Oncogene (2010) 29, 687–697 & 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 $32.00 www.nature.com/onc ORIGINAL ARTICLE Suppression of cellular proliferation and invasion by the concerted lipid and protein activities of PTEN

L Davidson1,4, H Maccario1,4, NM Perera1,4, X Yang2,3, L Spinelli1, P Tibarewal1, B Glancy1, A Gray1, CJ Weijer2, CP Downes1 and NR Leslie1

1Division of Molecular Physiology, College of Life Sciences, University of Dundee, Dundee, UK and 2Division of Cell and Developmental Biology, College of Life Sciences, University of Dundee, Dundee, UK

PTEN is a tumour suppressor with phosphatase activity in vivo, seems to inhibit cell proliferation, survival and in vitro against both lipids and proteins and other potential growth and, in some cell types, motility (Keniry and non-enzymatic mechanisms of action. Although the Parsons, 2008; Salmena et al., 2008). PTEN is a lipid importance of PTEN’s lipid phosphatase activity in phosphatase that inhibits phosphoinositide 3-kinase regulating the PI3K signalling pathway is recognized, (PI3K)-dependent signalling, by metabolizing the PI3K the significance of PTEN’s other mechanisms of action is product, PtdInsP3. Although PTEN’s PtdInsP3 phos- currently unclear. In this study, we describe the systematic phatase activity appears to mediate many of its cellular identification of a PTEN mutant, PTEN Y138L, with effects, PTEN has additional potential mechanisms of activity against lipid, but not soluble substrates. Using this action, including robust protein phosphatase activity mutant, we provide evidence for the interfacial activation and proposed non-enzymatic mechanisms (Myers et al., of PTEN against lipid substrates. We also show that 1997; Mahimainathan and Choudhury, 2004; Rafto- when re-expressed at physiological levels in PTEN null poulou et al., 2004; Okumura et al., 2005; Shen et al., U87MG glioblastoma cells, the protein phosphatase activity 2007). Several studies have reported lipid phosphatase- of PTEN is not required to regulate cellular PtdInsP3 levels independent effects of PTEN affecting cell migration, or the downstream protein kinase Akt/PKB. Finally, in and also contributing to other functions potentially three-dimensional Matrigel cultures of U87MG cells related to tumour suppression (Park et al., 2002; similarly re-expressing PTEN mutants, both the protein Freeman et al., 2003; Gildea et al., 2004; Raftopoulou and lipid phosphatase activities were required to inhibit et al., 2004; Leslie et al., 2007; Shen et al., 2007; invasion, but either activity alone significantly inhibited Trotman et al., 2007). A wealth of correlative data proliferation, albeit only weakly for the protein phospha- connects the regulation of PI3K signalling by PTEN tase activity. Our data provide a novel tool to address the with its physiological functions and tumour suppressor significance of PTEN’s separable lipid and protein activity (Sansal and Sellers, 2004). However, given the phosphatase activities and suggest that both activities apparent number and diversity of both PTEN’s normal suppress proliferation and together suppress invasion. functions and of the effects of PTEN loss on tumour Oncogene (2010) 29, 687–697; doi:10.1038/onc.2009.384; development in different tissues (Chow and Baker, 2006; published online 16 November 2009 Suzuki et al., 2008), it seems quite possible that other mechanisms of action may account for some of PTEN’s Keywords: PTEN; PI3 kinase; phosphoinositide; cancer; functions and tumour suppressor activities in some TPIP; Akt tissues. Experiments to reveal the mechanisms and signifi- cance of PTEN’s separable activities have been greatly assisted by the use of functionally selective mutants. In Introduction particular, many studies have used PTEN G129E, a mutation that was identified in two Cowden disease PTEN is one of the most frequently lost tumour families (Liaw et al., 1997) and shows remarkable selec- suppressors in human cancers, which in culture and tivity, displaying greatly reduced lipid phosphatase activity while retaining full protein phosphatase activity in vitro (Furnari et al., 1998; Myers et al., 1998). In this Correspondence: Dr NR Leslie, Division of Molecular Physiology, University of Dundee, Dow Street, Dundee, Scotland DD1 5EH, UK. study we describe the systematic generation of a PTEN E-mail: [email protected] mutant with the converse specificity, that is, retaining 3Current address: Key Laboratory for Regenerative Medicine of the lipid phosphatase while lacking significant protein phos- Ministry of Education, Medical College, Jinan University, Guangzhou phatase activity. We then describe experiments using 510632, PR China. 4These authors contributed equally to this work. both mutants to reveal the contribution of these two Received 1 April 2009; revised 29 September 2009; accepted 6 October separable activities to PTEN’s activity in several cell- 2009; published online 16 November 2009 based assays of proliferation, invasion and migration. PTEN’s lipid and protein phosphatase activities L Davidson et al 688 Results tyrosine/glutamic acid polymer polyGluTyr-P. Second, despite its extremely close sequence similarity with The identification of a PTEN mutant with lipid but not TPIP, TPTE-reactivated displayed robust activity protein phosphatase activity against both the lipid PtdInsP3 and this peptide The proteins encoded in the that are substrate polyGluTyr-P. This identification of a most closely related to PTEN are TPIP and TPTE. PTEN-related phosphatase lacking protein phosphatase These are expressed almost exclusively in the testis activity suggests that protein phosphatase activity may (Walker et al., 2001; Tapparel et al., 2003), although have been selected for during evolution. The protein their physiological functions are unclear. Our previous phosphatase activity displayed by TPTE-reactivated has studies of the lipid phosphatase activity of TPIP and interesting implications for TPTE function, but given its TPTE revealed that TPIP has similar PtdInsP3 phos- close relationship to TPIP, suggests a systematic phatase activity to PTEN, and that while TPTE lacks approach for the mutation of PTEN to produce a detectable activity in the assays used, it could be mutant enzyme with only lipid phosphatase activity. reactivated through mutation, producing an active Therefore, at each of the 11 residues at which TPIP phosphatase termed ‘TPTE-reactivated’ (Walker et al., differs from TPTE and PTEN through the phosphatase 2001; Leslie et al., 2007). However, when the protein domain and N-terminal region of the C2 domain, we phosphatase activity of these proteins was investigated, mutated the PTEN sequence to the corresponding we obtained two surprising results (Figure 1). First, in amino acid in TPIP (See Supplementary Figure S1). contrast to PTEN, TPIP was found to lack detectable In an initial screen, these 11 mutants were expressed protein phosphatase activity against the phosphorylated in bacteria as GST–PTEN fusion proteins, purified

PTEN 121 AIHCKAGKGRTGVMICA TPIP 240 AIHCKGGKGRTGTMVCA TPTE 297 AIHCKGGTDRTGTMVCA TPTE React 297 AIHCKGGKGRTGTMVCA

PIP3 polyGluTyr-P 12 40 10 8 30

6 20 4 10 2 Phosphatase activity Phosphatase activity

(% substrate conversion) 0 0 (% substrate conversion) TPIP TPTE PTEN TPIP TPTE- React TPTE PTEN TPTE- React

dePIP3 fP-Poly(Glu,Tyr) 35 35

30 30 PTEN wt PTEN wt 25 Y138L 25 Y138L

20 20

15 15

PTEN PTEN 10 10 wt Y138L 5 5 % Substrate conversion % Substrate conversion 0 0

0 20 40 60 80 100 120 140 0 20 40 60 80 100 120 140 Time (min) Time (min) Figure 1 Both TPIP and PTEN Y138L have lipid but not protein phosphatase activity. (a) An alignment of the active site P-loop of PTEN, TPIP and TPTE is shown, with the PTEN catalytic cysteine residue boxed in red. The black boxing shows the threonine and aspartic acid residues at which TPTE differs from the active PTEN and TPIP in this region, and the mutation of these residues performed in reactivated TPTE (called TPTE-R) (Leslie et al., 2007). (b and c) PTEN and the cytosolic regions of TPIP and TPTE (wild-type and reactivated) were expressed as GST-fusion proteins and purified. The activity of the indicated proteins was 33 assayed using either P radiolabelled PtdIns(3,4,5)P3 (b) or phosphorylated polyGluTyr (c) as substrate. Data are shown as mean activity and s.d. from triplicate assays. (d) Wild-type and Y138L PTEN proteins were expressed in bacteria as GST-fusion proteins, affinity purified and the tag removed by proteolysis before analysis of these preparations by electrophoresis and staining with 33 Coomassie. (e and f) These wild-type and Y138L PTEN proteins were assayed against P radiolabelled PtdIns(3,4,5)P3 (e)or phosphorylated polyGluTyr (f) for the indicated times and activity shown as the mean activity and s.d. from triplicate assays.

Oncogene PTEN’s lipid and protein phosphatase activities L Davidson et al 689 and assayed against both PtdInsP3 and polyGluTyr-P PTEN phosphatase and C2 domains, interfering with in vitro. This screen identified one mutant, PTEN membrane binding (Vazquez et al., 2001; Das et al., 2003; Y138L, which displayed similar PtdInsP3 phosphatase Ning et al., 2006; Gil et al., 2007). We therefore tested the activity to wild-type PTEN, but lacked detectable effects of phosphorylation of PTEN with CK2, and also activity against polyGluTyr-P (Supplementary Figure with GSK3, on the activity of wild-type PTEN and S2). Further kinetic analysis of PTEN Y138L revealed PTEN Y138L against both PtdInsP3 and Ins(1,3,4,5)P4. 33 that the lipid phosphatase activity against 3- P PtdInsP3 These experiments showed, as might be predicted from is essentially indistinguishable from wild-type PTEN, previous experiments using mutant proteins (Ning et al., whereas activity against polyGluTyr-P is extremely 2006), that phosphorylation of bacterially expressed weak, being approximately 3% of wild-type activity, PTEN with CK2 caused a modest reduction in activity and frequently undetectable (Figures 1, 2, Supplemen- against PtdInsP3 andalargerincreaseinactivityagainst tary Figure S2 and S3). the soluble headgroup Ins(1,3,4,5)P4. However, phos- phorylated PTEN Y138L showed a similar decrease in PTEN Y138L has activity against lipid but not soluble PtdInsP3 phosphatase activity, but still lacked detectable substrates activity against Ins(1,3,4,5)P4 (Figure 2b). We therefore The activity of PTEN Y138L was tested against the tested the activity of wild-type and Y138L mutant PTEN soluble phosphate substrate, Ins(1,3,4,5)P (the proteins lacking the C-terminal tail due to a stop 4 codon at position 354. This showed that PTEN Y138L headgroup of PtdInsP3) and although the wild-type enzyme was strongly active, PTEN Y138L lacked 354stop still lacked activity against PolyGluTyr-P and detectable activity (Figures 2b and c). This result Ins(1,3,4,5)P4, while retaining lipid phosphatase activity suggested that the Y138L mutation does not alter the (Figure 2c). Therefore, the substrate specificity of PTEN precise conformation of the active site to accommodate Y138L may suggest that its active site can only be available for substrate binding when the enzyme interacts the headgroup of PtdInsP3, yet excluding phospho- tyrosine, but might alter the conformation of the with a lipid surface. phosphatase on a larger scale, favouring lipid substrates presented at a membrane interface over soluble sub- Defining the requirements for the lipid and protein strates. It seemed possible that the unstructured phosphatase activities of PTEN in cell based assays C-terminal tail of PTEN might be involved in this To analyse the effects of PTEN’s separable lipid and selective activity, as especially when phosphorylated, this protein phosphatase activities on cellular signalling and tail is believed to bind into the basic surfaces of the behaviour, we first used the well-characterized PTEN

120 No Kinase CK2 CK2 + GSK3 100

80 = PIP3 PIP3 = PolyGluTyrP 60 = IP4 40 120

20 Phosphatase activity 100 0 80 WT WT WT Y138L Y138L Y138L 60

300 40

250 20

200 Relative Phosphatase activity 0

150 IP4 WT 1-353 Y138L 100

50 1-353 Y138L Phosphatase activity

0 WT WT WT Y138L Y138L Y138L

Figure 2 PTEN Y138L lacks activity against polyGluTyrP and Ins(1,3,4,5)P4, and this substrate specificity is not affected by the C-terminal phosphorylated tail. (a and b) Purified tagless PTEN wild-type and Y138L proteins were phosphorylated with purified recombinant CK2 and GSK3 as indicated, before assaying against PtdInsP3 or Ins(1,3,4,5)P4. Activity data are represented as the mean activity relative to unphosphorylated wild-type þ s.d. from triplicate assays. (c) Mutant GST-PTEN proteins (wt and Y138L) lacking the unstructured C-terminal tail, termed PTEN 1–353, were expressed, purified and assayed against radiolabelled PtdInsP3, polyGluTyrP and Ins(1,3,4,5)P4. Activity data are represented as the mean activity relative to wild-type þ the range/2 from duplicate assays.

Oncogene PTEN’s lipid and protein phosphatase activities L Davidson et al 690 null glioblastoma cell line, U87MG. We expressed Y138L proteins were indistinguishable in their ability to PTEN in these cells using a lentiviral delivery system suppress Akt phosphorylation (Figures 3 and Supple- able to engineer expression of PTEN in almost all cells mentary Figure S4 (Maccario et al., 2007)). As part of in target populations, even when used at titres required this project, we also addressed the effects of PTEN to express PTEN at levels broadly equivalent to those activity on the phosphorylation of some of its proposed in several cells expressing endogenous PTEN. We re- substrates, FAK, PDGFR and SHC. However, we expressed wild-type PTEN and the three mutant found that these proteins appeared to have the same proteins, PTEN C124S (phosphatase dead), PTEN level of tyrosine phosphorylation in U87MG G129E (protein phosphatase only) and PTEN Y138L cells regardless of PTEN expression (Supplementary (as described, lipid phosphatase only) in U87MG cells to Figure S5). analyse their effects on cellular PtdInsP3 levels. These PTEN has been previously shown in cells cultured in data showed a strict correlation with the lipid phospha- soft agar to suppress proliferation and colony formation tase activity of the expressed proteins, as wild-type and (Furnari et al., 1997; Cheney et al., 1998; Davies et al., Y138L enzymes both reduced cellular PtdInsP3 by 1998) and that this ability appeared to require lipid approximately 80%, whereas C124S and G129E had phosphatase activity. Sham transduced U87MG cells no significant effect (Figure 3a). We also similarly and those expressing green fluorescent protein (GFP) analysed the effects of these PTEN proteins on the form colonies suspended in agarose, whereas expression phosphorylation of the PtdInsP3-dependent protein of wild-type PTEN caused an almost complete suppres- kinase Akt/PKB. In these experiments, expression of sion of colony formation. Phosphatase dead PTEN PTEN C124S or PTEN G129E had no detectable effect C124S and PTEN G129E, each had no inhibitory effect on Akt phosphorylation, whereas expression of either on colony formation, whereas PTEN Y138L inhibited wild-type PTEN or PTEN Y138L at approximately colony formation in a very similar manner to the wild- physiological levels greatly reduced the phosphorylation type enzyme (Figure 4). Over four experiments, the of Akt (Figures 3b and Supplementary Figure S4). On mean number of colonies formed by cells expressing conducting these experiments many times over a range PTEN Y138L was slightly smaller than the number of expression levels it was observed that wild-type and formed by cells expressing wild-type PTEN, but this

20 200000 18 180000 16 160000 14 140000 12 120000 10 content

3 8 100000

PIP 6 80000 (arbitray units) 4 60000 2 40000 0 Akt/PKB activity (dpm) 20000 PTEN 0

P-T308 Akt PTEN P-Ser473 Akt P-S473 Akt α pGSK3 S21 P-Ser21/9 GSK3 β pGSK3 S9

WT GAPDH Y138L C124S G129E WT Y138L C124S G129E Assay blank Untransduced

Figure 3 PTEN Y138L reduces cellular PtdInsP3 levels and Akt activity and phosphorylation with similar efficacy to wild-type PTEN. (a and b) PTEN null U87MG cells were transduced with lentiviruses encoding GFP, PTEN or mutant PTEN (C124S, G129E or Y138L) as indicated, and lysed after 24 h. Cellular PtdInsP3 content was measured using a TR-FRET sensor complex by comparison with a standard curve of known PtdInsP(3,4,5)3 concentrations as previously described (Gray et al., 2003). Data are presented as the mean PIP3 content þ s.d. in arbitrary units from quadruplicate samples. PTEN expression and Akt phosphorylation were investigated by western blotting of total cell lysates using total and phospho-specific antibodies, and Akt enzyme activity in immunoprecipitates was assayed in vitro against a peptide substrate. Data are represented as the mean dpm radiolabelled phosphate incorporated into the peptide substrate þ s.d. from triplicate samples. Experiments to measure PtdInsP3 content and Akt phosphorylation were performed on two and eight occasions, respectively with similar results.

Oncogene PTEN’s lipid and protein phosphatase activities L Davidson et al 691 700

600

500 Control WT PTEN G129E PTEN 400

300

200 Number of colonies

100

0

PTEN

GFP C124S PTEN Y138L PTEN GAPDH GFP PTEN wt PTEN Y138L PTEN C124S PTEN G129E Untransduced Figure 4 The effects of PTEN re-expression on U87MG cell growth in soft agar. (a and b) U87MG cells were transduced with recombinant lentiviruses encoding wild-type PTEN and PTEN point mutants, tryprinized 24 later, and then grown in 0.35% agarose for 21 days. Cells were fixed with 2% paraformaldehyde and the number of colonies was counted. In parallel, samples of transduced cells were seeded onto standard culture dishes and lysed after 24 h to determine the expression of PTEN by western blotting. (a) The numerical data shown are derived from four independent experiments, each performed in triplicate (±s.e.m.). Data were analysed by analysis of variance and Tukey’s pairwise comparison, showing that the data for PTEN wild-type and Y138L differ significantly from that for GFP, PTEN C124S and PTEN G129E, but there were no significant differences within these two groups. (b) An image of a representative data set is shown.

difference was not significant. These data strongly sufficient to block the invasion of these cells into the correlate the ability of PTEN to suppress colony surrounding matrigel, even when these two mutants formation in soft agar with its lipid phosphatase were expressed additively together in the same cells activity. (Figures 5a–g). Consistent with the efficacy of only wild- In vivo, the invasion of glial tumours frequently type PTEN in the regulation of this invasive cell proceeds along white matter tracts and blood vessels, morphology, when colony size was measured after following the endothelial cell basement membrane 7 days in three-dimensional culture, addressing the (Goldbrunner et al., 1998; de Bouard et al., 2002). We spread of cells into the surrounding matrix, again, only therefore established an assay for the proliferation and wild-type PTEN significantly inhibited colony spread invasion of cultured U87MG cells suspended in the (Figures 5h). Microscopy also revealed other consistent reconstituted basement membrane extract, Matrigel, differences in the morphology of these colonies, in with cell morphology being visualized by staining for particular, cells expressing PTEN G129E generally DNA and filamentous actin. In this assay, after 3 days in formed a more dense central mass of cells, with fewer culture, sham-transduced cells and those expressing cells invading deep into the surrounding matrix GFP or phosphatase dead PTEN C124S proliferated (Figures 5j–n). We also investigated the effects of PTEN to form colonies that spread through the surrounding expression on net cell proliferation in these cultures, matrix (Figure 5) in contrast to cells proliferating in using an LDH enzyme assay-based method (Figure 5i). agarose where they grow as nearly spherical clusters. The reproducible differences in colony morphology The expression of wild-type PTEN at near physiological made unbiased counting of stained nuclei challenging, levels suppressed the invasion of these cells, causing although similar results were obtained (data shown in them to grow initially as rather tightly packed balls Supplementary Figure S6) and Matrigel also appeared of cells, with few invading the surrounding matrix to interfere with several other proliferation assay frequently as strings of cells, in a manner different from methods. This work showed that both wild-type PTEN other cells in these experiments (Figure 5). Neither and PTEN Y138L were able to strongly suppress expression of PTEN G129E nor of PTEN Y138L was cellular proliferation, but with PTEN G129E mediating

Oncogene PTEN’s lipid and protein phosphatase activities L Davidson et al 692 No virus PTEN wt PTEN C124S PTEN G129E

PTEN Y138L PTEN G129E + Y138L GFP PTEN wt PTEN C124S PTEN G129E PTEN Y138L PTEN GE + YL HEK293

PTEN

GAPDH

1.8 1000 1.6 M) 900 μ 800 1.4 700 1.2 GFP 600 1 500 0.8 400 300 0.6 200 0.4 100

Cell number (LDH activity) 0.2 Mean colony diameter ( 0 0 WT GFP WT Y138L GFP C124S G129E Y138L No virus C124S G129E

PTEN wt PTEN C124SPTEN G129E PTEN Y138L

Figure 5 The effects of PTEN re-expression on U87MG cell proliferation and invasion in three-dimensional culture in matrigel. U87MG cells were transduced for 24 h with recombinant lentiviruses encoding wild-type PTEN and PTEN point mutants. Cells were trypsinized, then either seeded at low density in matrigel (a–f and h–n) or plated for 24 h further two-dimensional culture on plastic to verify expression of PTEN by western blotting alongside a sample of HEK293 cells expressing endogenous PTEN (g). (a–f) Cells were grown for 3 days, fixed with 2% paraformaldehyde and stained with rhodamine phalloidin and 4,6-diamidino-2-phenylindole (DAPI) to visualize filamentous actin and DNA, respectively. Samples were examined and photographed by widefield deconvolution microscopy (a–e) or confocal microscopy (f). Similar data were obtained by both methods. White scale bars represent 30 mM.(j–n). Cells were grown for 5 days, fixed with 2% paraformaldehyde, stained with DAPI and examined by confocal microscopy. Numerical data shown in (h) represent the mean colony size of sets of 5 day colonies measured as the longest observable axis of colonies photographed at low magnification by phase contrast microscopy. An assessment of cell number shown in (i) was measured by an LDH enzyme assay of matrix/cell extracts after 10 days growth and is shown as mean þ s.e.m.

Oncogene PTEN’s lipid and protein phosphatase activities L Davidson et al 693 a weaker but still significant effect (Figures 5i and or protein phosphatase activity are sufficient to mediate Supplementary Figure S6). Overexpression of either a similar effect on cell migration in this assay (Figure 6). PTEN wild type or PTEN Y138L was able to cause an Interestingly in these experiments, migration of cells almost complete block in colony formation, apparently expressing wild-type PTEN was clearly observed, and through cell death, but this effect was not evident at this migration rate over the period of the experiment physiological expression levels (data not shown). appeared little different with cells expressing either We also investigated the described effects of PTEN re- PTEN G129E or PTEN Y138L, suggesting that, at least expression in U87MG cells on cell spreading and using this assay format, the inhibitory effects of PTEN’s migration (Tamura et al., 1998). For cell spreading activities against lipid and soluble substrates may not be assays, cells transduced with lentiviruses encoding additive. PTEN were trypsinized thoroughly before allowing We went on to analyse the effects of PTEN Y138L on attachment and spreading on culture substrates. More- cell migration in vivo, using a previously established over, as described, wild-type PTEN caused a delay in the assay for mesoderm cell migration away from the ability of cells to spread. In this assay, the ability to primitive streak in the early chick embryo (Yang et al., inhibit spreading correlated with PTEN’s lipid phos- 2002). In our previous studies, we found that the protein phatase activity, as PTEN Y138L was almost as efficient phosphatase activity of PTEN appeared to contribute to as the wild-type enzyme, but PTEN C124S and PTEN a block on the epithelial to mesenchymal transition that G129E had little or no effect (Figure 6). Cell migration these cells undergo before they exit the primitive streak, assays from a confluent monolayer into a 500 mm whereas the lipid phosphatase activity of PTEN had no defined cell-free gap showed that PTEN retards the rate effect on this epithelial to mesenchymal transition, but of migration into this gap. As has been previously interfered with the directed nature of the subsequent described, PTEN G129E was also able to inhibit migration of these cells through the embryo (Leslie glioblastoma cell migration (Tamura et al., 1998; et al., 2007). In this assay, in agreement with these Raftopoulou et al., 2004), but interesting, this ability previous conclusions, expression of GFP-PTEN Y138L was shared by PTEN Y138L, suggesting that either lipid did not affect primitive streak cell epithelial to

100%

80% = spread

60% = intermediate = round GFP PTEN wt PTEN C124S 40% % of cells

160 20% 140 120 100 0% 80 60

PTEN in wound 40 cell number 20 0 WT

GFP PTEN G129E PTEN Y138L WT Y138L C124S GFP G129E Y138L C124S G129E PTEN PTEN

0 Hr 20 Hr 0 Hr 20 Hr 0 Hr 20 Hr

1mm GFP GFP + LY PTEN Y138L

Figure 6 The regulation of cell spreading and cell migration by PTEN. (a) Trypsinized U87MG cells expressing GFP or different PTEN proteins were allowed to adhere onto collagen IV-coated culture slides for 30 min before fixation. Data are presented as the mean fractions of each cell population blind scored as spread, intermediate or round, þ s.e.m. from five randomly selected fields from a representative experiment. This experiment has been repeated twice on collagen and twice on plastic with similar results. Equal expression of the PTEN mutants was verified by western blotting and the same set of transduced cells was used in the following experiments addressing adherent cell migration. (b–f) Trypsinized U87MG cells expressing GFP or different PTEN proteins were seeded at confluence and allowed to adhere to matrigel-treated plastic before the culture inserts were removed to leave a defined 500 mm cell-free gap. Cells were then allowed to migrate over a 24 h period, before cells were fixed and photographed. (g) Quantitation represents the mean number of cells in a central area of the gap from six randomly selected fields±s.e.m. This experiment was performed three times with similar results. (h–j) PTEN Y138L interferes with directed cell migration in the developing chick embryo, but does not block cell motility. Cell migration assays were performed using cells expressing GFP or GFP-PTEN Y138L as previously described (Yang et al., 2002; Leslie et al., 2007).

Oncogene PTEN’s lipid and protein phosphatase activities L Davidson et al 694 mesenchymal transition, but interfered with the directed conformation and also fits well with data regarding the cell migration of cells through the embryo (Figures 6i-k). interfacial activation of PTEN on acidic membrane Finally, attempting to identify effects of PTEN on well- surfaces rich in PtdIns(4,5)P2 (Campbell et al., 2003; characterized signalling pathways, we used a panel of Das et al., 2003; McConnachie et al., 2003; Iijima et al., phospho-specific antibodies to investigate adherent 2004; Walker et al., 2004; Redfern et al., 2008). It also U87MG cell cultures expressing wild-type and mutant seems possible that only one active conformation exists PTEN proteins at approximately two- to threefold that is greatly stabilized by membrane surface interac- above the level seen in HEK293 cells (Supplementary tion. However, in this latter case it seems to us harder to Figure S8). These experiments showed robust effects of explain why if PTEN Y138L was less stable in a single either wild-type or Y138L PTEN expression, suppres- active conformation in solution, it would not also have sing Akt phosphorylation and also increasing IRS1 lower lipid phosphatase activity on a membrane surface. levels, as has been previously described in these cells It may also be relevant to these ideas that our expression (Lackey et al., 2007). However, other effects, and any experiments lead us to believe that the PTEN Y138L apparently mediated by PTEN G129E implicating protein could be generally slightly less stable than wild- protein phosphatase activity, were not observed. type PTEN, as we consistently required more plasmid or viral expression vector (usually two- to threefold) to achieve the same level of protein expression, and PTEN Y138L was slightly less stable to thermal denaturation Discussion than the wild-type enzyme (Supplementary Figure S7). Second, the similarity between wild-type PTEN and Given the significance of PTEN as a tumour suppressor PTEN Y138L in terms of their PtdInsP3 phosphatase and as a regulator of many physiological processes, it is activity in vitro and of their effects on both cellular important that we determine its mechanisms of action in PtdInsP3 levels and Akt/PKB phosphorylation strongly these specific functions. This is particularly true in suggests that the protein phosphatase activity of PTEN tumour suppression, as intense efforts are going into the is not required for PTEN to efficiently metabolize development of PI3K inhibitors for the treatment of cellular PtdInsP3. As autodephosphorylation has been cancer, with several currently in clinical trials. These proposed as a mechanism regulating PTEN conforma- agents are likely to be given to patients with tumours, in tion and activity that may be important in some which PTEN function is believed to be compromised, settings (Raftopoulou et al., 2004), this seems potentially although as yet, in many tumour types, it is still possible significant. that PI3K-independent pathways may be of great Finally, we have used the two functionally selective importance. In this study we have developed a PTEN mutants, PTEN G129E and Y138L, to address the mutant, PTEN Y138L, that retains wild-type PtdInsP3 relative importance of the lipid and protein phosphatase lipid phosphatase activity, but in contrast to the wild- activities of PTEN in the regulation of different cellular type enzyme, PTEN Y138L lacks activity against a processes, when expressed at physiological levels in model peptide substrate, polyGluTyrP and the soluble PTEN null U87MG cells. This combination of these two PtdInsP3 headgroup, Ins(1,3,4,5)P4. We have then used mutants has allowed us to identify not only processes this mutant to reveal several significant aspects to PTEN that can be affected by PTEN’s protein phosphatase function. activity, but also to identify processes that appear to The activity of the PTEN-related phosphatase, TPIP, require both activities acting together. This work has towards lipids but not the tested peptide substrate identified examples of process that appear to be indicates that the protein phosphatase activity of PTEN inhibited solely or largely by the lipid phosphatase is not a necessary but inconsequential by-product of a activity (proliferation in soft agar and cell spreading), PTP phosphatase domain with a large basic active site processes that can be significantly inhibited by either capable of accommodating PtdInsP3. Instead, this activity (proliferation and adherent cell migration) and suggests a functional requirement for PTEN’s protein those that require both activities in concert for efficient phosphatase activity. Similarly, the selective activity of inhibition (invasion in Matrigel). Also, together with PTEN Y138L against the lipid PtdInsP3, but not the previously published experiments (Leslie et al., 2007), identical soluble headgroup, Ins(1,3,4,5)P4, strongly although not in U87MG cells, we show that the suggests to us that the active site of PTEN Y138L is epithelial to mesenchymal transition of cells in the chick only available for substrate binding when the enzyme embryo primitive streak is inhibited solely by the protein interacts with a lipid surface. It is possible that this phosphatase activity of the expressed PTEN. conformational regulation does not apply to wild-type We find that the ability of PTEN to inhibit colony PTEN, but we feel this is unlikely, and that a more formation in a classic soft agar assay correlates closely probable case is that PTEN can exist in at least three with its lipid phosphatase activity. However, using a conformations: a ‘lipid active’ conformation in contact novel matrigel invasion/proliferation assay, we show with membrane surfaces, a ‘soluble active’ conformation that both the lipid and protein phosphatase activities in solution, that has evolved to use protein substrates, contribute significantly to the inhibition of both invasive and an inactive conformation (Leslie et al., 2008). This morphology and proliferation in this setting. Although hypothetical model would require that PTEN Y138L it has been well established that the protein phosphatase and TPIP do not stably take up the ‘soluble active’ activity of PTEN can mediate a suppression of glial cell

Oncogene PTEN’s lipid and protein phosphatase activities L Davidson et al 695 migration (Tamura et al., 1998; Park et al., 2002; Materials and methods Raftopoulou et al., 2004), effects on proliferation have not been previously described. This is perhaps because Cell culture these effects are not evident in assays such as the U87MG cells were originally purchased from the ECACC, and classical soft agar assays and because they appear much cultured in minimum essential medium and 10% fetal bovine weaker than those mediated by the lipid phosphatase serum (Sigma, Dorset, UK). 293T cells were kindly provided by Michelle West and Colin Watts (Division of Cell Biology and activity. Indeed, the contrasting results obtained from Immunology, University of Dundee) and grown in Dulbecco’s cells in soft agar/serum and those in Matrigel agree well modified Eagle medium, 10% fetal bovine serum. PTEN proteins with previous data implicating integrin signalling in the were expressed in U87MG cells using a lentiviral expression stimulation of signalling pathways inhibited by the system developed in Mary Collins’ laboratory (Ikeda et al., 2003). protein phosphatase activity of PTEN (Tamura et al., Non-replicative lentiviruses were produced by triple transfection 1998; Dey et al., 2008). Several proteins have been pro- of 1.5 million 293T cells in a 60-mm dish with 1.5 mgpHR-SIN- posed as potential substrates for PTEN, including FAK, based shuttle vectors, 1 mg pHR-CMV 8.2 deltaR packaging SHC, PDGF-R, b-catenin, 5-HT2c-R and SMAD3 vector and 1 mg pCMV VSV-G envelope vector using Trans-IT (Tamura et al., 1998; Gu et al., 1999; Mahimainathan transfection reagent (Mirus, Madison, WI, USA). Transfected and Choudhury, 2004; Hjelmeland et al., 2005; cells were maintained in category 2 containment following UK home office and local guidelines. Media were changed after Vogelmann et al., 2005; Ji et al., 2006), and several of 16 h and after a further 48 h, the 4 ml of viral supernatant was these would seem plausible mediators of effects on glial harvested, centrifuged for 5 min at 100  g to pellet any cells and cell migration, invasion and proliferation. However, to frozen at À80 1C in aliquots. Batches of viral supernatants were date none of these have been validated with great cross-compared by transduction of U87MG cells supplemented confidence, and our experiments failed to find support with 8 mg/ml polybrene and western blotting for their efficiency for these proteins as substrates. It will require significant in driving PTEN expression. Initial experiments transducing effort to investigate these possibilities conclusively and it U87MG cells with simultaneously prepared virus preparations seems most likely to us that significant substrates for encoding GFP or PTEN indicated that even at low levels of PTEN’s protein phosphatase activity remain uniden- virus at which resultant PTEN expression was approximately tified. The use of functionally selective mutants has that of several cultured cell lines (around 1% culture volume), expression of GFP appeared relatively homogeneous and evident identified biological processes that appear to be con- in at least 95% of the target cell populations. trolled independently by PTEN’s protein phosphatase activity (adherent cell migration) or require both activities (invasion in matrigel). These tools should Cell-based assays also assist in the difficult challenges of investigating For three-dimensional matrigel cultures, 50 ml of matrigel was these mechanisms and identifying the substrates res- spread on a 16 mm diameter coverslip in the well of a 12-well ponsible. plate and allowed to form a gel. 2000 U87MG cells were Although a clear picture of the significance of PTEN seeded onto the matrigel in 50 ml of growth medium and as a regulator of cellular inositol phosphates is yet to allowed to attach for 2 h. The medium was then removed and cells overlaid with a further 50 ml of matrigel. Once the second emerge (Orchiston et al., 2004; Deleu et al., 2006), the layer of matrigel had gelled, the culture was overlaid with 1 ml lack of activity of either PTEN G129E or PTEN Y138L of growth medium and grown for 3, 7 or 10 days. The medium against Ins(1,3,4,5)P4 would argue against these mole- was removed, cultures washed twice with phosphate-buffered cules as significant substrates, at least in mediating the saline (PBS), and cells fixed in 2% paraformaldehyde in PBS effects we observe here on proliferation and migration. for 30 min at room temperature. Cells were then washed twice In the crystal structure of PTEN, tyrosine 138 is located with PBS and permeabilized with TBS-T (25 mM Tris–HCl, in a pocket of the phosphatase domain, well away from 150 mM NaCl, 0.1% Tween-20, pH 7.5) for 10 min. Cells were the active site, giving few clues to explain the substrate stained with rhodamine phalloidin (10 nM) and 4,6-diamidino- selectivity of the Y138L mutant (Lee et al., 1999). To 2-phenylindole (1 mg/ml) for 10 min, washed twice with TBS-T, our knowledge, only one missense mutation has been once with TBS and mounted using Vectashield (Vector Labora- tories, Peterborough, UK). Cells were examined using either a described targeting tyrosine 138 (Y138C in a small-cell SP2 Leica confocal microscope or an Applied Precision DV3 lung cancer case) (http://www.sanger.ac.uk/genetics/ widefield deconvolution microscope (University of Dundee CLS CGP/cosmic). CHIPs microscopy facility) equipped with a  20 0.7 NA oil Our work provides a novel tool that in the future immersion lens. Two-dimensional projections were created and should allow the determination of the significance of nuclei counted using Volocity imaging software (Improvision, PTEN’s lipid and protein phosphatase activities in many Perkin Elmer, Beaconsfield, Bucks, UK). Data were analysed by assay systems. The data also show that when re- analysis of variance, and Tukey’s pairwise comparison was used expressed in PTEN null glioblastoma cells at a to identify groups that were significantly different. physiological expression level and these cells are grown Cell proliferation was also measured after 10 days growth in in Matrigel, both the lipid and protein phosphatase matrigel using an LDH activity assay kit (Roche, Burgess Hill, West Sussex, UK). Three-dimensional matrigel cultures were activities of PTEN independently contribute an inhibi- washed twice in PBS, scraped into Eppendorf tubes and lysed tion of cellular proliferation and that both activities are in 200 ml of PBS containing 1% Triton X-100 and complete required in concert to achieve efficient inhibition of protease inhibitors on ice for 30 min. Samples were then invasion. Considering this, it seems possible that the centrifuged for 5 min at 5000  g at 4 1C to pellet the matrigel importance of PTEN’s protein phosphatase activity may and the supernatant was recovered into a fresh tube. Five have been underestimated. microlitres of each sample was added to 50 ml of assay reagent

Oncogene PTEN’s lipid and protein phosphatase activities L Davidson et al 696 in the wells of 96-well plate, incubated for 1 h at room temper- Acknowledgements ature in the dark, and the absorbance measured at 490 nm using a spectrophotometer. Additional methods are described We thank Sam Swift and Paul Appleton from the Dundee in the Supplementary Information. Light Microscopy Facility for their help with image acquisition and analysis. We thank Hilary McLauchlan, James Hastie and their staff in the DSTT (University of Dundee) for provision of purified antibodies and recombinant protein kinases and Conflict of interest Steven Hubbard for advice regarding data analysis. NL is an RCUK Academic Fellow. Work in the Inositol Lipid The research funding of the Inositol Lipid Signalling Signalling laboratory was funded by the Medical Research laboratory includes a research grant from a pharmaceutical Council, the Association for International Cancer Research consortium (Astra Zeneca, Boehringer Ingelheim, Glaxo and the pharmaceutical companies of the DSTT consortium SmithKline, Merck Serono and Pfizer). These companies have (Astra Zeneca, Boehringer Ingelheim, GlaxoSmithKline, not been directly involved in, or influenced, this work. Merck KGaA and Pfizer).

References

Campbell RB, Liu F, Ross AH. (2003). Allosteric activation of PTEN Gu J, Tamura M, Pankov R, Danen EH, Takino T, Matsumoto K phosphatase by phosphatidylinositol 4,5-bisphosphate. J Biol Chem et al. (1999). Shc and FAK differentially regulate cell motility and 278: 33617–33620. directionality modulated by PTEN. J Cell Biol 146: 389–404. Cheney IW, Johnson DE, Vaillancourt MT, Avanzini J, Morimoto A, Hjelmeland AB, Hjelmeland MD, Shi Q, Hart JL, Bigner DD, Demers GW et al. (1998). Suppression of tumorigenicity of Wang XF et al. (2005). Loss of phosphatase and tensin homologue glioblastoma cells by adenovirus-mediated MMAC1/PTEN increases transforming growth factor beta-mediated invasion transfer. Cancer Res 58: 2331–2334. with enhanced SMAD3 transcriptional activity. Cancer Res 65: Chow LM, Baker SJ. (2006). PTEN function in normal and neoplastic 11276–11281. growth. Cancer Lett 241: 184–196. Iijima M, Huang YE, Luo HR, Vazquez F, Devreotes PN. (2004). Das S, Dixon JE, Cho W. (2003). Membrane-binding and activation Novel mechanism of PTEN regulation by its phosphatidylinositol mechanism of PTEN. Proc Natl Acad Sci USA 100: 7491–7496. 4,5-bisphosphate binding motif is critical for chemotaxis. J Biol Davies MA, Lu Y, Sano T, Fang X, Tang P, LaPushin R et al. (1998). Chem 279: 16606–16613. Adenoviral transgene expression of MMAC/PTEN in human Ikeda Y, Takeuchi Y, Martin F, Cosset FL, Mitrophanous K, Collins glioma cells inhibits Akt activation and induces anoikis. Cancer M. (2003). Continuous high-titer HIV-1 vector production. Nat Res 58: 5285–5290. Biotechnol 21: 569–572. de Bouard S, Christov C, Guillamo JS, Kassar-Duchossoy L, Palfi S, Ji SP, Zhang Y, Van Cleemput J, Jiang W, Liao M, Li L et al. (2006). Leguerinel C et al. (2002). Invasion of human glioma biopsy Disruption of PTEN coupling with 5-HT2C receptors suppresses specimens in cultures of rodent brain slices: a quantitative analysis. behavioral responses induced by drugs of abuse. Nat Med 12: J Neurosurg 97: 169–176. 324–329. Deleu S, Choi K, Pesesse X, Cho J, Sulis ML, Parsons R et al. (2006). Keniry M, Parsons R. (2008). The role of PTEN signaling per- Physiological levels of PTEN control the size of the cellular turbations in cancer and in targeted therapy. Oncogene 27: Ins(1,3,4,5,6)P(5) pool. Cell Signal 18: 488–498. 5477–5485. Dey N, Crosswell HE, De P, Parsons R, Peng Q, Su JD et al. (2008). Lackey J, Barnett J, Davidson L, Batty IH, Leslie NR, Downes CP. The protein phosphatase activity of PTEN regulates SRC family (2007). Loss of PTEN selectively desensitizes upstream IGF1 and kinases and controls glioma migration. Cancer Res 68: 1862–1871. insulin signaling. Oncogene 26: 7132–7142. Freeman DJ, Li AG, Wei G, Li HH, Kertesz N, Lesche R et al. (2003). Lee JO, Yang H, Georgescu MM, Di Cristofano A, Maehama T, Shi PTEN tumor suppressor regulates p53 protein levels and activity Y et al. (1999). Crystal structure of the PTEN tumor suppressor: through phosphatase-dependent and -independent mechanisms. implications for its phosphoinositide phosphatase activity and Cancer Cell 3: 117–130. membrane association. Cell 99: 323–334. Furnari FB, Huang HJ, Cavenee WK. (1998). The phosphoinositol Leslie NR, Batty IH, Maccario H, Davidson L, Downes CP. (2008). phosphatase activity of PTEN mediates a serum-sensitive G1 Understanding PTEN regulation: PIP2, polarity and protein growth arrest in glioma cells. Cancer Res 58: 5002–5008. stability. Oncogene 27: 5464–5476. Furnari FB, Lin H, Huang HS, Cavenee WK. (1997). Growth Leslie NR, Yang X, Downes CP, Weijer CJ. (2007). PtdIns(3,4,5)P(3)- suppression of glioma cells by PTEN requires a functional dependent and -independent roles for PTEN in the control of cell phosphatase catalytic domain. Proc Natl Acad Sci USA 94: migration. Curr Biol 17: 115–125. 12479–12484. Liaw D, Marsh DJ, Li J, Dahia PL, Wang SI, Zheng Z et al. (1997). Gil A, Andres-Pons A, Pulido R. (2007). Nuclear PTEN: a tale of Germline mutations of the PTEN gene in Cowden disease, many tails. Cell Death Differ 14: 395–399. an inherited breast and thyroid cancer syndrome. Nat Genet 16: Gildea JJ, Herlevsen M, Harding MA, Gulding KM, Moskaluk CA, 64–67. Frierson HF et al. (2004). PTEN can inhibit in vitro organotypic Maccario H, Perera NM, Davidson L, Downes CP, Leslie NR. (2007). and in vivo orthotopic invasion of human bladder cancer cells PTEN is destabilized by phosphorylation on Thr366. Biochem J even in the absence of its lipid phosphatase activity. Oncogene 23: 405: 439–444. 6788–6797. Mahimainathan L, Choudhury GG. (2004). Inactivation of platelet- Goldbrunner RH, Bernstein JJ, Tonn JC. (1998). ECM-mediated derived growth factor receptor by the tumor suppressor PTEN glioma cell invasion. Microsc Res Tech 43: 250–257. provides a novel mechanism of action of the phosphatase. J Biol Gray A, Olsson H, Batty IH, Priganica L, Peter Downes C. (2003). Chem 279: 15258–15268. Nonradioactive methods for the assay of phosphoinositide 3-kinases McConnachie G, Pass I, Walker SM, Downes CP. (2003). Interfacial and phosphoinositide phosphatases and selective detection kinetic analysis of the tumour suppressor phosphatase, PTEN: of signaling lipids in cell and tissue extracts. Anal Biochem 313: evidence for activation by anionic phospholipids. Biochem J 371: 234–245. 947–955.

Oncogene PTEN’s lipid and protein phosphatase activities L Davidson et al 697 Myers MP, Pass I, Batty IH, Van der Kaay J, Stolarov JP, Hemmings Sansal I, Sellers WR. (2004). The biology and clinical relevance BA et al. (1998). The lipid phosphatase activity of PTEN is critical of the PTEN tumor suppressor pathway. J Clin Oncol 22: for its tumor supressor function. Proc Natl Acad Sci USA 95: 2954–2963. 13513–13518. Shen WH, Balajee AS, Wang J, Wu H, Eng C, Pandolfi PP et al. Myers MP, Stolarov JP, Eng C, Li J, Wang SI, Wigler MH et al. (2007). Essential role for nuclear PTEN in maintaining chromoso- (1997). P-TEN, the tumor suppressor from human mal integrity. Cell 128: 157–170. 10q23, is a dual-specificity phosphatase. Proc Natl Acad Sci USA Suzuki A, Nakano T, Mak TW, Sasaki T. (2008). Portrait of PTEN: 94: 9052–9057. messages from mutant mice. Cancer Sci 99: 209–213. Ning K, Miller LC, Laidlaw HA, Burgess LA, Perera NM, Downes CP Tamura M, Gu J, Matsumoto K, Aota S, Parsons R, Yamada KM. et al. (2006). A novel leptin signalling pathway via PTEN inhibition (1998). Inhibition of cell migration, spreading, and focal adhesions in hypothalamic cell lines and pancreatic beta-cells. EMBO J 25: by tumor suppressor PTEN. Science 280: 1614–1617. 2377–2387. Tapparel C, Reymond A, Girardet C, Guillou L, Lyle R, Lamon C Okumura K, Zhao M, Depinho RA, Furnari FB, Cavenee WK. et al. (2003). The TPTE gene family: cellular expression, subcellular (2005). Cellular transformation by the MSP58 oncogene is inhibited localization and alternative splicing. Gene 323: 189–199. by its physical interaction with the PTEN tumor suppressor. Proc Trotman LC, Wang X, Alimonti A, Chen Z, Teruya-Feldstein J, Yang Natl Acad Sci USA 102: 2703–2706. H et al. (2007). Ubiquitination regulates PTEN nuclear import and Orchiston EA, Bennett D, Leslie NR, Clarke RG, Winward L, tumor suppression. Cell 128: 141–156. Downes CP et al. (2004). PTEN M-CBR3, a versatile and selective Vazquez F, Grossman SR, Takahashi Y, Rokas MV, Nakamura N, regulator of inositol 1,3,4,5,6-pentakisphosphate (Ins(1,3,4,5,6)P5). Sellers WR. (2001). Phosphorylation of the PTEN tail acts as an Evidence for Ins(1,3,4,5,6)P5 as a proliferative signal. J Biol Chem inhibitory switch by preventing its recruitment into a protein 279: 1116–1122. complex. J Biol Chem 276: 48627–48630. Park MJ, Kim MS, Park IC, Kang HS, Yoo H, Park SH et al. (2002). Vogelmann R, Nguyen-Tat MD, Giehl K, Adler G, Wedlich D, PTEN suppresses hyaluronic acid-induced matrix metalloprotei- Menke A. (2005). TGFbeta-induced downregulation of E-cadherin- nase-9 expression in U87MG glioblastoma cells through focal based cell-cell adhesion depends on PI3-kinase and PTEN. J Cell adhesion kinase dephosphorylation. Cancer Res 62: 6318–6322. Sci 118: 4901–4912. Raftopoulou M, Etienne-Manneville S, Self A, Nicholls S, Hall A. Walker SM, Downes CP, Leslie NR. (2001). TPIP: a novel (2004). Regulation of cell migration by the C2 domain of the tumor phosphoinositide 3-phosphatase. Biochem J 360: 277–283. suppressor PTEN. Science 303: 1179–1181. Walker SM, Leslie NR, Perera NM, Batty IH, Downes CP. (2004). Redfern RE, Redfern D, Furgason ML, Munson M, Ross AH, The tumour-suppressor function of PTEN requires an N-terminal Gericke A. (2008). PTEN phosphatase selectively binds phosphoi- lipid-binding motif. Biochem J 379: 301–307. nositides and undergoes structural changes. Biochemistry 47: Yang X, Dormann D, Munsterberg AE, Weijer CJ. (2002). Cell 2162–2171. movement patterns during gastrulation in the chick are controlled Salmena L, Carracedo A, Pandolfi PP. (2008). Tenets of PTEN tumor by positive and negative chemotaxis mediated by FGF4 and FGF8. suppression. Cell 133: 403–414. Dev Cell 3: 425–437.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene