REPS2 downregulates receptor signaling and endocytosis: REPS2/POB1 interacts with Eps15, epsin EPN1, 14-3-3 isoforms, Grb2, amphiphysin
The presence of EH domains in REPS2/POB1 is symptomatic of a role of this gene in receptor endocytosis. In fact, REPS2/POB1 EH domain binds to Eps15 and to epsin that are both proteins present in clathrin-coated pits, involved in receptor endocytosis and receptor down regulation. The EH domain interacts specifically with the three Asn-Pro-Phe (NPF) motifs in the C-terminal region of epsin and their binding regulates receptor-mediated endocytosis (Morinaka et al., 1999).
Augmented expression of full-length POB1 in A431 cells does not affect either binding or internalization of EGF, on the contrary, over-expression of either the EH domain or the C-terminal region of REPS2/POB1 affects the ligand dependent internalization pathway of EGF and insulin without interfering with the constitutive transferrin pathway (Nakashima et al., 1999).
Santonico et al. have demonstrated that the EH domain of REPS2/POB1 binds Eps15 through an unconventional recognition specificity, since it binds to both NPF and DPF (Asp-Prp-Phe) motifs. The region of Eps15 responsible for the interaction with the EH domain of REPS2/POB1 maps within a 18 amino acid peptide (residues 623-640) that includes three DPF repeats. Accordingly, the authors identify a cluster of solvent exposed Lys residues, which are only found in the EH domain of REPS2/POB1, and influence binding to both NPF and DPF motifs (Santonico et al., 2007).
RalBP1, REPS2/POB1, epsin, and Eps15 form a complex with alpha-adaptin of AP-2 in Chinese hamster ovary cells and this complex is reduced in mitotic phase, when REPS2/POB1 and epsin are found phosphorylated. They are both phosphoryated by p34 cdc2 kinase, in vitro. POB1 is found phosphorylated in Ser551 and Ser493, in vivo. Phosphorylation of epsin in Ser 357 inhibits binding to POB1, causing disassembly of the complex, thus inhibiting receptor mediated endocytosis (Kariya et al., 2000). This data explains the contribution of the EH domain of POB1 to the formation of a protein complex that favours receptor internalization and that it is dismantled in mitosis. It is suggested that EGF stimulation induces also tyrosine-phosphorylation of POB1 and subsequent formation of a EGFR-POB1 complex in COS cells (Ikeda et al., 1998).
REPS2/POB1 shorter isoform2 is downregulated during human prostate cancer progression from androgen-dependent to androgen-independent (Oosterhoff et al., 2003). It was observed that a high level of REPS2 correlates with reduced EGF-internalisation and signaling since the induced expression of REPS2 exerts an inhibiting effect on several EGF-responsive genes (EGF-receptor, EGR-1, Fos and Jun) (Oosterhoff et al., 2005). Accordingly, increased expression of POB1 isoform 2 correlates with a decrease of EGF-induced phosphorylation of Erk1-Erk2 and Shc (Tomassi et al., 2008). From these experiments, it can be concluded that increased REPS2/POB1 expression negatively affects EGF receptor internalisation and subsequent signaling. Therefore, the decreased REPS2 expression observed during prostate cancer progression, results in enhanced EGF receptor expression and signaling, which could add to the androgen-independent state of advanced prostate cancer.
The central region of REPS2/POB1 plays a regulatory role in epidermal growth factor receptor endocytosis and signaling. Overexpression of the central region of POB1 (447-504), inhibits EGF endocytosis, titrating essential proteins away, thus depauperating the receptor down-regulation machinery. In fact, this region of POB1/REPS2 plays its regulatory role in EGFR endocytosis by binding: (i) to 14-3-3 proteins in a phosphorylation dependent way (i.e., phospho-Ser493 of POB1), (ii) to the C-SH3 domain of Grb2 and (iii) to the SH3 of amphiphysin II. The target of the SH3 domain of amphiphysin and of the carboxy-terminal SH3 of Grb2 is a short peptide flanking Arg483 in POB1. These interactions are not EGF dependent and are probably exclusive, since the binding motifs are only nine amino acids apart. These findings suggest that 14-3-3 could work by bridging the EGF receptor and the scaffold protein POB1/REPS2. The 14-3-3 binding motif HSRASSLD, flanking the Ser493 of POB1, is conserved in the mouse orthologs and in the 14-3-3 binding motif that flanks the Ser510 of human REPS1 protein, found phosphorylated in vivo in A431 cells (Stover DR et al. Phosphosite). The POB1 Ser493 is predicted to be phosphorylated by Akt. In agreement, when cells are treated with PI3K/Akt inhibitor wortmannin, 14-3-3 binding to REPS2/POB1 is abolished (Tomassi et al., 2008). The 14-3-3 zeta has already been reported as associating with the EGFR, epidermal growth factor receptor, cytoplasmic tail and co-localizing along the plasma membrane with EGFR upon EGF stimulation (Jin et al., 2004). Thus a 14-3-3 dimer could bridge REPS2/POB1 to the EGFR upon EGF induction.

Cell migration and paxillin localization: REPS2/POB1 antagonises PAG2/ASAP1
POB1 forms a complex with PAG2/ASAP1 in intact cells. PAG2 is a paxillin-associated protein with ADP-ribosylation factor GTPase-activating protein activity, also called ASAP1 (ArfGAP with SH3 domain, ankyrin repeat and PH domain UniProtKB: Q9ULH1). The SH3 of PAG2 binds the proline motif (562PSKPIR567) at the carboxyl-terminal region of POB1. This motif is essential for PAG2-POB1 interaction since substitution of the two proline residues with alanines in mutant POB1(PA), impaired its binding to PAG2. POB1 may therefore form a complex with paxillin through PAG2. Paxillin is a focal adhesion-associated scaffolding protein that recruits signaling molecules to the focal adhesions and forms protein complexes that coordinate signaling, cell spreading and motility. PAG2 overexpression causes loss of endogenous paxillin recruitment to focal contacts and also impaires cell migratory activities. The ability to suppress fibronectin-dependent migration depends on the ArfGAP domain of PAG2, but not on the POB1-binding domain, of PAG2. On the other end, POB1, but not POB1(PA), can suppress the inhibitory action of PAG2 on paxillin localization to focal adhesion (Oshiro et al., 2002). These results suggest that POB1, by binding to PAG2, suppresses the inhibitory action of PAG2 on the paxillin recruitment to focal contacts. This suggests that POB1 may function as a scaffold protein that interacts with proteins involved in endocytosis and migration, thus regulating signaling and motility. PAG2/ASAP1 gene was found associated with prostate cancer metastasis since it is up-regulated in a human metastatic prostate subline and immunohistochemistry of xenograft sections show a significantly strong cytoplasmic ASAP1 protein staining in tumor-nonmetastatic PCa2 tissue, compared to a non-staining in benign tissue, and an even stronger staining in PCa1-metastatic tissue. Moreover, additional ASAP1 gene copies are detected in 58% of the primary prostate cancer clinical specimens. A small interfering RNA reducing ASAP1 protein expression, can suppress in vitro PC-3 cell migration and matrigel invasion. Therefore PAG2/ASAP1 represents a therapeutic target and a biomarker for metastatic disease (Lin et al., 2008) while REPS2/POB1 can suppress PAG2 oncogenic-metastatic activity.