CANCER GENOMICS & PROTEOMICS
Volume 5, Number
1, January-February 2008
|Global Comparative Gene Expression Analysis of Melanoma Patient Samples, Derived Cell Lines and Corresponding Tumor Xenografts. Y.-U. XI, A. RIKER, L.A. SHEVDE-SAMANT, R. SAMANT, C. MORRIS, E. GAVIN, O. FODSTAD, J. JU (Mobile, AL, USA; Oslo, Norway)
|CDH11 Expression is Associated with Survival in Patients with Osteosarcoma. DG. NAKAJIMA, A.-G., SKJALG BRUHEIM, Y. XI, M.-S. JULIAN, F. LECANDA, L. SIERRASESUMAGA, C. MÜLLER, O. FODSTAD, J. JU (Mobile, AL, USA; Pamplona, Spain; Oslo, Norway)
|Simultaneous Modeling of Concentration-effect and Time-course Patterns in Gene Expression Data from Microarrays. Y.-F. BRUN, R. VARMA, S.-M. HECTOR, L. PENDYALA, R. TUMMALA, W.-R. GRECO (Buffalo, NY, USA)
|Modelling Gene Regulation Networks via Multivariate Adaptive Splines.
X. CHEN, H. ZHANG (New Haven,CT, USA)
|The Proteome Profile of the Human Osteosarcoma U2OS Cell Line. K.-N. NIFOROU, A.-K. ANAGNOSTOPOULOS, K. VOUGAS, C. KITTAS, V.-G. GORGOULIS, G.- T. TSANGARIS (Athens, Greece)
CANCER GENOMICS & PROTEOMICS 5: 1-35 (2008)
Global Comparative Gene Expression Analysis of Melanoma Patient Samples, Derived Cell Lines and Corresponding Tumor Xenografts
YAGUANG XI1, ADAM RIKER1, LALITA SHEVDE-SAMANT1,
RAJEEV SAMANT1, CHRISTOPHER MORRIS1, ELAINE GAVIN1,
OYSTEIN FODSTAD1,2, JINGFANG JU1
1Mitchell Cancer Institute-USA, Mobile, AL, 36688, U.S.A.;
2Department of Tumor Biology, Norwegian Radium Hospital, Oslo, Norway
Abstract: Various in vitro and in vivo experimental models have been used for the discovery of genes and pathways involved in melanoma and other types of cancer. However, in many cases, the results from various tumor models failed to be validated successfully in clinical studies. Limited information is available on how closely these models reflect the in vivo physiological conditions. In this study, a comprehensive genomics approach was used to systematically compare the expression patterns of snap frozen samples obtained from patients with primary melanoma, lymph node metastasis, and distant metastases, and compare these patterns to those of their corresponding cell lines and tumor xenografts in nude mice. The GE Healthcare 20k human genome array was used and the expression data was normalized and analyzed using GeneSpring 7.2 software. Based on the expression analysis, the correlation rate between the snap frozen primary patient samples vs. derived cell lines was 66%, with 1687 differentially expressed genes. The correlation rate between the snap frozen primary patient samples and the tumor xenografts was 75%, with 1,374 differentially expressed genes, and the correlation rate comparing tumor xenografts to derived cell lines ranged between 58% and 84%. These results demonstrated significant gene expression differences between tumor materials with different in vitro and in vivo growth microenvironments. Such studies can help us to distinguish between genes up- or down-regulated as a result of the microenvironment and those stably expressed independently of the tumor milieu. With the extensive use of cell lines and xenografts in cancer research, the information obtained using our approach may help to better interpret results generated from different tumor models by understanding common differences, as well as similarities at the gene expression level, information that may have important practical and biological implications.
CANCER GENOMICS & PROTEOMICS 5: 37-42 (2008)
CDH11 Expression is Associated with Survival in Patients with Osteosarcoma
GO NAKAJIMA1*, ANA PATINO-GARCIA2*, SKJALG BRUHEIM3,
YAGUANG XI1, MIKEL SAN JULIAN2, FERNANDO LECANDA4,
LUIS SIERRASESUMAGA2, CHRISTOPH MULLER3,
OYSTEIN FODSTAD1,3, JINGFANG JU1
1Cancer Genomics Laboratory, Mitchell Cancer Institute-USA, Mobile, AL, U.S.A.;
2University Clinic of Navarra, Pamplona, Spain;
3Norwegian Radium Hospital, Oslo, Norway;
4Laboratory of Adhesion and Metastasis, Cancer for Medical Applied Medicine,
CIMA, Pamplona, Spain
Abstract: Previous studies have shown that cadherin-11 (CDH11) may be involved in the metastatic process of osteosarcoma. The correlation of the expression levels of CDH11 in osteosarcoma samples with the risk of disease progression and metastasis was examined. Real time qRT-PCR was used to quantify CDH11 expression in a set of newly established osteosarcoma cell lines, 11 primaries and five metastases, compared to the levels in 12 normal osteoblast cell lines established from healthy bone, and also in a set of 10 snap-frozen osteosarcoma samples. In all cases long term clinical follow-up data was available. The CDH11 expression level decreased gradually from the osteoblast to the primary cell lines (p=0.2184) and further to those established from the tumor metastases (p=0.0275). Importantly, the level of CDH11 expression correlated significantly (p=0.01) with patient survival (Kaplan-Meier survival analysis) in both sample sets (p=0.0128 for the cell lines, p=0.0492 for the biopsies). In conclusion, the results indicate that CDH11 may be useful as a prognostic marker of disease progression and survival in osteosarcoma.
CANCER GENOMICS & PROTEOMICS 5: 43-53 (2008)
Simultaneous Modeling of Concentration-effect and Time-course Patterns in Gene Expression Data from Microarrays
YSEULT F. BRUN1,2, RAM VARMA3, SUZANNE M. HECTOR3,
LAKSHMI PENDYALA3, RAMAKUMAR TUMMALA3,
WILLIAM R. GRECO1,2*
1Cancer Prevention and Population Sciences and 3Department of Medicine,
Roswell Park Cancer Institute, Buffalo, NY 14263; 2University at Buffalo,
School of Pharmacy, Buffalo, NY 14260, U.S.A.
Abstract: Background: Time-course and concentration-effect experiments with multiple time-points and drug concentrations provide far more valuable information than experiments with just two design-points (treated vs. control), as commonly performed in most microarray studies. Analysis of the data from such complex experiments, however, remains a challenge. Materials and Methods: Here we present a semi-automated method for fitting time profiles and concentration-effect patterns, simultaneously, to gene expression data. The submodels for time-course included exponential increase and decrease models with parameters, such as initial expression level, maximum effect, and rate-constant (or half-time). The submodel for concentration-effect was a 4-parameter Hill model. Results: The method was applied to an Affymetrix HG-U95Av2 dataset consisting of 51 arrays. The specific study focused on the effects of two platinum drugs, cisplatin and oxaliplatin, on A2780 human ovarian carcinoma cells. Replicates were available at most time points and concentrations. Eighteen genes were selected, and after selection, time-course and concentration-effect were modeled simultaneously. Conclusion: Comparisons of model parameters helped to distinguish genes with different expression patterns between the two drug treatments. This overall paradigm can help in understanding the molecular mechanisms of the agents, and the timing of their actions.
CANCER GENOMICS & PROTEOMICS 5: 55-62 (2008)
Modelling Gene Regulation Networks via Multivariate Adaptive Splines
XIANG CHEN, HEPING ZHANG
Department of Epidemiology and Public Health and Collaborative Center for
Statistics in Science Yale University School of Medicine,
New Haven, CT 06520-8034, U.S.A.
Abstract: After the completion of sequencing for dozens of genomes, as well as the draft of human genome, a major challenge is to characterize genome-wide transcriptional regulation networks. Identification of regulatory functions for transcription factor binding sites in eukaryotes greatly enhances our understanding of the networks, as it has been done extensively under various physiological conditions in yeast. We propose a novel approach based on multivariate adaptive splines to modelling regulatory roles of motifs in gene expression time series data. By applying the proposed approach on two meiotic datasets, we identified well-documented motifs as well as some novel putative motifs that are involved in the transcriptome reprogramming. In addition to identifying single regulatory motifs, we also modelled and unravelled motifs that manifest their effects through coupling with others in regulatory networks. Our findings reveal the potential of multivariate adaptive splines in deciphering complex and important transcriptional regulatory networks in eukaryotes.
CANCER GENOMICS & PROTEOMICS 5: 63-78 (2008)
The Proteome Profile of the Human Osteosarcoma U2OS Cell Line
KATERINA N. NIFOROU1, ATHANASIOS K. ANAGNOSTOPOULOS2,
KONSTANTINOS VOUGAS2, CHRISTOS KITTAS1,
VASSILIS G. GORGOULIS1, GEORGE T. TSANGARIS2
1Department of Histology-Embryology, School of Medicine, University of Athens;
2Proteomics Research Unit, Centre of Basic Research II,
Biomedical Research Foundation of the Academy of Athens, Athens, Greece
Abstract: The human osteosarcoma U2OS cell line is one of the first generated cell lines and is used in various areas of biomedical research. Knowledge of its protein expression is limited and no comprehensive study on the proteome of this cell line has been reported to date. Proteomics technology was used in order to analyse the proteins of the U2OS cell line. Total protein extracts were separated by two-dimensional gel electrophoresis (2-DE) and analysed by matrix-assisted laser desorption ionisation-mass spectrometry (MALDI-MS) and MALDI--MS-MS following in-gel digestion with trypsin and, finally, protein identification was carried out by peptide mass fingerprint (PMF) and post source decay (PSD), respectively. Approximately 3,000 spots were excised from two 2-DE gels and were analysed, resulting in the identification of 237 different gene products. The majority of the identified proteins were enzymes, regulatory proteins and RNA-associated proteins, while leukocyte markers and oncogenes were also present. Our findings include 11 protooncogenes (FKBP4, SRC8, PSD10, FUBP1, PARK7, NPM, PDIA1, OXRP, SET, TCTP and GRP75) related to the cancerous state of the U2OS cell line. The U2OS 2-DE database provides the basis for future protein studies.