Protein tyrosine phosphatases in higher plants

Authors


Summary

Reversible protein phosphorylation is the most common mechanism for cellular regulation in eukaryotic systems. Indeed, approximately 5% of the Arabidopsis genome encodes protein kinases and phosphatases. Among the thousands of such enzymes, only a small fraction has been examined experimentally. Studies have demonstrated that Ser/Thr phosphorylation and dephosphorylation play a key role in the regulation of plant physiology and development. However, function of tyrosine phosphorylation, despite the overwhelming importance in animals, has not been systematically studied in higher plants. As a result, it is still controversial whether tyrosine phosphorylation is important in plant signal transduction. Recently, the first two protein tyrosine phosphatases (PTPs) from a higher plant were characterized. A diverse group of genes encoding putative PTPs have been identified from the Arabidopsis genome sequence databases. Genetic analyses of various PTPs are underway and preliminary results have provided evidence that these PTPs serve critical functions in plant responses to stress signals and in plant development.

Introduction

Signal transduction regulates cellular processes in response to external/internal stimuli and is crucial for the growth and development of all organisms. It is now recognized that the reversible phosphorylation of proteins is an essential component of almost all signalling pathways in living cells. Changes in the phosphorylation state of a protein are conducted by two types of enzyme activities: protein kinases that catalyse the covalent attachment of a phosphate group to an amino acid side chain, and protein phosphatases that reverse this process. Whereas all protein kinases are structurally related to one another (Hunter, 1995; Johnson et al., 1996), protein phosphatases are defined by at least three distinct families (Stone & Dixon, 1994; Barford, 1995; Cohen, 1997; Neel & Tonks, 1997) (Table 1). The PPP and ppm families consist of Ser/Thr phosphatases, while the protein tyrosine phosphatase (PTPase) family includes both tyrosine-specific and dual-specificity phosphatases (Denu et al., 1996; Tonks & Neel, 1996). Even within the same family, significant structural diversity can be generated by the presence of unique regulatory and targeting domains or attachment of regulatory subunits to the catalytic subunits. These regulatory domains or subunits may localize the protein complexes to a specific subcellular compartment, modulate the substrate specificity, or alter the catalytic activity.

Table 1. Protein phosphatases and their distribution in yeast, plants, and animals
Ser/Thr Phosphatases
PPP family
 PPI: yeast, plants, and animals
 PP2A: yeast, plants, animals
 PP2B: yeast, plants (?), animals
 novel: PP4, PP5, PP6, RdgC/PP7 (yeast, plants? and animals)
PPM family
 PP2C yeast, animals, and plants (e.g. ABI, KAPP, MP2C)
Tyrosine phosphatases
Tyrosine-specific
 receptor-like: animals
 intracellular: yeast, animals, and plants (e.g. AtPTP1)
Dual-specificity
 VH1
 MKPs: yeast, animals, and plants (e.g. AtDsPTP1)
 CDC25: yeast, aniamals, plants (?)
 PTEN: animal, yeast, and plants

The attachment to, or removal from, a protein of a phosphate group often has profound effects on its structure and thereby modifies the functional property of the protein. For example, phosphorylation can regulate an enzyme activity by initiation of allosteric conformational changes, which may directly block access to the active site (Johnson et al., 1993; Johnson & O’Reilly, 1996). Another important function for phosphorylation is to regulate interactions among protein partners that must form complexes in order to function (Pawson, 1995). Nearly all aspects of cell function involve reversible phosphorylation. Examples include metabolism, cell cycle progression, ion transport, developmental control, and stress responses. This diverse spectrum of cellular functions is reflected in the large number of intracellular proteins that are subject to reversible phosphorylation. It may not be surprising that eukaryotic genomes encode approximately 2000 protein kinases and 1000 phosphatases, respectively, corresponding to 3% of these genomes (Hunter, 1995). Currently it is estimated that nearly 1000 genes encode proteins belonging to the eukaryotic protein kinase superfamily in Arabidopsis thaliana. An additional 300 genes encode protein phosphatases.

While the large number of protein kinases and their functions have been recognized for many years, the number, diversity, and function of protein phosphatases have only been appreciated more recently. Even less is known regarding protein phosphatases in higher plants. The large number and dynamic nature of protein kinases and phosphatases make it difficult to understand the function of each enzyme by conventional biological approaches. Recent developments in genomics and reverse genetics have made it possible to place all kinases and phosphatases in the functional network of a plant cell. Several recent papers (Smith & Walker, 1996; Luan, 2000) review the basic concepts and progress in the field of plant phosphatases. Here, we briefly describe what is known about protein tyrosine phosphatases in animals and plants.

Sequence, structure, catalytic mechanism, and regulation of PTPases

PTPases in animal and yeast systems

As research on plant PTPases is still in its infancy, the basic concepts and the paradigms have been established by studies on PTPases in animal and yeast cells, which produce hundreds of enzymes that are capable of dephosphorylating tyrosine residues in a protein substrate. Based on phosphoamino acid specificity, PTPases can be further divided into two large groups: tyrosine-specific PTPases and dual-specificity PTPases. Tyrosine-specific PTPases dephosphorylate phosphotyrosine – but not phosphoserine/threonine residues; the dual-specificity PTPases are able to dephosphorylate both phosphotyrosine and phosphoserine/threonine residues (Stone & Dixon, 1994). Table 1 lists the major groups of PTPases and members representative of each group. Although the overall protein sequence of tyrosine-specific PTPases and dual-specificity PTPases share little homology, all PTPases contain a signature motif in the catalytic core: (V/I)HCXAGXGR(S/T)G that harbors an essential cysteinyl residue required for the formation of a phosphoenzyme intermediate. In addition, the secondary and tertiary structure of all PTPases bears high similarity in the catalytic region. The substrate specificity is determined by sequences outside the catalytic domain (Denu et al., 1996; Tonks & Neel, 1996).

Tyrosine-specific PTPases are a diverse group of proteins. Based on their subcellular localization, these PTPases are classified into two subgroups (Fig. 1): receptor-like PTPases and intracellular PTPases (Stone & Dixon, 1994; Neel & Tonks, 1997). The receptor-like PTPases all contain an extracellular domain of variable length and composition, a single transmembrane region, and one or two cytoplasmic PTPase catalytic domains. The intracellular PTPases typically contain a single catalytic domain and various N- and C-terminal extensions that are involved in localization or other regulatory functions (Mauro & Dixon, 1994; Van Vactor et al., 1998). Of particular interest are several PTPases with Src-homology 2 (SH2) domains attached to the catalytic region. These PTPases have been shown to interact with activated receptor tyrosine kinases and other targets (Feng et al., 1993; O’Reilly & Neel, 1998; Timms et al., 1998).

Figure 1.

Structural domains of some representative protein tyrosine phosphatases (PTPases). Receptor-like PTPases often have two catalytic domains, whereas intracellular PTPases have only one. The dual-specificity PTPases are overall different in sequence but always contain the catalytic domain signature motif (as indicated by a open window in the centre of the molecule). Plant intracellular (AtPTP1) and dual-specificity PTPase (AtDsPTP1) are listed.

The dual-specificity PTPases are also a large family of intracellular proteins having a similar catalytic region but distinct noncatalytic regions. The prototype of this group is VH1, which is encoded by vaccinia virus open-reading frame H1 (Guan et al., 1991). The largest group of mammalian DsPTPs is those that dephosphorylate and inactivate mitogen-activated protein kinases (MAPKs). At least six of these MAPK phosphatases (MKPs) have been characterized. Each has distinct substrate specificity and is involved in the regulation of different MAPK pathway (Keyse, 1998). Another distinct class of DsPTPase is represented by CDC25, which is required for activation of CDC2 protein kinase and cell division (Dunphy, 1994). The primary sequence of CDC25 is only distantly related to other DsPTPases.

Crystal structures of both tyrosine-specific and dual-specificity PTPases have been solved and provide insights into the structural basis for catalysis and substrate specificity (Barford et al., 1994; Stuckey et al., 1994; Yuvaniyama et al., 1996; Hoffmann et al., 1997). Despite the sequence divergence between these two subfamilies of PTPases, all crystal structures solved so far reveal essentially the same core structural features: the central four-stranded parallel beta-sheet surrounded on both sides by one and four alpha-helices. The dual-specificity PTPases appear to be truncated versions of tyrosine-specific PTPases. The signature motif residues are found within a single loop, nestled at the base of a cleft on the surface of the protein. The essential cysteine is in the position for nucleophilic attack on an incoming phosphotyrosyl residue. The remaining residues of the core motif function to increase the nucleophilicity of the catalytic cysteine and to bind to and position the phosphate group. The arginyl residue in the signature motif is particularly important for this function. The depth of the catalytic cleft determines the substrate specificity and is set by an invariant tyrosyl residue in the tyrosine-specific PTPases. Therefore, only phosphotyrosine (not phosphoserine/threonine) is long enough to access the catalytic cysteine (Denu et al., 1996; Tonks & Neel, 1996).

Plant PTPases

So far, a typical tyrosine kinase has not been identified from yeast or plants. Nevertheless, at least four PTPases have been characterized from budding yeast, three tyrosine-specific (PTP1, PTP2, and PTP3) and one dual-specificity PTPase (Msg5) (Guan et al., 1992a; Guan et al., 1992b; Zhan et al., 1997). In higher plants, a number of studies have been directed towards isolating a bona fide PTPase. However, this has not been successful until recently.

Following the discovery of tyrosine phosphorylation in animal systems, a number of plant biologists searched for evidence of protein tyrosine phosphorylation in plants. In some cases, phosphotyrosine-containing proteins have been found (Elliot & Geytenbeek, 1985; Torruella et al., 1986). Efforts have also been made to purify PTPases from plant sources (Cheng & Tao, 1989; Guo & Roux, 1995), but none of these studies have led to molecular characterization of plant PTPases. Instead, some of these studies resulted in purification of other phosphatases unrelated to PTPases in sequence and structure (Guo et al., 1998). After many other unsuccessful efforts, it was generally thought that plants may not have PTPases similar to those in animal systems.

During our study of osmosensing in guard cells, we identified a MAPK activity that was regulated by hyperosmotic stress (Fig. 2a). Upon stress treatment, at least two MAPK activities were activated rapidly but were subsequently inactivated shortly after activation. The rapid inactivation was caused by dephosphorylation of MAPK by protein phosphatases. When phosphatase inhibitors including okadaic acid (OA) and vanadate were applied to the cells, we observed a hyper-activation of MAPK activities in guard cells (Fig. 2b). However, such a hyper-activation was not observed if OA or vanadate alone was applied to the cells. Because OA is a specific inhibitor of PP1 and PP2A-type phosphatases, and vanadate specifically inhibits PTPases, we hypothesized that both PP1/2 A-type phosphatases and PTPases are involved in the inactivation of MAPK in guard cells. This finding provided important evidence that plants may indeed contain MAPK phosphatases that are sensitive to vanadate and should belong to the typical PTPase family. Meanwhile, a typical DsPTPase was isolated from a green alga, Chlamydomonas (Haring et al., 1995), further supporting the possibility that higher plants may also have typical PTPases.

Figure 2.

Regulation of mitogen activated protein kinase (MAPK) activity by osmotic stress in guard cell. (a) Transient activation of MAPK isoforms by osmotic stress. (b) Hyperactivation of MAPK isoforms by phosphatase inhibitors. Total proteins were extracted from guard cells incubated in 600 mM mannitol for the time indicated. In-gel kinase (Zhang & Klessig, 1997) assay was performed to monitor the kinase activity in different samples. 

Sequence analyses generated several conserved regions in PTPases. Using a systematic PCR approach, we succeeded in isolating a cDNA from Arabidopsis referred to as AtPTP1 (Xu et al., 1998). Sequence alignment revealed that AtPTP1 contains a typical signature motif present in all other PTPases. Phosphatase assays unambiguously demonstrated that AtPTP1 encodes a bona fide PTPase. For example, AtPTP1 recombinant protein specifically dephosphorylates phosphotyrosine in protein substrates labelled by pp60 Src, a tyrosine-specific kinase, but does not dephosphorylate the same substrates labelled by PKA, a Ser/Thr-specific kinase. The PTPase activity was severely inhibited by submillimolar vanadate, a typical feature for all PTPases studied so far. Moreover, replacing the cysteine residue in the signature motif with a serine, only one atom change to the protein completely abolishes AtPTP1 activity (Xu et al., 1998). Following identification of AtPTP1, a tyrosine-specific PTPase, we identified a DsPTPase from Arabidopsis (AtDsPTP1, Gupta et al., 1998). Another study shows that AtPTP1 homologues are found in other plants (Fordham-Skelton et al., 1999). Some results of phosphatase assays on AtPTP1 and AtDsPTP1 are shown in Fig. 3.

Figure 3.

Activity assays of AtDsPTP1 and AtPTP1 (Gupta et al., 1998; Xu et al., 1998). (a) Phosphotyrosine-specific activity of AtPTP1 protein. Two substrates were labelled by tyrosine kinase and used for dephosphorylation experiments by AtPTP1. One is myelin basic protein (open circles), the other is casein (solid circles). The same substrates labelled by serine/threonine kinase were not dephosphorylated (open triangles and open squares). (b) AtDsPTP1 dephosphorylates both phosphotyrosine and phosphoserine/threonine in protein substrates. Myelin basic protein labelled by tyrosine kinase (open circles) or by serine/ threonine kinase (solid circles) were both dephosphorylated by AtDsPTP1, although the tyrosine phosphatase activity is stronger. Vanadate completely inhibited the activity of AtDsPTP1 (solid squares). 

An updated sequence search revealed at least a dozen more Arabidopsis genes that encode proteins with a catalytic core motif of PTPases (S.L. and R.G., unpublished results). Some are highly similar to AtDsPTP1 but others are divergent. Further experiments are needed to determine the enzymatic properties of their protein products. Identification of these PTPases is a step towards further understanding of the functional significance of tyrosine phosphorylation in higher plants.

Regulation of PTPases

A critical feature of all signal transduction molecules is that they must have the structural basis for sensitive regulation by extracellular and intracellular signals. Protein kinases and phosphatases are pivotal regulators of almost all signalling pathways in a eukaryotic cell. We discuss the various mechanisms by which protein tyrosine phosphatases are regulated in the cell. These include regulation at the level of expression, localization, substrate specificity, and activity of these enzymes.

Many PTPases are expressed in a tissue- or developmental stage-specific manner and their expression levels are often modulated by environmental conditions. These expression patterns are closely related to their specific functions. For example, CD45 has been shown to be expressed preferentially in the lymphocytic tissues in mammalian systems and is important in signal transduction in T and B cells (Charbonneau et al., 1988; Ledbetter et al., 1988). Several receptor-like PTPases are selectively expressed in the neuronal tissues in Drosophila and are involved in neural cell adhesion (Desai et al., 1994). A cytoplasmic SH2-containing PTPase, SHP-1, is expressed at a high level in the haematopoietic cells and plays a critical role in lymphoid cell differentiation (Neel & Tonks, 1997). Dual specificity PTPases play a role in regulation of MAPK pathways and their expression is often induced by mitogenic and stress signals (Keyse & Emslie, 1992; Sun & Tonks, 1994). This is also true in yeast where tyrosine-specific PTPases are involved in MAPK regulation and are responsive to stress signals (Wurgler-Murphy et al., 1997). In addition, Msg5, a dual-specificity PTPase involved in pheromone signal transduction, is also regulated by pheromone application (Doi et al., 1994). AtPTP1, the only plant tyrosine-specific PTPase identified so far is regulated by stress factors (Xu et al., 1998). The Arabidopsis dual-specificity PTPase (AtDsPTp1) is not regulated by stress signals (Gupta et al., 1998). Further studies are required to examine the detailed expression pattern of both AtPTP1 and AtDsPTP1 in Arabidopsis at various developmental stages and under other stress conditions.

PTPases are also regulated by phosphorylation. For example, several PTPases have been shown to interact with tyrosine kinases and are phosphorylated by these kinases, although the functional significance of these phosphorylation events are not understood (Lorenz et al., 1994; Shifrin et al., 1997; Van Vactor et al., 1998). Other modifications are exemplified by redox control of phosphatase activity. All PTPases contain a catalytic cysteine that must be in a reduced form for phosphatase activity. Oxidation of cysteine under oxidative stress reversibly inactivates the PTPase (Denu & Tanner, 1998). To date, little is known about the regulation of plant protein phosphatases by covalent or noncovalent modification of the protein.

Although the catalytic region of the same family of PTPases is highly homologous, these members always contain distinct noncatalytic regions that represent functional domains involved in the regulation of various aspects of a specific phosphatase. Perhaps the most common function is to direct the localization of the catalytic activity in a cell. The receptor-like PTPases all contain a transmembrane domain for plasma membrane targeting (Stone & Dixon, 1994; Neel & Tonks, 1997). Some dual-specificity PTPases such as MKP1, MKP2 are strictly localized to the nucleus (Rohan et al., 1993; Guan & Butch, 1995). Some cytoplasmic PTPases contain an SH2 domain for interacting with receptor tyrosine kinases in the plasma membrane (Feng et al., 1993). The noncatalytic regions also have other functions including the regulation of the catalytic activity. For example, receptor-like PTPases are regulated by ligand-binding to the extracellular domains. A recent hypothesis suggests that ligand-binding to the receptor-like PTPases may inactivate the PTPase activity (Neel & Tonks, 1997).

Function of PTPases in animal, yeast, and plants

We will introduce the well-studied functions of PTPases in other systems such as animal and yeast, and then present findings on the function of plant enzymes. As you will see, plant PTPases have unique functions in plant growth and development although their sequences in the catalytic domains are very similar to those in animals.

Animal and yeast enzymes and their functions

A diverse array of functions has been identified for PTPases in animal cells (Stone & Dixon, 1994; Neel & Tonks, 1997). The receptor-like PTPases play crucial roles in cell adhesion, consistent with the fact that their extracellular domains all contain regions with high similarity to cell adhesion molecules. Several studies using biochemical and genetic analyses all support this notion. For example, expression of receptor-like PTPases in insect cells makes the cells aggregate. Deletion analysis indicates that the extracellular domains mediate the cell–cell interaction (Gebbink et al., 1993; Sap et al., 1994). Mutant embryos lacking expression of some of these PTPases were generated (Desai et al., 1996; Krueger et al., 1996). Some mutants show defects in their growth cones that fail to recognize muscle targets, or follow pathways that bypass these targets altogether. Manipulation of some adhesion molecules had a similar phenotype, suggesting the overlapping function of receptor-like PTPases and cell adhesion molecules.

The SH2-containing PTPases (SHPs) have been identified in all animal systems studied including mammals, frogs, and flies. Two SHPs are present in vertebrates, SHP-1 is expressed in highest levels in haematopoietic cells and SHP-2 is ubiquitously expressed. The presence of an SH2 domain clearly indicates the possibility for these phosphatases to interact with receptor-like tyrosine kinases. Both biochemical and genetic approaches have validated this hypothesis. SHP1 has been shown to negatively regulate multiple haematopoietic signalling pathways, including those that are downstream of cytokine and immune recognition receptors (Neel & Tonks, 1997). For example, activation of signalling through the B cell antigen receptor is abrogated upon cross-linking of the inhibitory Fc receptor that interacts with SHP-1. Interaction analysis demonstrated that SHP-1 also interacts with and down regulates the receptor tyrosine kinase c-kit in vivo. An SHP-like PTPase from Drosophila, named Corkscrew, is required for proper embryonic head-tail development, which is regulated by a receptor tyrosine kinase (Perkins et al., 1992; Perkins et al., 1996). Corkscrew may also regulate the pathway mediated by daughter of sevenless gene product, another receptor tyrosine kinase (Allard et al., 1996).

A family of dual-specificity PTPases have been shown to play a critical role in the regulation of MAPK pathways. MAPK family members are essential components in many signalling pathways including responses to growth factors and stress. For activation, MAPK requires phosphorylation, by a dual-specificity kinase, of both a threonine and a tyrosine residue in the kinase activation domain (Ahn et al., 1992; Guan, 1994). A significant body of literature implicates dual-specificity PTPases in the dephosphorylation of MAPKs (Guan, 1994; Sun & Tonks, 1994; Keyse, 1998). In yeast, both tyrosine-specific and dual-specificity PTPases play a role in regulating MAPKs (Wurgler-Murphy et al., 1997; Wurgler-Murphy & Saito, 1997; Zhan et al., 1997). In mammalian cells, a variety of dual-specificity PTPases exhibit activity towards activated forms of MAPKs both in vivo and in vitro. In each case, dephosphorylation of the MAPK isoform by a dual-specificity PTPase results in loss of MAPK activity. It was initially puzzling that multiple isoforms of dual-specificity PTPases can be found in a single cell type, as can multiple forms of MAPKs. Studies show that dual-specificity PTPases have rather strict substrate specificity towards different isoforms of MAPKs (Keyse, 1998). Therefore various combinations of MAPK and dual-specificity PTPase isoforms may constitute modules in distinct signalling pathways in a single cell type (Keyse, 1998). In budding and fission yeast, tyrosine-specific PTPases are also important for MAPK regulation. For example, the osmosensing pathway in budding yeast is initiated by a two-component histidine kinase Sln1 that activates a kinase cascade involving an MAPK member, HOG1. Biochemical and genetic analyses indicate that HOG1 is inactivated by PTP2 (Wurgler-Murphy & Saito, 1997; Wurgler-Murphy et al., 1997). A similar MAPK-mediated pathway has been shown to link the environmental changes to the onset of mitosis in fission yeast. In this pathway, Spc/Sty1 MAPK is inactivated by typical PTPases Pyp1 and Pyp2 (Shiozaki & Russell, 1995).

An important subclass of dual-specificity PTPase is CDC25 protein phosphatase that is required in cell cycle progression. As its name indicates, CDC25 was initially identified as a cdc gene in which a loss of function mutation halts cell cycle in yeast. Homologues of CDC25 have since been found in all animal systems examined. Biochemical and genetic analyses have defined a role for CDC25 in activation of CDC2 protein kinase (Dunphy, 1994). Inactivation of CDC2 involves phosphorylation of a specific tyrosine residue in the protein. The phosphatase activity of CDC25 is responsible for dephosphorylating the inhibitory tyrosine residue and activating CDC2 kinase.

Most recently, a subgroup of dual-specificity PTPases has been identified for its function as a tumour suppressor. This subclass, represented by PTEN/MMC1, is a unique DsPTP that contains the PTP catalytic core motif and shares significant similarity with the cytoskeleton-interacting protein tensin (Li et al., 1997; Steck et al., 1997). PTEN not only functions as a DsPTP but also hydrolyses the 3′ position of phosphatidylinositol and inositol phosphates (Myers et al., 1997; Maehama & Dixon, 1998). In particular, phosphatidylinositol 3,4,5-triphosphate (PIP3) has been shown to be a physiological substrate of PTEN. PIP3 is the product of PI3 kinase and is essential for cell growth and cell cycle progression via activation of protein kinase B (Maehama & Dixon, 1999). PTEN converts PIP3 into PIP2 thereby countering the action of PI3K and inhibiting cell growth (Stambolic et al., 1998; Wu et al., 1998; Maehama & Dixon, 1999; Sun et al., 1999). Loss-of-function mutations in PTEN cause cancer in many cell types (Maehama & Dixon, 1999). Some studies have shown that PTEN also dephosphorylates protein substrates including focal adhesion kinase (FAK) and regulates MAPK pathways (Gu et al., 1998, 1999; Tamura et al., 1998). PTEN homologues are present and play critical roles in development of other animals such as Drosophila and C. elegans (Ogg & Ruvkun, 1998; Huang et al., 1999).

Plant enzymes and their functions

AtPTP1 and AtDsPTP1 in stress response and MAPK regulation

In plants, a typical protein tyrosine kinase has not been identified. The importance of tyrosine phosphorylation in plant cells had been controversial until the finding of MAPK pathways in numerous signal transduction pathways in higher plants (Mizoguchi et al., 1997; Meskiene & Hirt, 2000). As discussed earlier, MAPK is regulated by tyrosine phosphorylation and dephosphorylation in both yeast and animals. In higher plants, a number of MAPKs have been recently identified by molecular cloning procedures. Concerning the function of plant MAPKs, studies in the past several years have revealed a large array of signalling pathways that involve activation of MAPKs (Mizoguchi et al., 1997; Luan, 2000; Meskiene & Hirt, 2000). For example, MAPKs are activated by a variety of stress signals including cold, drought, wounding, and oxidative stress (Seo et al., 1995; Usami et al., 1995; Jonak et al., 1996; Stratmann & Ryan, 1997; Zhang & Klessig, 1998b; Cazale et al., 1999; Kovtun et al., 2000). In both tobacco and parsley cells, an MAPK is activated in response to pathogen elicitors (Suzuki & Shinshi, 1995; Ligterink et al., 1997; Zhang & Klessig, 1998; Zhang & Klessig, 1998a). Salicylic acid, a signalling molecule in the pathogen response, also activates an MAPK in tobacco cells (Zhang & Klessig, 1997). Besides stress responses, MAPKs also participate in hormonal responses and developmental processes (Chang & Meyerowitz, 1995; Wilson et al., 1997; Kovtun et al., 1998). Taken together, these findings demonstrate that MAPK activation is a common and an early step in the response to stress, hormones, and developmental signals, although diverse isoforms of MAPK may be involved in pathways initiated by different signals.

Several studies in higher plants have shown that MAPK activation is accompanied by tyrosine phosphorylation of the enzyme protein (Suzuki & Shinshi, 1995; Stratmann & Ryan, 1997; Zhang & Klessig, 1997; Romeis et al., 1999, Hunag et al., 2000; Nuhse et al., 2000). These studies provide strong evidence that plant MAPKs may be phosphorylated at tyrosine by autophosphorylation (Hunag et al., 2000), by MAPK kinases, or by other tyrosine kinases. In addition, they also suggest that plants may contain tyrosine phosphatases that are responsible for the dephosphorylation and inactivation of MAPKs during signalling processes. To date, one tyrosine-specific PTPase (AtPTP1) has been identified from Arabidopsis and its expression seems to be regulated by stress factors including high salt and cold stress (Xu et al., 1998). Further biochemical studies demonstrated that AtPTP1 dephosphorylates a MAPK from Arabidopsis in vitro and inactivates the kinase (Huang et al., 2000) (Fig. 4). A study also shows that other plants produce AtPTP1 homologues (Fordham-Skelton et al., 1999). A dual-specificity PTPase (AtDsPTP1) has also been identified from Arabidopsis and shown to dephosphorylate and inactivate an MAPK from the same plant (Gupta et al., 1998). These studies suggest that plant PTPases may function to regulate MAPK pathways in response to a number of extracellular signals. Further studies are being conducted to identify the in vivo targets of these phosphatases and their function in plant cellular and developmental processes.

Figure 4.

 AtPTP1 plays a role in a stress responsive mitogen activated protein kinase (MAPK) pathway. Stress signal is detected by sensor or receptor proteins in the plasma membrane and transduced into activation of MAPKKK, MAPKK, and MAPK in a kinase cascade. AtPTP1 dephosphorylates MAPK and inactivates it. 

One approach we are currently taking is a reverse genetics procedure that identifies insertional alleles for the gene of interest. This procedure takes advantage of random insertion of T-DNA (or a transposon) into the Arabidopsis genome, thereby generating an insertional mutant of genes in the genome (Azpiroz-Leehan & Feldman, 1997). From a large collection of T-DNA-tagged population, we have identified a T-DNA insertional allele of AtPTP1. We have begun to analyse the phenotypic and molecular changes in these plants. Some initial results suggested that AtPTP1 may be involved in MAPK regulation in response to stress factors (R. Gupta & S. Luan, unpublished results, Fig. 4).

Tyrosine phosphatase in cell cycle regulation

Cyclin-dependent protein kinases (CDKs) are positively regulated by cdc25, a dual-specificity tyrosine phosphatase (Dunphy, 1994). Because the cell cycle control mechanism is highly conserved among eukaryotes, one would expect that plants would also utilize tyrosine phosphorylation as a regulatory step for CDK activation. However, a bona fide cdc25 homologue has not been identified from plants. Several studies provide evidence that plant CDKs may be regulated by tyrosine phosphatases (Zhang et al., 1996; Meszaros et al., 2000). In these studies, it has been shown that plant CDKs can be activated by yeast or fly cdc25, indicating that CDK tyrosine phosphorylation may play a role as a negative regulation step. If plants do not have a typical cdc25, other PTPases may take its place to function in cell cycle control by regulating CDK activity.

Phosphotyrosine proteins in plants

Regarding other possible substrates of PTPases in plants (in addition to MAPKs), recent studies have shown that profilin is phosphorylated on the tyrosine residues (Guillen et al., 1999). Consistent with the notion that cytoskeleton organization may be regulated by tyrosine phosphorylation, another study provides evidence that actin seems to be phosphorylated at tyrosine residues and dephosphorylation of actin may be involved in leaflet movement in mimosa (Kameyama et al., 2000). Using antiphosphotyrosine antibody as a probe, a study shows that a number of putative tyrosine phosphorylated proteins in plants may be altered by developmental stages and conditions (Barizza et al., 1999). A high-throughput proteomics approach is being used to survey the phosphotyrosine protein profiles in Arabidopsis under various conditions (Hunt & Luan, unpublished). Such studies will most likely reveal more proteins that are phosphorylated at tyrosine residues and will provide molecular targets for further studies using genetics and genomics tools.

Some indirect evidence also suggests that tyrosine phosphorylation is important in higher plants. For example, a tyrosine kinase in some plant bacterial pathogens is important for the virulence of the pathogen (Ilan et al., 1999). Rhizogene transformation of roots is accompanied by the presence of a tyrosine kinase activity (Rodriguez-Zapata et al., 1998). These studies suggest that tyrosine phosphorylation plays a role in plant–microbe interaction.

Future perspectives

The biochemical properties of protein phosphatases have been highly conserved among eukaryotic organisms. Future efforts will certainly be directed towards the functional analysis of these enzymes in a context of plant cell and developmental biology. As genome sequencing projects move along, all the genes for protein phosphatases in model plants (e.g. Arabidopsis) will be identified. The question is how to make the connection from the sequence information to the functional significance. Several approaches have already been taken in the study of plant protein phosphatases and will continue to shed light on the subject. Our studies on AtPTP1 have provided examples of how the function of these genes can be explored by combination of molecular biology, genomics, and genetics approaches. As more genes for protein phosphatases are identified by genomic sequencing, the reverse genetics and functional genomics approaches will be increasingly important in dissecting the role of these signalling enzymes. Despite the structural similarities in all eukaryotes, protein phosphatases certainly play distinct functions in various organisms with different developmental programs. These unique functions in plant growth and development make the research on plant protein phosphatases necessary and exciting.

Acknowledgements

Research in the authors’ laboratory is supported by National Institute of Health and US Department of Agriculture.

Ancillary