|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 23, No 1 (January 1), 2005: pp. 58-69 © 2005 American Society of Clinical Oncology. DOI: 10.1200/JCO.2005.11.076 Gene Expression Profiling in Seminoma and NonseminomaFrom the Institute of Radiobiology of the Federal Armed Forces, Munich; Departments of Urology and Internal Medicine, Federal Armed Forces Hospital, Ulm; Department of Urology, Klinikum Kassel GmbH, Kassel; Department of Urology, Federal Armed Forces Hospital, Hamburg, Germany Address reprint requests to Hans Ulrich Schmelz, MD, Institute of Radiobiology, Federal Armed Forces, Ernst-von-Bergmann-Kaserne, Neuherbergstr 11, 80937 Munich, Germany; e-mail: michaelabend{at}bundeswehr.org
PURPOSE: Gene expression profiles of seminoma were compared with nonseminoma to get insights into tumorigenesis. MATERIALS AND METHODS: Eleven testicular tumor biopsies (five pure seminoma, six nonseminoma; pT1N0M0 to pT2N2M1) and biopsies from unaffected sites were analyzed once per patient using a macroarray (1,176 genes). On the same patients, six genes were validated using real-time quantitative (RTQ) polymerase chain reaction (PCR). Additionally, in a separate cohort of 19 patients, 24 genes selected from the macroarray were measured using RTQ-PCR. RESULTS: (1) The agreement in gene expression was 94% between the two methods and two different patient cohorts. (2) Two features in gene expression were independent of the tumor entity: Most changes of gene expression occurred in five functional groups like "cell cycle" and "apoptosis." Genes within these groups were almost similarly (> 80%) up- or downregulated. (3) Nonseminoma were characterized by downregulated genes (75%), but in seminoma, upregulated genes (64%) prevailed. Furthermore, 64.4% of those genes that were differentially expressed in both tumor entities were usually upregulated in seminoma but downregulated in nonseminoma. A reverse pattern was found in 24.4% of such genes. Eleven percent of these genes showed a similar up- or downregulation in gene expression in both tumor entities. CONCLUSION: Seminoma in this preliminary study can be differentiated from nonseminoma due to almost opposing gene expression profiles (89% of the significantly differentially expressed genes) and are in line with the histological discrimination of both tumor entities. Underlying mechanisms and implications regarding the origin and tumor progression of both entities are discussed.
A few years ago, it was merely possible to examine a limited number of genes or proteins simultaneously. Examinations of cell proliferation, cell cycle control, and apoptosis represented a focus in the field of tumorigenesis.1–6 This is also the focus of several publications concerning testis tumors of different origins.7–16 With the advent of array technology, it became possible to investigate changes in gene expression on a large scale. Gene expression analysis has been performed on many tissues, and especially on tumors.17–21 However, only a limited number of publications on testis tumors exist.22 To examine changes in gene expression not only of cell cycle control and apoptosis, but also of further functionalities, a macroarray was chosen, which enabled the detection of gene expression of these and other genes (1,176 genes) simultaneously. Testis germ cell tumors are classified in two main subtypes, seminoma and nonseminoma23 due to histologic and clinical prognosis and therapeutic criteria. Tissue samples from seminoma and nonseminoma were resected and, in addition, biopsies were taken far from the tumor region representing predominantly healthy tissue. These served as an individual control for each tumor. Hence, in a limited number of tissue samples, differential gene expression and its relationship to tumorigenesis was examined, as well as gene expression profiles of seminoma compared with nonseminoma.
Tissue Samples and Histological Examination Testicular tumor biopsies from 11 patients (average age, 33 years; range, 18 to 60 years), and corresponding biopsies of unaffected sites of the resected testis were analyzed. Five pure seminoma and six nonseminoma were found. TNM system classification24 ranged from pT1N0M0 to pT2N2M1. Nonseminoma consisted of mixed nonseminomatous germ cell tumors. On this cohort of patients, both macroarray analysis and real-time quantitative (RTQ) polymerase chain reaction (PCR) were performed. A separate cohort of 19 patients (average age, 35 years; range, 21 to 42 years) containing nine pure seminoma and 10 nonseminoma (three pure embryonal carcinomas and seven mixed nonseminomatous germ cell tumors) with pT1N0M0 to pT3N2M1 were examined utilizing RTQ-PCR only. Tissues were fixed in RNA Later solution (Qiagen, Hilden, Germany) immediately after surgery. All human samples were obtained with informed consent.
RNA Isolation Only total RNA with a ratio of A260/A280 greater than 1.8 (spectralphotometry), with 28S ribosomal bands that were present at approximately twice the amounts of the 18S RNA (agarose gel electrophoresis) and without detectable contamination of DNA (conventional PCR run over 40 cycles using ß-Actin primers for detection of DNA contamination; data not shown) were used for gene expression analysis.
Gene Expression Array
Hybridization signals were analyzed using Atlas Image 2.0 Software (Clontech) and others. Normalization was performed using total gene expression, and reproducibility was taken into account according to a method shown previously.25 Due to this procedure, background levels of gene expression arrays and the threshold of genes considered to represent differentially expressed genes were adjusted so that genes falling within that category were differentially expressed with a likelihood of  90%.
RTQ-PCR
After conversion of threshold cycles into ng cDNA with the use of the standard curve, a normalization step using 18SrRNA followed. Since 18SrRNA copy numbers are expressed in the same order of magnitude in every cell, this RNA species was measured in parallel to the gene of interest in tumor tissues as well as the associated normal tissue samples, thus allowing correction for the same amount of ng cDNA added to each PCR reaction. Then, the abundance of the gene of interest in the tumor tissue was calculated relative to the abundance of the same gene in the normal tissue. This ratio was called differential gene expression. A ratio of 1 corresponds to a gene expression that is similar in the tumor and the associated healthy tissue. A ratio greater 1 or lower than 1 refers to a several-fold over- or underexpression of the gene of interest in the tumor sample, relative to the corresponding normal tissue.
Statistics
RNA Extraction and General Findings Regarding the Array Data From 1 mg of tissue, 0.8 µg (± 0.76 µg) high-quality DNA-free total RNA was extracted. To run one array experiment, 50 to 100 mg of tissue was needed. Total RNA (for array analysis) was isolated from 20 patients. However, the criteria for quality and quantity of total RNA were met for only 11 patients. Gene expression array analysis (1,176 genes) of the tumor and the adjacent healthy tissue (Fig 1) were performed on this limited number. Gene expression over background was found for 35% (± 13%) of these 1,176 genes. When comparing genes (which in tumor samples appeared to be differentially expressed) with their appropriate healthy tissue, these differences were usually found not only in one but in all tumor biopsies examined (Fig 3).
Confirmation of Array Data Using RTQ-PCR According to array analysis, six differentially expressed genes were selected for further evaluation via RTQ-PCR using the same cohort of 11 patients. The genes were (1) clusterin precursor (CLU); (2) prothymosin alpha (PTMA); (3) ubiquitin (UBC); (4) GAPDH; (5) cytochrome P450, family 11 (CYP11A); and (6) PENK. The gene expressions of these genes drawn from the analysis of macroarrays (11 patients) were compared with the corresponding data using RTQ-PCR, and performed on the same patients. Differential gene expression (tumor v normal tissue) from five of six genes was comparable using a semiquantitative and a quantitative technique. In detail, the comparison of differential gene expression calculated by array data (first number in parentheses) and compared with RTQ-PCR measurements (second number in parentheses) revealed a downregulation for CLU (16-fold v 21-fold), UBC (1.9-fold v 1.9-fold), and CYP11A (10-fold v 28-fold). An upregulation in the tumor tissue was found for PTMA (2.0-fold v 2.6-fold) and GAPDH (2.1-fold v 3.4-fold). We found no significant differences between the results determined with the two different methods. However, for PENK, a 4.5-fold downregulation was found when using array analysis. This was in contrast to a 55-fold downregulation when utilizing RTQ-PCR and was statistically significant (P < .002, t test). Furthermore, differential gene expression of six genes measured by the macroarray approach in 11 patients (n = 66 genes) was compared with the patients' corresponding RTQ-PCR data (Fig 4A). The regression analysis implied a correlation with f(x) being close to x, with f(x) = 1.2x – 0.1 and r2 = 0.71 when considering the whole data set (solid line). Without PENK, the correlation was f(x) = 1.12x + 0.03 and r2 = 0.84 (dotted line). The mean ± SEM difference in gene expression between the semiquantiative macroarry and the quantitative RTQ-PCR was 1.6-fold ± 0.3. Moreover, the majority (62 of 66 genes; 94%) of up- and downregulated genes appeared similarly regulated using both methods. The disagreement between both methods was 6%.
Twenty-four selected genes that were differentially expressed in a cohort of 11 patients using the macroarray were chosen for examination with RTQ-PCR using a separate cohort of 19 patients (see Materials and Methods).The selected genes were chosen not because of the height of their differential gene expression, but because these genes represent important key regulators of biologic functions like apoptosis, cell cycle control, repair/stress, and signal transduction, which are supposed to be altered in tumors. Of these genes, 93.7% appeared similarly regulated in this separate cohort of 19 patients (Fig 4B). The disagreement between both methods was 6.3%.
Comparison of Differentially Expressed Genes Found in Seminoma and Nonseminoma
Seminoma, like nonseminoma, showed an almost similar up- or downregulation of genes within a functional group. For instance, 80% of genes corresponding to the groups "oncogenes" (four of five), "translation" (one of one), "cell cycle" (six of six), and "DNA synthesis" (five of six) showed an upregulation; however, a downregulation was found in the groups "transcription" (four of four) and "cell signaling" (four of five). Two functional groups ("stress response" and "protein turnover") were characterized by a comparable number of up- and downregulated genes. Furthermore, seminoma, unlike nonseminoma, predominantly showed an upregulation of their genes (63.9% of differentially expressed genes), with only a few genes being downregulated (36.1% of differential expressed genes). When comparing nonseminoma and seminoma, most changes in gene expression were found in six of 23 groups (Fig 5, an appropriate table can be requested from the authors). In detail, two oncogenes (myeloblastosis and sarcoma virus oncogenes) and the nonmetastatic gene (NME1) were upregulated in seminoma, but control levels were found in nonseminoma. With regard to cell cycle control, no changes in gene expression of cyclins D, E, and A (control of G1 and S transition) were found in either entity, but cyclin B1 was upregulated in nonseminoma. An upregulation of UbcH10 was found in seminoma and nonseminoma, which might be indicative of progression through mitosis.26,27 Most of the genes responsible for intracellular signaling (eg, integrin linked kinase, PKC-like 1, MAP3K, RAB 2, G-protein, or Rho GTPase activation protein 5) were downregulated in nonseminoma. Only a few intracellular transducers were upregulated in seminoma, namely tyrosine kinase 2, MAP3K 11, calcium/calmodulin-dependent protein kinase I, and junction plakoglobin. With regard to apoptosis, a silencing in both tumor entities evolved because proapoptotic (caspase 3, calpain, BAD, BAX) as well as antiapoptotic genes (adenosine A1 receptor, BAG-1, Bcl2-like 1, clusterin) were either downregulated or at control level. However, seminoma revealed an upregulation of caspase 3 but at the same time, an upregulation of Bcl2-like 1, with the other genes being either downregulated or at control level. Presumably, even in seminoma, on the gene expression level, apoptosis processes remained silent. DNA synthesis and repair mechanisms, as well as the transcriptional machinery, appeared silenced in nonseminoma because replication factors (3 and C), certain enzymes responsible for different kinds of repair (DNA ligase I and III, ERCC1, XRCC1, DNA-PK, RAD 23 and 51), as well as signal transduction of cytokines (STAT1, glucocorticoid receptor DNA binding factor 1) and a large selection of different DNA binding proteins (GATA binding protein 4, transcriptional adaptor 3-like, TATA box binding protein associated factor), transcription factors (nuclear factor I/X, nuclear transcription factor Y, alpha, heat shock transcription factor 1), regulators of chromatin (SWI/SNF A2 and B1), and splicing factor 1 were at control or downregulated. With regard to transcriptional activity, silencing was also found in seminoma, but functions related to DNA synthesis (PCNA), proliferation (PCNA, MCM2), and repair (DNA-PK and RAD51) were upregulated. Taken together, nonseminoma were characterized by a silencing of many functions (oncogenes, intracellular signaling, apoptosis machinery, DNA synthesis and repair, as well as transcription) while a positive input on cell cycle control was found. Seminoma were characterized by activated oncogenes, DNA synthesis, proliferation, cell cycle progression, and repair while transcriptional activity and apoptosis appeared silenced.
Comparison of Genes That Showed a Significant Difference in Their Expression in Seminoma Versus Nonseminoma
Between 50 and 100 mg of testis tissue was needed for the isolation of a sufficient amount of total RNA (20 to 50 µg) in order to hybridize it on the gene expression macroarray. The yield of total RNA per milligram of tissue was improved by homogenization (see Materials and Methods) and proteinase K digestion. Both the yield and the quality of RNA could be improved by the use of RNA Later solution. Without that solution, approximately 50% of the specimens could be not used due to degraded RNA. With the use of RNA Later solution in more than 95% of the specimens, no degradation of total RNA occurred. For the purpose of validating the semiquantitative macroarray data, a selected number of genes found to be differently expressed were measured with a second but quantitative method (namely, RTQ-PCR) on the same patients. With both methods, comparable levels of gene expression were found for five of six genes examined, thus confirming the kind of macroarray analysis performed. Moreover, these results are in agreement with previous findings.25 Only for PENK was a 12-fold discrepancy found between array analysis (4.5-fold downregulated) and RTQ-PCR results (55-fold downregulated). However, the expression of PENK was greater than background level in only one of 11 tumors examined using the macroarray; thus, measurements with array technology were inside the linear dynamic range of the method for normal tissue but outside the range in most cases, for tumor tissue (data not shown). This leads to uncertainties and an underestimation in the calculation of differential gene expression. Furthermore, RTQ-PCR in our experiments typically showed a linear dynamic range within six orders of magnitude (Fig 2) and was characterized by a higher specificity and sensitivity due to a gene-specific PPD and the method applied. All measurements for PENK with RTQ-PCR were within the linear dynamic range (data not shown), and, therefore, quantification with RTQ-PCR for PENK was judged to be superior to the results from array technology. Recently, Chu et al introduced a method called significance analysis of microarrays (SAM).28 SAM identified genes with statistically significant changes in expression relative to the standard deviation of repeated measurements for that gene. When comparing SAM with Northern blot analysis, a 10% disagreement was found.28 In our experiment, a comparable percentage of disagreement (6.1%) was found, and the regression lines showed almost similar values when comparing differential gene expression drawn from semiquantitative array analysis and quantitative RTQ-PCR results of the same patients (Fig 4A), thus supporting the array analysis. Moreover, 94% of the 66 genes (six genes examined in 11 patients) appeared similarly up- or downregulated using two different methods (Fig 4A). Besides the variations caused by the maroarray method, it is assumed that interindividual biologic variation probably represents the greatest source of variation in microarray studies.29,30 Therefore, based on the macroarray data of 11 patients, 24 differentially expressed genes were selected, and their expression was measured in a separate cohort of 19 patients (nine seminoma and 10 nonseminoma) using RTQ-PCR. The selected genes were chosen not because of the height of their differential gene expression, but because these genes represent important key regulators of biologic functions like apoptosis, cell cycle control, repair/stress, and signal transduction, which are supposed to be altered in tumors. When plotting mean differential gene expression of these different cohorts of patients and tumor entities (representing a total of 456 individual measurements), a similar regulation for 93.7% of the genes could be shown (Fig 4B), supporting the view that most of the selected genes of the macroarray indeed represent interindividually independent expressed genes. A few of the genes examined are already known to be differentially expressed in testis tumor cells. For instance, reduced amounts of clusterin protein were observed in testicular germ cell tumors.31 Our data correspond to these published findings, whereas an upregulation for cytochrome P450 (detected in leydig tumor cells)32 is in contrast to our findings because cytochrome P450 was found to be downregulated (1.5-fold; nonsignificant) as well as cytochrome P450, subfamily XIA (7.4-fold; P < .001). The published upregulation for zinc finger genes (found in seminoma)33 could not be confirmed. Instead, a downregulation for zinc finger protein 36 in seminoma (5.6-fold; P < .02) and zinc finger protein 161 in nonseminoma (1.2-fold; P = .03) was found. Interestingly, genes that were up- or downregulated in one individual, in general, were differently expressed in the same order of magnitude in other individuals, too (Figs 3, 5, and 6). This consistency of the data has also been described by others.22 Certain characteristics being independent of the tumor entity evolved: (1) Most changes of gene expression occurred in five functional groups, namely for genes involved in the control of the "cell cycle," "intracellular transducer," "apoptosis," "DNA synthesis and repair," and "transcription" (Fig 5). (2) Typically, genes within these groups were similarly up- or downregulated. (3) There was a tendency for an upregulation of the cell cycle, but transcriptional and apoptosis machinery appeared silenced. These data highlight the fact that changes in cell cycle control and apoptosis, a focus of research in the past,34–36 are not the only changes that take place. However, major changes in gene expression were found in only a limited number of other functions (see above). Furthermore, a feature of both entities was the silencing of certain functions (transcription and apoptosis). This principle (active silencing of certain functions) was also evident in irradiated MCF-7 cells (breast cancer cell line) and was associated with cell death.25 It could be hypothesized that tumor cells that are supposed to die but fail to do so have features (silencing of different functions) in common with cell death. Other characteristics evolved that were related to the tumor entities examined: (1) nonseminoma, in general, were characterized by a downregulation (75% of differentially expressed genes), while in seminoma, upregulated genes (63.9%) prevailed (Figs 5 and 6). The upregulation of one of these, junction plakoglobulin, was also found by others.22 Moreover, gene groups such as oncogenes, intracellular transducer, DNA synthesis, as well as proliferation and repair seemed to be activated in seminoma only. (2) In 29 (64.4%) of 45 genes, the expression of genes in seminoma was greater (upregulated) than the expression of the same genes in nonseminoma (Fig 6), which, in most cases, was downregulated. Again, this effect was related to functional groups. The opposite pattern was found in 11 genes (24.4%). There was an overlap of 11.1% of genes that showed a similar change in differential gene expression in both tumor entities, but, again, most genes (88.8%) revealed an opposing pattern. The differences of seminoma and nonseminoma found on the gene expression level correspond with the discrimination of these entities due to histologic and clinical criteria. Moreover, these data suggest that seminoma represent a tumor entity that in most respects (88.8%), appeared as a "mirror image" of nonseminoma. Several different mechanisms could be responsible for those findings: (1) it is assumed that seminoma and nonseminoma have the same origin.37 According to our data this suggests that during early stages of testis tumor genesis, two almost opposite ways of progression might exist. However, if we bear in mind the processes of genomic instability/selection taking place during tumorigenesis,38 it is unusual to find only two opposing gene expression profiles, because 11 patients probably represent different stages in tumor progression. (2) It is suggested by others that testis tumors develop as a continuum and that seminoma may progress further into nonseminoma.22,23,39,40 Again, one would expect a mixture of both characteristics typical for seminoma and nonseminoma. On the contrary, the individual gene expression profiles found reveal a remarkable consistency that is in line with findings by others22 and that fall within only two opposing patterns. (3) Interestingly, nonseminoma were characterized by an overall downregulation that resembled gene expression patterns associated with cell death as shown recently.25 With the occurrence of cell death, an "active silencing" of different cellular functions was described. It could be hypothesized that the processes of genomic instability and selection itself, in the end lead to only two "stable" states (equilibrium)—one characterized by features of cell death (without the activation of cell death machinery) and the other characterized by opposite features. Tumors with a progression lying within these two equilibriums might be eliminated by different modes of cell death (selection). This attitude could represent an inherent characteristic of testis tumors. These characteristics restrict the variability of testis tumors: from a gene expression point of view testis tumors reveal an almost homogeneous attitude. This uniformity could explain the success of testis tumor treatment. However, only a limited number of tissues have been examined, thus allowing only generation of hypotheses. Further experiments (eg, microdissections of nonseminomatous germ cell tumors and examinations on a larger cohort) are needed to specify and further validate our findings.
The authors indicated no potential conflicts of interest.
We thank Dr K. Greulich-Bode and Prof Dr D. van Beuningen for their valuable input; Dr R. Girgensohn and M. Mann (from the Institute of Medical Informatics/Statistics, University Gießen, Germany) for their support with statistical analysis; R. Obermair, I. Nuyken, and C. Baaske for their excellent technical assistance; and J. Chen for her assistance with the completion of the Tables.
Supported by the German Ministry of Defense. Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Darzynkiewic Z: Apoptosis in antitumor strategies: Modulation of cell cycle or differentiation. J Cell Biochem 58:151–159, 1995[CrossRef][Medline] 2. Fisher DE: Apoptosis in cancer therapy: Crossing the threshold. Cell 78:539–542, 1994[CrossRef][Medline] 3. Isaacs JT: Advances and controversies in the study of programmed cell death/ apoptosis in the development of and therapy for cancer. Curr Opin Oncol 6:82–89, 1994[Medline] 4. Kerr JFR, Wyllie AH, Currie AR: Apoptosis: A basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 26:239–257, 1972[Medline] 5. Lee JM, Bernstein A: Apoptosis, cancer and the p53 tumor suppressor gene. Cancer Metastasis Rev 14:149–161, 1995[CrossRef][Medline] 6. Rubin LL, Philpott KL, Brooks SF: The cell cycle and cell death. Curr Biol 3:391–394, 1993[CrossRef][Medline] 7. Abend M, Schmelz HU, Kraft K, et al: Intercomparison of apoptosis morphology with DNA cleavage on single cells in vitro and on testis tumors. J Pathol 185:419–426, 1998[CrossRef][Medline] 8. Abend M, Schmelz HU, Kraft K, et al: Modified approach for apoptosis detection reveals changes in apoptotic processes in the seminoma-associated tissue. Apoptosis 4:283–290, 1999[Medline] 9. Bartkova J, Rajpert de Meyts E, Skakkebak NE, et al: D-type cyclins in adult human testis and testicular cancer: Relation to cell type, proliferation, differentiation, and malignancy. J Pathol 187:573–581, 1999[CrossRef][Medline] 10. Cheville JC, Sebo TJ, Lager DJ, et al: Leydig cell tumor of the testis: A clinicopathologic, DNA content, and MIB-1 comparison of nonmetastasizing and metastasizing tumors. Am J Surg Pathol 22:1361–1367, 1998[CrossRef][Medline] 11. Heiskanen P, Billig H, Toppari J, et al: Apoptotic cell death in the normal and cryptorchid human testis: The effect of human chorionic gonadotropin on testicular cell survival. Pediatr Res 40:351–356, 1996[Medline]
12. Koji T, Hishikawa Y, Ando H, et al: Expression of Fas and Fas ligand in normal and ischemia-reperfusion testes: Involvement of the Fas system in the induction of germ cell apoptosis in the damaged mouse testis. Biol Reprod 64:946–954, 2001 13. Schmelz HU, Abend M, Kraft K, et al: Apoptosis in human embryonal cell carcinoma: Preliminary results. Urol Res 27:368–375, 1999[CrossRef][Medline] 14. Schmelz HU, Abend M, Kraft K, et al: Fas/Fas ligand system and apoptosis induction in testicular carcinoma. Cancer 95:73–81, 2002[CrossRef][Medline] 15. Ulbright TM, Orazim A, de Riese W, et al: The correlation of P53 protein expression with proliferative activity and occult metastases in clinical stage I nonseminomatous germ cell tumors of the testis. Mod Pathol 7:64–68, 1994[Medline]
16. Lee J, Richburg JH, Younkin S, et al: The Fas system is a key regulator of germ cell apoptosis in the testis. Endocrinology 138:2081–2088, 1997
17. Schoch C, Kohlmann A, Schnittger S, et al: Acute myeloid leukemias with reciprocal rearrangements can be distinguished by specific gene expression profiles. Proc Natl Acad Sci U S A 99:10008–10013, 2002 18. Kohlmann A, Schoch C, Schnittger S, et al: Neue Methode zur Diagnostik von Leukämien: DNA-Microarray. Dtsch Med Wochenschr 127:2216–2222, 2002[CrossRef][Medline]
19. Tanaka TS, Jaradat SA, Lim MK, et al: Genome-wide expression profiling of mid-gestation placenta and embryo using a 15,000 mouse developmental cDNA microarray. Proc Natl Acad Sci U S A 97:9127–9132, 2000
20. Sgroi DC, Teng S, Robinson G, et al: In vivo gene expression profile analysis of human breast cancer progression. Cancer Res 59:5656–5661, 1999 21. DeRisi J, Penland L, Brown PO, et al: Use of a cDNA microarray to analyse gene expression patterns in human cancer. Nat Genet 14:457–460, 1996[CrossRef][Medline]
22. Skotheim RI, Monni O, Mousses S, et al: New insights into testicular germ cell tumorigenesis from gene expression profiling. Cancer Res 62:2359–2364, 2002 23. Dieckmann KP, Skakkebaek NE: Carcinoma in situ of the testis: Review of biological and clinical features. Int J Cancer 83:815–822, 1999[CrossRef][Medline] 24. Laguna MP, Pizzocaro G, Klepp O, et al: EAU Working Group on Oncological Urology: EAU guidelines on testicular cancer. Eur Urol 40:102–110, 2001[CrossRef][Medline] 25. Stassen T, Port M, Nuyken I, et al: Radiation-induced gene expression in MCF-7 cells. Int J Radiat Biol 79:319–331, 2003[CrossRef][Medline]
26. Okamoto Y, Ozaki T, Miyazaki K, et al: UbcH10 is the Cancer-related E2 Ubiquitin-conjugating enzyme. Cancer Res 63:4167–4173, 2003
27. Townsley FM, Aristarkhov A, Beck S, et al: Dominant-negative cyclin-selective ubiquitin carrier protein E2-C/UbcH10 blocks cells in metaphase. Proc Natl Acad Sci U S A 94:2362–2367, 1997
28. Tusher VG, Tibshirani R, Chu G: Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 98:5116–5121, 2001 29. Simon R, Radmacher MD, Dobbin K: Design of studies using DNA microarrays. Genet Epidemiol 23:21–36, 2002[CrossRef][Medline] 30. Simon R: Bioinformatics and whole genome technologies. Technical Report 005. ftp://linus.nci.nih.gov/pub/techreport/TechReport005.pdf 31. Behrens P, Jeske W, Wernert N, et al: Downregulation of clusterin expression in testicular germ cell tumors. Pathobiology 69:19–23, 2001[CrossRef][Medline] 32. Young M, Lephart ED, McPhaul MJ: Expression of aromatase cytochrome P450 in rat H540 Leydig tumor cells. J Steroid Biochem Mol Biol 63:37–44, 1997[CrossRef][Medline] 33. Ogawa T, Poncelet DA, Kinoshita Y, et al: Enhanced expression in seminoma of human zinc finger genes located on chromosome 19. Cancer Genet Cytogenet 100:36–42, 1998[CrossRef][Medline] 34. Ashwell JD, Berger NA, Cidlowski JA, et al: Coming to terms with death: Apoptosis in cancer and immune development. Immunol Today 15:147–151, 1994[CrossRef][Medline] 35. Czader M, Mazur J, Pettersson M, et al: Prognostic significance of proliferative and apoptotic fractions in low grade follicle center cell-derived non-Hodgkin's lymphomas. Cancer 77:1180–1188, 1996[CrossRef][Medline] 36. Wheeler JA, Stephens LC, Tornos C, et al: ASTRO research fellowship: Apoptosis as a predictor of tumor response to radiation in stage IB cervical carcinoma. Int J Radiat Oncol Biol Phys 32:1487–1493, 1995[CrossRef][Medline] 37. Heidenreich A, Kuczyk M, Albers P: Molecular pathogenesis and prognostic factors in testicular tumor. Urologe A 37:593–608, 1998[CrossRef][Medline] 38. Meyn RE, Strasfeld L, Allen C: Testing the role of p53 in the expression of genetic instability and apoptosis in ataxia-telangiectasia. Int J Radiat Biol 66:141–149, 1994 (suppl 6) 39. Looijenga LH, Oosterhuis JW: Pathogenesis of testicular germ cell tumors. Rev Reprod 4:90–100, 1999[Abstract] 40. Skakkebaek NE, Berthelsen JG, Giwercman A, et al: Carcinoma in situ of the testis: Possible origin from gonocytes and precursor of all types of germ cell tumors except spermatocytoma. Int J Androl 10:19–28, 1987[Medline] Submitted November 12, 2003; accepted October 4, 2004. This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
