Plant Physiol. (1983) 73, 869-8730032-0889/83/73/0869/05/$00.50/0

A Pathway for Photosynthetic Carbon Flow to Mannitol inCelery Leaves'ACTIVITY AND LOCALIZATION OF KEY ENZYMES

Received for publication May 23, 1983 and in revised form July 18, 1983

MARY E. RUMPHO, GERALD E. EDWARDS2, AND WAYNE H. LOESCHERDepartments ofBotany (M. E. R., G. E. E.) and Horticuhture (W. H. L.), Washington State University,Pullman, Washington 99164-4230

ABSTRACT

In the polyol producing plant, celery (Apium graecokls L.), mannitolis a major photosynthetic product and a form in which carbohydrate istranslocated. Measurements of whole leaf extracts of celery indicatedsubstantial activity of the following enzymes: mann P reductase,mannose-6P isomerase, mannitol-l-P phosphatase, and nonreversibleglycermldehyde-3-P dehydrogenase. The activities of these enzymes wereeither undetectable or very low in the nonpolyol producing plants, Secalkcerealk L. (rye) and Vigna mango (L.) Hepper (black gram).

Mesophyll protoplasts were enzymically isolated from celery leaves,broken with a Yeda press and the intracellular lcization of the aboveenzymes for mannitol synthesis studied following differential and/orsucrose density gradient centrifugation of the protoplast extract. Thesedata suggested the enzymes involved in mannitol synthesis are exclusivelylocalized in the cytoplasm. Ninety-five to 100% of the activity of theseenzymes, along with the cytoplasmic marker enzyme phosphoenolpyrn-vate carboxylase, was found in the cytosolic fraction.We propose the pathway of photosynthetic carbon flow from triose-P

to mannitol in celery occurs via fructose-6-P, mannose--P, and mannitol-I-P, these final reactions being catalyzed by the cytoplasmic enzymes,mannose-6-P isomerase, NADPH-dependent mannosP reductase,and mannitol-l-P phosphatase, respectively. The requirement forNADPH may be met via the cytoplasmically located NADP-linkednonreversible glyceraldehyde-3-P dehydrogenase.

Within the plant kingdom the distribution of polyols is wide-spread, occurring in fungi, algae, lichens, and lower and highervascular plant species (4, 15). Although it has been estimated (4)that up to 30% of all primary production goes through polyolsrather than sugars, very little is known concerning their completedistribution, physiological role, or metabolism particularlyamong higher plants. In addition, most studies on polyols arecomplicated by the variable effects environment and plant de-velopment have on polyol production (1, 4, 7, 8, 16).Proposed roles for polyols in plants have included: (a) serving

as a source of carbohydrate (energy reserve and reducing power)(8, 20, 26); (b) being involved in the regulation of partitioning asthe major translocated compound (5, 25, 30); and (c) protectionagainst osmotic and freezing stress (1, 2, 21). Given that sucrosecan also perform these roles in various plants (9,26), the question

' This work was supported by National Science Foundation GrantPCM 82-04625.

2 To whom requests for reprints should be addressed.

of why and how polyols are used is still unanswered.At least 13 different polyols have been isolated from higher

plants, with mannitol being the most widespread and found inover 70 families (15). Mannitol is a major photosynthetic productand form of translocated carbohydrate in the higher plants inwhich it has been reported, comprising 10 to 60% of the solublecarbohydrate (4). Previous information on mannitol biosynthesisresulted largely from studies on brown algae (11). In this groupofplants, mannitol is produced by reduction ofF6P3 to mannitol-1-P (with NAD as cofactor) with subsequent dephosphorylationby a specific mannitol-l-P phosphatase. A NADP-dependentmannose-6-P:mannitol- 1-P oxidoreductase has been identified(17) in celery (Apium graveolens) leaves. In contrast to theenzyme in brown algae, this higher plant enzyme uses NADPHas cofactor and converts mannose-6-P rather than F6P to man-nitol- 1-P. It has also been shown with labeling studies (Redgwell,Loescher, and Bieleski, unpublished data), that mannitol is pro-duced in amounts roughly equal to sucrose and proceeds via theintermediates mannose-6-P and mannitol- 1-P in celery leaves.The presence ofmannitol in sieve tube exudate has been detectedin several higher plant species (30).Although in higher plants sucrose synthesis has been convinc-

ingly shown to occur outside the chloroplast in the cytoplasm (6,22, 23), neither the pathway to nor the localization of enzymesfor mannitol biosynthesis have been reported. In this study, wehave used whole leaves and enzymically isolated protoplasts todemonstrate the probable pathway and intracellular localizationof enzymes necessary for mannitol biosynthesis in celery leaves.

MATERIALS AND METHODSPlant Material. Celery (Apium graveolens L.) plants were

grown in a local greenhouse from November to March, withsunlight supplemented by metal halide lamps. Day/night tem-peratures were 21/17°C. Plants were watered daily and fertilizedweekly. Mature, fully expanded leaves were used from plantsapproximately 12 to 16 weeks old.

Vigna mungo (L.) Hepper (black gram) and Secale cereale L.(rye) were grown in growth chambers with a 16-h photoperiodand a day/night temperature regime of 25/20°C. The quantumflux density was about 1000 uE m-2 s-'. Plants used in theexperiments were approximately 3 weeks and 1 week old for V.mungo and rye, respectively.Enzyme Extraction. For the assay of enzymes from crude leaf

extracts, leaves (1 g fresh weight) were harvested in the light andimmediately frozen in liquid N2 and ground to a powder. The

3Abbreviations: F6P, fructose-6-P; DTE, dithioerythritol; G3P, glyc-eraldehyde-3-P; PEPCase, phosphoenolpyruvate carboxylase; G6P, glu-cose-6-P; PGA, 3-phosphoglycerate.

869 www.plantphysiol.orgon July 26, 2020 - Published by Downloaded from

Copyright © 1983 American Society of Plant Biologists. All rights reserved.

Plant Physiol. Vol. 73, 1983

powdered tissue was extracted with 10 ml of grinding mediumcontaining 450 mm sorbitol, 5 mm EDTA, 5 mm DTE,1% (w/v) PVP-40, and 50 mm Hepes-KOH, pH 7.8 (same as protoplastbreaking media). When assaying for mannose-6-P isomerase ormannitol- 1-P phosphatase, the EDTA was omitted from themedium. The extract was filtered through two layers of cheese-cloth, an aliquot taken for Chl determination, and then centri-fuged at 20,000g, 4°C, for 10 min. The supernatant was assayeddirectly except when assaying for mannitol- 1-P phosphatase, inwhich case the extract was rapidly desalted through a SephadexG-25 column. The column was pre-equilibrated with 50 mMHepes-KOH, 5 mm DTE, and 2 mM MgCI2 (pH 7.5).

Photosynthesis Measurements. Photosynthetic rates were de-termined on greenhouse-grown celery plants with an open IRgas analysis system as described (13). Leaf temperature wasmaintained at 25 ± 0.5°C, light intensity at 1800 gE m-2 sg',and the flow rate at 60 1/h. Photosynthetic rates were measuredat 21% 02 and 320 M1/1 CO2.

Protoplast Isolation. Protoplasts were enzymically isolatedfrom celery leaves harvested early in the morning. Leaf tissuewas sliced into 1-mm-wide segments with a double-edged razorblade. Approximately 8 g of tissue were cut per 100 ml digestionmedium. The enzyme medium contained 700 mM sorbitol, 5mM Mes-HCI, 1 mm CaCI2, 0.05% BSA, 2% Cellulase R-10, and0.25% Macerozyme R-10 (Yakult Biochemical Co., Ltd., Nishi-nomiyo, Japan), pH 5.5. Digestion took place for about 2.5 h ina 28°C water bath with slight agitation and an irradiance ofapproximately 200 ,E m-2 s-'. Excess heat was removed with a3-cm water filter between the leaf tissue and light source. Follow-ing digestion, the mixture was swirled and filtered through a 1-mm aperture tea strainer and 200-,um aperture nylon net toremove undigested tissue. The filtrate was centrifuged at 100gfor 5 min and the supernatant discarded.

Protoplast purification was carried out at 4°C by resuspendingthe pellet by shaking in a solution (about 2 ml) containing 5%(w/v) dextran T-20, 700 mm sucrose, 1 mM CaC12, 2 mM DTE,1% (w/v) PVP-40, and 50 mm Hepes-KOH (pH 7.8). Thesuspension was transferred into a 50-ml Babcock bottle and thevolume was brought to about 45 ml with the above medium.About 3 ml of a solution, containing 500 mm sucrose, 200 mMsorbitol, 1 mM CaCl2, 2 mM DTE, 1% (w/v) PVP-40, and 50 mMHepes-KOH (pH 7.8), was layered on top of the protoplastsuspension. A third layer of about 1.5 ml consisting of 700 mMsorbitol, 1 mM MgCl2, I mM EDTA, 10 mM DTE, 1% (w/v)PVP-40, and 50 mM Hepes-KOH (pH 7.8), then was added. Thegradient was centrifuged at 250g for 5 min and the protoplastscollected between the second and third layers. The protoplastswere washed once with the third layering solution and centrifugedat 100g for 5 min.

Preparation of Chloroplasts from Protoplasts. The pelletedprotoplasts were resuspended in breaking medium (about 35 zgChl ml-') containing 450 mm sorbitol, 5 mm EDTA, 5 mm DTE,1% (w/v) PVP-40, and 50 mm Hepes-KOH (pH 7.8). Theprotoplasts were ruptured by passing through a Yeda press underN2 at 65 psi.

Differential Centrifuption. The broken protoplast extracts(600 ,d aliquot/1.5 ml capacity centrifuge tube) were centrifugedat 820g, 4°C, for 2 min in a Fisher microfuge. The supernatantwas removed and the chloroplast pellet resuspended in an equalvolume of 25 mm Tris-HCI (pH 8.0) plus 5 mM DTE.

Sucrose Density Gradient Centrifugation. Two ml ofprotoplastextract (30-35 tsg Chl ml-'), prepared as above, was layered ontop of a sucrose gradient and centrifuged at 4,000 rpm for 5 minand then 10,000 rpm for 10 more min at 4°C in a Beckman SW28 rotor (18). Twenty-five fractions (1.2 ml) were collected pergradient with an ISCO gradient fractionator, beginning from thetop of the centrifuge tube. All fractions were kept on ice andenzymes assayed within 8 h. The gradient was similar to that ofMiflin and Beevers (18) and Usuda and Edwards (23) and

consisted of 4 ml 60% (w/w) sucrose, 2 ml each of 55, 50, and45% (w/w) sucrose, 5 ml 42% (w/w) sucrose, 2.5 ml each of 40,37.5, 35, and 32.5% (w/w) sucrose, and 3 ml of 30% (w/w)sucrose, all dissolved in 50 mm Hepes-KOH (pH 7.8), 2 mMEDTA, 5 mm DTE, and 1% (w/v) PVP-40. The gradient wasprepared 2 h before use and kept at4°C.Enzyme Assays. PEPCase (EC 4.1.1.31), NADP-G3P dehy-

drogenase, phosphorylating (EC 1.2.1.13), fumarase (EC 4.2.1.2),hydroxypyruvate reductase (EC 1.1.1.26), and NAD-malate de-hydrogenase (EC 1.1.1.37) were assayed as described (27). Forhydroxypyruvate reductase, 50 mm K-phosphate was used ratherthan Mes-KOH. Nonreversible G3P dehydrogenase (EC 1.2.1.9)was assayed according to Kelly and Gibbs (12). Other enzymeswere assayed as follows. Mannose-6-P:mannitol- 1-P oxidore-ductase (EC unassigned): 31 mM Tris-HCl (pH 7.5), 0.1 mMNADPH, 50 to 100 MI enzyme extract, and 5 mM mannose-6-Pin a total volume of1 ml; mannose-6-P isomerase (EC 5.3.1.8):31 mm Tris-HCI (pH 7.5), 0.5 mm NADP, 5mM MgC92, 3 IUG6P dehydrogenase, 6 IU G6P isomerase, 50 to 100 Al enzymeextract, and 2.5 mm mannose-6-P in a total volume of 1 ml.The above enzymes were measured at 25C by following the

change in A of the pyridine nucleotides at 340 nm. For fumarase,0.1% (v/v) Triton X-100 was routinely added to the extract. Allactivities reported were proportional to the amount of extractadded, dependent upon the addition of the appropriate substrateand linear for at least 3 min.

Mannitol-1-P phosphatase (EC 3.1.3.22) was assayed by themethod of Leigh and Walker (14). Thirty to 100 ul of desaltedextract was incubated for 10 min at,25°C with 35 mm Hepes-KOH (pH 7.5), 8.3mM MgCl2, and 3.3 mM mannitol-l-P in afinal volume of 300 Ml. The reaction was stopped by the additionof a solution of NH4-molybdate in TCA. Following precipitationand centrifugation, the pellet was washed with TCA and thendissolved in Tris buffer. To this was added the assay mixture forPi consisting of NH4-molybdate and SDS in H2SO4 and ascorbicacid. Following incubation at 37°C for 1 h, the change in A at700 nm was measured.Other Methods. Chl was determined by the method of Arnon

(3) or Wintermans and De Mots (28). The density ofeach sucrosefraction was determined by refractometry. The barium salts ofmannose-6-P and mannitol-l-P were converted to K salts byaddition of an equal molar amount of K2SO4 and subsequentremoval of BaSO4 by centrifugation.

RESULTS

Whole Leaf Enzyme Activities and Photosynthetic Rates. En-zyme activities from whole leaf extracts are shown in Table I forcelery and the nonpolyol producing plants, rye and V. mungo.Except for the enzymes involved in mannitol synthesis, compa-rable enzyme activities are found among the three species.NADPH-mannose-6-P reductase activity was only detected inextracts of celery leaves, while the activity of mannose-6-P isom-erase was about 50-fold greater in celery than either rye or V.mungo. This low activity in rye and V. mungo might be due toa nonspecific hexose epimerase or isomerase in these plants. Themannitol-l-P phosphatase activity in desalted crude extractsfrom celery leaves required Mg2e and was inhibited by EDTA.The pH of the reaction medium was varied from 5.5 to 9.0, andthe pH optimum for phosphatase activity was at pH 6.5. Thevalues at pH 7.5 are reported here (Table I) since there was someactivity at the lower pH values with rye and V. mungo whichmay be due to a nonspecific acid phosphatase. At all pH values,the activity of celery leaves was much greater than that of eitherV. mungo or rye, with that of V. mungo being barely detectableat any pH (data not shown). The activity of nonreversible G3Pdehydrogenase was 3.1- and 5.4-fold greater in celery extractsthan in V. mungo and rye extracts, respectively (Table I).The photosynthetic rate ofmature, fully expanded celery leaves

was measured for comparison with the above enzyme activitiesfor mannitol synthesis. Photosynthetic rates of 22 to 26 mg CO2

870 RUMPHO ET AL.

www.plantphysiol.orgon July 26, 2020 - Published by Downloaded from Copyright © 1983 American Society of Plant Biologists. All rights reserved.

PATHWAY FOR MANNITOL SYNTHESIS IN CELERY LEAVES

Table I. Enzyme Activitiesfrom Whole-LeafExtracts ofCelery, Rye,and Vigna mungo and Photosynthetic Ratesfrom Mature Leaves of

Celery

Measurement Celery V. Ryemungo

#mol mgr' Chi h-'NADP-G3P dehydrogenase 943.5 1160.5 744.3Fumarase 46.8 27.0 29.3Hydroxypyruvate reductase 929.0 1146.3 318.3NAD-Malate dehydrogenase 10937.3 6018.2 7977.1PEP carboxylase 24.5 32.2 31.2Nonreversible G3Pdehydrogenase 70.4 22.5 13.0

Mannose-6-P isomerase 240.1 3.9 5.7Mannose-6-P reductase 35.9 0 0Mannitol-l-P phosphatase 31.0 1.4 4.7Photosynthetic rate 136-160

dm-2 h-' or 136 to 160 ,umol mg-' Chl h-' (Table I) weremeasured.

Differential Centrifugation. The pattern of intracellular distri-bution of enzymes in celery leaves was first studied using proto-plasts broken with a Yeda press, and the extract centrifuged at820g for 2 min. Due to the small size ofthe protoplasts, repeatedpassage through a 20-;sm net or 25-gauge needle failed to dispersethe protoplast contents. Rupturing the protoplasts under N2 at65 p.s.i. with a Yeda press yielded complete breakage and dis-persal of the protoplast contents. Addition of EDTA and omis-sion of cations was required in the breaking medium to preventaggregation of the chloroplasts. Organelles were then separatedby differential centrifugation of protoplast extracts. The activityand distribution of marker enzymes for chloroplasts (NADP-G3P dehydrogenase), mitochondria (fumarase), peroxisomes(hydroxypyruvate reductase), and cytoplasm (PEPCase) are givenin Table II, along with those involved in mannitol synthesis.High chloroplast intactness was evident by the fact that 97% ofthe Chl and about 80% of the activity of the chloroplast markerenzyme, NADP-G3P dehydrogenase, were retained in the pellet(Table II), as well as by observation of the chloroplasts with aphase contrast microscope (data not shown). There was no

apparent pelleting of cytoplasmic enzymes with the chloroplasts,as PEPCase was found totally in the supernatant (Table II). Someaggregation of peroxisomes and mitochondria with the chloro-plasts occurred in the differential centrifugation step, based onsome activity of the marker enzymes fumarase and hydroxypyr-uvate reductase in the chloroplast pellet; however, these enzymeswere largely recovered in the supernatant fraction.The enzymes for mannitol synthesis (mannose-6-P isomerase

and reductase, and mannitol-1-P phosphatase) were exclusivelyrecovered in the supernatant fraction (Table II), indicating thatthe final steps for mannitol synthesis in celery occur outside thechloroplast. In addition, nonreversible G3P dehydrogenase(which can indirectly transfer NADPH from the chloroplast tothe cytoplasm [12]), was also found totally in the supernatant.The total activity ofeach enzyme in the protoplast extract (TableII) was similar to, although somewhat lower than, that obtainedfor crude whole leaf extracts of celery (Table I).

Sucrose Density Gradient Centrifugation. The localization ofenzymes in celery was further examined by sucrose densitygradient centrifugation (Fig. 1). Only one peak of Chl (Fig. la)was found, corresponding to the peak of the chloroplast markerenzyme, NADP-G3P dehydrogenase (Fig. lb). Hydroxypyruvatereductase peaked in fraction 6 (density = 1.20 g cm-3) with someactivity at the top of the gradient (Fig. ld), presumably due toslight breakage of the peroxisomes. The mitochondrial marker,fumarase, peaked in fraction 3 (Fig. ld) at a density of 1.18 gcm-3, with only residual activity at the top of the gradient. Noco-migration of the peroxisomes or mitochondria with the chlo-roplasts was observed in the gradient as occurred with differentialcentrifugation, as evidenced by the absence of peroxisomal ormitochondrial marker enzymes with the chloroplast marker at adensity of 1.29 g cm-3 (Fig. 1, b and d).PEPCase and the enzymes of interest relative to mannitol

synthesis (mannose-6-P isomerase and reductase and nonrevers-ible G3P dehydrogenase), were found only at the top of thegradient (Fig. 1, b and c), indicating a cytoplasmic localizationfor these enzymes. There was no activity of these enzymesassociated with the peaks of activity of peroxisomal, mitochon-drial, or chloroplastic marker enzymes. The total Chl and activityof each enzyme recovered from the gradient was at least 90% of

Table II. Distribution ofChl and Enzyme Activities after Differential Centrifugation ofCelery ProtoplastExtracts

Total Distribution of ActivityActivity

Measurement in 820g 820gProtoplast Pellet SupernatantbExtract

Asmol mgBChlh-'

Chl 97.0 3.0NADP-G3P dehydrogenase 426.9 76.3 23.7Fumarase 35.2 27.0 73.0Hydroxypyruvate

reductase 245.4 39.0 61.0PEP carboxylase 23.3 0 100Nonreversible G3Pdehydrogenase 31.5 0 100

Mannose-6-P isomerase 193.7 3.1 96.9Mannose-6-P reductase 16.4 0 100Mannitol-1-P phosphatase 12.3 0 100

'Total activity was calculated from the sum of activities in the supernatant and pellet.b The sum of the enzyme activities recovered in the supematant and pelleted fractions were greater than

90% of the activity in the total protoplast extract except for fumarase and hydroxypyruvate reductase, whichwere 66 and 71.5%, respectively.

871

www.plantphysiol.orgon July 26, 2020 - Published by Downloaded from Copyright © 1983 American Society of Plant Biologists. All rights reserved.

Plant Physiol. Vol. 73, 1983

E

,T

t040

orI?

0

10

E

0

0

;K

f-D

'EIL

OL0 5 10 15 20 25

*c

u

cX

r-

'E'5E

"k

a

x

I.o

CL

.cT'EzE'k

010.2

0lr

1.T10

0cc02

1.2

I.0

0.8

0.6-

0.4

0.2

e

,

n0_.0 E

0_

DU Ca 00

, e0e*

C X

z

Tz

E

-v

ir

x

2I0

0 5 10 15 20 25

Fraction Number

FIG. 1. Distribution of Chl (a) and activity of various enzymes (b-d)following sucrose density gradient centrifugation of protoplast extractsfrom celery. Fractions are numbered beginning with the top of thegradient. Presence of a particular symbol at or below the zero ordinateindicates no activity of the corresponding enzyme was present in thefraction.

the total protoplast extract added except for fumarase which was

about 80%.

DISCUSSION

Although polyols are found throughout the plant kingdom,their importance to photosynthetic carbon metabolism among

higher plants remains unanswered for several reasons. Amongthese are the difficulties encountered in analyzing for polyols inthe past, the variable effects environment and plant development

have on the synthesis and metabolism of polyols, and the vastliterature that exists on sucrose as a well known form of carbo-hydrate transported among the higher plants studied. In somehigher plants, however, polyols are the major form of translo-cated carbon in the phloem and storage compound in the sinktissue (4). The control mechanisms of source leaf partitioning ofcarbon into polyols versus sucrose are unknown; however, theyare likely to be regulated by environmental factors and subjectto close metabolic control by the developing sink.

In polyol synthesizing plants, an understanding of the depen-dence of productivity on partitioning of photosynthate requiresstudy of the pathway, site, and regulation of synthesis of polyols.We approached questions on the pathway and site of polyolsynthesis in celery leaves by using whole leaf extracts and enzym-ically isolated mesophyll protoplasts. Probable key enzymes lead-ing to mannitol synthesis were detected in celery leaf extractswith mannose-6-P reductase activity being specific for celery andnot detectable in extracts from the nonpolyol producing plants(Table I). In addition, the activities of mannose-6-P isomeraseand mannitol- I-P phosphatase were much greater in celery leavesthan either rye or V. mungo. Considering that the highest pho-tosynthetic rate we measured for celery leaves was about 160,gmol mg-' Chl h-' (Table I), the activity of the enzymes neededfor mannitol synthesis are adequate. As noted earlier, if abouthalf of the photosynthetic carbon goes into mannitol and halfinto sucrose, then about 80zmol CO2mg ' Chl h-' would befixed into mannitol. Mannitol, being a six-carbon compound,would then be made at a rate of about 13 pmol mg-' Chl h-'.The lowest activity we measured in celery leaf extracts for anyof the enzymes involved in mannitol biosynthesis was 31,umolmg-' Chl h-' for the phosphatase (Table I). If, as we propose,the reductant for mannose-6-P reductase activity is provided tothe cytoplasm by nonreversible G3P dehydrogenase, then foreach mannitol synthesized three triose-P would need to be ex-ported (two metabolized to mannitol and one to provide thereductant). This would result in a higher level of transport onthe Pi translocator than in sucrose producing plants.

Since mannitol is an early labeled product of '4CO2 fixationin celery leaves, involving the intermediates mannose-6-P andmannitol-l-P (Loescher, Redgwell, and Bieleski, unpublisheddata), and based on the activity and localization of enzymes formannitol synthesis as reported here, a pathway of photosyntheticcarbon flow from triose-P to mannitol can be proposed (Fig. 2).Triose-P is proposed to be exported from the chloroplast andmetabolized to F6P as in plants which synthesize sucrose as themain soluble photosynthetic product. F6P is then converted tomannose-6-P by an isomerase reaction. The mannose-6-P is thenreduced to mannitol-l-P with NADPH-dependent mannose-6-Preductase and dephosphorylated to mannitol with mannitol-l-Pphosphatase, all occurring in the cytoplasm. The possibility existsthat mannose-6-P could be formed by isomerization of F6P toG6P and then epimerization of G6P to mannose-6-P (see reac-tions 5 and 6, Fig. 2). We found the activity of the epimerase tobe very low compared to that of the mannose-6-P isomerase(data not shown). However, until enzymes are purified andcharacterized, the alternate route (reactions 5 and 6) cannot becompletely ruled out.

This is the first report of a mannitol-l-P phosphatase in ahigher plant species. The phosphatase has been purified andcharacterized from brown algae (1 1) and fungi (24). The onlyother polyol phosphatase reported in higher plants is a sorbitol-6-P phosphatase in apple leaves (9). We propose that the NADPHrequired for mannose-6-P reductase in the cytoplasm may beshuttled out of the chloroplast in the light and linked to thecytoplasmic, nonreversible G3P dehydrogenase (Fig. 2). Kellyand Gibbs ( 12) have demonstrated the cytopslamic localizationof this enzyme in photosynthetic tissue and suggested it may be

872 RUMPHO ET AL.

www.plantphysiol.orgon July 26, 2020 - Published by Downloaded from Copyright © 1983 American Society of Plant Biologists. All rights reserved.

PATHWAY FOR MANNITOL SYNTHESIS IN CELERY LEAVES

Mannitol

Pir @Cytoplasm

Mannitol- 1-P

NADPH NADP (

Mannose-6-P

DHAP

HAF-6-P -- G-6-P

Y AD DP DHAP DHA PPi

ChloroplastFIG. 2. A proposed pathway for photosynthetic carbon flow to man-

nitol in celery leaves. DHAP, produced as a result of photosynthetic CO2fixation, is exported from the chloroplast on the Pi translocator. TheDHAP is metabolized to mannitol in the cytoplasm (reactions 2-4), orused to indirectly shuttle NADPH from the chloroplast to the cytoplasm(reactions 1). Broken arrows indicate a possible alternate route of carbonflow. Circled numbers refer to the enzymes, catalyzing the steps indicated,as follows: (1), nonreversible G3P dehydrogenase; (2), mannose-6-Pisomerase; (3), mannose-6-P reductase; (4), mannitol-l-P phosphatase;(5), G6P isomerase; and (6), mannose-6-P epimerase. DHAP, dihydrox-yacetone phosphate; RuBP, ribulose bisphosphate.

linked to a tnose-P/PGA shuttle allowing the indirect transfer ofNADPH from the chloroplast to the cytoplasm. Substantiallyhigher activity of this enzyme was detected in celery leavesrelative to V. mungo and rye (Table I). Activity of this enzymein celery was also greater than the values reported by Kelly andGibbs (12) for green and nongreen tissues of a number of species.The pathway of mannitol biosynthesis in celery is similar to

sorbitol synthesis in higher plants in requiring NADPH as reduc-tant. Aldose-6-P reductase (sorbitol-6-P dehydrogenase) catalyzesthe conversion ofG6P to sorbitol-6-P using NADPH as cofactorin loquat fruits (10) and mature leaves of several Rosaceae species(19). The only reported work on the intracellular localization ofsorbitol or any other polyol synthesizing enzymes was by Yamaki(29) using protoplasts from apple cotyledons. He concludedsorbitol-6-P dehydrogenase was associated with the thylakoidmembranes. He also suggested the enzyme may be partly locatedin the cytoplasm. However, incomplete separation of organellesand lack of a cytoplasmic marker enzyme hinders the interpre-tation of these data.

Distribution of enzymes for sorbitol metabolism in photosyn-thetic versus nonphotosynthetic tissues ofapple suggests synthesisis restricted to source tissue using the enzymes aldose-6-P reduc-tase (16) and sorbitol-6-P phosphatase (9). Sink tissues, mean-while, contain sorbitol dehydrogenase, which is responsible forsorbitol oxidation and subsequent utilization (16). In contrast tothe polyol synthesizing enzymes which use NADPH, the degra-dative enzymes use NAD (16).The majority of higher plant polyol producers also synthesize

and translocate sucrose to some degree (4, 15). The regulation ofpolyol versus sucrose synthesis in these species, at the enzyme

level and by environmental factors, needs to be investigated.

Acknowledgments-We would like to thank A. Lansing, V. R. Franceschi, and

873J. Kobza for providing some of the plant material; B. Moore for assistance withenzyme assays on the sucrose gradient; and M. S. B. Ku and R. A. Kennedy forhelping with the photosynthetic measurements.

LITERATURE CITED

1. ADLER L, A PEDERSEN, I TUNBLAD-JOHANSSON 1982 Polyol accumulation bytwo filamentous fungi grown at different concentrations of NaCI. PhysiolPlant 56: 139-142

2. AHMAD T, F LARHER, GR STEWART 1979 Sorbitol, a compatible osmotic solutein Plantago maritima. New Phytol 82: 671-678

3. ARNON DI 1949 Copper enzymes in isolated chloroplasts. Polyphenoloxidasein Beta vulgaris. Plant Physiol 24: 1-15

4. BIELESKI RL 1982 Sugar alcohols. In F A Loewus, W Tanner, eds, PlantCarbohydrates I. Intracellular Carbohydrates, Encyclopedia of Plant Physi-ology, Vol 13A, New Series. Springer-Verlag, New York, pp 158-192

5. BIELESKI RL, Rl REDGWELL 1980 Sorbitol metabolism in nectaries fromflowers of Rosaceae. Aust J Plant Physiol 7: 15-26

6. BIRD IF, MJ CORNELIUS, AJ KEYs, CP WHITTINGHAM 1974 Intracellular siteof sucrose synthesis in leaves. Phytochemistry 13: 59-64

7. CAREY EE, DB DICKINSON, LY WEI, AM RHODES 1982 Occurrence of sorbitolin maize. Phytochemistry 21: 1909-1911

8. CHONG C 1971 Study of the seasonal and daily distribution of sorbitol andrelated carbohydrates within apple seedlings by analysis of selected tissuesand organs. Can J Plant Si 51: 519-525

9. GRANT CR, T AP REES 1981 Sorbitol metabolism by apple seedlings. Phyto-chemistry 20: 1505-151 1

10. HiRAi M 1979 Sorbitol-6-phosphate dehydrogenase from loquat fruit. PlantPhysiol 63: 715-717

1 1. IKAWA T, T WATANABE, K NLsizAwA 1972 Enzymes involved in the last stepsof the biosynthesis of mannitol in brown algae. Plant Cell Physiol 13: 1017-1029

12. KELLY GJ, M GiBBs 1973 A mechanism for the indirect transfer of photosyn-thetically reduced nicotinamide adenine dinucleotide phosphate from chlo-roplasts to the cytoplasm. Plant Physiol 52: 674-676

13. KENNEDY RA, JL EASTBURN, KG JENSEN 1980 C3-C4 photosynthesis in thegenus Mollugo: structure, physiology and evolution of intermediate charac-teristics.Am J Bot 67: 1207-1217

14. LEIGH RA, RR WALKER 1980 A method for preventing sorbitol interferencewith the determination of inorganic phosphate. Anal Biochem 106: 279-284

15. LEWIS DH, DC SMrrH 1967 Sugar alcohols (polyols) in fungi and green plants.I. Distribution, physiology, and metabolism. New Phytol 66: 143-184

16. LOESCHER WH, GC MARLOW, RA KENNEDY 1982 Sorbitol metabolism andsink-source interconversions in developing apple leaves. Plant Physiol 70:335-339

17. LOESCHER WH, R REDGWELL, R BIELESKI 1982 Mannitol biosynthesis inhigher plants: detection and characterization of a NADPH-dependent man-nose 6-phosphate reductase in mannitol-synthesizing plants. Plant Physiol69: S-51

18. MIFLIN BJ, H BEEVERS 1974 Isolation of intact plastids from a range of planttissues. Plant Physiol 53: 870-874

19. NEGM FB, WH LOESCHER 1981 Characterization of aldose-6-phosphate reduc-tase (alditol 6-phosphate: NADP l-oxidoreductase) from apple leaves. PlantPhysiol 67: 139-142

20. PERCIVAL F, RH McDoWELL 1967 Chemistry and Enzymology of MarineAlgal Polysaccharides. Academic Press, New York

21. RAESE JT, MW WILLIAMS, HD BILLINGSLEY 1977 Sorbitol and other carbo-hydrates in dormant apple shoots as influenced by controlled temperatures.Cryobiology 14: 373-378

22. ROBINSON SP, DA WALKER 1979 The site of sucrose synthesis in isolated leafprotoplasts. FEBS Lett 107: 295-299

23. USUDA H, GE EDWARDS 1980 Localization of glycerate kinase and someenzymes for sucrose synthesis in C3 and C4 plants. Plant Physiol 65: 1017-1022

24. WANG S-Y C, D LE TOURNEAU 1972 Mannitol biosynthesis in Sclerotiniascierotiorum. Arch Mikrobiol 81: 91-99

25. WEBB KL, JWA BURLEY 1962 Sorbitol translocation in apple. Science 137:766

26. WHETrER JM, CD TAPER 1963 Note on seasonal occurrence of sorbitol (D-glucitol) in buds and leaves of Malus. Can J Bot 41: 175-177

27. WINTER K, JG FOSTER, GE EDWARDS, JAM HOLTUM 1982 Intracellularlocalization of enzymes of carbon metabolism in Mesembryanthemum crys-tallinum exhibiting C3 photosynthetic characteristics or performing Crassu-lacean acid metabolism. Plant Physiol 69: 300-307

28. WINTERMANS JFGM, A DE MOTs 1965 Spectrophotometric characteristics ofchlorophyll and their pheophytins in ethanol. Biochim Biophys Acta 109:448-453

29. YAMAKI S 1981 Subcellular localization ofsorbitol-6-phosphatedehydrogenasein protoplast from apple cotyledon. Plant Cell Physiol 22: 359-367

30. ZIMMERMAN MH, H ZEIGLER 1975 List of sugars and sugar alcohols in sieve-tube exudates. In MH Zimmerman, JA Milburn, eds, Transport in Plants I.Phloem Transport, Encyclopedia of Plant Physiology, Vol 1, New Series.Springer-Verlag, New York, pp 480-503

www.plantphysiol.orgon July 26, 2020 - Published by Downloaded from Copyright © 1983 American Society of Plant Biologists. All rights reserved.