l-Fucose is accumulated via a specific transport system in eukaryotic cells

THE

Vol. 269, No. 36,Issue of September 9,pp. 22705-22711, 1994 Printed in U.S.A.

JOURNAL OF BIOLWXCAL CHEMISTRY

L-Fucose Is Accumulated via a Specific Transport System in Eukaryotic Cells* (Received for publication, April 13, 1994)

Thomas J. WieseS, Joyce A. Dunlap, and Mark A. Yorekg From the Department of Znternal Medicine. Diabetes-Endocrinology Research Center a n d Veterans Affairs Medical Center, University of Zowa, Iowa City, Iowa 52246

L-Fucose is a monosaccharide normally present at low myo-inositol uptake in cultured neuroblastoma cells(13), and dietary L-fucose causes nerve dysfunction in rats similar to concentrations in serum and is the only levorotatory sugar utilized by mammalian systems. The metabolism changes notedin diabetic neuropathy (14). Decreased myo-inoof L-fucose is only partially understood. Inthis report, sitol metabolism and content is thought to be a factor in the we characterize the uptake of L-fucose by four widely development of diabetic neuropathy and possibly other diabetic varying mammalian cell lines (murine neuroblastoma, complications in animal models of diabetes (15). bovine aortic endothelial, murine cerebral microvessel A function forL-fucose, the only levorotatory monosaccharide endothelial,andMadin-Darbycaninekidney cells). utilized in mammalian systems, in cellular metabolism has not Based on the criteria of saturability and specificity of been defined, though its presence on a multitude of glycoproL-fucose uptake, we conclude that L-fucose is accumu- teins implies that it may have a specific role. The existence of lated via a specific recognition mechanism. Accumula- hepatic lectins that recognize L-fucose (16-19) indicates that tion of L-fucoseat 4 "C and in the presence ofcolchicine L-fucose may playa role in cellular recognition or protein clearand cytochalasinD rules out receptor-mediated endocytosis as an uptake mechanism. Thus, the accumulation ance. L-Fucose is the immunodominant sugar on glycoproteins, appears to be via a carrier system. Using a variety of and its presence may increase the antigenic response (1). Liver, criteria, we determined that L-fucose is not taken upby kidney, and thyroid arethe only tissues reported to synthesize a glucose transporter system. Accumulation of ~-[5,6- L-fucose (20-27); however, L-fucokinase is hypothesized to be SH]fucose is Na+-independent and reduced by loading present in all cells that produce glycoproteins (28). Thus, the cells with L-fucose or depleting the cell of its phospho- accumulation of L-fucose is probably an important process in rylation capability, suggesting that the uptake of ~ - f u - the synthesis of glycoproteins. In this report, we describe the uptake, accumulation, and cose is by passive facilitative diffusion. A significant amount of the L-fucose taken up byeach of the four cell metabolism of L-fucose by four mammalian cell linesand protypes was incorporated into protein and secreted into vide evidence indicatingthat i t is probably taken up by means the medium. in a fashion similar to, of a passive facilitative diffusion system but nonetheless distinct from, the glucose transporter systems, i.e. Glutl-Glut4 (29). Each of t h e cell lines was capable of incorporating L-fucose into proteins,a significant portionof which were secreted into the medium.

L-Fucose i s a monosaccharide that is normally presentat low concentrations in serum and has received a significant amount of attention from researchers in a variety of fields (1). Although EXPERIMENTALPROCEDURES poorly understood, L-fucose metabolism seems to be profoundly altered in cancer and other conditions. In ovarian cancer, LMaterials-L-Glucose, cytochalasin B, cytochalasin D,colchicine, fucose contentof haptoglobin is increased 7-fold (2). Reactivity phloretin, phloridzin, D-fucose,D-fructose,D-mannose, L-arabinose, Dribose,D-galactose,myo-inositol,bovine serum albumin, D-xylose, Dand GalP1of lotuslectin,whichrecognizesFucal-2Gal 4(Fucal-3)GlcNAc, has been correlated with metastatic activ- allose, L-rhamnose, D-arabinose,L-glucose, L-galactose, phorbol 12-myristate 13-acetate (PMA),' 3-O-methylglucose, and N-ethylmaleimide ity in bladder carcinoma (3). Furthermore, the serum sialic were obtained from Sigma. Insulin was a gift from Eli Lily, Indianapacidfucose ratio and serum L-fucose levels have been investi- olis, IN. N-BromosuccinimidewasfromAldrich. L-Fucose wasfrom and breast carcinoma (5-7), Pfanstiehl Laboratories, Waukegon, IL.~-[5,6-~H]Fucose gated as markers of oral cancer (4) was from ICN and respectively. Reports have also appeared indicating altered L- (Costa Mesa, CA) orAmersham Corp., and ~-2-[2,6-~H]deoxyglucose fucose metabolism in cirrhosis (8)and rheumatoidarthritis (9). 3-O-[methyZ-3Hlglucosewere from Amersham Corp. Reagent ethanol, hydrochloric acid, methanol, chloroform, trichloroacetic acid, magne(10-12), Serum L-fucose levels are also increased in diabetes sium sulfate, calcium chloride,and potassium chloride, as well as Cornparticularly in association with the serum proteins a,-antitryp- ing 75- and 25-cm2tissue culture flasks, Falcon p35 and 6-well plates sin, &,-acid glycoprotein, and haptoglobin(12). The significance were from Fisher. Safety-Solve and scintillation vials were from RPI, of this observation is not known, although L-fucose inhibits Mount Prospect, IL. Cell culture media was obtained through the Diabetes-Endocrinology ResearchCenter, University of Iowa. Sodium chlo* This work is supported by Grants 45453 and 25295 from the Na- ride, HEPES, and sodium hydroxide wereobtained from EM Scientific, tional Institutes of Health, by a grantfrom the Juvenile Diabetes Foun- Gibbstown, NJ. Sodium dodecyl sulfate was from British Drug House dation, by a career development award from the Juvenile Diabetes Limited, Poole, United Kingdom. '251-Transferrin wasobtained through Foundation, and by a grant from the Veterans Administration. The the Diabetes-EndocrinologyResearch Center, University of Iowa. costs of publication of this article were defrayed in part by the payment Cell Linesand Culture-Neuroblastoma NB41A3 cells were grownin of page charges. This article must therefore be hereby marked "aduer- Ham'sF-10medium supplemented with 2.5% heat-inactivated fetal tisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Supported by Grant 07018 from the National Institutes of Health. The abbreviations used are: PMA, phorbol 12-myristate 13-acetate; 5 Towhom correspondence should be addressed: 3E-17 VA Medical CME, cerebral microvessel endothelial; MDCK, Madin-Darby canine Ctr., Iowa City, IA 52246. Tel.: 319-338-0581(ext. 7696); Fax: 319-339- kidney; NEM, N-ethylmaleimide; DNP, 2,4-dinitrophenol;IOA, iodoac7025. etate; NB, neuroblastoma; BAE, bovine aortic endothelial.

22705

22706

L-Fucose Dansport in Eukaryotic Cells

bovine serum, 15% horse serum, 100 unitslml penicillin, 100 pg/ml strepTABLEI tomycin, and 294 pg/ml glutamine. MB114 CME cells were grown in 48-h L-fucose accumulation M199 medium supplemented with10% heat-inactivated fetalbovine seCells were incubated in media containing 1pCi/well ~-[5,6-~H]fucose rum, 50 unitslml penicillin, 50 pglml streptomycin, and basal media in 10 PM unlabeled L-fucose for 48 h. Cell suspensions were then obEagle aminoacid and vitamin solutions. Bovine aortic endothelial(BAE) tained and analyzed for protein content, L-fucose accumulation, protein cells were grownin Dulbecco’smodified Eagle’s medium supplemented incorporation, lipid incorporation, and secretion of protein as previously with 10% heat-inactivated fetal bovine serum, 100 unitdm1 penicillin, described (see “Experimental Procedures”). Each value is a meanof at 100 pg/ml streptomycin, and 294 pg/ml glutamine. Madin-Darby canine leastnineindependentobservations.Accumulationisexpressedas nmol/mg of protein, and incorporation and secretionare expressed as a kidney (MDCK) cells were generously provided by Dr. Arthur Spector and grown in Dulbecco’s modified Eagle’s medium supplemented with percent of total accumulation of L-fucose &.E. 10% heat-inactivated fetal bovine serum, 100 unitdml penicillin, and Percent of total accumulation 100 pg/ml streptomycin. The cells were propagated in Corning75-cm2 Cells Accumulation Protein tissue culture flasks ina n incubator maintainedat 37 “C with 5% CO, Lipid Intracellular Secreted in humidified air as the gas phase. Cells were passed weekly at a 1:20 dilution and refed 3 timeslweek by replacing the media. In some studies, nmol l m g medium was supplemented with m 30 M o-glucose, or 1or 10 mM L-fucose, NB 1.53 ? 0.20 0.3 ? 0.1 22.4 i 4.0 44.5 2 6.7 in which case the cells were “conditioned for at least 1 week prior to CME 1.37 t 0.36 0.5 ? 0.2 17.9 2 5.4 49.2 ? 4.9 experimental manipulations.Before experimental use,cells were seeded BAE 1.60 t 0.59 0.6 t 0.5 22.1 t 4.1 55.7 t 2.6 onto Falcon6-well cluster plates, and assays were conducted in triplicate MDCK 1.37 2 0.40 1.7 2 1.0 24.7 2 4.8 43.8 2 10.2 when cells were in earlyconfluency. L-Fucose Accumulation a n d Metabolism-Medium as described above was modified by the addition of ~-[5,6-~Hlfucose and 10 PM unlabeled L-fucose. Cells were incubated in a medium containing 1 pCi of cipitated by centrifugation, and the pellet was washed with 10% trichloroacetic acid. The washed pellet was transferred to a scintillation ~-[5,6-~H]fucose/wellinFalcon6-well plates for 1-144 h. Following vial, 4.5 ml of Safety-Solve was added, and radioactivity measured as these incubations, cells were washed twice in ice-cold buffer (10 mM above. Media incubated in the absence of cells were treated in the same HEPES pH 7.4, 128 mM NaCl,5.2 rn KCl, 2.1 mM CaCI,,2.9 mM manner for use as a control for nonspecific absorption of ~-[5,6-~Hlfucose MgSO,, and 5 mM glucose) and collected in 1.5 ml of water. The cell in theprotein pellet. This value was subtractedfrom the experimental suspension was sonicatedfor 5 s, and samples were taken to determine values obtained. proteincontent,~-[5,6-~Hlfucoseaccumulation,andincorporation of Amino Acid Modification Experiments-Media containing 1pCi/well ~-[5,6-~H]fucose into protein and lipid. ~-[5,6-~HlFucose accumulation ~-[5,6-~H]fucose or ~-2-[2,6-~H]deoxyglucose were modified by the addiwas determined by taking duplicate aliquotsof cell suspension, adding tion of 0.5 mM NEM or 0.5 mM N-bromosuccinimide, andthe cells were 4.5 ml of Safety-Solve, and measuring the radioactivity with a Beckman incubated for 2 h. Cell suspensions were then obtained and analyzed for LS8100 liquid scintillation counter. Incorporation of ~-[5,6-~Hlfucose protein content and ~-[5,6-~H]fucose or ~-[2,6-~H]2-deoxyglucose accuinto protein was determinedby the additionof 300 plof 20% trichloromulation as described above. acetic acid to a 25O-pl aliquot of cell suspension using50 pl of 1 mg/ml Temperature Dependence of L-FucoseAccumulation-The accumulabovine serum albuminas a carrier protein. After vortexing and standing tion of L-fucose by CME and MDCK cells was examinedat 4 and 37 “C 5 min on ice, the protein was precipitated by centrifugation, and the by incubating cells in media containing 1 pCi/ml of ~-[5,6-~HIfucose, pellet was washed with10%trichloroacetic acid. The washed pellet was which had been temperature-equilibrated t o either 4 or 37 “C, and at transferred to a scintillation vial,4.5 ml of Safety-Solve was added, and various time intervals, examining accumulation as describedabove. radioactivity was measured as above. To determine incorporation of Receptor-mediated Endocytosis Assay-Uptake of ‘251-transferrin ~-[5,6-~H]fucose into lipid, an aliquot of cell suspension was placed 10 in was examined at 4 and 37 “C as a control for receptor-mediated endoml of CHClfleOH (2:l). The lipid fraction was isolated following phase cytosis. In these experiments,lo6 cpm of ‘251-transfen-inwas added per separation with 4 mlof 154 mM NaCl containing 0.6M HCl. The entire well and assayed for uptake 1 h later. A zero time control was used to allowed t o evapolipid fraction wascollected into a scintillation vial and determine nonspecific binding of radioactivity to cells, and this value rate, and then 10 ml of Safety-Solve was added, and the radioactivity was subtracted from the experimental value. Following experiments counted as above. Protein content was determined in duplicate aliquots designedtotestreceptor-mediated endocytosis, cells werequickly of cell suspension using a modification of the method of Lowry as dewashed 3 times with cold buffer, stripped of surface-bound material scribed by Lees and Paxman (30) using bovine serum albumin as a with 1 muwell of 0.2 M acetic acid, 0.5 M NaCl, pH 2.4, for 4 min, and standard. then quickly washed once more with cold buffer digested with 0.5 ml/ To examine the concentration dependence of ~-[5,6-~Hlfucose uptake, well of 1 N NaOH (31), and the total well contents was counted on a cells were grown to confluence in Falcon p35 plates and incubated for 30 Micromedic 4/2OOplus automatic gamma counter. s with increasing concentrations of L-fucose, while maintaining a conData Analysis-Data are reported as nmoVmg of protein, cpm, perstant specific activity of ~-[5,6-~Hlfucose. Following the incubation, upcent of total accumulation, oras percent of control. Where appropriate, take was stopped by the addition ofcold buffer and the wells were statistical comparisons were performed using Student’s t test. Thecalquickly aspirated and washed twice more with cold buffer. For each culation used to determine accumulation assumes that protein-bound concentration of L-fucose, zero time controls, representing nonspecific radioactive L-fucose must have been taken upby the cell, incorporated absorption of radioactivity, were subtractedfrom experimental values. into protein, and then secreted into the media. The calculation does not To examine the influence of selected inhibitors and competitors on take into account ~-[5,6-~Hlfucose taken upby the cells and secreted ~-[5,6-~H]fucose accumulation, compounds were added to the labeling unbound, and, thus, represents the lower oflimit L-fucoseaccumulation. media at concentrations listed in the table legends. I n cases where Kinetic values were calculated using Cleland’s method of weighting alternate solubilization of chemicals was required (e.g. cytochalasin B values (32). into ethanol), alternate controls were utilized; however, no difference was seen between unmodified control and alternate control values. IniRESULTS tial studies indicatedthat competition was maximal after 24 ofh incubation; therefore, this timeperiod was usedfor subsequent competition L-Fucose Accumulation and Incorporation-Previous work in experiments. Inhibitor experiments were for runshorter periodsof time our laboratory had indicated an accumulation of L-fucose by since the cells did not survive for 24 h in the presence of all inhibitors. cultured cells (10). To determine accumulation of 10 pM ~-[5,6To examine regulation of sugar accumulation, cells were serum- and 3H]fucose, accumulation wasinitially examined at various glucose-starved for 2 h in the absence or presence of insulin or PMA, followed by an additional 1-h incubation in the presence of o-2-[2,6- times up to 144 h. As shown in Table I, at 48 h the amount of incorporated into lipid was minimal, whereas 3H]deoxyglucose, 3-O-[methyl-3Hlglucose, or ~-[5,6-~HIfucose and ana- ~-[5,6-~H]fucose incorporation into protein was considerably higher. Some of lyzed as above. L-Fucose-containing Secreted Proteins-To examine the secretion of these L-fucose-containing proteins were present intracellularly ~-[5,6-~H]fucosylated proteins into culture media, cells were incubated at the time of harvest, whereas themajority were secretedinto as above. At the time points under investigation, aliquots of media were the medium. The remainderof the ~-[5,6-~Hlfucose takenis up removed from the cells and centrifuged (12,000 x g for 1mid, and the supernatant was treated with an equalvolume of 20% trichloroacetic presumed to be unbound and is certain t o represent various acid. After vortexing and standing 5 min on ice, the protein was pre- chemical species such as L-fucose, L-fucose 1-phosphate, GDP-

22707

L-Fucose Dunsport in Eukaryotic Cells TABLE I11

TABLEI1 Effects of various monosaccharides on L-fucose accumulation in NB, CME, BAE, and MDCK cells Monosaccharides were added to buffer containing 1 pCi/well ~-[5,63H]fucosein 10 VM unlabeled L-fucose and incubated 24 h. Cell suspensions werethen obtained and analyzed for protein content and L-fucose accumulation as previously described (see "Experimental Procedures"). Each value is a mean of at least nine independent observations expressed as a percent of the accumulation measured in the absence of comuetitors 2S.E. Conditions

L-Fucose accumulation NB

CME

BAE

MDCK

Effects of inhibitors of L-fucose accumulation Inhibitors were added to buffer containing 1 pCi/well~-[5,-6~Hlfucose and incubated 2 h. Cells were incubated inthe absence of extracellular sodium for 8 h. In the case of DNP and IOA,cells were pretreated for 1 h prior to the addition of the ~-[5,-6~H]fucose. Cell suspensions were then obtained and analyzedfor protein content and L-fucose accumulation as previously described (see "ExperimentalProcedures").Each value is a mean of at least six independent observations, expressedas a percent of the accumulation measured in the absence of competitors or presence of sodium ?S.E. L-Fucose uptake Inhibitor

3'% of control

None (control) 1mM L-fucose 5 mM D-fucose 10 mM D-glucose

100 100 100 100 49 t 8" 52 % 5" 49 t 5" 28 t 2" 9551 95 t 2 9 8 2 2 108 t 6 952 2 90 t 2 84 5 2" 86 t 3" 110 f 8 104 % 2 100 2 2 90 t 2 5 mM D-fructose 90 t 2 98 f 6 100%2 95 2 3 5 m~ D-mannose 94%2 9823 101 f 4 97 t 2 5 m~ L-arabinose 9852 91 t 5 5 mM D-ribose 9522 99 f 1 91 % 1 9453 106 t 3 9453 5 mM D-galactose 88 % 2" 9 2 %3 104 5 4 100 1.1~myo-inositol 108 f 2 108 f 2 9 7 % 3 9Of3 86 % 1" 5 mM D-xylose 97%4 9154 78 % 2" 101 f 2 5 mM D-allose 99%1 106 5 2 95 % 1 9552 5 mM L-rhamnose 94%3 9153 95 % 1 111 5 3 5 mM D-arabinose 99 % 1 110 f 6 104 5 1 95 % 1 5 mM L-glucose 100 2 1 9552 93 % 2 98f2 5 mM L-galactose

NB

CME

BAE

MDCK

% of control

None 100 (control) 100 100 100 0.5 mM phloretin 74 5 9" 0.5 mM phloridzin 94 t 2" 100 p~ cytochalasin B 64 f 2" 93 5 8 -Na 6 6 f 5" 0.5 m~ DNP 0.5 mM IOA 66 t 3" 32

57 % 8" 57 f 5" 3 4 %3" 80 f 3" 57 5 5" 63 2 4" 66 f 8" 46 % 4"45 f 4" 69 f 6' 8 5 %7 92 f 7 5 4 %3" 73 % 2" 57 t 2" % 2" 77 f 2" 9 4 t 2

" p < 0.001. b p < 0.01.

IOA in eachof the four cell lines except for MDCK cells in the presence of IOA. As discussed below, the decreased accumulation in thepresence of DNP or IOA was probably due todepletion of the cell's phosphorylation capability. " p < 0.01. Saturability of L-Fucose Uptake-Analysis of the concentration dependence of ~-[5,6-~H]fucose uptake indicated CME, that L-fucose, etc. (1).The identity of the ~-[5,6-~Hlfucosylated pro- NB, and MDCK cells exhibited saturation of L-fucoseuptake by following 30-s incubations (data not 200 p~ ~-[5,6-~H]fucose teins is currently being investigated. conditions was An important observation was that the accumulation of L- shown). Uptake of ~-[5,6-~H]fucose under these fucose is dependenton the confluency state of the cells, as has linear for longer than 1min. Table N shows the kinetic values and ~-[5,6-~H]fucose uptake derived been shown to be the case for D-glucose uptake (33).A greater of ~-2-[2,6-~H]deoxyglucose show that K , values are a is accumulated by subconfluent and from these experiments. These data amount of ~-[5,6-~Hlfucose early confluent cells than by cells in a prolonged state of con- minimum of 6-fold less for ~-[5,6-~Hlfucose uptake thanfor are also profoundly fluency (data not shown). Every effort was taken in these stud-~-[2,6-~H]glucose uptake.Values for V, different. ies to use early confluent cells to eliminate this variable. L-Fucose Dunsport Mechanism-Receptor-mediated endocyAccumulation of L-Fucose Is Specific and Inhibitable-As is shown in Table 11, accumulation of ~-[5,6-~H]fucose specific, tosis is generally accepted not to occur at 4 "C. Therefore, acas deduced by competition of accumulation by other monosac- cumulation of L-fucose was monitored during simultaneous incharides. Glucose a t a concentration of 10 m~ inhibited the cubation of cells at 4 and 37 "C. In CME cells, accumulation accumulation of 10 p~ L-fucose by up to 15%.Other monosac- was decreased at 4 "C, as one would expect, but it was not charides, a t 5 mM concentration, were less effective than D- eliminated. Infour separate experiments,accumulation was 52 glucose in reducing ~-[5,6-~Hlfucose accumulation. In contrast, t 3% of control values. '251-Transferrin uptake at 4 "C was 22 1 mM L-fucose decreased~-[5,6-~Hlfucoseaccumulation by t 1%of control values. Colchicine and cytochalasin D, which greater than 50%. The structural and stereochemical require- have both been reported to block receptor-mediated endocytoments of ~-[5,6-~Hlfucose for accumulation also seemed to be sis, perhapsby their effects on microtubule formation (35, 36), quite specific. L-Galactose, which differs from L-fucose only by did not block L-fucose accumulation (111 t 1 and 98 t 3% of having a hydroxyl group in its sixth position rather than a control, respectively, in CME cells). Cytochalasin D diminished hydrogen, did not inhibit L-fucose accumulation. L-Rhamnose, transferrin uptake by 32% at the concentrations used here. which has its 2 and 3 hydroxyls in "syn" rather than "anti" Similar results have been obtained in MDCK cells. orientation, did not consistently inhibit ~-[5,6-~Hlfucose accuTherefore, it seemed that L-fucose transport by mammalian mulation by all fourcell lines. MDCK cells seemed to be some- cells occurs via a carrier proteinmechanism. Although a priori what more sensitive to competition than the othercell lines. there wasno reason to believe that L-fucose would be accumuTable I11 shows the effects of various metabolic inhibitors on lated by a glucose transport system, the inhibitionof accumu~-[5,6-~Hlfucose accumulation. Although reported to be specific lation by phloridzin, phloretin, andcytochalasin B, as shown in inhibitors of glucose uptake (341, phloridzin, phloretin, and Table 111, necessitated ruling out the possibility. To ascertain accu- whether the accumulationof L-fucose was regulated ina fashcytochalasin B all effectively diminished ~-[5,6-~H]fucose mulation in each of the cell lines examined.Table I11 also shows ion similar to the glucose transporter system,we examined the that ~-[5,6-~H]fucose accumulation was not dependenton ex- influence of insulin and PMA on L-fucose accumulation. tracellular sodium, as isosmotic replacement of NaCl in the Insulin is known to stimulate glucose uptake in CME cells incubation media with CsCl did not diminish ~-[5,6-~H]fucose (37, 38). Therefore, experiments were performed using CME accumulation consistently in any of the four cell lines exam- cells t o examine theinfluence of insulin on 2-[2,6-3Hldeoxygluined.Cerebral microvessel endothelial cells singularly dis- cose and ~-[5,6-~Hlfucose accumulation. These studies comare played some degree of sodium dependence. That L-fucose accu- plicated by the fact thatmost cells grown in culture aremaximulationisinpartenergy-dependentwas shown by the mally activated with respectto glucose transport under normal decreased accumulation in the presence of 0.5 mM DNP and cell culture conditions. Therefore, CME and MDCK cells were

7

22708

L-Fucose Dansport in Eukaryotic Cells

TABLE IV Kinetics of 2-deoxy-o-g~ucose and L-fucose uptake Cells were incubated in 10-200 w L-fucose or 1-50 II~MD-glucose, containing constant specific activity of ~-[5,6-~Hlfucoseor 0-2[2,63Hldeoxyglucose, for 30 s. Cell suspensions were then obtained and analyzed for protein content and L-fucose or D-2-deoxyglucoseuptake as previously described(see “Experimental Procedures”).This experiment was done only one time, in duplicate, due to the prohibitive cost of maintaining radioactive sugars at constant specific activity. N D , not determined. L-Fucose uptake

Cell line

5.3 55.6

NB CME BAE MDCK 177.4

8-Deoxy-~-glucose uptake Cell line

Km

v,”,

K!”

V””

,UM

pmollmg min

m M

nmollmg min

256 251 ND 862

104.9 380.5 22.8

TABLE V Effectof insulin on 2-deoxy-~-glucose and L-fucose accumulation Cellswere serum- and glucose-starved for 2 h in the absence or presence of 1w insulin. 1pCi/well of ~-2-[2,6-~Hldeoxyglucose or ~-[5,63Hlfucose was then added and incubated for an additional hour. Cell suspensions were then obtained and analyzed for protein content and ~-[5,6~Hlfucose or ~-2-[2,6~Hldeoxyglucose accumulation as previously described (see “Experimental Procedures”). Each value is a mean of nine independent observations.Accumulation is expressed as percent of the accumulation by cells not exposed to insulin 2S.E.

10.3

Accumulation L-Fucose

2-Deox~-~-alucose % of control

CME MDCK

96 ? 2 98 ? 2

215 L 6” 108 L 4

o p < 0.001.

TABLE VI Altered L-fucose accumulation in cells conditioned in 10 mM L-fucose serum- and glucose-starved in theabsence or presence of 1 insulin for 2 h, at which time ~-2-[2,6-~H]deoxyglucose or ~-[5,6- Cells were grown in media containing 10 mM L-fucose for at least 1 week. Cells werethen incubated in media containing 1pCi/well ~-[5,63Hlfucose was added to the medium, and thecells were incu- 3Hlfucose and 10 1”unlabeled L-fucose for 24 h. Cell suspensions were bated for an additional 1h. As shown in Table V, insulin stimu- then obtained and analyzed for protein content and L-fucose accumulain CME cells tion as previouslydescribed (see “Experimental Procedures”). Each lated the transport of ~-2-[2,6-~H]deoxyglucose MDCK cells value is a mean of at least nine independent observations. Accumula2-fold without effect on ~-[5,6-~H]fucose transport. tion is expressed as percent of the accumulation by cells not grown in as a negative the presence of 10 mM L-fucose, ?S.E. lack the insulinreceptor (39,40), and thus serve control. The MDCK cells treated with insulinshowed no differCell line L-Fucose accumulation accumulaence in ~-[5,6-~Hlfucose or ~-2-[2,6-~H]deoxyglucose % of control tion from control. Although D-glucose is maximally stimulated NB 26 ? 6 after serum- and glucose-starving for 2 h, starving for as long 44 2 4 CME as 6 h did not lead to stimulation ofL-fucose accumulation. BAE 30 f 2 Another known regulator of D-glucose uptake, 100 nM phorbol MDCK 30 f 2 12-myristate 13-acetate (411, increased ~-2-[2,6-~H]deoxyglucose uptake inCME cells to 2552 10% of control, without effect on ~-[5,6-~Hlfucose transport in any of the cell lines tested. cells grown in 30m glucose for 1week took up ~-[5,6-~H]fucose Naftalin and Rist (42) have shown that phorbol ester causes at a rate identical to cells grown in normal media (data not protein-proteininteraction of hexokinase withthe glucose shown). Additionalstudies indicated that BAE cells cultured in transporter, leadingto an increased phosphorylation rate that 10 m L-fucose and exhibitingdown-regulation of ~-[5,6-~H]fumimics an increased uptake rate. To rule out this possible ar- cose accumulation show no decrease in uptake of ~-2-[2,6tifact, 3-O-[rnethyZ-3Hlglucose,which is not a phosphorylation 3Hldeoxyglucose and that reversal of down-regulation is possubstrate, was also examined and found to increase in uptake sible within 1week after replacingcells in normal media (data not shown). to 180% of control in CME cells. It was initially assumed that the down-regulation of L-fucose In order toprovide further evidence that a specific transport protein was responsible for ~-[5,6-~H]fucose accumulation, N- transport was due t o a decrease in plasma membrane transethylmaleimide was used to modify cysteine amino acid resi- porters. However, subsequent studies showed that culturing and ~-[5,6- cells in 10 mM L-fucose for 1h, followed by analysis of accumudues, and accumulationof ~-2-[2,6-~H]deoxyglucose 3Hlfucose was examined. Although accumulation of both D-2- lation, also showeda decrease of 3040%. Gas chromatographic [2,6-3Hldeoxyglucose and ~-[5,6-~Hlfucose were decreased, the analysis of cell metabolites indicatedthat the intracellularcon30-60 L-fucose after this 1-hincubationis was consistently more pro- centration of effect on ~-2-[2,6-~H]deoxyglucose nmol/mg of protein, corresponding to an internal concentration nounced, making it less likely that the same transporter is being modified. The possibility does exist, however, that differ- of L-fucose of 4-5 mM. This suggests thatdown-regulation may ent amino acids of the same transporter are required for the be due to intracellular accumulationof L-fucose, following extransport of different sugars. For CME cells, 0.5 mM NEM posure of cells to a high nonphysiological concentration of Ldecreased ~-[5,6-~H]fucose accumulation34% to of control, and fucose, thus blocking L-fucose uptake. This would be consistent accumulation to 1% of con- with a passive facilitative diffusion transporter mechanismfor decreased ~-2-[2,6-~H]deoxyglucose trol. N-Bromosuccinimide, which modifies aromatic amino ac- L-fucose uptake. The accumulation thatoccurs in L-fucose-conids, decreased accumulation of both sugars equally t o 80% of ditioned cells is probably due tononfacilitated diffusion across control values a t a concentration of 0.5 mM. Since these experi- the cell membrane. This hypothesis is supportedby results of ments were run for 2-h intervals, it is likely that the amino experiments with CME cells in which increasing amounts of of ~-[5,6-~Hlfucose acid-modifying compounds were acting on a transport protein cold L-fucosecompete with constant amounts for uptake, reachinga limiting valueby 1mM, which represents and not on some downstream metabolic event. In addition, exposure of cells to NEM, followed bywashout of the NEM and the amountof L-fucoseaccumulated via nonfacilitated diffusion in I1 subsequentanalysis of ~-[5,6-~H]fucoseaccumulation,indi- (Fig. 1).These data are nearly identical to the data Table cated that a 1-min exposure t o NEM resulted in a decrease in and indicate that 28-52% of the L-fucose accumulated enters mammalian cells by trans-bilayer diffusion. ~-[5,6-~Hlfucose accumulation. nrnouer of L-Fucose-Shown in Fig. 2 are the results of a As shown in Table VI, cells conditioned in 10 mM L-fucose 50% of the amountof ~-[5,6-~Hlfucose as representative pulse-chase experiment performed in the four accumulated less than cells grown in normal media. Endothelial cells cultured in 1II~M cell lines under examination. Inall cell lines, accumulation of over a 24-h period proceeded in a linear L-fucose showed a similar “down-regulation.” In addition,CME 10 PM ~-[5,6-~H]fUcose

L-Fucose Dunsport in Eukaryotic Cells

22709

media was about 90% of the total radioactivity present. The amount of unbound ~-[5,6-~H]fucose in the chase medium did not appreciably change during the remainderof the chase period. After the initial 6-h phase, the decrease in intracellular ~-[5,6-~H]fucose protein content during the 48-h chase period can be accounted for by the continuous secretionof ~-[5,6-~HIfucosylated proteins into the medium (Fig. 3, compare secreted protein with proteinincorporation). DISCUSSION

L-FUC(pM) FIG.1. Quantitation of rate of nonfacilitative diffusion of Lfucose. Cells were incubated in medium containing 1 pCi/well ~-[5,63H]fucose and increasing amounts of unlabled L-fucose for 2 h. Cell suspensions were then obtained and analyzed for protein content and L-fucose accumulation as described previously (see “Experimental Procedures’’). Each value represents experiments performed in duplicate. Uptake is expressed as percent of uptake in the presence of 10 p unlabled L-fucose. Similar results were obtained in all cell lines.

v

MDCK

0 BAE v

CME NE

Y-

O

c

C

a,

2

”..T......

n“

v

”

~

-...___

~~

12

24

36

48

60

Time (h) FIG.2. Turnover of L-fucose.Cells were incubated in medium containing 1 pCi/well ~-[5,6-~Hlfucose for 24 h, followed by incubation in medium containing no L-fucose for an additional 2 4 8 h. Cell suspensions were obtained at time points throughout the pulse and chase periods and analyzed for protein content, L-fucose accumulation, protein incorporation, lipid incorporation, and protein secretion as described previously (see “Experimental Procedures”).Each value is representative of experiments performed three times, in duplicate. Accumulation is expressed as percent of accumulation at the end of the pulse.

fashion. Afterwards,resuspending thecells in mediumcontaining no ~-[5,6-~Hlfucose, the ~-[5,6-~Hlfucose content of the cells decreased in a biphasic fashion.There was an initial rapid loss of ~-[5,6-~Hlfucose for about the first 6 h of the chase period followed by a slower loss to 50% of the pulsevalue after 24 hof incubation. During the entireperiod of the pulse-chase study, incorporation into lipid didnot change (data notshown). Incorporationof ~-[5,6-~H]fucose into protein increased steadily with time during the 24-h pulse period (not shown) and decreased gradually during the chase period (Fig. 3). Protein secretion was observed by 1h into thepulse period (not shown) and continued throughout the entire chase period (Fig. 3). The initial and rapid loss of radioactivity during the first6 h of the chaseperiod appears to be due to releaseof free ~-[5,6-~Hlfucose,lipid as incorporation did not change, nor did protein secretion suddenly change (not shown). Furthermore, after 4 h of incubation in the chasemedium, the portion of unbound (non-trichloroacetic acid-precipitable) radioactivity derived from~-[5,6-~H]fucose present in

There are four general mechanisms by which cells take up molecules: pinocytosis, channels or pores, endocytosis, and carrier proteins. Generally, uptake of classes of molecules can be ascribed t o an uptakemechanism. Ions such as Na+ aretypical of channel accumulation, proteins such as insulin are taken up by receptor-mediated endocytosis, and smallorganic molecules such as sugars are taken by upcarrier proteins. Glucose, which serves as a point of comparison in these studies, is known to gain entry into cells by trans-bilayer nonfacilitated diffusion and by carrier proteins known as Glutl-Glut4(29). Pinocytosis as a mechanism for ~-[5,6-~H]fucose accumulation is ruled out by the saturability of L-fucose uptake. In addition, modification of amino acids resulting in a decrease in L-fucose accumulationcannot be explained in a pinocytosis model, given that it takes only 1 min of exposure to NEM t o decrease accumulation by40-50%. Endocytosis as a mechanism for L-fucose accumulation is ruled outby accumulation of at 4 “C and in the presence of colchicine and ~-[5,6-~H]fucose cytochalasin D. Stein (43) outlines seven criteria for the identification of facilitated diffusion systems. His first criterion is energy independence of transport. We do not believe this to be contradictory with our finding that DNP or IOA reduces accumulation substantially. The data presented Table in I11 were collected by pretreating thecells with DNP and then testing accumulation. When DNP or IOA and ~-[5,6-~H]fucose areadded simultaneously, this decrease is not initially observed (data notshown). The decrease in accumulation requires 25-30 min of incubation. We interpret this to mean that a depletion of the phosphorylation capability of cells results in intracellular L-fucose accumulation, thus blocking uptake since incorporation of Lfucose into protein, which is the major pathway ofL-fucose utilization, requires phosphorylation by L-fucokinase (28). Stein’s second criterion is that the rate of penetration is higher thanwould be expected based on its molecular size. As Stein points out, thiscriterion is somewhat subjective. Third is saturation of uptake, which we have shown, and the fourth is competition of uptake by structurally analogous molecules. We are somewhat restricted in the examination of this point, as there are relatively fewL-fucose-derived chemicals commercially available, but as shown in Table 11, this does occur to some extent. The fifth criterion is inhibition of uptake by nonstructurally analogous molecules. We have shown this in the case of cytochalasin B and NEM. The sixth is net transfer of permeant across the cell membrane, as we show in Table I. In addition, we show in Fig. 3 that efflux can occur. The seventh criterion is that transport islinked to other molecules in some cases. We have not rigorously attempted to determine whether this occurs. There is one additional, anecdotal line of evidence indicating that L-fucose may be accumulated via a carrier protein. The glucose transporters are conserved from bacteria t o mammals (34), and anL-fucose permease, responsible for L-fucose uptake in E. coli, has been cloned (44). The datashowing that L-fucose accumulation is decreased in cells containing a high intracellular level of L-fucose and that the L-fucose uptake is decreased by depletion of cellular phospho-

22710

L-Fucose llansport in Eukaryotic Cells 6000

N0

5000

I

CME

4000

FIG.3. Release of protein-bound Lfucose. Cells were incubated in medium containing 1 pCi/well ~-[5,6-~Hlfucose for 24 hfollowedby incubation in medium containing noL-fucosefor an additional 2-48 h. Cell suspensions were obtainedat timepointsthroughout the pulse and chase periods and analyzedfor protein incorporation,andmedium was analyzed for protein secretion as describedpreviously (see “ExperimentalProcedures”). Shown are representative values for experiments performed three times, in duplicate. Results are expressed as total counts/min present as protein incorporation (open symbols) and protein secretion (filled symbols). The abscissa in this figure represents the number of hours that cells were exposed to chase media.

2oool/ 1000 0‘ 0

I

I

10

20

30

3000

I

3000

2000 1000

I

40

I 50

0 0

10

20

30

40 01%

40

50

5000 I ...~

1

2500

I

MDCK

3000

1000

500

m’

0‘ 0

I

1

I

I

10

20

30

40

I 50

0’ 0

I

1

I

1

I

10

20

30

40

50

Time (h) rylation capability support a passive facilitated diffusion mechanism for L-fucose uptake by mammalian cells. In addition, under the conditions used in these studies 2040%of L-fucose accumulationmay be dueto nonfacilitated trans-bilayer diffusion. The inhibition ofL-fucose uptake by the same compounds that inhibit glucose uptake suggests that the L-fucose transporter may be similar in structuret o the glucose transporters. This hypothesis would be supported by the recent discovery that fructose is takenup by the “Glut5” transporter (45). Though some data presented in this report could be construed as suggestingL-fucose is taken up by a glucose transporter (e.g. inhibition of uptake by cytochalasin B), the overwhelming majority would rule out this possibility. Most cultured cells contain only Glutl (461, and Glutldid not transport L-glucose or other levorotatorymonosaccharideswhenexpressed in Xenopus oocytes (47). The regulation of o-glucose and L-fucose uptake by insulin and PMA were different. Amino acid modification experiments and kinetic parameters indicated differing transporter sensitivities. Theobserved K, values, coupled with the observations that L-fucose uptakeissaturated by 200 PM, whereas D-glucose uptake is not saturated until20 mM, provide additional evidence that L-fucose is not taken upby a glucose transport system, as the rateof D-glucose uptake would obliterate any L-fucose uptake in competition experiments. Moreover, since the glucose transporter systems have been shown not to transportlevorotatory sugars, theK, values for L-fucose if it were being transcertainly would not be in the range ported by a glucose transporter. Taken together, these data provide considerable evidence that L-fucose is not taken up by a glucose transporter. However, the only direct proof would be purification or cloning of the L-fucose transporter, which awaits further studies. The similarity of the quantity of L-fucose taken up and the sensitivity to inhibitors would suggest that a similar transporter is probably present in all four cell lines, but the variations in percent inhibitions thatwere seen areconsistent with the possibility that the transporter variessome in way between

cells. The metabolism of L-fucose is remarkably similar in the four widely varying cell lines, also indicating a similar transport mechanism and enzymology of metabolism. To our knowledge, the secretion of L-fucose-containing proteins hasnot beendescribed for the cell types we have examined. Obviously secretion must occur from hepatocytes since serum L-fucose-containing proteins exist; however, it is somewhat surprising that endothelial and epithelialcells would secrete such proteins. We have not yet determined the identity of these proteins or whether the proteins present intracellularly are the same as those secreted but still in theprocess of being synthesized. Until the identityis known, no speculation can be made concerning the physiological relevance of this observation. The K, for L-fucose uptake obtained in these studies is in the physiological concentration range ofL-fucose. There is good agreement in the literature that the serum protein-bound Lfucose level is under 1mM (5-7,11,12,48). Serumvalues of free L-fucose are not in as good agreement; infour separate studies (11, 49-51), the average values of serum-free L-fucose varied from 60 to 2.2 m~ for normal subjects. Thus, the potential for L-fucose uptake by these cells to be physiologically relevant is quite good. In summary, L-fucose uptake by mammalian cells seemed to occur by a specific, facilitative diffusion transport system, and a large portion of the L-fucose taken upby these cells was used in glycoprotein synthesis. Alargeportion of these glycoproteins were secreted into the media. REFERENCES 1. Flowers, H. (1981)Adu. Carbohydc Chem. Biochem. 3 9 , 2 7 9 3 4 5 2. Thompson, S., Dargan, E., and Turner, G. A. (1992) Cancer Lett. 6 6 , 4 3 4 8 3. Shirahama, T., Ikama, M., Muramatsu, H., Muramatsu, T., and Ohi, Y. (1992) J . Urol. 147, 1659-1664 4. Ghosh, M., and Nayak, B. R. (1991) Ann. Dentistry 60,33-35 5. Macheth, R. A., and McBride, G. (1965) Cancer Res. 25, 1779-1780 6. Rosato, F. E., and Seltzer, M. H. (1969) Am. J . Surg. 1123, 6 1 4 4 7. Hadjivassilious, A,, Castanaki, A,, Hristou, G . , and Lissaios, B. (1975) Sorg. Gynecol. & Obstet. 140, 239-240 8. Sakal, T., Yamamoto, K., Yokota, H., Hakozaki-Usui, K., Hino, F., and Kato, I. (1990) Clin. Chen. 36,474-476

L-Fucose Dansport in Eukaryotic Cells ..

10.Yorek,M.A,, Dunlap, J. A., Stefani, M. R., and Davidson, E. P. (1992) J. Neurochem. 68, 1626-1636 11. Radhakrishnamurthy, B., Berenson, G. S., Parfaonkar, P. S., Voors, A.W., Srinivasan, S. R.,Plavidal, F., Dolan, P., and Dalferes, E. R. (1976) tab. Inuest. 34, 159-165 12. McMillan, D. E. (1972) Diabetes 21, 863-871 13. Stefani, M.R., Dunlap, J.A., and Yorek,M.A. (1993) J. Cell. Physiol. 163, 321-331 14. Yorek, M.A., Wiese,T. J.,Davidson, E. P., Dunlap, J. A., Stefani, M. R., Comer, C. E., Lattimer, S. A,, Kamijo, M., Greene, D. A,, and Sima, A. A. F. (1993) Diabetes 42, 1401-1406 15. Winegrad, A. I. (1987) Diabetes 36, 39&406 16. Lehrman, M. A., Pizzo, S. V., Imber, M. J., and Hill, R. L. (1986)J. Biol. Chem. 261, 7412-7418 17. Lehrman, M. A,, and Hill, R. L. (1986)J. Biol. Chem. 261, 7419-7425 18. Lehrman, M. A,, Haltiwanger, R. S., and Hill, R. L. (1986)J. B i d . Chem. 261, 7426-7432 19. Haltiwanger, R. S., Lehrman, M. A,, Eckhardt, A. E., and Hill, R. L. (1986)J. B i d . Chem. 261,7433-7439 20. Foster, D. W., and Ginsburg, V. (1961)Biochirn. Biophys. Acta 64,376-378 21. Nwokoro, N. A., and Schachter, H. (1975) J. Eiol. Chern. 260, 6191-6196 22. Nwokoro. N. A,, and Schachter, H. (1975) J. Biol. Chem. 260, 6185-6190 23. Chang, S., Duen; B., and Serif, G. (1988) J. Biol. Chem. 26.3, 1693-1697 24. Overton, K., and Serif, G. S. (1981) Biochim. Biophys. Acta 676,281-284 25. Broschat, K. O., Chang, S., and Serif, G. S. (1985) Eur J. Biochem. 163, 397401 26. Yuen, R., and Schachter, H. (1972) Can. J. Biochem. 60, 798-806 27. Grove, D. S., and Serif, G. S. (1981) Eiochim. Biophys. Acta 662,246255 28. Richards, W. L., and Serif, G. S. (1977) Biochim. Biophys. Acta 484, 353-367 29. Bell, G. I., Burant, C. F., Takeda, J., and Gould, G.W. (1993) J. Biol. Chem. 268, 19161-19164 30. Lees, M. B., and Paxman, S. (1972)Anal. Biochem. 47, 184-192 31. Hopkins, C. R., and Trowbridge, I. S. (1983) J. Cell B i d . 97, 508-521

22711 ..

John Wiley & Sons, Inc., New York 33. Yang, J., Clark, A. E., Kozka, I. J., Cushman, S. W., and Holman, G. D. (1992) J . Biol. Chem. 267, 10393-10399 34. Carmthers, A. (1990)Physiol. Rev. 70, 1135-1176 35. Gottlieb, T.A., Ivanov, I. E., Adesnik, M., and Sabatini, D. D. (1993) J, Cell Bid. 120, 695-710 36. Gronowicz, G., Masur, S. K , and Holtman, E. (1980) J . Memb. Biol. 63, 221235 37. Gemtsen, M. E., Burke, T. M., and Allen, L. A. (1988) Microvascular Res. 36, 153-166 38. King, G. L., Buzney,S. M., Kahn, C. R., Hetu, N., Buchwald, S., MacDonald, S. G., and Rand, L.I. (1983) J. Clin. Invest. 71, 974-979 39. Lotan, R. M., and Barzilai, D. (1990)Isr J. Med. Sci. 26, 5-11 40. Hauguel-de Mouzon, S., and Kahn, C. R. (1991) Mol. Endocrinol. 6,51-60 41. Longo,N., Griffin, L. D.,Langley, -~ S. D., and Elsas, L. J. (1992) Biochim. Biophys. Acta 1104,24-30 42. Naftalin, R. J., and Rist, R. J. (1989) Biochem. J . 260, 143-152 43. Stein, W. D. (1967) The Movement of Molecules across Cell Membranes, Academic Press, New York 44. Lu, Z., and Lin, E. C. C. (1989) Nucleic Acids Res. 17, 4883-4884 45. Burant, C. E , Takeda, J., Brot-Laroche, E., Bell, G. I., and Davidson, N. 0. (1992) J. Biol. Chem. 267, 14523-14526 46. Kaiser, N., Sasson, S., Feener, E. I?,Boukobza-vardi, N., Higashi, S., Moller, D. E., Davidheiser, S., Przybylski, R. J., and King, G. L. (1993) Diabetes 42, 80-89 47. Gould, G. W., Thomas, H. M.,Jess, T.J., and Bell, G. I. (1991)Biochemistry30, 5139-5145 48. Spellacy, W. N., Buhi, W. C., G., Staba, M. A,, and Birk, S. A. (1970) Obstet. Gynecol. 36, 3 9 4 3 49. Sirakov, L.M. (1971)Acta Diabetol. Lat. 8, 949-956 50. Gallardo, J. T., Lagunes, R. M. L., Castillo, J. S., and Aquime Gas, A. S. (1977) Prensa Med. Mex. 42, 120-126 51. Rahman, M. A., Zafar, G., and Shera, A. S. (1990)Biomed. Pharmacother: 44, 229-234