Inducible expression of a constitutively active mutant of mitogen-activated protein kinase kinase 7 specifically activates c-JUN NH2-terminal protein kinase, alters expression of at least nine genes, and inhibits cell proliferation.

MKK7 is a recently discovered mitogen-activated protein kinase (MAPK) kinase that is unique in that it specifically activates only the c-JUN NH(2)-terminal protein kinase (JNK) family of enzymes. Very little is known about the biological role of MKK7. We generated inducible cell lines from the human embryonal kidney carcinoma cell line, HEK293, by stable transfection with a constitutively active mutant of MKK7, MKK7(3E), fused to green fluorescent protein (GFP), under the control of an ecdysone-inducible promoter. Treatment of cells with the synthetic ecdysone analog ponasterone A induced expression of GFP-MKK7(3E) and resulted in sustained activation of endogenous JNK, but neither of the other endogenous MAPKs, ERK or p38. Red and green fluorescing cDNA copies of mRNA extracted from cells obtained before and after induction of GFP-MKK7(3E) were hybridized to microarrays containing more than 6,000 cDNAs in eight independent experiments. By selection criteria, 23 genes were differentially regulated after 24 h of induction of GFP-MKK7(3E) and 16 after 48 h. The expression of 9 genes was consistently changed after both 24 and 48 h of induction. These changes included down-regulation of three genes, c-myc, angiopoietin-2, and glucose-regulated protein 58, and up-regulation of 6 genes, tissue factor pathway inhibitor-2, GRP78, autotaxin, PPP1R7, the DKFZ cDNA p434D0818, and 1 unknown gene. Consistent with previously described roles of several of the altered genes, MKK7(3E) inhibited cell proliferation. These data implicate active MKK7 in the negative regulation of cell proliferation and provide evidence for a new role for this kinase in the regulation of a distinct, hitherto unrecognized set of genes.

The c-JUN NH 2 -terminal protein kinase (JNK) 1 family of enzymes regulates a broad spectrum of biological processes including inflammation, apoptosis, development, and tumorigenesis (1,2). In the absence of extracellular stimulation JNKs are inactive. For activation, JNKs require phosphorylation in the conserved motif Thr-Pro-Tyr (1,2). This phosphorylation is brought about by protein kinases of dual specificity, so-called MAP kinase kinases (MKK). Two MKKs have been shown to phosphorylate and, thereby, activate JNK, MKK4, and MKK7 (also called JNK kinase (JNKK) 1 and 2, respectively (1,2)). Whereas MKK4 also activates p38 MAPK, experiments with ectopically expressed or recombinant MKK7 revealed that it activates JNK but not ERK or p38 MAPKs in vivo and in vitro (3)(4)(5)(6). Thus, MKK7 is the only specific direct upstream activator of the JNK pathway identified to date (1,2). Recently, six closely related and highly conserved forms of MKK7 have been identified in mammalian cells. All six are derived from the MKK7 gene by alternative splicing, and their individual functions are not known (7).
Like JNK, MKK7 is activated by cytokines such as interleukin-1 and tumor necrosis factor ␣, as well as by stressors such as sorbitol, anisomycin, and UV light (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13). MKK7, therefore, is likely to serve as the upstream effector molecule that coordinates the cellular response to those extracellular stimuli that ultimately activate JNK. However, the normal physiological role of MKK7 is unknown, and its precise role within the JNK signaling pathway remains elusive. In most cell types, JNK activation occurs through both MKK4 and MKK7, which are ubiquitously expressed and may also synergize. In this situation, MKK7 preferentially phosphorylates the threonine, and MKK4 the tyrosine residue in the JNK activation motif, Thr-Pro-Tyr (3,14,15). Recently, scaffolding molecules, such as JNK-interacting proteins 1-3 (1,2,16), have been identified. These proteins tether JNK, MKK7, and MAPK kinase kinases such as MLK3 (16). In addition, protein interaction domains on JNKs (17) and on their substrates (18) provide mechanisms to recruit JNK to upstream activators as well as to downstream substrates, such as the transcription factor AP-1 (1,2). Collectively, these findings suggest that the specific signaling complexes that are assembled in a particular biological context determine the outcome of the activation of the JNK pathway in each situation (1,2). Consistent with this notion, genetic evidence for Drosophila MKK4 and 7, D-MKK4 and Hep/D-MKK-7, respectively, demonstrated that both JNK activators serve distinct and nonredundant functions in flies (1). Recently, this was confirmed in mice. While MKK4 Ϫ/Ϫ mice die because of liver cell apoptosis (19,20), loss of MKK7 results in embryonic death of unknown origin (21). Thus, biochemical evidence clearly indicates that MKK7 is an important component of the JNK signaling pathway, but the exact physiological consequences of its activation still remain to be defined. Despite the diversity of biological functions that have been ascribed to the JNK pathway, the number of direct JNK substrates identified to date is relatively small, consisting mainly of transcription factors such as c-JUN, D-JUN, ATF-2, ELK-1, and SAP-1 (1,2). This implies that one of the key functions of JNK activation is the regulation of gene expression. The MKK7 isoforms can be found in the nucleus as well as in the cytoplasm, suggesting that they do, indeed, activate nuclear JNK and contribute directly to JNK's activation of transcription factors (7).
Because there is no known extracellular stimulus that exclusively activates MKK7 (and JNK) without also activating MKK4, or other signaling pathways, we used an inducible MKK7 expression vector to activate JNK specifically in intact cells. In this way we were able to isolate the MKK7-JNK pathway and determine its effect on gene expression using high density cDNA microarrays.
Preparation of Cell Extracts-Cells were treated with ponasterone for the indicated times or left untreated. For the preparation of whole cell extracts, the medium was removed, and the cells were placed on ice, washed once in phosphate-buffered saline, and scraped in phosphatebuffered saline. Cells were collected at 500 ϫ g for 5 min and lysed in whole cell lysis buffer (10 mM Tris, pH 7.05, 30 mM NaPP i , 50 mM NaCl, 1% Triton X-100, 2 mM Na 3 VO 4 , 50 mM NaF, 20 mM ␤-glycerophosphate and freshly added 0.5 mM phenylmethylsulfonyl fluoride, 0.5 g/ml leupeptin, 0.5 g/ml pepstatin, 400 nM okadaic acid). After 10 min on ice, lysates were clarified by centrifugation at 10,000 ϫ g for 15 min at 4°C. Nuclear and cytosolic extracts were prepared as described previously (22,23). The protein concentration of cell extracts was determined by the method of Bradford, and samples were stored at Ϫ80°C.
Immune Complex Protein Kinase Assays-Whole cell extract (250 -500 g of protein) was diluted in 500 l of ice-cold immunoprecipitation buffer A (20 mM Tris, pH 7.4, 154 mM NaCl, 50 mM NaF, 1 mM Na 3 VO 4 , 1% Triton X-100). Samples were incubated for 3 h with 2 l of rabbit antibodies against JNK, p38, or ERK MAPK followed by the addition of 20 l of a 50% suspension of protein A-Sepharose beads and incubation for 1-2 h at 4°C. Beads were spun down, washed three times in 1 ml of immunoprecipitation buffer A, and resuspended in 10 l of the same buffer. Then 1 g of recombinant protein substrates (GST-JUN 1-135 , HIS-MK-2, or myelin basic protein) in 10 l of H 2 O and 10 l of kinase buffer (150 mM Tris, pH 7.4, 30 mM MgCl 2 , 60 M ATP, 4 Ci of [␥-32 P]ATP) were added. After 30 min at room temperature, SDS-PAGE sample buffer was added, and proteins were eluted from the beads by boiling for 5 min. After centrifugation at 10,000 ϫ g for 5 min, supernatants were separated on 10% or 12.5% SDS-PAGE. Phosphorylated proteins were visualized by autoradiography.
Western Blotting-Cell extract proteins were separated on 10% SDS-PAGE, and Western was blotting performed as described (22,23). Proteins were detected by using the Amersham enhanced chemiluminescence system.
FACS Analysis-Cells were cultured in six-well dishes, trypsinized after the indicated treatments, and fluorescence analyzed by a FACScan (Becton Dickinson) using the Lysis 2.1 software (Becton Dickinson).
Confocal Microscopy-Subcellular distribution of GFP-tagged MKK7 3E in clone E10 cells (Fig. 1D) and of GFP-tagged wild-type MKK7 (clone 28, data not shown) was examined by differential interference contrast and confocal microscopy after 0, 8, 16, and 24 h of ponasterone induction. Green fluorescence and differential interference contrast (i.e. Nomarski optics) images of the same cell were collected in separate channels using a transmitted light detector. Each fluorescent image represented a 0.5-m section through the cell.
Determination of Cell Number and DNA Synthesis-Cells were seeded in six-well plates and counted after the indicated treatments in a Neubauer chamber. For determination of DNA synthesis rates, 10 4 cells were seeded in 96-well plates and incubated with 0.5 Ci/well [ 3 H]thymidine (Hartmann Analytics) for the final 4 h of treatment. Radioactivity incorporated into cellular DNA was determined by liquid scintillation counting.
Northern Blotting-Northern blotting was performed exactly as described previously (23).
cDNA Array Experiments-Fluorescent cDNA copies of the mRNAs in clone E10 cells before and after induction of expression of activated MKK7 3E were prepared by reverse transcription, principally as described (24). To be specific, the MicroMax Direct Labeling Kit (PerkinElmer Life Sciences/Amersham Biosciences, Inc.) was used as directed by the manufacturer, with 10 -40 g of total RNA and 4 l of Cy3-dUTP or 2 l of Cy5-dUTP in each reaction.
The NCI 6.4 K "oncochip" contained PCR-generated copies of 6317 human cDNA clones, including 5,646 unique clones: 527 are expressed sequence tag clusters and 4,430 are "named" cDNAs (representing genes involved in numerous cellular processes) that had been spotted onto poly-L-lysine-coated slides according to Eisen and Brown (25) using an OmniGrid arrayer (GeneMachines). Hybridized arrays were scanned at 10-m resolution on a GenePix 4000 scanner (Axon Instruments) at variable PMT voltage settings to obtain maximal signal intensities with Ͻ 1% probe saturation. Resulting tiff images were analyzed via GenePix Pro v3.0.6.41 software (Axon Instruments) and Web-based algorithms on the National Cancer Institute/National Institutes of Health microarray (mAdb) Web site. Outliers were defined as the set of genes with expression ratios significantly greater than 2.0 or less than 0.5 for the 12 h and 48 h experiments or greater than 1.5 or less than 0.667 for the 24 h experiments. Additionally, ratio data from probes with signal intensities of Ͻ150 units in one or both channels were excluded from the analyzed data sets.
Initially, a comparison was made of Cy3-labeled cDNA from untreated clone E10 cells with Cy5-labeled cDNA from an independent preparation of RNA from a separate batch of untreated clone E10 cells. Thereafter, each array hybridization consisted of Cy3-labeled cDNA generated from a mixture of the two aforementioned preparations of RNA from untreated clone E10 cells and Cy5-labeled cDNA generated from clone E10 cells that had been treated with ponasterone for 12, 24, or 48 h. Cy3-or Cy5-specific artifactual hybridization was ruled out by reversing the labels in one duplicate hybridization experiment (not shown).
GenBank acquisition numbers and PubMed references for the cDNAs contained on the array used in this study can be found by entering IMAGE:NNNNNN at http://nciarray.nci.nih.gov/CR_query.shtml.

Inducible Expression of a Constitutively Active Mutant of MKK7-
To investigate the possible functions of MKK7, a previously described constitutively active form of MKK7, MKK7 3E (22), was fused at its amino terminus to EGFP and was expressed in HEK293 cells under the control of an ecdysoneinducible promoter. Initially, transient transfection experiments revealed that GFP did not compromise the function of MKK7 3E compared with a version of the enzyme-tagged NH 2terminally with a Myc epitope (data not shown). Treatment of the stable cell line E10 with the synthetic ecdysone analog ponasterone A showed time-and dose-dependent expression of GFP-MKK7 3E as analyzed by Western blot using anti-GFP and anti-MKK7 antibodies. Maximal GFP-MKK7 3E expression occurred between 2 and 10 M ponasterone (Fig. 1A). The protein encoded by the induced transgene was detectable after 6 h and increased to maximal levels between 24 h and 48 h of treatment (Fig. 1B). The maximal level of GFP-MKK7 3E attained was about twice that of the endogenous MKK7 protein (Fig. 1, B and C). Like endogenous MKK7, GFP-MKK7 3E was expressed in the cytosol and in the nucleus, demonstrating that the epitope tag and the mutations did not alter cellular distribution of GFP-MKK7 3E (Fig. 1, C and D). The subcellular distribution of GFP-MKK7 3E as analyzed over time by confocal microscopy was not different from that of inducibly expressed wild-type GFP-MKK7 (data not shown).
In correlation with its expression levels (Figs. 1B and 2A), GFP-MKK7 3E significantly phosphorylated endogenous JNK at late time points of induction ( Fig. 2B) but only weakly at early times after ponasterone treatment (data not shown). These results were also confirmed by measuring the activity of immunoprecipitated endogenous JNK in vitro (Fig. 2C). As expected, GFP-MKK7 3E did not change the activity of ERK or p38 MAPKs (Fig. 3). As shown previously by us for the Myctagged kinases (22), induced expression of the inactive mutants, GFP-MKK7 3A or GFP-MKK7 K149M did not activate JNK (data not shown). Importantly, reverse transcription-PCR (Fig.  4A), immunoprecipitation of metabolically labeled 35 S-GFP-MKK7 3E (Fig. 4B), confocal microscopy ( Fig. 1D), and FACS analysis (Fig. 4C) revealed no significant basal expression of GFP-MKK7 3E at the mRNA, protein, or cellular levels in clone E10 (Figs. 1C, 4, and data not shown). Essentially the same results were obtained for two other GFP-MKK7 3E -expressing clonal cell lines, E4 and E6 (data not shown). Furthermore, in clone E10 about 90% of cells expressed the transgene after induction with ponasterone (Fig. 4C). This cell line was therefore chosen to analyze changes in gene expression at the mRNA level before and after induction of GFP-MKK7 3E .
Expression Analysis of GFP-MKK7 3E -dependent Genes by cDNA Microarrays-Pairs of reverse transcribed, fluorescencelabeled cDNAs were hybridized to each slide that carried the National Cancer Institute cDNA oncochip microarray, which contained more than 6,000 human cDNAs, selected for possible relevance to cancer. Initially, to test for possible biological variations, cDNAs from two different RNA preparations from uninduced cells were labeled separately with red (Cy5) or green (Cy3) fluorophores and compared with one another by mixing and hybridizing to the same array. This experiment revealed no significant differences in expressed genes among noninduced cells (data not shown). A mixture of these two RNAs was used subsequently as the control for each hybridization. It was used to generate cDNA labeled with green fluorescence by reverse transcription with Cy3-tagged dUTP and mixed with red (Cy5)-fluorescing cDNAs from each RNA that had been isolated 24 or 48 h after ponasterone induction of GFP-MKK7 expression. Comparison of the two 48 h experiments revealed that 17 genes were expressed differentially at least 2-fold after induction with ponasterone, of which the ectopically expressed MKK7 was the most strongly induced (Fig. 5). Automated integration of total fluorescence intensity of each spot as well as visual spot inspection showed good reproducibility and sensitivity in the two individual array experiments. Nine genes had been up-regulated and seven down-regulated by more than 2-fold (Fig. 5). To include genes that were regulated differentially over time, additional experiments were performed comparing RNA from cells induced for 24 h to RNA from the noninduced state. In these experiments the detection threshold was deliberately lowered to detect genes with small changes in expression levels also. As shown in Table I this was achieved without compromising the reliability of the results by probing four arrays instead of two. This revealed revealed 24 spots, representing 23 genes (c-myc cDNA was on two spots) that were statistically significantly differentially expressed, some of which were different from those detected after 48 h of ponasterone induction (Table I). In all six experiments, however, the ectopically expressed MKK7 was found to be strongly induced (from 3-to 23-fold, mean 12.9 Ϯ 3.3) as summarized in Fig. 6. A total of nine genes was consistently changed in at least four of the six individual array experiments, i.e. c-myc, tissue factor pathway inhibitor 2 (TFPI-2), the 78-kDa glucose-regulated protein precursor (GRP78), angiopoietin-2, the 58-kDa glucoseregulated protein precursor (GRP58), protein phosphatase 1 regulatory subunit 7 (PPP1R7), ectonucleotide pyrophosphatase (autotaxin), the DKFZ cDNA p434D0818, and one unknown gene (Fig. 6). The most prominent consistent change after 24 h and 48 h of ponasterone induction was the downregulation of the c-myc gene. In good agreement with the repetitive cDNA array experiments, this effect was confirmed by Northern blot analysis of RNAs after 24 h and 48 h of induction (Fig. 7A).
To determine the onset of c-myc suppression, its mRNA was analyzed by Northern blotting at earlier time points after induction of GFP-MKK7 3E . c-myc down-regulation occurred at 8 -12 h of ponasterone treatment (Fig. 7B). Two additional cDNA microarray experiments perfomed at the 12 h time point revealed only a few genes whose expression levels had changed (Table II) regulation of the two myc spots were readily detectable, but smaller than at later time points (Fig. 7B and Table II). Clearly, myc-down-regulation starts at times where expression GFP-MKK7 3E has just begun to increase and may, therefore, be a direct consequence of JNK activation.
To our knowledge, the expression level of none of the differentially expressed genes has been reported to be regulated by MKK7 (or JNK) so far. However, as summarized in Fig. 5, the expression of several of these genes is known to be strongly regulated under stressful conditions or mitogenic stimuli. In addition, we noticed that several genes had been implicated in some form of growth control (Fig. 5, last column). 2 Of particular interest was the c-MYC protein, which plays a pivotal role in cell proliferation and malignant transformation (26 -29).
GFP-MKK7 3E Suppresses Cell Proliferation-To follow up the potential implications of these observations we sought to estimate the impact of the changed pattern of gene expression on cell growth, by analyzing the effects of GFP-MKK7 3E induction on cell proliferation. The addition of ponasterone reduced the increase in cell number (Fig. 8A) as well as [ 3 H]thymidine incorporation into DNA (Fig. 8B), indicating that induction of GFP-MKK7 3E affects basal cell growth of these cultured cells. Importantly, in the same experiments, the parental cells expressing the ecdysone receptor heterodimer did not show any significant ponasterone-or solvent-dependent change in cell growth or [ 3 H]thymidine incorporation (Fig. 8, A and B). Determination of the percentage of dead cells by trypan blue dye exclusion showed that induction of GFP-MKK7 3E expression did not increase apoptosis, but instead it reduced cell proliferation (Fig. 8C). Collectively, the results presented in this study reveal that by means of a thoroughly controlled inducible expression system, reproducible and time-dependent cDNA microarray hybridization experiments resulted in the discovery of novel FIG. 5. GFP-MKK7 3E induces expression of a discrete set of genes. Cells of clone E10 were treated for 48 h with 2 M ponasterone or left untreated. Total RNA was extracted and used as a template to prepare cDNA. cDNA from uninduced cells was labeled with Cy3 (green) and that of ponasterone-treated cells was labeled with Cy5 (red). Labeled cDNAs from untreated cells and ponasterone-treated cells were mixed and hybridized to the NCI cDNA microarray. Red and green fluorescence of each spot were measured at 532 nm and 435 nm, respectively, using an Axon4000 array reader and normalized to the average intensity of the entire array. Ratios of normalized red fluorescence intensity divided by the normalized intensity of green fluorescence were calculated for each spot of cDNA on the array using GenePix Pro software. Recorded data were analyzed further using the Web-based NCI microarray data base system (mAdb) expression query tool software, version 8.0. Genes expressed with a ratio м 2.0 or Ϲ 0.5 between uninduced and induced samples in both experiments are illustrated. Shown are the original spot images and the ratios of red to green fluorescence intensities from two experiments (Experiments 1 and 2). The relative binding of Cy3-or Cy5-labeled cDNAs is visualized by different colors. Spots of genes that are upregulated are visualized by progressively brighter shades from yellow to red, whereas genes that are down-regulated are visualized by progressively brighter shades from yellow to green. Gene names, accession numbers, gene description, and known functions for the encoded proteins as well as known conditions regulating their expression are indicated (see Footnote 2). (ϩ) indicates up-regulated, and (Ϫ) indicates down-regulated genes.
GFP-MKK7 3E -dependent changes in gene expression, which are likely to play a role in negatively regulating cell proliferation.
Because most extracellular stimuli that activate MKK7 also activate MKK4 in parallel, it is difficult to assess the function of MKK7 independently from MKK4. One way to achieve this is to express constitutively active MKK7 directly, bypassing the upstream signals that might activate other MKK molecules. This has been done successfully for MKK1 and MKK6, activa-  (Experiments 1, 2, 3, and 4) were performed as described in the legend of Fig. 5. Data were analyzed as described in Fig. 5, except that the threshold was set as the -fold change м or Ϲ 1.5 in at least three out of the four experiments. Shown is the -fold change in expression, comparing uninduced with induced cells for every single experiment as well as the mean -fold change Ϯ S.E. from the individual experiments. Note: two different cDNAs encoding c-myc were included in this array. The greatest changes in expression are found at the top and bottom for up-regulated and down-regulated genes, respectively. na indicates not analyzable in this particular experiment. Statistics were calculated using Student's t test. * indicates p Ͻ 0.05. tors of the ERK and p38 MAPK pathways, respectively (30,31). We and others have demonstrated recently that substitution of two or three phosphorylation sites by negatively charged amino acids results in a partially active form of MKK7 in transient transfection assays (22,32). We therefore generated stably transfected cell lines expressing an inducible constitutively active mutant of MKK7, MKK7 3E , in which the regulatory amino acids Ser-271, Thr-275, and Ser-277 were replaced by glutamic acid. We identified three overexpressing clones in which MKK7 3E activated endogenous JNK, but as expected, neither ERK nor p38 MAPK (Figs. 1-3). MKK7 3E expression was controlled by an ecdysone-responsive promoter. In this system, a constitutively expressed ecydsone receptor is activated by the addition of synthetic Drosophila steroids, such as ponasterone A or muristerone (33). We found that expression was maximal after 24 -48 h of induction, and at those times total cellular MKK7 levels increased by 2-3-fold ( Figs. 1 and 2).
Because we expressed an active signaling molecule, we carefully examined any evidence for "leakage" of expression because of basal promoter activity in the absence of inducer. However, we did not detect any significant basal MKK7 3E expression or activity (Figs. 1, 2, and 4). The active MKK7 3E was NH 2 -terminally fused to GFP, enabling monitoring of expression at the single cell level. We were able to establish a clone, E10, in which more than 90% of the cells expressed the transgene after ponasterone treatment. This clone was well suited for an analysis of the effects of active MKK7 on gene expression. Any changes observed upon ponasterone induction should depend on MKK7 3E .
We employed this biological system to analyze gene expression using hybridization to high density cDNA microarrays. Recent work emphasized that one could reliably detect small changes in altered gene expression by these methods if experiments were repeated carefully (34). Accordingly, we performed a total of eight independent experiments that compared RNA preparations from uninduced cells with RNA from parallel cultures that had been induced for two different times. RNA levels from the inducible, exogenous GFP-MKK7 3E transgene served as an internal control to document successful induction. Analysis of the relative hybridization to more than 6,000 cDNAs revealed that a small number of genes were up-or down-regulated after MKK7 3E induction. Expression of nine of them was consistently altered after both 24 and 48 h of ponasterone treatment.
Many of these genes are known to vary their expression after stressful or mitogenic stimuli. Of particular interest, some of these genes are critically involved in cell proliferation and apoptosis (see Fig. 5). 2 Our objective was not to test in more detail all of the genes identified by this array study. Rather, we focused on c-myc, which is well known for its association with cell proliferation. Its down-regulation after MKK7 activation was confirmed by Northern blot analysis and correlated with the increase of the of GFP-MKK7 3E (Fig. 7). This prompted us to analyze whether MKK7 3E exerted any effect on cell growth. As shown in Fig. 8, the induction of MKK7 3E inhibited cell proliferation, indicating that MKK7 3E -dependent signaling pathways play a negative regulatory role in this process. c-myc is known to play a pivotal role in the transcriptional regulators of cell growth, and this alone could account for all the effects reported in Fig. 8. However, it may be premature to ascribe the effects on cell growth solely to this gene. It is certainly possible that the observed effect on cell proliferation of MKK7 3E is the net result of the altered expression pattern of many of the genes identified in Fig. 5 and Table I. In addition, genes not represented on the cDNA array may also contribute to the effect of MKK7 on cell growth. Although purely speculative at present, it is theoretically possible that some genes are regulated by MKK7 independently from JNK. Furthermore, this type of gene expression pattern may be specific for the ␤1 FIG. 7. Induction of GFP-MKK7 3E down-regulates c-myc mRNA. Cells of clone E10 were treated for the indicated times with 2 M ponasterone (P) or left untreated. Thereafter expression of c-myc mRNA was examined by Northern blot analysis of 15 g of total RNA. In panel A, these RNAs were also used in the array hybridizations shown in Table I and Fig. 5. In panel B, c-myc mRNA expression was followed over time. Equal loading of samples was confirmed by hybridization of the same blot to the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) "housekeeping gene" probe.

TABLE II
GFP-MMK7 3E -induced changes in gene expression 12 h after induction cDNA array experiments from untreated cells or cells treated with ponasterone for 12 h (Experiments 1 and 2) were performed as described in the legend of Fig. 5. Data were analyzed as described in Fig. 5, except that the threshold was set as the average -fold change м or Ϲ 2.0. Shown is the -fold change in expression, comparing uninduced with induced cells for every single experiment as well as the mean -fold change Ϯ S.E. from the individual experiments. The greatest changes in expression are found at the top and bottom for up-regulated and down-regulated genes, respectively. One of the two c-myc spots (417226) did not fulfill selection criteria but is included for comparison with the values shown for c-myc expression in Table I  isoform of MKK7 used in this study. Nonetheless, the altered gene expression pattern observed in this study is likely to depend largely on MKK7-mediated JNK activation because we measured activation of this pathway after inducible expression of MKK7. This is the first report that MKK7 has a negative role in cell proliferation. It is likely that MKK7 achieves its growth inhi-bition through JNK, but it is difficult to compare this interpretation with published reports of the functions of JNK because in most systems the JNK pathway is activated in concert with other stress signaling pathways, such as the protein kinase cascades that lead to activation of nuclear factor-B or p38 MAPK (1,2,9,22,23). However, there are a few instances where proliferation has been shown to be JNK-dependent. For example, JNK2 is required for epidermal growth factor-stimulated cell growth, or for cell density-dependent growth arrest (35)(36)(37).
Another important consideration for relating the findings obtained in this study to known biological situations arises from the fact that the diverse stimuli that activate MAPK pathways vary greatly in their kinetics of MAPK activation. It is likely that the onset, the strength, and the duration of enzyme activation all contribute significantly to the diverse effects of MAPK stimulation. For example, the inflammatory cytokines interleukin-1 and tumor necrosis factor cause a rapid but very transient JNK activation, which is required for their profound effects on gene expression (1,23,38). Interleukin-1 and tumor necrosis factor strongly activate MKK7 and JNK, but the induction of constitutively active MKK7 3E did not change the expression of any of the known interleukin-1 and tumor necrosis factor-induced genes in our cDNA array experiments, such as those encoding cytokines or other proinflammatory proteins. This is likely because of the impossibilty of producing a rapid and transient kinetics of JNK activation by the ecdysone-inducible expression of GFP-MKK7 3E . This situation also applies to other transcriptional or post-translational conditional activation systems, as was demonstrated for tetracycline-inducible MEK1 (39) or tamoxifen-inducible MEK1 mutants (40). As in our experiments, these studies achieved only a delayed and prolonged activation of MAPKs.
On the other hand, the kinetics of activation achieved in experiments such as ours makes the study of effects of isolated MAPKs on gene expression a very appropriate model for another situation, the JNK activation caused by UV or ␥-radiation or by cytostatic drugs such as cisplatin or microtubuleinterfering agents (41)(42)(43)(44)(45)(46). The slow and sustained activation of the JNK pathway which we achieved with activated MKK7 resembles the kinetics of JNK activation followed by inhibition of cell growth, which results from administration of low doses of these drugs (41,43,44,46). From the results reported here and the aforementioned drug studies, it is tempting to speculate that a MKK7-dependent signaling pathway is involved in the growth inhibition that is seen when cells are treated with cytostatic drugs. In this context it is of interest that MKK7 3E induction for 48 h resulted in up-regulation of ERCC1, a major DNA repair enzyme, whose gene is known to be induced by DNA-damaging agents (41). Furthermore, transient expression of a dominant negative version of MKK7 efficiently inhibited cisplatin-stimulated JNK activation, providing more direct evidence that MKK7 indeed plays a role in cisplatin-mediated signal transduction. 3 In summary, the results presented here represent the first study employing high density cDNA microarrays to investigate the overall impact on gene activation by the MKK7 protein kinase. We report a hitherto undescribed role for the MKK7 molecule in negatively regulating cell proliferation, which correlates with a restricted set of differentially expressed genes, as detected by cDNA microarrays. Additional studies are required to find out if this observation also holds true for other MKK7 isoforms, in other cell lines, in vivo in tissues and intact animals, and in nonphysiological situations such as cytostatic 3 S. Wolter and M. Kracht, unpublished results. treatment of tumors. It will also be important to conclusively identify the proteins downstream of MKK7 (and JNK) which mediate this type of complex gene response.