Interactions of phocein with nucleoside-diphosphate kinase, Eps15, and Dynamin I.

Phocein, an intracellular protein interacting with striatin, bears a few homologies with the sigma-subunits of clathrin adaptor proteins (Baillat, G., Moqrich, A., Castets, F., Baude, A., Bailly, Y., Benmerah, A., and Monneron, A. (2001) Mol. Biol. Cell 12, 663-673). Using phocein as a bait in a yeast two-hybrid screen, we identified two novel interacting proteins, nucleoside-diphosphate kinase (NDPK) and Eps15. Immunoprecipitation and pull-down experiments involving native and/or recombinant phocein and, respectively, NDPK and Eps15, biochemically validated their interactions. NDPK and Eps15 were recently shown to be functional neighbors of dynamin. Dynamin I is shown here to directly interact with NDPK through its C-terminal proline-rich domain, whereas recombinant phocein associates with native dynamin I. Immunocytochemical studies of rat embryonic hippocampal neurons demonstrated partial co-localization of phocein and dynamin I. Phocein thus appears to be a component of the complexes involved in some steps of the vesicular traffic machinery.

Phocein, highly conserved throughout the animal kingdom, is a 26-kDa intracellular protein expressed in multiple tissues (1). The sequence of phocein contains within its N-and Cterminal regions several short stretches homologous to the -subunits of clathrin adaptor complexes, suggesting a role in vesicular traffic. Subcellular fractionation of HeLa cells and rat brain showed that phocein partitions between the cytosol and the detergent-soluble membrane fractions. In unpolarized cells, phocein is prominent in the Golgi area, whereas in mature neurons, it is found in the perikaryal-dendritic region (1).
Phocein is a direct partner of the members of the striatin family (1). This family includes striatin, SG2NA, and zinedin, which are multimodular, WD repeat, and calmodulin-binding proteins thought to act both as scaffolds and as signaling proteins (2)(3)(4)(5). In the adult, such proteins are mostly expressed in neurons, where they localize to the somato-dendritic compartments. Phocein co-localizes with striatin and/or SG2NA (1,6).
To further investigate the function of phocein, we searched for phocein partners using the two-hybrid screen in yeast. Several partners of phocein were found, including: the ␤-subunit of nucleoside-diphosphate kinase (NDPK, 1 EC 2.7.4.6) and Eps15. NDPKs are ubiquitous enzymes that exchange ␥-phosphates between nucleoside tri-and diphosphates (7). Eps15 is a multidomain protein involved in clathrin-mediated endocytosis (8,9). Recently, NDPK and Eps15 have been shown by genetic studies to be functional neighbors of dynamin, a GTPase that plays a critical role in endocytosis; NDPK mutations in Drosophila, as well as Eps15 mutations in Caenorhabditis elegans, enhance the phenotypes of dynamin mutations (10,11). Altogether, these findings strengthen the hypothesis that phocein participates in membrane traffic and more specifically in membrane budding reactions.

EXPERIMENTAL PROCEDURES
Two-hybrid Assay-Phocein fused to the LexA DNA-binding domain was used as a bait to search for fusion proteins expressed by a rat brain cDNA library encoding the activation domain of Gal4 (plasmid pGAD10, Matchmaker, CLONTECH, Palo Alto, CA) (12). A full-length phocein insert (1) was ligated into the pLex-11 vector (a gift from M.C. Dagher, CEA Grenoble, France) in-frame with the LexA DNA-binding domain, yielding plasmid pLex-pho. L40 yeast strain cells grown in minimal medium were transformed with pLex-pho, using the lithium acetate method (13). The Lex-pho fusion protein was stably expressed in L40 cells, as verified by immunoblotting using anti-phocein antibodies. L40 cells expressing Lex-pho were transformed with the plasmid library. From 8 ϫ 10 6 colonies obtained 5 days after cotransfection, 100 were His ϩ . They were tested for ␤-galactosidase activity by a color filter assay using the substrate 5-bromo-4-chloro-3-indolyl-D-galactoside (Xgal). Plasmids from the 68 His ϩ /LacZ ϩ colonies were prepared according to Kimmel and Berger (14). After electroporation in Escherichia coli HB101 cells of Leu Ϫ phenotype, the selected library plasmids were rescued by complementing the Leu Ϫ phenotype on minimal medium and subjected to restriction analysis. Selected inserts were sequenced (ESGS, Paris, France).
Antibodies-Anti-phocein antibodies have been described (1). Antidynamin goat antibodies (sc-6402, Santa Cruz Biotechnology, Santa Cruz, CA) were directed against the proline-rich, C-terminal domain (PRD) of human dynamin I. A rabbit anti-NDPK A antibody (sc-343) was from Santa Cruz Biotechnology; it cross-reacts with the ␤-subunit of NDPK. A goat anti-Eps15 antibody (sc-11716) was from Santa Cruz Biotechnology, and a monoclonal anti-Eps15 antibody was a gift from P. Di Fiore. They recognize at least two proteins of 135 and 120 kDa, which are believed to be diversely phosphorylated species of Eps15. For immunocytochemistry, the monoclonal anti-dynamin Hudy-1 antibody was used (Upstate Biotechnology, Lake Placid, NY). Fluorescent secondary antibodies used for confocal microscopy were Alexa-conjugated antibodies from Molecular Probes (Eugene, Oregon).
Brain Fractionation-Adult Wistar rats were deeply anesthetized using a mixture of 0.5 ml of ketamine (50 mg/ml, Rhône-Mérieux) and 0.37 ml of xylazine (2 mg/kg, Bayer). Brain homogenates were prepared * This work was supported by Center National de la Recherche Scientifique and by Grant ARC 9318 from the Association pour la Recherche sur le Cancer (to F. C. and A. M.) and Grant AFM FRN 210/7961 from the Association Française contre les Myopathies. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  in TBS buffer (Tris-HCl 50 mM, pH 7.4, NaCl 150 mM containing inhibitors of proteases, 10-ml final volume for one brain). Cytosol was obtained by centrifuging the homogenate for 1 h at 100,000 ϫ g, 4°C.
The pellet was homogenized in TBS-CHAPS buffer (TBS buffer containing 7.5 mM CHAPS and 0.5 M NaCl) and centrifuged for 1 h at 100,000 ϫ g. The supernatant was referred to as the CHAPS fraction. Protein was determined using the Schaffner and Weissmann assay (15).
Coimmunoprecipitation Assays-There was no need to overexpress the interacting proteins since they were represented in sufficient amounts in brain extracts. For coimmunoprecipitation assays, 300 -600 l of freshly obtained brain cytosol or CHAPS fraction (250 mM NaCl final concentration) were incubated with 40 g of affinity-purified rabbit anti-phocein antibodies or 40 g of affinity-purified rabbit anti-NDPK antibodies or 40 g of rabbit preimmune immunoglobulins (Sigma) overnight at either 4 or 30°C with gentle agitation. Batches of 40 l of Dynabeads (Dynal, Oslo, Norway) washed in TBS containing 0.1% bovine serum albumin were added to the samples and further incubated at the appropriate temperature for 4 h. Dynabeads were washed several times in the appropriate TBS or TBS-CHAPS buffers and boiled for 5 min in Laemmli sample buffer. The solubilized proteins were analyzed on 8 and 15% SDS-polyacrylamide gels and transferred onto Protran membranes (Schleicher and Schuell, Dassel, Germany).
The antibodies used to reveal Western blots were described above and diluted to 0.2-1 g/ml. Bound antibodies were detected using the ECL procedure (Pierce) or the phosphatase-alkaline procedure (Promega, Madison, WI).
In some cases, the amount of the relevant protein (phocein and Eps15) in the immunoprecipitates (respectively obtained with anti-NDPK and anti-phocein antibodies) was quantified and expressed as the percent of the protein present in the samples of cytosol prior to incubation with the antibody (three separate experiments). Precise quantification was achieved by densitometric analyses of the immunoreactive bands using the NIH Image 1.59 software.
Pull-down Assays-Recombinant rat NDPK was obtained as follows. A 786-bp NcoI-EcoRI fragment of a plasmid selected from the yeast library, pGAD-NDPK (see below), was subcloned in a modified pGEX-KT vector (Amersham Biosciences) carrying the PreScission Protease site and yielded plasmid pGEX-P-NDPK. This plasmid encodes the full-length NDPK sequence in-frame with glutathione S-transferase (GST). E. coli DH5 ␣ cells were transformed and, upon induction by 0.1 mM isopropyl-␤-D-thiogalactoside, expressed high levels of GST-P-NDPK (43 kDa). The cells were lysed, and the fusion protein contained in the soluble fraction was bound to glutathione-Sepharose 4B (Amersham Biosciences). Recombinant phocein was obtained by subcloning the phocein cDNA in the modified pGEX-KT vector, as described above. pGEX-6P-PRD was obtained by inserting the PRD fragment of dynamin I into pGEX-6P (16). GST-P-NDPK, GST-P-phocein, and GST-P-PRD were eluted from the glutathione-Sepharose resin using reduced glutathione and used as such. Alternatively, NDPK, phocein, and PRD were cleaved from GST using PreScission protease (Amersham Biosciences), as indicated by the supplier. The polyhistidine human dynamin I deletion construct missing the PRD domain (His-dynamin ⌬ PRD), as well as GST-amphiphysin construct (amino acids 545-695), were described (16). PGEX-5.1 vectors encoding domains I, II, and III of Eps15 were kindly provided by A. Benmerah (8).
From 4 to 6 g of GST-phocein (1), GST-P-NDPK, GST-P-PRD, or GST were incubated with 40 l of 80% glutathione-Sepharose for 2 h at 4°C in phosphate-saline buffer (PBS). After three washes in PBS containing 0.1% bovine serum albumin, beads were incubated with gentle agitation overnight at either 4 or 30°C with 300 l of rat brain cytosol or 300 l of CHAPS fraction or with 150 -300 l of 5 M solutions of recombinant NDPK or dynamin I PRD fragment or phocein. After extensive washes with PBS containing 7.5 mM CHAPS and 0.1% bovine serum albumin, the beads were boiled in Laemmli sample buffer and treated as above.
Cell Culture, Immunofluorescence, and Confocal Microscopy-Cultures of primary E18 hippocampal neurons were prepared as described (17). Neurons at stages 3 and 5 were fixed for 20 min in 4% paraformaldehyde at room temperature, washed in PBS, and incubated in PBS containing 0.1% Triton X-100 and 10% normal goat serum (NGS) for 15 min at room temperature. After several washes, cells were blocked with PBS containing 10% NGS (PBS/NGS) for 2 h at room temperature and then incubated in PBS/NGS containing 10 g/ml of either the monoclonal anti-dynamin Hudy-1 or mouse preimmune immunoglobulins (Sigma) and 5 g/ml of either rabbit anti-phocein antibodies or rabbit preimmune antibodies (Sigma) for 1 h at room temperature. After several washes in PBS, hippocampal neurons were incubated in PBS/ NGS containing Alexa 546-conjugated goat anti-mouse (1:800) and Alexa 488-conjugated goat anti-rabbit antibodies (1:400) for 1 h at room temperature followed by several washes in PBS. Coverslips were mounted in Mowiol.
Labeling was viewed with a confocal laser scanning microscope (Leica TCS) equipped with an argon-krypton laser (488-, 568-, and 657-nm excitation lines). For double staining, light emitted from the two fluorophores was detected sequentially. Band-pass filters were chosen to select each emission. Images were reconstructed from a series of optical sections taken in the x-y plane from consecutive z positions (0.45-0.5 m apart) using the standard microscope software (Leica Scanware). Original fields were made up of 512 ϫ 512 pixels. Images were processed with Adobe Photoshop.

Isolation of NDPK and Eps15 as Phocein Interactors in a
Yeast Two-hybrid Screen-To identify interacting proteins, phocein was used as a bait in a yeast two-hybrid screen of a rat brain cDNA library. Among the 68 positive clones obtained, 5 encoded NDPK, 2 encoded a fragment of Eps15, 26 encoded the ferritin H chain, and 35 contained plasmids of different insert sizes, not further studied. Four positive identical clones encoded an insert of 0.9 kb containing the 456-bp entire sequence of the ␤-subunit of NDPK (18); a fifth positive clone encoded a 1.3-kb insert encompassing the latter sequence and extending further into the 3Ј non-coding sequence. Rat NDPKs consist of homo-hexamers of ␣ or ␤ isoforms, which are 89% identical (18). Both isoforms are found in brain. Two identical positive clones contained a 1.7-kb insert encoding the last 317 C-terminal amino acids of Eps15 and part of the 3Ј non-coding sequence. The C-terminal domain of Eps15 contains several Asp-Pro-Phe (DPF) motifs and binds the ear domain of ␣-adaptin (8).
Biochemical Validation of the Interactions of Phocein with NDPK and Eps15-In vitro biochemical confirmation of the interactions revealed by the yeast two-hybrid screen was achieved by reciprocal coimmunoprecipitation of rat brain proteins and by pull-down experiments. Phocein and NDPK were found in brain cytosol (Fig. 1, A and B, lane 1) as well as in CHAPS-solubilized membrane fractions, abbreviated as CHAPS fractions (Fig. 1, A and B, lane 4). Anti-NDPK antibodies immunoprecipitated NDPK from cytosol (Fig. 1B, lane 2) or from CHAPS fractions (Fig. 1B, lane 5). Phocein endogenous to each fraction coimmunoprecipitated with NDPK (Fig. 1A, lanes  2 and 5). From 5 to 10% of the phocein contained in the cytosol sample was immunoprecipitated by the anti-NDPK antibodies (mean of three experiments). Control immunoprecipitates obtained with preimmune immunoglobulins contained neither phocein nor NDPK (Fig. 1, A and B, lanes 3 and 6). Reciprocally, anti-phocein antibodies incubated with rat brain cytosol fractions coimmunoprecipitated small amounts of NDPK together with phocein (Fig. 1C, lane 2), whereas control immunoprecipitates contained neither protein (Fig. 1C, lane 3). Performing the incubations at 30 versus 4°C resulted in slightly enhanced amounts of coimmunoprecipitated protein. Varying the Ca 2ϩ concentration of the incubated samples did not affect the latter parameter. We further demonstrated that the interaction between phocein and NDPK is direct using GST pulldown experiments with recombinant proteins. GST-phocein ( Fig. 2A, lane 1) pulled down recombinant NDPK (Fig. 2B, lane  1), whereas GST did not (Fig. 2A, lane 2). Reciprocally, GST-P-NDPK (Fig. 2A, lane 3) pulled down recombinant phocein (Fig. 2B, lane 3), whereas GST did not (Fig. 2A, lane 4).
Similarly, in vitro biochemical confirmation of the interaction between phocein and Eps15 was obtained. Eps15 contained in endogenous rat brain cytosol (Fig. 3, lane 1) was detected in small amounts in immunoprecipitates obtained with antibodies directed against phocein (Fig. 3, lane 2), whereas it was absent from immunoprecipitates obtained with preimmune immunoglobulins (Fig. 3, lane 3). From 1 to 6% of the Eps15 contained in the cytosol sample was immunoprecipi-tated by anti-phocein antibodies. Pull-down experiments showed that Eps15 contained in rat brain cytosol fractions was pulled down by GST-phocein (Fig. 3, lane 4) but not by GST (Fig. 3, lane 5). Attempts to express the correctly folded protein GST fused to the C-terminal fragment of Eps is in sufficient amount to study its putative interaction with recombinant phocein were unsuccessful.
Phocein Interacts with Dynamin I-Dynamin has been shown to be a direct partner of Eps15 (11) and a potential partner of NDPK (10). Since phocein was proposed to have a role in vesicular traffic, we investigated whether phocein interacts with dynamin I. As shown in Fig. 4, immunoprecipitates obtained by incubating anti-phocein antibodies with rat brain cytosol (lane 3) contained dynamin I. Likewise, pull-down experiments conducted with GST-phocein (Fig. 5A, lanes 1 and 2) incubated with rat brain cytosol (Fig. 5A, lane 1) or a CHAPS fraction (Fig. 5A, lane 2) showed that dynamin was pulled down by GST-phocein (Fig. 5B, lanes 1 and 2), whereas it was not pulled down by GST (Fig. 5, A and B, lane 3). However, attempts to demonstrate a direct interaction between dynamin and phocein using recombinant proteins (GST-P-PRD, Hisdynamin ⌬ PRD, and phocein) yielded only negative results, although GST-P-PRD was able to pull down recombinant amphiphysin (not shown) (16).

NDPK and Dynamin I Directly Interact in Vitro-Although
Eps15 was shown to interact not only genetically but also biochemically with dynamin (11), no physical association of NDPK with dynamin could be demonstrated despite a strong genetic interaction between the two proteins in Drosophila (10). We further investigated the interaction between NDPK and dynamin. Dynamin I was present in immunoprecipitates obtained by incubating anti-NDPK antibodies with rat brain cytosol (Fig. 4, lane 2) or a CHAPS fraction (Fig. 4, lane 5), whereas immunoprecipitates obtained with control antibodies were devoid of dynamin I (Fig. 4, lanes 4 and 6). We next incubated the fusion protein GST-P-NDPK with rat brain cytosol (Fig. 6A, lane 1) or a CHAPS fraction (Fig. 6A, lane 2); in both cases, GST-P-NDPK pulled down endogenous dynamin I (Fig. 6B, lanes 1 and 2), whereas GST did not (Fig. 6, A and B,  lane 3).
Compared Localizations of Phocein and Dynamin I-Hippocampal cells from rat E18 embryos, cultured for 3 and 10 days (Banker's stages 3 and 5), extended several branched neurites (Fig. 7, A and B). Phocein immunoreactivity (green fluorescence) was conspicuous all over the soma, except for nuclei, and involved all neurites. It appeared as small spots in neurites and perikarya; however, the staining was more diffuse in the perinuclear region. Dynamin I immunoreactivity (red fluorescence) also had a spotty appearance and involved all neurites and perikarya. Confocal microscopy showed that the distribution of the two proteins overlapped within several regions: in the Golgi area (Fig. 7A) at the emergence of neurites from the cell body, where the dynamin labeling however spread closer to the cell surface than the phocein labeling (Fig. 7, A1 and B2) and within neurites (Fig. 7, A1, B1, and B2). Colocalization was observed at Banker's stage 3 in growth cones (Fig. 7, A2). Immunolabeling of NDPK in these neurons revealed the presence of the protein in all subcellular compartments, nucleus, soma and neurites, down to the smallest branches (not shown).

DISCUSSION
Phocein was isolated previously as a partner of the members of the striatin family. Based on its homology to the -subunit of adaptor proteins, its subcellular localization, and its sensitivity to brefeldin A, phocein was proposed to be involved in vesicular traffic (1). Here, we have identified two novel binding partners for phocein, NDPK and Eps15. Both proteins have been implicated in membrane traffic, further strengthening a potential role of phocein in membrane dynamics. More specifically, all these proteins appear to be part of a protein network, which also includes dynamin.
As demonstrated both in vitro and in vivo, phocein directly interacts with NDPK. NDPKs are ubiquitous enzymes that participate in a variety of cell processes and appear to be important in the supply of local pools of GTP (for example, in the cytoplasm near the cell surface, where a large number of GTPases operate (19 -22)). Among them, dynamins are GT-Pases that play an essential role in the fission of clathrincoated vesicles from the plasma membrane and are also implicated in other steps of intracellular vesicular trafficking (reviewed in Refs. [23][24][25]. Although in the case of most GT-Pases, GTP binding is controlled by guanyl-nucleotide exchange factors, it appears that for dynamins, GTP loading is primarily dependent upon the local concentration of substrate, thus explaining the dependence of dynamin activity on NDPK. Dynamins are endowed with very low affinity for GTP, yet they have particularly high intrinsic (1-2 min Ϫ1 ) and stimulated (over 100 min Ϫ1 ) rates of GTP hydrolysis (24,26,27). Based on copurification experiments, Shpetner and Vallee (28) were the  1 and 2) and GST (lane 3) were incubated with rat brain cytosol (lane 1) or a CHAPS fraction (lanes 2 and 3). A, Ponceau red staining of the blotted fusion proteins. B, blots revealed with antidynamin I antibodies.
first to consider a possible role for NDPK in dynamin function. Recently, genetic studies conducted in Drosophila by Krishnan et al. (10) have shown that the activity of NDPK is critically required for the function of dynamin at synapses. They suggested that NDPK 'transiently associates with dynamin, thus being optimally positioned to provide a very high local concentration of soluble GTP.' However, using fly heads and techniques such as coimmunolocalization, coimmunoprecipitation, or pull-down experiments, they did not find physical evidence for such an association. The present study provides a clear indication that mammalian dynamin and NDPK physically interact.
Dynamin is present throughout the cell (29,30) and is particularly concentrated in nerve terminals (31)(32)(33). Phocein, according to our previous immunocytochemical data (1), is found in the perikaryal-dendritic region of neurons, down to the spines, whereas in non-neuronal cells, it localizes predominantly in the Golgi complex. By contrast, within neurons, immunoreactivity for NDPK is diffuse and ubiquitous. NDPK is thus present in the subcellular regions where phocein and dynamin are expressed, in agreement with their partial asso-ciation in brain extracts. The distributions of phocein and dynamin partially overlap in neurons, as well as in unpolarized cultured cells, in which dynamin isoforms are indeed detected at the level of the Golgi complex (34,35). Within neurons, phocein may be implicated in only a subset of the reactions assisted by dynamin I, more specifically in reactions that occur in dendrites.
The other phocein partner identified in the yeast two-hybrid assay, Eps15, has been implicated at several steps of the endocytic pathway. Eps15 is a major, regulated binding protein for the clathrin adaptor AP-2. Its enrichment at the neck of clathrin-coated pits, as determined by immunogold cytochemistry, has suggested a function somehow interconnected with the action of dynamin (8, 9, 36 -38). Clathrin-dependent endocytosis is selectively blocked when Eps15 function is perturbed by antibody or peptide microinjection or by the expression of constructs that function by dominant negative interference (Ref. 39 and references therein). Interestingly, Salcini et al. (11) very recently showed that mammalian Eps15 and dynamin genetically and biochemically interact, both in vitro and in vivo.
Phocein is thus a part of the multiprotein complexes com-  1, 2, and 4), GST-P-PRD (lane 6), and GST (lanes 3, 5, and 7). The fusion proteins had been incubated previously with cytosol (lane 1) or CHAPS fractions (lanes 2 and 3), the recombinant PRD fragment of dynamin I (lanes 4 and 5), and recombinant NDPK (lanes 6 and 7). B, the lower part of the blots revealed with anti-dynamin I (lanes 1-5) and anti-NDPK antibodies (lanes 6 and 7). 7. Dynamin and phocein partially co-localize in cultured hippocampal neurons. Stage 3 (A) and stage 5 (B) neurons were fixed and processed for immunofluorescence microscopy using rabbit anti-phocein (5 g/ml) (green) and mouse anti-dynamin I (10 g/ ml) (red) antibodies. Areas of co-localization appear yellow in the computer-generated composite image. A1, A2, B1, and B2 are 5ϫ enlargements of the boxed regions in panels A and B, respectively. In panels A and B: scale bar, 40 m.
prising NDPK, Eps15, dynamin, and the proteins of the striatin family (2,3,5). That these complexes are bulky has been shown by gel filtration and sucrose gradient centrifugation (1). Blotted immunoprecipitates obtained by antibodies directed against phocein can be sequentially shown to contain all these proteins. However, the analyzed immunoprecipitates are a mixture of complexes; as a result, it is not possible to know whether these proteins all coexist within one given complex at the same time. Nevertheless, even if phocein interacts with these proteins in a sequential manner, such results are clearly in favor of a likely role of phocein in vesicular trafficking, particularly endocytosis. Yet how do these interactions occur? In both its N-and C-terminal domains, phocein has homology to the -subunits of AP membrane adaptors and thus might be components of AP complexes. Preliminary experiments indicate that ␥and ␣-adaptins, respectively, components of the AP1 and AP2 adaptors, coimmunoprecipitate with phocein even in rather harsh conditions. The finding that ␥-adaptin coimmunoprecipitates with phocein is consistent with the localization of phocein at the Golgi complex (1). Likewise, coimmunoprecipitation of ␣-adaptin with phocein is consistent with a role in clathrinmediated endocytosis. Availability of antibodies directed against various adaptor subunits should help clarify whether phocein could actually be part of stable tetrameric adaptor complexes.
Phocein, however, may exist independently of known AP subunits. Phocein has a central domain containing a putative Src homology 3 (SH3)-binding motif not conserved in -subunits. Another protein, stonin 2, which has homology to the -subunit of AP adaptors, is not found in tetrameric complexes (40).
Whatever the way through which phocein is inserted within these NDPK, Eps15, dynamin, and striatin-containing complexes, phocein could help localize and/or stabilize their association. Since phocein has been shown to be a substrate of the protein phosphatase 2A, it is likely that some of its interactions are regulated by its phosphorylation state (6).
The multimodular WD repeat-containing and calmodulinbinding proteins that constitute the striatin family are likely to play scaffolding and Ca 2ϩ -dependent signaling roles. Phocein, their major interactor, which also interacts with proteins involved in vesicular trafficking, could be involved in the crosstalk between endocytosis and signaling; growing evidence is now documented for this cross-talk (41).