intracellular localization of fructose 1, gbisphosphate aldolase

8/14/2019 Intracellular Localization of Fructose 1, GBisphosphate Aldolase

1/6

THE JOURNAL OF BIOLOGICAL CHEMIS TRYVol. 50, No. 5 , Issue of March 10, PP. 1892-1897, 1975Printed in U.S.A.

Intracellular Localization of Fructose 1, GBisphosphate Aldolase*(Received for publication, June 18, 1974)

RICHARD S. FOEMMEL,~ ROBERT H. GRAY, AND ISADORE A. BERNSTEINFrom the Department of Biological Chemist ry and the Cellular Chemist ry Laboratory of the Department of En-vironmental and Industrial Health, The University of Michigan, Ann Arbor, ikfichigan &?10&

SUMMARYSubmission of a rat liver homogenate made in 250 mMsucrose-l mM EDTA to centrifugation between 9,500 x g for10 min and 105,000 X g for 60 min results in the sedimenta-tion of 60 to 70% of the total cellular fructose 1,6-bisphosphatealdolase (EC 4.1.2.13). Under these conditions only about

one-quarter of the total triose phosphate dehydrogenase andphosphoglyceratekinase appears n the microsomal fraction.Ultrastructural immunologic localization techniques havedemonstrated that the aldolase s associated with the endo-plasmic reticulum, in situ. The binding of this enzyme tothe membrane s sensitive to changes n pH with an optimumat 6.0, and to increasing concentrations of NaCl and fructose1,6-bisphosphate, being about loo-fold more sensitive tothe ester than to the inorganic salt.

Traditionally, the glycolyti c enzymes have been considered tobe located in the soluble portion of the cell (cf. Ref . 1) althoughthere are data which suggest that particular enzymes o f this se-quence are associated with specific subcellular particulate ele-ments (cf. Ref. 2).Green et al. (2) reported that when bovine erythrocytes weretreated with 250 mM sucrose and then lysed in distilled water,most of the glycolyti c enzymes remained at least partially associ-ated with t.he membranous fraction. Hexokinase, fructose-1, 6-PZ1 aldolase, and triose phosphate isomerase were recoveredin 507, or greater yield by sedimentation at 60,000 X g. How-ever, in the presence of 150 mM NaCl, these enzymes dissociatedfrom the membranes which sedimented at 60,000 X g and re-mained in the supernatant solution. Green and colleaguessuggested that all glycolyt ic enzymes were associated with mem-

* This work is supported by Research Grant ES-00339 from theNational Institutes of Health, United States Public Health Serv-ice.1 Predoctoral Research Trainee, Training Grant 5-TOl-GM-00187.9 from the National Institutes of Health, United StatesPublic Health Service, and Predoctoral Research Fellow, Institutefor Environmental Quality, The University of Michigan. Sub-mitted in partial fulfil lment of the requirements for the Ph.D.degree at the Univers ity of Michigan. Present address, Diagnos-t,ics Division, Abbott Laboratories, North Chicago, Illinois (50064.1 The abbreviations used are : fructose-Pn, fructose bisphos-phate; fructose-l,S-Pz, fructose 1,6-bisphosphate; glucose-1,6-Pz, glucose 1,6-bisphosphate.

branes in the bovine erythrocyte. These workers obtained sup-porting data for this concept with Xaccharonayces cerevisiae.

Hernandez and Crane (3) reported the sedimentation of 80to 90% of the hexokinasc act ivi ty when muscle f rom pig heartwas homogenized in 300 mM sucrose and submitted to centri fuga-tion at 105,000 X g. Addition of 400 rnfix KC1 solubilized aboutone-third of the act ivi ty, while increasing the pH from 5 to 9solubilized nearly all of the enzyme. The enzyme reassociatedwith the particulate fraction after removal of the salt by dialysis.The solubilization could also be effected by 0.5 mM glucose-6-I,the enzymes product. The authors postulated an equilibriumdependent upon salt and pH between the soluble and particulatestates of the enzyme with glucose-6-P as a modulator of thisequilibrium.

Rose and Warms (4) localized hexokinase in the mitochondriaof cells in an ascitcs tumor. Solubilization was accomplished byexposure to 1 M NaCl or 0.1 m&f glucose-6-P. Again reassoci-ation could be achieved after dialysis. Li and Chien (5) sug-gested that this equilibrium might play a regulatory role sincethe K, for ATP was about 3 times greater for the soluble thanfor the bound enzyme.

Roodyn (6, 7) found that when nuclei from rat hepatocyteswere ultrasonicated in 250 mM sucrose-O.018 mM CaC12, only 3%of the total fructose-P2 aldolase was released into solution. How-ever, addition of 150 mM NaCl to the sucrose-CaClz mixturecaused the solubilization of more than 90% of the enzyme with-out ultrasonication. Clarke et al. (8) found that 20 to 30% ofthe total fructose-P2 aldolase sedimented when a homogenate ofrat brain or muscle in 250 mM sucrose-10 mM Tris-HCl, pH 7.4,was centrifuged at 100,000 x g. A reduction of the pH from7.4 to 5.5 with lactic acid, but not with acetic acid, resulted insedimentation of 60% of the brain enzyme but only 10% of themuscle enzyme. These workers proposed that this eff ect was afunction of the differential effe ct of lactic acid on the binding ofthe two isozymes of this aldolase in these tissues.

Arnold and Pette (9, 10) found that fructose-P2 aldolase andtriose phosphate dehydrogenase from rabbit muscle could bereversibly bound in vitro to F-actin, one of the structural proteinsof the muscle contractile apparatus. The binding of the aldolasewas sensitive to alkaline pH, to ionic strengths great.er than 100mM, and to the presence of metabolites such as fructose-l, 6-Pz,glucose-l, 6-l*, dihydroxyacetone-I, and 2,3-Pz-glycerate. TheK, of the bound enzyme was nearly 10 times greater than the K,for the soluble enzyme while the Vmax was twice as great. Hist.o-chemical staining confirmed the relevance of this observation,in situ. Staining for fructose-Pz aldolase at the light microscopic

atS

G

U

S

,o

ece

be

,

009

jbcog

o

oaded

o
http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/


2/6

1893level indicated that in sections of rabbit muscle, the enzyme waslocated at the sites of the actin filaments.

Data in the present report show that in rat liver, fructose -Psaldolase is associated with the endoplasm ic reticulum. By dif-ferential centrifugation of a homogenate of this tissue made in250 m&f sucrose-l m M EDTA , 60 to 70% of the total cellularaldolase sedimented in the microsomal fraction. Ultrastructuralimmunolog ic localization techniques showed specific labeling forfructose-P2 aldolase on and around the endoplasm ic reticulum.The binding is sensitive to changes in pH, and to variations in theconcentration and identity of salts and metabolic intermediates.

I I IETHODS AND MATERIALSAl l work w as done with adult male rats of the CFN strain (Cars-

worth Farms) from a randomly inbred colony main tained in thislaboratory.

Subfrac tionation of 1Zat Liver Homog enate-Microsomal frac-tions were prepared from a 1O70 (w/v) homoge nate in 250 mMsucrose-l mM EDTA as described by Dallner (11). Rat liver ho-mogenates were subfractionated into nuclear, mitochondrial, mi-crosomal, and soluble fractions by a slight modification of themethod of Scha chter (12).

Ellzymatic Assays-The conversion of fructose-1,6-P* to 1,3-Pz-glycerate was measured by the conversion in 10 min at 37 of32Pi to organically bound [32P]phosphate in the presence of 1 mMADP, 1 rnM ATP , 1 rn~ NAD+, 1.7 mM MgC12, 1 mM fructose-1,6-Pz,33 mM Tris-HCl, pH 7.6, 1.8 mM sodium 32P;, and homogenate orextract (0.1 to 1.0 mg of protein) in a final volume of 3.0 ml. Afteraddition of 1 ml of 10% NH40H containing 300 mM Pi, all radio-active P; was precipitated at 0 by the addition of 1.0 ml of mag-nesia mixture (cf. Ref. 13) leaving organic ally bound 32P, primarilyATP, in solution. Radioactivity was measured by monitoringCerenkov radiation (14) at the normal setting for tritium in aPackard Tri-Carb 3375 liquid scintillatio n counter, corrected forquench by the channels ratio method (15).

Fructose-P% aldolase was assayed either spectrophotometrically(16) or calorim etrically (17) ; horserad ish peroxidase was assayedas described by Maehly and Chance (18), and glucose B-phospha-tase was assayed by the method of Swanson (19) as modified byHubscher and West (20).Chemical Analysis-Protein was determined by the method ofLowry (21). DNAw asextracted from cellul arsub fraction s by themethod of Schne ider (22), and analyzed by the diphen ylaminemethod of Dische (23). Calf thymus DNA was used as a standard.

Purijkatio n of Fructose-f ,6-Pp Aldo lase (EC 4.1.2.13) from RatLiver Microsomes for Preparation of Anti-Aldolase-A microsomalfraction was obtained from a homoge nate of rat liver, made in250 rnM sucrose-l mM EDTA , by sedimentation between 7,500 Xg for 10 min at 4 (the post-mito chond rial fraction) and 70,000 Xg for 105 min at 4. After rinsing in the sucrose-EDTA solution,the pellet was rehomogenized in one-half the original volume of250 mM sucro se-l mM EDT A-150 mM NaCl to liberate the boundaldolase w hich appeared in the supernatant solution after centri-fugation at 70,000 X g for 105 min at 4. Fructose-Pz aldolase wasthen precipitated by dialysis against saturated (NHI),SO~. Theprecipitate was resuspended in 1 mM EDTA-20 mM Tris-IICl, pH7.6, to a final concentration of 20 mg per ml, dialyzed against thisbuffer to remove a ll (NHa)#Oa, and fractionated two succ essi vetimes on columns of phosphoce llulose essentially as described byGracy et al. (24). Elution of the second column with fructose-1,6-Pn showed one peak with coincidence of protein and aldolaseactivity which exhibited only one band of protein by disc gel elec-trophoresis (25) on 77, polyacrylamide in 380 mM Tris-glycine atpH 8.3. The final recovery of enzyme, based upo n the level of ac-tivity observed in the homoge nate, was 10% at a purifica tion of62-fold with a specific activity of 6.03 units per mg of protein as-sayed (16) at 37. These characteristics agreed well with valuesobtaine d for a crystallin e preparation of this enzyme by Matsu-shima ei al. (26).

Preparation of Rabbit IgG Anti-Aldolase-Antiserum to purifiedfructose-P2 aldolase w as obtained from New Zealand white rab-bits which had received, at IO-day in tervals, three series of injec -tions (0.1 ml in each foot pad and 0.1 ml intradermally) of an emul-sified equal volume mixture of the aldolase in physiological saline

(0.9% NaCl solution) (2 mg per ml) and Freunds complete ad-juvant. Beginning 2 weeks after the last injection, bleedings fromthe ear vein were done once each 2 weeks for 2 months. The IgGfraction was isolated from antiserum by precipitation with NazSOrand purified on DEAE -cellulose (27). The purity and specificityof the antibody was examined against the supernatant fraction of ahomogen ate (57, w/v) of rat liver in 250 mM sucro se-l mM EDT A-150 mM NaCl centrifuged at 100,000 X g for 60 min. When theantiserum and anti-aldolase IgG were tested by immuno-diffusion,a single precipitin band with no spurring was observed. Controlrabbit serum and IgG obtained from the same rabbits prior to im-munization showed no precipitin bands. Immunoelectrophoresis(28) indicated a single band when tested against the rat liver ex-tract. Anti-aldolase IgG precipitated 95% of the aldolase fromthe extract when incuba ted at 4 for 4 days and centrifuged at10,000 X g for 20 min at 4. Control rabbit serum had no such ef-fect.

Purijkatio n of Peroxida se-labeled Goat IgG (Anti-Rabbit-ZgG)-Peroxidase was conjug ated to goat (anti-rabbit-IgG) IgG usingp,p-difluoro-m,m-dinitrodiphenylsulfone as the coupling reagentaccording to the procedure described by Nakane and Pierce (29).The conjugate was further purified by adsorption on activatedcharcoal and rat liver acetone powder.2 Ten milligrams of ratliver acetone powder, prepared accord ing to Coons (30), wereadded to 0.1 ml of stock conjugate previously diluted to 1.0 mlwith 50 mM phosphate-0.9oy0 NaCl, pH 7.2, and were shaken for 1hour at room temperature. The mixture was then submitted tocentrifugation at about 18,000 X g for 15 min at 4. The superna-tant was exposed to the acetone powder procedure twice morewith the addition each time of 10 mg of activated charcoal. Th elight yellow supernatant solution (about 0.7 ml) which finally re-sulted was dialyzed overnight against 100 mM Tris-HCl, pH 7.4and used immediately.

Tissu e Fixation-Liver was excised, diced into pieces of about2 mm3 while in 2.77, formaldehyde-0.2yn picric acid-250 mM su-crose-20 mM cacodyla te-HCl, pH 7.2, at 4, and incuba ted in theice-cold solution for 4 hours. After several rinses in 200 mM caco-dylate-HCl, pH 7.2, and wash ing overnight at 4 in the buffer withconstant mixing, the fixed tissue was embedded in 7.5% agar.Sections were cut at 60 to 100 pm using a Sorvall TC-2 tissue sec-tioner and were washed overnight in 200 mM Tris-HC l, pH 7.4, toremove the last traces of fixative.

Cytochemical Procedure-Ultrastructural immunologic localiza-tion of fructose-P2 aldolase was carried out by first attaching rab-bit anti-aldolase IgG to the aldolase in the fixed rat hepatocytictissue then exposing the combined aldolase-IgG to goat (anti-rabbit IgG) IgG conjug ated to horserad ish peroxidase. Th eentire complex was localized by incubating the tissue in 3,3-di-aminobenzidine, and H,Os. After preparation of tissues for elec-tron microsco py, the peroxidase reaction product was electronopaque and visible as a dark precipitate in the tissue.

The immunochem ical techniques which were used, were adaptedfrom methods described by Kraehenbuhl et al. (31). A t issue sec-tion was exposed to 0.25 ml of rabbit IgG (either control or anti-aldola se) for 5 to 6 hours at room tempe rature, rinsed, and washedovernight in 200 mM Tris-HC l, pH 7.4, at 4 with consta nt agita-tion. The specimen s were then incubated with 0.25 ml of goatIgG-peroxidase conjug ate in 100 mM Tris-HC l, pH 7.4, for 5 to 6hours at room temperature and again rinsed an d washe d overnightin Tris-HCl as before. After fixation for 3 hours in 37, glutaralde-hyde-100 mM cacodylate-H Cl, pH 7.4, at 4, and rinsin g for 4 hoursin 100 m M Tris-HCl, pH 7.4, the sections were stained for peroxi-dase by incubation in 0.05% 3,3-diaminobenzidine for 1 hourat room temperature and then for 10 min more a t room tempera-ture in 0.05y0 3,3-diaminobenzidine-O.OvO H202 in 200 mM Tris-HCl, pH 7.4. The stained sections were finally rinsed in theTris-HCl buffer, washed overnight with constant agitation in 200mM cacodyla te-HCl, pH 7.4, at 4, and prepared for electronmicroscopy.

Preparation of Tissu es for Electron l%!ficroscopy-Tissues werepostfixed in 2% OsOa-200 mM cacodyla te-HCl, pH 7.4, at 4 for 60min, rinsed three times in buffer, dehydrated in alcohol, washedtwo times in propylene oxide, and embedded in Epon. Sectionswere cut at 60 to 100 nm and mounted on grids coated with Form-

2 B. Baker, Department of Anatomy, The University of Michi-gan, personal communication.
http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/http://www.jbc.org/


3/6

1894var and carbon. The mounted sections were poststained in adrop of cacodylate-buffered 0~04 fo r 10 min at room temperatureand washed with distilled water. The specimens were examinedby an AEI Corinth 275 electron microscope.Reagents-Reagents were obtained from the following sources.Fructose-l ,6-P2 heptahydrate was from Wessex Biochemicals,adenine nucleotides and 3,3-diaminobenzidine from Sigma Chem-ical Co., NADH from Boehringer Mannheim Corp., ultrapure(NH~)~SOI from Schwarz-Mann, phosphocellulose from Mann Bio-chemicals, horseradish peroxidase (type II) from Sigma ChemicalCo., goat (anti-rabbit-IgG) IgG from Pentex, Freunds completeadjuvant from Difco Laboratories, agar and p,p-difluoro-m,m-dinitrodiphenylsulfone from General Biochemicals, activatedcharcoal from Atlas Power Co. , 0~04 from Merck and Co., p-for-maldehyde from Mallinckrodt Chemical Works, and glutaralde-hyde (as 70yo solution in sealed ampules) was from Ladd ResearchIndustries.

RESULTSIn a preliminary effort to delineate the extent to which certain

of the glyco lytic enzymes (fructose-Pa aldolase, triose phosphatedehydrogenase, and phosphoglycerate kinase) are present in themicrosomal fraction from rat liver, the conversion of a2Pi to or-ganic P, dependent upon added fructose-l, 6-Pa, was measured.When assaying for the level o f each of the three enzymes, excessesof the other two enzymes, as commercial preparations, wereadded. The microsomal fraction contained 65% of the aldolase,but only 3% of the dehydrogenase, and 10% of the kinase whichwerepresent n the post-mitochondrialsupernatantsolution.

TABLE ISubcellular localization of fructose-l ,6-PZ aldolase activ ity innuclear, mitochondrial, microsomal, and soluble fractions

For procedural details, see Methods and Materials in text.Fraction Aldolase Protein

%Crude homogenate 100 100Nuclear.. . . 19 9Mitochondrial.. 2 29Microsomal. 58 25Soluble. 13 30

Relativespecificactivity DNA

1.02.10.12.30.4

%100 1002 9113 470 Cl4


4/6

FIG. 4. Electron microscop ic demonstration of fructose-Pz aldolase speci fic staining in the endoplasmic reticulum of fixed liver tis-sue. A and C, anti-aldolase IgG; B, control IgG. Post-stained in 2% 0~0~. Magnification: A and B, X 30,000;C, X 12,000.total cellular aldolase eattached to the microsomes. The datain Fig. 1 indicate that not only was he enzyme almostcompletelysolubilized rom the particulate fractions by addition of 40 mMNaCl but that the total recovered activity exceeded100% ofthat present n the homogenaten the absence f salt. While ittook 5 mM NaCl to solubilize 50% of the microsomalaldolase,as little as 0.05 mM fructose-1,6-I2 produced the sameeffect(Fig. 2). The binding of aldolase o the microsomes as alsofound to be sensitive o changesn pH being maximal at pH 6.0and 20% at pH 8.0 (Fig. 3).Since he enzyme can be reversibly solubilizedby changing he

pH or the concentrationsof various salts, he binding might bean artifact and not exist in wivo. The best available proof forthe physiological elevanceof the binding would be the immuno-logical ultrastructural localization of the enzyme in tissuesec-tions. Fig. 4A demonstrates y this technique hat the enzymeis associatedwith the endoplasmic eticulum. Control IgGproducedno labeling (Fig. 4B). Although by biochemical rac-tionation, the nucleusappeared o contain aldolase,no nuclearlabeling was seen (Fig. 4C) by the immunological echnique.Fig. 4C showshat the enzyme s associated ith both the smoothand rough endoplasmic eticular membranes. When the im-


5/6

1896

F IG . 5. Specific staining of microsomal membranes for fructose-P2 aldolase. Anti-aldolase IgG is used with no post-staining inosoa. Magnification: X 30,000. A, microsomes isolated in 250 mM sucrose-l mM EDTA; B, microsomes isolated in 250 mM sucrose-1 mM EDTA-150 mM N&l.muno-electron microscopic technique was applied to a micro-somal fraction made in sucrose-EDTA, labeling was seen (Fig.5A), whereas addition of 150 mM NaCl resulted in a preparationin which no labeling is observed (Fig. 5B).

DISCUSSIONAlthough there are published data, obtained in vitro, which sug-gest that certain of the glycolyt ic enzymes are localized in an

organelle or on a cellular structural element, most of these datacould also be interpreted as representing an artifactual bindingwhich occurred during the experimental procedure employed.The most compelling data favoring particulate localizations ofthese enzymes, in situ, would make use of ultrastructural im-munological localization techniques. The present study utilizedsuch techniques to demonstrate the association of fructose-Pzaldolase with the endoplasmic reticulum, in situ. Rabbit anti-fructose-P2 aldolase IgG was shown by electron microscopy tolocalize on or. in close apposition to these membranes in tissuesections of rat liver. The specifi city of the anti-aldolase IgG wasshown b y its ability to precipitate all of the aldolase from a crudeextract of rat liver which had been made in the presence of 150mM NaCl; the purity was shown b y the formation of only oneprecipitin band when tested by immunoelectrophoresis againstthe extract. Although it is not presently possible to quantitatecytoimmunologically the percentage of the total aldolase whichis present on the endoplasmic reticulum, biochemical fractiona-tion showed that at least 607, of the enzyme is associated withthese membranes.

The increase in total cellular enzymatic act ivi ty seen when allof the fructose-Pz aldolase was solubilized (Fig. 1) could be ex-.plained by an increase in the Vm,, of the enzyme as was found forfructose-P2 aldolase in rabbit muscle (10). Further investiga-tion of this point is in progress.

Recent studies in this laboratory3 have demonstrated an in-crease of at least 3507 in the total act ivi ty of microsomal fruc-tose-12 aldolase when 80 mM NaCl was added to a preparation ofmicrosomes which had been isolated by centrifugation from a ratliver homogenate and resuspended in 250 mM sucrose-l mMEDTA. Simply adding 80 mM NaCl to a soluble dialyzed prepa-ration of the aldolase does not result in an increase of act ivi ty.These observations and the finding that the association of fruc-tose-Pz aldolase with the endoplasmic reticular membranes issensitive to relatively small changes in the substrate suggest thatreversible association of the enzyme with the membranes maybe a mechanism for controlling the rate of glyco lysis in rat liver.Similar results have been reported in rat brain where the additionof 2 mM fructose-l , 6-PZ completely solubilizes fructose-12 aldol-ase from microsomes isolated in sucrose-EDTA, whereas about150 mM NaCl is necessary to accomplish the same eff ect (32).Arnold and Pette (10) found that the binding of aldolase to F-actin was sensitive to low concentrations of fructose-l ,6-Pz,dihydroxyacetone-I, 2,3-Pz-glycerate, and glucose-l, 6-PZ butmuch less sensitive or insensitive to other glycolyt ic intermedi-ates.

Support for the proposal that modulation of the act ivi ty offructose-Pz aldolase by interaction with the endoplasmic reticu-lum represents a regulatory mechanism for the glyco lytic path-way would include evidence that (a) the enzyme can reversiblybind to the membranes, (b) the percentage of the enzyme boundto the membranes or free in the cytosol is affected by biochemicalor physiological parameters; (c) the effec tive acti vity of solublealdolase is different from bound aldolase; and (d) physiological ly,this enzyme limits the rate of glyco lysis in either direction by

3 I. Weiss, Department of Environmental and IndustrialHealth, The University of Michigan, personal communication.


6/6

1897virtue of its transit between the particulate and soluble phases ofthe cell.

The reversible association of fructose-P2 aldolasc with themembranous cellular fraction has now been documented for cellsin the brain and muscle (8) as well as in the liver of the rat; inthe muscle of the rabbit (9, lo), in the bovine erythrocyte (a),and in yeast (a), in vitro, and for muscle actin in the rabbit (10)and the cndoplasmic reticulum in the hepatocytc of the rat, insitu. In rat brain, OIE of two isozymcs of aldolasc is prefercn-tially bound to the microsomcs while the other is preferentiallylocated in the cytosol (8). This binding has been showrl to besensitive to lactic acid in the casr of the brain of the rat (8), tosome glycolyt ic intcrmcdi~tes in the muscle of the rabbit (lo),and to fructose-l, 6-ly in the liver o f the rat. Insrcascs in totalcellular act ivi ty of the enzyme up011 solubilization have bee11demonstrated in the lattcxr two systems. That the rcvcrs ibk dis-sociation of fructosc-l 2 aldolasc f rom the membranes can b c nrlallostcric control rnc&~~lisrn is suggcstcd by the obscxrvation tllatupon adsorption 011 pl~os~~l~occllulosc, the enzyme (xhibits thesigmoidal kinetics of allostcric enzymes whw enzymatic activityis plotted against the concentration of fructose-l , G-l, (33).Ilowcvcr , tlic crucial demonstration that the incrcasc~ in act ivi ty,resulting from solubilizatioll, is associated with an alteration ofthe physiological rate of glycolys is, has not ~1syet been achieved.

REFERENCES1. DE DUVE, C., W.~TTI.~U X, It., AND B.YUDIIUIN, P. (19G2) Adv.

Etlzymol. Rclat. Subj. Biochem. 24, 291-3582. GIO:I,:N, I>. I

intracellular localization of fructose 1, gbisphosphate aldolase

Documents