MINIREVIEW

Sequence and Functional Analysis of GLUT10: A GlucoseTransporter in the Type 2 Diabetes-Linked Region

of Chromosome 20q12–13.11

Paul A. Dawson,‡ Josyf C. Mychaleckyj,‡,†,§ Sallyanne C. Fossey,\ S. John Mihic,†Ann L. Craddock,‡ and Donald W. Bowden*,‡,\ ,2

*Department of Biochemistry, †Department of Physiology and Pharmacology, ‡Department of Internal Medicine, §Department of Public

Molecular Genetics and Metabolism 74, 186–199 (2001)doi:10.1006/mgme.2001.3212, available online at http://www.idealibrary.com on

Health Sciences, and the \Program in Molecular Genetics, Wake Forest University School of Medicine,North

uly 12

a

TW

tat

Winston-Salem,

Received June 20, 2001, and in revised form J

We have carried out a detailed sequence and func-tional analysis of a novel human facilitative glucosetransporter, designated GLUT10, located in theType 2 diabetes-linked region of human chromo-some 20q12–13.1. The GLUT10 gene is located be-tween D20S888 and D20S891 and is encoded by 5exons spanning 26.8 kb of genomic DNA. The humanGLUT10 cDNA encodes a 541 amino acid proteinthat shares between 31 and 35% amino acid identitywith human GLUT1–8. The predicted amino acidsequence of GLUT10 is nearly identical in length tothe recently described GLUT9 homologue, but islonger than other known members of the GLUTfamily. In addition, we have cloned the mousecDNA homolog of GLUT10 that encodes a 537 aminoacid protein that shares 77.3% identity with humanGLUT10. The amino acid sequence probably has12 predicted transmembrane domains and sharescharacteristics of other mammalian glucose trans-porters. Human and mouse GLUT10 retain severalsequence motifs characteristic of mammalianglucose transporters including VP497ETKG in the

1 The sequence reported in this paper has been deposited inGenBank Database (Accession Nos. AF248053 and AY029579)and given the HGMW approved symbol SLC2A10.

iG

2 To whom correspondence should be addressed at Departmentof Biochemistry, Wake Forest University School of Medicine,Medical Center Boulevard, Winston-Salem, NC 27157. Fax: (336)716-7200. E-mail: dbowden@wfubmc.edu.

1861096-7192/01 $35.00Copyright © 2001 by Academic PressAll rights of reproduction in any form reserved.

Carolina 27157

, 2001; published online September 17, 2001

cytoplasmic C-terminus, G73R[K,R] between TMD2nd TMD3 (PROSITE PS00216), VD92RAGRR between

TMD8 and TMD9 (PROSITE PS00216), Q242QLTG inMD7, and tryptophan residues W430 (TMD10) and454 (TMD11), that correspond to trytophan residues

previously implicated in GLUT1 cytochalasin B bind-ing and hexose transport. Neither human nor mouseGLUT10 retains the full P[E,D,N]SPR motif afterLoop6 but instead is replaced with P186AG[T,A]. APROSITE search also shows that GLUT10 has lost theSUGAR_TRANSPORT_2 pattern (PS00217), a result ofthe substitution G113S in TMD4, while all otherknown human GLUTs retain the glycine and the pat-tern match. The significance of this substitution isunknown. Sites for N-linked glycosylation are pre-dicted at N334ATG between TMD8 and TMD9 andN526STG in the cytoplasmic C-terminus. Northern hy-bridization analysis identified a single 4.4-kb tran-script for GLUT10 in human heart, lung, brain, liver,skeletal muscle, pancreas, placenta, and kidney. ByRT-PCR analysis, GLUT10 mRNA was also detected infetal brain and liver. When expressed in Xenopus oo-cytes, human GLUT10 exhibited 2-deoxy-D-glucoseransport with an apparent Km of ;0.3 mM. D-Glucosend D-galactose competed with 2-deoxy-D-glucose andransport was inhibited by phloretin. The gene local-

zation and functional properties suggest a role forLUT10 in glucose metabolism and Type 2 diabetes.

© 2001 Academic Press

TYPE

Glucose is an important source of energy for mostliving organisms. The movement of glucose acrossmembranes is accomplished by two classes of trans-porters, the energy-dependent Na1-glucose cotrans-porters (1) and the facilitative glucose transporters.In humans, the facilitative glucose transporter fam-ily consists of at least six glucose transporters (2–10)and a fructose transporter (11,12). These facilitativetransporters regulate the movement of glucose be-tween extra- and intracellular spaces to maintain aconstant supply of circulating glucose (13).

Defects in facilitative glucose transporters havebeen implicated in several metabolic disorders, in-cluding GLUT13 deficiency syndrome (14), Fanconi-Bickel syndrome (15) and Type 2 diabetes (16–18).Type 2 diabetes is one of the most prevalent meta-bolic diseases, characterized by peripheral insulinresistance, impaired insulin production, and in-creased hepatic glucose production all contributingto hyperglycemia. Despite intensive investigation,the etiology of the disease remains unknown. Thefirst glucose transporters identified (GLUT1–5)were extensively analyzed for mutations contribut-ing to Type 2 diabetes, but to date no common caus-ative mutation has been identified. Several novelglucose transporters, GLUT6, GLUT8, and GLUT9(7–10) have recently been identified and additionalglucose transporters may exist. The key metabolicrole of glucose transport suggests that the identifi-cation of novel transporters may lead to new in-sights into the underlying biological processes ofglucose homeostasis and Type 2 diabetes.

The results of several recent genetic linkage stud-ies suggest that Type 2 diabetes in Caucasian pa-tients is linked to the q12–q13.1 region of humanchromosome 20 (19–22). Evidence of linkage dis-equilibrium with Type 2 diabetes has also been ob-served with alleles of two genetic markers withinthis linked region, adenosine deaminase (ADA) andD20S888, markers separated by approximately 6 cM(23). A survey of the published human sequence inthis region of chromosome 20 identified a partialtranscript as “similar to membrane transport pro-tein” (Sanger Centre; ftp://ftp.sanger.ac.uk/pub/

3 Abbreviations used: BAC, bacterial artificial chromosome;2-DOG, 2-deoxy-D-glucose; EST, expressed sequence tag; GLUT,

GLUT10 AND

glucose transporter; MBS, modified Barth’s saline; RACE, rapidamplification of cDNA ends; RT-PCR, reverse transcriptase-poly-merase chain reaction; TMD, transmembrane domain; YAC,yeast artificial chromosome.

human/sequences/Chr 20/). Homology searches (24)of the nonredundant nucleotide database NCBI(www.ncbi.nlm.nih.gov/BLAST) with this transcriptsequence revealed significant similarities with bac-terial and mammalian hexose transporters. Theseobservations led to identification of the GLUT10gene (HGMW approved gene symbol SLC2AI0; 25)within the Type 2 diabetes susceptibility region onhuman chromosome 20. Independently McVie et al.(26) identified and cloned the same candidate glu-cose transporter gene. Here we describe a detailedsequence analysis of GLUT10 and demonstrate thatthe expressed protein carries out facilitative trans-port of glucose in vitro.

MATERIALS AND METHODS

Cloning and Localization of Human GLUT10

Analysis of published human sequence data pro-duced by the Human Chromosome 20 SequencingGroup at the Sanger Centre (ftp://ftp.sanger.ac.uk/pub/human/sequences/Chr 20/) identified a partialtranscript described as a “membrane transporter-likeprotein” (GenBank Accession Number AL031055).This sequence was used in homology searches (24)of the nonredundant nucleotide and EST data-bases at NCBI (http://www.ncbi.nlm.nih.gov/).Primers (GLUT10 No. 11, 12, see Table 1) de-signed from this sequence were used to amplify afull-length GLUT10 transcript from Marathon-Ready human placental cDNA using the Advan-tage cDNA PCR and 59/39 RACE Adapter kits(Clontech Laboratories Inc., Palo Alto, CA). Allamplified products were sequenced bidirectionallyon an ABI Prism 377 automated DNA sequencer(Applied Biosystems, Foster City, CA) by theWake Forest University School of Medicine DNASequencing Core Laboratory. PCR primers specificfor GLUT10 (GLUT10 No. 2, 4; Table 1) were usedto screen a panel of overlapping Human YAC li-brary and CITB BAC library clones (Version 4.0,Research Genetics, Huntsville, AL) that spanneda contiguous 6 Mb region of human chromosome20q12–13.1 (27,28). The localization of GLUT10was inferred from the retention pattern withinindividual BAC clones.

Identification and Cloning of the Mouse GLUT10

1872 DIABETES

Homolog

Mouse Glut10 oligonucleotide primers were de-signed based on the sequence of the mouse EST

s

TGCA

ation c

clones (GenBank Accession Numbers AA461907 andAI614852) and used for 59 and 39 RACE (Smart RaceKit; Clontech Laboratories, Inc.,). For 39 RACE,first-strand cDNA was synthesized from 1 mg ofBalb/c mouse brain poly(A) RNA (Clontech) using anoligo(dT) adapter primer (39 CDS). The primer pairUPM/mGlut10 No. 4 (Table 1) was then employed inthe primary PCR amplification of the cDNA accord-ing to the manufacturer’s instructions. The 1941nucleotide PCR-amplified product was isolated froma 1.0% (w/v) agarose gel, subcloned into pBluescriptII KS. For 59 RACE, first-strand cDNA was synthe-ized from 1 mg of Balb/c mouse brain poly(A) RNA

TAHuman and Mouse GLUT10 O

Oligonucleotide

GLUT10 No. 1 AGGAGCAGGCTGLUT10 No. 2 AGTGGCATAGGGLUT10 No. 3 CCCAGTTGAAGGLUT10 No. 4 CTGCAACAGCTGLUT10 No. 5 CTATCTCTGCCGLUT10 No. 6 CCAGTCGTTGGGLUT10 No. 7 AAGGCCAGTCGGLUT10 No. 8 CTGAGGAATCCGLUT10 No. 9 CCATGGCGAGCGLUT10 No. 10 CTCTTCATTTTGGLUT10 No. 11 AGTCCCGCTCGGLUT10 No. 12 CAGGCAGACGGGLUT10 No. 13 CGTCCCGCCTCGLUT10 No. 14 GGCGGTGTCTAGLUT10 No. 15 GGGAACCCCAGGLUT10 No. 16 CAAACCACCTTGLUT10 No. 17 CTCAACTATGCGLUT10 No. 18 AACTGGCTGGGGLUT10 No. 19 TCGTGAACTCCGLUT10 No. 20 GGAGGCCAACAGLUT10 No. 21 CCATGCGTGCAGLUT10 No. 22 GTGAGTTGAACmGLUT10 No. 1b AGACTGTCCTTmGLUT10 No. 2 AGGCCAAGGACmGLUT10 No. 3 GTCCAAGCCAAmGLUT10 No. 4 GATCGGTGCCAmGLUT10 No. 5 CTATGGGCTGAmGLUT10 No. 8 CTCGCTAGCTAmGLUT10 No. 9 CAGACGGATGCmGLUT10 No. 10 CGTTCATCAAGmGLUT10 No. 11 GATTACAGGCA

a Location within GLUT10 cDNA sequence. A of the ATG initib Mouse GLUT10.

188 DAWS

(Clontech) using an oligo(dT) adapter primer (59CDS). The primer pairs UPM/mGlut10 No. 2 andNUP/mGlut10 No. 3 were employed in the primary

and secondary PCR amplifications, respectively, ofthe cDNA. The 1492 nucleotide PCR-amplified prod-uct was isolated from a 1.0% (w/v) agarose gel, sub-cloned into pBluescript II KS. The mouse Glut10 59and 39 RACE clones were sequenced using mouseGlut10 primers (Nos. 1 to 11; Table 1) with an ABIPrism 377 automated sequencer (Perkin Elmer, Fos-ter City, CA). The mouse Glut10 cDNA has beensubmitted to the GenBank database under Acces-sion Number AY029579.

Human Tissue Expression Analysis

A poly(A) RNA multiple-tissue Northern blot (Cat-

1ucleotide Primer Sequences

ce Locationa

ACCA 166–148GTCAG 933–952TGCAG 1368–1349CTGGG 1349–1368

CTGG 1524–1505GATAG 1505–1524CAGAG 1501–1520GCCTG 1622–1641CT (1)3–(2)13TCAGGC 1799–1819

GG (2)13–(1)3CTCAG 1641–1622CCT (2)196–(2)179CTGG Intron 1AGGT Intron 4GGCAGG 1755–1735

CTGG 456–475TTTCG 2033–2052CTCAAG 2281–2301TGCTTG 2642–2662TGGGTG 3012–3032TCATCCG 3611–3631TGAGGAACAAG 1500–1474GTCAGCCCATAG 1457–1431ATGGCACCGATC 1421–1395CTTGGCTTGGAC 1395–1421TGTCCTTGGCCT 1431–1457CCTTCGCCCAG (2)9–16GGAGGCTGAG 1623–1599CTCAGAGGAC 1882–1907

CCAGCACATAG 2983–2958

odon is designated 11.

T AL.

BLEligon

Sequen

GCCCCCTCCTGTTCAAAACGACAGATTGGGTCTGGGAAGTC

CCATATTCCAGGCACCTGGA

TGGGACTGGACATTGAGAACCCATAACTGTGGTTAGCGGCCGTCGGCCGCTGGGCTCATGCTT

ON E

alog Number 7760-1) was purchased from ClontechLaboratories and hybridization was carried out ac-cording to manufacturer’s recommendations. PCR

pow1

TYPE

amplification was conducted using GLUT10 primers3 and 17 (Table 1) to screen 24 tissues in the Rapid-Scan gene expression panel (OriGene Technologies,Inc., Rockville, MD) according to the manufacturer’sguidelines.

Functional Analysis of Human GLUT10 inXenopus Oocytes

A full-length GLUT10 cDNA was cloned into theexpression vector pCMV-Tag4a (Stratagene, LaJolla, CA). The pcDNA3.1/GS plasmids expressingthe full-length GLUT3 and GLUT4 cDNA were pur-chased from Invitrogen Corporation (Carlsbad, CA).Capped mRNA was generated using the mMES-SAGE mMACHINE transcription kit (Ambion Inc.,Austin, TX).

Adult female Xenopus laevis were obtained fromXenopus Express (Homosassa, FL) and housed at17–19°C on a twelve 12-h light/dark cycle. StageV-VI oocytes were removed from anesthetized frogsand placed in isolation media (108 mM NaCl, 1 mMEDTA, 2 mM KCl, 10 mM Hepes, pH 7.5). The oocytefollicular layer was removed by immersion in 0.5mg/ml collagenase buffer (83 mM NaCl, 2 mM KCl,1 mM MgCl2, and 5 mM Hepes, pH 7.5). Isolatedoocytes were maintained in modified Barth saline(MBS:88 mM NaCl, 1 mM KCl, 10 mM Hepes, 0.82mM MgSO4, 2.4 mM NaHCO3, 0.91 mM CaCl2, and0.33 mM Ca(NO3)2, pH 7.5). Visually healthy oocyteswere injected with 30 nl of either an mRNA solution(1.0 mg/ml) or sterile H2O at the interface betweenthe animal and vegetal poles. Individual oocyteswere placed in 96-well microtiter plates (CorningCostar Corporation, Cambridge, MA) containingMBS plus 2 mM sodium pyruvate, 0.5 mM theoph-ylline, 10 U/ml penicillin, 10 mg/L streptomycin, and50 mg/L gentamycin (Sigma, St. Louis, MO) at 22°C.

2-Deoxy-D-glucose (2-DOG) uptake assays wereerformed at 22°C using pools of 4 to 12 healthyocytes 3 days postinjection. Oocyte pools wereashed in MBS and incubated for 30 min at 22°C in00 ml MBS in the presence of 25 or 250 mM 2-DOG

[1,2-3H] DOG (final specific activity 5 0.5 Ci/mmol)(NEN Life Science Products, Inc., Boston, MA;American Radiolabeled Chemicals Inc., St. Louis,MO). Competitors, inhibitors, or insulin was addedto the incubation as indicated. Uptake was termi-

GLUT10 AND

nated by removal of the radioactive solution andthree 500 ml washes with ice-cold MBS containing0.1 mM phloretin (Sigma). Pools of oocytes were

dissolved in 10% SDS, mixed with scintillation fluid,and internalized radioactivity was measured byscintillation spectrometry using an LS-600 scintilla-tion counter (Beckman Coulter, Inc., Fullerton, CA).Data are expressed as the arithmetic mean or me-dian of duplicate or triplicate pools of 4 to 12 oocytesat each data point. Transport kinetics were analyzedby best-fit analysis of data points (Cricket Graph III,Computer Associates International, Inc., Islandia,NY) and Eadie-Hofstee transformation.

RESULTS

Identification and Chromosomal Localization ofGLUT10

A survey of the published human sequence data in20q12–13.1 identified a partial transcript with sim-ilarity to membrane transport proteins within theannotated BAC clone 28H20 (GenBank AccessionNumber AL031055). Specific PCR primer sets (Table1) were then designed to locate this transcript se-quence within a contiguous 7-Mb physical map ofoverlapping BAC clones spanning the two linkagedisequilibrium maxima (23) of the chromosome 20Type 2 diabetes susceptibility region. The mem-brane transporter-like transcript, SLC2A10/GLUT10, was mapped approximately 100 kb telomeric tothe genetic marker D20S888 (28).

Genomic Structure and Sequence Analysis

59 and 39 RACE was performed using primer setsspecific for the membrane transporter-like tran-script and human placental mRNA to isolate thefull-length 4382-bp cDNA. The human GLUT10cDNA encodes a 541 amino acid protein with a cal-culated molecular mass of 56,875 Da. The predictedinitiator methionine lies within an appropriate con-sensus for initiation of translation (29) and is pre-ceded by a 250-bp 59-untranslated region with nu-merous in-frame stop codons. The 1626-bp codingsequence is followed by a 2509-bp 39-untranslatedsequence that extended to the poly(A) tail. The 39sequence of the cDNA matched UniGene clusterHs.305971 that had been mapped to chromosome 20between genetic markers D20S119 and D20S197.The complete cDNA sequence was submitted to bothGenBank and HUGO and assigned the approved

1892 DIABETES

gene symbol SLC2A10, alias GLUT10 (GenBank Ac-cession Number AF248053). The human GLUT10cDNA sequence was used to identify several mouse

humaning ind

GLUT10 EST clones; those sequences were thenused for 39 and 59 RACE to obtain the full-lengthmouse GLUT10 cDNA (GenBank Accession NumberAY029579). The predicted mouse GLUT10 proteinsequences shares 77.4% identity with the humanGLUT10 (Fig. 1). The human GLUT10 cDNA se-quence was aligned with the genomic DNA of humanBAC clone 28H20 (GenBank Accession NumberAL031055) to reveal the gene organization. The sizeand sequence of the GLUT10 exons were also con-firmed through multiple RT-PCR and RACE exper-iments as well as alignments with existing ESTclones (e.g., GenBank Accession Number BE237601,AA313045, AW028359, AA628914, W02942). TheGLUT10 gene is organized in 5 exons spanning 26.8kb of genomic DNA.

The predicted amino acid sequence of GLUT10 is

FIG. 1. Comparison of the deduced amino acid sequences ofprogram and BOXSHADE (UCSD Biology Workbench). Blue shadbelonging to the same conservation group.

190 DAWS

nearly identical in length to the recently publishedGLUT9 homologue (541 aa vs 540 aa) (9), but islonger than other known members of the GLUT

family. Pairwise local alignment (GCG best-fit pro-gram, BLOSUM50, gap function 210/22) revealedthat GLUT10 shares from 31% (GLUT1) to 33%(GLUT8) and 35% (GLUT6) identity with the previ-ously identified human GLUTn (n 5 1–6, 9). Ahuman GLUT family multiple-sequence alignmentand the predicted transmembrane domain (TMD)organization for GLUT10 are shown in Fig. 2A. Thehypothetical TMD structure was analyzed usinga number of most recent computer prediction pro-grams that use a range of algorithms: SOSUI(30), DAS (31), HMMTOP (32), TMMHMM (33),TMPRED (34), and was correlated with visual hy-dropathy inspection. We conclude that GLUT10probably has a 12 transmembrane domain structurelike other members of the GLUT subfamily, but theautomated prediction programs return a range of

and mouse GLUT10. Alignment was generated using the alignicates identical residues, and gray shading indicates amino acids

T AL.

ON E

conflicting results: 9 TMDs (TMPRED); 11 TMDs(SOSUI, DAS); 12 TMDs (HMMTOP, TMMHMM).For comparison, identical analyses with GLUT1,

gaps, atstrapUT9 aUT6 a

FIG. 2. (A) Protein multiple sequence alignment of the humathe GCG pileup and clustalx (56) programs (variable BLOSUM seare shaded with decreasing matrix similarity score relative to thenonidentical score; yellow, conserved residue, but less than meapositions shown overlined are for GLUT10 and were determinedNP_006507 (GLUT1), NP_000331 (GLUT2), NP_008862 (GLUTNP_064425 (GLUT9), AF2458053 (GLUT10, this work), and EMfamily. The phylogram was generated from the alignment in A uand Nei (57). Distances were calculated excluding positions withmetric is the estimated number of substitutions per site, and booIn this manuscript we refer to the “GLUT9” of reference (9) as GLreceived the HGMW designation of SLC2A6, and the original GL

GLUT10 AND

GLUT8, and GLUT6 in all cases unambiguouslypredict a TM protein with 12 domains, albeit withslight differences in the exact domain positions. The

T family (n 5 1–6, 8, 9, 10). The alignment was generated usingatrices, gap function 212/22) with manual adjustment. Residuesnsensus: blue, identical to site consensus; green, score $ to meanidentical score; white, others. Putative transmembrane domaincribed in the main text. GenBank Accession Numbers used were_001033 (GLUT4), NP_003030 (GLUT5), NP_055395 (GLUT8),B96996 (GLUT6). (B) Unrooted phylogram of the human GLUTustalx, which implements the neighbor-joining method of Saitound with a Kimura multiple substitution correction. The distancesupport values are for 10,000 replicates. (Note on nomenclature:nd the “GLUT9” of reference (10) as GLUT6 since the latter haslias has been withdrawn.)

1912 DIABETES

n GLUries msite con nonas des3), NPBL CAsing cl

TYPE

main source of the GLUT10 discrepancies appears tobe the inability to resolve TMD 11/12, and TMD 3/4into separate domains because of extended contigu-

c(rcs1

i1(yseso

mqdpEatss

ON E

ous blocks of hydrophobic residues (positions 457–498 and 78–126, respectively). The predicted do-main boundaries for GLUT10 vary betweenprograms by 0–7 bases, mean 3. Figure 2A showsthe best estimate of the putative GLUT10 12 TMDpositions derived from the joint TMD prediction andvisual hydropathy analysis. Figure 2B shows a phy-logram for the human GLUT family that quantifiesthe distinction of GLUT10 from the rest of the trans-porters. The exclusion of column positions contain-ing gaps means that this phylogram reflects inferredrelationships from the conserved TMD blocks of res-idues and generally omits the more divergent loopsand termini. As expected from the pairwise align-ment identities, GLUT10 forms a clade with GLUT8and GLUT6 but is an outlier relative to these se-quences. This is also reflected in the gene organiza-tion of these GLUTs. GLUT6 has been shown tohave a quite different structure from GLUT10 with10 exons, spaced by moderate intron lengths of 164–1012, transcription start site at position 2474. Nei-ther Ibberson et al. (6) or Doege et al. (7) published agenomic structure for GLUT8/X1, so we performed acomputational analysis to determine the putativestructure. A perfect Smith-Waterman alignment(GCG best fit, gap penalty 250/21, adjusted foronsensus donor/acceptor splice sites) of the mRNAGenBank Accession Number NM_014580) with theeference human genomic sequence (GenBank Ac-ession Number AL445222) revealed a structureimilar to GLUT6 of 10 exons (sizes 59, 163, 207,00, 197, 144, 109, 174, 146, 557) with intervening

4 All nucleotide positions in this report are quoted relative to

FIG. 2—Continued

192 DAWS

the A of the initiation codon ATG as position, 11, and by conven-tion, no nucleotide is present at position 0. All amino acid posi-tions are superscripted and numbered relative to the first methi-onine residue.

ntrons (99, 366, 175, 1950, 906, 155, 750, 428,546), and transcription start site at position 23data not shown). Therefore genomic structure anal-sis definitively places GLUT8 and GLUT6 as aubgroup with GLUT10 as an outlier with a differ-nt structure. This simple phylogram model broadlyupports previous alignments for this family of hex-se transporters (7).Analysis of the human GLUT10 proximal pro-oter region using the human genomic clone se-

uence (GenBank Accession Number AL031055)oes not show an obvious TATA box, but there is autative weak CCAAT box at 2479, and GRAIL-XP 3.0 (35) predicts a CpG Island between 2197nd 2249 (74% GC). A MatInspector (36) search ofhe Transfac database with the proximal 1 kb ofequence reveals two putative HNF3beta bindingites at 2628 (sense) and 2645 (antisense). Unfor-

tunately, the corresponding mouse genomic clone isnot yet publicly available to corroborate these putativeelements through comparative genomic alignment.

Human and mouse GLUT10 retain several se-quence motifs characteristic of the mammalian glu-cose transporters including VP497ETKG in the cyto-plasmic C-terminus, G73R[K,R] between TMD2 andTMD3 (PROSITE PS00216), VD92RAGRR betweenTMD8 and TMD9 (PROSITE PS00216), Q242QLTGin TMD7, and tryptophan residues W430 (TMD10)and W454 (TMD11), that correspond to W388 and W412

of GLUT1 (all positions are for the human ortholog).These tryptophan residues have previously been im-plicated in GLUT1 cytochalasin B binding and hex-ose transport (37). Neither human nor mouseGLUT10 retain the full P[E,D,N]SPR motif afterLoop6 but instead is replaced with P186AG[T,A]. APROSITE search (http://www.expasy.ch/prosite)also shows that GLUT10 has lost the SUGAR_TRANSPORT_2 pattern (PS00217), a result of thesubstitution G113S in TMD4, while all other humanGLUTs retain the glycine and the pattern match.The significance of this substitution is unknown.Sites for N-linked glycosylation are predicted atN334ATG between TMD8 and TMD9 and N526STG inthe cytoplasmic C-terminus.

A FASTA search (ktup 5 2) of the nonredundanttranslated GenPept database at NCBI using GLUT10protein sequence reveals that the highest scoringmatches are to metabolite transport-like proteins in

T AL.

Bacillus subtilis (CAB15600, CAB07473), and a hypo-thetical protein in Arabidopsis thaliana (AAG50560)which have better similarity scores and lower E values

oePtwbaNsalaspi(r(

pawlmsmoutGGpotscab

(rustllcrof

F

tcfGFGGutGf

TYPE

than the best match to a heterologous mammalianGLUT, human GLUT8. This is in contrast to analo-gous searches with human GLUT8 which matchesGLUT6 (and vice versa) and GLUT1 which matchesGLUT3 and GLUT4 with better alignment and E val-ues than matches to orthologous nonmammalianGLUT proteins. However, these prima facie candidaterthologs in B. subtilis and A. thaliana still lack thextended Loop9 sequence of GLUT10 and retain theESPR motif. A sensitive FASTA search (ktup 5 1) of

he extended Loop9 against NCBI GenPept reveals aeak similarity to the carboxyl end of gorilla car-oxyl-ester lipase (Accession Number AAF17700)nd human bile salt-activated lipase (Accessionumber AAA63211), 35.9 and 37.0% identity, re-

pectively, as well as human carboxyl ester lipasend cholesterol esterase. This region of the humanipase includes 16 proline-rich repeat units thatre extensively O-glycosylated (38,39). The Loop9imilarity to the C-terminal end of the lipase isrobably a reflection of the local compositional biasn the loop. Loop9 is relatively rich in proline16/86 5 18.6%) and contains 14 serine/threonineesidues excluding the putative N-glycosylation siteN334ATG).

Tissue Expression Analysis

The GLUT10 cDNA was used to probe a poly(A)RNA multiple-tissue Northern blot. A single 4.4-kbtranscript was detected in all tissues examined, withlevels highest in heart, lung . liver, skeletal muscle,

ancreas . brain, placenta . kidney (Fig. 3A). Inddition to the 4.4-kb transcript, two larger bandsere also present in pancreas. The origin of these

arger bands has not yet been investigated, but theyay represent alternatively spliced GLUT10 tran-

cripts or cross-hybridization with related GLUTs. Aore extensive tissue expression survey was carried

ut using RT-PCR analysis. PCR was performedsing first-strand cDNA prepared from 24 humanissues and oligonucleotide primers specific forLUT10 or actin as a control. As shown in Fig. 3B,LUT10 mRNA was readily detected in liver, lung,lacenta, salivary gland, thyroid, adrenal, pancreas,vary, prostate, and skin. This widespread distribu-ion of GLUT10 mRNA was also observed in aearch of existing ESTs. The 39 end of the GLUT10

GLUT10 AND

DNA matches the UniGene cluster Hs178603 thatre represented by 44 ESTs from various tissue li-raries including aorta (1 EST), bone (3 ESTs), brain

lci

1 EST), foreskin (4 ESTs), heart (2 ESTs), parathy-oid (6 ESTs), prostate (1 EST), testis (1 EST),terus (8 ESTs), and whole embryo (3 ESTs). Inearching the EST database, several other represen-ative ESTs were also identified from liver, kidney,ung, pancreas, neuron, fetal brain, fetal heart, fetaliver, and fetal lung libraries. A search of the mouseDNA sequence against the mouse EST databaseevealed only 3 ESTs with high identify scores: tworiginated from mouse mammary gland and onerom a mixed-tissue library.

unctional Analysis of GLUT10

The X. laevis oocyte expression system was usedo determine whether GLUT10 is a functional glu-ose transporter. Capped mRNA were transcribedrom the full-length human GLUT3, GLUT4, orLUT10 cDNA and used to inject Xenopus oocytes.igure 4A shows the 2-DOG transport mediated byLUT10 with respect to human GLUT3 andLUT4. GLUT10-injected oocytes exhibited 2-DOGptake that was 5-fold over the water-injected con-rols and was similar to the uptake in GLUT3 andLUT4-injected oocytes. As shown in Fig. 4B, a 100-

old excess of either 2-DOG or D-glucose effectivelycompeted with radioactive 2-DOG for uptake. D-Ga-lactose also inhibited 2-DOG uptake but less effec-tively. In contrast, a 100-fold excess of fructose didnot inhibit 2-DOG uptake (data not shown). Glucoseuptake was also inhibited by phloretin (Fig. 4C), ageneral inhibitor of mammalian glucose transport-ers. To determine the affinity of GLUT10 for glucose,GLUT10 mRNA-injected oocytes were incubated for30 min with increasing concentrations of [3H]2-DOG. Preliminary studies had shown that transportof 25 mM [3H]2-DOG was linear with time for up to60 min at 22°C (data not shown). The transport of[3H]2-DOG by GLUT10 mRNA-injected oocytes wassaturable and of relatively high affinity (Fig. 5). Asdetermined by Eadie-Hofstee analysis (Fig. 5, inset),GLUT10 mediated 2-DOG uptake with an apparentKm of approximately 0.3 mM and a Vmax of 0.85 pmoloocyte21 30 min21.

DISCUSSION

Glucose transport plays a central role in metabo-

1932 DIABETES

ism. Defects in glucose transport have been impli-ated in the reduced insulin sensitivity and in thencreased insulin resistance observed in Type 2 di-

can getissu

r hum

abetes (16–18,40,41). In addition, GLUT4 knockoutmice appear to have normal glucose homeostasis butdecreased insulin sensitivity (42,43). GLUT1–5 havebeen extensively analyzed for mutations contribut-ing to Type 2 diabetes, but no common causativemutation has been identified. The speculation thatadditional mammalian glucose transporters existhas been supported by the recent identification oftwo novel human glucose transporters, GLUT8(6–8) and GLUT9 (9). We report the characteriza-tion of an additional human glucose transporter,GLUT10.

FIG. 3. Tissue expression of GLUT10 mRNA in human tissuesa commercially available Northern blot (Clontech 7760-1) containiof GLUT10 expression in 24 human tissues. A human Rapid-Samplification with GLUT10-specific oligonucleotide primers. TheParallel control reactions were performed with the use of water oon a 1% agarose gel and visualized with ethidium bromide.

194 DAWS

GLUT10 was identified through a survey of ex-pressed sequences in the Type 2 diabetes-linked re-gion of human chromosome 20q12–13.1. The gene is

encoded by 5 exons spanning 26.8 kb of genomicDNA and is present in the finished sequence of theHuman Genome Project clone RP1-28H20 (Gen-Bank Accession Number AL031055). The GLUT10cDNA encodes a 541 aa protein that shares signifi-cant identity with other known members of the hu-man GLUTn (n 5 1–6, 8, 9) family. Analysis of theGLUT10 structure using a range of TMD predictionprograms, hydropathy analysis, and the presence ofkey conserved motifs suggests that GLUT10 is ahuman sugar transporter with 12 TMDs, but thereis a residual possibility that the structure is variant

full-length human GLUT10 cDNA was labeled and used to probeg of poly(A) RNA from the indicated tissues. (B) RT-PCR analysisne expression panel (OriGene Technologies) was used for PCRe sources of the human cDNA are indicated within each panel.an genomic DNA (Gen DNA). The PCR products were separated

T AL.

. (A) Ang 2 m

ON E

from the family consensus. The GLUT10 amino acidsequence is nearly identical in length to the recentlypublished GLUT9 homolog (541 aa versus 540 aa)

al(a

toshfa

w3tiApi

ssdtitspttdcbtmamthl

dGmEw2

wp

TYPE

but ranges from 3% (GLUT2; 524 aa) to 13%(GLUT8; 477 aa) longer than other known membersof the human GLUTn family (n 5 1–6, 8). Thedditional residues in GLUT10 are found in theonger exofacial Loop9 between TMD9 and TMD1091 aa), as compared to the loops for GLUT8 (32 aa),nd GLUT9 (11 aa).In contrast, the extra GLUT9 residues comprise

he longest cytoplasmic N-terminus sequence of anyf the known GLUTs. McVie et al. have reportedimilar findings in their independent evaluation of

FIG. 4. GLUT10 mediated 2-DOG uptake in Xenopus oocytes.(A) Uptake of 2-DOG (25 mM, 30 min at 22°C) was measured 3

ays after injection of water, or 30 ng of GLUT3, GLUT4, orLUT10 RNA into Xenopus oocytes. Each bar represents theean of duplicate determinations using 10 oocytes per assay. (B)ffect of competitors on 2-DOG uptake. Xenopus oocytes injectedith GLUT10 RNA (30 ng) or water were incubated for 30 min at2°C with 250 mM [3H]2-DOG in the presence of 25 mM 2-DOG,

D-glucose, or D-galactose. (C) Effect of phloretin on 2-DOG uptake.Xenopus oocytes injected with GLUT10 RNA (30 ng) or water

ere incubated for 30 min at 22°C with 25 mM [3H]2-DOG in theresence of 100 mM phloretin.

GLUT10 AND

uman GLUT10 (26), although there are a few dif-erences. We have cloned the full cDNA sequencend report exon 1 to be 254 bp in length, 250 bp of

hich is 59 UTR. Our Exon 5 is 2588 in size and the9 UTR region displays single base differences rela-ive to their cDNA cloned sequence (AF321240) thats identical to the reference genomic sequence inL031055. These could be SNPs manifested in ourlacenta library and will be the subject of furthernvestigation.

As expected, the corresponding murine GLUT10hows high sequence similarity to human, with pre-umably a similar TMD/loop structure, although theistance between the protein sequences (77.4% iden-ity) is greater than for murine/human GLUT1 (97%dentity) or GLUT8 (86% identity). When consideredogether, the multiple sequence alignment and re-ulting phylogram, gene organization and resultingrotein structure, and motif analysis all suggesthat GLUT10 is a member of a distinct subfamily ofhe human GLUTs. The GenPept protein sequenceatabase search across all species revealed signifi-ant matches in nonmammalian species that scoredetter than intra-human GLUT family matches, buthese protein sequences do not display the sameotif or a special TMD/loop structure that might

llow us to confidently propose putative nonmam-alian orthologous members of this subfamily. With

he murine protein as the only apparent ortholog touman GLUT10, it currently appears that the evo-

utionary event(s) that distinguish the distinct

FIG. 5. Concentration-response curve for 2-DOG uptake. Xe-nopus oocytes injected with GLUT10 RNA (30 ng) or water wereincubated with the indicated concentration of [3H]2-DOG for 30min at 22°C. The oocytes were washed and lysed to determineassociated radioactivity. Each point represents the average of 4 to

1952 DIABETES

10 oocytes. The uptake values were corrected for the backgrounduptake in water-injected oocytes. Inset: Eadie-Hofstee analysis ofuptake data revealed an apparent Km for 2-DOG uptake of 280mM and a Vmax of 0.85 pmol oocyte21 30 min21

as

aGitPiiK

kt

ttGb

ON E

GLUT10 paralog occurred relatively recently in amammalian or premammalian ancestor.

GLUT10 carries two N-glycosylation sites, one inexofacial Loop9, a feature that is shared withGLUT6 and GLUT8, and uniquely, a second one inthe cytoplasmic C-terminus. In the case of GLUT1,N-glycosylation has been shown to increase trans-porter affinity for glucose in leukemic cells (44), andis required for transporter activity, but is nonessen-tial for GLUT1 translocation (45). In addition, thePro/Ser/Thr composition of Loop9 suggests thatGLUT10 could be O-glycosylated, possibly confer-ring altered stability and/or conformational rigidityon the protein (46).

The most striking difference in human and mu-rine GLUT10 motifs relative to the other GLUTs isthe loss of one of the four characteristic motifs of thesugar transporter family, P[E,D,N]SPR in the cyto-plasmic end of TMD6/beginning of Loop6, and re-placement with P186AG[T,A]. Although the exactfunction of this PESPR motif is unknown, it is be-lieved to be a key structural component for trans-porter function. The moiety SPR is potentially aprotein kinase C phosphorylation site, and so it isalso possible that the motif may combine a TMD6proline structural component that is retained, whilelosing the phosphorylation site. GLUT10 also car-ries two other possible PKC sites; T194HK in Loop6,nd T370 AK in Loop9; and a PKA phosphorylationite R515RFT in the cytoplasmic carboxyl tail se-

quence. Experimental evidence exists from studiesof other human GLUT family members that serine/threonine sites in the C-tail are phosphorylated un-der appropriate conditions. GLUT4 has been shownto be phosphorylated in vivo (47–49) at the serine inthe unique TPS488LL C-tail sequence, but does notppear to be of prime importance in the movement ofLUT4 to the plasma membrane under conditions of

nsulin recruitment (50). However, the phosphoryla-ion may still play a role in intracellular sorting.KA-dependent phosphorylation has also been ver-

fied for GLUT2 in pancreatic beta cells in vivo andn vitro at the C-tail serines in KGKS491 andKS503GS505 (51, residue numbers adjusted)GLUT10 demonstrates replacement of most of the

ey residues in TMD2 shown to be critical to GLUT1ransport activity, including (GLUT1 positions) F72,

G76, and G79, as well as N288 in TMD7 (52). These

196 DAWS

residues may line one edge of the TM helices andpromote tighter packing of the helices forming thechannel, or exofacial cleft, for transport of sugar

moieties. However, the replacement Leu and Alaresidues in GLUT10 is also relatively small, andmay preserve the steric environment. In commonwith GLUT6 and GLUT8, GLUT10 TMD8 appearsto have lost its amphipathic character relative toGLUT1, while TMD10 appears to gain amphipath-icity. The amphipathic and highly conserved char-acter of TMD8 (GLUT1–4) was critical in earliermodels of the transporters as a putative lining for anaqueous sugar transport channel (53), but alaninemutagenesis studies of this domain (54) question thevalidity of this claim. GLUT10 and other recentlydiscovered GLUTs 5, 6, 9 show more divergence inthis TMD, Mutagenesis of the proline-rich TMD10GPGPIP motif in GLUT3 showed that the replace-ments P381A or P385A significantly reduced Km for2-DOG transport (54). Both GLUT10 and GLUT8show higher affinities than the other GLUTs andhave one or both of these prolines replaced. It istheorized that this proline-rich motif confers confor-mational flexibility for the reversible endo-to-exofa-cial conformation change for transport of the sugar.

Functional analysis of the expressed GLUT10gene strongly suggests that GLUT10 is a glucosetransporter. When expressed in Xenopus oocytes,GLUT10 functioned as a glucose transporter with anapparent Km of ;0.3 mM. This affinity is higherhan that of GLUT4 and GLUT8, the highest affinityransporters of the GLUT family identified to date.LUT10-mediated glucose transport was inhibitedy D-glucose and D-galactose but not fructose. The

level of glucose transport in our Xenopus system wassignificantly lower than that seen in some otherpublished descriptions of glucose transporters butwas similar to that recently reported for GLUT8(6,8). Analysis of the GLUT10 mRNA tissue distri-bution suggests that GLUT10 is expressed in a widevariety of glucose-metabolizing tissues such as liver,muscle, and pancreas. As such, GLUT10 may func-tion similarly to GLUT4 and be responsible formaintaining the relatively normal glucose ho-meostasis in the homozygous GLUT4 knockout mice(42,43).

The presence of HNF3beta sites in GLUT10 isconsistent with the confirmed regulation of GLUT2by the HNF3beta factor (55); GLUT2 is predomi-nantly expressed in hepatocytes, pancreatic betacells, intestine, and kidney, a subset of the tissues in

T AL.

which GLUT10 is also expressed. This suggests pos-sible targets for future studies of the transcriptionalregulation of GLUT10.

TYPE

In summary, sequence and functional analysis ofthe GLUT10 gene indicate that it codes for a novelhigh-affinity glucose transporter. The glucose trans-port properties of GLUT10, combined with the local-ization of this gene to a region of chromosome 20that is both linked and associated with Type 2 dia-betes in genetic studies, make GLUT10 an intrigu-ing candidate Type 2 diabetes susceptibility gene.

ACKNOWLEDGMENTS

We thank Drs. Greg Shelness and Mark Lively for helpfuldiscussion and assistance with the manuscript. This work wassupported by NIH Grants R01 DK41269, R01 DK56289, and R01DK53591 to DWB, and R01 DK47987 to P.A.D.

REFERENCES

1. Hediger MA, Roads DB. Molecular physiology of sodium-glucose transporters. Physiol Rev 74:993–1026, 1994.

2. Mueckler M, Caruso C, Baldwin SA, Panico M, Blench I,Morris HR, Allard WJ, Lienhard GE, Lodish HF. Sequenceand structure of a human glucose transporter. Science 229:941–945, 1985.

3. Fukumoto H, Seino S, Imura H, Seino Y, Eddy RL, Fuku-shima Y, Byers MG, Shows TB, Bell GI. Sequence, tissuedistribution, and chromosomal localization of mRNA encod-ing a human glucose transporter-like protein. Proc NatlAcad Sci USA 85:5434–5438, 1988.

4. Kayano T, Fukumoto H, Eddy RL, Fan YS, Byers MG,Shows TB, Bell GI. Evidence for a family of human glucosetransporter-like proteins: sequence and gene localization ofa protein expressed in fetal skeletal muscle and other tis-sues. J Biol Chem 263:15245–15248, 1988.

5. Fukumoto H, Kayano T, Buse JB, Edwards Y, Pilch PF, BellGI, Seino S. Cloning and characterization of the major insu-lin-responsive glucose transporter expressed in human skel-etal muscle and other insulin-responsive tissues. J BiolChem 264:7776–7779, 1989.

6. Ibberson M. Uldry M. Thorens B. GLUTX1, a novel mam-malian glucose transporter expressed in the central nervoussystem and insulin-sensitive tissues. J Biol Chem 275:4607–4612, 2000.

7. Doege H, Schurmann A, Bahrenberg G, Brauers A, JoostHG. GLUT8, a novel member of the sugar transport facili-tator family with glucose transport activity. J Biol Chem275:16275–16280, 2000.

8. Carayannopoulos MO, Chi MM-Y, Cui Y, Pingsterhaus JM,McKnight RA, Mueckler M, Devaskar SU, Moley KH.GLUT8 is a glucose transporter responsible for insulin-stim-ulated glucose uptake in the blastocyst. Proc Natl Acad SciUSA 97:7313–7318, 2000.

GLUT10 AND

9. Phay JE, Hussain HB, Moley JF. Cloning and expressionanalysis of a novel member of the facilitative glucose trans-porter family, SLC2A9 (GLUT9). Genomics 66:217–220,2000.doi:10.1006/geno.2000.6195

10. Doege H, Bocianski A, Joost HG, Schurmann A. Activity andgenomic organization of human glucose transporter 9(GLUT9), a novel member of the family of sugar-transportfacilitators predominantly expressed in brain and leuco-cytes. Biochem J 350:771–776, 2000.

11. Burant CF, Takeda J, Brot-Laroche E, Bell GI, DavidsonNO. Fructose transporter in human spermatozoa and smallintestine is GLUT5. J Biol Chem 267:14253–14256, 1992.

12. Davidson NO, Hausman AML, Ifkovits CA, Buse JB, GouldGW, Burant CF, Bell GI. Human intestinal glucose trans-porter expression and localization of GLUT5. Am J Physiol262:C795–C800, 1992.

13. Olson AL, Pessin JE. Structure, function, and regulation ofthe mammalian facilitative glucose transporter gene family.Annu Rev Nutr 16:235–256, 1996.

14. Seidner G, Abrarez MG, Yeh JI, Driscoll KR, Klepper J,Stump TS, Wang D, Spinner NB, Birnbaum MJ, De VivoDC. GLUT-1 deficiency syndrome caused by haploinsuffi-ciency of the blood-brain barrier hexose carrier. NatureGenet 18:188–191, 1998.

15. Santer R, Schneppenheim R, Dombrowski A, Gotze H, Stein-mann B, Schaub J. Mutations in GLUT2, the gene for theliver-type glucose transporter, in patients with Fanconi-Bickel syndrome. Nature Genet 17:324–326, 1997.

16. Butler PC, Kryshak EJ, Marsh M, Rizza RA. Effect of insu-lin on oxidation of intracellularly and extracellularly de-rived glucose in patients with NIDDM. Evidence for primarydefect in glucose transport and/or phosphorylation but notoxidation. Diabetes 39:1373–1380, 1990.

17. Rothman DL, Magnusson I, Cline G, Gerard D, Kahn CR,Shulman RG, Shulman GI. Decreased muscle glucose trans-port/phosphorylation is an early defect in the pathogenesisof non-insulin-dependent diabetes mellitus. Proc Natl AcadSci USA 92:983–987, 1995.

18. Cline GW, Petersen KF, Krrsak M, Shen J, Hundal RS,Trajanoski Z, Inzucchi S, Dresner A, Rothman DL, ShulmanGI. Impaired glucose transport as a cause of decreased in-sulin-stimulated muscle glycogen synthesis in type 2 diabe-tes. N Engl J Med 341:240–246, 1999.

19. Bowden DW, Sale MM, Howard TD, Qadri A, Spray BJ,Rothschild CB, Akots G, Rich SS, Freedman BI. Linkage ofgenetic markers on human chromosome 20 and 12 toNIDDM in Caucasian sib pairs with a history of diabeticnephropathy. Diabetes 46:882–886, 1997.

20. Ghosh S, Watanabe RM, Hauser ER, Valle T, Magnuson VL,Erdos MR, Langefeld CD, Balow J, Ally DS, Kohtamaki K,Chines P, Birznieks G, Kaleta H-S, Musick A, Te C, Tan-nenbaum J, Eldridge W, Shapiro S, Martin C, Witt A, So A,Chang J, Shurtleff B, Porter R, Kudelko K, Unni A, Segal L,Sharaf R, Blaschak-Harvan J, Eriksson J, Tenkula T,Vidgren G, Ehnholm C, Tuomilehto-Wolf E, Hagopian W,Buchanan TA, Tuomilehto J, Bergman RN, Collins FS,Boehnke M. Type 2 diabetes: Evidence for linkage on chro-mosome 20 in 716 Finnish affected sib pairs. Proc Natl Acad

1972 DIABETES

Sci USA 96:2198–2203, 1999.

21. Ji L, Malecki M, Warram JH, Wang Y, Rich SS, KrolewskiAS. New susceptibility locus for NIDDM is localized to hu-man chromosome 20q. Diabetes 46:876–881, 1997.

ON E

22. Zouali H, Hani EH, Philippi A, Vionnet N, Beckman JS,Demenais F, Froguel P. A susceptibility locus for early-onsetnon-insulin dependent (type 2) diabetes mellitus maps tochromosome 20q, proximal to the phosphoenolpyruvate car-boxykinase gene. Hum Mol Genet 6:1401–1408, 1997.

23. Price JA, Brewer CS, Howard TD, Fossey SC, Rich SS,Freedman BI, Weurth J-P, Bowden DW. Construction of aphysical map of chromosome 20q12-q13.1 and linkage dis-equilibrium analysis in diabetic nephropathy patients. Am JHum Genet 58:A241, 1997.

24. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ.Basic local alignment search tool. J Mol Biol 215:403–410,1990.

25. Dawson PA, Fossey SC, Mihic SJ, Craddock AL, Mycha-leckyj J, Snyder JR, Bowden DW. GLUT 10: A novel glucosetransporter in the Type 2 diabetes susceptibility linked re-gion of chromosome 20q12-13.1. Am J Hum Genet 67:A50,2000.

26. McVie AJ, Lamson DR, Chen YT. Molecular cloning of anovel member of the GLUT family of transporters, SLC2A10(GLUT10), localized on chromosome 20q13.1: A candidategene for NIDDM susceptibility. Genomics 72:113–117,2001.doi:10.1006/geno.2000.6457

27. Price JA, Brewer CS, Howard TD, Fossey SC, Sale MM, Ji L,Krolewski AS, Bowden DW. A physical map of the 20q12-q13.1 region associated with type 2 diabetes. Genomics 62:208–215, 1999. doi:10.1006/geno.1999.6007

28. Fossey SC, Mychaleckyj J, Snyder JR, Pendleton JA, Bow-den DW. A high resolution 6 Mb transcript map of the type2 diabetes susceptibility locus on human chromosome20q12-13.1. Genomics 76, 45–57, 2001.

29. Kozak M. Interpreting cDNA sequences: Some insight fromstudies on translation. Mamm Genome 7:563–575, 1996.

30. Hirokawa T, Boon-Chieng S, Mitaku S. SOSUI: Classifica-tion and secondary structure prediction system for mem-brane proteins. Bioinformatics 14:378–379, 1998.

31. Cserzo M, Wallin E, Simon I, von Heijne G, Elofsson A.Prediction of transmembrane alpha-helices in prokaryoticmembrane proteins: the dense alignment surface method.Protein Eng 10:673–676, 1997.

32. Tusnady GE, Simon I. Principles governing amino acidscomposition of integral membrane proteins: Application totopology prediction. J Mol Biol 283:489–506, 1998.

33. Sonnhammer EL, von Heijne G, Krogh A. A hidden Markovmodel for predicting transmembrane helices in protein se-quences. Proc Int Conf Intell Syst Mol Biol 6:175–182, 1998.

34. Hofmann K, Stoffel W. TMBASE: A database of membranespanning protein segments. Biol Chem Hoppe-Seyler 374:166, 1993.

35. Hyatt D, Snoddy J, Schmoyer D, Chen G, Fischer K, ParangM, Vokler I, Petrov S, Locascio P, Olman V, Land M, ShahM, Uberbacher E. Improved Analysis and Annotation Toolsfor Whole-Genome Computational Annotation and Analysis:

198 DAWS

GRAIL-EXP Genome Analysis Toolkit and Related AnalysisTools. Genome Sequencing & Biology Meeting, May 2000.

36. Quandt K, Frech K, Karas H, Wingender E, Werner T.MatInd and MatInspector: New fast and versatile tools for

detection of consensus matches in nucleotide sequence data.Nucleic Acids Res 23:4878–4884, 1995.

37. Mueckler M. Facilitative glucose transporters. Eur J Bio-chem 219:713–725, 1994.

38. Baba T, Downs D, Jackson KW, Tang J, Wang CS. Structureof human milk bile salt activated lipase. Biochemistry 30:500–510, 1991.

39. Wang CS, Dashti A, Jackson KW, Yeh JC, Cummings RD,Tang J. Isolation and characterization of human milkbile salt-activated lipase C-tail fragment. Biochemistry 34:10639–10644, 1995.

40. Shepherd PR, Kahn BB. Glucose transporters and insulinaction. Implications for insulin resistance and diabetes mel-litus. New Engl J Med 341:248–257, 1999.

41. Tirosh A, Rudich A, Bashan N. Regulation of glucose trans-porters. Implications for insulin resistance states. J PediatrEndocrinol Metab 13:115–133, 2000.

42. Katz EB, Stenbit AE, Hatton K, DePinho R, CharronMJ. Cardiac and adipose tissue abnormalities but not dia-betes in mice deficient in GLUT4. Nature 377:151–155 1995.

43. Stenbit AE, Tsao TS, Li J, Burcelin R, Geenen DL, FactorSM, Houseknecht K, Katz EB, Charron MJ. GLUT4 het-erozygous knockout mice develop muscle insulin resistanceand diabetes. Nature Med 10:1096–1101, 1997.

44. Ahmed N, Berridge MV. N-glycosylation of glucose trans-porter-1 (Glut-1) is associated with increased transporteraffinity for glucose in human leukemic cells. Leukocyte Res23:395–401, 1999.

45. Samih N, Hovsepian S, Aouani A, Lombardo D, Fayet G.Glut-1 translocation in FRTL-5 thyroid cells: Role of phos-phatidylinositol 3-kinase and N-glycosylation. Endocrinol-ogy 141:4146–4155, 2000.

46. Van den Steen P, Rudd PM, Dwek RA, Opdenakker G.Concepts and principles of O-linked glycosylation. Crit RevBiochem Mol Biol 33:151–208, 1998.

47. James DE, Hiken J, Lawrence JC. Isoproterenol stimulatesphosphorylation of the insulin-regulatable glucose trans-porter in rat adipocytes. Proc Natl Acad Sci USA 86:8368–8372, 1989.

48. Lawrence JC, Hiken JF, James DE. Phosphorylation of theglucose transporter in rat adipocytes. Identification of theintracellular domain at the carboxyl terminus as a target forphosphorylation in intact-cells and in vitro. J Biol Chem265:2324–2332, 1990.

49. Lawrence JC, Hiken JF, James DE. Stimulation of glucosetransport and glucose transporter phosphorylation by oka-daic acid in rat adipocytes. J Biol Chem 265:19768–19776,1990.

50. Marsh BJ, Martin S, Melvin DR, Martin LB, Alm RA, GouldGW, James DE. Mutational analysis of the carboxy-terminalphosphorylation site of GLUT-4 in 3T3-L1 adipocytes. Am JPhysiol 275:E412–422, 1998.

51. Thorens B, Deriaz N, Bosco D, DeVos A, Pipeleers D., Schuit

T AL.

F, Meda P, Porret A. Protein kinase A-dependent phosphor-ylation of GLUT2 in pancreatic beta cells. J Biol Chem271:8075–8081, 1996.

52. Olsowski A, Monden I, Krause G, Keller K. Cysteine scan-

TYPE

ning mutagenesis of helices 2 and 7 in GLUT1 identifies anexofacial cleft in both transmembrane segments. Biochem-istry 39:2469–2474, 2000.

53. Zeng H, Parthasarathy R, Rampal AL, Jung CY. Proposedstructure of putative glucose channel in GLUT1 facilitativeglucose transporter. Biophys J 70:14–21, 1996.

54. Seatter MJ, Kane S, Porter LM, Arbuckle MI, Melvin DR,Gould GW. Structure-function studies of the brain-type glu-

GLUT10 AND

cose transporter, GLUT3: alanine-scanning mutagenesis ofputative transmembrane helix VIII and an investigation ofthe role of proline residues in transport catalysis. Biochem-istry 36:6401–6407, 1997.

55. Cha JY, Kim H, Kim KS, Hur MW, Ahn Y. Identification oftransacting factors responsible for the tissue- specific ex-pression of human glucose transporter type 2 isoform gene.Cooperative role of hepatocyte nuclear factors 1alpha and3beta. J Biol Chem 275:18358–18365, 2000.

56. Thompson JD, Gibson TJ, Plewniak F, Jeanmougin F, Hig-gins DG. The CLUSTAL_X windows interface: Flexiblestrategies for multiple sequence alignment aided by quality

1992 DIABETES

analysis tools. Nucleic Acids Res 25:4876–4882, 1997.57. Saitou N, Nei M. The neighbor-joining method: A new

method for reconstructing phylogenetic trees. Mol Biol Evol4:406–425, 1987.