Intracellular phospholipase A₂ expression and location in human macrophages: Influence of synthetic material surface chemistry

Preparation of material surfaces

The model PCNUs used in this study were synthesized with a PCN soft segment, BD chain extender and either HDI or MDI as the diisocyanate as described in detail previously (Tang et al., 2001). Three PCNUs were used for this study (Table 1); (1) a PCNU containing a linear aliphatic diisocyanate (HDI) with a stoichiometric ratio of 4:3:1 (HDI/PCN/BD—referred to as HDI431), (2) a HDI-containing PCNU with a stoichiometric ratio of 3:2:1 (HDI/PCN/BD—referred to as HDI321), and (3) a PCNU containing an aromatic diisocyanate (MDI) with a stoichiometric ratio of 3:2:1 (MDI/PCN/BD—referred to as MDI321). The final partial structure of each model PCNU used, in addition to the structure of the reference TCPS surface are presented in Figure 1.

The basis for selection of these model PCNU materials arises from previous studies analyzing these model HDI- and MDI-based PCNUs with varying ratios of hard segment (HDI or MDI and BD) to soft segment (PCN) in addition to comparing hard segment chemistry differences (HDI vs. MDI). Although HDI- and MDI-based PCNUs are relatively similar in chemical structure (see Fig. 1), surface characterization data previously reported on may reflect relative differences in stability of these materials and “activatibility” towards macrophages illustrated previously (Labow et al., 2001a, 2005a; Matheson et al., 2002, 2004; Dinnes et al., 2005). X-ray photoelectron spectroscopy (XPS) analysis of HDI431, HDI321, and MDI321 (Tang et al., 2002) using C1s high resolution at an incident angle of 15°, generated results regarding the content of carbonate (OCOO) and urethane (NHCOO) groups for the top 1 nm of each material. Comparing the NHCOO/OCOO ratio can estimate the potential for interacting groups via hydrogen bonding. NHCOO groups have a greater potential to form hydrogen bonds. Therefore, MDI-based PCNU, which showed a greater NHCOO/OCOO ratio than HDI-based PCNUs, will be more stable due a greater number of NHCOO groups from the hard segment that is also composed of the large aromatic moieties providing steric hinderance to hydrolytic cleavage. Data supporting that MDI321 is a more structured material at the surface relative to HDI431 and HDI321 include atomic force microscopy (AFM) phase images (Revenko et al., 2001). AFM images suggest greater phase mixing with an MDI-based PCNU therefore allowing hydrogen bonding to occur, making MDI321 more resistant to degradation by hydrolysis. Furthermore, the smaller aliphatic HDI moieties allow access to enzymes for hydrolytic cleavage. Studies have illustrated that HDI-based PCNU are highly degradable by both hydrolytic enzymes (Labow et al., 2002; Tang et al., 2002) and macrophages (Matheson et al., 2004; Labow et al., 2005a). In order to compare these degradable material surfaces to a “reference” material, TCPS (non-degradable) was analyzed in this study since this is a material surface used most often in the scientific community for growth and cultivation of numerous cell types. Water contact angle data generated for HDI431, HDI321, and MDI321 (Tang et al., 2003) were not significantly different, however, indicated that each of these materials is relatively hydrophobic in comparison to typical TCPS. Contact angle data have not been previously reported on by our group, however are included here to indicate the relative hydrophobicity of these model PCNUs in comparison to TCPS surfaces used in standard cell culture. Water contact angle analysis was performed as outlined in detail previously (Tang et al., 2003), introducing a bead of HPLC-grade water to the TCPS cover slip surface and using a standard goniometer apparatus (NRL Goniometer, Rame-Hart, Inc., Mountain Lake, NJ). The contact angle data of TCPS relative to previously published contact angle results for HDI431, HDI321, and MDI321 (Tang et al., 2003) are presented in Table 2.

The synthesized PCNUs were dissolved overnight in a 10% dimethylacetamide solution (w/v), centrifuged and filtered (0.45 µm Teflon filter; Chromatographic Specialties, Toronto, ON, Canada), then coated as a thin layer (100 µl) onto 15 mm diameter glass cover slips (Fisher Scientific, Ottawa, ON, Canada) under sterile conditions in a laminar flow hood, as previously described (Labow et al., 2001b). The slips were dried overnight at 50°C, purged for 24 h and then dried under vacuum for 72 h. Prior to using the PCNU slips or TCPS multiwell plates for an experiment, they were hydrated by incubating with phosphate buffered saline (PBS; 0.0027 M KCl, 0.0015 M KH₂PO₄, 0.137 M NaCl, 0.008 M Na₂HPO₄) at 37°C and 5% CO₂ with 100% humidity overnight. This process equilibrates the polymer slips with an aqueous environment and removes trace solvent that may possibly interfere with the experiment.

U937 cell culture

U937 cells (American Type Culture Collection (ATCC™ catalogue # CRL—1,593.2)) were employed as a well characterized model for MDM-mediated PCNU biodegradation (Matheson et al., 2002). U937 cells were maintained as a promonocytic cell suspension and fed every second day with RPMI 1640 medium containing 10% fetal bovine serum (Invitrogen Canada, Inc., Burlington, ON, Canada), 1 mM sodium pyruvate (Invitrogen) with 100 U/ml penicillin and 0.1 mg/ml streptomycin (Invitrogen), as described in detail previously (Matheson et al., 2002). Only cells that were between passage 7 and 20 were used for differentiation into macrophage-like cells by culturing a 1 × 10⁶ cells/ml suspension in TCPS 12-well plates with phorbol 12-myristate 13-acetate (PMA) (1 × 10⁻⁷ M) for 72 h. At this time point cell division had ceased as determined by DNA analysis (see below), the cells were adherent and the parameters (maximal protein synthesis and esterase activity) that characterize a fully differentiated U937 cell for this in vitro model system occurred (Matheson et al., 2002). The cells were removed from the TCPS surface by gentle pipetting and were resuspended at a concentration of 1 × 10⁶ or 2 × 10⁶ cells/ml and reseeded onto the required material surface for the experiment.

Effect of aristolochic acid (ARIST) on protein expression

Differentiated U937 cells were re-seeded (1 ml of a 2 × 10⁶ cells/ml suspension) into 24-well plates alone (TCPS control) or into wells containing HDI431-, HDI321-, or MDI321-coated glass slips. The cells were allowed to adhere for 1 h after which the medium was replaced with or without the sPLA₂ inhibitor ARIST [200 µM] (Rosenthal et al., 1989; Dinnes et al., 2005). The cells were then incubated further for 48 h after which the cells were lysed for DNA determination (see below) and prepared for immunoblot analysis.

DNA determination

Whole cell lysates were prepared and analyzed for DNA as described in detail previously (Labow et al., 2001b; Matheson et al., 2004). Cold Triton X-100 buffer (180 µl) (0.05% Triton-X 100, 10 mM EDTA in PBS) was added to each well and the cells were lysed for 1 h on ice. Hoechst dye (number 33258; Amersham Biosciences., Baie d'Urfe, QC, Canada) was diluted with TRIS buffer (0.01 M Tris, 0.001 M EDTA, 0.2 M NaCl, pH 7.4) immediately prior to analysis. Cell lysates collected for DNA (10 µl) were added to the dye (100 µl) in 96 well microtiter plates (Microfluor 2 Black; VWR, Mississauga, ON, Canada). The lysates were compared to a DNA standard (Calf Thymus DNA) and each standard well to be analyzed also contained the same amount of 0.05% Triton X-100 buffer as in the lysate samples. The samples were analyzed using a fluorescence microplate reader (POLARstar Galaxy, BMG Labtechnologies, Durham, NC) with an excitation and emission wavelength of 360 and 460 nm respectively.

Immunoblotting

Proteins from whole cell lysate samples were separated by SDS–PAGE using a Protean II Cell System (Bio-Rad Laboratories Ltd., Mississauga, ON, Canada) with 5% stacking and 10% (cPLA₂), 12% (β-actin, glyceraldehyde-3-phosphate dehydrogenase (GAPDH), MSE) or 15% (sPLA₂) separating gel. U937 cells stimulated with PMA are non-dividing, terminally differentiated cells, therefore samples for electrophoresis were loaded into each well based on the amount of DNA in each sample. Two hundred and fifty nanograms of DNA were loaded into each well except for gels that were immunoblotted for cPLA₂ and sPLA₂, which required 1,000 ng of DNA to be loaded per well (as a result of the low abundance of these proteins). GAPDH was used as an additional control to ensure equal loading of protein in each well. Proteins from the resulting gels were transferred to nitrocellulose membranes (0.45 µm, Bio-Rad) for 75 min at 130 V. The membranes were then blocked for 1 h in Tris-buffered saline (TBST; 10 mM Tris, 0.1 M NaCl, 0.1% Tween-20, pH 7.5) containing either 5% non-fat dry skim milk or 5% bovine serum albumin. The membranes were then probed for 1 h with rabbit anti-cPLA₂ (Research Diagnostics, Inc., Concord, MA), mouse anti-sPLA₂ (group V) (Cayman Chemical Co., Ann Arbor, MI), rabbit anti-porcine esterase (Rockland Immunochemicals, Inc., Gilbertville, PA—reacts with MSE), mouse anti-β-actin or mouse anti-GAPDH (Chemicon International, Inc., Temecula, CA) and subsequently washed eight times with TBST. The membranes were then incubated with either anti-rabbit IgG (Rockland) or anti-mouse IgG (Pierce Biotechnology, Inc., Rockford, IL) conjugated to horseradish peroxidase, for 1 h followed by eight washes with TBST. Protein bands were visualized using a chemiluminescence detection kit (Pierce). The intensity of the protein bands were quantified using Quantity One® software (Bio-Rad).

Immunofluorescence and laser scanning confocal microscopy

Differentiated U937 cells, as described above, were re-seeded at a concentration of 1 × 10⁶ cells/ml into 24-well plates containing either HDI431-coated glass slips or TCPS cover slips (Sarstedt, Inc., Montreal, QC, Canada) and allowed to adhere for 30 min or 2 h. The adherent cells were washed three times with warmed PBS and fixed at room temperature with 4% paraformaldehyde (Fisher Scientific) in PBS for 10 min. The cells were then blocked for 1 h with 10% rabbit serum (Cedarlane Laboratories Ltd., Hornby, ON, Canada) in PBS, followed by a 1 h incubation with mouse anti-cPLA₂ (group IVA)(Santa Cruz Biotechnology, Inc., Santa Cruz, CA) in 5% rabbit serum in PBS. The slips were then washed three times in 5% rabbit serum in PBS followed by incubation for 45 min with rabbit anti-mouse Alexa Fluor® 488 (Invitrogen). The slips were finally washed three times in PBS and mounted onto microscope slides with 90% glycerol. The cells were visualized with an Olympus 100X oil immersion objective with a numerical aperture of 1.4, on an Olympus IX80 laser scanning confocal microscope operated by FV1000 software (version 1.4a). The Alexa Fluor® 488 fluorochrome was excited with the 488 nm line of the argon multi-line laser.

Statistical analysis

Statistical analyses were performed using a one-way ANOVA. The SAS® program (version 9.1) for windows was used and values were considered significant with a P-value of <0.05.

Specific inhibition of intracellular sPLA₂ protein expression by ARIST

Previous studies compared several chemical inhibitors of PLA₂ to assess their effect on U937 cell-mediated ¹⁴C-PCNU degradation and inhibition of PCNU-mediated U937 cell ³H-AA release (Dinnes et al., 2005). The sPLA₂ inhibitor ARIST demonstrated the greatest reduction of both U937 cell-mediated ¹⁴C-PCNU degradation and PCNU-induced ³H-AA release by U937 cells. To assess the specificity of ARIST towards sPLA₂ versus cPLA₂, immunoblot analysis for intracellular cPLA₂ and sPLA₂ protein from lysates of cells cultured with and without ARIST was performed to determine the effect of ARIST on intracellular protein expression. Figure 2A shows representative blots of cell lysates from U937 cells cultured on each material surface with and without ARIST for 48 h. Lanes 1, 3, 5, and 7 of each blot represent U937 cell lysates from TCPS, HDI431, HDI321, and MDI321 respectively from cells cultured with media alone. Lanes 2, 4, 6, and 8 of each blot represent U937 cell lysates from TCPS, HDI431, HDI321, and MDI321 respectively from cells cultured with ARIST. The first blot demonstrates lysates probed with anti-cPLA₂ and the second blot probed with anti-sPLA₂. ARIST had no significant effect on cPLA₂ protein expression as indicated by quantification of blots from multiple experiments (Fig. 2B). However, as seen in the second blot (Fig. 2A), ARIST had a significant effect on sPLA₂ protein expression in U937 cells adherent to TCPS, HDI321 and MDI321 each demonstrating 50–60% banding intensity of sPLA₂ protein relative to cells cultured with media alone (Fig. 2B). ARIST had no significant effect on sPLA₂ protein expression in U937 cells adherent to HDI431.

In order to ensure that ARIST was not eliciting non-specific effects on key intracellular proteins, immunoblot analysis was performed on these U937 cell lysates to probe for β-actin and MSE. Immunoblot analysis determined that ARIST did not have significant effects on intracellular MSE protein as seen in Figure 2A (third blot) and quantified in Figure 2C. No significant effect on β-actin expression (fourth blot) was detected as outlined in the quantification of blots from multiple experiments (Fig. 2C).

Effect of PCNU materials on intracellular cPLA₂ and sPLA₂ protein expression

Immunoblotting analysis of lysates from U937 cells cultured for 2 or 48 h on PCNU (HDI431, HDI321, and MDI321) and TCPS was performed to probe for group IVA cPLA₂ and group V sPLA₂. These time points were chosen based on established biodegradation assays (Labow et al., 2001a, b; Matheson et al., 2002; Dinnes et al., 2005) in which cells were re-seeded onto the candidate surface for 1 h (since the majority of cells have become adherent by approximately 1 h post re-seeding). The culture medium was then replaced with or without the reagent to be tested and then incubated further for 48 h for degradation analysis. Representative immunoblots showed that at 2 h post re-seeding to PCNU surfaces, cPLA₂ protein expression was increased relative to cells cultured on TCPS (Fig. 3A) for all three PCNUs. Two hours after U937 cell attachment to each PCNU surface, cPLA₂ protein was significantly increased by 3- to 3.5-fold on each PCNU relative to TCPS. By 48 h post cell attachment the expression of cPLA₂ protein, although still increased on PCNU relative to TCPS, was no longer significant on all surfaces (data not shown). Representative sPLA₂ blots (Fig. 3A), verified by quantification from multiple experiments (Fig. 3B), indicated that intracellular sPLA₂ protein expression was not significantly altered by the material surface to which U937 cells were adherent.

PCNU materials induce early changes in cPLA₂ protein expression

Group IVA cPLA₂ protein expression was clearly altered by the material surface to which U937 cells were adherent (Fig. 3A). Since there was such a large increase (∼3-fold) in cPLA₂ protein expression at 2 h post cell attachment, earlier time points were investigated to determine how early this response was initiated by cPLA₂ as a result of cell attachment to PCNU. At 2 h, the altered cPLA₂ protein expression was comparable on each model PCNU material (Fig. 3B), therefore further experiments exploring earlier time points of PLA₂ protein expression were carried out between only HDI431 and TCPS. Previous studies which demonstrated that PCNU materials induced increased ³H-AA release compared to TCPS illustrated that the greatest PLA₂ activation occurred within 10 min of U937 cell attachment (Dinnes et al., 2005). Accordingly, a time course of protein expression between 10 and 60 min post cell attachment to HDI431 and TCPS was analyzed to assess the initial point at which PCNU had the ability to induce detectable changes in either cPLA₂ or sPLA₂ protein expression. Further to the results from 2 h, sPLA₂ intracellular protein expression (Fig. 4A) was not significantly altered in HDI431 adherent U937 cells compared to TCPS (Fig. 4B). However, a slight increase in cPLA₂ protein expression was detectable as early as 10 min post cell attachment to HDI431 compared to TCPS (Fig. 4A) with significantly increased expression (∼2- to 2.5-fold increase) at 20 and 30 min (Fig. 4B). cPLA₂ protein expression returned to comparable expression levels seen in TCPS-adherent U937 cells by 45 min post cell attachment (Fig. 4A,B).

cPLA₂ location was altered by U937 cell adherence to a PCNU surface

Considering the significant changes in group IVA cPLA₂ expression in response to material chemistry differences of the substrate on which U937 cells were cultured, cellular location of cPLA₂ was assessed using laser scanning confocal microscopy to investigate material surface influences on cPLA₂ protein mobilization. U937 cells that were cultured for either 30 min or 2 h on TCPS or HDI431 were immunofluorescently labeled with anti-group IVA cPLA₂ and visualized by laser scanning confocal microscopy. These time points, 30 min and 2 h respectively, were chosen based on the fact that a peak of cPLA₂ expression was detected at 30 min (Fig. 4B) and 2 h (Fig. 3B) on PCNU relative to TCPS. Fluorescent images illustrated a change in location of cPLA₂ in response to adherence to HDI431 compared to TCPS. At both 30 min and 2 h, cPLA₂ was located diffusely throughout the cytoplasm in TCPS adherent cells (Fig. 5A,C respectively). In contrast, U937 cells that were adherent to HDI431 for both 30 min and 2 h exhibited a change in location of cPLA₂ towards the cell membrane (Fig. 5B,D respectively) compared to U937 cells adherent to TCPS (Fig. 5A,C).

Specific inhibition of intracellular sPLA₂ protein expression by ARIST

Chemical inhibitors are often not as specific as claimed therefore experiments were carried out to determine if the inhibition of U937-cell mediated ¹⁴C-PCNU degradation elicited by ARIST (Dinnes et al., 2005) was a consequence of non-specific effects on protein expression. Particularly, expression of proteins implicated in macrophage-mediated degradation processes in addition to both sPLA₂ and cPLA₂ protein expression was evaluated. β-actin is a key cytoskeletal protein in many cell types but particularly motile cells such as macrophages (Buccione et al., 2004). Since β-actin is so central to cell shape and attachment, as well as cell spreading and adhesion (Dadsetan et al., 2004; Dinnes et al., 2007), protein expression of β-actin was assessed by immunoblotting to ensure no significant influences on β-actin protein expression were occurring in response to this inhibitor. Target substances can be released from macrophages into the extracellular space formed between the cells and the material surface that may contribute to extracellular matrix/biomaterial degradation (Wright and Silverstein, 1984). Candidate enzymes that have been well studied and implicated in macrophage-mediated PU degradation are esterases, specifically MSE (Labow et al., 2001a, 2002, 2005b). MSE is an esterase that has demonstrated high degradative capacity toward PCNU materials, in addition to being newly synthesized and secreted from MDM in response to cell attachment to these surfaces (Matheson et al., 2004). Since this enzyme is central to macrophage-mediated degradation of PCNU, MSE protein expression was assessed in cells cultured with ARIST to ensure inhibition of degradation seen previously was not a result of effects on MSE protein expression.

Since both β-actin and MSE protein expression were unaltered in the presence of ARIST (Fig. 2), it can be concluded that inhibition of ¹⁴C-PCNU degradation elicited by U937 cells was not a result of non-specific effects on β-actin or intracellular MSE and was due to the effect of ARIST on PLA₂. Supporting this result was the previous study where the pattern of inhibition by ARIST of ¹⁴C-PCNU degradation by U937 cells (Dinnes et al., 2005) was similar to the ARIST inhibition of sPLA₂ protein expression in HDI321 and MDI321 adherent cells (Fig. 2B). The more prominent effect on both ¹⁴C-PCNU degradation (Dinnes et al., 2005) and sPLA₂ protein expression (Fig. 2A,B) was demonstrated on PCNU with a stoichiometric ratio of 3:2:1 of the starting components, regardless of the hard segment chemistry (HDI or MDI), in comparison to HDI431. This suggests the presence of a hard segment dependant response, possibly related to the strong polar chemistry contained within the hard segment, since polar groups have been shown to influence both degradation and adhesion (Santerre et al., 2005).

The comparable results between ARIST influence on U937 cell-mediated ¹⁴C-PCNU degradation and inhibition of sPLA₂ protein expression without non-specific effects on cPLA₂ expression indicate that ARIST inhibition of ¹⁴C-PCNU degradation was indeed via sPLA₂ down-regulation. It is important to note that although ARIST inhibited U937 cell-mediated PCNU degradation, this was not due to the enzymatic activity of sPLA₂ on the material itself. No PCNU degradation, as measured by radiolabel release, occurred when sPLA₂ from bovine liver was incubated with ¹⁴C-PCNUs (data not shown). It is unclear what the inhibitory effect of ARIST on sPLA₂ expression had on the cells' response to the PCNU surfaces, since sPLA₂ expression was unchanged in U937 cells adherent to PCNUs (Fig. 3). Although ARIST is generally classified as inhibiting sPLA₂, at the concentration used in this study (Rosenthal et al., 1989; Dinnes et al., 2005) it may be a more general inhibitor of PLA₂ activity. In a study where monocyte chemotaxis to MCP-1 was assessed in the presence of 100 µM ARIST, >80% inhibition was observed (Carnevale and Cathcart, 2001). This process was then shown to involve cPLA₂ and iPLA₂ activity (Carnevale and Cathcart, 2001).

Effect of PCNU materials on intracellular cPLA₂ and sPLA₂ protein expression

As illustrated in Figure 3, each model PCNU stimulated an increase in cPLA₂ protein expression 2 h post-cell attachment in U937 cells, relative to cells on TCPS (control surface), with no significant effect on intracellular sPLA₂ protein expression. In addition, sPLA₂ protein could not be detected in conditioned media by immunoblot analysis (data not shown). Although sPLA₂ is typically a protein that is secreted in response to stimulation, sPLA₂ has also been indicated to be internalized via a PLA₂ receptor shown to be located on the plasma membrane (Ohara et al., 1995; Hanasaki and Arita, 2002), therefore potentially contributing again to the intracellular pool of sPLA₂ protein.

PCNU materials induce early changes in cPLA₂ protein expression

It is possible that the lack of induced expression of sPLA₂ protein observed at 2 h (Fig. 3) could have been due to an early burst of AA release occurring within minutes of cell attachment to the PCNU surface (Dinnes et al., 2005). At 2 h, there was no significant difference in the altered cPLA₂ protein expression in the cell lysates between each model PCNU material (Fig. 3B). Therefore further experiments exploring earlier time points of PLA₂ protein expression were only carried out for one PCNU (HDI431) and TCPS. HDI431 has been shown to be the most biodegradable model PCNU (Labow et al., 2002) in addition to inducing the greatest amount of macrophage fusion and multinucleation (Matheson et al., 2004), therefore likely to be the model PCNU able to initiate the greatest inflammatory response. As mentioned previously, surface characterization of the model PCNUs used here by XPS and AFM analysis (discussed above) illustrated that HDI-based PCNUs are more phase-separated than MDI-based PCNUs, with a greater ratio of OCOO groups at the surface of HDI-based PCNUs and smaller aliphatic HDI groups all contributing to less stability of this material to cleavage. Although these surface characteristics reflect the susceptibility of these materials to degradation by enzymes and macrophages, these surface differences between the PCNUs do not appear to be directing any differences in cPLA₂ protein expression in PCNU-adherent macrophages. However, the differences in cPLA₂ expression agree with the differences in water contact angle (relative hydrophobicity) on PCNUs versus TCPS (advancing and receding angle TCPS << PCNU), with greater cPLA₂ expression on the hydrophobic PCNU surfaces relative to the hydrophilic TCPS. TCPS is a hydrophilic surface, however, gas plasma treatment typical of TCPS surfaces for cell culture would further contribute to the large difference in water contact angle between TCPS and these PCNU surfaces.

The significant induction of cPLA₂ protein expression that was detected at 30 min post cell attachment to HDI431 versus TCPS (Fig. 4) that then diminishes again until 2 h can be explained by an immediate and delayed AA release response that has been reported due to cell activation (Fonteh and Wykle, 2004). This effect has been suggested to be mainly initiated by cPLA₂ and possibly sustained or enhanced by sPLA₂ (Murakami and Kudo, 2001). Regardless, these results have clearly demonstrated a role for cPLA₂ in the inflammatory signaling pathways induced in the macrophage foreign body response to PCNU materials.

cPLA₂ location was altered by U937 cell adherence to a PCNU surface

The role of cPLA₂, upon cell activation, is to hydrolyze membrane phospholipids to release AA as an initial precursor to several potent inflammatory mediators. cPLA₂ requires phosphorylation for activation in addition to possessing a requirement for µM levels of Ca²⁺ necessary for translocation to cell membranes (Qiu et al., 1998). Under resting conditions cPLA₂ is typically constitutively expressed and located throughout the cytoplasm (Schievella et al., 1995), as was indicated in TCPS adherent U937 cells (Fig. 5A,C). Although not confirmed through co-localization to cytosolic compartments, upon adherence to PCNU (HDI431), cPLA₂ appeared to localize to the edges of the cytoplasm perhaps clustering towards the plasma membrane (Figs. 4D and 5B). This would allow access to cell membrane phospholipids for AA release as has been extensively reported on (Dennis, 2000; Balsinde et al., 2002; Dinnes et al., 2005). Previous studies assessing various forms of cell activation have indicated localization of cPLA₂ to the plasma membrane (Durstin et al., 1994; Schievella et al., 1995; Klapisz et al., 1999), nuclear envelope (Fatima et al., 2005), endoplasmic reticulum (Schievella et al., 1995), in addition to the golgi apparatus (Evans and Leslie, 2004).