Phosphorylation by Protein Kinase A Inhibits Nuclear Import of 5-Lipoxygenase*

The enzyme 5-lipoxygenase initiates the synthesis of leukotrienes from arachidonic acid. Protein kinase A phosphorylates 5-lipoxygenase on Ser523, and this reduces its activity. We report here that phosphorylation of Ser523 also shifts the subcellular distribution of 5-lipoxygenase from the nucleus to the cytoplasm. Phosphorylation and redistribution of 5-lipoxygenase could be produced by overexpression of the protein kinase A catalytic subunit α, by pharmacological activators of protein kinase A, and by prostaglandin E2. Mimicking phosphorylation by replacing Ser523 with glutamic acid caused cytoplasmic localization; replacement of Ser523 with alanine prevented phosphorylation and redistribution in response to protein kinase A activation. Because Ser523 is positioned within the nuclear localization sequence-518 of 5-lipoxygenase, the ability of protein kinase A to phosphorylate and alter the localization of green fluorescent protein fused to the nuclear localization sequence-518 peptide was also tested. Site-directed replacement of Ser523 with glutamic acid within the peptide impaired nuclear accumulation; overexpression of the protein kinase A catalytic subunit α and pharmacological activation of protein kinase caused phosphorylation of the fusion protein at Ser523, and the phosphorylated protein was found chiefly in the cytoplasm. Taken together, these results indicate that phosphorylation of Ser523 inhibits the nuclear import function of a nuclear localization sequence, resulting in the accumulation of 5-lipoxygenase enzyme in the cytoplasm. As cytoplasmic localization can be associated with reduced leukotriene synthetic capacity, phosphorylation of Ser523 serves to inhibit leukotriene production by both impairing catalytic activity and by placing the enzyme in a site that is unfavorable for action.

Previous studies have shown that the subcellular localization of soluble 5-LO before cell stimulation can affect the amount of LT secreted following cell stimulation. For example, import of 5-LO into the nucleus in neutrophils upon adherence increases LTB 4 secretion upon subsequent stimulation (9). On the other hand, nuclear import of 5-LO following adherence rapidly inhibits LTC 4 synthetic capacity in eosinophils (10). Whereas adherence can change 5-LO localization and LT production rapidly, cytokines alter these parameters more slowly. For example, interleukin-3 has been shown to increase the nuclear localization of 5-LO and increase LTC 4 synthetic capacity, in eosinophils treated for 6 h (11). Also, differentiation of human cord blood-derived mast cells with interleukin-3 or interleukin-5 for 5 days increased both nuclear localization of 5-LO and LTC 4 production upon cell stimulation (12). These results demonstrate that different factors can alter 5-LO localization and that 5-LO redistribution affects LT generation upon subsequent cell activation. However, little is known about the intracellular signaling pathways that alter 5-LO localization and consequent LT production.
We recently demonstrated that 5-LO can be phosphorylated by protein kinase A (PKA) on Ser 523 (13). Numerous studies have demonstrated that factors that elevate cellular cAMP levels rapidly inhibit LT synthesis (e.g. Refs. 14 and 15). We found that phosphorylation of 5-LO, in cells or in vitro, as well as mimicking phosphorylation by substituting Ser 523 with Glu, reduced the enzymatic activity of 5-LO. Thus, the direct phosphorylation of 5-LO by PKA may contribute to the reduction in LT synthesis that occurs following elevation of cellular cAMP.
Interestingly, Ser 523 is embedded in one of the three nuclear localization sequences (NLS) of 5-LO, NLS 518 (16). Because the function of a classical NLS is to bind with a karyopherin (importin) protein (17) to initiate import, we asked whether phosphorylation at Ser 523 would inhibit the nuclear import of 5-LO. We report here, for the first time to our knowledge, that phosphorylation of 5-LO on Ser 523 results in an accumulation of 5-LO in the cytoplasm.

EXPERIMENTAL PROCEDURES
Plasmids, Mutagenesis, and DNA Construction-Plasmids containing the wild type (WT) 5-LO or the S523A mutant of 5-LO, alone in pcDNA or fused to green fluorescent protein (GFP) in pEGFP, have been described previously (13,18). Plasmid containing the NLS 518 peptide of 5-LO (encoding Val 514 -Leu 535 ) in fusion with GFP in pEGFP has also been described previously (16); substitution of Ser 523 with Glu in 5-LO or the NLS 518 peptide was performed using the QuikChange sitedirected mutagenesis kit (Stratagene) following the manufacturer's directions. The double mutant, mutNLS 112 ϩS271A, was produced by site-directed mutagenesis of Ser 271 in the previously described and characterized mutNLS 112 of GFP/5-LO (19), where mutNLS 112 is R115Q/ K117Q/R120Q. Mutations and protein frame reading were verified by DNA sequence analysis (DNA Sequencing Core, University of Michigan). All oligonucleotides were synthesized by Integrated DNA Tech-nologies Inc. (Coralville, IA). Plasmid and oligonucleotide sequences are available upon request. The plasmid encoding the mouse PKA catalytic subunit ␣ (C␣) was a gift of Dr. Michael D. Uhler (Department of Biological Chemistry, University of Michigan).
Indirect Immunofluorescence and Confocal Microscopy-NIH 3T3 cells grown on two-well chamber slides (BD Falcon TM ) were transfected with 0.5 g of plasmid DNA overnight, then washed, and treated as described. For immunostaining, cells were washed with PBS, fixed in 4% paraformaldehyde in PBS for 25 min at room temperature, and permeabilized with 0.3% Triton X-100 in PBS containing 0.1% bovine serum albumin. The cells were then blocked with 1% bovine serum albumin in PBS, 0.3% Triton X-100 and incubated with rabbit phospho-5-LO antibody (p5-LO, titer 1: 200) for 1 h at room temperature. Cells were washed in permeabilization buffer and incubated with rhodamine-conjugated goat anti-rabbit secondary antibody (titer 1:250) in blocking buffer for 1 h. Mounting was done in mounting medium containing 4Ј,6-diamidino-2-phenylindole (DAPI, Vector Laboratories, Inc., Burlingame, CA). Cells were visualized and imaged using a Nikon E600 microscope equipped for epifluorescence and digital image capture using a SPOT RT camera. Confocal microscopy was performed with a Bio-Rad MRC-600 laser confocal microscope.
Quantitation of Subcellular Distribution following PKA Activation-As described previously (16), 3T3 cells at 16 h posttransfection were treated with 1 mM 8-br-cAMP for various time points from 0 to 6 h. After fixation with 4% paraformaldehyde, 100 positive cells were scored as to whether nuclear fluorescence was greater than, equal to, or less than cytosolic fluorescence. Care was taken to avoid damaged, dead, or autofluorescent cells. Results from at least three independent transfections were used for statistical analysis. As a second approach, 100 individual cells from each time point after 8-br-cAMP treatment were scored for cytosolic and nuclear fluorescence intensity. Using Adobe Photoshop 6.0, grayscale digital images were adjusted to include the full black-to-white range, and representative gray values, from 0 (white) to 100 (black), were obtained for the cytoplasm and nucleoplasm. Cytoplasmic and nuclear values for each cell were summed to give total cellular fluorescence, and the percent fluorescence values for the nuclear compartment were calculated.
Measurement of Intracellular cAMP Production-After (20), 3T3 cells were plated until confluent in 6-well tissue culture dishes in complete medium. The medium was then replaced with serum-free medium, and the cells were exposed to prostaglandin (PG) E 2 , the EP2selective agonist butaprost, the EP4-selective agonist ONO-AE1-329 (final concentration of each 1 M), or vehicle for the times indicated. Culture supernatants were aspirated, and the cells were lysed by incubation for 20 min with 0.1 M HCl (22°C), followed by disruption using a cell scraper. Intracellular cAMP levels were determined by enzymelinked immunosorbent assay kit according to the manufacturer (Cayman Chemical, Ann Arbor, MI). PGE 2 and butaprost (supplied as the free acid) were from Cayman Chemical; ONO-AE1-329 was a generous gift from ONO Pharmaceutical. Compounds were dissolved in Me 2 SO 4 , and stock solutions were stored at Ϫ80°C until used in assays. Required dilutions of all compounds were prepared immediately before use, and equivalent quantities of vehicle were added to the appropriate controls.
Statistical Analysis-Statistical significance was evaluated by oneway analysis of variance, using p Ͻ 0.05 as indicative of statistical significance. Pairs of group means were analyzed using the Tukey-Kramer posttest.

RESULTS
Persistent Phosphorylation on Ser 523 of 5-LO-As phosphorylation of proteins can be transient, the effects of phosphorylation may also be transitory. To clearly evaluate the effects of phosphorylation of Ser 523 on the localization of 5-LO, we co-transfected NIH 3T3 cells with plasmids encoding GFP/5-LO with or without C␣. In cells expressing only GFP/ 5-LO, the majority of the fluorescence was in the nucleus, co-localizing with DAPI-stained DNA (Fig. 1, A-C). As expected, these cells were negative for p5-LO. In cells expressing GFP/5-LO with C␣, the fluorescence was outside of the nucleus, in the cytoplasmic compartment ( Fig.  1D). Staining for p5-LO matched the fluorescence pattern of GFP/5-LO almost exactly (Fig. 1E). These results demonstrated that the overexpression of the active catalytic subunit of PKA with GFP/5-LO resulted in the accumulation of 5-LO in the cytoplasm, rather than in the nucleus.
The phosphorylation of specific residues on proteins can be mimicked by substitution of the residue with an acidic amino acid, such as glutamic acid. The replacement of Ser 523 with Glu on GFP/5-LO resulted in a striking redistribution from the nucleus (WT, Fig. 2, A and B) to the cytoplasm (S523E, Fig. 2, C and D). Thus, a single amino acid change, like co-transfection with C␣, produced cytoplasmic localization of GFP/5-LO.
Activation of PKA Leads to Phosphorylation and Redistribution of 5-LO-To study the effects of PKA activation on 5-LO localization, we first tested different agonists for their ability to phosphorylate 5-LO. The adenylyl cyclase activator forskolin, with or without the phosphodiesterase IV inhibitor 3-isobutyl-1-methylxanthine, as well as PKAspecific agonist 6-bnz-cAMP and the cAMP analogues 8-br-cAMP and dibutyryl-cAMP, were all able to elicit phosphorylation of 5-LO on Ser 523 , as indicated by positive immunostaining using the p5-LO antibody (Fig. 3A). However, the B site activator of PKA, 8-n-hexylaminoadenosine-cAMP, appeared to be a poor agonist. The treatment of cells overexpressing 5-LO with Ser 523 replaced with Ala did not show phos-phorylation in response to the PKA-specific agonist 6-bnz-cAMP (Fig.  3B). Based on these results, additional experiments focused on the effects of 6-bnz-cAMP and 8-br-cAMP on 5-LO phosphorylation and localization.
The kinetics of phosphorylation of 5-LO by PKA was evaluated using 6-bnz-cAMP, a potent, selective activator of PKA (21). Although significant phosphorylation of 5-LO was evident within 1 h after the addition of 6-bnz-cAMP, phosphorylation continued to increase over 6 h (Fig. 4). Modest phosphorylation of Ser 523 , as indicated by positive staining with the p5-LO antibody, was evident in untreated cultures. Enhanced phos-phorylation could be seen as early as 15 min after treatment (data not shown). Also, treatment with the phosphatase inhibitors okadaic acid or calyculin A significantly increased phosphorylation at earlier time points (data not shown). Taken together, these results indicate that PKA activation results in a relatively slow but continuous phosphorylation of 5-LO, countered in part by dephosphorylation by endogenous phosphatase activity, with phosphorylated 5-LO accumulating over time.
Because persistent phosphorylation altered 5-LO localization, it was of interest to determine whether a time-dependent redistribution of 5-LO paralleled the time-dependent phosphorylation. Live cell imaging of individual cells indicated that treatment with either 6-bnz-cAMP or 8-br-cAMP resulted in a significant change in the subcellular localization of 5-LO within 1 h in some cells, whereas other cells did not respond at all (data not shown). To quantitate the change in localization within a population of cells, cultures were treated with 1 mM 8-br-cAMP for various times, then fixed, photographed, and scored for subcellular distribution of GFP/5-LO. In untreated cells overexpressing GFP/5-LO, essentially all cells had fluorescent nuclei (Fig. 5A). Following PKA activation, the difference in the fluorescence between the nucleus and cytoplasm diminished, with some cells having nuclei that were darker than the cytoplasm (Fig. 5, B and C). Quantitative analysis of nuclear fluorescence of untreated cells indicated that most cells exhibited moderate (Fig. 5D, N 1 ) or strong (Fig. 5D, N 2 ) nuclear accumulation of WT 5-LO. PKA activation resulted in a decrease in the    Phosphorylation by PKA Inhibits Nuclear Import of 5-LO DECEMBER 9, 2005 • VOLUME 280 • NUMBER 49 percentage of cells with strong nuclear fluorescence and an increase in cells with predominantly cytoplasmic fluorescence. Cells overexpressing GFP/5-LO with the S523A mutation showed moderate to strong nuclear fluorescence that was not affected by PKA activation (data not shown). A statistically significant decrease in the percentage of cells with predominantly nuclear fluorescence, paralleled by a significant increase in cytoplasmic predominant cells, was documented at 6 h after 8-br-cAMP treatment (Fig. 6A). Immunofluorescent staining of these cells with the p5-LO antibody revealed strong positive staining within the cytoplasm (Fig. 6B). Thus, PKA activation produced a time-dependent redistribution of 5-LO with accumulation of phosphorylated 5-LO in the cytoplasm.
From these results, it was not clear whether PKA phosphorylation of 5-LO occurred in the cytoplasm, in the nucleus, or both. Further examination of cells treated with 8-br-cAMP for 6 h and stained for p5-LO revealed that several cells with more nuclear than cytoplasmic GFP/ 5-LO also stained positive for p5-LO. Confocal analysis of these cells indicated that the majority of the p5-LO was in the cytoplasm (Fig. 7, top  and bottom), although some was evident within the nucleus, particularly at the periphery (Fig. 7, middle). To further test whether cytoplasmic 5-LO is phosphorylated more than nuclear 5-LO, we compared the ability of PKA to target WT GFP/5-LO versus a mutant developed to alter 5-LO localization but not activity. This mutant, with substitutions to inactivate NLS 112 and block phosphorylation of Ser 271 , has significantly stronger cytoplasmic localization than the WT protein (Fig. 8A). Activation of PKA with 8-br-cAMP produced greater phosphorylation of the mutant than WT GFP/5-LO, as determined by immunoblot analysis (Fig. 8B). These results indicate that although PKA may be able to phosphorylate 5-LO within the nucleus there appears to be greater targeting of 5-LO within the cytoplasm.
Effect of p38 Mitogen-activated Protein Kinase (MAPK) Inhibition on PKA Modulation of 5-LO-5-LO can be phosphorylated on Ser 271 in response to stresses that activate p38 MAPK (22). We asked whether phosphorylation of Ser 523 was independent of p38 MAPK activity in   vivo by treating cells with the selective p38 MAPK inhibitor SB203580. The PKA/PKG inhibitor H-89 demonstrated a dose-dependent inhibition of phosphorylation of 5-LO by 8-br-cAMP, whereas SB203580 was without effect up to 30 M (Fig. 9A). It has been reported that the IC50 of SB203580 on p38 MAPK is 0.5 M (23). In addition, 3T3 cells overexpressing GFP/5-LO retained nuclear localization of 5-LO and did not stain with the p5-LO antibody when treated with 8-br-cAMP (1 mM) plus H-89 (30 M) (Fig. 9B). The inhibition of p38 MAPK with SB203580 (30 M) did not block cytoplasmic localization and phosphorylation of Ser 523 on 5-LO in response to 8-br-cAMP (Fig. 9B).
PKA Can Phosphorylate and Alter the Localization of the NLS 518 Peptide-We previously reported that the NLS 518 peptide was sufficient to drive nuclear accumulation of GFP (16). We asked whether phosphorylation of GFP-tagged NLS 518 on Ser 523 would alter nuclear import. As expected, GFP alone was distributed diffusely throughout the cell (Fig.  10A), whereas GFP/NLS 518 was strongly accumulated within the nucleus (Fig. 10C). Substitution of Ser 523 with Glu, to produce a phosphorylation mimic within the GFP/NLS 518 construct, resulted in a relatively diffuse distribution, consistent with little or no active nuclear import (Fig. 10E). Co-transfection of 3T3 cells with GFP/NLS 518 and C␣ resulted in more cytoplasmic than nuclear GFP fluorescence, with strong positive staining with the p5-LO antibody, particularly in the cytoplasm (Fig. 11, A-C). Finally, activation of PKA with 8-br-cAMP also resulted in strong positive staining with p5-LO, again in the cyto-plasm (Fig. 11, D-F). In this treatment, the effect of phosphorylation on localization becomes most apparent, as abundant fluorescent protein is able to be accumulated in the nucleus when NLS 518 is not phosphorylated and the majority of phosphorylated NLS 518 is found in the cytoplasm. Taken together, these results suggest that the NLS 518 peptide itself can be targeted by PKA and that when phosphorylated its ability to drive nuclear import is impaired. 2 can inhibit leukotriene synthesis via a cAMP-dependent process (24). PGE 2 acts through four distinct G protein-coupled receptor subtypes (EP1-4), with distinct signaling pathways. Because the G s -coupled EP2 and EP4 receptors activate adenylate cyclase activity (25), we measured changes in intracellular cAMP in response to PGE 2 (Fig. 12A). Treatment of 3T3 cells with PGE 2 (1 M) provoked an immediate elevation of cAMP, resulting in a 50% increase in cAMP 5 min after PGE 2 exposure. Levels remained Phosphorylation by PKA Inhibits Nuclear Import of 5-LO DECEMBER 9, 2005 • VOLUME 280 • NUMBER 49 elevated for ϳ60 min. To determine the contribution of either EP2 or EP4 receptors to this increase, cells were treated with equimolar amounts of the EP2-selective agonist butaprost or the EP4-selective agonist ONO-AE1-329 for various times. Both butaprost and ONO-AE1-329 evoked an ϳ20% increase in intracellular cAMP within 15 min that returned to baseline levels by 60 min (data not shown). Treatment with the same concentration of PGE 2 also resulted in phosphorylation of 5-LO within 15 min, as indicated by immunoblot analysis (Fig. 12B). Phosphorylation of 5-LO in response to PGE 2 continued to increase over 6 h, although the level of phosphorylation obtained with PGE 2 was less than that induced by 6-bnz-cAMP. The vehicle, Me 2 SO, did not induce phosphorylation. After 6 h of treatment with PGE 2 , 5-LO was clearly cytosolic in some cells, as indicated by redistribution of GFP/ 5-LO (Fig. 12C). Some cells showed little redistribution following PGE 2 treatment, with 5-LO remaining largely within the nucleus; other cells had a balanced distribution between nucleus and cytoplasm. Following fixation and staining for p5-LO, cells with cytoplasmic 5-LO based on GFP fluorescence (Fig. 12D) also had positive staining for p5-LO in the cytoplasm (Fig. 12E). Cells that had predominantly nuclear 5-LO following PGE 2 treatment did not typically stain positive for p5-LO (not shown).

DISCUSSION
As 5-LO serves the key function of initiating LT synthesis from arachidonic acid, it represents a primary point for regulating the generation of these potent proinflammatory mediators. Several studies have demonstrated that changes in the subcellular localization of 5-LO in resting leukocytes significantly affects the amount of LT produced when those leukocytes are activated. For example, we have recently used molecular modification of the import sequences of 5-LO to show that LTB 4 production decreased, as less 5-LO could be imported into the nucleus (26). In that study and others, the positioning of 5-LO in the cytoplasm of resting leukocytes typically correlated with reduced production of LTs upon cell stimulation. In this study, we have examined the effects of phosphorylation of Ser 523 on the localization of 5-LO. This residue is nested in NLS 518 , identified as 518 RGRKSSGFPKSVK 530 (16). Previous work demonstrated that phosphorylation of Ser 523 significantly reduced the intrinsic enzymatic activity of 5-LO (13). We report here, for the first time to our knowledge, that phosphorylation of Ser 523 also blocks nuclear import, resulting in the accumulation of 5-LO in the cytoplasm. Thus, phosphorylation of 5-LO achieves two effects that both serve to reduce cellular LT generation: a direct molecular effect on the intrinsic catalytic activity of 5-LO and a slower, cellular effect of placing 5-LO in a subcellular compartment that is less favorable for arachidonic acid metabolism.
A very surprising result is the apparently slow rate of phosphorylation of 5-LO following PKA activation. One contributing factor is the activity of endogenous phosphatases. Elevated cAMP levels increase activity of phosphatases (27,28), suggesting that our results might underestimate the true rate of phosphorylation. Consistent with this interpretation, we found that pretreatment with the phosphatase inhibitors okadaic acid or calyculin A resulted in stronger phosphorylation of 5-LO at earlier time points (data not shown). Thus, phosphorylation of 5-LO by PKA may occur in minutes, as observed in response to PGE 2 treatment.  Dephosphorylation by phosphatases, then, would serve to limit the time of phosphorylation and its impact on LT synthesis.
The positioning of Ser 523 within NLS 518 suggests that this NLS is functional in promoting import when Ser 523 is not phosphorylated, with phosphorylation inhibiting import perhaps by interfering with docking of karyopherins with the NLS. Activation of PKA, as following PGE 2 treatment, typically resulted in strong cytoplasmic localization of 5-LO in some cells, with normal import in others. These results are strongly reminiscent of those obtained with mutation of NLS 518 , which left two functional but regulated NLSs intact (16,19). This suggests that 5-LO that is phosphorylated in the cytoplasm may conceivably be imported through the action of other NLSs. That is, the presence of three NLSs allows several grades of import of 5-LO (29). Silencing of NLS 518 leaves two functional sequences, NLS 112 and NLS 158 , which can mediate import of 5-LO. As nuclear export appears also to be regulated (30,31), other factors besides phosphorylation of NLS 518 must be able to influence the final positioning of 5-LO.
Activated PKA can phosphorylate nuclear targets, like the cAMP response element binding protein, and so it may phosphorylate 5-LO within the nucleus. This raises the possibility that, in addition to reducing 5-LO activity, phosphorylation may promote the export of 5-LO. We cannot rule out this possibility. However, our results using the GFP/ NLS 518 fusion peptide, which showed that unphosphorylated protein was found in the nucleus and that phosphorylated protein remained cytoplasmic (Fig. 11), strongly indicated that phosphorylation inhibits the import function of this NLS. PGE 2 generally down-regulates the function of leukocytes, and this is mediated by elevation of cAMP through activation of the EP2 and EP4 receptors. For example, PGE 2 inhibits macrophage phagocytosis (20) and neutrophil phospholipase D (32) through EP2, monocyte IL-12 production through EP4 (33), and major histocompatibility complex class II expression in dendritic cells (34) and the release of tumor necrosis factor-␣, endothelin-1, and interleukin-1␣ by macrophages via both EP2 and EP4 (35). Through these suppressive effects on leukocytes, PGE 2 can contribute to the resolution of inflammation. PGE 2 has been shown to inhibit LT synthesis through the elevation of cAMP (24,36). Our finding that PGE 2 promotes phosphorylation of 5-LO on Ser 523 and causes redistribution of 5-LO to the cytoplasm, both effects that can reduce the synthesis of proinflammatory LTs, provides insight into the mechanism by which this effect is mediated.
Previously, we reported that PKA modulators did not alter the subcellular distribution of 5-LO in the first 30 min after treatment (13). As shown in Fig. 5, the changes in 5-LO localization are slow, with only modest changes observable at 2-h posttreatment. We also found that the effects of PKA modulators were heterogeneous, with phosphorylation as well as redistribution of 5-LO occurring faster in some cells, slower in others, and not at all in some cells (data not shown). This may be because of the fact that a non-synchronized cell culture is heterogeneous. The lag between phosphorylation and redistribution of 5-LO suggests that phosphorylation may be the initial factor affecting 5-LO activity, with redistribution serving to increase or prolong the effects of phosphorylation.
It is important to note that elevation of intracellular cAMP also inhibits the release of arachidonic acid (37)(38)(39). This effect is rapid, whereas the effects of PKA activation on 5-LO phosphorylation and subcellular distribution appear to be slower. Whether arachidonic acid release remains impaired after prolonged exposure to cAMP-elevating agents is, to our knowledge, not known. It is possible that at early time points inhibition of LT synthesis by elevated cAMP results primarily from decreased arachidonic acid release. At later time points, 5-LO redistribution to the cytosolic compartment and its sustained phosphorylation on Ser 523 will likely lead to a dramatic decrease in the capacity to synthesize LTs even by trans-cellular metabolism (i.e. where arachidonic acid originates from other cell types unaffected by the cAMP-elevating agent). Additional kinetic experiments with leukocytes are necessary to assess this possible sequential cascade of molecular events involved in the down-regulation of LT synthesis by elevated cAMP concentrations.
The experiments presented in this study take advantage of the easeof-transfection of 3T3 cells to address molecular aspects of the regulation of 5-LO. However, an important question is the relevance of these findings to primary leukocytes. Cell transformation can be achieved by reduced phosphatase activity, (40), and it is likely that the transformed 3T3 cells used here have reduced phosphatase activity when compared with primary, differentiated leukocytes. Consistent with this possibility, we found that inhibition of endogenous phosphatases is required to demonstrate phosphorylation of Ser 523 on 5-LO in leukocytes (data not shown). Additional studies on the regulation of the phosphorylation of Ser 523 on 5-LO in leukocytes, in vitro and in vivo, are in progress.
In summary, we report here that elevation of cAMP results in redistribution of 5-LO from the nucleus to the cytoplasm through phosphorylation of Ser 523 and inhibition of nuclear import through NLS 518 . As both phosphorylation on Ser 523 and positioning 5-LO in the cytoplasm FIGURE 12. Effect of PGE 2 on cAMP production and the phosphorylation and localization of 5-LO. A, time course of cAMP generation in 3T3 cells exposed to prostaglandin E 2 . 3T3 cells were treated with PGE 2 (1 M) or vehicle control for the times indicated followed by determination of intracellular cAMP as described under "Experimental Procedures." Experiments were performed in duplicate on two separate occasions, and a representative result is shown. Data are expressed as relative to the vehicle-treated control (basal cAMP level ϭ 4.29 pmol/ml). B, cells overexpressing 5-LO were stimulated with vehicle (DMSO), PGE 2 (1 M), or 6-bnz-cAMP (1 mM) for the indicated times, then analyzed by immunoblot for p5-LO or total 5-LO. C, localization of GFP/5-LO in 3T3 cells after treatment with PGE 2 (1 M) for 6 h. Arrows, cells with cytoplasmic fluorescence. In a parallel experiment, cells overexpressing GFP/5-LO were stimulated with PGE 2 for 6 h and then were fixed, stained, and imaged for GFP/5-LO (D) or p5-LO (E). DECEMBER 9, 2005 • VOLUME 280 • NUMBER 49 serve to reduce LT synthesis, these events may be important in the resolution of inflammation.