INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

PROSTANOIDS (prostaglandins and thromboxane) are generated by cyclooxygenases (COX) acting on the substrate arachidonic acid, which is released from membrane phospholipids upon stimulation of cells with growth factors and cytokines. Two cyclooxygenase isoforms have been isolated, COX-1 and COX-2. Although they are both involved in the generation of prostaglandins and thromboxane, they seem to have different biological roles. COX-1 is constitutively expressed in most tissues and is involved in maintaining cellular homeostasis, including regulation of vascular tone (37). In contrast, COX-2 is usually undetectable in cells under normal conditions but is readily expressed in response to inflammatory cytokines in many cells, including myocytes (17, 20, 37). Moreover, COX-2 is induced in the myocardium of failing human hearts and may contribute to cardiac dysfunction during this process (38). COX-2 also seems to play a key role in the regulation of tumorigenesis (33) and cellular growth (24). Because the growth of myocytes in response to pathological stimuli results in left ventricular hypertrophy, and hypertrophy is a major risk factor for development of heart failure, we investigated the impact of COX-2 expression on the growth of neonatal ventricular myocytes (NVM).

One of the inflammatory cytokines regulating the synthesis of COX-2 is interleukin-1 (IL-1). We have previously shown that IL-1 regulation of COX-2 involves the p42/44 and p38 MAPK-signaling pathways (17). The phorbol ester phorbol-12-myristate-13-acetate (PMA) has been shown to regulate expression of COX-2 in many types of cells, including fibroblasts, endothelial cells, epithelial cells, and medullary interstitial cells (8, 11, 30, 34). Phorbol ester activates protein kinase C (PKC) and mitogen-activated protein kinases (MAPKs), which also mediate the action of IL-1 in many types of cells (12, 15, 17, 26). Thus we investigated whether the combined stimuli would enhance COX-2 expression and prostaglandin production. We treated NVM with PMA, which has been shown to induce hypertrophy by activation of protein kinase C (PKC) (4), in the presence and absence of IL-1, and we also measured COX-2 protein levels and PGE₂ and PGF₂ production. In addition, we measured PMA and PMA + IL-1-stimulated protein synthesis as assessed by [³H]leucine incorporation with and without COX inhibitors to test the effect of endogenous COX-2 products on growth of NVM. Finally, we tested whether exogenous PGE₂ and PGF₂ stimulate NVM growth.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell culture. Primary cultures of NVM were derived from the digestion of 1- to 2-day-old neonatal Sprague-Dawley rat hearts (Charles River, Kalamazoo, MI) as described previously (18). This protocol was approved by the Henry Ford Hospital Committee for Care and Use of Experimental Animals. To minimize contamination by fibroblasts, cells were preplated for 30 min. Myocytes were plated at high density [10⁵ cells/cm² (10⁶ cells/well of a 6-well dish)] in DMEM (GIBCO-BRL) plus 10% fetal bovine serum (HyClone) and 0.1 mM bromodeoxyuridine for 40 h. Serum-free (SF) medium supplemented with glutamine, insulin, selenium, and transferrin was then added for 24 h (18). Bromodeoxyuridine and serum-free (SF) medium were used to inhibit proliferation of contaminating fibroblasts in the myocyte culture. All treatments were added to SF-DMEM. IL-1 (5 ng/ml or 3 × 10¹⁰ M) and other chemicals were used at concentrations previously shown to be effective in studies on the regulation of inducible nitric oxide synthase, COX-2, and B-natriuretic peptide (9, 17-20).

Enzyme immunoassay for measurement of PGE₂ and PGF₂. 1 × 10⁶ cells per well of a six-well plate were treated with PMA or IL-1 in 1 ml of SF-DMEM. At the end of the treatment period, a 1-ml aliquot of medium from each well was dried down and resuspended in 0.15 ml of Ultrapure H₂O (Cayman, Ann Arbor, MI). Aliquots were diluted 1:20 (for controls) to 1:5,000 (IL-1 treatment) and assayed for PGE₂ and PGF₂ using enzyme immunoassay kits (Cayman). Values from triplicate wells were averaged. Means ± SE were calculated from the data generated in multiple experiments.

Isolation of protein and Western blot analysis. Protein was isolated from NVM using buffers and protease inhibitors as described previously (19). Lysate protein (50 µg) was separated out by electrophoresis on an 8% SDS-polyacrylamide gel and transferred to an Immobilon-P PVDF membrane (Millipore). To detect the 72-kDa COX-2 protein, we used 0.0001 mg/ml of an anti-goat COX-2 polyclonal antibody (Santa Cruz). The appropriate secondary antibody linked to horseradish peroxidase was used for chemiluminescent detection using ECL Western blotting detection reagents (Amersham Pharmacia Biotech). The signal was detected by exposure to Fuji RX film and analyzed by scanning densitometry. Results (in densitometry units) were normalized to either control or IL-1 treatment. We then determined the means ± SE for the fold increase for all experiments.

Isolation of RNA and Northern blot analysis. Total RNA was isolated from NMVs and analyzed for 4.2 kb COX-2 mRNA. GAPDH mRNA was used to correct for variation in loading of samples onto gels as described previously (18). Electrophoresis, blotting, preparation of radiolabeled cDNA probes, hybridization, and washing of blots were described previously (18, 19). Either 1.8 kb ovine COX-2 cDNA (Biomol) or 2.1 kb rat cDNA (35) was used to detect COX-2 mRNA.

Measurement of protein synthesis. Protein synthesis was determined by incorporation of [³H]leucine into trichloroacetic acid (TCA)-insoluble material. Myocytes were grown under SF conditions for 48 h, after which they were treated with PMA (0.1 µmol/l) or PMA + IL-1 and incubated with 2.5 µCi [³H]leucine (6.2 TBq/mmol = 168 Ci/mmol) in SF-DMEM for 48 h. Protein was precipitated by adding 1 ml ice-cold 10% TCA to each well, and the precipitate was vacuum filtered through GF/C filters (Whatman). Filters were washed three times with 5 ml of 5% TCA and once with 5 ml of 70% ethanol and counted in 3 ml of scintillation fluid (Insta-gel XF, Packard Instruments) using a Packard 3320/3330 Tri-Carb -scintillation counter. For each treatment, counts-per-minute (CPM) values from triplicate filters were averaged. The control CPM was set to 100%, and all other treatments were normalized to it. We then determined the means ± SE for all of the experiments.

Statistics. Values are presented as means ± SE; n represents the number of experiments. When two treatments were compared, statistical significance was evaluated by Student's t-test. When more than two treatments were compared, we used ANOVA with the Student-Newman-Keuls correction for multiple comparisons. P < 0.05 was considered significant.

Chemicals. Indomethacin and PMA were obtained from Sigma; NS-398 (NS), PGE₂, and PGF₂ were from Cayman; L-[3,4,5-³H(N)]leucine was from DuPont/NEN (Boston, MA); IL-1 was from Promega; GF-109203X (GF) and curcumin (Curc) came from Biomol (Plymouth Meeting, MA); and SB-203580 (SB) and PD-98059 (PD) were from Calbiochem (San Diego, CA). Routine laboratory supplies and chemicals were obtained from Fisher and Sigma.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effect of PMA on COX-2 and prostanoid production. Treatment of NVM with 0.1 µmol/l PMA for 24 h increased COX-2 mRNA (Fig. 1A) and protein (Fig. 1B). Regarding COX-2 protein, analysis of blots from 12 separate experiments indicated that PMA stimulated COX-2 protein 2.8 ± 0.4-fold. Using enzyme immunoassay, we measured PGE₂ and PGF₂ secreted into the culture medium. Under control conditions, PGE₂ levels were ~2.5 times higher than PGF₂. PMA stimulated PGE₂ and PGF₂ 3.7-fold and 2.9-fold, respectively (Fig. 1C). In a separate study, we determined the effect of chronic PMA treatment on COX-2 synthesis and PGE₂ production and found that COX-2 protein and PGE₂ did not increase further from 24 to 48 h of treatment (data not shown).

View larger version (33K): [in this window] [in a new window]   Fig. 1.   Effect of phorbol-12-myristate-13-acetate (PMA) on cyclooxygenase-2 (COX-2) and prostanoids. A: representative Northern blot showing PMA stimulation of COX-2 mRNA. Similar results were obtained in 4 separate experiments. B: representative Western blot showing PMA stimulation of 72-kDa COX-2 protein. C: PGE2 and PGF2 production (ng/ml produced by 1 × 106 cells). Under control conditions, 1 × 106 cells produced 0.9 ± 0.1 ng/ml PGE2 (n = 11) and 0.4 ± 0.1 ng/ml PGF2 (n = 6) in 24 h. *P < 0.01 vs. control; #P < 0.05 vs. control.

Effect of PKC and MAPK inhibition on PMA regulation of COX-2. To test the involvement of PKC, p42/44 MAPK, p38 MAPK, and c-Jun NH₂ terminal kinase (JNK) in PMA regulation of COX-2 synthesis, we pretreated NVM with (in µmol/l) 10 GF, 25 PD, 10 SB, and 10 Curc (2), respectively, for 1 h before treatment with PMA for 24 h. The PKC inhibitor GF and the p38 MAPK inhibitor SB totally inhibited PMA-stimulated COX-2 synthesis, whereas the p42/44 inhibitor PD and JNK inhibitor Curc were only partially effective (Fig. 2).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Effect of kinase inhibitors on PMA-stimulated COX-2 synthesis. Densitometry data from 4-12 separate Western blots. CTL, control; GF, protein kinase C (PKC) inhibitor GF-109203X; PD, mitogen-activated protein kinase kinase inhibitor PD-98059; SB, p38 mitogen-activated protein kinase (MAPK) inhibitor SB-203580; Curc, c-Jun NH2 terminal kinase (JNK) inhibitor curcumin. *P < 0.01 vs. PMA. #P < 0.05 vs. PMA.

Effect of COX-2 inhibition on PMA-stimulated protein synthesis. NVM were treated with PMA for 48 h in the presence of [³H]leucine, resulting in a 1.6 ± 0.1-fold increase in protein synthesis versus untreated cells (Fig. 3). We next tested whether COX-2 products were involved in PMA-stimulated protein synthesis. Neither the COX-1/COX-2 inhibitor indomethacin (Indo, 10 µmol/l) nor the COX-2-selective inhibitor NS (10 µmol/l) had any significant effect on PMA-stimulated [³H]leucine incorporation (PMA + Indo = 1.7 ± 0.1-fold and PMA + NS = 1.6 ± 0.2-fold). Neither inhibitor had any effect on basal protein synthesis in untreated cells (data not shown). Indo and NS inhibited IL-1-stimulated PGE₂ production by 97.8% and 99.8%, respectively (data not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Effect of COX inhibitors on PMA-stimulated protein synthesis. Relative [3H]leucine content data from 12 separate experiments. Control incorporation was set to 100%. *P < 0.01 vs. control but there were no significant differences between PMA and PMA + indomethacin (Indo) or NS-398 (NS).

2

2

2

Effect of treatment with PMA and IL-1 on COX-2 and prostanoid production. IL-1 has been shown to induce COX-2 in NVM (17) as well as intestinal myofibroblasts (10), chondrocytes (23), and rat mesangial cells (5), resulting in large increases in PGE₂. In NVM, IL-1 stimulation increases PGE₂ production much more than PMA stimulation. We hypothesized that PMA-induced increases in PGE₂ might not be sufficient to affect myocyte growth. We tested whether treatment with PMA and IL-1 for 24 h would increase COX-2 and PGE₂ production and thus affect protein synthesis. IL-1 increased COX-2 protein 42.5 ± 5.8-fold and PGE₂ production 1,608 ± 295-fold versus control, whereas cotreatment with PMA + IL-1 increased COX-2 protein 31.8 ± 3.1-fold and PGE₂ production 1,093 ± 236-fold. Thus treatment with IL-1 and PMA resulted in a slight decrease in both COX-2 protein (by 25%) and PGE₂ production (by 32%) versus IL-1 alone (Fig. 4, A-C); however, only the COX-2 protein reduction was significantly different versus IL-1 alone.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   Effect of interleukin-1 (IL-) and PMA on COX-2 and PGE2. A: representative Western blot showing combined PMA and IL-1 stimulation of 72-kDa COX-2 protein. B: densitometry data from 12 separate Western blots. *P < 0.01 vs. IL-1 stimulation. C: PGE2 production (ng/ml) from 12 separate experiments. 1 × 106 cells stimulated with IL-1 produced 1,521 ± 279 ng/ml PGE2 in 24 h. P values not significant (IL-1 vs. PMA + IL-1).

2

Effect of PKC and MAPK inhibition on IL-1 regulation of COX-2. LaPointe and Isenovic (17) have previously shown that IL-1 regulation of COX-2 involves p38 and p42/44 MAPK. To test the involvement of PKC and JNK in IL-1 regulation of COX-2 synthesis, we pretreated NVM with 10 µmol/l GF and 10 µmol/l Curc, respectively, for 1 h before treatment with IL-1 for 24 h. The PKC inhibitor GF totally inhibited IL-1-stimulated COX-2 synthesis, whereas the JNK inhibitor Curc was only partially effective (Fig. 5).

View larger version (14K): [in this window] [in a new window]   Fig. 5.   Effect of kinase inhibitors on IL-1-stimulated COX-2 synthesis. Densitometry data from 4 separate Western blots. IL-1 (5 ng/ml). *P < 0.01 vs. IL-1. #P < 0.05 vs. IL-1.

Effect of PMA and IL-1 on cardiac myocyte growth. Harding et al. (9) have previously shown that IL-1 has no effect on [³H]leucine incorporation into total protein in NVM. These results were confirmed in our present study (Fig. 6). In addition, cotreatment with PMA and IL-1 had no effect on protein synthesis versus PMA alone. Inhibition of COX activity with either Indo or NS had no effect on protein synthesis stimulated by PMA + IL-1.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Effect of COX inhibitors on PMA + IL-1-stimulated protein synthesis. Relative [3H]leucine content data from 12 separate experiments. *P < 0.01 vs. control but no significant difference between PMA + IL-1 vs. treatment with Indo and NS.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The major new finding of this study is that the phorbol ester PMA induces COX-2 synthesis and prostanoid production in NVM by a mechanism involving PKC and p38 MAPK. PGE₂ production was higher than PGF₂ production in both control and treated myocytes. Unlike the ability of endogenous COX-2 products to promote growth in several types of cells (24, 33), COX-2 products had no effect on PMA-stimulated protein synthesis in NVM. However, exogenous administration of PGF₂, but not PGE₂, stimulated protein synthesis. Endogenous production of PGF₂ in control and PMA-treated myocyte cultures results in nanomolar amounts of the prostanoid, and this amount seems insufficient to induce growth.

Our study indicates that PMA regulates COX-2 mRNA; however, we have not investigated whether this occurs via transcriptional or posttranscriptional mechanisms. Inoue et al. (11) and Fletcher et al. (6) have shown that the transcriptional effect of PMA is targeted to regulatory elements in the COX-2 promoter. Most likely, it is mediated by a number of kinase signaling cascades. PMA is known to activate two groups of PKC isoforms, both classic and novel PKCs, which have a number of nuclear and cytosolic substrates (26). The -isoform of PKC has been shown to directly phosphorylate Raf-1, providing a link to the p42/44 MAPK pathway (13). PKC-mediated activation of the MEKK-SEK-JNK cascade (where MEKK is MAPK kinase kinase that activates SEK, and SEK is MAPK kinase activating JNK) has also been described as a consequence of G protein-coupled receptor activation (3). Our studies implicate PKC and MAPKs in the effect of PMA, and these kinases can phosphorylate and activate transcription factors such as c-Jun and activating transcription factor, which are known to be involved in regulation of COX-2 (34, 39, 40). On the other hand, MAPKs have been reported to increase the stability of COX-2 mRNA (31). Thus it is possible that PMA regulates COX-2 expression at both transcriptional and posttranscriptional levels in cardiac myocytes.

IL-1 signaling involves numerous mediators depending on the type of cell; these may include the sphingomyelin-ceramide pathway, serine-threonine kinases, PKC, tyrosine kinases, and phosphatidylinositol 3-kinase (12, 14, 17, 21, 29). Although IL-1 preferentially activates the JNK and p38 kinase pathways (15), it has also been reported to activate p42/44 MAPK (17). Our data indicate that both PMA and IL-1 stimulation of COX-2 requires activation of PKC and p38, whereas inhibition of either p42/44 or JNK only partially decreases COX-2. Because IL-1 and PMA apparently activate the tested signaling pathways in a very similar way, we hypothesized that they would have additive or synergistic effects on regulation of COX-2 expression. However, our data indicate that either IL-1 or PMA can induce COX-2 expression but that their combined effect is smaller than the effect of IL-1 alone. One possible explanation for this is that chronic treatment with PMA downregulates a critical PKC isoform involved in IL-1 regulation of COX-2 gene expression, reducing COX-2 protein levels. This is probably not the case, because preliminary data from four experiments indicate that COX-2 mRNA levels are approximately equivalent in IL-1- and IL-1 + PMA-treated cells. Other possible explanations of our data are that PMA induces a factor that partially inhibits the activity of COX-2 or that interferes with binding of IL-1 to its receptor. Additional experiments are required to address this issue.

As expected, PMA increased protein synthesis in NVM. IL-1 alone had no effect on protein synthesis, in accord with our previous publication (9). In addition, IL-1 had no effect on PMA-stimulated protein synthesis, similar to our previous study in which phenylephrine-stimulated protein synthesis was unaffected by IL-1 treatment (9). In contrast to our studies, investigators have detected an effect of IL-1 on protein synthesis when NVM are cultured at low density and under different conditions than used in our laboratory (28). Although a number of mitogens and growth factors can induce both protein synthesis and gene expression in NVM, there are differences in the signaling cascades leading to these end points (32). Thus it is not surprising that combined treatment with PMA and IL-1 had different effects on COX-2 expression and protein synthesis.

PMA stimulation of COX-2 and prostanoid production had no effect on the growth of NVM. Two different COX inhibitors were used in these experiments, one nonselective and one selective for COX-2, and neither affected PMA-stimulated protein synthesis even when COX-2 and prostanoid production were stimulated by cotreatment with PMA + IL-1 or by IL-1 alone. COX inhibition with nonsteroidal anti-inflammatory drugs has proven effective in inhibiting colon cancer growth in vivo and colon cancer cell proliferation in vitro (33), as well as serum- or Ras-stimulated proliferation of fibroblasts (24).

There are several possible explanations for the difference between our data and those cited. The most likely explanation is that prostanoids have cell-specific effects on growth. In contrast to studies using fibroblasts or cancer cell lines (24, 33), Matsell et al. (25) have shown that IL-1-stimulated PGE₂ production inhibits proliferation of mesangial cells. Another recent study indicates that PGA₂ mediates programmed cell death (apoptosis) in colorectal carcinoma cells (7), which would act to decrease cell number. Also, different classes of prostanoids are produced by different cells. Studies indicate that under control conditions, cardiac myocytes mostly produce PGI₂ and PGE₂ (20, 27), whereas PGF₂ and thromboxane A₂ are produced at a much lower level (27). Previous data from LaPointe and Sitkins (20) plus the current study support this. In addition, treatment with PMA results in a greater absolute amount of PGE₂ than PGF₂, although both are still in the nanomolar range of concentration in the culture medium. On the other hand, IL-1 results in massive stimulation of PGE₂ such that micromolar levels can be measured in the medium, and levels can be one to two orders of magnitude greater than PGF₂. Interestingly, several studies indicate that only PGF₂ promotes growth of cardiac myocytes in vitro and that the maximal stimulatory effect is seen at 1 µM (1, 16). Cardiac-derived endothelial cells and fibroblasts synthesize prostanoids, including PGF₂ (22, 36), and PGF₂ production is enhanced in the infarcted heart (16, 36). Thus a reasonable conclusion of our data is that COX-2 expression in myocytes does not result in sufficient PGF₂ production to affect their growth. However, PGF₂ produced within the heart by endothelial cells and fibroblasts could act on myocytes to induce growth.

In conclusion, PMA stimulates hypertrophic growth, COX-2 synthesis, and PGE₂ and PGF₂ production in NVM. Exogenous PGF₂ is involved in the growth process, but endogenous COX-2 products are not. Both IL-1 and PMA need PKC to induce COX-2 synthesis. It would be interesting to learn the consequences of COX-2-induced prostaglandin production on other aspects of cardiac myocyte function, because COX-2 has been implicated in a number of inflammatory diseases as well as in heart failure (6).