Advertisement

The Phox Homology (PX) Domain Protein Interaction Network in Yeast*

      The phox homology (PX) domain is a phosphoinositide-binding domain that is conserved from yeast to human. Here we show for the first time by genome-wide two-hybrid screens and in vitro binding assays that the PX domain is a bona fide protein interaction domain. The yeast PX domain-only proteins Grd19p (YOR357C) and Ypt35p (YHR105W), as well as the isolated PX domains from Mvp1p (YMR004W), Snx42p/Cvt20p/Atg20p (YDL113C), Vam7p (YGL212W), and Vps17p (YOR132W), yielded a total of 40 reproducible two-hybrid interactions. Thirty-five interactions were found for the full-length proteins of Bem1p (YBR200W), Snx42p, Snx4p/Cvt13p (YJL036W), Vam7p, Vps5p (YOR069W), and Vps17p, but these appear not to require the PX domain, because these interactions could not be reproduced with PX-only baits. Interactions of Grd19p, Vam7p, Vps5p, Vps17p, and Ypt35p with members of the Yip1p family of proteins were detected consistently and were verified by in vitro binding assays. The N-terminal cytoplasmic domain of Yip1p and Yif1p mediates these interactions with PX domains. A mutation in the lipid-binding pocket of Ypt35p that reduces lipid binding markedly does not affect these PX domain protein interactions, arguing that lipid binding uses a different interaction surface than protein binding.
      Phosphoinositides (PIs)
      The abbreviations used are: PI, phosphoinositide; PtdIns, phosphatidylinositol; PX, phox homology domain; SH3, Src homology 3; ER, endoplasmic reticulum.
      1The abbreviations used are: PI, phosphoinositide; PtdIns, phosphatidylinositol; PX, phox homology domain; SH3, Src homology 3; ER, endoplasmic reticulum.
      act as signaling molecules in a variety of cellular processes including vesicular trafficking and cell motility (
      • Simonsen A.
      • Wurmser A.E.
      • Emr S.D.
      • Stenmark H.
      The role of phosphoinositides in membrane transport..
      ,
      • Toker A.
      Phosphoinositides and signal transduction..
      ,
      • Gruenberg J.
      Lipids in endocytic membrane transport and sorting..
      ). PI signaling is mediated by proteins that contain one or several lipid-binding domains such as PH, FYVE, ENTH, ANTH, FERM, PHD, Tubby, and C2 domains, all of which have been reported to bind PIs. Recently, the PX domain was shown to represent another PI-binding domain (reviewed in Refs.
      • Hurley J.H.
      • Meyer T.
      Subcellular targeting by membrane lipids..
      ,
      • Itoh T.
      • Takenawa T.
      Phosphoinositide-binding domains: Functional units for temporal and spatial regulation of intracellular signalling..
      ,
      • Lemmon M.A.
      Phosphoinositide recognition domains..
      ,
      • DiNitto J.P.
      • Cronin T.C.
      • Lambright D.G.
      Membrane recognition and targeting by lipid-binding domains..
      ).
      Phox homology (PX) domains were originally identified as a common motif of about 120 amino acids found in the p40phox and p47phox subunits of the neutrophil NADPH oxidase (Phox) complex (
      • Ponting C.P.
      Novel domains in NADPH oxidase subunits, sorting nexins, and PtdIns 3-kinases: Binding partners of SH3 domains?.
      ). This domain is conserved from yeast to human, and more than 400 examples have been annotated in sequence databases (e.g. SMART (
      • Letunic I.
      • Copley R.R.
      • Schmidt S.
      • Ciccarelli F.D.
      • Doerks T.
      • Schultz J.
      • Ponting C.P.
      • Bork P.
      SMART 4.0: Towards genomic data integration..
      )). PX domain-containing proteins have been implicated in highly diverse functions such as cell signaling, vesicular trafficking, protein sorting, and lipid modification (reviewed in Refs.
      • Odorizzi G.
      • Babst M.
      • Emr S.D.
      Phosphoinositide signaling and the regulation of membrane trafficking in yeast..
      and
      • Worby C.A.
      • Dixon J.E.
      Sorting out the cellular functions of sorting nexins..
      ).
      PX domains show relatively little sequence conservation, yet their structure appears to be highly conserved. All published crystal structures exhibit a clearly defined PI-binding pocket (
      • Bravo J.
      • Karathanassis D.
      • Pacold C.M.
      • Pacold M.E.
      • Ellson C.D.
      • Anderson K.E.
      • Butler P.J.
      • Lavenir I.
      • Perisic O.
      • Hawkins P.T.
      • Stephens L.
      • Williams R.L.
      The crystal structure of the PX domain from p40(phox) bound to phosphatidylinositol 3-phosphate..
      ,
      • Hiroaki H.
      • Ago T.
      • Ito T.
      • Sumimoto H.
      • Kohda D.
      Solution structure of the PX domain, a target of the SH3 domain..
      ,
      • Lu J.
      • Garcia J.
      • Dulubova I.
      • Sudhof T.C.
      • Rizo J.
      Solution structure of the Vam7p PX domain..
      ,
      • Zhou C.Z.
      • de La Sierra-Gallay I.L.
      • Quevillon-Cheruel S.
      • Collinet B.
      • Minard P.
      • Blondeau K.
      • Henckes G.
      • Aufrere R.
      • Leulliot N.
      • Graille M.
      • Sorel I.
      • Savarin P.
      • de la Torre F.
      • Poupon A.
      • Janin J.
      • van Tilbeurgh H.
      Crystal structure of the yeast Phox homology (PX) domain protein Grd19p complexed to phosphatidylinositol-3-phosphate..
      ). Although phosphatidylinositol-3-phosphate (PtdIns(3)P) is the primary target of PX domains, binding to phosphatidic acid, phosphatidylinositol-3,4-bisphosphate (PtdIns(3,4)P2), phosphatidylinositol-3,5-bisphosphate (PtdIns(3,5)P2), phosphatidylinositol-4,5-bisphosphate (PtdIns(4,5)P2), and phosphatidylinositol-3,4,5-trisphosphate (PtdIns(3,4,5)P3) has been reported as well (
      • Karathanassis D.
      • Stahelin R.V.
      • Bravo J.
      • Perisic O.
      • Pacold C.M.
      • Cho W.
      • Williams R.L.
      Binding of the PX domain of p47(phox) to phosphatidylinositol 3,4-bisphosphate and phosphatidic acid is masked by an intramolecular interaction..
      ,
      • Xu Y.
      • Seet L.F.
      • Hanson B.
      • Hong W.
      The Phox homology (PX) domain, a new player in phosphoinositide signalling..
      ).
      The yeast genome encodes 15 PX domain-containing proteins that are involved in vesicular trafficking, cell signaling, control of bud emergence and cell polarity, or are of largely unknown function (Table I). All yeast PX domains (except the sequence outlier Ypr097p) bind specifically to PtdIns(3)P with different affinities (
      • Yu J.W.
      • Lemmon M.A.
      All phox homology (PX) domains from Saccharomyces cerevisiae specifically recognize phosphatidylinositol 3-phosphate..
      ).
      Although most PX domains contain additional domains—or at least additional protein sequences—two yeast proteins, Grd19p and Ypt35p, consist almost exclusively of the PX domain (Table I). Assuming that signaling proteins are not likely to bind to a membrane for its own sake suggested that these PtdIns(3)P-specific PX domain-only proteins must also be involved in protein-protein interactions in order to transduce a signal from the membrane to some effector. This is further supported by several genetic observations. For example, yeast strains with a deletion of their GRD19 gene mis-sort the TGN membrane proteins A-ALP, Pep12p, and Kex2p to the vacuole (
      • Voos W.
      • Stevens T.H.
      Retrieval of resident late-Golgi membrane proteins from the prevacuolar compartment of Saccharomyces cerevisiae is dependent on the function of Grd19p..
      ,
      • Hettema E.H.
      • Lewis M.J.
      • Black M.W.
      • Pelham H.R.
      Retromer and the sorting nexins Snx4/41/42 mediate distinct retrieval pathways from yeast endosomes..
      ).
      A direct involvement of the PX domain in protein-protein interactions has been observed in the p47phox subunit of the neutrophil NADPH oxidase complex. p47phox is held in an inactive state by an intramolecular interaction between its C-terminal Src homology 3 (SH3) domain and its PX domain. Upon activation by invading microorganisms, p47phox becomes phosphorylated and the PX-SH3 interaction is disrupted (
      • Sumimoto H.
      • Kage Y.
      • Nunoi H.
      • Sasaki H.
      • Nose T.
      • Fukumaki Y.
      • Ohno M.
      • Minakami S.
      • Takeshige K.
      Role of Src homology 3 domains in assembly and activation of the phagocyte NADPH oxidase..
      ,
      • Leto T.L.
      • Adams A.G.
      • de Mendez I.
      Assembly of the phagocyte NADPH oxidase: Binding of Src homology 3 domains to proline-rich targets..
      ). The C-terminal SH3 domain interacts with a proline-rich region of the p47phox-PX domain (
      • Hiroaki H.
      • Ago T.
      • Ito T.
      • Sumimoto H.
      • Kohda D.
      Solution structure of the PX domain, a target of the SH3 domain..
      ).
      Moreover, a recent large-scale two-hybrid screen has shown that one PX-only protein from yeast, Ypt35p (YHR105W), interacts with several proteins (
      • Uetz P.
      • Giot L.
      • Cagney G.
      • Mansfield T.A.
      • Judson R.S.
      • Knight J.R.
      • Lockshon D.
      • Narayan V.
      • Srinivasan M.
      • Pochart P.
      • Qureshi-Emili A.
      • Li Y.
      • Godwin B.
      • Conover D.
      • Kalbfleisch T.
      • Vijayadamodar G.
      • Yang M.
      • Johnston M.
      • Fields S.
      • Rothberg J.M.
      A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae..
      ). In order to investigate a possible role of the PX domain as a bona fide protein interaction module, we used PX-domain only bait proteins (for all yeast PX domain-containing proteins) in a genome-wide two-hybrid screen to search for binding partners of this domain. Our results argue that the PX domain is a bona fide protein-protein interaction module and binds predominantly to membrane proteins that are involved in vesicular transport.

      EXPERIMENTAL PROCEDURES

       Plasmids and Strains—

      pOBD2 and pOAD (
      • Cagney G.
      • Uetz P.
      • Fields S.
      High-throughput screening for protein-protein interactions using two-hybrid assay..
      ) low-copy plasmids were used as two-hybrid bait and prey vectors, respectively. For the Grd19p, Spo14p, and Snx42p/Cvt20p baits, the high-copy plasmid pGBD (
      • James P.
      • Halladay J.
      • Craig E.A.
      Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast..
      ) has been used instead (kindly provided by E. Conibear). These vectors express fusions of the Gal4 protein, Gal4AD-ORF (prey vector) and Gal4DBD-ORF (bait vectors).
      Construction of the Gal4DBD-ORF fusions was performed by means of PCR and recombination (
      • Ma H.
      • Kunes S.
      • Schatz P.J.
      • Botstein D.
      Plasmid construction by homologous recombination in yeast..
      ) as described in Ref.
      • Uetz P.
      • Giot L.
      • Cagney G.
      • Mansfield T.A.
      • Judson R.S.
      • Knight J.R.
      • Lockshon D.
      • Narayan V.
      • Srinivasan M.
      • Pochart P.
      • Qureshi-Emili A.
      • Li Y.
      • Godwin B.
      • Conover D.
      • Kalbfleisch T.
      • Vijayadamodar G.
      • Yang M.
      • Johnston M.
      • Fields S.
      • Rothberg J.M.
      A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae..
      . Transformation was performed using the lithium acetate procedure (
      • Ito H.
      • Fukuda Y.
      • Murata K.
      • Kimura A.
      Transformation of intact yeast cells treated with alkali cations..
      ). Bait constructs were transformed into yeast strain PJ69-4α (
      • Uetz P.
      • Giot L.
      • Cagney G.
      • Mansfield T.A.
      • Judson R.S.
      • Knight J.R.
      • Lockshon D.
      • Narayan V.
      • Srinivasan M.
      • Pochart P.
      • Qureshi-Emili A.
      • Li Y.
      • Godwin B.
      • Conover D.
      • Kalbfleisch T.
      • Vijayadamodar G.
      • Yang M.
      • Johnston M.
      • Fields S.
      • Rothberg J.M.
      A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae..
      ) and preys into PJ69-4a (
      • James P.
      • Halladay J.
      • Craig E.A.
      Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast..
      ).
      For strains expressing the Gal4DBD-ORF fusions, successful cloning was confirmed by DNA sequencing.
      PX domains cloned into pGSTag (
      • Ron D.
      • Dressler H.
      pGSTag—A versatile bacterial expression plasmid for enzymatic labeling of recombinant proteins..
      ) for expressing them as GST fusions were kindly provided by Jong W. Yu and Mark A. Lemmon (
      • Yu J.W.
      • Lemmon M.A.
      All phox homology (PX) domains from Saccharomyces cerevisiae specifically recognize phosphatidylinositol 3-phosphate..
      ).
      The plasmid pGSTag-Ypt35Y123A was constructed using the Quick-Change™ site-directed mutagenesis kit (Stratagene, La Jolla, CA) using the following primers: CCA AAC CTA TAA ACG AGC CTC CGA TTT CGT CAG (forward), CTG ACG AAA TCG GAG GCT CGT TTA TAG GTT TGG (reverse), with mutated bases shown in bold.

       Two-Hybrid Screens and Retests—

      An array containing most of the ∼6,000 Saccharomyces cerevisiae ORFs expressed as Gal4AD fusions was used to screen for interacting proteins (
      • Uetz P.
      • Giot L.
      • Cagney G.
      • Mansfield T.A.
      • Judson R.S.
      • Knight J.R.
      • Lockshon D.
      • Narayan V.
      • Srinivasan M.
      • Pochart P.
      • Qureshi-Emili A.
      • Li Y.
      • Godwin B.
      • Conover D.
      • Kalbfleisch T.
      • Vijayadamodar G.
      • Yang M.
      • Johnston M.
      • Fields S.
      • Rothberg J.M.
      A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae..
      ). Haploid transformants expressing either a full-length Gal4DBD-ORF fusion protein or a Gal4DBD-PX-domain-only fusion protein were mated to the array. The resulting diploids were pinned with a Biomek 2000 Laboratory Automation Workstation (Beckman-Coulter, Fullerton, CA) onto selective media. Positive prey clones from a first-round screen were re-arrayed as quadruplicate copies and tested again for reproducibility. All retests were carried out at least eight times, and their results were used to assign “quality” scores to our positives: positives that were reproducible in at least eight out of 12 or more tests were assigned values of 3, positives that were found at least four times out of 12 (or more) tests were assigned scores of 2, and positives that were found two to three times were assigned values of 1. Some highly reproducible positives were classified as 1 if the results were difficult to interpret due to high background or other reasons. Two-hybrid positives were classified by Gene Ontology descriptors as provided by the Saccharomyces Genome Database (www.yeastgenome.org).

       Protein Expression and GST-Pulldowns—

      Expression clones of pGSTag-GST-PX domain fusion proteins have been described previously (
      • Yu J.W.
      • Lemmon M.A.
      All phox homology (PX) domains from Saccharomyces cerevisiae specifically recognize phosphatidylinositol 3-phosphate..
      ,
      • Ron D.
      • Dressler H.
      pGSTag—A versatile bacterial expression plasmid for enzymatic labeling of recombinant proteins..
      ). The PX domains fused to GST corresponded to the following fragments: Bem1p- (amino acids 269–418), Bem3p- (502–637), Grd19p- (1–162), Mdm1p- (776–908), Mvp1p- (121–258), Snx4p- (20–170), Snx42p- (150–316), Vps5p- (271–397), Vps17p- (88–234), and Ypt35- (65–214). Proteins were expressed in Escherichia coli and purified on glutathione-Sepharose as recommended by the manufacturer (Amersham Pharmacia Biotech, Piscataway, NJ).
      Modified primers for Yip1p, Yif1p, and Yip4p, containing a T7 promoter and an eukaryotic translation inititation sequence, were used for in vitro transcription and translation using the TnT™-coupled reticulocyte system in the presence of [35S]methionine (Promega, Madison, WI). PCR primers were as follows: Yip1p forward, GGA TCC TAA TAC GAC TCA CTA TAG GGA ACA GCC ACC ATG TCT TTC TAC AAT ACT AGT; Yip1p reverse, TCA GAC AAA AAT TAC CAT GAG; Yip1p-N terminus reverse, GAT CCC CGG GAA TTG CCA TGC CCA GCC AAA TCT GAA TCG TT; Yif1p forward, GGA TCC TAA TAC GAC TCA CTA TAG GGA ACA GCC ACC ATG TCT TAT AAT CCG TAC GCA; Yif1p-reverse, TCA ACC CAT TAA CCA CAT TAG; Yif1p-N terminus reverse, GAT CCC CGG GAA TTG CCA TGG TCC ATA ATT CGT TGC CAG TT; Yip4p forward, GGA TCC TAA TAC GAC TCA CTA TAG GGA ACA GCC ACC ATG TCA TAC GGA AGA GAA GAC; Yip4p reverse, TCA GAA CTT TCT GCC GTG GCT.
      GST-PX domain fusion proteins or GST alone were coupled to glutathione-Sepharose beads (Amersham Pharmacia Biotech) and incubated with 5 μl of in vitro-translated proteins in pulldown buffer (20 mm Tris pH 7,6; 1 mm mercaptoethanol; 3 mm EDTA, 150 mm NaCl; 1% Nonidet P-40) for 2 h at 4 °C under rotation. Bound proteins were washed six times (for 10 min under rotation at 4 °C) with pulldown buffer at decreasing detergence concentration (1 × 1%, 4 × 0.1% and 1 × 0%), harvested in SDS-sample buffer, separated by SDS-PAGE, and analyzed by autoradiography.

       Lipid Overlay Assay—

      Protein-lipid overlay assays were performed as previously described (
      • Dowler S.
      • Kular G.
      • Alessi D.R.
      Protein lipid overlay assay..
      ). Briefly, 1 μl of lipid solution containing 6.5 to 200 pmol of phospholipids dissolved in a mixture of chloroform/methanol/water (1:2:0.8) was spotted on to Hybond-C extra membrane and allowed to dry at room temperature for 1 h. The membrane was blocked in 3% (w/v) fatty acid-free BSA in TBST [50 mm Tris/HCl, pH 7.5, 150 mm NaCl, and 0.1% (v/v) Tween 20] for 1 h. The membrane was then incubated overnight at 4 °C with gentle stirring in the same solution containing 5 μg of the indicated GST-fusion protein. The membrane was washed six times for 30 min in TBST, and then incubated for 1 h with 1:1,000 dilution of mouse anti-GST monoclonal antibody (Sigma-Aldrich, St. Louis, MO), washed again as before, then incubated for 1 h with 1:5,000 dilution of anti-mouse-horseradish peroxidase conjugate (Bio-Rad, Hercules, CA). Finally, the membrane was washed 12 times for 1 h in TBST, and membrane-bound GST-fusion protein was detected by ECL (Amersham).

       Interaction Visualization Software—

      Interactions were visualized with Cytoscape (www.cytoscape.org) (
      • Ideker T.
      • Ozier O.
      • Schwikowski B.
      • Siegel A.F.
      Discovering regulatory and signalling circuits in molecular interaction networks..
      ).

       Detailed Protocols—

      More detailed protocols are available at itgmv1.fzk.de/www/itg/uetz/protocols/.

      RESULTS

      Our previous two-hybrid data (
      • Uetz P.
      • Giot L.
      • Cagney G.
      • Mansfield T.A.
      • Judson R.S.
      • Knight J.R.
      • Lockshon D.
      • Narayan V.
      • Srinivasan M.
      • Pochart P.
      • Qureshi-Emili A.
      • Li Y.
      • Godwin B.
      • Conover D.
      • Kalbfleisch T.
      • Vijayadamodar G.
      • Yang M.
      • Johnston M.
      • Fields S.
      • Rothberg J.M.
      A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae..
      ) suggested that PX domains may act as protein interaction domains, although it should be noted that even the yeast “PX domain-only” proteins Grd19p and Ypt35p contain some additional sequences beyond the PX domain itself. To extend our previous findings, we screened all yeast PX domain-containing proteins and their isolated PX domains against a S. cerevisiae whole-genome two-hybrid array that contains most yeast ORFs as Gal4 activation domain fusions (
      • Uetz P.
      • Giot L.
      • Cagney G.
      • Mansfield T.A.
      • Judson R.S.
      • Knight J.R.
      • Lockshon D.
      • Narayan V.
      • Srinivasan M.
      • Pochart P.
      • Qureshi-Emili A.
      • Li Y.
      • Godwin B.
      • Conover D.
      • Kalbfleisch T.
      • Vijayadamodar G.
      • Yang M.
      • Johnston M.
      • Fields S.
      • Rothberg J.M.
      A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae..
      ). Because two-hybrid interactions often suffer from incomplete reproducibility, we reanalyzed all positive interactions in quadruplicate to ensure that none of them had arisen by random activation events. Representative results of these screens are shown in Fig. 1. A summary of all interactions is shown in Fig. 2 and Table II, respectively. We assigned one of three “S” values to the interactions in our datasets in order to reflect various degrees of reproducibility (Table II, see “Experimental Procedures” for details). We found 27 highly reproducible interactions, denoted by S values of 3 in Table II, 18 well-reproducible interactions (S = 2), and 30 interactions that were reproduced several times but not consistently (S = 1). This classification serves as a rough indication for the reliability of our interactions and may be roughly correlated with their interaction strength.
      Figure thumbnail gr1
      Fig. 1Yeast two-hybrid screens and results.A, two-hybrid array screen using Ypt35p as a bait. Shown is only plate 12 (out of 16), which had one positive, Rbd2p (see “Experimental Procedures” for details). B, retest plate with each prey in quadruplicate. Note that genome-wide screens as shown in A were repeated twice but only preys that were positive in both screens were retested to ensure reproducibility (see text for details).
      Figure thumbnail gr2
      Fig. 2The PX domain interaction network in yeast. PX-domain containing proteins (thick border) and isolated PX domains (red arrows) in yeast. Thickness of the arrows represents the reproducibility of the interactions (i.e. scores 1–3 in ). Colors represent the localization of the proteins and the shape their Gene Ontology (GO) terms (see “Experimental Procedures”).
      Table IITwo-hybrid interactions of the yeast PX proteins
      PX proteinPX domain
      No symbol, interaction has been obtained only with the full length protein but note that Grd19p and Ypt35p are essentially PX-only proteins; o, interaction obtained with both full length protein and its PX domain; px, interaction has been obtained only with its PX domain.
      PreyTwo-hybrid quality score
      3, highly reproducible with strong signal; 2, reproducible; 1, reproduced several times but not consistently.
      GO keyword
      Asterisks (*) indicate proteins with multiple functions.
      Ref.
      I, Ito et al. 2001 (51); U, Uetz et al. 2000 (23); S, Seaman et al. 1998 (31).
      Bem1pCdc24p2Cell polarity*I
      Grd19pAdh1p1Fermentation
      Akr2p1Unknown
      Apg17p1Autophagy
      Rbd2p3Unknown
      Rvs167p1Endocytosis*
      Sec15p1Cell polarity*
      Yif1p3ER-Golgi transport
      Yip1p3Vesicular transport
      Yip5p3Vesicular transport
      Ylr031p1Unknown
      Ylr424p1Splicing
      Ypt53p2Vacuolar targeting*
      Mvp1ppxStd1p1Pol II transcript.*
      pxYra2p2mRNA export
      Snx42ppxRrn10p2Pol I transcriptionU
      Snx4p3Vesicular transport*U
      Yip1p1Vesicular transport
      Snx4pApg17p1AutophagyI
      Ask10p2Transcription
      Cog2p3ER to Golgi transport*
      Cog6p3Intra-Golgi transport
      Dam1p1Spindle assembly
      Ecm11p3Cell wall biogenesis
      Lcp5p1rRNA modification*
      Nip100p3Mitotic anaphase B
      Sds3p3Histone deacetylation*
      Sec15p3Cell polarity*
      Sgt1p1Complex assembly*
      Snx41p3Protein transportI
      Snx42p3Vacuolar targeting*
      Swf3p1Mitochondria
      Vma8p1Vacuolar acidification
      Vps17p3Endosome-Golgi tr.
      Ylr031p3Unknown
      Ylr424p1mRNA splicing
      Ynl086p3Unknown
      Ypl201p1Unknown
      Vam7poAkr2p1Unknown
      Avt2p2Transport
      oBsp1p3Actin cytoskeleton*
      Ecm11p2Cell wall biogenesis
      oRbd2p1UnknownU
      Rpt5p2Protein catabolism
      Spc72p2Chromos. segregation*U
      Vps21p1Vacuolar targeting*
      Ygr223p3Unknown
      oYif1p3ER to Golgi transportU
      oYip1p3Vesicular transport
      Yip5p2Vesicular transportU
      Vps17ppxAkr2p1Unknown
      pxBsp1p1Actin cytoskeleton*
      Vps5p3Golgi retention*U
      pxYif1p2ER to Golgi transport
      pxYip1p2Vesicular transport
      pxYpt53p1Vacuolar targeting*
      Vps5pBub1p1Protein phosphoryl.*
      Vps17p2Endosome -> GolgiU
      Vps35p2Golgi retention*S
      Ynl086p1Unknown
      Ypt35pAkr2p3UnknownU
      Atr1p1Multidrug transport
      Bet1p1ER to Golgi transport*
      Bos1p2ER to Golgi transport
      Ktr3p1Cell wall biogenesis*U
      Pep12p1Golgi-vacuole transport
      Rbd2p3UnknownU
      Sps18p1Sporulation
      Tvp23p2Unknown
      Yif1p3ER to Golgi transportU
      Yip1p3Vesicular transport
      Yip4p3Vesicular transport
      Yip5p3Vesicular transportU
      Ymd8p2Nucleotide-sugar transp.
      Yor292p2Unknown
      a No symbol, interaction has been obtained only with the full length protein but note that Grd19p and Ypt35p are essentially PX-only proteins; o, interaction obtained with both full length protein and its PX domain; px, interaction has been obtained only with its PX domain.
      b 3, highly reproducible with strong signal; 2, reproducible; 1, reproduced several times but not consistently.
      c Asterisks (*) indicate proteins with multiple functions.
      d I, Ito et al. 2001 (
      • Ito T.
      • Chiba T.
      • Ozawa R.
      • Yoshida M.
      • Hattori M.
      • Sakaki Y.
      A comprehensive two-hybrid analysis to explore the yeast protein interactome..
      ); U, Uetz et al. 2000 (
      • Uetz P.
      • Giot L.
      • Cagney G.
      • Mansfield T.A.
      • Judson R.S.
      • Knight J.R.
      • Lockshon D.
      • Narayan V.
      • Srinivasan M.
      • Pochart P.
      • Qureshi-Emili A.
      • Li Y.
      • Godwin B.
      • Conover D.
      • Kalbfleisch T.
      • Vijayadamodar G.
      • Yang M.
      • Johnston M.
      • Fields S.
      • Rothberg J.M.
      A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae..
      ); S, Seaman et al. 1998 (
      • Seaman M.N.
      • Williams H.P.
      Identification of the functional domains of yeast sorting nexins Vps5p and Vps17p..
      ).
      Out of the 75 interactions listed in Table II, 17 have been published before. The majority of those, 14, belong to our classes 2 and 3, which confirms their validity.

       Full-length Versus Domain Interactions—

      Of the 15 PX domain proteins in yeast, nine showed a total of 67 two-hybrid interactions when they were screened as full-length proteins (Table II). We did not find any reproducible interactions for five proteins (Bem3p, Mdm1p, Spo14p, Ykr078p, and Ypr097p). Snx41p was not investigated further because it turned out to be a very strong transcriptional activator, even under high 3AT concentrations.
      When we used the isolated PX domains as bait, only those from Mvp1p, Snx42p, Vam7p, and Vps17p showed a total of 13 reproducible interactions (Table II).
      Bem1p, Snx4p, and Vps5p interacted in our assays only as full-length proteins, that is, their PX domain did not interact with any other protein. This indicates that other domains of these baits mediate their interactions. Grd19p and Ypt35p were tested only as full-length proteins as they consist almost exclusively of their PX domains. However, their two-hybrid interactions were confirmed with PX-only constructs in vitro (see below). In one case, Mvp1p, we did not find any interactions for the full-length protein, while its PX domain surprisingly interacted with two nuclear proteins, Std1p and Yra2p.
      Finally, for Snx42p, Vam7p, and Vps17p we found different, but specific, interactions with both the full-length protein and the isolated domain. Interestingly, only Vam7p yielded interactions that were found with both the full-length protein and its PX domain.
      Since we found eight interactions only with the PX domain but not the corresponding full-length protein, we conclude that full-length proteins often prevent interactions, possibly due to their fusion with the Gal4 domain fusion. Whatever the mechanistic reason may be, this indicates that we may have missed a significant number of interactions despite our comprehensive approach. Such false negatives may be prevented by more extensive usage of domains as baits.

       The Nature of PX Protein Interactors—

      As expected from the role of many PX domain-containing proteins such as sorting nexins, the preys that we identified were enriched in proteins known to be involved in vesicular transport (36% of preys as opposed to 5% in the whole genome). Interestingly, a relatively large fraction of interactors were uncharacterized proteins (20 versus 30.5%). All other functional classes combined made up 44% of all preys (whole genome, 64%).
      Similarly, when localization data was analyzed as opposed to functional annotation, the majority (49%) of preys has been localized to the secretory or endocytic pathway, that is, to the endoplasmic reticulum (ER), Golgi, endosomes, or the vacuole. Hence 53% of preys represented membrane or membrane-associated proteins, which was somewhat unexpected because membrane proteins are often said to resist detection in two-hybrid screens. Another unexpected finding was that Snx4p, Snx42p, and Mvp1p all interacted with a significant number of nuclear proteins.
      No single domain or motif was common among our preys. However, some domains were significantly enriched. Four PX domain-containing proteins bind to another PX domain protein, resulting in four pairs: Vps17p-Vps5p, Vps17p-Snx4p, Snx4p-Snx42p, and Snx42p-Snx41p. However, because the PX domain-only baits do not reproduce these interactions, this reflects characteristics of other parts of these sorting nexins—particularly their coiled-coil regions, which have been shown to drive homo- and hetero-oligomerization in other studies. This finding has been confirmed independently in a study of the Vps17p-Vps5p interaction (
      • Seaman M.N.
      • Williams H.P.
      Identification of the functional domains of yeast sorting nexins Vps5p and Vps17p..
      ).
      A notable class of preys consists of the Yip1 family of proteins (
      • Calero M.
      • Winand N.J.
      • Collins R.N.
      Identification of the novel proteins Yip4p and Yip5p as Rab GTPase interacting factors..
      ). Five out of the nine productive PX proteins (Ypt35p, Vam7p, Grd19p, Vps17p, Snx42p) interacted with at least one member of the Yip1 family, and these interactions were defined by the PX domain as assessed using PX domain-only baits and/or by in vitro interaction studies (see below).
      Other domains that were found repeatedly among PX protein interactors were the SNARE domain (in Ypt35p interactions with Bet1p, Bos1p, and Pep12p), the Rab domain (Vps17p-Ypt53p, Grd19p-Ypt53p, Vam7p-Vps21p), and PH domains (Snx4p-Ask10p, Bem1p-Cdc24p). However, none of these interactions seems to involve the PX domain as only the full-length baits but not the PX-only baits detect those preys.

       PX Domains Interact with Members of the Yip1 Family In Vitro—

      Given the surprisingly strong interactions of PX domains with the highly hydrophobic Yip1 family members in our two-hybrid screens, we wanted to verify these results by in vitro binding assays. All yeast PX domains that were positive in our two-hybrid screens were expressed as GST fusion proteins in E. coli and purified on glutathione-Sepharose beads. Note that the Mvp1p and Vam7p PX domains could not be tested because they were either unstable or insoluble. Yip1p, Yip4p, and Yif1p were translated and 35S-labeled in vitro and incubated with the GST-PX fusion proteins. Yip5p could not be tested because its in vitro translation repeatedly failed. Our in vitro binding assays clearly show that Yip1 family members bind specifically to certain PX domains (Fig. 3). The strongest interaction was shown for Ypt35-PX, which interacted strongly with Yif1p, and somewhat less with Yip1p and Yip4p. Vps17-PX showed a similar binding pattern, although the overall binding strength appears to be weaker. Snx4-PX only bound significantly to Yip4p. The PX domains from Snx42p preferred Yip1p over Yif1p, while the Vps5p PX domain appeared to bind most strongly to Yif1p. The interactions of isolated PX domains with Yip1 family members are summarized in Table III.
      Figure thumbnail gr3
      Fig. 3PX domains bind to Yip1 family members in vitro. Various GST-PX domain fusion proteins were tested for Yip1p, Yif1p, and Yip4p binding in vitro. Pictures on top of each panel show Coomassie blue-stained GST fusion proteins, while bottom rows show autoradiography of bound Yip proteins (see “Experimental Procedures” for details). Yip1p and Yif1p bound equally strong to the Ypt35-PX domain, while Yip4p bound only weakly. Yif1p and Yip1p were also bound by Vps17- and Vps5-PX domains and weakly by Snx42-PX domain. Yip4p was additionally bound by the PX domains Snx4 and Vps17. I, input; L, ladder (molecular mass marker).
      Table IIIResults of GST-pulldown experiments
      GSTPreys
      Yip1pYip4pYif1p
      Ypt35-PX+++++
      Vps17-PX+(+)+
      Snx4-PX++
      Mdm1-PX
      Grd19p(–)
      Snx42-PX+(+)
      Vps5-PX(–)(+)
      Bem1-PX
      Bem3-PX

       Comparison of Two-Hybrid and In Vitro Binding Data—

      The in vitro binding data presented in Fig. 3 largely supported our two-hybrid results. However, there were a few notable differences: Yif1p clearly bound to Vps5-PX in vitro, whereas our two-hybrid screens did not identify any Yip1 family interactions with Vps5p as a bait. Similarly, Snx4-PX appeared to interact with Yip4 in vitro, but not in our two-hybrid screen.
      By contrast, Grd19p interacted strongly with Yif1p, Yip1p, and Yip5p in a two-hybrid assay, but binding was barely detectable in vitro (Yip5 was not tested).
      Finally, we observed one two-hybrid interaction, Snx42p-Yip1p, only with the full-length bait (but not the PX-only bait), although Snx42-PX interacted weakly with Yip1p and Yif1p in vitro.

       Ypt35-PX Binds to Yip1 and Yif1 N Termini—

      Yip1 family members are extremely hydrophobic proteins with two to five predicted transmembrane domains (
      • Calero M.
      • Winand N.J.
      • Collins R.N.
      Identification of the novel proteins Yip4p and Yip5p as Rab GTPase interacting factors..
      ). The only significant cytoplasmic part is an N-terminal domain of ∼100–150 amino acids that is conserved among Yip proteins. We wondered if this domain interacts with PX domains. In vitro-translated Yif1p and Yip1p N termini indeed bound to GST-Ypt35-PX (Fig. 4). Binding assays with the remaining C-terminal half also showed some weak binding but were not conclusive (data not shown). Hence it remains unclear if the small cytoplasmic loops in the C termini of Yip1p and Yif1p contribute to PX binding.
      Figure thumbnail gr4
      Fig. 4Yip1p and Yif1p N-terminal fragments bind in vitro to Ypt35-PX. Ypt35-PX domain binds in vitro to the N-terminal domain of Yip1p (1–112 aa) and Yif1p (1–138 aa). I, input; L, ladder (molecular mass marker); FL, full-length protein; N-Term, N-terminal fragment.

       PI Binding Is Not Necessary for Protein-Protein Interactions—

      We next asked whether the PtdIns(3)P binding site also contributes to the protein-protein interactions that we have observed for PX domains. We addressed this possibility by introducing a point mutation (Y123A) into the lipid-binding pocket of Ypt35-PX, which we predicted should impair PtdIns(3)P binding based on results with an analogous mutation in Vam7p, Snx3p, Cvt20p, and Cvt13p (
      • Cheever M.L.
      • Sato T.K.
      • de Beer T.
      • Kutateladze T.G.
      • Emr S.D.
      • Overduin M.
      Phox domain interaction with PtdIns(3)P targets the Vam7 t-SNARE to vacuole membranes..
      ,
      • Xu Y.
      • Hortsman H.
      • Seet L.
      • Wong S.H.
      • Hong W.
      SNX3 regulates endosomal function through its PX-domain-mediated interaction with PtdIns(3)P..
      ,
      • Nice D.C.
      • Sato T.K.
      • Stromhaug P.E.
      • Emr S.D.
      • Klionsky D.J.
      Cooperative binding of the cytoplasm to vacuole targeting pathway proteins, Cvt13 and Cvt20, to phosphatidylinositol 3-phosphate at the pre-autophagosomal structure is required for selective autophagy..
      ). The Y123A mutation did not completely abolish PtdIns(3)P binding by Ypt35p as shown by a protein-lipid overlay assay, although its binding affinity was significantly reduced (Fig. 5A). Importantly, this mutation did not prevent in vitro binding of Ypt35-PX to any of the Yip1 family members (Fig. 5B), although the amount of bound protein appeared to be slightly reduced compared with that seen for wild-type protein. Thus, lipid binding does probably not interfere with protein-protein interactions in PX domains (although the two ligands may cooperate). In addition, this suggests that interacting proteins also use a different interaction surface than lipids or at least do not require the critical tyrosine-123 that is involved in binding the phosphate head group of PIs.
      Figure thumbnail gr5
      Fig. 5Protein-protein interaction is independent of PX domain phospholipid binding.A, the affinity of Ypt35-PX domain for PtdIns(3)P and its mutant Ypt35pY123A-PX were assessed by a protein-lipid overlay assay. PtdIns were spotted onto a nitrocellulose membrane at different concentrations and incubated with 5 μg of GST-fusion protein as well as GST alone (data not shown). Binding of the proteins was detected with anti-GST antibodies. The Ypt35pY123A-PX mutant exhibits significantly reduced binding affinity as no binding was detected at a lipid concentration of 12.5 pmol when compared with the wild-type protein that still bound at a concentration of 6.25 pmol. The overall binding to PtdIns(3)P of the mutant was therefore reduced by about 3-fold. B, mutant GST-Ypt35pY123A with partially abolished PtdIns(3)P binding interacts with similar strength to Yip1p, Yif1p, and Yip4p as wild-type Ypt35-PX. I, input; L, ladder (molecular mass marker).

      DISCUSSION

       PI Binding Domains as Protein Interaction Domains—

      We describe here the first systematic study of PX domain interactions with other proteins. The most striking result of our PX domain two-hybrid screens was the repeated appearance of Yip1 family proteins (Table II). Yif1p, Yip1p, and Yip5p were consistently found with the PX domain-only proteins Grd19p and Ypt35p as baits and with Vam7p (PX domain alone and full-length protein). Yip1p and Yif1p were additionally found with the PX domain from Vps17p.
      Yif1p and Yip1p have been reported to be localized to the Golgi (
      • Yang X.
      • Matern H.T.
      • Gallwitz D.
      Specific binding to a novel and essential Golgi membrane protein (Yip1p) functionally links the transport GTPases Ypt1p and Ypt31p..
      ,
      • Matern H.
      • Yang X.
      • Andrulis E.
      • Sternglanz R.
      • Trepte H.H.
      • Gallwitz D.
      A novel Golgi membrane protein is part of a GTPase-binding protein complex involved in vesicle targeting..
      ), and the two proteins have been postulated to be part of a complex (
      • Matern H.
      • Yang X.
      • Andrulis E.
      • Sternglanz R.
      • Trepte H.H.
      • Gallwitz D.
      A novel Golgi membrane protein is part of a GTPase-binding protein complex involved in vesicle targeting..
      ). Recently, several studies showed that Yip1p/Yif1p proteins are necessary for COPII vesicle budding from the ER membrane (
      • Heidtman M.
      • Chen C.Z.
      • Collins R.N.
      • Barlowe C.
      A role for Yip1p in COPII vesicle biogenesis..
      ) and for membrane fusion with the Golgi (
      • Barrowman J.
      • Wang W.
      • Zhang Y.
      • Ferro-Novick S.
      The Yip1p.Yif1p complex is required for the fusion competence of endoplasmic reticulum-derived vesicles..
      ). Yip1p is thought to cycle between the ER and Golgi compartments (
      • Heidtman M.
      • Chen C.Z.
      • Collins R.N.
      • Barlowe C.
      A role for Yip1p in COPII vesicle biogenesis..
      ).
      It was therefore surprising that the strongest and most consistent PX domain-containing interactors of Yif1p/Yip1p were proteins that have been localized to the endocytic pathway. For example, Ypt35p is mainly found associated with endosomes (
      • Huh W.K.
      • Falvo J.V.
      • Gerke L.C.
      • Carroll A.S.
      • Howson R.W.
      • Weissman J.S.
      • O’Shea E.K.
      Global analysis of protein localization in budding yeast..
      ), and Grd19p is localized to both the vacuolar membrane (
      • Huh W.K.
      • Falvo J.V.
      • Gerke L.C.
      • Carroll A.S.
      • Howson R.W.
      • Weissman J.S.
      • O’Shea E.K.
      Global analysis of protein localization in budding yeast..
      ) and the cytosol (
      • Voos W.
      • Stevens T.H.
      Retrieval of resident late-Golgi membrane proteins from the prevacuolar compartment of Saccharomyces cerevisiae is dependent on the function of Grd19p..
      ), although it becomes associated with the prevacuolar compartment in Δvps27 null mutant cells (
      • Voos W.
      • Stevens T.H.
      Retrieval of resident late-Golgi membrane proteins from the prevacuolar compartment of Saccharomyces cerevisiae is dependent on the function of Grd19p..
      ). Vps17p also localizes to the endosome, as reported by a C-terminally tagged GFP fusion protein (
      • Huh W.K.
      • Falvo J.V.
      • Gerke L.C.
      • Carroll A.S.
      • Howson R.W.
      • Weissman J.S.
      • O’Shea E.K.
      Global analysis of protein localization in budding yeast..
      ) as well as to the prevacuolar compartment (
      • Burda P.
      • Padilla S.M.
      • Sarkar S.
      • Emr S.D.
      Retromer function in endosome-to-Golgi retrograde transport is regulated by the yeast Vps34 PtdIns 3-kinase..
      ). So do Snx42p and Snx4p, which have additionally been localized to the post-Golgi endosomes and the pre-autophagosomes (
      • Hettema E.H.
      • Lewis M.J.
      • Black M.W.
      • Pelham H.R.
      Retromer and the sorting nexins Snx4/41/42 mediate distinct retrieval pathways from yeast endosomes..
      ). Snx42p is also found at a perivacuolar structure (
      • Nice D.C.
      • Sato T.K.
      • Stromhaug P.E.
      • Emr S.D.
      • Klionsky D.J.
      Cooperative binding of the cytoplasm to vacuole targeting pathway proteins, Cvt13 and Cvt20, to phosphatidylinositol 3-phosphate at the pre-autophagosomal structure is required for selective autophagy..
      ). And finally, Vam7p is usually seen in association with vacuolar membranes (
      • Boeddinghaus C.
      • Merz A.J.
      • Laage R.
      • Ungermann C.
      A cycle of Vam7p release from and PtdIns 3-P-de pend ent rebinding to the yeast vacuole is required for homotypic vacuole fusion..
      ,
      • Sato T.K.
      • Darsow T.
      • Emr S.D.
      Vam7p, a SNAP-25-like molecule, and Vam3p, a syntaxin homolog, function together in yeast vacuolar protein trafficking..
      ).
      There are three possible explanations for our finding that a group of ER/Golgi proteins (the Yip1 family) interact with endosomal/vacuolar proteins (the PX domain-containing proteins). First, a significant fraction of at least some PX proteins can be found in the cytoplasm, and could be recruited in small amounts to the ER membrane and act there during COPII vesicle formation. Although there is no direct evidence for this, while 5% of Vam7p associates with vacuolar membranes, the rest is soluble (
      • Sato T.K.
      • Darsow T.
      • Emr S.D.
      Vam7p, a SNAP-25-like molecule, and Vam3p, a syntaxin homolog, function together in yeast vacuolar protein trafficking..
      ). Soluble and therefore diffusely stained cytoplasmic fractions of PX proteins could easily escape detection if there is at least one other membrane-related localization. In addition, sites of COPII vesicle formation do not seem to be visible easily by light microscopy and neither are the proteins involved in such highly dynamic membrane spots. Second, it remains likely that Yif1p/Yip1p proteins have other, yet unnoticed functions (independent of COPII vesicle formation) in post-Golgi sorting or endocytosis. This idea is supported by the requirement for the Yip1p/Yif1p-binding Rab GTPases in docking and fusion of transport vesicles, but not for budding (
      • Heidtman M.
      • Chen C.Z.
      • Collins R.N.
      • Barlowe C.
      A role for Yip1p in COPII vesicle biogenesis..
      ). Furthermore, Sivars et al. 2003 (
      • Sivars U.
      • Aivazian D.
      • Pfeffer S.R.
      Yip3 catalyses the dissociation of endosomal Rab-GDI complexes..
      ) showed that Yip3, a mammalian Yip family member, can act as a GDI-displacement factor for the targeting of Rab GTPases, particularly Rab9. It may be interesting to determine the effect of PX proteins on the recruitment of GDFs, their activity, or some downstream event such as the recruitment of GEFs. Third and last, we cannot completely rule out the possibility that at least some of our interactions are experimental artifacts with no physiological relevance. However, the specificity, apparent strengths of the interactions, and confirmation by independent methods argue against this possibility.
      Interestingly, Yif1p has been shown to localize to the vacuole under certain circumstances, for example in a Btn2 null mutant, which does not appear to disrupt vesicular transport per se (
      • Chattopadhyay S.
      • Roberts P.M.
      • Pearce D.A.
      The yeast model for Batten disease: A role for Btn2p in the trafficking of the Golgi-associated vesicular targeting protein, Yif1p..
      ). In addition, Yif1p was localized to the cytoplasm and may be free to diffuse to locations other than the ER (
      • Kumar A.
      • Agarwal S.
      • Heyman J.A.
      • Matson S.
      • Heidtman M.
      • Piccirillo S.
      • Umansky L.
      • Drawid A.
      • Jansen R.
      • Liu Y.
      • Cheung K.H.
      • Miller P.
      • Gerstein M.
      • Roeder G.S.
      • Snyder M.
      Subcellular localization of the yeast proteome..
      ).
      Even if Yif1p and Yip1p do not have a role in endocytosis or post-Golgi sorting, such a function may have been taken over by Yip4p and Yip5p as their localization remains unclear. Yip5p localizes to the cytoplasm, as determined for a C-terminally tagged GFP fusion protein (
      • Huh W.K.
      • Falvo J.V.
      • Gerke L.C.
      • Carroll A.S.
      • Howson R.W.
      • Weissman J.S.
      • O’Shea E.K.
      Global analysis of protein localization in budding yeast..
      ), while Yip4p does not appear to be localized in any study.

       Nuclear PX Interactions—

      We were surprised to find that several PX domain-containing proteins showed preferential interactions with nuclear proteins. Although this was initially unexpected, it is clear that there are several relationships of inositol phosphates to nuclear functions (reviewed in Ref.
      • Jones D.E.
      • Divecha N.
      Linking lipids to chromatin..
      ). Mvp1p, a protein required for sorting proteins to the vacuole, interacted weakly with Std1p, a suppressor of a TBP deletion and thus likely a transcription factor (although its biochemical function is unknown). Mvp1p also interacted with Yra2p, a protein that is involved in mRNA export from the nucleus. Although the significance of both interactions is unclear, it is notable that both of these Mvp1p interactors have been found both in the nucleus as well as in the cytoplasm. Std1p has also been localized to the plasma membrane (
      • Schmidt M.C.
      • McCartney R.R.
      • Zhang X.
      • Tillman T.S.
      • Solimeo H.
      • Wolfl S.
      • Almonte C.
      • Watkins S.C.
      Std1 and Mth1 proteins interact with the glucose sensors to control glucose-regulated gene expression in Saccharomyces cerevisiae..
      ). Vps5p and Snx42p also have interactions with nuclear proteins of which Snx42p’s interaction with Rrn10p is remarkable because it appears to be specific for the PX domain of Snx42p. Additional support for an Snx42p-Rrn10p interaction comes from the observation that Rrn10 null mutants are hypersensitive to wortmannin, an inhibitor of PI3 kinase (
      • Zewail A.
      • Xie M.W.
      • Xing Y.
      • Lin L.
      • Zhang P.F.
      • Zou W.
      • Saxe J.P.
      • Huang J.
      Novel functions of the phosphatidylinositol metabolic pathway discovered by a chemical genomics screen with wortmannin..
      ).
      The most remarkable example of nuclear protein interactions was seen with Snx4p, which has seven interactions with nuclear proteins. Two of these were highly reproducible, namely with Ecm11p and Sds3p. Ecm11p is an unusual protein in that its mutants exhibit phenotypes related to extracellular matrix functions such as zymolyase hypersensitivity. Nevertheless, the protein has been localized to the nucleus. Its molecular function remains obscure. The other nuclear Snx4p interactor, Sds3p, is a suppressor of defective silencing, i.e. it is involved in silencing the HMR mating cassette locus. Its precise biochemical activity is also unknown but, interestingly, it binds phosphoinositol 4-phosphate in vitro (
      • Zhu H.
      • Bilgin M.
      • Bangham R.
      • Hall D.
      • Casamayor A.
      • Bertone P.
      • Lan N.
      • Jansen R.
      • Bidlingmaier S.
      • Houfek T.
      • Mitchell T.
      • Miller P.
      • Dean R.A.
      • Gerstein M.
      • Snyder M.
      Global analysis of protein activities using proteome chips..
      ). Rrn10 and Sds3 are two of 38 genes whose deletion results in hypersensitivity to wortmannin but the mode of action remains to be determined.
      In this article, we have shown that PX domains can interact with a number of different proteins. Most striking was their preference for membrane proteins of the Yip1 family. Given the much greater diversity of both Yip and PX proteins in vertebrate genomes, we hope that our findings will encourage further studies on these proteins and so improve our understanding of both their diversity, specificity, and, eventually, biological function.

      Acknowledgments

      We would like to thank Tanja Kuhn for excellent technical assistance, Sebastian Heucke for helping with two-hybrid assays, Jong W. Yu and Mark A. Lemmon for PX domain GST fusion constructs and very helpful comments on the manuscript, and Xiaoping Yang for stimulating discussions and for a Yip1 mutant strain. Massimiliano Mazza and Michael Knop kindly helped with lipid overlay experiments. Elizabeth Conibear provided the Mvp1, Grd19, Spo14, and Vps5 bait constructs. Ben Chun initially helped with constructing and sequencing some PX baits. Several PX screens have been carried out first in the laboratory of Stan Fields.

      Supplementary Material

      REFERENCES

        • Simonsen A.
        • Wurmser A.E.
        • Emr S.D.
        • Stenmark H.
        The role of phosphoinositides in membrane transport..
        Curr. Opin. Cell Biol. 2001; 13: 485-492
        • Toker A.
        Phosphoinositides and signal transduction..
        Cell Mol. Life Sci. 2002; 59: 761-779
        • Gruenberg J.
        Lipids in endocytic membrane transport and sorting..
        Curr. Opin. Cell Biol. 2003; 15: 382-388
        • Hurley J.H.
        • Meyer T.
        Subcellular targeting by membrane lipids..
        Curr. Opin. Cell Biol. 2001; 13: 146-152
        • Itoh T.
        • Takenawa T.
        Phosphoinositide-binding domains: Functional units for temporal and spatial regulation of intracellular signalling..
        Cell Signal. 2002; 14: 733-743
        • Lemmon M.A.
        Phosphoinositide recognition domains..
        Traffic. 2003; 4: 201-213
        • DiNitto J.P.
        • Cronin T.C.
        • Lambright D.G.
        Membrane recognition and targeting by lipid-binding domains..
        Sci. STKE. 2003; 213: re16
        • Ponting C.P.
        Novel domains in NADPH oxidase subunits, sorting nexins, and PtdIns 3-kinases: Binding partners of SH3 domains?.
        Protein Sci. 1996; 5: 2353-2357
        • Letunic I.
        • Copley R.R.
        • Schmidt S.
        • Ciccarelli F.D.
        • Doerks T.
        • Schultz J.
        • Ponting C.P.
        • Bork P.
        SMART 4.0: Towards genomic data integration..
        Nucleic Acids Res. 2004; 32: 142-D144
        • Odorizzi G.
        • Babst M.
        • Emr S.D.
        Phosphoinositide signaling and the regulation of membrane trafficking in yeast..
        Trends Biochem. Sci. 2000; 25: 229-235
        • Worby C.A.
        • Dixon J.E.
        Sorting out the cellular functions of sorting nexins..
        Nat. Rev. Mol. Cell. Biol. 2002; 3: 919-931
        • Bravo J.
        • Karathanassis D.
        • Pacold C.M.
        • Pacold M.E.
        • Ellson C.D.
        • Anderson K.E.
        • Butler P.J.
        • Lavenir I.
        • Perisic O.
        • Hawkins P.T.
        • Stephens L.
        • Williams R.L.
        The crystal structure of the PX domain from p40(phox) bound to phosphatidylinositol 3-phosphate..
        Mol. Cell. 2001; 8: 829-839
        • Hiroaki H.
        • Ago T.
        • Ito T.
        • Sumimoto H.
        • Kohda D.
        Solution structure of the PX domain, a target of the SH3 domain..
        Nat. Struct. Biol. 2001; 8: 526-530
        • Lu J.
        • Garcia J.
        • Dulubova I.
        • Sudhof T.C.
        • Rizo J.
        Solution structure of the Vam7p PX domain..
        Biochemistry. 2002; 41: 5956-5962
        • Zhou C.Z.
        • de La Sierra-Gallay I.L.
        • Quevillon-Cheruel S.
        • Collinet B.
        • Minard P.
        • Blondeau K.
        • Henckes G.
        • Aufrere R.
        • Leulliot N.
        • Graille M.
        • Sorel I.
        • Savarin P.
        • de la Torre F.
        • Poupon A.
        • Janin J.
        • van Tilbeurgh H.
        Crystal structure of the yeast Phox homology (PX) domain protein Grd19p complexed to phosphatidylinositol-3-phosphate..
        J. Biol. Chem. 2003; 278: 50371-50376
        • Karathanassis D.
        • Stahelin R.V.
        • Bravo J.
        • Perisic O.
        • Pacold C.M.
        • Cho W.
        • Williams R.L.
        Binding of the PX domain of p47(phox) to phosphatidylinositol 3,4-bisphosphate and phosphatidic acid is masked by an intramolecular interaction..
        EMBO J. 2002; 21: 5057-5068
        • Xu Y.
        • Seet L.F.
        • Hanson B.
        • Hong W.
        The Phox homology (PX) domain, a new player in phosphoinositide signalling..
        Biochem. J. 2001; 360: 513-530
        • Yu J.W.
        • Lemmon M.A.
        All phox homology (PX) domains from Saccharomyces cerevisiae specifically recognize phosphatidylinositol 3-phosphate..
        J. Biol. Chem. 2001; 276: 44179-44184
        • Voos W.
        • Stevens T.H.
        Retrieval of resident late-Golgi membrane proteins from the prevacuolar compartment of Saccharomyces cerevisiae is dependent on the function of Grd19p..
        J. Cell Biol. 1998; 140: 577-590
        • Hettema E.H.
        • Lewis M.J.
        • Black M.W.
        • Pelham H.R.
        Retromer and the sorting nexins Snx4/41/42 mediate distinct retrieval pathways from yeast endosomes..
        EMBO J. 2003; 22: 548-557
        • Sumimoto H.
        • Kage Y.
        • Nunoi H.
        • Sasaki H.
        • Nose T.
        • Fukumaki Y.
        • Ohno M.
        • Minakami S.
        • Takeshige K.
        Role of Src homology 3 domains in assembly and activation of the phagocyte NADPH oxidase..
        Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 5345-5349
        • Leto T.L.
        • Adams A.G.
        • de Mendez I.
        Assembly of the phagocyte NADPH oxidase: Binding of Src homology 3 domains to proline-rich targets..
        Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 10650-10654
        • Uetz P.
        • Giot L.
        • Cagney G.
        • Mansfield T.A.
        • Judson R.S.
        • Knight J.R.
        • Lockshon D.
        • Narayan V.
        • Srinivasan M.
        • Pochart P.
        • Qureshi-Emili A.
        • Li Y.
        • Godwin B.
        • Conover D.
        • Kalbfleisch T.
        • Vijayadamodar G.
        • Yang M.
        • Johnston M.
        • Fields S.
        • Rothberg J.M.
        A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae..
        Nature. 2000; 403: 623-627
        • Cagney G.
        • Uetz P.
        • Fields S.
        High-throughput screening for protein-protein interactions using two-hybrid assay..
        Methods Enzymol. 2000; 328: 3-14
        • James P.
        • Halladay J.
        • Craig E.A.
        Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast..
        Genetics. 1996; 144: 1425-1436
        • Ma H.
        • Kunes S.
        • Schatz P.J.
        • Botstein D.
        Plasmid construction by homologous recombination in yeast..
        Gene. 1987; 58: 201-216
        • Ito H.
        • Fukuda Y.
        • Murata K.
        • Kimura A.
        Transformation of intact yeast cells treated with alkali cations..
        J. Bacteriol. 1983; 153: 163-168
        • Ron D.
        • Dressler H.
        pGSTag—A versatile bacterial expression plasmid for enzymatic labeling of recombinant proteins..
        BioTechniques. 1992; 13: 866-869
        • Dowler S.
        • Kular G.
        • Alessi D.R.
        Protein lipid overlay assay..
        Sci. STKE. 2002; 129: PL6
        • Ideker T.
        • Ozier O.
        • Schwikowski B.
        • Siegel A.F.
        Discovering regulatory and signalling circuits in molecular interaction networks..
        Bioinformatics. 2002; 18: S233-S240
        • Seaman M.N.
        • Williams H.P.
        Identification of the functional domains of yeast sorting nexins Vps5p and Vps17p..
        Mol. Biol. Cell. 2002; 13: 2826-2840
        • Calero M.
        • Winand N.J.
        • Collins R.N.
        Identification of the novel proteins Yip4p and Yip5p as Rab GTPase interacting factors..
        FEBS Lett. 2002; 515: 89-98
        • Cheever M.L.
        • Sato T.K.
        • de Beer T.
        • Kutateladze T.G.
        • Emr S.D.
        • Overduin M.
        Phox domain interaction with PtdIns(3)P targets the Vam7 t-SNARE to vacuole membranes..
        Nat. Cell Biol. 2001; 3: 613-618
        • Xu Y.
        • Hortsman H.
        • Seet L.
        • Wong S.H.
        • Hong W.
        SNX3 regulates endosomal function through its PX-domain-mediated interaction with PtdIns(3)P..
        Nat. Cell Biol. 2001; 3: 658-666
        • Nice D.C.
        • Sato T.K.
        • Stromhaug P.E.
        • Emr S.D.
        • Klionsky D.J.
        Cooperative binding of the cytoplasm to vacuole targeting pathway proteins, Cvt13 and Cvt20, to phosphatidylinositol 3-phosphate at the pre-autophagosomal structure is required for selective autophagy..
        J. Biol. Chem. 2002; 277: 30198-30207
        • Yang X.
        • Matern H.T.
        • Gallwitz D.
        Specific binding to a novel and essential Golgi membrane protein (Yip1p) functionally links the transport GTPases Ypt1p and Ypt31p..
        EMBO J. 1998; 17: 4954-4963
        • Matern H.
        • Yang X.
        • Andrulis E.
        • Sternglanz R.
        • Trepte H.H.
        • Gallwitz D.
        A novel Golgi membrane protein is part of a GTPase-binding protein complex involved in vesicle targeting..
        EMBO J. 2000; 19: 4485-4492
        • Heidtman M.
        • Chen C.Z.
        • Collins R.N.
        • Barlowe C.
        A role for Yip1p in COPII vesicle biogenesis..
        J. Cell Biol. 2003; 163: 57-69
        • Barrowman J.
        • Wang W.
        • Zhang Y.
        • Ferro-Novick S.
        The Yip1p.Yif1p complex is required for the fusion competence of endoplasmic reticulum-derived vesicles..
        J. Biol. Chem. 2003; 278: 19878-19884
        • Huh W.K.
        • Falvo J.V.
        • Gerke L.C.
        • Carroll A.S.
        • Howson R.W.
        • Weissman J.S.
        • O’Shea E.K.
        Global analysis of protein localization in budding yeast..
        Nature. 2003; 425: 686-691
        • Burda P.
        • Padilla S.M.
        • Sarkar S.
        • Emr S.D.
        Retromer function in endosome-to-Golgi retrograde transport is regulated by the yeast Vps34 PtdIns 3-kinase..
        J. Cell Sci. 2002; 115: 3889-3900
        • Boeddinghaus C.
        • Merz A.J.
        • Laage R.
        • Ungermann C.
        A cycle of Vam7p release from and PtdIns 3-P-de pend ent rebinding to the yeast vacuole is required for homotypic vacuole fusion..
        J. Cell Biol. 2002; 157: 79-89
        • Sato T.K.
        • Darsow T.
        • Emr S.D.
        Vam7p, a SNAP-25-like molecule, and Vam3p, a syntaxin homolog, function together in yeast vacuolar protein trafficking..
        Mol. Cell. Biol. 1998; 18: 5308-5319
        • Sivars U.
        • Aivazian D.
        • Pfeffer S.R.
        Yip3 catalyses the dissociation of endosomal Rab-GDI complexes..
        Nature. 2003; 425: 856-859
        • Chattopadhyay S.
        • Roberts P.M.
        • Pearce D.A.
        The yeast model for Batten disease: A role for Btn2p in the trafficking of the Golgi-associated vesicular targeting protein, Yif1p..
        Biochem. Biophys. Res. Commun. 2003; 302: 534-538
        • Kumar A.
        • Agarwal S.
        • Heyman J.A.
        • Matson S.
        • Heidtman M.
        • Piccirillo S.
        • Umansky L.
        • Drawid A.
        • Jansen R.
        • Liu Y.
        • Cheung K.H.
        • Miller P.
        • Gerstein M.
        • Roeder G.S.
        • Snyder M.
        Subcellular localization of the yeast proteome..
        Genes Dev. 2002; 16: 707-719
        • Jones D.E.
        • Divecha N.
        Linking lipids to chromatin..
        Curr. Opin. Genes Dev. 2004; 14: 1-7
        • Schmidt M.C.
        • McCartney R.R.
        • Zhang X.
        • Tillman T.S.
        • Solimeo H.
        • Wolfl S.
        • Almonte C.
        • Watkins S.C.
        Std1 and Mth1 proteins interact with the glucose sensors to control glucose-regulated gene expression in Saccharomyces cerevisiae..
        Mol. Cell. Biol. 1999; 19: 4561-4571
        • Zewail A.
        • Xie M.W.
        • Xing Y.
        • Lin L.
        • Zhang P.F.
        • Zou W.
        • Saxe J.P.
        • Huang J.
        Novel functions of the phosphatidylinositol metabolic pathway discovered by a chemical genomics screen with wortmannin..
        Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 3345-3350
        • Zhu H.
        • Bilgin M.
        • Bangham R.
        • Hall D.
        • Casamayor A.
        • Bertone P.
        • Lan N.
        • Jansen R.
        • Bidlingmaier S.
        • Houfek T.
        • Mitchell T.
        • Miller P.
        • Dean R.A.
        • Gerstein M.
        • Snyder M.
        Global analysis of protein activities using proteome chips..
        Science. 2001; 293: 2101-2105
        • Ito T.
        • Chiba T.
        • Ozawa R.
        • Yoshida M.
        • Hattori M.
        • Sakaki Y.
        A comprehensive two-hybrid analysis to explore the yeast protein interactome..
        Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 4569-4574