ARTICLE IN PRESS. Building 22, Room 3342, Silver Spring, MD , USA

Transcription

1 CLS-06477; No of Pages 14 ARTICLE IN PRESS + MODEL Cellular Signalling xx (2007) xxx xxx Loss of PTEN expression does not contribute to PDK-1 activity and PKC activation-loop phosphorylation in Jurkat leukaemic T cells Michael Freeley a, Jongsun Park b, Keum-Jin Yang b, Ronald L. Wange c,d, Yuri Volkov a, Dermot Kelleher a, Aideen Long a, a Department of Clinical Medicine, Institute of Molecular Medicine, Trinity College, Dublin, Ireland b Cell Signaling Laboratory, Cancer Research Institute, Department of Pharmacology, College of Medicine, Chungnam National University, 6 Munhwa-dong, Jung-gu, Taejon, , South Korea c Laboratory of Cellular and Molecular Biology, National Institute on Aging, National Institutes of Health, Baltimore, MD 21224, USA d Division of Metabolism and Endocrinology Products, Food & Drug Administration, New Hampshire Ave, Building 22, Room 3342, Silver Spring, MD , USA Received 21 February 2007; received in revised form 13 July 2007; accepted 23 July 2007 Abstract Unopposed PI3-kinase activity and 3 -phosphoinositide production in Jurkat T cells, due to a mutation in the PTEN tumour suppressor protein, results in deregulation of PH domain-containing proteins including the serine/threonine kinase PKB/Akt. In Jurkat cells, PKB/Akt is constitutively active and phosphorylated at the activation-loop residue (Thr308). 3 -phosphoinositide-dependent protein kinase-1 (PDK-1), an enzyme that also contains a PH domain, is thought to catalyse Thr308 phosphorylation of PKB/Akt in addition to other kinase families such as PKC isoforms. It is unknown however if the loss of PTEN in Jurkat cells also results in unregulated PDK-1 activity and whether such loss impacts on activation-loop phosphorylation of other putative PDK-1 substrates such as PKC. In this study we have addressed if loss of PTEN in Jurkat T cells affects PDK-1 catalytic activity and intracellular localisation. We demonstrate that reducing the level of 3 -phosphoinositides in Jurkat cells with pharmacological inhibitors of PI3-kinase or expression of PTEN does not affect PDK-1 activity, Ser241 phosphorylation or intracellular localisation. In support of this finding, we show that the levels of PKC activation-loop phosphorylation are unaffected by reductions in the levels of 3 -phosphoinositides. Instead, the dephosphorylation that occurs on PKB/Akt at Thr308 following reductions in 3 -phosphoinositides is dependent on PP2A-like phosphatase activity. Our finding that PDK-1 functions independently of 3 -phosphoinositides in T cells is also confirmed by studies in HuT-78 T cells, a PTEN-expressing cell line with undetectable levels of 3 -phosphoinositides. We conclude therefore that loss of PTEN expression in Jurkat T cells does not impact on the PDK-1/PKC pathway and that only a subset of kinases, such as PKB/Akt, are perturbed as a consequence PTEN loss Elsevier Inc. All rights reserved. Keywords: PTEN; PDK-1; PKB/Akt; PKC; PP2A; Phosphorylation Abbreviations: PtdIns(3,4)P 2, phosphatidylinositol-(3,4)-biphosphate; PtdIns(3,4,5)P 3, phosphatidylinositol-(3,4,5)-triphosphate; PtdIns(4,5)P 2, phosphatidylinositol-(4,5)-biphosphate; PLC, phospholipase C; PH, pleckstrin homology; PI3-K, phosphoinositide 3-kinase; PTEN, phosphatase and tensin homolog deleted on chromosome 10; PDK-1, 3 -phosphoinositide-dependent protein kinase-1; PKA, Protein kinase A; PKB, protein kinase B; PKC, protein kinase C; PP2A, protein phosphatase 2A; PBL, peripheral blood lymphocyte; DAG, diacylglycerol; ELISA, enzyme-linked immunosorbant assay; ERK, extracellular signal-related kinase; LFA-1, lymphocyte function-associated antigen-1; p90 RSK, p90 ribosomal S6 kinase; p70 S6 kinase, p70 ribosomal S6 kinase; SGK, serum and glucocorticoid kinase; Sos, son of sevenless; TCR, T cell receptor; ZAP-70, zeta-chain associated protein kinase of 70 kda; MAPK, mitogen-activated protein kinase; MEK, MAPK kinase; NF-κB, nuclear factor of κb; IKK, IκB kinase; GFP, green fluorescent protein; sirna, small interfering RNA; HRP, horseradish peroxidase; Itk, inducible T cell kinase; Hsp-70, heat shock protein-70, STRAP, serine-threonine kinase receptor-associated protein; PVDF, polyvinylidene difluoride; FBS, foetal bovine serum; AKAP, A-kinase anchoring protein; PMA, phorbol 12-myristate 13-acetate; DMSO, dimethyl sulfoxide. Corresponding author. Tel.: ; fax: addresses: freeleym@tcd.ie (M. Freeley), insulin@cnu.ac.kr (J. Park), Ronald.Wange@FDA.HHS.GOV (R.L. Wange), yvolkov@tcd.ie (Y. Volkov), dermot.kelleher@tcd.ie (D. Kelleher), longai@tcd.ie (A. Long) /$ - see front matter 2007 Elsevier Inc. All rights reserved. doi: /j.cellsig

2 2 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx 1. Introduction Phosphoinositide 3-kinase (PI3-kinase) is a family of enzymes that play important roles in cellular proliferation, survival, adhesion, cytoskeletal reorganisation and motility [1,2]. PI3-kinases phosphorylate the 3 position of the inositol ring of inositol phospholipids, producing 3 -phosphoinositides such as phosphatidylinositol-(3,4)-biphosphate (PtdIns(3,4)P 2 ) and phosphatidylinositol-(3,4,5)-triphosphate (PtdIns(3,4,5)P 3 ). PtdIns(3,4)P 2 and PtdIns(3,4,5)P 3 promote the membrane recruitment and/or activation of a number of proteins that contain pleckstrin homology domains (PH domains), including the serine/threonine kinase Protein Kinase B (PKB: also known as Akt) [3,4]. Following PI3-kinase activation, PKB/Akt translocates from the cytosol to the plasma membrane where it binds PtdIns(3,4)P 2 /PtdIns(3,4,5)P 3 via its N-terminal PH domain. The interaction of 3 -phosphoinositides with the PH domain of PKB/Akt also induces a conformational change on the enzyme that facilitates phosphorylation of PKB/Akt on two key residues [5]. These residues are Thr308, which is located on the kinase domain and is known as the activation-loop site, and Ser473, which is located outside the kinase domain and is known as the C-terminal hydrophobic-motif site. Strong evidence suggests that Thr308 phosphorylation is catalysed by 3 -phosphoinositide-dependent protein kinase-1 (PDK-1) [5,6], whilst the mechanism of Ser473 phosphorylation is not fully understood at present, with many enzymes (including PKB/Akt itself) identified as the hydrophobic-motif kinase(s) [7]. Following Thr308/Ser473 phosphorylation, activated PKB/Akt detaches from the plasma membrane, translocates to the cytosol and nucleus and phosphorylates substrates that contribute to cell proliferation and survival [3,4]. PDK-1 is known as a master kinase, because in addition to PKB/Akt, this enzyme has been also shown to phosphorylate the conserved activation-loop residue of many other protein kinase families (equivalent to Thr308 of PKB/Akt) in different cell types, including protein kinase C (PKC), protein kinase A (PKA), p70 ribosomal S6 kinase (p70 S6K), p90 ribosomal S6 kinase (p90 RSK) and serum and glucocorticoid regulated (SGK) kinases [5,6]. Recent studies have also shown that PDK-1 controls the MAPK and NF-κB pathways via direct phosphorylation of the activation-loop residues of MEK 1/2 [8] and IKKβ respectively [9]. In T lymphocytes, PDK-1 recruits PKC θ and various adaptor proteins to the plasma membrane that assemble to promote NF-κB activation [10]. The expression of PDK-1 is important for T cell development, as complete loss of PDK-1 in the T cell lineage using Cre-loxP technology completely blocks the differentiation of α/β pre-t cells in the thymus [11], whilst deleting a single PDK-1 allele in the T cell lineage also significantly impairs α/β T cell differentiation [12]. Like PKB/Akt, PDK-1 also contains a PH domain, but whether 3 -phosphoinositides influence the intracellular localisation and/or activity of this enzyme is unresolved [5,6]. The PI3-kinase pathway is negatively regulated by the activity of a lipid phosphatase called PTEN (phosphatase and tensin homolog deleted on chromosome 10), which dephosphorylates the 3 inositol ring of PtdIns(3,4)P 2 and PtdIns(3,4,5)P 3 [13,14]. PTEN is a tumour suppressor gene located on chromosome 10q23, a region that suffers loss of heterozygosity in many human cancers [15,16]. Many leukaemic T cell lines, including the widely used Jurkat T cell line, do not express functional PTEN protein, due to naturally occurring mutations in both alleles of the PTEN gene [17,18]. Tumours or cell lines that have lost expression of PTEN have unopposed PI3-kinase activity, resulting in elevated basal levels of PtdIns(3,4)P 2 /PtdIns(3,4,5)P 3. In accordance with this finding, 50 % of the total cellular pool of the PH domaincontaining tyrosine kinase Itk is localised in the plasma membranerich fraction of unstimulated PTEN-null Jurkat T cells [17]. Furthermore, PKB/Akt is present in Jurkat cells as a constitutively active (Thr308/Ser473 phosphorylated) and membrane-localised enzyme in the absence of any exogenous stimulus [17,18]. Treating Jurkat cells with pharmacological inhibitors of PI3-kinase or expressing PTEN in this cell line, to reduce the levels of 3 - phosphoinositides, inhibits the association of Itk with the plasma membrane [17]. These same treatments also promote the dephosphorylation of PKB/Akt at Thr308/Ser473 and abrogate the association of the enzyme with the plasma membrane [17 19]. The constitutive phosphorylation levels of PKB/Akt at Thr308 in Jurkat T cells are potentially attributable to aberrant PDK-1 signalling, an enzyme that also contains a PH domain, although this has never been formally tested. We have previously demonstrated that PKCs are also phosphorylated at the activation-loop residue in unstimulated Jurkat T cells and that, similar to PKB/Akt phosphorylation, short-term cellular stimulation with anti-tcr/ CD28 antibodies does not affect the levels of PKC phosphorylation [20]. Because of the importance of PDK-1 in regulating many of the pathways that control T cell function, we investigated if and how PTEN loss impacts on the PDK-1 pathway in Jurkat T cells. Our findings demonstrate that PDK-1 functions independently of 3 -phosphoinositides in T cells. 2. Experimental procedures 2.1. Cell culture The Jurkat E6.1 leukaemic T cell line and HuT-78 T cell lines were obtained from the American Type Culture Collection (Manassas, VA) and were cultured in RPMI-1640 containing 10% (v/v) heat-inactivated foetal calf serum (FCS), 50 U/ml penicillin, 50 μg/ml streptomycin and 2 mm L-glutamine (complete medium) in a humidified chamber at 37 C containing 5% CO 2. The PTENinducible Tet-on Jurkat clones have previously been described [21] and were cultured in complete RPMI-1640 medium with freshly added antibiotics G418 (100 μg/ml) and Hygromycin B (100 μg/ml) (Clontech, CA). Three PTENinducible Tet-on Jurkat clones were used in this study; clones 12 and 17 are PTEN-inducible whilst clone 18 is a non-pten-expressing control clone. The expression of PTEN was induced by the addition of 1 μg/ml doxycycline (Clontech, CA) to the cells for 48 h. Peripheral blood lymphocytes (PBLs) were isolated from the blood of healthy volunteers on Lymphoprep buffy coats (Nycomed, Norway), washed twice with sterile PBS and incubated overnight in complete RPMI-1640 medium at 37 C with 5 % CO 2. Non-adherent cells were then harvested by centrifugation and processed for total cell lysates Reagents and antibodies The non-radioactive PtdIns(3,4,5)P 3 (PIP 3 ) mass assay ELISA kit was obtained from Echelon Biosciences (Bryce Canyon, UT). Okadaic acid was obtained from Merck Biosciences (Nottingham, UK). LY was obtained from Alexis (Nottingham, UK). Wortmannin was obtained from Sigma (St. Louis, MO).

3 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx 3 Antibodies to the phosphorylated (pthr308 and pser473) and total forms of PKB/Akt were obtained from Cell Signalling Technology (Beverly, MA). Antibodies to phosphorylated PDK-1 (pser241), PTEN, phosphorylated ERK 1/2, total ERK 1/2, phosphorylated PKC θ (pthr538) and HRP-conjugated secondary antibodies were also obtained from Cell Signalling Technology. The anti-cd11a/cd18 (LFA-1) antibody was obtained from BD Biosciences (Oxford, UK). Antibodies to PKC β were obtained from Seikagaku Corporation (Tokyo, Japan) and Zymed (San Francisco, CA. Anti-PKC δ antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA) and BD Biosciences (Oxford, UK). Anti-PKC θ antibodies were obtained from BD Biosciences (Oxford, UK). The antibody that recognises the phosphorylated activation-loop threonine of all PKC isoforms (P500 antibody) was generously provided by Professor Alexandra Newton (University of California, San Diego) [22]. Sheep anti-human PDK-1 antiserum and the GFP-tagged PDK-1 construct were kindly donated by Dr. Dario Alessi (University of Dundee, Scotland). A second anti-pdk-1 antibody, used to immunoprecipitate endogenous PDK-1 for kinase assays was obtained from Upstate Biotech (Charlottesville, VA). HRP-labelled anti-sheep secondary antibody was from Sigma (St. Louis, MO). SMARTpool small interfering RNA oligonucleotides (sirnas) to human PDK-1 (Entrez Gene ID 5170) and non-targeting control pool sirnas (sicontrol) were obtained from Dharmacon (Chicago, IL) (catalogue numbers M and D , respectively). The sirnas to PDK-1 in this SMARTpool are designed to target both PDK-1 mrna transcript variants (NM_ and NM_031268) that are deposited in the Entrez database ( &val= and nucleotide&val= respectively) Estimation of PtdIns(3,4,5)P 3 levels PtdIns(3,4,5)P 3 was extracted from Jurkat, HuT-78 and PBL cells according to the manufacturers instructions and quantified by a non-radioactive competitive ELISA kit (Echelon Biosciences). Lipid extract from Jurkat and HuT-78 T cells was analysed in duplicate in these assays while extract from PBLs was analysed in duplicate. This is because PBLs are much smaller in size than Jurkat or HuT-78 T cells and therefore the amount of PBLs used in the assay were normalised to Jurkat and HuT-78 T cells based on a protein assay rather than total cell number. Total PtdIns(3,4,5)P 3 levels were estimated from PtdIns(3,4,5)P 3 standards T cell transfection and sirna-mediated knockdown of PDK-1 expression Jurkat T cells were transfected with a GFP-tagged PDK-1 construct, or with sirnas to human PDK-1 or non-silencing controls, using the Amaxa Biosystems nucleofector electroporation system as described elsewhere [23]. Briefly, Jurkat cells were resuspended in 100 μl of nucleofection solution R with 10 μg of GFP-PDK-1 or with the indicated amounts of sirnas. After electroporation, 0.5 ml serum and antibiotic-free RPMI-1640 medium was added and the cells were incubated for 15 min at 37 C containing 5% CO 2. Complete RPMI-1640 medium (5 mls) was then added and the cells were incubated overnight at 37 C containing 5% CO 2. GFP-PDK-1 transfected Jurkat cells were treated with DMSO or LY for 6 h, harvested and placed on Superfrost slides (VWR International, Leuven) and fixed in 4% paraformaldehyde. Slides were visualised on a Nikon Eclipse E800 microscope. Jurkat cells that were electroporated with sirnas were washed three times in complete medium after overnight incubation and lysates were prepared 24, 48, 72 and 96 h post-electroporation. HuT-78 T cells were transfected with the GFP-PDK-1 construct according to standard protocols (Amaxa Biosystems) Cell lysis, immunoprecipitation and cell fractionation Stimulation of T cells with anti-tcr/cd28 antibodies and preparation of total cell lysates, immunoprecipitates and subcellular fractions was performed as described in [20] SDS-PAGE and Western blotting Total lysates, immunoprecipitates and fractionated lysates were resolved on 10 % SDS-PAGE gels, transferred to PVDF membranes and probed with the relevant antibodies according to the manufacturers instructions. Immunoreactive bands were visualised with the Phototope-HRP detection system (Cell Signalling) and subsequent exposure to Kodak light-sensitive film (Rochester, NY). Where indicated, the PVDF membrane was stripped of its antibodies by incubation in stripping buffer (62.5 mm Tris HCl (ph 6.8), 2% SDS and 100 mm β-mercaptoethanol) for 30 min at 65 C, followed by washing, blocking and reprobing with antibodies as outlined above. Densitometric analysis of selected immunoblots was performed using the Kodak 1D Image Analysis Software program (New Haven, CT) PDK-1 activity assays Assays for PDK-1 activity were performed as outlined in [24], using 100 μm Suntide (RRKDGATMKTFCGTPE) as substrate, except that endogenous PDK-1 was immunoprecipitated with a commercially available anti-pdk-1 antibody (Upstate, VA) and assayed under the specific conditions. 3. Results 3.1. Inhibitors of PI3-kinase do not affect the catalytic activity or Ser241 phosphorylation status of PDK-1 in Jurkat T cells Jurkat leukaemic T cells lack expression of the lipid phosphatase PTEN, an effect that results in unopposed PI3-kinase activity and therefore high basal levels of PtdIns(3,4)P 2 and PtdIns(3,4,5)P 3 in these cells [17,18]. Preliminary experiments also confirmed that our Jurkat T cell line grown in culture lacked any detectable expression of PTEN (Fig. 1A). Although the expression of this phosphatase was absent in the Jurkat leukaemic T cell line, PTEN was readily detectable in lysates derived from unstimulated PBLs. In agreement with other reports [19,25], we found that PTEN was also expressed in the HuT-78 lymphoma T cell line (Fig. 1A). The phosphorylation status of PKB/Akt at the activation-loop residue (Thr308) was analysed in the same lysates by Western blotting. Consistent with the finding that Jurkat cells contain elevated levels of PtdIns(3,4)P 2 /PtdIns(3,4,5)P 3, it was observed that PKB/Akt was phosphorylated at Thr308 in the absence of any exogenous stimulation (Fig. 1A). This is in contrast to normal PBLs and unstimulated HuT-78 T cells, where PKB/Akt was expressed at similar levels to Jurkat cells but was not phosphorylated at Thr308 (Fig. 1A). PKB/Akt was also robustly phosphorylated on Thr308 when Jurkat cells were cultured overnight in the absence of serum (data not shown). Furthermore, PKB/Akt phosphorylation was non-inducible in Jurkat cells as stimulation with anti-tcr/cd28 antibodies did not modulate the levels of phosphorylated PKB/Akt (Fig. 1B) but nevertheless induced robust ERK 1/2 phosphorylation (data not shown and Fig. 1C). In contrast, anti-tcr/cd28 stimulation of HuT-78 T cells induced transient phosphorylation of PKB/Akt at Thr308 and an inhibitor of PI3-kinase (LY294002) blocked this phosphorylation (Fig. 1B). We have also recently demonstrated non-inducible PKC activation-loop phosphorylation in Jurkat T cells [20]. Non-inducible phosphorylation of both PKC and PKB/Akt in Jurkat cells implies that the upstream kinase, PDK- 1, could be unregulated in these cells due to PTEN loss and unopposed PI3-kinase activity. In support of this, we observed that PDK-1 in Jurkat T cells was maximally phosphorylated on Ser241, the activation-loop residue that is analogous to Thr308

4 4 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx of PKB/Akt and is absolutely required for PDK-1 activity (Fig. 1C) [26]. While anti-tcr/cd28 stimulation did not modulate the levels of phosphorylated PDK-1, we consistently observed a slight reduction in the levels of PDK-1 following stimulation (Fig. 1C) which could indicate degradation and/or translocation of the kinase to a detergent-insoluble fraction following stimulation (e.g. nuclear, cytoskeletal or lipid raft fraction). However, the loss of PDK-1 from the detergentsoluble fraction did not correlate with the appearance of the kinase in the detergent-insoluble fraction following stimulation (data not shown). To investigate whether unopposed PI3-kinase activity results in deregulation of PDK-1 activity and intracellular localisation in Jurkat T cells, the cells were incubated with LY294002, a pharmacological inhibitor of PI3-kinase, to reduce the levels of 3 -phosphoinositides and the catalytic activity of endogenous PDK-1 was analysed. As shown in Fig. 2A, endogenous PDK-1 displayed high basal activity towards a peptide substrate based on the activation-loop sequence of PKB/Akt, when isolated from unstimulated Jurkat T cells. Treating Jurkat T cells with 50 μm LY for 30 min or 3 h did not affect the catalytic activity or the phosphorylation status of PDK-1 at Ser241 (Fig. 2A). Since PDK-1 has been reported to catalyse activation-loop phosphorylation of PKC isoforms [5,6], the phosphorylation status of PKC isoforms β I, δ and θ at the activation-loop residue was analysed following PI3-kinase inhibition or PTEN expression, as an additional readout of PDK-1 activity. As shown in Fig. 2B, treating Jurkat T cells with LY or wortmannin (an unrelated inhibitor of PI3-kinase) did not affect the phosphorylation status of PKC β I, δ and θ at the activation-loop site. Incubation of Jurkat cells with LY for longer periods (up to and including 24 h) also had no effect on the phosphorylation status of PKC isoforms in Jurkat T cells (data not shown). The phosphorylation status of PKC θ at the activation-loop site (Thr538) was confirmed using a commercially available phosphorylation-specific antibody following PI3-kinase inhibition; inhibitors of PI3-kinase did not Fig. 1. (A) Unstimulated Jurkat T cells, peripheral blood lymphocytes (PBL) and HuT-78 T cell lysates were probed for the expression of PTEN, phosphorylated (pthr308) PKB/Akt and total PKB/Akt by immunoblotting. (B) Resting and anti-tcr/cd28 stimulated Jurkat and HuT-78 T cell lysates were probed for phosphorylated (pthr308) or total PKB/Akt by immunoblotting. HuT-78 T cells were also pre-treated with 50 μm LY (LY) for 30 min before being stimulated for an additional 30 min with anti-tcr/cd28 antibodies. (C) Resting and anti-tcr/cd28 stimulated Jurkat T cell lysates were probed for phosphorylated PDK-1 (pser241), total PDK-1, or phosphorylated (perk 1/2) and total ERK 1/2 by immunoblotting.

5 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx 5 Fig. 2. Inhibitors of PI3-kinase do not affect the catalytic activity or Ser241 phosphorylation status of endogenous PDK-1 in Jurkat cells. (A) Jurkat cells were treated with 50 μm LY (+) or DMSO control ( ) for 30 min and 3 h. Lysates were immunoprecipitated with an antibody to PDK-1 and immunoprecipitates were tested for PDK-1 activity in an in-vitro kinase assay. A portion of each lysate was probed for the expression of phosphorylated (pser241) PDK-1, total PDK-1 or actin (as a loading control) by immunoblotting. (B) Jurkat T cells were treated for 3 h with 200 nm wortmannin (W), 50 μm LY (LY) or DMSO control ( ) and lysates were immunoprecipitated (IP) with antibodies to PKC β I, δ and θ. Immunoprecipitates were immunoblotted with an antibody that recognises the phosphorylated activation-loop threonine of PKC isoforms. As a relative loading control, the blots were stripped and reprobed for total PKCs. (C) PtdIns(3,4,5)P 3 lipid were extracted from Jurkat cells treated with DMSO or 50 μm LY for 3 h and from unstimulated HuT-78 and PBLs. PtdIns(3,4,5)P 3 mass was quantified by ELISA. Extract from Jurkat and HuT-78 T cells was analysed in these assays while extract from PBLs was analysed for reasons outlined in the Experimental procedures section. (D) Jurkat cells were treated with DMSO (0), 200 nm wortmannin or 50 μm LY for various times and lysates were immunoblotted for phosphorylated (pthr308) and total PKB/Akt.

6 6 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx modulate the phosphorylation status of PKC θ at Thr538 in Jurkat cells (Supplementary Fig. 1). To demonstrate that inhibition of PI3-kinase was effective in Jurkat T cells, we quantified PtdIns (3,4,5)P 3 mass in Jurkat T cells both before and after LY treatment. Unstimulated Jurkat cells contained high basal levels of PtdIns(3,4,5)P 3 and treating the cells with 50 μm LY for 3 h reduced PtdIns(3,4,5)P 3 mass by 65 % (Fig. 2C). Furthermore, PKB/Akt underwent almost complete dephosphorylation at Thr308 in Jurkat T cells following LY or wortmannin treatment (Fig. 2D).

7 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx Expression of PTEN in Jurkat T cells does not affect the catalytic activity or Ser241 phosphorylation status of PDK-1 The levels of 3 -phosphoinositides in Jurkat cells were also reduced by reconstituting the cells with PTEN [21]. PTEN was expressed in Jurkat T cells under the control of a tetracyclineinducible (Tet-on) expression system. For this study, three PTEN-inducible Tet-on Jurkat clones were used; clone 12 and clone 17 can be induced to express PTEN by the addition of doxycycline, whilst clone 18 is a negative control clone that does not express PTEN under the same induction conditions. As shown in Fig. 3A, addition of doxycycline to the culture medium of Jurkat Tet-on clones 12 and 17 for 48 h resulted in the upregulation of PTEN in these cells (but not in clone 18). The expression of PTEN in Jurkat Tet-on cells did not significantly affect PDK-1 activity or the levels of phosphorylation of PDK-1 at Ser241 (Fig. 3A). The effect of PTEN expression on the phosphorylation status of PKC isoforms at the activation-loop residue in Jurkat T cells was also analysed. PTEN expression did not affect the phosphorylation status of PKC isoforms PKC β I, δ and θ (Fig. 3B). The phosphorylation status of PKC θ following PTEN expression was also confirmed in Jurkat cells using a commercially available phosphorylation-specific (Thr538) activation-loop antibody. The expression of PTEN did not modulate the phospho-thr538 status of PKC θ (data not shown). To confirm that PTEN expression in Jurkat T cells was effective in reducing PtdIns(3,4,5)P 3 levels, the phosphorylation status of PKB/Akt at Thr308 was analysed before and after PTEN expression. As shown in Fig. 3C, expression of PTEN in Jurkat Tet-on clones 12 and 17 resulted in the dephosphorylation of PKB/Akt at Thr308 whilst not affecting the total levels of expression of the enzyme. As expected, PTEN was not induced in control clone 18 cells and the phosphorylation status of PKB/ Akt was not affected by the addition of doxycycline (Fig. 3C) Inhibitors of PI3-kinase do not affect the subcellular localisation of PDK-1 in Jurkat T cells Since PDK-1 contains a PH domain, it is expected that the levels of 3 -phosphoinositides in cells would influence its subcellular localisation. The effect of PI3-kinase inhibitors on the intracellular localisation of endogenous PDK-1 in Jurkat T cells was therefore investigated by immunofluorescence and Western blotting of subcellular fractions. Because the sheep antihuman PDK-1 antibody was not suitable for immunofluorescence staining of endogenous PDK-1 in our hands, Jurkat cells were transfected with a GFP-tagged PDK-1 construct and the intracellular distribution of this protein was monitored before and after LY treatment. In drug vehicle control-treated Jurkat cells, GFP-PDK-1 was highly concentrated at the plasma membrane with less intense, albeit detectable, fluorescence observed throughout the cytoplasm (Fig. 4A). Treating the cells with LY for 6 h did not substantially affect the distribution of PDK-1 at the plasma membrane. Subcellular fractionation experiments also revealed that reducing the levels of 3 -phosphoinositides in Jurkat cells with LY for up to 6 h did not affect the proportion of endogenous PDK-1 that associated with the membrane-rich fraction (Fig. 4B). To demonstrate that reducing 3 -phosphoinositide levels with inhibitors of PI3-kinase were effective in Jurkat T cells, the subcellular distribution of PKB/Akt was monitored before and after LY treatment by cell fractionation and Western blotting. As a first step it was observed that PKB/Akt was present in both the cytosol and membrane-rich fraction in PTEN-null Jurkat T cells in comparison to PTEN-expressing HuT-78 T cells, where the majority of the kinase was recovered in the cytosolrich fraction only (Fig. 4C). Furthermore, PKB/Akt was equally phosphorylated on Thr308 and Ser473 in both the cytosolic and membrane fractions of Jurkat cells (see Fig. 4D). Localisation of phosphorylated PKB/Akt in the cytosol, as well as the membrane fraction of Jurkat T cells is consistent with reports in other cell types that activated (phosphorylated) PKB/Akt disengages from the plasma membrane and translocates through the cytosol to the nucleus in response to PI3-kinase activation [27,28]. We next investigated the effect of inhibitors of PI3-kinase on the intracellular localisation of PKB/Akt in Jurkat T cells. As expected, PKB/Akt was recovered from both the cytosol and membrane-rich fractions of unstimulated Jurkat T cells and was phosphorylated on Ser473 in both fractions (Fig. 4D). Treatment with LY for up to three hours resulted in the dephosphorylation of PKB/Akt at Ser473 and virtually abrogated the association of this kinase with the membrane-rich fraction of Jurkat cells (Fig. 4D and E). Taken together, these experiments demonstrate that although PDK-1 contains a PH domain, alterations in the levels of 3 -phosphoinositides in Jurkat cells do not influence its catalytic activity, activation-loop phosphorylation status or intracellular localisation An inhibitor of PP2A phosphatases abrogates the dephosphorylation of PKB/Akt following reductions in 3 - phosphoinositide levels in Jurkat cells Because PI3-kinase inhibitors or expression of PTEN does not affect the activity, Ser241 phosphorylation or subcellular Fig. 3. Expression of PTEN in Jurkat T cells does not affect the catalytic activity or Ser241 phosphorylation status of PDK-1. (A) PTEN-inducible Tet-on Jurkat T clones (indicated with #) were left untreated or treated with doxycycline (Dox) for 48 h to induce expression of PTEN. Clones 12 and 17 are PTEN-inducible while Clone 18 is a PTEN-negative control. Lysates were immunoprecipitated with an antibody to PDK-1 and immunoprecipitates were tested for PDK-1 activity in an invitro kinase assay. A portion of each lysate was immunoblotted for the expression of PTEN, phosphorylated (pser241) PDK-1, total PDK-1 or actin as a loading control. (B) PTEN-inducible Tet-on Jurkat T clones cells were left untreated or treated with doxycycline (Dox) for 48 hours to induce PTEN expression and lysates were immunoprecipitated with antibodies to PKC β I, δ, θ or a non-specific monoclonal antibody. Immunoprecipitates were probed by Western blotting with an antibody that recognises the phosphorylated activation-loop threonine of PKC isoforms. As a relative loading control, the blots were stripped and reprobed for total PKCs. (C) PTEN-inducible Tet-on Jurkat T clones were left untreated or treated with doxycycline for 48 h to induce expression of PTEN. Lysates were immunoblotted for the expression of PTEN, phosphorylated ( pthr308) PKB/Akt and total PKB/Akt. Also included in the analysis, as controls for immunoblotting, were wild-type (WT) Jurkat T cells (PTEN-negative) and HuT-78 T cells (PTEN-positive).

8 8 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx

9 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx 9 Fig. 5. An inhibitor of PP2A phosphatases abrogates the dephosphorylation of PKB/Akt following reductions in 3 -phosphoinositide levels in Jurkat cells. (A) Jurkat cells were pretreated or not for 1 h with 300 nm okadaic acid before adding 200 nm wortmannin (for 6 h) or 50 μm LY (for 3 h). Lysates were immunoblotted for phosphorylated (pthr308) and total PKB/Akt. Also included in the analysis were HuT-78 T cells that were treated with DMSO or 300 nm Okadaic acid for one hour. (B) Quantification of phosphorylated and total PKB/Akt levels in (A). localisation of PDK-1, there exists the possibility that the dephosphorylation of PKB/Akt on the activation-loop site is due to (a) redistribution of this enzyme back to the cytosol (and away from constitutively active and membrane-localised PDK- 1) (Fig. 4D) and (b) sensitivity to site-specific phosphatases. Therefore, the possibility that pre-treating the cells with a serine/threonine phosphatase inhibitor could abrogate or retard the wortmannin or LY mediated dephosphorylation of PKB/Akt in Jurkat T cells was investigated. Treating Jurkat cells with 300 nm okadaic acid (an inhibitor of PP2A-like phosphatases) prior to addition of wortmannin/ly partially inhibited the dephosphorylation of PKB/Akt (Fig. 5; for example compare lanes 2 and 5 or lanes 3 and 6). Importantly, treating Jurkat cells with this concentration of okadaic acid alone did not increase the basal levels of phosphorylation of PKB/Akt in this cell line or promote the phosphorylation of non-phosphorylated PKB/Akt in HuT-78 T cells at the activation-loop residue (Fig. 5; lanes 4 and 8). This experiment therefore reveals that the dephosphorylation of PKB/Akt following reductions in 3 -phosphoinositides is ultimately controlled by PP2A phosphatase activity. The ability of okadaic acid to abrogate the PTEN-induced dephosphorylation of PKB/ Akt could not be performed in Jurkat Tet-on cells due to the toxicity associated with okadaic acid in long-term culture Fig. 4. Inhibitors of PI3-kinase do not affect the subcellular localisation of PDK-1 in Jurkat T cells. (A) Jurkat cells were transfected with a GFP-PDK-1 construct and after overnight incubation the cells were treated with DMSO or 50 μm LY for 6 h. GFP-transfected cells were then analysed by immunofluorescence. (B) Jurkat cells were treated with 50 μm LY for up to 6 h or with DMSO for 6 h, cells were fractionated into cytosol (C) and membrane-rich (M) fractions and immunoblotted for PDK-1 expression. (C) Unstimulated Jurkat or HuT-78 T cells were fractionated into cytosol (C), membrane (M) and cytoskeletal-rich (S) fractions and immunoblotted for total PKB/Akt. The quality of the fractions was assessed by probing the same fractions for ERK 1/2 (cytosol) and LFA-1 (membrane). (D) Jurkat cells were treated with 50 μm LY for 30 min and 3 h or with DMSO as a control for 3 h. The cells were fractionated into cytosol (C), membrane (M) and cytoskeletal-rich (S) fractions and immunoblotted for phosphorylated ( pser473) and total PKB/Akt. The quality of the fractions was assessed by probing the same fractions for ERK 1/2 (cytosol) and LFA-1 (membrane). (D) Densitometric analysis of the percentage of total PKB/Akt in the membrane fraction from Fig. 4C.

10 10 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx (doxycycline-induced PTEN expression in Jurkat Tet-on cells required 48 h) Phosphorylation and subcellular localisation of PDK-1 in HuT-78 T cells While we have shown in Jurkat T cells that reductions in the levels of 3 -phosphoinositides does not influence PDK-1 activity, Ser241 phosphorylation or intracellular localisation, it should be highlighted that inhibitors of PI3-kinase did not completely abolish PtdIns(3,4,5)P 3 mass after three hours of treatment (Fig. 2C). Whilst reducing 3-phosphoinositide levels by 65% with inhibitors of PI3-kinase was sufficient to abolish PKB/Akt phosphorylation and localisation at the membranerich fraction, the remaining 3 -phosphoinositides could potentially still activate PDK-1, as the affinity of the PH domain of PDK-1 for 3 -phosphoinositides is 20-fold higher than that of the PH domain of PKB/Akt [29 31]. We therefore analysed the Ser241 phosphorylation status and subcellular distribution of PDK-1 in HuT-78 T cells i.e. PTEN-expressing cells that possess undetectable PtdIns(3,4,5)P 3 levels (Fig. 2C). As shown in Fig. 6A, PDK-1 was also phosphorylated at Ser241 in unstimulated HuT-78 T cells and stimulation of the cells with anti- TCR/CD28 antibodies did not modulate the levels of phosphorylation. Furthermore, transfection of HuT-78 T cells with GFP-tagged PDK-1 demonstrated that the enzyme also predominantly localised to the plasma membrane (Fig. 6B) and that PDK-1-regulated substrates such as PKC isoforms were also phosphorylated in this cell line (Fig. 6C). The phosphorylation status of PKC isoforms was also investigated in PTENexpressing PBLs (Fig. 6D). We determined that PKCs were phosphorylated to similar extents in these cells when compared to Jurkat T cells, again demonstrating that PDK-1 functions independently of 3 -phosphoinosides in T lymphocytes. 4. Discussion Fig. 6. Phosphorylation and subcellular localisation of PDK-1 in HuT-78 T cells. (A) Resting and anti-tcr/cd28 stimulated HuT-78 T cell lysates were probed for phosphorylated PDK-1 (pser241), total PDK-1, or phosphorylated (perk 1/2) and total ERK 1/2 by immunoblotting. (B) HuT-78 and Jurkat T cells were transfected with GFP-PDK-1 and after overnight incubation PDK-1 expressing cells were analysed by immunofluorescence. (C) Jurkat (1) or HuT- 78 (2) T cell lysates were immunoprecipitated with antibodies to PKC β I, δ and θ or irrelevant control antibody. Immunoprecipitates were immunoblotted with an antibody that recognises the phosphorylated activation-loop threonine of PKC isoforms. As a relative loading control, the blots were stripped and reprobed for total PKCs. (D) Jurkat (1) or PBL (2) lysates were also probed for phosphorylated PKC isoforms as described in (C). The central importance of PDK-1 signalling in-vivo is highlighted by the finding that mouse embryos lacking PDK-1 die at day 9.5 during development [32]. To circumvent embryonic lethality, mutant mice with tissue-specific deletion of PDK- 1 using Cre-loxP technology have been generated. Although initially viable, these mice soon develop a number of defects [33 36]. As expected, activation of PDK-1-regulated substrates, such as PKB/Akt or p70 S6K, is defective in these tissues. The role of PDK-1 in T lymphocyte biology has also recently been addressed, demonstrating roles for the enzyme in T cell differentiation and proliferation [11,12]. Although the central importance of PDK-1 expression in different tissues has been explored, what is less understood is the regulation of PDK-1 by the PI3-kinase pathway. For example, it has been reported that cellular stimulation and PI3- kinase activation modulates PDK-1 activity and/or localisation [24, 37 43], whilst other studies have shown that PDK-1 is a constitutively active enzyme whose intracellular localisation is regulated independently of the PI3-kinase pathway [29,44]. Nonetheless, because PDK-1 contains a PH domain, its intracellular localisation or activity is potentially unregulated in Jurkat cells as a consequence of PTEN loss. In this study we found that restoring the PI3-kinase/PTEN balance in Jurkat cells with pharmacological inhibitors of PI3-kinase or expression of PTEN to reduce 3 -phosphoinositide levels did not affect the catalytic activity of endogenous PDK-1 or its Ser241 activationloop phosphorylation status. Since PDK-1 has also been reported to catalyse activation-loop phosphorylation of PKC isoforms, the phosphorylation status of PKCs at this residue was analysed as an additional readout of PDK-1 activity. It was found that reducing 3 -phosphoinositide levels did not affect PKC activation-loop phosphorylation. The intracellular

11 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx 11 localisation of PDK-1 in Jurkat T cells following restoration of the PI3-kinase/PTEN balance was also investigated. While immunofluorescence and subcellular fractionation studies demonstrated that the majority of PDK-1 was localised at or near the plasma membrane in PTEN-null Jurkat cells, reducing the levels of 3 -phosphoinositides did not affect the proportion of PDK-1 that associated with the plasma membrane. To confirm that the PI3-kinase/PTEN balance was restored in Jurkat cells, 3 -phosphoinositide levels were directly measured after PI3-kinase inhibition and it was demonstrated that treating cells with LY reduced PtdIns(3,4,5)P 3 levels by 65 %. This reduction in 3 -phosphoinositide levels led to the dephosphorylation of PKB/Akt at Thr308 and abolished its localisation at the membrane-rich fraction in Jurkat T cells. Because the dephosphorylation of PKB/Akt was not due to reduced PDK-1 activity, we investigated whether protein phosphatases controlled PKB/Akt inactivation. It was observed that the dephosphorylation of PKB/Akt was sensitive to PP2Alike inhibitors in Jurkat cells, a finding that is in agreement with reports elsewhere that PP2A phosphatases catalyse the dephosphorylation of PKB/Akt in-vitro and in-vivo [45 48]. Although both PDK-1 and PKB/Akt possess PH domains, these domains on the respective proteins differ both in terms of their primary protein sequence and their affinity for 3 - phosphoinositides: two independent studies demonstrated that the PH domain of PDK-1 binds 3 -phosphoinositides with a 20- fold higher affinity than the PH domain of PKB/Akt in-vitro [29 31]. Furthermore, it appears that the PH domain of PDK-1, like PKB/Akt, negatively regulates the enzyme, as evidence suggests that engagement of this module relieves autoinhibition [38,49]. While reducing the levels of PtdIns(3,4,5) P 3 by 65 % in Jurkat cells following treatment with LY was sufficient to abolish PKB/Akt phosphorylation at Thr308 and membrane localisation, we could not rule out that the PtdIns (3,4,5)P 3 that remained after PI3-kinase inhibition could potentially still retain PDK-1 at the plasma membrane in a high activity form (phosphorylated at Ser241 and relieved of autoinhibition). We therefore analysed the phosphorylation status and intracellular distribution of PDK-1 in HuT-78 T cells that express endogenous PTEN and possess undetectable PtdIns (3,4,5)P 3 levels in the resting state (Fig. 2C). It was observed that PDK-1 in HuT-78 T cells, similar to Jurkat cells with high basal levels of PtdIns(3,4,5)P 3, was also constitutively phosphorylated on its own activation-loop residue and that the enzyme was localised at the plasma membrane in the absence of PtdIns(3,4,5)P 3. Furthermore, PKC isoforms also retained activation-loop phosphorylation in both HuT-78 and PBLs in unstimulated cells. Our studies demonstrate therefore that PDK- 1 activity, Ser241 phosphorylation status and intracellular distribution at the plasma membrane is regulated independently of 3 -phosphoinositide levels in T lymphocytes. Recent studies in antigen-specific mouse D10 T cells however have demonstrated that a proportion of PDK-1 transiently translocates to the immunological synapse (a sub-domain of the plasma membrane enriched in specific signalling complexes that corresponds to the contact point between the T cell and antigen-presenting cell) in a PI3-kinase-dependent manner following stimulation. This translocation is necessary for PKA phosphorylation/activation and interleukin-4 production [50]. This implies that antigen stimulation and 3 -phosphoinositide production could influence the intracellular distribution of PDK-1 at defined sub-compartments of the plasma membrane in T lymphocytes, rather than the entire plasma membrane. While we found that PDK-1 was not influenced by the levels of 3 -phosphoinositides in T lymphocytes, PDK-1 activity and Ser241 phosphorylation may be regulated by 3 -phosphoinositides in other cell types, which could explain the discrepancies that have been reported in the apparent requirement of PDK-1 for these lipids [26,43]. Alessi and colleagues have reported that entities others than 3 -phosphoinositides, such as other membrane lipids or scaffolding proteins, might also regulate PDK-1 plasma membrane distribution, as quantitatively eliminating 3 -phosphoinositide mass using chronic PI3-kinase inhibition did not abolish PDK-1 membrane localisation [29]. Furthermore, if 3 -phosphoinositides solely controlled PDK-1 membrane localisation, then it could be surmised that all of the cellular pool of PDK-1 would be present at the plasma membrane of cells following PtdIns (3,4)P 2 /PtdIns(3,4,5)P 3 production. This is not the case, as studies to date have all reported that only a small proportion of PDK-1 either constitutively resides in the plasma membrane of cells or translocates following PI3-kinase activation, with substantial amounts always present in the cytosol [24,29,37 42]. Distribution of PDK-1 at the plasma membrane may instead be influenced by PtdIns(4,5)P 2, the lipid substrate of PI3-kinase that is constitutively present in the plasma membrane of quiescent cells, rather than 3 -phosphoinositides [29]. A very recent study has demonstrated that phosphatidic acid, produced via phospholipase D, interacts with a defined site in the PH domain of the Ras guanine nucleotide-exchange factor Sos to regulate its plasma membrane recruitment [51]. It is therefore tempting to speculate that phosphatidic acid, rather than 3 -phosphoinositides, regulates PDK-1 plasma membrane localisation. Other inositol phospholipids such as Ins(1,3,4,5,6)P 5 and IP 6 are important for localising PDK-1 to areas other than the plasma membrane, including the cytosol, where it catalyses activation-loop phosphorylation of substrates that do not contain membrane-binding modules, such as p70 S6K and p90 RSK [52]. A number of PDK-1-interacting proteins have also been identified in other cell types that influence its activity, stability and/or spatial distribution, including proteins [53], Hsp-90 [54], Grb-14 [55], STRAP [56] AKAPs [57] and caveolin-1 [58]. Whether these lipids and proteins modulate PDK-1 activity and localisation in T cells is unknown however. Our finding that PKC activation-loop phosphorylation levels are constitutive in T lymphocytes and not a consequence of aberrant PDK-1 activity due to PTEN loss is consistent with a report from Newton and colleagues that the levels of PKC α/β II activation-loop phosphorylation are not influenced by the levels of 3 -phosphoinositides in adherent COS-7 cells [49]. Furthermore, embryonic stem cells that express a form of PDK-1, with a mutation in its PH domain that abolishes 3 -phosphoinositide binding, have unaltered PKC activation-loop phosphorylation levels [59]. Studies from Alessi and colleagues and the Newton laboratory have shown that PDK-1 instead docks on the region

12 12 M. Freeley et al. / Cellular Signalling xx (2007) xxx xxx that encompasses the C-terminal hydrophobic-motif of newly synthesised PKCs (equivalent to the region that surrounds the Ser473 site of PKB/Akt) and catalyses activation-loop phosphorylation in a co-translational manner [60,61]. Studies in other cell types however have demonstrated inducible and/or PI3-kinase-dependent PKC activation-loop phosphorylation [62,63 69]. Whether this PI3-kinase-dependency in PKC activation-loop phosphorylation is channelled directly through PDK-1 or whether it is mediated via other intermediary PH domain-containing proteins in unknown. In HEK-293 cells however, the 3-phosphoinositide requirement lies with the PH domain of PDK-1, since deleting this domain on PDK-1 abolishes PKC ζ phosphorylation [63]. These studies highlight the fact that the regulation of PDK-1 and PKC activation-loop phosphorylation by 3 -phosphoinositides may be cell-type specific. A recent report demonstrated inducible activationloop phosphorylation of PKC θ in both Jurkat and primary T cells following T cell stimulation [10] but we (M.F. and A.L), and our collaborators have never observed such inducible phosphorylation at the activation-loop [70]. As well as directly measuring PDK-1 catalytic activity, we also monitored PKB/Akt and PKC activation-loop phosphorylation as additional readouts of PDK-1 activity. To demonstrate that PDK-1 was indeed the in-vivo activation-loop kinase of PKB/Akt and PKC isoforms in Jurkat T cells, knockdown of PDK-1 expression levels with PDK-1-specific sirnas was carried out. Unfortunately, knockdown of PDK-1 protein levels by approximately 90% did not result in a change in the levels of phosphorylation of PKB/Akt at the activation-loop residue (Supplementary Fig. 2). Unlike other kinase families in the human genome, only a single PDK-1 gene has been described and the sirnas that we used in this study are predicted to target both PDK-1 mrna transcript variants. It is unlikely that PDK-1 is not the activation-loop kinase of PKB/Akt and PKC isoforms in human Jurkat T cells, as PKB/Akt is not phosphorylated at Thr308 following insulin stimulation whilst PKC ζ activationloop phosphorylation levels are reduced in PDK-1-null human embryonic cells [71]. These cells also have reduced expression levels of other PKC isoforms (possibly due to a nonphosphorylated activation-loop residue causing destabilisation) [71]. Furthermore, PDK-1 phosphorylates PKB/Akt and PKCs both in-vitro and in cells and co-expression of kinase-inactive PDK-1 blocks PKC phosphorylation [reviewed in Refs. 5,6,72]. A recent study also demonstrated that a pharmacological inhibitor of PDK-1 reduces activation-loop phosphorylation of both PKB/Akt and PKC isoforms in human cells [73]. A plausible explanation as to why knockdown of PDK-1 protein in Jurkat cells does not result in a change in activation-loop phosphorylation levels of PKB/Akt and PKC isoforms is that the low levels of PDK-1 that remain after sirna-mediated knockdown may still be capable of supporting phosphorylation of these substrates. Along this line, activation of PKB/Akt, p70 S6K and p90 RSK following insulin stimulation is not affected in PDK-1 hypomorphic mice that express 10 % of the normal levels of PDK-1 [32]. Similarly, Bilanges and Stokoe reported that knockdown of PDK-1 protein levels by approximately 90% with sirnas in U87-MG glioblastoma cells was insufficient to cause a change in PKB/Akt and p70 S6K activities, with occasional knockdown of PDK-1 by greater than 95% only causing a 40% reduction in the activity of these substrates [74]. We cannot rule out also that PDK-1-regulated kinases undergo compensatory phosphorylation at the activation-loop residue following PDK-1 knockdown, as expression of a catalytic domain PKC θ construct in bacteria is phosphorylated at the activation-loop residue [75]. Our data on PDK-1, PKC and PKB/Akt phosphorylation in T lymphocytes is consistent with the following model: in normal T cells (i.e. PTEN-expressing cells such as PBLs), PKCs are phosphorylated at the activation-loop by PDK-1 in a cotranslational manner that is independent of 3 -phosphoinositides. Phosphorylation does not activate PKCs but rather primes them for activation, allowing them to be activated in response to diacylglycerol or phorbol esters. In contrast, PKB/Akt is first synthesized as a non-phosphorylated protein in PTEN-expressing T cells and remains as such until T cell stimulation occurs and 3 -phosphoinositide production promotes PKB/Akt plasma membrane translocation and a conformational change on the enzyme that permits Thr308 phosphorylation (by PDK-1) and Ser473 phosphorylation. Should loss of PTEN expression occur, it is unlikely that this loss affects PDK-1 function and its ability to phosphorylate its substrates (i.e. PKCs). Instead, PTEN loss impacts only on a relatively few number of PH domain-containing proteins such as PKB/Akt, as this enzyme is strictly dependent on 3 -phosphoinositides for its plasma membrane translocation and the conformational change that allows phosphorylation at Thr308/Ser473 to occur. The activities of many other PH domain-containing enzymes are not perturbed as a consequence of PTEN loss in Jurkat T cells, as most of these kinases require a secondary signal of some sort in addition to PI3-kinase activation. For example, although 50 % of the total cellular pool of the PH domain-containing enzyme Itk is membrane-bound in Jurkats [17], the enzyme is not active because catalytic activation of Itk also requires ZAP-70- dependent activation-loop phosphorylation, which only occurs upon TCR stimulation. Similarly, PLCγ, which also contains a PH domain and regulates PKC translocation and activation, is catalytically inactive in Jurkat cells, as the enzyme also requires Lck-dependent phosphorylation on different tyrosine residues, including Tyr783, for optimal activity [76]. Whilst most PH domain-containing enzymes are only partially dependent on PI3-kinase for activity, we have shown that another PH domaincontaining enzyme, PDK-1, is not influenced by the level of 3 - phosphoinositides in Jurkat T cells at all. Because the PDK-1 pathway has been proposed as a potential target of PI3-kinase inhibitors [77,78], we suggest that these compounds would not be effective in inhibiting the PDK-1 pathway in T lymphocytes. Acknowledgements We thank Prof. Alexandra Newton for generously providing the P500 phosphorylation-specific activation-loop PKC antibody and Dr. Dario Alessi for donating the sheep anti-human PDK-1 antiserum and the GFP-tagged PDK-1 construct. This work was supported by the Health Research Board of Ireland