Search

© 2015 Macmillan Publishers Limited.Defects in primary cilium biogenesis underlie the ciliopathies, a growing group of genetic disorders. We describe a whole-genome siRNA-based reverse genetics screen for defects in biogenesis and/or maintenance of the pri

Abstract Background Classification and variable selection play an important role in knowledge discovery in high-dimensional data. Although Support Vector Machine (SVM) algorithms are among the most powerful classification and prediction methods with a wide range of scientific applications, the SVM does not include automatic feature selection and therefore a number of feature selection procedures have been developed. Regularisation approaches extend SVM to a feature selection method in a flexible way using penalty functions like LASSO, SCAD and Elastic Net. We propose a novel penalty function for SVM classification tasks, Elastic SCAD, a combination of SCAD and ridge penalties which overcomes the limitations of each penalty alone. Since SVM models are extremely sensitive to the choice of tuning parameters, we adopted an interval search algorithm, which in comparison to a fixed grid search finds rapidly and more precisely a global optimal solution. Results Feature selection methods with combined penalties (Elastic Net and Elastic SCAD SVMs) are more robust to a change of the model complexity than methods using single penalties. Our simulation study showed that Elastic SCAD SVM outperformed LASSO (L1) and SCAD SVMs. Moreover, Elastic SCAD SVM provided sparser classifiers in terms of median number of features selected than Elastic Net SVM and often better predicted than Elastic Net in terms of misclassification error. Finally, we applied the penalization methods described above on four publicly available breast cancer data sets. Elastic SCAD SVM was the only method providing robust classifiers in sparse and non-sparse situations. Conclusions The proposed Elastic SCAD SVM algorithm provides the advantages of the SCAD penalty and at the same time avoids sparsity limitations for non-sparse data. We were first to demonstrate that the integration of the interval search algorithm and penalized SVM classification techniques provides fast solutions on the optimization of tuning parameters. The penalized SVM classification algorithms as well as fixed grid and interval search for finding appropriate tuning parameters were implemented in our freely available R package 'penalizedSVM'. We conclude that the Elastic SCAD SVM is a flexible and robust tool for classification and feature selection tasks for high-dimensional data such as microarray data sets.

Abstract Background Interpreting the results of high-throughput experiments, such as those obtained from DNA-microarrays, is an often time-consuming task due to the high number of data-points that need to be analyzed in parallel. It is usually a matter of extensive testing and unknown beforehand, which of the possible approaches for the functional analysis will be the most informative Results To address this problem, we have developed the Flexible Annotation and Correlation Tool (FACT). FACT allows for detection of important patterns in large data sets by simplifying the integration of heterogeneous data sources and the subsequent application of different algorithms for statistical evaluation or visualization of the annotated data. The system is constantly extended to include additional annotation data and comparison methods. Conclusion FACT serves as a highly flexible framework for the explorative analysis of large genomic and proteomic result sets. The program can be used online; open source code and supplementary information are available at http://www.factweb.de.

Stress response pathways are critical for cellular homeostasis, promoting survival through adaptive changes in gene expression and metabolism. They play key roles in numerous diseases and are implicated in cancer progression and chemoresistance. However, the underlying mechanisms are only poorly understood. We have employed a multi-omics approach to monitor changes to gene expression after induction of a stress response pathway, the unfolded protein response (UPR), probing in parallel the transcriptome, the proteome, and changes to translation. Stringent filtering reveals the induction of 267 genes, many of which have not previously been implicated in stress response pathways. We experimentally demonstrate that UPR-mediated translational control induces the expression of enzymes involved in a pathway that diverts intermediate metabolites from glycolysis to fuel mitochondrial one-carbon metabolism. Concomitantly, the cells become resistant to the folate-based antimetabolites Methotrexate and Pemetrexed, establishing a direct link between UPR-driven changes to gene expression and resistance to pharmacological treatment.

The cilium is an essential organelle at the surface of mammalian cells whose dysfunction causes a wide range of genetic diseases collectively called ciliopathies. The current rate at which new ciliopathy genes are identified suggests that many ciliary components remain undiscovered. We generated and rigorously analyzed genomic, proteomic, transcriptomic and evolutionary data and systematically integrated these using Bayesian statistics into a predictive score for ciliary function. This resulted in 285 candidate ciliary genes. We generated independent experimental evidence of ciliary associations for 24 out of 36 analyzed candidate proteins using multiple cell and animal model systems (mouse, zebrafish and nematode) and techniques. For example, we show that OSCP1, which has previously been implicated in two distinct non-ciliary processes, causes ciliogenic and ciliopathy-associated tissue phenotypes when depleted in zebrafish. The candidate list forms the basis of CiliaCarta, a comprehensive ciliary compendium covering 956 genes. The resource can be used to objectively prioritize candidate genes in whole exome or genome sequencing of ciliopathy patients and can be accessed at http://bioinformatics.bio.uu.nl/john/syscilia/ciliacarta/., Competing Interests: The authors have declared that no competing interests exist.

Background: Recent research into ciliary structure and function provides important insights into inherited diseases termed ciliopathies and other cilia-related disorders. This wealth of knowledge needs to be translated into a computational representation to be fully exploitable by the research community. To this end, members of the Gene Ontology (GO) and SYSCILIA Consortia have worked together to improve representation of ciliary substructures and processes in GO., Methods: Members of the SYSCILIA and Gene Ontology Consortia suggested additions and changes to GO, to reflect new knowledge in the field. The project initially aimed to improve coverage of ciliary parts, and was then broadened to cilia-related biological processes. Discussions were documented in a public tracker. We engaged the broader cilia community via direct consultation and by referring to the literature. Ontology updates were implemented via ontology editing tools., Results: So far, we have created or modified 127 GO terms representing parts and processes related to eukaryotic cilia/flagella or prokaryotic flagella. A growing number of biological pathways are known to involve cilia, and we continue to incorporate this knowledge in GO. The resulting expansion in GO allows more precise representation of experimentally derived knowledge, and SYSCILIA and GO biocurators have created 199 annotations to 50 human ciliary proteins. The revised ontology was also used to curate mouse proteins in a collaborative project. The revised GO and annotations, used in comparative 'before and after' analyses of representative ciliary datasets, improve enrichment results significantly., Conclusions: Our work has resulted in a broader and deeper coverage of ciliary composition and function. These improvements in ontology and protein annotation will benefit all users of GO enrichment analysis tools, as well as the ciliary research community, in areas ranging from microscopy image annotation to interpretation of high-throughput studies. We welcome feedback to further enhance the representation of cilia biology in GO.

Cellular organelles provide opportunities to relate biological mechanisms to disease. Here we use affinity proteomics, genetics and cell biology to interrogate cilia: poorly understood organelles, where defects cause genetic diseases. Two hundred and seventeen tagged human ciliary proteins create a final landscape of 1,319 proteins, 4,905 interactions and 52 complexes. Reverse tagging, repetition of purifications and statistical analyses, produce a high-resolution network that reveals organelle-specific interactions and complexes not apparent in larger studies, and links vesicle transport, the cytoskeleton, signalling and ubiquitination to ciliary signalling and proteostasis. We observe sub-complexes in exocyst and intraflagellar transport complexes, which we validate biochemically, and by probing structurally predicted, disruptive, genetic variants from ciliary disease patients. The landscape suggests other genetic diseases could be ciliary including 3M syndrome. We show that 3M genes are involved in ciliogenesis, and that patient fibroblasts lack cilia. Overall, this organelle-specific targeting strategy shows considerable promise for Systems Medicine.

Ciliopathies are Mendelian disorders caused by dysfunction of cilia, ubiquitous organelles involved in fluid propulsion (motile cilia) or signal transduction (primary cilia). Retinal dystrophy is a common phenotypic characteristic of ciliopathies since photoreceptor outer segments are specialized primary cilia. These ciliary structures heavily rely on intracellular minus-end directed transport of cargo, mediated at least in part by the cytoplasmic dynein 1 motor complex, for their formation, maintenance and function. Ninein-like protein (NINL) is known to associate with this motor complex and is an important interaction partner of the ciliopathy-associated proteins lebercilin, USH2A and CC2D2A. Here, we scrutinize the function of NINL with combined proteomic and zebrafish in vivo approaches. We identify Double Zinc Ribbon and Ankyrin Repeat domains 1 (DZANK1) as a novel interaction partner of NINL and show that loss of Ninl, Dzank1 or both synergistically leads to dysmorphic photoreceptor outer segments, accumulation of trans-Golgi-derived vesicles and mislocalization of Rhodopsin and Ush2a in zebrafish. In addition, retrograde melanosome transport is severely impaired in zebrafish lacking Ninl or Dzank1. We further demonstrate that NINL and DZANK1 are essential for intracellular dynein-based transport by associating with complementary subunits of the cytoplasmic dynein 1 motor complex, thus shedding light on the structure and stoichiometry of this important motor complex. Altogether, our results support a model in which the NINL-DZANK1 protein module is involved in the proper assembly and folding of the cytoplasmic dynein 1 motor complex in photoreceptor cells, a process essential for outer segment formation and function.

Defects in primary cilium biogenesis underlie the ciliopathies, a growing group of genetic disorders. We describe a whole-genome siRNA-based reverse genetics screen for defects in biogenesis and/or maintenance of the primary cilium, obtaining a global resource. We identify 112 candidate ciliogenesis and ciliopathy genes, including 44 components of the ubiquitin-proteasome system, 12 G-protein-coupled receptors, and 3 pre-mRNA processing factors (PRPF6, PRPF8 and PRPF31) mutated in autosomal dominant retinitis pigmentosa. The PRPFs localize to the connecting cilium, and PRPF8- and PRPF31-mutated cells have ciliary defects. Combining the screen with exome sequencing data identified recessive mutations in PIBF1, also known as CEP90, and C21orf2, also known as LRRC76, as causes of the ciliopathies Joubert and Jeune syndromes. Biochemical approaches place C21orf2 within key ciliopathy-associated protein modules, offering an explanation for the skeletal and retinal involvement observed in individuals with C21orf2 variants. Our global, unbiased approaches provide insights into ciliogenesis complexity and identify roles for unanticipated pathways in human genetic disease.

Analyzing the molecular architecture of native multiprotein complexes via biochemical methods has so far been difficult and error prone. Protein complex isolation by affinity purification can define the protein repertoire of a given complex, yet, it remains difficult to gain knowledge of its substructure or modular composition. Here, we introduce SDS concentration gradient induced decomposition of protein complexes coupled to quantitative mass spectrometry and in silico elution profile distance analysis. By applying this new method to a cellular transport module, the IFT/lebercilin complex, we demonstrate its ability to determine modular composition as well as sensitively detect known and novel complex components. We show that the IFT/lebercilin complex can be separated into at least five submodules, the IFT complex A, the IFT complex B, the 14-3-3 protein complex and the CTLH complex, as well as the dynein light chain complex. Furthermore, we identify the protein TULP3 as a potential new member of the IFT complex A and showed that several proteins, classified as IFT complex B-associated, are integral parts of this complex. To further demonstrate EPASIS general applicability, we analyzed the modular substructure of two additional complexes, that of B-RAF and of 14-3-3-ε. The results show, that EPASIS provides a robust as well as sensitive strategy to dissect the substructure of large multiprotein complexes in a highly time- as well as cost-effective manner.

Invasion is a critical step in lung tumor progression. The interaction between tumor cells and their surroundings may play an important role in tumor invasion and metastasis. To better understand the mechanisms of tumor invasion and tumor-microenvironment interactions in lung tumors, total RNA was isolated from the inner tumor, tumor invasion front, adjacent lung, and distant normal lung tissue from 17 patients with primary squamous cell lung carcinoma using punch-aided laser capture microdissection. Messenger RNA expression profiles were obtained by microarray analysis, and microRNA profiles were generated from eight of these samples using TaqMan Low Density Arrays. Statistical analysis of the expression data showed extensive changes in gene expression in the inner tumor and tumor front compared with the normal lung and adjacent lung tissue. Only a few genes were differentially expressed between tumor front and the inner tumor. Several genes were validated by immunohistochemistry. Evaluation of the microRNA data revealed zonal expression differences in nearly a fourth of the microRNAs analyzed. Validation of selected microRNAs by in situ hybridization demonstrated strong expression of hsa-miR-196a in the inner tumor; moderate expression of hsa-miR-224 in the inner tumor and tumor front, and strong expression of hsa-miR-650 in the adjacent lung tissue. Pathway analysis placed the majority of genes differentially expressed between tumor and nontumor cells in intrinsic processes associated with inflammation and extrinsic processes related to lymphocyte physiology. Genes differentially expressed between the inner tumor and the adjacent lung/normal lung tissue affected pathways of arachidonic acid metabolism and eicosanoid signaling., (Copyright © 2012 Wiley Periodicals, Inc.)

Microtubule plus-end tracking proteins (+TIPs) are structurally and functionally diverse factors that accumulate at the growing microtubule plus-ends, connect them to various cellular structures, and control microtubule dynamics [1, 2]. EB1 and its homologs are +TIPs that can autonomously recognize growing microtubule ends and recruit to them a variety of other proteins. Numerous +TIPs bind to end binding (EB) proteins through natively unstructured basic and serine-rich polypeptide regions containing a core SxIP motif (serine-any amino acid-isoleucine-proline) [3]. The SxIP consensus sequence is short, and the surrounding sequences show high variability, raising the possibility that undiscovered SxIP containing +TIPs are encoded in mammalian genomes. Here, we performed a proteome-wide search for mammalian SxIP-containing +TIPs by combining biochemical and bioinformatics approaches. We have identified a set of previously uncharacterized EB partners that have the capacity to accumulate at the growing microtubule ends, including protein kinases, a small GTPase, centriole-, membrane-, and actin-associated proteins. We show that one of the newly identified +TIPs, CEP104, interacts with CP110 and CEP97 at the centriole and is required for ciliogenesis. Our study reveals the complexity of the mammalian +TIP interactome and provides a basis for investigating the molecular crosstalk between microtubule ends and other cellular structures., (Copyright © 2012 Elsevier Ltd. All rights reserved.)

Medulloblastoma is the most common malignant brain tumor in children. A subset of medulloblastoma originates from granule cell precursors (GCPs) of the developing cerebellum and demonstrates aberrant hedgehog signaling, typically due to inactivating mutations in the receptor PTCH1, a pathomechanism recapitulated in Ptch1(+/-) mice. As nitric oxide may regulate GCP proliferation and differentiation, we crossed Ptch1(+/-) mice with mice lacking inducible nitric oxide synthase (Nos2) to investigate a possible influence on tumorigenesis. We observed a two-fold higher medulloblastoma rate in Ptch1(+/-) Nos2(-/-) mice compared to Ptch1(+/-) Nos2(+/+) mice. To identify the molecular mechanisms underlying this finding, we performed gene expression profiling of medulloblastomas from both genotypes, as well as normal cerebellar tissue samples of different developmental stages and genotypes. Downregulation of hedgehog target genes was observed in postnatal cerebellum from Ptch1(+/+) Nos2(-/-) mice but not from Ptch1(+/-) Nos2(-/-) mice. The most consistent effect of Nos2 deficiency was downregulation of growth-associated protein 43 (Gap43). Functional studies in neuronal progenitor cells demonstrated nitric oxide dependence of Gap43 expression and impaired migration upon Gap43 knock-down. Both effects were confirmed in situ by immunofluorescence analyses on tissue sections of the developing cerebellum. Finally, the number of proliferating GCPs at the cerebellar periphery was decreased in Ptch1(+/+) Nos2(-/-) mice but increased in Ptch1(+/-) Nos2(-/) (-) mice relative to Ptch1(+/-) Nos2(+/+) mice. Taken together, these results indicate that Nos2 deficiency promotes medulloblastoma development in Ptch1(+/-) mice through retention of proliferating GCPs in the external granular layer due to reduced Gap43 expression. This study illustrates a new role of nitric oxide signaling in cerebellar development and demonstrates that the localization of pre-neoplastic cells during morphogenesis is crucial for their malignant progression., Competing Interests: The authors have declared that no competing interests exist.

Traditionally, protein-protein interactions were thought to be mediated by large, structured domains. However, it has become clear that the interactome comprises a wide range of binding interfaces with varying degrees of flexibility, ranging from rigid globular domains to disordered regions that natively lack structure. Enrichment for disorder in highly connected hub proteins and its correlation with organism complexity hint at the functional importance of disordered regions. Nevertheless, they have not yet been extensively characterised. Shifting the attention from globular domains to disordered regions of the proteome might bring us closer to elucidating the dense and complex connectivity of the interactome. An important class of disordered interfaces are the compact mono-partite, short linear motifs (SLiMs, or eukaryotic linear motifs (ELMs)). They are evolutionarily plastic and interact with relatively low affinity due to the limited number of residues that make direct contact with the binding partner. These features confer to SLiMs the ability to evolve convergently and mediate transient interactions, which is imperative to network evolution and to maintain robust cell signalling, respectively. The ability to discriminate biologically relevant SLiMs by means of different attributes will improve our understanding of the complexity of the interactome and aid development of bioinformatics tools for motif discovery. In this paper, the curated instances currently available in the Eukaryotic Linear Motif (ELM) database are analysed to provide a clear overview of the defining attributes of SLiMs. These analyses suggest that functional SLiMs have higher levels of conservation than their surrounding residues, frequently evolve convergently, preferentially occur in disordered regions and often form a secondary structure when bound to their interaction partner. These results advocate searching for small groupings of residues in disordered regions with higher relative conservation and a propensity to form the secondary structure. Finally, the most interesting conclusions are examined in regard to their functional consequences.

Background: Classification and variable selection play an important role in knowledge discovery in high-dimensional data. Although Support Vector Machine (SVM) algorithms are among the most powerful classification and prediction methods with a wide range of scientific applications, the SVM does not include automatic feature selection and therefore a number of feature selection procedures have been developed. Regularisation approaches extend SVM to a feature selection method in a flexible way using penalty functions like LASSO, SCAD and Elastic Net.We propose a novel penalty function for SVM classification tasks, Elastic SCAD, a combination of SCAD and ridge penalties which overcomes the limitations of each penalty alone.Since SVM models are extremely sensitive to the choice of tuning parameters, we adopted an interval search algorithm, which in comparison to a fixed grid search finds rapidly and more precisely a global optimal solution., Results: Feature selection methods with combined penalties (Elastic Net and Elastic SCAD SVMs) are more robust to a change of the model complexity than methods using single penalties. Our simulation study showed that Elastic SCAD SVM outperformed LASSO (L1) and SCAD SVMs. Moreover, Elastic SCAD SVM provided sparser classifiers in terms of median number of features selected than Elastic Net SVM and often better predicted than Elastic Net in terms of misclassification error.Finally, we applied the penalization methods described above on four publicly available breast cancer data sets. Elastic SCAD SVM was the only method providing robust classifiers in sparse and non-sparse situations., Conclusions: The proposed Elastic SCAD SVM algorithm provides the advantages of the SCAD penalty and at the same time avoids sparsity limitations for non-sparse data. We were first to demonstrate that the integration of the interval search algorithm and penalized SVM classification techniques provides fast solutions on the optimization of tuning parameters.The penalized SVM classification algorithms as well as fixed grid and interval search for finding appropriate tuning parameters were implemented in our freely available R package 'penalizedSVM'.We conclude that the Elastic SCAD SVM is a flexible and robust tool for classification and feature selection tasks for high-dimensional data such as microarray data sets.

To identify novel glioma-associated pathomechanisms and molecular markers, we performed an array-based comparative genomic hybridization analysis of 131 diffuse astrocytic gliomas, including 87 primary glioblastomas (pGBIV), 13 secondary glioblastomas (sGBIV), 19 anaplastic astrocytomas (AAIII) and 12 diffuse astrocytomas (AII). All tumors were additionally screened for IDH1 and IDH2 mutations. Expression profiling was performed for 74 tumors (42 pGBIV, 11 sGBIV, 13 AAIII, 8 AII). Unsupervised and supervised bioinformatic analyses revealed distinct genomic and expression profiles separating pGBIV from the other entities. Classifier expression signatures were strongly associated with the IDH1 gene mutation status. Within pGBIV, the rare subtype of IDH1 mutant tumors shared expression profiles with IDH1 mutant sGBIV and was associated with longer overall survival compared with IDH1 wild-type tumors. In patients with IDH1 wild-type pGBIV, PDGFRA gain or amplification as well as 19q gain were associated with patient outcome. Array-CGH analysis additionally revealed homozygous deletions of the FGFR2 gene at 10q26.13 in 2 pGBIV, with reduced FGFR2 mRNA levels being frequent in pGBIV and linked to poor outcome. In conclusion, we report that diffuse astrocytic gliomas can be separated into 2 major molecular groups with distinct genomic and mRNA profiles as well as IDH1 gene mutation status. In addition, our results suggest FGFR2 as a novel glioma-associated candidate tumor suppressor gene on the long arm of chromosome 10., (Copyright © 2010 UICC.)

Background: Chronic lymphocytic leukemia cells show prolonged survival in vivo, but rapidly die by spontaneous apoptosis in vitro, unless they are co-cultured with stromal cells or non-malignant leukocytes. The objective of this study was to characterize the survival-inducing cross-talk of chronic lymphocytic leukemia cells with their microenvironment to identify novel therapeutic targets., Design and Methods: We analyzed and compared microarray-based expression profiles of chronic lymphocytic leukemia cells before and after three different survival-inducing culture conditions: (i) stromal cell co-culture, (ii) stromal cell conditioned medium and (iii) high cell density cultures of unsorted peripheral blood mononuclear cells. Cytokine antibody arrays were applied to study the composition of soluble factors present in these cultures., Results: The different survival-supportive culture conditions induced distinct gene expression changes, the majority of which were common to all three conditions. Pathway analyses identified - in addition to known signaling networks in chronic lymphocytic leukemia - novel pathways, of which Toll-like receptor signaling, nuclear respiratory factor-2 (NRF2)-mediated oxidative stress response, and signaling via triggering receptor expressed on myeloid cells-1 (TREM1) were the most relevant. A high proportion of up-regulated genes were inflammatory cytokines, of which chemokine (C-C motif) ligand 2 (CCL2) was shown to be induced in monocytes by the presence of chronic lymphocytic leukemia cells in vitro. In addition, increased serum levels of this chemokine were detected in patients with chronic lymphocytic leukemia., Conclusions: Our data provide several lines of evidence that an inflammatory microenvironment is induced in survival-supportive cultures of chronic lymphocytic leukemia cells which might be directly or indirectly involved in the prolonged survival of the malignant cells.

Gene copy number aberrations are involved in oral squamous cell carcinoma (OSCC) development. To delineate candidate genes inside critical chromosomal regions, array-CGH was applied to 40 OSCC specimens using a microarray covering the whole human genome with an average resolution of 1 Mb. Gene copy number gains were predominantly found at 1q23 (9 cases), 3q26 (11), 5p15 (13), 7p11 (7), 8q24 (17), 11q13 (15), 14q32 (8), 19p13 (8), 19q12 (7), 19q13 (8), and 20q13 (9), whereas gene copy number losses were detected at 3p21-3p12 (15), 8p32 (11), 10p12 (8), and 18q21-q23 (10). Subsequent mRNA expression analyses by quantitative real time polymerase chain reaction found high mRNA expression of candidate genes SOX2 in 3q26.33, FSLT3 in 19p13.3, and CCNE1 in 19q12. Tissue microarray (TMA) analyses in a representative OSCC collection found gene copy number gain for SOX2 in 52% (115/223) and for CCNE1 in 31% (72/233) of the tumors. Immunohistochemical analyses on TMA sections of the corresponding proteins detected high expression of SOX2 in 18.1% (49/271) and of CyclinE1 in 23.3% (64/275) of tumors analyzed. These findings indicate that SOX2 and CCNE1 might be activated via gene copy number gain and participate in oral carcinogenesis. The combination of array-CGH with TMA analyses allows rapid pinpointing of novel promising candidate genes, which might be used as therapeutic stratification markers or target molecules for therapeutic interference.

Precursor T-cell acute lymphoblastic leukemia (T-ALL) in children represents a clinical challenge, because relapses are usually fatal. It is thus necessary to identify high-risk patients as early as possible to effectively individualize treatment. We aimed to define novel molecular risk markers in T-ALL and performed array-based comparative genomic hybridization (array-CGH) and expression analyses in 73 patients. We show that DNA copy-number changes are common in T-ALL and affect 70 of 73 (96%) patients. Notably, genomic imbalances predicted to down-regulate the TGF-beta or up-regulate the PI3K-AKT pathways are identified in 25 of 73 (34%) and 21 of 73 (29%) patients, suggesting that these pathways play key roles in T-ALL leukemogenesis. Furthermore, we identified a deletion at 6q15-16.1 in 9 of 73 (12%) of the patients, which predicts poor early treatment response. This deletion includes the CASP8AP2 gene, whose expression is shown to be down-regulated. The interaction of CASP8AP2 with CASP8 plays a crucial role in apoptotic regulation, suggesting a functional link between the clinical effect of the deletion and the molecular mode of action. The data presented here implicate the TGF-beta and PI3K-AKT pathways in T-ALL leukemogenesis and identify a subgroup of patients with CASP8AP2 deletions and poor early treatment response.

Summary: Support vector machine (SVMs) classification is a widely used and one of the most powerful classification techniques. However, a major limitation is that SVM cannot perform automatic gene selection. To overcome this restriction, a number of penalized feature selection methods have been proposed. In the R package 'penalizedSVM' implemented penalization functions L(1) norm and Smoothly Clipped Absolute Deviation (SCAD) provide automatic feature selection for SVM classification tasks., Availability: The R package 'penalizedSVM' is available from the Comprehensive R Archive Network (http://cran.r-project.org/) under GPL-2 or later.

The molecular pathogenesis of pediatric astrocytomas is still poorly understood. To further understand the genetic abnormalities associated with these tumors, we performed a genome-wide analysis of DNA copy number aberrations in pediatric low-grade astrocytomas by using array-based comparative genomic hybridization. Duplication of the BRAF protooncogene was the most frequent genomic aberration, and tumors with BRAF duplication showed significantly increased mRNA levels of BRAF and a downstream target, CCND1, as compared with tumors without duplication. Furthermore, denaturing HPLC showed that activating BRAF mutations were detected in some of the tumors without BRAF duplication. Similarly, a marked proportion of low-grade astrocytomas from adult patients also had BRAF duplication. Both the stable silencing of BRAF through shRNA lentiviral transduction and pharmacological inhibition of MEK1/2, the immediate downstream phosphorylation target of BRAF, blocked the proliferation and arrested the growth of cultured tumor cells derived from low-grade gliomas. Our findings implicate aberrant activation of the MAPK pathway due to gene duplication or mutation of BRAF as a molecular mechanism of pathogenesis in low-grade astrocytomas and suggest inhibition of the MAPK pathway as a potential treatment.

Unlabelled: Hepatocellular carcinoma (HCC) is one of the most common cancers worldwide and is characterized by aggressive tumor behavior coupled with poor prognosis. Various etiologies have been linked to HCC development, most prominently chronic hepatitis B and C virus infections as well as chronic alcohol consumption. In approximately 10% of HCCs, the etiology remains cryptic; however, recent epidemiological data suggest that most of these cryptogenic HCCs develop due to nonalcoholic steatohepatitis. To identify etiology-dependent DNA copy number aberrations and genes relevant to hepatocarcinogenesis, we performed array-based comparative genomic hybridization of 63 HCCs of well-defined etiology and 4 HCC cell lines followed by gene expression profiling and functional analyses of candidate genes. For a 10-megabase chromosome region on 8q24, we observed etiology-dependent copy number gains and MYC overexpression in viral and alcohol-related HCCs, resulting in up-regulation of MYC target genes. Cryptogenic HCCs showed neither 8q24 gains, nor MYC overexpression, nor target gene activation, suggesting that tumors of this etiology develop by way of a distinct MYC-independent pathomechanism. Furthermore, we detected several etiology-independent small chromosome aberrations, including amplification of MDM4 on 1q32.1 and frequent gains of EEF1A2 on 20q13.33. Both genes were overexpressed in approximately half the HCCs examined, and gene silencing reduced cell viability as well as proliferation and increased apoptosis rates in HCC cell lines., Conclusion: Our findings suggest that MDM4 and EEF1A2 act as etiology-independent oncogenes in a significant percentage of HCCs.

Supratentorial primitive neuroectodermal tumors (stPNETs) and medulloblastomas have long been thought to arise from a common cell type in the subventricular germinal matrix. Because of the infrequent occurrence of stPNETs, little is known about their genetic background. Here, we performed a genome-wide screening for DNA copy-number aberrations in 10 supratentorial PNETs using array-based comparative genomic hybridization (array-CGH). Comparing our findings with data from a previous array-CGH study on 47 medulloblastomas, we identified differences in the frequency of copy-number losses at chromosome regions 1p12-22.1 and 9p, and gains at 19p, all of them more frequently occurring in stPNETs. In contrast to previous reports, we detected chromosome 17 aberrations by array-CGH in 2/10 stPNETs. To validate our findings obtained by array-CGH, we analyzed the loci of interest by fluorescence in situ hybridization in an independent set of 11 stPNETs and found deletions of 9p21 in 5/11 tumors of the second set, three of them being homozygous. All 9p21 deletions were associated with loss of CDKN2A protein expression. Altogether, CDKN2A deletions were detected in 7/21 stPNETs including four homozygous deletions, whereas such deletions were only found in 4/112 medulloblastomas, all of these being heterozygous (P < 0.001). Gains of 19p (14% vs. 0% in medulloblastomas, P = 0.02) were found to be significantly more frequent in stPNETs, whereas gains of 17q (14% vs. 45% in medulloblastomas, P = 0.02) were confirmed to be more frequent in medulloblastomas. These data further support the hypothesis of two different tumor entities of embryonal neuroepithelial tumors with characteristic genetic aberrations. (c) 2007 Wiley-Liss, Inc.

We have comparatively analyzed mechanisms associated with chromosomal and microsatellite instability in giant cell glioblastoma multiforme (gcGBM) and classic GBM. This included microsatellite instability (MSI), loss of expression of four major mismatch repair (MMR) proteins, aberrations of five chromosomes, EGFR copy number, and TP53 mutations. MSI was more frequent among gcGBM (30 vs. 7.8%, P = 0.054). TP53 mutations were more commonly observed in gcGBM (83.3%), whereas EGFR was amplified in just one gcGBM (8.3%). By tumor cell phenotype-specific cytogenetic analysis of gcGBM, increased chromosome copy numbers were identified in 72-84% of giant cells but in only 4-14% of nongiant cells; in classic GBM, intermediate frequencies were noted (11-49%). Chromosome 10 deletions were found in nongiant cells of all gcGBM cases but in only approximately 45% of the cell population in classic GBM. The present study shows a distinct pattern of cytogenetic alterations in nongiant and giant cell phenotypes in gcGBM and suggests that multinuclear giant cells evolve from nongiant tumor cells at an early tumor stage. Furthermore, the data point to differences in the profile of chromosomal and microsatellite instability in gcGBM and classic GBM that might underscore the distinct pathological features of both tumor subtypes.

Loss of heterozygosity (LOH) on chromosomal arms 1p and 19q is the most common genetic alteration in oligodendroglial tumors and associated with response to radio- and chemotherapy as well as favorable prognosis. Using microsatellite analysis, we previously identified the chromosomal regions 1p36.22-p36.31 and 19q13.3, as candidate tumor suppressor gene regions being commonly deleted in these tumors. To identify genes within these regions that are downregulated in oligodendroglial tumors with LOH 1p/19q, we performed cDNA microarray-based RNA expression profiling of 35 gliomas with known allelic status on 1p and 19q, including 7 oligodendrogliomas and 8 diffuse astrocytomas of World Health Organization (WHO) grade II, as well as 14 anaplastic oligodendrogliomas and 6 anaplastic oligoastrocytomas of WHO grade III. The microarrays used for expression profiling carried approximately 7,000 gene-specific cDNAs, with complete coverage of the genes located in 1p36.13-p36.31 and 19q13.2-q13.33. Microarray analysis identified 8 genes from these regions (MGC4399, SRM, ICMT, RPL18, FTL, ZIN, FLJ10781 and DBP), which all showed significantly lower expression in 1p/19q-deleted gliomas when compared to gliomas without 1p/19q losses. Quantitative real-time reverse transcription-PCR analyses were performed for the MGC4399, ICMT and RPL18 genes and confirmed the microarray findings. In addition, we found that the cytosolic phospholipase A2 (PLA2G4C) gene at 19q13.3 demonstrated significantly lower expression in anaplastic oligodendrogliomas (WHO grade III) when compared to well-differentiated oligodendrogliomas (WHO grade II). Taken together, our study provides a set of interesting novel candidate genes that may play important roles in the pathogenesis of oligodendroglial tumors., (Copyright 2006 Wiley-Liss, Inc.)

Purpose: Primary systemic therapy (PST) with gemcitabine (G), epirubicin (E), and docetaxel (Doc) has resulted in a pathologic complete response (pCR) in 26% of primary breast cancer patients. This study was aimed at the identification of a gene expression signature in diagnostic core biopsy tissue samples that predicts pCR., Patients and Methods: Core biopsy samples from patients with operable primary breast cancer, T2-4N0-2M0, enrolled onto two phase I and II trials evaluating GEDoc (n = 48) and GE sequentially followed by Doc (GEsDoc; n = 52) as PST were snap frozen and subjected to RNA expression profiling. A signature predicting pCR was discovered in the training set (GEsDoc) applying a support vector machine algorithm, and performance of this classifier was validated on the independent test set (GEDoc) by receiver operator characteristics analysis., Results: We identified a signature consisting of 512 genes, which was enriched in genes involved in transforming growth factor beta and RAS-mediated signaling pathways, that predicts pCR with a sensitivity of 78%, a specificity of 90%, and an overall accuracy of 88% (95% CI, 75% to 95%). Apart from our signature, only HER2 overexpression was an independent predictor of pCR in multivariate analysis., Conclusion: In conclusion, our gene expression signature allows prediction of pCR to PST containing G, E, and Doc with unprecedented high overall accuracy and robustness.

Medulloblastoma is a highly malignant embryonal tumor of the cerebellum that accounts for 20%-25% of all intracranial pediatric tumors. The most frequent chromosomal rearrangement in medulloblastoma is isochromosome 17, or i(17q). Its frequency suggests that it serves an important role in tumor pathogenesis, possibly mediated by the disruption or permanent activation of a gene at the breakpoint. To address this question, we performed a detailed analysis of chromosome 17 DNA copy number from 18 medulloblastomas previously shown to carry an apparent i(17q). We identified two breakpoint regions, one well within band 17p11.2 (n = 16) and a second within the pericentromeric region (n = 2). To map the breakpoints more precisely, we constructed a tiling-path matrix-CGH array covering chromosomal band 17p11.2 to the centromere and utilized it to delineate two small breakpoint intervals mapping at Mb 19.0 and 21.7 in seven of the medulloblastomas and in nine hematological neoplasias with i(17q). The former interval contains two breakpoint clusters that each colocalize with a pair of head-to-head inverted DNA sequence repeats, and the latter maps close to a region of alpha-satellite repeats. No consensus coding sequence localizes in these regions. Together, these data strongly suggest that the effects of i(17q) in medulloblastoma are mediated by gene-dosage effects of genes on 17p or 17q rather than by the disruption or deregulation of a "breakpoint" gene. Furthermore, we identified artifacts introduced in DNA copy number data by cross-hybridization of low-copy repeat sequences and discuss the challenge these can pose in the interpretation of diagnostic microarrays.