Regulation of the inducible nuclear dual-specificity phosphatase DUSP5 by ERK MAPK
Introduction
Mitogen-activated protein kinase (MAPK) modules constitute a highly conserved group of signal transduction pathways that are responsible for the transduction of diverse extracellular signals to a large number of cytoplasmic and nuclear protein substrates [1], [2], [3], [4]. One of the most studied and best understood MAPK pathways is the extracellular signal-related kinase (ERK) pathway, which lies downstream of the cellular proto-oncogene Ras and is involved in a diverse range of cellular activities including cell proliferation, differentiation and survival [5], [6]. Both the ERK1 and ERK2 isoforms are activated by phosphorylation of both a threonine and tyrosine residue in the conserved T–X–Y motif found in the activation loop of the kinase [7]. Since the abnormal activation of this pathway can lead to uncontrolled proliferation and has been implicated in the genesis and progression of cancer [8], [9], ERK is also subject to negative regulation. This can be mediated, at least in part, by protein phosphatases that dephosphorylate either or both of the phosphorylated residues within the T–X–Y motif [10].
The specific dephosphorylation of MAPKs on both threonine and tyrosine residues is mediated by a subfamily of cysteine-dependent dual-specificity protein phosphatases (DUSPs) also known as MAPK phosphatases (MKPs) [11]. There are 10 bona-fide DUSPs/MKPs in mammals and these can be subdivided into three groups according to amino-acid sequence homology, substrate specificity and/or subcellular localisation [12], [13]. The largest group consists of four inducible nuclear phosphatases DUSP1/MKP-1, DUSP2/PAC-1, DUSP4/MKP-2 and DUSP5. DUSPs 1, 2 and 4 display phosphatase activity towards both ERK1 and 2 and the stress-activated kinases p38 and JNK [14]. In contrast, DUSP5 is unique within this group in that it appears to be ERK1/2 specific [15]. The amino terminal non-catalytic domain of DUSP5 contains both a conserved kinase interaction motif (KIM), which is essential for binding to ERK1/2, and a nuclear localisation signal (NLS). Furthermore, expression of DUSP5 leads to both inactivation and nuclear translocation of ERK1/2. Translocation requires both a functional KIM and NLS in DUSP5 and is mediated via the conserved common docking (CD) site in ERK1/2 [15]. This indicates that DUSP5 can also act as a nuclear anchor for inactive ERK1/2 and its expression may account for the observation that ERK1/2 is retained in the nucleus in its inactive form during prolonged exposure to growth factor stimulation [15], [16].
In addition to acting as phosphatases towards MAPKs, certain DUSPs/MKPs have also been demonstrated to be MAPK substrates. The best-characterised effects of DUSP phosphorylation are to modify the ubiquitin-dependent proteosomal degradation of the phosphatase. Such effects may be complex. For example ERK-mediated phosphorylation of the extreme C-terminus (Ser359 and Ser364) of DUSP1/MKP-1 has been shown to result in increased protein stability, thus functioning as a positive feedback loop to reinforce phosphatase activity [17]. In contrast, ERK-mediated phosphorylation of distinct sites within DUSP1/MKP-1 (Ser296 and Ser323) leads to recruitment of the ubiquitin ligase SCFSkp2 and increases the rate at which this phosphatase is degraded [18]. Erk- and mTor-mediated phosphorylation of the cytoplasmic ERK-specific MKP DUSP6/MKP-3 on Ser159 and Ser197 also results in an increased rate of proteosomal degradation, thus reducing DUSP6/MKP-3 phosphatase activity [19], [20].
Here we have studied the relationship between ERK activation and DUSP5 and find that the growth factor-dependent expression of DUSP5 mRNA and protein is ERK dependent and that DUSP5 is also a bona fide ERK substrate. Both the inactivation of ERK by DUSP5 and the ability of DUSP5 to act as an ERK substrate are mediated by the conserved kinase interaction motif (KIM) within DUSP5. While phosphorylation of three conserved sites within DUSP5 is induced by growth factor stimulation and blocked by a specific inhibitor of the ERK MAPK pathway, we can find no evidence of any effect of these modifications on the stability, activity or localisation of DUSP5. However DUSP5 protein levels are increased by co-expression of both wild-type and kinase dead ERK1/2, but are not increased by expression of a mutant (sevenmaker) which is unable to bind to DUSP5. Our results suggest that DUSP5 protein is stabilised by interaction with its physiological substrate and that this may reinforce its activity as both an ERK-specific phosphatase and nuclear anchor for this MAPK.
Section snippets
Materials
Cell culture reagents were purchased from Invitrogen. The antibodies used in this study are as follows. The polyclonal anti-DUSP5 antibody was raised in sheep using full length recombinant DUSP5 protein as antigen, and affinity purified using the recombinant protein. Anti-Myc and HA mouse monoclonal antibodies were supplied by Cancer Research UK. Anti-phospho- and total ERK1/2 antisera were obtained from Cell Signalling Technologies; and anti-HA and anti-tubulin were from Santa Cruz
DUSP5 mRNA and protein are induced by ERK signalling
Previous studies have indicated that DUSP5 is a growth factor inducible gene with reports of increased mRNA expression in response to EGF, serum, and TPA [29], [30]. Consistent with these studies, we recently showed that DUSP5 protein levels are increased in response to serum stimulation in NIH3T3 mouse fibroblasts [15] and we can also detect increased levels of DUSP5 protein in serum-stimulated HeLa and Cos-1 cells (data not shown). To explore the links between DUSP5 expression and ERK
Discussion and conclusions
Our initial finding that DUSP5 expression is strongly induced by serum stimulation and can be blocked by inhibition of the ERK signalling pathway suggests that DUSP5 functions as part of a negative feedback mechanism to control the duration and magnitude of nuclear ERK activation. This is reminiscent of the regulation of the cytoplasmic ERK-specific phosphatase DUSP6/MKP-3, which is inducible in response to activation of the ERK pathway by fibroblast growth factor (FGF) [24], [34], [35]. Our
Acknowledgements
We thank Dr. Simon Cook (Babraham Institute, Cambridge) for providing pBABE puro-Δ-Raf-1:ER*, Professor Ron Hay (College of Life Sciences, University of Dundee) for the plasmid encoding 6×hisUb and Professor Walter Kolch (Beatson Institute, Glasgow) for the plasmid encoding kinase dead (K52R) ERK2. We also thank Dr. Mark Saville (Ninewells Hospital, Dundee) for helpful advice and Dr. Philippe Lenormand (CNRS UMR 6543, Université de Nice) for useful discussions during the course of this work.
References (36)
J. Biol. Chem.
(1993)- et al.
Curr. Opin. Cell Biol.
(1997) - et al.
Gene
(2003) - et al.
Biochim. Biophys. Acta
(2007) Curr. Opin. Genet. Dev.
(1994)Curr. Opin. Cell Biol.
(2000)- et al.
J. Biol. Chem.
(2006) - et al.
J. Biol. Chem.
(2001) - et al.
Gene
(1989) - et al.
Mol. Cell Proteomics
(2006)
Biochim. Biophys. Acta
J. Biol. Chem.
J. Biol. Chem.
J. Biol. Chem.
Curr. Biol.
FEBS Lett.
Nature
Semin. Cancer Biol.
Cited by (0)
- 1
These authors made an equal contribution to the work presented here.