Nature's Regulation of PTP Activity
Aberrant tyrosine phosphorylation levels have been linked both to cancer and metabolic disease states. Hence, tight regulation and specificity of PTPs is required. Nature may achieve control of PTP activity by a number of different mechanisms, as exemplified in the following: (1) Site-specific tyrosine phosphorylation and dephosphorylation may result from inherent substrate specificity of kinases and phosphatases. (2) The interaction between PTPs and their cognate substrates may be regulated by subcellular localization which can be controlled either by targeting domains of the PTPs ('zip codes') (Mauro and Dixon 1994), alternative splicing (Shifrin and Neel 1993), or proteolytic cleavage (Frangione et al. 1993; Gil-Henn et al. 2000, 2001; Ragab et al. 2003). (3) Different isoforms of PTPs may be synthesized from the same gene due to alternative usage of isoform-specific 5' exons and promoters (Elson and Leder 1995; Tanuma et al. 1999; Amoui et al. 2003). (4) The enzymatic activity of PTPs may further be influenced by covalent modification such as phosphorylation of specific tyrosine and serine/threo-nine residues (Flint et al. 1993; Moeslein et al. 1999; Ravichandran et al. 2001), or by dimerization (Bilwes et al. 1996; Majeti et al. 1998; Jiang et al. 1999; Blanchetot and Den Hertog 2000; Jiang et al. 2000; Majeti et al. 2000; Blanchetot et al. 2002a), or by oxidation of the active site cysteine (see below). (5) As with other enzymes, the cellular levels of PTPs may be further controlled by specific transcription factors (Fukada and Tonks 2003), or by the half-life of the mRNA or protein of the PTP in question.
Although most of the above principles for regulation of PTP activity and specificity could potentially be utilized for development of drugs that modulate the activity of PTPs, the most straightforward approach—and also the focus of most laboratories—is to develop active site-directed inhibitors. It should be emphasized that covalent modification of PTPs on residues in the proximity of the active site, e.g., phosphorylation of Ser50 in PTP1B (Ravichandran et al. 2001), might potentially influence the activity of inhibitors that are developed against the non-modified enzyme. It is therefore an inherent risk in a drug optimization program to rely exclusively on recombinant, non-modified proteins expressed in prokaryotic systems.
Recombinant PTPs show relatively high catalytic activity compared to protein tyrosine kinases (Denu and Dixon 1998). Therefore, it appears that there is a need for transient, reversible inactivation of PTPs to ensure proper cellular tyrosine phosphorylation levels and thereby efficient propagation of signaling. In the present context, it is of particular relevance that nature seems to regulate PTP activity by H2O2-mediated reversible oxidation of the active site cysteine. It took about three decades from the initial biological observations to unravel the structural basis for the effects of H2O2 on PTPs. In the early 1970s, Michael Czech and his colleagues made the intriguing ob servation that thiols mimic the action of insulin on isolated fat cells (stimulating glucose uptake and metabolism and inhibiting lipolysis) by Cu2+-de-pendent production of H2O2 (Czech et al. 1974a,c). Further, it was shown that extracellular addition of H2O2 can mimic the action of insulin in fat cells (Czech et al. 1974c), and evidence was provided for involvement of sulf-hydryl oxidation in this process (Czech et al. 1974b). The physiological relevance of these findings became apparent when it was demonstrated that insulin stimulates intracellular formation of H2O2 in rat epididymal fat cells (May and de Haen 1979).
A large number of other agents have also been shown to mimic the action of insulin in isolated cells, including vanadate (Dubyak and Kleinzeller 1980), which was shown to act synergistically with H2O2 to stimulate the insulin receptor kinase activity in intact cells in an apparently direct manner (Tamura et al. 1984; Kadota et al. 1987; Koshio et al. 1988; Heffetz et al. 1990; Heffetz and Zick 1992). However, it was not realized until 1989 that the synergistic effects of H2O2 and vanadate were due to the formation of "pervana-date", which increases the phosphorylation and activity of the IRTK, not directly, but by inhibiting counteracting PTP(s) (Fantus et al. 1989). This seminal publication was the first to demonstrate that perhaps it would be possible to mimic the action of insulin by inhibiting PTP(s) that negatively regulate the insulin signaling pathway. Indeed, this publication played a significant role for initiation of the PTP inhibitor drug discovery program at Novo Nordisk. In the present context, it is intriguing that recent studies have demonstrated that insulin-stimulated formation of H2O2 leads to reversible inhibition of PTP1B and enhancement of the insulin signaling in 3T3-L1 cells (Mahadev et al. 2001a,b).
In 1998, it was demonstrated at the biochemical level that PTPs could be negatively regulated by H2O2 (Denu and Tanner 1998). PTPs react rapidly with low micromolar concentrations of H2O2, specifically oxidizing the active site cysteine and involving a sulfenic acid intermediate. Importantly, the inactivation took place in the presence of reducing agents and it was reversible, thus providing support for the view that PTP function may be regulated in vivo by reactive oxygen species. In an attempt to determine the structure of PTP1B with the catalytic cysteine oxidized to the sulfenic acid state, Barford, Tonks, and colleagues unexpectedly observed a modification of the active site cysteine, termed "sulphenyl-amide", which was indicated by a well-defined continuous electron density between the Sg atom of Cys215 and the main chain nitrogen atom of Ser216 in the X-ray structure (Salmeen et al. 2003). As a result, the P-loop undergoes a major conformational change (with Gly218 moving about 7 A outwards) and the oxidized cysteine is now exposed to the reducing environment in the cell, thus facilitating the regeneration of the active state of the enzyme. Moreover, the conformational change imposed by the sulphenyl-amide bond formation results in disruption of the hydrogen bond between Ser216 and Tyr46 in the pTyr loop, and the latter tyrosine residue adopts a solvent-exposed position. Interestingly, this repositioning of Tyr46 makes PTP1B susceptible to phosphorylation by the insulin receptor, and the authors propose that such redox-dependent phosphorylation may represent an additional level of control over the signaling specificity of PTP1B. These structural studies together with the above biochemical studies provide not only a plausible novel mechanism for PTP regulation, but also a unique possibility for designing compounds that only influence signal transduction pathways that are already activated (see Sect. 10). Of note, the structural changes of PTP1B observed by addition of H2O2 may also be induced by addition of putative inhibitors of the enzyme (van Montfort et al. 2003; unpublished observations).
Interestingly, UV irradiation causes inactivation of PTPs and converts PTPa into a substrate-trapping mutant which can coprecipitate the platelet-derived growth factor (PDGF)-b receptor, similarly to the PTPaC433S mutant (Gross et al. 1999). Further, den Hertog et al. have recently shown that H2O2 treatment of cells leads to rotational coupling and inactivation of PTPa di-mers (Blanchetot et al. 2002b; van der Wijk et al. 2003).
Using a modified 'in-gel' PTP assay to allow visualization of oxidized PTPs, Tonks and coworkers recently provided evidence for the notion that reversible oxidation may be a general mechanism for regulation of PTP activity (Meng et al. 2002). Specifically, these authors showed that ligand-in-duced, transient oxidation of SHP-2 in the PDGF-R complex coincided with autophosphorylation of the receptor, thus demonstrating that PTPs not only respond to oxidative stress in the environment, but also to reactive oxygen species generated to physiological responses, as in the case of insulin signaling.
Was this article helpful?