Quick viewing(Text Mode)

The Yeast SNF3 Gene Encodes a Glucose Transporter Homologous to the Mammalian Protein (1Acz Fusions/Membrane Proteins/Immunofluorescence/Glucose Repression) JOHN L

The Yeast SNF3 Gene Encodes a Glucose Transporter Homologous to the Mammalian Protein (1Acz Fusions/Membrane Proteins/Immunofluorescence/Glucose Repression) JOHN L

Proc. Nail. Acad. Sci. USA Vol. 85, pp. 2130-2134, April 1988 Biochemistry The yeast SNF3 gene encodes a transporter homologous to the mammalian protein (1acZ fusions/membrane proteins/immunofluorescence/glucose repression) JOHN L. CELENZA, LINDA MARSHALL-CARLSON, AND MARIAN CARLSON* Department of Genetics and Development and Institute for Cancer Research, Columbia University College of Physicians and Surgeons, New York, NY 10032 Communicated by David Botstein, December 23, 1987

ABSTRACT The SNF3 gene is required for high-affinity Construction of SNF3-lacZ Fusions. SNP'3-lacZ gene fu- glucose transport in the yeast and sions were constructed by inserting SNF3 DNA fragments has also been implicated in control of gene expression by containing at least 0.8 kilobases of 5' noncoding sequence glucose repression. We report here the nucleotide sequence of into the 2-Am plasmid vectors YEp353 and YEp357R (11). the cloned SNF3 gene. The predicted sequence The SNF3(3)-lacZ fusion was constructed by inserting an shows that SNF3 encodes a 97-kilodalton protein that is EcoRI-BamHI fragment into YEp353, SNF3(321)-lacZ by homologous to mammalian glucose transporters and has 12 inserting a Sal I-Spe I fragment into YEp357R, and SNF3- putative membrane-spanning regions. We also show that a (797)-lacZ by inserting a Sal I-EcoRI fragment into YEp- functional SNF3-lacZ gene-fusion product cofractionates with 357R. membrane proteins and is localized to the surface, as Preparation of Glucose-Derepressed Cells. Wild-type judged by indirect immunofluorescence microscopy. Expres- (SNF3) yeast cells carrying each of the SNF3-lacZ gene sion of the fusion protein is regulated by glucose repression. fusions were grown to midlogarithmic phase in synthetic complete medium containing 2% glucose, with selection for the plasmid. Cells were derepressed by shifting to medium Transport of glucose in the yeast Saccharomyces cerevisiae, containing 0.05% glucose for 3 hr. as in most mammalian cell types, occurs by carrier-mediated Protein Fractionation and Enzyme Assays. Crude lysates facilitated diffusion (1-3). Bisson and Fraenkel (4, 5) showed were prepared from glucose-derepressed cells as described by kinetic analysis of glucose uptake that yeast has a (12), except that protease inhibitor was omitted, and were constitutive low-affinity transport system (Km 20 mM for fractionated by centrifugation at 13,000 x g in a microcen- glucose) and a glucose-repressible high-afflinity system (Km trifuge (Fisher) at 4°C. Protein concentrations were deter- 1 mnM). Previously, Neigeborn and Carlson (6) described mined with the Bio-Rad protein-assay kit. p-Galactosidase the isolation of mutations in the SNF3 (sucrose nonferment- (13), dipeptidyl aminopeptidase (14), and glucose-6-phos- ing) gene that caused defects in regulation of glucose-repres- phate dehydrogenase (15) were assayed as described. sible genes, and Neigeborn et al. (7) reported the cloning of Indirect Immunofluorescence. Cells were fixed with form- SNF3 and the construction of snJ3 null mutations. Mutations aldehyde, treated with Glusulase (DuPont) and Zymolyase in SNF3 cause defects in growth on glucose at low concen- lOOT (Seikagaku Kogyo, Tokyo), immobilized on polyly- tration (0.1%) and on sugars such as sucrose, which is sine-coated slides, and stained with antibody and 4',6- hydrolyzed extracellularly by invertase to yield glucose and diamidino-2-phenylindole (DAPI), essentially as described fructose. Subsequent studies showed that snJ3 mutants lack (16). Mouse monoclonal antibody to f3-galactosidase was the high-affinity glucose-uptake system, whereas low- purchased from Promega Biotec (Madison, WI), and fluo- affinity uptake appears to be unimpaired (8). These findings rescein isothiocyanate (FITC)-conjugated sheep antibody to suggested the possibility that SNF3 encodes a structural mouse immunoglobulin G was purchased from Sigma. Cells component of the high-affinity glucose-uptake system. We were examined with a Zeiss photomicroscope III equipped therefore determined the sequence of the cloned SNF3 for epi-illumination fluorescence with standard FITC and gene,t and we report here that SNF3 encodes a protein that DAPI filter sets. Cells were photographed with Kodak is homologous to mammalian glucose transporters. We also T-MAX 400 film and a x 40 objective. Exposure times were show that a functional SNF3-lacZ gene-fusion product cof- 3 min for FITC fluorescence and 2 min for DAPI fluores- ractionates with membrane proteins and is localized to the cence. cell surface. Expression of the gene fusion is regulated by glucose repression. RESULTS Sequence Homology Between the SNF3-Encoded Protein and Mammalian Glucose Transporters. We determined the MATERIALS AND METHODS nucleotide sequence of a 3690-base-pair region of the previ- Nucleotide Sequence Analysis. Intact and BAL-31 nucle- ously cloned (7) SNF3 gene. Fig. 1 shows the nucleotide ase-treated restriction fragments from the SNF3 gene were sequence and the inferred amino acid sequence. The se- cloned in the vectors M13mpl8 and M13mpl9 (9). The quence of the predicted 96,713-dalton protein (884 amino sequence of both strands was determined by the method of acids) was compared to the sequences of the glucose trans- Sanger et al. (10) with the 17-nucleotide sequencing primer porters from human HepG2 hepatoma cells (17) and rat brain (Amersham). (18). Significant homology (28% identity) was found between Computer Methods. Amino acid sequences were aligned with the assistance of the program ALIGN (30). Abbreviations: DAPI, 4',6-diamidino-2-phenylindole; FITC, fluo- rescein isothiocyanate. *To whom reprint requests should be addressed. The publication costs of this article were defrayed in part by page charge tThis sequence is being deposited in the EMBL/GenBank data base payment. This article must therefore be hereby marked "advertisement" (Bolt, Beranek, and Newman Laboratories, Cambridge, MA, and in accordance with 18 U.S.C. §1734 solely to indicate this fact. Eur. Mol. Biol. Lab., Heidelberg) (accession no. J03246).

Downloaded by guest on September 29, 2021 2130 Biochemistry: Celenza et al. Proc. Natl. Acad. Sci. USA 85 (1988) 2131

-853 ~~~~~~~~~~~~~~~~~~~~~~~~~~GTCGACTTCTTrGA -840AAAGATG'CGTAACTGCCAAGA'ATkACTTAGATTAAAGTAAA'CTTTCA''GAAACGATATAC -720 ACACACATTTCTACCTGATAGAGAAAAAkAAGTTGCTC'GTAArGTC--kC'G'CtGGGG''ATG

Met Asp Pro Asn Ser Asn Ser Ser Ser Gin Thr Leo Arg Gin Gin Lys Gin Gly Phe Leu Asp Lys Ala Leo Gin Arg Val Lvs Cv Ilie 30 I ATG CAT CCT AAT AGT NAG ACT TCT AGC GA.A ACA TTA CCC CAA GAG AMA CAG ATA GAC GAG AAC TCA ATT TIA TTT TCA GAG CCT CCt CAC Ala Leu Ar gAr g Asn Asn Set Asn Lys As p His Thr Thr As pAs Thr Thr GIy Ser Ilie Arg TFr Pro Thr Ser Leol Gin Arg Gln Asn 60 91 AAA CMA TCT ATG ATG ATG TGT MAC AIA CTCAT ACA ACA CAT CAT AGC ACA CCI AGC ATA CGA ACG CCI ACC AGC TTG GAG CCC CMA MT 5cr As p Gin Ser -Asn Met Thr Set Val Phe Thr As p As pIlie Ser Thr Ilie As As p sn Set Ilie Leo Phe Set Gin Pro Pro Gin 90 181 TCT GAC ArmAGC CM TCT MAT AIG ACA TCG GTG TTT ACG CAT GAC ATT TCT ACC ATA GAC CACMCI TCA AlT TTA TTT TCA GAG CCI CCT GAG Lys Gin Set Met Met Met Set tIle Cys Vat ClyV Val The Val l a l l Te.tensThelytrM ItGy eUlejAsn Set 12u 271 AAA CMATCTATG AIGATG TCTIATAITUO GTA CIGil TT GT MCA CIT CCC IMG !TTTAXIIGGT AUCT ACA' GTt GTGATCMAC ACT lie Thr Set Met Asn It- Val Lys Set His Val Ala Pro Asn His As pSet Phe Thr Ala Gin Gin Met Set'l eTru Val erPhe Leu 150 3o1 ATT ACA TCT ATG MAC iAr CTG MAG TCA GACGTCA GCA CCT MAT CAC CAT TCA TTT ACC CCC CMA CMA ATC TCC TTflGTOGTATTTT2 G SerLeuGlyThrPhePheGI laLen Thr Ala Pro Phe Ile -Set As p Set lyr Gly ArgLys Pro Thr leT IlePhe Ser Thr-lieiPh 180 451 C TGGGA ACT TTT TIT GC OCT TA ACT OUCAOA ITT AlATICGAT TCC IAT CCC AGG MkG CCI ACT C5 2TTIGCAGTACA ATIT T IlePheSerIlGl As Se Lu G-n al lyAla Gly Gly Ilie Thr Leo Le Ilie Vai Gly Arg a Il~e Sex Gly Ile GI Ile GIy 213 541 AT T C T G A TTACGGAGT=G ATC ACA TTA TIC ATT GIG CCA AG GT ATTA GGT ATOC CC AlA. CC AlaIle Set Al# Vat Vat Pro Len Ty-r Gin Ala Gin AlaiThr His Lys Set Leu Argr ati leSlet ThrT Gh IrAe i 24 631 GCA TOTTC TTGTOCA TTA TAC CAA GGA GAA GC1 AGA CAT AMA TGA TTA AGA 001 MT AnT ATT ITT ACT TAG CAA TOXGCCC ATT iThr:TIrQy ten tan, Vat Set- Set Ala: Set Gin Gly STh His Aia Ar g Ason As 1)Ala Set Set Tyr Arg-7le Pr Ol Le Gin 270 721 AO CCCCIr,T=TGCTGI GTG [CA ACT OGA GTj TCG CAA CCC AGCA CAL GCA AGAAAMG GAL GGA TCT TCG PATlG CCCA±TACCCITRG CA t Vat Itr Set Set Phe Len Ala Ile Gly Met Phe Phe Len Pro 'in Set Pro Arg Ivr Tvr Val Leo Lys Asp Lys Leo As GIn Aia 3100) 811 T1T1Ir TOG10 QTC1 TT CCG CC ATC GGC AIG TTG TIT GIG CCT 'AGAT CGA CCC T'AT TAG GTT TTG AMA GAG MAG CTA CAT GMA GGA Aia Lvs Set Leo Set Phe Leo Arg Civ Val Pro Vai His As p Set Gly Leo Len Gin Glu Leo Val Gin TIle Lvs Ala Thr Tyr As p Tvr 330 901 GGT AAA TGT TTA TGG TTT TTA AGA CCI GTA GGA GTG CAT CAT TGT CCC TTA CTC GMA GM- GTA GTT CMA ATA ACG GGA ACA TAT CAT T.AG Ciu Ala Set Phe Clv Set Set Asn Phe Iie Asp Cys Phe Iie Set Set Lvs Set Atg Pro Lys Gin Thr Leo Arg et Ph ITt Gly 13$ 360 991 GAG CGA TCT ITT CCi TGT TGC AAG TTG ATT CAT TCT ITT ATT TGA ACT AAA ACT AGA GGA MAG CMA ACT GTA AG6CATG TTT AGO GGA AIT Ala Len Gin Ala Phe Gin Gin Phe Se' 01 Ilie Asan he tIle Pha Iyr Tsr Civ Val As n Phe Phe [K T7Yhr Ciy Val 5cr Asn Set 390 1081 CCC CTI CIA OGA TTT CAA CAA TTT ICA COT ITO 1A0 ITI ATA ITTT Tk k ~ TC MAT TITC TIrC MAT MGAGA GCA GTC AGT MAT AGT TreVat Set Pa tIle Thr yr Ate Vat IAn Vai Vat The IAn Vat Pro Gly Lau Phe Phe Val-lGin Phe Phe Clv Arg Arg LyVat1 s2U 1171 TAT 10011 IGA TTT ATA AGO TAIT CCI OTT AAI OTT GIG TIT AAT GTT GOT CCI TIC TTT TTI GTG0CM ITT TTT CCI AGA CCI AGLG: Len Val Val Giv Gly Vel Ilie Met Thr Ilie Ala IAsn The TIle Val Ala Ilie Val CiCa Set Leol Lys Thr Gal Ala Ala Ala Lys Vai1 1450 1261 CGI01OTT CCGG QGT CIT AIC AIC ACT ATA GCC lAG TTT All GIG, GCC All GTT GaC i~ IC TAAMG ACT CIA CCC CCC GCA AA T Met lie Ala Pe tIle Gys Len Phe Ile Ate Ala Pha Ser Ate Thr Trp Gy GI Vat Vat ITrp Vat li~e Set Ala Ginu Leo Tsr Pro LeoL 480 1351 AIC AlA GCA TTT AlA 101 CIA TIC AlA CCI GCG TTT TOT OCCI ACA TOG 001001 IIGIT TR_OTTIC llPT TCA GCA GMA GIG TAG CCA TIC Gla Lys Ges Thr Ala Ilie Ces Ala tla Ala Msn Itr Leu Vat Asan Phe Ile C a Ala Len lie ITt Pro r lie Vat Asp 5~10 141GG{ GTG AGuSetIIA ITAGCITAT&C OCT OCT OCT lAACGI TTGTlCA lAO TTT AllTtrOTIIAlAOGT~IAIGl Thr Ciy Set His Thr Set Set Len Gly Ala Lys l~e t Pte Ilie Tp G1l Ser LUeu Isa Ala Met GIy VaI le Val VaIyT Ln k 54o 1531 ACT CCC ICC CAT ACA ICA TGA TTA CCI GGA MAIAAIC TIC AT! IC CC TCO TA AAT- CCC ATG CCC GIG LTA OTl GTT TI-C IT ICC VlTrGin 1Tr Lys Civ Leo Thr Len Gin Gin Ilie As p Gin Leoi Tsr Ilie Lvs Set Set Tlir CGI Vai Gal Set Pro Lvs Phe Aan Lys SIC 1621 TT CMA ACG MAG GC1 TIC ACA TTA GMA GAG ATl CAT CMA TIA TAP All A:AC ICA ICC ACT GGI GIG GIG ICA CCA AAAk III AAI MA~ As p Ilie Arg GI Arg Ala Leu Lys Phe Gin Tyr Asp Pro Leo Gin Arg Len Gin As Clyv Lys Aen Thr Phe Gal Ala Lys, Arg Mnr Asn 6J30 1711 CATI All AC C GM' CCC GCA CIT AkA FTIC CMA TACGCAT CCI TIC CAA AGA TIA GMA GA GCGA MAG AAC ACT ITT GTT GCC MkA AGAAMI MT Phe As As Gin Thr Pro Arg An As pPhe Are4 As n Thr lie Set Giv G in Ilie Asp His 11cr Pro A-nr GIn Lyes GIn11'.Ga1 nis Set IIe 630 1801III C~~~~~~GACCA GMA ACA CGA AGA M'T" CTITT CCXAAMT ACC AlA ICC GGt CMA AlA CAT CAT ACT CCC MAT GM- AXA C;AA CIT CAl ICT ATC Pro Gin Arg Gal As p Lie' Pro Thr Set Th r Gin lie Leo G Lin Set Pro Asn Lvs Set Set Clv% Met Thr VolI Pro Vol1 Set Pro Set Lenl 60 1891 CGA GM- CCI CIT C"AT~ AllF CCI ACT ACT ACA GMA AlT CII GAA ACC CGGe MAC AAAGAT ACT CCI ATG ACA GIG CCI GIG TCA CL'I PCI CGI Gin As paGa Pro Ilie Pro Gi1n Th r 1St G lu Pro Ala G I iiLe Arg TI-1r Lys Tv r Vol1 As p Leo CvAsnI ClIv Leo CL 1v Leo Aso- 'Thr Tsr 690 1981 CAAMGAGCT CCA AIC CCC CAA ACA ACA GAG CCT CCI GAA AT1 CG'A ACC AAkA TAT CGIC GAG CIA CCGA MT GUC CIT CCI CIT 441 ACC TAT Aen Arg Clv Pro Pro Set Leo Set.r Set Asp Set Set Gin Asp Ivr 1Th G I As p Gin Ilie Glis i Pro Set Set G-In (Civ Asp Gin Set 720 2071 AAT AGA CCC CCI CCC IGA GIGCICA AGC GAG, IGACL A;C GA, CAT TAG ACA CAA GAT GAA AlA GHt [GG CCC IGA TCT CAA GG~ GAG GM ACT Asn Arg Set Th r Met Asan As pIlie AS As' Tv r MIet A\Ia Ari Leoulie His Set Tntr Set 1Th AlIaSSet Asnr Thr Th r As p Lvs Phe Set lI 21o1 MAT AGA ACT ACT ATC AAT GAIAlT AAI GAA~ G I~A l GGATAIW ~ G ACT MC ACG AC AillM TIC ICC

GCIv Asn Gi1n Set Th r LeoL At' ITy u ~ Ala Set Set His -S)et Asp 1Sti T',ir G i Li Ass) Set As n Le% Met Asp1 Leuo Civ Aen Clv 7t80 GIG GGA XAC 2251 CCI MAC CMA ACT ACC CIFT C)~TAG-k, CAC 'AC C GCI ICC TCA CAT ICC CAT ACA ACT GAA CCiA GAG AGG M.T TIC 410 G"-AG CCL~ Lenu Ala LeoL Asn Ala Iycr Aen Arg Civ Pro Pr'oc Set Ilie Lel MAt an Set Set -~ss via Go Al1a Asn GCIv Civ G- i niTr Setr Asp Aen 81I0 2341 CII CCC TIC MAT CCIr TrAP MG AGA CCI CCA CII IGA All 114 410G ANAT ICC ACT (Ak o'A- oX' UGA MiT CCI; CCGI GAG ACCIG CAT MT Leo Asen STh AlIa Gin As *LeoL haIGI v Mlet Lys GIn Arg; Met AbIs Gi1n Pie Ala aTA*Lyi s A, I v iv Len G in Pro 0-PJ 2431 TTG MAC AGA CCI CMCAAL Tic; ULT' GGT ATC AC~ CAN CG(,~ATGO1CCCAG TIPr~ CCG AGAC TAT t CL CAGAU CCG CICC A C

1 Gin Th r Gin Set hAns i II e ii S('r Thr Satr anL Set a1M tA1 As C. ,r ANs G in. His Asn As: Gin I Leni Hi s Set Set Gin GiI II ) 2521 GAA ACT CAA ICT MT AlT~ ETC ACC ACTr ICF CEC TICC nUr AG -ACLATIGA A T NAT GAA ATIC CTC IN G C MG A~sn Ala 1Tr As n GIn Pro. Gal NAn Ci hu'Ahn 'As's. ane 1y *5 * 0 261 AAC GCCC ACT MAT C-AA CCI CITA AAT CGAA NAT A41 C;AY FECGAl' IA'STTAAiTTKrAICTIGTTATTLCGATCPIATTAAIIAAGLCGCACGIGC CTNATTK

FIG. 1. Nucleotide sequence of the SNF3 gene and predicted amino acid sequence of the gene product. Putative membrane-spanning regions are shaded. The 23 amino acid duplication is underlined; solid lines indicate conserved residues. The unshaded box indicates a potential N-linked site that is predicted to lie on the extracellular face of the membrane. Asterisks indicate termination codons. Nucleotides are numbered on the left and amino acids on the right. The SNF3-lacZ fusion constructions confirm that the first ATG codon can serve as a translational start site and that the SNES coding sequence extends to codon 797. The size of the coding region is consistent with the size of the 3-kilobase SNF3 RNA (7).

amino acids 86-581 of the SNF3 protein and the HepG2 is also homologous, as are the mammalian glucose transport- (Fig. 2), which is 97.6% identical to the ers, to the arabinose-H+ (AraE) and rat transporter (18) (sequence not shown). The SNF3 protein xylose-H + (Xy1E) transporters (19) (Fig. 2). Downloaded by guest on September 29, 2021 2132 Biochemistry: Celenza et al. Proc. Natl. Acad. Sci. USA 85 (1988)

SNF 3 86 S SPQS M MESCG VF A F- T_nF D HLVPINIS M Y V K S A H DRSF HGT M[ S S K L T G - R L H L A GA V- LGSIL PQ K VI E F Y NQ T WV Y XylE --- MNT YNSSYQFS ITL L GILILI-IF OYD V II -WG V E S LNTVF A QNLYENL II T P R S L R D A A GA L P F I T AraE T RrJN M F[iSE |A LV L Lv D F V

141 T A P S Y - P T SNF 3 Q Q M - TFT F F F IlD IWF S T Q HGT 56 1 L P TT L T T L W IF G IYG -S V L FVNFG -NSMLMMLL Az S A V L M G XyIE 53 A N S L LG F C VA S A LI CII I GAL G G Y C N R F G RR D S 1 K EJA A V F V GO2A W AraE 59 R L Q E W V S S M MGA A I G A LF N W L SF LRGR Y S L M A G A I-VLES I G S G SNF 3 19 -.AGO IT L -RV I S IflA A Y I A T H K S 1) HOT 112 PS.-K -K SFE M|LI FGR _ VY C:G L TT C F YP _ V S P T A F Xyll 105 P E L C F T S I N P D NT V P V YL V P E -- F V I YR I I G & V (GI L A M T, P Y I 1 A P A li 1 07 - Ara; A F S V E M - - = A A R L .I V1 &~ I A SY T A Z LY 1.SM A SiE' hik

SNF 3 229 R& AI T WLVA T]W L -V S S A V SQ CT H A R-[-I-JY V w s HGT LGTLQLGVV GI----LI A Q V F G L D S I J W P LL LS I IFI XylE 160) |R C-L-F-QWFINDGIF|G|Q - - --L|LI V Y C N Y F I A R SV -RY---JWL NT D - M F A 146 R G K V S D T S G M L A AraH MM L M VLTLL I L.A FPL JFUS1Y WWEA - V 1 1.1AVL

SNF3 276 Y V KS- 1) F- F[9Afl TSPRY L[D[2L DEA SFL G|VP I1 . HGT - - -WV L IFCE------IP E S P RIF L L I N R N E K7N RA KJVILI K K|LL T A DV T H E}LJ : KF1 Xy IF. 206 E C I P A LLI|F M LL Y T VIP E SP RIW 1. M S R GWQ E II]N- R K I T L A rQT AV Ara: 191 -QIM L V V FL .- Ki-GA H I ]E E V -j R M T S K K A F. INjI 1. SNF3 325 T D F R Q JA Y[DY EWSI S S N I C I S SES K Q T[IRBT A-t--Qh--( A F Q S G IN F HGT 249 R Q MM R E K K V T I 1.E, 1 R S - A YR Q P I L IT --~ A V L Q Q W: S G I N A K ---LJ VJF Kyl 257 JH S L7[ H G R K.G- REJ L V G V I V I RV M L F Q 0 v G I N V V AraE 240) K Q GG W L.JVLIV L -- GINR L IQM

SNF 3 3 7f y T IG V N F VSNN S Y - -9V S F Y An V V N G - - 7 F G KR K VQYV V T G V I HGT 292 x Y s 'r S I F EL A G VQ Q P V Y A T I G S G IT A F T V S S V y A Gv K T i 1.L I G L A G; XylE 298 Y Y A P E V F K T L G [STT) I A LMQ T I G V I JL T F TV L A I - - - M TIV[W Kj K P 1, ( A f; AraE 279 APR 11F K M AG F TT E - - Q Q M I ALUL VE L TJ A T F I A VWT K KP A, K I;

428 M i SNF3 TWA N I V A VWC S KIElV A -- K - A FI L I AA F S T HGT 344 A GCMA I L M T1A L A LLEQL P W I S- Y 11 S V A F G V A[F F G G P W VAIV] 1. Xy1E 352 M Af G M jSS1 - - -DT A F Y[IQ P I A - L L S M L Y V A A A S W P W V Ara; 334 [MJALGTLVLCYCLMQFDNG TSSOSWLSWO-MTMM[EIA AP--

SNF3 478 YP R SEC T JC A NW.T LVNFI FC A L I T Yl EV T G S H TinSIS K I W S HGT 395 F S QUJPERP A AIV A G. F A|NNWTISV GMC FJV E Q L C - Y VI - F T VL XylE 399 F A I R GE]A VST SWSW HFHNFSYWYSL(cM 388 K C D F G I T C T T T N S T L - L9 I G W - Y TA Ara; LIAA.S MlII PE]D -bKNC[

SNF 3 53(0 W1A G II V VY L T [E T K G L [LE F I 1) KiE Y I K SS T G V V S P KF N K 1) I R A 1. K F EY HGT 441 LVLFF FT K VI PITKGRI A SO EQ GASQS r P LL)F H SPVP 1. G A KylE 452 G V L AA IP F M W K FPjE T K GK T L E EIL E AE]W E P E T K K T QQ T A TI. AraLE 4 3, MI1 A FEP G I T F W L I P1 T K N V T 1L El HI E R K L M A G E K L R N C; V FIG. 2. Sequence homology of the predicted SNF3 protein, human HepG2 glucose transporter (HGT) (17), and E. coli xylose-H + (XyJE) and arabinose-H + (AraE) transporters (19). Sequences (given in standard single-letter amino acid symbols) were aligned to maximize identity to the SNF3 sequence. Dashes indicate gaps introduced to optimize the alignment. Identities with the SNF3 protein are boxed. Putative transmembrane regions in the SNF3 and HepG2 transporters are shaded. Amino acids are numbered on the left. The hydrophobicity profile suggests that the SNF3 protein for the mammalian transporters, the SNF3 protein has no contains 12 membrane-spanning regions (Fig. 3), as has been apparent signal sequence adjacent to the amino terminus. found for the mammalian and E. coli transporters. The Two lines of evidence indicate that the carboxyl-terminal 87 spacing of the putative membrane-spanning regions (shaded amino acids are not required for SNF3 function: both the in Figs. 1 and 2) is similar for the yeast and mammalian EcoRI fragment ending at codon 797 (7, 8) and a gene fusion glucose transporters, suggesting a similar structure. The between SNF3 and the E. coli lacZ gene at this site produce membrane-spanning regions are arranged in two groups of a functional transporter and complement an snJ3 null muta- six, separated by a region that is rich in charged residues (22 tion (see below). An interesting feature of the carboxyl- of 70 and 25 of 65 residues are charged for the SNF3 and terminal region is a sequence of 23 amino acids that occurs at HepG2 transporters, respectively). In several cases, un- two positions (680 and 776) with 17 of 23 residues identical charged polar residues within predicted membrane-spanning (underlined in Fig. 1). regions are conserved between the SNF3 and HepG2 pro- Another difference between the SNF3 and HepG2 trans- teins-for example, the glutamine and asparagine residues in porters is that the N-linked glycosylation site assigned to hydrophobic regions 7 and 11, respectively. Mueckler et al. Asn-45 of the human glucose transporter (17, 22) is not (17) suggested that these residues may be involved in form- conserved in the yeast protein, although there is a potential ing amphipathic a-helices that might function in a glucose glycosylation site (Asn-383) at an analogous position in the binding site or pore. second set of six membrane-spanning regions. Maiden et al. The most striking difference between the SNF3 protein (19) suggested that the two sets of six membrane-spanning and the mammalian glucose transporters is that the former is regions arose by gene duplication. much larger, with an additional 85 amino acids at the amino Cofractionation of SNF3-lacZ Gene-Fusion Proteins with terminus and 303 amino acids at the carboxyl terminus. We Membrane Proteins. To confirm that SNF3 encodes a mem- did not detect homology between these terminal regions and brane protein, we constructed gene fusions between codons the proteins in the Protein Identification Resource of the 3, 321, or 797 of SNF3 and codon 8 of the E. coli /3- National Biomedical Research Foundation.t As was found galactosidase gene lacZ (see Materials and Methods), which are named according to the fusion point in SNF3. For tProtein Identification Resource (1987) Protein Sequence Database simplicity, the protein products of these gene fusions will be (Nati. Biomed. Res. Found., Washington, DC), Release V14.0. referred to as SNF3-lacZ fusion proteins. /B-Galactosidase is Downloaded by guest on September 29, 2021 Biochemistry: Celenza et al. Proc. Natl. Acad. Sci. USA 85 (1988) 2133

-00. -o0 I

100 200 300 400 500 600 700 800 Residue number FIG. 3. Hydrophobicity profile of the predicted SNF3 protein. The profile was determined using the algorithm of Kyte and Doolittle (20) with the normalized consensus hydrophobicity values of Eisenberg (21) and a window of 21 amino acids. Twelve 21-residue segments have hydrophobicity values greater than 0.42; these putative membrane-spanning regions are numbered. a soluble protein but can be targeted to the membrane if it is SNF3(3)-lacZ fusion protein cofractionated with the soluble fused to a membrane protein (23). If SNF3 encodes a mem- marker glucose-6-phosphate dehydrogenase. Immunoblot brane protein, the SNF3(797)-1acZ fusion protein should co- analysis, using anti-j3-galactosidase antibody to detect the fractionate with membrane markers because it includes all 12 fusion proteins, verified that the ,B-galactosidase activity hydrophobic regions and, moreover, complements an snJ3 corresponded to the level of SNF3-lacZ protein (data not null mutation. The SNF3(321)-1acZ fusion protein might also shown). be expected to cofractionate with membrane proteins, since Localization of a Functional SNF3-lacZ Fusion Protein to it contains the first 6 hydrophobic regions; Mueckler and the Cell Surface. Indirect immunofluorescence microscopy Lodish (22) showed that the first 8 membrane-spanning re- was used to localize the SNF3-lacZ fusion proteins within gions of the HepG2 glucose transporter allow insertion into the yeast cell. The SNF3(797)-IacZ fusion protein was local- microsomes in vitro. The SNF3(3)-lacZ fusion protein, how- ized to the cell surface, presumably to the plasma mem- ever, should remain soluble. brane, since it functions to complement an snJ3 mutation To test these predictions, crude protein lysates prepared (Fig. 4a). The control, the SNF3(3)-lacZ gene product, was from cells carrying each of the three gene fusions were distributed throughout the cytoplasm but was excluded from fractionated by centrifugation at 13,000 x g, and the super- the vacuole (data not shown). Interestingly, the SNF3(321)- natant and pellet fractions were assayed for f3-galactosidase, lacZ protein was not localized to the plasma membrane but glucose-6-phosphate dehydrogenase, and the membrane- rather was found in a cap-like structure around the nucleus associated enzyme dipeptidyl aminopeptidase. The X- (Fig. 4 b and c). This finding suggests that the first six galactosidase activity associated with the SNF3(797)-lacZ membrane-spanning regions are not sufficient for correct and SNF3(321)-lacZ fusion proteins cofractionated with localization; however, we cannot exclude the possibility that dipeptidyl aminopeptidase (Table 1), whereas the control the presence of the p-galactosidase moiety at this position interferes in some way. Table 1. Cofractionation of SNF3-lacZ fusion proteins with Regulation of SNF3-lacZ Fusion Protein by Glucose Repres- membrane proteins sion. We next examined the regulation of the SNF3(797)- Enzyme activity lacZ fusion protein by glucose repression. Previous studies showed that expression of the SNF3 RNA is about 5-fold p-Galactosidase DPAPase G6PD glucose-repressible (7) and that high-affinity glucose trans- Fusion gene S P P/S P/S P/S port is glucose-repressible (5). Glucose-repressed and dere- pressed cells carrying the SNF3(797)-1acZ gene fusion were SNF3(797)-lacZ 7.5 37.0 4.9 5.3 0.2 assayed for,-galactosidase (24); repressed and derepressed SNF3 (321)-lacZ 7.5 55.0 7.3 5.0 0.2 cells produced, respectively, 0.8 and 12 units of activity SNF3(3)-lacZ 6.3 2.3 0.4 4.8 0.4 normalized to cell density as described by Miller (13) (aver- Lysates from glucose-derepressed cells carrying each of the ages of seven determinations; standard error <25%). Immu- SNF3-lacZ gene fusions were prepared and supernatant (S) and noblot analysis verified that the increase in enzyme activity pellet (P) fractions were obtained as described in Materials and reflected an increase in the amount of fusion protein. In Methods. DPAPase, dipeptidyl aminopeptidase. G6PD, glucose-6- cells carrying this gene fusion phosphate dehydrogenase. P3-Galactosidase activities in the S and P addition, glucose-repressed fractions are expressed as nmol of o-nitrophenyl ,3-D-galactoside did not stain with antibody to 3-galactosidase by immuno- cleaved per min per mg of protein. P/S, ratio of specific activity in fluorescence microscopy (data not shown). Preliminary data pellet to that in supernatant. Similar results were obtained in a suggest that expression of the SNF3(797)-IacZ fusion is duplicate experiment. affected by trans-acting genes that control expression of Downloaded by guest on September 29, 2021 2134 Biochemistry: Celenza et al. Proc. Natl. Acad. Sci. USA 85 (1988) physically, can be identified by isolating extragenic suppres- sor mutations that restore function in an snJ3 missense or null mutant. Genetic analysis should also yield insight into the regulation of SNF3 expression in response to glucose availability. We thank A. Myers and A. Tzagoloff for providing vectors and T. H. Steinberg for advice on microscopy. We thank M. Gottesman, S. Goff, C. Manoil, A. Mitchell, and K. Drickamer for helpful criticism of the manuscript. This work was supported by Grant GM34095 from the National Institutes of Health and by a Feasibility Grant from the American Diabetes Association. 1. Cirillo, V. P. (1962) J. Bacteriol. 84, 485-491. 2. Lang, J. M. & Cirillo, V. P. (1987) J. Bacteriol. 169, 2932-2937. 3. Bisson, L. F. & Fraenkel, D. G. (1983) J. Bacteriol. 155, 995-1000. 4. Bisson, L. F. & Fraenkel, D. G. (1983) Proc. Natl. Acad. Sci. USA 80, 1730-1734. FIG. 4. Localization of SNF3-lacZ fusion proteins by indirect 5. Bisson, L. F. & Fraenkel, D. G. (1984) J. Bacteriol. 159, immunofluorescence. Glucose-derepressed cells were stained with 1013-1017. mouse monoclonal antibody to 8-galactosidase and FITC-conju- 6. Neigeborn, L. & Carlson, M. (1984) Genetics 108, 845-858. gated sheep antibody to mouse immunoglobulin G, as described in 7. Neigeborn, L., Schwartzberg, P., Reid, R. & Carlson, M. Materials and Methods. Cells were also stained with DAPI to (1986) Mol. Cell. Biol. 6, 3569-3574. localize nuclei and mitochondria. (a) Cells carrying SNF3(797)-lacZ 8. Bisson, L. F., Neigeborn, L., Carlson, M. & Fraenkel, D. G. fusion; FITC fluorescence. (b and c) Cells carrying SNF3(321)-lacZ (1987) J. Bacteriol. 169, 1656-1662. fusion; FITC and DAPI fluorescence, respectively. No staining was 9. Norrander, J., Kempe, T. & Messing, J. (1983) Gene 26, detected when anti-3-galactosidase antibody was omitted or when 101-106. cells lacking an SNF3-lacZ gene fusion were examined. (Bar = S 10. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467. /Am.) 11. Myers, A. M., Tzagoloff, A., Kinney, D. M. & Lusty, C. J. other glucose-repressible genes. Regulation of glucose- (1986) Gene 45, 299-310. 12. Celenza, J. L. & Carlson, M. (1986) Science 233, 1175-1180. transporter levels in response to glucose availability also 13. Miller, J. L. (1972) Experiments in Molecular Genetics (Cold occurs in mammalian and avian cells (25, 26). Spring Harbor Lab., Cold Spring Harbor, NY). 14. Julius, D., Blair, L., Brake, A., Sprague, G. & Thorner, J. DISCUSSION (1983) Cell 32, 839-852. 15. Clifton, D., Weinstock, S. B. & Fraenkel, D. G. (1978) Genet- Three lines of evidence argue strongly that SNF3 encodes a ics 88, 1-11. glucose transporter: previous genetic evidence that SNF3 is 16. Adams, A. E. M. & Pringle, J. R. (1984) J. Cell Biol. 98, transport the se- 934-945. required for high-affinity glucose (7, 8); 17. Mueckler, M., Caruso, C., Baldwin, S. A., Panico, M., quence homology between the SNF3 protein and mamma- Blench, I., Morris, H. R., Allard, W. J., Lienhard, G. E. & lian glucose transporters; and the localization of a functional Lodish, H. F. (1985) Science 229, 941-945. SNF3-lacZ fusion protein to the plasma membrane. In 18. Birnbaum, M. J., Haspel, H. C. & Rosen, 0. M. (1986) Proc. addition to its transport function, the SNF3 protein may also Natl. Acad. Sci. USA 83, 5784-5788. be involved in regulatory mechanisms for monitoring the 19. Maiden, M. C. J., Davis, E. O., Baldwin, S. A., Moore, availability of glucose; certain snJ3 missense mutations D. C. M. & Henderson, P. J. F. (1987) Nature (London) 325, cause defects in and of 641-643. repression derepression glucose- 20. Kyte, J. & Doolittle, R. F. (1982) J. Mol. Biol. 157, 105-132. repressible genes (6, 7). The SNF3 protein is not, however, 21. Eisenberg, D. (1984) Annu. Rev. Biochem. 53, 595-623. absolutely essential for regulation, since glucose repression 22. Mueckler, M. & Lodish, H. F. (1986) Cell 44, 629-637. is normal in an snf3 null mutant (7). It is noteworthy that the 23. Hagen, D. C., McCaffrey, G. & Sprague, G. F., Jr. (1986) dependence of high-affinity glucose transport on the pres- Proc. Natl. Acad. Sci. USA 83, 1418-1422. ence of hexose (2-4) suggests interaction between 24. Guarente, L. (1983) Methods Enzymol. 101, 181-191. the SNF3 protein and kinases. Hexokinase P11 (B) has been 25. Haspel, H. C., Wilk, E. W., Birnbaum, M. J., Cushman, implicated in control of glucose repression (27-29); thus, the S. W. & Rosen, 0. M. (1986) J. Biol. Chem. 261, 6778-6789. regulatory effects of the SNF3 protein might be mediated by 26. Shawver, L. K., Olson, S. A., White, M. K. & Weber, M. J. (1987) Mol. Cell. Biol. 7, 2112-2118. interaction with hexokinase PII. 27. Entian, K.-D. & Frolich, K.-U. (1984) J. Bacteriol 158, 29-35. Mutational analysis of the structure and function of this 28. Entian, K.-D., Hilberg, F., Opitz, H. & Mecke, D. (1985) Mol. glucose transporter by use of the cloned SNF3 gene will be Cell. Biol. 5, 3035-3040. facilitated by the availability of snj3 null mutants of yeast. 29. Ma, H. & Botstein, D. (1986) Mol. Cell. Biol. 6, 4046-4052. Moreover, proteins that are functionally related to the 30. Dayhoff, M. O., Barker, W. C. & Hunt, L. T. (1983) Methods SNF3-encoded transporter, including those that interact Enzymol. 91, 524-545. Downloaded by guest on September 29, 2021

Top View