Requirement of Protein Kinase C (cid:1) Activation and Calpain-mediated Proteolysis for Arachidonic Acid-stimulated Adhesion of MDA-MB-435 Human Mammary Carcinoma Cells to Collagen Type IV*

Arachidonic acid (AA) stimulation of adhesion of human metastatic breast carcinoma cells to collagen type IV depends on the protein kinase C (PKC) pathway(s) and is associated with the translocation of PKC (cid:1) from the cytoplasm to the membrane. In the present study, we have further explored the role of PKC (cid:1) in AA-stimulated adhesion. PKC (cid:1) activation site serines 738/742 and autophosphorylation site serine 910 are rapidly phosphorylated, and in vitro PKC (cid:1) kinase activity is enhanced in response to AA treatment. Inhibition of PKC (cid:1) activation blocks AA-stimulated adhesion. A phosphorylated, truncated species of PKC (cid:1) was detected in AA-treated cells. This 77-kDa protein contains the kinase domain but lacks a significant portion of the regulatory domains. Inhibition of calpain protease activity blocks generation of the truncated protein, promotes accumulation of the activated, full-length protein in the membrane, and blocks the AA-mediated increase in adhesion. p38 MAPK activity is also required for AA-stimulated adhesion. Activation of PKC (cid:1) and p38 are independent events. How-ever, inhibition of p38 activity reduces calpain-medi-ated proteolysis of PKC (cid:1) and in vivo calpain activity,

lular adhesion to ECM proteins, such as collagen type IV, often increases metastatic potential (1). Cell-ECM interactions are dependent on cell-surface adhesion molecules, such as integrins and selectins, and are influenced by the microenvironment. We are interested in elucidating pathways through which factors in the microenvironment regulate adhesive properties of human tumor cells.
Arachidonic acid (AA) and its precursor, linoleic acid, are common dietary n-6 cis-polyunsaturated fatty acids (PUFA) that may be present in the extracellular microenvironment. Although there is controversy over the effects of fat in the human diet on the development of breast cancer, there is substantial evidence for an effect of cis-PUFAs in animal models of mammary tumorigenesis and metastasis (2)(3)(4)(5)(6). AA is also present in an esterified form in cell membrane phospholipids and can be liberated by the actions of phospholipases. Studies have shown that fatty acids or their metabolites can alter the adhesion of breast tumor and other cell lines to components of the ECM (7)(8)(9). For example, we have shown previously (10) that AA stimulates adhesion of metastatic human breast carcinoma MDA-MB-435 cells to collagen type IV and also to vitronectin and fibronectin.
In our earlier studies, we have also shown that adhesion to collagen type IV requires ␣ 1 ␤ 1 and ␣ 2 ␤ 1 integrins and functional protein kinase C (PKC) and mitogen-activated protein kinase (MAPK) signaling pathways (11,12). AA stimulation of adhesion is associated with the translocation of PKC⑀ and PKC from the cytosolic to the membrane fraction of cell lysates, indicative of enzyme activation (11). PKC⑀ and PKC belong to the calcium-independent, diacylglycerol-responsive novel PKC (nPKC) family (13). However, the lack of a typical pseudosubstrate domain and the addition of an amino-terminal hydrophobic region and large pleckstrin homology domain-containing region within the regulatory domain make PKC unique (14). In addition, the PKC consensus substrate site is quite different from that of other PKCs (15,16). Members of the PKC family are regulated, in part, through phosphorylation events, including phosphorylation at an activation site (17). Activation of PKC requires phosphorylation at serines 738/ 742 (mouse homologue PKD serines 744/748) (16,18). PKC can be activated through direct phosphorylation by PKC⑀ or PKC (19,20). Interestingly, it has been reported that PKC, but not PKC⑀, is present in prostate cancer cells in which linoleic acid acts as a pro-cancer agent (21). It has also been reported that activation of PKC⑀ is necessary for muscle cell adhesion to fibronectin (22). PKC has also been detected in complexes with cortactin and paxillin in ␤ 1 integrin-containing breast cancer cell invadopodia into ECM (23).
Members of the calpain family of calcium-dependent neutral proteases can activate conventional PKC (cPKC) family members (␣, ␤, and ␥) by limited proteolysis (24,25). PKC⑀ has also been shown to be proteolyzed by calpain in vitro (26). cPKCs are cleaved in the third variable region, releasing the kinase domain from the inhibitory pseudosubstrate domain (25). Cytoskeletal proteins are also calpain substrates, and calpain has been found to co-localize with focal adhesions (27). It is therefore not surprising that calpain inhibition or overexpression affects cell adhesion and migration. However, the effects observed following manipulation of calpain activity appear to be cell type-specific. For example, treatment of Chinese hamster ovary cells with calpain inhibitor results in increased adhesiveness, whereas calpain inhibition results in decreased adhesiveness of mast cells and appears to have no effect on adhesiveness of vascular smooth muscle cells (28 -30). Up-regulation of calpain protein expression or mRNA has been observed in metastatic prostate tumors and has been associated with lymph node metastases of malignant renal cell carcinomas (31,32).
MAPK signaling pathways are activated by a wide variety of stimuli and regulate a plethora of cellular responses. We have reported that of the three MAP kinase subfamilies, the extracellular signal-responsive kinases (ERK), the c-Jun amino-terminal kinases (JNK), and the p38 MAP kinases, only the p38 MAP kinase pathway is required for AA-stimulated adhesion of MDA-MB-435 cells to collagen type IV (12). p38 is regulated by phosphorylation that increases upon exposure to AA. Concomitant with the AA-induced increase in p38 phosphorylation, there is an increase in the phosphorylation and activity of the downstream effector, MAPKAPK2, and phosphorylation of the MAPKAPK2 substrate, HSP27 (12). PKC has been shown to play a role in the regulation of both the ERK and JNK pathways, but no association has been made between PKC and the p38 pathway (19,20,33,34).
In this study, we have examined the involvement of PKC in the AA-mediated enhancement of MDA-MB-435 cell adhesion to collagen type IV. We have also investigated the possible involvement of calpain in modulation of breast cancer cell adhesion and in proteolysis of PKC. Finally, because we previously found that multiple signaling pathways are required for the AA effect, we have begun to explore potential mechanisms of cross-talk between these pathways.
Cell Culture and Preparation-The human mammary adenocarcinoma cell line MDA-MB-435 was obtained from Dr. Janet Price (M.D. Anderson Cancer Center, Houston, TX). Cells were maintained in Eagle's minimal essential medium (Invitrogen), supplemented with 5% fetal bovine serum (Hyclone, Logan, UT), 1 mM sodium pyruvate, 2 mM L-glutamine, and 2ϫ minimal essential medium vitamin solution (Invitrogen). Subconfluent cells were harvested by a brief incubation with Versene (Invitrogen), washed twice with serum-free medium, and resuspended in serum-free medium at 3 ϫ 10 5 cells/ml (adhesion assay) or 2 ϫ 10 6 cells/ml. Cells were allowed to equilibrate in serum-free medium for 20 min at 37°C under 5% CO 2 .
Adhesion Assays-Adhesion assays were carried out essentially as described (10). 96-Well tissue culture plates were pre-coated with collagen type IV, poly-D-lysine, or BSA. Wells were washed with PBS (137 mM NaCl, 2.7 mM KCl, 1 mM KH 2 PO 4 , 7.4 mM Na 2 HPO 4 ), and nonspecific binding sites were blocked with BSA. Inhibitors were added to cells prepared as described above, and cells were incubated for an additional 30 min. AA (30 M) or vehicle was added to the cells, and 100 l of the cell suspension was added to each well. Cells were incubated for 45 min, and non-adherent cells were removed by washing with serum-free medium. Adherent cells were fixed with 6% (v/v) glutaraldehyde and stained with 0.5% (w/v) crystal violet. The dye in adherent cells was solubilized with 1% (w/v) SDS, and the absorbance at 595 nm was determined. After background binding to BSA was subtracted, results were standardized as a percentage of adhesion to poly-D-lysine.
Western Blotting-AA (30 M) or vehicle was added to cells prepared as described above, and cells were incubated for an additional 1 min to 2 h. Inhibitors were added to the cell suspension, and the cells were incubated for an additional 30 min prior to the addition of AA. Cells were collected, washed with PBS, and lysed in 50 mM Tris-HCl, pH 7.4, 1% (v/v) Nonidet P-40, 0.25% (w/v) deoxycholate, 150 mM NaCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml aprotinin, 1 g/ml leupeptin, 1 g/ml pepstatin, 1 mM Na 3 VO 4 , 1 mM NaF. Lysates were centrifuged at 16,000 ϫ g for 10 min at 4°C. The samples were resolved by SDS-PAGE and transferred to Immobilon-P membranes. Membranes were blocked in TBS-T (20 mM Tris, 137 mM NaCl, pH 7.6, plus 0.1% (v/v) Tween 20) with 5% (w/v) milk, incubated with primary antibody for 1 h to overnight, and incubated with an horseradish peroxidase-conjugated secondary antibody for 1 h. Immunoreactivity was detected using SuperSignal chemiluminescent substrate. Blots were stripped by incubation in 100 mM glycine, pH 2.9, for 30 min; they were subsequently rinsed with TBS-T, blocked, and probed.
In Vitro Kinase Assay-Cells were treated and lysed as above. PKC was isolated by immunoprecipitation with the PKD/PKC antibody (1:100 dilution) for 2 h at 4°C in lysis buffer. Immune complexes were recovered using protein A-Sepharose. Kinase assays were performed essentially as described by Waldron et al. (36). Immune complexes were washed once with lysis buffer and twice with kinase buffer (30 mM Tris, pH 7.4, 10 mM MgCl 2 , 1 mM dithiothreitol). Syntide-2 substrate peptide (5 g), [␥-32 P]ATP (0.4 Ci), and inhibitors, if applicable, were added in a final volume of 30 l of kinase buffer. Reactions were incubated for 10 min at 30°C and stopped by addition of 100 l of 75 mM H 3 PO 4 . Reactions (75 l) were spotted on Whatman P-81 phosphocellulose, washed with 75 mM H 3 PO 4 , dried, and quantitated in a scintillation counter. Results were normalized to levels of PKC immunoprecipitated from each sample.
Cell Fractionation-Cell fractionation was carried out as described previously (11). Briefly, cells were treated as above and washed with PBS. Cells were resuspended in lysis buffer (20 mM Tris, pH 7.5, 5 mM dithiothreitol, 250 mM sucrose, 2 mM EDTA, 10 mM EGTA, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin), sonicated with three 10-s pulses at 22.5 kHz, and centrifuged at 100,000 ϫ g for 45 min at 4°C. Supernatants were collected and saved as the cytoplasmic fraction. Pellets were resuspended in lysis buffer plus 1% Triton X-100, sonicated again, and centrifuged. The supernatants were collected and used as the membrane fraction. PKC was examined by Western blotting using equal protein, as described above; loading equal protein resulted in a 2-fold difference in cell equivalents between the cytosolic and membrane fraction, with the greater number of cell equivalents in the membrane fraction.
In Vivo Calpain Assay-Cells were harvested as above, resuspended in serum-free medium, and treated with inhibitors, as appropriate. Cells were incubated for 20 min with 10 M t-Boc-Leu-Met-CMAC, a cell-permeable fluorogenic calpain substrate. Aliquots (200 l) of cell suspension were loaded into wells of a 96-well plate, and fluorescence generated through calpain cleavage of t-Boc-Leu-Met-CMAC was analyzed using an F max microplate reader (Molecular Devices, Sunnyvale, CA) with a 355 nM excitation filter and a 460 nM emission filter.
In Vitro Calpain Assay-Cells were transfected with pcDNA-PKC (2 g/60-mm dish) using FuGENE. After 48 h, cells were treated and lysed as above, and PKC was immunoprecipitated as above. Immune complexes were washed 3 times, with the final wash in 50 mM Tris, pH 7.4. Calpain assays were performed essentially as described by Carragher et al. (37). Calpain I or calpain II (0.05 units) was added to immune complexes in calpain buffer (50 mM Tris, pH 7.4, 10 mM CaCl 2 , 30 mM NaCl, 5 mM ␤-mercaptoethanol; 40 l final volume) in the presence or absence of inhibitors. Reactions were incubated at 30°C for 30 min and stopped by the addition of SDS sample buffer. Cleavage of PKC was analyzed by Western blot as described above.

AA Promotes Activation of PKC in MDA-MB-435 Cells-
Subsequent to PKC/PKD cloning, many aspects of PKC biochemistry have been elucidated, but less has been ascertained about its biological functions. One characteristic of PKC is that activation requires phosphorylation at serines 738/742. To investigate the role of PKC activation in AA-mediated adhesion, we used an antibody specific for serine 738/742 phosphorylation. With this antibody, we detect multiple species of PKC, the origins of which are described below. Treatment of MDA-MB-435 cells with AA led to a rapid increase in serine 738/742 phosphorylation (Fig. 1A). Within 1 min of exposure to AA, phosphorylation of PKC increased, and PKC phosphorylation was stable for approximately 1 h. AA did not induce a change in the levels of total PKC protein during this time course (Fig. 1, A and B).
The activation site serines 738/742 are predominantly transphosphorylated, in this instance probably by PKC⑀ (36). However, PKC contains additional phosphorylation sites, including an autophosphorylation site at serine 910 (PKD serine 916) (38,39). Phosphorylation at serine 910 has been correlated with and used as a marker for PKC activity (38,40,41). A serine 910 phosphorylation-specific antibody revealed AA-induced autophosphorylation of PKC that is stimulated as rapidly as activation site phosphorylation (Fig. 1B).
To further establish AA as a modulator of PKC activity, we examined the activity of PKC purified from MDA-MB-435 cells. PKC from AA-stimulated cells showed an increase in in vitro kinase activity within 2 min of AA treatment. The enhancement of activity peaked at an average of 207% of the activity of PKC from vehicle-treated cells at ϳ5 min and extended to at least 10 min. In addition, treatment with higher concentrations of AA led to further stimulation of PKC (data not shown). These results, along with the translocation of PKC, support the conclusion that AA activates PKC.
Effect of PKC Inhibitors on AA-mediated Adhesion to Collagen Type IV-PKC inhibitors Gö6976 and Gö6983 inhibit the activity of PKC in vitro at vastly different concentrations. Gö6976 inhibits recombinant PKC from baculovirus-infected cells in the nanomolar range, whereas Gö6983 inhibits PKC in the micromolar range; similar concentrations inhibit in vitro activity of PKC purified from MDA-MD-435 cells (42) (data not shown). The capacity of these inhibitors to block PKC activity in MDA-MB-435 cells was examined by treating the cells with inhibitor and then stimulating with AA. The results in Fig. 2 demonstrate that both inhibitors block PKC activity, as measured by autophosphorylation. These inhibitors did not affect levels of total PKC protein (see Fig. 7B). Subsequently, we examined the abilities of Gö6976 and Gö6983 to alter AAstimulated adhesion of MDA-MB-435 cells to collagen type IV. Cells were incubated with inhibitor for 30 min prior to treatment with AA and plating on collagen type IV. Both inhibitors blocked the AA-mediated increase in adhesion (Fig. 3). A significant decrease in the response to AA was obtained with 20 nM Gö6976, and nearly complete inhibition was seen at 80 nM. Similar levels of inhibition of adhesion required 10 -20 M Gö6983. The inhibitor solvent (Me 2 SO) did not decrease the AA-stimulated adhesion (data not shown).
Phosphorylated PKC Is Subject to Limited Proteolysis by Calpain at the Cell Membrane-The predominant serine 738/ 742-phosphorylated form of PKC that is detected after stimulation with AA is smaller than the full-length protein (Fig. 4). Given previous reports of calpain proteolysis of cPKC isoforms and PKC⑀ and the stability of this form of PKC (Fig. 1A), we examined the hypothesis that the truncated PKC is generated through limited proteolysis by calpain (24 -26). MDA-MB-435 cells were treated with a cell-permeable calpain inhibitor, either calpeptin (Fig. 4) or MDL-28170 (Fig. 5), prior to AA exposure. In the absence of calpain inhibitor, a 77-kDa activation site phosphorylated protein was detected, and the amount of phosphorylated protein significantly increased upon addition of AA (Fig. 4). In the presence of calpain inhibitor, the fulllength, phosphorylated protein was detected, and AA treatment increased the levels of phosphorylated, full-length PKC. Concomitant with the increase in full-length, phosphorylated protein observed when AA-treated cells were pretreated with calpain inhibitor, there was a decrease in the level of proteolyzed, phosphorylated PKC, indicating that the truncated protein was produced via calpain cleavage of the full-length protein (Fig. 4).
We have previously shown translocation of total PKC from the cytosolic fraction to the membrane fraction in lysates of MDA-MB-435 cells treated with AA (11). Cell fractionation revealed that proteolyzed, activated PKC was localized predominantly in the cytosolic fraction of cells treated with AA (Fig. 5). Pretreatment of cells with calpain inhibitor allowed detection of phosphorylated, full-length PKC; this form was also localized to the cytosolic fraction. When calpain inhibitortreated cells were exposed to AA, full-length, activated PKC accumulated in the membrane, and levels of the proteolyzed PKC decreased relative to the levels observed in cells not pretreated with calpain inhibitor (Fig. 5). These results indicate that activated, full-length PKC translocates to the membrane where it is converted to the 77-kDa form through limited proteolysis by calpain.
Calpain Activity Is Required for the AA-mediated Enhancement of MDA-MB-435 Cell Adhesion to Collagen Type IV-Calpain is active in MDA-MB-435 cells following AA stimulation; however, calpain activity may be unrelated to the increased adhesion that is triggered by AA. To determine whether calpain is required for enhancement of MDA-MB-435 cell adhesion, we analyzed cells that had been pretreated with calpain inhibitor for adhesion to collagen type IV upon exposure to AA. The addition of calpain inhibitor MDL-28170 decreased the AA-induced enhancement of cell adhesion to collagen type IV in a dose-dependent manner, with a significant decrease in adhesion seen at 75 M (Fig. 6A). The concentrations of calpain inhibitor used in these assays did not affect the ability of vehicle-treated cells to adhere to collagen type IV nor the ability of AA-or vehicle-treated cells to adhere to poly-Dlysine ( Fig. 6A and data not shown). As shown in Fig. 6B, the decrease in adhesion to collagen type IV parallels the increase in full-length, phosphorylated PKC that accumulates when cells that have been pretreated with calpain inhibitor are stimulated with AA. In fact, an approximate 50% increase in fulllength, phosphorylated PKC corresponds to an approximate 50% decrease in AA-stimulated adhesion.

Active PKC Is Not Required for Activation of p38 MAPK and Active p38 MAPK Is Not Required for Activation of PKC-We
reported previously (12) that in addition to the PKC pathway, p38 MAPK is required for the AA-induced increase in adhesion. Given the recent reports (19,20,33,34) of PKC effects on other MAPK pathways, we examined the effects of PKC inhibition on activation of p38 during AA stimulation. Cells were treated with inhibitor followed by AA, and cell lysates were analyzed for p38 activity by immunoblotting with an antibody that recognizes active (phosphorylated) p38. Pretreatment with the p38 inhibitor PD169316 led to inhibition of AA-stimulated p38 activation and to inhibition of phosphorylation of the downstream target HSP27, whereas pretreatment with the PKC inhibitor Gö6976 or Gö6983 did not affect p38 activation nor HSP27 phosphorylation (Fig. 7). Similarly, the non-functional analogue of the p38 inhibitor SB202474 did not affect p38 activation (Fig. 7). In a parallel experiment, we examined the potential effects of inhibition of p38 activity on the activation of PKC. p38 inhibitor PD169316 pretreatment of cells did not affect AA-induced autophosphorylation of PKC, whereas the PKC inhibitor Gö6976 or Gö6983 blocked PKC activity, indicating that activation of PKC and activation of p38 are independent events (Fig. 7).
Active p38 Is Required for Full Calpain Activity and Complete Proteolysis of PKC-The PKC and p38 pathways appear not to interact upstream from activation of these kinases. However, a downstream step in the adhesion process where signals from each pathway could merge is the cleavage of PKC by calpain. To examine this possibility, we analyzed the effects of p38 inhibition on cleavage of phosphorylated PKC. Incubation of cells with p38 inhibitor PD169316 prior to AA exposure resulted in a significant increase in the level of serine 738/742 phosphorylated, full-length PKC and a decrease in the level of phosphorylated, truncated PKC, compared with cells not ex-posed to inhibitor or cells treated with the non-functional analogue SB202474 (Fig. 8A). The amount of full-length, activated PKC detected in the presence of p38 inhibitor is not equivalent to the amount of full-length, phosphorylated protein observed when calpain is directly inhibited by using MDL-28170; however, these results indicate that active p38 plays a role in regulating PKC cleavage by calpain. In vitro calpain assays were used to examine the specificity of the p38 inhibitor. Both purified calpain I and purified calpain II proteolyze purified PKC in vitro (Fig. 8B). In the presence of calpain inhibitor MDL-28170, proteolysis is blocked; however, in the presence of p38 inhibitor PD169316 or the non-functional analogue SB202474, PKC is proteolyzed by calpain (Fig. 8B). These results indicate that the p38 inhibitor is not directly inhibiting calpain but is blocking the ability of p38 to regulate calpain.
Further evidence for a role of p38 in modulation of calpain activity was obtained by using an in vivo assay for calpain activity with a fluorogenic substrate. The cell-permeable calpain substrate, t-Boc-Leu-Met-CMAC, emits fluorescent light upon cleavage. Incubation of MDA-MB-435 cells with calpain inhibitor or with p38 inhibitor prior to addition of calpain substrate reduced the resulting fluorescence relative to control cells (Table I). Calpain inhibitor reduced fluorescence to levels approaching those seen in cells containing no substrate. Inhibition of p38 reduced calpain activity by 57%, whereas addition of the non-functioning analogue of the 38 inhibitor led to only a 23% reduction, suggesting that p38 is one regulator of calpain activity in MDA-MB-435 cells (Table I). DISCUSSION In these studies we have examined the signal transduction pathways that relay the information required for cis-PUFA promotion of breast cancer cell adhesion to ECM. We had shown previously (11) that PKC pathways were involved and that PKC⑀ and PKC were probable mediators of this signaling. Biological functions of PKC have not yet been well documented, but PKC is involved in transport through the Golgi and in regulation of apoptosis, DNA synthesis, and some MAPK signaling pathways (20,35,40,41,(43)(44)(45). In this report, we have further implicated PKC in regulation of the adhesion process, which is an important aspect of the metastatic cascade. We show that PKC rapidly becomes activated upon exposure of MDA-MB-435 breast cancer cells to AA and present the first evidence of a requirement for PKC in AAenhanced cell adhesion to ECM. We show for the first time that phosphorylated PKC is a target for calpain-mediated proteolysis and that calpain activity is also required for AA-enhanced breast cancer cell adhesion to ECM. Finally, we present data that suggest the regulation of calpain is partially controlled by p38 MAPK.
In response to AA, PKC is rapidly phosphorylated at activation site serines and acquires the capacity for autophosphorylation. These events lead to increased PKC kinase activity, as measured by in vitro assays. PKC inhibitors that block activity of PKC inhibit the enhancement of MDA-MB-435 cell adhesion imparted by AA. The inhibitor Gö6976 affects cPKCs as well as PKC; however, we have no evidence for activation of cPKCs by AA (42). Gö6976 does not affect other nPKCs, such as PKC⑀, at low concentration (reviewed in Ref. 38). Therefore, these results indicate that active PKC is required for adhesion. The specificity and concentration requirements of Gö6983 in vivo have been less well documented. This inhibitor is active on PKC in vitro only at micromolar concentrations, whereas it is active on cPKCs and some other nPKCs in vitro at nanomolar concentrations; PKC⑀ was not examined in that study (42). The high concentration of Gö6983 required for inhibition of adhesion of the MDA-MB-435 cells in response to AA implicates PKC as responsible for signal transduction. Immunoblotting for autophosphorylated PKC reveals that this concentration of Gö6983 does, indeed, block in vivo activity of PKC. However, it remains possible that the block in PKC activity is due to inhibition of activation of PKC through a block in the activity of PKC⑀. PKC is an in vitro substrate for PKC⑀; PKC and PKC⑀ form co-immunoprecipitable complexes in vivo, and overexpression of PKC⑀ or co-expression of PKC⑀ with the upstream phosphoinositide-dependent kinase leads to activation of PKC in HEK293 cells (20,46). PKC that has been activated by phosphorylation at activation site serines appears to translocate to the cell membrane and become rapidly proteolyzed. In some cell types, activated calpain localizes to the cell membrane (47,48). Inhibition of calpain blocks PKC proteolysis. Calpain cleaves cPKCs, releasing a truncated, active kinase, which has been termed PKM (25,49). More recently, calpain has been shown to cleave PKC⑀ in vitro, also resulting in release of a catalytically active fragment (26). Experiments have suggested that the activated form of cPKC is the preferred target for calpain-mediated proteolysis   (24,25). A consensus cleavage site for calpain has not been defined. Amino acid preferences determined by examination of calpain cleavage of peptides and histones do not account for calpain cleavage of all substrates; however, these studies have also suggested that calpain recognizes higher order structures (50). The region containing the calpain cleavage site of cPKCs does not have an exact PKC counterpart. cPKCs are cleaved in the V3 "hinge region" between the C2 calcium-binding domain region and the C3 kinase domain region (25). However, amino-terminal to the PKC kinase domain is the pleckstrin homology domain rather than a classical V3 region (14). Based on the size of the truncated protein, it appears that PKC is cleaved in or near the second cysteine-rich region. Given that calpain recognizes inter-domain regions of its protein substrates, we speculate that the calpain cleavage site for PKC resides at the amino or carboxyl terminus of the second cysteine-rich "zinc finger" region and that the truncated protein would contain the entire kinase domain. PKD mutants that lack one or both cysteine-rich domains are constitutively active kinases (51). There is some discrepancy between these findings and results of experiments performed with PKC; however, different experimental systems were employed (52). In addition, a 14-3-3 protein binding domain, which has been associated with inhibition of PKC in T-cells, would be removed upon truncation (53). These findings lead us to believe that the calpain-proteolyzed PKC is catalytically active. Our hypothesis is supported by the report that PKC that has been proteolyzed by caspase-3 is a constitutively active kinase. Caspase-3 cleaves PKC at amino acid 378, between the second cysteine-rich region and the pleckstrin homology domain, generating a smaller 60-kDa protein (54). As has been reported for caspase inhibitors, we have found the caspase inhibitor benzyloxycarbonyl-VAD-fluoromethyl ketone (Z-VAD-fmk) (caspases 1, 3, 4, and 7) to be nonspecific in vitro, i.e. this inhibitor will block calpain I or calpain II activity (55). 2 However, the appearance of the 77-kDa phosphorylated form of PKC/PKD in both MCF-7 cells, which lack caspase-3, and in NIH-3T3 cells, in which PKD lacks the caspase-3 cleavage site, indicates that the 77-kDa isoform is a distinct species (56). 2 Calpain activity is required for the AA-mediated enhancement of MDA-MB-435 cell adhesion to collagen type IV. Calpain inhibitor blocks the increase in adhesion of cells that have been treated with AA but does not affect the adhesive capacity of cells that have been treated with vehicle nor the ability of cells to bind to poly-D-lysine. Therefore, the decrease in adhesion due to calpain inhibition is specific for the AA-stimulated signaling pathway and not a result of nonspecific alterations in calpain-mediated effects on cytoskeletal proteins. Adhesion of human umbilical vein endothelial cells to prothrombin substrate is also blocked by treatment with either PKC inhibitor or calpain inhibitor; however, the calpain target is unknown (57). In addition, T-cell adhesion to fibronectin that has been induced by T-cell receptor cross-linking or 12-O-tetradecanoylphorbol-13-acetate treatment is sensitive to calpain inhibition. In this case, the decrease in adhesion has been linked to a block in proteolysis of protein-tyrosine phosphatase 1B (58).
PKC regulates the ERK and JNK pathways in some cell types but has not yet been reported to affect p38 (19,20,33,34). Given our previous observations that p38 is activated on exposure to AA and that both the p38 and PKC pathways are required for AA-enhanced adhesion, we examined these pathways for cross-talk (11,12). Activation of either kinase is not dependent on activity of the other, but proteolysis of PKC by calpain is regulated, in part, by active p38. In MDA-MB-435 cells, active p38 is required for at least a portion of calpain activity. This represents one step in the AA-stimulated adhesion process where signals from each pathway could merge. Calpain can be phosphorylated in vitro; however, it is not known how this modification affects activity (27,59).
In this report, we have presented evidence to support our hypothesis that PKC is involved in AA-stimulated adhesion of metastatic human mammary carcinoma cells to collagen type IV, and we have introduced calpain as a mediator of PKC signaling. In our current model (Fig. 9), AA stimulates activation of PKC⑀ via signaling by a lipoxygenase metabolite of AA. Active PKC⑀ then phosphorylates PKC at activation site serines. PKC translocates to the cell membrane where it is proteolyzed by calpain into a catalytically active 77-kDa molecule that is localized to the cytosol. p38 may play a role in funneling signals from AA to the PKC pathway through regulation of calpain activity. p38, which is also activated via AA signaling, may also mediate adhesion through activation of MAPKAPK2 and phosphorylation of HSP27, or through interactions with cytoskeletal proteins at focal adhesions. 3 Few in vivo PKC substrates are known, and the targets for  9. Model of the signaling pathways involved in the AAinduced increase in mammary tumor cell adhesion to ECM proteins. AA or its metabolite, hydroxyeicosatetraenoic acid (HETE), induces activation of PKC⑀, which, in turn, activates PKC. Activated PKC translocates to the cell membrane, where it undergoes proteolysis by calpain to become an active cytosolic kinase. Calpain is regulated, in part, by p38, which is also activated through AA signaling. Each of these enzymes, PKC, p38, and calpain, may be involved in the activation of ␤ 1 integrins and cytoskeletal rearrangements required for cell adhesion to ECM. PKC in this system have not yet been identified. We speculate that both the full-length and the truncated PKC isoforms may act as important signaling molecules in the adhesion process. Both full-length and truncated cPKC isoforms are catalytically active, and there is precedent for distinct activities of each kinase as described for PKC␣/PKM (60). PKC␣ and PKM function in alternatively activated signal transduction pathways that lead to similar phosphorylation events but have different signaling requirements (60). We expect that identification of the targets of PKC will further the understanding of both the AA-stimulated adhesion process and the biological functions of PKC. As we progress toward defining the signal transduction mechanisms that regulate AA-mediated increases in cell adhesion to ECM, and potentially, mechanisms that regulate metastasis, we hope to uncover steps that could be targeted for inhibition.