Constitutive Activation of Akt/Protein Kinase B in Melanoma Leads to Up- Regulation of Nuclear Factor-␬B and Tumor Progression1

Home , Protein kinase B

[CANCER RESEARCH 62, 7335–7342, December 15, 2002] Constitutive Activation of Akt/Protein Kinase B in Melanoma Leads to Up- Regulation of Nuclear Factor-␬B and Tumor Progression1

Punita Dhawan, Amar B. Singh, Darrel L. Ellis, and Ann Richmond2 Departments of Veterans Affairs [A. R.], Cancer Biology [P. D., A. R.], Nephrology [A. B. S.], and Medicine, Division of Dermatology [D. L. E.], Vanderbilt University School of Medicine, Nashville, Tennessee 37232

ABSTRACT iting a cumulative melanoma risk of 1.9% compared with 0.04% of patients without dysplastic nevi (7). The current lifetime incidence The serine/threonine kinase Akt/protein kinase B and the pleiotropic of melanoma in the United States is estimated at 1:74 (8). The high transcription factor nuclear factor-␬B [NF-␬B (p50/p65)] play important incidence of dysplastic nevi and melanoma presents a large public roles in the control of cell proliferation, apoptosis, and oncogenesis. Pre- vious studies from our laboratory have shown the constitutive activation health problem. of NF-␬B in melanoma cells. However, the mechanism of this activation is Whereas histopathology is the current gold standard for the diag- not clearly understood. The purpose of this study was to explore the role nosis of atypical melanocytic lesions, interobserver correlation in the of Akt in the activation of NF-␬B during melanoma tumor progression. diagnosis of dysplastic nevi is variable (9, 10). The malignant poten- Based on our observation that two of the five melanoma cell lines exam- tial of dysplastic nevi and early melanomas is difficult to predict with ined exhibit constitutive Akt activation, we evaluated Akt activation by standard histological staining. Therefore, better techniques are de- immunohistochemistry in a series of pigmented skin lesions using an sired. Standard immunohistochemical markers such as HMB-45 have antibody specific for phospho-Akt Ser-473. Normal and slightly dysplastic not been useful in differentiating a dysplastic nevus from melanoma in nevi exhibited no significant Akt expression, in marked contrast to the situ (11). Some newer immunohistochemical markers such as Ki 67, dramatic Akt immunoreactivity seen in severely dysplastic nevi and melanomas (66.3% positive). When these same lesions were stained for nu- proliferating cell nuclear antigen, cyclin A, matrix metalloprotein- clear p65, a similar expression pattern was observed. In addition, inter- ase-2, osteonectin, p16, and gp100 may be helpful prognostically in ruption of Akt activation resulted in increased apoptosis and decreased some subgroups of melanoma patients (12). However, the molecular NF-␬B promoter activity. These results indicate that activation of Akt events that correlate with the progression of benign melanocytic nevi kinase is linked to enhanced NF-␬B nuclear localization and transactiva- to dysplastic nevi to malignant melanomas currently are not well tion. We propose that activation of Akt may be an early marker for tumor understood and are vital for clarifying ambiguities in histopathology progression in melanoma. and clinical prognosis as well as offering potential new therapeutic interventions. INTRODUCTION Many human solid tumor cell lines display increased nuclear NF- ␬B3 levels and/or increased NF-␬B transcriptional activity (13–15). The incidence of cutaneous melanoma has risen logarithmically in the NF-␬B is activated in Hodgkin’s lymphoma, head and neck squamous last few decades, and melanoma accounts for the most deaths from skin cell carcinoma, non-small cell lung cancer, colorectal cancer, thyroid cancer (1–3). Melanoma affects both relatively young and older popula- cancer, pancreatic carcinoma, leukemia, multiple myeloma, prostrate tions, metastasizes rapidly, and is highly resistant to chemotherapy and cancer, and breast cancer (16–18). Our laboratory has previously other treatments. To develop new diagnostic strategies, it is very impor- shown constitutive NF-␬B expression in melanoma cells (19). In most tant to understand the events involved in the progression from dysplasia nontransformed cell types, NF-␬B complexes (a heterotrimer com- to malignancy. The transition from benign lesions to invasive, metastatic posed of p50 and RelA/p65 subunits bound to an inhibitor, I␬B) are tumors occurs through a stepwise process involving changes in expres- largely cytoplasmic. On activation, the I␬B proteins become phos- sion and/or function of oncogenes or tumor suppressor genes, as well as phorylated, ubiquitinated, and subsequently degraded. This liberates constitutive expression of chemotactic cytokines, growth factors, and NF-␬B, allowing it to accumulate in the nucleus, where it enhances metalloproteinases (reviewed in Ref. 4). the transcription of specific genes. The I␬B kinase, or IKK complex Clinically, melanocytic lesions may be divided into benign mela- ␣ (Mr 700,000–900,000), consists of two catalytic units, IKK- and nocytic nevi (i.e., junctional nevi, compound nevi, and dermal nevi), IKK-␤, bound to a regulatory subunit IKK-␥ or NEMO (20–23). lesions with possible malignant potential (i.e., dysplastic nevi, con- Akt/PKB, a serine/threonine kinase, is a core component of the genital nevi, and lentigo maligna), and malignant lesions [i.e., primary PI3K signaling pathway that is activated through phosphorylation of melanoma such as superficial spreading melanoma, nodular mela- Ser-473/474 and Thr-308/309 (24). Several studies have shown that noma, lentigo maligna melanoma, and acral lentiginous melanoma as Akt/PKB activates the transcription of a wide range of genes, espe- well as metastatic melanoma (5, 6)]. For this study, we selected the cially those involved in immune activation, cell proliferation, apo- following types of clinical lesions to study melanocytic tumor pro- ptosis, and cell survival (25). Mechanisms associated with the ability gression: (a) benign melanocytic nevi; (b) dysplastic nevi; (c) lentigo of Akt to suppress apoptosis include the phosphorylation and inacti- maligna; and (d) metastatic melanomas. vation of many proapoptotic proteins such as Bad (26), caspase-9 The incidence of dysplastic nevi in the general population has (27), and the forkhead family of transcription factors (28) and acti- been estimated at 7%, with patients having dysplastic nevi exhib- vation of NF-␬B (29). Based on its role as a key regulator of cellular survival, Akt is Received 5/16/02; accepted 10/18/02. emerging as a central player in tumorigenesis. In the last decade, 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 evidence from different studies suggests that Akt perturbations play 18 U.S.C. Section 1734 solely to indicate this fact. an important role in human malignancy. Akt1 and Akt2 amplification 1 Supported by a Career Scientist Award from the Department of Veterans Affairs (to A. R.), by National Cancer Institute Grants CA56704 (to A. R.) and CA34590 (to A. R.), and NIH Center Grants CA 68485 and P30 AR 41943. 3 The abbreviations used are: NF-␬B, nuclear factor-␬B; PKB, protein kinase B; PI3K, 2 To whom requests for reprints should be addressed, at Department of Cancer Biology, phosphatidylinositol 3Ј-kinase; GST, glutathione S-transferase; WM, wortmannin; RPE, Vanderbilt University School of Medicine, MCN T-2212, Nashville, TN 37232. Phone: retinal pigment epithelial; NHEM, normal human epidermal melanocyte; IKK, I␬B (615) 343-7777; Fax: (615) 343-4539; E-mail: [email protected]. kinase. 7335

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2002 American Association for Cancer Research. ROLE OF AKT IN MELANOMA TUMOR PROGRESSION and/or mRNA overexpression was observed in human gastric cancer was used to visually capture the immunolocalization of antigen using Novared and 10–20% of human ovarian, breast, and pancreatic cancers, re- substrate, and cell contents were counterstained with hematoxylin. For immu- spectively (30, 31). Activation of the Akt2 kinase is reported in 40% nofluorescence, primary antibodies were used at the following dilutions: rabbit of ovarian cancers (32, 33), whereas Akt1 kinase activity is often anti p-Akt (1:50; Cell Signaling Technology); and mouse anti RelA/p65 increased in prostate and breast cancers (34). Amplification of AKT3 (1:100; Chemicon). All primary antibody incubations were performed in a moisture chamber overnight at 4°C. Secondary antibodies were donkey anti- has not been described. Furthermore, loss of the negative regulator of rabbit Texas red and goat antimouse Alexa 488. The slides were fixed with this pathway, the tumor suppressor PTEN, correlates with increased Vectashield and sealed with clear nail polish. Stained sections were viewed and Akt activity in many types of cancers (35, 36). photographed using the fluorescence microscope. The excitation wavelength The present challenge is to make sense of the network that controls was 549 nm for Texas red and was 488 nm for Alexa 488. apoptosis during melanoma tumorigenesis, which requires evaluating Measurement of Apoptosis. Subconfluent melanoma Hs294T, SKMel28, the Akt status of typical and atypical pigmented lesions, as well as cell and WM115 cells and normal RPE cells were transfected with either vector lines. Although high expression of Akt and/or Rel/NF-␬B has been control or Akt dominant negative constructs. Alternatively, cells were incu- demonstrated in several different tumors, the relevance of constitutive bated with the PI3K inhibitor, WM (100 nM). The cells were then collected and expression of these genes to the progression and metastasis of human stained with fluorescein-conjugated annexin V and propidium iodide, using the melanomas is poorly defined. In this report, we investigated the TACS-Annexin V-FITC kit in accordance with the manufacturer’s instructions (Boehringer Mannheim, Basel, Switzerland). After staining, the cells were activation of Akt and NF-␬B in melanoma cell lines and melanocytic sorted by flow cytometry with a FACScan (Becton Dickinson, Franklin Lakes, lesion tissues. NJ). Annexin V-positive stained cells were counted as apoptotic cells. Cells that were propidium iodide positive but annexin V negative were not counted MATERIALS AND METHODS as apoptotic cells. Statistical Analysis. Student’s t test for paired samples was used to deter- Plasmids and Reagents. The NF-␬B luciferase reporter vector contains mine statistical significance of the transfection data. Differences were consid- five tandem repeats of the NF-␬B element 5Ј to the transcription initiation site ered statistically significant at P Յ 0.05. and is contained in pLUC-MCS reporter vector (Stratagene, La Jolla, CA). Rabbit anti-phospho-Akt was obtained from Cell Signaling Technology (Bev- RESULTS erly, MA), and rabbit anti-RelA/p65 and rabbit anti-Akt were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). GST-I␬B purified protein was Akt/PKB Phosphorylation and Activity Are Constitutive in also obtained from Santa Cruz Biotechnology. WM and LY294002 were Some Melanoma Cells. To determine whether Akt/PKB is acti- obtained from Calbiochem (La Jolla, CA). vated in melanoma cells, five different metastatic melanoma cell Cell Culture and Transfection. The human melanoma cell lines Hs294T, lines obtained from American Type Culture Collection were stud- SKMel5, SKMel28, WM115, and WM164, normal RPE cells, and NHEMs ied. Cells were deprived of serum overnight, and then expression were obtained and cultured as described previously (37). One day before of activated Akt was assessed by immunoblotting cell lysates using transfection, the cells were seeded in 6-well cell culture plates to provide a final density of 40–60% confluence (ϳ3 ϫ 105 cells/well). Cells were trans- phospho-specific antibodies against phosphorylated Ser-473 fected using effectene transfection reagent (Qiagen, Valencia, CA) as de- (Fig. 1A). Two of the five melanoma cell lines, Hs294T and scribed previously (37). Fold stimulation was calculated for each sample by WM115, were highly phosphorylated at Ser-473 as compared with dividing the normalized luciferase activity by the value obtained from the normal controls. In vitro kinase assays using glycogen synthase control transfection containing empty parental expression vectors (pCMV). kinase-3␣/␤ as a substrate were performed to verify that phospho- Immunoblot Analysis. Whole cell extracts were obtained according to our rylated Akt/PKB is enzymatically active. Fig. 1B shows that the standard protocol using radioimmunoprecipitation assay buffer, resolved on highest levels of Akt/PKB enzymatic activity are found in the 10% SDS-PAGE, transferred to the nitrocellulose membrane, blocked with 5% milk in Tris-buffered saline containing 0.1% Tween 20, and probed with appropriate antibodies as described previously (37). The p-Akt antibody was used at a dilution of 1:500, whereas Akt was used at a 1:1000 dilution. The antibodies were visualized with either horseradish peroxidase-conjugated goat antimouse or antirabbit IgG (Boehringer Mannheim, Indianapolis, IN) using enhanced chemiluminescence (Pierce, Rockford, IL). Immunoprecipitation and Kinase Assays. Normal and melanoma cells were grown in 100-mm cell culture dishes. Cells were placed in serum-free media for 18–24 h. Thereafter, cells were lysed, and extracts were prepared. Samples were then immunoprecipitated with IKK-␣␤ antibody, and a kinase assay was performed using full-length GST-I␬B␣ substrate, as described previously (37). The same blot was normalized with the antibody used for immunoprecipitation. Immunohistochemistry and Immunofluorescence. Paraffin-fixed tissue samples were obtained from melanoma and normal tissues of patients using protocols approved by the Vanderbilt University Institutional Review Board. All melanocytic lesion diagnoses on tissues used in these studies were made independently by dermatopathologists who were not involved in the study. Criteria for the diagnosis of dysplastic nevi were those adopted by the 1992 NIH Consensus Conference (38). This diagnosis requires both (a) architectural disorder and (b) melanocytic atypia, which is classified as mild, moderate, or Fig. 1. Akt is constitutively activated in Hs294T and WM115 melanoma cell lines. severe. Sections were placed on glass slides and deparaffinized with xylene. RPE, NHEMs and melanoma cells were cultured in the absence of serum overnight, and The antigen was unmasked by heating samples in 10 mM sodium citrate buffer the next morning, whole cell extracts were prepared for immunoblotting. A, the expression (pH 6.0) for 5 min, the reaction was quenched using hydrogen peroxide, and of both activated Akt and total Akt was determined using anti-phospho-Akt (Ser-473) and samples were then immunostained for p-Akt using rabbit anti-p-Akt antibody anti-Akt antibodies. B, in vitro kinase assays of AKT immunoprecipitates from normal and melanoma cells. Akt activity was determined using Akt kinase assay with glycogen (1:50; Cell Signaling Technology) or for activated Rel A/p65 with mouse-anti synthase kinase-3␣/␤ as a substrate. This figure is representative of three separate RelA/p65 (1:50, Chemicon, Temecula, CA). The ABC biotin/avidin reagent kit experiments. 7336

Fig. 2. Paraffin-embedded sections from biopsy specimens of patients were immunostained with antibody for activated Akt. The red color indicates a positive immunolocalization of Akt phosphorylated on Ser-473. Melanomas stain strongly for activated Akt [patient 19 (A and B, ϫ20 and ϫ100, respectively) and patient 25 (C and D, ϫ20 and ϫ100, respectively)]. Severely dysplastic nevi also stain strongly for activated Akt [patient 14 (E, ϫ20)]. Activated Akt was not observed in nevi with mild dysplasia [patient 11 (F, ϫ20; G, ϫ100)]. The negative control for immunostaining in B (patient 19) is shown in H (ϫ100). I shows an example of a melanoma with no staining for Akt [patient 23 (ϫ20)]. The brown color is due to the pigment in the tumor cells and does not represent activated Akt. Arrows indicate p-Akt staining.

Hs294T and WM115 cells, which have the highest levels of phos- Relationship between p-Akt and RelA/p65 Expression in phorylated Akt/PKB. When combined, the data from the phospho- Melanoma Lesions. To determine whether p-Akt and activated rylation and kinase assay experiments show that there is constitu- RelA/p65 colocalized, double immunofluorescence staining was used. tive activation of Akt/PKB in two of the five melanoma cell lines. This was accomplished using Texas red-conjugated donkey antirabbit Immunostaining of p-Akt in Normal and Melanoma Tissues. antibody and Alexa 488-conjugated goat antimouse to stain p-Akt and To determine whether activated Akt is present in the biopsies of RelA/p65 antibody, respectively. Colocalization of p-Akt and acti- pigmented lesions at various stages of tumor progression, immu- vated RelA/p65 was observed in tumor tissues (Fig. 4) and melanoma nohistochemical staining of 29 paraffin-embedded tumor speci- cell lines (Fig. 5), and this colocalization was largely nuclear because mens in various stages of tumor progression was performed using p-Akt and activated p65 translocate to nucleus (40). In addition, we a phospho-specific anti-AKT antibody. As shown in Fig. 2, mel- found that the normal nevi exhibited no staining, but in cultured anoma and severely dysplastic nevi showed enhanced Akt im- NHEMs, a low level of staining was observed in a few cells, probably munoreactivity (Fig. 2, AϪE), in contrast to a mildly dysplastic due to the in vitro proliferation of these cells. These observations specimen (Fig. 2, F and G). Fig. 2H represents a negative control. suggest that Akt activation is associated with NF-␬B activation in Some melanoma lesions did not exhibit immunostaining for Akt melanoma. Moreover, this event parallels the degree of atypia ob- (Fig. 2I), suggesting that there may be subgroups of melanoma that served in pigmented lesions (Figs. 4 and 5). do not exhibit activation of this pathway. Interestingly, phospho- PI3K Is Required for Activation of Akt and Thus NF-␬Bin rylated Akt was located in the tumor cell membrane, the cyto- Melanoma Cells. To address the roles of the PI3K pathway in NF-␬B plasm, and the nucleus, as supported by the previously reported activation by Akt, the effects of pharmacological PI3K inhibitors observation that Akt transiently associates with plasma membrane LY294002 and WM on Akt phosphorylation and NF-␬B activity were of stimulated cells, which is followed by translocation to the examined. Fig. 6A shows that both LY294002 (5, 10, and 20 ␮M) and nucleus (39). The normal nevus lesions were largely negative for WM (100 and 200 nM) almost completely inhibited the phosphoryl- phospho-Akt. ation of Ser-473 on Akt in a dose-dependent manner in the Hs294T Constitutive NF-␬B (RelA/p65) Activation Is Observed in cell line, whereas there was no effect in RPE cells (data not shown). Melanoma Cells. Immunostaining for RelA/p65 was done using an Native Akt/PKB levels were not affected by these treatments. Thus, antibody that is directed against the nuclear localization signal, and Akt/PKB phosphorylation in the Hs294T melanoma cell line is PI3K thus immunofluorescence is able to detect nuclear NF-␬B as well as dependent. Furthermore, to test whether NF-␬B activation in mela- RelA/p65 not bound to I␬B in the cytoplasm. The active RelA/p65 noma cell lines is PI3K dependent, the effect of these inhibitors was was displayed by melanoma tumor tissues and severely dysplastic examined on NF-␬B promoter luciferase activity in Hs294T and nevi (Fig. 3, AϪE) and not by normal nevi or in nevi with mild to WM115 melanoma cell lines (Fig. 6B). A dose-dependent decrease in moderate dysplasia (Fig. 3, F and G). A melanoma lesion negative for reporter luciferase activity in both melanoma cell lines was observed. RelA/p65 immunostaining is shown in Fig. 3H. Thus, for these two melanoma cell lines, NF-␬B activity is constitu- The Akt and RelA/p65 activation status of the 29 patients tested is tively activated, and this activation is dependent on the PI3K-Akt summarized in Table 1. The normal patients were mainly negative for pathway. both RelA/p65 and phosphorylated Akt staining. Of 11 melanoma IKK Activity Is Not Affected by PI3K Inhibitors. To explore the cases examined, 7 (63.6%) were positive for both p-Akt and nuclear possibility of IKK being involved in NF-␬B activation, the effect of p65. Interestingly, the p-Akt and RelA/p65 staining was almost neg- treatment with the PI3K inhibitors LY294002 and WM on IKK ative for normal nevi, increased with progression of melanocytic activity was evaluated in Hs294T melanoma and RPE control cells. lesion atypia, and was most apparent in metastatic melanoma lesions. IKK-␣ and IKK-␤ were immunoprecipitated from the treated samples, 7337

Fig. 3. Paraffin-embedded sections from biopsy specimens were immunostained with antibody for activated p65. The red color indicates a positive immunolocalization of activated p65. Melanomas stain strongly for activated p65 [patient 19 (A and B, ϫ20 and ϫ100, respectively) and patient 25 (C and D, ϫ20 and ϫ100, respectively)]. Severely dysplastic nevi also stained strongly for activated p65 [patient 14 (E, ϫ20)]. Activated p65 was not observed in nevi with mild dysplasia [patient 11 (F, ϫ20)]. The negative control for immunostaining is shown in G (ϫ20). H shows an example of a melanoma (patient 23) negative for p65 immunostaining. Arrows indicate the activated p65.

and then an IKK kinase assay was performed using GST-I␬B␣ as a used to alter Akt/PKB activity and assess apoptosis. Melanoma cells substrate. As shown in Fig. 7, there was no difference between control expressing constitutively active Akt/PKB were either transiently and treated samples in the phosphorylation of GST-I␬B␣ for either transfected with dominant negative Akt or treated with WM (100 nM) Hs294T or RPE cells. Thus, it appears that the PI3K-Akt activation for 48 h. Typical morphological changes of apoptosis were observed, pathway does not affect the IKK activity leading to I␬B␣ phospho- including membrane blebbing, increased refractoriness, and chroma- rylation, ubiquitination, and degradation. tin condensation (data not shown). Other evidence for WM-induced PI3K-Akt Pathway Promotes Survival of Melanoma Cells. To apoptosis included increased Annexin V binding to cells as assessed test whether active Akt/PKB promotes cellular survival of melanoma by flow cytometry. Apoptosis was potentiated when Akt activation cell lines, transient transfection and pharmacological approaches were was inhibited in the Hs294T and WM115 cells by about 30–35%, but neither dominant negative Akt nor WM (100 nM) had substantial effects on the SKMel28 cell line, which exhibited the lowest level of Table 1 Immunostaining for phosphorylation of Akt and activated p65/RelA in normal and melanoma patients Akt/PKB activity (Table 2). Thus, WM not only inhibits Akt phosphorylation and activity but also increases melanoma cell apoptosis in p-Akt Patient no. Histological diagnosisa staining P65 staining relative proportion to the level of endogenous Akt/PKB activity. 1 Intradermal nevus Ϫb ϩ Therefore, the PI3K-Akt pathway is involved in promoting cell sur- 2 Intradermal nevus ϩϪvival by activating NF-␬B and preventing apoptosis. 3 Intradermal nevus Ϫϩ 4 Intradermal nevus ϩϩ 5 Intradermal nevus ϪϪDISCUSSION 6 Intradermal nevus Ϫ ND 7 Intradermal nevus ϪϪ 8 Intradermal nevus Ϫ ND In recent years, there has been an explosion of information describ- 9 Compound nevus ϪϪing the roles of Akt (PKB) in signaling pathways essential in the 10 Compound nevus Ϫϩregulation of cell growth, survival, and apoptosis. It is overexpressed 11 Lentigo maligna ϩϩ/Ϫ 12 Dysplastic nevus mild ϩϩϩas well as activated in numerous human malignancies (41). However, 13 Dysplastic nevus moderate ϩϪwhether Akt overexpression and/or constitutive activation has any 14 Dysplastic nevus moderate ϩϩ 15 Dysplastic nevus moderate/severe ϩϩ ND role in the development of melanomas is not known. In this study we 16 Dysplastic nevus moderate/severe ϩϩϩ ϩϩ have shown not only that Akt is constitutively activated in human 17 Severely dysplastic nevi ϩϩ ND melanomas but also that it enhances cell survival through NF-␬B 18 Metastatic melanoma ϩϩ ϩϩ 19 Metastatic melanoma ϩϩϩ ϩϩϩ activation, hence playing a key role in melanoma development. In 20 Metastatic melanoma ϩϩϩ ϩ addition, we have provided suggestive evidence that Akt can be used 21 Metastatic melanoma ϩϩϩ Ϫ as a prognostic marker of melanoma progression. Our immunohisto- 22 Metastatic melanoma Ϫ ND 23 Metastatic melanoma ϪϪchemical studies of various stages of human melanomas suggest a 24 Metastatic melanoma ϪϪpossible correlation between enhanced Akt activation, tumor progres- 25 Metastatic melanoma ϩϩϩ ϩϩϩ 26 Metastatic melanoma ϩϩϩ ϩϩ sion, and malignancy. 27 Metastatic melanoma ϩϩϩ ϩ Of the five human melanoma cell lines tested, two had a high level 28 Metastatic melanoma ϩϩof constitutive Akt phosphorylation and activity (Hs294T and 29 Metastatic melanoma ϩϩϩ ϩϩϩ WM115). To determine whether the differences in phospho-Akt ob- a Represents patients in different stages of melanoma tumor progression. b Ϫ, negative; ϩ, slightly positive; ϩϩ, moderately positive; ϩϩϩ, highly positive; served among cell lines are a cell-specific phenomenon or whether ND, not determined. they reflects possible differences among human melanoma lesions, 29 7338

Fig. 4. Colocalization of Akt and p65 in the section of melanoma tumors. Tumors tissues (patient 19) were deparaffinized and then subjected to immunofluorescence using rabbit anti-p-Akt antibody (activated Akt) and mouse anti-p65 (activated p65). The secondary antibody was labeled with either Texas red (rabbit) or Alexa 488 (mouse). The section was then observed by fluorescence microscopy. The green and red col- ors demonstrate that the NF-␬B subunits and Akt were local- ized throughout both the cytosol and nucleus, respectively (A and B). C shows the overlap of A and B, and yellow appears wherever colocalization occurs.

melanocytic lesions providing a spectrum of tumor progression were the melanocytic lesions based on immunostaining with the phospho- examined by immunostaining using phospho-specific anti-Akt: junc- specific anti-Akt antibody. Our data suggest a possible role for Akt in tional, compound, and dermal nevi (benign); dysplastic nevi (with the transformation of nevi into melanoma at the critical juncture cytologic atypia graded as mild, moderate, or severe; precancerous); where the nevus becomes atypical. In addition, as observed with lentigo maligna; and metastatic melanoma. When evaluating precan- melanoma cell lines, the melanoma lesions fell into two groups: (a) cerous lesions (dysplastic nevi), it is often assumed that the greater the those with high phospho-Akt; and (b) those negative for phospho-Akt. degree of melanocytic cytologic atypia, the greater the risk for ma- Taken together, the observations suggest the possibility for high and lignant transformation of a given lesion, although the significance of low phospho-Akt subtypes of melanoma. The low phospho-Akt group the degree of atypia has not been established for prognosis of the risk of melanoma lesions would presumably use Akt-independent signals of a single lesion progressing to a melanoma or of the patient devel- in association with NF-␬B activation and transformation. In this oping a melanoma (42). Interestingly, the phosphorylation and acti- regard, we have previously demonstrated NF-␬B-inducing kinase vation of Akt increased in a stage-specific manner, with maximum (NIK)-dependent NF-␬B constitutive activation in certain human mel- activation in metastatic melanoma lesions and melanoma cell lines. anoma cell lines (37). In support of our report, others have shown that Thus, we found good correlation between the progression of atypia of in primary glial cells, Akt activation converts anaplastic astrocytoma

Fig. 5. Colocalization of Akt and p65 in normal NHEM (AϪC) and melanoma cell lines Hs294T (DϪF) and SKMel28 (GϪI). Cells were plated on coverslips and then subjected to immunofluorescence using rabbit anti-p-Akt antibody (activated Akt) and mouse anti-p65 (activated p65). The secondary antibody was conjugated with either Texas red (rabbit) or Alexa 488 (mouse). Localization of activated Akt and p65 was then observed by fluorescence microscopy. The double staining was lo- calized mainly in nucleus, and it appears yellow wherever colocalization occurs.

7339

melanoma cell line (G361) demonstrated that overexpression of the dominant negative p85 subunit of PI3K induced apoptosis (45). In addition, N-Cadherin dependent Akt activation has also been shown to enhance cell growth and survival in melanoma cells, though the mechanism involves signaling through beta-catenin (46). Based on our observation that Akt and NF-␬B activation in melanoma tissues showed a similar pattern, our data suggest that NF-␬Bis a significant downstream target of the signaling pathway initiated by constitutive up-regulation of the PI3K/Akt pathway in melanoma cells. The IKK/I␬B␣/NF-␬B pathway is the major mechanism for NF-␬B activation (Fig. 8). To determine whether PI3K/Akt-dependent NF-␬B activation also modulates NF-␬B activation in melanoma lesions, we investigated the role of IKK phosphorylation of I␬B. No change in IKK phosphorylation of I␬B was observed after treatment with PI3K inhibitors, suggesting that IKK/I␬B degradation is not required for NF-␬B activation (Fig. 8). Similar observations showing that activated Akt does not stimulate endogenous IKK activity directed toward I␬B degradation have been reported (47). Thus, it is possible that IKK activity (possibly independent of I␬B phosphorylation) or some change in a structural aspect of IKK might be required for Akt Fig. 6. A, Akt/PKB is constitutively activated in a PI3K-dependent manner in mela- stimulation of RelA/p65 phosphorylation and transactivation. noma cells. Hs294T cells were either left untreated (control) or cultured for3hinthe Direct phosphorylation of p65 by Akt has been the subject of presence of increasing concentrations of the PI3K inhibitor LY294002 or WM. Cells were controversy in literature. The transactivation domain (TAD1) region lysed, and the activity of Akt/PKB was determined by immunoblotting with the anti- phospho-Akt (Ser-473) antibody. The same blot was reprobed with anti-Akt antibody as of p65 does not contain a strong consensus sequence for Akt phos- a control for protein loading. This figure is a representative of three separate experiments. B, Akt activates NF-␬B in a PI3K-dependent manner. The Hs294T and WM115 cells were cotransfected with NF-␬B luciferase reporter construct and the RSV-␤-galactosidase expression construct and either left untreated or treated with increasing concentrations of Table 2 Exposure of dominant negative Akt (AKT DN) or WM leads to increased apoptosis in melanoma cells LY294002 (10 and 20 ␮M) or WM (50 and 100 nM) for 10–12 h. Cells were then harvested, and luciferase and ␤-galactosidase activity were measured. The relative lucif- The table represents the fold increase in apoptosis. erase activity represents the luciferase activity of the sample that was normalized by Cell lines Control AKT DN WM ␤-galactosidase activity from three different experiments. The results are reported as the mean Ϯ SD of inhibition, considering 1.0 as the relative luciferase activity of the cells SKMe128 1 1.04 Ϯ 0.26 1.16 Ϯ 0.33 P Յ 0.05. WM 115 1 5.91 Ϯ 0.43 4.31 Ϯ 0.35 ,ء .transfected with corresponding empty vector Hs294T 1 2.95 Ϯ 0.61 3.32 Ϯ 0.50

Fig. 7. IKK activity is not affected by the PI3K inhibitors LY294002 or WM. Whole cell lysates of cells treated with LY294002 and WM were immunoprecipitated with IKK-␣␤, and then IKK kinase assays were performed on these immunoprecipitates using GST-I␬B␣ substrate (1 ␮g/reaction). This figure is a representative of three separate experiments. to glioblastoma multiforme (43). Also, the Akt activation in mammary glands imparts survival signals in association with tumor progression (44). Taken together, these observations imply that increased Akt activity might be a common requirement for oncogenic transformation of primary cells to metastatic tumors. To determine whether constitutive Akt activation in our system is PI3K dependent and also to test its role as a potential cell Fig. 8. A model for Akt-induced activation of NF-␬B. Akt can activate NF-␬B through survival pathway, we used PI3K-specific pharmacological inhibi- three potential and overlapping mechanisms (growth factor and/or cytokine activation of the PI3K/Akt pathway, constitutive activation of the PI3K/Akt pathway can lead to NF-␬B tors LY294002 and WM as well as overexpression of a dominant activation, and PTEN can be mutated and/or absent). NF-␬B can be activated by direct negative mutant construct of Akt. We show here that Akt activation is phosphorylation by Akt or by phosphorylation of IKK-␣, followed by IKK-␣ phosphorylation of p65, and Akt can activate NF-␬B through the IKK/I␬B/NF-␬B pathway. Upon PI3K dependent; moreover, Akt activation increases the survival of activation, NF-␬B can activate a wide variety of genes, including those involved in melanoma cells. In support of our results, a previous study in one apoptosis and cell survival. 7340

phorylation (RXRXXS/T-bulky hydrophobic; Ref. 48). However, 11. Ahmed, I., Piepkorn, M., Goldgar, D. E., Cannon-Albright, L. A., Meyer, L. J., IKK-␣ does have a perfect Akt phosphorylation consensus sequence; Skolnick, M. H., and Zone, J. J. HMB-45 staining of dysplastic melanocytic nevi in melanoma risk groups. J. Cutan. Pathol., 18: 257–260, 1991. therefore, Akt may potentially phosphorylate IKK-␣, which subse- 12. Ruiter, D. J., Spatz, A., van den Oord, J. J., and Cook, M. G. Microstaging in quently phosphorylates RelA/p65 at Ser-529. Currently, it is not clear cutaneous melanoma. J. Pathol., 195: 525–529, 2001. 13. Sharma, H. W., and Narayanan, R. The NF-␬B transcription factor in oncogenesis. whether Akt can directly phosphorylate p65 or whether Akt only Anticancer Res., 16: 589–596, 1996. phosphorylates IKK-␣ to phosphorylate RelA/p65 (Fig. 8). Additional 14. Luque, I., and Ge´linas, C. Rel/NF-␬B, and I␬B factors in oncogenesis. Semin. Cancer studies are required to answer these questions. Biol., 8: 103–111, 1997. 15. Gilmore, T. D., Koedood, M., Piffat, K. A., and White, D. W. Rel/NF-␬B/I␬B The constitutive Akt activation in severely dysplastic nevi and proteins and cancer. Oncogene, 13: 1367–1378, 1996. melanoma lesions could be due to either autocrine cytokine stimula- 16. Bargou, R. C., Emmerich, F., Krappmann, D., Bommert, K., Mapara, M. Y., tion, loss of function PTEN mutation, or gain of functional mutation Arnold, W., Royer, H. D., Grinstein, E., Greiner, A., Scheidereit, C., and Dorken, B. Constitutive nuclear factor-␬B-RelA activation is required for proliferation and in Akt and PI3K. However, previous studies have shown that in survival of Hodgkin’s disease tumor cells. J. Clin. Investig., 100: 2961–2969, human melanoma, pleckstrin homology domain mutations or muta- 1997. 17. Sunwoo, J. B., Chen, Z., Dong, G., Yeh, N., Crowl Bancroft, C., Sausville, E., Adams, tions of the activation-associated phosphorylation sites at codons 308 J., Elliott, P., and Van Waes, C. Novel proteasome inhibitor PS-341 inhibits activation and 473 of PKB/Akt do not play a major role in melanoma carcino- of nuclear factor-␬B, cell survival, tumor growth, and angiogenesis in squamous cell genesis (49, 50). We have observed loss of PTEN in melanoma cell carcinoma. Clin. Cancer Res., 7: 1419–1428, 2001. 4 18. Sovak, M. A., Bellas, R. E., Kim, D. W., Zanieski, G. J., Rogers, A. E., Traish, A. M., lines with higher Akt expression, and this could very well be the case and Sonenshein, G. E. Aberrant nuclear factor-␬B/Rel expression and the pathogen- in human melanoma samples with high Akt activation. These studies esis of breast cancer. J. Clin. Investig., 100: 2952–2960, 1997. are currently under way. We hypothesize that activation of Akt, 19. Shattuck-Brandt, R. L., and Richmond, A. Enhanced degradation of I␬B␣ contributes to endogenous activation of NF-␬B in Hs294T melanoma cells. Cancer Res., 57: whether mediated by PTEN inactivation or growth factor signaling 3032–3039, 1997. activation, might represent an important common pathway in the 20. Karin, M. How NF-␬B is activated: the role of the I␬B kinase (IKK) complex. pathogenesis or progression of melanoma. Oncogene, 18: 6867–6874, 1999. 21. Woronicz, J. D., Gao, X., Cao, Z., Rothe, M., and Goeddel, D. V. I␬B kinase-␤: In summary, the studies in this report suggest that constitutive Akt NF-␬B activation and complex formation with I␬B kinase-␣ and NIK. Science activation plays an important role in human melanoma. Although it (Wash. DC), 278: 866–869, 1997. 22. Israel, A. The IKK complex: an integrator of all signals that activate NF-␬B? Trends may not be essential for initiation of melanoma, Akt activation facil- Cell. Biol., 10: 129–133, 2000. itates melanoma progression, possibly by enhancing cell survival 23. Li, X. H., Fang, X., and Gaynor, R. B. Role of IKK␥/nemo in assembly of the I␬B through up-regulation of NF-␬B and escape from apoptosis. These kinase complex. J. Biol. Chem., 276: 4494–4500, 2001. 24. Harlan, J. E., Yoon, H. S., Hajduk, P. J., and Fesik, S. W. Structural characterization studies also provide compelling evidence for use of activation of Akt of the interaction between a pleckstrin homology domain and phosphatidylinositol as an important prognostic marker of melanoma progression. In ad- 4,5-bisphosphate. Biochemistry, 34: 9859–9864, 1995. ␬ dition, Akt activation in severely dysplastic nevi and metastatic mel- 25. Karin, M., and Lin, A. NF- B at the crossroads of life and death. Nat. Immunol., 3: 221–227, 2002. anoma and not in normal or mild to moderate dysplastic nevi suggests 26. Datta, S. R., Dudek, H., Tao, X., Masters, S., Fu, H., Gotoh, Y., and Greenberg, M. E. that a therapeutic window may exist in patients whereby antagonism Akt phosphorylation of BAD couples survival signals to the cell-intrinsic death machinery. Cell, 91: 231–241, 1997. of Akt activity might block progression of premalignant lesions. Akt 27. Cardone, M. H., Roy, N., Stennicke, H. R., Salvesen, G. S., Franke, T. F., Stanbridge, activates a number of diverse downstream proliferative and anti- E., Frisch, S., and Reed, J. C. Regulation of cell death protease caspase-9 by apoptotic pathways and thus is a promising target for future molecular phosphorylation. Science (Wash. DC), 282: 1318–1321, 1998. 28. Brunet, A., Bonni, A., Zigmond, M. J., Lin, M. Z., Juo, P., Hu, L. S., Anderson, M. J., based therapy. Arden, K. C., Blenis, J., and Greenberg, M. E. Akt promotes cell survival by phosphorylating and inhibiting a Forkhead transcription factor. Cell, 96: 857–868, 1999. REFERENCES 29. Romashkova, J. A., and Makarov, S. S. NF-␬B is a target of AKT in anti-apoptotic 1. Ries, L. A. G., Wingo, P. A., Miller, D. S., Howe, H. L., Weir, H. K., Rosenberg, PDGF signalling. Nature (Lond.), 401: 86–90, 1999. H. M., Vernon, S. W., Cronin, K., and Edwards, B. K. The annual report to the nation 30. Bellacosa, A., de Feo, D., Godwin, A. K., Bell, D. W., Cheng, J. Q., Altomare, on the status of cancer, 1973–1997, with a special section on colorectal cancer. D. A., Wan, M., Dubeau, L., Scambia, G., Masciullo, V., et al. Molecular Cancer (Phila.), 88: 2398–2424, 2000. alterations of the AKT2 oncogene in ovarian and breast carcinomas. Int. J. 2. Rigel, D. S., Friedman, R. J., and Kopf, A. W. The incidence of malignant melanoma Cancer, 64: 280–285, 1995. in the United States: issues as we approach the 21st century. J. Am. Acad. Dermatol., 31. Cheng, J. Q., Ruggeri, B., Klein, W. M., Sonoda, G., Altomare, D. A., Watson, D. K., 34: 839–847, 1996. and Testa, J. R. Amplification of AKT2 in human pancreatic cells and inhibition of 3. Parker, S. L., Tong, T., Bolden, S., and Wingo, P. A. Cancer statistics. CA Cancer AKT2 expression and tumorigenicity by antisense RNA. Proc. Natl. Acad. Sci. USA, J. Clin., 47: 5–27, 1997. 93: 3636–3641, 1996. 4. Schneider, G. P., Salcedo, R., Welniak, L. A., Howard, O. M., and Murphy, W. J. The 32. Cheng, J. Q., Godwin, A. K., Bellacosa, A., Taguchi, T., Franke, T. F., Hamilton, diverse role of chemokines in tumor progression: prospects for intervention. Int. J. T. C., Tsichlis, P. N., and Testa, J. R. AKT2, a putative oncogene encoding a member Mol. Med., 8: 235–244, 2001. of a subfamily of protein-serine/threonine kinases, is amplified in human ovarian 5. Rhodes, A. R. Benign neoplasias and hyperplasias of melanocytes. Dysplastic mela- carcinomas. Proc. Natl. Acad. Sci. USA, 89: 9267–9271, 1992. nocytic nevi. In: I. M. Freedberg, A. Z. Eisen, K. Wolff, K. F. Austen, L. A. 33. Yuan, Z. Q., Sun, M., Feldman, R. I., Wang, G., Ma, X., Jiang, C., Coppola, D., Goldsmith, S. I. Katz, and T. B. Fitzpatrick (eds.), Dermatology in General Medicine, Nicosia, S. V., and Cheng, J. Q. Frequent activation of AKT2 and induction of pp. 1018–1059 and 1060–1079. New York: McGraw-Hill, 1999. apoptosis by inhibition of phosphoinositide-3-OH kinase/Akt pathway in human 6. Langley, R. G. B., Barnhill R. L., Mihm, M. C., Jr., Fitzpatrick, T. B., and Sober, A. J. ovarian cancer. Oncogene, 19: 2324–2330, 2000. Neoplasms: cutaneous melanoma. In: I. M. Freedberg, A. Z. Eisen, K. Wolff, K. F. 34. Sun, M., Wang, G., Paciga, J. E., Feldman, R. I., Yuan, Z. Q., Ma, X. L., Shelley, Austen, L. A. Goldsmith, S. I. Katz, and T. B. Fitzpatrick (eds.), Dermatology in S. A., Jove, R., Tsichlis, P. N., Nicosia, S. V., and Cheng, J. Q. AKT1/PKB␣ General Medicine, pp. 1080–1116. New York: McGraw-Hill, 1999. kinase is frequently elevated in human cancers and its constitutive activation is 7. Schneider, J. S., Moore, D. H., and Sagebiel, R. W. Risk factors for melanoma required for oncogenic transformation in NIH3T3 cells. Am. J. Pathol., 159: incidence in prospective follow-up. The importance of atypical (dysplastic) nevi. 431– 437, 2001. Arch. Dermatol., 130: 1002–1007, 1994. 35. Haas-Kogan, D., Shalev, N., Wong, M., Mills, G., Yount, G., and Stokoe, D. Protein 8. Rigel, D. S., and Carucci, J. A. Malignant melanoma: prevention, early detection, and kinase B (PKB/Akt) activity is elevated in glioblastoma cells due to mutation of the treatment in the 21st century. CA Cancer J. Clin., 50: 215–236, 2000. tumor suppressor PTEN/MMAC. Curr. Biol., 8: 1195–1198, 1998. 9. Duncan, L. M., Berwick, M., Bruijn, J. A., Byers, H. R., Mihm, M. C., and Barnhill, 36. Dahia, P. L., Aguiar, R. C., Alberta, J., Kum, J. B., Caron, S., Sill, H., Marsh, D. J., R. L. Histopathologic recognition and grading of dysplastic melanocytic nevi: an Ritz, J., Freedman, A., Stiles, C., and Eng, C. PTEN is inversely correlated with the interobserver agreement study. J. Investig. Dermatol., 100: 318SϪ321S, 1993. cell survival factor Akt/PKB and is inactivated via multiple mechanisms in haema- 10. Duray, P. H., DerSimonian, R., Barnhill, R., Stenn, K., Ernstoff, M. S., Fine, J., and tological malignancies. Hum. Mol. Genet., 8: 185–193, 1999. Kirkwood, J. M. An analysis of interobserver recognition of the histopathologic 37. Dhawan, P., and Richmond, A. A novel NF-␬B-inducing kinase-MAPK signaling features of dysplastic nevi from a mixed group of nevomelanocytic lesions. J. Am. pathway up-regulates NF-␬B activity in melanoma cells. J. Biol. Chem., 277: 7920– Acad. Dermatol., 27: 741–749, 1992. 7928, 2002. 38. National Institutes of Health Consensus Development Conference Statement on Diagnosis and Treatment of Early Melanoma (January 27–29, 1992). Am. J. Der- 4 Dhawan, P., Richmond, A., unpublished data. matopathol. 15: 34–43, 1993. 7341

39. Andjelkovic, M., Alessi, D. R., Meier, R., Fernandez, A., Lamb, N. J., Frech, M., 45. Tachiiri, S., Sasai, K., Oya, N., and Hiraoka, M. Enhanced cell killing by overex- Cron, P., Cohen, P., Lucocq, J. M., and Hemmings, B. A. Role of translocation in the pression of dominant-negative phosphatidylinositol 3-kinase subunit, ⌬p85, follow- activation and function of protein kinase B. J. Biol. Chem., 272: 31515–31524, 1997. ing genotoxic stresses. Jpn. J. Cancer Res., 91: 1314–1318, 2001. 40. Camper-Kirby, D., Welch, S., Walker, A., Shiraishi, I., Setchell, K. D., Schaefer, 46. Li, G., Satyamoorthy, K., and Herlyn, M. N-Cadherin-mediated intercellular interac- E., Kajstura, J., Anversa, P., and Sussman, M. A. Myocardial Akt activation and tions promote survival and migration of melanoma cells. Cancer Res., 61: 3819– gender: increased nuclear activity in females versus males. Circ. Res., 25: 3825, 2001. 1020–1027, 2001. 47. Madrid, L. V., Wang, C. Y., Guttridge, D. C., Schottelius, A. J., Baldwin, A. S., Jr., 41. Tsatsanis, C., and Spandidos, D. A. The role of oncogenic kinases in human cancer. and Mayo, M. W. Akt suppresses apoptosis by stimulating the transactivation potential of the RelA/p65 subunit of NF-␬B. Mol. Cell. Biol., 20: 1626–1638, 2000. Int. J. Mol. Med., 5: 583–590, 2000. 48. Alessi, D. R., Caudwell, F. B., Andjelkovic, M., Hemmings, B. A., and Cohen, P. 42. Farmer, E. R., and Hood, A. F. Melanocytic tumors. In: E. R. Farm A. F. Hood (eds.) Molecular basis for the substrate specificity of protein kinase B; comparison with Pathology of the Skin, 2nd Ed., pp. 1105–1155. New York: McGraw-Hill, 2000. MAPKAP kinase-1 and p70 S6 kinase. FEBS Lett., 399: 333–338, 1996b. 43. Sonoda, Y., Ozawa, T., Aldape, K. D., Deen, D. F., Berger, M. S., and Pieper, R. O. 49. Waldmann, V., Wacker, J., and Deichmann, M. Absence of mutations in the pleck- Akt pathway activation converts anaplastic astrocytoma to glioblastoma multiforme strin homology (PH) domain of protein kinase B (PKB/Akt) in malignant melanoma. in a human astrocyte model of glioma. Cancer Res., 61: 6674–6678, 2001. Melanoma Res., 12: 45–50, 2002. 44. Hutchinson, J., Jin, J., Cardiff, R. D., Woodgett, J. R., and Muller, W. J. Activation 50. Waldmann, V., Wacker, J., and Deichmann, M. Mutations of the activation-associated of Akt (protein kinase B) in mammary epithelium provides a critical cell survival phosphorylation sites at codons 308 and 473 of protein kinase B are absent in human signal required for tumor progression. Mol. Cell. Biol., 21: 2203–2212, 2001. melanoma. Arch. Dermatol. Res., 293: 368–372, 2001.

7342

Downloaded from cancerres.aacrjournals.org on September 25, 2021. © 2002 American Association for Cancer Research. Constitutive Activation of Akt/Protein Kinase B in Melanoma Leads to Up-Regulation of Nuclear Factor- κB and Tumor Progression

Punita Dhawan, Amar B. Singh, Darrel L. Ellis, et al.

Cancer Res 2002;62:7335-7342.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/62/24/7335

Cited articles This article cites 47 articles, 13 of which you can access for free at: http://cancerres.aacrjournals.org/content/62/24/7335.full#ref-list-1

Citing articles This article has been cited by 40 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/62/24/7335.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/62/24/7335. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.