Plant Physiol. (1984) 76, 170-1740032-0889/84/76/01 70/05/$0 1.00/0

Synthesis of L-( )-Tartaric Acid from L-Ascorbic Acid via5-Keto-D-Gluconic Acid in Grapes

Received for publication March 6, 1984 and in revised form April 24, 1984

KAZUMI SAITO* AND ZENZABURO KASAIThe Radioisotope Research Center, Kyoto University, Kyoto 606, Japan (K. S.); and The ResearchInstitutefor Food Science, Kyoto University, Uji 611, Japan (Z. K.)

ABSTRACT

5-Keto-L-idionic acid (a6-keto-Dngluconic acid, D-xylo-5-hexulosonicacid) was found as a metabolic product of L-ascorbic acid in sLices ofimmature grapes, Vitis labrusca L. cv 'Delaware'. Specifically labeledcompounds, recognized as metabolic products of L-ascorbic acid in grapes,were fed to young grape tissues to investigate the metabolic pathwayfrom L-ascorbic acid to L4+)-tartaric acid.

Label from dehydro-L-l-"Cjascorbic acid, 2-keto-L-1-`Clidonic acid(L-xylo-2-hexulosonic acid), L-l-'Cqidonic acid, or 5-keto-L-1-'4Cjidonic acid was incorporated into L-(+)-tartaric acid in high yields as itwas in the L1-4Clascorbic acid experiment. In a double label experimentinvolving a mixture of L-1-`4qidonic acid and L-2-AHjidonic acid, the3H/'4C ratios of 5-keto-L-idonic acid and L-(+)-tartaric add synthesizedin young grape leaves were almost the same as the value of the L-idonicacid fed. Label from 5-keto-L-16-4Cidonic acid was incorporated intosugars and insoluble residue in the same way as L{-6-'Cjascorbic acidwas metabolized in grapes.

These results provide strong evidence that in grapes L-(+)-tartaric acidis synthesized from the C4 fragment that corresponds to the Cl to C4group of the 5-keto-L-idonic acid derived from L-ascorbic acid via 2-keto-L-idonic acid and L-idonic acid.

Since the first report that [ I-'4C]AA' is a precursor ofcarboxyl-labeled TA in grapes (9), several kinds of specifically labeled AAhave been fed to TA-accumulating higher plants to investigatethe metabolic route ofAA to TA. When plants belonging to theVitaceae have been the material used, results have indicatedclearly that the C4 fragment corresponding to the Cl to C4 groupof AA is metabolized to TA and that the remaining C2 moietythat corresponds to the C5 to C6 group of AA contributes tosugar metabolism (3, 7, 10, 11, 15, 17, 18). Little informationon the metabolic pathway connecting AA and TA has beenobtained because intact tissues of plants which belong to theVitaceae accumulate only small amounts of AA metabolitesother than TA under natural conditions.To trace the conversion ofAA to TA, [1 -'4C]AA was vacuum-

infiltrated into slices of immature grapes and incorporated intolabeled metabolites over a 5-h period (12). Four labeled products

I Abbreviations: AA, L-Ascorbic acid; dehydro-AA, dehydro-L-ascor-bic acid; TA, L-(+) tartaric acid; 2-keto-IA, 2-keto-L-idonic acid; IA, L-idonic acid; 5-keto-IA, 5-keto-L-idonic acid ( 5-keto-D-gluconic acid);FID, hydrogen flame ionization detector; TMS, trimethylsilyl. IUPACrules of nomenclature identify 2-keto-L-idonic acid as L-xylo-2-hexulo-sonic acid and 5-keto-L-idonic acid (-5-keto-D-gluconic acid) as D-xylo-5-hexulosonic acid.

were formed. Three, identified as 2-keto-IA, IA, and L-idono-y-lactone, were labeled rapidly before incorporation of 14C into TAbegan to increase linearly. Therefore, we concluded that at leastone ofthese compounds is the effective precursor ofTA in grapesand is located on the metabolic pathway between AA and TA.The fourth product, which was an anionic substance referred toas Peak-III in a previous paper (12), is identified as 5-keto-IA.This compound had long been supposed to be the precursor ofTA in microorganisms (6) and higher plants (2), but there is littleexperimental data to show that 5-keto-IA really is the precursorof TA.Here we present results of feeding experiments in which spe-

cifically labeled 2-keto-IA, IA, L-idono-'y-lactone, or 5-keto-IAwas introduced into slices or intact tissues of grapes in order todetermine the synthetic pathway between AA and TA.

MATERIALS AND METHODSPlant Materials. Grape tissues were obtained from a vine of

Vitis labrusca L. cv 'Delaware' growing in a vineyard at KyotoUniversity. Slices of young grapes (38 d after anthesis) 1-mmthick were prepared by the procedure described previously ( 12).Grape apices containing a growing point, two folded leaves, andthree small leaves (about 10 cm long, 1.5 g fresh weight) wereexcised when new shoots were growing actively, and solutionscontaining one or more of the labeled compounds were fed tothe freshly severed stem. Young grape leaves (7-8 cm wide) alsowere harvested from the tips of actively growing new shoots.

Chemicals. Methyl-2-keto-L-gulonate (imethyl-2-keto-L-idonate) was purchased from ICN Pharmaceuticals Inc. Potas-sium 5-keto-D-gluconate (mpotassium 5-keto-L-idonate), thestandard used to identify Peak-III, was purchased from SigmaChemicals Corp.

Labeled Compounds. [1-'4C]AA (8.4 Ci/mol) was purchasedfrom New England Nuclear Corp. Dehydro-[1-_4C]AA was pre-pared by Br2 oxidation of a [1-'4C]AA solution. When oxidationwas complete, the excess Br2 was expelled by bubbling N2 throughthe vessel; the colorless solution produced was run throughsuccessive columns of Dowex 50W x 8(H+) and Dowex I x8(0H-). The resulting liquid containing dehydro-[1-'4C]AA wasconcentrated under reduced pressure at 30C. 2-Keto-[l-14CJIA(0.2 Ci/mol) was synthesized in grape slices which had beenvacuum-infiltrated with [1-'4C]AA and iodoacetic acid then iso-lated by Dowex ion-exchange column chromatography and pa-per chromatography as reported previously (12).

[1-'4CJIA was prepared by condensing K'4CN (8 Ci/mol,purchased from NEN) with an excess amount of L-xylose, thenhydrolyzing the cyanhydrin synthesized with acid. The [1-'4C]IAthat formed was separated from L-[ 1-'4C]gulonic acid, which alsowas synthesized simultaneously, by paper chromatography usingthe solvent system described below after lactonizing the acids at60°C for 3 h under reduced pressure.

170 www.plantphysiol.orgon May 8, 2020 - Published by Downloaded from

Copyright © 1984 American Society of Plant Biologists. All rights reserved.

TARTARIC ACID SYNTHESIS FROM 5-KETOGLUCONIC ACID

[2-3H]IA was synthesized by the reduction ofan excess amountof 2-keto-IA with KB3H4 (410 Ci/mol, purchased from NEN).The [2-3H]IA produced was separated from L-[2-3H]gulonic acidby paper chromatography as described below. 5-Keto-[ 1-, or 6-'4C]IA was prepared from D-[1-, or 6-'4C]glucose (6.5 and 5.8Ci/mol, purchased from NEN) by two synthetic steps. First, D-

['4C]glucose was oxidized to -['4C]gluconic acid by glucoseoxidase (EC 1.1.3.4), then the D_['4C]gluconic acid formed wastreated with a suspension of Gluconobacter suboxydans IFO12520 at 4°C, for 12 h in 0.3 M citrate buffer (pH 3.0) to obtain5-keto-['4C]IA (14). The respective specific radioactivities of the5-keto-[1-'4C]IA and 5-keto-[6-'4C]IA synthesized were deter-mined to be 10.9 Ci/mol and 8.7 Ci/mol by spectrophotometricmicroanalysis of 5-keto-IA (13). The final yield of 5-keto-['4C]IA was about 70% based on the amount of starting D_[14C]glucose.

Administration of Labeled Compounds. A small amount of a

solution of the labeled compound (1 uCi/10 ,l) was vacuum-infiltrated into three grape slices as described previously (12). Inthe feeding experiment with young leaves or apices, the solutionof the labeled compound (1-7 ,uCi/25 ul) was fed through thefreshly severed surface of the petiole or stem by placing thesevered part in a small feeding vial made from the plastic tip ofa micropipette. In both cases, the labeled solution of the acidiccompound (IA, 2-keto-IA, or 5-keto-IA) was neutralized byNaOH prior to its administration. When most of the labeledsolution had been absorbed, a small amount of water was addedrepeatedly (50 ,d x 3) to rinse out the remaining labeled com-pound. Then the leaves or apices were transferred to a smallbeaker filled with water, and their cut surfaces submerged in thewater for the remainder of the metabolic period. Light wasprovided continuously by ten 15-w fluorescent lamps at an

average distance of 40 cm.In some cases, feeding procedures were performed under a

glass bell-jar to collect for respired "4CO2. Two, series-connected400-ml gas dispersion bottles, each containing 140 ml of N KOH,were used. More than 95% of the 14C02 was trapped in the firstbottle.

Analytical Procedures Used to Identify Metabolic Products.Most of the analytical procedures used in this study have beendescribed in th preceding paper (12). Labeled tissues were groundin a mortar, then extracted with 0.1% oxalic acid solution. Afterprecipitating off the oxalate as its Ca-salt by adding Ca-formate,the remaining liquid was fractionated into cationic, anionic, andneutral compounds by Dowex ion-exchange column chromatog-raphy. In the double label experiment, samples from the gasdispersion bottle and the neutral effluent were lyophilized torecover 3H20. Descending paper chromatography was performedin ethyl acetate-pyridine-water-glacial acetic acid-methanol(14:10:3:2:2, v/v) (15) when necessary, especially in the analysisof neutral compounds. The 14C radioactivity in the insolubleresidue was counted after treating this residue in a sampleoxidizer (type 206, Packard Instrument Corp.). Radio gas chro-matography of Peak-III was performed as described elsewhere(12). A Shimadzu, GC-MS-6020 instrument with NH3 as thereacting gas at an accelerating voltage of 3.5 kv and an ionizingelectron energy of 30 ev was used for GC-MS analysis. A glasscolumn (2 mm i.d. x 0.5 m) packed with 3% OV-17 was usedin the chromatographic separation.

Determination of Radioactivity. Radioactivity was countedwith a Tri-Carb 2003 liquid scintillation spectrometer (PackardInstrument Corp.) as described elsewhere (12), but in the doublelabel experiment the dpm values of 3H and 14C were determinedwith an Aloka 903 liquid scintillation spectrometer by the exter-nal ratio method.

RESULTS

Metabolism of [1-'4CJIA and L-1-'4CjIdono--y-Lactone inGrape Slices (Experiment 1). [1-'4C]IA or L-[l-'4C]idono-y-lactone was vacuum-infiltrated into grape slices and left to me-tabolize for 5 h. After freezing the slice in a bath of methanol-dry ice to stop the metabolic reaction, the '4C metabolites derivedfrom each labeled substrate were analyzed. Analytical results aresummarized in Table I which also gives the results ofthe [1-'4C]AA experiment for comparison.A previous paper (12) reported that both substrates (IA and L-

idono-y-lactone) might be precursors of TA in grape slices. Butin this experiment, incorporation of '4C into TA from eachlabeled compound was very low (respectively, 1.5 and 3.0% for5 h of metabolism) compared to the [1-"C]AA experiment inwhich 20% of the total infiltrated '4C was incorporated into TAunder the same conditions. In both experiments, substantial '4Cwas recovered as IA and L-idono-y-lactone (84% of the '4C inthe [1-'4C]IA experiment and 73% in the L-[1-'4C]idono-y-lac-

Table I. Incorporation of14C into Metabolic Products in Slices ofImmature Grapes Vacuum-Infiltrated with [1-"CJIA, L-[J-"CJIdono-'y-

Lactone, and [J-'4CJAAOne ,Ci (about 0.03 mg) ofeach labeled compound was administered.

The metabolic period was 5 h.

0.1% Oxalic Acid-Soluble Radioactivity

[1-'4C]AA [1-_4C]IA L-[l-"4CJIdono-%y-lactone

Neutral fractionL-Idono-y-lactone 1.6 11.5 27.6Other compounds 7.1 12.8 22.1

Cationic fraction 0 0 0Anionic fractionIA 5.7 72.4 44.92-Keto-IA 5.3 0.5 0.6Peak-III 1.2 0.3 0.5TA 20.0 1.5 3.0Other compounds 59.la 1.0 1.3

a This fraction contained unmetabolized [1-'4C]AA.

Table II. Incorporation of'4C into Metabolic Products in Grape ApicesLabeled by[l-'4CIAA, Dehydro_[1-'4CJAA, 2-Keto1[-'4CJIA, and

[1-'4CJIAWith the exception of 2-keto-[1-'4C]IA, I jsCi (about 0.03 mg) of each

compound was administered. The metabolic period was 24 h.

0.1% Oxalic Acid-Soluble Radioactivity

[1-'4C]AA Dehydro- 2-Keto- [1-14C]IA

Neutral fractionGlucose + fructose 1.0 4.3 0.3 1.7Sucrose 1.3 1.3 0.4 1.0Other compounds 0.9 1.5 2.7 0.8

Cationic fraction 0.4 1.0 0.6 0.3Anionic fractionIA 3.1 2.0 1.6 5.9AA 0.6 0.6 0 0Peak-II 0.7 0.8 0.1 1.12-Keto-IA 0.8 1.4 1.3 0TA 86.9 81.2 89.8 85.6Other compounds 4.3 5.9 3.2 3.6

' One pCi (1.1 mg) was administered.

171

www.plantphysiol.orgon May 8, 2020 - Published by Downloaded from Copyright © 1984 American Society of Plant Biologists. All rights reserved.

Plant Physiol. Vol. 76, 1984

tone experiment) but a small amount of 14C of both labeledsubstrates was incorported into the characteristic metabolites ofAA in grape slices, including 2-keto-IA and Peak-III as reportedpreviously (12). The amounts of '4C incorporated into thosecomponents were not, however, as great as in the accompanying([1-'4CJAA) experiment.Metabolism of Dehydro-1-4CIAA, 2-Keto-1-'4CIIA, and 11-

'4CIIA in Grape Apices (Experiment 2). Dehydro-[1-'4C]AA, 2-keto-[1-'4C]IA, or [1-'4C]IA was fed to grape apices throughfreshly cut stem surfaces then allowed to be metabolized for 24h. The analytical results of these experiments are summarized inTable II. This table also contains results of a [I-'4C]AA feedingexperiment for comparative purposes.Almost all the radioactivity was present in the anionic com-

ponents, especially TA. Incorporation of '4C into TA in the 2-keto-[1-_4C]IA and [1-'4C]IA experiments was comparable toresults from the [1-'4C]AA experiment, evidence that 2-keto-IAand IA are intermediates between AA and TA in grapes. Distri-bution of '4C among metabolites of dehydro-[1-'4C]AA wassimilar to that of [1-'4C]AA. Some '4C appeared in Peak-III ineach experiment although the amount was very small. No labelwas detected in dehydro-AA or in L-idono-y-lactone in eachexperiment.Metabolism of 11-'4CjIA and 12-3HIIA in Young Grape Leaves

(Experiment 3). As judged by their chemical structures, both 2-keto-IA and IA are metabolically close to AA. To decide whichis the metabolite first formed from AA, a double label experimentwas performed. A mixture of [1-_4C]IA and [2-3H]IA (3H/'4C =17.9) was fed to young grape leaves and left to be metabolizedfor 24 h (Table III). Metabolism of [1-'4C]IA in young leavesresembled that found in grape apices. Moreover, metabolism of[2-3H]IA was similar to that of [ 1-'4C]IA. About 92% of the 3Hwas present from IA, Peak-III, and TA, and the 3H/'4C ratios ofthese compounds were very similar to the original value of theIA supplied to the leaves. Neither 3H nor '4C were incorporatedinto 2-keto-IA.

Identification of Peak-III. Labeling ofPeak-III (Ref. 12; TablesI, II, and III in this paper) always accompanied the labeling ofTA in grapes. This suggested to us that Peak-III was possibly anintermediate in the conversion of AA to TA in grapes andprompted us to determine the identity of Peak-Ill.

['4C]Peak-III was isolated from grape slices that had beenvacuum-infiltrated with [I -'4C]AA for 3 h and partially purifiedby Dowex I x 8 (formate) ion-exchange column chromatography

Table III. Metabolism of[2-3H, 1-'4CJIA in Young Grape LeavesThe 3H/'4C ratio of the IA originally fed to leaves was 17.9 ([2-3HJIA,

6.7 #Ci; [1-'4C]IA, 0.38 ;&Ci; totally, about 0.03 mg). The metabolicperiod was 24 h.

RadioactivityFound in the0.1% Oxalic 3H/14CAcid Extract Incorporatedand in CO23H 14C

% RatioCO2 2.2H20 3.2Neutral fraction 0.8 1.4 10.3Cationic fraction Trace 0.2 1.4Anionic fractionIA 4.9 4.9 17.8Peak-III 1.9 2.1 16.4TA 85.4 85.0 18.0Other compounds 3.8 4.2 16.3

FIG. 1. Radio gas chromatogram of the TMS derivative of Peak-III.(a), Main FID peak of the TMS derivative of Peak-III: (b), radioactivepeak of the TMS derivative of Peak-III; (c), FID peak of the TMSderivative of authentic 5-keto-IA.

Pentokis-O-trimethylsilyl- 1M TMS) NH

5 -keto-D-gluconic acid M * NHI

00o 300 S00500 60mle

TMS-derivative of PK-E 512

.tL,.II,2

200 300 mleWie

500

FIG. 2. GC-MS (chemical ionization spectrum) ofthe TMS derivativeof Peak-III. NH3 was the reacting gas.

Table IV. Metabolism ofS-Keto-[1-'4CJIA and S-Keto-[6-'4CJIA inGrape Apices

One pCi (about 0.03 mg) ofeach labeled compound was administered.The metabolic period was 24 h.

Total Radioactivity Recovered

5-Keto-[1-"'C]IA 5-Keto-[6-'4C]IA

CO2 0 5.6Neutral fraction 0.6 15.9Cationic fiaction 0.5 7.1Anionic frction

5-Keto-IA 1.4 1.7TA 91.8 I.5Other compounds 4.8 13.4

Insoluble residue 0.9 54.8

and paper chromatography. It contained about 1% of the 14Cused in the infiltration. Since Peak-II was a metabolite of IA ingrapes (Table III), we compared its chromatographic behavioragainst several compounds thought to be metabolites of IA. 5-Keto-IA gave the same chromatographic behavior as Peak-III byion-exchange column chromatography and paper chromatogra-phy. Therefore, the relation between Peak-III and 5-keto-IA wasinvestigated in detail by radio gas chromatography and GC-MS

172 SAITO AND KASAI

2

15

www.plantphysiol.orgon May 8, 2020 - Published by Downloaded from Copyright © 1984 American Society of Plant Biologists. All rights reserved.

TARTARIC ACID SYNTHESIS FROM 5-KETOGLUCONIC ACID

techniques.A radio gas chromatogram of the TMS derivative of Peak-III

is shown in Figure 1. Peak-III still contained several componentsas impurities, but both the retention times of the main peak(peak a) detected by FID and of the radioactive peak (peak b)coincided with the retention time (1 1.5 min) for the FID peak(peak c) of the TMS derivative of authentic 5-keto-IA. To con-firm the chemical structure of the radioactive constituent con-tained in Peak-Ill, we used GC-MS analysis of peaks a and c.The mass spectrum obtained from peak a by chemical ionizationwith NH3 as the reacting gas was identical to peak c shown inFigure 2. Consequently, we concluded that Peak-III was identicalto 5-keto-IA.Metabolism of 5-Keto11-[l-"IA and 5-Keto-6-`CIIA in Grape

Apices (Experiment 4). To confirm whether TA is synthesizedfrom AA via 5-keto-IA, specifically labeled 5-keto-IA (5-keto-[ 1-1"C]IA or 5-keto-[6-"4C]IA) was fed to grape apices and allowedto metabolize for 24 h. In the 5-keto-[ I-'4C]IA experiment (TableIV), almost all ofthe "C (92%) was recovered as TA. By contrast,very little label from 5-keto-[6-"C]IA was incorporated into theTA although a substantial amount did appear in neutral com-pounds and insoluble residue. More "C from 5-keto-[6-"C]IAwas respired as CO2 and converted into cationic compoundsthan from 5-keto-[1-"C]IA. No 14C appeared as IA or 2-keto-IAin either experiment.

DISCUSSION

In the preceding paper (12), we showed that vacuum-infiltra-tion of [I-_4C]AA into slices of immature grapes resulted in theincorporation of "C into 2-keto-IA, IA, and L-idono-y-lactoneand this incorporation preceded linear accumulation of labeledTA. To determine whether these compounds participate as in-termediates, we have prepared specifically labeled 2-keto-IA, IA,and its lactone and present here the results of a study in whichthese labeled compounds were administered to young grapetissues to ascertain their relative roles during conversion of AAto TA.

Incorportion of "C of 2-keto-[ 1-`C]IA and [ -`'C]IA into TAwas as comparable to results with [1-`C]AA (Table II). Appar-ently, 2-keto-IA and IA are intermediates between AA and IA ingrapes. To determine their relative positions in this metbolicpathway, a double label experiment with [-1-'C, 2-3H]IA (exper-iment 3) was run. The 3H/"C ratio in IA was preserved in TA(Table III), a clear indication that TA was produced from IAwithout loss of 3H. Either 2-keto-IA precedes IA during the AAto TA conversion or there is a stereospecific oxidation/reductionthat involves retention of 3H on carbon 2. The former appearsmore probable and conceivably proceeds by hydrolysis of theAA 2-KETO-IA IA 5-KETO-IA TA

0

cHCHOCJ

HOCHOCH

6H20H

I I0

oc

HCL

&H20H

DEHYDRO-AA

COOHC-=O

HOCHHCOH

HOCH

6H20H

COOH

HCOH

, HOCH

4 HIOH0HOCH

CH2OH

It0

HCOH

HOCHHC-

HOCH

CH2OH

COOH

HCOH

HOCH

HCOH

CH2OH

COOH

HCOH

HOCH

HCOH

CH20H

C

CICI

I-

C

I

SUGAR

L-IDONO-T- D-GLUCONICLACTONE

FIG. 3. Synthetic pathway for TA starting from AA in grapes.(), Known pathway; (--), supposed pathway.

lactone of AA, tautomeric rearrangement to 2-keto-IA and ster-eospecific reduction to IA.

Distribution of label in products from 2-keto-[1-"C]IA and[ -'4C]IA was similar to that obtained with [I-14C]AA (Table II).Most notable was the large amount (86-90%) of label thatappeared in TA from each labeled substrate. In intact younggrape tissues, each intermediate metabolized principally in thedirection of TA synthesis (Table II); AA -X 2-keto-IA -+ IA -.-X TA. Slices of immature grapes, however, failed to utilizeseither 2-keto-IA or IA effectively (Table I). The reason for alower conversion to TA with [1-`C]IA or L-(I-'4Cjidono-'y-lactone than with [I-'4C]AA is not yet known.As reported in a previous paper, the maturation ofgrape leaves

was accompanied by an increase in the ratio of '"C in dehydro-AA to that in AA when labeled AA was supplied to detachedleaves (1 1). In experiment 2, no "4C appeared in dehydro-AAwhen either [1-_4C]AA or dehydro-[l-'4C]AA was fed to grapeapices and the labeling pattern in each case was similar (TableII). Apparently, the young tissues used in the present studyrapidly converted dehydro-AA to AA. This may explain whydehydro-[ 1-'4C]AA was as effective a precursor ofTA as [I-14C]AA.

Characterization of the labeled unknown acid obtained from[I-'4C]AA-labeled immature grape slices (Peak-III in Ref. 12) as5-keto-IA provided a very important clue in establishing theoverall pathway from AA to TA in the grapes. When 5-keto-[ 1-'4C]IA and 5-keto-[6-'4C]IA were introduced into grape apices,their metabolic behavior was exactly the same as that displayedby [1-'4C]AA and [6-'4C]AA (7,1 1). Results strongly suggest that5-keto-IA is an intermediate between AA and TA in grapes. Itsposition in the pathway is thought to be between IA and TA forthe following reasons. (a) In intact grapes TA synthesis is theprincipal product of AA metabolism. (b) When [I-'C]AA, de-hydro-[1_-4C]AA, or 2-keto-[1'-"C]IA was supplied to younggrape leaves, both 2-keto-IA and IA were labeled; but, when [1-'4C]IA was supplied, no "4C appeared in 2-keto-IA (Table II). (c)Carbon-14 was incorporated into Peak-III (5-keto-IA), albeit insmall amounts, when grape apices were labeled with [I-'C]AA,dehydro-[1-14C]AA, 2-keto-[1-'4C]IA, or [1-'4C]IA, but no "4Cwas detected in 2-keto-IA and IA when 5-keto-['4C]IA (eitherCl- or C6-labeled) was introduced into grape apices (Table IV).Similar experimental results were obtained after feeding [1-'4C]IA and 5-keto-[1-4C]IA to Pelargonium (K. Saito, S. Morita, Z.Kasai, unpublished studies), although the metabolic activity of5-keto-IA was much lower than that observed in grapes.From these observations, we propose the scheme as shown in

Figure 3. Intact young tissues (detached leaves or apices) failedto incorporate 14C from [1-'4C]AA, 2-keto-[1-'4C]IA, or [1-_4C]IA into L-idono-y-lactone; therefore, L-idono-y-lactone probablyis not a metabolite of AA under natural conditions. In thescheme, the stereospecificity of AA-derived TA (L-[+]-form) ingrapes (16) is defined absolutely by enzymic reduction by 2-keto-IA to IA.

5-Keto-IA (5-keto-D-gluconic acid) is oxidatively cleaved toTA by chemical means under both acidic and alkaline conditions(1, 4). In 1949, Gakhokidze first proposed that TA is synthesizedin plants from the C4 freed by a split between C4 and C5 of 5-keto-IA (2). Since then, experimental support for this as it appliesto TA synthesis in the grape has accumulated. Ribereau-Gayonshowed that D-[ I-'4C]glucose is a better precursor ofTA in grapesthan is D-[6-"4C]glucose (8). He suggested that 5-keto-IA mightbe a precursor ofTA in grapes. In microorganisms, TA is knownto be synthesized by the oxidation of sugars via D-gluconic acidand 5-keto-IA (6).To the best of our knowledge, ours is the first report to

demonstrate the actual conversion of 5-keto-IA to TA in higherplants. Our results also have made it clear that 5-keto-IA is a

173

www.plantphysiol.orgon May 8, 2020 - Published by Downloaded from Copyright © 1984 American Society of Plant Biologists. All rights reserved.

Plant Physiol. Vol. 76, 1984

metabolic product ofAA in grapes. Thus, it is possible to explainthe conversion of AA to TA in grapes clearly. A new question,however, is posed as to whether TA actually is synthesized viaAA in intact grapes. D-[U-'4C]Gluconic acid has been reportedto be a somewhat better precursor ofTA than [U-'4C]sucrose inyoung grapes (10). This experimental result has been explainedonly by the supposition that TA is synthesized from the C4moiety (Cl to C4) ofAA in grapes. The meaning of the experi-mental results presented by the previous paper (10) should bereconsidered in the light of the fact that TA also can be synthe-sized from 5-keto-IA in grapes and that 5-keto-IA is a putativeproduct of 1-glucose via D-gluconic acid in microorganisms andpossibly in higher plants as well.One report suggests a synthetic pathway for TA from 5-keto-

IA in a microorganism. Kotera et al. (5) found a new oxidationproduct of 5-keto-IA, 1,2-dihydroxyethyl hydrogen L-(+)-tartaricacid, which they named pretaric acid, in the cell culture liquidof a high TA-producing mutant of Gluconobacter suboxydans.On the basis of its chemical structure and properties, they pro-posed that pretaric acid is located on the metabolic pathwaybetween 5-keto-IA and TA. Experiments to clarify the metabolicsteps from 5-keto-IA to TA in microorganisms and in higherplants are now needed.

Acknowledgments-We thank Dr. 0. Adachi for the supply of the microorgan-ism and for his valuable suggestion on preparing 5-keto-L-['4C]idonic acid from D>['4CJglucose and Dr. T. Tomana for providing the plant materials used. We aregrateful to Dr. T. Suzuki and to the Kyoto Analytical Center of Shimadzu Corp.for the GC-MS analysis of Peak-III. Thanks are also due to Dr. F. Loewus for hisvaluable advise in revising this manuscript.

LITERATURE CITED

1. BARCH WE 1933 Oxidation of 5-ketogluconic acid with nitric acid in thepresence of vanadium. J Am Chem Soc 55: 3653-3658

2. GAKHOKIDZE AM 1949 Mechanism of formation of organic acids in plants.Soobshcheniya Akad Nauk Gruzinskoi SSSR 10: 25-3 1, C.A. 46: l0307e

3. HELSPER JPFG, K SArro, FA LOEWUS 1981 Biosynthesis and metabolism ofL-ascorbic acid in virginia creeper (Parthenocissus quinquefolia L.). Planta152: 171-176

4. ISBELL HS, NB HOLT 1945 Oxidation ofgalacturonic acid and of5-ketogluconicacid in alkaline solution. J Res Natl Stand 35: 433-438

5. KOTERA U, T KODAMA, Y MINODA, K YAMADA 1972 Isolation and chemicalstructure ofnew oxidation product of 5-ketogluconic acid, and a hypotheticalpathway from glucose to tartaric acid through this new compound. AgricBiol Chem 36: 1315-1325

6. LOCKWOOD LB, GEN NELSON 1951 d-Tartaric acid. US Patent 2,559,6507. LOEWUS FA, HA STAFFORD 1958 Observation on the incorporation of 14C into

tartaric acid and the labeling pattern of D-glucose from an excised grape leafadministered L-ascorbic acid-6-'4C. Plant Physiol 33: 155-156

8. RIBEREAU-GAYON G 1968 A study of the mechanism of synthesis and conver-sion of malic acid, tartaric acid and citric acid in Vitis vinifera L. Phytochem-istry7: 1471-1482

9. SAITO K, Z KASAI 1969 Tartaric acid synthesis from L-ascorbic acid-1-'C ingrape berries. Phytochemistry 8: 2177-2182

10. SAITo K, Z KASAI 1978 Conversion of labeled substrates to sugars, cell wallpolysaccharides, and tartaric acid in grape berries. Plant Physiol 62: 215-219

11. SAITo K, FA LOEWUS 1979 The metabolism of L-[6-14CJascorbic acid indetached grape leaves. Plant Cell Physiol 20: 1481-1488

12. SAITo K, Z KASAI 1982 Conversion of L-ascorbic acid to L-idonic acid, L-idono-y-lactone and 2-keto-L-idonic acid in slices ofimmature grapes. PlantCell Physiol 23: 499-507

13. SCHRAMM M 1956 Spectrophotometric determination of5-ketogluconate. AnalChem 28: 963-965

14. SHINAGAWA E, K MATSUSHITA, 0 ADACHI, M AMEYAMA 1983 Selectiveproduction of 5-keto-D-gluconate by Gluconobacter strains. J Ferment Tech-nol 61: 359-363

15. WAGNER G, FA LOEWUS 1974 L-ascorbic acid metabolism in Vitaceae. Con-version to (+)-tartaric acid and hexoses. Plant Physiol 54: 784-787

16. WAGNER G, JC YANG, FA LoEwus 1975 Stereoisomeric characterization oftartaric acid produced during L-ascorbic acid metabolism in plants. PlantPhysiol 55: 1071-1073

17. WILLIAMS M, FA LOEWUS 1978 Biosynthesis of (+)-tartaric acid from L-[4-'4C]ascorbic acid in grape and geranium. Plant Physiol 61: 672-674

18. WILLIAMS M, K SAITo, FA LoEwuS 1979 Ascorbic acid metabolism in grapeand geranium. Phytochemistry 18: 953-956

174 SAITO AND KASAI

www.plantphysiol.orgon May 8, 2020 - Published by Downloaded from Copyright © 1984 American Society of Plant Biologists. All rights reserved.