Aberrant regulation of the Ras/Raf/MEK1/2/ERK1/2 pathway frequently occurs in human carcinogenesis (Downward, 2003
). Our work and the work of others has shown that MKP-3, a putative tumor suppressor and regulator of ERK1/2, is often upregulated in cells and tissues that exhibit abnormal activation of the Ras/Raf/MEK1/2/ERK1/2 pathway (Croonquist et al., 2003
; Furukawa et al., 2003
; Warmka et al., 2004
; Sweet-Cordero et al., 2005
). This observation led us to investigate the action of this phosphatase within the context of oncogenic Ras expression. MKP-3 has been mainly studied in in vitro
and exogenous over-expression systems. Most studies that report on endogenous MKP-3 have focused on understanding the regulation of FGF-stimulated signaling during development in vivo
(Kawakami et al., 2003
; Tsang et al., 2004
; Gomez et al., 2005
; Li et al., 2007
). To our knowledge, the roles and regulation of endogenous MKP-3 in cells that express oncogenic Ras have not been previously explored. This is an important area of investigation because expression of oncogenic Ras results in chronic stimulation of the Ras/Raf/MEK1/2/ERK1/2 pathway; under such conditions, which frequently occur in human carcinogenesis, MKP-3 may be a key modulator.
The results presented here revealed two types of roles for endogenous MKP-3 in suppressing ERK1/2 in cells that express oncogenic Ras. First, MKP-3 is involved in a negative feedback pathway that suppresses ERK1/2 when expression of oncogenic Ras is the primary stimulus of the Ras/Raf/MEK1/2/ERK1/2 pathway (). We used two tools, the pharmacological inhibitor U0126 and siRNA targeted against MKP-3, to demonstrate that in serum-starved H-ras MCF10A and DLD-1 cells, ERK1/2 drives the upregulation of MKP-3, which in turn dephosphorylates and thus inactivates ERK1/2. This suggests that MKP-3 keeps the basal level of ERK1/2 activity in check even under the pressure of stimulation by oncogenic Ras. Knockdown of MKP-3 by siRNA also results in an increase in ERK1/2 phosphorylation when the cells are maintained in complete media (data not shown), indicating that MKP-3 inhibits ERK1/2 even in the presence of serum. Second, our studies indicate that MKP-3 can act in a negative feedback pathway that modulates the duration of ERK1/2 activity when cells that express oncogenic Ras undergo further stimulation by extracellular agents. This was revealed when we found that knockdown of MKP-3 by siRNA prolonged the EGF- and TNF-α-stimulated elevation of phospho-ERK1/2 in DLD-1 cells. In EGF-stimulated DLD-1 cells, the decrease in phospho-ERK1/2 preceded the recovery of MKP-3, which suggests that other phosphatases may also be involved in the initial dephosphorylation of ERK1/2 (see ). We were not able to detect an effect of MKP-3 knockdown on TPA- or EGF-stimulated ERK1/2 phosphorylation in H-ras MCF10A cells under the conditions of our studies. TPA stimulated prolonged, highly elevated MEK1/2 activity in H-ras MCF10A cells (see ), which could counteract and mask the action of MKP-3 and thus explain the prolonged, elevated activation of ERK1/2. Further research is required to determine why the regulation of EGF-stimulated ERK1/2 differs so dramatically between H-ras MCF10A cells and DLD-1 cells, however. The two cell lines could certainly differ with respect to the balance between and access to ERK1/2 regulators because of differences in the following: 1) mechanism of oncogenic Ras expression (engineered overexpression in H-ras MCF10A cells versus endogenous expression in DLD-1); 2) expression of the oncogenic Ras family member (H-Ras versus Ki-Ras); and 3) the cell type (breast versus colon). Altogether, our studies indicate that the persistent elevation of MKP-3 can help restrict the ERK1/2 response of cells that express oncogenic Ras. The differences we observed between H-ras MCF10A and DLD-1 indicate, however, that the effectiveness of MKP-3 in suppressing ERK1/2 is likely to vary depending on many conditions.
Fig. 7 Reciprocal regulation of ERK1/2 and MKP-3. Phosphorylation and activation of ERK1/2 typically occurs through stimulation of the Ras/Raf/MEK1/2 protein kinase cascade. The GTPase Ras activates the protein kinase Raf, which phosphorylates and activates (more ...)
The various roles of MKP-3 in regulating ERK1/2 are countered by the multiple mechanisms by which ERK1/2, in turn, regulates MKP-3 (). Stimulation of both cell lines revealed a striking ERK1/2-dependent biphasic modulation of MKP-3. The initial downregulation of MKP-3 appears to involve ERK1/2-dependent protein destabilization. Our results are consistent with those of Marchetti et al., who used hamster fibroblast cells that overexpress exogenous MKP-3 to show that ERK1/2 phosphorylates MKP-3 and targets it for proteasomal degradation (Marchetti et al., 2005
). The function of the initial ERK1/2-dependent downmodulation of MKP-3 is not yet clear. It has been suggested that MKP-3, which is localized to the cytoplasm, may regulate the subcellular localization of ERK1/2 (Karlsson et al., 2004
). Further studies are required to determine if the initial loss of MKP-3 in stimulated H-ras
MCF10A and DLD-1 cells liberates ERK1/2 from its cytoplasmic anchor so that this kinase can translocate to other subcellular compartments. Our studies indicate that transcription is required for MKP-3 protein levels to rebound following its initial downmodulation, and furthermore that ERK1/2 plays a major role in the regulation of both basal and stimulated MKP-3 gene expression. The rapid rebound in MKP-3 levels following its initial downmodulation suggests that cells that express oncogenic Ras are poised to maintain elevated MKP-3 levels, and thus limit ERK1/2 activity.
The regulation of MKP-3 gene expression varies widely depending on the system. In H-ras
MCF10A cells the induction of MKP-3 is detected relatively rapidly (within 60 minutes). TPA and EGF induce MKP-3 gene expression in a similar manner in the parental MCF10A cells, indicating that the ability to stimulate MKP-3 gene expression in this cell line does not depend on the expression of oncogenic Ras (data not shown). In other systems, the induction of MKP-3 is only detected after prolonged (greater than 12 hours) exposure to stimuli (Woods and Johnson, 2006
). Finally, MKP-3 has been reported to be mainly constitutively expressed and not strongly inducible in other systems, such as human skin fibroblasts (Groom et al., 1996
; Dowd et al., 1998
). Such differences may be due to cell type dependent differences in the expression of transcriptional machinery and the operation of signaling networks.
The data presented here generally show greater levels of phospho-ERK2 (pp42 ERK) than phospho-ERK1 (pp44) in both resting and simulated H-ras
MCF10A and DLD-1 cells. This apparent preferential phosphorylation of ERK2 over ERK1 has been observed in several cell lines (Pelech, 2006
; Vantaggiato et al., 2006
). Interestingly, stimulation of the cells with EGF or TPA resulted in increases in both ERK1 and ERK2 phosphorylation, whereas knockdown of MKP-3 in serum-starved cells, in which oncogenic Ras is the primary stimulus, mainly resulted in an increase in ERK2 phosphorylation. The mechanism underlying such apparent preferential phosphorylation has not been established, but it is likely to be due, at least in part, to the interactions between and subcellular localization and levels of several proteins, including ERK1, ERK2, MEK1, MEK2, protein scaffolds, and protein phosphatases (Pelech, 2006
; Lefloch et al., 2008
The observation that knockout of ERK2 in vivo is lethal, whereas knockout of ERK1 is not indicates that ERK1 and ERK2 have different functions (Pages et al., 1999
; Yao et al., 2003
). This suggests that differential dephosphorylation and inactivation of ERK1 and ERK2 could affect the cellular response to stimuli that activate the Ras/Raf/MEK/1/2/ERK1/2 pathway. Further research is required to determine whether the specificity of MKP-3 differs for ERK1 versus ERK2. Although knockdown of MKP-3 in serum-starved H-ras
MCF10A and DLD-1 cells appears to primarily result in an increase in phospho-ERK2 levels, we have found that knockdown of MKP-3 in other cell lines results in a clear increase in both phospho-ERK1 and phospho-ERK2 (data not shown). Muda et al, reported that MKP-3 binds equally well to both ERK1 and ERK2, as indicated by the ability of a GST-MKP-3 fusion protein to precipitate ERK1 and ERK2 from cell lysates prepared from cells in which the different ERK isoforms were overexpressed (Muda et al., 1998
). To our knowledge a direct comparison of the specificity of MKP-3 for dephosphorylating ERK1 versus ERK2 has not been published, however. The balance between the phosphorylated and desphosphorylated states of ERK1 and ERK2 is likely to depend on a complex set of interactions and conditions, as discussed above,
It might be expected that expression of oncogenic Ras would result in superactivation of ERK1/2. Instead, our studies suggest that cells may adapt to the expression of oncogenic Ras, and the chronic stimulation of the Ras/Raf/MEK1/2/ERK1/2 pathway, by upregulating MKP-3, which then suppresses ERK1/2. This hypothesis is consistent with the concept of MKP-3 as a tumor suppressor, and might help explain, at least in part, why activation of Ras alone is not sufficient to induce tumors. That is, cells respond to the expression of oncogenic Ras by upregulating MKP-3, which serves to rein in ERK1/2 activity; disruption of MKP-3, in turn, unleashes ERK1/2 activity and thus contributes another step along the pathway of carcinogenesis. Previous results from our laboratory on the novel skin tumor promoter palytoxin support this hypothesis (Warmka et al., 2002
; Zeliadt et al., 2003
; Warmka et al., 2004). In the classic multi-stage mouse skin model of carcinogenesis, the first stage, known as initiation, typically involves activation of the oncogene Ras; subsequent repeated stimulation by tumor promoters results in tumor development (Balmain and Pragnell, 1983
; Yuspa, 1998). We previously demonstrated that palytoxin, a non phorbol ester tumor promoter in the multi-stage mouse skin model of carcinogenesis, stimulates ERK1/2 activation by triggering the downmodulation of MKP-3 in mouse keratinocytes derived from initiated mouse skin that express oncogenic Ras (Warmka et al., 2004). Accordingly, we also showed that palytoxin modulates several targets that have been implicated in carcinogenesis, including c-Fos, AP-1, and matrix metalloprotease-13, through ERK1/2-dependent pathways in these cells (Warmka et al., 2002). Further research is required to determine whether knockdown of MKP-3 mimics palytoxin action and whether MKP-3 is a target for other carcinogenic agents.
Interestingly, studies that implicate MKP-3 as a tumor suppressor in pancreatic cancer, which is characterized by a high frequency of activating Ras mutations, reported an association between elevated expression of MKP-3 at early stages of pancreatic cancer, but a loss of MKP-3 at advanced stages (Furukawa et al., 2003
). It has been estimated that oncogenic Ras is expressed in over 40% of colon tumors (Bos et al., 1987
; Downward, 2003). In breast cancer the regulation of the Ras/Raf/MEK/1/2/ERK1/2 pathway is frequently undermined by the aberrant regulation of growth factor receptors (Santen et al., 2002). Limited information has been published concerning the expression of MKP-3 in these cancers (Cui et al., 2006). Further research is therefore required to determine how the expression of MKP-3 changes during colon and breast carcinogenesis and whether MKP-3 functions as a tumor suppressor in these types of cancers.
Given that activation of Ras frequently occurs in human cancers (Downward, 2003
), there is an urgent need to understand the biochemical characteristics of cells that express activated Ras that make them particularly susceptible to agents and events that advance the process of carcinogenesis. The unstable nature of MKP-3 makes it a potentially vulnerable target in such cells. Altogether, the evidence that aberrant ERK1/2 activity plays a role in carcinogenesis together with the identification of MKP-3 as a potential tumor suppressor, underscores the importance of understanding the nature of the repartee between this centrally important protein kinase and its modulator.