Cinnamyl and pCoumary1 Esters as Intermediates in the ...

3 downloads 35041 Views 1018KB Size Report
number of quinic acid esters, including p-coumaryl-quinic acid, have been previously ..... ter. Chlorogenic acid and coumaryl es- ter. Coumaryl and cin- namyl esters. -. C g. -. 6 oncentra tion of .... labeled z-phenylalanine and glucose- Cl4 into.
THEJOURNAL Vol.

BIOLOGICAL CHEMISTRY

OF 235, No.

8, August 1960 Printed in U.S.A.

Cinnamyl and pCoumary1 Esters as Intermediates in the Biosynthesis of Chlorogenic Acid* CARL C. LEVY~ AND MILTON ZUCKER From the Departments of Biochemistry and of Plant Pathology and Botany of The Connecticut Agricultural Experiment Station, New Haven, Connecticut (Received for publication, February 5, 1960)

2418

Downloaded from http://www.jbc.org/ by guest on August 13, 2017

The pathway of biosynthesis of chlorogenic acid, a major pheaction. The absorption maximum of the solution is at 549 mp. nolic component present in a wide variety of plants, is, to a large Quinic acid, which differs in structure only in lacking the hyextent, unknown. Its structure was fully elucidated in 1932 by droxyl group at position 6, was found to give the same reaction, Fischer and Dangshat (l), who showed that caffeic acid (3,4-diand an assay procedure modified slightly from that of Srinivasan hydroxycinnamic acid) is esterified by its carboxyl group to the and Sprinson was accordingly developed. To a conical centrihydroxyl at position 3 of quinic acid. Other workers (2-4) have fuge tube containing 0.25 ml of a quinic acid solution (0.01 to found that phenylalanine and truns-cinnamic acid, derived pre0.08 pmole), 0.25 ml of 0.025 N HI04 in 0.125 N HzSOa was added. sumably by way of the shikimic acid pathway of Davis (5), give The solution was shaken and the reaction allowed to continue rise to the caffeyl moiety of a number of compounds. Since for 20 minutes at room temperature. The excess periodate was previous work (6) had shown that potato tuber tissue offers an then removed by the addition of 0.5 ml of a 2 y0 solution of sodium ideal material for biosynthetic studies of chlorogenic acid, an arsenite in 0.5 N HCl. The tube was shaken vigorously for a few attempt was made to assess the role of phenylalanine and trans- seconds and, after 2 minutes, 2 ml of a 0.3% solution of 2-thiocinnamic acid in this system. During this investigation two barbituric acid were added. After being heated in a boiling esters, in addition to chlorogenic acid, were found to accumulate water bath for 10 minutes, the tube was cooled in a water bath in the tissue. These esters have been identified as cinnamylheld at 40”. The colored product was transferred to an organic quinic and p-coumaryl-quinic acids, and evidence is presented phase for measurement according to Waravdekar and Saslaw that they are direct precursors of chlorogenic acid. Although a (13), by addition of 3 ml of a 1 to 1 mixture of isoamyl alcohol number of quinic acid esters, including p-coumaryl-quinic acid, and 12% HCl. The tube was shaken vigorously for 20 seconds have been previously encountered in plants (7-lo), this is the and centrifuged. The absorbancy of the organic layer was measured at 549 rnp and was found to be a linear function of concenfirst time, insofar as we are aware, of the occurrence of a cinnamyl-quinic ester in nature. tration in the range from 0.01 to 0.08 pmole of quinic acid (Fig. 1). Shikimic acid gives no color at this range of concentration EXPERIMENTAL PROCEDURE (13). A second less specific assay of quinic acid was used in some Methods-Chlorogenic acid synthesis was studied by floating experiments. Periodate oxidation was accomplished as described disks cut from plugs of Kennebec potato tubers in shallow soluabove, and the aldehyde formed was measured by the method of tions of various organic acids. The cultural conditions and the Friedemann and Haugen (14). After removal of the excess perchlorogenic acid assay previously described (6) were used. iodate, 1 ml of a 0.1 y. solution of 2,4-dinitrophenylhydrazine Assay of Aromatic Acids-To a tube containing 0.5 ml of the in 2 N HCl was added. The stoppered tube was shaken and solution of the ester, 0.5 ml of 10 N KOH was added. The mixafter which ture was shaken and the hydrolysis was allowed to proceed for allowed to stand 10 minutes at room temperature, 3 ml of 2.5 N KOH were added and the solution was shaken for 15 minutes at room temperature. The hydrolysate was then the absorbancy was measacidified with 1 ml of 6 N HCl, and the aromatic moiety was ex- a few seconds. After centrifugation tracted by shaking the solution five times with equal volumes of ured at 470 rnp. Quinic acid can be determined by this method in the range from 0.1 to 0.5 pmole. ether saturated with 6 N HCl. The combined ether extracts Preparation of Cell-free Extract-Potato tubers were peeled were evaporated to dryness and the residue was dissolved in 10 ml of 0.1 M phosphate buffer at pH 3.5. The optical density of and 37 g of inner tissue were ground in 100 ml of chilled 0.1 M phosphate buffer (pH 7.3) in a blender. The suspension was the solution was measured at the appropriate absorption maxifiltered through three layers of cheesecloth and centrifuged for mum. 15 minutes at 20,000 x g at 0”. To each 100 ml of the clear Quinic Acid Assay-The compound 1,3,4,5,6-pentahydroxypink supernatant solution, 50 g of ammonium sulfate were added. cyclohexanecarboxylic acid, upon oxidation with periodic acid After 10 minutes in an ice bath, the mixture was centrifuged, and subsequent treatment with thiobarbituric acid, has been shown by Srinivasan and Sprinson (12) to give a pink color re- and the protein precipitate was dissolved in 8 ml of 0.1 M phosphate buffer at pH 7.3. * This study was supported in part by funds from the National Preparation of Pectinase-To each 500 mg of Pectinase (NuScience Founadtion. tritional Biochemical Corporation), 5 ml of 0.01 M phosphate t Present address, Department of Pharmacology, Tufts Unibuffer (pH 7.5) were added and the mixture was stirred for 2 versity School of Medicine, Boston, Massachusetts.

August

1960

C. C. Levy and M. Zucker

hours in an ice bath. It was then centrifuged at 0” and the brown supernatant solution was dialyzed for 16 hours against 3 changes (2 liters each) of 0.001 M phosphate buffer (pH 7.5). Fractionation of Phenolic Esters by Lead Acetate PrecipitationTo 5 ml of a concentrated alcoholic extract, representing about 20 g of cultured potato tissue, were added 2 ml of a saturated After centrifugation, the supernasolution of lead subacetate. tant solution was freed from lead with an excess of sulfuric acid. The basic lead acetate precipitate was decomposed by suspending it in 2 ml of ethanol containing 0.1 ml of 10 N H&O+ RESULTS

Neish et al. (2, 15) proposed a pathway of synthesis for the caffeyl moiety of rutin and lignin precursors involving L-phenylalanine, phenylpyruvate, phenyllactate, and trans-einnamic acid. To determine whether these organic acids can stimulate the net synthesis of chlorogenic acid by serving as precursors to its caffeyl moiety, potato tuber disks were floated for 48 hours in shallow solutions of each of these compounds. Table I shows that only L-phenylalanine and trans-cinnamic acid gave a pronounced stimulation of net synthesis. Previous work (6) had Thus, to obindicated that quinic acid was itself stimulatory. tain maximal stimulation the organic acids were used in conjunction with the sodium salt of quinic acid. Here again, only the culture solutions containing phenylalanine or cinnamic acid in Neither addition to quinic acid gave rise to marked stimulation. phenylpyruvate nor phenyllactate alone or with quinic acid gave any stimulation. Although chromatographic evidence indicated that p-hydroxy cinnamic acid (p-coumaric acid), a logical intermediate in chlorogenic acid synthesis, entered the tissue, no stimulation was seen in these cultures at concentrations as high as 0.05 M, nor could any evidence be found for the hydroxylation of p-coumaric acid to caffeic acid. Under the present conditions of culture, caffeic acid was found to be toxic to the tissue (6). At the concentrations employed, culture solutions of phenylalanine stimulated the synthesis of chlorogenic acid to a greater extent than those of trans-cinnamic acid. However, when so-

dium quinate was added to the culture solution, cinnamic acid was as effective as phenylalanine even though it was present at a much smaller concentration. This suggests that the synthesis of some intermediate between cinnamic and chlorogenic acids is Since p-coumaric acid did not serve as a precursor rate-limiting. to chlorogenic acid, experiments were designed to find intermediates between cinnamic acid and the caffeyl moiety of chlorogenie acid which might accumulate in the tuber tissue. The alcoholic extracts of both 0.05 M L-phenylalanine and 0.01 M trans-cinnamate cultures were chromatographed in the butanolacetic acid-water solvent. Examination of the dried chromatograms under ultraviolet light revealed the presence of a number One of these bands had of bands not found in control cultures. When exposed to aman RF of 0.67 and was quite prominent. monia vapors, it changed from a dark violet quenching band to a deep blue fluorescent one. This type of response is characteristic of p-coumaric acid and coumaryl-containing compounds. The band, too slow moving to be the free acid, was subsequently shown to contain a coumaryl ester. The accumulation of a coumaryl ester, in addition to the caffeyl ester (chlorogenic acid) normally present, suggested that an ester of cinnamic acid might be accumulating as well. Accordingly, the section of the paper where such an ester might be found was examined carefully. At an RF of 0.75, a faint quenching band could be seen in light predominantly of 2537 A, but which disappeared in light of 3660 A. No change occurred when the paper was exposed to ammonia vapors. These properties are associated with cinnamic acid and cinnamyl esters. Both esters were chromatographically purified as described below. Preparation of Esters-The cinnamyl and coumaryl esters were prepared by culturing large thin potato sections, cut from peeled tubers, in a shallow solution of 0.005 M trans-cinnamic acid and 0.05 M sodium quinate (pH 6.0), or 0.01 M sodium cinnamate and 0.1 M glucose. After 48 hours, the sections were ground in a The alcoblender with 95% ethanol as the grinding medium. holic extract, separated by filtration, was concentrated to dryness by vacuum distillation. After washing the residue several times I of chlorogenic

TABLE

Stimulation

of net synthesis

acid

in disks

of potato

tuber Chlorogenic acid synthesized* Culture solution

Water control Sodium phenylpyruvate, 0.05 dl-Phenyllactic acid, 0.05 M nL-phenylalanine, 0.05 M Sodium cinnamate, 0.005 M Sodium p-coumarate, 0.01 M

I

M

No additions

+0.05 II Na quinate, pH 6

0.15 0.10 0.05 0.77 0.43 0.14

0.68 0.48 0.51 1.27 1.26 0.58

* Disks of tuber tissue, representing about 1.5 g fresh weight were cultured 48 hours at room temperature in the dark in 5 ml of the above solutions. Figures represent the average of duplicate determinations in several replicated experiments. Each sample of disks initially contained 0.15 rmole of chlorogenic acid. In the phenylalanine and cinnamic acid cultures containing quinic acid, stimulation of chlorogenic acid synthesis could be detected within 6 hours. Exposure of the water cultures to light also produced a large stimulation of synthesis.

Downloaded from http://www.jbc.org/ by guest on August 13, 2017

I I I I I I J .04 .06 .02 p MOLES OF QUINIC ACID FIG. 1. Standard curve for determination of quinic acid by the periodate and thiobarbituric acid reagents.

2419

Intermediates

2420

in Chlorogenic

TABLE II RF values of esters and their component moieties

Type of component

-

Compound

1

Solvent* 2

3

0.60 0.67 0.75

unstable 0.49 0.67

0.61 0.74 0.90

Caffeic acid or moiEther-soluble aromatic comety from chlorogenie acid ponent of hydrolyzed ester-E p-Coumaric acid or moiety from coumaryl ester trans-Cinnamic acid or moiety from cinnamyl ester

0.84

unstable

0.39

0.89

0.27

0.47

0.90

0.44

0.60

Nonaromatic moiety

From chlorogenic acid From coumaryl ester From cinnamyl . ester Quinic acid

0.40 :O.28 0.40 :o. 28 0.40 :o. 28 0.40

Alkali-treated quinic acid

:0.28

Unhydrolyzed esters

Chlorogenic acid Coumaryl ester Cinnamyl ester

no migration

0.56 0.39) 0.56 ‘0.39) 0.56 :O.39) 0.56 :O.39)

* Solvent 1, n-butanol-acetic acid-water, 4: 1:5; Solvent 2, n-butanol-1.5 N NH,OH-0.8 M (NHSZCO~, 2: 1: 1; Solvent 3, 5oj, acetic acid. Chlorogenic acid was also chromatographed in redistilled methyl isobutyl ketone-formic acid-water (3: 1: 2), RF 0.42 (24). The Rp values in parentheses were those of compounds found only under conditions specified in the text.

Synthesis

Vol. 235, No. 8

of the cinnamyl ester. Because of the small amount of material available and of the presence of contaminants extracted by ethanol from the paper, attempts to crystallize these esters failed. The contaminants gave none of the reactions ascribed to the esters. A small absorbance in the ultraviolet was applied as a correction on the absorbancy of the esters. The preparation of a sample of the coumaryl ester with a constant specific activity, in experiments with isotopes described below, indicated that, except for this contaminating material, the coumaryl ester was 85 to 90% pure after chromatography in 2 solvents. Table II gives the RF values in different solvents of the compounds described. Identijication of Esters-The coumaryl and cinnamyl compounds both gave positive ferric hydroxamate tests before hydrolysis (16). Alkaline hydrolysis of both compounds, under conditions which hydrolyzed chlorogenic acid, gave rise to two components, bothacidic in nature. As with chlorogenic acid, one of these components was readily soluble in ether. The nonaromatic nature of the other acid component was indicated by the fact that it could be transferred to ether only after prolonged liquid-liquid extraction. These facts suggest the presence of an ester linkage similar to that found in chlorogenic acid. The ethersoluble moieties were identified as follows: After alkaline hydrolysis of the supposed cinnamyl and coumaryl esters, the hydrolysate was acidified and extracted with ether three times. The ether extracts were combined, evaporated to a small volume, and aliquots were chromatographed in 3 different solvents with transcinnamic acid, p-coumaric acid, and the unhydrolyzed esters as markers. In all cases, the ether-soluble component of the ester believed to contain cinnamic acid had the same RF as a known sample of trans-cinnamic acid, and differed considerably in movement from the unhydrolyzed material. Cochromatography with trans-cinnamic acid in these solvents further substantiated the identity of the ether-soluble component of this ester as transcinnamic acid (see Table II for RF values). Similarly, the ether-soluble component from the ester believed to contain p-coumaric acid had the same Rp as an authentic sample of this compound and gave a single spot on cochromatography. Here, too, the unhydrolyzed material differed considerably in mobility from its hydrolysis product. The aromatic moiety of this ester was thus provisionally identified as p-coumaric acid (see Table II). In the cinnamyl ester, it is apparent that the ester linkage can occur only through the carboxyl group of cinnamic acid to some functional group on the nonaromatic moiety. In the coumaryl ester, on the other hand, the ester linkage may be either through the para hydroxyl or the carboxyl group of the p-coumaric acid. Spectral data indicate that the linkage is through the latter group. When measured at an acid pH (3.5) in phosphate buffer, the spectrum of p-coumaric acid is identical with that of the ethersoluble component of the coumaryl ester. Both have peaks at 312 rnp and troughs at 250 rnp. Jurd (11) has shown that with free p-coumaric acid, spectral shifts to shorter wave lengths occur on neutralizing the acid solution. As can be seen in Fig. 2, this shift occurs with the ether-soluble component of the coumaryl ester. Jurd has further shown that when the carboxyl group is esterified, there is no difference in the spectra at acid or neutral pH. As shown in Fig. 3, the absorption spectrum of the unhydrolyzed coumaryl ester does not shift in this pH range, indicating that the carboxyl group of the coumaric acid moiety is esterified. This is entirely analogous to chlorogenic acid which

Downloaded from http://www.jbc.org/ by guest on August 13, 2017

with small amounts of ether, it was dissolved in a small volume of 95% ethanol. This solution was streaked on water-washed chromatographic papers (Whatman 3 MM) and chromatographed overnight with n-butanol-acetic acid-water (4: 1: 5) as irrigating solvent. The presence of a number of bands with either absorbing or fluorescent properties was revealed by examining the dried chromatograms under ultraviolet light. The bands with Rp values of 0.67 and 0.75 were cut out and the esters were The eluates were concentrated eluted in boiling 60% ethanol. to a small volume and rechromatographed overnight in an nbutanol-1.5 N NH40H-0.8 M (NH&CO3 (2: 1: 1) solvent system. The procedures for locating and eluting the esters were repeated and the separate eluates were chromatographed for 5 hours in 5% acetic acid. The esters, free from any apparent impurity, were eluted from this final chromatogram, the solutions were concentrated to dryness under reduced pressure, and the residues Both esters were aswere dissolved in water or 95% ethanol. sayed spectrophotometrically in phosphate buffer (pH 3.5) at an absorption maximum of 280 rnp for the cinnamyl ester and 312 rnp for the coumaryl compound. The molar extinction coefficient for cinnamic acid (19,200) and that for p-coumaric acid (23,000), (ll), were used as approximations of the coefficients of estimates are probthe esters. As a result, spectrophotometric ably somewhat high. A kilogram of cultured potato tissue yielded 30 to 60 pmoles of coumaryl ester and only trace amounts

Acid

August

1960

2421

C. C. Levy and M. Zucker

I

I

I

I

I

I 380

240

280 320 360 WAVE LENGTH (MJ ) FIQ. 3. Absorption spectrum of the coumaryl ester: O-0, measured at pH 3.5 and 7.0; and l - - -0, measured at pH > 10. behaves similarly under the same conditions. Additional evidence that it is not the hydroxyl group of the coumaric acid which is esterified comes from the fact that, under alkaline conditions, there is a pronounced bathochromic shift in the absorption maximum of the ester to a longer wave length. This could not occur if the hydroxyl group were esterified and therefore unable to ionize. A free phenolic hydroxyl group in the ester is also indicated by a positive Folin phenol test (17). The absorption spectrum of the ether-soluble component of the cinnamyl ester was the same as that of a known sample of trans-cinnamic acid. The cinnamyl ester had a similar absorption spectrum (see Fig. 4). Identification of Nonaromatic Moiety-Both esters were hy-

I 240 FIG.

and 7.0.

4.

,Absorption

I

I I I 280 320 WAVE LENGTH (up)

spectrum

of the cinnamyl

I

ester at pH 3.5

Downloaded from http://www.jbc.org/ by guest on August 13, 2017

320 380 280 WAVE LENGTH (MI) ) FIG. 2. Absorption spectrum of the ether-soluble component isolated from the hydrolysate of the coumaryl ester: O--O, measured at pH 3.5; and 0-0, measured at pH 7.0. 240

drolyzed with alkali and the aromatic moieties were removed by brief extraction with ether after acidification. To separate the nonaromatic moieties from the salts in the hydrolysate, the solution was subjected to liquid-liquid extraction with ether for a 24-hour period. The ether was evaporated and the samples were chromatographed in 3 solvents. Authentic quinic acid, quinic acid subjected to the treatment described above, and quinic acid derived from hydrolyzed chlorogenic acid were used as markers. Quinic acid was chosen to determine whether these esters are related to chlorogenic acid. The samples from both esters showed acidic spots which had the same RJ= as authentic quinic acid (Table II). The substance in these spots when eluted gave a positive thiobarbituric acid test. In addition, a second acid spot with a slower RF than that of quinic acid appeared in the samples from both esters, and from chlorogenic acid, as well as from alkali-treated quinic acid. Eluates of the material in these slower moving spots (including that from alkali-treated quinic acid) failed to give a positive thiobarbituric acid test. To avoid this complication, another means of hydrolyzing the esters was sought. Reid (4) had used a commercial esterase preparation to hydrolyze chlorogenic acid. An extract from a similar material (Pectinase) was found to hydrolyze both esters as well as chlorogenie acid at a neutral pH. The reaction mixture consisted of 2.5 pmoles of ester in 0.5 ml of solution, 0.1 ml of enzyme, and 0.4 ml of 0.1 M phosphate buffer at pH 7.0. After 2 hours at room temperature, the reaction mixture was acidified and extracted several times with ether to remove the aromatic moiety. The remaining solution was concentrated and chromatographed in 2 different solvents with a sample of quinic acid and enzymatically hydrolyzed chlorogenic acid as markers. A single acid spot appeared in samples from the cinnamyl and coumaryl esters as well as in that from the hydrolyzed chlorogenic acid, and these corresponded in RF to quinic acid. After oxidation with periodate, the eluates gave a positive test with thiobarbituric acid. These findings indicate that the nonaromatic moiety of both esters is quinic acid. In view of this, the compounds are considered to be cinnamyl-quinic acid and p-coumaryl-quinic acid.

2422

Intermediates TABLE

Stoichiometry

Experiment

in Chlorogenic

III

of components

of isolated

esters

Ester

pmoles

/.l?deS

p?noles

1*

Coumaryl Cinnamyl

2.1 1.0

1.6 0.82

1.8 0.86

0.89 0.95

11t

Coumaryl Cinnamyl

1.0 1.0

0.89 0.79

0.90 0.90

0.99 0.88

* In Experiment I, the esters were hydrolyzed enzymatically and quinic acid was determined with the thiobarbituric acid reagent. t In Experiment II, the esters were isolated independently of those in Experiment I and were hydrolyzed with alkali. Quinic acid was assayed after oxidation by the method of Friedemann and Haugen (14). All values shown are the average of 2 determinations.

Synthesis

Vol. 235, No. 8

over, the cyclic compound differed from quinic acid in that it gave no test for an (y. hydroxy acid with ferric chloride (20). Indeed, the lack of an a! hydroxy group, as in shikimic acid, could explain the failure of this substance to give a pink color with thiobarbituric acid (12). Thus, the one significant difference between this compound and quinic acid appears to be the absence of an cx hydroxy group. A compound with a similar RF has been obtained by other workers during the course of the isolation of quinic acid from chlorogenic acid (4) and from a coumaryl ester found in pineapples (21). Evidence that the quinic acid-like substance which lacks an a! hydroxy group is an artifact of the chromatographic isolation of the cinnamyl and coumaryl esters was obtained as follows. If these esters originally contained quinic acid, there should be more esterified quinic acid in the original alcoholic extracts of the tissue before chromatography than can be accounted for by chlorogenic acid alone. With this in mind, the esters in the original alcoholic extract were transferred to ethyl acetate and the amount of esterified quinic acid was assayed by the thiobarbituric acid procedure. Experiment I of Table IV shows that chlorogenic acid can account for only one-half of the esterified quinic acid present before chromatography. In a second experiment, chlorogenic acid was precipitated quantitatively from the extract by lead acetate. This procedure does not remove all the coumaryl or cinnamyl esters from solution. Consequently 2 fractions were obtained, one rich in chlorogenic acid and the other devoid of it, but rich in the coumaryl and cinnamyl esters. As Table IV shows, there is a striking difference in the amount of esterified quinic acid present and the amount accounted for by the chlorogenic acid in the supernatant solution. Even in the lead acetate precipitate, the amount of esterified quinic acid is approximately twice that of the chlorogenic acid found. The fact that, upon chromatography of these fractions rich in esterified quinic acid, an altered form of the acid has appeared at times suggests that this alteration is an artifact of the chromatographic procedure rather than that two types of each ester exist as such in the tissue. Conversion of Coumaryl Ester into Chlorogenic Acid-To demTABLE

Concentration

Experi malt

of

esterified

IV

quinic acid ltato tissue*

Preparation

in

alcoholic -

extracts

of

C oncentra - ceoncentration of chloro-

Type of ester present

tion of :sterified acid

g enic acid I q: linic 6tmolelnl

P

moles jml

I

Ethyl acetate extract

Chlorogenic acid and coumaryl ester

0.45

0.80

II

Lead acetate pre-

Chlorogenic acid and coumaryl ester Coumaryl and cinnamyl esters

0.41

1.0

0.06

1.2

cipitate

Supernatant

solu-

tion

* Potato sodium

tissue was cultured

cinnamate

and 0.1

M

48 hours in a solution

glucose

before

acid

present

in each

sample

of 0.01 1~

extraction.

quinic acid was found in any of the preparations. than chlorogenic matographically.

Downloaded from http://www.jbc.org/ by guest on August 13, 2017

Stoichiometry-To establish the ratio of quinic acid to the aromatic constituents of both esters, these compounds were hydrolyzed both enzymatically and with alkali. In the former case, quinic acid was assayed by the periodate-thiobarbituric acid reaction and in the latter by an adaptation of the Friedemann-Haugen procedure. Aromatic moieties were assayed spectrophotometrically. The results in Table III clearly show that in both the cinnamyl and coumaryl esters, quinic acid exists in a 1 to 1 ratio with its respective aromatic partner. In quinic acid there are 4 hydroxyl groups available for esterification to the aromatic moiety. The 2 cis hydroxyls at positions 4 and 5 can be excluded from consideration for the following reasons. Periodate oxidation of either unhydrolyzed ester under acid conditions for 20 minutes at room temperature yielded a product which formed an intensely colored 2,4-dinitrophenylhydrazone. An equimolar quantity of the aromatic moiety of the ester produced only a weakly colored product under the same conditions. Consequently, the formation of a 2,4-dinitrophenylhydrazone indicates unesterified adjacent hydroxyl groups in the quinic acid moiety. The rapidity of the periodate oxidation further suggests that the hydroxyls are cis to each other. That this is the case is substantiated by a positive cupric acetate test for cis hydroxyls (18) in the coumaryl ester. The only remaining hydroxyls available for esterification are those at positions 1 and 3. Evidence is presented below that the coumaryl ester can be converted to chlorogenic acid in vitro and in vivo. Therefore, by analogy, it is probable that the ester linkage occurs at the hydroxyl in position 3 of quinic acid. On a number of occasions, the nonaromatic moiety of each ester failed to give a positive test with the thiobarbituric acid reagent even when the ester was hydrolyzed enzymatically. Chromatography showed that this nonreacting moiety had the same RF as the slower moving component of alkali-treated quinic acid, although no alkali had been used. This moiety gave positive tests described above for adjacent cis hydroxyl groups. That it is cyclic in nature is deduced from the fact that it was converted to protocatechuic acid by fusion with KOH, a procedure which converts quinic acid to this product as well (19). Protocatechuic acid was identified chromatographically with use of nitrous acid or of ferric chloride as spray reagents. More-

Acid

were

No

free

Esters other detected

chro-

August

C. C. Levy and M. Zucker

1960 TABLE

Conversion Experiment

of p-coumaryl-quinic

V acid

to chlorogenic

acid

Culture medium

in

vivo

Chlorogenie acid formed imole

I*

IIi

Water Coumaryl-quinic acid, 1.7 X 1e3 M Sodium cinnamate and sodium quinate, both 1.7 x lo-3 M

0.38 0.70 0.44

Water Coumaryl, quinic acid, 3.4 X 1c3 M + potassium ascorbate, 0.025 M Potassium ascorbate, 0.025 M

0.16 0.27 0.10

onstrate that the coumaryl ester is, in fact, a direct precursor of chlorogenic acid, potato tissue was incubated with a solution of the ester for 48 hours in the dark. As shown in Table V, the coumaryl ester at 0.0017 M produced about twice as much chlorogenie acid as the water controls. At this same concentration, the combination of trans-cinnamic and quinic acids was virtually ineffective in raising the chlorogenic acid levels above that of the controls. At somewhat higher concentrations (0.0034 M), the stimulating properties of the coumaryl ester is again evident (Experiment 2). At these concentrations, however, ascorbic acid itself played no apparent role in stimulating the synthesis of chlorogenic acid. Equally effective in demonstrating the conversion of the coumaryl ester to chlorogenic acid was the use of cell-free extracts. In these experiments, the reaction mixture consisted of coumaryl ester, ascorbic acid, cell-free extract, and Tris buffer (pH 8.3). Chlorogenic acid was identified chromatographically as a product of the reaction. Quantitative data given in Table VI show that in the complete system there was a 16% conversion of the coumaryl ester to chlorogenic acid. If ascorbic acid were omitted, the chlorogenic acid formed was rapidly oxidized and polymerized to brown pigments. There is a distinct possibility that the enzyme involved in the With hydroxylation of the coumaryl ester is polyphenoloxidase. use of the conversion of p-coumaric acid to caffeic acid as a model system, it was found that the activity of the cell-free extract was completely inhibited in the presence of either 1O-4 M phenylthiourea or 4-chlororesorcinol, both potent inhibitors of polyphenoloxidase (22, 23). Attempts to stimulate chlorogenic acid synthesis with the cinnamyl ester in z&o have been unsuccessful because of the toxicity of the preparations available. Raclioaetive Substrates-The use of randomly labeled L-phenylalanineprovides additional evidence that the coumaryl ester is a direct precursor of chlorogenic acid. In one set of experiments, 10 potato disks were placed in a Petri dish containing 1 ml of 0.05 M L-phenylalanine-Cl4 which had a specific activity of

12,600 c.p.m. per pmole. After 48 hours in the dark, an alcoholic extract of the tissue containing both the coumaryl and chlorogenie esters was chromatographed in n-butanol-acetic acid-water (4: 1: 5), the esters were eluted and rechromatographed in 5% acetic acid. After elution from this second chromatogram, the total activities and concentration of both esters were determined, and the chlorogenic acid was chromatographed to constant specific activity in the methyl isobutyl ketone solvent (24). In a similar manner, the coumaryl ester was chromatographed to constant specific activity in a solvent system consisting of nbutanol-1.5 N NHaOH-0.8 M (NH&COI (2:l:l). The specific activities of both esters were essentially the same, and were equal to the specific activity of L-phenylalanine-Cl4 (Table VII). When the coumaryl ester was hydrolyzed, the activity was found entirely in the coumaric acid moiety; conse-

Enzymatic

conversion

TABLE VI of p-coumaryl-quinic

to chlorogenic

acid

Reaction mixture*

Complete system Complete system (boiled enzyme) No coumaryl-quinic acid

p~OlC

%

0.23 0 0

16

* Complete reaction mixture contained 20 rmoles of phosphate buffer at pH 8.3,20 pmoles of potassium ascorbate, 1.33 pmoles of the coumaryl ester, and 0.25 ml of enzyme in a total volume of 1.0 ml. The mixture was incubated 3 hours at room temperature and the reaction was stopped by addition of alcohol. TABLE

Incorporation

of randomly

VII

labeled z-phenylalanine into phenolic esters

and

glucose-Cl4

Specific activity Experiment

Culture medium’ Chlorogenic acid c.).m./p?rtole

I II

III

0.05 M L-Phenylalanine-C” 0.05 M L-Phenylalanine-Cl4 0.05 M L-Phenylalanine-C” + 0.005 M sodium cinnamate 0.05 M L-Phenylalanine-Cl4 + 0.05 M sodium quinate 0.05 M L-Phenylalanine-Cl4 + 0.05 M sodium quinate + 0.005 M sodium cinnamate 0.1 M Glucose-Cl4

12,600

#-Coumar;l-quinic c.p.m./~mole

13,800(14,2OO)t 12,500

9,700

8,2c@

12,800

13,200

2,800

3,800

39,700(11,900)’

* Specific activity of L-phanylalanine-Cl4 administered 12,600 c.p.m. per pmole. Specific activity of glucose-CL4 administered 28,000 c.p.m. per rmole. t Figures in parentheses give the specific activity of the aromatic moiety isolated after hydrolysis of the ester.

Downloaded from http://www.jbc.org/ by guest on August 13, 2017

* In Experiment I, 20 disks, 12 mm in diameter and 1 mm thick were cultured in 2 ml of solution for 48 hours. Values listed are averages of duplicate 10 disk samples used for each assay. The initial concentration of chlorogenic acid per sample was 0.19 pmole. t Similar conditions were used in Experiment II except that culture was for 24 hours only. The initial concentration of chlorogenic acid was 0.22 rmole per sample (about 1.5 g fresh weight).

2423

2424

Intermediates

in Chlorogenic Acid Synthesis

of both the cinnamyl and the coumaryl esters possess all the properties that could be expected of intermediates one step removed from each other and from the final product. In addition, coumaryl-quinic acid at an extremely low concentration (1.7 x 10M3 M) stimulated chlorogenic acid synthesis above that of contral cultures. At this low concentration, no other compound or combination of compounds exerted such a stimulation. The isotope data also substantiate the role of these esters as intermediates in the biosynthesis of chlorogenic acid. Despite large differences in specific activities produced by variations in the composition of the culture solutions, the activities of both the coumaryl and chlorogenic acid esters were always equal to each other. The ability of cinnamic acid to dilute the activity of both esters formed in the presence of labeled phenylalanine strongly suggests that cinnamic acid is a precursor used for their On the other hand, the fact that quinic acid did not synthesis. cause any dilution in specific activity indicates that it is not a precursor to the aromatic moieties. The substantial dilution in activity of both esters, upon addition of cinnamic acid to culture solutions containing labeled phenylalanine and quinic acid, is indeed what might be expected if a cinnamyl-quinic acid ester were involved in chlorogenic acid synthesis. That is, the presence of preformed quinic acid should increase the rate of esterification of cinnamic acid, and consequently increase its utilization at the expense of n-phenylalanine-C14. An increased rate of formation of a cinnamyl-quinic acid ester could also explain the ability of quinic acid to stimulate chlorogenic acid formation even though it is not a precursor to the aromatic moiety. Thus, the dilution of activity in both esters, under the conditions described, is not only evidence that coumaryl-quinic acid is involved as an intermediate in chlorogenic acid synthesis, but is also good presumptive evidence that cinnamyl-quinic acid plays a comparable role. GLUCOSE

\

PHENYLALANINE DISCUSSION

Based on isotopic studies of the formation of phenolic acids in salvia cuttings, McCalla and Neish (15) have proposed that by a series of hydroxglations, cinnamic acid (derived from phenylalanine) is converted through p-coumaric acid to caffeic acid. These authors suggested that, although p-coumaric and caffeic acids could be detected only as esters in the tissue, nevertheless the series of interconversions occur at the free acid level before esterification. The present data substantiate their findings that n-phenylalanine and trans-cinnamic acid play a key role in the biosynthesis of phenolic acids. In the potato tissue, as in salvia, neither free p-coumaric nor caffeic acids could be detected chromatographically, but were found only as esters. In this connection, the occurrence of a cinnamyl-quinic-acid ester, in addition to the p-coumaryl and caffeyl-quinic esters, would suggest that the hydroxylations of these various acids in potato tissue occur not at the free acid level, but rather after the parent compound, In fact, repeated experiments with cinnamic acid, is esterified. free p-coumaric acid in concentrations as high as 0.05 M showed that this substance failed to stimulate the biosynthesis of chlorogenie acid. From a sequential point of view, the structural characteristics

OUINIC

ACID (0)

CINNAMYL-OUINIC

ACID

I

CH=CHC

-0-Q

f+Tla p-COUNARYL-OUINIC

ACID

OH 1 GH.OH+--O 0 0

OH OH

CHLOROGENIC FIG.

ACID 5

Downloaded from http://www.jbc.org/ by guest on August 13, 2017

quently the phenylalanine was incorporated into this moiety Although attempts were made to without dilution of activity. hydrolyze the labeled chlorogenic acid, the amounts of caffeic acid recovered were insufficient to allow accurate determination of its specific activity. In any case, Reid (4), working with tobacco leaves, has shown that in chlorogenic acid derived from labeled phenylalanine, all of the label is in the caffeic acid moiety. In another series of experiments, designed to test the precursor properties of cinnamic acid, potato tissue was incubated for 48 hours in the dark with combinations of phenylalanine-04 and the other substrates shown in Table VII. Despite the large differences in the specific activities resulting from differences between culture conditions, the specific activity of the coumaryl ester was equal to that of chlorogenic acid. In the cultures with phenylalanine alone and of phenylalanine with quinic acid, the specific activities of both esters remained unchanged from that of the original n-phenylalanine-CY. When trans-cinnamic acid was added to the phenylalanine culture solution, there was a noticeable drop in the specific activities of both esters. Most striking, however, is the dilution in specific activity caused by the addition of truns-cinnamic acid to the phenylalanine-quinic acid culture solution. In this system, quinic acid is known to stimulate chlorogenic acid formation markedly above that of the control cultures (6). When, however, quinic acid is added to randomly labeled Lphenylalanine-Cl4 solutions, no dilution in specific activities from phenylalanine cultures can be seen. Apparently, then, quinic acid is not on the pathway to caffeic acid. That quinic acid is actively synthesized by the tissue is shown from the results of experiments in which randomly labeled 0.1 M glucose-Cl4 is incubated with potato tissue. Here the specific activity of the labeled chlorogenic acid on a molar basis is about one and onehalf times that of the original glucose solution, but on a carbon basis, there is a 2-fold dilution in activity. Since caffeic acid has three carbons more than glucose, this moiety could account for Upon hydrolysis of the ester, all of the activity in the molecule. however, some 6570 of the activity was found in the quinic acid moiety.

Vol. 235, No. 8

August

C. C. Levy and M. Zucker

1960

SUMMARY

1. L-Phenylalanine and trans-cinnamic acid stimulate the net synthesis of chlorogenic acid in disks of potato tuber. 2. In response to treatment with n-phenylalanine or transcinnamic acid, two esters of quinic acid in addition to chlorogenic acid arise in potato tissue. These have been identified as cinnamyl-quinic and p-coumaryl-quinic acids. 3. The conversion of the coumaryl ester into chlorogenic acid

was demonstrated directly, both in vivo and in vitro. Data obtained with isotopes strongly support the role of the ester as well as cinnamyl-quinic acid in chlorogenic acid synthesis. 4. A suggested pathway of chlorogenic acid formation from phenylalanine has been outlined and discussed. Acknowledgments-Appreciation Vickery for his aid in preparation

is expressed to Dr. of this manuscript.

H. B.

REFERENCES 1. FISCHER, H. 0. L., AND DANGSCHAT, G., Ber. Chem. Ges., 65, 1037 (1932). 2. UNDERHILL, E. W., WATKINS, J. E., AND NEISH, A. C., Can. J. Biochem. Physiol., 35, 219 (1957). 3. GEISSMAN, T. A., AND SWAIN, T., Chem. & Ind., 984 (1957) 4. REID, W. W., Chem. & Znd.. 1439 (1958). 5. DAVIS, B. D:, in W. D. MCELROY AND B. GLASS (Editors), Amino acid metabolism, The Johns Hopkins Press, Baltimore, 1955, p. 799. 6. ZUCKER, M., AND LEVY, C. C., Plant Phusiol.. 34, 108 (1959). 7. CARTWRIGHT, R. A., R&BERT& E. A. B., F~bod, A. I?., AND WILLIAMS, A. H.. Chem. & Ind.. 1062 11955‘1. 8. SCH~~TTE, H’. R., ~ANGENBECK, g,. AND‘ ~-&ME, H., Naturwissenschaften, 44, 63 (1957). 9. HERZMANN, H., Phytopath. Zeit., 34, 109 (1958). 10. WILLIAMS, A. H., Chem. and Ind., 1200 (1958). 11. JURD, L., Arch. Biochem. Biophys., 66, 284 (1957). 12. SRINIVASAN, P. R., AND SPRINSON, D. B., J. Biol. Chem., 234, 716 (1959). 13. WARAVDEKAR, V. S., AND SASLAW, L. D., J. Biol. Chem., 234, 1945 (1959). 14. FRIEDEMANN, T. E., AND HAUGEN, G. E., J. Biol. Chem., 147, 415 (1943). 15. MCCALLA, D. R., AND NEISH, A. C., Can. J. Biochem. Physiol., 37. 537 (1959). 16. CH&ONIS; N. b., AND ENTRIKIN, J. B., Semimicro qualitative organic analysis, 2nd edition, Interscience Publishers, Inc., New York, 1957, p. 229. 17. FOLIN, O., AND CIOCALTEU, V., J. Biol. Chem., 73, 627 (1929). 18. WEISS, U., AND MINGIOLI, E. S., J. Am. Chem. Sot., 78, 2894 (1956). 19. HESSE, O., Ann. Chem. Liebigs, 200, 232 (1880). 20. WEISS, U., DAVIS, B. D., AND MINGIOLI, E. S., J. Am. Chem. Sot., 76, 5572 (1953). 21. SUTHERLAND, G. K., AND GORTNER, W. A., Australian J. Chem., 12, 240 (1959). 22. BERNHEIM, F., AND BERNHEIM, M. L. C., J. Biol. Chem., 145, 213 (1942). 23. KULL, F. C., GRIMM, M. R., AND MAYER, R. L., Proc. SOC. Exvtl.

24. HuL&,

Biol.

Med.

A. C., Chem.

Sci.. 86, 330 (19541. &‘I&., 217 (1958).

Downloaded from http://www.jbc.org/ by guest on August 13, 2017

These findings suggest a scheme (Fig. 5) outlining the probable sequence of events leading to chlorogenic acid. In the present study, the administration of randomly labeled glucose-Cl4 resulted in the formation of a chlorogenic acid molecule with activity in both caffeic and quinic acids. Phenylalanine and quinic acid are placed on divergent pathways from glucose because of the negative results obtained in the isotope dilution experiments, and because the administration of shikimic acid, a key intermediate in the conversion of quinic acid to phenylalanine in other systems (5), had no apparent effect on chlorogenic acid formation (6). The role of phenylpyruvate or phenyllactate in the conversions outlined in Fig. 5 cannot be defined as yet. The data do not offer any positive justification for placing them in the scheme. On the other hand, the negative results obtained in the relatively long term feeding experiments described above do not eliminate them conclusively as intermediates, for explanations such as the impermeability of the tissue, or instability in the culture solution can be offered for their failure to stimulate the synthesis of chlorogenic acid. With respect to the mechanism of hydroxylation of the esters, some evidence was presented that the coumaryl ester could be enzymatically converted to chlorogenic acid. Studies on the inhibition of the hydroxylation of p-coumaric acid to caffeic acid indicated that polyphenoloxidase was mediating this conversion in vitro. However, since p-coumaric acid is apparently not hydroxylated to caffeic acid in viva, other enzyme systems may be operative. This is certainly the case with the nonphenolic ester, cinnamyl-quinic acid. At the present time little can be said about the general importance of esterification as a preliminary step to hydroxylation reactions in other tissues. Experiments with tobacco leaf disks have shown that both coumaryl and cinnamyl esters do arise in response to administration of phenylalanine and trans-cinnamic acid. However, the nonaromatic moieties have not, as yet, been identified.

2425

Cinnamyl and p-Coumaryl Esters as Intermediates in the Biosynthesis of Chlorogenic Acid Carl C. Levy and Milton Zucker J. Biol. Chem. 1960, 235:2418-2425.

Access the most updated version of this article at http://www.jbc.org/content/235/8/2418.citation Alerts: • When this article is cited • When a correction for this article is posted Click here to choose from all of JBC's e-mail alerts Downloaded from http://www.jbc.org/ by guest on August 13, 2017

This article cites 0 references, 0 of which can be accessed free at http://www.jbc.org/content/235/8/2418.citation.full.html#ref-list-1