THE JOWAL OF B I O ~ I C A L CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 15, Issue of April 15, pp. 11578-11583, 1994 Printed in U S A .

Substitution of Tyrosine 293 of GLUTl Locks the Transporter into an Outward Facing Conformation*

(Received for publication, October 12, 1993, and in revised form, January 10, 1994)

Hiroyuki MoriS, Mitsuru HashiramotoS§, Avril E. Clarkn, Jing Yangn, Akihiro MuraokaS, Yoshikazu TamoriS, Masato KasugaS, and Geoffrey D. Holmann From the $Second Department of Internal Medicine, Kobe University School of Medicine, 7-5-1 Kusunoki-Cho, Chuo-Ku, Kobe 650, Japan and the Wepartment of Biochemistry, University of Bath, Claverton Down, Bath BA2 7Ak: United Kingdom

Tyrosines 292 and 293 in the mammalian glucose transporter GLUTl have been substituted by either iso- leucine or phenylalanine. Chinese hamster ovary clones that were transfected with Tyr-292 -* Ile, Tyr-292 -* Phe, Tyr-293 -* Ile, and Tyr-293 -* Phe constructs of GLUTl were shown, by Western blotting and cell surface carbo- hydrate labeling, to have expression levels that were comparable with the wild type. The V,, for 2-deoxy-~- glucose transport was markedly reduced only as a result of the Tyr-293 -* ne mutation. The ability of the Tyr-293 -* Ile mutated GLUTl to bind the exofacial ligand 2-~-4- (1-azi-2,2,2-trifluoroethyl)benzoyl-1,3-bis-(~-mannos-4- yloxy)-2-propylamine (ATB-BMPA) and the endofacial li- gand cytochalasin B were assessed by photolabeling procedures. The ability to bind the bis-mannose com- pound was unimpaired, whereas the ability to bind cy- tochalasin B was totally abolished, and the level of la- beling was lower than in the nontransfected clone. Affinities of the wild-type and Tyr-293 -+ Ile GLUTl for D-glucose, the exofacial ligands (ATB-BMPA and 4,6-0- ethylidene-D-glucose), and the endofacial ligand (cy- tochalasin B) were assessed by the ability of these agents to displace the radioactive ATB-BMPA photola- bel. These data indicated that the Tyr-293 - Ile substi- tution produced no change in the affinity for D-glucose, a relatively small enhancement in the affinity for exofa- cial ligands, but a large 400-fold reduction in affinity for cytochalasin B, suggesting that the mutated GLUTl is locked in an outward facing conformation. The obser- vation that the ”yr-293 -* Ile mutant transporter can bind nontransported C4 and C6 substituted hexose ana- logues but cannot catalyze transport is interpreted as indicating that Tyr-293 is involved in closing the exofa- cial site around C4 and C6 of D-glucose in the transport catalysis process.

The mechanism by which glucose transporters catalyze glu- cose transport is thought to involve alternating conformational changes that sequentially expose the hexose binding site to the external and internal surfaces of the protein. A combination of ligand binding and site-directed mutagenesis studies have been used to investigate the protein domains responsible for this mechanism. Photolabeling studies involving active site ligands

* This work was supported by the Monbusho International Research Program: Joint Research and by grants from the Otsuka Pharmaceuti- cal Co., Ltd. for diabetes research (to M. K.) and Science and Engineer- ing Research Council (United Kingdom) (to G. D. H.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

0 To whom correspondence and reprint requests should be addressed. Fax: 078-382-2080.

suggest that the bis-mannose compound, ATB-BMPA1 (an exo- facial ligand), cytochalasin B (an endofacial ligand), and 3-iOdO- 4-azidophenethylamino-7-O-succinyldeacetyl-forskolin inter- act only with the C-terminal half of GLUTl (in transmembrane helices (TM) 7-11) (Holman and Rees, 1987; Cairns et al., 1987; Wadzinski et al., 1988). Domain assembly experiments suggest that the N- and C-terminal halves of sugar transporters are both necessary for function (Kaback, 1992).’ The N-terminal half of the GLUTl may be responsible for providing a packing surface for stabilizing the C-terminal half in a ligand binding conformation. A reduction in exofacial ATB-BMPA binding and poor cell surface expression (Asano et al., 1991,1993) have been observed in mutants in which the glycosylated amino acid Asn- 45, in the N-terminal half of GLUT1, is substituted. Experi- ments in which a C-terminally truncated GLUTl is expressed in CHO cells suggest that the C-terminal region is required for exposing the exofacial ATB-BMPA binding site. C-terminal truncation produces a locking of GLUTl into a stable inwardly directed conformation3 (Oka et al . , 1990).

Mutagenesis has been carried out on Pro-385 in TM 10 (Ta- mori et al., 1994). Reduction in transport and exofacial ATB- BMPA photolabeling occur as the result of substitution with isoleucine but not with glycine at this position. These results suggest that retention of flexibility in this region is required for the transport catalysis process. This study (Tamori et al., 1994) and molecular dynamic simulations of a model of the three- dimensional structure of GLUTL4 Gould and Holman (1993) have suggested that the proline- and glycine-rich region in TM 10 provides a pivotal point for a conformational change, which allows TM 11 and TM 12 to either pack against the outside (ATB-BMPA binding) site in TM 7-9 or against the inner (cy- tochalasin B binding) site at the base of TM 10.

Mutagenesis studies have been used for examining amino acids that may be important in binding at the endofacial (cy- tochalasin B/forskolin binding) site. Tryptophan fluorescence changes occur on the binding of these ligands (Appleman and Lienhard, 1985; Carruthers, 19861, and tryptophan residues have been postulated to be involved in the photochemical la- beling mechanism (Deziel et al., 1984; Cairns et al., 1987). Therefore, tryptophan 412 and 388 have been mutated. Muta- tion of Trp-412 in TM 11 of GLUTl results in reduced transport

The abbreviations used are: ATB-BMPA, 2-~-4-(l-azi-2,2,2-trifluoro- ethyl)benzoyl-l,3-bis(o-mannos-4-yloxy)-2-propylamine; TM, trans- membrane segment; PBS, phosphate-buffered saline; CHO, Chinese hamster ovary.

D. L. Cope, G. D. Holman, and A. J. Wolstenholme, unpublished results.

A. Muraoka, M. Hashiramoto, A. E. Clark, T. Kadowaki, H. Sakura, G. D. Holman, and M. Kasuga, unpublished results.

P. A. Hodgson, D. J. Osguthorpe, and G. D. Holman, unpublished results.

11578

97K -

66K -

45K -

11580 Functional Role of Qr-293 in GLUTl TABLE I

Analysis of the expression, transport actiuity, and ligand labeling of GLUTl tyrosine 292 and 293 mutants Results are the mean a S.E. of the indicated number of experiments except where the results are from two separate experiments, as noted.

CHO-K1 Wild type Y292F Y292I Y293F Y293I

Western blotting (% of wild type; n = 3) 14.4 k 2.2 100 100.3 29.7 129.0 a 35.3 97.4 f 8.9 134.8 t 3.4 Borohydride labeling (% of wild type; n = 4) 14.8 k 3.4 100 Transport K,,, (mM; n = 3) 2.6 (3.4, 1.8)” 1.9820.1 2.46k0.32 2.17a0.33 2.0t0.12 1.38a0.15

133.5 a 5.7 131.9 f 31.0 137.3 k 18.5 112.7 t 29.7

ATB-BMPA labeling (% of wild type; n = 3) Transport Vmmb (nmolhnidmg of protein; n = 3) 7.1 (7.0, 7.2)“ 74.9 a 18.2 72.7 a 23.6 75.1 a 19.2 83.5 t 11.7 10.3 k 2.8

16.2 a 5.3 100 147.6 a 60.7 118.7 k 15.8 105.8 f 27.1 126.2 a 34.9 Cytochalasin B labeling (% of wild type; n = 4) 18.2 a 2.3 100 103.5 a 8.6 53.8 a 7.3 107.3 f 8.1 4.5 a 2.1

a Results are from two separate experiments.

does not alter any of the conclusions reached in that study. The V,, in our previous study (Hashiramoto et al., 1992) is incorrect due to an error in calculation of 2-deoxy-~-glucose specific activity. This

RPR-100 photochemical reactor with 300-nm lamps. Following irradia- tion, the dishes were washed five times in PBS, directly solubilized in SDS-electrophoresis sample buffer, and subjected to electrophoresis on 10% acrylamide gels. The radioactivity associated with the gel peaks was determined as described above for the sodium borohydride labeling procedure. The cytochalasin B labeling procedure was a modification of the previously described procedure (Hashiramoto et al . , 1992; Tamori et al . , 1994) in which plasma membrane fractions were isolated and la- beled. The direct labeling of intact cells and of membrane fractions was compared for the clones described, and the two procedures gave essen- tially the same result.

RESULTS The four mutant GLUTl cDNAs encoding Tyr-292 + Phe

(Y292F), Tyr-293 --j Ile (Y2921), Tyr-293 + Phe (Y293F), and Tyr-293 - Ile (Y2931) were transfected into CHO-K1 cell lines by the calcium phosphate method, and the stable transfor- mants were selected by their resistance to neomycin. Western blot analysis of crude membranes isolated from the neomycin- resistant clones using an anti-peptide antibody directed against residues 478492 of the C terminus of GLUTl enabled clones that had levels of expression comparable with the wild- type clone previously isolated to be selected (Hashiramoto et al., 1992) (Fig. 1). Using Western blot analysis, the expression levels in these clones were shown to be -7-fold higher than that observed for a nontransfected CHO-K1 clone. The levels of ex- pression were calculated as 100.3 2 29.7% (n = 3) for Y292F, 129.0 f 35.3% (n = 3) for Y2921,97.4 ? 8.9% (n = 3) for Y293F, and 134.8 f 3.4% (n = 3) for Y293I of that occurring in the wild-type clone.

The substitution of tyrosine 292 by neither isoleucine nor phenylalanine altered cell surface labeling using sodium boro- hydride, 2-deoxy-~-glucose transport activity, or the labeling by ATB-BMPA. In comparison with the wild-type GLUT1, the Y292I GLUTl showed slightly reduced cytochalasin B labeling. (Table I).

The results of cell surface labeling the Y293I and Y293F clones using sodium borohydride are compared with the label- ing of the wild-type and the nontransfected clones in Fig. 2. Using this technique, the Tyr-293 mutant transporters were confirmed to be expressed at the cell surface to levels of 112.7 2 29.7% (n = 4) for Y293I and 137.3 2 18.5% (n = 4) for Y293F, respectively, of that occurring in the wild-type clone. By con- trast, the CHO-K1 clone was labeled to a level that was only 14.8 2 3.4% (n = 4) of the wild-type clone.

2-Deoxy-~-ghcose uptake measurements in the Y293I clone showed that a marked reduction in the V,, for transport that was only 14% of that observed in the wild type (Fig. 3). By contrast, the Y293F clone retained a relatively high level of transport activity. The V,, was similar to that of the wild-type clone (Table I).

We initially assessed the ability of the Tyr-293 mutated clones to bind exofacial and endofacial ligands using photola- beling procedures. As shown in Fig. 4A, both the Y293I and Y293F mutant GLUTls were efficiently labeled with ATB-

12000 r 10000 I. /I

0

0 2 4 6 8 1 0

Gel Slice Number

FIG. 2. Cell surface carbohydrate labeling of mutated GLUTl transporters. Cells of the wild-type (A), nontransfeded (A), Y293I (O), and Y293F (0) CHO-K1 clones were grown to confluence in 35-mm

neuraminidase at 18 “C, and then 1 mCi of tritiated sodium borohydride dishes. Cells were simultaneously treated with galactose oxidase and

was added. The cells were maintained at 18 “C for 10 min, washed four times in PBS, and then solubilized in detergent buffer. Immunoprecipi- tation of the labeled transporters was then carried out using C-terminal antiserum. The labeled proteins were then analyzed by electrophoresis. The positions of the radioactive peaks in the gel lanes were compared with molecular mass markers (arrowheads). Results are from an ex- periment representative of four similar determinations (see “Results”).

BMPA to levels of 126.2 2 34.9% (n = 3) and 105 27.1% (n = 31, respectively, of that obtained in the wild-type clone. Strik- ingly, the Y293I mutant transporter was labeled with cytocha- lasin B to a level of only 4.5 2 2.1% (n = 4) of the wild type and at a level that was lower than that observed in the nontrans- fected clone and that was 18.2 f 2.3% (n = 4) of the level for the wild type (Fig. 4B). The levels of cytochalasin B labeling ob- served in the Y293F clone were 107.3 2 8.1% (n = 4) of the wild-type clone.

To further characterize the changes in affinity for side-spe- cific ligands that occur in the Y293I mutant GLUT1, we have determined the extent of displacement of the ATB[2-3HlBMPA by either exofacial or endofacial ligands. Fig. 5, A and B , shows the displacement of the tritiated photolabel from the wild-type and Y293I mutated GLUTl by a range of concentrations of nonlabeled ATB-BMPA and 4,6-O-ethylidene-~-glucose, respec- tively. The affinity constant or K, values are calculated by fit- ting,

LJL = 1 + IIK, (Eq. 1)

where Lo and L are the levels of labeling observed without and with the competing ligand ( I ) . The K, values in the wild-type and Y293I mutated GLUTl are 8.5 2 0.5 n m and 3.0 f 0.2 mM for 4,6-O-ethylidene-~-glucose and 118.8 2 3.8 1.1~ and 80.0 f 4.3 1.1~ for ATB-BMPA, respectively, indicating an -2-fold higher affinity of the mutated GLUTl for these exofacial ligands.

Functional Role of 5r-293 in GLUTl 11581

Fig. 6, A and B , compares the displacement of ATB[2- 3H]BMPA by D-glucose and cytochalasin B in the wild-type and Y293I mutated GLUT1. The competition between 10 mM D-

glucose and ATB[2-3H]BMPA is similar in the wild-type and Y293I mutated GLUT1. By contrast, a range of concentrations of cytochalasin B that markedly inhibit ATB[2-3H]BMPA pho- tolabeling in the wild-type GLUTl has virtually no inhibitory effect on the extent of labeling in the Y293I mutated GLUT1. The approximate Ki for cytochalasin B in the wild-type and Y293I mutated GLUTl are 0.5 p~ and 144 p ~ , respectively. These values are only approximations, because in neither case was the range of concentrations of cytochalasin B ideal for

1-5 r Y2931 mutated GLUTl

1 .o

0.5

0

I/ Wild-type GLUTl

0 1 2 3 4 5

S (rnM) FIG. 3. Kinetic analysis of 2-deoxy-o-glucose transport. The

wild-type (0) and Y293I mutated (A) GLUTl were analyzed for trans- port activity as described in "Experimental Procedures." The results are from a representative experiment. The results of three kinetic experi- ments for each of the mutated GLUTl clones are shown in Table I. The kinetic parameters were calculated from least squares regression of the Michaelis-Menten equation.

A

50000

40000

0 2 30000 E \

n E 20000

10000

0

66KD1 45KDl v v

0 2 4 6 8 1 0

Gel Slice Number

determining the affinity constant. Nevertheless, the Ki for cy- tochalasin B is at least 250-fold higher in the Y293I mutated GLUT1.

A simple analysis (Stein, 1986; Clark and Holman, 1990) suggests that the K, for binding exofacial and endofacial li- gands are determined according to Eq. 2,

OUtK, = OUtKd (1 +A); '"K, = inKd (1 + UA) (Eq. 2)

where Kd are the dissociation constants, and A is the ratio of distribution of alternately exposed sites at inside compared with outside surfaces. If, in the wild-type clone, the value of A is approximately unity, i.e. if the alternately exposed binding sites are equally available at inside and outside surfaces, then it follows that the effects of mutagenesis of Tyr-293 could be entirely due to a redistribution of available sites from inside to outside so that A is very low. This explanation would account for the small -2-fold decrease in OUtK, (the 4,6-O-ethylidene-~- glucose and ATB-BMPA affinity constants) and a large increase in '"K, (the cytochalasin B affinity constant).

DISCUSSION

D-Glucose analogues in which bulky alkyl groups have been substituted have provided information on the spatial require- ments for hexose binding at the outer and inner sites of the hexose transporters, GLUTl and GLUT4 (Barnett et al., 1973a, 1973b, 1975; Holman and Rees, 1982). Substitutions into the C-1 position of D-glucose markedly reduce interaction with the exofacial site, suggesting that there is a close approach to the front of the glucose molecule as it interacts at the outer binding site. By contrast, analogues with hydrophobic alkyl substitu- tions into the C-4 and C-6 positions of D-glucose were found to have high affinity for the transporter, in some cases exceeding that of D-glucose. The high affinity of the hydrophobic C-6 sub- stituted analogues led to the suggestion that there is a hydro- phobic patch on the transporter, which is close to the C-6 of the sugar as it locates in the exofacial site. Although binding well to the exofacial site, the C4 and C6 substituted analogues were not transported substrates. Therefore, it was suggested that on initial binding of the sugar, there was not a close approach to the C4 and C6 positions but that in the absence of a bulky substitution at C-6, the transporter could close the hydrophobic patch around the back of the sugar and so facilitate the trans-

B

1200

600 0 .- a \

400 f

0

0 2 4 6 8 10

Gel Slice Number

andY293F (0) clones in 35-mm dishes were labeled with 100 pCi ofATB[2-'HlBMPA(A) or 1.4 pCi of [4-'H]cytochalasin B with 100 p cytochalasin FIG. 4. Photolabeling of -293 -t Ile and Tyr-293 + Phe substituted GLUTl. Cells of the wild-type (A), nontransfected (A), Y293I (O),

E ( B ) in 250 pl PBS at 18 "C by irradiation for 1 min in a Rayonet photochemical reactor. Following irradiation, the cells were washed five times in PBS and directly solubilized in SDS-electrophoresis sample buffer and subjected to electrophoresis. The positions of the radioactive peaks were compared with molecular mass markers (arrowheads). Results are from an experiment representative of three similar determinations.

11582 Functional Role of fir-293 in GLUTl

A B

6 - 12 - 5 . 10 .

4 - 8 .

A \

-I o 6 -

0 0

0 100 200 300 400 0 5 10 15 20 25 30

ATE-BMPA lpM) 4-6-0-sthylidene-D-glucose ImM)

FIG. 5. Exofacial displacement of ATB[2-'H]BMF'A photolabel in wild-type and -293 + ne GLUT1. Cells of the wild-type (0) and Y293I (A) clones were grown to confluence in 35-mm dishes and then labeled with 100 pCi of ATB[2-3HlBMPA in the presence of increasing concentrations of either nonlabeled ATB-BMPA (A) or 4,6-O-ethylidene-~-glucose (B). Following irradiation, the cells were washed five times in PBS and directly solubilized in SDS-electrophoresis sample buffer and subjected to electrophoresis on 10% acrylamide gels. The total radioactivity associated with the labeled transporter peak either in the presence (L) or absence (Lo) of competing ligand were calculated. Results are the mean of two similar experiments. K, values were calculated by least squares regression of the equation LIL, = 1 + UK,, where I is the concentration of displacing ligand.

A. B

1.0 -

0.8 .

lpM 5pM 20pM lOmM 50mM lpM 5pM 20pM 10mM 5OmM

CYTOCHALASIN B D-GLUCOSE CYTOCHALASIN B D-GLUCOSE

FIG. 6. Glucose and cytochalasin B displacement of ATB[2-'HIBMPA photolabel in wild-type and '&~293 + ne GLUT1. Cells of the wild-type (A) and Y293I (B) clones were grown to confluence in 35-mm dishes and then labeled with 100 pCi of ATB[2-3HlBMPA in the presence of increasing concentrations of either D-glucose or cytochalasin B. Following irradiation, the cells were washed five times in PBS and directly solubilized in SDS-electrophoresis sample buffer and subjected to electrophoresis on 10% acrylamide gels. The total radioactivity associated with the labeled transporter peak either in the presence (L) or absence (Lo) of competing ligand were calculated. Results are the mean of two similar experiments, except the D-glucose displacement from the wild-type GLUTl are the mean S.E. of three experiments. K, values were calculated by least squares regression of the equation LIL, = 1 + UKi, where I is the concentration of displacing ligand.

port catalysis process. There is a relatively minor perturbation of cytochalasin B

labeling in the Y292I mutated GLUT1, but the cytochalasin B labeling is totally abolished as a result of the Y293I substitu- tion. We interpret these data as indicating that the residues F291, Y292, and Y293 constitute part of the hydrophobic patch, previously postulated from studies involving hexose analogues (Fig. 7). Size and hydrophobicity of the tyrosine residue are both important properties, as the relatively small but hydrophobic isoleucine group is insuficient to allow the normal conformational change that accompanies the transport catalysis process. The bulky phenylalanine substitution does, however, allow this change. If the residues of this hydrophobic patch move around the back of the sugar to separate it from the external solution, then this process may occlude the sugar within the center of the protein and strengthen the interac- tion with hydrogen bonding amino acid side chains in this re- gion. The sequence Q282QLSGINAVFYY93 in TM 7 of GLUTl is therefore suggested to have a role in both hydrogen bonding

to D-glucose and closing the exofacial site during the transport catalysis process (Fig. 7). This interpretation is consistent with the observation that substitution of Q282 by leucine abolishes ATB-BMPA binding. The glutamines in this region may be responsible for providing some of the hydrogen bond- ing groups to the C1, C3, and ring-oxygen positions that pre- vious studies on fluoro- and deoxy-hexose analogue have sug- gested are the hydrogen bonding positions on the sugar (Barnett et al., 1973a).

The Y293I mutant GLUTl shows a decrease in glucose trans- port activity that is ,associated with a complete loss of affinity for the inside specific ligand cytochalasin B but a retention of high affinity for the outside specific ligand ATB-BMPA. These results are in marked contrast with those observed in C-termi- nally truncated GLUT1, where the decreased glucose transport activity is associated with corresponding loss of labeling by ATB-BMPA but with retention of high affinity for cytochalasin B (Oka et al., 1990). Oka et al. (1990) interpreted these results as indicating that the truncated GLUTl is locked into an in-

Functional Role of Qr-293 in GLUTl 11583

6 H

, FIG. 7. Proposed role of QT-ZSS in closing the exofacial hexose

binding site. The sequence Q2"QLSGINAV"lTg3 in TM 7 of GLUTl is suggested to have a role both in hydrogen bonding to o-glucose and closing the exofacial site necessary for facilitating the occlusion of this transported sugar. Only the hydrogen bonding is necessary for interac- tion with the exofacial ligands ATB-BMPA and 4,6-O-ethylidene-wglu- cose, which are substituted in the C4 and C6 positions of the hexopyr- anose ring. Replacement of the critical Tyr-293 prevents the conformational change necessary for transport catalysis and locks the transporter into an outward facing conformation.

ward facing conformation. By contrast, the results presented here suggest that the Y293I transporter is locked in an outward facing conformation.

From a combination of ligand binding and mutagenesis stud- ies, it is now apparent that conformational flexibility is neces- sary for the transport catalysis process. The glycine- and pro- line-rich region of TM 10 may act as a pivotal point in this process that alternately allows opening and closing of the outer and inner sites (Tamori et al., 1994). Nontransported ligands may stabilize the outward and inward directed conformations (Barnett et al, 1973b; Holman and Fkes, 1982, 1987), but mu- tagenesis and proteolytic enzymes can also lock transporters in these stable conformations (Oka et al., 1990; Clark and Hol- man, 1990). It is particularly striking that a completely stable and outward directed conformation can be induced by just a single amino acid substitution, Tyr + Ile a t position 293.

REFERENCES

Asano, T., Katagiri, H., Takata, K, Lin, J-L., Ishihara, H., Inukai, K, Tsukuda, K, Appleman J. R., and Lienhard, G. E. (1985) J. Biol. Chern. 260,45754778

Kikuchi. M., Hirano, H., Yazaki. Y., and Oka, Y. (1991) J. Biol. Chern. 266, 24632-24636

Asano, T., Takata, K., Katagiri, H., Ishihara, H., Inukai, K., Anai, M., Hirano, H., Yazaki, Y.. and Oka, Y. (1993) FEBS Lett. 324,258-261

Barnett. J. E. G., Holman, G. D., and Munday. K. A. (1973a) Biochern. J. 131,

Barnett, J. E. G., Holman, G. D.. and Munday, K A. (1973b) Biochern. J. 136, 211-221

539-541 Barnett, J. E. G., Holman, G. D., Chalkley, R. A, and Munday, K. A. (1975)

Biochern. J. 145,417-429 Cairns, M. T., Alvarez, J., Panico, M., Gibbs, A. F., Moms, H. R., Chapman, D., and

Carruthers,k (1986) J. Biol. Chern. 261,11028-11037 Baldwin. S. A. (1987) Biochirn. Biophys. Acta 906,295-310

Clark, A. E., and Holman, G. D. (1990) Biochern. J. 269,615-622 Deziel, M. R., Pegg, W.. Mack, E., Rothstein, A., and Klip, A. (1984) Biochirn.

Garcia, J. C., Strube, M., Leingang, K, Keller, K, and Mueckler, M. M. (1992) J.

Gould, G. W., and Holman, G. D. (1993) Eiochern J. 296,329-341 Hashiramoto, M., Kadowaki, T., C1ark.A. E., Muraoka, A,. Momomura, K. Sakura,

Biol. Chern. 267, 17502-17507 H.. Tobe, K., Akanuma, Y., Yazaki, Y., Holman, G . D., and Kasuga, M. (1992) J.

Biophys. Acta 772,403-406

Biol. Chern. 267,7770-7776

Henderson, P. J. F. (1990) Res. Micrubiol. 14.316328 Holman, G. D., and Rees, W. D. (1982) Biochirn. Biophys. Acta 686.78-86 Holman, G. D., and Rees, W. D. (1987) Biochirn. Biophys. Acta 897,395-405 Kaback, H. R. (1992) Int. Rev. Cytol. 1374 97-125 Katagiri, H., Asano, T., Ishihara, H., Tsukuda, K, Lin, L-J., Inukai, K, Kikuchi,

M., Yazaki, Y., and Oka, Y. (1992) J. Biol. Chern. 267,22550-22555 Katagiri. H., Asano, T., Ishihara, H., Lin, J-L., Inukai, K, Shanahan, M. F.,

Tsukuda, K., Kikuchi, M., Yazaki, Y., and Oka, Y. (1993) Biochern. J. 291, 861-867

Kunkel, T. A. (1985) P m . Natl. Acad. Sci. U. S. A. 82,488492 Mueckler, M. M., Caruso, C., Baldwin, S. A, Panico, M., Blench, I., Morris, R. H.,

Oka, Y., Asano, T., Shiba8aki.Y.. Lin, J-L.. Tsukuda, K, Katagiri, H., Akanuma, Y.,

Schiirmann. A., Keller, IC, Monden. I., Brown, F. M., Wandel, S., Shanahan, M. F.,

Stein, W. D. (1986) in 7hznsport andDiflusion acruss Cell Membmnes, pp. 231461,

Tamori, Y., Hashiramoto, M., Clark, A. E., Mori, H.. Muraoka. A,, Kadowaki, T.,

Wadzinski, B. E., Shanahan, M. P., Clark, R. B., and Ruoho, A. E. (1988) Biochern.

Wellner, M., Monden, I., and Keller, K. (1992) FEES Lett. 309,293-296

Allard, W. J.. Lienhard. G. E., and Lodish, H. F. (1985) Science 229,941-945

and Takaku, F. (1990) Nature 346,227-233

and Joost, H. G . (1993) Biochern. J. 290,497401

Academic Press, Inc., London

Holman, G. D., and Kasuga, M. (1994) J. Biol. Chern. 269,2982-2986

J. 255,983-990