The dual specificity phosphatases M3/6 and MKP-3 are highly selective for inactivation of distinct mitogen-activated protein kinases.

The mitogen-activated protein (MAP) kinase family includes extracellular signal-regulated kinase (ERK), c-Jun NH2-terminal kinase/stress-activated protein kinase (JNK/SAPK) and p38/RK/CSBP (p38) as structurally and functionally distinct enzyme classes. Here we describe two new dual specificity phosphatases of the CL100/MKP-1 family that are selective for inactivating ERK or JNK/SAPK and p38 MAP kinases when expressed in COS-7 cells. M3/6 is the first phosphatase of this family to display highly specific inactivation of JNK/SAPK and p38 MAP kinases. Although stress-induced activation of p54 SAPKβ, p46 SAPKγ (JNK1) or p38 MAP kinases is abolished upon co-transfection with increasing amounts of M3/6 plasmid, epidermal growth factor-stimulated ERK1 is remarkably insensitive even to the highest levels of M3/6 expression obtained. In contrast to M3/6, the dual specificity phosphatase MKP-3 is selective for inactivation of ERK family MAP kinases. Low level expression of MKP-3 blocks totally epidermal growth factor-stimulated ERK1, whereas stress-induced activation of p54 SAPKβ and p38 MAP kinases is inhibited only partially under identical conditions. Selective regulation by M3/6 and MKP-3 was also observed upon chronic MAP kinase activation by constitutive p21ras GTPases. Hence, although M3/6 expression effectively blocked p54 SAPKβ activation by p21rac (G12V), ERK1 activated by p21ras (G12V) was insensitive to this phosphatase. ERK1 activation by oncogenic p21ras was, however, blocked totally by co-expression of MKP-3. This is the first report demonstrating reciprocally selective inhibition of different MAP kinases by two distinct dual specificity phosphatases.


EXPERIMENTAL PROCEDURES
Materials-Dulbecco's modified Eagle's cell culture medium was obtained from Life Technologies, Inc., protein A-Sepharose 4 fast flow was from Pharmacia Biotech Inc., murine EGF was from Promega (Madison, WI), and [␥-32 P]ATP (5000 Ci/mmol) was from DuPont de Nemours International S. A. (Regensdorf, Switzerland). Anti-HA monoclonal antibodies 12CA5 or HA.11 were purchased from Boehringer Mannheim (Rotkreuz, Switzerland) or Ruwag Diagnostics (Zurich, Switzerland), respectively. Anti-Myc monoclonal antibody 9E10 was from Dr. Glaser AG (Basel), protein A/G-horseradish peroxidase conjugate from Pierce (Zurich, Switzerland), whereas horseradish peroxidase conjugates of * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. , and pXJ40-HA-p21 rac1 (G12V) from E. Manser (Glaxo-IMCB, Singapore). Plasmids expressing Myctagged M3/6 and MKP-3 were constructed by subcloning into pMT-SM as described (40,44).
Immunoprecipitation and Immune Complex Kinase Assays-Cells were homogenized and prepared for immunoprecipitation as described (40). The supernatant (800 l) was then mixed with 75 l of a preformed immunoprecipitating complex (100 l of HA epitope specific monoclonal antibody 12CA5-I or HA.11 preincubated with 900 l of 50% (v/v) protein A-Sepharose beads in 10 mM Tris-HCl, pH 7.5, for 2 h at 4°C) by rotary mixing for 2 h at 4°C. Beads were then sedimented, and immune complex kinase assays were performed as before (40).
Western Blotting-For immunodetection of heterologously expressed MAP kinases precipitated for immune complex assay, 10 l of bead suspension (see above) was diluted with 20 l of 10 ϫ Laemmli sample buffer and heated for 5 min at 95°C followed by centrifugation at 10,000 ϫ g for 5 min. Western analysis was then performed using supernatant fractions (20 l) by SDS-polyacrylamide gel electrophoresis and electrotransfer to nitrocellulose membranes as described (40). Immunoprecipitated HA-ERK1, HA-p54-SAPK␤, HA-JNK1, and HA-p38 MAP kinases were detected using HA epitope-specific monoclonal antibody 12CA5 modified by biotinylation together with avidin-horseradish peroxidase conjugate and enhanced chemiluminescence.

RESULTS AND DISCUSSION
To establish a system to allow assessment of MAP kinase regulation by M3/6 and MKP-3, HA-tagged ERK1, p54 SAPK␤, and p38 MAP kinase were expressed in COS-7 cells and activated by exposure to a number of acute stimuli. Although ERK1 is stimulated by EGF, H 2 O 2 , and menadione (vitamin K), p54 SAPK␤ and p38 are activated following cellular exposure to anisomycin, sodium arsenite, H 2 O 2 , UV light, and sorbitol. p38 MAP kinase but not p54 SAPK␤ also undergoes activation by menadione (Fig. 1). EGF (ERK1), UV (p54 SAPK␤), and H 2 0 2 (p38) were selected as stimuli to test inhibitory regulation by dual specificity phosphatases.
M3/6 has been expressed previously in COS cells and shown to be ineffective as an inhibitor of serum-stimulated ERK2 phosphorylation (44). In accordance with these observations, although increasing the concentration of Myc-M3/6 plasmid in transfections from 0.1 to 2.0 g resulted in a dose-dependent increase in immunodetectable Myc-tagged protein (not shown), EGF-dependent enzymatic activation of ERK1 was inhibited less than 30% even at the maximum levels of M3/6 obtained (Fig. 2). Interestingly, however, parallel experiments demonstrated clearly that M3/6 blocks stress-induced activation of both p54 SAPK␤ and p38 MAP kinases with maximal inhibition observed when cells were transfected using only 0.5-1.0 g of plasmid (Fig. 2). Identical inhibition by M3/6 was observed when p54 SAPK␤ and p38 MAP kinases were activated by UV, anisomycin, or H 2 O 2 . Also, p46 HA-SAPK␥ (JNK1) activated by anisomycin is inhibited identically over the range of M3/6 expression levels obtained in COS-7 cells (not shown). Together, these observations demonstrate that M3/6 displays highly selective inactivation of JNK/SAPK and p38 MAP kinases when expressed in mammalian cells.
We have reported previously that MKP-3 expressed in COS-7 cells blocks totally both endogenous and heterologously expressed ERK2 activation following stimulation with growth factors (40). In a more detailed analysis we have now co-transfected ERK1 with a range of Myc-MKP-3 plasmid levels (0.1-2.0 g), which results in a dose-dependent increase in immunodetectable protein (not shown). This approach demonstrates that MKP-3 also blocks EGF-stimulated ERK1 activation and that this inhibition is maximal when cells are transfected with 0.5-1.0 g of MKP-3 plasmid (Fig. 3). Basal ERK1 activity is also abolished when cells are transfected with 1.0 g or more of MKP-3 plasmid (Fig. 3). In contrast to observations with ERK1 and ERK2, stress-activated p54 SAPK␤ or p38 MAP kinases were suppressed only partially using 0.5-1.0 g of plasmid with near complete blockade observed only when cells were co-transfected with 2.0 g of MKP-3 plasmid (Fig. 3). These observations demonstrate that MKP-3 appears highly selective for inactivation of the ERK family of MAP kinases.
It is now clear that several members of the p21 ras superfamily of GTPases are linked to the activation of different MAP kinase family members. For instance, constitutively active p21 ras (G12V) stimulates activation of ERK (8), and this may underlie mitogenesis and cellular transformation by this oncogene (19,46). In addition, recent studies have shown that mutationally activated versions of the p21 rho GTPase family members p21 cdc42 and p21 rac elicit enzymatic activation of both JNK/SAPK and p38 but not ERK MAP kinases (10,11). Interestingly, p21 cdc42 and p21 rac also stimulate DNA synthesis and appear to play a critical role mediating mitogenesis and transformation by oncogenic p21 ras (47,48). To test whether M3/6 or MKP-3 retain their activity and selectivity for inhibiting MAP kinases undergoing oncogenic activation, COS-7 cells were triple transfected with constitutively activate p21 ras (G12V) or p21 rac1 (G12V) together with different MAP kinases and in-creasing concentrations of plasmid for MKP-3 or M3/6. This experiment shows clearly that as with acute stimulation, M3/6 inhibited effectively p54 SAPK␤ activation by constitutive p21 rac (G12V) (Fig. 4). Also, as observed with short term exposure to growth factor, MKP-3 blocked ERK1 activation by oncogenic p21 ras (G12V) (Fig. 4), whereas M3/6 was completely ineffective (not shown). Importantly, co-transfection with increasing concentration of either MKP-3 or M3/6 plasmid does not alter the level of immunodetectable p54 HA-SAPK␤ or HA-ERK1 (Fig. 4). The ability of MKP-3 to suppress chronic MAP kinase activation by p21 ras (G12V) together with its clear selectivity for ERK family members could indicate a physiological role as an inhibitor of proliferation or even a tumor suppressor.
This is the first account of two dual specificity phosphatases, M3/6 and MKP-3, displaying reciprocal selectivity for inactivating ERK or JNK/SAPK and p38 MAP kinases. The remarkable inactivity of M3/6 against ERK family members indicates a high degree of specificity between MAP kinase family members and has not been demonstrated previously for any phosphatase of this class. These observations with M3/6 and MKP-3 are distinct from experiments using the dual specificity phosphatases PAC1, MKP-2, and CL100/MKP-1, which appear moderately selective when expressed in mammalian cells (49). Our data on selective enzymatic inhibition indicate that regu- O 2 (p38) followed by immunoprecipitation and immuno complex assays using myelin basic protein (ERK1), GST-c-Jun (SAPK␤), or GST-ATF-2 (p38) as substrate. A, autoradiograph of phosphorylated substrates separated using a 15% gel. Substrate bands were excised for counting by scintillation spectrometry, and the data were used to calculate kinase activity. This is indicated numerically below each lane expressed as fold stimulation over basal activity (defined as 1.0) measured in unstimulated cells. B, Western blot of HA-ERK1, HA-SAPK␤, and HA-p38 MAP kinases used for immune complex assays shown in A. Kinases were detected as described in the legend to Fig. 1. The data are typical of more than five separate co-expression and immune complex experiments. lated expression of MKP-3 and M3/6 could be a critical parameter in both short and long term control of cell function by different MAP kinases. MKP-3 and M3/6 are unique amongst dual specificity phosphatases insofar that they are both localized in cytosolic compartments (40,44), whereas other members of this gene family are nuclear (33,35,37,38). The M3/6 gene also possess a translated complex trinucleotide repeat resulting in multiple serine and glycine residues within the COOH-terminal third of the protein (44). Whether either of these novel characteristics underly their selectivity for MAP kinase inactivation is currently under investigation in our laboratories.
FIG. 4. M3/6 and MKP-3 inhibit SAPK␤ and ERK1 activation by constitutive p21 rac (G12V) and p21 ras (G12V). COS-7 cells were triple transfected with p21 rac (RacV12) (0.5 g of plasmid) or p21 ras (RasV12) (0.25 g of ,plasmid) as well as HA-SAPK␤ or HA-ERK1 (1.0 g of plasmid) together with 0.1, 0.25, 0.5, 1.0, or 2.0 g of plasmid expressing Myc-M3/6 or Myc-MKP-3 as indicated. Following 24 h of growth and 16 h of serum starvation, cells were homogenized, and the MAP kinases were immunoprecipitated for immune complex assays using GST-ATF2 (SAPK␤) or myelin basic protein (ERK1) as substrate. A, autoradiograph of phosphorylated substrates separated using a 15% gel. Bands were excised for counting by scintillation spectrometry and calculation of relative kinase activity, which is indicated numerically below each lane. Basal MAP kinase activity in cells not expressing p21 rac (G12V) or p21 ras (G12V) is defined as 1.0. B, Western blot of immunoprecipitated HA-SAPK␤ or HA-ERK1 used for immune complex assays shown in A. The data are typical of three identical coexpression and immune complex experiments.