburg. droxylation of 5-0-(4-coumaroyl)shikimate The cells were grown in B-5 medium (16) at 25°C in can be mediated by a microsomal cyto- the dark. Two and one half to three grams of 7-day- chrome P-450-dependent mixed-function old cultures was propagated in 40 ml of fresh medium. Light induction Seven-day-old cells were washed monooxygenase (12). A hydroxycinnamoyl- with fresh medium and an inoculum of 10 g was CoA:shikimate hydroxycinnamoyl trans- transferred to 40 ml medium. After 10 to 12 h in the ferase (CST) is induced concomitantly. This dark, the cells were irradiated for 20 h with Osram L suggests the possibility of a flux of 4-cou- 40 W/73 fluorescent tubes (A,,,,, 350 nm, half band- marate into the ester pathway, at which width 40 nm, 7.8 W m-a) (1’7). stage 3-hydroxylation of the coumaroyl Chromatography. TLC on cellulose plates (Merck, moiety occurs. In the case of chlorogenic Darmstadt) was performed with 2% formic acid (I). acid biosynthesis, a similar pathway for 4- HPLC was carried out on a lo-pm Lichrosorb RP-18 coumaroyl 3-hydroxylation might be ex- column (0.5 X 25 cm) (Merck, Darmstadt) using water/ pected, since some earlier tracer studies acetonitrile/acetic acid (87/12/l, v/v/v) (II) and (84/ 15/l, v/v/v) (III) as solvents. have shown 5-o-(4-coumaroyl)-D-quinate Enzymatic preparation of 5-0-(&coumaroyl)-D- to be an intermediate (13,14). PiniC acid. Preparation of 5-@(4-coumaroyl)-D-quinic Carrot (Duucus carob L.) cell cultures acid was carried out according to (12) using 300 nmol are known to produce chlorogenic acid (15). I-coumaroyl-CoA and 1 rmol D-quinate in a volume We have recently found that the formation of 500 ~1. The product was extracted twice with 250 of this compound can be increased by ir- /.d of l-butanol and evaporated to dryness. After TLC, radiation with blue/uv light. the product was purified further by HPLC with solvent In this paper, we report on a particulate III. The product was identified by ‘H NMR spectros- hydroxylase from carrot cell suspension copy in methanol-d4. The data were compared with cultures responsible for the conversion of those obtained from chlorogenic acid. 5-O-(4-coumaroyl)-D-quinate to chloro- 5-@Caffeoyl-D-quinate: 7.56 (d, H-0, 15.9), 7.04 (d, H-2’,ZO), 6.95 (dd, H-6,2.0/8.2), 6.77 (d, H-5,8.2), 6.26 genate (Fig. l), which is induced by blue/ (d, H-a, 15.9), 5.34 (m, H-5,4.3/9.7/9.7), 4.17 (m, H-3, uv light. The relationship between the 4- 5.0/3.3/3.3), 3.72 (dd, H-4,3.2/8.6), 2.24 (m, H&q, 2.1/ coumaroyl3-hydroxylases from carrot and 4.5/13.4), 2.17 (dd, H-Zax, 3.3/14.1), 2.07 (dd, H-Gax, parsley cells is discussed. 9.7/13.3), 2.06 (m, H-Zeq, 2.1/5.0/14.1); (cf. (18)). 228 KUHNL ET AL.

5-0-(4-Coumaroyl)-D-quinate: 7.63 (d, H-8, 15.8), (1.7 TBq/mol). In the presence of an excess of micro- 7.42 (d, H-Z/H-6’, 6.5), 6.76 (d, H-3’/H-5’, ‘7.2), 6.33 (d, somal protein, cinnamate I-hydroxylase yielded up to H-a, 14.9), 5.40 (H-5), 4.24 (H-3), 3.74 (H-4). 1.5 nmol4-[3-‘%]coumarate in the assay. Preparation of enzymes. Crude extracts and micro- Product ident@ztion Chlorogenate prepared by somal fractions were prepared from freshly harvested enzymatic hydroxylation of 5-o-(d-COUUXirOyl)-D- carrot cells as described earlier (12), but with 0.1 M quinate was identified by TLC (solvent I) and HPLC potassium phosphate buffer, pH 7.5, with 1.5 IUM di- (solvent II). thioerythritol and 10% (w/v) sucrose. The microsomal Assaysfor other enzymes. CQT was measured spec- preparations were frozen in liquid nitrogen and stored trophotometrically as described by Ulbrich and Zenk at -70°C. (4). Cinnamate 4-hydroxylase was measured as de- Synthesis of hydrvxycinnamoyl-D-quinic acids. (a) scribed previously (12). The 3-O and 4-O isomers of I-coumaroyl- and caffeoyl- Other analytical methods. Protein was determined D-quinic acids were prepared according to the method by a modified Lowry procedure (12). ‘H NMR spectra of Hanson (19) by keeping a neutral solution of 5-O were recorded on a Bruker WM 300 spectrometer. (I-coumaroyl) quinic acid and chlorogenic acid, respectively, at 90°C for 30 min. The products were ex- RESULTS AND DISCUSSION tracted with I-butanol. (b) cis-5-0-(4-Coumaroyl)-rr- quinic acid and cis-chlorogenic acid were obtained by Detection of the .Y-Hydroxylation of 5-o- irradiation of the corresponding trans isomers dis- solved in methanol with uv light at 254 nm. Assay for the 5-O-(,+xumar&)-D-quinate/shiki- When trans-5-0-(4-coumaroyl)-D-quin- mate f-hydroxylase. The hydroxylating activity with ate was incubated with microsomes pre- 5-0-(4-coumaroyl)-D-quinate and 5-O-(4-couma- pared by Mgz+ precipitation from blue/uv royl)shikimate, respectively, as substrates was mea- light-induced carrot cell suspension cul- sured at 2O’C in the dark in a final volume of 50 ~1. tures in the presence of NADPH, a new The reaction mixture contained 1 nmol of the I-cou- compound could be detected as a blue-white maric acid ester, 50 nmol NADPH, 5 pmol potassium fluorescing spot by TLC on cellulose with phosphate, pH 7.5, and up to 6.5 I.rg protein (micro- solvent I. On treatment with ammonia, the somal pellet). The reaction was terminated after 20 min by addition of 50 ~1 ethanol. After centrifugation fluorescence color changed to yellow-green at 25,OOOgfor 10 min, 40 ~1 of the clear supernatant indicating the presence of a caffeoyl deriv- was analyzed by HPLC (Waters, Konigstein) under ative. The product showed the same Rf isocratic conditions with solvent system II. Elution value as 5-O-caffeoyl-D-quinic acid (chlo- was performed at a flow rate of 2 ml/min and moni- rogenic acid), which was clearly distinct tored at 340 nm. For quantification, the peak areas of from the Rfvalues of the isomeric 3-O- and substrate and product were corrected according to 4-O-caffeoyl-D-quinic acids (Table I). their extinction coefficients at 340 nm (6200 and 15,700 The product was further characterized M-l cm-‘, respectively). by reversed-phase HPLC with solvent II. Assays under exclusion of oxygen and inhibition Under the conditions chosen it eluted at experiments were performed as previously described 4.30 min as did authentic chlorogenic acid (12). The pH optimum was determined with 0.25 M (Table I). When the assay mixture was ex- potassium phosphate (pH 6.0-7.5) and 0.25 M Tris- HCl (pH 7.2-8.8) in the assay. posed to uv light (254 nm) after incubation, Substrate speti@ity. The substrate specificity to- new signals were detected at 12.45 and 6.70 ward the 4-coumaroyl-D-quinate isomers was tested min. These represent the respective ois iso- in mixture experiments. Analysis was carried out by mers of substrate and product. HPLC (solvent II) (Table I). (a) trans-3-o-(4-Coumaroyl)-D-quinate and trans- General Properties of the 5-0- 4-O-(4-coumaroyl)-D-quinate were incubated in a (&Coumaroyl)-D-quinateishikimate standard assay with trans-5-0-(4-coumaroyl)-D- 3’-Hydroxylase quinate (2 nmol each). (b) A mixture of 1.4 nmol bans- and 0.6 nmol cis- Using tra?‘2S-5-O-(4-coumaroyl)-D-quin- 5-O-(4-coumaroyl)-D-quinate was prepared by irra- ate as substrate, crude enzyme prepara- diation with uv light and incubated under standard tions from blue/uv irradiated cells exhib- conditions. ited hydroxylase activities of approx 3 Conversion of free I-coumarate was investigated in pkat/kg. The specific activities could be in- a combined hydroxylase assay with [3-‘%]cinnamate creased lo-fold using microsomes prepared Daucus carota 5-0-(4-COUMAROYL)-D-QUINATE/SHIKIMATE 3’-HYDROXYLASE 229

TABLE I

CHROMATOGRAPHIC PROPERTIES OF THE ISOMERIC HYDROXYCINNAMOYL-D-QUINIC AND -SHIKIMIC ACIDS

HPLC (RP-18, TLC (cellulose, solvent II) RT solvent I) R,

Compound eis trans cis trans

3-@(4-Coumaroyl)-D-quinic acid n.d.” 4.10 0.78 0.64 4-o-(4-Coumaroyl)-D-quinic acid n.d. 13.20 0.74 0.52 5-O-(4-Coumaroyl)-D-quinic acid 12.45 7.90 0.75 0.56 3-0-Caffeoyl-D-quinic acid n.d. 2.85 0.71 0.55 4-0-Caffeoyl-D-quinic acid n.d. 4.60 0.66 0.42 5-O-Caffeoyl-D-quinic acid 6.70 4.30 0.69 0.46 (= chlorogenic acid) 5-0-(4-Coumaroyl)shikimic acid n.d. 20.15 n.d. 0.50b 5-OCaffeoylshikimic acid n.d. 9.30 n.d. 0.41b

a Not determined. *Values taken from (12). by M$+-precipitation, thus demonstrating of 4-coumaroylquinate, only 5-O-caffeoyl- the particulate localization of the enzyme. quinate could be detected as a product by When the enzyme was kept at 3O”C, a HPLC. The 3-O and 4-O isomers of 4-cou- half-life of 10 min for activity was ob- maroylquinate remained unchanged. served. On storage at -7O”C, the enzyme In a further mixture experiment, the was stable over a period of several months. tram and cis isomers of 5-0-(4-coumaroyl)- The standard assay was linear with time D-quinate were used as substrates in a 2:l up to 30 min and with protein up to 6.5 pg. ratio under standard assay conditions. Sc0.5jfor 5-0-(4-coumaroyl)-D-quinate was Only the tram isomer was converted to determined to be approx 4.5 PM. Optimum trana-chlorogenate. Not even traces of cis- activity was observed at pH 7.5 with half- chlorogenate could be detected by HPLC maximal activities at pH 6.3 and 8.6. Sim- analysis. Free 4-coumarate could not serve ilar pH optima have been reported for sev- as a substrate for 3-hydroxylation. eral cytochrome P-450 monooxygenases These data show that the 4-coumaroyl3- (12,20-22). hydroxylase is specific for the 5-O isomers of 4-coumaroylquinate and -shikimate, Sub.strate SpecQicity suggesting that these esters are interme- The 5-0-(4-coumaroyl)-D-quinate/shi- diates in the biosynthesis of caffeoyl con- kimate 3’-hydroxylase converts 5-0-(4- stituents. This has recently been demon- coumaroyl)-D-quinate and 5-O-(4-couma- strated for 5-0-(4-coumaroyl)shikimate in royl)shikimate to the corresponding caffeic parsley (12). acid esters. The activity toward the quinate Since the enzyme from carrot cells is ca- ester is about 30% higher. When both sub- pable of hydroxylating both 4-coumaroyl- strates were mixed in a standard assay 1: quinate and -shikimate esters, the hydrox- 1, the hydroxylating activities were re- ylase from parsley was reinvestigated duced to about 50% as compared with the using the new quantification procedure separate incubations. This suggests that based on HPLC. With 5-0-(4-coumaroyl)- there is one active site responsible for both D-ClUinate as substrate, a hydroxylation reactions. activity of approx 20% as compared to the In a standard assay performed with a 1: respective shikimate ester could be ob- 1:l mixture of the 3-0,4-O, and 5-O isomers served. 230 KUHNL ET AL.

Cofactor Dependence D-quinate and -shikimate, virtually iden- tical effects were observed. Interestingly, 5-O-(4-Coumaroyl)-D-quinate/shikimate both KCN and diethyldithiocarbamate 3’-hydroxylase activity was strictly depen- stimulate the enzyme activity by a factor dent on NADPH as a reducing cofactor. of about 2. Stimulation of cytochrome P- With 5-0-(4-coumaroyl)-o-quinate as sub- 450-dependent monooxygenases by KCN strate S~o,5jwas determined to be 19 PM. has frequently been observed (21-23), al- With NADH and ascorbate, both at final though in the hydroxylase from parsley concentrations of 1 IrIM, no enzyme activity (12) both KCN and diethyldithiocarbamate could be observed. were without effect. Some inhibition was The reaction requires molecular oxygen observed with diethylpyrocarbonate and (Table II). When the microsomes were p-chloromercuribenzoate, indicating the preincubated with an Oz-consuming system participation of histidyl and cysteinyl res- under a nitrogen atmosphere, the activity idues, respectively, in the reaction. was reduced to 5% compared with a control Furthermore, several specific inhibitors incubation without additions in the pres- for cytochrome P-450-dependent enzyme ence of air. Nitrogen alone caused a reduc- systems were tested. Inhibition with cy- tion in activity to 84% .This oxygen depen- tochrome c is of comparable magnitude to dence is of a similar magnitude as that re- that determined for other hydroxylases of ported for the hydroxylase from parsley this type (12, 21, 24, 25). Tetcyclacis with 5-0-(4-coumaroyl)shikimate as sub- was found to inhibit the 5-O-(4-couma- strate (12). A complete incubation but us- royl)quinate/shikimate 3’-hydroxylase ing boiled glucose oxidase showed an ac- substantially. A concentration of 10 ELM tivity increase of 46% as compared to the caused a reduction of the activity to about control assay without additions. 60%) 100 /IM Tetcyclacis nearly completely inhibited 3’-hydroxylation of both sub- Inhibition Experiments strates. Ketoconazole and Ancymidol at Table III shows the effects of various en- concentrations of 100 PM exhibited no zyme inhibitors on 5-O-(4-COUIrX?ZOyl)-D- marked effect, as has been observed with quinate/shikimate 3’-hydroxylase activity. the enzyme of parsley (26) With both substrates, 5-0-(4-coumaroyl)- In an atmosphere containing carbon

TABLE II EFFECTOFOXYGENDEPLETIONON~-O-(~-COUMAROYL)-D-QUINATE/SHIKIMATE 3'-HYDROXYLASEANDCINNAMATE~HYDROXYLASE

Enzyme activity (% )

5-O-(4-Coumaroyl)- quinate/shikimate Cinnamate 3’-hydroxylasen I-hydroxylase

Additions* Air Nitrogen Air Nitrogen

None 100 84 loo 70 10 mM Glucose + 5 U glucose oxidase + 10 U catalase 22 5 11 4 10 mM Glucose + boiled glucose oxidase + 10 U catalase 146 58 123 30

a With 5-0-(4-coumaroyl)-D-quinate as substrate. * After preincubation for 5 min in closed vials under nitrogen with the relevant additions, the reaction was started with NADPH. The incubations were continued under either nitrogen or air. Daucus co/rota 5-0-(4-COUMAROYL)-D-QUINATE/SHIKIMATE 3’-HYDROXYLASE 231

TABLE III

EFFECTOFENZYME INHIBITORSON~'-HYDROXYLASEACTIVITY USING 5-o-(4-COUMAROYL)-D-QUINATE AND&O-(&COUMAROYL)SHIKIMATE ASSUBSTRATES

Enzyme activity (%) Concentration Inhibitor m4 5-0-(4-Coumaroyl)-D-quinate 5-0-(4Coumaroyl)sbikimate

None - 100 100 KCN 1 235 247 10 219 194

Diethyldithiocarbamate 0.2 191 187 2 205 205

EDTA 1 102 114 5 105 99

Diethylpyrocarbonate 2 69 74 10 50 55 pChloromercuribenzoate 0.2 100 n.d. 0.5 68 n.d. 1 54 n.d. Cytochrome c 0.01 63 n.d. 0.05 44 n.d. 0.1 24 n.d

Tetcyclacis 0.091 106 122 0.01 62 57 0.1 2 9

Ketoconazole 0.1 78 99

Ancymidol 0.1 91 102

monoxide and oxygen (9/l, v/v), a reduc- again supports the view that a single en- tion in enzyme activity to 66% as compared zyme species is induced. to a control assay with nitrogen instead of Blue/uv light also stimulates the activity carbon monoxide was found. Irradiation of CQT and the accumulation of chloro- with light at 450 nm during incubation re- genie acid (unpublished work), thus dem- stored the activity to 81% of that of the onstrating that this pathway is under pho- nitrogen-containing control. tocontrol in carrot cells as has already been shown for other plant systems (5-7). The Light Induction light induction of the 4-coumaroyl 3-hydroxylase suggests that this enzyme is in- The activity of the 5-0-(4-coumaroyl)-D- volved in chlorogenic acid biosynthesis. quinate/shikimate 3’-hydroxylase is in- We conclude that the 5-O-(4-coumaroyl)- ducible by blue/uv light. Carrot cells ir- D-quinate/shikimate 3’-hydroxylase cata- radiated for 20 h exhibited four times the lyzes the last step of the chlorogenic acid activity of control cells grown in the dark. pathway in carrot cells. This enzyme, a This induction was observed for both sub- particulate monooxygenase, appears to be strates 5-0-(4-coumaroyl)-D-quinate (from dependent on a cytochrome P-450 system 8 to 33 pkat/kg) and 5-0-(4-couma- as is the case for other hydroxylases spe- royl)shikimate (from 6 to 22 pkat/kg). This cifically involved in phenylpropanoid me- 232 KUHNL ET AL.

tabolism (20, 21, 24, 25, 27, 28). The sub- 3. VILLEGAS, R. J. A., AND KOJIMA, M. (1985) Agric. strate specificity is restricted to the 5-O Biol Chzem..49,263-265. isomers of 4-coumaroyl-D-quinate and 4- 4. ULBRICH, B., AND ZENIC, M. H. (1979) Phytochem- coumaroylshikimate. Free 4-coumarate istry 18.929-933. 5. ULBRICH, B., ST~CKIGT, J., AND ZENK, M. H. (1976) cannot serve as a substrate. Since all en- NaturwissenschoJten 63,484. zymes which have been described as hy- 6. LAMB, C. J. (1977) FEBS Lett. 75,37-41. droxylating 4-coumarate (9, 10) have not 7. ULBRICH, B., AND AMRHEIN, N. (1978) Planta 138, been conclusively shown to be specifically 69-71. involved in caffeate formation, we suggest 8. RHODES, M. J. C., AND WOOLTORTON, L. S. L. (1978) that the 3-hydroxylation may well occur Phytochemistry 17,1225-1229. exclusively at the level of the 4-cou- 9. Burr, V. S., AND LAMB, C. J. (1981) in The Bio- maroyl conjugates, particularly 5-0-(4- chemistry of Plants (Stumpf, P. K., and Corm, coumaroyl)-D-quinate and 5-0-(4-couma- E. E., Eds.), Vol. 7, pp. 627-665, Academic Press, royl)shikimate. Thus, at least in carrot New York. 10. BOLWELL, G. P., AND Burr, V. S. (1983) Phgto- cells, 5-0-(4-coumaroyl)-D-quinate can be chemistry 22,37-45. seen as the last intermediate in the bio- 11. HANSON, K. R., AND ZUCKER, M. (1963) J. Biol synthesis of chlorogenic acid. This confirms Chewz. 238,1105-1115. earlier tracer studies by Hanson with po- 12. HELLER, W., AND K~HNL, T. (1985) Arch. Biochem tato (13). Biophys. 241,453-460. Results from competition experiments as 13. HANSON, K. R. (1966) Phytochemistry 5,491-499. well as induction and inhibition effects 14. NAGELS, L., AND PARMENTIER, F. (1976) Phyb suggest that there is only one enzyme spe- chxnzistq/ 15,703-706. cies responsible for the hydroxylation of 15. HEINZMANN, U., SEITZ, U., AND SEITZ, U. (1977) Plonta 135,313-318. the 5-O-(4-coumaric) esters of D-qUir&& 16. GAMBORG, 0. L., MILLER, R. A., AND OJIMA, K. and shikimate. A further restricted sub- (1968) Exp. CeU Res. 50,151-158. strate specificity for the 3-hydroxylation 17. BEGGS, C. J., AND WELLMANN, E. (1985) Ph~tocher~ step is not necessary, since the CQT, which Photobiol. 41,481-486. is absolutely specific for the esterification 18. MORISHITA, H., IWAHASHI, H., OSAKA, N., AND of D-quinate (4), controls the preceding step KIDO, R. (1984) .J. Chromatogr. 315,253-260. channeling the 4-coumaroyl moiety via 4- 19. HANSON, K. R. (1965) Biochemistry 4,2719-2731. coumaroyl-CoA into the chlorogenic acid 20. GRAND, C. (1984) FEBS Lett. 169,7-11. pathway. 21. HAGMANN, M., HELLER, W., AND GRISEBACH, H. (1983) Eur. J. Biochem 134,547-554. 22. PETERSEN, M., AND SEITZ, H. U. (1985) FEBSLett. ACKNOWLEDGMENTS 188,11-14. 23. FUJITA, M., OBA, K., AND URITANI, I. (1982) Plant We are indebted to T. Heigl and U. Wolfsperger, Phgsiol. 70.573-578. Giidecke AG Freiburg, for proton magnetic resonance 24. HAGYANN, M., AND GRISEBACH, H. (1984) FEBS spectroscopy. We thank C. Beggs, H. Grisebach, and L&t. 175,199-202. U. Matern for critical reading of the manuscript. 25. HAGMANN, M., HELLER, W., AND GRISEBACH, H. (1984) Eur. J. Biochenz. 142,127-131. REFERENCES 26. FRITSCH, H., KOCHS, G., HAGMANN, M., HELLER, W., JUNG, J., AND GRISEBACH, H. (1986) in Sixth 1. HERRMANN, K. (1978) in Progress in The Chem- International Congress of Pesticide Chemistry, istry of Organic Natural Products (Herz, W., pp. 313-316, IUPAC, August 10-15, Ottawa, Grisebach, H., and Kirby, G. W., Eds.), Vol. 35, abstract. pp. ‘73-132, Springer-Verlag, Vienna/Heidel- 27. RUSSELL, D. W. (1971) J. Bd Chem. 246, 3870- berg/New York. 3878. 2. KOJIMA, M., AND VILLEGAS, R. J. A. (1984) Agric. 28. WENDORFF, H., AND MATERN, U. (1986) Eur. J. Biol. Chew. 48,2397-2399. B&hem 161,391-398.