INTRODUCTION

Cytosolic Ca²⁺ signals activate numerous rapid cellular responses including contraction, secretion, and excitation; Ca²⁺ signals also mediate control over longer term responses such as cell division and growth (1, 2). A significant source of Ca²⁺ for these signals is the Ca²⁺ stored within intracellular pools, which can be released through the activation of intracellular release Ca²⁺ channels (1-3). Intracellular pools exist within the endoplasmic reticulum or subfractions thereof (1-4) and accumulate Ca²⁺ via the action of Ca²⁺ pumps of the sarcoplasmic/endoplasmic reticulum calcium ATPase (SERCA)¹ family, which are widely distributed within the endoplasmic reticulum of most cells (5, 6). In addition to serving as a source of Ca²⁺ for the generation of cytosolic Ca²⁺ signals, the Ca²⁺ accumulated within pools may effect control over a number of cellular functions. Thus, the intraluminal Ca²⁺ level appears to be the primary trigger for activating Ca²⁺ influx channels in the plasma membrane following Ca²⁺ pool release (1-3, 7). Intraluminal Ca²⁺ levels also appear to exert control over fundamental endoplasmic reticulum functions including the folding, processing, and assembly of proteins (8-10). These actions may be mediated by intraluminal Ca²⁺ binding proteins, which can function as molecular chaperones (11, 12).

In recent studies we have shown that the content of agonist-sensitive Ca²⁺ pools also exerts a profound effect upon the ability of cells to progress through the cell cycle (2, 4, 13-17). Experiments reveal that the SERCA pump inhibitors thapsigargin (18) and 2,5-di-tert-butylhydroquinone (19) deplete intracellular Ca²⁺ pools and cause the accompanying entry of DDT₁MF-2 smooth muscle cells into a stable quiescent G₀-like growth state (13, 14). Although growth-arrested, these Ca²⁺ pool-depleted cells remain intact and viable, and maintain normal cellular and subcellular morphology and mitochondrial function for up to one week (13, 14). Whereas SERCA pump inhibition by thapsigargin is essentially irreversible (20, 21), we previously demonstrated that treatment of thapsigargin-arrested cells with high (20%) serum induces reappearance of Ca²⁺ pools and activation of a transition of cells back into the cell cycle (14, 15). High serum induces expression of new functional Ca²⁺ pump protein within 1-3 h and the reappearance of agonist-releasable Ca²⁺ pools within 6 h (15). Cells begin to enter the S phase 16 h later and thereafter continue to divide normally (14).

Recently we revealed that each of the essential fatty acids, arachidonic acid, linoleic acid, and linolenic acid (22), can mimic the action of high serum treatment on Ca²⁺ pool-depleted growth-arrested cells, inducing reappearance of agonist-sensitive Ca²⁺ pools and re-entry of cells into the cell cycle. It was shown that either high serum or EFA treatment can induce recovery of Ca²⁺ pools, the actions of both being dependent upon protein synthesis (22). Experiments indicated that the EFAs within serum could be the component responsible for stimulating pool reappearance and growth recovery. Thus, the EC₅₀ values for the action of each of the three EFAs on growth induction of thapsigargin-arrested cells were similar, approximately 5 µM, corresponding with the total EFA concentration present in the 20% serum treatment conditions (22). Phospholipase A₂ inhibitors did not inhibit high serum-induced recovery suggesting that serum does not stimulate phospholipase A₂-induced formation of free EFAs. We postulated that the EFAs themselves may not be the agents directly involved in recovery, but instead, active metabolite(s) of these lipid species might be responsible. This was supported by studies showing that the nonmetabolizable analogue of arachidonic acid, 5,8,11,14-eicosatetraynoic acid (ETYA), was unable to induce recovery (22). ETYA has been shown to mimic the actions of arachidonic acid in systems where metabolism of the fatty acid is not required (23-25). Additionally, ETYA is an effective blocker of the entry of EFAs into each of the pathways through which eicosanoids are formed, including the cyclooxygenase, lipoxygenase, and epoxygenase (or monooxygenase) pathways (26). Our observations that ETYA is effective in blocking arachidonic acid-induced recovery provided further support for the possible requirement for metabolism of arachidonic acid in inducing cell recovery (22).

The metabolism of EFAs within cells is complex, and at least three major pathways are known (27). It was important to determine whether metabolism of EFAs through one of these pathways was required for the activation of recovery of pools and cell growth following Ca²⁺ pool emptying. In addition, it was important to try to identify whether any of the many products of metabolism might be active in mediating this response. Presented here are studies examining the role of each of the three major pathways of eicosanoid synthesis in mediating the action of EFAs. The results indicate that the cytochrome P-450 epoxygenase pathway is required to give rise to active products. Studies further reveal that the effect is induced by only two specific epoxyeicosatrienoic acid metabolites of arachidonic acid and not observed with other cytochrome P-450 metabolites or mimicked by products from other eicosanoid pathways. The results also provide evidence indicating that the EFAs contained within serum are indeed the active components mediating serum-induced recovery of Ca²⁺ pool-depleted cells. The results not only provide further information on the link between Ca²⁺ pools and cell growth but also shed light on a possible new signaling pathway controlling transition from a stationary to proliferative growth state.

EXPERIMENTAL PROCEDURES

DDT₁MF-2 smooth muscle cells derived from hamster vas deferens were cultured in Dulbecco's modified Eagle's medium supplemented with 2.5% serum (Calf-Plus, Inovar, Gaithersburg, MD) as described previously (28). Calf-Plus is new-born calf serum supplemented with additional growth factors and is referred to as "serum" in this report. All fatty acids and derivatives were added to cells in the presence of 1% (w/v) fatty acid free bovine serum albumin. This condition more closely resembled those under which free serum fatty acids interact with cells and prevented the membrane-destabilizing action of added fatty acids which occurs in the absence of the carrier protein (22).

DDT₁MF-2 smooth muscle cells were grown in 24-well dishes (1 × 10⁵ cells/well). Thapsigargin-treated cells were prepared by adding 3-8 µM thapsigargin in DMEM to otherwise normally cultured cells in 2.5% serum for 3 h. Thapsigargin-treatment was followed by 3 washes in thapsigargin-free DMEM with 2.5% serum followed by recovery under the conditions specified. Treatment with inhibitors of arachidonic acid metabolism began 30 min prior to the addition of recovery agents (arachidonic acid or high serum) at the concentrations specified and continued throughout the experiment. All recovery media contained DMEM with 2.5% serum together with additions as specified. For all cell proliferation experiments, the total time from the end of thapsigargin treatment until determination of cell number was 72 h. At the end of 72 h, cells were resuspended by pipette, transferred to cuvettes, and counted spectrophotometrically as described previously (22). For each experiment, standard curves were obtained for cell number (by direct counting) and light scattering was measured by absorbance at 600 nm. Absorbance values obtained at different dilutions of cells were compared with the linear portion of the standard curve and values for cell number were obtained. All measurements for cell number were obtained in quadruplicate, and the results presented are typical of at least three different experiments.

Cultured DDT₁MF-2 cells were allowed to attach to poly-L-lysine-coated 25-mm glass coverslips in culture for at least 4 h prior to use. Attached cells were treated with 3 µM thapsigargin for 3 h as described above, then, after washing in thapsigargin free DMEM with 2.5% serum,they were transferred to recovery medium as indicated in figures. Incubation with inhibitors, where noted, began 30 min prior to treatment with recovery medium and continued throughout the course of the experiment. Cells were incubated for 24 h under appropriate recovery conditions in DMEM with 2.5% serum, at the end of which Ca²⁺ levels were measured. Measurements of free cytosolic Ca²⁺ were similar to those previously described (29). Attached cells were transferred to Hepes-buffered Kreb's medium (107 mM NaCl, 6 mM KCl, 1.2 mM MgSO₄, 1 mM CaCl₂, 1.2 mM KH₂PO₄, 11.5 mM glucose, 0.1% bovine serum albumin, 20 mM Hepes-KOH, pH 7.4) and loaded with fura-2 acetoxymethylester (2 µM) for 10 min at 20 °C in the dark. Cells were then washed in the same medium and loaded dye was allowed to de-esterify for 15 min at 20 °C in the dark. Under these conditions, approximately 95% of the dye was confined to the cytoplasm as determined by the signal remaining after saponin permeabilization (29). Coverslips were inserted into a Dvorak-Stotler chamber (Nicholson Precision Instruments, Gaithersburg, MD) and cells were viewed with a Nikon Diaphot microscope equipped with epifluorescence optics (Nikon 40X UV-Fluor objective). Additions were made by aspiration and addition of fresh bathing solution in an open chamber configuration. Full agonist-mediated Ca²⁺ release from pools was activated by the addition of 10 µM bradykinin. Responses shown are for fields containing groups of between 10 and 15 cells. Excitation at 340 and 380 nm was obtained with a PTI D103 filter-based 75 W xenon light source (PTI, S. Brunswick, NJ) and fluorescence emission at 505 nm (10 nm band pass interference filter; Omega Optical, Brattleboro, VT) was measured with a D104 microscope photometer (PTI). Free intracellular Ca²⁺ concentrations were calculated from 340/380 fluorescence intensity ratios using the calculations of Grynkiewicz, et al. (30) with a K_d of 135 nM. Maximum fluorescence ratio was determined in the presence of 40 µM ionomycin with 10 mM Ca²⁺ and minimum ratio in the presence of 40 µM ionomycin and 10 mM EGTA.

The additions of all fatty acids and derivatives were made to cells in DMEM with 2.5% serum containing 1% (w/v) fatty acid free BSA. Thapsigargin was from LC Services, Corp., Woburn, MA. Fura-2 acetoxymethylester was from Molecular Probes, Eugene, OR. Arachidonic acid, all epoxyeicosatrienoic acids, all dihydroxyeicosatrienoic acids, valeryl salicylate, 5,8,11-eicosatriynoic acid, 5,8,11,14-eicosatetraynoic acid, and octadeca-9,12-diynoic acid were from Cayman Chemical, Ann Arbor, MI. BW755c was a kind gift from Burroughs-Wellcome Corp. All other compounds were from Sigma. The DDT₁MF-2 cell line was originally obtained from Drs. James Norris and Lawrence Cornett, University of Alabama.

RESULTS AND DISCUSSION

The intracellular Ca²⁺ pump blockers, thapsigargin, 2,5-di-tert-butylhydroquinone, and cyclopiazonoic acid, induce profound changes in the growth of DDT₁MF-2 smooth muscle cells in culture (2, 4). In previous studies, we described how Ca²⁺ pool emptying with these agents induced entry of the normally rapidly dividing DDT₁MF-2 cells into a growth-arrested state (13-15). After treatment of cells with pump blockers, cells can progress through S phase before entering a stable G₀-like quiescent growth state (13). In this state, the cells remain viable with normal morphology and mitochondrial function for up to 1 week; protein synthesis continues but at a substantially reduced rate, approximately 20% of that in normal dividing cells (13, 14). Thapsigargin induces an essentially irreversible blockade of intracellular Ca²⁺ pump activity (20, 21, 31); a brief (30-min) treatment of cells with thapsigargin results in permanently emptied pools being retained in nondividing cells for 7 days even if the cells are maintained in thapsigargin free medium (13). The observation that treatment of the pool-depleted growth-arrested cells with high serum (20%) could rescue the cells (14) lead to a search for active components within serum responsible for this effect. We determined that the essential fatty acids, linolenic acid, linoleic acid, and arachidonic acid, each were able to induce similar recovery of Ca²⁺ pools and the resumption of growth (22). The levels of free EFAs present in effective serum concentrations (approximately 5 µM) were close to the EC₅₀ values for EFAs in inducing recovery, suggesting, but not proving, that EFAs might be the active component within serum inducing recovery. This view was supported by the lack of effect of phospholipase A₂ inhibitors on serum-induced recovery indicating that serum does not activate breakdown of cellular phospholipids, but more likely supplies the fatty acids directly. The observation that the nonmetabolizable structural analogue of arachidonic acid, ETYA, did not mimic the action of arachidonic acid (22) provided further support for the view that metabolism of arachidonic acid was required for the recovery-inducing action. Important to investigate was the basis of action of EFA-induced pool recovery and reentry of pool-depleted cells into the cell cycle, specifically, to assess whether metabolism is in fact required, through which pathway(s) active metabolites might be produced, and which of the many possible metabolic products of EFA metabolism might be the active species.

Initial experiments were designed to determine the role of essential fatty acid metabolism through dissection of the three major components of the eicosanoid synthesis pathway (26). Studies utilized a range of specific inhibitors of the cyclooxygenase, lipoxygenase, and cytochrome P-450 epoxygenase pathways. Thapsigargin-treated cells were either induced to recover with 100 µM arachidonic acid or pretreated with an eicosanoid synthesis pathway inhibitor followed by arachidonic acid in the continued presence of the inhibitor. All fatty acids and derivatives were utilized in the presence of 1% (w/v) fatty acid free bovine serum albumin, and measurements of cell number were undertaken after 72 h in culture (see "Experimental Procedures").

The lipoxygenase and cyclooxygenase pathways have been extensively studied, and the actions of a number of well characterized inhibitors of these pathways were examined for their effects on arachidonic acid-induced growth recovery in Ca²⁺ pool-depleted growth-arrested cells. Their effects on the growth of untreated control cells were also examined. As shown in Fig. 1A, the cyclooxygenase inhibitors, aspirin (32, 33), indomethacin (34), NS-398 (35, 36), and valeryl salicylate (37), in each case did not prevent arachidonic acid-induced growth recovery of thapsigargin-arrested cells. Each of these inhibitors did have a small retarding effect on the growth of control cells and this limited growth-slowing effect was also evident on the growth of pool-depleted cells induced to recover with arachidonic acid.

[View Larger Version of this Image (28K GIF file)]

Investigations also examined inhibitors of the second eicosanoid synthesizing pathway, the lipoxygenase pathway. As shown in Fig. 1A, the lipoxygenase inhibitors, 5,8,11-eicosatriynoic acid (ETI) (38, 39) and baicalein (40, 41), induced little significant change in either control growth of cells or arachidonic acid-induced recovery. Similarly, the dual cyclooxygenase and lipoxygenase inhibitors, BW755c (42, 43), and octadeca-9,12-diynoic acid (ODYA) (44), were ineffective on control growth or arachidonic acid-induced recovery. These results indicate that the more thoroughly studied eicosanoid synthesizing pathways, the cyclooxygenase and lipoxygenase pathways, are unlikely to be involved in the arachidonic acid induced recovery mechanism.

In contrast, experiments utilizing inhibitors of the third eicosanoid pathway, the cytochrome P-450 epoxygenase pathway, consistently indicated that the actions of arachidonic acid on inducing recovery of pool-depleted cells, were dependent on this pathway. As shown in Fig. 1B, the cytochrome P-450 inhibitors, metyrapone (45) and SKF525A (45, 46), each prevented arachidonic acid-induced growth recovery but had only marginal effects on the growth of control cells. Interestingly, the "classic" lipoxygenase inhibitor, nordiydroguaiuretic acid (NDGA), also prevented growth recovery induced by arachidonic acid. However, there is an interesting consistency in these results; NDGA in addition to acting as a lipoxygenase inhibitor, is also a potent cytochrome P-450 epoxygenase inhibitor (47, 48), therefore the inhibitory effect is consistent with the involvement of only the epoxygenase pathway in arachidonic acid-induced recovery. It is important to note that, whereas each of these cytochrome P-450 inhibitors greatly reduced (by 80-90%) the ability of arachidonic acid to induce growth recovery of quiescent pool-depleted cells, each had little or no effect upon control cell growth. This indicates that there is a specific requirement for arachidonic acid-derived products of the P-450 epoxygenase pathway in mediating the transition from growth arrest back into the cell cycle and that this effect does not reflect a generalized action of these products on cell growth per se.

As mentioned above, we previously reported that the nonmetabolizable arachidonic acid analogue, 5,8,11,14-eicosatetraynoic acid (ETYA), is unable to induce growth recovery of pool-depleted cells (22). This analogue also acts as an inhibitor of all three eicosanoid synthesizing pathways (23, 26, 27). Experiments revealed that ETYA up to 10 µM did, at least partially inhibit the growth recovery of thapsigargin-arrested cells induced by arachidonic acid (data not shown). However, we also noted that above 10 µM, ETYA has a significant inhibitory effect on normal cell growth therefore it was not possible to effectively gauge its inhibitory action on growth recovery. The basis of the inhibitory action of ETYA on normal cell growth is unknown but may be related to the inhibition of normal cell growth seen with high concentrations (>100 µM) of arachidonic acid, as described earlier (22).

From the above experiments it is apparent that EFA-induced growth recovery of cells that have entered a stable, quiescent growth state following Ca²⁺ pool depletion can be blocked by inhibitors of the cytochrome P-450 epoxygenase pathway of eicosanoid synthesis. However, these growth experiments measured only the end-result of recovery after 3 days. It was important, therefore, to assess the role of eicosanoid synthesis inhibitors on the earlier events that precede growth recovery. We previously demonstrated that growth recovery of thapsigargin-arrested cells following either EFA or high serum treatment occurs after a protein synthesis-dependent induction of new SERCA pump protein and the recovery of inositol 1,4,5 trisphosphate-releasable Ca²⁺ pools (14, 15, 22), which occurs within 6 h. Obviously it was important to ascertain whether the blockade of EFA-induced recovery was at a step before or after the recovery of new Ca²⁺ pools releasable in response to receptor-mediated activation. In other words, arachidonic acid could still be inducing the recovery of pools but reentry of cells into the cell cycle could be inhibited by the epoxygenase inhibitors at a step downstream from the appearance of Ca²⁺ pools. Experiments therefore measured effects of different eicosanoid synthesis inhibitors on the appearance of functional (that is agonist-releasable) Ca²⁺ pools.

As shown in Fig. 2A, the cyclooxygenase inhibitor, aspirin, had no effect upon the arachidonic acid-induced recovery of InsP₃-sensitive Ca²⁺ pools; the function of which was determined by measuring the Ca²⁺ release response to application of the phospholipase C-activating agonist, bradykinin. Similarly, the lipoxygenase inhibitors, ETI and baicalein, were also unable to inhibit the arachidonic acid-mediated pool recovery (Fig. 2B). In contrast, it is clear from the data in Fig. 2C that metyrapone, SKF525A, and NDGA, were each able to completely inhibit the recovery of functional Ca²⁺ pools induced by arachidonic acid. Thus it appears that the metabolism of arachidonic acid through the cytochrome P-450 epoxygenase pathway is necessary for the induction of new Ca²⁺ pools and that growth recovery ensues as a result of the new pools becoming functional.

[View Larger Version of this Image (31K GIF file)]

Whether any of the known eicosanoid products of the P-450 epoxygenase pathway could be identified as being active in mediating recovery of Ca²⁺ pool-depleted cells was important to determine. The primary cytochrome P-450 epoxygenase metabolites of arachidonic acid are the epoxyeicosatrienoic acids (EETs). The epoxygenase enzyme catalyzes an NADPH-dependent addition of oxygen across the double bonds in arachidonic acid to form any one of four regiospecific cis-EET isomers, namely, the 5,6-, 8,9-, 11,12-, and 14,15-EETs, shown in Fig. 3 (45). The epoxy group on the epoxyeicosatrienoic acids are relatively unstable and as a result these EETs are converted to dihydroxyeicosatrienoic acids (DHTs) either enzymically or nonenzymically (45, 49, 50). Adding further to the instability of the epoxy group in the 5,6-EET molecule is the proximity of the epoxy oxygen to the carboxyl group (the 1-carbon on the original arachidonic acid molecule). For this reason, experiments with this molecule were repeated using the more stable, methyl ester form.

[View Larger Version of this Image (14K GIF file)]

Results revealed an important action of only certain members of the EET family of molecules in inducing recovery of cells from thapsigargin-induced Ca²⁺ pool depletion and growth arrest. Addition of the four regiospecific EETs at a concentration of 1.5 µM in the presence of 1% fatty acid free BSA revealed an intriguing structural specificity requirement for the molecules inducing recovery. As shown in Fig. 4, only the 8,9- and 11,12-EETs were able to induce the recovery of agonist sensitive Ca²⁺ pools after thapsigargin treatment. The other two major cytochrome P-450 metabolites of arachidonic acid were ineffective. The two effective molecules both contain an epoxy group across either one of two pre-existing central double bonds within the arachidonic acid molecule (see Fig. 3). Studies have shown that all four of the EET molecules are potent vasodilators (45, 51); the 5,6-EET molecule may be more active, and this molecule may be converted by cyclooxygenase into the corresponding prostaglandin (51, 52). We have shown (Figs. 1 and 2) that the action of arachidonic acid is not blocked by cyclooxygenase inhibitors therefore it is unlikely that cyclooxygenase products are involved in the recovery that we observed. As described above, the 5,6-EET is somewhat more unstable than the other EET molecules, however, the more stable methyl derivative also did not give rise to any recovery of pools or growth.

[View Larger Version of this Image (30K GIF file)]

It was also important to investigate whether the two active EET molecules, 8,9-EET and 11,12-EET, were by themselves sufficient to induce recovery or whether further metabolism of these EETs was necessary. The data in Fig. 5 indicate that the effectiveness of the active EETs in inducing recovery of Ca²⁺ pools is not blocked by the cytochrome P-450 inhibitor, SKF525A. This indicates that there is no further need for epoxygenation in the function of the active EET molecules. Importantly, it also indicates the specificity of the action of SKF525A, which clearly is blocking the conversion of arachidonic acid to active EET metabolites through the P-450 epoxygenase pathway. In other words, the blocking action of SKF525A on arachidonic acid-induced recovery of growth (Fig. 1B) or pools (Fig. 2C) is not through some nonspecific action of SKF525A on the recovery process.

[View Larger Version of this Image (34K GIF file)]

The role of hydration of the epoxy group on each of the EET molecules either by cytosolic epoxide hydrolase activity (49) or by nonenzymatic mechanisms leading to DHT molecules also was an important metabolic change to be addressed. Fig. 6 shows the results of treatment with 8,9- and 11,12-DHTs, the respective hydration products of the two active EET molecules, on the recovery of Ca²⁺ pools in thapsigargin-treated cells. Clearly, neither of the dihydroxy metabolites was effective in inducing pool recovery in comparison to the effect of 8,9- or 11,12-EET under identical conditions. Therefore, the metabolism of the functional EETs to their respective DHTs is not necessary for their activity on inducing new Ca²⁺ pools. From these data, it appears therefore that cytochrome P-450 metabolites of arachidonic acid, 8,9- and 11,12- EET, are each by themselves the necessary and sufficient agent responsible for inducing the reappearance of functional Ca²⁺ pools in previously pool-depleted cells.

[View Larger Version of this Image (29K GIF file)]

Obviously it was important also to determine if the recovery of growth of thapsigargin-treated cells was modified by EET molecules with the same specificity as recovery of pools. As shown in Fig. 7, cells arrested with thapsigargin and treated with the four primary arachidonic acid metabolites of cytochrome P-450 epoxygenase showed a specificity for recovery of growth that exactly mirrored the specificity for pool recovery. Whereas both 5,6- and 14,15-EET were ineffective in inducing growth recovery of thapsigargin-arrested cells, 8,9- and 11,12-EET at 1.5 µM, both induced growth that was close to the effectiveness of 100 µM arachidonic acid. Neither 8,9- nor 11,12-DHT were effective in inducing growth recovery. Significantly, each of the EET and DHT compounds used in these studies was without any significant effect upon the growth of control cells, that is, cells that had not been treated with thapsigargin (data not shown), indicating again that the recovery-inducing action of the EETs is quite distinct from any generalized effects on the growth of cells.

[View Larger Version of this Image (56K GIF file)]

These results provide compelling evidence for the role of cytochrome P-450 metabolism of arachidonic acid in the recovery of Ca²⁺ pools and growth in cells that have been pool-depleted and arrested. From our earlier work we suggested that essential fatty acids contained in serum might be the active agents responsible for inducing pool and growth recovery (22), but no proof for this was provided. Here we provide further evidence to support this hypothesis. Experiments similar to those conducted on arachidonic acid to dissect the metabolic pathway responsible for its induction of recovery were performed on high serum-induced growth recovery (Fig. 8). As with arachidonic acid-induced recovery, neither the cyclooxygenase inhibitors, aspirin or indomethacin, nor the lipoxygenase inhibitors, ETI or baicalein, were able to inhibit high serum-induced recovery of thapsigargin-arrested cells. However, each of the cytochrome P-450 epoxygenase inhibitors, NDGA, metyrapone, and SKF525A, almost completely inhibited the high serum-induced recovery of growth. In addition, as would be expected, SKF525A and metyrapone completely prevented the recovery of bradykinin-releasable Ca²⁺ pools (data not shown). Hence, these results provide clear evidence that high serum-induced recovery is mediated through the same mechanism(s) as the recovery induced by arachidonic acid. This data in conjunction with the earlier reports (22) indicating that the levels of essential fatty acids in serum are sufficient to induce recovery, together provide compelling evidence to support the view that serum directly supplies the essential fatty acids necessary for thapsigargin-arrested cells to synthesize the required 8,9- and 11,12-EETs and hence induce recovery.

[View Larger Version of this Image (41K GIF file)]

Our previous work has revealed an intriguing relationship between the function of Ca²⁺ pools and the ability of cells to either continue through the cell cycle or enter a quiescent growth state (13-17, 22). This report provides evidence for an important action of cytochrome P-450 metabolites of essential fatty acids in inducing the synthesis and expression of new Ca²⁺ pump and functional Ca²⁺ pools within cells that have been induced to undergo entry into a stable quiescent G₀-like state as a result of Ca²⁺ pool emptying with thapsigargin. As a result of this induction of new pools, the cells are able to reenter the cell cycle and continue to grow and divide normally. Indeed, as previously shown, the recovery of pools is a requirement for reentry into the cell cycle since Ca²⁺ pump blockade by thapsigargin still blocks growth recovery resulting from serum-induced recovery of functional Ca²⁺ pools (13, 14). The specific action of the 8,9- and 11,12-EETs in inducing this response may reflect a potentially important pathway by which these epoxides mediate transition from a stationary to a proliferative growth state. Since the appearance of newly synthesized pump protein is the initial response that can be measured following serum-treatment of thapsigargin-treated cells, it is possible that early events mediated by the epoxides are on protein synthesis. Recently it has been observed that thapsigargin- or ionophore-induced pool emptying causes inhibition of protein synthesis through activation of double-stranded RNA-dependent protein kinase (PKR) and subsequent phosphorylation of initiation factor eIF-2 (53, 54). Phosphorylation of eIF-2 represents a key step in translational control and may be important in the regulation of cell growth (54). It is therefore possible that 8,9- and 11,12-EET may reverse these events to permit synthesis of new functional Ca²⁺ pump protein to replenish pools, and through increasing endoplasmic reticulum Ca²⁺ levels, allow the events following pool depletion to be reversed.

Much interest surrounds the possible actions of EETs. Even though much less studied than the products of the cyclooxygenase and lipoxygenase pathway, considerable information exists on the physiological actions of EETs (45). The effects of these molecules are diverse, mediating changes in secretion, vasodilation, and platelet aggregation; such effects may be related to EET-induced alteration of specific transport mechanisms including inhibition of Na⁺ pump activity or activation of intracellular Ca²⁺ release or Ca²⁺ entry (45). Specificity among the four EET subtypes has frequently been observed although not necessarily identical to the selectivity between the isomers observed in the current study. Generally sensitivity in the low or submicromolar range has been reported for EETs (45); in the present study, although we have not measured the exact sensitivity to each molecule, the activity of 8,9- and 11,12-EET appears to be maximal at approximately 1 µM. Little is known regarding the mechanism of action of EETs. Reports have suggested that EETs may act to increase cyclic AMP levels as well as cytosolic Ca²⁺ levels (45, 55). We have not observed significant changes in cytosolic Ca²⁺ as a direct result of EET addition to cells; nor could we measure significant changes in cyclic AMP levels in response to either 8,9-EET, 11,12-EET, or high serum, in normal or thapsigargin-arrested cells.² Additionally, there appear to be no changes in cyclic GMP levels or protein kinase C activity during the recovery of pool-depleted cells.² Clearly, the specific action of EETs in mediating this defined change in the function of Ca²⁺ pools and growth state of cells represents an important area of investigation not only in providing new information on a specific mechanism activated by EETs but also as a potentially significant signaling pathway that may control protein synthesis and/or the transition from a stationary to a proliferative growth state.