|
- 积分
- 0
- 威望
- 0
- 包包
- 0
|
作者：Tatiana Blasco, José J. Aramayona, Ana I. Alcalde, Julia Catalán, Manuel Sarasa, Victor Sorribas作者单位：Departments of Toxicology, Pharmacology, Physiology, Genetics, and Anatomy,University of Zaragoza, Zaragoza E5001 Spain
; V" p( @' j4 k9 ^- F4 M
& V! d: U3 G' s5 i- Y 4 t# Q5 I% l' t
# K# R/ ]7 l( y' }( b ! L/ q- ?1 W Q6 c( G$ i
g+ x9 B5 @$ i$ }
4 S" `. t7 e. v6 J- \2 X
* |4 q8 ?* Y6 B( [% m) h0 v
/ V$ R1 l* Q7 j# Y& B: N
% @) e3 s0 b L' w
5 S4 T4 [9 F( H6 f; X/ V7 ~( H& ~3 } 5 `1 g# t2 S4 `2 ^# c: c$ y# v
3 \! A" t1 z* p+ s/ r8 w w) u
【摘要】: R7 e! M- a( S5 t6 x& Z( h: J6 V
Renal reabsorption is the main mechanism that controls mannose homeostasis.This takes place through a specific Na-coupled uphill transport system, themolecular identity of which is unknown. We prepared and screened asize-selected rat kidney cortex cDNA library through the expression of mannosetransport in Xenopus laevis oocytes. We have identified a membraneprotein that induces high-affinity and specific Na-dependent transport of D -mannose and D -glucose in X. laevis oocytes,most likely through stimulation of the capacity of an endogenous transport system of the oocyte. Sequencing has revealed that the cDNA encodes thecounterpart of the human membrane-associated protein MAP17, previously knownby its overexpression in renal, colon, lung, and breast carcinomas. We showthat MAP17 is a 12.2-kDa nonglycosylated membrane protein that locates to thebrush-border plasma membrane and the Golgi apparatus of transfected cells andthat it is expressed in the proximal tubules of the kidney cortex and in thespermatids of the seminiferous tubules. It spans twice the cell membrane, withboth termini inside the cell, and seems to form homodimers throughintracellular Cys 55, a residue also involved in transportexpression. MAP17 is responsible for mannose transport expression in oocytesby rat kidney cortex mRNA. The induced transport has the functionalcharacteristics of a Na-glucose cotransporter (SGLT), because D -glucose and -methyl- D -glucopyranoside are alsoaccepted substrates that are inhibited by phloridzin. The correspondingtransporter from the proximal tubule remains to be identified, but it is different from the known mammalian SGLT-1, -2, and -3.
1 d. x8 g* J1 B0 S, A- p* y 【关键词】 expression cloning kidney cortex mannose reabsorption y- v! j I; \3 ^. K6 J
DIETARY SUPPLY IS THE MAIN source of D -mannose inhumans. Once absorbed, the fate of mannose depends on the cell type, but it ismainly focused on membrane protein glycosylation. In fact, mutations inenzymes involved in mannose handling for glycosylation and metabolism causeserious syndromes, such as several congenital disorders of glycosylation( 1 ) and -mannosidosis ( 3 ). In addition toglycosylation, sperm has the special feature of being able to directly usemannose as an energy source ( 23 ). D -Mannosehomeostasis seems to be controlled by the reabsorption activity of the kidney,because most of the ultrafiltrated mannose is readily reabsorbed in theproximal tubule ( 25, 27, 29, 39 ). This important role hasbeen studied for more than 30 years using in vivo and in vitro experiments anddifferent animal species ( 5, 10, 22, 25, 27, 29, 30, 33, 39 ). These studies haveconcluded that the brush-border membrane of the renal tubular cells contains ahigh-affinity ( K m 0.1 mM) and very specificNa-dependent D -mannose transport system. Stoichiometricdeterminations have evidenced Na- D -mannose relationships of 1:1( 10, 22 ) and 2:1( 5 ), a difference that can beexplained by the use of different animal species, as well as experimental anddata analysis approaches. With respect to specificity, despite being inhibitedby D -glucose, -methyl- D -glycoside, andphloridzin, the Na-mannose cotransport system seems to be different from theknown Na-dependent D -glucose transport system( 5, 10, 22, 25, 27, 29, 30, 39 ). Several groups have alsoreported a strong inhibition of renal mannose transport by D -fructose, which could be explained by direct competition for thesame transport system ( 5, 22, 25, 27, 39 ).# ?7 M7 c1 O6 \8 n
! k( M7 U0 D. y n% o v- h" z x
The molecular characterization of this renal transport was initiated by ourgroup a few years ago ( 5 ). Wereported that the size of the rat kidney RNA responsible for D -mannose transport expression in Xenopus laevis oocyteswas exceptionally small ( 1 kb) compared with all other known transporters(2-5 kb). The transport induced by the 1-kb-fraction-enriched mRNA inoocytes was also very small ( 100% greater than in water-injectedoocytes), but it exhibited kinetic characteristics similar to the transport measured by the renal brush-border membrane vesicles.
/ M1 }7 T6 [/ S5 A( u2 F% }8 u* E; ?( {. f1 L$ V5 E6 i+ Z
To determine the physiological relationship of this small RNA to mannosetransport in the kidney, we have identified by expression cloning the cDNAresponsible for the Na-coupled D -mannose transport induction in X. laevis oocytes. The kinetic behavior is similar to that in thedata published for mannose transport in rat kidney and other animal models.The cDNA sequence reveals that it is the rat counterpart to the human 17-kDamembrane-associated protein MAP17/DD96( 16, 17 ), whose molecularcharacteristics are very different from those of all known transporters. MAP17 was first cloned as an mRNA that was overexpressed in most carcinomas, whichthen led to the identification of type 1 PSD95-Dlg-zona occludens-1 (PDZK1), aPDZ domain-containing globular protein that interacts with the COOH terminusof MAP17 ( 18 ). In this work,we also show evidence of the need for interaction with an additionalprotein(s) in the oocyte, which is functionally similar to, but differentfrom, the known Na-glucose cotransporters SGLT-1, -2, and -3( 38 ).
% F i/ F5 t& }4 d! q3 F" M2 ^# D6 ^6 C8 X$ P& @1 Q0 I
MATERIALS AND METHODS
7 h% S. I2 X% y g' D, M5 ]% ?4 Y6 `' W
Preparation, injection, and uptake assays with oocytes. The general methods for handling X. laevis and their oocytes and fortransport assays have been described previously( 37 ). Oocytes were injectedwith different amounts of capped cRNA, using an automatic nanoliter injector(World Precision Instruments, Hertfordshire, UK). In vitro transcriptions wereperformed with an mMESSAGE mMACHINE kit (Ambion, Austin, TX). Transport assayswere performed at room temperature for 30 or 60 min using D -[ 3 H(2,6)]mannose (Amersham Biosciences,Buckinghamshire, UK) at 40-80 µCi/ml as a tracer, according to the cold substrate concentration. For Na-dependent uptake, groups of 10 oocyteswere incubated in an uptake medium [(in mM) 100 NaCl, 2 KCl, 1CaCl 2, 1 MgCl 2, and 10 HEPES-Tris, pH 7.5] containing the cold and radiolabeled substrates. For Na-independent uptake, NaCl wasequimolecularly substituted by choline chloride. Before the uptake assay, theoocytes were washed for 2 min in uptake solution without Na at roomtemperature, and after the specific incubation time the radioactive medium wasrecovered, and the oocytes were washed four times with 4 ml each of ice-coldwashing solution. Except where indicated to the contrary, all data show transport in the presence of NaCl./ S4 u& E# r* j9 B
9 w T: {' b7 O* xCloning of MAP17. Directional cDNA library construction and expression cloning were performed as published( 14, 31 ) using the SuperscriptPlasmid System (Invitrogen, Paisley, UK). The first cDNA strand was methylatedusing 5-methyl-dCTP, transformed into Epicurian Coli XL-Gold ultracompetentcells (Stratagene, La Jolla, CA), and screened at 500 colonies/plate.Sequencing was automatically performed using AbiPrism 377 (PerkinElmer LifeSciences, Boston, MA) and Vector NTI Suite software.
( [% _+ q5 u3 `8 w+ I
- H: w7 k9 g8 l6 ?8 g4 ?Culture and transfections of opossum kidney cells. Cell culture and uptake assays were performed as reported( 32 ). For transient transfections, the cDNAs in pCMV Script (Stratagene) were transfected usingLipofectamine Plus reagent (Invitrogen) at 80% confluence for 4 h. Eachsix-well vessel received 2 µg DNA, 6 µl Plus reagent, and 8 µllipofectamine, and an additional 24 h were allowed for protein expression.Permanently transfected clones were selected using 500 µg/ml G418 andmaintained at 200 µg/ml( 12 ).
6 \+ w8 w7 e4 k7 D# U' v# w$ t2 S4 [) D# X/ V7 v
Metabolic labeling and in vitro translation. The metabolic labeling of oocytes was performed using 5 ng cRNA/oocyte and, 8 h later, 0.5µCi of a L -[ 35 S]methionine/cystein mixture (Amersham). After 12 h, the oocyte cell membranes were purified as published( 31 ). After SDS-PAGE, the gelwas fixed, dried, and exposed to Kodak Biomax MR film at -80°C using Kodak Biomax TranScreen LE intensifying screens (both from Amersham Biosciences). In vitro translation of rat MAP17 was done using rabbitreticulocyte lysate in the absence or presence of canine pancreatic microsomalmembranes (both from Promega, Madison, WI). Posttranslation modifications wereanalyzed by a combination of microsome addition and endoglycosidase Hdigestion exactly as published( 15 ).- N% @* m" T4 @. r8 f' F' C6 Y
$ w( o" x' z \ S& g/ U' t
Northern blot analysis and in situ hybridization. Total RNA waspurified using a QuickPrep Total RNA Extraction kit (Amersham Biosciences).For Northern blot analysis, 20 µg total RNA were electrophoresed informaldehyde-denaturing agarose gels as explained elsewhere( 2 ) and vacuum-blotted ontonylon membranes (Biodyne, Pall Gelman, Ann Arbor, MI). After hybridization inan ULTRAhyb solution (Ambion) with a 660-bp MAP17-derived 32 P-labeled riboprobe, signals were obtained by exposure to KodakBiomax MS film at -80°C for 4-8 h.6 [' i) f; n4 `! F& N7 n: |
) L L$ K1 V. n+ c' _. W
In situ hybridization histochemistry was done as published elsewhere ( 34 ). Five-micrometercryosections, thaw-mounted onto gelatin-coated slides, were air dried, fixedin paraformaldehyde, and dehydrated in ethanol. Forty-mer antisenseoligonucleotides (5'-AGATGGCTGTGATTCAAGAGAGGTGAGAGGTCAGCTTGTT) were3'-labeled with [32 P]dATP (New England Nuclear, Boston, MA)using terminal deoxynucleotidyl-transferase and hybridized with the tissuesections overnight under Nescofilm coverslips in a humid chamber at 42°Cin a hybridization buffer. The slides were washed at high stringency,dehydrated with ethanol, dipped into LM-1 emulsion (Amersham Biosciences), andexamined after development, using darkfield and Nomarski microscopy (Olympus BX60). As controls for in situ hybridization histochemistry, Northernblotting, the melting temperature of the hybrids formed, competition incohybridization experiments, and hybridization with probes for other mRNAswere all used as explained elsewhere ( 34 ). The melting temperatureof the hybrids was determined during the washing procedure and found to beclose to the values predicted by Primer Analysis software (Oligo 6.0,MedProbe).9 W' h: k, Y: [7 G/ _7 s! h1 R
, G+ O3 a/ j' o( q2 gMutant constructions. The hemagglutinin antigen (HA; YPYDVPDYA) was introduced in positions 5, 23, and 66 of the MAP17protein by site-directed mutagenesis using a QuikChange kit (Stratagene) andthe following primers (HA-encoding sequences are underlined):' K7 b; \# |% R6 L$ f
4 E' y& F4 i5 }% o1 ZHA5, 5'-TTGGCCCTC TACCCATACGACGTCCCAGACTACGCG AGTCTGCTC" I }/ w$ P* d
" S5 `: ^. c. L6 a
HA23, 5'-GCTGTCAACAA TACCCATACGACGTCCCAGACTACGCG GGTCTAGGGAACC1 Q9 N, ?3 M; O; _$ Y0 r' T
( Q7 o, j W4 N5 F: dHA66, 5'-GAGCACGATG TACCCATACGACGTCCCAGACTACGCG ATGATCACTG.% q/ O" i9 a' _) D% a
9 \% G/ `' k5 ]( m$ mFor cystein mutants, they were mutated to serines using the followingprimers (point mutations for Cys conversion are underlined): Cys 20 (C20S), CCGAAGTGGCACCTGCC A GCAGTCAACA; and Cys 55 (C55S), GCTCTTCCTGGCTCCAGA A GTGGTTGAC.
4 W" ?9 `6 P; H. ?& h1 M& k. y' |1 t3 E7 y, U0 l \$ A* L
Immunoblots, immunofluorescence, and confocal imaging. Crude membrane preparations were obtained from the oocytes expressing the HA-taggedMAP17 as published ( 31 ).Western blot analysis was performed using a rabbit polyclonal anti-HA antibody(Clontech, Palo Alto, CA), a goat anti-rabbit IgG secondary antibodyconjugated to horseradish peroxidase (Chemicon, Temecula, CA), and enhanced chemiluminescence detection (ECL Plus, Amersham Biosciences). Opossum kidney(OK) cells expressing HA-tagged MAP17 were immunodecorated basically asdescribed ( 21 ). Anti-HAantibody was diluted 1:100 in PBS and incubated for 2 h and then localizedwith a goat anti-rabbit IgG secondary antibody, FITC-conjugated (Sigma, St.Louis, MO), diluted 1:50, and incubated for 1 h. Mowiol 4-88(Calbiochem, San Diego, CA) was used as an antifading agent in the mountingmedia (Merck, Darmstadt, Germany). Cells were visualized by epifluorescencewith either an Olympus BX60 or a confocal Carl Zeiss LSM 310 microscope. Golgistaining and colocalization were performed using the ceramide analog BODIPY-TR(Molecular Probes, Eugene, OR) following the manufacturer's guidelines. Inshort, BODIPY-TR (1 mM) was dissolved in ethanol and incubated with the fixedcells as BSA-complexes, for 10 min at 5 µM, after a washing of thesecondary antibody. When necessary, the cells were incubated for 30 min at37°C with 10 µg/ml brefeldin A (Molecular Probes) before fixation, froma stock of 5 mg/ml in ethanol.
! h8 J7 C' J" E; X8 ]) r1 P" D. G7 D8 {+ D: J- b
Hybrid depletion in oocytes. Hybrid depletion experiments were performed as described previously( 31 ) using sixoligodeoxyribonucleotides derived from the rat MAP17 sequence as follows:sense oligos: 105-124 ATGTTGGCCCTCAGTCTGCT (S105), 251-270CGTCAACCACTTCTGGTGCC (S251), and 407-426 TGTTCTGGAGGAAGAGGGCA (S407);and antisense oligonucleotides: 105-124 AGCAGACTGAGGGCCAACAT (AS124),335-352 TATCTGCCATCCATGCCC (AS335), and 430-449 TCACATGGGTGTGCTGCGGA (AS430). Rat kidney cortex (poly)A RNA waspurified from total RNA using an mRNA Purification Kit (Amersham Biosciences).Mannose transport was assayed 5 days after injection.9 d! `+ t$ j- o! m2 b" v$ J
A8 V. [4 v4 MBiochemical procedures. Treatment with p -chloromercuribenzoate (pCMB) was done as published elsewhere( 11 ). Oocytes were incubated for 5 min in uptake medium (without substrates) containing 1 mM pCMB, rinsed,and then incubated 5 min in the same medium with or without 5 mM -mercaptoethanol. Before substrate uptake, the oocytes were again rinsedthree times. The methanethiosulfonate (MTS) reagents (Toronto ResearchChemicals, Downsview, Ontario, Canada) [2-(trimethylammonium)ethyl]methanethiosulfonate (MTSET) and 2-aminoethyl methanethiosulfonate (MTSEA)were dissolved in DMSO and used at 1 or 5 mM for 5 or 10 min of preincubation with the oocytes.( m* S1 z7 `. M6 A, X
4 c$ a- C2 T/ x% g7 Z) w8 e" Y1 v
Coimmunoprecipitation. Groups of 30 oocytes were injected with 10ng HA23-MAP17 cRNA, alone or in combination with 20 ng of rat kidney cortexpoly(A) RNA, and metabolically labeled, as explained above. Forcross-linking assays, groups of oocytes were washed twice inPBS 2 containing (in g/l) 8 NaCl, 1.15 Na 2 HPO 4, 0.2 KCl, 0.2 KH 2 PO 4,0.132 CaCl 2 x 2H 2 O, and 0.1 MgCl 2 x 6H 2 O and incubated in 3 ml glycerol buffer (10% glycerol, 0.1% saponin, 1 mM orthovanadate, in PBS 2 ) for 20 minat 4°C. Oocytes were then washed once in ice-cold glycerol buffer and incubated with 2 ml cross-linker buffer (0.25 M sucrose, 1 mM orthovanadate inPBS 2 ) containing the cleavable cross-linker 3,3'-dithio- bis (sulfosuccinimidyl propionate) (DTSSP; Pierce, Rockford, IL). This was added from a freshly made 100 x stock solution inDMSO to a final concentration at 1 mM. Cross-linking was allowed for 2 h at4°C. Then, the oocytes were washed once in PBS 2 ,and a buffer containing 100 mM glycine in PBS 2 wasadded to stop the reaction. After a final wash in PBS 2 , the oocytes were lysed by vortexing for 20 s in 20 µl/oocyte of anondenaturing homogenization buffer (120 mM NaCl, 50 mM Tris·HCl, pH 8,and 0.5% Nonidet P-40) supplemented with protease inhibitors (Complete-Mini;Roche Diagnostics, Mannheim, Germany) and incubated briefly on ice. Thelysates were then centrifuged at 16,000 g for 10 min and 4°C, and the supernatants were stored at -80°C. Incorporated radioactivitywas determined by scintillation counting of trichloroacetic acid precipitatesfrom aliquots.
) L% d% w4 m% D B5 A+ r- W4 u+ k0 ]* [2 z ?; B
Immunoprecipitation was performed with polyclonal anti-HA antibody (Clontech) bound to protein G plus/protein A-agarose (Calbiochem); 10 µl of0.1 µg/µl anti-HA were coupled to 30 µl protein G/A-agaroseslurry/group of oocytes by slow rotation for 2 h at 4°C. Equal amounts (2 x 10 6 cpm/µl) of precleared lysates were mixed with thecoupled beads and shaken slowly overnight at 4°C. The beads were washedfour times in buffer containing 100 mM NaCl, 20 mM Tris·HCl, pH 8.0, 1mM EDTA, 0.5% Nonidet P-40, and 500 mM LiCl and four times in the same bufferwithout LiCl. The beads were resuspended in an SDS-PAGE loading buffer andheated at 65°C for 15 min. When necessary, -mercaptoethanol (5%final) was added to the samples. After electrophoresis in 15%SDS-polyacrilamide, the gels were fixed in 3% glycerol, 10% acetic acid, and20% methanol for 30 min, then treated with Amplify (Amersham) for 15 min,dried, and exposed to Eastman Kodak Biomax films for 4 days at-80°C.. e+ g3 H6 t0 a0 w6 Y$ K2 W2 Q6 W2 ~
$ A6 ~0 A2 |4 t4 N2 F8 hRT-PCR. The cloning of an 1,808-bp fragment of rat PDZK1, thereby encoding the full open reading frame, was done by two-step RT-PCR using theSuperscript Preamplification System and Platinum Taq DNA PolymeraseHigh Fidelity (both from Invitrogen). The following primers were designed fromthe GenBank sequence: sense 5'-GTTCCAAGACTAGTAGTGTTCA-3' andantisense 5'-CAGCTAAGCTGAGCATAAAT-3'. The product was cloned intoPCR-Script (Stratagene), and the capped cRNA was tailed using a Poly(A)Tailing kit (Ambion).6 o: h4 R! w& a1 m! m
, `! \8 `9 Y! N5 a. ?. H* k1 ^
Statistics and fits. The data are shown as means ± SE. Thestatistical significance was determined by one-way ANOVA and the Tukeymultiple-comparison test, whereby P was consideredsignificant. For the transport kinetic studies in oocytes, Michaelis-Mentenand generalized Hill equations ( 6, 28 ) were used to calculate the K m, V max, and n coefficients by iterative, nonlinear regression. The fits were accepted when twoconsecutive iterations changed the sum of squares by As anindication of the goodness of the fits, the coefficient correlation( r ) and degrees of freedom are shown (degrees of freedom representthe number of data points minus the number of fitted parameters). Given thatthe total uptake measured in RNA-injected oocytes also includes the endogenousor water-injected uptake, an effort was made to calculate the difference, or the net expressed transport. This was done, despite the low expression level,by subtracting the mean values of water-injected oocytes for every substrateconcentration from the corresponding mean value of the RNA-injected oocytes.GraphPad Prism 3.0cx software for Macintosh was used for statistical andkinetic analysis. All experiments were repeated between three and five timeswith similar results, including saturation and Na-activation kinetics;nevertheless, the low level of transport expression prevented an accuratekinetic analysis, as shown by the asymptotic SE of the fits.. A& a0 X1 v1 l2 I. o5 s9 I
, g: g3 t8 b \
RESULTS5 B9 S0 J3 B- K- T: \" ~6 n
0 H) X( D- }6 u x9 j/ w
Expression cloning of MAP17. A cDNA library was prepared from size-selected rat kidney cortex poly(A) RNA (1 kb as the mean) according to our previous results( 5 ), and it was screened byexpression cloning in X. laevis oocytes. A single cDNA clone of 0.8kb was obtained out of 40,000 colonies, which induced 150% Na-coupled D -mannose transport above the endogenous, water-injected level ofthe cell when assayed at 0.1 mM D -mannose for 1 h ( Fig. 1 A ).Dose-response and expression-time course experiments revealed that the maximaltransport rate was already obtained using 0.5 ng cRNA/oocyte after 1 day ofexpression time and that the uptake was linear for at least 4 h (see Fig. 10 A; other datanot shown). Despite the low level of expression that prevented an accuratekinetic characterization, we found that MAP17 induces high-affinity Na-coupled D -mannose transport ( Fig.1 B 2 for Na activation ( Fig. 1 C ), which is inagreement with our previous results ( 5 ). The fit analysis issummarized in Table 1, wherebywe conclude that the effect of MAP17 most likely consists of an increase inthe capacity of the endogenous uphill transport system of the oocyte.
. |# N& C" q/ q9 O1 `3 m
. W* T2 F1 C. i/ O9 u+ F0 r& @Fig. 1. Characteristics of the expressed transport. A : Na dependence of0.1 mM mannose transport induced with 1 ng cRNA/oocyte and 2 days ofexpression; stippled bars, NaCl; hatched bars, choline chloride. * P B and C : mannose saturation ( B ) andsodium activation kinetics ( C ) of water ( )- and cRNA( )-injected oocytes;, net transport. Solid lines represent thenonlinear regression of each transport, as explained in MATERIALS AND METHODS. Values are means ± SE ( n = 10 oocytes) of arepresentative experiment due to uptake variation among frog oocytes.
" U5 @& Y n% `9 x- V8 x/ }- _" L' {, l. v; X) ?, N b" y
Fig. 10. Dose-response relationships. Increasing amounts of HA23-MAP17 cRNA wereinjected into oocytes to compare mannose transport induction ( A ) vs.membrane protein expression, shown as a Western blot ± DTT( B ), and densitometries of the signals in reducing conditions( C ). Whereas transport saturation started with 0.01 ng/oocyte, theprotein plateau started after 0.1 ng/oocyte, in either the 12- or 24-kDaband.1 h# Z' ^ ]- A7 L2 N
# m; W. [% b V
Table 1. Kinetic analysis of MAP17-induced transport in Xenopus laevisoocytes
# O1 W, W! ~8 k' M7 R( B' V$ _, A5 }7 z1 ~
The specificity of the induced transport was assayed in two ways. First,the uptake of 0.05 mM D -mannose was completely inhibited only by D -mannose, D -glucose, and phloridzin( Fig. 2 A ). However, nosignificant effects were observed with sugars such as L -mannose, D -fructose, phloretin, the -mannosidase inhibitor1-deoxymannojirimycin (DMM), and the amino acids and ions indicated in Fig. 2 A. Second,according to the inhibition results, a direct measure of the uptake of severalradioactive sugars showed that 0.5 mM D -[ 14 C]glucose and -methyl- D -[ 14 C]glucopyranoside were transportedsimilarly to D -mannose by MAP17-expressing oocytes( Fig. 2 B ). However,0.5 mM D -[ 14 C]galactose,3- O -methyl- D -[ 14 C]glucopyranose, and D -[ 14 C]fructose were excluded. Therefore, a kineticcharacterization of expressed glucose transport was also performed( Table 1 ), thereby obtaining areduced affinity for glucose compared with mannose( Fig. 2 C ) and asimilar stoichiometry ( Fig.2 D ). As with mannose, the effect of MAP17 was mainly onthe V max of glucose transport. Nevertheless, these resultsshould be considered as an approximation due to the low level of transport expression and therefore the general error of the fits( Table 1 ).6 U _: A/ V( r B3 H
+ {5 l% g6 H7 z
Fig. 2. Specificity of expressed transport. A : inhibition of 0.05 mM D -mannose transport expressed as percentages with respect to water(open bars); filled bars, cRNA-injected oocytes; and hatched bars, netexpressed transport. *Significant effect compared with the correspondingcontrol condition (water-injected or cRNA-injected oocytes or net transport),* P D -mannose, L -mannose, D -glucose, D -fructose, L -alanine, L -leucine, L -threonine, and L -lysine were present at 50 mM, whereas glutamic acid, inorganicsulfate and phosphate, tetraethylammonium (TEA), and p -aminohippurate(PAH) were at 10 mM; phloridzin and phloretin at 0.1 mM; and1-deoxymannojirimycin (DMM) at 1 mM. B : transport is specific for D -mannose (Man), D -glucose (Gluc), and -methyl- D -glucopyranoside (MDG); D -galactose(Gal), 3- O -methyl- D -glucopyranose (OMG), and D -fructose (Fruc) are excluded. Bars are as in A.* P C and D : D -glucosesaturation ( C ) and sodium activation kinetics ( D ) of water( )-, cRNA ( )-injected oocytes and net transport ( ). Solidlines represent nonlinear regression of each transport.9 r. N. @( m @" M
4 r/ n1 @- Q5 a( k, \8 {( P
Sequencing and secondary structure. Two-direction sequencing revealed an 816-bp cDNA that encodes an open reading frame of 114 amino acidproteins and 12,243 Da ( Fig.3 ). This was confirmed by SDS-PAGE of in vivo( Fig. 4 A ) and in vitro( Fig. 4 B ) translated cRNA. In addition, a thiol-dependent dimerization seems to occur according tothe 24-kDa signal observed under nonreducing conditions ( Fig. 4 B, lanes1-3, and see Fig.10 B ). A comparison of the nucleotide and amino acidsequences via BLAST found 82% identity with the human MAP17/DD96( 16, 17; GenBank accession NM_005764 ), an orphan protein selectively up-regulated in human carcinomas. Awide variety of weak identities was also found with several ATP-synthases,ATPases, Na-solute symporters, and transferases. The rat MAP17 cDNA wascommunicated to GenBank with accession no. AF402772 .
' T+ f( l* A1 m3 S' i9 e8 ~- J# }8 S
Fig. 3. Sequence analysis of human membrane-associated protein (MAP17). Nucleotideand single-letter code amino acid sequences are shown. The 2 transmembranedomains are indicated as TM1 and TM2, the PDZ-binding site at the COOHterminus is underlined, and the potential phosphorylation sites are labeledwith asterisks and letters in bold. The amino acids encoding aphosphomannomutase block (PMM/PGM) are framed, and the 3 alternativehemmaglutinin antigen (HA) insertion sites are indicated with arrowheads. Apolyadenylation signal is in bold and underlined./ Q7 F9 X9 d4 V" [
4 U: I2 N, }0 ^Fig. 4. Molecular analysis of MAP17. A : 35 S metabolic labelingof Xenopus laevis oocytes injected with water or MAP17 cRNA andelectrophoresis in reducing conditions, showing the predicted 12.2-kDamolecular size (arrow). B : in vitro translation of MAP17 withreticulocyte lysates ( lanes 1 and 4 ), in the presence ofmicrosomes ( lanes 2 and 5 ), and digested with Endo H( lanes 3 and 6 ). Lanes 4-6 contain 100 mMDTT. C : opossum kidney (OK) cells transiently transfected with theMAP17 cDNA tagged in positions 5, 23, and 66 wereimmunolabeled with anti-HA antibody ± saponin. Without detergent, theantibody only finds the corresponding antigen in the HA23 mutant, therebyconfirming the 2-membrane hydropathy model.) g1 C; @, S/ l: X8 v
/ t- a; T. O+ E6 a2 A7 L
Hydropathy analysis ( 20 )confirmed that rat MAP17 is an integral membrane protein with several possiblemodels. These include one membrane-spanning domain (SAPS andPredictProtein-PHD servers) that, in addition, could contain anNH 2 -terminal signal peptide (SOSUIsignal, TMHMM, and SignalP), ortwo membrane-spanning domains without signal sequence (DAS, SosuiTM, andTMpred), a prediction problem that is common in membrane proteins with hydrophobic NH 2 ends. In vitro translation using rabbitreticulocyte lysates in the presence of canine microsomes( Fig. 4 B, lanes 2 and 5 ) showed no evidence of a signal sequence, as no change in mobility was observed compared with the absence of microsomes ( lanes1 and 4 ). The number of transmembrane domains was also directlydetermined by tagging MAP17 with HA in positions 5, 23, and 66 (see positions in Fig.3 ). Given that all three mutants were fully functional in oocytes,they were transiently expressed in OK cells, and the proteins wereimmunodetected using an anti-HA primary and FITC-conjugated secondaryantibodies, with or without saponin ( Fig.4 C ). Only HA23-expressing cells were immunodecoratedwithout permeabilization of the membrane, therefore suggesting that this partof the protein is located extracellularly and that MAP17 contains twotransmembrane domains, with both NH 2 and COOH termini inside thecell. In addition, MAP17 is nonglycosylated (NetOGlyc 2.0 and the lack of aneffect by endoglycosidase H, Fig.4 B, lanes 3 and 6 ) and contains severalpotential phosphorylation sites in serines, threonine, and tyrosine( Fig. 3 ). Additional searchesfound an anion-exchangers family 1 alignment at residues 56-103(ProfileScan) and a phosphomannomutase/phosphoglucomutase block at 73-82 ( Fig. 3 ). The PDZ-binding siteat the COOH terminus (amino acids 111-114) for interaction with theglobular protein PDZK1/diphor1( 9, 18 ) is also shown underlinedin Fig. 3.
, d% P/ N5 U% L
* P v+ J4 N3 X6 g* S: XTissue and cell distribution of MAP17. Northern blot analysis showed that, as expected, MAP17 mRNA is very abundant in the rat kidneycortex, but also in the testis, and less so in the urinary bladder( Fig. 5 A ). We did notfind expression in the duodenum, jejunum, ileum, colon, liver, spleen, lung,heart, brain cortex, brain stem, cerebellum, skeletal muscle, or adipose tissue. We also assayed several renal cell line RNAs, obtaining hybridizationsignals in Madin-Darby canine kidney (dog), LLC-PK 1 (pig), and MCT(mouse) cells; OK cells do not seem to express MAP17.
! b/ [$ C! A+ e$ w) U1 ?: Q
, l7 l9 f) H* G, o. bFig. 5. Analysis of MAP17 RNA expression in rat tissues and cell lines. A :Northern blot analysis showing the hybridization signal in superficial renalcortex (SC), juxtamedullar cortex (JM), urine bladder (BL), and testis (TS).No expression was found in the duodenum (DU), jejunum (JE), ileum (IL), colon(CO), spleen (SP), brain cortex (BC), brain stem (BS), cerebellum (CE), lung(LU), heart (HE), liver (LI), skeletal muscle (SM), or fat (FT). Thehybridization signal was also found in Madin-Darby canine kidney,LLC-PK 1 and MCT, but not in the OK cell line. Top and bottom : gel photomontage of the corresponding 18S ribosomal bands tothe scale of the corresponding blot, respectively. B : darkfield viewof MAP17 expression in the superficial and juxtamedullar cortex of the kidney,labeling the proximal tubules of the outer and deep nephrons. C :darkfield view of the seminiferous tubule walls of the testis labeled with aMAP17-derived probe. D : further analysis with Nomarski imaging, whichrevealed that the expression is exclusive of the spermatids.
( M q# e" e$ X3 E+ A: p0 w- b" T* Q1 R9 p/ s+ b) w
Further localization studies were performed by in situ hybridization. Inthe kidney, the expression was similar to the human MAP17, that is, restrictedto the proximal tubules, but from both superficial and deep nephrons( Fig. 5 B ). In thetestis, MAP17 cRNA was expressed in seminiferous tubules( Fig. 5 C ), and Nomarski microscopy revealed a precise location in the spermatids ( Fig. 5 D ).8 d8 a! {) S$ a
; C; z/ O. h' M, [1 p! w" }
A more detailed (subcellular) analysis of the epithelial expression ofMAP17 was made by confocal laser immunofluorescence in OK cells permanentlytransfected with HA23-tagged MAP17. Figure6 shows that, in the absence of saponin (A in figure; nopermeabilization), MAP17 is located at the apical-most end of the cellmembrane. Permeabilization with saponin, however, evidenced an additional immunodecoration in the cytoplasm, reminiscent of Golgi network staining, anda subtle staining of the nuclear membrane( Fig. 6 B ). Doublestaining of the cells with the Golgi marker BODIPY-TR showed that MAP17colocalizes with a specific subset of the Golgi apparatus( Fig. 7, A - F ). Finally, this specific staining wasalso dispersed after treatment with the Golgi toxin brefeldin A( Fig. 7 G ).
: M1 e9 \/ c5 ~; B3 v: C* K: A2 j
Fig. 6. Subcellular expression of MAP17 in OK cells. Confocal sections wereobtained every 1.5 µm from the apical to the basal membrane of OK cellspermanently transfected with HA23-tagged MAP17 and immunodecorated withanti-HA antibody in the absence ( A ) or presence ( B ) ofsaponin. Without permeabilization, MAP17 is located at the apical membrane;however, the use of saponin revealed an uneven staining of the cytoplasm, witha pattern reminiscent of the Golgi network.' F) s+ d6 g' ?* E+ x
. e, i9 A+ S" R! }6 C7 W# `9 jFig. 7. MAP17 colocalizes with specific regions of the Golgi apparatus.Permeabilized OK cells permanently expressing HA23-tagged MAP17 were doublelabeled with anti-HA antibody (green, A ) and with the Golgi markerBODIPY-TR (red, C ). E : colocalization is shown by theemergence of the yellow staining of MAP17/Golgi. B, D, and F : Z -sections along the axis shown in A, C, and E, respectively. The anti-HA staining pattern waslost by pretreatment with brefeldin A ( G ).
& ]3 ^3 E: {) I! H# O
, U. n) M8 K' o; ?; j3 RStructure-function relationship. Pretreatment of MAP17-expressing oocytes for 10 min with 1 mM of thiophilic pCMB inhibited the netNa- D -mannose cotransport by 60%( Fig. 8 A ). This wasnot reversed on washout but rather by using -mercaptoethanol, a resultexpected from the covalent modification of cysteins ( Fig. 8 A ). Treatmentwith thiol-oxidating methylthiosulfonates MTSEA or MTSET showed no effect ateither 1 or 5 mM and for 5, 10, or 30 min, most likely because the reagentswere not accessible to the cysteins ( Fig.8 A ). Next, we mutated the two cysteins, Cys 20 (extracellular) and Cys 55 (intracellular), of HA23-tagged MAP17( Fig. 8 B ) to checkwhether they were involved in the pCMB effect. We found that onlyCys 55 and the double mutant Cys-less exhibited reduced mannosetransport, similar to the pCMB effect. Moreover, Western blot analysis showedthat these mutants also could not form homodimers( Fig. 8 C ). Thistherefore suggests a relationship between the quaternary conformation of MAP17and transport induction, given that the abundance (measured by densitometry)of the homodimer conformation signal in these experiments was only 20%that of the monomer band, but it was responsible for 60% of the inducedtransport.0 n; c+ Q) H, W% n' s3 _3 g% c
9 \1 G- _. ]9 S6 Z ~1 g# G$ XFig. 8. Role of cysteins on function and structure. A : induced mannosetransport (filled bars) is inhibited by the mercurial p -chloromercuribenzoate (pCMB), and its effect is reversed by -mercatoethanol ( M). [2-(Trimethylammonium)ethyl]methanethiosulfonate (MTSET) and 2-aminoethyl methanethiosulfonate (MTSEA)have no effect in either water (open bars)- or cRNA-injected oocytes. B : only the mutation of Cys 55 inhibited the inducedtransport. *Significantly different from wild MAP17-induced transport( P C : Cys 55 mutants are not able toshow the 24-kDa band under nonreducing conditions.. T! E: f3 |% u1 ~& G
; H% q# T$ B! @; aHybrid depletion of MAP17. To determine the role of MAP17 in mannose transport expressed in X. laevis oocytes by total kidney cortex mRNA, we performed hybrid depletion with three sense and threeantisense 20-mer oligonucleotides. The small mRNA-induced Na- D -mannose cotransport in oocytes was abolished by the three antisense oligos (all located inside the open reading frame, at either the5'- and 3'-ends or the center), with no effect provoked by thesense oligos ( Fig. 9 ).Therefore, MAP17 is responsible for mannose transport expressed by rat kidneycortex mRNA in oocytes.
( L' X: S0 Q" a+ q0 r+ u
! b& ]! n4 T- t: lFig. 9. Hybrid depletion experiments on mRNA-induced Na- D -mannosecotransport in oocytes. Oocytes were injected with water or rat kidney cortexmRNA (30 ng/oocyte), alone or annealed with rat MAP17-derived 20-meroligonucleotides. Six oligos (within the open reading frame) were used: senseoligos S105, S251, and S407; and antisense oligos AS105, AS335, and AS430 (see MATERIALS AND METHODS ). D -Mannose transport was measured5 days after injection. *Significant compared with MAP17-induced transport( P* O- s; V! P- v
8 W2 `# p/ J: N% H1 r4 L
Evidence for the involvement of additional proteins. The small size of MAP17 and the low expression level of sugar transport induced inoocytes make it unlikely that this protein by itself represents a hexosetransport system. Indirect evidence for the participation of additionalproteins in the oocyte expression system arose from dose-response experimentsof HA23-tagged MAP17 expression. Figure10 A shows that transport saturation appeared after 0.01ng of MAP17 cRNA/oocyte. This could be explained, for example, by a saturationof the translation and/or by processing of the protein. However, aquantification of the synthesized protein by Western blotting anddensitometric analysis showed that the protein plateau started after 0.1 ng/oocyte, in either the 12- or 24-kDa band( Fig. 10, B and C ). As MAP17 does not have extracellular free aminogroups, biotinylation of MAP17 to exclusively determine the protein insertedinto the plasma membrane could not be performed. One possible interpretationof these results is that the expression of mannose transport in oocytes byMAP17 is restricted by the need for one or more endogenous proteins to bepresent in a limited amount, and therefore the increasing amounts of MAP17 would saturate the activity.7 y- G/ K1 B% J9 D
; O& J; q. P- w$ O
We then assayed a direct approach to find possible collaboration betweenMAP17 and the characterized accumulative glucose transporters. The fact thatMAP17, in addition to mannose, induces Na- D -glucose andNa- -methyl- D -glucopyranoside cotransport( Fig. 2, B and C ) and that it is inhibited by phloridzin( Fig. 2 A ) indicates that a member of the SGLT family could be interacting or being modified byMAP17. As the SGLT-1 substrates D -galactose and 3- O -methyl- D -[ 14 C]glucopyranose were nottransported ( Fig. 2 B ),SGLT-1 should be excluded( 38 ). Despite this, wedirectly tested all three SGLT members characterized to date, namely, ratSGLT-1 and SGLT-2 and pig SGLT-3, expecting that if one of them wereinteracting with MAP17 in the kidney to transport D -mannose, thencoexpression of both in the oocyte would further increase the uptake shown byMAP17 alone. The corresponding cRNAs were injected separately or coinjectedwith either MAP17 by itself or in combination with MAP17 plus PDZK1 cRNAs( Fig. 11 A ). Theglobular, PDZ-containing protein PDZK1 was assayed, given that it was thefirst protein reported to interact with MAP17. As a result, all SGLT cRNAsinduced Na- D -glucose cotransport at the expected intensities( 38 ). However, none of themwas able to induce significant net Na- D -mannose cotransport, either alone or in different combinations with MAP17 and/or PDZK1. Subsequently, wesimilary coinjected MAP17 and total rat kidney cortex mRNA into X.laevis oocytes. As Fig.11 B shows, again poly(A) RNA and MAP17induced a similar level of Na-mannose cotransport, and the coinjection of bothelicited a slight but significant stimulation above the mRNA injectionlevel.5 G7 n) F1 ]/ o5 s
) ~1 z: H6 R: d! P/ v; s
Fig. 11. Coexpression in X. laevis oocytes. A : MAP17, the SGLT-1,-2, and -3 carrier cRNAs, and PDZK1 cRNA were coexpressed as indicated toinduce D -glucose (filled bars) and mannose (open bars) transport.Although all SGLTs expressed Na-glucose cotransport at the expectedintensities, only MAP17 was able to induce D -mannose transportabove the level in water-injected oocytes 2 days postinjection. *Significantlydifferent mannose uptake vs. water ( P B : oocyteswere injected with 20 ng of rat kidney cortex messenger RNA, 1 ng of MAP17cRNA, or a combination of both. Uptake was measured 5 days after theinjection. *Significantly different from water-injected oocytes ( P
m1 ]7 j4 _# ^9 h8 `. X- Q/ P6 T& [ g8 d$ @3 Q8 n3 Q
Finally, several coimmunoprecipitations were performed in oocytes expressing either HA23-MAP17 alone or in combination with rat kidney cortexmRNA. In the absence of the cross-linker DTSSP (see MATERIALS AND METHODS ), only the 12- and 24-kDa bands were precipitated( Fig. 12, arrows). However,with DTSSP additional proteins of 15, 19, and 28 kDa (arrowheads) were pulled.Of these, the 15- and 19-kDa bands represent oocyte proteins, because theywere already precipitated by MAP17 injection alone. The 28-kDa band seems torepresent a renal protein because it appeared exclusively in the MAP17/mRNAoocyte extracts. All other proteins were also precipitated in thewater-injected oocytes.
( t z* ]+ Q, Z9 b7 R" j5 d9 _! b6 a. ]! d
Fig. 12. Coimmunoprecipitation of MAP17. Oocytes were metabolically labeled andinjected with either water, HA23-MAP17 cRNA, or a combination of HA23-MAP17and rat kidney cortex polyA RNA. Before the lysis, some oocyteswere treated with the cross-linker3,3'-dithio- bis (sulfosuccinimidyl propionate) (DTSSP), and thenMAP17 was immunoprecipitated with anti-HA in nondenaturing conditions. In theabsence of DTSSP ( A ), only the 12- and 24-kDa bands were seen(arrows). However, with DTSSP ( B ), HA antibody additionally pulledbands of 15, 19, and 28 kDa (arrowheads), most of which disappeared with -mercaptoethanol. The 28-kDa band is only seen in the mRNA-injectedoocytes.
8 S: f) _% [: D% y% k& r. p! A; \: D b7 a+ T+ J* |
DISCUSSION# l i% ]2 ^& W2 N7 Y6 C
1 [, }$ Q: `- P( j$ N7 hBy expression cloning, we have identified MAP17 as a cDNA that inducesNa-activated transport of D -glucose and D -mannose when expressed in X. laevis oocytes( Fig. 1 A ).Furthermore, we have shown that MAP17 is necessary for rat kidney cortex(poly)A RNA to induce mannose transport in oocytes( Fig. 9 ). Nevertheless, thequestion of whether MAP17 is directly involved in the renal handling of D -mannose is not yet clear. Both the structural simplicity of MAP17(see below) and kinetic analysis of the induced transport suggest that MAP17is not solely responsible for the expressed transport, but rather it is actingas an activator of the capacity of an oocyte's endogenous transporter ( Table 1 ). The expressedtransport fits the general kinetic characteristics observed in both in vitroand in vivo renal mannose uptake assays: Na-dependent mannose transport ofhigh affinity and low capacity( 5, 10, 22, 27, 29, 30; Fig. 1, B and C ) and strong inhibition by the classic substrates of theSGLT family of transporters( 38 ), D -glucose, -methyl- D -glucopyranoside, and phloridzin( 5, 10, 22, 39; Fig. 2 A ). The maindifference in the mammalian system arises from the specificity of the transport, given that most of the previous studies indicate that, even if D -glucose inhibits D -mannose transport, both sugars donot share the same route. This conclusion was deduced not only from the typeof inhibition ( 22, 30 ) but also from the factthat D -mannose was not able to successfully inhibit D -glucose transport( 22, 39 ). However, we have shownthat MAP17 also induces Na-coupled D -glucose transport in X.laevis oocytes ( Fig. 2, B - D ), a result that could still be in agreement with our data if, as we postulate, MAP17 is interacting/activating an SGLT-related transporter other than SGLT-1, -2, or -3. In this event, D -mannose would not completely inhibit the transport of D -glucose, because additional SGLT members that do not collaborate with MAP17 are present along the nephron( 38 ). In any case, it wouldnot be surprising if the putative SGLT-like protein of the X. laevis oocyte had some functional differences with respect to its mammalian kidneycounterpart in accordance with their evolutionary distance.5 ?' `6 Y9 b7 b
8 c$ F" \1 B6 K
MAP17 was previously cloned in a search for mRNAs upregulated in renalcarcinomas, but it also turned out to be overexpressed in most colon, lung,and breast carcinomas ( 16, 17 ). Now, we have also shown aphysiological expression in the testis, more precisely in the spermatids( Fig. 5 ). Initial hydropathic analysis suggested the existence of a signal peptide and a transmembrane domain, because the software tools for molecular analysis cannot differentiatebetween signaling peptides and transmembrane domains when the amino acidsequence starts with a hydrophobic NH 2 terminus. However, we havenow shown by mutation analysis that the predicted signal peptide is actually afirst transmembrane domain, and therefore that MAP17 spans twice the brushborder of the tubular cells ( Fig.4 C ). In addition, we have also shown that MAP17 seems toform homodimers of 24 kDa (e.g., Fig. 4 B ), most likely through a disulfide bond betweenintracellular Cys 55 residues( Fig. 8 C ). Thepresence of Cys 55 also seems to be necessary for the completeexpression of mannose transport by MAP17( Fig. 8 B ), whichindicates that either the homodimer conformation is implicated in thetransport or, alternatively, Cys 55 participates directly in theactivation.
, @ r1 g+ I/ P5 E3 n1 i' C& {) e" n, N3 }$ f: P" H
To understand the function of MAP17, human PDZK1 was initially identifiedas a globular protein containing four PDZ domains, the protein of whichinteracted with the COOH terminus of MAP17 ( 19 ). PDZK1 also interactswith the organic anion transporter cMOAT( 18 ) and the CFTR chloridechannel ( 7 ), whereas the rat(diphor1) ( 9 ) and mouse(NaPi-Cap1) ( 13 ) counterparts do so with the Na-phosphate cotransporter (NaPi-2). More recently, a directinteraction between MAP17 and NaPi-IIa has also been reported( 26 ). Therefore, the challengenow is to identify the renal protein that interacts with MAP17 and transports D -mannose. In addition to the structural simplicity of MAP17, experimental evidence for the collaboration of MAP17 with another protein toproduce the transport expression arises from dose-response experiments, giventhat the induction of mannose transport with increasing amounts of MAP17 cRNAdoes not parallel the expression of MAP17 protein in the oocyte( Fig. 10 ). A similar lack ofcorrelation was reported, for example, for the amino acid transporter 4F2hc( 4, 36 ), which encodes the heavychain of a heteromultimeric complex. In this case, the low expression inoocytes was due to the use and depletion of the endogenous light chain of theoocyte ( 8 ). The function ofPDZK1 could consist of either simply maintaining MAP17 in the plasma membrane or acting as an intermediary that joins MAP17 and a third protein with thecharacteristics of SGLT. Interestingly, the SGLT members known to date do notcontain PDZ-binding domains, a characteristic that agrees with the lack ofincrease in the induction of mannose transport by MAP17 when MAP17/PDZK1 iscoexpressed with SGLT-1, -2, and -3 ( Fig.11 A ). Independently of the sugar transporter implicated,the interactions with MAP17 seem to be very weak, as deduced from the lack ofhigh bands coimmunoprecipitated with MAP17( Fig. 12 ). The same result hasbeen observed with the Na-phosphate cotransporter NaPi-IIa( 26 ), a protein that interactswith MAP17 through its NH 2 terminus, as shown in yeast two-hybridassays: the interaction is so weak that pull-down experiments did notcorroborate the interaction.0 S5 m' Y- _* ~$ d, u
9 ^; K! @; i8 @& ~) M2 g
The interaction of MAP17 with the NaPi-IIa cotransporters is, at first,puzzling. The physiological meaning of this interaction is not yet known, butit adds additional appeal to the cell biology of this little protein, becauseits functions must not be restricted to sugar handling. For example, MAP17could modulate the activity or organization of membrane transporters by directinteraction as the RS1 modifier( 35 ), or through competitionfor PDZ-binding places to alter the stoichiometry of the transporters-PDZproteins. In this way, MAP17 could force other PDZK1-associated proteins toreorganize in a PDZK1-independent way. The structural simplicity of MAP17supports this simple regulatory role: first, MAP17 is integrated into themembrane with 2 spanning domains, followed by a single cytosolic region consisting of the last 48 amino acids. Second, we have inserted the 9 aminoacid-HA tags in 3 different positions (MAP17 is only 114 amino acids long),and, surprisingly, all the resultant constructs are still able to induce thesame mannose and glucose transport as the native protein. It is thereforeevident that the structural requirements of MAP17 to induce its effects are not as sophisticated as they seem to be for the real transporters, even in ahomodimer conformation. On the other hand, the binding of membrane proteins todifferent PDZ proteins seems to be a new mechanism of regulation, given thatit is the CFTR chloride channel that binds either to PDZK1, which remains inthe plasma membrane, or to CFTR-associated ligand, which continues to accumulate in the Golgi ( 7 ).Because MAP17 is expressed in both the plasma membrane and the Golgi, itsexpression could also be regulated in a similar way.1 U: c% T* ?/ r8 Y& O$ x y
) q7 A7 g8 g3 f$ N; ~, D
In summary, to our knowledge we are the first group to report a functionalrole of MAP17 in the cell, i.e., the induction of uphill transport of mannoseand glucose in oocytes, most likely through the interaction or modulation ofat least one other protein. We have also clarified its molecular structure andits location in the cell and tissues. Identification of all proteinsinteracting with MAP17 will help to understand the complete function of MAP17in the kidney.
6 J- A, E/ ]5 E5 ^6 C; G8 l) k! e; C
DISCLOSURES
; J& O9 o F# P1 G
, U+ e V8 S+ J2 W4 c- F. _$ i4 wThis work was supported in part by two grants to V. Sorribas, one from theGovernment of Aragón (CONSID-P078/99) and the other from the SpanishMinistry of Science and Technology (MCYT-BIO2000/1608). T. Blasco wassupported by a fellowship from the Government of Aragón (CONSIDB197/98).& H6 D* \$ h; X0 v5 X
0 v! }6 m5 b$ {# fACKNOWLEDGMENTS
( S0 ^( D; [8 ]6 t* O, f% j- u0 l0 g. q( g8 Q
We are indebted to M. A. Hediger (Harvard Institutes of Medicine, Boston,MA) for the kind gifts of rat SGLT-1 and -2 cDNAs and E. M. Wright (UCLASchool of Medicine, Los Angeles, CA) for providing pig SGLT-3 and usefulcomments on the manuscript. We also thank J. Biber for critiquing themanuscript and making helpful suggestions and H. Murer for supplyingMadin-Darby canine kidney and LLC-PK 1 cells (Institut ofPhysiology, Univ. of Zurich). Finally, we are grateful to M. Levi (Univ. ofColorado Health Sciences Center, Denver, CO) for providing MCT cells.
/ I9 e: b% [' n( h( Y& ~ 【参考文献】 p* R" d# N& I& \2 `) g
Aebi M andHennet T. Congenital disorders of glycosylation: genetic model systemslead the way. Trends Cell Biol 11: 136-141,2001.
: a7 F# C3 `: i, P- b$ ~
0 e. Q: r! `7 W& s4 ^1 ^
) M+ j5 l. G3 S- b7 e1 ` G# e3 \: t; i. y
Alcalde AI,Sarasa M, Raldua D, Aramayona J, Morales R, Biber J, Murer H, Levi M, andSorribas V. Role of thyroid hormone in regulation of renal phosphatetransport in young and aged rats. Endocrinology 140: 1544-1551,1999.$ ?1 l7 a) |: P2 ?
1 r8 P4 g2 ^4 C" _2 }! V) o
, O3 H5 U/ ]* f4 o9 a* K& t5 S
% b. m' {! E Y
Berg T, RiiseHM, Hansen GM, Malm D, Tranebjaerg L, Tollersrud OK, and Nilssen O. Spectrum of mutations in -mannosidosis. Am J HumGenet 64:77-88, 1999.
3 j( u4 |! o B0 ^2 f& O4 @4 x. N* `) |
9 L. s" z) I& U8 U" ^0 a @$ q' ~
8 P. t: A9 P s
Bertran J,Magagnin S, Werner A, Markovich D, Biber J, Testar X, Zorzano A, Jun LC,Palacin M, and Murer H. Stimulation of system y -like aminoacid transport by the heavy chain of human 4F2 surface antigen in Xenopuslaevis oocytes. Proc Natl Acad Sci USA 89: 5606-5610,1992.
; N, d$ X7 R* ^! S" l% [& c3 D& L2 G+ k5 ? ?7 R( c6 {. n- t
9 |; x' b- E4 \5 e9 D1 [3 x. P T' A
Blasco T,Aramayona JJ, Alcalde AI, Halaihel N, Sarasa M, and Sorribas V. Expressionand molecular characterization of rat renal D -mannose transport in Xenopus oocytes. J Membr Biol 178: 127-135,2000.
/ u7 W1 }9 q) t! T% ?2 m
1 Y2 W( n9 R) u }# M) [) F0 e& }( u2 T, S
4 [; q! r2 ` w$ ]
Brot-Laroche E,Serrano MA, Delhomme B, and Alvarado F. Temperature sensitivity andsubstrate specificity of two distinct Na -activated D -glucose transport systems in guinea pig jejunal brush bordermembrane vesicles. J Biol Chem 261: 6168-6176,1986.- [; _4 Q5 Z6 z/ D
1 y$ d t9 p1 x" u' ? o4 f9 Z5 ]- W4 S. i) [2 s$ C
{4 ]6 a: S3 @
Cheng J, MoyerBD, Milewski M, Loffing J, Ikeda M, Mickle JE, Cutting GR, Li M, Bruce A,Stanton BA, and Guggino WB. A Golgi-associated PDZ domain proteinmodulates cystic fibrosis transmembrane regulator plama membrane expression. J Biol Chem 277:3520-3529, 2002.
5 u5 r1 B4 [- B+ h" T" {4 J6 G
% |) O8 X1 H8 ~& T- B/ i: W1 @/ H1 W5 E' b1 I9 n6 _& E
* K F! V% W# o& A+ @
Chillarón J, Roca R, Valencia A, Zorzano A, and PalacínM. Heteromeric amino acid transporters: biochemistry, genetics, andphysiology. Am J Phyiol Renal Physiol 281: F995-F1018,2001.7 u, m9 W, [3 B- C, G! T6 G& |4 y3 I
1 N3 G/ |; J- T& }- P, C/ M1 v0 z
: X( q2 B! J+ o2 [4 z5 q" L
. N& R S1 e& s) Q" V) D- J& b( w
Custer M,Spindler B, Verrey F, Murer H, and Biber J. Identification of a new geneproduct (diphor-1) regulated by dietary phosphate. Am J PhysiolRenal Physiol 273:F801-F806, 1997.
2 ^6 q% C8 N( F3 ?& V! k% i) a) y6 S$ |7 j& c6 s z* ^, Z
8 X8 W& s: Z& H& n
% E+ {4 K; C. o
De la Horra MC,Cano M, Peral MJ, Garcia-Delgado M, Duran JM, Calonge ML, and Ilundain AA. Na -dependent D -mannose transport at the apical membraneof rat small intestine and kidney cortex. Biochim BiophysActa 1512:225-230, 2001.
; i% B* Y% ~ n8 G0 P: |7 Y C1 C5 h0 q% j1 f; L" l7 ^+ ^3 v
! x/ I& ~* p$ x# A6 H
* P* o, E3 C% u; \( o
Estévez R, Camps M, Rojas AM, Testar X, Deves R, Hediger MA,Zorzano A, and Palacin M. The amino acid transport systemy L/4F2hc is a heteromultimeric complex. FASEBJ 12:1319-1329, 1998.
; A I( C( f1 x
" T0 V) A1 ^& ? c- e9 \8 N8 f% `+ `6 t, O: A. o
D6 k6 {7 E+ P, }Gallagher SR. Current Protocols in MolecularBiology, CD-ROM. New York: Wiley, 1995.
7 [( L( R) p3 ?7 g! Y4 X$ Y
2 n% I! Y' C6 R1 K
4 U! X8 e1 D) h. S5 L" e" h w( V1 X& @3 l- X1 V
Gisler SM,Stagljar I, Traebert M, Bacic D, Biber J, and Murer H. Interaction of thetype IIa Na/P i cotransporter with PDZ proteins. J BiolChem 276:9206-9213, 2001.
' O' U% T! s! X* V, ^
4 z1 p1 \* U: z* W W% I6 _# ^/ l) e" D* }- T
+ `* g9 U+ j$ X2 ~/ T, Z6 l
Hediger MA,Coady MJ, Ikeda TS, and Wright EM. Expression cloning and cDNA sequencingof the Na /glucose co-transporter. Nature 330: 379-381,1993.3 T! p" _/ X9 G; b$ p
; X- E. ~) _' B+ S; K3 n/ }; |
1 _) H9 S1 ]& c S' v; _7 C& U
7 ?3 T8 `/ [9 x/ l: b/ UJagus R, JoshiB, Miyamoto S, and Beckler GS. Current Protocols in CellBiology. New York: Wiley, 1998, p.11.2.1-11.2.25." f6 }3 K' f7 O9 q) G2 `
W" B% U! I) W6 y+ n4 S3 B) P9 m0 A7 w
7 R: C, M3 W2 K% s# n" t5 \( {$ f. Q! `: h
Kocher O,Cheresh P, Brown LF, and Lee SW. Identification of a novel gene,selectively up-regulated in human carcinomas, using the differential displaytechnique. Clin Cancer Res 1:1209-1215, 1995.& P/ `1 o2 m; ~) X0 d' ^0 ]
: [) \6 j- L" |2 Q) W
" p( I. T. ?, m T0 U3 R) `0 S2 }( ~/ l# k$ _& c
Kocher O,Cheresh P, and Lee SW. Identification and partial characterization of anovel membrane-associated protein (MAP17) up-regulated in human carcinomas andmodulating cell replication and tumor growth. Am JPathol 149:493-500, 1996.0 D# B# q8 f0 w& n' f H: ~
4 U. S# F6 `3 O5 H# o
. z. v9 `& m% K7 W
7 n$ T* U5 W9 T5 _3 ~6 Y: ~8 QKocher O,Comella N, Gilchrist A, Pal M, Tognazzi K, Brown LF, and Knoll JHM. PDZK1,a novel PDZ domain-containing protein up-regulated in carcinomas and mapped tochromosome 1q21 , interacts with cMOAT (MRP2), the multidrugresistance-associated protein. Lab Invest 79: 1161-1170,1999.7 x( ]3 C, ?3 A, V4 ?8 I& Q2 m: e
5 F! n. \8 ~% h1 k; ~
2 ?* T: ?6 k S' Z" S y) t% [4 v. ?: V5 [0 D1 j6 k
Kocher O,Comella N, Tognazzi K, and Brown LF. Identification and partialcharacterization of PDZK1: a novel protein containing PDZ interaction domains. Lab Invest 78:117-125, 1998.
8 J6 ^+ V# ~& m: s2 l2 Y: Z5 e, \4 b" v$ C
# Z8 X0 z! N( S! Y
' {& n+ Y2 M- I- F8 A7 BKyte J andDoolittle RF. A simple method for displaying the hydropathic character ofa protein. J Mol Biol 157:105-132, 1982.8 H# u+ a! H. `5 K
( P, \2 n. Q* O+ T" P9 X
) l& M3 F' J$ h) A9 J
5 `* q6 A# g X. DLederer ED,Sohi SS, Mathiesen JM, and Klein JB. Regulation of expression of type IIsodium-phosphate cotransporters by protein kinases A and C. Am JPhysiol Renal Physiol 275:F270-F277, 1998.' ~* k! {. E6 _$ e( J$ A" v1 \
. N7 c" e: K$ L {
6 [9 E/ p5 X3 |! n2 Z1 J: Y8 W& v' D! b3 L" w8 a, A9 {5 g
Mendelssohn DC and Silverman M. A D -mannose transport system in renalbrush-border membranes. Am J Physiol Renal Fluid ElectrolytePhysiol 257:F1100-F1107, 1989.
' R6 K6 D' k. _6 D
4 r1 B9 r1 E/ o2 j8 u F* p" G( p- `& N" u' ^# A; f
) h7 q; h0 R/ j1 h; U, v
Niwa K andIritani A. Effect of various hexoses on sperm capacitation and penetrationof rat eggs in vitro. J Reprod Fertil 53: 267-271,1978.
& _% g6 Y: w# r# Y. |
$ s0 l" ?% W+ f- O* l
" B: `3 C6 Q' a k5 y: N3 U1 F2 H; V5 @
Panneerselvam K and Freeze HH. Mannose enters mammalian cells using a specific transporterthat is insensitive to glucose. J Biol Chem 271: 9417-9421,1996.8 l7 L3 y+ {; Z7 f4 S$ B
9 x: p1 ?/ K* r( O ?
4 }9 t. x. ]6 ^- ^
& m+ A- I' `* k/ g* HPitkänken E and Pitkänen OM. Renal tubularreabsorption of 1,5-anhydro- D -glucitol and D -mannose invivo in the rat. Pflügers Arch 420: 367-375,1992.
" L( ]5 E# d! ^. \2 s/ a
- ~( ]# G- Y% o9 K, N( g; Y
& a! E3 _" Z' C* ]) R" o8 B. L, l l& Y+ g4 f" ~, l% g! l
Pribanic S,Hernando N, Gisler S, Bacic D, Sorribas V, Biber J, and Murer H. MAP17 ispart of a proximal tubular apical complex containingNa/P i -cotransporter NaPi-IIa and the PDZ protein NaPi-Cap1. J Am Soc Nephrol 13:483a, 2002.
2 C) f% L* s( Y; g- z( V4 V$ }$ G. W& h3 y% q" t/ E
' w' N% _) P, z* q) x/ t) I/ _
; I( U: ?& Y/ ] x$ ~1 [/ H; {/ g) s+ X* V
Pritchard JB,Booz GW, and Kleinzeller A. Renal sugar transport in the winter flounder.VI. Reabsorption of D -mannose. Am J Physiol Renal FluidElectrolyte Physiol 242:F415-F422, 1982.
, i9 _% X s% J5 V0 V% E, A, b- t; e" r; x+ A' u6 W
/ T" i S! ^ W3 `! X- g$ {2 V0 D. _! Y5 D
Restrepo D andKimmich GA. Kinetic analysis of mechanism of intestinalNa -dependent sugar transport. Am J Physiol CellPhysiol 248:C498-C509, 1985.
8 `) \$ R% f& N& k$ P
- [2 L) p/ Y0 a' [. f/ X. h- j* v- K" N& c
$ B; Y8 e& k% y* A1 VSilverman M,Aganon MA, and Chinard FP. Specificity of monosaccharide transport in dogkidney. Am J Physiol 218:743-750, 1970.
$ h: u! P% A0 m" r* D/ J3 z
* j$ @7 }' E# h" D! H8 h9 P3 i$ Y: a) y6 v! Q: }
! N$ V( g# v1 s9 c% d9 A" I6 x
Silverman M andHo L. Kinetic characterization of Na / D -mannosecotransport in dog kidney: comparison with Na / D -glucosecotransport. Biochim Biophys Acta 1153: 34-42,1993.- @& w3 h; w6 U
! B7 V- A1 \7 e8 {0 s `1 P: f' v+ \% C
o3 B2 l) y. Q6 ASorribas V,Markovich D, Hayes G, Stange G, Forgo J, Biber J, and Murer H. Cloning ofa Na/P i cotransporter from opossum kidney cells. J BiolChem 269:6615-6621, 1994.
" g) V+ a; E0 S# [3 p; V- a" i/ H) ~
3 N6 O, X5 v& E6 b
# _, I- w- t3 m! j" zSorribas V,Markovich D, Verri T, Biber J, and Murer H. Thyroid hormone stimulation ofNa/P i -cotransport in opossum kidney cells. PflügersArch 431: 266-271, 1995.7 z% [: [, w/ O. I
/ H& [8 s. Y; x
; |; \1 i/ L$ X% Q6 p3 r: E$ m' c; X) P) _1 F
Soyama K. Enzymatic determination of D -mannose in serum. ClinChem 30:293-294, 1984.
7 x6 q+ A- {: U% K1 o0 N+ i
+ g$ z& |; m& c- e' G4 k5 w# n0 P: e7 h0 M
* ]& C0 X/ n; {+ DTerrado J,Gerrikagoitia I, Martínez-Millán L, Pascual F, Climent S,Muniesa P, and Sarasa M. Expression of the genes for -type and -type calcitonin gene-related peptide during postnatal rat braindevelopment. Neuroscience 80:951-970, 1997.$ t/ f* F E( Y+ l: o- [; |
$ A6 p/ r9 Y/ U) m3 N* w2 t1 C
; A; Z# m" J1 Q7 S- I+ l- x# d: b' [0 w
2 c: h+ D+ g. ]8 K% qVeyhl M,Spangenberg J, Puschel B, Poppe R, Dekel C, Fritzsch G, Haase W, and KoepsellH. Cloning of a membrane-associated protein which modifies activity andproperties of the Na - D -glucose cotransporter. J Biol Chem 268:25041-25053, 1993.
5 X1 ~4 h2 ~6 G" N
9 z3 w' ~% K% i
! Y0 q& G/ a$ T0 K C H# C7 z) M' D' Z5 g" K' A& k
Wells RG, LeeWS, Kanai Y, Leiden JM, and Hediger MA. The 4F2 antigen heavy chaininduces uptake of neutral and dibasic amino acids in Xenopus oocytes. J Biol Chem 267:15285-15288, 1992.9 o9 A* M/ V1 |, s) r) y
8 _- ]' K4 v4 H6 J/ i+ A& ?
. j7 A0 i# ]9 y" h- ]4 Q! w* u7 N0 K G+ `+ R: O- \% f
Werner A, BiberJ, Forgo J, Palacín M, and Murer H. Expression of renal transportsystems for inorganic phosphate and sulfate in Xenopus laevis oocytes. J Biol Chem 265:12331-12336, 1990.
5 m' t8 W: B- X/ S. V! N" i; Q0 c; K4 e6 u" a" C; K1 v9 ]
J6 v+ }9 \: Y% I5 l H8 W s. l
4 a/ @# J1 w: Y' l$ ]0 y8 w
Wright EM. Renal Na -glucose cotransporters. Am J Physiol RenalPhysiol 280:F10-F18, 2001.
' p. L, f1 n+ U+ E/ r2 {% H9 F1 ~2 Q( G7 V' p, I1 s
" @$ E5 K8 F! | D, k
' m0 U/ A: p! a- |- Y& o7 W7 a3 wYamanouchi T,Shinohara T, Ogata N, Tachibana Y, Akaoka I, and Miyashita H. Commonreabsorption system of 1,5-ahydro- D -glucitol, fructose, and mannosein rat renal tubule. Biochim Biophys Acta 1291: 89-95,1996. |
|
发表于 2009-4-21 13:45
