Array Comparative Genomic Hybridization Reveals Distinct DNA Copy Number Differences between Gastrointestinal Stromal Tumors and Leiomyosarcomas

Array-CGH reveals recurrent genomic changes in Merkel cell carcinoma including amplification of L-Myc

Merkel cell carcinoma (MCC) is an aggressive neuroendocrine skin cancer with poorly characterized genetics. We performed high resolution comparative genomic hybridization on 25 MCC specimens using a high-density oligonucleotide microarray. Tumors frequently carried extra copies of chromosomes 1, 3q, 5p, and 6 and lost chromosomes 3p, 4, 5q, 7, 10, and 13. MCC tumors with less genomic aberration were associated with improved survival (P=0.04). Tumors from 13 of 22 MCC patients had detectable Merkel cell polyomavirus DNA, and these tumors had fewer genomic deletions. Three regions of genomic alteration were of particular interest: a deletion of 5q12-21 occurred in 26% of tumors, a deletion of 13q14-21 was recurrent in 26% of tumors and contains the well-characterized tumor suppressor RB1, and a previously unreported focal amplification at 1p34 was present in 39% of tumors and centers on L-Myc (MYCL1). L-Myc is related to the c-Myc proto-oncogene, has transforming activity, and is amplified in the closely related small cell lung cancer. Normal skin showed no L-Myc expression, whereas 4/4 MCC specimens tested expressed L-Myc RNA in relative proportion to the DNA copy number gain. These findings suggest several genes that may contribute to MCC pathogenesis, most notably L-Myc.

Genetic aberrations of gastrointestinal stromal tumors

Gastrointestinal stromal tumor (GIST) is the most common mesenchymal neoplasm in the gastrointestinal tract and is associated with mutations of the KIT or PDGFRA gene. In addition, other genetic events are believed to be involved in GIST tumorigenesis. Cytogenetic aberrations associated with these tumors thus far described include loss of 1p, 13q, 14q, or 15q, loss of heterozygosity of 22q, numeric chromosomal imbalances, and nuclear/mitochondrial microsatellite instability. Molecular genetic aberrations include loss of heterozygosity of p16(INK4A) and p14(ARF), methylation of p15(INK4B), homozygous loss of the Hox11L1 gene, and amplification of C-MYC, MDM2, EGFR1, and CCND1. GISTs in patients with neurofibromatosis type 1 appear to lack the KIT and PDGFRA mutations characteristic of GISTs and may have a different pathogenetic mechanism. Gene mutations of KIT or PDGFRA are critical in GISTs, because the aberrant versions not only are correlated with the specific cell morphology, histologic phenotype, metastasis, and prognosis, but also are the targets of therapy with imatinib and other agents. Furthermore, specific mutations in KIT and PDGFR appear to lead to differential drug sensitivity and may in the future guide selection of tyrosine kinase inhibitors. Activation of the receptor tyrosine kinases involves a signal transduction pathway whose components (mitogen-activated protein kinase, AKT, phosphoinositide 3-kinase, mammalian target of rapamycin, and RAS) are also possible targets of inhibition. A new paradigm of classification, integrating the standard clinical and pathological criteria with molecular aberrations, may permit personalized prognosis and treatment.

Phylogenetic modeling of heterogeneous gene-expression microarray data from cancerous specimens

The qualitative dimension of gene expression data and its heterogeneous nature in cancerous specimens can be accounted for by phylogenetic modeling that incorporates the directionality of altered gene expressions, complex patterns of expressions among a group of specimens, and data-based rather than specimen-based gene linkage. Our phylogenetic modeling approach is a double algorithmic technique that includes polarity assessment that brings out the qualitative value of the data, followed by maximum parsimony analysis that is most suitable for the data heterogeneity of cancer gene expression. We demonstrate that polarity assessment of expression values into derived and ancestral states, via outgroup comparison, reduces experimental noise; reveals dichotomously expressed asynchronous genes; and allows data pooling as well as comparability of intra- and interplatforms. Parsimony phylogenetic analysis of the polarized values produces a multidimensional classification of specimens into clades that reveal shared derived gene expressions (the synapomorphies); provides better assessment of ontogenic pathways and phyletic relatedness of specimens; efficiently utilizes dichotomously expressed genes; produces highly predictive class recognition; illustrates gene linkage and multiple developmental pathways; provides higher concordance between gene lists; and projects the direction of change among specimens. Further implication of this phylogenetic approach is that it may transform microarray into diagnostic, prognostic, and predictive tool.

Genetic aberrations in soft tissue leiomyosarcoma

Leiomyosarcoma is a malignant mesenchymal tumor composed of cells showing smooth muscle differentiation. This tumor usually occurs in middle-aged or older adults, and forms a significant percentage of retroperitoneal, vascular, extremity, and uterine sarcomas. Leiomyosarcomas are most often associated with complex karyotypes with numerous chromosomal gains and losses. Some of these cytogenetic and molecular genetic aberrations correlate with histopathologic features and clinical outcomes. Identification of genetic alterations with specific identification of oncogenes and tumor suppressor genes may lead to additional insights into the tumorigenesis of leiomyosarcoma and the opportunity to confer the benefits of targeted therapy.

Insulin-like growth factor 1 receptor is a potential therapeutic target for gastrointestinal stromal tumors

A subset of gastrointestinal stromal tumors (GISTs) lack gain-of-function mutations in c-KIT and PDGFRalpha. These so-called wild-type (WT) GISTs tend to be less responsive to imatinib-based therapies and have a poor prognosis. We identified amplification of IGF1R in a SNP analysis of GIST and thus studied its potential as a therapeutic target in WT and mutant GIST. Expression of IGF1R and downstream effectors in clinical GIST samples was examined by using immunoblots and immunohistochemistry. The roles of IGF1R signaling in GIST and viability were analyzed by using NVP-AEW541, an inhibitor of IGF1R, alone and in combination with imatinib, or via siRNA silencing of IGF1R. IGF1R was strongly overexpressed, and IGF1R amplification was detected at a significantly higher frequency in WT GISTs, including a pediatric WT GIST, compared with mutant GISTs (P = 0.0173 and P = 0.0163, respectively). Inhibition of IGF1R activity in vitro with NVP-AEW541 or down-regulation of expression with siIGF1R led to cytotoxicity and induced apoptosis in GIST cell lines via AKT and MAPK signaling. Combination of NVP-AEW541 and imatinib in GIST cell lines induced a strong cytotoxicity response. Our results reveal that IGF1R is amplified and the protein is overexpressed in WT and pediatric GISTs. We also demonstrate that the aberrant expression of IGF1R may be associated with oncogenesis in WT GISTs and suggest an alternative and/or complementary therapeutic regimen in the clinical management of all GISTs, especially in a subset of tumors that respond less favorably to imatinib-based therapy.

DNA copy number changes in high-grade malignant peripheral nerve sheath tumors by array CGH

Background

Malignant peripheral nerve sheath tumors (MPNSTs) are rare and highly aggressive soft tissue tumors showing complex chromosomal aberrations. In order to identify recurrent chromosomal regions of gain and loss, and thereby novel gene targets of potential importance for MPNST development and/or progression, we have analyzed DNA copy number changes in seven high-grade MPNSTs using microarray-based comparative genomic hybridization (array CGH).

Results

Considerable more gains than losses were observed, and the most frequent minimal recurrent regions of gain included 1q24.1-q24.2, 1q24.3-q25.1, 8p23.1-p12, 9q34.11-q34.13 and 17q23.2-q25.3, all gained in five of seven samples. The 17q23.2-q25.3 region was gained in all five patients with poor outcome and not in the two patients with disease-free survival. cDNA microarray analysis and quantitative real-time reverse transcription PCR were used to investigate expression of genes located within these regions. The gene lysyl oxidase-like 2 (LOXL2) was identified as a candidate target for the 8p23.1-p12 gain. Within 17q, the genes topoisomerase II-α (TOP2A), ets variant gene 4 (E1A enhancer binding protein, E1AF) (ETV4) and baculoviral IAP repeat-containing 5 (survivin) (BIRC5) showed increased expression in all samples compared to two benign tumors. Increased expression of these genes has previously been associated with poor survival in other malignancies, and for TOP2A, in MPNSTs as well. In addition, we have analyzed the expression of five micro RNAs located within the 17q23.2-q25.3 region, but none of them showed high expression levels compared to the benign tumors.

Conclusion

Our study shows the potential of using DNA copy number changes obtained by array CGH to predict the prognosis of MPNST patients. Although no clear correlations between the expression level and patient outcome were observed, the genes TOP2A, ETV4 and BIRC5 are interesting candidate targets for the 17q gain associated with poor survival.

GeneCount: genome-wide calculation of absolute tumor DNA copy numbers from array comparative genomic hybridization data

Absolute tumor DNA copy numbers can currently be achieved only on a single gene basis by using fluorescence in situ hybridization (FISH). We present GeneCount, a method for genome-wide calculation of absolute copy numbers from clinical array comparative genomic hybridization data. The tumor cell fraction is reliably estimated in the model. Data consistent with FISH results are achieved. We demonstrate significant improvements over existing methods for exploring gene dosages and intratumor copy number heterogeneity in cancers.

Differential gene expression of 36-kDa microfibril-associated glycoprotein (MAGP-36/MFAP4) in rat organs

By using quantitative Western blot analysis and the real time polymerase chain reaction technique, we investigated the differential gene expression of microfibril-associated glycoprotein (MAGP-36) in rat organs. The gene was expressed highly in sites rich in elastic fibers, such as aorta, skin, and esophagus. However, MAGP-36 was also expressed highly in some other sites containing no elastic fibers. In lung and trachea, the expression levels of MAGP-36 mRNA were about seven times higher than those in other elastic tissues, although the protein abundances were almost at the same levels as other elastic tissues. MAGP-36 seemed to be secreted outside these organs. In brain, kidney, and spleen, although the expression levels of MAGP-36 mRNA were low, substantial amounts of MAGP-36 protein were detected. An immunohistochemical study revealed that MAGP-36 was present at the brush border of the S3 segment of proximal tubules in kidney. Since MAGP-36 is known to bind to mannan, MAGP-36 might be involved in mannose transport in the S3 segment. Thus, MAGP-36 might be multifunctional and present in a wide variety of sites in various organs.

Aberrant methylation status of known methylation-sensitive CpG islands in gastrointestinal stromal tumors without any correlation to the state of c-kit and PDGFRA gene mutations and their malignancy

To identify additional alterations to c-kit or platelet-derived growth factor receptor alpha (PDGFRA) genes in gastrointestinal stromal tumors (GIST), we investigated the methylation status of nine known methylation-sensitive CpG islands (p15, p16, p73, 0-6-methylguanine-DNA methyltransferase, E-cadherin, mutL homolog 1, colon cancer nonpolyposis type 2 (escherichia), methylated in tumors [MINT]1, MINT2, and MINT31), and compared the results with the malignant potential and gain-of-function mutation types of GIST. Thirty-five GIST (c-kit mutations in 25 cases, PDGFRA mutations in seven cases, and lacking either mutation in three cases) were subjected to methylation-specific polymerase chain reaction to detect the methylation status of the nine methylation-sensitive CpG islands. Aberrant DNA methylation of these loci was found in 94% of all GIST. The rates of DNA methylation at each locus were as follows: hMLH1, 60%; MINT2, 51%; MGMT, 49%; p73, 49%; p16, 20%; E-cadherin, 14%; MINT1, 9%; p15, 6%; and MINT31, 0%. CpG islands methylator phenotype, which was defined as methylation involving more than three gene promoters, was found in 57% of GIST with c-kit or PDGFRA gene mutations. According to the risk categories, CpG islands methylator phenotype was present in 55% of low-risk GIST, and in 58% of high-risk GIST. Our results suggested that in addition to c-kit or PDGFRA mutations, the aberrant methylation of CpG islands, especially of mismatch-repair genes, may have a role in the tumorigenesis of GIST.

CGH microarrays and cancer

Genetic alterations are a key feature of cancer cells and typically target biological processes and pathways that contribute to cancer pathogenesis. Array-based comparative genomic hybridization (aCGH) has provided a wealth of new information on copy number changes in cancer on a genome-wide level and aCGH data have also been utilized in cancer classification. More importantly, aCGH analyses have allowed highly accurate localization of specific genetic alterations that, for example, are associated with tumor progression, therapy response, or patient outcome. The genes involved in these aberrations are likely to contribute to cancer pathogenesis, and the high-resolution mapping by aCGH greatly facilitates the subsequent identification of these cancer-associated genes.

Molecular and transcriptional characterization of the novel 17p11.2-p12 amplicon in multiple myeloma

Multiple myeloma (MM) is a malignancy of clonal bone marrow plasma cells characterized by a high genomic instability increasing with disease progression. We describe here a genomic amplification at 17p11.2-p12, an unstable chromosomal region characterized by a large number of low-copy repeats, which have been proven to mediate deletion and duplication in several genomic disorders and amplifications in solid tumors. An approximately 5 Mb 17p11.2-p12 amplified region was detected in the KMS-26 myeloma cell line by SNP microarray analysis. Further fluorescence in situ hybridization mapping showed two unidentified amplified chromosomes as well as a complex pattern of rearranged chromosomes 17. The analysis of transcriptional profiles in a proprietary database of myeloma cell lines identified 12 significantly overexpressed genes in the KMS-26 amplified region, including TNFRSF13B/TACI, COPS3, and NCOR1. The evaluation of their expression levels in a database including 141 plasma cell dyscrasia primary tumors showed a significant overexpression of at least one gene in 13 patients. FISH analyses of these patients identified one MM carrying a 3.8 Mb amplified region and two MMs with gains specifically involving the TACI locus. Interestingly, the complete inactivation of TP53 at 17p13.1 was found in the KMS-26, whereas a monoallelic loss was identifiable in two of the three patients carrying gain/amplification. Our data suggest that, similarly to solid tumors, amplification/gain of the 17p11.2-p12 region in MM could be mediated by the presence of repeats located in this region and may provide insights for defining novel candidate myeloma-associated genes.

Detection of DNA copy number alterations in cancer by array comparative genomic hybridization

Over the past few years, various reliable platforms for high-resolution detection of DNA copy number changes have become widely available. Together with optimized protocols for labeling and hybridization and algorithms for data analysis and representation, this has lead to a rapid increase in the application of this technology in the study of copy number variation in the human genome in normal cells and copy number imbalances in genetic diseases, including cancer. In this review, we briefly discuss specific technical issues relevant for array comparative genomic hybridization analysis in cancer tissues. We specifically focus on recent successes of array comparative genomic hybridization technology in the progress of our understanding of oncogenesis in a variety of cancer types. A third section highlights the potential of sensitive genome-wide detection of patterns of DNA imbalances or molecular portraits for class discovery and therapeutic stratification.

Novel clusters of highly expressed genes accompany genomic amplification in breast cancers

Breast cancer is the most common cancer in women worldwide. To identify novel amplicons involved in the mammary carcinogenesis, we constructed gene expression maps of chromosomes in 35 human breast cancer cell lines and extracted six candidate amplicons containing highly expressed gene clusters on chromosomes 8, 17, and X. We also confirmed the presence of the identified amplicons in clinical specimens by Southern blot analysis. Highly expressed genes identified in the amplicons will contribute to the characterization of breast cancer phenotypes, thereby providing novel targets for anticancer therapies.

Array comparative genomic hybridization analysis of chromosomal imbalances and their target genes in gastrointestinal stromal tumors

Gastrointestinal stromal tumor (GIST) is the most common mesenchymal tumor of the gastrointestinal tract. The tumors characteristically harbor KIT or PDGFRA mutations, and mutant tumors respond to imatinib mesylate (Glivectrade mark). Chromosomal imbalances resulting in altered gene dosage are known to have a role in the molecular pathogenesis of these tumors, but the target genes remain to be identified. The present study aimed to identify some of these genes. In total, 35 GIST samples were screened for chromosomal imbalances by array-based comparative genomic hybridization. A cDNA array was used to define the minimal common overlapping areas of DNA copy number change. Eight confirmative, replicate hybridizations were performed using an oligonucleotide array. The most recurrent copy number losses were localized to 14q, 22q, and 1p. Gains were less common with 8q being the most recurrent. Two recurrent deleted regions of 14q were 14q11.2 harboring the PARP2, APEX1, and NDRG2 genes and 14q32.33 harboring SIVA. Additional target candidates were NF2 at chromosome 22, CDKN2A/2B at 9p, and ENO1 at 1p for copy number losses, and MYC at 8q for copy number gains. Array CGH proved to be an effective tool for the identification of chromosome regions involved in the development and progression of GISTs.

Distinguishing gastrointestinal stromal tumors from their mimics: an update

Since the discovery of activating KIT mutations in gastrointestinal stromal tumors (GISTs) in 1998 and the subsequent demonstration that some malignant GISTs respond to targeted therapy with imatinib, it has become increasingly important for pathologists to correctly diagnose GISTs and separate them from their potential mimics in the gastrointestinal tract and abdominal cavity. Some mesenchymal tumors, such as leiomyomas of the muscularis mucosae, are easily distinguished from GIST on the basis of their anatomic location and morphologic appearance. Others, such as gastrointestinal schwannomas, can significantly overlap with GIST in their gross appearance and morphology and require a panel of immunostains for correct diagnosis. This article will review the most common mimics of GISTs: desmoid tumors, smooth muscle tumors (leiomyomas and leiomyosarcomas), gastrointestinal schwannomas, inflammatory fibroid polyps, and solitary fibrous tumors. Pertinent differences between each of these tumors and GIST in terms of gross appearance, histologic features, and immunophenotype will be emphasized. It is important to separate GISTs from these potential mimics because their treatment and prognosis can differ markedly.

Adult human sarcomas. II. Medical oncology

Human sarcoma cells can be killed by radio- and chemotherapy, but tumor cells acquiring resistance frequently kill the patient. A keen understanding of the intracellular course of oncogenic cascades leads to the discovery of small molecular inhibitors of the involved phosphorylated kinases. Targeted therapy complements chemotherapy. Oncogene silencing is feasible by small interfering RNA. The restoration of some of the mutated or deleted tumor-suppressor genes (p53, Rb, PTEN, hSNF, INK/ARF and WT) by demethylation or reacetylation of their histones has been accomplished. Genetically engineered or naturally oncolytic viruses selectively lyse tumors and leave healthy tissues intact. Adeno- or retroviral vectors deliver genes of immunological costimulators, tumor antigens, chemo- or cytokines and/or tumor-suppressor proteins into tumor (sarcoma) cells. Suicide gene delivery results in apoptosis induction. Genes of enzymes that target prodrugs as their substrates render tumor cells highly susceptible to chemotherapy, with the prodrug to be targeted intracellularly. It will be combinations of sophisticated surgical removal of the nonencapsulated and locally invasive primary sarcomas, advanced forms of radiotherapy to the involved sites and immunotherapy with sarcoma vaccines that will cure primary sarcomas. Adoptive immunotherapy with immune lymphocytes will be operational in metastatic disease only when populations of regulatory T cells are controlled. Targeted therapy with small molecular inhibitors of oncogene cascades, the driving forces of sarcoma cells, alteration of the tumor stroma from a supportive to a tumor-hostile environment, reactivation or replacement of wild-type tumor-suppressor genes, and radio-chemotherapy (with much reduced toxicity) will eventually accomplish the cure of metastatic sarcomas.

Highly accurate two-gene classifier for differentiating gastrointestinal stromal tumors and leiomyosarcomas

Gastrointestinal stromal tumor (GIST) has emerged as a clinically distinct type of sarcoma with frequent overexpression and mutation of the c-Kit oncogene and a favorable response to imatinib mesylate [also known as STI571 (Gleevec)] therapy. However, a significant diagnostic challenge remains in the differentiation of GIST from leiomyosarcomas (LMSs). To improve on the diagnostic evaluation and to complement the immunohistochemical evaluation of these tumors, we performed a whole-genome gene expression study on 68 well characterized tumor samples. Using bioinformatic approaches, we devised a two-gene relative expression classifier that distinguishes between GIST and LMS with an accuracy of 99.3% on the microarray samples and an estimated accuracy of 97.8% on future cases. We validated this classifier by using RT-PCR on 20 samples in the microarray study and on an additional 19 independent samples, with 100% accuracy. Thus, our two-gene relative expression classifier is a highly accurate diagnostic method to distinguish between GIST and LMS and has the potential to be rapidly implemented in a clinical setting. The success of this classifier is likely due to two general traits, namely that the classifier is independent of data normalization and that it uses as simple an approach as possible to achieve this independence to avoid overfitting. We expect that the use of simple marker pairs that exhibit these traits will be of significant clinical use in a variety of contexts.