PED/PEA-15 interacts with the 67 kD laminin receptor and regulates cell adhesion, migration, proliferation and apoptosis

Abstract Phosphoprotein enriched in diabetes/phosphoprotein enriched in astrocytes-15 kD (PED/PEA-15) is an anti-apoptotic protein whose expression is increased in several human cancers. In addition to apoptosis, PED/PEA-15 is involved in the regulation of other major cellular functions, including cell adhesion, migration, proliferation and glucose metabolism. To further understand the functions of this protein, we performed a yeast two-hybrid screening using PED/PEA-15 as a bait and identified the 67 kD high-affinity laminin receptor (67LR) as an interacting partner. 67 kD laminin receptor is a non-integrin cell-surface receptor for the extracellular matrix (ECM), derived from the dimerization of a 37 kD cytosolic precursor (37LRP). The 67LR is highly expressed in human cancers and widely recognized as a molecular marker of metastatic aggressiveness. The molecular interaction of PED/PEA-15 with 67LR was confirmed by pull-down experiments with recombinant His-tagged 37LRP on lysates of PED/PEA-15 transfected HEK-293 cells. Further, overexpressed or endogenous PED/PEA-15 was co-immunoprecipitated with 67LR in PED/PEA-15-transfected HEK-293 cells and in U-373 glioblastoma cells, respectively. PED/PEA-15 overexpression significantly increased 67LR-mediated HEK-293 cell adhesion and migration to laminin that, in turn, determined PED/PEA-15 phosphorylation both in Ser-104 and Ser-116, thus enabling cell proliferation and resistance to apoptosis. PED/PEA-15 ability to induce cell responses to ECM-derived signals through interaction with 67LR may be of crucial importance for tumour cell survival in a poor microenvironment, thus favouring the metastatic spread and colonization.

Because of its functional role in ERK signalling and apoptosis, increased PED/PEA-15 levels may affect tumourigenesis and cancer progression as well as sensitivity to anti-cancer agents. We demonstrated that overexpression of PED/PEA-15 in a transgenic mouse model increases the susceptibility to chemically induced skin cancer [9]. Increased PEA-15 levels inhibit apoptosis in non-small cell lung cancer (NSCLC) [10], B-cell chronic lymphocytic leukaemia [11] and thyroid cancer [12]. In astrocytic tumours, PED/PEA-15 suppresses apoptosis [13] and prevents glucose deprivation-induced cell death via the ERK pathway [14], suggesting that PED/PEA-15 promotes tumour cell survival in a poor microenvironment.
PED/PEA-15 also plays a role in the regulation of cell adhesion and migration; indeed, its binding to ERK1/2 regulates the affinity for fibronectin (FN) of integrin adhesion receptors [15]. In astrocytes, PEA-15 prevents cell migration through a PKC deltadependent pathway [16]. It has been recently reported that PED interacts with Rac1 and regulates cell migration/invasion processes in human NSCLC cells [17].
To further understand the functions of PED/PEA-15 in cancer, we performed a yeast two-hybrid screening using PED/PEA-15 as a bait and identified the 67LR as an interacting partner. 67 kD laminin (LM) receptor was originally identified as a non-integrin cell surface receptor for LM, an extracellular matrix molecule [18]. Laminins, other glycoproteins, collagen IV and proteoglycans constitute a tight network to form the basement membrane. Laminin-1, a 900-kD glycoprotein, is the major component of basement membranes and contains many bioactive domains involved in binding both integrin and non-integrin receptors [19]. Interactions between the non-integrin 67LR and LM play a major role in mediating changes in the cellular environment that affect cell adhesion [20], neurite outgrowth [19] and tumour growth and metastasis [21]. 67 kD LM receptor derives from homo-or hetero-dimerization of a 37LRP, by fatty acid acylation [22,23]. 67 kD LM receptor binds LM through different binding domains [24,25]. Laminin conformation changes upon binding 67LR, thus interacting more efficiently with integrins [26] and becoming more sensitive to the action of proteolytic enzymes [27], with the release of motility fragments [28]. 67 kD LM receptor is co-expressed and can physically interact with the ␣6-integrin chain [29]. 67 kD LM receptor expression is increased in neoplastic cells as compared to their normal counterparts and directly correlates with an enhanced invasive and metastatic potential [30], mediated by high-affinity interactions between 67LR and LM [31]. Thus, 67LR overexpression is considered a molecular marker of metastatic aggressiveness in cancers of many tissues, including breast, lung, ovary, prostate and also in leukaemia and lymphomas [32][33][34]. For these reasons, the specific targeting of 67LR with small-interfering RNAs (siRNAs), blocking antibodies and Sindbis viral vectors confers antitumour effects [35,36].
Herein, we show 67LR interaction with both overexpressed and endogenous PED/PEA-15 and investigate the functional consequences of this interaction in the regulation of cell adhesion, migration, proliferation and apoptosis.

Transformation of yeast strains and ␤-galactosidase assay
Plasmid DNA transformations were performed with high-efficiency lithium acetate procedure [38]. Co-transformants were propagated on Trp ϩ , Leu Ϫ plates and potential interacting clones selected in Trp Ϫ , Leu Ϫ , His Ϫ , Ade Ϫ media. After 4 days of incubation at 30ЊC, positive clones were further tested for ␤-galactosidase activity by liquid culture assays using the substrate o-nitrophenyl-␤-d-galactopyranoside as described by Miller et al. [39]. Clones of interest were analysed by DNA sequencing and BLAST analysis.

Cell culture
Two ninety-three kidney embryonic cells, featuring very low levels of the anti-apoptotic protein PED/PEA-15 [40], and human U-373 glioma cells were grown in DMEM medium supplemented with 10% heat-inactivated FCS. Transfected cells were grown in DMEM supplemented with 10% FBS and antibiotics.

Transfection
PED/PEA-15 cDNA was cloned in a pcDNA3 vector with resistance to Geneticin, and the resulting plasmid was named PED/PEA-15-pcDNA3. A total of 5 ϫ 10 6 293 cells were transfected with 10 g of PED/PEA-15-pcDNA3 or with the empty vector pcDNA3 by 60 l of Lipofectamine, for 5 hrs at 37ЊC (5% CO2). Transfected cells, named PED-293 and V-293 respectively, were selected by Geneticin at 1.5 mg/ml for 15 days, pooled and cultured in the presence of 0.5 mg/ml Geneticin.

Flow cytometric analysis of surface molecules
Flow cytometric analysis of cell surface molecules was performed as previously described [33]. Briefly, cells were incubated for 2 hrs at 4ЊC with 20 g/ml anti-67LR or isotype control antibodies. This step was followed by a second incubation for 1 hr at 4ЊC with an anti-rabbit fluoresceinconjugated antibody. Finally, cells were washed and analysed with a FACSCalibur Cytofluorometer using Cell Quest software (Becton & Dickinson, San Fernando, CA, USA). A total of 10 4 events for each sample were acquired in all cytofluorometric analyses.

Western blot
V-293 and PED-293 cells were lysed in lysis buffer [50 mM HEPES (pH 7.5), 150 mM NaCl, 4 mM EDTA, 10 mM Na4PO7, 2 mM Na3VO4, 100 mM NaF, 10% glycerol, 1% Triton X-100] containing protease and phosphatase inhibitors, for 120 min. at 4ЊC. Cell lysates were clarified at 5000 ϫ g for 15 min. and the protein content was measured by a colorimetric assay. Fifty micrograms of protein was electrophoresed on a 15% SDS-PAGE and transferred onto a polyvinylidene fluoride (PVDF) membrane. The membrane was blocked with 5% non-fat dry milk, and probed with the primary and secondary antibodies; immunoreactive bands were detected by ECL according to the manufacturer's instructions.
For the detection of phosphorylated PED/PEA-15, ERK1/2, Akt, CaMKII and PKC, 5ϫ10 5 V-293 and PED-293 cells were plated onto 35 mm plates previously coated for 24 hrs with LM (20 g/ml in PBS). After the indicated times, cells were washed with PBS, lysed and subjected to Western blot, as described earlier.

PED/PEA-15 interaction with 67LR
To investigate the interaction of PED/PEA-15 with 67LR, 37LRP-His-tag fusion protein was generated. To this end, wild-type 37LRP cDNA (33) was cloned into the pTrcHis B expression vector (Invitrogen, San Diego, CA, USA) and expressed in TOP-10 bacteria (Invitrogen). According to the procedures specified by Invitrogen, transformed bacteria were lysed in a dena-turing lysis buffer (20 mM sodium phosphate, 500 mM sodium chloride, pH 7.8) containing 6M guanidium and His-tagged 37LRP (His-37LRP) was bound to nickel-NTA agarose beads, through its His-tagged N-terminus, in the same denaturing buffer containing 8M urea. Beads were washed several times at pH 6.0 and 5.3, to dissociate contaminating proteins and finally, His-37LRP bound to agarose beads was more than 90% pure, as assessed by SDS-PAGE and Coomassie stain. Thus, His-37LRP conjugated beads were washed in 50 mmol/l Tris (pH 7.5)-0.1% Triton X-100, to remove urea, and resuspended in the same buffer.
Lysates from PED/PEA-15-transfected 293 cells (500 g) were incubated in the presence of 50 l of agarose-bound His-37LRP (approximately 2 g) for 2 hrs at 4ЊC. Beads were washed four times with 50 mmol/l Tris (pH 7.5)-0.1% Triton X-100, and then resuspended in Laemmli buffer followed by boiling for 6 min. and centrifugation at 25,000 ϫ g for 3 min. Supernatants were analysed by SDS-PAGE and blotted with anti-PED-PEA/15 antibodies.

Co-immunoprecipitation
Cells were harvested into 1 ml of lysis buffer, containing protease and phosphatase inhibitors. Total cell lysates were pre-cleared with 20 l protein A-Sepharose beads (~50% slurry; Amersham Biosciences) for 30 min. at room temperature (RT). Five micrograms of anti-67LR, anti PED/PEA-15 or non-immune antibodies was added to 500 g of cleared lysate, and incubated for 2 hrs at 4ЊC. Then, 20 l of protein A-Sepharose beads was added and the lysates were incubated 30 min. at RT. Immunoprecipitates were washed six times in lysis buffer and Western blotting with anti-67LR antibodies was used to detect 67LR co-immunoprecipitated with PED/PEA-15 in V-293 and PED-293 cells. Viceversa, Western blotting with anti-PED/PEA-15 antibodies was used to detect PED/PEA-15 co-immunoprecipitated with 67LR in U-373 cells. Separately, 25 g of total cell lysate was immunoblotted for 67LR and PED/PEA-15, as a loading control.

Rac1 pull-down assay
V-293 and PED-293 cells were starved for 24 hrs and then plated on LM-coated wells (10 g/ml) or BSA-coated wells, as a negative control, for the indicated times. After a quick wash with ice-cold PBS, cells were lysed with GST-Fish buffer (50 mM Tris-HCl pH 7.4, 2 mM MgCl2, 1% NP-40, 10% glycerol, 100 mM NaCl, 1 mg/ml leupeptin, 1 mg/ml pepstatin, 1 mg/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride and 2 mM dithiothreitol). After 10 min. at 4ЊC under agitation, cells were scraped and lysates were cleared by centrifugation in a pre-cooled rotor. Five hundred micrograms of total protein extract was mixed with 10 g of GST-PAK-CRIB domain coupled to glutathione-sepharose beads (Upstate Biotechnology, Danvers, MA, USA) and incubated 30 min. at 4ЊC under agitation. Beads were then rinsed three times rapidly with 1 ml of ice-cold GST-Fish buffer. The amounts of total Rac and Rac-GTP were estimated by immunoblot against Rac1 (Upstate Biotechnology).

Adhesion assay
Ninety-six-well flat-bottom microtitre plates were coated with 10 g/ml of LM, vitronectin, FN, YIGSR peptide or 1% heat-denatured BSA-PBS as a negative control, and incubated overnight at 4ЊC. The plates were then blocked 1 hr at room temperature with 1% heat-denatured BSA-PBS. Cells were harvested by 2 mM EDTA-PBS and resuspended in Ca 2ϩ /Mg 2ϩ -containing PBS.
A total of 10 5 cells were plated in each coated well and incubated for 1 hr at 37ЊC. Attached cells were fixed with 3% paraformaldehyde in PBS for 10 min. and then incubated with 2% methanol for 10 min. Cells were finally stained for 10 min. with 0.5% crystal violet in 20% methanol. Stain was eluted by 0.1M sodium citrate in 50% ethanol, pH 4.2 and the absorbance at 540 nm was measured by a spectrophotometer.

Cell proliferation assay
V-293 and PED-293 cells were serum-starved overnight using DMEM 0.1% BSA, plated at 10 4 cells/well in 96-well plates coated with 10 g/ml LM in the same medium and incubated for 1, 24, 48 and 72 hrs or 7 days at 37ЊC, 5% CO2. The growth medium was then removed and 20 l/well of CellTiter 96 AQueous One Solution Reagent was added. After incubation at 37ЊC for 4 hrs, the absorbance was determined by an ELISA reader (Bio-Rad) at a wavelength of 490 nm.

Cell apoptosis assay
V-293 and PED-293 cells were serum-starved overnight using DMEM 0.1% BSA. A total of 10 6 cells were harvested by 2 mM EDTA-PBS and plated in 100 mm dishes pre-coated with 10 g/ml LM. After 24, 48 and 72 hrs at 37ЊC, 5% CO2, apoptosis was analysed using a colorimetric assay that measures the proteolytic cleavage of the chromophore p-nitroanilide (pNA) by caspase-3. Some experiments were performed in the presence of the anti-67LR antibody or non-immune control Ig (10 g/ml).

Cell migration and invasion assays
Cell migration assays were performed in Boyden chambers using 8 m pore size PVPF polycarbonate filters coated with 50 g/ml collagen, as an adhesion substrate. V-293 and PED-293 cells (2 ϫ 10 5 ) were plated in the upper chamber in DMEM 0.1% BSA. DMEM 0.1% BSA alone or containing 50 g/ml LM was added in the lower chamber. Cells were allowed to migrate for 4 hrs at 37ЊC, 5% CO2. For the invasion assay, filters were coated with 70 g/ml Matrigel TM (BD Biosciences, San Jose, CA, USA) and incubated for 3 hrs for gelling. A total of 2 ϫ 10 5 cells were allowed to migrate towards DMEM medium supplemented with 10% FCS or 0.1% BSA, as a control, for 18 hrs at 37ЊC, 5% CO2.
At the end of both experiments, cells on the lower surface of the filter were fixed in ethanol, stained with haematoxylin, and counted at 200ϫ magnification (10 random fields/filter).
In a separate set of experiments, transfected cells were pre-incubated for 1 hr at 37ЊC with a polyclonal antibody directed to 67LR (10 g/ml) or non-immune antibodies as a negative control, or with different pharmacological Rac1 and ERK1/2 inhibitors (10 M NSC23766, 25 M PD98059) or diluents as a control.
Cell migration and invasion were expressed as percent increase over control.

Statistical analysis
Differences between groups were evaluated by the Student's t-test using PRISM software (GraphPad, San Diego, CA, USA). P Յ 0.05 was considered statistically significant.

Isolation and identification of the 37 kD LM receptor precursor (37LRP) as a novel PED/PEA-15 interacting protein
To search for proteins specifically interacting with PED/PEA-15, a yeast two-hybrid system was established using the full-length PED/PEA-15 gene (pG-BKT7-ped/pea-15) as a bait to screen a human HeLa library (Clontech). Upon HIS3 selection, 45 ␤-galactosidase positive cDNA clones were detected. Based upon sequence analysis and BLAST searching, three of these clones were shown to match the 37LRP sequence [22]. Moreover, two clones coding for phospholipase A2 and three clones coding for FADD protein, both well-known as PED/PEA-15 interactors [1,7,8], were also found.
Dimerization of 37LRP [22,23] generates the mature form of the receptor, the 67LR. 37LRP is localized mostly in the cytoplasm acting as a precursor for 67LR, associated to ribosomes and in the nucleus, where it interacts with histones [30]. 67 kD LM receptor is localized in the plasma membrane, where it binds LM with high affinity, thus playing a crucial role in tumour invasion and metastasis [22][23][24][25]30]. We sought to define by co-immunoprecipitation experiments the form of the receptor interacting with PED/PEA-15 in PED-293 cells, which express endogenous 37LRP and 67LR (Fig. 1C). Precipitation of lysates from PED-293 cells with anti-PED/PEA-15 antibodies, followed by blotting with antibodies able to recognize both 37LRP and 67LR, revealed that the mature membrane-bound form of the 67LR and not the 37LRP co-precipitates with PED/PEA-15, thus indicating a PED/PEA-15 interaction with 67LR. These results suggest that the PED/PEA-15 binding domain is exposed in purified recombinant 37LRP and in mature cellular 67LR whereas it is not available in cellular 37LR ( Fig. 1B  and C), as it occurs for LM, which binds bacterial recombinant 37LRP (our unpublished data and 41) while interacts only with cellular 67LR [18,23,42,43].

PED/PEA-15 overexpression increases 67LR-mediated cell adhesion and migration to LM and extracellular matrix invasion
PED/PEA-15 is overexpressed in different human tumours, including breast and lung cancer, in which it determines refractoriness to anticancer therapy and resistance to apoptosis [14][15][16][17]. Therefore, we investigated the functional effects of PED/PEA-15 interaction with 67LR in PED/PEA-15 overexpressing cells.
Firstly, we sought to investigate whether PED/PEA-15 overexpression could influence 67LR-mediated cell functions, namely, cell adhesion and migration to LM. Thus, 293 cells transfected with the empty pcDNA3 vector (V-293) and 293 cells transfected with the PED/PEA-15-pcDNA3 plasmid (PED-293) were allowed to adhere to LM, FN and vitronectin (VN). PED-293 cell adhesion to LM was significantly increased, as compared to V-293; conversely, PED-293 cell adhesion to VN and FN was not affected ( Fig. 2A), indicating that PED/PEA-15 overexpression in 293 cells specifically regulates the adhesion to LM. To investigate whether PED/PEA-15 overexpression could increase cell adhesion to LM through 67LR activation, V-293 and PED-293 cells were plated on wells coated with a peptide derived from the B1 chain of LM which specifically binds 67LR, the Tyr-Ile-Gly-Ser-Arg (YIGSR) pentapeptide [45]. Both V-293 and PED-293 cells were able to adhere to YIGSR coated wells; however, PED-293 cells showed a significantly increased binding to the YIGSR peptide, similar to that observed to full length LM (Fig. 2B).
Then, to further discriminate 67LR and integrin contribution to the increased adhesion to LM of PED-293 cells, adhesion assays were performed in the presence of specific antibodies. V-293 and PED-293 cell treatment with anti-67LR antibodies showed that endogenous 67LR strongly contributes to V-293 cell adhesion to LM, even in the absence of PED/PEA-15, and confirmed that increased PED-293 cell adhesion to LM is mainly mediated by the 67LR (Fig. 2C). Accordingly, antibodies directed to the LM-binding site of ␣3 and ␣6 chains, the major LM-binding integrin subunits, did not affect significantly V-293 and PED-293 cell adhesion. Surprisingly, antibodies directed against the signal transducing ␤1-integrin chain, almost completely abolished the adhesion to LM of both V-293 and PED-293 cells, similarly to anti-67LR antibodies. These results suggest that, even integrins are not directly involved in the increased binding of PED-293 cells to LM, anyhow they contribute to generate a downstream signal, likely because 67LR is associated to the ␣6 ␤1-integrin on the cell membrane [29]. We thus hypothesize that 67LR mediates 293 cell binding to LM, but adhesion signals are mediated by its lateral association to ␤1-integrins.
We finally assessed whether PED/PEA-15 overexpression regulated 293 cell adhesion to LM by increasing 67LR expression. Cytofluorimetric analysis showed that V-293 and 293-PED cells expressed on their surface comparable levels of 67LR (Fig. 2D).
Then, the effect PED/PEA-15 overexpression on 67LR-dependent cell migration to LM was examined. V-293 and PED-293 cells were allowed to migrate towards LM in Boyden chambers. Anti-67LR antibodies completely impaired PED-293 cell migration to LM; anti-67LR antibodies also impaired V-293 cell migration to LM, likely by blocking the endogenous 67LR, as in adhesion assays (Fig. 3A).
Further, V-293 and PED-293 cells were plated on Matrigel, whose main component is LM, and allowed to migrate towards a generic chemoattractant, such as foetal bovine serum. Also in this case, PED/PEA-15 overexpression increased cell ability to invade Matrigel and anti-67LR antibodies abolished the observed increase (Fig. 3B).
These results suggest that PED/PEA-15 overexpression increases cell adhesion to LM by interacting with the 67LR and that 67LR requires ␤1-integrin partnership for an efficient intracellular signalling. Moreover, PED/PEA-15 overexpression, increases migration to LM and extracellular matrix invasion specifically via 67LR.

Adhesion to LM affects cell signalling in PED/PEA-15 overexpressing cells
Therefore, we investigated LM effects on the cell signalling in PED/PEA-15 overexpressing cells. PED/PEA-15 activates ERK/MAP Kinases through a Ras-dependent pathway [46]. It has been recently reported that, in an NSCLC cell line, PED/PEA-15 overexpression promotes the activation of Rac1, a member of the mammalian Rho GTPase protein family, in response to growth factor stimulation; in turn, active Rac1 increases ERK1/2 phosphorylation [17,47].
Control V-293 and PED-293 cells were plated on LM; then, ERK1/2 and Rac1 activation were evaluated by Western blot analysis and by pull-down assay, respectively. PED/PEA-15 overexpressing cells, following LM stimulation, showed increased ERK1/2 activation, as compared to control V-293 cells (Fig. 4A, upper panel). Cell adhesion to LM also promoted Rac1 activation in PED/PEA-15 overexpressing cells, as compared to vector transfected cells (Fig. 4B, upper panel). Both ERK1/2 and Rac1 activation were specifically induced by the LM stimulus, as shown by plating PED-293 cells on uncoated or LM coated wells ( Fig. 4A and  B, lower panels).
These results show that LM exerts its downstream effects in PED/PEA-15 overexpressing cells by activating Rac1 and ERK1/2. Because both Rac1 and ERK1/2 activation can mediate cell migration [47,48], their involvement in the migration of PED/PEA-15 overexpressing cells towards LM was evaluated. The pharmacological inhibition of Rac1 and ERK1/2 by NSC23766 and PD98059, respectively, impaired PED-293 cell migration towards LM (Fig. 4B), demonstrating their involvement in LM-induced PED-293 cell migration.

67LR-mediated cell adhesion to LM regulates proliferation and apoptosis in PED/PEA-15 overexpressing cells by modulating PED/PEA-15 phosphorylation status
We then investigated whether 67LR can modulate PED/PEA-15 effects on cell proliferation and apoptosis. Firstly, we analysed V-293 and PED-293 cell proliferation after adhesion to LM by a colorimetric assay (Fig. 5A). As expected, LM induced a strong proliferative response in V-293 cells, whereas PED-293 cell proliferation was inhibited. However, after 24 hrs, PED-293 cells recovered their proliferative capacity, reaching control cells at 7 days. V-293 and PED-293 cell number, after adhesion to LM, was assessed at the indicated times and confirmed the results of the colorimetric assay (Fig. 5B). PED/PEA-15 can bind and retain into cytosol ERK1/2, independently of their phosphorylation status, thus impairing cell proliferation in PED/PEA-15 overexpressing cells (17,49). PED/PEA-15 phosphorylation on Serine-104 inhibits its interaction with ERK1/2, thus allowing cell proliferation even in cells expressing high levels of PED/PEA-15 [49]. Therefore, we investigated whether 67LR-mediated cell adhesion to LM could alter PED/PEA-15 phosphorylation status, thus allowing the observed recovery of PED-293 cell proliferation. PED/PEA-15 phosphorylation after PED-293 cell adhesion to LM or BSA, as a negative control, was analysed. After 1 hr of cell adhesion to LM, PED/PEA-15 phosphorylation occurred both in Ser-104 and in Ser-116 and further increased at 24 hrs (Fig. 5C). Similar results were obtained after PED-293 cell adhesion to YIGSR, the LM peptide which specifically binds 67LR (Fig. 5D), demonstrating that PED/PEA-15 phosphorylation, likely responsible for the proliferation recovery of PED-293 cells on LM, can be specifically induced by 67LR-mediated binding to LM.
Then, we investigated whether increased 67LR-mediated binding to LM in PED/PEA-15 overexpressing cells could generate the signal leading to PED/PEA-15 phosphorylation. PED/PEA-15 is an endogenous substrate for PKC, calcium/calmodulin-dependent protein kinase II (CAM kinase II), and Akt. Therefore, PKC, CAMKII and Akt phosphorylation, after V-293 and PED-293 cells adhesion to LM, was evaluated by Western blot analysis. Akt was not activated whereas CAMKII was phosphorylated after 1 hr and still at 24 hrs; PKC was transiently phosphorylated at 1 hr (Fig. 6A).
Therefore, 67LR binding to LM stimulates PED/PEA-15 phoshorylation in Ser-104, through PKC activation, thus enabling cell proliferation in response to LM in PED/PEA-15 overexpressing cells.
Because PED/PEA-15 phosphorylation in Ser-116, through CAMKII activation, induces its recruitment to the death-inducing signalling complex (DISC) causing the inhibition of the apoptotic process [6], we investigated the apoptotic response to serum deprivation in PED-293 cells, following adhesion to LM, which induces PED/PEA-15 phosphorylation. V-293 and PED-293 cells were plated on LM in the absence of serum and apoptosis was evaluated by ELISA assay. No apoptosis could be detected after 24 hrs of serum deprivation (not shown); at 48 hrs, V-293 cells underwent apoptosis whereas PED-293 cells showed a significant resistance (Fig. 6B). Noteworthy, cell-treatment with anti-67LR antibodies restored, at least in part, PED-293 cell sensitivity to serum deprivation-induced apoptosis. Resistance to apoptosis after 48 hrs of PED-293 cell adhesion to LM was also documented by the strong reduction of caspase 3 cleavage, as compared to V-293 cells (Fig. 6C).
Therefore, 67LR-dependent binding to LM, through PED/PEA15 phosphorylation, besides restoring cell proliferation, is also able to inhibit the apoptotic process in PED/PEA-15 overexpressing cells.

Discussion
PED/PEA-15 is an anti-apoptotic protein whose expression is increased in several human cancers. PED/PEA-15, being a member of the DED protein family, induces resistance to TRAIL-, FasLand TNF␣-mediated death in cancer, making therapy often ineffective. PED/PEA-15 is also a potent modulator of MAPK-signalling cascades. It binds ERK1/2 in the cytosol, avoiding their nuclear translocation and thereby reducing the ERK1/2-mediated transcriptional activity and cell proliferation.
To give new insights in its role in cancer, we aimed to find new PED/PEA-15 interactors. We performed a yeast two hybrid screening and identified several new candidate partners of PED/PEA-15, one of which was the 67LR. 67LR is a non-integrin cell surface receptor for the extracellular matrix whose expression is highly increased in human cancers and widely recognized as a molecular marker of metastatic aggressiveness [30].
We confirmed the interaction between PED/PEA-15 and 67LR through pull down and immunoprecipitation experiments, using both exogenous and endogenous proteins.
Then, we investigated the effects of PED/PEA-15 overexpression, occurring in different human tumours, such as breast and lung cancer, on 67LR-mediated cell functions. To this end, we evaluated cell adhesion and migration to LM in PED/PEA-15 transfected HEK 293 cells. PED/PEA-15 overexpression promoted cell adhesion and migration to LM, the major component of basement membranes [19], via the 67LR. The specific involvement of 67LR was unequivocally demonstrated by cell adhesion assays to an LM-derived pentapeptide, YIGSR, which specifically binds 67LR and by inhibition experiments with anti-67LR antibodies. Antibodies directed to LM-specific ␣6 and ␣3 integrin chains failed to abolish PED-293 cell adhesion to LM, confirming that increased binding to LM was mostly mediated by 67LR.
However, we also propose the involvement of ␤1-integrins in the generation of the downstream signal, because we observed that anti-␤1-integrin antibodies exerted an effect similar to 67LR inhibition. In fact, 67LR is co-expressed and physically interacts with ␣6-containing integrins on the cell membrane [29] and PED/PEA-15 has been largely shown to functionally regulate integrin activation and integrin-dependent functions [6,15,50]. Therefore, PED/PEA-15 overexpression increases cell adhesion to LM through its interaction with 67LR, which is both able to bind LM and to recruit integrins, thus leading to an increased intracellular signalling. This mechanism regulates also the activity of other non-integrin cell surface receptors for the extracellular matrix, for instance the urokinase receptor, which is also a vitronectin receptor and requires integrins as signalling partners in cell adhesion and migration [51].
Then, we investigated the effects of cell adhesion to LM on the cell signalling. In NSCLC cells, PED/PEA-15 is associated with Rac1 and promotes its activation in response to growth factor stimulation [17]; in turn, Rac1 increases ERK1/2 phosphorylation [47]. We showed that LM stimulation of PED/PEA-15 overexpressing cells activates Rac1 and ERK1/2. Therefore, a PED/PEA-15dependent signal transduction pathway, usually activated in response to growth factor stimulation, can be also activated after 67LR-mediated cell adhesion to basement membranes, a prerequisite for tumour invasion and metastasis.
Accordingly with our results, it has been recently reported that 67LR is overexpressed in astrocytoma and its down-regulation reduces the migratory activity of human glioma cells [52]. Findings on PED/PEA-15 are more controversial; it has been described a PED/PEA-15-mediated inhibition of cell migration/invasion in astrocytoma [16] whereas an increase has been shown in NSCLC [17]. Our study is in agreement with the last report; it is conceivable that this apparent discrepancy could be due to different regulatory mechanisms selectively controlling motility in a cell type-specific manner.
We then investigated whether the 67LR could also be involved in an LM-dependent modulation of PED/PEA-15 cellular functions, such as cell proliferation and apoptosis. It has been reported that PED/PEA-15 overexpression inhibits ERK1/2 nuclear localization, thus blocking cell proliferation; PED/PEA-15 phosphorylation on Ser-104, induces ERK release to the nucleus and enables cell proliferation, even in the presence of high levels of PED/PEA-15 expression [49]. We demonstrated that 67LR-mediated cell adhesion to LM induces an intracellular signal that, through PCK and CaMKinase II activation, promotes PED/PEA-15 phosphorylation on Ser-104, thus enabling cell proliferation even in PED/PEA-15 overexpressing cells.
Interestingly, 67LR, such as PED/PEA-15, is involved in the regulation of cell proliferation and survival. Indeed, reduction of 67LR expression results in apoptosis [53,54]; on the contrary, 67LR-dependent cell signalling pathways are important for cell survival [55,56]. In agreement with these reports, we showed that 67LR-mediated cell adhesion to LM also determines PED/PEA-15 phosphorylation on Ser-116, which promotes resistance to apoptosis induced by growth factor deprivation [44].
In summary, when overexpressed, PED/PEA-15 increases cell adhesion and migration to LM via the 67LR, thus promoting LMderived motility signals, which are crucial in intra-vasation and extra-vasation of cancer cells, during invasion and metastasis. Moreover, in PED/PEA-15 overexpressing cells, increased 67LRmediated cell adhesion to LM activates a signal transduction pathway that restores cell proliferation and promotes resistance to apoptosis, thus also contributing to cancer cell survival in a poor microenvironment, as it occurs during metastatic spread and colonization.