Design of a multi-signature ensemble classifier predicting neuroblastoma patients' outcome.

BACKGROUND: Neuroblastoma is the most common pediatric solid tumor of the sympathetic nervous system. Development of improved predictive tools for patients stratification is a crucial requirement for neuroblastoma therapy. Several studies utilized gene expression-based signatures to stratify neuroblastoma patients and demonstrated a clear advantage of adding genomic analysis to risk assessment. There is little overlapping among signatures and merging their prognostic potential would be advantageous. Here, we describe a new strategy to merge published neuroblastoma related gene signatures into a single, highly accurate, Multi-Signature Ensemble (MuSE)-classifier of neuroblastoma (NB) patients outcome. METHODS: Gene expression profiles of 182 neuroblastoma tumors, subdivided into three independent datasets, were used in the various phases of development and validation of neuroblastoma NB-MuSE-classifier. Thirty three signatures were evaluated for patients' outcome prediction using 22 classification algorithms each and generating 726 classifiers and prediction results. The best-performing algorithm for each signature was selected, validated on an independent dataset and the 20 signatures performing with an accuracy > = 80% were retained. RESULTS: We combined the 20 predictions associated to the corresponding signatures through the selection of the best performing algorithm into a single outcome predictor. The best performance was obtained by the Decision Table algorithm that produced the NB-MuSE-classifier characterized by an external validation accuracy of 94%. Kaplan-Meier curves and log-rank test demonstrated that patients with good and poor outcome prediction by the NB-MuSE-classifier have a significantly different survival (p < 0.0001). Survival curves constructed on subgroups of patients divided on the bases of known prognostic marker suggested an excellent stratification of localized and stage 4s tumors but more data are needed to prove this point. CONCLUSIONS: The NB-MuSE-classifier is based on an ensemble approach that merges twenty heterogeneous, neuroblastoma-related gene signatures to blend their discriminating power, rather than numeric values, into a single, highly accurate patients' outcome predictor. The novelty of our approach derives from the way to integrate the gene expression signatures, by optimally associating them with a single paradigm ultimately integrated into a single classifier. This model can be exported to other types of cancer and to diseases for which dedicated databases exist.

Bone marrow of neuroblastoma patients shows downregulation of CXCL12 expression and presence of IFN signature.

BACKGROUND: At diagnosis, children with neuroblastoma (NB) present with either localized or metastatic disease. Since the mechanisms responsible for BM invasion are not well known, we investigated the transcriptome of resident BM cells from NB patients as compared to healthy children. PROCEDURE: Ninety-two and 88 children with localized and metastatic NB, respectively, and 15 healthy children were included in the study. BM resident cells recovered from BM aspirates by immunomagnetic bead manipulation were subjected to genome-wide microarray analysis. After validation in an independent set of samples, the genes significantly modulated in resident BM cells from NB patients were tested for their diagnostic/prognostic values. RESULTS: BM resident cells, irrespective of neoplastic cell invasion, significantly overexpressed genes involved in innate immune responses, and interferon (IFN) and IFN-DRS signatures were enriched. Genes coding for metallothioneins and zinc finger proteins, and involved in histone and nucleosome/chromatin organization were also overexpressed. Resident BM cells from NB patients significantly downregulated genes involved in cell adhesion, and in erythrocyte, myeloid, and platelet differentiation pathways. Among downregulated genes, CXCL12 expression reached near complete silencing in patients with metastatic disease. The downregulation of CXCL12 expression was independent of contact between NB cell and resident BM cell. CONCLUSIONS: We demonstrated that NB tumor growth at the primary site can alter the BM microenvironment, and the presence of BM-infiltrating NB cells makes the alterations more pronounced. Therefore, the restoration of a BM physiological state by means of IFN-α monoclonal antibody, Sifalimumab, and selective noradrenaline receptor blockers should be further studied to ameliorate patients' clinical management.

A biology-driven approach identifies the hypoxia gene signature as a predictor of the outcome of neuroblastoma patients.

BACKGROUND: Hypoxia is a condition of low oxygen tension occurring in the tumor microenvironment and it is related to poor prognosis in human cancer. To examine the relationship between hypoxia and neuroblastoma, we generated and tested an in vitro derived hypoxia gene signature for its ability to predict patients' outcome. RESULTS: We obtained the gene expression profile of 11 hypoxic neuroblastoma cell lines and we derived a robust 62 probesets signature (NB-hypo) taking advantage of the strong discriminating power of the l1-l2 feature selection technique combined with the analysis of differential gene expression. We profiled gene expression of the tumors of 88 neuroblastoma patients and divided them according to the NB-hypo expression values by K-means clustering. The NB-hypo successfully stratifies the neuroblastoma patients into good and poor prognosis groups. Multivariate Cox analysis revealed that the NB-hypo is a significant independent predictor after controlling for commonly used risk factors including the amplification of MYCN oncogene. NB-hypo increases the resolution of the MYCN stratification by dividing patients with MYCN not amplified tumors in good and poor outcome suggesting that hypoxia is associated with the aggressiveness of neuroblastoma tumor independently from MYCN amplification. CONCLUSIONS: Our results demonstrate that the NB-hypo is a novel and independent prognostic factor for neuroblastoma and support the view that hypoxia is negatively correlated with tumors' outcome. We show the power of the biology-driven approach in defining hypoxia as a critical molecular program in neuroblastoma and the potential for improvement in the current criteria for risk stratification.

Identification of multiple hypoxia signatures in neuroblastoma cell lines by l1-l2 regularization and data reduction.

Hypoxia is a condition of low oxygen tension occurring in the tumor and negatively correlated with the progression of the disease. We studied the gene expression profiles of nine neuroblastoma cell lines grown under hypoxic conditions to define gene signatures that characterize hypoxic neuroblastoma. The l(1)-l(2) regularization applied to the entire transcriptome identified a single signature of 11 probesets discriminating the hypoxic state. We demonstrate that new hypoxia signatures, with similar discriminatory power, can be generated by a prior knowledge-based filtering in which a much smaller number of probesets, characterizing hypoxia-related biochemical pathways, are analyzed. l(1)-l(2) regularization identified novel and robust hypoxia signatures within apoptosis, glycolysis, and oxidative phosphorylation Gene Ontology classes. We conclude that the filtering approach overcomes the noisy nature of the microarray data and allows generating robust signatures suitable for biomarker discovery and patients risk assessment in a fraction of computer time.

Detection of cell-free RNA in children with neuroblastoma and comparison with that of whole blood cell RNA.

BACKGROUND: Since there is no validated assay to monitor disease in children with neuroblastoma (NB), we tested whether NB specific cell-free RNA could be detected in their plasma samples. Moreover, with the aim of reducing patients' discomfort, we compared this assay to a recently standardized procedure that uses a larger amount of whole blood. PROCEDURES: Using conditions that excluded RNA recovery from contaminating tumor cells, the total amount of cell-free RNA present in healthy children and patients with NB was quantified. Expression of tyrosine hydroxylase (TH) was assayed by quantitative RT-PCR. RESULTS: In patients with NB the amount of cell-free RNA was higher than in healthy children. However, it was less and more degraded than in healthy adults. The median amount of cell-free RNA that was reverse transcribed, measured through the use of standard curves for reference genes, was 0.03 (range 0-30) pg of input RNA, that is, always less than 1/10,000 of that reverse transcribed from total RNA extracted from whole cells. Despite the presence of disease and the positive results obtained with RNA extracted from peripheral blood cells, few cell-free RNA samples tested positive by the TH assay. Similar results were obtained also with TH primers specifically designed to amplify 50 bp RNA fragments. CONCLUSION: These findings suggest that for monitoring disease status detection of cell-free tumor-specific RNAs in patients with NB is not a reliable alternative to whole cell RNA.

The l1-l2 regularization framework unmasks the hypoxia signature hidden in the transcriptome of a set of heterogeneous neuroblastoma cell lines.

BACKGROUND: Gene expression signatures are clusters of genes discriminating different statuses of the cells and their definition is critical for understanding the molecular bases of diseases. The identification of a gene signature is complicated by the high dimensional nature of the data and by the genetic heterogeneity of the responding cells. The l1-l2 regularization is an embedded feature selection technique that fulfills all the desirable properties of a variable selection algorithm and has the potential to generate a specific signature even in biologically complex settings. We studied the application of this algorithm to detect the signature characterizing the transcriptional response of neuroblastoma tumor cell lines to hypoxia, a condition of low oxygen tension that occurs in the tumor microenvironment. RESULTS: We determined the gene expression profile of 9 neuroblastoma cell lines cultured under normoxic and hypoxic conditions. We studied a heterogeneous set of neuroblastoma cell lines to mimic the in vivo situation and to test the robustness and validity of the l1-l2 regularization with double optimization. Analysis by hierarchical, spectral, and k-means clustering or supervised approach based on t-test analysis divided the cell lines on the bases of genetic differences. However, the disturbance of this strong transcriptional response completely masked the detection of the more subtle response to hypoxia. Different results were obtained when we applied the l1-l2 regularization framework. The algorithm distinguished the normoxic and hypoxic statuses defining signatures comprising 3 to 38 probesets, with a leave-one-out error of 17%. A consensus hypoxia signature was established setting the frequency score at 50% and the correlation parameter epsilon equal to 100. This signature is composed by 11 probesets representing 8 well characterized genes known to be modulated by hypoxia. CONCLUSION: We demonstrate that l1-l2 regularization outperforms more conventional approaches allowing the identification and definition of a gene expression signature under complex experimental conditions. The l1-l2 regularization and the cross validation generates an unbiased and objective output with a low classification error. We feel that the application of this algorithm to tumor biology will be instrumental to analyze gene expression signatures hidden in the transcriptome that, like hypoxia, may be major determinant of the course of the disease.

Transcriptome of hypoxic immature dendritic cells: modulation of chemokine/receptor expression.

Hypoxia is a condition of low oxygen tension occurring in inflammatory tissues. Dendritic cells (DC) are professional antigen-presenting cells whose differentiation, migration, and activities are intrinsically linked to the microenvironment. DCs will home and migrate through pathologic tissues before reaching their final destination in the lymph node. We studied the differentiation of human monocytes into immature DCs (iDCs) in a hypoxic microenvironment. We generated iDC in vitro under normoxic (iDCs) or hypoxic (Hi-DCs) conditions and examined the hypoxia-responsive element in the promoter, gene expression, and biochemical KEGG pathways. Hi-DCs had an interesting phenotype represented by up-regulation of genes associated with cell movement/migration. In addition, the Hi-DC cytokine/receptor pathway showed a dichotomy between down-regulated chemokines and up-regulated chemokine receptor mRNA expression. We showed that CCR3, CX3CR1, and CCR2 are hypoxia-inducible genes and that CCL18, CCL23, CCL26, CCL24, and CCL14 are inhibited by hypoxia. A strong chemotactic response to CCR2 and CXCR4 agonists distinguished Hi-DCs from iDCs at a functional level. The hypoxic microenvironment promotes the differentiation of Hi-DCs, which differs from iDCs for gene expression profile and function. The most prominent characteristic of Hi-DCs is the expression of a mobility/migratory rather than inflammatory phenotype. We speculate that Hi-DCs have the tendency to leave the hypoxic tissue and follow the chemokine gradient toward normoxic areas where they can mature and contribute to the inflammatory process.

Normalization of low-density microarray using external spike-in controls: analysis of macrophage cell lines expression profile.

BACKGROUND: The normalization of DNA microarrays allows comparison among samples by adjusting for individual hybridization intensities. The approaches most commonly used are global normalization methods that are based on the expression of all genes on the slide and on the modulation of a small proportion of genes. Alternative approaches must be developed for microarrays where the proportion of modulated genes and their distribution are unknown and they may be biased towards up- or down-modulated trends. RESULTS: The aim of the work is to study the use of spike-in controls to normalize low-density microarrays. Our test-array was designed to analyze gene modulation in response to hypoxia (a condition of low oxygen tension) in a macrophage cell line. RNA was extracted from controls and cells exposed to hypoxia, mixed with spike RNA, labeled and hybridized to our test-array. We used eight bacterial RNAs as source of spikes. The test-array contained the oligonucleotides specific for 178 mouse genes and those specific for the eight spikes. We assessed the quality of the spike signals, the reproducibility of the results and, in general, the nature of the variability. The small values of the coefficients of variation revealed high reproducibility of our platform either in replicated spots or in technical replicates. We demonstrated that the spike-in system was suitable for normalizing our platform and determining the threshold for discriminating the hypoxia modulated genes. We assessed the application of the spike-in normalization method to microarrays in which the distribution of the expression values was symmetric or asymmetric. We found that this system is accurate, reproducible and comparable to other normalization methods when the distribution of the expression values is symmetric. In contrast, we found that the use of the spike-in normalization method is superior and necessary when the distribution of the gene expression is asymmetric and biased towards up-regulated genes. CONCLUSION: We demonstrate that spike-in controls based normalization is a reliable and reproducible method that has the major advantage to be applicable also to biased platform where the distribution of the up- and down-regulated genes is asymmetric as it may occur in diagnostic chips.

Growth arrest-inducing genes are activated in Dbl-transformed mouse fibroblasts.

The Dbl oncogene is a guanine nucleotide exchange factor for Rho GTPases and its activity has been linked to the regulation of gene transcription. Dbl oncogene expression in NIH3T3 cells leads to changes in morphological and proliferative properties of these cells, inducing a highly transformed phenotype. To gain insights into Dbl oncogene-induced transformation we compared gene expression profiles between Dbl oncogene-transformed and parental NIH3T3 cells by cDNA microarray. We found that Dbl oncogene expression is associated with gene expression modulation involving upregulation of 51 genes and downregulation of 49 genes. Five of the overexpressed genes identified are known to exert antiproliferative functions. Our observations suggest that the expression of Dbl oncogene in NIH3T3 may lead to the induction of genes associated with cell cycle arrest, possibly through the activation of stress-induced kinases.

Analysis of HLA-G expression in breast cancer tissues.

Among the different mechanisms by which cancer can elude the immune system, alterations in the expression of human leukocyte antigen (HLA) class I molecules on tumor cells may play a crucial role by impairing the HLA molecules interaction with T and natural killer (NK) cells specific receptors. More recently, aberrant expression of HLA-G has been described in different tumor tissues in addition to HLA class I downregulation. The HLA-G molecule is a nonclassical HLA class I antigen selectively expressed by trophoblast and thymic epithelial cells. Several studies reported that the HLA-G function might represent an additional mechanism of tumor immune escape, mainly inhibiting NK and cytotoxic T-cell activity. Here we report the analysis of HLA-G expression both at RNA level by reverse transcriptase-polymerase chain reaction and at protein level by Western blot and immunohistochemistry in 25 breast cancer patient tissues. The aim of this study was to elucidate the HLA-G gene expression pattern in breast tumor tissues and correlate it with HLA class I alterations. Our results demonstrated that HLA-G molecules expression was never found even in a group of patients revealing HLA class I total loss, and that HLA-G is not expressed in breast cancer tissue with a low-tumor grade (G1-G2) and minimal stromal contamination.

Bone marrow-infiltrating human neuroblastoma cells express high levels of calprotectin and HLA-G proteins.

Metastases in the bone marrow (BM) are grim prognostic factors in patients with neuroblastoma (NB). In spite of extensive analysis of primary tumor cells from high- and low-risk NB patients, a characterization of freshly isolated BM-infiltrating metastatic NB cells is still lacking. Our aim was to identify proteins specifically expressed by metastatic NB cells, that may be relevant for prognostic and therapeutic purposes. Sixty-six Italian children over 18 months of age, diagnosed with stage 4 NB, were included in the study. Metastatic NB cells were freshly isolated from patients' BM by positive immunomagnetic bead manipulation using anti-GD2 monoclonal antibody. Gene expression profiles were compared with those obtained from archived NB primary tumors from patients with 5 y-follow-up. After validation by RT-qPCR, expression/secretion of the proteins encoded by the up-regulated genes in the BM-infiltrating NB cells was evaluated by flow cytometry and ELISA. Compared to primary tumor cells, BM-infiltrating NB cells down-modulated the expression of CX3CL1, AGT, ATP1A2 mRNAs, whereas they up-regulated several genes commonly expressed by various lineages of BM resident cells. BM-infiltrating NB cells expressed indeed the proteins encoded by the top-ranked genes, S100A8 and A9 (calprotectin), CD177 and CD3, and secreted the CXCL7 chemokine. BM-infiltrating NB cells also expressed CD271 and HLA-G. We have identified proteins specifically expressed by BM-infiltrating NB cells. Among them, calprotectin, a potent inflammatory protein, and HLA-G, endowed with tolerogenic properties facilitating tumor escape from host immune response, may represent novel biomarkers and/or targets for therapeutic intervention in high-risk NB patients.

Induction of epithelial mesenchimal transition and vasculogenesis in the lenses of Dbl oncogene transgenic mice.

BACKGROUND: The Dbl family of proteins represents a large group of proto-oncogenes involved in cell growth regulation. The numerous domains that are present in many Dbl family proteins suggest that they act to integrate multiple inputs in complicated signaling networks involving the Rho GTPases. Alterations of the normal function of these proteins lead to pathological processes such as developmental disorders and neoplastic transformation. We generated transgenic mice introducing the cDNA of Dbl oncogene linked to the metallothionein promoter into the germ line of FVB mice and found that onco-Dbl expression in mouse lenses affected proliferation, migration and differentiation of lens epithelial cells. RESULTS: We used high density oligonucleotide microarray to define the transcriptional profile induced by Dbl in the lenses of 2 days, 2 weeks, and 6 weeks old transgenic mice. We observed modulation of genes encoding proteins promoting epithelial-mesenchymal transition (EMT), such as down-regulation of epithelial cell markers and up-regulation of fibroblast markers. Genes encoding proteins involved in the positive regulation of apoptosis were markedly down regulated while anti-apoptotic genes were strongly up-regulated. Finally, several genes encoding proteins involved in the process of angiogenesis were up-regulated. These observations were validated by histological and immunohistochemical examination of the transgenic lenses where vascularization can be readily observed. CONCLUSION: Onco-Dbl expression in mouse lens correlated with modulation of genes involved in the regulation of EMT, apoptosis and vasculogenesis leading to disruption of the lens architecture, epithelial cell proliferation, and aberrant angiogenesis. We conclude that onco-Dbl has a potentially important, previously unreported, capacity to dramatically alter epithelial cell migration, replication, polarization and differentiation and to induce vascularization of an epithelial tissue.

Early response of gene clusters is associated with mouse lung resistance or sensitivity to cigarette smoke.

We have investigated the effects of cigarette smoke exposure in three different strains of mice. DBA/2 and C57BL/6J are susceptible to smoke and develop different lung changes in response to chronic exposure, whereas ICR mice are resistant to smoke and do not develop emphysema. The present study was carried out to determine early changes in the gene expression profile of mice exposed to cigarette smoke with either a susceptible or resistant phenotype. The three strains of mice were exposed to smoke from three cigarettes per day, 5 days/wk, for 4 wk. Microarray analysis was carried out on total RNA extracted from the lung using the Affymetrix platform. Cigarette smoke modulates several clusters of genes (i.e., proemphysematous, acute phase response, and cell adhesion) in smoke-sensitive DBA/2 or C57BL/6J strains, but the same genes are not altered by smoke in ICR resistant mice. Only a few genes were commonly modulated by smoke in the three strains of mice. This pattern of gene expression suggests that the response to smoke is strain-dependent and may involve different molecular signaling pathways. Real-time quantitative PCR was used to verify the pattern of modulation of selected genes and their potential biological relevance. We conclude that gene expression response to smoke is highly dependent on the mouse genetic background. We speculate that the definition of gene clusters associated, to various degrees, with mouse susceptibility or resistance to smoke may be instrumental in defining the molecular basis of the individual response to smoke-induced lung injury in humans.

Hypoxia modifies the transcriptome of primary human monocytes: modulation of novel immune-related genes and identification of CC-chemokine ligand 20 as a new hypoxia-inducible gene.

Peripheral blood monocytes migrate to and accumulate in hypoxic areas of inflammatory and tumor lesions. To characterize the molecular bases underlying monocyte functions within a hypoxic microenvironment, we investigated the transcriptional profile induced by hypoxia in primary human monocytes using high-density oligonucleotide microarrays. Profound changes in the gene expression pattern were detected following 16 h exposure to 1% O(2), with 536 and 677 sequences showing at least a 1.5-fold increase and decrease, respectively. Validation of this analysis was provided by quantitative RT-PCR confirmation of expression differences of selected genes. Among modulated genes, 74 were known hypoxia-responsive genes, whereas the majority were new genes whose responsiveness to hypoxia had not been previously described. The hypoxic transcriptome was characterized by the modulation of a significant cluster of genes with immunological relevance. These included scavenger receptors (CD163, STAB1, C1qR1, MSR1, MARCO, TLR7), immunoregulatory, costimulatory, and adhesion molecules (CD32, CD64, CD69, CD89, CMRF-35H, ITGB5, LAIR1, LIR9), chemokines/cytokines and receptors (CCL23, CCL15, CCL8, CCR1, CCR2, RDC1, IL-23A, IL-6ST). Furthermore, we provided conclusive evidence of hypoxic induction of CCL20, a chemoattractant for immature dendritic cells, activated/memory T lymphocytes, and naive B cells. CCL20 mRNA up-regulation was paralleled by increased protein expression and secretion. This study represents the first transcriptome analysis of hypoxic primary human monocytes, which provides novel insights into monocyte functional behavior within ischemic/hypoxic tissues. CCL20 up-regulation by hypoxia may constitute an important mechanism to promote recruitment of specific leukocyte subsets at pathological sites and may have implications for the pathogenesis of chronic inflammatory diseases.