The dysregulation of PI3K/AKT/mTORC1 signalling and/or hyperactivation of MYC are observed in a high proportion of human cancers, and together they form a ‘super signalling’ network mediating malignancy. A fundamental downstream action of this signalling network is up-regulation of ribosome biogenesis and subsequent alterations in the patterns of translation and increased protein synthesis, which are thought to be critical for AKT/MYC-driven oncogenesis. We have demonstrated that AKT and MYC cooperate to drive ribosomal DNA (rDNA) transcription and ribosome biogenesis, with AKT being essential for rDNA transcription and in vitro survival of lymphoma cells isolated from a MYC-driven model of B-cell lymphoma (Eμ-Myc) [Chan JC et al., (2011) Science Signalling 4, ra56]. Here we show that the allosteric AKT inhibitor MK-2206 rapidly and potently antagonizes rDNA transcription in Eμ-Myc B-cell lymphomas in vivo, and this is associated with a rapid reduction in indicators of disease burden, including spleen weight and the abundance of tumour cells in both the circulation and lymph nodes. Extended treatment of tumour-bearing mice with MK-2206 resulted in a significant delay in disease progression, associated with increased B-cell lymphoma apoptosis. Our findings suggest that malignant diseases characterized by unrestrained ribosome biogenesis may be vulnerable to therapeutic strategies that target the PI3K/AKT/mTORC1/MYC growth control network.
terminal deoxynucleotide transferase dUTP nick end labelling
The PI3K/AKT/mTORC1 signalling hub plays an essential role in malignant transformation [1-5], and is dysregulated in many cancers, with multiple components of this pathway being shown to act as oncogenes or tumour suppressors [3, 6-8]. Most studies examining the mechanism(s) by which this signalling hub contributes to malignancy have focused on the processes of pro-survival, cell-cycle progression, angiogenesis and metabolic rate. However, it is becoming increasingly apparent that the downstream actions of this pathway in regulation of ribosome biogenesis and translation are essential for its oncogenic effects [9-11]. The oncogene MYC is dysregulated in 15–20% of human malignancies, and, like the PI3K/AKT/mTORC1 pathway, recent studies have demonstrated that MYC plays a critical role in regulating rDNA transcription [12-15]. Indeed PI3K/AKT/mTORC1 and MYC signalling may cooperate, converging at a number of steps to establish master control of ribosome biogenesis, and thus protein synthesis [4, 5, 11]. The availability of sufficient functional ribosomes is a fundamental rate-limiting step for growth and proliferation in mammalian cells [16, 17], and diseases, such as cancer, that are associated with uncontrolled growth and proliferation are frequently characterized by increased ribosome synthesis . Importantly for this study, increased ribosome synthesis and subsequent modulation of the efficiency of translation of key cell proliferation, growth and survival proteins is an essential mechanism by which MYC promotes tumorigenesis [10, 18-22].
We recently demonstrated that RNA polymerase I transcription may be selectively targeted by the small molecule CX-5461 to treat MYC-driven B-cell lymphoma . This model of spontaneous B-cell lymphoma is a prototypical example of a tumour driven by uncontrolled cell growth, as evidenced by the increased rates of rRNA and protein synthesis and the larger size of B cells from equivalent stages of development [22, 23]. Given the key role of AKT in MYC-dependent rDNA transcription and ribosome biogenesis , we hypothesized that specific inhibition of AKT using the allosteric inhibitor MK-2206, currently in phase I/II clinical trials , would antagonize MYC-driven rRNA synthesis in the Eμ-MYC B-cell lymphoma model and delay lymphoma progression. MK-2206 [8-(4-(1-aminocyclobutyl)phenyl)-9-phenyl-[1, 2, 4]triazolo[3,4-f] [1, 6]naphthyridin-3(2H)-one] is a highly specific inhibitor of AKT, with low nanomolar IC50 values for all three isoforms of AKT (AKT1, 8 nM; AKT2, 12 nM; AKT3, 65 nM), and 100-fold selectivity for AKT over a large panel of other protein kinases . By specifically targeting the pleckstrin homology domain of AKT isoforms, MK-2206 prevents recruitment of the kinase to the plasma membrane and subsequent activating phosphorylation events at the Thr308 and Ser473 residues . MK-2206 at doses between 120 and 480 mg·kg−1 has demonstrated therapeutic properties in pre-clinical cancer models in vivo, and robustly inhibited AKT activity and tumour cell growth in ovarian, lung, breast and neuroblastoma cancer xenograft models [25-28]. In this study, MK-2206 rapidly and potently antagonized rDNA transcription, restored spleen weight to normal levels, and reduced the number of tumour cells in the circulation and lymph nodes. Extended inhibition of AKT activity and rRNA synthesis by MK-2206 treatment of mice with Eμ-Myc B-cell lymphoma caused a significant delay in disease progression associated with increased tumour cell apoptosis.
AKT signalling is required for rRNA gene transcription in vivo
To establish the ability of MK-2206 to inhibit AKT activity and rDNA gene transcription in vivo, C57Bl/6 mice were transplanted intravenously with MSCV GFP Eμ-Myc B-cell lymphoma cells. Once disease was established (14 days post-injection), mice were given a single dose of vehicle (30% captisol, 0.1 mL·kg−1) or 200 mg·kg−1 MK-2206, which was established as the maximum tolerated dose of MK-2206 suitable for extended treatment of these mice (Fig. S1). Axillary lymph nodes were removed at 6, 24 and 48 h post-treatment. Western analysis confirmed that a single dose of MK-2206 robustly inhibited AKT activity in the lymph nodes within 6 h, and inhibition was maintained at 24 h, as measured by the abundance of AKT phosphorylated at Ser473 (Fig. 1A). In contrast, total AKT abundance was not altered (Fig. 1A). Furthermore, AKT inhibition was associated with suppression of 47S rDNA gene transcription, which was evident at 6 h and statistically significant by 24 h (45% decrease; Fig. 1B).
Acute administration of MK-2206 inhibits Eμ-Myc B-cell lymphoma progression
Following a single dose of MK-2206 at 6, 24 and 48 h, the spleen, axillary lymph nodes and cardiac blood were examined for disease. An enlarged spleen is characteristic of Eμ-Myc B-cell lymphoma-bearing mice  and reflects the disease burden. Spleen weight was significantly reduced 24 and 48 h post-administration of MK-2206 (Fig. 2A). MK-2206 treatment also prevented the increase in GFP-expressing circulating lymphoma cells after 24 h of treatment, and this population was decreased at 48 h compared to time-matched vehicle controls (Fig. 2B). MK-2206 treatment also prevented the typical expansion of the white blood cell population (Fig. 2C) and the increase in lymph node size observed in control mice at 24 and 48 h post- treatment (Fig. 2D).
MK-2206 treatment induces apoptosis of Eμ-Myc B-cell lymphoma cells in vivo
We have previously demonstrated that inhibition of AKT activity leads to apoptosis of Eμ-Myc B-cell lymphoma cells in vitro . To determine whether the MK-2206-mediated delay of lymphoma progression in vivo also correlated with apoptosis, TUNEL staining was performed on fixed inguinal lymph nodes and spleens from mice treated with vehicle (control) or MK-2206 for 24 h. MK-2206 increased the proportion of apoptotic (TUNEL-positive) cells in the lymph nodes (Fig. 2E) and spleen (Fig. S2A), compared to the small proportion of apoptotic cells observed in the control mice, which is consistent with the known pro-apoptotic effects of MYC over-expression [30, 31] in the Eμ-Myc B-cell lymphoma model. This demonstrates that inhibition of AKT signalling promotes apoptosis of the lymphoma cells in vivo, which most likely accounts for the delay in lymphoma progression observed.
MK-2206 treatment prolongs survival and delays Eμ-Myc B-cell lymphoma progression in vivo
To determine the efficacy of repeated dosing with MK-2206 on Eμ-Myc B-cell lymphoma progression, C57Bl/6 mice bearing GFP Eμ-Myc tumours were treated at day 8 post-tumour inoculation with either vehicle (30% captisol, 0.1 mL·kg−1) or MK-2206 (200 mg·kg−1) three times weekly over the course of the experiment. MK-2206 treatment significantly prolonged survival of mice relative to the control group (Fig. 3A) and, as was observed following single-dose treatment (Fig. 2), this was associated with a significant reduction in both spleen weight (Fig. 3B) and the population of circulating tumour cells (Fig. 3C) compared to the control group. While MK-2206 treatment prolonged survival, the mice eventually succumbed to nodal and extra-nodal disease (Fig. 3D). The extended survival of these mice compared to the control group is consistent with our observation in the single-dose experiment (Fig. 2) that MK-2206 treatment delays lymphoma progression. In addition, TUNEL staining showed robust increases in apoptosis in both the inguinal lymph nodes (Fig. 3E) and spleens (Fig. S2B) of mice repeatedly dosed with MK-2206, compared to control mice. Furthermore, MK-2206 treatment also induced an increase in the apoptotic sub-G1 cell population in the axillary lymph nodes (Fig. S2C), as determined by FACS analysis. These data are consistent with induction of apoptosis as the major therapeutic mechanism through which pharmacological AKT inhibition delays lymphoma progression and prolongs survival in this B-cell lymphoma model.
Dysregulation of ribosome biogenesis is no longer a passive indication of malignant transformation but is now a realistic therapeutic target [22, 32, 33]. Our previous in vitro analyses found a critical role for AKT in the control of rDNA transcription at multiple levels, including RNA polymerase I transcription elongation and/or processing of the rRNA, together with its control of mTORC1-dependent RNA polymerase I transcription initiation . Consequently, inhibiting AKT activity reduces rRNA synthesis more rapidly and potently than the mTORC1-specific inhibitor rapamycin alone . Importantly for the current study, this was most notable in MYC-driven B-cell lymphoma cells in vitro, where reduced AKT activity resulted in potent inhibition of rRNA synthesis and cell death, but mTORC1 inhibition had little effect on either . Here we show that this reliance of rRNA synthesis and cell survival on AKT activity in vitro may be targeted in an in vivo model of MYC-driven B-cell lymphoma, and is associated with induction of apoptosis of the lymphoma cells, delayed disease progression and thus prolonged survival. This therapeutic response is similar to the protection observed in neuroblastoma xenografts treated with MK-2206 . Taken together with our previous study demonstrating selective apoptosis of Eμ-Myc B-cell lymphomas by inhibition of RNA polymerase I , these data raise the possibility that at least part of the mechanism by which AKT inhibition prolongs survival of these mice is through inhibition of RNA polymerase I transcription.
Inhibition of ribosome biogenesis may be an important determinant of the anti-cancer effects of PI3K/AKT/mTORC1 pathway inhibitors, particularly in cancers driven by dysregulation of the critical cell growth regulatory pathways downstream of MYC, PI3K and RAS. For example, inhibition of mTORC1 using everolimus resulted in a similar robust improvement in overall survival in Eμ-Myc B-cell lymphoma (1.3–2-fold)  compared with MK-2206 (1.3-fold); however, this was associated with increased cellular senescence rather than apoptosis. Given the lack of effect of mTORC1 inhibition on rRNA synthesis in lymphoma cells , this difference in cellular response (senescence versus apoptosis) may be due to everolimus-induced inhibition of translation of specific mRNAs required to antagonize senescence .
Furthermore, it is possible that the ability of MK-2206 to target ribosome biogenesis and function at multiple steps, including long-term inhibition of mTORC1 , may be an important element in its ability to drive cells down the apoptotic pathway. Consistent with this hypothesis, there is emerging evidence that targeting multiple members of signalling pathways or key cellular process may have additive or even synergistic effects on cancer cell proliferation and/or survival. It is becoming clear that approaches that provide more robust inhibition of the pathway favour cytotoxic over cytostatic responses . For example, targeting multiple members of the PI3K pathway such as PI3K, mTOR and mTORC1 resulted in synergistic treatment of hepatocellular carcinoma . Similarly, targeting AKT and mTORC1 in combination therapy of mice bearing neuroblastoma xenografts resulted in enhanced therapeutic efficacy , and combined inhibition of the critical growth control pathways PI3K/mTOR and RAS cooperates, to inhibit ovarian cancer cell growth in vitro and in vivo  (K. E. Sheppard and R. B. Pearson, unpublished results).
While specific targeting of RNA polymerase I  and AKT signalling may prolong survival in the Eμ-Myc model of B-cell lymphoma, the mice succumb to disease in both cases over time. This is a common outcome in response to the majority of cancer-targeted therapies; despite profound improvements in survival by targeting oncogenes, resistance develops in many patients [38, 39]. Inhibition of AKT results in potent inhibition of rRNA synthesis by targeting rDNA transcription initiation and elongation/processing . It is thus possible that combinations of PI3K/AKT/mTOR and RNA polymerase I inhibitors may enhance the tumour response by cooperative inhibition of ribosome biogenesis. In addition, it will be important to further define the mechanisms by which inhibition of PI3K/AKT antagonizes lymphoma cell survival, as extra-ribosomal targeting of cell survival pathways may also lead to cooperative therapeutic effects. Indeed, AKT plays a key role in promoting cell survival via a number of diverse mechanisms including regulation of p53 by MDM2 as well as pro- and anti-apoptotic members of the Bcl-2 family of proteins such as the BH3-only protein BAD [40-43]. Disruption of these pathways may play a significant role in MK-2206-mediated apoptosis of MYC-driven lymphoma cells, and it will be important to test this hypothesis in future in vivo studies.
In conclusion, our findings raise the exciting possibility that malignant diseases driven by dysregulation of key controllers of cell growth, including the MYC, PIK3CA and RAS oncoproteins, may be vulnerable to therapeutic strategies that target AKT signalling.
All animal experiments were performed according to protocols approved by the institutional animal experimentation ethics committee.
MK-2206 was a gift from Merck (Darmstadt, Germany). MK-2206 was dissolved using sonication (50 Hz) in 30% w/v captisol (CyDEX, La Jolla, CA, USA) prepared in sterile water.
Preparation and transplantation of Eμ-Myc B-cell lymphoma into C57Bl/6 recipient mice
Eμ-Myc B-cell lymphoma cells (clone 4242) derived from the lymphoma of an Eμ-Myc mouse and transduced with a MSCV GFP construct were generated as described previously . Cells were thawed, washed (5 min, room temperature) and resuspended in phosphate-buffered saline prior to intravenous injection into recipient C57Bl/6 mice (~ 2 × 105 cells/mouse). Disease was monitored by determining white blood cell counts from peripheral whole blood collected into 10 mm EDTA, and analysed using the Adiva120 hematology system (GMI, Minneapolis, MN, USA) as described previously .
Acute dose and repeated-dose therapy
Mice were randomized into two groups: vehicle control or MK-2206-treated. For the acute dose therapy, 15 mice per group were treated with 30% captisol (0.1 mL·kg−1) or MK-2206 (200 mg·kg−1) by oral gavage, and the spleens, lymph nodes and circulating white blood cells were collected after 6, 24 or 48 h. For the repeated-dose therapy, mice were treated with 30% captisol (0.1 mL·kg−1) or MK-2206 (200 mg·kg−1) by oral gavage three times per week until an experimental end point was reached. At the end point, the spleens, lymph nodes and circulating blood were collected.
Spleens and inguinal lymph nodes were fixed in 10% neutral buffered formalin, embedded in paraffin wax, serially sectioned and processed for haematoxylin and eosin staining or TUNEL analysis using a Millipore Apoptag peroxidase in situ apoptosis kit (S7100, Billerica, MA, USA) as described previously . Sections were analysed using the an Olympus (Center Valley, PA, USA) BX-51 microscope at 40 × magnification, and quantification of TUNEL staining was performed using MetaMorph microscopy automation and image analysis software (Molecular Devices, Sunnyvale, CA, USA).
Cell suspensions from axillary lymph nodes
Cell suspensions were generated from the axillary lymph nodes by grinding the tissue, filtering the cells through a 0.7 μm filter, and washing with 2% fetal bovine serum in phosphate-buffered saline. Cell number was determined using a Z2 counter (Beckman Coulter, Brea, CA, USA), and a proportion of cells were used for FACS analysis of GFP/B220 expression or propidium iodide staining, and for extraction of protein or RNA.
FACS analysis for GFP/B220 expression
White blood cells isolated using red blood cell lysis buffer (144 mM NH4Cl, 17 mM Tris/HCl, pH 7.65) and 1 × 106 axillary lymph node cells were analysed for GFP and B220 + (CD45R) surface marker expression using B220-APC (BD Pharmigen, Franklin Lakes, NJ, USA) and a FACS Canto II flow cytometer (BD Pharmigen). The results were analysed using fcsexpress software (De Novo, Los Angeles, CA, USA).
Protein was extracted from axillary lymph node cells using SDS lysis buffer (0.5 mm EDTA, 20 mm HEPES, 2% w/v SDS, pH 7.9), boiling at 95 °C, shearing with a 26 gauge needle, and centrifugation at 15 700 g for 10 min at room temperature. The protein concentration was determined using the DC Protein Assay (Bio-Rad, Hercules, CA, USA) with bovine serum albumin as a standard according to the manufacturer's instructions. Equal concentrations of protein were separated by SDS/PAGE, transferred to poly(vinylidene difluoride membrane and immunoblotted with antibodies against phospho-AKT S473 (Cell Signaling, #4058, Danvers, MA, USA), AKT (Cell Signaling, #9272) or tubulin (Sigma-Aldrich, T9026, St. Louis, MO, USA) and the respective horseradish peroxidase-conjugated antibodies against rabbit or mouse IgG (Bio-Rad). Protein abundance was visualized using a Western Lighting Plus enhanced chemiluminescence kit (Perkin Elmer, Waltham, MA, USA) and X-ray film (Kodak, Rochester, NY, USA). Quantification was performed using imagequant TL software (GE Healthcare, Little Chalfont, Bucks, UK).
Quantitative real-time PCR analysis for rDNA transcription
RNA was extracted from axillary lymph node cells using a Bioline Isolate RNA kit with addition of a 32P-labelled riboprobe to determine RNA recovery. cDNA was prepared from RNA equivalent to equal cell number using SuperScript III (Invitrogen, Carlsbad, CA, USA) and random hexamer primers (Promega, Fitchburg, WI, USA) according to the manufacturers' instructions. The abundance of 47S rRNA 5′ external transcribed spacer (5′ETS) was measured by quantitative real-time polymerase chain reaction using Fast SYBR Green Master Mix (Applied Biosystems, Foster City, CA, USA) and an Applied Biosystems StepOne Plus quantitative real-time-PCR machine according to the manufacturer's instructions. The primer sequences used were 5′-CCAAGTGTTCATGCCACGTG-3′ (5′ETS Forward) and 5′-CGAGCGACTGCCACAAAAA-3′ (5′ETS Reverse).
Propidium iodide analysis of the sub-G1 cell population
Axillary lymph nodes (1 × 106 cells) from the repeated-dose experiment were incubated in ice in 95% ethanol for 30 min, and stained with 50 μg·mL−1 propidium iodide and 0.1 mg·mL−1 RNase A in phosphate-buffered saline supplemented with 5% v/v fetal bovine serum. Samples were analysed on a FACS Canto II flow cytometer (BD Pharmigen), and the proportion of cells in the sub-G1 region was determined using fcsexpress software.
The survival curve data were analysed using Mantel–Cox and Gehan–Breslow–Wilcoxin tests using graphpad prism software (GraphPad, La Jolla, CA, USA). All the remaining data were assessed using Student's t test (two-tailed) using graphpad prism Software.
We would like to thank Merck for supplying MK-2206. This work was supported by National Health and Medical Research Council of Australia project grants 1043884, 251608, 566702, 166908, 251688, 509087, 400116, 400120 and 566876. The researchers were funded by National Health and Medical Research Council of Australia fellowships (R.D.H., R.B.P. and R.W.J.), the Cancer Council of Victoria (Sir Edward Weary Dunlop Fellowship to G.A.M.) and the Leukaemia Foundation of Australia (PhD scholarship to J.R.D.). We thank Jeannette Schreuders, Susan Jackson and Rachael Walker for animal technical assistance.