452 Original article
High expression of glycolytic and pigment proteins is associated with worse clinical outcome in stage III melanoma Johan Falkeniusa, Joakim Lundebergb, Hemming Johanssona, Rainer Tuominena, Marianne Frostvik-Stolta, Johan Hanssona and Suzanne Egyhazi Bragea There are insufficient numbers of prognostic factors available for prediction of clinical outcome in patients with stage III malignant cutaneous melanoma, even when known adverse pathological risk factors, such as macrometastasis, number of lymph node metastases, and ulceration are taken into consideration. The aim of this study was therefore to identify additional prognostic factors to better predict patients with a high risk of relapse, thus enabling us to better determine the need for adjuvant treatment in stage III disease. An RNA oligonucleotide microarray study was performed on first regional lymph node metastases in 42 patients with stage III melanoma: 23 patients with short-term survival (r 13 months) and 19 with long-term survival (Z 60 months), to identify genes associated with clinical outcome. Candidate genes were validated by real-time PCR and immunohistochemical analysis. Several gene ontology (GO) categories were highly significantly differentially expressed including glycolysis (GO: 0006096; P < 0.001) and the pigment biosynthetic process (GO: 0046148; P < 0.001), in which overexpression was associated with short-disease-specific survival. Three overexpressed glycolytic genes, GAPDHS, GAPDH, and PKM2, and two pigment-related genes, TYRP1 and OCA2, were selected for validation. A significant difference in
GAPDHS protein expression between short- and long-term survivors (P = 0.021) and a trend for PKM2 (P = 0.093) was observed in univariate analysis. Positive expression of at least two of four proteins (GAPDHS, GAPDH, PKM2, TYRP1) in immunohistochemical analysis was found to be an independent adverse prognostic factor for disease-specific survival (P = 0.011). Our results indicate that this prognostic panel in combination with established risk factors may contribute to an improved prediction of patients with a high risk of c 2013 Wolters relapse. Melanoma Res 23:452–460 Kluwer Health | Lippincott Williams & Wilkins.
Introduction
that genes related to immune response and proliferation have an impact on melanoma progression [4–8], but few candidate genes have been overlapping between reports. These studies have predominantly focused on comparing different stages of the disease to identify genes involved in progression. So far, there have been few transcriptome studies investigating stage III melanoma patients [9,10] and only one study focusing specifically on lymph node metastases [10], which may be because of difficulties in accessing fresh frozen regional lymph node metastasis specimens.
The incidence of malignant cutaneous melanoma has steadily increased over many decades in white populations [1]. Patients in stage III with regional nodal metastases have variable clinical outcomes [2,3], with a reported 5-year survival of 23–87% depending on the substage (A–C) [3]. The already established risk factors, on the basis of histopathology, such as micrometastasis or macrometastasis, number of lymph node metastases, and ulceration of the primary tumor are not sufficient to accurately predict clinical outcome in stages IIIA–C. There is thus a need to find better prognostic markers to identify stage III melanoma patients with a high risk of relapse. Molecular characteristics of the tumor probably play an important role in clinical outcome but are still insufficiently studied in stage III melanoma. Over the past few years, microarray technology has been used to search for gene expression profiles that correlate with progression or survival. Several studies have demonstrated c 2013 Wolters Kluwer Health | Lippincott Williams & Wilkins 0960-8931
Melanoma Research 2013, 23:452–460 Keywords: glycolysis, melanoma, pigment biosynthetic process, prognosis, stage III a Department of Oncology-Pathology, Karolinska Institutet, Cancer Center Karolinska, Karolinska University Hospital and bScience for Life Laboratory, Department of Gene Technology, Royal Institute of Technology, Stockholm, Sweden
Correspondence to Johan Falkenius, MD, Department of Oncology–Pathology, Karolinska Institutet, Cancer Center Karolinska, Karolinska University Hospital, Solna, Stockholm 171-76, Sweden Tel: + 46 8 51779237; e-mail: [email protected] Received 18 January 2013 Accepted 6 September 2013
It is of great importance to prevent progression from a locally advanced malignant melanoma to a generalized disease as treatment options for patients with disseminated, stage IV, disease are limited and the prognosis is poor [11]. There are, however, novel therapies such as inhibitors of BRAF-mutated protein, MEK inhibitors, and antibodies against CTLA4 and PD-1, as well as the PD-1 ligand, which have shown promising initial clinical results [12–14]. Some of these drugs are presently being DOI: 10.1097/CMR.0000000000000027
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Prognostic markers in stage III melanoma Falkenius et al. 453
investigated as potential adjuvant therapies in targeted patient populations, providing greater opportunities to treat stage III melanoma. The aim of the present study was to identify prognostic biomarkers that can be used to improve the identification of stage III melanoma patients with a high risk of relapse. This transcriptome study on first regional lymph node metastases has been performed in stage IIIB–C melanoma patients with either short- or long-term survival after lymphadenectomy.
Material and methods Patients and specimens
Two cohorts were selected on the basis of survival time from regional lymph node dissection to death or last follow-up. Forty-two patients with stage IIIB–C disease who underwent lymphadenectomy between 1994 and 2002 at the Karolinska University Hospital, Solna, were included in this study: 19 patients with long-term survival (Z 60 months) and 23 with short-term survival (r 13 months). One specimen from the first-detected regional lymph node macrometastasis of each patient was selected. The specimens were fresh frozen in liquid nitrogen and kept at – 701C in a biobank until analysis. The last follow-up was in June 2013. The cause of death was generalized melanoma for all patients in the shortterm survival group. Three of the long-term survivors (16%) showed relapses with locoregional recurrence within 5 years after lymphadenectomy and one of them started palliative chemotherapy 5 years after lymphadenectomy. Patient characteristics and pathological characteristics of primary tumors and lymph node metastases are shown in Table 1. The proportion of tumor cells was estimated by a pathologist. Only samples with greater than 50% tumor cells were included for further analysis. The majority of the specimens (67%) contained 70% or more malignant cells. In total, 56 specimens were included and among them 12 specimens were excluded because of inappropriate RNA quality (n = 8), insufficient amount of RNA (n = 2), and low proportion of tumor cells (n = 2). Two specimens were excluded because of technical failure during the array procedure. This study has been approved by the Research Ethics Committee of Karolinska Institutet (Dnr: 2006/1373-31/3). All involved individuals have provided informed consent to be included in the study. RNA preparation of tumor samples
The tumor biopsies were embedded in cryogel (OCT, Histolab, Gothenburg, Sweden) before sectioning. Sections were placed in lysis buffer (RNeasy Mini Kit, Qiagen, Hilden, Germany). Proteinase K was added and extraction was performed according to Qiagen’s RNeasy
Mini Kit protocol (Qiagen); DNase treatment was also performed (Qiagen DNase Kit). RNA quality was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, California, USA). RNA amplification
RNA was amplified using the RiboAmp II kit (Applied Biosystems, Foster City, California, USA). The kit is based on in-vitro transcription using a T7 promoter with a corresponding RNA polymerase (MEGAscript; Applied Biosystems), which yields amplified antisense RNA (aRNA). For all tumors, 500 ng of total RNA was used as the input for the amplification. RNA reference
Universal Human Reference RNA (Stratagene, Agilent Technologies, Santa Clara, California, USA) was used as reference RNA. The reference RNA was amplified and labeled according to the same procedure used for the patient material. Labeling and cDNA synthesis
Labeling was performed using the ULS aRNA Fluorescent Labeling Kit (Kreatech Diagnostics, Amsterdam, the Netherlands), which allows for labeling of unmodified aRNA. A 2-mg sample was used as the input in the labeling reaction. The ULS molecule consists of a platinum complex with a bound detectable molecule (cy5 or cy3), which under the right buffering conditions will form a co-coordinative bond at position N7 of the guanine base. Oligonucleotide microarray
The microarray used in this work is an in-house printed array manufactured at the KTH core facility. The spotted 70-mer oligonucleotides originate from version 3.03 of Operons Human Genome Oligo Set (Operon, Huntsville, Albuquerque, USA), and the microarray contains 35 344 features (spots), representing 28 948 Entrez Gene Ids [15], of which 17 972 are unique. Hybridization
Prehybridization of microarray slides has been described previously [16]. Fragmentation was performed after labeling of the aRNA with RNA fragmentation reagent, AM870 (Applied Biosystems), yielding aRNA fragments of 60–200 nucleotides. The cy5-labeled and cy3-labeled aRNA fragments were then pooled in equivalent amounts and mixed with the hybridization buffer [5 SSC, 25% formamide (Sigma-Aldrich, St Louis, Missouri, USA), 0.1% SDS, 25% Kreablock (Kreatech Diagnostics), 10 mg herring sperm DNA (Invitrogen, Carlsbad, California, USA), and 10 mg yeast tRNA (Invitrogen)]. The hybridization mixture was then heat denatured at 701C for 3 min and subsequently placed on ice until injection in a MAUI hybridization mixer (BioMicro Systems, Salt Lake City,
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
454
Melanoma Research 2013, Vol 23 No 6
Table 1
Baseline patient and pathological characteristics
Characteristics
Short-term (%)
Sex Male 14 (61) Female 9 (39) Age at diagnosis (years) Median (range)] 68 (23–85) Age at lymphadenectomy (years) Median (range) 68 (24–86) OS after lymphadenectomy (months) Median (range) 6 (2–13) Type of melanoma SSM 12 (52) Other than SSM 11 (48) Primary tumor site Trunk 18 (78) Other than trunk 5 (22) Breslow thickness (mm) Median (range) 3.1 (0.5–30.0) Ulceration Present 15 (65) Absent 8 (35) Number of metastatic nodes 1 5 (22) 2–3 14 (61) >4 4 (17) Periglandular tumor growth Present 8 (35) Absent 15 (65) Adjuvant therapy Radiotherapy 6 (26) Interferon 2 (9) No therapy 13 (56) Missing data 2 (9) Palliative therapy Chemotherapy 9 (39) Mutation status BRAF mutated 13 (57) NRAS mutated 3 (13) Wild-type BRAF and NRAS 7 (30)
Long-term (%)
P-value
13 (68) 6 (32)
0.75
53 (17–86)
0.10
58 (18–91)
0.34
a
141 (60–226) 9 (47) 10 (53)
1.00
12 (63) 7 (37)
0.32
1.2 (0.4–7.0)
0.016
3 (16) 16 (84)
0.002
13 (68) 3 (16) 3 (16)
0.005
5 (26) 14 (74)
0.74
1 (5) 2 (11) 15 (79) 1(5) 0 6 (32) 8 (42) 5 (26)
viously [16]. On average, 64% of all features remained after the filtering steps. All microarrays were normalized with print-tip loess normalization [20]. Before investigating differences between long- and short-term survivors, features with low variance (interquartile range of the M-values across all samples for each gene was 1). b Odds ratios and confidence intervals are estimated using exact logistic regression. c P-values from the observed sufficient statistics. d Selected genes: GAPDHS, GAPDH, PKM2, and TYRP1. e Selected genes: GAPDHS, GAPDH, and PKM2. f Selected genes: GAPDHS and PKM2.
associated with prognosis in any type of cancer. TYRP1, the catalase involved in the melanogenesis, has recently been correlated with an adverse clinical outcome in stage III–IV melanoma [29]. In our microarray study, GO categories were highly significantly different between the good and poor prognosis groups, indicating the importance of specific biological functions such as glycolysis, the pigment biosynthetic process, and immune response for clinical outcome. We have chosen to focus on and verify a subset of genes belonging to the two top GO categories, glycolysis and pigment biosynthesis, in which overexpression was associated with an adverse clinical outcome. It is has been well established that reprogramming of metabolism is an important hallmark of cancer [30]. Hence, metabolism could be an interesting therapeutic strategy against melanoma. Today we have insufficient knowledge on the details even though several recent studies on melanoma have demonstrated the importance of different metabolic pathways [31,32]. The clinical application of fluorodeoxyglucose positron emission tomography (18F-FDG-PET) in cancer imaging shows that upregulated glycolysis is common in primary and metastatic cancers [33]. The accumulation of lactate within tumors due to upregulation of glycolysis has been reported to be associated with a poor clinical outcome [34]. Results from our study show that this also applies to cutaneous melanoma.
GAPDHS is a sperm-specific isoform of somatic GAPDH, a well-known glycolytic key enzyme, and the genes show 68% sequence identity in humans [35]. GAPDHS may be a substitute for GAPDH because of the high grade of sequence identity and may activate the same pathway. There is limited documentation of the functions of GAPDHS, besides its role in sperm metabolism. The possible overlap between GAPDH and GAPDHS may facilitate more efficient metabolism in melanoma cells. A metabolic symbiosis of oxidative phosphorylation and glycolysis has been suggested to occur in stage IV melanoma with normal serum LDH levels, whereas in tumors with high serum LDH levels, the main source of energy is glycolysis [32]. Elesclomol is a potential drug with proapoptotic activity, which induces reactive oxygen species and apoptosis through oxidative phosphorylation [36]. Results from a phase III study on elesclomol plus paclitaxel versus paclitaxel alone as treatment for patients with stage IV melanoma showed a significant improvement in progression free survival for the combination in the subgroup with normal LDH levels, supporting the importance of oxidative phosphorylation [37]. One of our top GO categories is oxidative phosphorylation (data not shown), which is more highly expressed among short-term survivors, indicating that both glycolysis and oxidative phosphorylation together may have an impact on the clinical outcome in stage III melanoma.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Prognostic markers in stage III melanoma Falkenius et al.
There are several potential candidate targeted drugs available against glycolysis. 2-Deoxyglucose (2-DG) is an interesting inhibitor of glucose metabolism. Because of tumor dependence on glycolysis, 2-DG has been considered as a potential anticancer agent, especially in combination with other cytotoxic agents [38]. Both GAPDH and GAPDHS would be suitable targets for 2-DG. In addition, recent in-vitro findings have shown that GAPDH overexpression reduces the effect of the BRAF inhibitor, vemurafenib, in melanoma, suggesting that inhibition of GAPDH/GAPDHS may improve the efficacy of vemurafenib [39]. It would therefore be interesting to investigate whether GAPDH/GAPDHS contributes to the acquired clinical BRAF inhibitor resistance. It is also well known that pigment genes, involved in melanocyte development, play an important role in the pathogenesis of cutaneous melanoma [40]. Hence, not surprisingly, the pigment biosynthetic process (GO: 0046148; P < 0.001) was one of the top GO categories in our microarray, which was associated with poor outcome. At some point in early melanoma pathogenesis changes in genotype and phenotype of melanocytes involving reactive oxygen species and oxidative stress occur, which are crucial steps in the oncogenesis process [41]. TYRP2, which downregulates oxidative stress and is correlated with antiapoptotic functions [42], belongs to the same GO group as TYRP1 and has previously been linked to adverse clinical outcome in melanoma [4]. TYRP1 has become an interesting target for monoclonal antibodies and inhibition of melanoma progression has been demonstrated in xenograft models [43]. In other transcriptome studies on stage III–IV melanoma patients, there is a limited overlap between the genespecific results among studies, possibly related to differences in study design, selection of cohorts of patients from different melanoma stages, as well as differences in array platforms. The most recent study to date [10] has a study design similar to ours; it compares patients with macroscopic stage III disease with short-term survival of below 1 year with those with long-term survival of above 4 years. In accordance with our study, that study reported differential expression of immune-related genes and a few metabolic and pigment genes. In our study, the presence of NRAS mutation was correlated with a better prognosis, which is in contrast to the findings of a previous report on stage IV disease [44]. One confounding factor could be that in our study the patients with NRAS mutation have a lower median age (55 years) compared with patients with wild-type NRAS (66 years). In contrast, the presence of BRAF mutation was associated with adverse clinical outcome, which supports the results from other studies on stage III–IV melanomas [45]. We have a limited number of samples and need to study a larger cohort for validation of these findings. The stronger correlation on combining results from four genes compared with fewer genes probably reflects the
459
heterogeneity of gene and protein expression both within and between tumors, which is a well-known phenomenon in melanoma [46]. These findings, together with other results, indicate that it may be more relevant to focus on activity in pathways rather than expression of single genes or proteins [4–10]. We have conducted a study of two distinctly separated prognostic groups of stage III melanoma patients to identify candidate prognostic factors. Microarray RNA expression analysis identified some GO categories associated with poor prognosis: in particular glucose metabolism and melanin synthesis, whereas immunerelated genes were associated with a good prognosis. In subsequent validation analyses, we focused on glycolysisassociated and melanin synthesis-associated gene products. GAPDHS, together with GAPDH, PKM2, and TYRP1, was found to be an independent adverse prognostic factor for overall survival in stage III melanoma. GAPDHS, in particular, is an interesting potential prognostic candidate for further study, which is significantly associated with adverse clinical outcome in the univariate analysis. The intermediary prognostic group with a survival of 14–59 months represents a large patient group that was not included in this study, which could be considered as a limitation. For subsequent validation of our identified set of IHC markers, an unselected population of patients will be analyzed. This panel, in combination with the known pathological risk factors, may improve the identification of patients with a high risk of relapse and in need of adjuvant treatment.
Acknowledgements The authors thank Diana Linde´n for valuable cooperation in providing clinical and pathological data, Lena Kanter for valuable evaluation of hematoxylin-stained sections, Johan Lindberg for excellent performance of microarray and data analysis, and Liss Garberg for excellent technical assistance. This study was funded by research grants from the Swedish Cancer Society, Radiumhemmet Research Funds, Karolinska Institutet Research Funds, and Stockholm County Council (ALF). Conflicts of interest
There are no conflicts of interest.
References 1
2
3
Erdmann F, Lortet-Tieulent J, Schuz J, Zeeb H, Greinert R, Breitbart EW, et al. International trends in the incidence of malignant melanoma 1953–2008 – are recent generations at higher or lower risk? Int J Cancer 2013; 132:385–400. Pilko G, Besic N, Zgajnar J, Hocevar M. Prognostic heterogeneity after the excision of lymph node metastases in patients with cutaneous melanoma. Surg Oncol 2011; 20:26–34. Balch CM, Gershenwald JE, Soong SJ, Thompson JF, Ding S, Byrd DR, et al. Multivariate analysis of prognostic factors among 2,313 patients with stage
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
460 Melanoma Research 2013, Vol 23 No 6
4
5
6
7
8
9
10
11 12
13
14
15 16
17 18 19
20
21
22
23
24 25
III melanoma: comparison of nodal micrometastases versus macrometastases. J Clin Oncol 2010; 28:2452–2459. Mandruzzato S, Callegaro A, Turcatel G, Francescato S, Montesco MC, Chiarion-Sileni V, et al. A gene expression signature associated with survival in metastatic melanoma. J Transl Med 2006; 4:50. Jonsson G, Busch C, Knappskog S, Geisler J, Miletic H, Ringner M, et al. Gene expression profiling-based identification of molecular subtypes in stage IV melanomas with different clinical outcome. Clin Cancer Res 2010; 16:3356–3367. John T, Black MA, Toro TT, Leader D, Gedye CA, Davis ID, et al. Predicting clinical outcome through molecular profiling in stage III melanoma. Clin Cancer Res 2008; 14:5173–5180. Bogunovic D, O’Neill DW, Belitskaya-Levy I, Vacic V, Yu YL, Adams S, et al. Immune profile and mitotic index of metastatic melanoma lesions enhance clinical staging in predicting patient survival. Proc Natl Acad Sci USA 2009; 106:20429–20434. Winnepenninckx V, Lazar V, Michiels S, Dessen P, Stas M, Alonso SR, et al. Gene expression profiling of primary cutaneous melanoma and clinical outcome. J Natl Cancer Inst 2006; 98:472–482. Journe F, Id Boufker H, Van Kempen L, Galibert MD, Wiedig M, Sales F, et al. TYRP1 mRNA expression in melanoma metastases correlates with clinical outcome. Br J Cancer 2011; 105:1726–1732. Mann GJ, Pupo GM, Campain AE, Carter CD, Schramm SJ, Pianova S, et al. BRAF mutation, NRAS mutation, and the absence of an immune-related expressed gene profile predict poor outcome in patients with stage III melanoma. J Invest Dermatol 2013; 133:509–517. Bhatia S, Tykodi SS, Thompson JA. Treatment of metastatic melanoma: an overview. Oncology (Williston Park) 2009; 23:488–496. Flaherty KT, Infante JR, Daud A, Gonzalez R, Kefford RF, Sosman J, et al. Combined BRAF and MEK inhibition in melanoma with BRAF V600 mutations. N Engl J Med 2012; 367:1694–1703. Hodi FS, O’Day SJ, McDermott DF, Weber RW, Sosman JA, Haanen JB, et al. Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med 2010; 363:711–723. Simeone E, Ascierto PA. Immunomodulating antibodies in the treatment of metastatic melanoma: the experience with anti-CTLA-4, anti-CD137, and anti-PD1. J Immunotoxicol 2012; 9:241–247. Maglott D. Entrez Gene: gene-centered information at NCBI. Nucleic Acids Res 2005; 33:D54–D58. Lindberg J, af Klint E, Ulfgren AK, Stark A, Andersson T, Nilsson P, et al. Variability in synovial inflammation in rheumatoid arthritis investigated by microarray technology. Arthritis Res Ther 2006; 8:R47. Yang YH, Speed T. Design issues for cDNA microarray experiments. Nat Rev Genet 2002; 3:579–588. Team R: A language and environment for statistical computing. Available at: http://www.R-project.org [Accessed 27 February 2008]. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, et al. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 2004; 5:R80. Yang YH, Dudoit S, Luu P, Lin DM, Peng V, Ngai J, et al. Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res 2002; 30:e15. Smyth GK. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol 2004; 3:1–25, Article 3. Reiner A, Yekutieli D, Benjamini Y. Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics 2003; 19:368–375. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 2000; 25:25–29. Kim SY, Volsky DJ. PAGE: parametric analysis of gene set enrichment. BMC Bioinformatics 2005; 6:144. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA 2005; 102:15545–15550.
26
27
28
29
30 31
32
33 34
35
36
37
38
39
40 41 42
43
44
45
46
Edlundh-Rose E, Egyhazi S, Omholt K, Mansson-Brahme E, Platz A, Hansson J, et al. NRAS and BRAF mutations in melanoma tumours in relation to clinical characteristics: a study based on mutation screening by pyrosequencing. Melanoma Res 2006; 16:471–478. Peng XC, Gong FM, Zhao YW, Zhou LX, Xie YW, Liao HL, et al. Comparative proteomic approach identifies PKM2 and cofilin-1 as potential diagnostic, prognostic and therapeutic targets for pulmonary adenocarcinoma. PLoS One 2011; 6:e27309. Revillion F, Pawlowski V, Hornez L, Peyrat JP. Glyceraldehyde-3-phosphate dehydrogenase gene expression in human breast cancer. Eur J Cancer 2000; 36:1038–1042. El Hajj P, Journe F, Wiedig M, Laios I, Sales F, Galibert MD, et al. Tyrosinaserelated protein 1 mRNA expression in lymph node metastases predicts overall survival in high-risk melanoma patients. Br J Cancer 2013; 108:1641–1647. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011; 144:646–674. Scott DA, Richardson AD, Filipp FV, Knutzen CA, Chiang GG, Ronai ZA, et al. Comparative metabolic flux profiling of melanoma cell lines: beyond the Warburg effect. J Biol Chem 2011; 286:42626–42634. Ho J, de Moura MB, Lin Y, Vincent G, Thorne S, Duncan LM, et al. Importance of glycolysis and oxidative phosphorylation in advanced melanoma. Mol Cancer 2012; 11:76. Gatenby RA, Gillies RJ. Glycolysis in cancer: a potential target for therapy. Int J Biochem Cell Biol 2007; 39:1358–1366. Kennedy KM, Dewhirst MW. Tumor metabolism of lactate: the influence and therapeutic potential for MCT and CD147 regulation. Future Oncol 2010; 6:127–148. Welch JE, Brown PL, O’Brien DA, Magyar PL, Bunch DO, Mori C, et al. Human glyceraldehyde 3-phosphate dehydrogenase-2 gene is expressed specifically in spermatogenic cells. J Androl 2000; 21:328–338. Blackman RK, Cheung-Ong K, Gebbia M, Proia DA, He S, Kepros J, et al. Mitochondrial electron transport is the cellular target of the oncology drug elesclomol. PLoS One 2012; 7:e29798. O’Day SJ, Eggermont AM, Chiarion-Sileni V, Kefford R, Grob JJ, Mortier L, et al. Final results of phase III SYMMETRY study: randomized, double-blind trial of elesclomol plus paclitaxel versus paclitaxel alone as treatment for chemotherapy-naive patients with advanced melanoma. J Clin Oncol 2013; 31:1211–1218. Raez LE, Papadopoulos K, Ricart AD, Chiorean EG, Dipaola RS, Stein MN, et al. A phase I dose-escalation trial of 2-deoxy-D-glucose alone or combined with docetaxel in patients with advanced solid tumors. Cancer Chemother Pharmacol 2013; 71:523–530. Hall A, Meyle KD, Lange MK, Klima M, Sanderhoff M, Dahl C, et al. Dysfunctional oxidative phosphorylation makes malignant melanoma cells addicted to glycolysis driven by the (V600E)BRAF oncogene. Oncotarget 2013; 4:584–599. Uong A, Zon LI. Melanocytes in development and cancer. J Cell Physiol 2010; 222:38–41. Meyskens FL Jr, Farmer P, Fruehauf JP. Redox regulation in human melanocytes and melanoma. Pigment Cell Res 2001; 14:148–154. Pak BJ, Chu W, Lu SJ, Kerbel RS, Ben-David Y. Lineage-specific mechanism of drug and radiation resistance in melanoma mediated by tyrosinase-related protein 2. Cancer Metastasis Rev 2001; 20:27–32. Saenger YM, Li Y, Chiou KC, Chan B, Rizzuto G, Terzulli SL, et al. Improved tumor immunity using anti-tyrosinase related protein-1 monoclonal antibody combined with DNA vaccines in murine melanoma. Cancer Res 2008; 68:9884–9891. Jakob JA, Bassett RL Jr, Ng CS, Curry JL, Joseph RW, Alvarado GC, et al. NRAS mutation status is an independent prognostic factor in metastatic melanoma. Cancer 2012; 118:4014–4023. Long GV, Menzies AM, Nagrial AM, Haydu LE, Hamilton AL, Mann GJ, et al. Prognostic and clinicopathologic associations of oncogenic BRAF in metastatic melanoma. J Clin Oncol 2011; 29:1239–1246. Harbst K, Staaf J, Masback A, Olsson H, Ingvar C, Vallon-Christersson J, et al. Multiple metastases from cutaneous malignant melanoma patients may display heterogeneous genomic and epigenomic patterns. Melanoma Res 2010; 20:381–391.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
High expression of glycolytic and pigment proteins is associated with worse clinical outcome in stage III melanoma Johan Falkeniusa, Joakim Lundebergb, Hemming Johanssona, Rainer Tuominena, Marianne Frostvik-Stolta, Johan Hanssona and Suzanne Egyhazi Bragea There are insufficient numbers of prognostic factors available for prediction of clinical outcome in patients with stage III malignant cutaneous melanoma, even when known adverse pathological risk factors, such as macrometastasis, number of lymph node metastases, and ulceration are taken into consideration. The aim of this study was therefore to identify additional prognostic factors to better predict patients with a high risk of relapse, thus enabling us to better determine the need for adjuvant treatment in stage III disease. An RNA oligonucleotide microarray study was performed on first regional lymph node metastases in 42 patients with stage III melanoma: 23 patients with short-term survival (r 13 months) and 19 with long-term survival (Z 60 months), to identify genes associated with clinical outcome. Candidate genes were validated by real-time PCR and immunohistochemical analysis. Several gene ontology (GO) categories were highly significantly differentially expressed including glycolysis (GO: 0006096; P < 0.001) and the pigment biosynthetic process (GO: 0046148; P < 0.001), in which overexpression was associated with short-disease-specific survival. Three overexpressed glycolytic genes, GAPDHS, GAPDH, and PKM2, and two pigment-related genes, TYRP1 and OCA2, were selected for validation. A significant difference in
GAPDHS protein expression between short- and long-term survivors (P = 0.021) and a trend for PKM2 (P = 0.093) was observed in univariate analysis. Positive expression of at least two of four proteins (GAPDHS, GAPDH, PKM2, TYRP1) in immunohistochemical analysis was found to be an independent adverse prognostic factor for disease-specific survival (P = 0.011). Our results indicate that this prognostic panel in combination with established risk factors may contribute to an improved prediction of patients with a high risk of c 2013 Wolters relapse. Melanoma Res 23:452–460 Kluwer Health | Lippincott Williams & Wilkins.
Introduction
that genes related to immune response and proliferation have an impact on melanoma progression [4–8], but few candidate genes have been overlapping between reports. These studies have predominantly focused on comparing different stages of the disease to identify genes involved in progression. So far, there have been few transcriptome studies investigating stage III melanoma patients [9,10] and only one study focusing specifically on lymph node metastases [10], which may be because of difficulties in accessing fresh frozen regional lymph node metastasis specimens.
The incidence of malignant cutaneous melanoma has steadily increased over many decades in white populations [1]. Patients in stage III with regional nodal metastases have variable clinical outcomes [2,3], with a reported 5-year survival of 23–87% depending on the substage (A–C) [3]. The already established risk factors, on the basis of histopathology, such as micrometastasis or macrometastasis, number of lymph node metastases, and ulceration of the primary tumor are not sufficient to accurately predict clinical outcome in stages IIIA–C. There is thus a need to find better prognostic markers to identify stage III melanoma patients with a high risk of relapse. Molecular characteristics of the tumor probably play an important role in clinical outcome but are still insufficiently studied in stage III melanoma. Over the past few years, microarray technology has been used to search for gene expression profiles that correlate with progression or survival. Several studies have demonstrated c 2013 Wolters Kluwer Health | Lippincott Williams & Wilkins 0960-8931
Melanoma Research 2013, 23:452–460 Keywords: glycolysis, melanoma, pigment biosynthetic process, prognosis, stage III a Department of Oncology-Pathology, Karolinska Institutet, Cancer Center Karolinska, Karolinska University Hospital and bScience for Life Laboratory, Department of Gene Technology, Royal Institute of Technology, Stockholm, Sweden
Correspondence to Johan Falkenius, MD, Department of Oncology–Pathology, Karolinska Institutet, Cancer Center Karolinska, Karolinska University Hospital, Solna, Stockholm 171-76, Sweden Tel: + 46 8 51779237; e-mail: [email protected] Received 18 January 2013 Accepted 6 September 2013
It is of great importance to prevent progression from a locally advanced malignant melanoma to a generalized disease as treatment options for patients with disseminated, stage IV, disease are limited and the prognosis is poor [11]. There are, however, novel therapies such as inhibitors of BRAF-mutated protein, MEK inhibitors, and antibodies against CTLA4 and PD-1, as well as the PD-1 ligand, which have shown promising initial clinical results [12–14]. Some of these drugs are presently being DOI: 10.1097/CMR.0000000000000027
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Prognostic markers in stage III melanoma Falkenius et al. 453
investigated as potential adjuvant therapies in targeted patient populations, providing greater opportunities to treat stage III melanoma. The aim of the present study was to identify prognostic biomarkers that can be used to improve the identification of stage III melanoma patients with a high risk of relapse. This transcriptome study on first regional lymph node metastases has been performed in stage IIIB–C melanoma patients with either short- or long-term survival after lymphadenectomy.
Material and methods Patients and specimens
Two cohorts were selected on the basis of survival time from regional lymph node dissection to death or last follow-up. Forty-two patients with stage IIIB–C disease who underwent lymphadenectomy between 1994 and 2002 at the Karolinska University Hospital, Solna, were included in this study: 19 patients with long-term survival (Z 60 months) and 23 with short-term survival (r 13 months). One specimen from the first-detected regional lymph node macrometastasis of each patient was selected. The specimens were fresh frozen in liquid nitrogen and kept at – 701C in a biobank until analysis. The last follow-up was in June 2013. The cause of death was generalized melanoma for all patients in the shortterm survival group. Three of the long-term survivors (16%) showed relapses with locoregional recurrence within 5 years after lymphadenectomy and one of them started palliative chemotherapy 5 years after lymphadenectomy. Patient characteristics and pathological characteristics of primary tumors and lymph node metastases are shown in Table 1. The proportion of tumor cells was estimated by a pathologist. Only samples with greater than 50% tumor cells were included for further analysis. The majority of the specimens (67%) contained 70% or more malignant cells. In total, 56 specimens were included and among them 12 specimens were excluded because of inappropriate RNA quality (n = 8), insufficient amount of RNA (n = 2), and low proportion of tumor cells (n = 2). Two specimens were excluded because of technical failure during the array procedure. This study has been approved by the Research Ethics Committee of Karolinska Institutet (Dnr: 2006/1373-31/3). All involved individuals have provided informed consent to be included in the study. RNA preparation of tumor samples
The tumor biopsies were embedded in cryogel (OCT, Histolab, Gothenburg, Sweden) before sectioning. Sections were placed in lysis buffer (RNeasy Mini Kit, Qiagen, Hilden, Germany). Proteinase K was added and extraction was performed according to Qiagen’s RNeasy
Mini Kit protocol (Qiagen); DNase treatment was also performed (Qiagen DNase Kit). RNA quality was assessed using the Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, California, USA). RNA amplification
RNA was amplified using the RiboAmp II kit (Applied Biosystems, Foster City, California, USA). The kit is based on in-vitro transcription using a T7 promoter with a corresponding RNA polymerase (MEGAscript; Applied Biosystems), which yields amplified antisense RNA (aRNA). For all tumors, 500 ng of total RNA was used as the input for the amplification. RNA reference
Universal Human Reference RNA (Stratagene, Agilent Technologies, Santa Clara, California, USA) was used as reference RNA. The reference RNA was amplified and labeled according to the same procedure used for the patient material. Labeling and cDNA synthesis
Labeling was performed using the ULS aRNA Fluorescent Labeling Kit (Kreatech Diagnostics, Amsterdam, the Netherlands), which allows for labeling of unmodified aRNA. A 2-mg sample was used as the input in the labeling reaction. The ULS molecule consists of a platinum complex with a bound detectable molecule (cy5 or cy3), which under the right buffering conditions will form a co-coordinative bond at position N7 of the guanine base. Oligonucleotide microarray
The microarray used in this work is an in-house printed array manufactured at the KTH core facility. The spotted 70-mer oligonucleotides originate from version 3.03 of Operons Human Genome Oligo Set (Operon, Huntsville, Albuquerque, USA), and the microarray contains 35 344 features (spots), representing 28 948 Entrez Gene Ids [15], of which 17 972 are unique. Hybridization
Prehybridization of microarray slides has been described previously [16]. Fragmentation was performed after labeling of the aRNA with RNA fragmentation reagent, AM870 (Applied Biosystems), yielding aRNA fragments of 60–200 nucleotides. The cy5-labeled and cy3-labeled aRNA fragments were then pooled in equivalent amounts and mixed with the hybridization buffer [5 SSC, 25% formamide (Sigma-Aldrich, St Louis, Missouri, USA), 0.1% SDS, 25% Kreablock (Kreatech Diagnostics), 10 mg herring sperm DNA (Invitrogen, Carlsbad, California, USA), and 10 mg yeast tRNA (Invitrogen)]. The hybridization mixture was then heat denatured at 701C for 3 min and subsequently placed on ice until injection in a MAUI hybridization mixer (BioMicro Systems, Salt Lake City,
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
454
Melanoma Research 2013, Vol 23 No 6
Table 1
Baseline patient and pathological characteristics
Characteristics
Short-term (%)
Sex Male 14 (61) Female 9 (39) Age at diagnosis (years) Median (range)] 68 (23–85) Age at lymphadenectomy (years) Median (range) 68 (24–86) OS after lymphadenectomy (months) Median (range) 6 (2–13) Type of melanoma SSM 12 (52) Other than SSM 11 (48) Primary tumor site Trunk 18 (78) Other than trunk 5 (22) Breslow thickness (mm) Median (range) 3.1 (0.5–30.0) Ulceration Present 15 (65) Absent 8 (35) Number of metastatic nodes 1 5 (22) 2–3 14 (61) >4 4 (17) Periglandular tumor growth Present 8 (35) Absent 15 (65) Adjuvant therapy Radiotherapy 6 (26) Interferon 2 (9) No therapy 13 (56) Missing data 2 (9) Palliative therapy Chemotherapy 9 (39) Mutation status BRAF mutated 13 (57) NRAS mutated 3 (13) Wild-type BRAF and NRAS 7 (30)
Long-term (%)
P-value
13 (68) 6 (32)
0.75
53 (17–86)
0.10
58 (18–91)
0.34
a
141 (60–226) 9 (47) 10 (53)
1.00
12 (63) 7 (37)
0.32
1.2 (0.4–7.0)
0.016
3 (16) 16 (84)
0.002
13 (68) 3 (16) 3 (16)
0.005
5 (26) 14 (74)
0.74
1 (5) 2 (11) 15 (79) 1(5) 0 6 (32) 8 (42) 5 (26)
viously [16]. On average, 64% of all features remained after the filtering steps. All microarrays were normalized with print-tip loess normalization [20]. Before investigating differences between long- and short-term survivors, features with low variance (interquartile range of the M-values across all samples for each gene was 1). b Odds ratios and confidence intervals are estimated using exact logistic regression. c P-values from the observed sufficient statistics. d Selected genes: GAPDHS, GAPDH, PKM2, and TYRP1. e Selected genes: GAPDHS, GAPDH, and PKM2. f Selected genes: GAPDHS and PKM2.
associated with prognosis in any type of cancer. TYRP1, the catalase involved in the melanogenesis, has recently been correlated with an adverse clinical outcome in stage III–IV melanoma [29]. In our microarray study, GO categories were highly significantly different between the good and poor prognosis groups, indicating the importance of specific biological functions such as glycolysis, the pigment biosynthetic process, and immune response for clinical outcome. We have chosen to focus on and verify a subset of genes belonging to the two top GO categories, glycolysis and pigment biosynthesis, in which overexpression was associated with an adverse clinical outcome. It is has been well established that reprogramming of metabolism is an important hallmark of cancer [30]. Hence, metabolism could be an interesting therapeutic strategy against melanoma. Today we have insufficient knowledge on the details even though several recent studies on melanoma have demonstrated the importance of different metabolic pathways [31,32]. The clinical application of fluorodeoxyglucose positron emission tomography (18F-FDG-PET) in cancer imaging shows that upregulated glycolysis is common in primary and metastatic cancers [33]. The accumulation of lactate within tumors due to upregulation of glycolysis has been reported to be associated with a poor clinical outcome [34]. Results from our study show that this also applies to cutaneous melanoma.
GAPDHS is a sperm-specific isoform of somatic GAPDH, a well-known glycolytic key enzyme, and the genes show 68% sequence identity in humans [35]. GAPDHS may be a substitute for GAPDH because of the high grade of sequence identity and may activate the same pathway. There is limited documentation of the functions of GAPDHS, besides its role in sperm metabolism. The possible overlap between GAPDH and GAPDHS may facilitate more efficient metabolism in melanoma cells. A metabolic symbiosis of oxidative phosphorylation and glycolysis has been suggested to occur in stage IV melanoma with normal serum LDH levels, whereas in tumors with high serum LDH levels, the main source of energy is glycolysis [32]. Elesclomol is a potential drug with proapoptotic activity, which induces reactive oxygen species and apoptosis through oxidative phosphorylation [36]. Results from a phase III study on elesclomol plus paclitaxel versus paclitaxel alone as treatment for patients with stage IV melanoma showed a significant improvement in progression free survival for the combination in the subgroup with normal LDH levels, supporting the importance of oxidative phosphorylation [37]. One of our top GO categories is oxidative phosphorylation (data not shown), which is more highly expressed among short-term survivors, indicating that both glycolysis and oxidative phosphorylation together may have an impact on the clinical outcome in stage III melanoma.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Prognostic markers in stage III melanoma Falkenius et al.
There are several potential candidate targeted drugs available against glycolysis. 2-Deoxyglucose (2-DG) is an interesting inhibitor of glucose metabolism. Because of tumor dependence on glycolysis, 2-DG has been considered as a potential anticancer agent, especially in combination with other cytotoxic agents [38]. Both GAPDH and GAPDHS would be suitable targets for 2-DG. In addition, recent in-vitro findings have shown that GAPDH overexpression reduces the effect of the BRAF inhibitor, vemurafenib, in melanoma, suggesting that inhibition of GAPDH/GAPDHS may improve the efficacy of vemurafenib [39]. It would therefore be interesting to investigate whether GAPDH/GAPDHS contributes to the acquired clinical BRAF inhibitor resistance. It is also well known that pigment genes, involved in melanocyte development, play an important role in the pathogenesis of cutaneous melanoma [40]. Hence, not surprisingly, the pigment biosynthetic process (GO: 0046148; P < 0.001) was one of the top GO categories in our microarray, which was associated with poor outcome. At some point in early melanoma pathogenesis changes in genotype and phenotype of melanocytes involving reactive oxygen species and oxidative stress occur, which are crucial steps in the oncogenesis process [41]. TYRP2, which downregulates oxidative stress and is correlated with antiapoptotic functions [42], belongs to the same GO group as TYRP1 and has previously been linked to adverse clinical outcome in melanoma [4]. TYRP1 has become an interesting target for monoclonal antibodies and inhibition of melanoma progression has been demonstrated in xenograft models [43]. In other transcriptome studies on stage III–IV melanoma patients, there is a limited overlap between the genespecific results among studies, possibly related to differences in study design, selection of cohorts of patients from different melanoma stages, as well as differences in array platforms. The most recent study to date [10] has a study design similar to ours; it compares patients with macroscopic stage III disease with short-term survival of below 1 year with those with long-term survival of above 4 years. In accordance with our study, that study reported differential expression of immune-related genes and a few metabolic and pigment genes. In our study, the presence of NRAS mutation was correlated with a better prognosis, which is in contrast to the findings of a previous report on stage IV disease [44]. One confounding factor could be that in our study the patients with NRAS mutation have a lower median age (55 years) compared with patients with wild-type NRAS (66 years). In contrast, the presence of BRAF mutation was associated with adverse clinical outcome, which supports the results from other studies on stage III–IV melanomas [45]. We have a limited number of samples and need to study a larger cohort for validation of these findings. The stronger correlation on combining results from four genes compared with fewer genes probably reflects the
459
heterogeneity of gene and protein expression both within and between tumors, which is a well-known phenomenon in melanoma [46]. These findings, together with other results, indicate that it may be more relevant to focus on activity in pathways rather than expression of single genes or proteins [4–10]. We have conducted a study of two distinctly separated prognostic groups of stage III melanoma patients to identify candidate prognostic factors. Microarray RNA expression analysis identified some GO categories associated with poor prognosis: in particular glucose metabolism and melanin synthesis, whereas immunerelated genes were associated with a good prognosis. In subsequent validation analyses, we focused on glycolysisassociated and melanin synthesis-associated gene products. GAPDHS, together with GAPDH, PKM2, and TYRP1, was found to be an independent adverse prognostic factor for overall survival in stage III melanoma. GAPDHS, in particular, is an interesting potential prognostic candidate for further study, which is significantly associated with adverse clinical outcome in the univariate analysis. The intermediary prognostic group with a survival of 14–59 months represents a large patient group that was not included in this study, which could be considered as a limitation. For subsequent validation of our identified set of IHC markers, an unselected population of patients will be analyzed. This panel, in combination with the known pathological risk factors, may improve the identification of patients with a high risk of relapse and in need of adjuvant treatment.
Acknowledgements The authors thank Diana Linde´n for valuable cooperation in providing clinical and pathological data, Lena Kanter for valuable evaluation of hematoxylin-stained sections, Johan Lindberg for excellent performance of microarray and data analysis, and Liss Garberg for excellent technical assistance. This study was funded by research grants from the Swedish Cancer Society, Radiumhemmet Research Funds, Karolinska Institutet Research Funds, and Stockholm County Council (ALF). Conflicts of interest
There are no conflicts of interest.
References 1
2
3
Erdmann F, Lortet-Tieulent J, Schuz J, Zeeb H, Greinert R, Breitbart EW, et al. International trends in the incidence of malignant melanoma 1953–2008 – are recent generations at higher or lower risk? Int J Cancer 2013; 132:385–400. Pilko G, Besic N, Zgajnar J, Hocevar M. Prognostic heterogeneity after the excision of lymph node metastases in patients with cutaneous melanoma. Surg Oncol 2011; 20:26–34. Balch CM, Gershenwald JE, Soong SJ, Thompson JF, Ding S, Byrd DR, et al. Multivariate analysis of prognostic factors among 2,313 patients with stage
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
460 Melanoma Research 2013, Vol 23 No 6
4
5
6
7
8
9
10
11 12
13
14
15 16
17 18 19
20
21
22
23
24 25
III melanoma: comparison of nodal micrometastases versus macrometastases. J Clin Oncol 2010; 28:2452–2459. Mandruzzato S, Callegaro A, Turcatel G, Francescato S, Montesco MC, Chiarion-Sileni V, et al. A gene expression signature associated with survival in metastatic melanoma. J Transl Med 2006; 4:50. Jonsson G, Busch C, Knappskog S, Geisler J, Miletic H, Ringner M, et al. Gene expression profiling-based identification of molecular subtypes in stage IV melanomas with different clinical outcome. Clin Cancer Res 2010; 16:3356–3367. John T, Black MA, Toro TT, Leader D, Gedye CA, Davis ID, et al. Predicting clinical outcome through molecular profiling in stage III melanoma. Clin Cancer Res 2008; 14:5173–5180. Bogunovic D, O’Neill DW, Belitskaya-Levy I, Vacic V, Yu YL, Adams S, et al. Immune profile and mitotic index of metastatic melanoma lesions enhance clinical staging in predicting patient survival. Proc Natl Acad Sci USA 2009; 106:20429–20434. Winnepenninckx V, Lazar V, Michiels S, Dessen P, Stas M, Alonso SR, et al. Gene expression profiling of primary cutaneous melanoma and clinical outcome. J Natl Cancer Inst 2006; 98:472–482. Journe F, Id Boufker H, Van Kempen L, Galibert MD, Wiedig M, Sales F, et al. TYRP1 mRNA expression in melanoma metastases correlates with clinical outcome. Br J Cancer 2011; 105:1726–1732. Mann GJ, Pupo GM, Campain AE, Carter CD, Schramm SJ, Pianova S, et al. BRAF mutation, NRAS mutation, and the absence of an immune-related expressed gene profile predict poor outcome in patients with stage III melanoma. J Invest Dermatol 2013; 133:509–517. Bhatia S, Tykodi SS, Thompson JA. Treatment of metastatic melanoma: an overview. Oncology (Williston Park) 2009; 23:488–496. Flaherty KT, Infante JR, Daud A, Gonzalez R, Kefford RF, Sosman J, et al. Combined BRAF and MEK inhibition in melanoma with BRAF V600 mutations. N Engl J Med 2012; 367:1694–1703. Hodi FS, O’Day SJ, McDermott DF, Weber RW, Sosman JA, Haanen JB, et al. Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med 2010; 363:711–723. Simeone E, Ascierto PA. Immunomodulating antibodies in the treatment of metastatic melanoma: the experience with anti-CTLA-4, anti-CD137, and anti-PD1. J Immunotoxicol 2012; 9:241–247. Maglott D. Entrez Gene: gene-centered information at NCBI. Nucleic Acids Res 2005; 33:D54–D58. Lindberg J, af Klint E, Ulfgren AK, Stark A, Andersson T, Nilsson P, et al. Variability in synovial inflammation in rheumatoid arthritis investigated by microarray technology. Arthritis Res Ther 2006; 8:R47. Yang YH, Speed T. Design issues for cDNA microarray experiments. Nat Rev Genet 2002; 3:579–588. Team R: A language and environment for statistical computing. Available at: http://www.R-project.org [Accessed 27 February 2008]. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, Dudoit S, et al. Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 2004; 5:R80. Yang YH, Dudoit S, Luu P, Lin DM, Peng V, Ngai J, et al. Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res 2002; 30:e15. Smyth GK. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol 2004; 3:1–25, Article 3. Reiner A, Yekutieli D, Benjamini Y. Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics 2003; 19:368–375. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet 2000; 25:25–29. Kim SY, Volsky DJ. PAGE: parametric analysis of gene set enrichment. BMC Bioinformatics 2005; 6:144. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA 2005; 102:15545–15550.
26
27
28
29
30 31
32
33 34
35
36
37
38
39
40 41 42
43
44
45
46
Edlundh-Rose E, Egyhazi S, Omholt K, Mansson-Brahme E, Platz A, Hansson J, et al. NRAS and BRAF mutations in melanoma tumours in relation to clinical characteristics: a study based on mutation screening by pyrosequencing. Melanoma Res 2006; 16:471–478. Peng XC, Gong FM, Zhao YW, Zhou LX, Xie YW, Liao HL, et al. Comparative proteomic approach identifies PKM2 and cofilin-1 as potential diagnostic, prognostic and therapeutic targets for pulmonary adenocarcinoma. PLoS One 2011; 6:e27309. Revillion F, Pawlowski V, Hornez L, Peyrat JP. Glyceraldehyde-3-phosphate dehydrogenase gene expression in human breast cancer. Eur J Cancer 2000; 36:1038–1042. El Hajj P, Journe F, Wiedig M, Laios I, Sales F, Galibert MD, et al. Tyrosinaserelated protein 1 mRNA expression in lymph node metastases predicts overall survival in high-risk melanoma patients. Br J Cancer 2013; 108:1641–1647. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011; 144:646–674. Scott DA, Richardson AD, Filipp FV, Knutzen CA, Chiang GG, Ronai ZA, et al. Comparative metabolic flux profiling of melanoma cell lines: beyond the Warburg effect. J Biol Chem 2011; 286:42626–42634. Ho J, de Moura MB, Lin Y, Vincent G, Thorne S, Duncan LM, et al. Importance of glycolysis and oxidative phosphorylation in advanced melanoma. Mol Cancer 2012; 11:76. Gatenby RA, Gillies RJ. Glycolysis in cancer: a potential target for therapy. Int J Biochem Cell Biol 2007; 39:1358–1366. Kennedy KM, Dewhirst MW. Tumor metabolism of lactate: the influence and therapeutic potential for MCT and CD147 regulation. Future Oncol 2010; 6:127–148. Welch JE, Brown PL, O’Brien DA, Magyar PL, Bunch DO, Mori C, et al. Human glyceraldehyde 3-phosphate dehydrogenase-2 gene is expressed specifically in spermatogenic cells. J Androl 2000; 21:328–338. Blackman RK, Cheung-Ong K, Gebbia M, Proia DA, He S, Kepros J, et al. Mitochondrial electron transport is the cellular target of the oncology drug elesclomol. PLoS One 2012; 7:e29798. O’Day SJ, Eggermont AM, Chiarion-Sileni V, Kefford R, Grob JJ, Mortier L, et al. Final results of phase III SYMMETRY study: randomized, double-blind trial of elesclomol plus paclitaxel versus paclitaxel alone as treatment for chemotherapy-naive patients with advanced melanoma. J Clin Oncol 2013; 31:1211–1218. Raez LE, Papadopoulos K, Ricart AD, Chiorean EG, Dipaola RS, Stein MN, et al. A phase I dose-escalation trial of 2-deoxy-D-glucose alone or combined with docetaxel in patients with advanced solid tumors. Cancer Chemother Pharmacol 2013; 71:523–530. Hall A, Meyle KD, Lange MK, Klima M, Sanderhoff M, Dahl C, et al. Dysfunctional oxidative phosphorylation makes malignant melanoma cells addicted to glycolysis driven by the (V600E)BRAF oncogene. Oncotarget 2013; 4:584–599. Uong A, Zon LI. Melanocytes in development and cancer. J Cell Physiol 2010; 222:38–41. Meyskens FL Jr, Farmer P, Fruehauf JP. Redox regulation in human melanocytes and melanoma. Pigment Cell Res 2001; 14:148–154. Pak BJ, Chu W, Lu SJ, Kerbel RS, Ben-David Y. Lineage-specific mechanism of drug and radiation resistance in melanoma mediated by tyrosinase-related protein 2. Cancer Metastasis Rev 2001; 20:27–32. Saenger YM, Li Y, Chiou KC, Chan B, Rizzuto G, Terzulli SL, et al. Improved tumor immunity using anti-tyrosinase related protein-1 monoclonal antibody combined with DNA vaccines in murine melanoma. Cancer Res 2008; 68:9884–9891. Jakob JA, Bassett RL Jr, Ng CS, Curry JL, Joseph RW, Alvarado GC, et al. NRAS mutation status is an independent prognostic factor in metastatic melanoma. Cancer 2012; 118:4014–4023. Long GV, Menzies AM, Nagrial AM, Haydu LE, Hamilton AL, Mann GJ, et al. Prognostic and clinicopathologic associations of oncogenic BRAF in metastatic melanoma. J Clin Oncol 2011; 29:1239–1246. Harbst K, Staaf J, Masback A, Olsson H, Ingvar C, Vallon-Christersson J, et al. Multiple metastases from cutaneous malignant melanoma patients may display heterogeneous genomic and epigenomic patterns. Melanoma Res 2010; 20:381–391.
Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
