11
Review article PTEN: The down side of PI 3-kinase signalling Nick R. Leslie*, C. Peter Downes Division of Cell Signalling, School of Life Sciences, Medical Sciences Institute, University of Dundee, Dundee, DD1 5EH Scotland, UK Received 6 June 2001; accepted 14 August 2001 Abstract The PTEN tumour suppressor protein is a phosphoinositide 3-phosphatase that, by metabolising phosphatidylinositol 3,4,5-trisphosphate (PtdIns(3,4,5)P 3 ), acts in direct antagonism to growth factor stimulated PI 3-kinases. A wealth of data has now illuminated pathways that can be controlled by PTEN through PtdIns(3,4,5)P 3 , some of which, when deregulated, give a selective advantage to tumour cells. Early studies of PTEN showed that its activity was able to promote cell cycle arrest and apoptosis and inhibit cell motility, but more recent data have identified other functional consequences of PTEN action, such as effects on the regulation of angiogenesis. The structure of PTEN includes several features not seen in related protein phosphatases, which adapt the enzyme to act efficiently as a lipid phosphatase, including a C2 domain tightly associated with the phosphatase domain, and a broader and deeper active site pocket. Several pieces of data indicate that PTEN is a principal regulator of the cellular levels of PtdIns(3,4,5)P 3 , but work is only just beginning to uncover mechanisms by which the cellular activity of PTEN can be controlled. There also remains the vexing question of whether any of PTEN’s cellular functions reflect its evolutionary roots as a member of the protein tyrosine phosphatase superfamily. D 2002 Elsevier Science Inc. All rights reserved. Keywords: Phosphoinositide; Tumour suppressor; Phosphatase; Lipid signalling 1. Introduction 1.1. Lipid signalling and the role of PtdIns(3,4,5)P 3 Many different phosphorylated derivatives of the lipid, phosphatidylinositol (PtdIns), exist in cells and play diverse roles in cellular signalling. These inositol phospholipids differ in both the number and distribution of phosphate groups around the inositol ring, resulting in at least eight naturally occurring phosphoinositide species. The versatility of these molecules as cellular signals results from the ability of lipid-binding protein domains (such as pleckstrin homo- logy and FYVE zinc finger domains) specifically to recog- nise particular configurations of the inositol phosphate headgroup. Each inositol lipid signal molecule is thus able to recruit and, in some cases, activate a range of protein- binding partners. Moreover, since these lipids are likely to be limited in their distribution to the membranes in which they are made, both temporal and spatial changes in their abundance are likely profoundly to influence the signalling outcomes, (see Refs. [1–5] for recent reviews). The approx- imate levels of the naturally occurring inositol phospholi- pids and some proposed routes of production in a ‘model’ mammalian cell are indicated in Fig. 1. Much attention has focused on PtdIns(3,4,5)P 3 (Fig. 1) as an intracellular second messenger produced rapidly in response to many divergent cellular stimuli, including many hormones and growth factors (reviewed by Refs. [4,6 – 8]). It is one of the least abundant inositol phospholipids, being present only at approximately 0.005% of the abundance of PtdIns in most cell types analysed (this corresponds to a cellular concentration in the tens of nanomolar range, but presumably far higher in the membrane compartments within which they are found), broadly similar to PtdIns(3,4)P 2 and PtdIns(3,5)P 2 (Fig 1). The general mech- anism for the production of PtdIns(3,4,5)P 3 appears to be through the regulated action of Class I PI 3-kinase enzymes on the relatively abundant phosphoinositide, PtdIns(4,5)P 2 . The production of PtdIns(3,4,5)P 3 stimulates many downstream 0898-6568/02/$ – see front matter D 2002 Elsevier Science Inc. All rights reserved. PII:S0898-6568(01)00234-0 * Corresponding author. Tel.: +44-1382-345156; fax: +44-1382- 201063. E-mail addresses: [email protected] (N.R. Leslie), [email protected] (C.P. Downes). www.elsevier.com/locate/cellsig Cellular Signalling 14 (2002) 285 – 295

PTEN: The down side of PI 3-kinase signalling

Embed Size (px)

Citation preview

Review article

PTEN:

The down side of PI 3-kinase signalling

Nick R. Leslie*, C. Peter Downes

Division of Cell Signalling, School of Life Sciences, Medical Sciences Institute, University of Dundee, Dundee, DD1 5EH Scotland, UK

Received 6 June 2001; accepted 14 August 2001

Abstract

The PTEN tumour suppressor protein is a phosphoinositide 3-phosphatase that, by metabolising phosphatidylinositol 3,4,5-trisphosphate

(PtdIns(3,4,5)P3), acts in direct antagonism to growth factor stimulated PI 3-kinases. Awealth of data has now illuminated pathways that can

be controlled by PTEN through PtdIns(3,4,5)P3, some of which, when deregulated, give a selective advantage to tumour cells. Early studies

of PTEN showed that its activity was able to promote cell cycle arrest and apoptosis and inhibit cell motility, but more recent data have

identified other functional consequences of PTEN action, such as effects on the regulation of angiogenesis. The structure of PTEN includes

several features not seen in related protein phosphatases, which adapt the enzyme to act efficiently as a lipid phosphatase, including a C2

domain tightly associated with the phosphatase domain, and a broader and deeper active site pocket. Several pieces of data indicate that

PTEN is a principal regulator of the cellular levels of PtdIns(3,4,5)P3, but work is only just beginning to uncover mechanisms by which the

cellular activity of PTEN can be controlled. There also remains the vexing question of whether any of PTEN’s cellular functions reflect its

evolutionary roots as a member of the protein tyrosine phosphatase superfamily. D 2002 Elsevier Science Inc. All rights reserved.

Keywords: Phosphoinositide; Tumour suppressor; Phosphatase; Lipid signalling

1. Introduction

1.1. Lipid signalling and the role of PtdIns(3,4,5)P3

Many different phosphorylated derivatives of the lipid,

phosphatidylinositol (PtdIns), exist in cells and play diverse

roles in cellular signalling. These inositol phospholipids

differ in both the number and distribution of phosphate

groups around the inositol ring, resulting in at least eight

naturally occurring phosphoinositide species. The versatility

of these molecules as cellular signals results from the ability

of lipid-binding protein domains (such as pleckstrin homo-

logy and FYVE zinc finger domains) specifically to recog-

nise particular configurations of the inositol phosphate

headgroup. Each inositol lipid signal molecule is thus able

to recruit and, in some cases, activate a range of protein-

binding partners. Moreover, since these lipids are likely to

be limited in their distribution to the membranes in which

they are made, both temporal and spatial changes in their

abundance are likely profoundly to influence the signalling

outcomes, (see Refs. [1–5] for recent reviews). The approx-

imate levels of the naturally occurring inositol phospholi-

pids and some proposed routes of production in a ‘model’

mammalian cell are indicated in Fig. 1.

Much attention has focused on PtdIns(3,4,5)P3 (Fig. 1) as

an intracellular second messenger produced rapidly in

response to many divergent cellular stimuli, including many

hormones and growth factors (reviewed by Refs. [4,6–8]).

It is one of the least abundant inositol phospholipids, being

present only at approximately 0.005% of the abundance of

PtdIns in most cell types analysed (this corresponds to a

cellular concentration in the tens of nanomolar range, but

presumably far higher in the membrane compartments

within which they are found), broadly similar to

PtdIns(3,4)P2 and PtdIns(3,5)P2 (Fig 1). The general mech-

anism for the production of PtdIns(3,4,5)P3 appears to be

through the regulated action ofClass I PI 3-kinase enzymes on

the relatively abundant phosphoinositide, PtdIns(4,5)P2. The

production of PtdIns(3,4,5)P3 stimulates many downstream

0898-6568/02/$ – see front matter D 2002 Elsevier Science Inc. All rights reserved.

PII: S0898 -6568 (01 )00234 -0

* Corresponding author. Tel.: +44-1382-345156; fax: +44-1382-

201063.

E-mail addresses: [email protected] (N.R. Leslie),

[email protected] (C.P. Downes).

www.elsevier.com/locate/cellsig

Cellular Signalling 14 (2002) 285–295

cellular processes through its interactions with proteins that

contain modular ligand-binding domains (pleckstrin homo-

logy or PH domains) able to bind with some specificity to

PtdIns(3,4,5)P3 [3,9,10].

2. PTEN is a tumour suppressor gene that encodes

a PtdIns(3,4,5)P3 3-phosphatase

PTEN (Phosphatase and TENsin homologue deleted on

chromosome TEN) is one of the few human tumour sup-

pressor genes for which this status is supported by several

strong lines of evidence. PTEN, also called MMAC1 and

TEP1, was first identified in 1997 as a candidate tumour

suppressor locus found to be mutated in sporadic glioblas-

tomas, prostate and breast cancers and located at chro-

mosome 10q23, a region deleted at very high frequency in

many tumour types [11,12]. Since then it has become clear

that PTEN mutation is associated with many human tumour

types, but at particularly high frequency in endometrial

tumours and glioblastomas (reviewed in Refs. [13,14]).

Human germline mutations in PTEN are associated with

the rare autosomal dominant hamartomatous syndromes,

Cowden Disease and Bannayan–Riley–Ruvalcaba syn-

drome, the former including an increased risk of breast

and thyroid tumours. Mice homozygous for targeted dele-

tions within the PTEN gene (PTEN� /�) die during

embryonic development (between Days 6.5 and 9.5), while

heterozygous mice develop normally, but are prone to a

wide range of tumour types. Evidence supporting the

tumour suppressor status of PTEN has been reviewed

recently [15–17].

The tumour suppressor gene, PTEN, encodes a 403

amino acid protein, PTEN, that is a member of the protein

Fig. 1. (A) Pathways of cellular inositol phospholipid synthesis. The inositol phospholipids identified in mammalian cells are shown, along with the proposed

pathways for their synthesis from PtdIns. We have indicated the relative abundances of these lipids in resting cells by the size of the represented box and text.

The data for these relative levels are based on the well-studied 1321N1 astrocytoma cell line, but are representative of many mammalian cell types and are

shown as the percentage abundance relative to PtdIns. PtdIns(4)P and PtdIns(4,5)P2 each comprise about 5% of PtdIns; PtdIns(3)P and PtdIns(5)P are about

0.25% of PtdIns; and PtdIns(3,4)P2, PtdIns(3,5)P2, and PtdIns(3,4,5)P3 are the least abundant of the identified phosphoinositides comprising about 0.005%

under basal conditions. Of the many pathways of phosphoinositide dephosphorylation, only those thought to be catalysed by PTEN are shown. (B) Structure of

PtdIns(3,4,5)P3. The structure of sn-1:2 phosphatidyl D-myo-inositol 3,4,5-trisphosphate [elsewhere abbreviated to PtdIns(3,4,5)P3] is shown. The fatty acid

side chains have been omitted. Most cellular PtdIns appears to be sn-1-stearoyl:2-arachidonyl PtdIns and it is assumed that other phosphoinositides derived

from PtdIns share this composition.

N.R. Leslie, C.P. Downes / Cellular Signalling 14 (2002) 285–295286

tyrosine phosphatase family. Soon after the identification of

PTEN, data were published making a strong connection

between the phosphatase activity of PTEN and its tumour

suppressor function and this conclusion has been supported

by extensive studies of tumour-derived PTEN mutants

[18–20]. It now appears that almost all pathological PTEN

mutations remove or impair phosphatase activity in vitro

[18,21,22]. This indicated that the tumour suppressor func-

tion of PTEN requires the dephosphorylation of a particular

substrate or substrates. The poor activity of PTEN against

phosphoproteins and peptide substrates, and its preference

for highly acidic substrates suggested that the physiological

substrates of PTEN might be acidic nonphosphoprotein

molecules, such as phospholipids [23,24]. This led to

the significant finding that PTEN dephosphorylates

PtdIns(3,4,5)P3 in vitro and in vivo [23–25]. Additionally,

the analysis of a tumour-derived PTEN mutant (PTEN

G129E) that retained activity against phosphotyrosine sub-

strates, but lacked detectable activity against phosphoinosi-

tides, established the link between the tumour suppressor

activity of PTEN and this lipid phosphatase activity

[24,26]. Many forms of evidence have now shown that

PtdIns(3,4,5)P3 is a physiological substrate of PTEN and

that PTEN functions to antagonise growth factor stimulated

PI 3-kinase signalling (see Fig. 2).

2.1. Biochemical analyses of substrate specificity

PTEN will dephosphorylate several phosphoinositide

signalling molecules in vitro, specifically removing phos-

phate from the D-3 position of the inositol ring in each case

[23,24]. PtdIns(3,4,5)P3 and PtdIns(3,4)P2 are the most

efficient substrates for PTEN in vitro, although it will also

dephosphorylate PtdIns(3)P and the soluble headgroup

of PtdIns(3,4,5)P3, inositol 1,3,4,5-tetrakisphosphate

(Ins(1,3,4,5)P4). There is also a report of activity against

PtdIns(3,5)P2 [27]. However, PtdIns(3)P and Ins(1,3,4,5)P4have been tested and are poorer substrates in vitro, and

overexpression of PTEN in cells has little effect on the

levels of PtdIns(3)P or PtdIns(3,5)P2 (McConnachie, Ben-

nett, Pass, Leslie and Downes, unpublished information).

Therefore, it is believed that the relevant physiological

substrates of PTEN are PtdIns(3,4,5)P3 and possibly

PtdIns(3,4)P2. Importantly, this phosphoinositide phospha-

tase activity of PTEN is not a feature of other related protein

tyrosine phosphatase family members, such as PTP1D,

VHR, Cdc25 and Cdc14 [23,24].

2.2. PTEN reduces cellular levels of PtdIns(3,4,5)P3 and

antagonises PI 3-kinase signalling

The effects of PTEN activity on cellular PtdIns(3,4,5)P3levels have been studied in a variety of cell types including

several PTEN null cell lines [23,25,28,29]. Murine fibro-

blasts and embryonic stem cells with targeted deletion of

both Pten alleles have greatly elevated unstimulated levels

of PtdIns(3,4,5)P3 relative to isogenic Pten wild-type cells

[25,29]. Similar results have also been obtained with human

tumour cell lines lacking PTEN, in which PtdIns(3,4,5)P3levels appear higher than other PTEN expressing cells and

are reduced dramatically upon reexpression of the phospha-

tase [28]. These experiments involving HPLC analysis of

labelled cellular phosphoinositides have also shown that

levels of PtdIns(3,4)P2 appear similarly high and are

reduced by PTEN reexpression [28]. In addition to this

direct evidence, a wealth of data has shown greatly

increased activity of the downstream PtdIns(3,4,5)P3 regu-

lated kinase, protein kinase B (PKB, also known as Akt), in

many PTEN null cell types, and that this activity is reduced

by expression of PTEN [24,25,29–36].

The conclusion that PTEN is a functional antagonist of

PI 3-kinase has been greatly strengthened by genetic ana-

lyses in nematodes (Caenorhabditis elegans), and insects

(Drosophila melanogaster). Previous results from these

organisms established that several components of the human

PI 3-kinase signalling pathway are conserved in both

sequence and function, including the activation of Class I

PI 3-kinases by growth factor receptors, the production of

PtdIns(3,4,5)P3, and the activation of PKB [37–39]. The

conservation of this signalling module has allowed genetic

analysis in flies and worms to demonstrate the role of

orthologues of PTEN as functional antagonists of PI

3-kinase upstream of PKB [40–46]. Stimulation of this

signalling pathway transduces rather different signals in

these organisms, however, inducing survival, proliferation,

and many of the metabolic actions of insulin in mammals;

Fig. 2. Model for the inhibition of PI 3-kinase signalling by PTEN and SHIP. PTEN is a phosphoinositide 3-phosphatase, able to remove phosphate from the 3

position of the inositol ring of several cellular signalling molecules including PtdIns(3,4,5)P3, converting it to PtdIns(4,5)P2. The reverse reaction is catalysed

by Type I PI 3-kinase enzymes activated by numerous cellular stimuli, including growth factors. SHIPs (two isoforms, SHIP1 and SHIP2 are currently known)

are 5-phosphatases that convert PtdIns(3,4,5)P3 to PtdIns(3,4)P2.

N.R. Leslie, C.P. Downes / Cellular Signalling 14 (2002) 285–295 287

increasing proliferation and cell size in flies; and inhibiting

dauer arrest, fat storage, and longevity in worms.

Final evidence for the mutually antagonistic functions of

PTEN and PI 3-kinase arises from the frequent finding that

many of the cellular effects of PTEN are mimicked by

pharmacological inhibitors of PI 3-kinase. Two structurally

unrelated inhibitors have been extensively used, the revers-

ible ATP-competitive, LY294002 and the irreversible fungal

metabolite, wortmannin [4,9,47]. While many cellular

effects of these inhibitors are shared with PTEN, in addition

to the Class I PI 3-kinases, these compounds are also able to

inhibit a number of other related enzymes, although often

only at higher concentrations. This can lead to inhibition, for

example, of processes mediated by Type III PI 3-kinases via

PI(3)P production, that would not be affected by PTEN. As

with all pharmacological agents care must be exercised both

in the dose of inhibitor employed and in the understanding

that specificity is rarely absolute.

2.3. Is PTEN the principal cellular regulator of basal

PtdIns(3,4,5)P3 levels?

Currently, there is little evidence that other enzymes or

mechanisms reduce basal cellular PtdIns(3,4,5)P3 levels in

most cells in the absence of PTEN. Several PTEN null cell

types from both tumours and knockout mice have been

demonstrated to have greatly elevated basal levels of

PtdIns(3,4,5)P3 [25,28,29] and PKB activity [25,29–31].

However, since elevated PtdIns(3,4,5)P3 is likely to confer a

selective advantage on cultured cells, it is possible that

analysis of both tumours and mouse embryos misses cell

types in which loss of PTEN does not lead to elevated levels

of PtdIns(3,4,5)P3 due to functional redundancy, and there-

fore does not confer a selective advantage. This is an

important point, as it may be a factor in determining the

occurrence of PTEN mutations in different tumour types.

2.4. Other PtdIns(3,4,5)P3 phosphatases— the SHIP

proteins and PTEN homologues

Other PtdIns(3,4,5)P3 phosphatases exist, for example,

the 5-phosphatases, SHIP and SHIP2 [48], and at least two

PTEN homologues, TPTE and PTEN2 [27,49]. However,

while there is strong evidence that the SHIP proteins do

metabolise PtdIns(3,4,5)P3 in several cell types (to produce

PtdIns(3,4)P2), they appear not to regulate basal

PtdIns(3,4,5)P3 levels, but rather may control the duration

and magnitude of stimulated increases in this lipid [50]. This

conclusion is derived from data showing that SHIP null

mice develop normally, and that bone marrow-derived mast

cells from wild-type and mutant mice appear to have similar

basal levels of cellular PtdIns(3,4,5)P3 and PKB activation.

However, upon IL-3 stimulation, the cells from SHIP null

mice gave a larger and more prolonged response than wild-

type cells in terms of both PtdIns(3,4,5)P3 levels and PKB

activation [50]. Mice lacking SHIP2, the more widely

expressed 5-phosphatase, also develop normally up to birth

but exhibit insulin hypersensitivity, supporting the idea that

these enzymes may not be regulators of basal PtdIns(3,4,5)P3levels in many cell types [51]. However, currently, no work

has been published investigating downstream signalling in

cells from these mice.

Another functional difference between PTEN and the

SHIP proteins is that whereas PTEN converts PtdIns(3,4,5)P3to the abundant PtdIns(4,5)P2, the SHIP proteins convert

PtdIns(3,4,5)P3 to PtdIns(3,4)P2 (see Fig. 2). Because

PtdIns(3,4)P2 appears to be of similar abundance to

PtdIns(3,4,5)P3 in most cell types, this reaction probably

has a significant effect on the cellular abundance of

PtdIns(3,4)P2. The recent identification of PH domain con-

taining proteins able to bind specifically to PtdIns(3,4)P2 has

strengthened the argument that PtdIns(3,4)P2 is a second

messenger with roles independent from PtdIns(3,4,5)P3[52]. Therefore, the production of PtdIns(3,4)P2 from

PtdIns(3,4,5)P3 by SHIP proteins may convert one signal to

another, whereas PTEN action appears to mediate a simple

inactivation of the PtdIns(3,4,5)P3 signal.

There are several genes in the human genome very

similar to PTEN, and there is significant evidence that at

least two should express functional proteins and at least one

is expected to be able to metabolise PtdIns(3,4,5)P3 (Refs.

[27,49], and Walker, Leslie, and Downes, unpublished).

However, current data suggest that the tissue distribution

of these proteins may be tightly restricted, possibly being

expressed only in the testis.

2.5. Is all PtdIns(3,4,5)P3 functionally equivalent?

Current data suggest that in many cell types PTENmay be

a principal negative regulator of PtdIns(3,4,5)P3 levels. How-

ever, phospholipids are clearly not uniformly distributed

throughout the cell, and are probably not even distributed

uniformly throughout topologically continuous membrane

compartments. For example, PtdIns(4,5)P2 is found in many

different membranes within the cell and even associated with

nonmembranous structures within the nucleus [1,5,53]. It

seems likely that the cellular distribution of PtdIns(3,4,5)P3within the plasma membrane is often not uniform, depending

upon the type and environment of the cell, and that the effects

of this nonuniformity on downstream processes may be

significant. Strong evidence, particularly from studies of

cellular chemotaxis supports this idea [54,55]. In both the

slime mold Dictyostelium discoideum and in mammalian

neutrophil-differentiated HL-60 cells, PtdIns(3,4,5)P3and/or PtdIns(3,4)P2 accumulates at the leading edge of the

migrating cell as indicated by the localised recruitment of a

fluorescently tagged PH domain [56,57].

It seems quite possible, therefore, that precise targeting of

PTEN to particular cellular pools of PtdIns(3,4,5)P3 could

have significant functional consequences. Several cellular

proteins have been identified that interact with PTEN through

PDZdomains and it has been shown that interferingwith these

N.R. Leslie, C.P. Downes / Cellular Signalling 14 (2002) 285–295288

interactions affects some functions of PTEN, but not others.

This will be discussed in more detail below in the Section 5.

2.6. Is the protein phosphatase activity of PTEN

functionally significant?

In addition to its lipid phosphatase activity, human PTEN

also has significant protein tyrosine phosphatase activity in

vitro [19,58]. While strong evidence connects the lipid

phosphatase activity of PTEN with most of its identified

downstream functions, including tumour suppressor activity,

it remains possible that there are also physiological protein

substrates for PTEN. PTEN is a dual-specificity protein

tyrosine phosphatase in vitro, as it has detectable activity

against phosphoserine- and phosphothreonine-containing

substrates in addition to phosphotyrosine. However, the best

peptide substrates that have been described for PTEN in

vitro are acidic phosphotyrosine-containing polypeptides

[19]. The signalling proteins, focal adhesion kinase (FAK)

and Shc, are substrates for PTEN in vitro that have been

proposed as potential cellular targets of the enzyme. It has

also been suggested that dephosphorylation of these sub-

strates could mediate the inhibition of cell motility caused

by PTEN [59,60]. However, two pieces of evidence argue

against this. Firstly, neither phosphorylation of FAK, nor

cell motility, appear to be affected by expression of a PTEN

mutant that lacks lipid phosphatase activity, but retains

activity against phosphotyrosine substrates in vitro (PTEN

G129E) [61]. Secondly, evidence has been presented iden-

tifying signalling pathways leading from PtdIns(3,4,5)P3,

the known lipid substrate of PTEN, to phosphorylation of

both FAK and Shc, that could explain the effects of PTEN

on the phosphorylation states of these proteins [62,63].

More recent data, however, have implicated the protein

phosphatase activity of PTEN in reducing the cellular levels

of cyclin D1 in the breast cancer cell line MCF7 [64]. In the

latter experiments, PTEN G129E failed to inhibit PKB

activity, but appeared fully capable of down-regulating

cyclin D1. It will be interesting to see whether these effects

are found in other cell types, and to identify specific

substrates upon which the protein phosphatase activity of

PTEN might be acting. In this regard, a mutation in PTEN

that mirrors the effect of the G129E mutation, by blocking

protein phosphatase activity while retaining activity against

lipid substrates, would be most useful. Such a mutation has

not yet been reported.

3. The PTEN protein

3.1. Linking structure and function

Initial analysis of its amino acid sequence showed that

PTEN has an N-terminal phosphatase domain with homo-

logy to both the protein tyrosine phosphatase family and the

nonphosphatases tensin and auxilin [11,12] (Fig. 3). The

extreme C-terminus of PTEN was first predicted, then

shown experimentally, to contain a binding motif for a class

of PDZ protein–protein interaction domains [65–67]. How-

ever, significant understanding of the possible functions of

the remainder of the molecule was provided only by the

solution of its crystal structure [21].

The determination of the structure of PTEN has

identified several characteristics that are not shared by

other PTP enzymes and which direct specificity towards

inositol phospholipid substrates [21]. Most enzymes that

metabolise membrane lipids have other membrane-binding

motifs in addition to the active site, allowing the enzyme

to remain associated with a membrane during the com-

pletion of multiple catalytic cycles [68]. In addition to its

N-terminal phosphatase domain, PTEN has a calcium-

independent C2 domain that has been shown to bind to

lipid vesicles in vitro [21]. In other proteins, C2 domains

have been shown to mediate lipid binding and membrane

localisation [2,69]. The functional significance of the

PTEN C2 domain has also been demonstrated, as muta-

tions that reduce the lipid binding of the domain, without

interfering with in vitro phosphatase activity against a

soluble substrate, dramatically to impair all aspects of

downstream function investigated in cells, including the

reduction of cellular PtdIns(3,4,5)P3 levels [21,70,71].

Recent results also show that these C2 domain mutations

are only partially rescued by artificial membrane targeting

of the phosphatase [70]. It therefore seems possible that

this C2 domain may not only allow transient association

of PTEN with cellular membranes, but may play a role in

ensuring the correct productive orientation of the active

site [21,70]. This idea is also supported by the extensive

tight interface between the phosphatase domain and the

Fig. 3. The PTEN protein. The 403 amino acid PTEN protein is represented. The N-terminal phosphatase domain (amino acids 7–185), including the catalytic

cysteine (C) and the C2 domain (186–351), Are hatched and both required for enzymatic activity. The extreme C-terminal PDZ-binding sequence is also

shown and although it is represented as a small region, the extent of further sequences required for optimal specificity and affinity of binding is not known. The

existence of probable phosphorylation sites in the C-terminal tail is represented by a circled letter P. The amino acid numbering of domains is derived from the

crystal structure of the protein [21].

N.R. Leslie, C.P. Downes / Cellular Signalling 14 (2002) 285–295 289

C2 domain, that is one of the most conserved regions of

the protein between divergent species, and is also a

frequent target of mutation in tumours. Extensive muta-

tion studies seem to show that together, the phosphatase

and C2 domains form a minimal catalytic unit, as

truncation mutations removing only a few residues from

the C-terminus of the C2 domain ablate all detectable

phosphatase activity, even against soluble substrates. On

the other hand, mutations in the C-terminal tail, beyond

the C2 domain, seem to have little or no effect upon

enzyme activity [18,21,22,72,73].

A further structural adaptation of PTEN, contrasting with

other PTPases, is the size of the active site pocket. Phos-

photyrosine-specific classical PTPases for which structures

have been determined, such as PTP1B, have a deep narrow

active site cleft, whereas the dual-specificity phosphatases

(phosphotyrosine and phosphoserine/phosphothreonine-spe-

cific), such as VHR have a more shallow active site [74].

The active site pocket of PTEN, however, is deep, but also

very broad, allowing access to the bulky phosphorylated

inositol headgroup. It is also highly basic, consistent with

the preference for acidic substrates [21].

Unfortunately, in order to obtain a crystal structure, the

C-terminal 49 amino acids of PTEN needed to be deleted (as

well as the N-terminal 7 amino acids and a 24 residue C2

domain internal loop) [21]. While this C-terminal region is

not required for catalytic activity, or for most cellular

functions analysed, evidence suggests that it may play an

important regulatory role. The C-terminal sequence, beyond

the end of the C2 domain, has now been shown to contain

several phosphorylation sites, a PDZ domain-binding

sequence and two possible PEST sequences (Fig. 3). The

mechanisms and functional significance of these regulatory

mechanisms remain rather unclear, and will be discussed in

greater detail below, but greatly influence the stability of the

protein. It should be noted that the significance of the

homology between PTEN and the nonphosphatase mole-

cules, tensin and auxilin, is not yet clear.

3.2. Downstream of PtdIns(3,4,5)P3—cellular effects

of PTEN

Conserved signalling pathways can be used to transduce

very different information in different organisms and in

different cells in the same organism [75,76]. Therefore,

the cellular effects of PTEN action, and possibly any

selective advantage of PTEN loss in a tumour, will depend

on the predominant functions for which the PI 3-kinase

signalling pathway is used in any particular cell type.

However, it has been known for some years that in many

and possibly most mammalian cell types, PI 3-kinase is

activated by many growth factors and acts as a strong

survival and proliferative signal. We will not rigorously

review all the identified signalling events and biological

endpoints downstream of PI 3K as these have been reviewed

recently by others [6,7,77,78] or those shown to be regu-

lated by PTEN [15,79]. However, we will give a brief

account of some recent work that has shed significant light

on the mechanism of tumour suppression by PTEN.

Currently, several biological processes have been iden-

tified that can be controlled by PtdIns(3,4,5)P3 signalling,

many of which can be down-regulated by PTEN and are

likely to be pathologically significant in tumours with PTEN

mutations. As previously mentioned, PTEN has been shown

to inhibit cell cycle progression and survival [25,29,80], but

can also inhibit cell spreading and motility [60,61] and the

release of angiogenic factors [81–84] (Fig. 4). Additionally,

many other cellular processes have some requirement for

Type I PI 3-kinase signalling, such as endocytic pathways

[85–87], regulation of translation [88,89], and the activation

of certain transcription factors. While the signalling path-

ways linking PtdIns(3,4,5)P3 with many of these complex

biological processes are still rather poorly understood,

considerable progress has been made in the identification

of direct effectors of PtdIns(3,4,5)P3 signals, particularly

those containing PH domains shown to bind directly to this

signalling lipid (reviewed in Refs. [3,7,9,90]).

Much attention has focused on the serine/threonine kinase

PKB (also known as Akt) as a downstream target of PTEN.

This is largely because the roles of PtdIns(3,4,5)P3 and

PtdIns(3,4)P2 in the activation of this enzyme are well

established and because, as a proto-oncogene, PKB has been

shown to play an important role in the transduction of

survival and proliferative signals in mammalian cells

(reviewed in Refs. [78,79,91,92]). However, it is also critical

that elevated levels of cellular PtdIns(3,4,5)P3 seem suf-

ficient for the activation of PKB in the absence of other

stimuli, as several PTEN null cell types have been shown to

Fig. 4. Signalling pathways downstream of PtdIns(3,4,5)P3. Some of the

known signalling pathways downstream of the second messenger,

PtdIns(3,4,5)P3, are shown. The synthesis of this lipid is stimulated by

many extracellular stimuli, including growth factors, hormones, and

extracellular matrix components. PtdIns(3,4,5)P3 can bind to PH domains

within many target proteins, of which the best understood are shown, and

plays a role in the activation of the indicated kinases (PKB, some PKC

isoforms, S6K1, SGK, and the Tec family), PLCg, and exchange factors for

the GTPase, Rac. A guide to some of the biological endpoints regulated by

these pathways is also shown.

N.R. Leslie, C.P. Downes / Cellular Signalling 14 (2002) 285–295290

have almost maximal PKB activity in the absence of acute

growth factor stimulation [24,25,30]. A number of potential

cellular PKB substrates have been identified that may be

responsible for the survival and proliferative effects of PKB

activity. The most convincing at present are the proapoptotic

regulatory protein, Bad, and transcription factors of the

Forkhead family [77,78,92]. More substrates of PKB sig-

nificant in the transduction of these signals probably remain

to be identified.

Recent work studying the downstream pathways

through which PTEN induces cell cycle arrest and pro-

motes apoptosis has highlighted the importance of Fork-

head transcription factors and the cyclin-dependent kinase

inhibitor p27Kip1. PKB-mediated phosphorylation of some

Forkhead transcription factors has been shown to inhibit

their transcriptional activity through cytoplasmic localisa-

tion and to be constitutive in cells lacking PTEN. Sig-

nificantly, expression of a Forkhead protein that could not

be phosphorylated by PKB was able to induce cell cycle

arrest and apoptosis with similar efficiency compared with

the effects of PTEN expression, suggesting that the func-

tion of the Forkhead proteins could be sufficient to

mediate many of the downstream effects of PTEN [80].

One of the significant targets for transcriptional regulation

by these factors is the cyclin-dependent kinase inhibitor

p27Kip1 and evidence suggests that increases in p27 may

be required for the induction of a G1 cell cycle arrest and

tumour suppression by PTEN [93–95]. PTEN appears able

to affect both the transcription and protein stability of p27

so that it may exert a two-pronged approach to regulating

the cellular amounts of this key cell cycle regulator

[80,96–98].

4. Regulation of PTEN

Given the critical role of PTEN in antagonising PI 3-

kinase signalling pathways, it might be anticipated that the

enzyme would be the target of complex control mecha-

nisms. Until recently, few studies had addressed this issue,

but evidence is now emerging for functional regulation of

PTEN through effects on protein stability, localisation, and

transcription of the PTEN gene. The regulation of stability

and localisation appears to be achieved through the C-

terminal tail of PTEN beyond the C2 domain, a region not

required for either enzymatic activity or reduction of the

major pool of cellular PtdIns(3,4,5)P3. Interestingly,

although mutations do occur in this region of the protein

in tumours, they do so at lower frequency than other regions

[13,14]. The evidence suggests that this is because most

tumour-derived mutations within the phosphatase domain or

the C2 domain (up to amino acid 351) remove all detectable

PTEN phosphatase activity [18,72]. Mutations in the C-

terminal tail appear only partially to interfere with PTEN

function, possibly through effects on protein stability/

abundance or localisation [22,67,71,99] and in tumours

appear not to lead to the dramatic increase in PKB activity

seen with null mutations [71]. It is therefore interesting that

mutations in the C-terminal tail have never been described

in either Cowden disease and other inherited conditions, or

in endometrial tumours, where PTEN mutation appears to

be a very early event in tumour progression [13,14,100]. By

contrast, such mutations have been found rather frequently

in gliomas, where PTEN mutation is associated with late-

stage aggressive tumours (Table 1).

4.1. Phosphorylation of PTEN

It has recently become clear through the work of two

research groups that PTEN exists in cells as a phosphopro-

tein [22,99]. When cells are labelled with radioactive

phosphate, both endogenous and expressed PTEN proteins

incorporate label, predominantly into phosphoserine, but

also phosphothroenine residues. Although indirect, the data

can be interpreted to suggest that most cellular PTEN is

constitutively phosphorylated. Deletion and mutational

studies indicate that phosphorylation sites are probably

located in a cluster in the C-terminal tail of PTEN, beyond

the end of the C2 domain, possibly at amino acids S370,

S380, T382, T383 and S385. Casein kinase 2 (CK2) has

been proposed as the responsible kinase, based largely on in

vitro data [99], but this needs confirmation by more tech-

nically demanding in vivo labelling approaches.

Analysis of the functional consequences of phosphory-

lation revealed that wild-type PTEN or mutant PTEN

proteins with phosphorylation sites replaced by acidic

residues were found to be more stable in cells than proteins

in which phosphorylation sites had been replaced by alanine

residues, or where the whole C-terminus had been deleted

[22,99]. This agrees well with previous work, noted above,

demonstrating a role for the C-terminal tail in controlling

PTEN stability [72]. Evidence also suggests that PTEN may

be degraded by a proteosome-mediated pathway [99]. Phos-

phorylation appears to reduce the biological activity of trans-

fected PTEN as measured by the relative activities of

phosphorylation site mutants in the induction of growth arrest

and in a Forkhead transcriptional activation assay [22]. This

raises the interesting possibility that phosphorylation of

Table 1

Mutations in the C-terminal tail of PTEN

Total

mutations analysed

Mutations

in C-terminus

Glioblastomas 159 6

116 7

Endometrial carcinoma

and precancerous lesions

187

136

0

0

Germline 110 0

The frequency of occurrence of mutations in the whole coding region and

specifically the C-terminal tail (beyond the C2 domain, i.e., aa 352–

Stop404) of PTEN is shown. Two estimates for data from glioblastomas

and for endometrial tumours are shown from two reviews of PTEN

mutational data [13,14].

N.R. Leslie, C.P. Downes / Cellular Signalling 14 (2002) 285–295 291

PTEN in the C-terminal tail is able to stabilise but reduce

the activity of the protein. At present, there is no evidence of

conditions that regulate the degree of phosphorylation of

PTEN, and hence, the functional significance of these

observations remains unclear. It seems likely that phosphor-

ylation may not greatly affect the in vitro activity of the

enzyme against soluble substrates, since deletion of the

C-terminal phosphorylation sites appeared not to greatly

affect the activity of protein immunoprecipitated from

human 293T cells or expressed and purified from insect

cells [21,72]. Taken together with the activity of bacterially

expressed PTEN, this suggests that the protein has constitu-

tive phosphatase activity, but raises interest in the mech-

anism by which phosphorylation could be inhibiting the

biological actions of PTEN.

4.2. Regulation through localisation—PTEN

targeting proteins

It appears that in most cell types analysed, PTEN is

largely cytosolic [20,58,60]. Since its principal substrate is a

membrane phospholipid, and artificial membrane targeting

enhances effects on downstream signalling [66,70,71], sig-

nificant effort has been invested in investigating mecha-

nisms by which PTEN might become membrane associated

and/or incorporated in signalling complexes assembled at

membrane loci.

The extreme C-terminus of PTEN can mediate interac-

tions with PDZ domains in several proteins [65–67]. These

include the membrane associated guanylate kinase with

inverted orientation (MAGI) proteins 1, 2, and 3, the related

protein, discs large (DLG/SAP97), itself a tumour sup-

pressor in flies, and MAST205 (microtubule-associated

serine/threonine kinase of 205 kDa). An interaction between

endogenous proteins, however, has thus far only been

demonstrated between PTEN and MAGI-2 [67]. Unfortu-

nately, little is currently known about other binding partners

of these PDZ domain proteins and the signalling complexes

they are presumed to coordinate.

Two groups found that overexpression of MAGI proteins

enhanced PTEN’s ability to block PKB activation [66,67],

while others failed to observe any effect on PKB activity of

deleting the PDZ-binding sequence from PTEN [71,73]. On

the other hand, the C-terminal PDZ-binding sequence is

required for efficient inhibition of membrane ruffling and

cell spreading [71,101]. Since it was found that deletion of

the PDZ-binding sequence did not measurably affect the

ability of PTEN to reduce cellular levels of PtdIns(3,4,5)P3,

it seems that PDZ domain-mediated targeting of PTEN may

play a role in the regulation of processes that are highly

sensitive to small changes in the levels of PtdIns(3,4,5)P3(which include cell spreading) [71]. This in turn may

suggest that a small localised pool of PtdIns(3,4,5)P3 is

dedicated to regulating lamellipodia formation, but not the

activation of PKB.

Since the PDZ-binding sequence is tightly conserved

within vertebrate PTEN orthologues (from Mammals,

Zebrafish, and Xenopus), but absent in those from flies

and worms, it seems likely that these regulatory mecha-

nisms might have evolved for the ‘fine tuning’ of

PtdIns(3,4,5)P3 signalling in higher organisms. This could

include, for example, the signalling complexes on mem-

brane subdomains in which PDZ scaffolds have been

proposed to play roles in polarised epithelial cells and

synaptic membranes [102–105]. In other proteins, it has

been proposed that phosphorylation of a serine or threonine

residue within the C-terminal sequence bound by some PDZ

domains can modulate these protein–protein interactions.

Although it has been shown that phosphorylation of Thr401

can alter the pattern of PDZ proteins bound by PTEN-

derived peptides, there is currently no evidence for phos-

phorylation of this residue of PTEN in cells [65]. Human

tumours have been described with missense mutations in the

PTEN PDZ-binding motif, in the stop codon and in threo-

nine 401. These mutations interfere with the interaction of

PTEN with PDZ domain containing proteins and this would

argue that functions requiring PDZ targeting of PTEN may

be important in the development of these tumours [71].

However, it is also possible that these mutations have other

additional effects, such as influencing the stability of the

enzyme [67].

It should be noted that in some cell types PTEN appears

to be present in the nucleus [106–108]. Since there is no

strong evidence yet for the existence of PtdIns(3,4,5)P3 in

the nucleus, the role of nuclear PTEN remains unclear. It is

possible that it has functions other than PtdIns(3,4,5)P3metabolism, or that the protein is simply functionally

inactivated by sequestration in the nucleus away from

its substrates.

4.3. Regulation of expression

Although PTEN was initially named TEP1 by Li and Sun

[58] because of a dramatic decrease in its expression that

occurs in human keratinocytes treated with TGFb, relativelylittle is still known about mechanisms that regulate PTEN

expression. Evidence from breast, prostate, thyroid, pancre-

atic, and haematopoietic tumours suggests that PTEN

expression can be lost or greatly reduced in some tumours

without mutation of the coding sequence of the PTEN gene

[106,107,109–111]. These tumours may have mutation or

methylation of noncoding regions of PTEN required for

expression, particularly since the PTEN promoter has not

yet been identified. It is also possible that there is an

alteration in these tumours of mechanisms further upstream

regulating the expression of PTEN, and information is just

emerging concerning the normal regulation of PTEN

expression. It has been shown that PTEN expression is

greatly increased during either granulocytic or monocytic

differentiation of myeloid leukaemic cells [112]. Very

recently, it has been reported that PTEN expression is

N.R. Leslie, C.P. Downes / Cellular Signalling 14 (2002) 285–295292

profoundly regulated by agonists acting on the peroxisome

proliferator-activated receptor g (PPARg) in both macro-

phages and Caco-2 tumour cells [113]. PPARg appears

directly to regulate transcription of the PTEN gene leading

to increased mRNA, PTEN protein, PtdIns(3,4,5)P3 3-phos-

phatase activity and a concomitant reduction in the activity

state of PKB. These results suggest the exciting possibility

that up-regulation of PTEN might be one of the mechanisms

through which PPARg agonists exert their recently

described anti-inflammatory and anticancer actions [113].

They appear to present a paradox, however, since PPARg

agonists are currently proving effective in the treatment of

Type II diabetes, where, by antagonising the PI 3-kinase

pathway, up-regulation of PTEN would be expected to be

harmful. Critically, however, PTEN appears to up-regulate

expression of insulin receptor substrate (IRS)-2, which is

one of four related adaptor proteins through which insulin

and certain other stimuli activate PI 3-kinase [114]. This

may be a reflection of a feedback pathway in which a

persistently low basal level of PtdIns(3,4,5)P3 leads to up-

regulation of IRS-2 and an enhanced ability to respond to

insulin via IRS-2. Since transgenic mice lacking IRS-2, but

not those lacking IRS-1, exhibit many features of Type II

diabetes, it is likely that up-regulation of IRS-2 will be

beneficial in this disease. The idea that PTEN can be

regulated through changes in its level of expression, either

via transcriptional control pathways or by controlling the

turnover of the expressed protein, is in tune with an

important feature of the enzyme in that it appears to be a

constitutively active phosphatase. Currently, there are only

hints that PTEN may be acutely regulated through covalent

modifications or translocation to signalling complexes

(unlike the SHIP proteins, which are membrane recruited

upon many cellular stimuli).

5. Future perspectives

Changes in levels of expression of PTEN, such as in

response to pharmacological agents, stimuli such as TGFb,or in heterozygotes with only one functional PTEN allele,

appear to directly influence basal levels of PI 3-kinase-

dependent lipid signals. Bearing in mind that the up-regu-

lation of PTEN expression induced by PPARg agonists may

explain the therapeutic effects of these compounds not just

in cancer, but also in inflammatory conditions and non-

insulin-dependent diabetes, the identification of the PTEN

promoter must be a key step in future work. Further analysis

of the transcriptional regulation of PTEN may also go some

way to explaining the types of tumours in which PTEN

mutations occur. Recent work has highlighted the range of

processes that PI 3-kinase signalling can influence, and the

many relevant processes deregulated in tumours carrying

PTEN mutations. It will be interesting to see this list

expand, and to know whether the regulation of other

processes that require some input from PI 3-kinase signal-

ling, such as vesicle trafficking and translation, play roles in

PTEN-related tumorigenesis.

Another area of future interest will be the interplay

between PtdIns(3,4,5)P3 metabolism via PTEN and other

enzymes that fulfil a similar role. There are already hints of

PTEN homologues and it will be important to confirm not

only that these proteins possess lipid phosphatase activity,

but also to examine their functional roles in cells. Perhaps

more significant is the existence of a distinct pathway for

PtdIns(3,4,5)P3 metabolism via selective removal of the 5-

phosphate, catalysed by SHIP proteins. Since these enzymes

generate a product, PtdIns(3,4)P2, that is considered to be a

second messenger, metabolism via SHIP or PTEN should

have functionally distinct outcomes with implications for

these enzymes as targets for novel therapies.

Acknowledgments

We would like to thank Ian Batty and Ian Fleming for

helpful discussions and critical reviews of the manu-

script. Work in the CPD laboratory is supported by the

Medical Research Council and the Division of Signal

Transduction Therapy.

References

[1] Divecha N, Irvine RF. Cell 1995;80:269–78.

[2] Hurley JH, Meyer T. Curr Opin Cell Biol 2001;13:146–52.

[3] Lemmon MA, Ferguson KM. Biochem J 2000;350(Pt 1):1–18.

[4] Wymann MP, Pirola L. Biochim Biophys Acta 1998;1436:127–50.

[5] Hinchliffe KA, Ciruela A, Irvine RF. Biochim Biophys Acta 1998;

1436:87–104.

[6] Toker A. Mol Pharmacol 2000;57:652–8.

[7] Cantrell DA. J Cell Sci 2001;114:1439–45.

[8] Alessi DR, Downes CP. Biochim Biophys Acta 1998;1436:151–64.

[9] Vanhaesebroeck B, Waterfield MD. Exp Cell Res 1999;253:239–54.

[10] Fruman DA, Rameh LE, Cantley LC. Cell 1999;97:817–20.

[11] Li J, Yen C, Liaw D, Podsypanina K, Bose S, Wang SI, et al. Science

1997;275:1943–7.

[12] Steck PA, Pershouse MA, Jasser SA, Yung WK, Lin H, Ligon AH,

et al. Nat Genet 1997;15:356–62.

[13] Ali IU, Schriml LM, Dean M. J Natl Cancer Inst 1999;91:1922–32.

[14] Bonneau D, Longy M. Hum Mutat 2000;16:109–22.

[15] Simpson L, Parsons R. Exp Cell Res 2001;264:29–41.

[16] Vazquez F, Sellers WR. Biochim Biophys Acta 2000;1470:M21–35.

[17] Dahia PL. Endocr Relat Cancer 2000;7:115–29.

[18] Han SY, Kato H, Kato S, Suzuki T, Shibata H, Ishii S, et al. Cancer

Res 2000;60:3147–51.

[19] Myers MP, Stolarov JP, Eng C, Li J, Wang SI, Wigler MH, Parsons

NK, Tonks NK. Proc Natl Acad Sci USA 1997;94:9052–7.

[20] Furnari FB, Lin H, Huang HS, Cavenee WK. Proc Natl Acad Sci

USA 1997;94:12479–84.

[21] Lee JO, Yang H, Georgescu MM, Di Cristofano A, Maehama T,

Shi Y, et al. Cell 1999;99:323–34.

[22] Vazquez F, Ramaswamy S, Nakamura N, Sellers WR. Mol Cell Biol

2000;20:5010–8.

[23] Maehama T, Dixon JE. J Biol Chem 1998;273:13375–8.

[24] Myers MP, Pass I, Batty IH, Van der Kaay J, Stolarov JP, Hemmings

BA, et al. Proc Natl Acad Sci USA 1998;95:13513–8.

N.R. Leslie, C.P. Downes / Cellular Signalling 14 (2002) 285–295 293

[25] Stambolic V, Suzuki A, de la Pompa JL, Brothers GM, Mirtsos C,

Sasaki T, et al. Cell 1998;95:29–39.

[26] Furnari FB, Huang HJ, Cavenee WK. Cancer Res 1998;58:5002–8.

[27] Wu Y, Dowbenko D, Pisabarro M, Dillard-Telm L, Koeppen H,

Lasky LA. J Biol Chem 2001;2:2.

[28] Taylor V, Wong M, Brandts C, Reilly L, Dean NM, Cowsert LM,

Moodie S, Stokoe D. Mol Cell Biol 2000;20:6860–71.

[29] Sun H, Lesche R, Li DM, Liliental J, Zhang H, Gao J, et al. Proc Natl

Acad Sci USA 1999;96:6199–204.

[30] Haas-Kogan D, Shalev N, Wong M, Mills G, Yount G, Stokoe D.

Curr Biol 1998;8:1195–8.

[31] Ramaswamy S, Nakamura N, Vazquez F, Batt DB, Perera S, Roberts

TM, Sellers WR. Proc Natl Acad Sci USA 1999;96:2110–5.

[32] Li DM, Sun H. Proc Natl Acad Sci USA 1998;95:15406–11.

[33] Li J, Simpson L, Takahashi M, Miliaresis C, Myers MP, Tonks N,

Parsons R. Cancer Res 1998;58:5667–72.

[34] Persad S, Attwell S, Gray V, DelcommenneM, Troussard A, Sanghera

J, Dedhar S. Proc Natl Acad Sci USA 2000;97:3207–12.

[35] Nakashima N, Sharma PM, Imamura T, Bookstein R, Olefsky JM.

J Biol Chem 2000;275:12889–95.

[36] Wu X, Senechal K, Neshat MS, Whang YE, Sawyers CL. Proc Natl

Acad Sci USA 1998;95:15587–91.

[37] Stocker H, Hafen E. Curr Opin Genet Dev 2000;10:529–35.

[38] Vanhaesebroeck B, Leevers SJ, Panayotou G, Waterfield MD. Trends

Biochem Sci 1997;22:267–72.

[39] Coelho CM, Leevers SJ. J Cell Sci 2000;113:2927–34.

[40] Gao X, Neufeld TP, Pan D. Dev Biol 2000;221:404–18.

[41] Gil EB, Malone Link E, Liu LX, Johnson CD, Lees JA. Proc Natl

Acad Sci USA 1999;96:2925–30.

[42] Goberdhan DC, Paricio N, Goodman EC, Mlodzik M, Wilson C.

Genes Dev 1999;13:3244–58.

[43] Huang H, Potter CJ, Tao W, Li DM, Brogiolo W, Hafen E, Sun H,

Xu T. Development 1999;126:5365–72.

[44] Mihaylova VT, Borland CZ, Manjarrez L, Stern MJ, Sun H. Proc

Natl Acad Sci USA 1999;96:7427–32.

[45] Ogg S, Ruvkun G. Mol Cell 1998;2:887–93.

[46] Rouault JP, Kuwabara PE, Sinilnikova OM, Duret L, Thierry-Mieg

D, Billaud M. Curr Biol 1999;9:329–32.

[47] Walker HE, Pacold EM, Perisic O, Stephens L, Hawkins TP,

Wymann PM, Williams LR. Mol Cell 2000;6:909–19.

[48] Rohrschneider LR, Fuller JF, Wolf I, Liu Y, Lucas DM. Genes Dev

2000;14:505–20.

[49] Chen H, Rossier C, Morris MA, Scott HS, Gos A, Bairoch A,

Antonarakis SE. Hum Genet 1999;105:399–409.

[50] Liu Q, Sasaki T, Kozieradzki I, Wakeham A, Itie A, Dumont DJ,

Penninger JM. Genes Dev 1999;13:786–91.

[51] Clement S, Krause U, Desmedt F, Tanti JF, Behrends J, Pesesse X.

Nature 2001;409:92–7.

[52] Dowler S, Currie RA, Campbell DG, Deak M, Kular G, Downes CP,

Alessi DR. Biochem J 2000;351:19–31.

[53] D’Santos CS, Clarke JH, Divecha N. Biochim Biophys Acta 1998;

1436:201–32.

[54] Dekker LV, Segal AW, Science 2000;287:982–3, 985.

[55] Watton SJ, Downward J. Curr Biol 1999;9:433–6.

[56] Meili R, Ellsworth C, Lee S, Reddy TB, Ma H, Firtel RA. EMBO J

1999;18:2092–105.

[57] Servant G, Weiner OD, Herzmark P, Balla T, Sedat JW, Bourne HR.

Science 2000;287:1037–40.

[58] Li DM, Sun H. Cancer Res 1997;57:2124–9.

[59] Gu J, Tamura M, Pankov R, Danen EH, Takino T, Matsumoto K,

Yamada KM. J Cell Biol 1999;146:389–404.

[60] Tamura M, Gu J, Matsumoto K, Aota S, Parsons R, Yamada KM.

Science 1998;280:1614–7.

[61] Liliental J, Moon SY, Lesche R, Mamillapalli R, Li D, Zheng Y, Sun

H, Wu H. Curr Biol 2000;10:401–4.

[62] Lopez-Ilasaca M, Crespo P, Pellici PG, Gutkind JS, Wetzker R.

Science 1997;275:394–7.

[63] Casamassima A, Rozengurt E. J Biol Chem 1998;273:26149–56.

[64] Weng LP, Brown JL, Eng C. Hum Mol Genet 2001;10:599–604.

[65] Adey NB, Huang L, Ormonde PA, Baumgard ML, Pero R, Byreddy

DV, Tavtigian SV, Bartel PL. Cancer Res 2000;60:35–7.

[66] Wu Y, Dowbenko D, Spencer S, Laura R, Lee J, Gu Q, Lasky LA.

J Biol Chem 2000;275:21477–85.

[67] Wu X, Hepner K, Castelino-Prabhu S, Do D, Kaye MB, Yuan XJ,

et al. Proc Natl Acad Sci USA 2000;97:4233–8.

[68] Gelb MH, Min JH, Jain MK. Biochim Biophys Acta 2000;1488:

20–7.

[69] Rizo J, Sudhof TC. J Biol Chem 1998;273:15879–82.

[70] Georgescu MM, Kirsch KH, Kaloudis P, Yang H, Pavletich NP,

Hanafusa H. Cancer Res 2000;60:7033–8.

[71] Leslie NR, Bennett D, Gray A, Pass I, Hoang-Xuan K, Downes CP.

Biochem J 2001;357:427–35.

[72] Georgescu MM, Kirsch KH, Akagi T, Shishido T, Hanafusa H. Proc

Natl Acad Sci USA 1999;96:10182–7.

[73] Tolkacheva T, Chan AM. Oncogene 2000;19:680–9.

[74] Tonks NK, Neel BG. Curr Opin Cell Biol 2001;13:182–95.

[75] Chang L, Karin M. Nature 2001;410:37–40.

[76] Simon MA. Cell 2000;103:13–5.

[77] Chan TO, Rittenhouse SE, Tsichlis PN. Annu Rev Biochem 1999;

68:965–1014.

[78] Vanhaesebroeck B, Alessi DR. Biochem J 2000;346:561–76.

[79] Leslie NR, Biondi RM, Alessi DR. Chem Rev 2001;101:2365–80.

[80] Nakamura N, Ramaswamy S, Vazquez F, Signoretti S, Loda M,

Sellers WR. Mol Cell Biol 2000;20:8969–82.

[81] Wen S, Stolarov J, Myers MP, Su JD, Wigler MH, Tonks NK,

Durden DL. Proc Natl Acad Sci USA 2001;98:4622–7.

[82] Zhong H, Chiles K, Feldser D, Laughner E, Hanrahan C, Georgescu

MM, Simons JW, Semenza GL. Cancer Res 2000;60:1541–5.

[83] Zundel W, Schindler C, Haas-Kogan D, Koong A, Kaper F, Chen E,

et al. Genes Dev 2000;14:391–6.

[84] Jiang BH, Zheng JZ, Aoki M, Vogt PK. Proc Natl Acad Sci USA

2000;97:1749–53.

[85] Jackson TR, Kearns BG, Theibert AB. Trends Biochem Sci

2000;25:489–95.

[86] Li E, Stupack D, Klemke R, Cheresh DA, Nemerow GR. J Virol

1998;72:2055–61.

[87] Amyere M, Payrastre B, Krause U, Van Der Smissen P, Veithen A,

Courtoy PJ. Mol Biol Cell 2000;11:3453–67.

[88] Aoki M, Blazek E, Vogt PK. Proc Natl Acad Sci USA 2001;98:

136–41.

[89] Gingras AC, Raught B, Sonenberg N. Annu Rev Biochem 1999;

68:913–63.

[90] Leevers SJ, Vanhaesebroeck B, Waterfield MD. Curr Opin Cell Biol

1999;11:219–25.

[91] Toker A, Newton AC. Cell 2000;103:185–8.

[92] Datta SR, Brunet A, Greenberg ME. Genes Dev 1999;13:2905–27.

[93] Gottschalk AR, Basila D, Wong M, Dean NM, Brandts CH, Stokoe

D, Haas-Kogan DA. Cancer Res 2001;61:2105–11.

[94] Di Cristofano A, De Acetis M, Koff A, Cordon-Cardo C, Pandolfi

PP. Nat Genet 2001;27:222–4.

[95] Dijkers PF, Medema RH, Pals C, Banerji L, Thomas NS, Lam EW,

et al. Mol Cell Biol 2000;20:9138–48.

[96] Matsushima-Nishiu M, Unoki M, Ono K, Tsunoda T, Minaguchi T,

Kuramoto H, et al. Cancer Res 2001;61:3741–9.

[97] Mamillapalli R, Gavrilova N, Mihaylova VT, Tsvetkov LM, Wu H,

Zhang H, Sun H. Curr Biol 2001;11:263–7.

[98] Medema RH, Kops GJ, Bos JL, Burgering BM. Nature 2000;

404:782–7.

[99] Torres J, Pulido R. J Biol Chem 2001;276:993–8.

[100] Mutter GL, Lin MC, Fitzgerald JT, Kum JB, Baak JP, Lees JA, Weng

LP, Eng C. J Natl Cancer Inst 2000;92:924–30.

[101] Leslie NR, Gray A, Pass I, Orchiston EA, Downes CP. Biochem J

2000;346(Pt 3):827–33.

[102] Garner CC, Nash J, Huganir RL. Trends Cell Biol 2000;10:274–80.

N.R. Leslie, C.P. Downes / Cellular Signalling 14 (2002) 285–295294

[103] Craven SE, Bredt DS. Cell 1998;93:495–8.

[104] Fanning AS, Anderson JM. J Clin Invest 1999;103:767–72.

[105] Kim SK. Curr Opin Cell Biol 1997;9:853–9.

[106] Gimm O, Perren A, Weng LP, Marsh DJ, Yeh JJ, Ziebold U, et al.

Am J Pathol 2000;156:1693–700.

[107] Perren A, Komminoth P, Saremaslani P, Matter C, Feurer S, Lees JA,

Heitz PU, Eng C. Am J Pathol 2000;157:1097–103.

[108] Sultana N, Sultana N, Schonhoff CM, Mitra P, Poluha W, Lambert

S, et al. J Neurosci 2000;20:1404–13.

[109] Aguiar RC, Alberta J, Kum JB, Caron S, Sill H, et al. Hum Mol

Genet 1999;8:185–93.

[110] Wu X, Suzuki H, Reiter RE, Tran C, Vessella RL, et al. Proc Natl

Acad Sci USA 1998;95:5246–50.

[111] Weng LP, Boag AH, Ziebold U, Thakore K, Dahia PL, et al. Am J

Pathol 1999;155:1253–60.

[112] Hisatake J, O’Kelly J, Uskokovic MR, Tomoyasu S, Koeffler HP.

Blood 2001;97:2427–33.

[113] Pass I, Coxon P, Downes CP, Smith SA, Macphee CH, et al. Curr

Biol 2001;11:764–8.

[114] Simpson L, Li J, Liaw D, Hennessy I, Oliner J, Christians F, Parsons

R. Mol Cell Biol 2001;21:3947–58.

N.R. Leslie, C.P. Downes / Cellular Signalling 14 (2002) 285–295 295