|
- 积分
- 0
- 威望
- 0
- 包包
- 0
|
作者：Sandra Pribanic, Serge Mike Gisler, Desa Bacic, Caveh Madjdpour, Nati Hernando, Victor Sorribas, Andrea Gantenbein, Jürg Biber, Heini Murer作者单位：Institutes of Physiology and Anatomy, University of Zürich, 8057 Zürich,Switzerland; and Department of Toxicology, Universityof Zaragoza, E-50.013 Zaragoza, Spain % H* f9 T; {1 h0 \2 T
* [- k1 m V* W! b5 b9 F% Z
0 Y5 h& |' j* P( A1 L
4 q7 J0 {& V# y# `
9 T- S& {7 P- i2 v Q( w/ R7 `
+ _, U7 R' ?3 L _& s1 C5 M( b
3 D( H- T( }5 _: K7 H2 w $ J, N; G K9 S- d
( [- X, s5 r' _
; a8 a0 a" Y. T3 A; [) G! S; Z; h6 ^! i # F3 d" @3 M1 t; J! k- M( |
/ u6 Y3 U1 u- g4 O
+ y: s8 `2 {& |6 u( u' t 【摘要】
0 ?: b, I6 X& s: G1 ?8 C An essential role in phosphate homeostasis is played by Na/Pi cotransporterIIa that is localized in the brush borders of renal proximal tubular cells.Recent studies identified several PDZ proteins interacting with theCOOH-terminal tail of NaPi-IIa, such as PDZK1 and NHERF-1. Here, by usingyeast two-hybrid screen of mouse kidney cDNA library, we attempted to findproteins interacting with the NH 2 -terminal part of NaPi-IIa. Weidentified MAP17, a 17-kDa membrane protein that has been described to beassociated with various human carcinomas, but it is also expressed in normalkidneys. Results obtained by various in vitro analyses suggested that MAP17interacts with the fourth domain of PDZK1 but not with other PDZ proteinslocalized in proximal tubular brush borders. As revealed byimmunofluorescence, MAP17 was abundant in S1 but almost absent in S3 segments. No alterations of the apical abundance of MAP17 were observed after maneuversundertaken to change the content of NaPi-IIa (parathyroid hormone treatment,different phosphate diets). In agreement, no change in the amount of MAP17mRNA was observed. Results obtained from transfection studies using opossumkidney cells indicated that the apical localization of MAP17 is independent ofPDZK1 but that MAP17 is required for apical localization of PDZK1. In summary,we conclude that MAP17 1 ) interacts with PDZK1 only, 2 )associates with the NH 2 terminus of NaPi-IIa within thePDZK1/NaPi-IIa/MAP17 complex, and 3 ) acts as an apical anchoring sitefor PDZK1.
( E; a9 f* [& G 【关键词】 interacting proteins Na/Pi cotransport PDZ proteins NHERF opossum kidney cells
) r0 T$ q- W( d5 B3 P THE N a - DEPENDENT PHOSPHATE transport protein NaPi-IIa (SLC34A1) is the major mediator inrenal reabsorption of inorganic phosphate (P i )( 2; for review, see Ref. 24 ). In proximal tubules,NaPi-IIa is localized in the brush border membrane and is a part of heteromultimeric complexes scaffolded by the PDZ proteins PDZK1 and NHERF-1( 4, 9, 11, 12, 17, 19, 26, 30 ). Interaction of NaPi-IIawith these PDZ proteins was shown to occur via the COOH-terminal PDZ bindingmotif (TRL) of NaPi-IIa. Moreover, these interactions occur only with distinctPDZ domains of PDZK1 and NHERF-1( 11 ). Recent data suggestedthat these interactions play important roles for the apical positioning ofNaPi-IIa. Overexpression of single PDZ domains in opossum kidney (OK) cellsresulted in an impairment of apical sorting/positioning of NaPi-IIacotransporters ( 12 ), and thecomplete lack of NHERF-1, as demonstrated with a mouse knockout, resulted in reduced abundance of NaPi-IIa and urinary wasting of phosphate ( 26 ). On the other hand,recent experiments on targeted PDZK1 gene disruption did not result in asignificant renal phenotype, suggesting that functional compensation of thelack of PDZK1 might occur by other PDZ proteins( 19 ).
' N, L# r) Q& l' {/ `2 X: k/ H: Z/ u* R: \' T1 A9 A
To explore the interactions of NaPi-IIa in more detail, we attempted tofind proteins that interact with the NH 2 terminus of NaPi-IIa. Results obtained by a yeast two-hybrid screen against a mouse kidney cDNAlibrary suggested that MAP17, a 17-kDa membrane-associated protein, interactswith the NH 2 terminus of NaPi-IIa. MAP17 was earlier identifiedbased on differential display performed with mRNA isolated from normal andcarcinogenic tissues and was documented to be upregulated in various humancarcinomas originating from the colon, breasts, lung, and kidney( 15, 16 ). Interestingly, MAP17 wasalso detected in normal tissue, such as renal proximal tubules( 15 ), and was shown tointeract with the PDZ protein PDZK1( 10, 15 ). Moreover, MAP17 was shown to be part of the heteromultimeric protein complex formed by the PDZ proteinPDZK1 and the multidrug resistance-associated protein MRP2( 17 ). The interaction withPDZK1 is suggested to be via the last three amino acids (TPM) contained in the COOH terminus of MAP17, representing a PDZ binding domain( 17 ).
! p8 |1 @, x& ^" I4 j
0 z! j$ |. X( l3 fThe physiological role of MAP17 in proximal tubules is not known. In thepresent study, we therefore further investigated the interaction pattern ofMAP17 with known PDZ proteins located in the apical/subapical site of proximaltubular cells. Our results demonstrated that MAP17 interacts only with PDZK1and that within this protein complex MAP17 may interact with the NH 2 terminus of NaPi-IIa. Furthermore, based on results obtained from transfection studies using OK cells, we conclude that the apical locationof MAP17 is not dependent on its interaction with PDZK1.9 ]0 q2 r1 V$ {
! N/ h0 x1 B: \' n
EXPERIMENTAL PROCEDURES
1 ~6 M$ S1 y( Q. [9 b% i
' |- X+ P) |, I* S7 ^. z5 ?Yeast two-hybrid screen and trap assays. A cDNA library of adult mice (Clonetech, Basel, Switzerland) was screened against the entireNH 2 terminus (Nt, aa 1-109) of the NaPi-IIa cotransporter (SLC34A1, Acc. No. AAC52361 ). NaPi-IIa-Nt was inserted into the vector pFBL23( 3 ) using the restriction sites Eco RI and Sal I (Promega, Madison, WI). The same sites wereused to insert the full-length MAP17 (Acc. No. AK002288 .1) plus two additional NH 2 -terminal glycines into pBTM116. For various yeast two-hybrid trap assays, the following preys were inserted into pACT2, as described( 11 ): full-lengths or singlePDZ domains of PDZK1, NaPi-Cap2 and NHERF-1. The vector pBTM116, empty orcontaining the subunit p51 of HIV reverse transcriptase (1RTp51)( 27 ), was used as controlbaits.! R# z; Z- ]* l, R2 r; y; [3 {
# ?6 F3 I3 n7 s/ s* K8 q6 SAll baits were expressed in Saccharomyces cerevisiae (L40) containing his3 and lacZ reporter genes( 29 ). Transformations wereperformed by the LiAc/SS-DNA/PEG procedure using 1 µg of plasmid( 8, 11 ). Yeast transformants weregrown on synthetic medium lacking leu, trp, and his for 4days, and expression of LacZ was assayed by a filter assay( 7 ) using5-bromo-4-chloro-3-indolyl - D -galactopyranoside (Alexis,Lausen, Switzerland). Single, positive clones were rescued by electroporationinto E. coli KC8 cells as described( 13 ).3 M8 s& e4 l& Z" W% ^
0 o0 t) G3 G9 m q; h1 MGST fusion constructs and pull-down experiments. The construction of GST fusions with full-length PDZK1, NaPi-Cap2, and NHERF-1 has beenpreviously described ( 11 ).Pull-downs were performed using proximal tubular brush border membranevesicles (BBMV) isolated from mice kidney( 5 ). BBMV were solubilized inbinding buffer (50 mM Tris·HCl, 120 mM NaCl, 0.5% Igepal CA-630; pH 8)for 5 min at 4°C and centrifuged at 16,000 g for 3 min.Preequilibrated glutathione-agarose beads (Sigma, Buchs, Switzerland) wereloaded with GST-fused proteins ( 2 µg) and incubated by rocking with0.05-0.3 mg of solubilized BBMV protein for 1 h at 4°C. After fourwashing steps with binding buffer, the samples were analyzed by Western blotanalysis.: R+ V9 m; {- T5 h- I8 I
7 z' m% w- k2 q- j/ T1 v% `! Z l
Immunofluorescence and Western blots. Tissue distributions of various proteins in kidneys of 8- to 12-wk-old mice (NMRI, Janvier, France)were assayed by immunofluorescence as described( 23 ). All animals used werekept on a standard chow with free access to tap water. Some animals were fedfor 5 days with low- or high-phosphate diet or injected with parathyroidhormone (bovine, 1-34 PTH; Bachem, Bubendorf, Switzerland) as described by Bacic et al. ( 1 ).$ v, ~4 B* \) w/ l* u
- y a0 p t, {0 s; h
Cryosections of 5-µm thickness were incubated with the following primaryantibodies. MAP17 was detected by a custom-made rabbit antiserum raisedagainst a cytoplasmatic COOH-terminal peptide (17-mer) of mouse MAP17 or byaffinity-purified polyclonal antibodies directed against the COOH terminus ofhuman MAP17 (kindly provided by Dr. O. Kocher, see Ref. 16; both used with a dilution1:1,000). NaPi-IIa (1:1,000) and PDZK1 (1:500) were detected using polyclonalantisera raised against the NH 2 terminus of each( 6, 11 ).* {/ U6 k, D7 ~- b6 {& }+ b, d
( u; x Y: z- f+ D
For Western blots, isolated proximal tubular BBMV proteins were separatedon SDS-PAGE gels, transferred to nitrocellulose membranes, and processed aspreviously described ( 6 ). Theblots were incubated overnight at 4°C with primary antibody (MAP17 1:3,000) and subsequently with secondary HRP-coupled IgGs (Amersham PharmaciaBiotech Europe GmbH, Dübendorf, Switzerland). Immunoreactivity wasvisualized by enhanced chemiluminescence (Pierce, Socochim SA, Lausanne,Switzerland).
u9 @4 @" Y' Y$ X0 W! d. y% `
! H1 @9 G" B, S/ q3 zCell culture, transfections, and immunostainings. OK cells (clone 3B/2) were cultured, transfected, and analyzed by confocal microscopy asdescribed previously ( 12, 25 ). Where indicated, cells were transfected with NH 2 -terminally tagged myc -PDZK1( 12 ), HA-MAP17 (HA-tagpositioned between Met 65 and Met 66 of the original ratclone, accession no. AF402772 ) or HA-MAP17 lacking the last three amino acids(TPM). HA-MAP17-TPM was produced by site-directed mutagenesis (Quick change,Stratagene GmbH, Basel, Switzerland), introducing a stop codon TGA at theposition -3. All transfections were done in the presence ofLipofectamine (GIBCO BRL, Basel, Switzerland). After reaching confluency, transfected cells were processed for staining with antibodies against theendogenus NaPi-IIa cotransporter NaPi-4 (1:100) ( 21 ), MAP17 (1:100), HA(1:1,000), or myc (1:5,000) epitopes (Sigma and Invitrogen AG, Basel,Switzerland, respectively), as well as phalloidin-AlexaFluor (1:20; MolecularProbes, Juro, Luzern, Switzerland) for actin. Swine anti-rabbit FITC-IgG (1:50; Sigma) and goat or donkey anti-mouse Cy3-IgG (1:500) and Texas Red(1:200) (Jackson, Milan, Italy) were used as secondary markers.
: _3 o2 Z0 }- M5 e% L8 l) a! A; d2 e) i% J0 w' r3 _2 v4 g
mRNA content in nephron segments. Mice (males, NMRI, 8 wk old) were fed a high (1.2% P)- or a low (0.1% P)-phosphate diet for 5 days. Eachgroup consisted of three animals. Animals were killed, and single kidneys wereprepared for laser microdissection as described by Gisler et al.( 9 ). Briefly, both S1 and S3 segments of superficial and juxtamedullary nephrons were identified byphase-contrast microscopy. A total area corresponding to 100,000µm 2 ( /- 1%) was microdissected for each sample, and RNAwas extracted (Absolutely RNA Nanoprep Kit, Stratagene GmbH). First-strandcDNA synthesis in a reaction volume of 50 µl was made from total RNA using Taq Man Reverse Transcription Reagents (Applied Biosystems, Rotkreutz, Switzerland) with random hexameres according to the manufacturer's protocol.Relative quantization of MAP17 mRNA was achieved by using ABI PRISM 7700Sequence Detection System (Applied Biosystems) with -actin as aninternal standard. Expression means obtained from three animals per samplegroup were averaged, and one-way ANOVA with Student-Newman-Keuls multiplecomparisons test was applied for statistical analysis between S1 and S3 segments. A probability of P significant.2 l4 b" u' o/ R
$ a1 ~+ i2 N7 U4 o+ \ q* L ?
The sequences of Taq Man probes (Biosearch Technologies, Novato, CA) and primers (Microsynth, Balgach, Switzerland) were as follows: 5'-(6-FAM) TCCCAGGCTCCGGCTCCTCCT (BHQ-1)-3' (probe),5'-AGGACCCCATCTGCCTTGTT-3' (forward),5'-CTTCGCCGTCAACCACTTCT-3' (reverse) for MAP17, 5'-(6-FAM)CCATGAAGATCAAGATCATTGCTCCTCCT (BHQ-1)-3' (probe), 5'-GACAGGATGCAGAAGGAGATTACTG-3' (forward), and5'-CCACCGATCCACACAGAGTACTT-3' (reverse) for -actin. Taq Man probes were chosen to be located across exon-exon boundariesto exclude any amplification of genomic DNA. PCR reaction was performed asdescribed previously in detail( 9 ). Relative MAP17 expressionwas calculated with the standard curve method (ABI PRISM 7700 SequenceDetection System, user bulletin #2). Standard curves for MAP17 and -actin were generated using total kidney mouse RNA.5 |9 {& n+ d5 B2 J- D9 @1 {- ^
( b$ L1 l" X/ D7 M2 D+ h
RESULTS
+ H$ P+ _7 [0 r. ]9 e
* o; [, H3 k/ U X" q- MBy performing a yeast two-hybrid screen against the cytoplasmatically oriented NH 2 terminus of the type IIa Na/Pi-cotransporter one clone, encoding of the small membrane protein MAP17 was identified. MAP17 wasoriginally identified by differential display using mRNA isolated from normaland carcinogenic renal tissue( 15 ). Interestingly, MAP17 isalso found in normal tissue, specifically in the apical membranes of renalproximal tubule cells, where it was described as part of the heteromultimericcomplex comprising PDZK1 and MRP2( 17, 18 ). In addition to PDZK1,other PDZ proteins, such as NHERF-1 and NaPi-Cap2, were shown to be localized in the brush borders of proximal tubules as well( 11, 30 ). It was therefore ofinterest to establish if, besides with PDZK1, MAP17 may interact with theother PDZ proteins. As listed in Tablel, in yeast two-hybrid trap assays, interaction of MAP17 wasobserved only with PDZK1 but not with NHERF-1 or NaPi-Cap2. Furthermore, theinteraction with PDZK1 was found to be specific for the PDZ domain numberfour. These findings were confirmed by pull-down experiments using isolatedrenal proximal brush border membranes and GST fusions of PDZK1, NHERF-1, andNaPi-Cap2. In agreement with the former data, a pull-down of MAP17 wasobserved with GST-PDZK1 only but not with the other GST constructs( Fig. 1 ).) C' m( [: {2 l2 o& n4 Z$ i
! v8 b4 ^# | r" `9 [6 v8 L/ ITable 1. Interaction of MAP17 with different PDZ proteins1 w0 S. Y5 X }1 `* ?" f' i
; x! u3 f$ a2 h3 P3 I9 z2 mFig. 1. GST pull-downs of MAP17 from mouse kidney brush border membrane vesicles(BBMVs). MAP17 was recovered from solubilized mouse proximal tubular BBMVs bythe GST-fusion construct of PDZK1 but not by GST/NaPi-Cap2 or GST/NHERF-1 orby GST alone. After pull-down, MAP17 was detected by Western blot analysis.Positive controls for GST/NaPi-Cap2 and GST/NHERF-1 constructs (ability topull down NaPi-IIa from BBMV) were described by Gisler et al.( 11 ).) O, Z+ H6 ?! g1 a5 i( \) O2 `
9 N* O# U M6 e6 EFurther analyses concerning the renal distribution and expression of MAP17were performed with renal tissue obtained from animals kept under normalconditions. As revealed by immunofluorescence ( Fig. 2 ), MAP17 was localizedin kidney cortex only, exhibiting similar abundances in proximal tubules ofsuperficial and juxtamedullar nephrons, and was absent in medullary rays andin the outer stripe of the outer medulla. Highest intensity was observed in S1segments and gradually decreased toward the S3 segments. Often, in lattersegments, the presence of MAP17 was not detectable. Overall, the distributionof MAP17 was found to be similar to that of the type IIa Na/Pi-cotransporter( Fig. 2 A ). In contrast to MAP17 and NaPi-IIa, PDZK1-mediated immunostaining was observed of equalintensity along the entire proximal segments of all nephron generations( Fig. 2 A ). As seen athigher magnification, MAP17 entirely overlapped with NaPi-IIa and PDZK1 in thebrush borders of S1 segments ( Fig.2 B ). Additionally, we determined the content of MAP17mRNA relatively to that of -actin in S1 and S3 segments, which werecollected by laser microdissection. As illustrated in Fig. 3, highest amount of MAP17mRNA was found in the S1 segments of both superficial and deep nephrons. Although MAP17 was not detectable in S3 segments by immunofluorescence, themRNA content was still around 50% of that found in the S1 segments.8 h2 y8 Z, a9 `+ h' |2 I4 g( V
7 }& Q, F: s) T& r3 X) S
Fig. 2. Immunolocalization of MAP17 in mouse kidney. A S3), whereas PDZK1 wasequally distributed throughout the whole proximal tubule. M, medulla; MR,medullary ray. Bar size = 200 µm. B : colocalization of MAP17 andNaPi-IIa in the brush borders of S1 cells. The specificity of the anti-MAP17antiserum was tested in the absence ( C ) or presence ( D ) ofthe antigenic peptide (50 µg/ml).
; t3 c" r P0 ~7 Z/ D' H
, F! r1 m q0 o0 CFig. 3. MAP17 mRNA expression in mouse kidney proximal tubules. S1 and S3 segmentsof both superficial (sup) and juxtamedullary (juxt) nephrons of animals fed ahigh- or a low-phosphate diet for 5 days were isolated by laser-assistedmicrodissection. Relative quantification of MAP17 mRNA was performed byRT-PCR, with -actin as an internal control. No significant differencesof the Ct values for -actin were observed between the different samples.Bars represent means ± SD of 3 individual experiments using samplesobtained from 3 individual animals. * P P i, inorganic phosphate., B& P- \) g7 i* m1 Q
$ M: c# K# S) H- ]6 Y+ B
Several factors were shown to regulate the brush border content of NaPi-IIa( 14, 22, 24, 28 ). To investigate whetherthe abundance of MAP17 may be influenced similarly, mice were treated withparathyroid hormone or fed diets containing a high or a low amount ofphosphate. As indicated in Fig.4, NaPi-IIa was to a large extent downregulated by parathyroidhormone, yet the distribution of MAP17 remained unchanged. Similarly, thedistribution of PDZK1 was not affected by parathyroid hormone. A comparison ofthe abundances of these proteins between renal tissues obtained from mice fedeither a high- or a low-P i diet also showed a change of theNaPi-IIa abundance but no changes of MAP17 or PDZK1. Real-time PCR experimentsdemonstrated that the amount of MAP17 mRNA in different proximal tubularsegments did not change by a low-phosphate diet compared with a high-phosphate diet ( Fig. 3 ).
2 z& ~6 O k$ h/ T/ d+ U( B' O5 V. [7 d/ J# P( q
Fig. 4. Apical abundance of MAP17, NaPi-IIa, and PDZK1 in proximal tubular S1segments after different treatments. The abundance of NaPi-IIa was found to bealtered by the dietary content of phosphate and after a treatment withparathyroid hormone (PTH), whereas no such alterations were observed for theabundance of MAP17 and PDZK1.
. Q. F; Z- u% ?- @* k' \. f& u
7 U- Z* e" w! U m( g7 n- DThe interaction of MAP17 with PDZK1 was further investigated in OK cells.These cells were chosen because of their endogenous apical expression of thetype IIa Na/Pi-cotransporter and NHERF-1 ( 12, 25 ), whereas neither thepresence of endogenous PDZK1 nor MAP17 could be demonstrated( 12; Sorribas V, unpublished data). As shown in Fig.5 A, after transfection of HA-tagged MAP17 (HA-MAP17), itlocalized in the apical membrane and colocalized with NaPi-IIa within theactin-containing apical patches. After truncation of the COOH-terminalPDZ-binding domain (TPM) of MAP17, the apical appearance was not completelyprevented (HA-MAP17-TPM; Fig.5 B2 ) and in most transfected cells HA-MAP17-TPM wasobserved in the cytoplasm ( Fig.5 B1 ). Whether the presence of MAP17 in the apicalmembrane of OK cells (see Fig.5 A ) would affect the inhibition of Na/Pi cotransport byparathyroid hormone was investigated by phosphate uptake measurements. Results 50%) did notinfluence the kinetic of parathyroid hormone-mediated inhibition of Na/Picotransport (not shown).
0 s( z8 b3 E' A* c; {8 V, `
# |% |; Z3 {4 o8 hFig. 5. Cellular location of MAP17 after transfection in opossum kidney (OK) cells.OK cells were transfected with HA-fused MAP17 ( A ) or MAP17-TPM( B ) and stained for HA (red, A2 ), MAP17 (green, B1 and B2 ), endogenously expressed NaPi-IIa (green, A3 ), and -actin (white, A1 ). HA-MAP17 was consistently observed in theapical patches of OK cells, a pattern reminiscent to -actin clusters andNaPi-IIa, as reflected by the yellow signal of the merge composite( A4 ). After transfection of a COOH termininally truncated MAP17(HA-MAP17-TPM; B1 and B2 ), MAP17 was observed to be mostlyintracellular ( B1 ). B2 : some cells also exhibited an apicallocalization of MAP17.
/ Y* b4 {6 {- P1 F; K/ J7 Q
' [4 t6 ~) P/ A+ M5 D# ]: pIn contrast to MAP17, myc -tagged PDZK1 was, after transfection, found to be distributed throughout the cytoplasm of OK cells (Ref. 12 and Fig. 6 A1 ), suggestingthat apical anchoring sites for PDZK1 in OK cells are missing. To test thispossibility, OK cells were cotransfected with HA-MAP17 and myc -PDZK1.As shown in Fig. 6, A2 - A4, in the presence of HA-MAP17, myc -PDZK1 colocalized with HA-MAP17 at the apical site and was almost absent in the cytoplasm. After cotransfection with the truncated formHA-MAP17-TPM and myc -PDZK1, two different cell populations wereidentified ( Fig. 6 B ).In one population of cells in which HA-MAP17-TPM was localized apically, myc -PDZK1 also exhibited an apical localization( Fig. 6, B1 - B3 ). However, in cells showing a morecytoplasmic localization of HA-MAP17-TPM, myc -PDZK1 remained in thecytoplasm as well ( Fig. 6, B4 - B6 )." M) w9 G! h# T2 h
3 N9 C- t& W. h5 p$ u( @' f0 y
Fig. 6. PDZK1 distribution in OK cells after coexpression with MAP17 or MAP17-TPM.When expressed alone in OK cells, myc -tagged PDZK1 showed a strongcytoplasmatic distribution (stained for myc in red; A1 ). Incontrast, when coexpressed ( A2 to A4 ) with HA-MAP17, PDZK1underwent a shift in its cellular distribution from the cytoplasm( A1 ) toward a patchy, apical localization ( A2 ). A3 :merge of A2 with A3. B : in the case of acoexpression of PDZK1 with HA-MAP17-TPM, the distribution of PDZK1 was foundto be consistent with that of HA-MAP17-TPM. Cells exhibiting apicalHA-MAP17-TPM (green, B2 ) also showed apical location of PDZK1( B1 ); in cells exhibiting a cytoplasmatic distribution ofHA-MAP17-TPM, PDZK1 remained in cytoplasm as well ( B4 to B6 ). Squares represent apical focal planes, and rectangles representconfocal cross sections., t! p2 B4 i5 ^. t0 q; r1 o
0 K4 _. F& h/ ?% v; B
DISCUSSION
+ F6 y% i7 j( I& f; d4 @! w. ?* b
& C2 Y& q, v. |. ^5 M, ~In renal proximal tubular cells, the Na/Pi cotransporter NaPi-IIa wasdescribed to interact with the multiple PDZ protein PDZK1 via a PDZ-bindingmotif contained in its COOH terminus( 11 ). On the basis of theyeast two-hybrid screen, we obtained evidence that NaPi-IIa, via itsNH 2 terminus, interacts with MAP17, which was described to interactalso with PDZK1 ( 16, 18 ). The interaction of MAP17was specific for the NH 2 -terminal tail because no positive reactioncould be observed when either the cytoplasmatically oriented COOH terminus ofNaPi-IIa or a proposed intracellular loop (ICL3, see Ref. 20 ) was used as bait in yeasttwo-hybrid trap assays (data not shown). However, by applying several in vitroassays (GST pull-downs or gel overlays; data not shown), we could not confirman in vitro interaction of the NH 2 terminus of NaPi-IIa with MAP17.Therefore, we conclude that the NaPi-IIa/MAP17 interaction may be of a weaknature but could be stabilized in the complex MAP17/NaPi-IIa/PDZK1.7 H2 s. u# Q, e; @9 a, D
# g0 C1 N6 z/ K, C0 qApart from PDZK1, NaPi-IIa was shown to interact with other PDZ proteinssuch as NHERF-1 and NaPi-Cap2( 11 ). As demonstrated in thisstudy, MAP17 exclusively interacted with PDZK1, because by yeast two-hybridtrap assays and GST pull-downs, no interaction of MAP17 with either NHERF-1 orwith NaPi-Cap2 was observed. In contrast to interaction described for humanisoforms of MAP17 and PDZK1, where MAP17 was shown to interact with both thefirst and the fourth PDZ domain of PDZK1 (using yeast two-hybrid trap assay;see Ref. 17 ), we show that byusing the same method the mouse isoform of MAP17 interacts exclusively withthe fourth PDZ domain of PDZK1.( a1 {8 [- q! ^6 Z) J, j
# P$ l1 E; o5 Z8 C; p1 K3 b- [The abundance of MAP17 has been associated with a number of differenttumors, but MAP17 is also expressed in proximal tubules of normal, healthykidneys ( 15, 16 ). As shown in the present study, the abundance of MAP17 in adult mice kidney was found to be highest inthe S1 proximal segments and gradually decreased along the proximal tubule,being almost absent in the S3 segments. No differences were observed betweensuperficial and juxtamedullary nephrons. The observed intranephronheterogeneity of MAP17 entirely matched the distribution of NaPi-IIa but notthat of PDZK1, which was uniformly abundant along the entire proximal tubule.Uniform intranephron abundance of NaPi-IIa can be achieved by alow-P i diet ( 22 ).Under such conditions, we could not observe a parallel increase of MAP17 inthe S3 segments. Thus these findings suggest that the proposedheteromultimeric protein complex comprising PDZK1, NaPi-IIa, and MAP17 is,with regard to the stochiometry MAP17/NaPi-IIa, inhomogeneous (afterupregulation of NaPi-IIa) along the entire proximal tubular segment. Whereas in the S1 segments MAP17 is part of the PDZK1/NaPi-IIa complex, in S3 segmentsthis appeared not to be the case. K7 a4 t% z3 `5 U+ D
9 p" q1 H7 c( F- z9 _9 d6 W
On the basis of the observations that unlike NaPi-IIa, PDZK1 is notinternalized by various stimuli, it is assumed that the NaPi-IIa/PDZK1interaction is dynamic. As the distribution of MAP17 was not affected aftertreatment with parathyroid hormone, it is concluded that the interaction ofMAP17 with PDZK1 is not regulated, at least not by parathyroidhormone-activated signaling cascades.
# f2 Y2 \/ V7 W/ \: U! }% R. ^
5 |: v$ D* c( |Because OK cells do not endogenously express MAP17 (Sorribas V, unpublisheddata) nor could the presence of PDZK1 at the protein level be demonstrated inthese cells ( 12 ), OK cells were used as a model for transfection studies aiming to investigate theinteraction of MAP17 and PDZK1. When transfected alone, MAP17 was found to belocalized in the apical membrane and showed the same distribution pattern(patches) as the NaPi-IIa cotransporter. Therefore, this observation suggestedthat PDZK1 is not necessary for the apical sorting and/or positioning of MAP17. Interestingly, in proximal tubules of a PDZK1 knockout mouse model, theabundance of MAP17 was not affected by the lack of PDZK1( 19 ) supporting our hypothesisthat the apical positioning/sorting of MAP17 does not depend on PDZK1 but that other, yet unknown, interactions may be required. Although NHERF-1 was shownto be apical in OK cells and proximal tubules( 12, 30 ), an interaction of MAP17with NHERF-1 can be excluded because by in vitro studies and yeast two-hybridtrap assays no evidence for such an interaction was obtained. In contrast toMAP17, when transfected alone, PDZK1 did not exhibit an apical distribution but was found uniformly distributed throughout the cytoplasm, which may be dueto the lack of an appropriate interaction site at the apical membrane. Ourresults obtained by cotransfecting OK cells with MAP17 and PDZK1 providedevidence that such an apical anchoring site could be provided by MAP17.
# ?( V, q; Y; f% E
/ @* u5 K, s- p5 H' [Despite the lack of endogenously expressed MAP17 and PDZK1, regulation ofthe Na/Pi-IIa cotransporter in OK cells closely resembles the one observed inproximal tubules ( 14, 25 ). That discrepancy withrespect to possible roles of PDZK1 and/or MAP17 in the regulation of NaPi-IIacould be explained by the existence of compensatory mechanisms in OK cells,such as the known interaction between NaPi-IIa and endogenously expressed NHERF-1 ( 12 ) or interactionswith other not known proteins. In future, it will be of interest to explorethe impact of deficiencies of MAP17 and/or PDZK1 on the hormonal regulation ofNa/Pi cotransporter under in vivo conditions.
* A5 N8 {. z' F( ^( |+ |: q
% H! U2 T8 S0 |1 }5 A% C0 F1 I% i4 yIn summary, our studies provided evidence that MAP17 is part of thePDZK1/NaPi-IIa complex in brush border membranes of proximal tubular cells (S1segments), that within this complex MAP17 may weakly interact with NaPi-IIa aswell, and that the abundance of MAP17 is not affected by maneuvers known toalter the content of NaPi-IIa. Furthermore, our data suggest that the apicallocalization of MAP17 does not depend on the presence of PDZK1 and that arecruitment of PDZK1 to the apical membrane is at least partially dependent onthe presence of MAP17. This finding could be of interest regarding the stillunknown role of MAP17 in various tumors. As the abundance of MAP17 isincreased in a number of carcinomas, our results imply that by increasing theamount of MAP17, additional PDZK1 may be positioned to distinct cellularsites. In fact, it was shown that in carcinomas the abundance of PDZK1 isincreased as well ( 17, 18 ). Because PDZK1 encompassesseveral PDZ domains that, in addition to NaPi-IIa, interact with numerousmembrane transporters such as MRP2( 17 ) and CFEX, URAT1 or OCTN1( 10 ), it could be speculated that increased amounts of MAP17 as observed in numerous cancers could recruitor stabilize more PDZK1, resulting in a cancer-typical organization ofdifferent membrane transporters.& q# q9 [, s# G4 _; W6 W# L
5 ?" S, h; t3 x2 PDISCLOSURES- D) a5 B8 Z- h
: t: Y5 G3 s' v, MThis work was supported by the Swiss National Foundations (Grant 31-65397.01 to H. Murer) and the Fridericus-Stiftung (Vaduz, FL).) v R: a7 j% k8 O3 d; C0 p; y5 }
! e; I8 c$ @1 [0 KACKNOWLEDGMENTS' {3 U5 ~3 B4 A
& c6 m" I) h5 b' a5 x# ]# M) ?
RTp51 was a gift from Dr. U. Hübscher, Veterinary Biochemistry, University of Zürich. The antibody against the endogenous NaPi-4 from theOK cells was kindly provided by Dr. E. D. Lederer. We thank C. Gasser forassistance in preparing the figures.9 X- X. b8 N# A
【参考文献】. a5 P0 y, M' c8 P! I) S
Bacic D, SchulzN, Biber J, Kaissling B, Murer H, and Wagner CA. Involvement of theMAPK-kinase pathway in the PTH mediated regulation of the proximal tubule typeIIa Na/Pi cotransporter in mouse kidney. PflügersArch 446:52-60, 2003.
4 A: K6 I1 s9 Q
; x4 R3 O* U/ a! X6 B8 y4 a/ a
' j1 a0 _' _- t) ]. }8 Q8 ]5 R+ w
% ~" A; k2 g4 KBeck L,Karaplis AC, Amizuka N, Hewson AS, Ozawa H, and Tenenhouse HS. Targetedinactivation of Npt2 in mice leads to severe renal phosphate wasting,hypercalciuria, and skeletal abnormalities. Proc Natl Acad SciUSA 95:5372-5377, 1998.
$ z8 |) U8 J1 E# G* c1 V# x+ @" p
4 D% H P- j! U4 h/ V6 I8 h( q0 E: h0 u9 q) o
. ]& w F- O! c5 R
Béranger F, Aresta S, de Gunzburg J, and Camonis J. Getting more from the two-hybrid system: N-terminal fusions to LexA areefficient and sensitive baits for two-hybrid studies. Nucleic AcidsRes 25:2035-2036, 1997.9 C4 F7 _' O& O( l, w8 a) K
; e! n L0 N1 U3 N' j$ C
/ S8 K5 y6 T/ g2 f3 [7 m9 [ z& x) W
Biber J. Emerging roles of transporter-PDZ complexes in renal proximal tubularreabsorption. Pflügers Arch 443: 3-5,2001.1 h/ H. m* c) _2 z) m' |/ \4 _
9 m; i- w4 A; y( A4 X
- R3 z' {8 Q9 b
( x* Z) {$ N3 i" Q( p( w2 H+ sBiber J,Stieger B, Haase W, and Murer H. A high yield preparation for rat kidneybrush border membranes. Different behaviour of lysosomal markers. Biochim Biophys Acta 647:169-176, 1981.
6 T; J" N. q S' ~ h4 u
/ a V4 m+ g9 {$ X! o, ^! Q" `
; N& x! n. C7 \( Q1 R& |
1 x) C+ W" t" w7 ICuster M,Lotscher M, Biber J, Murer H, and Kaissling B. Expression of Na-Picotransport in rat kidney: localization by RT-PCR and immunohistochemistry. Am J Physiol Renal Fluid Electrolyte Physiol 266: F767-F774,1994.! A/ V7 |3 s& W2 T5 b; s+ v
7 p: L4 {/ q @; A7 i
) w6 j; k; s7 D# o
$ }6 T, P: G, u: C* ?& MDalton S andTreisman R. Characterization of SAP-1, a protein recruited by serumresponse factor to the c- fos serum response element. Cell 68:597-612, 1992.3 b. \4 {* y4 d
% ` p1 H, R+ t, ~, H3 r& Q
4 o2 G) [2 E* m: r6 t
4 ^; z$ x1 i) L! pGietz RD,Schiestl RH, Willems AR, and Woods RA. Studies on the transformation ofintact yeast cells by the LiAc/SS-DNA/PEG procedure. Yeast 11:355-360, 1995./ n9 h/ B1 T# ]& G9 p( B
: `9 l& E9 T; _, G: g
; z' i& o! {/ ?2 h8 x; q5 S
7 N& p r, t) I& n- X: W. z' b
Gisler SM,Madjdpour C, Bacic D, Pribanic S, Taylor SS, Biber J, and Murer H. PDZK1.II. An anchoring site for the PKA-binding protein D -AKAP2 in renalproximal tubular cells. Kidney Int Inpress.9 s" I3 E# g$ I" v! h4 E4 x( \/ F
$ O- Y% a8 P% J3 W. D% ]' I5 D- q' g
( Y; |4 {7 g( j3 W5 U" |% d
( `0 B+ @' h3 i8 Q8 QGisler SM,Pribanic S, Bacic D, Forrer P, Gantenbein A, Sabourin LA, Tsuji A, Zhao Z,Manser E, Biber J, and Murer H. PDZK1. I. A major scaffolder in brushborders of proximal tubular cells. Kidney Int Inpress.- t/ B3 J c" Y( c% g
2 I4 L" H0 e- B+ a: P
1 I; C' n/ n& X5 L" Y9 ~5 M, }* `' o5 b D; O
Gisler SM,Stagljar I, Traebert M, Bacic D, Biber J, and Murer H. Interaction of thetype IIa Na/Pi cotransporter with PDZ proteins. J BiolChem 276:9206-9213, 2001.) P; v: l& q4 k7 q5 C, ?
# Z% k7 M! Y; G5 S: e( m$ i8 j
1 _. \8 J. p+ Y# N# O, m, f. E- u( l+ m; Z& C' u `/ X9 K
Hernando N,Déliot N, Gisler SM, Lederer E, Weinman EJ, Biber J, and Murer H. PDZ-domain interactions and apical expression of type IIa Na/Picotransporters. Proc Natl Acad Sci USA 99: 11957-11962,2002.1 ?4 f! T1 `7 i: J3 ]9 v) E
! Z. X/ A: N; F8 g! g6 W2 a: T
- t+ N3 }2 q" J" w8 G
6 J# y$ l6 M+ a( {9 C5 |Inoue H, NojimaH, and Okayama H. High efficiency transformation of Escherichiacoli with plasmids. Gene 96: 23-28,1990.$ D% n2 J$ t; Y, ]4 I/ f
% j& C( W1 \9 i
% Z! @) [& q% ~7 S9 T0 v/ v
7 N. M# N% E, T
Keusch I,Traebert M, Lötscher M, Kaissling B, Murer H, and Biber J. Parathyroid hormone and dietary phosphate provoke a lysosomal routing of theproximal tubular Na/Pi-cotransporter type II. KidneyInt 54:1224-1232, 1998.7 u) n1 `; g8 l3 ]9 E
" h r' g5 `$ ~; r
; p1 H* U' u+ K2 I% L' ^+ ?" ^; v5 o& l* s4 z- l
Kocher O,Cheresh P, Brown LF, and Lee SW. Identification of a novel geneselectively upregulated in human carcinomas, using the differential displaytechnique. Clin Cancer Res 1:1209-1215, 1995.1 |& k1 r& U9 u1 |
5 P# b) m7 {+ K2 k. b# k( A. U% X; R9 Z) b1 ~: B( H
4 ?4 e, {4 p' n' l% i% k; k% q
Kocher O,Cheresh P, and Lee SW. Identification and partial characterization of anovel membrane-associated protein (MAP17) upregulated in human carcinomas andmodulating cell replication and tumor growth. Am JPathol 149:493-500, 1996.% T3 S) `8 m7 S3 s. M
1 F5 |$ x- R; z* f
- ~ I! p6 J6 N1 T) t
: f/ Y ^2 w/ M: i7 \/ P1 q
Kocher O,Comella N, Gilchrist A, Pal R, Tongazzi K, Brown LF, and Knoll JHM. PDZK1,a novel PDZ domain-containing protein upregulated in carcinomas and mapped tochromosome 1q21 , interacts with cMOAT (MRP2), the multidrugresistance-associated protein. Lab Invest 79: 1161-1170,1999." N) ~. T9 d9 n V1 h0 r! f
/ l6 y! T- w+ q+ v
2 e" M5 A7 D0 m( l
& r5 ?8 I$ d& d* ?) QKocher 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, 1998.3 b' r, X6 x" W% D9 f+ j4 C
/ M' _* |( G$ P/ x7 g+ u
# P8 L+ Z# @# r0 T: E
3 C' X! L5 X) mKocher O, PalR, Roberts M, Cirovic C, and Gilchrist A. Targeted disruption of the PDZK1gene by homologous recombination. Mol Cell Biol 23: 1175-1180,2003.
# ]/ o3 B. ^6 L7 l6 J, x, {- x- O- i, h" d2 a" S
0 E2 D* t, X. s* J. i& y: M
2 N9 ?* ], N; t; O
Lambert G,Traebert M, Hernando N, Biber J, and Murer H. Studies on the topology ofthe renal type II NaPi-cotransporter. PflügersArch 437:972-978, 1999.
2 R- m3 g1 x3 P: h& k( M5 P
% L+ ]: @( ]5 ]7 A; P8 |/ V$ @ q8 Q: o
3 k" }0 a! U! @$ C$ w$ q. [, tLederer 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.
7 W- @ n0 Q0 K7 V3 ~1 m" [5 W6 C5 F
) t2 r' E7 j9 v, v$ y4 l% F7 v8 b0 \$ o0 d0 S/ N
Levi M,Lötscher M, Sorribas V, Custer M, Arar M, Kaissling B, Murer H, and BiberJ. Cellular mechanisms of acute and chronic adaptation of rat renalP i transporter to alterations in dietary P i. Am J Physiol Renal Fluid Electrolyte Physiol 267: F900-F908,1994.2 F% z$ g; ~. @- b
7 X" L3 b9 V0 F2 U" g
8 j* u k, c- N5 w
- F ~: Z0 a! M$ R) n/ d+ r# tLoffing J,Loffing-Cueni D, Valderrabano V, Klausli L, Hebert SC, Rossier BC, HoenderopJG, Bindels RJ, and Kaissling B. Distribution of transcellular calcium andsodium transport pathways along mouse distal nephron. Am J PhysiolRenal Physiol 281:F1021-F1027, 2001.
9 {: G: E0 w) t4 e, P0 n! Y; c" ~
^! D5 z0 ]) f. e$ L+ }
. M K/ J+ Z. F& u7 t+ v6 S( ] B- _* x) R8 J4 z
Murer H,Hernando N, Forster I, and Biber J. Proximal tubular phosphatereabsorbtion: molecular mechanisms. Physiol Rev 80: 1373-1400,2000.% }! |/ F1 C4 e/ N. Y
8 g+ o2 X& z6 w' _( `
6 _; }3 d6 V6 f ^& G n0 ~" }# q9 v: o# R$ ]
Pfister MF,Lederer E, Forgo J, Ziegler U, Lötscher M, Quabius ES, Biber J, and MurerH. Parathyroid hormone-dependent degradation of type IINa /P i cotransporters. J BiolChem 272:20125-20130, 1997.) w) n; u4 k( i
% B) L) f% d8 X; D& |+ U7 W' F6 \8 u
; v/ u8 l7 M5 T2 A, a' m! bShenolikar S,Voltz JW, Minkoff CM, Wade JB, and Weinman EJ. Targeted disruption of themouse NHERF-1 gene promotes internalization of proximal tubulesodium-phosphate cotransporter type IIa and renal phosphate wasting. Proc Natl Acad Sci USA 99:11470-11475, 2002.
% }% p6 O+ L5 W2 }, o0 g
2 E4 e0 J8 o& o: l9 T2 z
' o+ q. u" ?- @5 w4 Z U
0 V- }' B$ o/ B# L. yTasara T,Amacker M, and Hübscher U. Intramolecular chimeras of the p51 subunitbetween HIV-1 and FIV reverse transcriptases suggest a stabilizing functionfor the p66 subunit in the heterodimeric enzyme. Biochemistry 38:1633-1642, 1999.9 W/ I$ Z% D* L( L+ ^
. T+ r# E0 p' j
1 ^% F! _) v" G4 w3 E
0 u* N2 v S2 y+ h/ T2 \
Traebert M,Roth J, Biber J, Murer H, and Kaissling B. Internalization of proximaltubular type II Na-Pi cotransporter by PTH: immunogold electron microscopy. Am J Physiol Renal Physiol 278:F148-F154, 2000.' [ q; l- R9 f1 H% L* \
. e, s: K8 \( @; U8 ~; g2 ^
8 t5 K* e: M, \5 c2 M, S, g/ M
# I, l$ p. h/ H& XVojtek AB,Hollenberg SM, and Cooper JA. Mammalian Ras interacts directly with theserine/threonine kinase Raf. Cell 74: 205-214,1993.
- _5 f* L) z. n. F% J- D: K6 `. I* F6 u' L$ g+ g+ @
, k7 G/ X G4 K4 B4 u. J+ ?7 I1 L
Weinman EJ,Steplock D, Wang Y, and Shenolikar S. Characterization of a proteincofactor that mediates protein kinase A regulation of the renal brush bordermembrane Na -H exchanger. J ClinInvest 95:2143-2149, 1995. |
|
发表于 2009-4-21 13:45
