Plant Physiol. ( 1960) 44, 242-250

Glycolate and Glyoxylate Metabolism by IsolatedPeroxisomes or Chloroplasts'

T. Kisaki2 and N. E. TolbertDepartment of Biochemistry, Michigan State University, East Lansing, Michigan 48823

Receiv,ed October 14, 1968.

Abstract. Chloroplasts, mitochondria, and peroxisomes from leaves were separated byisopycnic sucrcse density gradient centrifugation. The peroxisomes converted glycolate-14Cor glyoxylate-14C to glycine, and contained a glutamate: glyoxylate aminotransferase as indi-cated by an investigation of substrate specificity. The pH optimum for the aminotransferasewas between 7.0 and 7.5, and the Km for L-glutamate was 3.6 mN' and for glyoxylate, 4.4 mM.The reactioni of glutamate plus glyoxylate was not reversible. The isolated peroxisomes didnot convert glycine to glyoxylate nor glycine to serine.

Peroxisomes did not oxidize gly-colate or glyoxylate to CO.. Chloroplasts could veryslowly oxidize glyoxylate, but not glycolate, to CO,,. Chloroplast oxidation of glyoxylate washeat labile and widely distributed among plants. Oxidation was stimulated by light and oxygen.but was not inhibited by 3-(3,4-dichlorophenyl)-1,1-dimethylurea (DCMU).A scheme for the distribution of enzymes associated with glycolate metabolism and photo-

respiration is presented. Glycolate biosynthesis occurs in the chloroplasts. In the peroxisomes,glycolate is oxidized with 0 uptake to glyoxylate by glycolate oxidase, and the glyoxylate isconverted to glycine by glutamate:glyoxylate aminotransferase. Further metabolism of glycinedoes not occur in the peroxisomes. It is possible that excess glyoxylate from the peroxisomescould return to the chloroplasts to be reduced to glycolate or oxidized to account for part ofthe CO. loss during photorespiration.

Leaf peroxisomes have beeni isolatedl bvw isopycnicsucrose density gradient centrifugation acnd showni tocontain glycolate oxidase, glyoxylate reductase. cata-lase, and malate dehvdrogenase (19, 20). Thus, theperoxisome should be associated with photoresp)ira-tion since this process is thought to be the result ofthe rapid metabolismni of glycolate as it is formlledduring photosynthesis (8, 14, 28). Glycolate oxidasein the l)articles woould account for the oxygen uptake.The carboxvl roup of glycolate has been implicatedas the souirce of CO., evolved during photorespiration(29), but wvhere and how this decarboxvlation occuirs

in the cell has nlot been determined. Consequently.the metabolic fate of glycolate-1-4C and glvoxVlate-1,2-14C wlheln adde(d to isolate p)eroxisomnes and chloro-plasts has been investigated. An objective of thispaper is to describe another enzyme. glutamnate-glyoxylate aminlotransferase. wvhich is to be fotunid inleaf peroxisomes. Whereas peroxisomles conlvertedglyoxvlate to glvcine. the chlol-oplasts slowly oxidlizedglyvxvlate to CO.

1 Supported in part by NSF Grant GB 4154 andl pub-lished as Journal article No. 4525 of the Michigan Agri-cultural Experiment Station. Parts of this work were

abstracted in Plant Physiol. 43: S12 (1968).2 On leave from the Central Research Institute, Japan

Monopoly Corporation, 1-28-3, Nishishinagawa, Shina-gawa-ku, Tokyo, Japan.

242

Materials and Methods

Planit lIatterial. The leaves tise(I in these experi-menlts wN-ere similar to those employed durinig a surveyof plants for peroxisomes (20). In most cases, theperoxisomlle preparations were nmade from spinachleaves which wA-ere obtained from local grocery stores.Swiss chard, tobacco, bean. sunflower. pigweed, andcorn were groxvn in the greenhouise. Fresh sugar-cane leaves were shipped in ice by air express fromFlorida.

Labeled Sz'bstratcs. For experimentet iinvolviin1g'4CO, release, the substrates wvere useld without car-rier. In the trainsamninase assay, the labeled sub-strates were diluted 1000-fold witlh unlabeled carrierso that the specific radioactivity of the labeled sulb-strate ; were the following: 2.45 >Y 10 cpnm/niolefor glycolate-1-14C, 1.45 X10_ cpni/pmole for gly-oxvlate-1,2-14C. 1.54 X 101 cpm/tmiole fol- glycine-2-'IC. 3.51 X 10n cp)m/mole for DL-glutamnate-2-14C,3.78 X 10-2 cpm/pmole for aspartate-3-1 4C, and6.6.3 X 1(); cpnm/imole for wt_-seriine- 1 14C.

PreParation, of Particles. These procedures havebeenldetailed elsewhere (19, 20). The essential stepswere as follow A) homogenizatio) of the leavesin 0.5 Nr sucrose containiing 0.025 M\ glycylglyciniebuffer at pH /7.5 ; B) differential cenitrifugation into4 fractions designated as cell debris with some wlholechloroi)lasts. broken chloroplast.s\with tle peroxi-

www.plantphysiol.orgon August 6, 2019 - Published by Downloaded from Copyright © 1969 American Society of Plant Biologists. All rights reserved.

KISAKI AND TOLBERT-GLYCOLATE METABOLISM BY PEROXISOMIES

somes, mitochoindria, and supernate; C) isopycnicnonlinear sucrose density gradient centrifugation ofthe broken chloroplast plus peroxisome fraction. Allthe data were obtained with fractions from thegradient centrifugation which were drained from thebottom of the centrifuge tube and number 1 to 9 asindicated in figure 1. The corresponding sucrose

molarity is also shown in this figure. In one case

(fig 2), the chloroplast band (fraction 7) was pel-leted by recentrifugation at 34,000g for 20 min,resuspended in grinding medium, and then su:bjectedagain to isopycnic centrifugation on a linear gradientbetween 0.9 and 1.7 M sucrose.

Enzyme Assays. Glycolate oxidase (EC 1.1.3.1)was measured by the rate of dichlorophenolindo-phenol reduction as indicated by a decrease in opticaldensity at 600 nM (19, 20), Cytochrome c oxidase(EC 1.9.3.1) was measured by the decrease in opticaldensity at 550 nM (19). Protein was measuredaccording to the Lowry procedure and chlorophyllby Arnon's procedure.

Decarboxylation of labeled substrates at 250 was

measured in Warburg flasks by trapping the 14CO.0with a solution of KOH on filter paper wicks in thecenter well. At the end of the assay, 4 N H2SO4was added to stop the reaction an.d liberate boundCO., Light experiments were run in a Gilson photo-synthetic respirometer which provided 750 ft-c of

light from photoflood bulbs to the bottom of theflasks.

Glutamate :glyoxylate aminotransferase (EC 2 6.1.4) was assayed by following the formation ofglycine-14C. In a final volume of 1.25 ml were

20 jumoles glyoxylate-1,2-' 4C. 25 umoles aminodonors, 0.1 ,umole pyridoxal phosphate, the enzyme.

and 0.02 M phosphate buffer i(pH 7.5). Reactionswere carried out at 30° for 15 min and terminatedby boiling. Unreacted glyoxylate-14C was removedby passage of the reaction mixture down a Dowex-1acetate coltumn (6 X 50 mm) which was then washedwith 2 ml of water. The comlbined effluents werecounted for glycine-14C. It was determined experi-mentally that all the 14C in the effluent from theDowex-1 acetate column was absorbed by Dowex-50H+ column and eluted with ammoniac. On thin layersilica gel chromatograms, all the 14C-eluant cochro-matographed with carrier glycine. For routineanalysis, only- the "1C from the Dowex-1 acetatecolumns were counted. When glutamate-2-_4C was

ulsed as the aminoi donor, the reaction mixture was

stubjected to Dovex-50 H-. and the fraction in whicha-ketoglutarate was eluted was counted for 14C.

Aspartate-a-ketoglutarate amninotransferase (EC2.6.1.1) was also assayed wvith labeled substrates.The reactioni mixture contained aspartate-3-14C. a-

ketoglutarate, pyridoxal phosplhate, enzyme, andphosphate buffer. '4C in the oxalacetate productwas counted after separation from aspartate on a

Dowex-50 H- coltumniii.Serine lhv droxy-methy-ltransfer alse (EC 2.1 2.1)

was assaved chroniatographically by the conversion

of serine-1-14C to glycine-14C (25). Assays wererun in phosphate buffer at pH 7.0 with pyridoxalphosphate, tetrahydrofolic acid, MgCl2, and TritonX-100 for as long as 1 hr to establish the absence of

E ,v 0.3-x ,

Co_ >i

E 0.2-

_ :00 01-

E=(

V

0

x

0

4DE20

D0

0

60

40

20

4-

2-

0.4

003

E 0.2

E0

:..... _--

A. Chlorophyll

_ _ __ _ I t

I..............

r-- n FmCyt c Glycolate

Oxidase Oxidase

I1 11

f191 8 7 6 151 4 1 31 2 THW

Fraction Number

-0.3 E

E-0.2 -

0

0.1

1.3 1.5 18 2.3 12.51Sucrose Molority

B.Glyoxylate -1,2 -14r

Glycolate ~I-1-1

r - -

Glutomate-Glyoxy lateAminotransferose

li-

I

7 1 6 151 4 1 312 111Fraction Number

111.3 1.5 18 1 2.3 12.5Sucrose Molarity

FIG. 1. Distribution of particles and enzymatic ac-tivities in fractions after sucrose density gradient cen-trifugation. Horizontally is indicated the gradient com-position in sucrose molarity and relative volume foreach concentration. The fractions are numbered as re-moved from the bottom of the centrifuge tube. A par-

ticulate fraction from spinach leaves was used whichhad first been fractionated between 1000 to 10,000g.Part A) Peroxisomes were mainly in fraction 3 as

indicated by glycolate oxidase activity. mitochondria andsome whole chloroplasts were in fraction 5 as indicatedby cytochrome c oxidase and chlorophyll, and brokenchloroplasts were in fraction 7 as indicated by chloro-phyll. Part B) Distribution of activity for decarboxy-lation of glycolate-1-14C anid glyoxylate-1 2-14C in thefractions after sucrose density gradient centrifugation.Part C) Aminotransferase activities.

243

www.plantphysiol.orgon August 6, 2019 - Published by Downloaded from Copyright © 1969 American Society of Plant Biologists. All rights reserved.

4PLANT PHYSIOLOGY

activity in the peroxisomes. After stol)p)ing the re-action by heat, the amino acids were separated oniDowex-50 (H+), as in the glutanmate :glyoxvlateaminotransferase reaction, and subjected to paperchromatographic analysis.

Results

]4CO., Evolution by Particles. Fromi isopycnicnonlinear stucrose density gradient centrifugation ofthe particulate fractioni from spinach leaves, separatebands were detected for peroxisomes in fraction 3as indicated by glycolate oxidase, for mitochondriain fraction 5 by cytochrome c oxidase, and for chloro-plasts in fractions 7 and 8 by chlorophyll (fig 1.part A). To these fractions from the gradient wereadded radioactive sulbstrates and 14CO., evolutionmonitored (fig 1, part B). [n the peroxisome frac-tion numnber 3 with active glycolate oxidase, littleconversion of glvcolate- 1-14C or glyoxvlate- 1 ,2-14Cto 14CO., occurred. Some 14CO2 was formed fromglycolate-1-'4C by fraction 5 which contained somewhole chloroplasts and. peroxisomes as well as mito-chondria. GIvoxvlate- 1,2-14C wN-as decarboxylated bychloroplasts in fractions 7 and 8. Significantamounts of 14CO0, vere not formed from oxalate-14Cor acetate-14C by any of these particles. The evolu-tion of 14CO. from glycolate-] 4C or glyoxylate-14C by the particles was heat labile (table I). Thedecarboxylatioln of glvcolate was enhanced by therecomtbination of the peroxisome and chloroplastfractions, presumably becatuse the peroxisomes oxi-dized glvcolate to glyoxylate and the chloroplastsoxidized glyoxylate to CO_. The decarboxylation ofexcess glyoxvlate by the chloroplasts was not en-hanced by the addition of peroxisomes. Also, theaddition of unlabeled glycolate Nvithi glv.oxylate-1.2-14C did not inhibit the decarhoxyla tion of glyoxylateby chloroplasts.

The enzynmatic system in the chloroplast whichcatalyzed glAoxylate oxidation was bound tightlyenouglh to the broken chloroplasts that the plarticlescould be recentrifuged without great loss of activity.The particles in fraction 7 after the first nonlinearsucrose gradient centrifugation as in figutre 1 wveresedimeinted anid then suibjecte(d to a seconid linear

- 0.2

E

3.r00

i~0.

r6)-2

x

E

CL

0cm

0

0

Volume (ml)

i11 1101 9181 7161 51 41 3 121Fraction Number

0.9 1.1 1.3 1 1.4 1 1.5 11.7Sucrose Molarity

FIG. 2. Distribution of activity for decarboxylationof glyoxylate-1,2-14C by broken chloroplast particlesafter further purification by isopycnic recentrifugationon a linear sucrose gradient.

sucrose gradienit centrifugationi as described inMIethods. Most of the glyoxylate-14C oxidationcapacity remained with the chlorophyll-containingparticles (fig 2).

The optimum pH for the decarboxylation ofglyoxylate by the chloroplast fraction 7 was about10 i(fig 3). Although this is a very alkaline pHoptimum, there was some decrease in activity atpH 11. The rate of oxidation was linear for atleast 30 min (data not shown).

The oxidation of glyoxylate by the chloroplastswas stimuilated by air and by light (table II). Alower level of oxidation in a nitrogen atmosphere wasnot stimulated by light. Aerobic oxidation, whichwas at least 250 % of the anaerobic rate, was stimui-late(d 250 % further by 750 ft-c of light. The natureof this oxidation of glyoxylate is not known. Glv-oxvlate ancd H.,O, non-enzymatically react rapidlv toprodluice CO. and HCOOH fromii the carboxyl anda-carbon, respectively. H1O10 production by theclhloroplasts in a Mehler-type reaction has not beenexcluided. H.1O. produiction by glycolate oxidase was

Table I. Decarbo.xylation. Y Fractions Froiii Density Gradicnt Ccotrifugatio)nReaction mixture contained 295,000 cpm of glycolate-1-14C or 28,900 cpmn of glyoxylate-1,2-14C, 1 ml of 1 -Ni

sucrose containing 0.1 M pyrophosphate buffer (pH 8.5), and 200 Al of each sucrose density gradient fraction fromSwiss chard leaves. Reaction time was 30 min.

Substrates

Glycolate- 1-14C Glyoxylate-1,2-14C Glycolate oxidase

3 (Peroxisotmie)7 (Chloroplasts)3 Plus 77 Plus boiledl 3Boiled 7

3,84713,88128,726

0

0

cpm CO., evolved5,610

11,66511,43711,500

0

nin0ole X min-' X nib-320

6326

0

0

Fraction No.

244

www.plantphysiol.orgon August 6, 2019 - Published by Downloaded from Copyright © 1969 American Society of Plant Biologists. All rights reserved.

KISAKI AND TOLBERT-GLYCOLATE 'METABOLISM BY PEROXISOMES

31

21

FIG. 31,2-14C bsity grad5.78 X:pared in0.2 ml oftionation

not likelfractionther, theinhibitedmethaneoxidase,upon thplasts.ruled ou

a concen9 6A- lbl,

phosphate - pyrophosphate

.-0

0

0

j.

Table III. Inihibitiont of Glyoxylate Decarboxylationby Keto Acids

Reaction solution contained 57,800 cpm of glyoxylate-1,2-14C, 1 ml of 1 M sucrose solution containing 0.1 Mpyrophosphate buffer (pH 8.5), the added compound ata final concentration of 3 X 10-3 M, and 200 ,AI of thesucrose density gradient fraction 7 from Swiss chard.Reaction time was 30 min.

Added compound CO, evolved

NonePyruvatea-KetoglutarateOxalateGlycine

cpn't

973295493784829

Glyoxylate oxidation was inhibited by other ketoacids, specifically, pyruvate and a-ketoglutarate(table III). This data is in contrast to a synergistic

I i 9 I I effect of pyruvate and a-ketoglutarate on the de-7 8 9 11 carboxylation of glyoxylate by pyruvate decarboxy-

pH lase (6). Oxalate and glycine, products formedfrom glyoxylate, did not significantly inhibit gly-

Effect of pH on decarboxylation of glyoxylate- ylate decar

)y broken chloroplast fraction from sucrose den- OXate bt0 oxy'lation. Also, -hloromer tibnlient centrifugation. Reaction mixture contained zoate 0.1 mM did not decarboxylation

104 Cpm of substrate, 1 ml of 1 M sucrose pre- by thechloroplastsi KCN at 0.25 mM inhibited the

0.1 M pyrophosphate or phosphate buffer, and reaction 60 %.

fraction 7 from a sucrose density gradient frac- Distribution. of Glyoxylate Decarboxylation Ac-of particles from sunflower leaves. tivity. Since the oxidation of glyoxvlate to CO2.

may reflect the activity of photorespiration (29), it

~y since tha-t enzyme was in the peroxisomalwas of interest to survey chloroplast fractions, as.

rather than the chloroplast fraction. Fur- prepared by sucrose density gradient centrifugation,ratherathan the chloroplastperactisomesFs from a number of plant species. The specific activityoxidation of glyc-olate by peroxisomes was

o hoohi ai aidwdl tbeT~1 84 % by 2.5 mM a-hydroxy-2-pyridine- corophyll basis varied widely (table IV)

sulfon,ate, because it inhibited glvcolatebut the sulfonate had no inhibitory effect Table IV. Decarboxylation of Glyoxylate by Chioroplast

e decarboxylation of glvoxylate by chloro- Fractions From Various PlantsPeroxide production by photosystenm II was Reaction solution contained 57,800 cpm of glyoxylate-rtby DCMU experiments. DCMU at 4p v I

1,2-14C, 1 ml of 1 M sucrose solution containing 0.1 M

itration which inhibited a Hill reaction with pyrophosphate buffer (pH 8.5), and 200 pl of fractionnumber 7 from the sucrose density gradient.nrnrn'hfnn1;1nrlnn1iPnn1 nQchvtIrnoen qrrnPrtnr_did not inhibit glyoxylate oxidation by the chloroplastfraction. A non-specific decarboxylation of a-ketoacids with H202 generated in a photochemicalprocess cannot account for all the results since thereaction also occurred in the dark.

Table II. Decarboxylationt of Glyoxylate byChloroplasts

Experimental conditions similar to those listed withtable I. All results are with fraction 7 (chloroplasts)from spinach leaves.

Atmosphere CO, evolved

cprnLight Air 3050Dark Air 1320Light N, 450Dark N.2 500

Plant CO2 evolved

cPmo X 103 X iiic1r chlorophyllSpinach 14Swiss chard 28Tobacco 3Sunflower 137Bean 268Pigweed 36Corn (young) 3" (mature) trace

Sugar cane (young) 4IY "py ((mature) trace

Sunflower, which exhibits photorespiration and whichgave excellent yields of peroxisomes (20) had chloro-plasts capable of rapid glyoxylate oxidation. Spinachchloroplasts were less active. Highest activity was

found with bean leaves, which photorespire but from

5

1(0x

E0

0)

'p

0w

00

245

4t

I

www.plantphysiol.orgon August 6, 2019 - Published by Downloaded from Copyright © 1969 American Society of Plant Biologists. All rights reserved.

26PLANT PHY SIOLOGY

wvhich a very poor yield of peroxisomiies were obtained('20). Decarhoxylation of glyoxylate by chloroplastsfrom corn or sugarcalne leaves, wh\lich are plantsxvithotit photorespiration, was low. This result isconsistenit with the concept thlat these plants shouldhave a low rate of glyoxyvlate decarboxylation l)ychloropla-lsits becauise of the absence of phlotorespira-tion. However, chloroplasts fromii tobacco leaves,wvhich have been tised for nhotorespl)iltion studies,were no more active for glyoxylate oxidation thanchloroplasts fronm corn leaves. It is possible that thetobacco chloroplasts wvere inactivated during the iso-lation procedure as the tobacco sal) was darklycolored fronm rapid oxidati(jn. Chloroplasts frompigweed, whiclh has no p-hotorespiration, were quiitecapable of oxidizing glyoxylate to CO_. Thus, noexact correlation was appa rent between the presenceof photorespiratioin and -lyoxylate oxidation to CO.,by isolated chloroplasts.

Gittamatc :glyoxylate Amintiotra iisfcrase in LeafPeroxisomes. In the peroxisonmes. glyoxylate oxida-tion to CO., was slow and linmited (fig 1) Addedglyoxylate-14C wvas nmainily converted to glycine-14C.As described in the Methods section, glycine-'4C wasisolated by chromatography. and kinetic studies ofits formation were run. The only radioactive aminoacid which was formed was glvcine; no serine wasdetected. Addition of tetrahvdrofolic acid aind pyri-doxal phosphate didi not alter the results. Someglycine was formed by isolated peroxisomes in theabsence of added aamino donors, but additioln of anamino donor such as glutatmate greatly increased theyield of glycine-14C.

Transarminase reactionis were measured in theparticulate fractions froml spinach after separationby sucrose gradient cenitrifugatioin. Glutamate :gl-oxylate aminotransferase activity was located in theperoxisomes along witlh glvcoh-te oxidase (fig 1,part C). An aspartate-a-ketoglutarate aimiinotrans-ferase was found in the miitochondria fraction. Someof this activity in the peroxisome fraction as wellprobably reflects the lack of substrate specificity forthese reactions of isolated enzymnes. The presence ofan active and specific transaminase in the peroxi-somes for converting glvoxvlate to glycine is con-sistent with in vi-'o studies on glycolate conversionto glycine (reviewed in ref 18).

It was not possible to (letermine accurately thedistribution of glutamate :glyoxylate aminotransferasebetween the peroxisomes and cytoplasmll because ofthe fragility of the peroxisoimies (19. 20) and the lackof specificity of the several aminotransferases in thecytoplasm A particulate fraction (peroxisomes plusbroken chloroplasts from differeential centrifugation)which contain-ed 28 % of the total glycolate oxidaseactivity also conitained 20 % of the total g1lutamiate:glyoxylate aminotransferase activity. Tt lhas beenproposed that all tlle glvcolate oxidase would befountd in the peroxisolmies if they could he isolatedintact (19,20). Likewise, it is probable that theg-lutam-ate :glvoxvlate aminotransferase hlas tlle s;allmespecific site, the l)eroxisomes, in the cells of learves.

Table V. Specificity of Amino Grolup Doitor for GlycineFormiation bv Peroxisomies

Fraction

i.-Glutamiiaten-GlutamateL-AlanineD-AlaniiueDL-SerineDL-GlutamineDL-a-AminobutyrateDL-AspartateDL-AsparagineL-Argiiiinei,-OrnithineDL-Lysine$-AlanineGlycinei)L-Valinei.-LeucineDL-PhenylalanineI)L-TryptophaneAmmonium sulfate

Crude extract Peroxisome

(relative rate)I1000225094572

3861352433814818854151

1105

15321

glYcinc foriiied1000

1000

5501891981602231881431319219813012813510384

Relative rate of glycine formation with L-glutamatewas taken as 1000. Reaction mixture contained 0.1,gmole of pyridoxal phosphate, 221,000 cpm of gly-oxylate, amino donor as indicated at a final concen-tration of 0.025 IA except phenylalanine and trypto-phane at 0.005 a\i, 1 ml of 0.01 M phosphate buffer(pH 7.3), 40, of peroxisome fraction from spinach,and 0.1 ml of Triton X-100 (0.5 %). Reaction periodwas 15 min at 300.

Substrate Specificity of 'eroxisomte A minotrans-ferase. The best naturally occurring amino groupdonor was L-glutamate, although L-alanine was nearlyas good (table V). Rates with glutamate and alaninetogether were similar to .-ates with only glutamate.DL-Serine was about 39 % as active as L-glutamate,and other amino acids were less active. DL-Gluta-mine, DL-asparta te, and L-ornithine, previously usedas amino donors in other test systems, were not asactive for the leaf peroxisomal aminotransferase withglyoxvlate. The unnatural n-glutamate was 2225 %more active than T-glutamate, hut D-alanine wasnearlv inactive. No reason is known for D-glutamatespecificity; likewise, liver peroxisomes containD-amino acid oxidase for unknown reasons.

Glvoxvlate was the stuperior amino group acceptorfor the leaf peroxisome transaminase (table VI).Although the enzyme was also active with pyruvate.activity w%vith the 2 substrates, gl oxylate and pyru-vate, was not additive. Consequently, only 1 trans-aminase is indicated. The availability of largeamounts of glycolate fromii photosynthesis and thepresence of an active gly-colate oxidase for glyo-xviateformation in the peroxisomiies strongly suggest thatthe peroxisome transaminase functions primarily forthe formation of glycine via the glycolate pathway\.An equally rapid translocation or mlovemenlt of gltl-tamate and a-ketoglutarate in alnd out of the peroxi-somes wvould be demanded b\- this sclhemle.

246

www.plantphysiol.orgon August 6, 2019 - Published by Downloaded from Copyright © 1969 American Society of Plant Biologists. All rights reserved.

KISAKI AND TOLBERT-GLYCOLATE METABOLIS'M BY PEROXISOMIES

Table Vi. Specificity of Aiiiino Groutp Acceptor FrontL-Glutaniiate by! Tranisaminhase i1 PeroxisoiniesReaction solutioin contained 193,500 cpm of glutamate-

2-14C, 0.1 ml of Triton X-100 (0.5 %), 0.1 ,umole ofpyridoxal phosphate, 20 ,ul of spinach peroxisome fraction,and 1 ml of 0.01 AI phosphate buffer (pH 7.3). Finalconcentration of each amino group acceptor was 0.01 M.Final volume was 1.3 ml. Reaction was carried out at300 for 15 min.

a-Keto acid acceptor a-Ketoglutarate formed

cp!IIGlyoxylate 2540Pyruvate 918Hydroxypyruvate 518Oxaloacetate 440Phenylpyruvate 75Glyoxylate + pyruvate 2310

pH Optimnumn and Kinietics for Peroxisonmal Glu-tamtate :giyoxylate Aininotransferase. Intact peroxi-somes in 1 M sucrose solution had a glutamate:gly-oxylate transaminase activity with a pH optimumbetween 7.0 and /7.5 (fig 4). In the supernatantfractions, glutamate :glyoxylate aminotransferase ac-tivity, which is believed to be the solubilized enzyme,also showed maximum activity above pH /7.0 but nooptimum, as also noted by Cossins et al. (4). Thedifference in the 2 pH optimum curves may reflecteither a pH effect on the permeation of substratesinto the particle or special structural conditions whichare sensitive to pH in the intact particle.

The reaction rate was proportional to the amountof peroxisome used (data not shown). As assayed.the rates decreased with time after the initial 5 to10 min (data not shown). The Kmi for L-glutamatewas 3.6 mM anid 4.4 mi for glvoxylate. Thesevalues are comparable witlh those wvhich Nakada (12)obtained with a partially plurified liver glutamate:glyoxylate aminotransferase.

Reversibility and Inihibitioni of the Aminiiotrans-ferase. The reverse reaction. the formation of glu-tamate from a-ketoglutarate with glycine as the

pHFIG. 4. Efiect of pH on glycine formation by iso-

lated spinach peroxisomes or the soluble cytoplasmicfraction.

amino donor, could not be detected experimentallywith *the peroxisome system. Similar results havebeen reported for the transanlinase reaction purifiedfrom liver (12), from Escheric/tia coli (15), or fromwheat leaves (11). Metzler et al. (13) have pre-sented thermodynamic argunments for the irreversi-bility of this reaction in noln-enzymliatic transaminasereactions. However, Cammarata and Cohen (3)have reported a transaminase reaction betweenglycine and a-ketoglutarate by lyophilized liver prep-aration,s, and Wilson et al. (24) have also observedthe reverse reactioni in soltuble extracts of tobaccoleaves. To reconcile these results, Nakada (12)suggested that another trailsaminase is involved intransferring the aminio grotup from glycine to a-keto-glutarate. The essential irreversibility of the per-oxisome transaminase indicate that the flow of carbonin the glycolate pathway is from glyoxylate toglycine, and indeed when 1'4C-labeled glycine has beenadded to whole leaves, no labeled glycolate lhas beenreported.

Isonicotinyl hydrazide at high concentraitionis hasbeen used as a transamlinase inhibitor (16, 26) andwhen used to block the glycolate pathwva) in 1Z,VO,glycine and glycolate accumulated (1, 16). Like-wvise, high concentrations (10 ntiM or mliore) of thehydrazide were required to inhibit the glyoxylate-glutamate transaminase activity of the isolated per-oxisomes (table II ). The effect of the inhibitor

Table VII. Inhibitioni of Glycinc Formzationt FromtGlycolate in Peroxisomiie by Isonicotinl HydrazideThese experiments were similar to those of table VI

except for the addition of the hydrazide.

InhibitionIsonicotinyl Assayed without Assayed withhydrazide detergent Triton X-100

1 X 10-3 81 X 10-2 522 X 10-2 62

1-67682

was less on the more intact particles in the absenceof the detergent, Triton X-100, in the assay. In thisexperiment it was necessary to use glycolate as thestarting substrate since the hydrazide and excessglyoxylate as substrate would react non,-enzymati-cally. Thus, the resulting inhibition by isonicotinylhy-drazide could result both from inhibition of theauminotransferase as well as reaction with glyoxvlateformed by the peroxisoimes.

Ghlciue-serine Jn tercoiversioii. In accordlancewith the glycolate p)athway, glycolate is oxidized toglvoxylate by peroxisomes and the glyoxylate furtherconverted to glvcine by the peroxisomal glutamate:olvoxv-late aminotranlsferase. In many experimentsstalrtinlg with 14C-labeled glycolate, glyoxvlate, org-lycine, no significant amotunts of serin-e-14C wereever formiied by the isolated peroxisoimies with supple-ments of cofactors. In addition. the serine hivdroxv-

247

www.plantphysiol.orgon August 6, 2019 - Published by Downloaded from Copyright © 1969 American Society of Plant Biologists. All rights reserved.

2'LANT PHYSIOLOGY

metlhyltranisferase-catalyzed conv-ersion of serine toglycine was measured with serine-1-14C, as describedin Methods, and no activity was found in the peroxi-some fraction after an hr of incubation. Rather,very active serine h-droxymethyltransferase activitywas found in other particles and in the soluble frac-tions. The enzy-me, serine hydroxymethvltransferase.has been extensively investigated from many tissues.and Davies has reported the presence of the enzymein plant extracts (5). Our data indicate that in theperoxisomes, glycolate is converted to glycine whichis an end product of peroxisome metabolismi.

Table VIII. Spceific Activity of Rclated Enkymnes inSpintach Particles

Enizyvne

Glycolate oxidaseGlutamate-glyoxvlate

aminotransferaseGlyoxylate decar-

boxylationNADH :Glvoxv late

reductaseMalate dehvdrogenaseCatalase

inmole X min-'Particle X mg-1 protein

Peroxisome 300 to 1000Peroxisome 5 to 16

Chloroplast 0.002 to 0.006

Peroxisome 300 to 1000

Peroxisome 2X104 to 3X104Peroxisome 5X 106 to 6X 106

Discussion

A working hypothesis is shown in figure 5 toaccounit for the distribution of the enzymes associatedwith glycolate metabolism and photorespirationamong the leaf cellular particles. Glycolate is bio-syn,thesized in the chloroplasts (2, /7. 9, 21, 23). Chlo-roplasts also contain the specific P-glycolate phos-phatase (17, 19, 27) which has been postulated to beinvolved with glvcolate excretion (10). In theperoxisome, glycolate is oxidized to glyoxylate andH,O0, the latter being completely destroyed by thepresence of excess catalase which is located in thesemicrobodies !(19, 20). This step would account foroxygen uptake in the peroxisonmes during photores-piration. No CO. evolution was observ-ed by iso-lated peroxisomes with either glycolate or glyoxvlate.but rather the glyoxylate was coinvertedl to glycineby a glutamate :glyoxylate aminotransferase of theperoxisome. If excess glyoxylate left the peroxi-sonmes and entered the chloroplasts, it would be eitherslowly oxidized to CO. or reduced to glycolate ascatalyzed by NADPH :glvoxvlate reductase. Thenmagnitude and ftun-ction of this recycling aspect of theglycolate pathway is poorly understood aind largelvconjecture. If CO2 were produced from the gl\ox\-late in the chloroplasts, it would be recorded asphotorespiration. If the glyoxylate were reduced toglycolate, a shuttle system would be completed fortransfer of reducing power out of the chloroplasts.

From 14C in vivo labeling experiments with 14CO.and glycolate-14C, it is known that most of the glvco-late and glvoxylate is converted to glvcine and thento serine (18). The glycine must be the end prodtuctof peroxisomal activity, for the next enzyme. seriniehydroxymethvltransferase, was not found in theperoxisomiies. This fact indicates that glycin,e shouldbe the carbon carrier from the peroxisomes. In theconversioin of 2 glycine molecules to 1 CO. and 1serine, most of CO, observed in photorespirationshould occur, but this site is not in the peroxisomes.

The specific activity of some of these enzvmes inthe peroxisomes as isolated are cited in table VIII.Since these peroxisomes are not pure (19, 20), higherspecific activities should be possible, but the ratiosshould not varv. The peroxisomes contain activeglycolate oxidase and a 10 to 100 times higher

specificity activity for nmalate dehydrogenase andlcatalase. The specific activity of glutamate:glyoxy-late aminotransferase is about one-sixtieth that ofthe glycolate oxidase. Nevertheless, glycine is theend product of peroxisomal metabolism, although anundetermined amount of glyoxylate may be excretedby the peroxisomes in vivo. The specific activityfor glyoxylate decarboxylation by the chloroplasts isnearly 2500-fold less than the values for peroxisomalglutamate :glyoxylate aminotransferase. Thus theobserved rates of glvoxylate decarboxylation bychloroplast was very small compared to the otherreactions in the peroxisomes. Since the chloroplastscontain much protein, these specific activities are forcomparison only to indicate that it appears thatglyoxylate decarboxylation by the chloroplasts is notpotentially as dynamic as glyoxvlate conversion toglycine by the peroxisomes. The supply of aminodonors. such as glutamate, to the peroxisomes wouldexert a marked regulatory effect upon these 2 alter-natives for glyoxylate utilization. The presence ofthe aminotransferase in the peroxisomles necessitatesan amino donor slhuttle across the peroxisome miiem-brane. as indicated in figure 5.

I PEEROXISOME

V! v _o-v

I_

cool:4

I- - ,- -i<ecglultarnite

C()(! ...

(.'!t ,1'.-,~

t-+

I

FIG. 5. Distribution of the glycolate pathlway inleaves amlonig the cellular- fI-actioins.

Using av,alue of 1 ftmole CO. fixation in vivoper mni per g tissue, when 10 to 100 % of theproducts p-ass through the glycolate pathway, 50 to500 nmoles of glycolate could be metabolized per min.The specific activity of glycolate oxidase and DPNH-glyoxylate reduictase wals reported in our accompainy-

248

I

il ()

Z_

www.plantphysiol.orgon August 6, 2019 - Published by Downloaded from Copyright © 1969 American Society of Plant Biologists. All rights reserved.

KISAKI AND TOLBERT-GLYCOLATE METABOLISMI BY PEROXISONIES

ing paper (20) to be 500 to 1500 nmoles per minper g leaf tissue for plants with photorespiration,and muclh larger amounts of catalase and malatedehydrogenase were found. Thus the in vitro assaysindicate adequate levels of these enzymes to supportphotorespiration. However the glutamate-glyoxylateaminotran.sferase activity was about one-sixtieth ofthe glycolate oxidase activity and thus equivalent toless than 10 to 100 nmoles per min per g tissue. Thelevel of activity of the aminotransferase is most likelyto be limiting in the glycolate pathway. As men-tioned above, the observed rates of glyoxvlate de-carboxylation by chloroplasts are so much less thatthe int vivo significance of CO.. release from glyoxy-late during photorespiration is not established.

Althouigh the existence of a glyoxylate-glutamateaminotransferase in plant tissue had been indicated,an isolation of this specific transaminase has notbeen reported. Separation of the peroxisomes bysucrose density gradient appears to have achievedthis isolation of glyoxylate-glutamic aminotransferasefrom among other transaminase activities of the cell.Althotugh the aminotransferase in the peroxisoilieshas not been further purified, experiments withmultiple substrates indicated that only 1 transaminasewas present in the particle. It is known that sev-eraltransaminases can form glvcine from glvoxylate byamino group transfer. The existence, how-ever. ofan aminotransferase preferentially favoring the glu-tamate :glvoxylate substrates seems to be consistenitwith a specific glycolate pathway in leaf peroxisomes.Thus, this specific aminotransferase may not occuirin other similar microbodies suichl as liver peroxi-somes or castor bean glyoxysomes. In exploratoryexperiments. glutamate :glyoxVlate aniinotransferaseactivity in fractions obtained by sucrose densitygradient centrifugation of particles from castor beancotyledons was compared with spinach leaf peroxi-somes. The leaf peroxisomes had a specific activityof 3.2 X 10" cpm X mg-1 protein for glycine forma-tion from glyoxylate-1,2-14C and L-glutamate, whereasthe fraction)s for castor bean glyoxysomes had aspecific activity of 5.9 X 103. The results suggestthe absence of aminotransferase in glyoxvsomes, asthe value of 103 were the order of mlagnitude cf theendogenotus non-enzymatic reactionl for glycine for-mation from glyoxylate.

A serine-glyoxvlate aminotransferase has been ofspecial interest since its postulation by XVang andWVaygood (22), because such a reaction would allowthe glvcolate pathway to function more independentlyof other amino donors such as glutamate. King aindlWaygood's recent report (11) on this transaminasefrom wheat leaves did not establish a substratespecificity for the serine :glyoxylate couple nor thecellular location of the enzyme. In the peroxisomes.the transaminase with glyoxylate is about 3-fold morespecific for L-glutamate than serine. However, at-tractive as a serine-glyoxylate transamination nmaybe, we have so far obtained no evidence for such aspecific enzyme in the peroxisomes. In fact. since

serine was not formed by the peroxisomes, there isno reason to thinik that a transaminase reaction withserine would selectively be involved in that part ofthe glycolate pathway which is in the peroxisomesand which involves glycine formation.

Literature Cited

1. ASADA, K., K. SAITO, S. KITOH, AND Z. KASAI.1964. Photosynthesis of glycine and serine in greenplants. Plant Cell Physiol. 6: 47-59.

2. BRADBEER, J. W. AND C. M. A. ANDERSON. 1967.Glycolate formation in chloroplast preparations.In: Biochemistry of Chloroplasts. Academic Press.p 175-79.

3. CAMMARATA, P. S. AND P. P. COHEN. 1950. Thescope of the transamination reactions in animaltissues. J. Biol. Chem. 187: 439-52.

4. Cossixs, E. A. AND S. K. SINHA. 1965. Occur-rence and properties of L-aminoacid :2-glyoxylateaminotransferase in plants. Can. J. Biochem. 43:495-506.

5. DAVIEs, D. D., J. HANFORD, AND A. P. WILKINSON.1959. The metabolism of aspartic acid, serine, andglycine in higher plants. Sym. Soc. Exptl. Biol.13: 353-64.

6. DAVIES, D. D. 1967. Glyoxylate as a substratefor pyruvic decarboxylase. Biochem. J. 104: 50 p.

7. ELLYARD, P. W. AND M. GIBBS. 1967. The effectof oxygen on photosynthesis in cholroplasts. PlantPhysiol. 42: S-33.

8. GOLDSWORTIHY, A. 1966. Experiments on the originof CO., released by tobacco leaf segments in thelight. Phytochemistry 5: 1013-19.

9. JENSEN, R. G. AND J. A. BASSHANM. 1966. Pho-tosynthesis by isolated chloroplasts. Proc. Natl.Acad. Sci. 56: 1095-1101.

10. KEARNEY, P. C. AND N. E. TOLBERT. 1962. Ap-pearance of glycolate and related products of pho-tosynthesis outside of chloroplasts. Arch. Biochem.Biophys. 98: 164-71.

1 1. KING, J. AND E. R. WAYGOOD. 1968. Glyoxylateaminotransferase from wheat leaves. Can. J.Biochem. 46: 771-79.

12. NAKADA, H. I. 1964. Glutamic-glycine transami-nase from rat liver. J. Biol. Chem. 239: 468-71.

13. LMETZLER, D. E., J. OLIV-ARD, AND E. E. SNELL.1954. Transamination of pyridoxamine and aminoacids with glyoxylic acid. J. Am. Chem. Soc.76: 644-47.

14. Moss, D. N. 1968. Photorespiration and glycolatemetabolism in tobacco leaves. Crop. Sci. 8: 71-76.

15. PITTS, J. D., J. A. STEWART, AND G. W. CROSBIE.1961. Observations on glycine metabolism inEscherichia coli. Biochim. Biophys. Acta 50: 361-63.

16. PRITCHARD, G. G., C. P. WHITTINGHAM, W. J.GRIFFIN. 1961. Effect of isonicotinyl hydrazideon the path of carbon in photosynthesis. Nature190: 553-54.

17. RICHARDSON, K. E. AND NX . E. TOLBERT. 1961.Phosphoglycolic acid phosphatase. J. Biol. Chem.236: 1285-91.

18. TOLBERT, N. E. 1962. Glycolate pathway. In:Photosynthetic Mechanisms in Green Plants.NSF-NRC publication 1145. p 648-62.

249

www.plantphysiol.orgon August 6, 2019 - Published by Downloaded from Copyright © 1969 American Society of Plant Biologists. All rights reserved.

PLANT PHYSIOLOGY

19. TOLBERT, N. E., A. OESER, T. KISAKI, R. H. HAGE-MAN, AND R. K. YAMAZAKI. 1968. Peroxisonmesfrom spinach leaves containing enzymes related toglycolate metabolism. J. Biol. Chem. 243: 5179-84.

20). TOLBERT, N. E., A. OESER, R. K. YANIAZAKI, R. H.HAGEMAN, AND T. KISAKI. 1969. A survey ofplants for leaf peroxisomes. Plant Physiol. 44:135-47.

21. VANDOR, S. L. AND N. E. TOLBERT. 1968. Gly-colate biosynthesis by isolated chloroplasts. PlantPhysiol. 43: S-12.

22. WANG, D. AND E. R. WAYGOOD. 1962. Carbonmetabolism of 14C-labeled amino acids in wheatleaves. I. A pathway of glyoxylate-serine me-

tabolism. Plant Physiol. 37: 826-32.23. WALKER, D. A. 1967. Photosynthetic activity of

isolated pea chloroplasts. In: Biochemistry ofChloroplasts. Academic Press. p 53-69.

24. WILsoN, D. G., K. W. KING, AND R. H. BURRIS.1954. Transamination reactions in plants. J.Biol. Chem. 208: 863-74.

25. WILKINSON, A. P. AND D. D. DAVIES. 1958.Serine-glycine interconversion in plant tissue. Na-ture 181: 1070-71.

26. YOUATT, J. 1958. The action of isoniazid on thetransaminases of Mycobacteriunt tuberculosis(B.C.G.). Biochem. J. 68: 193-97.

27. Yu, Y. L., N. E. TOLBERT, AND G. M. ORTH. 1964.Isolation and distribution of phosphoglycolatephosphatase. Plant Physiol. 39: 643-47.

28. ZELITCH, I. 1958. The role of glycolic acid oxi-dase in the respiration of leaves. J. Biol. Chem.233: 1299-1303.

29. ZELITCH, I. 1964. Organic acids and respirationin photosynthetic tissues. Ann. Rev. Plant Phy-siol. 15- 121-42.

250

www.plantphysiol.orgon August 6, 2019 - Published by Downloaded from Copyright © 1969 American Society of Plant Biologists. All rights reserved.