Identification of an Amino Acid Residue That Lies between the Exofacial Vestibule and Exofacial Substrate-binding Site of the Glut1 Sugar Permeation Pathway

A valine-to-isoleucine mutation at amino acid residue 197 of Glut2 or the equivalent residue 165 of Glut1 has been shown to impair glucose transport activity. This mutation was originally discovered in the Glut2 gene of a patient with type 2 diabetes. We investigated the mechanism of the effect of this mutation on transport activity via the analysis of Glut1 mutants expressed inXenopus oocytes combined with cysteine substitution mutagenesis and the use of cysteine-reactive chemical probes. Aliphatic side chain substitutions at position 165 that were bulkier than the native valine residue inhibited glucose transport activity, whereas substitutions of less bulky side chains had little effect on transport, suggesting a role for steric hindrance. A cysteine residue was introduced at position 165 of a functional, cysteine-less Glut1 construct, and this mutant was then tested for inhibition of transport activity by a membrane-impermeant sulfhydryl-specific reagent (p-chloromercuribenzenesulfonate).p-Chloromercuribenzenesulfonate inhibited activity of the Cys165 mutant when it was added to the external buffer but not when it was injected directly into oocytes, indicating that this residue is accessible from the external solvent but not from the cytoplasm. Competition experiments indicated that Cys165lies near the exofacial substrate-binding site or directly in the sugar permeation pathway. These data provide evidence that the side chain of Val165, which resides in the middle of transmembrane helix 5, juts into the aqueous permeation pathway of Glut1, probably between the exofacial substrate-binding site and the outer vestibule of the pathway. A valine-to-isoleucine mutation at amino acid residue 197 of Glut2 or the equivalent residue 165 of Glut1 has been shown to impair glucose transport activity. This mutation was originally discovered in the Glut2 gene of a patient with type 2 diabetes. We investigated the mechanism of the effect of this mutation on transport activity via the analysis of Glut1 mutants expressed inXenopus oocytes combined with cysteine substitution mutagenesis and the use of cysteine-reactive chemical probes. Aliphatic side chain substitutions at position 165 that were bulkier than the native valine residue inhibited glucose transport activity, whereas substitutions of less bulky side chains had little effect on transport, suggesting a role for steric hindrance. A cysteine residue was introduced at position 165 of a functional, cysteine-less Glut1 construct, and this mutant was then tested for inhibition of transport activity by a membrane-impermeant sulfhydryl-specific reagent (p-chloromercuribenzenesulfonate).p-Chloromercuribenzenesulfonate inhibited activity of the Cys165 mutant when it was added to the external buffer but not when it was injected directly into oocytes, indicating that this residue is accessible from the external solvent but not from the cytoplasm. Competition experiments indicated that Cys165lies near the exofacial substrate-binding site or directly in the sugar permeation pathway. These data provide evidence that the side chain of Val165, which resides in the middle of transmembrane helix 5, juts into the aqueous permeation pathway of Glut1, probably between the exofacial substrate-binding site and the outer vestibule of the pathway. Glut1 is the prototype member of the Glut family of membrane glycoproteins, which comprises four glucose transporter isoforms (Glut proteins 1–4) and a fructose transporter (Glut5) (1Baldwin S.A. Biochim. Biophys. Acta. 1993; 1154: 17-49Crossref PubMed Scopus (280) Google Scholar, 2Mueckler M. Eur. J. Biochem. 1994; 219: 713-725Crossref PubMed Scopus (967) Google Scholar). Glut1 is perhaps the most extensively studied of all of the facilitated diffusion-type membrane transport systems. Its kinetic properties have been studied for over four decades in the human erythrocyte membrane (3Carruthers A. Physiol. Rev. 1990; 70: 1135-1176Crossref PubMed Scopus (331) Google Scholar, 4Widdas W.F. Biochim. Biophys. Acta. 1988; 947: 385-404Crossref PubMed Scopus (54) Google Scholar), and more recently in a variety of cell types, includingXenopus oocytes (5Keller K. Strube M. Mueckler M. J. Biol. Chem. 1989; 264: 18884-18889Abstract Full Text PDF PubMed Google Scholar, 6Nishimura H. Pallardo F.V. Seidner G.A. Vannucci S. Simpson I.A. Birnbaum M.J. J. Biol. Chem. 1993; 268: 8514-8520Abstract Full Text PDF PubMed Google Scholar) and the muscle of transgenic mice overexpressing Glut1 (7Hansen P. Gulve E. Gao J. Schluter J. Mueckler M. Holloszy J. Am. J. Physiol. 1995; 268: C30-C35Crossref PubMed Google Scholar). Most of the kinetic observations on Glut1 are consistent with a simple alternating conformation mechanism for glucose transport (8Lowe A.G. Walmsley A.R. Agre P. Parker J.C. Red Blood Cell Membranes. 11. Marcel Dekker, Inc., New York1989: 597-634Google Scholar), although important transport anomalies have been observed in the human red blood cell that have not been adequately explained in terms of the simple carrier model (3Carruthers A. Physiol. Rev. 1990; 70: 1135-1176Crossref PubMed Scopus (331) Google Scholar). The amino acid sequence of Glut1 was deduced from analysis of cDNA clones isolated from human HepG2 (9Mueckler M. Caruso C. Baldwin S.A. Panico M. Blench I. Morris H.R. Allard W.J. Lienhard G.E. Lodish H.F. Science. 1985; 229: 941-945Crossref PubMed Scopus (1213) Google Scholar) and rat brain (10Birnbaum M.J. Haspel H.C. Rosen O.M. Proc. Natl. Acad. Sci. U. S. A. 1986; 83: 5784-5788Crossref PubMed Scopus (477) Google Scholar) libraries. The human polypeptide comprises 492 amino acid residues, exhibits a molecular mass of 54,117, and bears a single N-linked oligosaccharide at Asn45. The 12-transmembrane helix topological model for Glut1, originally postulated on the basis of hydrophobicity analysis (9Mueckler M. Caruso C. Baldwin S.A. Panico M. Blench I. Morris H.R. Allard W.J. Lienhard G.E. Lodish H.F. Science. 1985; 229: 941-945Crossref PubMed Scopus (1213) Google Scholar), is strongly supported by enzymatic (9Mueckler M. Caruso C. Baldwin S.A. Panico M. Blench I. Morris H.R. Allard W.J. Lienhard G.E. Lodish H.F. Science. 1985; 229: 941-945Crossref PubMed Scopus (1213) Google Scholar, 11Cairns M.T. Alvarez J. Panico M. Gibbs A.F. Morris H.R. Chapman D. Baldwin S.A. Biochim. Biophys. Acta. 1987; 905: 295-310Crossref PubMed Scopus (66) Google Scholar) and immunologic (12Davies A. Ciardelli T.L. Lienhard G.E. Boyle J.M. Whetton A.D. Baldwin S.A. Biochem. J. 1990; 266: 799-808PubMed Google Scholar) studies, and by a comprehensive glycosylation-scanning mutagenesis study (13Hresko R.C. Kruse M. Strube M. Mueckler M. J. Biol. Chem. 1994; 269: 20482-20488Abstract Full Text PDF PubMed Google Scholar). A 12-transmembrane helix model has also been experimentally verified for the Lac permease ofEscherichia coli (14Calamia J. Manoil C. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 4937-4941Crossref PubMed Scopus (232) Google Scholar), a protein that belongs to the same superfamily as the mammalian Glut proteins, lending indirect but additional evidence in support of the 12-transmembrane helix model for Glut1 and other members of the superfamily. We noted that 5 of the 12 transmembrane segments of Glut1 could form amphipathic α-helices and postulated that several of these helices might cluster together in the membrane to form the walls of an aqueous pathway through which sugar substrates traverse the fatty acyl core of the lipid bilayer (9Mueckler M. Caruso C. Baldwin S.A. Panico M. Blench I. Morris H.R. Allard W.J. Lienhard G.E. Lodish H.F. Science. 1985; 229: 941-945Crossref PubMed Scopus (1213) Google Scholar). Additionally, we suggested that hydroxyl and amide side chains within these amphipathic helices (helices 3, 5, 7, 8, and 11) might form the glucose-binding site or sites within the membrane via hydrogen bonding to glucose hydroxyl groups. In support of this hypothesis, experimental evidence has been presented that Gln282, within helix 7 (15Hashiramoto M. Kadowaki T. Clark A.E. Muraoka A. Momomura K. Sakura H. Tobe K. Akanuma Y. Yazaki Y. Holman G.D. Kasuga M. J. Biol. Chem. 1992; 267: 17502-17507Abstract Full Text PDF PubMed Google Scholar), and Gln161, within helix 5 (16Mueckler M. Weng W. Kruse M. J. Biol. Chem. 1994; 269: 20533-20538Abstract Full Text PDF PubMed Google Scholar), form part of the exofacial substrate-binding site. We previously reported that a mutation discovered in the Glut2 gene of a diabetic patient (Val197 → Ile) severely disrupts activity of the Glut2 protein expressed in Xenopus oocytes, and that the corresponding mutation in Glut1 (Val165 → Ile) also abolished transport activity (17Mueckler M. Kruse M. Strube M. Riggs A.C. Chiu K.C. Permutt M.A. J. Biol. Chem. 1994; 269: 17765-17767Abstract Full Text PDF PubMed Google Scholar). This valine residue lies within transmembrane helix 5, approximately one helical turn distant from Gln161. The effect of this mutation was completely unexpected, given the highly conservative nature of the substitution, and the unreactive nature of the valine side chain, especially in the context of a hydrophobic environment where valine cannot participate in hydrophobic interactions with other amino acid side chains that might otherwise stabilize transporter structure in the context of an aqueous environment. To investigate the role of Val165 in the structure and function of Glut1, we used an experimental approach involving cysteine substitution mutagenesis combined with sulfhydryl-reactive chemical probes. Khorana and colleagues (18Altenbach C. Flitsch S.L. Khorana H.G. Hubbell W.L. Biochemistry. 1989; 28: 7806-7812Crossref PubMed Scopus (252) Google Scholar, 19Flitsch S.L. Khorana H.G. Biochemistry. 1989; 28: 7800-7805Crossref PubMed Scopus (41) Google Scholar) were the first to use this approach to explore the structure and function of a membrane transporter, and the approach has subsequently been exploited in various permutations to investigate the structure-function relationships of a variety of membrane transporters (18Altenbach C. Flitsch S.L. Khorana H.G. Hubbell W.L. Biochemistry. 1989; 28: 7806-7812Crossref PubMed Scopus (252) Google Scholar, 19Flitsch S.L. Khorana H.G. Biochemistry. 1989; 28: 7800-7805Crossref PubMed Scopus (41) Google Scholar, 20Yan R.T. Maloney P.C. Cell. 1993; 75: 37-44Abstract Full Text PDF PubMed Scopus (72) Google Scholar, 21Yan R.T. Maloney P.C. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 5973-5976Crossref PubMed Scopus (79) Google Scholar, 22Lebendiker M. Schuldiner S. J. Biol. Chem. 1996; 271: 21193-21199Abstract Full Text Full Text PDF PubMed Scopus (25) Google Scholar, 23Frillingos S. Kaback H.R. Biochemistry. 1996; 35: 3950-3956Crossref PubMed Scopus (80) Google Scholar, 24Weitzman C. Kaback H.R. Biochemistry. 1995; 34: 9374-9379Crossref PubMed Scopus (21) Google Scholar) and ion channels (25Akabas M.H. Stauffer D.A. Xu M. Karlin A. Science. 1992; 258: 307-310Crossref PubMed Scopus (597) Google Scholar, 26Akabas M.H. Kaufmann C. Archdeacon P. Karlin A. Neuron. 1994; 13: 919-927Abstract Full Text PDF PubMed Scopus (359) Google Scholar, 27Akabas M.H. Kaufmann C. Cook T.A. Archdeacon P. J. Biol. Chem. 1994; 269: 14865-14868Abstract Full Text PDF PubMed Google Scholar, 28Akabas M.H. Karlin A. Biochemistry. 1995; 34: 12496-12500Crossref PubMed Scopus (159) Google Scholar). Our data suggest that Val165 lies within the aqueous sugar translocation pathway of Glut1, in close proximity to both the exofacial substrate-binding site and the outer vestibule of the transporter. These observations lend additional support to our model concerning the importance of helix 5 in transporter catalytic activity (9Mueckler M. Caruso C. Baldwin S.A. Panico M. Blench I. Morris H.R. Allard W.J. Lienhard G.E. Lodish H.F. Science. 1985; 229: 941-945Crossref PubMed Scopus (1213) Google Scholar). Procedures for the site-directed mutagenesis and sequencing of human Glut1 cDNA and the in vitro transcription and purification of Glut1 mRNAs (29Hresko R.C. Murata H. Marshall B.A. Mueckler M. J. Biol. Chem. 1994; 269: 32110-32119Abstract Full Text PDF PubMed Google Scholar), isolation, microinjection, and incubation of Xenopus oocytes (30Marshall B.A. Murata H. Hresko R.C. Mueckler M. J. Biol. Chem. 1993; 268: 26193-26199Abstract Full Text PDF PubMed Google Scholar), preparation of total oocyte membranes and laser confocal immunofluorescence microscopy of sectioned oocytes (31Garcia J.C. Strube M. Leingang K. Keller K. Mueckler M.M. J. Biol. Chem. 1992; 267: 7770-7776Abstract Full Text PDF PubMed Google Scholar), SDS-polyacrylamide gel electrophoresis and immunoblotting with Glut1 C-terminal antibody (16Mueckler M. Weng W. Kruse M. J. Biol. Chem. 1994; 269: 20533-20538Abstract Full Text PDF PubMed Google Scholar), and 2-deoxyglucose uptake measurements (5Keller K. Strube M. Mueckler M. J. Biol. Chem. 1989; 264: 18884-18889Abstract Full Text PDF PubMed Google Scholar) have been described in detail previously. Stage 5 Xenopus oocytes were injected with 50 ng of wild-type or mutant Glut1 mRNA. Three days after injection, groups of ∼30 oocytes were incubated for 15 min in the presence or absence of the indicated concentrations of p-chloromercuribenzoate (pCMB), 1The abbreviations used are: pCMB,p-chloromercuribenzoate; pCMBS,p-chloromercuribenzenesulfonate; C-less, Glut1 molecule in which all six native cysteine residues were changed to either glycine or serine. p-chloromercurisulfonate (pCMBS), iodomethane, or iodoacetate in Barth's saline at 22 °C. The 100-fold concentrated reagent stocks were prepared in 50% dimethyl sulfoxide, and control oocytes were treated with the appropriate concentration of vehicle alone. In some experiments, sugars (2-deoxyglucose, ethylidene glucose,d-glucose) were included in the incubation solution at 200 mm to test for their ability to delay the sulfhydryl reaction. After the 15-min incubation, the oocytes were washed three times in Barth's saline and then used for the determination of 3H2-deoxyglucose uptake (50 μm, 30 min at 22 We reported previously that the highly conservative substitution of an for Val165 of Glut1, a residue that lies in the middle of transmembrane helix 5, transport activity (17Mueckler M. Kruse M. Strube M. Riggs A.C. Chiu K.C. Permutt M.A. J. Biol. Chem. 1994; 269: 17765-17767Abstract Full Text PDF PubMed Google Scholar). Val165 lies one helical turn distant from Gln161, a residue that to in the exofacial substrate-binding site of Glut1 (16Mueckler M. Weng W. Kruse M. J. Biol. Chem. 1994; 269: 20533-20538Abstract Full Text PDF PubMed Google Scholar) We that the side chain, which bears an with might between glucose and the amide side chain of as a of We tested this by the effect of other amino acid substitutions at position 165 on transport activity. the is bulkier side chain substitutions at position 165 activity, whereas less bulky side chains have little effect on transport The of amino acid substitutions at position 165 of human Glut1 that were and are in The mutant were into mRNAs that were then injected into Xenopus and function were days after of total oocyte membranes that the of protein the mutants and wild-type Glut1 and also from one to as is the with the oocyte laser confocal immunofluorescence microscopy of oocyte that all of the mutants were expressed in the membrane as as wild-type Glut1 2 uptake measurements that amino acid side chain substitutions at position 165 that were bulkier than valine inhibited activity of Glut1 by whereas side chain substitutions less bulky than valine had little or effect on transport activity was a for the substitution mutant to a activity than wild-type Glut1, the was not introduced into wild-type or acid → → → → → → → → → → → → in → → → → → → → → → → were by site-directed mutagenesis as described previously (16Mueckler M. Weng W. Kruse M. J. Biol. Chem. 1994; 269: 20533-20538Abstract Full Text PDF PubMed Google Scholar). The first six were introduced into wild-type human Glut1 cDNA to cysteine-less The mutation was introduced into Glut1 cDNA to the The four were introduced into wild-type human Glut1 cDNA to test the effect of side chain on transport activity. in a uptake activity within the of substitution mutants at uptake (50 μm, 30 min at 22 was days after of the of three experiments, with control analysis of test for of the mutants to wild-type The between the and mutant and wild-type uptake were not were by site-directed mutagenesis as described previously (16Mueckler M. Weng W. Kruse M. J. Biol. Chem. 1994; 269: 20533-20538Abstract Full Text PDF PubMed Google Scholar). The first six were introduced into wild-type human Glut1 cDNA to cysteine-less The mutation was introduced into Glut1 cDNA to the The four were introduced into wild-type human Glut1 cDNA to test the effect of side chain on transport activity. The with the that approximately one helical turn distant from a residue that to directly with glucose through the suggest that Val165 into the aqueous pathway by which glucose the lipid and that bulky substitutions at this position either with or with transport activity in some other To more directly Val165 lies in the aqueous transmembrane pathway by Glut1, we first a Glut1 molecule of cysteine residues by site-directed mutagenesis to all six native cysteine residues to either glycine or residues The cysteine-less Glut1 molecule of wild-type Glut1 transport activity when expressed inXenopus oocytes as indicated by 3H2-deoxyglucose uptake four that of the native Glut1 cysteine residues is for transport function M. I. Keller K. 1995; PubMed Scopus Google Scholar). We then introduced a single cysteine into Glut1 at position 165 by site-directed The Cys165 mutant of the glucose transport activity of wild-type Glut1 four a of activity to an of the effect of on transport activity. that residue 165 resides in the middle of transmembrane helix 5, the Cys165 mutant with a this highly that this residue lies within an aqueous transmembrane accessible from the was used as the membrane-impermeant sulfhydryl-specific and the as a control for of Cys165 A that of oocytes in the presence of mm of either reagent inhibited 2-deoxyglucose uptake activity of wild-type Glut1 by consistent with a that this inhibition is to with the M. I. Keller K. 1992; PubMed Scopus Google Scholar). In the activity of Glut1 was by or when oocytes the Cys165 mutant were with or transport activity was inhibited by and that inhibition of by both was although the inhibition observed with either reagent from to in experiments not To Cys165 is accessible from the and were injected into oocytes to of 2-deoxyglucose 5 that was to the activity of Glut1 or the Cys165 mutant when injected into the of These data the transport Cys165 is accessible from the external solvent but not from the suggesting that this residue within or near the exofacial vestibule of the aqueous permeation of and on transport activity. days after of groups of oocytes were incubated in the presence or absence of mm or in Barth's saline at 22 °C. were washed three times in Barth's saline and then to 2-deoxyglucose uptake measurements the described in the to are expressed as The data shown one of three with control human the cysteine-less the Val165 → mutant in the oocytes the mutant were incubated for 15 min in the indicated concentration of or in Barth's washed three times in Barth's and then to 2-deoxyglucose uptake measurements as described in the to groups of oocytes the mutant or the Cys165 mutant in the were injected with 50 of a solution of either mm mm or dimethyl and 15 min to 2-deoxyglucose uptake measurements as described in the to The concentration of or within the oocyte was are expressed as The data shown one of four Cys165 lies in close proximity to the exofacial substrate-binding site or directly in the sugar translocation then sugar substrates might with and delay the of with A inhibition of the not sugar is a that sugar substrates with and delay the of with was used as an control that for The of the sugars at the their for 2-deoxyglucose ethylidene glucose (8Lowe A.G. Walmsley A.R. Agre P. Parker J.C. Red Blood Cell Membranes. 11. Marcel Dekker, Inc., New York1989: 597-634Google Scholar). Additionally, the that the Glut1 exofacial ethylidene inhibited the of with Cys165 that with the exofacial substrate-binding site is to and that inhibition was not the of transport into the or to the substrate-binding site. and were used as sulfhydryl in the experiments, in part these some to in and To are on that have to we the ability of other sulfhydryl-reactive both in than either or to with A that iodoacetate inhibited transport activity of the Cys165 this of inhibition could to the of the to to or to the of the or cysteine residue to transport activity, we the shown in 7 In this oocytes the Cys165 mutant were first incubated with either or with buffer to the and then incubated with that inhibition of transport activity. the with Cys165 but the not transport, then incubation with not in transport inhibition of the Cys165 the sulfhydryl have with the The data in 7 that incubation with in inhibition of transport activity, indicating that the not have to Cys165 within the permeation pathway of The experiments described an amino acid residue that to within the sugar translocation pathway of This residue was originally of in gene of a patient with type Y. Riggs A.C. Chiu K.C. R.C. J.M. R.T. Permutt M.A. 1994; PubMed Scopus Google Scholar). The valine-to-isoleucine mutation at position 197 of Glut2 and at the equivalent position 165 of Glut1 transport function of the isoforms (17Mueckler M. Kruse M. Strube M. Riggs A.C. Chiu K.C. Permutt M.A. J. Biol. Chem. 1994; 269: 17765-17767Abstract Full Text PDF PubMed Google Scholar). The effect of this mutation on transport activity the to the role of this residue in transporter Val165 is to directly in the of sugar substrates or in transporter given unreactive side chain and in of the described it to within the sugar translocation pathway in the of the exofacial site. have that Gln161, which lies one helical turn distant from Val165 within transmembrane helix 5, in both exofacial and in transporter (16Mueckler M. Weng W. Kruse M. J. Biol. Chem. 1994; 269: 20533-20538Abstract Full Text PDF PubMed Google Scholar). side chains at position 165 either of substrates with the amide side chain of or transporter and Maloney R.T. Maloney P.C. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 5973-5976Crossref PubMed Scopus (79) Google Scholar), a cysteine substitution experimental have three within transmembrane helix 7 of the glucose discovered that amino acid residues were either accessible to in accessible in membrane or accessible in both on their within the transmembrane close to the membrane into the first residues close to the membrane into the and residues near the middle of the helix into the their observations in terms of the membrane carrier model (3Carruthers A. Physiol. Rev. 1990; 70: 1135-1176Crossref PubMed Scopus (331) Google Scholar, 4Widdas W.F. Biochim. Biophys. Acta. 1988; 947: 385-404Crossref PubMed Scopus (54) Google Scholar), where one amino acid residues directly in the middle of helix 7 in the of that to both aqueous the of the transport that on the exofacial side of this to the and residues that to this to the data provide evidence for a model of membrane transport on the basis of kinetic transport The that Cys165 in Glut1 was accessible to from the external but not from the that this residue lies in the outer vestibule of Glut1 that not to the transporter the transport The after the reagent into oocytes with is that the not with the cysteine residue at position 165 for a other than of the side a of groups that for with the Cys165 lies one helical turn distant from a residue that to directly in exofacial and also to in a given that and iodoacetate not have to this one that the chain is to as suggested by the then Val165 most lies at an position between the outer vestibule of Glut1 and the of helix 5 directly in of the mechanism of the inhibition at Val165, the reported lend additional support to our model for Glut1 in which helix 5, with other amphipathic in an aqueous transmembrane accessible from one membrane at and one or more substrate-binding sites M. Agre P. Parker J.C. Red Blood Cell Membranes. Marcel Dekker, Inc., New York1989: Scholar). helix 5 of the Lac permease of E. been to a role in the transport of The Lac permease belongs to the 12 transmembrane helix transporter superfamily with the mammalian Glut proteins, but more is the relationships of the Lac permease with on a variety of Kaback H.R. Proc. Natl. Acad. Sci. U. S. A. PubMed Scopus Google Scholar) has that is the permease through a or by helices 5, 7, and 8, and that and within helix 5 are in with is to that helix is in all members of the 12 transmembrane helix and that these same helices the sugar translocation pathway in This consistent with data that in helix 7 and helix 5 participate in exofacial We the members of the Mueckler at for the of these