Characterization of an Akt Kinase Inhibitor with Potent Pharmacodynamic and Antitumor Activity

Akt is a serine-threonine protein kinase that plays a critical role in cellular survival, metabolism, proliferation, and growth (1, 2). The Akt family consists of three isoforms (Akt1, 2, and 3) that are regulated in a similar fashion and have overlapping functions. However, knockout of individual isoforms in mice suggest a role of Akt1 in overall growth (3, 4), Akt2 in insulin signaling (5, 6), and Akt3 in brain development (7, 8). In adults, Akt1 expression is ubiquitous, expression of Akt2 is elevated in insulin-responsive tissues, and expression of Akt3 is also ubiquitous with low levels in skeletal muscle and liver.

Hyperactivation of Akt kinases is one of the most common molecular findings in human malignancies (9, 10). Many oncogenes and tumor suppressor genes result in activation of phosphatidylinositol-3-OH kinase (PI3K)/Akt kinase activity. Amplification (ErbB2 and Met) and mutation [epidermal growth factor receptor (EGFR), PI3K, platelet-derived growth factor receptor, and c-Kit] of various receptor tyrosine kinases, mutation in downstream signaling molecules (ras, src, and PI3K), loss or mutation of tumor suppressor protein (PTEN), as well as mutation and amplification of Akt itself result in increased Akt signaling in tumor cells. Increased Akt1 activity has been observed in ∼40% of breast and ovarian cancers and >50% of prostate carcinomas. Activation of Akt2 kinase has been observed in 30% to 40% of ovarian and pancreatic cancers (11, 12). Increased Akt3 enzymatic activity was found in estrogen receptor–deficient breast cancer and androgen-insensitive prostate cancer cell lines, suggesting that Akt3 may contribute to the aggressiveness of steroid hormone–insensitive cancers (2, 13). Recently, an activating mutation in the pleckstrin homology domain of Akt1 has been identified in human breast, ovarian, and colorectal cancers, suggesting a direct role of Akt1 in human cancers (14).

Akt signaling promotes cell survival and proliferation. Constitutively active Akt has been shown to protect cells from PTEN-mediated apoptosis and also to reduce the sensitivity of tumor cells to proapoptotic cytotoxic agents (15). A dominant-negative mutant of Akt inhibited tumor growth in a mouse model and selectively induced apoptosis of tumor cells expressing activated Akt (16). Genetic ablation of Akt1 inhibited mammary tumor growth in transgenic mouse models (17, 18). Simultaneous inhibition of Akt1 and Akt2 was shown to be superior to inhibition of a single isozyme for induction of caspase-3 activity in tumor cells (15). The antitumor activity of rapamycin and its analogues, which target mammalian target of rapamycin (mTOR) kinase, a downstream substrate of Akt signaling, provide evidence that targeting of the Akt pathway is a rational approach to cancer therapy (13, 19).

The present study describes a novel, ATP-competitive, pan-Akt kinase inhibitor with potent enzyme and cellular activity. GSK690693 also inhibits phosphorylation of Akt substrates in human tumor xenografts in mice and reduces the growth of various human tumor xenografts in vivo. Based on these results, GSK690693 is currently being investigated in a phase I clinical trial in cancer patients.

GSK690693 preparation. GSK690693 is a novel Akt kinase inhibitor derived from the aminofurazan chemical series (Table 1) synthesized at GlaxoSmithKline. The discovery and synthesis of GSK690693 will be described in a separate article.⁷

Animals. Female CD1 Swiss nude mice were obtained from Taconic and C.B-17 severe combined immunodeficient (SCID) mice were obtained from Charles River. All animal studies were performed in compliance with federal requirements, GlaxoSmithKline policy on the Care and Use of Animals, and with related codes of practice.

Kinase assays. The ability of GSK690693 to inhibit the activity of a wide variety of kinases was tested in vitro. His-tagged full-length Akt1, 2, or 3 were expressed and purified from baculovirus. Activation was carried out with purified PDK1 to phosphorylate Thr³⁰⁸ and purified MK2 to phosphorylate Ser⁴⁷³. To more accurately measure time-dependent inhibition of Akt, activated Akt enzymes were incubated with GSK690693 at various concentrations at room temperature for 30 min before the reaction was initiated with the addition of substrate. Final reaction contains 5 to 15 nmol/L Akt1, 2, and 3 enzymes; 2 μmol/L ATP; 0.15 μCi/μL [γ−³³P]ATP; 1 μmol/L Peptide (Biotin-aminohexanoicacid-ARKRERAYSFGHHA-amide); 10 mmol/L MgCl₂; 25 mmol/L MOPS (pH 7.5); 1 mmol/L DTT; 1 mmol/L CHAPS; and 50 mmol/L KCl. The reactions were incubated at room temperature for 45 min, followed by termination with Leadseeker beads in PBS containing EDTA (final concentration, 2 mg/mL beads and 75 mmol/L EDTA). The plates were then sealed and the beads allowed to settle for at least 5 h, and product formation was quantitated using a Viewlux Imager (PerkinElmer).

To determine the selectivity of GSK690693, we characterized the inhibition of 89 protein kinases at GlaxoSmithKline. GSK690693 was also tested at 10 μmol/L concentration in the IC₅₀ Profiler Express panel (Upstate) against 209 protein kinases using filter-binding activity assays, and Ambit Biosciences panel against 180 protein kinases measuring binding with phage display technology. A subset of kinases that showed strong inhibition at 10 μmol/L were followed up with a full dose response and IC₅₀ generation. As much as possible, the assays were configured so that the IC₅₀ values approximate the intrinsic binding constant (K_i or K_d) of GSK690693 to each enzyme and can therefore be compared for selectivity against these kinases. However, the selectivity in cells may be different because inhibitor potency will be affected to differing degrees based on the ATP K_m values for each kinase.

The data for dose responses were plotted as % activity calculated with the data reduction formula in Eq. 1:

where U is the signal measured at indicated compound concentrations, C1 is the average of signal in the absence of compound (with 1% DMSO), and C2 is the average of signal in the no reaction control (defined by quenching the reaction with 75 mmol/L EDTA before adding enzyme). Inhibition curves were generated by plotting percentage control activity versus the concentration of inhibitor in a semilog fashion. The IC₅₀ values were calculated by nonlinear regression fitting to Eq. 2:

where y is the % activity, y_min is the minimum value of y, y_max is the maximum value of y, [I] is the concentration of inhibitor, and h is the Hill coefficient.

Inhibition of Akt substrate phosphorylation in cells. An ELISA was developed for the analysis of GSK3β phosphorylation in cells. Tumor cells were treated with GSK690693 at various concentrations in a 96-well plate for 1 h. Cell lysates were analyzed for phospho-GSK3β using anti-GSK3β antibody (BD Biosciences) for capture and antiphospho-GSK3α/β antibody (Cell Signaling) for detection. IC₅₀ values were obtained by fitting data to Eq. 2.

For Western blot analysis of various substrates of Akt phosphorylation, BT474 cells were treated with GSK690693 at concentrations ranging from 10 μmol/L to 1 nmol/L. Five hours later, the cells were then lysed in radioimmunoprecipitation assay buffer (RIPA) containing protease and phosphatase inhibitors. The lysates were diluted with sample loading buffer (Invitrogen), and 10 μg total protein were run on 4% to 20% SDS-PAGE gels. The samples were then transferred to polyvinylidene difluoride membranes and probed with the antibodies for phospho-Forkhead (FKHR/FKHRL1), phospho-p70S6K, phospho-PRAS40, phospho-GSK3α/β, phospho-EGFR, phospho-ErbB2, phospho–extracellular signal-regulated kinase (ERK), and phospho-Akt at 1:1,000 dilution. Secondary, fluorescently labeled antibodies were used at 1:5,000 dilutions. Tubulin was used as a loading control. All phosphosite specific antibodies were obtained from Cell Signaling, except phospho-PRAS40 (BioSource) and phospho-ERK (Santa Cruz Biotechnology). Antitubulin antibody was obtained from Sigma. All blots were imaged on a LiCor Odyssey instrument as directed by the manufacturer.

FOXO3A–green fluorescent protein translocation assay. FOXO3A was cloned from human lung and brain cDNA. The FOXO3A cDNA (amino acids 1–664) was subcloned into the vector pEGFP-N1 (Clontech). The resulting construct had the green fluorescent protein (GFP) fused to the carboxyl terminus of the FOXO3A sequence. U2OS cells stably transfected with pEGFP-N1 FOXO3a were plated in 6-well polystyrene plates at 250,000 cells per well and allowed to adhere overnight. GSK690693 was added using a 10-fold serial dilution at concentrations ranging from 10 μmol/L to 1 nmol/L (and untreated control). After 90 min in culture, the cells were analyzed by fluorescence microscopy.

Proliferation assay. Cell proliferation assays were performed for a number of cell lines as described earlier with some modifications (20). For these assays, cells were plated at densities that allowed untreated cells to grow logarithmically during the course of a 3-d assay. Briefly, cells were plated in 96- or 384-well plates in culture medium containing 10% fetal bovine serum and incubated overnight at 37° C in 5% CO₂. Cells were then treated with GSK690693 (ranging from 30 μmol/L–1.5 nmol/L) and incubated for 72 h. Cell proliferation was measured using the CellTiter Glo (Promega) reagent according to the manufacturer's protocol. Data were analyzed using the XLFit (IDBS Ltd.) curve-fitting tool for Microsoft Excel. IC₅₀ values were obtained by fitting data to Eq. 2.

Reverse-phase protein microarray. Protein arrays were constructed as described previously (21, 22). Briefly, serially diluted protein lysates from cell growing logarithmically were printed in duplicate onto nitrocellulose-coated glass slides. Total protein was quantified in selected arrays that were stained with Sypro Ruby Protein Blot Stain (Molecular Probes) according to the manufacturer's instructions. The lysate arrays were incubated for at least 5 h in blocking solution [1g I-block (Tropix); 0.1% Tween 20 in 500 mL PBS] at room temperature with constant rocking. Blocked arrays were stained with phospho-Akt (Ser⁴⁷³) antibody on an automated slide stainer (Dako Cytomation) using the Catalyzed Signal Amplification System kit according to the manufacturer's recommendation (CSA; Dako Cytomation). Stained slides were scanned individually on a UMAX PowerLook III scanner (UMAX) at 600 dpi and saved as TIF files in Photoshop 6.0 (Adobe). The TIF images for antibody-stained slides and Sypro-stained slide images were analyzed with MicroVigene image analysis software, version 2.200 (Vigenetech) and Microsoft Excel 2000 software. Images were imported into Microvigene, which performed spot finding, local background subtraction, replicate averaging, and total protein normalization, producing a single value for each sample at each end point.

Pharmacodynamic assay. Female SCID mice (3 mice per dosing group) with BT474 xenograft tumors (200–400 mm³) were dosed i.p. with 10, 20, or 40 mg/kg GSK690693 in 4% DMSO/40% HP-β-CD in water (pH 6.0). Four hours after administration of GSK690693, tumors were harvested. Frozen tumor samples were homogenized in RIPA lysis buffer containing protease and phosphatase inhibitors. Equal amounts of protein lysate (50 μg per sample) were analyzed by Western blots using antibodies against total GSK3α/β (Invitrogen) and phospho-GSK3β (Ser⁹; Cell Signaling). Time course of the pharmacodynamic effect was measured using a 20 mg/kg single i.p. dose of GSK690693. Blood glucose, plasma insulin, phospho-GSK3β in BT474 tumors, and drug concentration in blood and tumor were analyzed at baseline, 1, 2, 4, 8, and 24 h after compound administration. Blood glucose was measured from tail vein nicks using an Accu-Chek Compact glucometer (Roche Diagnostics). Circulating insulin levels were measured in plasma from mice using an ELISA kit (Crystal Chem Inc.).

Drug concentration analysis in blood and tumor tissue. Blood samples were hemolyzed with high performance liquid chromatography (HPLC) grade water before analysis. Tumor samples were mixed with HPLC grade water and homogenized using a probe-type homogenizer. Samples were assayed for GSK690693 using protein precipitation with acetonitrile followed by HPLC/MS/MS analysis using positive TurboIonSpray ionization or atmospheric pressure chemical ionization. This assay method was sufficiently accurate and precise for the determination of GSK690693 with a limit of detection as low as 10.0 ng/mL using 25 μL of whole blood or 50 μL of tumor homogenate. Tumor homogenate concentrations (ng/mL homogenate) were converted to tissue concentrations (ng/g tissue) by multiplying the homogenate concentrations with the dilution factor, which resulted from the addition of water during homogenization of the samples. This method assumes complete recovery of the analyte from the sample matrix during bioanalysis.

Tumor xenografts. Tumors were initiated by injection of tumor cell suspension (HCC1954, MDA-MB-453, and LNCaP) or tumor fragments (BT474, SKOV-3, and PANC1) s.c. in 8- to 12-wk-old CD1 Swiss Nude mice (LNCaP, SKOV-3, and PANC1) or SCID mice (HCC1954, MDA-MB-453, and BT474). When tumors reached a volume of 100 to 200 mm³, mice were randomized and divided into groups of 8 to 12 mice per group. GSK690693 was administered once daily at 10, 20, and 30 mg/kg by i.p. administration. Animals were euthanized by inhalation of CO₂ at the completion of the study. Tumor volume was measured twice weekly by calipers, using the equation: tumor volume (mm³) = (length × width²)/2. Results are reported as % inhibition on day 21 of treatment = 100 × [1-(average growth of the drug-treated population/average growth of vehicle-treated control population)]. Statistical analysis was done using two-tailed t test.

Immunohistochemistry. Before immunohistochemically (IHC) staining the study samples, protocols were optimized and antibodies were validated histologically by reproducing Western blot data from BT474 cells treated with 1 μmol/L GSK690693. Antibody specificity was confirmed by pretreating tissue sections with λ phosphatase to remove phosphorylation sites before application of primary antibody. Four-micrometer paraffin sections were prepared from the human tumor xenograft samples and affixed to glass slides. After tissue deparaffinization, antigens were retrieved using citrate buffer, and endogenous biotin and peroxidase were blocked. Nonspecific antibody binding was addressed through application of appropriate sera. Tissues were incubated with antibody against pPRAS40 (Calbiochem) and pFKHR-L (Cell Signaling) for 1 or 4 h, respectively, followed by a biotinylated goat anti-rabbit IgG secondary antibody. A streptavidin reagent conjugated to a reporter enzyme (peroxidase) was applied to complete the multilayer sandwich method. Immunoreactive areas were visualized using 3,3′-diaminobenzidine (DAB), and the sections were counterstained with hematoxylin before coverslipping with a permanent mounting medium. Areas of immunoreactivity are brown (DAB) and nuclei are blue (hematoxylin). Photomicrographs were taken using Leica DMRXA2 microscope (Leica Microsystems, Inc.) at ×20 magnification. Images were captured on Nuance spectral imaging camera (CRI) and a composite was made in Photoshop 7.01 (Adobe).

Kinase activity of GSK690693. GSK690693 is an ATP-competitive, low nanomolar inhibitor of Akt kinases with IC₅₀ values of 2, 13, and 9 nmol/L for Akt1, 2, and 3, respectively (Table 1). The apparent K_i^*s for full-length Akt1, 2, and 3 were determined as 1, 4, and 12 nmol/L, respectively.

To estimate the selectivity of GSK690693, we measured its ability to inhibit >250 in vitro expressed human protein kinases in either activity assays or binding assays. IC₅₀ values were generated against 95 kinases. Kinases with IC₅₀ values <100 nmol/L are shown in Table 1. GSK690693 is very selective for the Akt isoforms versus the majority of kinases in other families; however, it is less selective for members of the AGC kinase family including PKA, PrkX, and PKC isozymes. Other kinases inhibited by GSK690693 are AMPK and DAPK3 from the CAMK family, and PAK4, 5, and 6 from the STE family (Table 1).

Crystallography and biochemical mechanism of action analysis show that GSK690693 is an ATP competitive inhibitor for the Akt enzymes. We have also discovered that GSK690693 inhibits Akt1 and 2 in a time-dependent and reversible manner with a half life for dissociation (t_1/2) at 38 and 30 minutes for Akt1 and Akt2, respectively. Details of these studies and structure activity relationship of the series will be published in a separate article.⁷

Cellular activity of GSK690693. To measure inhibition of Akt kinase activity in cells by GSK690693, the levels of phosphorylated GSK3β (Ser⁹) from cell lysates were determined using an ELISA. GSK690693 inhibited the phosphorylation of GSK3β in tumor cells with average IC₅₀s ranging from 43 to 150 nmol/L (Table 2). Because GSK3β can be phosphorylated on Ser⁹ by kinases other than Akt, notably PKA (23), other Akt substrates were evaluated as additional measures of intracellular Akt inhibition. The phosphorylation of FKHR/FKHRL1, p70S6K, GSK3α/β, and PRAS40 in BT474 breast tumor cells was also inhibited by GSK690693 in a dose-dependent manner (Fig. 1A). Decreases in phosphorylation are evident for these Akt substrates at concentrations >100 nmol/L. An increase in Akt (Ser⁴⁷³) phosphorylation was observed at all GSK690693 concentrations except the highest (10 μmol/L), as has been described with an Akt inhibitor from a different chemical series (24). No effect on EGFR, ErbB2, and ERK phosphorylation was observed in BT474 cells, suggesting a lack of activity on upstream receptors or mitogen-activated protein kinase (MAPK) pathway. To evaluate the functional consequence of inhibiting Akt substrate phosphorylation in cells by GSK690693, cellular localization of a FOXO3A-GFP fusion protein was determined (Fig. 1B). It has been previously shown that phosphorylation of FOXO3A by Akt results in nuclear accumulation consistent with its role in apoptotic signaling (25). GSK690693 induced accumulation of the FOXO3A fusion protein in the nucleus at concentrations of 1 μmol/L or greater.

The effect of GSK690693 on cell proliferation was evaluated against normal fibroblasts and a variety of tumor cell lines (Table 2). Of the eight breast cancer cell lines tested, five of them had IC₅₀ values <1 μmol/L. The LNCaP (prostate) tumor cell line, which has constitutively active Akt due to the loss of PTEN (26), also showed sensitivity to GSK690693 (IC₅₀ = 147 nmol/L). None of the other cell lines tested was very sensitive to the Akt inhibitor (IC₅₀ >1 μmol/L), including the PTEN-null or mutant tumor cell lines 786-O, H157, MDA-MB-468, and PC-3 (27–30). Normal human foreskin fibroblast (HFF) cells were also insensitive to GSK690693 (IC₅₀ >7 μmol/L). Phospho-Akt levels were higher in cells that were sensitive to GSK690693, although all cell lines with high phospho-Akt levels were not very sensitive, suggesting that other signaling pathways are likely activated in these cells.

Because Akt is an important regulator of cell survival, apoptosis as a mechanism of growth inhibition by GSK690693 was evaluated in LNCaP and BT474 cells after treatment with various concentrations of GSK690693 for 24 or 48 hours, respectively. As measured by histone-complexed DNA fragments, GSK690693 was found to induce apoptosis at concentrations >100 nmol/L in both LNCaP and BT474 cells (Supplementary Fig. S1).

In vivo activity of GSK690693. To investigate the pharmacodynamic effect of GSK690693 in vivo, phosphorylation of GSK3β was measured from tumor lysates. In immune-compromised mice implanted with human breast carcinoma (BT474) xenografts, a single i.p. administration of GSK690693 inhibited GSK3β phosphorylation in a dose-dependent manner (Fig. 2). After a single 20-mg/kg i.p. dose of GSK690693, drug concentrations were measured in blood and BT474 tumor tissue (Fig. 2). Drug concentrations >3 μmol/L (∼1,500 ng/g) in BT474 tumor xenografts correlated with a sustained decrease (>60%) in GSK3β phosphorylation up to 8 hours (Fig. 2).

Because Akt is a critical regulator of insulin signaling, blood glucose and insulin levels were measured in mice after administration of GSK690693. Treatment with GSK690693 resulted in an acute increase in blood glucose that returned to baseline levels as circulating drug concentration decreased (Fig. 2). A rapid elevation of plasma insulin levels was also observed, reaching a maximum of 643 ng/mL 4 hours postdosing compared with the baseline level of 0.73 ng/mL (Fig. 2). The increase in insulin preceded the increase in blood glucose and decreased in parallel with the decrease in circulating drug concentration. Both glucose and insulin levels had declined by 8 hours and returned to baseline by 24 hours (Fig. 2). The treated mice had no sign of clinical distress, and the compound was well-tolerated with once daily administration up to 3 weeks.

GSK690693 was evaluated for its ability to inhibit growth of various human tumor xenografts in immunocompromised mice. Repeated i.p. administration (once daily for 21 days) produced significant antitumor activity in mice bearing established SKOV-3 ovarian, LNCaP prostate, and BT474 and HCC-1954 breast carcinoma xenografts (Fig. 3). Maximal inhibition of 58% to 75% was observed at the end of dosing period with 30 mg/kg/day dose. Daily administration of GSK690693 for 21 days was well-tolerated in mice with <10% body weight change with no overt clinical sign of toxicity. Effect of GSK690693 treatment on phosphorylation of Akt substrates was evaluated in BT474 tumor xenografts. Similar to the reduction of GSK3β phosphorylation observed after a single dose of Akt inhibitor (Fig. 2), IHC analysis of BT474 tumor xenografts after repeat dosing with GSK690693 showed a reduction in phosphorylation of the Akt substrates, PRAS40, and FKHR/FKHRL1 (Fig. 4). No significant change in proliferation (Ki-67) or apoptosis (activated caspase-3) were observed between vehicle and drug-treated samples using IHC (data not shown).

The present study describes the pharmacologic characterization of a novel pan Akt kinase inhibitor, GSK690693, which belongs to the aminofurazan compound class. GSK690693 is ATP competitive and exhibits time-dependent inhibition of the Akt isoforms. Selectivity of GSK690693 was assessed in a broad kinase panel. In addition to Akt, 13 kinases of the >250 unique kinases tested show IC₅₀ values ≤100 nmol/L, suggesting a high degree of kinase selectivity. Most of the kinases inhibited by GSK690693 belong to the AGC kinase family that includes Akt enzymes. AMPK and DAPK3, members of the CAMK family, showed 25- and 40-fold lower potency compared with Akt1. GSK690693 also inhibited members of group II PAK kinase family (31).

Akt1, 2, and 3 is activated in tumor cells by a number of upstream stimuli, such as activation of various receptor tyrosine kinases, PI3K, Ras, or by ligand-induced cell signaling. Akt, in turn, mediates its effect via regulation of a variety of downstream substrates (2). We measured the phosphorylation status of several Akt substrates (Fig. 1; Table 2) as a way to evaluate inhibition of Akt kinase activity in cells. GSK690693 caused a dose-dependent reduction in the phosphorylation state of multiple proteins downstream of Akt such as GSK3β, PRAS40, p70S6K, and FKHR/FKHRL1 in tumor cells. Treatment of tumor cells with GSK690693 led to a dose-dependent increase in the nuclear accumulation of the transcription factor FOXO3A, which shows a functional effect of Akt inhibition by a change in the phosphorylation status of FOXO3A protein (25).

Several recent studies have shown that mTOR inhibition results in an up-regulation of PI3K/Akt/mTOR signaling by relieving the negative p70S6K-mediated feedback inhibition on IRS-1 (32–34). Furthermore, Han et al. (24) described mTORC1-independent regulation of Akt phosphorylation upon treatment with the Akt kinase inhibitor, A-443654. Consistent with these previous studies, GSK690693 treatment resulted in a dose-dependent increase in the phosphorylation of Akt at both Ser⁴⁷³ (Fig. 1) and Thr³⁰⁸ (data not shown). Phosphorylation of proteins that are either upstream of Akt or involved in Akt-independent pathways, e.g., EGFR, ErbB2, MAPK, etc., was not altered on treatment with GSK690693, suggesting an Akt-specific feedback mechanism. Up-regulation of Akt activity after treatment with a mTOR inhibitor has been hypothesized to attenuate the antiproliferative effect of these agents due to mTOR-independent effects of Akt (32, 35). However, as shown by the reduction in the phosphorylation of multiple Akt substrates, GSK690693 effectively inhibited Akt kinase activity in cells (Fig. 1) regardless of any feedback hyperphosphorylation of Akt.

GSK690693 showed an antiproliferative effect in a subset of tumor cell lines tested, suggesting that Akt is one of several factors effecting the growth and survival of tumor cells. Our data show that the lack of antiproliferative effects were not due to a lack of Akt inhibition in the insensitive tumor cells, as the IC₅₀ for inhibition of GSK3β phosphorylation was similar in both sensitive and insensitive cell lines (Table 2). This observation is not surprising given the abundance of genetic alterations associated with tumorigenicity and redundancy in cellular signaling. Akt1 and 2 siRNA inhibited the proliferation of BT474 and LNCaP cells, with no significant effect on SKOV-3 and OVCAR-3 cells (Supplementary Fig. S2). These results are consistent with the effects observed with GSK690693, although others have shown ∼50% decrease in OVCAR-3 cell proliferation with Akt2 siRNA (36). The differences between our results and earlier reports can be due to differences in reagents and/or experimental conditions. Furthermore, potential off target effects of siRNAs have also been observed in some cases (37, 38). The correlation between in vitro sensitivity and activity in tumor xenografts with GSK690693 was reasonable with the exception of SKOV-3, which seems to be more sensitive as a xenograft. These differences can be attributed to differences in availability of growth factors and cellular signaling between cell culture and in vivo anchorage-independent growth as well as differences in the drug exposure. We are currently investigating the molecular markers of sensitivity and resistance to GSK690693. Combination of GSK690693 with other targeted agents and standard chemotherapeutic agents is ongoing to further explore the full therapeutic potential of Akt inhibition.

In addition to the potent inhibition of Akt kinases, GSK690693 also inhibit novel PKCs (η, 𝛉, δ, and ε); PKCβ1; PAK-4,5,6; PKA; PKG1β; and PrkX, which can potentially contribute to the observed antitumor effect. Several PKC isozymes have been shown to play an important role in cell proliferation and tumor growth; however, certain isozymes are also involved in differentiation and inhibition of proliferation (39). PKCα, βII, 𝛉,ι, and η are pro-oncogenic in various tissues, whereas PKCδ likely has tumor inhibitory potential in most tissues, except in the brain (40). Recent studies have elucidated the role of PAK kinases in cellular signaling through various oncogenes as well as altered expression of PAK1 and PAK4 in various cancers (31). Modulation of PKA levels by antisense has also been shown to effect cell proliferation and transformation; however, there are conflicting data in the literature on the role of PKA in human neoplasms (41, 42).

In addition to its role in cell survival and proliferation, the Akt pathway, particularly Akt2, is an integral part of the insulin signaling pathway (11, 13). In vitro experiments in tumor cells suggest that inhibition of Akt1 and Akt2 is necessary for antitumor effects (15). Therefore, it is expected that interruption of insulin signaling will result in hyperglycemia during treatment with drugs that are pan-Akt inhibitors. Treatment with GSK690693 resulted in acute hyperglycemia with blood glucose levels returning to baseline 8 to 10 hours after drug administration (Fig. 2). The concomitant-measured increase in circulating insulin observed after GSK690693 treatment is likely a homeostatic response to elevated glucose levels. Consistent with this is the observation that insulin level returned to baseline as the blood glucose levels declined. Daily administration of GSK690693 at the doses used in this study was well-tolerated and had no overt clinical effect in mice such as loss of body weight.

In vivo administration of GSK690693 resulted in decreased phosphorylation of Akt substrates in tumor xenografts (Figs. 2 and 4) and normal tissue (data not shown), indicating a clear pharmacodynamic effect in mice. Furthermore, daily administration of the compound inhibited growth of multiple human tumor xenografts in mice (Fig. 3), consistent with the antiproliferative effects observed in tissue culture (Table 2). The overall pharmacologic profile of GSK690693 is consistent with a selective Akt kinase inhibitor and it is currently being evaluated in clinical trials on human cancer patients.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

All authors are present or former employees of GlaxoSmithKline.