MKK6 activates myocardial cell NF-kappaB and inhibits apoptosis in a p38 mitogen-activated protein kinase-dependent manner.

In cardiac myocytes the stimulation of p38 mitogen-activated protein kinase activates a hypertrophic growth program and the induction of the cardiac-specific genes associated with this program. This study focused on determining whether these novel growth-promoting effects are accompanied by the p38-mediated inhibition of apoptosis, and if so, what signaling pathways might be responsible. Primary neonatal rat ventricular myocytes were driven into apoptosis by treatments known to induce apoptosis in other cell types, e.g. incubation with anisomycin or overexpression constitutively active MEKK-1 (MEKK-1COOH), a protein that strongly activates extracellular signal-regulated kinase and N-terminal c-Jun kinase, but not p38. Overexpression of constitutively active MKK6, MKK6 (Glu), which selectively activates p38 in cardiac myocytes, protected cells from either anisomycin- or MEKK-1COOH-induced apoptosis. This protection was blocked by SB 203580, a selective p38 inhibitor. MKK6 (Glu) also activated transcription mediated by NF-kappaB, a factor which protects other cell types from apoptosis. The activation of NF-kappaB and the protection from apoptosis mediated by MKK6 (Glu) were both blocked by SB 203580. Interestingly, overexpression of a mutant form of I-kappaBalpha, which inhibits nuclear translocation of NF-kappaB, completely blocked MKK6 (Glu)-activated NF-kappaB but had little effect on MKK6s anti-apoptotic effects. These findings suggest that, in part, the overexpression of MKK6 (Glu) may foster growth and survival of cardiac myocytes by protecting them from apoptosis in a p38-dependent manner. Additionally, while NF-kappaB is activated in myocardial cells by p38, this does not appear to be the major mechanism by which MKK6 (Glu) exerts its anti-apoptotic effects in this cell type, suggesting a novel pathway for p38-mediated protection from apoptosis.

The proper growth, development, and function of most tissues requires the appropriate balance between cell growth and death (1). Although a great deal is known about the molecular aspects of cell growth, the nuances of the mechanisms used by cells to regulate death have only recently begun to be appreciated. Among the most widely studied forms of cell death is apoptosis, a form of programmed cell death, or cell "suicide" that is believed to be responsible for the deletion of unwanted cells during organ and tissue development, as well as pathologically induced tissue damage. Apoptosis is an energy-requiring molecular suicide program characterized by cytoplasmic shrinkage, nuclear condensation, and DNA fragmentation into 200-base pair fragments (2)(3)(4)(5)(6). Another form of cell death, necrosis, is usually the result of massive cell injury characterized by decreased mitochondrial function and ATP depletion, cytoplasmic swelling, membrane permeabilization, and DNA fragmentation into random sizes. In addition to being activated during development-related cell reduction, apoptosis can be triggered in many cell types by various stresses, including ionizing radiation, osmotic stress, expression of viral proteins such as E1A, and exposure to certain cytokines (7).
Myocardial cell apoptosis is believed to contribute to cardiac dysfunction, including ischemia/reperfusion injury, vascular wall remodeling, heart failure, and myocardial infarction (8). For example, in humans the cardiomyopathy resulting from intractable congestive heart failure has been attributed in part to the loss of myocytes through apoptosis (9). Mimicking cardiac overload by stretching cultured myocardial cells initially activates myocyte growth (10,11) but later induces myocyte apoptosis (12). Accordingly, it is believed that although the initial response of the myocardium to overload may be cardiac myocyte hypertrophy, prolonged overload results in a remodeling of the myocardium that includes increases in fibroblast growth and collagen deposition, accompanied by decreased numbers of cardiac myocytes, a result of apoptosis (13).
Myocardial cell apoptosis may also serve important developmental roles (14). For example, in the postnatal heart there is a decrease in right ventricular muscle mass that takes place as an adaptation to the decreased load in that chamber just after birth; this selective decrease in tissue mass, which takes place while the left ventricle continues to grow, is the result of chamber-specific apoptosis of cardiac myocytes (12,15). Apoptosis has also been shown to account for cell death in the conotruncal cushions; this programmed cell death mediates remodeling of the bulbus cordis during early development (16,17). Also, a cardiac-specific growth factor, cardiotrophin 1, which is believed to play important roles during cardiac development, is believed to protect cardiac myocytes from apoptosis (18). Accordingly, a better understanding of apoptosis in the myocardium is required to establish a clearer picture of cardiac pathology as well as normal growth and development of the heart.
Studies of signaling mechanisms have revealed that the mitogen-activated protein kinases (MAP 1 kinases) may play im-portant roles in the pathways that regulate both growth and apoptosis. For example, the MAP kinase, extracellular signalregulated kinase (ERK), has been implicated as a growth promoter, displaying an apoptosis-protective effect in some cell types, while N-terminal c-Jun kinase (JNK) and/or p38 MAP kinases have been observed to foster apoptosis in certain cells. In NIH 3T3 and REF52 fibroblasts, overexpression of MEKK1 COOH , a constitutively active form of MEKK1 and a well characterized upstream activator of JNK, leads to apoptosis, which can be attenuated by ERK activation (19,20). In PC-12 cells and sympathetic neurons, nerve growth factormediated ERK activation is crucial for cell survival and neurite outgrowth. The removal of nerve growth factor from PC-12 cells leads to apoptosis (21,22) as well as the activation of JNK and p38 and the inhibition of ERK (19,23). In PC-12 cells the activation of JNK and p38 and the inhibition of ERK are critical for induction of apoptosis. Thus, PC-12 cell survival, neurite extension, and differentiation require the proper balance between the activities of ERK, JNK, and p38 MAP kinases.
Like differentiated neurons, cardiac myocytes are post-mitotic, electrically excitable cells that respond to growth factor treatment with an augmentation of the differentiated phenotype. In response to growth factors, primary cardiac myocytes undergo a hypertrophic growth program typified by an increase in cell size, enhanced sarcomeric organization, activation of spontaneous contractile activity, and the up-regulation of a group of cardiac genes that are generally expressed during early embryonic development (e.g. ANF, brain natriuretic peptide, ␣-skeletal actin), or at very low levels in the adult myocardium (e.g. ␤-myosin heavy chain, myosin light chain-2) (e.g. Ref. 24). This growth program not only recapitulates cardiac development, but probably also represents some of the cellular processes that contribute to the maintenance of a properly functioning myocardium in the adult. Given this background it seems possible that myocardial cell ERK activation might promote cell growth, while the activation of JNK and p38 might result in apoptosis. However, recent observations suggest that in cultured myocardial cells, p38 activation leads to hypertrophic growth (25) and might, therefore, protect against apoptosis, while ERK activation, although required for growth, is not in itself sufficient to support the hypertrophic phenotype (26). Accordingly, the present study was undertaken to address the hypothesis that p38 activation can protect primary myocardial cells from apoptosis.

MATERIALS AND METHODS
Cell Culture-Primary ventricular myocytes were prepared from 1-4 day-old neonatal rats as described previously (27,28). Following the enzymatic dissociation of ventricular tissue the cells were plated onto uncoated plastic dishes in DMEM/F-12 (1:1) (Life Technologies, Inc.) containing 10% fetal bovine serum for 1 h, during which time most of the fibroblasts adhered to the dish. The recovered cells were then either transfected (see below) and then plated or, in some cases, plated directly without transfection. Cultures were plated onto fibronectin-coated glass slides for TUNEL analyses, 100-mm fibronectin-coated plastic culture dishes for DNA fragmentation analyses, or onto fibronectin-coated 16or 24-mm plastic dishes for reporter enzyme analyses. After 18 h in DMEM/F-12 (1:1), 10% fetal bovine serum, the cultures were washed briefly with DMEM/F-12 (1:1) and then refed with serum-free DMEM/ F-12 (1:1), which contained phenylephrine (10 M), 1 M propranolol, transferrin (5 g/ml), insulin (1 g/ml), selenium (0.1 ng/ml), and triiodothyronine (0.5 ng/ml), unless otherwise stated. After 24 h, various test agents (e.g. anisomycin or sphingosine) were then added, and the cultures were maintained for an additional 48 to 96 h in the test agents followed by TUNEL analysis or DNA extraction. For reporter enzyme assays cells were maintained in DMEM/F-12 (1:1) for 48 h after transfection (see below).
Transfections-After preplating (see above), myocardial cells were usually resuspended at a density of 30 million cells/ml of minimal medium (DMEM/F-12 containing 1 mg/ml bovine serum albumin), and transfections were carried out as described previously (27,28). Briefly, for each transfection, 300 l, between 5 and 9 million cells, were mixed with 1.5-45 g of test construct (see below) and 9 g of CMV-␤galactosidase to normalize for transfection efficiency or to serve as a transfection marker in the TUNEL analyses (see below). The levels of plasmid used in each culture within an experiment were equalized using empty vector DNA, such as pCEP. Each 300-l aliquot was then electroporated in a Bio-Rad gene pulser at 500 V, 25 microfarads, 100⍀ in a 0.2-cm gap cuvette, a protocol that allows for the selective transfection of only cardiac myocytes (25,27). Under these conditions only cardiac myocytes are transfected, and the resultant viability is approximately 30% (27); accordingly, the 1-3 million viable cells were plated into four-chamber Lab Tek chamber slides at 0.15-0.9 ϫ 10 6 cells/2-cm 2 well. In two experiments (DNA ladder, Fig. 1B; ERK, JNK, and p38 kinase assays, Fig. 1C) cultures were plated into 100-mm fibronectincoated culture dishes, and in one experiment, cultures were also transfected with NF-B/Luc (29) and plated onto 35-mm fibronectin-coated plastic culture dishes (Fig. 5A).
Test Expression Constructs-To assess the effects of various signaling proteins, the following constructs were used: RSV-Raf-1 BXB (codes for activated Raf-1 kinase; from U. Rapp and U. Wurzburg, Wurzburg, Germany), pCMV5 MEKK COOH (codes for activated MEKK-1; from G. TUNEL-TUNEL of fragmented DNA was performed on transfected myocardial cells plated on fibronectin-coated glass slides essentially as described previously (30) and according to the manufacturer's protocol (Boehringer Mannheim), with the exception that either an ANF or a ␤-galactosidase monoclonal antibody was added to the TUNEL incubation mixture. This allowed for the co-staining of cells for ANF, which allowed the positive identification of cardiac myocytes, and/or ␤-galactosidase, which allowed the identification of transfected cells, and DNA strand breaks. In previous studies we have carried out double staining experiments on cultures transfected with luciferase and ␤-galactosidase-expressing reporters and shown that the co-transfections efficiency is 85% (25). This provided positive confirmation that the observed TUNEL-positive cells were cardiac myocytes (ANF-positive) and were transfected with the appropriate test constructs (␤-galactosidase-positive). In each experiment 100 ␤-galactosidase or ANF-positive cells were assessed for nuclear staining using the TUNEL assay.
DNA Fragmentation-To assess the effects of anisomycin or sphingosine on the structural integrity of DNA, myocardial cells were plated at 2 ϫ 10 6 cells/100-mm plastic culture dish, treated as described under the "Cell Culture" section of "Materials and Methods," above, and then lysed by suspending the cells in 10 mM Tris (pH 8), 100 mM NaCl, 25 mM EDTA, 0.5% SDS, and 0.5 mg/ml DNase-free proteinase K and incubated for 5 h at 55°C. Following two phenol extractions and one chloroform extraction, the DNA was precipitated with isopropyl alcohol and incubated with RNase (10 units of DNase-free RNase) for 1 h at 37°C. Following a final ethanol precipitation, 5 g of DNA derived from each treatment were fractionated on a 2% agarose gel, and the DNA was viewed following staining of the gel with ethidium bromide.
p38 Inhibits Myocardial Cell Apoptosis were incubated for 2 h at 4°C with anti-ERK (raised against the C-terminal 16 amino acids of ERK-1; Santa Cruz SC-093) or anti-JNK (raised against the C-terminal 17 amino acids of JNK-1; Santa Cruz SC-474) bound to protein A-Sepharose (Amersham Pharmacia Biotech), and immune complex kinase assays were carried out using the appropriate substrates, as described previously (26,31). Briefly, reactions were initiated by the addition of 1 g of myelin basic protein for ERK, GST-c-Jun for JNK, and 6 M [␥-32 P]ATP (5000 Ci/mmol) in a final volume of 30 l of kinase buffer (20 mM HEPES (pH 7.4), 20 mM MgCl 2 , 20 mM ␤-glycerophosphate, 2 mM dithiothreitol, 20 M ATP). After 30 min at 25°C, the reactions were terminated by the addition of Laemmli sample buffer, and the phosphorylation level of substrate proteins was evaluated by SDS-polyacrylamide gel electrophoresis followed by autoradiography and phosphorimage analyses. p38 assays were carried out as described previously (25). Briefly, myocardial cells (1.5 ϫ 10 6 cells/ 35-mm culture well) were extracted in 80 l of Laemmli buffer containing 1 mM p-nitrophenyl phosphate, 100 M sodium o-vanadate, 10 g/ml aprotinin, 10 g/ml leupeptin, and 1 mM phenylmethylsulfonyl fluoride. Sixty-l aliquots of each extract were fractionated on a 10% SDSpolyacrylamide gel followed by Western blotting using a p38 antibody specific for the phosphorylated/activated form of the kinase (specific for phosphothreonine 180/phosphotyrosine 182; New England Biolabs, Inc., catalog number: NE 92115). In each experiment three identically treated cultures (1.5 ϫ 10 6 cells/35-mm dish) were used for each treatment, and following densitometric analyses of the exposed phosphorimage plates, values for each treatment were averaged.
Reporter Enzyme Assays-To test the effects of various test constructs on NF-B activation, each culture of approximately 3 ϫ 10 6 myocardial cells was transfected with 10 g of NF-B/Luc and 9 g of CMV-␤-galactosidase and 10 g of the appropriate test construct and then plated into a 35-mm fibronectin-coated plastic culture dish. Following a 48-h incubation in serum-free DMEM/F-12 (1:1), the cultures were extracted and luciferase and ␤-galactosidase assays were performed as described (27). Luciferase activity was measured for 30 s on FIG. 1. Effects of anisomycin or sphingosine on DNA integrity and ERK, JNK, and p38 activities in myocardial cells. A, primary neonatal rat ventricular myocytes were prepared and maintained for 48 h Ϯ test agent as described under the "Cell Culture" section of "Materials and Methods." TUNEL analyses were carried out as described previously (30) and under "Materials and Methods." The TUNEL procedure was carried out following the instructions for the in situ fluorescein cell death kit (Boehringer Mannheim) with the exception that a mouse anti-ANF antibody (32) was added to the TUNEL reaction mixture. Upon viewing using a rhodamine-compatible fluorescence filter, cardiac myocytes were identified as ANF-positive cells. 100 ANF-positive cells were scored for apoptosis using the TUNEL assay. The percentage of ANF-positive cells that were TUNEL-positive is shown as the mean of three identical cultures Ϯ S.E.; the anisomycin results are shown using the lefthand y axis, and the sphingosine results are shown using the righthand y axis. **, p Ͻ 0.01; *, p Ͻ 0.05 compared with control (no anisomycin or sphingosine). B, DNA integrity was evaluated as described previously (30) and under "Materials and Methods." Shown are the results of an agarose gel on which 10-g aliquots of the DNA derived from cells receiving either no test agent (Con), 0.20 M anisomycin (An), or 10 M sphingosine (Sph) are viewed. For comparison a sample of DNA comprised of fragments differing in size by 100-base pair increments is shown (Ladder). These results are representative of at least three experiments. C, endogenous ERK and JNK activity levels were evaluated as described previously (25) and under "Materials and Methods." Cultures plated identically to those described in B received either no test agent (Con), 0.20 M anisomycin (An), or 10 M sphingosine (Sph), three cultures per treatment. Following either 30-min exposures to anisomycin or sphingosine, previously determined to be maximal exposure times for each of these kinases, cultures were extracted and ERK or JNK activity levels were determined using immune complex assays and either myelin basic protein or GST-c-Jun for the substrates. Shown are the phosphorimager analyses of the gels for ERK (myelin basic protein) and JNK (GST-c-JUN); lanes 1-3 are control; lanes 4 -6 are anisomycin, and lanes 7-9 are sphingosine. The bar graphs represent the densitometric conversions of the phosphorimager results and are plotted as the mean kinase values of three cultures per treatment Ϯ S.E. These results are representative of at least three experiments. **, p Ͻ 0.01 compared with control (no anisomycin or sphingosine). D, endogenous p38 activity levels were evaluated as described previously (25) and under "Materials and Methods." Cultures plated identically to those described in B received either no test agent (Con), 0.20 M anisomycin (An), or 10 M sphingosine (Sph), three cultures per treatment. Following either 30-min exposures to anisomycin or sphingosine, p38 activity levels were determined by Western blotting using a phospho-p38-specific antibody (shown as P-p38), while total p38 levels were determined using a non-phospho-specific p38 antibody (shown as p38) (see "Materials and Methods"). The relative densities of these blots were determined using a phosphorimager, and the P-p38 values were divided by the total p38 values to obtain the level of p38 Thr/Tyr phosphorylation and thus activation. The results of the phosphorimager analyses are shown as the bar graph depicting the mean kinase values of three cultures per treatment Ϯ S.E. These results are representative of at least three experiments. **, p Ͻ 0.01 compared with control (no anisomycin or sphingosine).
a Bio Orbit 1251 Luminometer (Amersham Pharmacia Biotech). Data are expressed as "relative luciferase (Rel Luc)" ϭ arbitrary integrated luciferase units/␤-galactosidase units, representative of at least three independent experiments performed with two different plasmid preparations, and represent the mean and S.E. of triplicate 35-mm wells.
Replicates and Statistical Treatment-All of the results shown are representative of at least three different experiments identical, or similar to, the one shown. Each value represents the mean (n ϭ 3) Ϯ S.E. Statistical analyses were performed using a one-way Newman-Keuls post hoc analysis of variance.

RESULTS AND DISCUSSION
To establish a model for studying the role of the p38 pathway on myocardial cell apoptosis, experiments were carried out to evaluate the effects of several known apoptotic agents on primary neonatal rat ventricular myocytes. Cultures were treated with either anisomycin or sphingosine and following fixation they were stained for ANF expression using an ANF monoclonal antibody (32) and for DNA fragmentation using the TUNEL assay (30,33,34). Only the cells that stained positively for ANF were evaluated for DNA fragmentation, thus ensuring the evaluation of apoptosis in cardiac myocytes only.
Both anisomycin and sphingosine displayed dose-dependent abilities to induce myocardial cell apoptosis (Fig. 1A). Within 48 h of treatment with 0.2 M anisomycin, approximately 20% of the transfected cells scored positive for apoptosis compared with about 1% of the untreated cells, representing a 20-fold induction of apoptosis. In other experiments levels of anisomycin greater than 0.2 M did not significantly increase the percentage of cells driven into apoptosis (not shown). As expected (30), sphingosine also strongly induced apoptosis within the same time frame, with greater than 80% of the cells exposed to 10 M sphingosine scoring positive for apoptosis compared with about 1% of the untreated cells.
Although anisomycin was a less powerful apoptotic agent of these two compounds, we found that it produced more consistent results than sphingosine. In contrast, the effects of sphingosine were more variable, resulting sometimes in such severe cell death, that by the end of 48 -72 h, most of the cells had detached from the dish. This detachment made it impossible to score them for ANF expression, transfection, or TUNEL staining. Accordingly, in most of the experiments shown in later figures, the soluble apoptotic agent of choice was anisomycin.
As further characterization of the apoptotic response, DNA was isolated from cells treated with 0.20 M anisomycin or 10 M sphingosine and analyzed on an agarose gel. The 200-base Following maintenance for 72 h, as described under the "Cell Culture" section of "Materials and Methods," the cells were fixed and submitted to the TUNEL procedure following the instructions for the in situ fluorescein cell death kit (Boehringer Mannheim), with the exception that a mouse anti-␤-galactosidase antibody (Promega) was added to the TUNEL reaction mixture, and ␤-galactosidase expression was detected using a Texas Red-conjugated anti-mouse IgG. Cells were then examined by fluorescence microscopy using a rhodamine-compatible filter (A and C), which allowed them to be scored for ␤-galactosidase production (i.e. transfected), and the same cells were then examined using a fluoresceincompatible filter (B and D), which allowed them to be scored for DNA strand breaks indicative of apoptotis. In E cells were transfected with the test constructs shown and after 72 h, between 50 and 100 ␤-galactosidase-positive cells were scored for positive nuclei assessed using the TUNEL assay. In F myocardial cells were transfected with CMV-␤-galactosidase and either an empty vector control (pCEP) or a plasmid encoding a constitutively active form of MKK6. Following maintenance for 24 h Ϯ anisomycin, Ϯ SB 203580 (10 M), as described under the "Cell Culture" section of "Materials and Methods," the cells were fixed and submitted to the TUNEL procedure and stained for ␤-galactosidase expression as described in the legend for Fig. 2. The percentage of the transfected cells that were TUNEL-positive is shown as the mean of three identical cultures Ϯ S.E. E and F, these results are representative of at least three experiments. **, p Ͻ 0.01 compared with all other treatments.
p38 Inhibits Myocardial Cell Apoptosis pair fragments characteristic of apoptotic cells (35) were evident in the anisomycin-and sphingosine-treated cells, but not in the control cells (Fig. 1B). Also, consistent with the relative potencies of the compounds in the TUNEL analysis, the fragmentation pattern was much stronger in the sphingosinetreated cultures than in those treated with anisomycin.
To begin addressing how MAP kinases might be involved in apoptosis in cardiac myocytes in response to anisomycin or sphingosine, the activity levels of ERK, JNK, and p38 were assessed. Anisomycin activated myocardial cell JNK by 10-fold and p38 by 11-fold; however, it had no effect on ERK, while sphingosine activated all three MAP kinases by about 5-10fold (Fig. 1, C and D). Thus, it did not seem likely that ERK could be involved in apoptosis mediated by both of these compounds; however, either JNK or p38 might be involved.
To assess the effects of each of the MAPK pathways on myocardial cell apoptosis, cultures were transfected with plasmids encoding constitutively active forms of Raf-1 kinase, MEKK1 or MKK6, which preferentially activate ERK, JNK, or p38, respectively (25). Only the myocytes that overexpressed constitutively active MEKK1 (MEKK1 COOH ) displayed significant increases in the number of TUNEL-positive cells (Fig. 2,  A-D), which amounted to approximately 30-fold over cells transfected with a control plasmid (Fig. 2E). Moreover, the TUNEL-positive nuclei in the MEKK1 COOH -transfected cells were pyknotic (Fig. 2D), consistent with the double strand DNA breaks characteristic of apoptosis (34). Since MEKK1 COOH serves as such a strong activator of JNK in myocardial cells, while BXB and MKK6 activate mainly ERK and p38 (25), these results were consistent with the hypothesis that activation of components of the JNK pathway could foster apoptosis in cardiac myocytes.
Given the recent finding that MKK6-activated p38 contributes to myocardial cell hypertrophic growth (25), and since MKK6 overexpression did not appear to induce apoptosis (Fig.  2E), we investigated whether the overexpression of constitutively active MKK6, MKK6 (Glu), could protect against apoptosis. As expected, compared with untreated controls, cultures transfected with the empty vector control, pCEP, displayed about 15-fold more TUNEL-positive cells in response to anisomycin treatment (Fig. 2F). However, the overexpression of MKK6 (Glu) decreased anisomycin-induced apoptosis to nearly control levels, supporting the view that in contrast to PC-12 cells, p38 is anti-apoptotic in cardiac myocytes. To further assess the anti-apoptotic effects of the p38 pathway, the p38specific inhibitor, SB 203580 (36), was employed. SB 203580 completely blocked the anti-apoptotic effects of MKK6 (Glu) overexpression (Fig. 2F). Since MKK6 (Glu) is a selective p38 activator in cardiac myocytes (25), and since SB 203580 also displays p38 specificity in cardiac myocytes, 2 these results strongly suggest that it is through p38 that MKK6 confers this protection against anisomycin-mediated programmed cell death.
We next evaluated whether MKK6 (Glu) overexpression could effect protection against MEKK1 COOH -induced apoptosis. As expected, the TUNEL-positive, MEKK1 COOH -transfected cells frequently displayed the pyknotic nuclei that are typical of cells undergoing apoptosis (Fig. 3D), while most of the cells transfected with MKK6 (Glu) alone (Fig. 3F) and many of the cells co-transfected with MKK6 (Glu) and MEKK1 COOH (Fig.  3H) did not score positive in the TUNEL assay. The assessment of 100 cells from each transfection revealed that MKK6 (Glu) 2 Even though SB 203580 is selective for the inhibition of p38 in other cell types (74), we carried out control experiments to establish the p38 specificity of SB 203580 in cardiac myocytes. In one experiment SB 203580 was shown to completely block MKK6 (Glu)-stimulated MEF2C transcriptional activation in cardiac myocytes (25); this mode of transcriptional activation has previously shown to be mediated by p38 and not JNK or ERK (73). In another experiment we transfected primary myocardial cells with ⌬MEKK4 (75), a constitutively active selective stimulator of JNK. The cultures were incubated Ϯ 10 M SB 203580 for 48 h, and after extraction, a JNK assay was performed; the cultures incubated with SB 203580 displayed the same level of JNK as those incubated without SB 203580. Additionally, in one experiment 10 M SB 203580 was added to the immunoprecipitated JNK derived from ⌬MEKK4-transfected myocardial cells, and there was no effect of the compound on JNK enzyme activity.

FIG. 3. Effects of MKK6 (Glu) on MEKK1 COOH -induced apoptosis in myocardial cells. Myocardial cells were transfected with CMV-
␤-galactosidase and/or constitutively active forms of MEKK1, MKK6, and/or the control plasmid, pCEP, as shown. Following maintenance for 72 h, as described under the "Cell Culture" section of "Materials and Methods," the cells were fixed and submitted to the TUNEL procedure and staining for ␤-galactosidase expression as described in the legend for Fig. 2. Cells were then examined by fluorescence microscopy using either a rhodamine-compatible filter (A, C, E, and G) and scored as ␤-galactosidase-positive (i.e. transfected) and then the same cells were re-examined using a fluorescein-compatible filter (B, D, F, and H) and scored as TUNEL-positive. 100 ␤-galactosidase-positive cells were scored for positive nuclei assessed using the TUNEL assay. The percentage of the transfected cells that were TUNEL-positive is shown as the mean of three identical cultures Ϯ S.E. (I). **, p Ͻ 0.01 compared with all other treatments.
p38 Inhibits Myocardial Cell Apoptosis 8236 could decrease the numbers of TUNEL-positive cells in MEKK1 COOH -transfected cells from about 6-fold over control to about 2-3-fold over control (Fig. 3I). The partial inhibition of MEKK1 COOH -induced apoptosis by MKK6 (Glu), which amounted to about 50%, suggests that the apoptotic effects of MEKK1 COOH may be mediated by a number of pathways, only a subset of which is functionally interrupted by MKK6. This is consistent with recent reports demonstrating the MEKK1 is a large, membrane-associated protein with various functional domains that appear to confer apoptosis via several different pathways (37)(38)(39).
One possible mechanism by which MKK6 (Glu) could protect myocardial cells from apoptosis could be through the activation of NF-B, a transcription factor shown to protect several other cell types from cytokine-induced apoptosis (29, 40 -44). In support of this hypothesis is a study showing that p38-specific inhibitors block tumor necrosis factor-mediated NF-B activation in L929 cells (45) and another report that p38 and NF-B are often activated by the same stimuli in 293 cells (46).
Accordingly, experiments were carried out to determine whether overexpression of MKK6 (Glu) can activate NF-B in primary myocardial cells. Cultures were co-transfected with a luciferase reporter construct possessing two copies of a consensus NF-B binding site and with various test constructs. Two different known activators of p38, MKK6 (Glu) (47), and MKK6b (Glu) (48) were able to activate NF-B-mediated transcription by about 10-fold in an SB 203580-sensitive manner (Fig. 4A). A kinase-dead form of MKK6, MKK6 (K82R) (47), was unable to activate NF-B-mediated transcription. Additionally, overexpression of a form of I-B␣, that cannot dissociate from NF-B (I-B␣M), thus rendering it inactive in the cytosol (42,49), was a potent blocker of MKK6 (Glu)-activated NF-B-mediated transcription. These results clearly showed that using this assessment of NF-B activation, the p38 pathway is able to activate NF-B in primary myocardial cells.
Further experiments were carried out to test whether it is through NF-B that MKK6 (Glu) protects cardiac myocytes from apoptosis. Anisomycin-induced apoptosis was blocked by MKK6 (Glu) in an SB 203580-sensitive manner; however, transfecting cells with I-B␣M, which inhibits NF-B activation, had no effect on MKK6 (Glu)-mediated protection from anisomycin-induced apoptosis (Fig. 4B). This suggested that the protective effects of MKK6 (Glu) were not due to its ability to activate NF-B. Interestingly, SB 203580 treatment of cells that had been transfected with both MKK6 (Glu) and I-B␣M augmented anisomycin-induced apoptosis. This finding suggests that when p38 is blocked, NF-B can serve to protect against apoptosis. To further test this hypothesis, apoptosis was assessed in myocardial cells in the absence of anisomycin. Under these conditions, SB 203580 or I-B␣M alone had no statistically significant effect on the number of apoptotic cells; however, SB 203580 and I-B␣M together increased apoptosis by about 5-fold (Fig. 4C). These results support the view that NF-B can protect cardiac myocytes against apoptosis, but only upon inhibition of p38.
Taken together, the results of the present study indicate that MKK6, a well characterized p38 MAP kinase kinases (47,48), can activate NF-B-mediated transcription in a manner dependent on p38. Moreover, it appears that MKK6 can protect cells from apoptosis via p38 activation; however, this protection does not seem to involve NF-B activation, but must function through alternate anti-apoptotic pathways. To our knowledge this is the first study demonstrating the anti-apoptotic effects of MKK6 (Glu) in any cell type, 3 suggesting a potential widespread importance of the p38 pathway in maintaining the appropriate balance between cell growth and cell death.
While it is unknown how p38 could protect myocardial cells from apoptosis, an understanding of the recently discovered downstream targets of p38 might provide some clues. For example, the MAP kinase-activated protein kinases (MAP-KAPs)-1, -2, -3 (MAPKAP-3 a.k.a. 3pK) have been shown to be activated by p38 (36, 50 -56). MAPKAPs phosphorylate several proteins, including CREB and ATF-1 (53) and heat-shock protein hsp27 (57,58). Accordingly, it may be through one or more of these MAPKAP-activated pathways that p38 confers protection from apoptotis in cardiac myocytes.
MAPKAP-mediated hsp27 phosphorylation is a prime candidate pathway by which p38 could serve an anti-apoptotic function in cardiac myocytes. For instance, MAPKAP-2 has been shown to be activated during ischemic preconditioning of isolated rat hearts (59,60). Such preconditioning is known to serve as a myocardial stress adaptation, resulting in enhanced protection from ischemia-induced myocardial cell death (61)(62)(63)(64). In part, it is believed that this cardioprotection is derived from the induction and activation of heat-shock proteins 27 and 70 (62,65,66), both of which are known to protect cells from apoptosis (67)(68)(69). Interestingly, after phosphorylation induced by either heat-shock or mitogen stimulation, hsp27 has been shown to bind to and stabilize actin filaments in mouse fibroblasts (70). Such hsp27-mediated filament stabilization in cardiac myocytes could be a major contributor to the striking sarcomeric organization observed upon MKK6-mediated p38 activation. Intracellular signaling pathways leading to hsp activation and/or phosphorylation in cardiac myocytes could also be mediated by ␣ 1 -adrenergic receptors. Supporting this view are studies showing that adrenergic receptor stimulation activates heat-shock proteins in a variety of cell types, most notably in rat aortic cells (71) and in rat cardiac cells (72).
Alternatively, or in combination with heat-shock proteins, muscle cell-enriched transcription factors, the MEFs, which also serve as downstream targets of p38 (73), could function to protect cardiac myocytes against apoptosis. For example, many of the genes that are induced during the hypertrophic growth program possesses important A/T-rich sequences that are required for inducibility by growth stimuli and that may bind MEFs. It is conceivable that, in part, the induction of some of these genes play roles as growth promoters by protecting the cell from apoptosis. Interestingly, p38 activates MEF-mediated transcription in primary myocardial cells (25), which is consistent with a role for this transcription factor in protecting from apoptotis.
In summary, the results presented in this paper demonstrate that MKK6-mediated activation of p38 MAP kinase leads to the protection of cardiac myocytes from apoptosis induced by either anisomycin or MEKK1 COOH . This is the first demonstration of an anti-apoptotic role for the p38 MAP kinase pathway, and it correlates with the recently discovered role that p38 plays in mediating the hypertrophic growth response in myocardial cells (25). These results also compliment another recent study using the same cell model system which showed that cardiotro-phin 1 protects cells from apoptosis, but through a pathway that appears to require ERK (18). Thus, it appears that in myocardial cells ERK and p38 are anti-apoptotic, and JNK may promote apoptosis. Future studies aimed at determining the molecular mechanism by which p38 exerts anti-apoptotic, growth-promoting effects in cardiac myocytes will likely reveal important new aspects of this stress kinase pathway in the regulation of the delicate balance between cell growth and cell death.