Relationship Between MYCN Copy Number and Expression in Rhabdomyosarcomas and Correlation With Adverse Prognosis in the Alveolar Subtype
http://www.100md.com
《临床肿瘤学》
the Molecular Cytogenetics, Section of Molecular Carcinogenesis and Section of Paediatric Oncology, The Institute of Cancer Research, Sutton, Surrey
Manchester Children's Hospital, Manchester
Department of Histopathology, the Royal Marsden National Health Service Trust
Unit of Molecular Haematology and Oncology, Institute of Child Health, London, United Kingdom
Department of Pathology, Catholic University of Leuven, Leuven, Belgium
Institut fur Pathologie, Universitatsklinikum der Heinrich-Heine-Universitat, Dusseldorf, Germany
ABSTRACT
PATIENTS AND METHODS: Using quantitative polymerase chain reaction, we measured MYCN copy number and expression levels in rhabdomyosarcoma samples from 113 and 92 individuals with a confirmed diagnosis of rhabdomyosarcoma, respectively.
RESULTS: Increased copy number of MYCN was found to be a feature of both the embryonal and alveolar subtypes. The copy number and expression levels were significantly greater in the alveolar subtype, although the range of expression in both subtypes spanned several orders of magnitude. MYCN copy number showed a significant correlation with expression in the alveolar subtype; this relationship between copy number and expression could be modeled as a logarithmic function. It is notable that relatively high expression frequently occurred in embryonal rhabdomyosarcoma without high copy number and that low expression was found in some cases with high copy number. In patients with alveolar rhabdomyosarcoma, overexpression (greater than median) or gain of genomic copies of MYCN were significantly associated with adverse outcome.
CONCLUSION: MYCN deregulation is a feature of rhabdomyosarcoma tumorigenesis, defines groups of patients with a poor prognosis, and is a potential target for novel therapies.
INTRODUCTION
The gene MYCN is a member of the MYC family of highly conserved oncogenes that activate transcription from E-box promoters as part of a heterodimeric complex with MAX. The MYC protein can exert an effect on multiple pathways promoting cell-cycle progression and inhibiting cell differentiation; it is perhaps unsurprising then that MYC genes are both amplified and overexpressed in many cancers. The best characterized example of MYCN amplification is in neuroblastoma, where a core 130-kbp region including MYCN is amplified.3 MYCN amplification of up to 700-fold has been observed in neuroblastoma cell lines with concomitant overexpression and increased protein levels.4 MYCN amplification is a demonstrated adverse prognostic factor in neuroblastoma and is used clinically for treatment stratification of neuroblastomas.5 Amplification of MYCN has also been described in retinoblastoma, small-cell lung cancer, medulloblastoma, and small numbers of RMS.6-8
A study of seven ARMS and six ERMS showed that amplification, as detected by Southern blot analysis, was present in 43% of ARMS but no ERMS and stated there was no significant difference in clinical outcome.9 A more recent study using semi-quantitative reverse transcriptase polymerase chain reaction (PCR) to study 19 primary samples suggested that there was no clear relationship between expression of MYCN and histology or clinical features.10 Analysis of 15 ARMS by fluorescence in situ hybridization (FISH) showed amplification of MYCN in nine (60%) of 15 ARMS and claimed a significant correlation with survival.11 Our previous studies of RMS using comparative genomic hybridization and comparative expressed sequence hybridization indicate that the 2p24 region where MYCN is located is frequently amplified, and even more frequently, genes from this region are overexpressed.12,13
In this study, we aimed to measure the expression and genomic copy number of MYCN in a series of well-characterized RMS and to identify any associations with histology and clinical characteristics. We used the TaqMan (Applied Biosystems, Foster City, CA) quantitative PCR method that has a much greater sensitivity and dynamic range than either Northern or Southern blots and semi-quantitative PCR. TaqMan analysis has previously been validated to detect single copy number changes in MYCN.14 Furthermore, by measuring quantitatively (rather than semi-quantitatively) expression and copy number from the same sample, we aimed to study the relationship between copy number and expression.
PATIENTS AND METHODS
DNA and RNA Extraction
DNA and RNA were extracted from tumor and normal tissues using a phenol-chloroform and Trizol (Invitrogen, Carlsbad, CA) method, respectively. RNA was reverse transcribed using random hexamers and Superscript II reverse transcriptase (Invitrogen). RNA and DNA were also extracted from formalin-fixed paraffin-embedded samples cut from blocks that contained predominantly tumor material. Seven of these formalin-fixed samples were judged to be of a sufficient quality to be informative when used in a previous study that tested the amount of product produced in house-keeper control PCR reactions.17 Furthermore, the paraffin samples used gave TaqMan readings of less than 35 threshold cycles and produced amplification curves parallel to the frozen samples used.
TaqMan Quantification of MYCN
Three sets of primers and probes were designed to measure the amount of genomic and mRNA copies of MYCN in RMS samples (Table 2). All primers and probes were designed in accordance with Applied Biosystems' TaqMan standard requirements. Using GenBank sequence Y00664, primers and a probe were designed within exon 3 of the MYCN gene to quantify genomic copies of MYCN. To correct for aneuploidy, the gene POLR2D was chosen as an endogenous control. POLR2D is a gene that is not believed to be involved in tumorigenesis and is located in a region of chromosome 2 (2q21) rarely altered in RMS other than through whole gain of chromosome 2.13 Using GenBank sequence NM_004805, the primers and a probe were designed within exon 4. To measure the amount of mature spliced mRNA copies of MYCN, a probe was designed across the exon 2–exon 3 boundary. Applied Biosystems' predeveloped glyceraldehyde 3-phosphate dehydrogenase was used as an endogenous control. Triplicate 25-μL multiplex PCR reactions using 2x Universal TaqMan Master Mix (Applied Biosystems; part No. 4304437), the concentration of primers and probes shown in Table 2 and 10 ng of DNA or cDNA were run under standard operating conditions on an ABI7700 SDS TaqMan Machine (Applied Biosystems). Limiting primer conditions were determined, and template titration showed that the reactions were equally efficient, and hence the comparative method was appropriate for both genomic and expression reactions (data not shown). The amount of MYCN was measured relative to either normal genomic DNA in the case of genomic measurements and relative to normal muscle pooled from 11 normal skeletal muscle samples for expression measurements.
FISH
FISH and interphase FISH analysis was performed as described previously.18 Probes used were for MYCN (Vysis Inc, Chicago, IL) and the centromere region of chromosome 2. Analysis was carried out after counterstaining with 4',6-diamidino-2-phenylindole using a Zeiss Axioplan microscope (Zeiss, Oberkochen, Germany) with a Photometrics digital camera (Photometrics, Tucson, AZ) and SmartCapture 2.0 software (Digital Scientific, Cambridge, United Kingdom).
Statistics
All statistics tests were performed using the SPSS 10.0 (SPSS Inc, Chicago, IL) package and tested to the 5% significance level. Failure-free survival was defined as the time from diagnosis to relapse, progression, death, or, if event free, to the date of last contact. Time to death was defined as the time from diagnosis to death or, if event free, to the date of last contact.
RESULTS
Genomic copy number was measured in primary tumor samples taken from 113 individuals with a diagnosis of RMS, of which 48 were ARMS, 58 were ERMS, and seven were RMS not otherwise specified. The distribution of MYCN genomic copies relative to normal genomic DNA is shown in Figure 1. The definition of clinically relevant MYCN amplification in neuroblastoma is the presence of four times as many MYCN signals as chromosome 2 centromere signals by FISH.19 As an approximation to the situation in neuroblastoma, we have defined samples that give a TaqMan result of ≥ four times that of normal DNA as having high-level MYCN copy number changes. Using this definition, 23 (20.4%) of 113 RMS show high-level MYCN copy number changes; by subtype, 12 (25%) of 48 ARMS and nine (16%) of 58 ERMS show high-level MYCN copy number changes. Among these RMS, the genomic copy number of the ARMS is significantly greater than that of the ERMS (Mann-Whitney U = 13.00; n = 21; P = .002). Where data was available for the presence of the PAX/FOXO1A (nine of 12), each ARMS showing high-level MYCN copy number changes also tested positive for the presence of a PAX/FOXO1A fusion (8 x PAX3/FOX01A 1 x PAX7/FOX01A). Anaplasia was present in only two embryonal cases of 44 RMS reviewed for this feature, although neither of these samples had high-level MYCN copy number changes. It is noteworthy that low-level gains of MYCN were more frequent than high level, as defined earlier. Eighty-nine (79%) of 113 RMS samples give a relative measurement greater than 1.5 (the equivalent of three copies in a diploid cell; by subtype, 38 [79%] of 48 ARMS and 45 [78%] of 58 ERMS).
The genomic copy number of MYCN was also investigated by FISH analyses of a cell line and tumor imprints made from several tumor samples (Fig 2). Analysis of chromosomes from the cell line SCMC-RM2 revealed two copies of apparently normal chromosome 2 and two copies of a chromosome derived from chromosome 4 material (determined through 24-color karyotype analysis, data not shown) but carrying multiple copies of MYCN. One of the signals on this chromosome was consistently stronger, indicating the presence of more than one copy at a particular locus (Fig 2B). Interphase analysis of this cell line indicated eight to 12 copies of MYCN in a tetraploid background, which is consistent with the TaqMan PCR data of a three-fold increase of MYCN relative to the long arm of chromosome 2. Interphase FISH analysis of tumor samples of both alveolar and embryonal histologies was similarly supportive (Fig 2). Furthermore, TaqMan analysis performed on 26 samples that had previously been analyzed by comparative genomic hybridization showed concordance with the presence or absence of amplification of 2p24 ≥ a four-fold change13 (Fig 2A and data not shown).
MYCN Expression
Expression of MYCN was measured in primary tumor samples from 92 individuals with a diagnosis of RMS, of which 43 were ARMS, 44 were ERMS, and five were RMS not otherwise specified. Samples showed a huge range in the amount of MYCN mRNA relative to normal skeletal muscle, up to five orders of magnitude in extreme cases (an amount well within the dynamic range of the TaqMan PCR method). The median value was 490 times greater than normal muscle, but the expression of MYCN was significantly different between ARMS (median, 1,209) and ERMS (median, 246; Mann-Whitney U = 647.5; n = 87; P = .013).
Copy Number and Expression Correlation
There is significant correlation between genomic copy number and expression, where DNA and RNA were taken from the same sample within the ARMS but not the ERMS (ARMS, {rho} = 0.500, n = 37, P = .002; ERMS, {rho} = –0.119, n = 36, P = .313). The relationship between MYCN copy number and expression in the ARMS can be best modeled by a logarithmic regression with no constant: 22603.5Ln(Genomic) = expression. This function significantly matches the real data (analysis of variance between regression and residuals, F = 9.13, P = .005). This relationship is illustrated in Figure 3.
Clinical Correlation
There was determined to be a significant difference in the duration of failure-free survival between those ARMS with high (greater than the median value) and low (lower than the median value) MYCN expression (log-rank statistic = 5.82; df = 1; n = 35; P = .0158), but this does not hold true for the ERMS. Furthermore, there is a significant difference in the time to death between high and low MYCN expressing ARMS (log-rank statistic = 4.39; df = 1; n = 34; P = .036; Fig 4).
There is a significant difference in the length of failure-free survival and time to death in ARMS that exhibit MYCN gain (> 1.5 times greater than normal DNA; log-rank statistic = 4.19, df = 1, n = 38, P = .0408; and log-rank statistic = 5.30, df = 1, n = 38, P = .0213, respectively). Again, this does not hold true for ERMS (Fig 4). Neither expression nor genomic copy number was correlated with age, sex, site, or International Society of Pediatric Oncology stage.
Pleomorphic RMS
In addition to the ARMS/ERMS samples, seven samples of DNA and four samples of cDNA from the pleomorphic RMS were measured. One case of amplification (7.2-fold change relative to 2q) was observed, but overall expression was relatively low (median, 4.15 times greater than normal muscle).
DISCUSSION
As seen with the level of copy number change of MYCN, expression is significantly greater in ARMS compared with ERMS. Patients with high-expressing ARMS (greater than the median 490 times greater than normal skeletal muscle) are more likely to experience relapse and less likely to survive than patients with low-expressing ARMS. The clinical and biologic precedent for this adverse correlation has been set in neuroblastoma, albeit that the level at which gain is prognostic in RMS is different in neuroblastoma. The reason for this difference is unclear but may be due to a differing relationship between genomic copy number, rate of transcription, rate of translation, and/or biologic activity of MYCN between the two tumors. Overexpression of MYCN is likely to promote cell growth and repress differentiation in myoblastic progenitor cells of RMS as it does in the neuronal progenitors of neuroblastoma.22 Indeed, one study has already examined the biologic behavior of high MYCN-expressing RMS cell lines. This study demonstrated that high MYCN-expressing RMS cell lines had a more invasive growth pattern in xenografts compared with that of low MYCN expressing RMS cell lines, although growth rate in vitro was similar.23
Expression of MYCN in RMS regardless of histologic subtype is frequently though not always greater than that of normal skeletal muscle, and the range of this expression spans several orders of magnitude. Overexpression is more common than high-level genomic gains, and relatively high expression can occur without similarly high genomic gain, particularly in ERMS. This is consistent with our previous CESH results showing overexpression from 2p24.13 Conversely, there are some ERMS that, despite high-level MYCN copy number gain, show relatively low expression. Consequently, ERMS do not show significant correlation between copy number and expression, indicating that genomic gain is not the primary method of MYCN deregulation in ERMS.
Increased copy number without overexpression is not wholly without precedent. Fan et al24 quote unpublished results of medulloblastomas with MYCN amplification but with relatively low expression in their article on the amplified oncogene hTERT, a gene which also occasionally shows amplification without overexpression. Furthermore, Grotzer et al25 found that only a subset of medulloblastoma showed C-MYC overexpression after amplification. This suggests either that only small increases in MYCN expression are required to produce a tumorigenic effect or that in some ERMS, MYCN is not the critical gene at 2p24.
Previous studies of the relationship between copy number of amplified oncogenes and their concomitant expression levels have used semi-quantitative techniques. One study of particular note showed a linear correlation between copy number of hTERT in embryonal CNS tumors and expression and used quantitative PCR to measure expression but a semi-quantitative technique to measure copy number.24 The fact that our study uses the quantitative TaqMan method to measure both copy number and expression from the same sample, rather than semi-quantitative methods like reverse transcriptase PCR, Southern blot analysis, and Northern blot analysis, is unique among studies of amplified oncogenes. Examining those cases of ARMS where it was possible to measure both expression and copy number reveals a logarithmic relationship of the form {alpha}Ln(Genomic copies) = expression (Fig 3). This phenomenon has not been previously described but suggests that expression of an oncogene is ultimately limited by factors other than the amount of genomic template (ie, availability of transcription factors). The implication of this relationship is a law of diminishing expression returns for increasing genomic gains. Potentially, these effects could act to limit the copy number of genes by ultimately limiting the selective clonal advantage produced by additional copy number gain. It is noteworthy that in culture, even after many passages, amplicons do not gain copies ad infinitum. Furthermore, neuroblastoma tend to be associated with higher copy numbers of MYCN than the RMS studied here. This is also true of neuroblastoma and RMS cell lines.4 The fact that the ERMS do not demonstrate the same relationship suggests that this may apply only under certain conditions in a particular cellular environment.
In conclusion, we have demonstrated that gain of genomic copies and overexpression of MYCN is a common phenomenon in both ARMS and ERMS, although the copy number gain and expression tend to be greater in ARMS. Furthermore, gain of genomic copies and overexpression of MYCN are associated with an adverse prognosis in patients with ARMS. An immunostain for the MYCN protein may also be of clinical relevance. Although the role of this gene in the different RMS subtypes requires further investigation, it is has been suggested that MYCN offers a target for novel therapies.26 This study suggests that such therapies may be of relevance in RMS.
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank Richard Carter, Andy Pearson, Annabel Foot, Richard Grundy, Ananth Shankar, Sue Picton, Jan Kohler, Syuiti Abe, Takayuki Nojima, Penny Flohr, and the United Kingdom Children's Cancer Study Group. We also thank Brenda Summersgill, Beata Grygalewicz, and Dyanne Rampling for their excellent technical assistance.
NOTES
Supported by Cancer Research United Kingdom.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Crist WM, Anderson JR, Meza JL, et al: Intergroup rhabdomyosarcoma study-IV: Results for patients with nonmetastatic disease. J Clin Oncol 19:3091-3102, 2001
2. Anderson J, Gordon A, Pritchard-Jones K, et al: Genes, chromosomes, and rhabdomyosarcoma. Genes Chromosomes Cancer 26:275-285, 1999
3. Reiter JL, Brodeur GM: MYCN is the only highly expressed gene from the core amplified domain in human neuroblastomas. Genes Chromosomes Cancer 23:134-140, 1998
4. Kohl NE, Kanda N, Schreck RR, et al: Transposition and amplification of oncogene-related sequences in human neuroblastomas. Cell 35:359-367, 1983
5. Brodeur G, Castleberry RP: Neuroblastoma, in Pizzo PA, Poplack DG (eds): Principles and Practice of Pediatric Oncology (ed 3). Philadelphia, PA, Lippincott-Raven, 1997, pp 761-798
6. Lee WH, Murphree AL, Benedict WF: Expression and amplification of the N-myc gene in primary retinoblastoma. Nature 309:458-460, 1984
7. Nau MM, Brooks BJJ, Carney D N, et al: Human small-cell lung cancers show amplification and expression of the N-myc gene. Proc Natl Acad Sci U S A 83:1092-1096, 1986
8. Wasson JC, Saylors RL, Zeltzer P, et al: Oncogene amplification in pediatric brain tumors. Cancer Res 50:2987-2990, 1990
9. Driman D, Thorner PS, Greenberg ML, et al: MYCN gene amplification in rhabdomyosarcoma. Cancer 73:2231-2237, 1994
10. Toffolatti L, Frascella E, Ninfo V, et al: MYCN expression in human rhabdomyosarcoma cell lines and tumour samples. J Pathol 196:450-458, 2002
11. Hachitanda Y, Toyoshima S, Akazawa K, et al: N-myc gene amplification in rhabdomyosarcoma detected by fluorescence in situ hybridization: Its correlation with histologic features. Mod Pathol 11:1222-1227, 1998
12. Gordon AT, Brinkschmidt C, Anderson J, et al: A novel and consistent amplicon at 13q31 associated with alveolar rhabdomyosarcoma. Genes Chromosomes Cancer 28:220-226, 2000
13. Lu JY, Williamson D, Clark J, et al: Comparative expressed sequence hybridization to chromosomes for tumor classification and identification of genomic regions of differential gene expression. Proc Natl Acad Sci U S A 98:9197-9202, 2001
14. De Preter K, Speleman F, Combaret V, et al: Quantification of MYCN, DDX1, and NAG gene copy number in neuroblastoma using a real-time quantitative PCR assay. Mod Pathol 15:159-166, 2002
15. Garvin AJ, Stanley WS, Bennett DD, et al: The in vitro growth, heterotransplantation, and differentiation of a human rhabdomyosarcoma cell line. Am J Pathol 125:208-217, 1986
16. Hayashi Y, Sugimoto T, Horii Y, et al: Characterization of an embryonal rhabdomyosarcoma cell line showing amplification and over-expression of the N-myc oncogene. Int J Cancer 45:705-711, 1990
17. Anderson J, Gordon T, McManus A, et al: Detection of the PAX3-FKHR fusion gene in paediatric rhabdomyosarcoma: A reproducible predictor of outcome Br J Cancer 85:831-835, 2001
18. McManus AP, O'Reilly MA, Jones KP, et al: Interphase fluorescence in situ hybridiza-tion detection of t(2;13)(q35;q14) in alveolar rhabdomyosarcoma: A diagnostic tool in minimally invasive biopsies. J Pathol 178:410-414, 1996
19. Ambros IM, Benard J, Boavida M, et al: Quality assessment of genetic markers used for therapy stratification. J Clin Oncol 21:2077-2084, 2003
20. Corvi R, Savelyeva L, Schwab M: Duplication of N-MYC at its resident site 2p24 may be a mechanism of activation alternative to amplification in human neuroblastoma cells. Cancer Res 55:3471-3474, 1995
21. Valent A, Le Roux G, Barrois M, et al: MYCN gene overrepresentation detected in primary neuroblastoma tumour cells without amplification. J Pathol 198:495-501, 2002
22. Knoepfler PS, Cheng PF, Eisenman RN: N-myc is essential during neurogenesis for the rapid expansion of progenitor cell populations and the inhibition of neuronal differentiation. Genes Dev 16:2699-2712, 2002
23. Kouraklis G, Triche TJ, Wesley R, et al: Myc oncogene expression and nude mouse tumorigenicity and metastasis formation are higher in alveolar than embryonal rhabdomyosarcoma cell lines. Pediatr Res 45:552-558, 1999
24. Fan X, Wang Y, Kratz J, et al: HTERT gene amplification and increased mRNA expression in central nervous system embryonal tumors. Am J Pathol 162:1763-1769, 2003
25. Grotzer MA, Hogarty MD, Janss A, et al: MYC messenger RNA expression predicts survival outcome in childhood primitive neuroectodermal tumor/medulloblastoma. Clin Cancer Res 7:2425-2433, 2001
26. Lu X, Pearson A, Lunec J: The MYCN oncoprotein as a drug development target. Cancer Lett 197:125-130, 2003(Daniel Williamson, Yong-J)
Manchester Children's Hospital, Manchester
Department of Histopathology, the Royal Marsden National Health Service Trust
Unit of Molecular Haematology and Oncology, Institute of Child Health, London, United Kingdom
Department of Pathology, Catholic University of Leuven, Leuven, Belgium
Institut fur Pathologie, Universitatsklinikum der Heinrich-Heine-Universitat, Dusseldorf, Germany
ABSTRACT
PATIENTS AND METHODS: Using quantitative polymerase chain reaction, we measured MYCN copy number and expression levels in rhabdomyosarcoma samples from 113 and 92 individuals with a confirmed diagnosis of rhabdomyosarcoma, respectively.
RESULTS: Increased copy number of MYCN was found to be a feature of both the embryonal and alveolar subtypes. The copy number and expression levels were significantly greater in the alveolar subtype, although the range of expression in both subtypes spanned several orders of magnitude. MYCN copy number showed a significant correlation with expression in the alveolar subtype; this relationship between copy number and expression could be modeled as a logarithmic function. It is notable that relatively high expression frequently occurred in embryonal rhabdomyosarcoma without high copy number and that low expression was found in some cases with high copy number. In patients with alveolar rhabdomyosarcoma, overexpression (greater than median) or gain of genomic copies of MYCN were significantly associated with adverse outcome.
CONCLUSION: MYCN deregulation is a feature of rhabdomyosarcoma tumorigenesis, defines groups of patients with a poor prognosis, and is a potential target for novel therapies.
INTRODUCTION
The gene MYCN is a member of the MYC family of highly conserved oncogenes that activate transcription from E-box promoters as part of a heterodimeric complex with MAX. The MYC protein can exert an effect on multiple pathways promoting cell-cycle progression and inhibiting cell differentiation; it is perhaps unsurprising then that MYC genes are both amplified and overexpressed in many cancers. The best characterized example of MYCN amplification is in neuroblastoma, where a core 130-kbp region including MYCN is amplified.3 MYCN amplification of up to 700-fold has been observed in neuroblastoma cell lines with concomitant overexpression and increased protein levels.4 MYCN amplification is a demonstrated adverse prognostic factor in neuroblastoma and is used clinically for treatment stratification of neuroblastomas.5 Amplification of MYCN has also been described in retinoblastoma, small-cell lung cancer, medulloblastoma, and small numbers of RMS.6-8
A study of seven ARMS and six ERMS showed that amplification, as detected by Southern blot analysis, was present in 43% of ARMS but no ERMS and stated there was no significant difference in clinical outcome.9 A more recent study using semi-quantitative reverse transcriptase polymerase chain reaction (PCR) to study 19 primary samples suggested that there was no clear relationship between expression of MYCN and histology or clinical features.10 Analysis of 15 ARMS by fluorescence in situ hybridization (FISH) showed amplification of MYCN in nine (60%) of 15 ARMS and claimed a significant correlation with survival.11 Our previous studies of RMS using comparative genomic hybridization and comparative expressed sequence hybridization indicate that the 2p24 region where MYCN is located is frequently amplified, and even more frequently, genes from this region are overexpressed.12,13
In this study, we aimed to measure the expression and genomic copy number of MYCN in a series of well-characterized RMS and to identify any associations with histology and clinical characteristics. We used the TaqMan (Applied Biosystems, Foster City, CA) quantitative PCR method that has a much greater sensitivity and dynamic range than either Northern or Southern blots and semi-quantitative PCR. TaqMan analysis has previously been validated to detect single copy number changes in MYCN.14 Furthermore, by measuring quantitatively (rather than semi-quantitatively) expression and copy number from the same sample, we aimed to study the relationship between copy number and expression.
PATIENTS AND METHODS
DNA and RNA Extraction
DNA and RNA were extracted from tumor and normal tissues using a phenol-chloroform and Trizol (Invitrogen, Carlsbad, CA) method, respectively. RNA was reverse transcribed using random hexamers and Superscript II reverse transcriptase (Invitrogen). RNA and DNA were also extracted from formalin-fixed paraffin-embedded samples cut from blocks that contained predominantly tumor material. Seven of these formalin-fixed samples were judged to be of a sufficient quality to be informative when used in a previous study that tested the amount of product produced in house-keeper control PCR reactions.17 Furthermore, the paraffin samples used gave TaqMan readings of less than 35 threshold cycles and produced amplification curves parallel to the frozen samples used.
TaqMan Quantification of MYCN
Three sets of primers and probes were designed to measure the amount of genomic and mRNA copies of MYCN in RMS samples (Table 2). All primers and probes were designed in accordance with Applied Biosystems' TaqMan standard requirements. Using GenBank sequence Y00664, primers and a probe were designed within exon 3 of the MYCN gene to quantify genomic copies of MYCN. To correct for aneuploidy, the gene POLR2D was chosen as an endogenous control. POLR2D is a gene that is not believed to be involved in tumorigenesis and is located in a region of chromosome 2 (2q21) rarely altered in RMS other than through whole gain of chromosome 2.13 Using GenBank sequence NM_004805, the primers and a probe were designed within exon 4. To measure the amount of mature spliced mRNA copies of MYCN, a probe was designed across the exon 2–exon 3 boundary. Applied Biosystems' predeveloped glyceraldehyde 3-phosphate dehydrogenase was used as an endogenous control. Triplicate 25-μL multiplex PCR reactions using 2x Universal TaqMan Master Mix (Applied Biosystems; part No. 4304437), the concentration of primers and probes shown in Table 2 and 10 ng of DNA or cDNA were run under standard operating conditions on an ABI7700 SDS TaqMan Machine (Applied Biosystems). Limiting primer conditions were determined, and template titration showed that the reactions were equally efficient, and hence the comparative method was appropriate for both genomic and expression reactions (data not shown). The amount of MYCN was measured relative to either normal genomic DNA in the case of genomic measurements and relative to normal muscle pooled from 11 normal skeletal muscle samples for expression measurements.
FISH
FISH and interphase FISH analysis was performed as described previously.18 Probes used were for MYCN (Vysis Inc, Chicago, IL) and the centromere region of chromosome 2. Analysis was carried out after counterstaining with 4',6-diamidino-2-phenylindole using a Zeiss Axioplan microscope (Zeiss, Oberkochen, Germany) with a Photometrics digital camera (Photometrics, Tucson, AZ) and SmartCapture 2.0 software (Digital Scientific, Cambridge, United Kingdom).
Statistics
All statistics tests were performed using the SPSS 10.0 (SPSS Inc, Chicago, IL) package and tested to the 5% significance level. Failure-free survival was defined as the time from diagnosis to relapse, progression, death, or, if event free, to the date of last contact. Time to death was defined as the time from diagnosis to death or, if event free, to the date of last contact.
RESULTS
Genomic copy number was measured in primary tumor samples taken from 113 individuals with a diagnosis of RMS, of which 48 were ARMS, 58 were ERMS, and seven were RMS not otherwise specified. The distribution of MYCN genomic copies relative to normal genomic DNA is shown in Figure 1. The definition of clinically relevant MYCN amplification in neuroblastoma is the presence of four times as many MYCN signals as chromosome 2 centromere signals by FISH.19 As an approximation to the situation in neuroblastoma, we have defined samples that give a TaqMan result of ≥ four times that of normal DNA as having high-level MYCN copy number changes. Using this definition, 23 (20.4%) of 113 RMS show high-level MYCN copy number changes; by subtype, 12 (25%) of 48 ARMS and nine (16%) of 58 ERMS show high-level MYCN copy number changes. Among these RMS, the genomic copy number of the ARMS is significantly greater than that of the ERMS (Mann-Whitney U = 13.00; n = 21; P = .002). Where data was available for the presence of the PAX/FOXO1A (nine of 12), each ARMS showing high-level MYCN copy number changes also tested positive for the presence of a PAX/FOXO1A fusion (8 x PAX3/FOX01A 1 x PAX7/FOX01A). Anaplasia was present in only two embryonal cases of 44 RMS reviewed for this feature, although neither of these samples had high-level MYCN copy number changes. It is noteworthy that low-level gains of MYCN were more frequent than high level, as defined earlier. Eighty-nine (79%) of 113 RMS samples give a relative measurement greater than 1.5 (the equivalent of three copies in a diploid cell; by subtype, 38 [79%] of 48 ARMS and 45 [78%] of 58 ERMS).
The genomic copy number of MYCN was also investigated by FISH analyses of a cell line and tumor imprints made from several tumor samples (Fig 2). Analysis of chromosomes from the cell line SCMC-RM2 revealed two copies of apparently normal chromosome 2 and two copies of a chromosome derived from chromosome 4 material (determined through 24-color karyotype analysis, data not shown) but carrying multiple copies of MYCN. One of the signals on this chromosome was consistently stronger, indicating the presence of more than one copy at a particular locus (Fig 2B). Interphase analysis of this cell line indicated eight to 12 copies of MYCN in a tetraploid background, which is consistent with the TaqMan PCR data of a three-fold increase of MYCN relative to the long arm of chromosome 2. Interphase FISH analysis of tumor samples of both alveolar and embryonal histologies was similarly supportive (Fig 2). Furthermore, TaqMan analysis performed on 26 samples that had previously been analyzed by comparative genomic hybridization showed concordance with the presence or absence of amplification of 2p24 ≥ a four-fold change13 (Fig 2A and data not shown).
MYCN Expression
Expression of MYCN was measured in primary tumor samples from 92 individuals with a diagnosis of RMS, of which 43 were ARMS, 44 were ERMS, and five were RMS not otherwise specified. Samples showed a huge range in the amount of MYCN mRNA relative to normal skeletal muscle, up to five orders of magnitude in extreme cases (an amount well within the dynamic range of the TaqMan PCR method). The median value was 490 times greater than normal muscle, but the expression of MYCN was significantly different between ARMS (median, 1,209) and ERMS (median, 246; Mann-Whitney U = 647.5; n = 87; P = .013).
Copy Number and Expression Correlation
There is significant correlation between genomic copy number and expression, where DNA and RNA were taken from the same sample within the ARMS but not the ERMS (ARMS, {rho} = 0.500, n = 37, P = .002; ERMS, {rho} = –0.119, n = 36, P = .313). The relationship between MYCN copy number and expression in the ARMS can be best modeled by a logarithmic regression with no constant: 22603.5Ln(Genomic) = expression. This function significantly matches the real data (analysis of variance between regression and residuals, F = 9.13, P = .005). This relationship is illustrated in Figure 3.
Clinical Correlation
There was determined to be a significant difference in the duration of failure-free survival between those ARMS with high (greater than the median value) and low (lower than the median value) MYCN expression (log-rank statistic = 5.82; df = 1; n = 35; P = .0158), but this does not hold true for the ERMS. Furthermore, there is a significant difference in the time to death between high and low MYCN expressing ARMS (log-rank statistic = 4.39; df = 1; n = 34; P = .036; Fig 4).
There is a significant difference in the length of failure-free survival and time to death in ARMS that exhibit MYCN gain (> 1.5 times greater than normal DNA; log-rank statistic = 4.19, df = 1, n = 38, P = .0408; and log-rank statistic = 5.30, df = 1, n = 38, P = .0213, respectively). Again, this does not hold true for ERMS (Fig 4). Neither expression nor genomic copy number was correlated with age, sex, site, or International Society of Pediatric Oncology stage.
Pleomorphic RMS
In addition to the ARMS/ERMS samples, seven samples of DNA and four samples of cDNA from the pleomorphic RMS were measured. One case of amplification (7.2-fold change relative to 2q) was observed, but overall expression was relatively low (median, 4.15 times greater than normal muscle).
DISCUSSION
As seen with the level of copy number change of MYCN, expression is significantly greater in ARMS compared with ERMS. Patients with high-expressing ARMS (greater than the median 490 times greater than normal skeletal muscle) are more likely to experience relapse and less likely to survive than patients with low-expressing ARMS. The clinical and biologic precedent for this adverse correlation has been set in neuroblastoma, albeit that the level at which gain is prognostic in RMS is different in neuroblastoma. The reason for this difference is unclear but may be due to a differing relationship between genomic copy number, rate of transcription, rate of translation, and/or biologic activity of MYCN between the two tumors. Overexpression of MYCN is likely to promote cell growth and repress differentiation in myoblastic progenitor cells of RMS as it does in the neuronal progenitors of neuroblastoma.22 Indeed, one study has already examined the biologic behavior of high MYCN-expressing RMS cell lines. This study demonstrated that high MYCN-expressing RMS cell lines had a more invasive growth pattern in xenografts compared with that of low MYCN expressing RMS cell lines, although growth rate in vitro was similar.23
Expression of MYCN in RMS regardless of histologic subtype is frequently though not always greater than that of normal skeletal muscle, and the range of this expression spans several orders of magnitude. Overexpression is more common than high-level genomic gains, and relatively high expression can occur without similarly high genomic gain, particularly in ERMS. This is consistent with our previous CESH results showing overexpression from 2p24.13 Conversely, there are some ERMS that, despite high-level MYCN copy number gain, show relatively low expression. Consequently, ERMS do not show significant correlation between copy number and expression, indicating that genomic gain is not the primary method of MYCN deregulation in ERMS.
Increased copy number without overexpression is not wholly without precedent. Fan et al24 quote unpublished results of medulloblastomas with MYCN amplification but with relatively low expression in their article on the amplified oncogene hTERT, a gene which also occasionally shows amplification without overexpression. Furthermore, Grotzer et al25 found that only a subset of medulloblastoma showed C-MYC overexpression after amplification. This suggests either that only small increases in MYCN expression are required to produce a tumorigenic effect or that in some ERMS, MYCN is not the critical gene at 2p24.
Previous studies of the relationship between copy number of amplified oncogenes and their concomitant expression levels have used semi-quantitative techniques. One study of particular note showed a linear correlation between copy number of hTERT in embryonal CNS tumors and expression and used quantitative PCR to measure expression but a semi-quantitative technique to measure copy number.24 The fact that our study uses the quantitative TaqMan method to measure both copy number and expression from the same sample, rather than semi-quantitative methods like reverse transcriptase PCR, Southern blot analysis, and Northern blot analysis, is unique among studies of amplified oncogenes. Examining those cases of ARMS where it was possible to measure both expression and copy number reveals a logarithmic relationship of the form {alpha}Ln(Genomic copies) = expression (Fig 3). This phenomenon has not been previously described but suggests that expression of an oncogene is ultimately limited by factors other than the amount of genomic template (ie, availability of transcription factors). The implication of this relationship is a law of diminishing expression returns for increasing genomic gains. Potentially, these effects could act to limit the copy number of genes by ultimately limiting the selective clonal advantage produced by additional copy number gain. It is noteworthy that in culture, even after many passages, amplicons do not gain copies ad infinitum. Furthermore, neuroblastoma tend to be associated with higher copy numbers of MYCN than the RMS studied here. This is also true of neuroblastoma and RMS cell lines.4 The fact that the ERMS do not demonstrate the same relationship suggests that this may apply only under certain conditions in a particular cellular environment.
In conclusion, we have demonstrated that gain of genomic copies and overexpression of MYCN is a common phenomenon in both ARMS and ERMS, although the copy number gain and expression tend to be greater in ARMS. Furthermore, gain of genomic copies and overexpression of MYCN are associated with an adverse prognosis in patients with ARMS. An immunostain for the MYCN protein may also be of clinical relevance. Although the role of this gene in the different RMS subtypes requires further investigation, it is has been suggested that MYCN offers a target for novel therapies.26 This study suggests that such therapies may be of relevance in RMS.
Authors' Disclosures of Potential Conflicts of Interest
Acknowledgment
We thank Richard Carter, Andy Pearson, Annabel Foot, Richard Grundy, Ananth Shankar, Sue Picton, Jan Kohler, Syuiti Abe, Takayuki Nojima, Penny Flohr, and the United Kingdom Children's Cancer Study Group. We also thank Brenda Summersgill, Beata Grygalewicz, and Dyanne Rampling for their excellent technical assistance.
NOTES
Supported by Cancer Research United Kingdom.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Crist WM, Anderson JR, Meza JL, et al: Intergroup rhabdomyosarcoma study-IV: Results for patients with nonmetastatic disease. J Clin Oncol 19:3091-3102, 2001
2. Anderson J, Gordon A, Pritchard-Jones K, et al: Genes, chromosomes, and rhabdomyosarcoma. Genes Chromosomes Cancer 26:275-285, 1999
3. Reiter JL, Brodeur GM: MYCN is the only highly expressed gene from the core amplified domain in human neuroblastomas. Genes Chromosomes Cancer 23:134-140, 1998
4. Kohl NE, Kanda N, Schreck RR, et al: Transposition and amplification of oncogene-related sequences in human neuroblastomas. Cell 35:359-367, 1983
5. Brodeur G, Castleberry RP: Neuroblastoma, in Pizzo PA, Poplack DG (eds): Principles and Practice of Pediatric Oncology (ed 3). Philadelphia, PA, Lippincott-Raven, 1997, pp 761-798
6. Lee WH, Murphree AL, Benedict WF: Expression and amplification of the N-myc gene in primary retinoblastoma. Nature 309:458-460, 1984
7. Nau MM, Brooks BJJ, Carney D N, et al: Human small-cell lung cancers show amplification and expression of the N-myc gene. Proc Natl Acad Sci U S A 83:1092-1096, 1986
8. Wasson JC, Saylors RL, Zeltzer P, et al: Oncogene amplification in pediatric brain tumors. Cancer Res 50:2987-2990, 1990
9. Driman D, Thorner PS, Greenberg ML, et al: MYCN gene amplification in rhabdomyosarcoma. Cancer 73:2231-2237, 1994
10. Toffolatti L, Frascella E, Ninfo V, et al: MYCN expression in human rhabdomyosarcoma cell lines and tumour samples. J Pathol 196:450-458, 2002
11. Hachitanda Y, Toyoshima S, Akazawa K, et al: N-myc gene amplification in rhabdomyosarcoma detected by fluorescence in situ hybridization: Its correlation with histologic features. Mod Pathol 11:1222-1227, 1998
12. Gordon AT, Brinkschmidt C, Anderson J, et al: A novel and consistent amplicon at 13q31 associated with alveolar rhabdomyosarcoma. Genes Chromosomes Cancer 28:220-226, 2000
13. Lu JY, Williamson D, Clark J, et al: Comparative expressed sequence hybridization to chromosomes for tumor classification and identification of genomic regions of differential gene expression. Proc Natl Acad Sci U S A 98:9197-9202, 2001
14. De Preter K, Speleman F, Combaret V, et al: Quantification of MYCN, DDX1, and NAG gene copy number in neuroblastoma using a real-time quantitative PCR assay. Mod Pathol 15:159-166, 2002
15. Garvin AJ, Stanley WS, Bennett DD, et al: The in vitro growth, heterotransplantation, and differentiation of a human rhabdomyosarcoma cell line. Am J Pathol 125:208-217, 1986
16. Hayashi Y, Sugimoto T, Horii Y, et al: Characterization of an embryonal rhabdomyosarcoma cell line showing amplification and over-expression of the N-myc oncogene. Int J Cancer 45:705-711, 1990
17. Anderson J, Gordon T, McManus A, et al: Detection of the PAX3-FKHR fusion gene in paediatric rhabdomyosarcoma: A reproducible predictor of outcome Br J Cancer 85:831-835, 2001
18. McManus AP, O'Reilly MA, Jones KP, et al: Interphase fluorescence in situ hybridiza-tion detection of t(2;13)(q35;q14) in alveolar rhabdomyosarcoma: A diagnostic tool in minimally invasive biopsies. J Pathol 178:410-414, 1996
19. Ambros IM, Benard J, Boavida M, et al: Quality assessment of genetic markers used for therapy stratification. J Clin Oncol 21:2077-2084, 2003
20. Corvi R, Savelyeva L, Schwab M: Duplication of N-MYC at its resident site 2p24 may be a mechanism of activation alternative to amplification in human neuroblastoma cells. Cancer Res 55:3471-3474, 1995
21. Valent A, Le Roux G, Barrois M, et al: MYCN gene overrepresentation detected in primary neuroblastoma tumour cells without amplification. J Pathol 198:495-501, 2002
22. Knoepfler PS, Cheng PF, Eisenman RN: N-myc is essential during neurogenesis for the rapid expansion of progenitor cell populations and the inhibition of neuronal differentiation. Genes Dev 16:2699-2712, 2002
23. Kouraklis G, Triche TJ, Wesley R, et al: Myc oncogene expression and nude mouse tumorigenicity and metastasis formation are higher in alveolar than embryonal rhabdomyosarcoma cell lines. Pediatr Res 45:552-558, 1999
24. Fan X, Wang Y, Kratz J, et al: HTERT gene amplification and increased mRNA expression in central nervous system embryonal tumors. Am J Pathol 162:1763-1769, 2003
25. Grotzer MA, Hogarty MD, Janss A, et al: MYC messenger RNA expression predicts survival outcome in childhood primitive neuroectodermal tumor/medulloblastoma. Clin Cancer Res 7:2425-2433, 2001
26. Lu X, Pearson A, Lunec J: The MYCN oncoprotein as a drug development target. Cancer Lett 197:125-130, 2003(Daniel Williamson, Yong-J)