Although reactive oxygen species play important roles in cellular physiology as signalling molecules, their molecular targets are largely unknown. A probable group of targets for mediating many of the effects of reactive oxygen species on cell signalling is the large diverse family of cysteine-dependent phosphatases, which includes the protein tyrosine phosphatases. Our work and that of others suggest that the oxidative inactivation of protein and lipid phosphatases plays an important part in signalling, downstream of many cellular stimuli. Future studies should give us a clearer picture of the role of phosphatase inactivation in cellular behaviour and explain how specificity is achieved in redox signalling.
Regulated reversible phosphorylation of proteins and other cellular molecules plays a ubiquitous role in the control of cellular behaviour. Several different classes of enzymes have evolved that catalyse the removal of phosphate from different substrates using different catalytic mechanisms. Among the phosphatase classes, it appears that several groups of enzymes have evolved independently that utilize a nucleophilic cysteine residue in catalysis. It now seems probable that at least some of these enzymes are important targets of the ROS (reactive oxygen species) produced physiologically in cell signalling and/or pathologically during oxidative stress.
Cellular ROS, such as superoxide and H2O2, are ubiquitously produced in the cells of higher organisms. ROS not only cause irreversible damage to cellular components, but it has become clear that they also act as signals in the regulation of normal cellular processes [1,2]. However, despite intensive research and tens of thousands of publications studying oxidative stress and redox signalling, the cellular targets or receptors for ROS that change their behaviour in response to these signals remain obscure. Under very few circumstances has a particular molecular target been identified that mediates a specific outcome of redox signalling. Particular interest in redox signalling has arisen due to the findings that many growth factors and cytokines stimulate the generation of cellular ROS and that the activation of downstream signalling and outcomes such as mitogenesis can be inhibited by antioxidant intervention . These studies support a model for growth factor receptor signalling in which receptor activation induces both the activation of downstream kinases and the oxidative inactivation of inhibitory phosphatases, and both components are required for efficient signalling [3,4] (Figure 1).
Model for PTEN oxidation during stimulation
Cysteine-dependent phosphatases as mediators of redox signalling
The PTP (protein tyrosine phosphatase) family and other cysteine-dependent phosphatases have been recognized for many years as potential targets of ROS in cell signalling. The specific environment of the active-site cysteine residue in these enzymes makes this cysteine residue, present as a thiolate anion at neutral pH, a highly reactive nucleophile, but also makes it sensitive to oxidation [5,6]. It has been suggested that the family as a whole represents mediators of redox signalling through the oxidative inactivation of phosphatase activity [4–7]. However, many of these studies have used unphysiologically high doses of H2O2 that clearly inactivate many or most cellular cysteine-dependent phosphatases. Under all but the most extreme circumstances, physiological redox signalling through phosphatase inactivation seems unlikely to involve global inactivation of cellular phosphatases. It seems more probable that regulated localized production of ROS can inactivate a subpopulation of phosphatase molecules closely co-localized with the site of ROS generation. Evidence for such specificity in redox signalling has come from several studies, of which many of the most exciting have addressed the oxidation of PTPs by the ROS produced by cells in response to stimulation with growth factors and cytokines, including epidermal growth factor, platelet-derived growth factor and insulin [8–10].
PTEN and other lipid phosphatases
One reason why it has been difficult to identify specific phosphatases that are targets for ROS responsible for mediating specific downstream signalling outcomes is the apparent functional redundancy within the large PTP family. Although it seems probable that individual PTP enzymes have tight physiological specificity for specific cellular tyrosine residues, this has been difficult to address experimentally, and identifying specific targets for each PTP with any confidence has proved very difficult. In contrast, several cysteine-dependent phosphatases have been shown to have poor protein phosphatase activity but potent activity against inositol lipids. The limited number of cellular phosphoinositides and the existence of an established methodology for addressing the abundance and localization of these molecules has meant that relatively rapid progress has allowed the identification of specific lipid substrates for several cysteine-dependent phosphatases. Thus physiological substrates have been identified with some confidence for PTEN (phosphatase and tensin homologue deleted on chromosome 10) [PtdIns(3,4,5)P3] and some members of the myotubularin [PtdIns(3)P and PtdIns(3,5)P2] phosphatase family and, with a lesser degree of confidence, for enzymes such as the Sac phosphatases [PtdIns(4)P and probably other substrates], type I and II 4-phosphatases [PtdIns(3,4)P2] and TPIP [TPTE (transmembrane phosphatase with tensin homology) and PTEN homologous inositol lipid phosphatase] [PtdIns(3,4,5)P3]. PTEN and PtdIns(3,4,5)P3 signalling have been well studied due to their common deregulation in tumour development . This and the non-redundant role of PTEN in regulating this signalling pathway make it a good target to investigate a potential role in redox signalling.
Oxidation of PTEN in vitro with H2O2 leads to the formation of a disulphide bond between the active-site cysteine (Cys-124) and another cysteine residue (Cys-71), which is close by in the three-dimensional crystal structure of PTEN . This is in contrast with the classical PTP member PTP1B (and probably other classical PTPs) that forms a sulphenyl amide linkage between the active-site cysteine and an adjacent main-chain nitrogen . PTEN is highly sensitive to oxidation and inactivation in vitro by H2O2 and, similarly in cells, evidence from several assays indicates that cellular oxidative stress induced by H2O2 treatment also inactivates most cellular PTEN [12,13]. It is probably significant that there appears to be some specificity in the re-activation through reduction of PTEN, since thioredoxin has been found to bind to PTEN in cells and to reduce oxidized PTEN very efficiently, in contrast with other cellular electron donors . The significance of cellular PTEN oxidation has been indicated by a previous study showing the accumulation of PtdIns(3,4,5)P3, and the activation of downstream signalling induced by H2O2 is found to occur only in cells expressing PTEN, but not in cells lacking the enzyme, indicating a non-redundant role for PTEN in the redox regulation of PI3K (phosphoinositide 3-kinase) signalling . Additionally, oxidation of PTEN by endogenously generated ROS has also been identified in stimulated macrophages, accompanied by an ROS-dependent activation of signalling downstream of PTEN. These results imply that PTEN oxidation can play a role in the activation of PtdIns(3,4,5)P3-dependent signalling by stimuli that cause the generation of ROS (Figure 1).
Most of these studies of cysteine-dependent phosphatase oxidation indicate that specificity must exist, such that cellular stimulation with different stimuli leads to the inactivation of only certain functionally significant phosphatases. How this specificity is generated is the subject of the present study and there are several possibilities. It seems probable that different phosphatases have different intrinsic sensitivities to oxidation. Additionally, evidence already exists supporting a role for co-localization of phosphatases with the signalling complexes inducing ROS generation in the generation of specificity [9,10,14]. Finally, as described above, there is already evidence for a very significant role in maintaining phosphatases in the inactive state (i) for specificity at the level of phosphatase reduction, (ii) for co-localization with electron donors such as thioredoxin and even (iii) for the stimulated inactivation/oxidation of specific electron donors [12,15]. This area is currently the focus of considerable research effort, and it seems probable that this will improve our understanding of the role of oxidative phosphatase inactivation in cellular signalling.
Energy: Generation and Information: A Focus Topic at BioScience2004, held at SECC Glasgow, U.K., 18–22 July 2004. Edited by J. Arthur (Rowett Research Institute, Aberdeen, U.K.), P. Newsholme (University College Dublin, Ireland), M. Murphy (MRC-Dunn Human Nutrition Unit, Cambridge, U.K.) and R. Reece (Manchester, U.K.).