AKT1 mutations in bladder cancer: identification of a novel ... - Nature

Oncogene (2010) 29, 150–155

& 2010 Macmillan Publishers Limited All rights reserved 0950-9232/10 $32.00 www.nature.com/onc

SHORT COMMUNICATION

AKT1 mutations in bladder cancer: identification of a novel oncogenic mutation that can co-operate with E17K JM Askham1, F Platt1, PA Chambers2, H Snowden2, CF Taylor2 and MA Knowles1 1 Cancer Research UK Clinical Centre, Leeds Institute of Molecular Medicine, St James’s University Hospital, Leeds, UK and 2Cancer Research UK Genome Variation Laboratory, St James’s University Hospital, Leeds, UK

The phosphatidylinositol-3-kinase (PI3 kinase)-AKT pathway is frequently activated in cancer. Recent reports have identified a transforming mutation of AKT1 in breast, colorectal, ovarian and lung cancers. We report here the occurrence of this mutation in bladder tumours. The AKT1 G49A (E17K) mutation was found in 2/44 (4.8%) bladder cancer cell lines and 5/184 (2.7%) bladder tumours. Cell lines expressing mutant AKT1 show constitutive AKT1 activation under conditions of growth factor withdrawal. We also detected a novel AKT1 mutation G145A (E49K). This mutation also enhances AKT activation and shows transforming activity in NIH3T3 cells, though activity is weaker than that of E17K. Enhanced activation of AKT1 when E17K and E49K mutations are in tandem suggests that they can co-operate. Oncogene (2010) 29, 150–155; doi:10.1038/onc.2009.315; published online 5 October 2009 Keywords: AKT1; mutation; bladder cancer

Introduction The phosphatidylinositol-3-kinase (PI3 kinase)-AKT pathway promotes cell growth, proliferation and survival, and is the most frequently mutated pathway in human cancer (Brugge et al., 2007). Activated receptor tyrosine kinases recruit the p85/p110 PI3 kinase complex to the membrane through the p85 regulatory subunit either directly or through insulin receptor substrate adapter proteins. The active p110 (catalytic) subunit then phosphorylates phosphatidylinositol-4,5-bisphosphate (PIP2) to phosphatidylinositol-3,4,5-trisphosphate (PIP3). PTEN (phosphatase and tensin homologue deleted on chromosome 10) catalyses the reverse reaction. PIP3 recruits protein-dependent kinase 1 (PDK1) and AKT to the plasma membrane where AKT is phosphorylated at Thr308 by PDK1 and at Ser473 by mTORC2 (Sarbassov et al., 2005). AKT1 can also be phosphorylated at Ser473 by DNA-PK in the nucleus Correspondence: Professor MA Knowles, Cancer Research UK Clinical Centre, Leeds Institute of Molecular Medicine, St James’s University Hospital, Beckett Street, Leeds LS9 7TF, UK. E-mail: [email protected] Received 25 July 2008; revised 24 June 2009; accepted 31 July 2009; published online 5 October 2009

(Bozulic et al., 2008). AKT is an evolutionarily conserved kinase, also known as protein kinase B. There are three members of the AKT family (AKT1-3), encoded by separate genes, but with over 80% amino-acid sequence identity. AKT occupies a key regulatory node in the PI3K pathway, below which the pathway branches significantly to influence a wide range of cellular processes that promote cell cycle progression, cell growth, cellular energy metabolism and resistance to apoptosis. Genes encoding many of the components of the PI3 kinase pathway are targeted by germline and somatic mutations, amplifications, rearrangements, over-expression, methylation and aberrant splicing (Hennessy et al., 2005; Kumar and Hung, 2005; Manning and Cantley, 2007), which generally result in increased AKT activity and aberrantly elevated downstream signalling. In bladder cancer, mutations have been identified in PIK3CA (Lopez-Knowles et al., 2006) and TSC1 (which lies downstream of AKT in the PI3 kinase-AKT-mTOR branch of the PI3 kinase pathway) (Knowles et al., 2003; Pymar et al., 2008), and loss of heterozygosity, homozygous deletion and inactivating mutations of PTEN have been found (Cappellen et al., 1997; Cairns et al., 1998; Aveyard et al., 1999; Wang et al., 2000). Given the central function of activated AKT1 in tumorigenesis, it was somewhat surprising that mutations were not detected in this gene until recently. A single activating point mutation in AKT1 has now been described in breast, colorectal and ovarian cancers (Carpten et al., 2007; Bleeker et al., 2008; Kim et al., 2008) and squamous cell carcinoma of the lung (Malanga et al., 2008). This mutation, G49A (E17K), in the pleckstrin homology (PH) domain of the protein results in its recruitment to the plasma membrane in the absence of PI3 kinase signalling and confers transforming activity in vitro and in vivo (Carpten et al., 2007). As mutations in other components of the PI3 kinase pathway are found in bladder cancer, we screened bladder tumours and cell lines for this mutation. Results and discussion Initially, we screened a panel of 42 bladder cell lines by PCR and sequencing of AKT1 exon 4 and found that 2/42 (4.8%) cell lines (KU19-19 and MGH-U3) contained the G49A (E17K) mutation (Supplementary Figure 1). In both cell lines, the mutation was heterozygous.

AKT1 mutations in bladder cancer JM Askham et al

151

During sequencing of exon 4, we identified a second heterozygous mutation, G145A (E49K), in KU19-19 (Supplementary Figure 2), suggesting the possibility of other activating mutations in the PH domain. The mutation screen was then extended to a panel of 184 well-characterized bladder tumours. These were analysed by a combination of high-resolution melting curve analysis to screen for mutations in AKT1 exon 4 and pyrosequencing as a specific assay for the E17K mutation (Supplementary Methods). Positive results were confirmed by PCR and sequencing (Supplementary Figure 1). E17K was the only mutation detected (5/184; 2.7%). This frequency is within the range published for breast, ovarian, colorectal and lung cancers (Carpten et al., 2007; Bleeker et al., 2008; Kim et al., 2008; Malanga et al., 2008). The mutation was not detected in DNA extracted from blood samples from the corresponding patients. Four of the five tumours were heterozygous for the mutation and one homozygous. Mutant tumours included low-grade, noninvasive (Ta grade 2) and muscle invasive (T2 grade 3) tumours. Mutation and loss of heterozygosity status for several genes that are commonly mutated in bladder cancer is known for this tumour series (Platt et al., 2009). AKT1 mutation was mutually exclusive with respect to PIK3CA mutation and PTEN loss of heterozygosity, but not FGFR3 and TSC1 mutation. Analysis of a much larger series of tumours will be needed to allow conclusions to be drawn about these relationships. The E17K substitution in the PH domain of AKT1 allows localization of the protein to the plasma membrane in the absence of upstream signalling (Carpten et al., 2007). Membrane-associated AKT is activated by phosphorylation at Thr308 by PDK1 and at Ser473 by mTORC2. Consistent with this, while phospho-AKT levels (Thr308) were markedly reduced by growth factor withdrawal in telomerase-immortalized normal human urothelial cells (TERT-NHUC), levels remained the same in KU19-19 and MGH-U3 (Figure 1a). Specific examination of AKT1 showed that phosphorylation at both Ser473 and Thr308 was elevated in growth factor-deprived KU19-19 and MGH-U3 compared with TERT-NHUC (Figure 1b), showing complete activation of AKT1. Although this experiment does not exclude the possibility of AKT1 activation through other means in these tumour cell lines, the results are entirely consistent with the published effects of the E17K mutation. To investigate the possible functional importance of the novel E49K mutation, AKT1 cDNA was amplified from KU19-19 cells by RT–PCR and cloned. During this process, it became apparent that KU19-19 cells express three AKT1 transcripts, one wild type, one E17K and one E17K þ E49K (data not shown). Consistent with this, KU19-19 contains three copies of chromosome arm 14q (data not shown). Retroviruses driving expression of wild-type AKT1, AKT1-E17K, AKT1-E49K and AKT1-E17K þ E49K were created and used to transduce primary NHUC. AKT activation was determined by assessing phosphorylation at Ser473 and

Figure 1 AKT1 is constitutively active in cells with an AKT1 mutation. (a) Immunoblot analysis of lysates of sub-confluent cells grown in the presence or absence of growth factors with antibodies to AKT1 and phospho-AKT Thr308. In TERT-NHUC cells, phospho-AKT levels are reduced under conditions of growth factor deprivation, but remain elevated in the cell lines with mutant AKT1. (b) Immunoblot analysis of AKT1 immunoprecipitates from lysates of sub-confluent growth factor-deprived cells with antibodies to phospho-AKT Ser473 and Thr308. No phosphoAKT signal was detected after immunoprecipitation with the nonspecific antibody. Phospho-AKT Ser473/Thr308 signals are greatly elevated in KU19-19 and MGH-U3 cells compared with TERTNHUC showing increased AKT1 activation in these cell lines in the absence of growth factors.

Thr308 (Figures 2a and b). Levels of phosphorylated Thr308 were modestly increased in cells expressing both AKT1-E17K and AKT1-E49K under growth factorstarved conditions. Under the same conditions, cells expressing AKT1-E17K and AKT1-E49K showed markedly increased levels of phosphorylated Ser473, AKT1-E17K having the greatest effect. Growth factorstarved cells expressing AKT1-E17K þ E49K showed increases in phosphorylation of both Thr308 and Ser473 to levels greater than those produced by either AKT1E17K or AKT1-E49K alone, suggesting possible cooperation or synergy of the two mutations in AKT activation. Notably, cells expressing AKT1-E17K þ E49K showed the highest levels of phosphorylated Thr308 and Ser473 on growth factor reintroduction. The AKT substrate GSK3b also showed elevated phosphorylation during starvation in AKT1-E17K and AKT1-E17K þ E49K expressing cells compared with controls, although constitutive activation was lower in AKT1-E49K cells. To determine the effects of the two mutations on AKT1 activation more clearly, immunoprecipitates of lysates of NIH3T3 cells expressing AKT1, AKT1-E17K, AKT1-E49K and AKT1-E17K þ E49K were analysed for AKT Thr308 and Ser473 phosphorylation during serum starvation and after subsequent serum Oncogene


152

Figure 2 AKT1 mutations cause constitutive and enhanced AKT activation in primary normal human urothelial cells and NIH3T3 cells. Immunoblot analysis (a, c) of AKT1, P-AKT Ser473 and Thr308, GSK3b, P-GSK3b Ser9 in lysates of sub-confluent NHUC (a, b) or immunoprecipitates of lysates of NIH3T3 cells (c, d) transduced with empty virus or virus expressing wild-type AKT1 or AKT1 mutants under conditions of growth factor starvation and subsequent stimulation. Phospho-AKT band intensities were normalized to the corresponding AKT1 expression level and are represented graphically relative to values obtained from cells expressing only endogenous AKT1 (pFB) (b, d). (a, b) AKT Ser473 phosphorylation is barely detectable in growth factordeprived NHUC with empty virus or expressing wild-type AKT1. Serum-starved cells expressing AKT1-E17K and AKT1-E49K show constitutive phosphorylation at Ser473 and to a more limited extent at Thr308 compared with wild-type AKT1-expressing cells. Cells expressing AKT1-E17K þ E49K show enhanced levels of Ser473 and Thr308 phosphorylation. GSK3b Ser9 phosphorylation is also constitutively elevated in serum-starved cells expressing AKT1-E17K and AKT1-E17K þ E49K, and to a limited extent AKT1-E49K compared with wild-type AKT1-expressing cells. (c, d) Serum-starved NIH3T3 cells expressing AKT1-E17K show increased phosphorylation at both Thr308 and Ser473. Cells expressing AKT1-E49K show a modest increase in Ser473 phosphorylation with no increase at Thr308. Ser473 and Thr308 phosphorylation is greatest in cells expressing AKT1-E17K þ E49K.

reintroduction (Figures 2c and d). Serum-starved cells expressing AKT1-E17K showed a dramatic increase in phosphorylation at both Thr308 and Ser473. Under similar conditions, cells expressing AKT1-E49K showed a more modest increase in Ser473 phosphorylation with Oncogene

no increase at Thr308. Serum-starved cells expressing AKT1-E17K þ E49K showed increases in phosphorylation at Thr308 and Ser473 greater than those produced by either AKT1-E17K or AKT1-E49K alone, again suggesting possible synergy in AKT1 activation. Again,


153

cells expressing AKT1-E17K þ E49K showed the highest levels of phosphorylated Thr308 and Ser473 on growth factor reintroduction. NIH3T3 cells transduced with the AKT1 and control viruses were used to assess transforming ability. In no case was morphological transformation induced (data not shown). This is compatible with earlier studies of activated AKT in NIH3T3 cells in which membranetargeted AKT1 or an activated mutant (E40K) induced anchorage-independent growth and tumourigenicity in nude mice, but with minimal morphological changes (Sun et al., 2001). Growth curve analysis showed that cells expressing AKT1-E49K, AKT1-E17K and AKT1E17K þ E49K have a significant proliferative advantage over controls (Figures 3a). AKT1-E49K cells were able to form colonies in soft agar, although these were smaller and fewer than those of AKT1-E17K and AKT1-E17K þ E49K cells (Figure 3b and c). This indicates that the E49K mutation is not as strongly transforming as the E17K mutation. There was no evidence for synergism between the two mutations in these assays. This may reflect the use of optimum growth conditions in these assays, in which differences in AKT1 activation were less striking than in growth factor-depleted conditions (Figures 2c and d). Our data show that the E17K mutation results in constitutive activation of AKT1 as shown by increased phosphorylation at Thr308 and Ser473. This is in contrast with the E49K mutation, which results only in constitutively elevated Ser473 phosphorylation. It has been shown that the E17K mutation results in enhanced translocation of AKT1 to the membrane in which phosphorylation by PDK1 occurs (Carpten et al., 2007). Recent results indicate that the effect of E17K is to broaden the lipid-binding specificity of AKT such that it can bind physiological concentrations PIP2 found in the plasma membrane (Landgraf et al., 2008). It has also been shown that in the inactive state, AKT is bound to PDK1, but activation is suppressed through interaction of the PH domain with the kinase domain, preventing PDK1 access to Thr308 and preventing phosphorylation (Calleja et al., 2007). Structural studies of the AKT PH domain, both uncomplexed (Milburn et al., 2003) and complexed to PIP3 (Thomas et al., 2002) have been reported. Comparison of these structures indicates a large conformational change on binding of PIP3 such that the variable loop 2 region, which seems flexible when uncomplexed, shows clustering of 3–4 negative charges (contributed by Asp44, Asp46, Glu49 and to a lesser extent Glu40) when complexed with PIP3 to form an acidic patch facing the solvent. It was suggested that mutation of residues within this patch may allow activation of AKT in the presence of lower concentrations of PIP3 (Milburn et al., 2003). Indeed, it has been shown that the mutation G40L results in increased activity of AKT in cells (Bellacosa et al., 1998). We hypothesize that the E49K mutation may have a similar effect, but that when present in isolation, this mutation does not facilitate membrane translocation and phosphorylation of Thr308. However, Ser473 phosphorylation is accomplished by mTORC2, present in the

Figure 3 AKT1 mutations are transforming. (a) NIH3T3 cells transduced with empty virus or virus containing wild-type AKT1 or AKT1 mutants were seeded in triplicate in six-well dishes at 15 000 cells per well and counted at days 1, 3, 6 and 8. AKT1 mutant expressing cells show increased proliferation. Experiments were repeated in triplicate. (b) 3T3 cells expressing AKT1 mutants show increased anchorage-independent growth. Cells were seeded in triplicate in agar at 50 000 per well and allowed to grow for 14 days before staining with p-iodotetrazolium violet. AKT1 mutantexpressing cells formed more large colonies than controls, although AKT1-E49K formed fewer colonies than AKT1-E17K and AKT1E17K þ E49K. Experiments were repeated in triplicate. (c) Typical colonies from each experiment. Bar ¼ 250 mM.

cytoplasm. Thus, synergy of E49K with E17K could result from enhanced membrane localization combined with a reduced ability of the PH domain of the protein Oncogene


154

to interact with the kinase domain and increased ability to adopt an active conformation. The reduced AKT activating capacity of E49K compared with E17K and the ability of E49K to augment the activating activity of E17K may explain why the E49K mutation was not detected independently of the E17K mutation. Indeed, the identification of alleles with E17K and both E17K and E49K in the cell line KU-19-19 indicates that the E17K mutation probably preceded the E49K mutation. As larger numbers of tumours with E17K are identified, additional mutations involving residues Asp44, Asp46, Glu49 and Glu40 in the ‘acidic patch’ may be identified. The PH domain of AKT1 also interacts with a number of proteins and these interactions impact on activation. These include Ca2 þ calmodulin (Dong et al., 2007), CKIP-1 (Tokuda et al., 2007) and TCL1 (Laine et al., 2000). Thus, it is possible that the E49K mutation may also modulate these interactions to enhance AKT1 activity. In conclusion, we have identified an additional mechanism by which PI3 kinase pathway activation can occur in bladder cancer. Taken together, the

frequencies of involvement of PIK3CA, PTEN, TSC1 and AKT1 implicate this pathway in the majority of bladder tumours. Importantly, we have identified a novel mutation in this gene that confers increased activation, transforming ability in rodent fibroblasts and shows evidence of co-operation with a known activating mutation. Examination of the structural consequences of this alteration and its effect on protein–protein interactions may improve understanding of the mechanism of action of this critical signalling protein.

Conflict of interest The authors declare no conflict of interest.

Acknowledgements This work was funded by a Programme grant from Cancer Research UK (C6228/A5433).

References Aveyard JS, Skilleter A, Habuchi T, Knowles MA. (1999). Somatic mutation of PTEN in bladder carcinoma. Br J Cancer 80: 904–908. Bellacosa A, Chan TO, Ahmed NN, Datta K, Malstrom S, Stokoe D et al. (1998). Akt activation by growth factors is a multiple-step process: the role of the PH domain. Oncogene 17: 313–325. Bleeker FE, Felicioni L, Buttitta F, Lamba S, Cardone L, Rodolfo M et al. (2008). AKT1(E17K) in human solid tumours. Oncogene 27: 5648–5650. Bozulic L, Surucu B, Hynx D, Hemmings BA. (2008). PKBalpha/Akt1 acts downstream of DNA-PK in the DNA double-strand break response and promotes survival. Mol Cell 30: 203–213. Brugge J, Hung MC, Mills GB. (2007). A new mutational AKTivation in the PI3K pathway. Cancer Cell 12: 104–107. Cairns P, Evron E, Okami K, Halachmi N, Esteller M, Herman JG et al. (1998). Point mutation and homozygous deletion of PTEN/ MMAC1 in primary bladder cancers. Oncogene 16: 3215–3218. Calleja V, Alcor D, Laguerre M, Park J, Vojnovic B, Hemmings BA et al. (2007). Intramolecular and intermolecular interactions of protein kinase B define its activation in vivo. PLoS Biol 5: e95. Cappellen D, Gil Diez de Medina S, Chopin D, Thiery JP, Radvanyi F. (1997). Frequent loss of heterozygosity on chromosome 10q in muscle-invasive transitional cell carcinomas of the bladder. Oncogene 14: 3059–3066. Carpten JD, Faber AL, Horn C, Donoho GP, Briggs SL, Robbins CM et al. (2007). A transforming mutation in the pleckstrin homology domain of AKT1 in cancer. Nature 448: 439–444. Dong B, Valencia CA, Liu R. (2007). Ca(2+)/calmodulin directly interacts with the pleckstrin homology domain of AKT1. J Biol Chem 282: 25131–25140. Hennessy BT, Smith DL, Ram PT, Lu Y, Mills GB. (2005). Exploiting the PI3K/AKT pathway for cancer drug discovery. Nat Rev Drug Discov 4: 988–1004. Kim MS, Jeong EG, Yoo NJ, Lee SH. (2008). Mutational analysis of oncogenic AKT E17K mutation in common solid cancers and acute leukaemias. Br J Cancer 98: 1533–1535. Knowles MA, Habuchi T, Kennedy W, Cuthbert-Heavens D. (2003). Mutation spectrum of the 9q34 tuberous sclerosis gene TSC1 in transitional cell carcinoma of the bladder. Cancer Res 63: 7652–7656. Oncogene

Kumar R, Hung MC. (2005). Signaling intricacies take center stage in cancer cells. Cancer Res 65: 2511–2515. Laine J, Kunstle G, Obata T, Sha M, Noguchi M. (2000). The protooncogene TCL1 is an Akt kinase coactivator. Mol Cell 6: 395–407. Landgraf KE, Pilling C, Falke JJ. (2008). Molecular mechanism of an oncogenic mutation that alters membrane targeting: Glu17Lys modifies the PIP lipid specificity of the AKT1 PH domain. Biochemistry 47: 12260–12269. Lopez-Knowles E, Hernandez S, Malats N, Kogevinas M, Lloreta J, Carrato A et al. (2006). PIK3CA mutations are an early genetic alteration associated with FGFR3 mutations in superficial papillary bladder tumors. Cancer Res 66: 7401–7404. Malanga D, Scrima M, De Marco C, Fabiani F, De Rosa N, De Gisi S et al. (2008). Activating E17K mutation in the gene encoding the protein kinase AKT1 in a subset of squamous cell carcinoma of the lung. Cell Cycle 7: 665–669. Manning BD, Cantley LC. (2007). AKT/PKB signaling: navigating downstream. Cell 129: 1261–1274. Milburn CC, Deak M, Kelly SM, Price NC, Alessi DR, Van Aalten DM. (2003). Binding of phosphatidylinositol 3,4,5-trisphosphate to the pleckstrin homology domain of protein kinase B induces a conformational change. Biochem J 375: 531–538. Platt FM, Hurst CD, Taylor CF, Gregory WM, Harnden P, Knowles MA. (2009). Spectrum of phosphatidylinositol 3-kinase pathway gene alterations in bladder cancer. Clin Can Res (in press). Pymar LS, Platt FM, Askham JM, Morrison EE, Knowles MA. (2008). Bladder tumour-derived somatic TSC1 missense mutations cause loss of function via distinct mechanisms. Hum Mol Genet 17: 2006–2017. Sarbassov DD, Guertin DA, Ali SM, Sabatini DM. (2005). Phosphorylation and regulation of Akt/PKB by the rictor-mTOR complex. Science 307: 1098–1101. Sun M, Wang G, Paciga JE, Feldman RI, Yuan ZQ, Ma XL et al. (2001). AKT1/PKBalpha kinase is frequently elevated in human cancers and its constitutive activation is required for oncogenic transformation in NIH3T3 cells. Am J Pathol 159: 431–437. Thomas CC, Deak M, Alessi DR, van Aalten DM. (2002). Highresolution structure of the pleckstrin homology domain of protein


155 kinase b/akt bound to phosphatidylinositol (3,4,5)-trisphosphate. Curr Biol 12: 1256–1262. Tokuda E, Fujita N, Oh-hara T, Sato S, Kurata A, Katayama R et al. (2007). Casein kinase 2-interacting protein-1, a novel Akt pleckstrin homology domain-interacting protein, down-regulates PI3K/Akt

signaling and suppresses tumor growth in vivo. Cancer Res 67: 9666–9676. Wang DS, Rieger-Christ K, Latini JM, Moinzadeh A, Stoffel J, Pezza JA et al. (2000). Molecular analysis of PTEN and MXI1 in primary bladder carcinoma. Int J Cancer 88: 620–625.

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

Oncogene