Sjoblom et al. (Research Article, 13 October 2006, p. 268) reported nearly 200 novel cancer genes said to have a 90% probability of being involved in colon or breast cancer. However, their analysis raises two statistical concerns. When these concerns are addressed, few genes with significantly elevated mutation rates remain. Although the biological methodology in Sjoblom et al. is sound, more samples are needed to achieve sufficient power.

Whole genome expression profiles are widely used to discover molecular subtypes of diseases. A remaining challenge is to identify the correspondence or commonality of subtypes found in multiple, independent data sets generated on various platforms. While model-based supervised learning is often used to make these connections, the models can be biased to the training data set and thus miss inherent, relevant substructure in the test data. Here we describe an unsupervised subclass mapping method (SubMap), which reveals common subtypes between independent data sets. The subtypes within a data set can be determined by unsupervised clustering or given by predetermined phenotypes before applying SubMap. We define a measure of correspondence for subtypes and evaluate its significance building on our previous work on gene set enrichment analysis. The strength of the SubMap method is that it does not impose the structure of one data set upon another, but rather uses a bi-directional approach to highlight the common substructures in both. We show how this method can reveal the correspondence between several cancer-related data sets. Notably, it identifies common subtypes of breast cancer associated with estrogen receptor status, and a subgroup of lymphoma patients who share similar survival patterns, thus improving the accuracy of a clinical outcome predictor.

MicroRNAs are recently discovered regulators of gene expression and are becoming increasingly recognized as important regulators of heart function. Genome-wide profiling of microRNAs in human heart failure has not been reported previously. We measured expression of 428 microRNAs in 67 human left ventricular samples belonging to control (n = 10), ischemic cardiomyopathy (ICM, n = 19), dilated cardiomyopathy (DCM, n = 25), or aortic stenosis (AS, n = 13) diagnostic groups. miRNA expression between disease and control groups was compared by ANOVA with Dunnett's post hoc test. We controlled for multiple testing by estimating the false discovery rate. Out of 428 microRNAs measured, 87 were confidently detected; 43 were differentially expressed in at least one disease group. In supervised clustering, microRNA expression profiles correctly grouped samples by their clinical diagnosis, indicating that microRNA expression profiles are distinct between diagnostic groups. This was further supported by class prediction approaches, in which the class (control, ICM, DCM, AS) predicted by a microRNA-based classifier matched the clinical diagnosis 69% of the time (P 0.001). These data show that expression of many microRNAs is altered in heart disease and that different types of heart disease are associated with distinct changes in microRNA expression. These data will guide further studies of the contribution of microRNAs to heart disease pathogenesis.

MicroRNAs (miRNAs) are a new class of small noncoding RNAs that post-transcriptionally regulate the expression of target mRNA transcripts. Many of these target mRNA transcripts are involved in proliferation, differentiation and apoptosis, processes commonly altered during tumorigenesis. Recent work has shown a global decrease of mature miRNA expression in human cancers. However, it is unclear whether this global repression of miRNAs reflects the undifferentiated state of tumors or causally contributes to the transformed phenotype. Here we show that global repression of miRNA maturation promotes cellular transformation and tumorigenesis. Cancer cells expressing short hairpin RNAs (shRNAs) targeting three different components of the miRNA processing machinery showed a substantial decrease in steady-state miRNA levels and a more pronounced transformed phenotype. In animals, miRNA processing-impaired cells formed tumors with accelerated kinetics. These tumors were more invasive than control tumors, suggesting that global miRNA loss enhances tumorigenesis. Furthermore, conditional deletion of Dicer1 enhanced tumor development in a K-Ras-induced mouse model of lung cancer. Overall, these studies indicate that abrogation of global miRNA processing promotes tumorigenesis.

Acute lymphoblastic leukemia (ALL) is the most common childhood cancer, whereas acute myeloid leukemia (AML) is the most common acute leukemia in adults. In general, ALL has a better prognosis than AML. To understand the distinct mechanisms in leukemogenesis between ALL and AML and to identify markers for diagnosis and treatment, we performed a large-scale genome-wide microRNA (miRNA, miR) expression profiling assay and identified 27 miRNAs that are differentially expressed between ALL and AML. Among them, miR-128a and -128b are significantly overexpressed, whereas let-7b and miR-223 are significantly down-regulated in ALL compared with AML. They are the most discriminatory miRNAs between ALL and AML. Using the expression signatures of a minimum of two of these miRNAs resulted in an accuracy rate of >95% in the diagnosis of ALL and AML. The differential expression patterns of these four miRNAs were validated further through large-scale real-time PCR on 98 acute leukemia samples covering most of the common cytogenetic subtypes, along with 10 normal control samples. Furthermore, we found that overexpression of miR-128 in ALL was at least partly associated with promoter hypomethylation and not with an amplification of its genomic locus. Taken together, we showed that expression signatures of as few as two miRNAs could accurately discriminate ALL from AML, and that epigenetic regulation might play an important role in the regulation of expression of miRNAs in acute leukemias.

Nonkeratinizing nasopharyngeal carcinoma (NPC) is 100% associated with Epstein-Barr Virus (EBV) and divided into two subtypes (WHO types II and III) based on histology. We tested whether these subtypes can be distinguished at the molecular genetic level using an algorithm that analyzes sets of related genes (gene set enrichment analysis). We found that a class of IFN-stimulated genes (ISG), frequently associated with the antiviral response, was significantly activated in type III versus type II NPC. Consistent with this, replication of the endogenous EBV was suppressed in type III. A strong association was also seen with a subset of ISGs previously identified in systemic lupus erythematosus, another disease in which 'normal' EBV biology is deregulated, suggesting that this pattern of ISG expression may be linked to the increased EBV activity in both diseases. In contrast, unsupervised hierarchical clustering of the complete expression profiles failed to distinguish the two subsets. These results suggest that type II and III NPC have not originated from obviously distinct epithelial precursors; rather, the histologic differences may be a consequence of a differential antiviral response, involving IFNs, to chronic EBV infection.

In spite of the therapeutic importance of endoderm derivatives such as the pancreas, liver, lung, and intestine, there are few molecular markers specific for early endoderm. In order to identify endoderm-specific genes as well as to define transcriptional differences between definitive and visceral endoderm, we performed microarray analysis on E8.25 definitive and visceral endoderm. We have developed an early endoderm gene expression signature, and clarified the transcriptional similarities and differences between definitive and visceral endoderm. Additionally, we have developed methods for flow cytometric isolation of definitive and visceral endoderm. These results shed light on the mechanism of endoderm formation and should facilitate investigation of endoderm formation from embryonic stem cells.

Gene expression mapping using microarray analysis has identified useful gene signatures for predicting outcome. However, little of this has been translated into clinically effective diagnostic tools as microarrays require high quality fresh-frozen tissue samples. We describe a methodology of multiplexed in situ hybridization (ISH) using a novel combination of quantum dot (QD)-labeled oligonucleotide probes and spectral imaging analysis in routinely processed, formalin-fixed paraffin embedded human biopsies. The conditions for QD-ISH were optimized using a poly d(T) oligonucleotide in decalcified bone marrow samples. Single and multiplex QD-ISH was performed in samples with acute leukemia and follicular lymphoma using oligonucleotide probes for myeloperoxidase, bcl-2, survivin, and XIAP. Spectral imaging was used for post hybridization tissue analysis, enabling separation of spatially colocalized signals. The method allows quantitative characterization of multiple gene expression using non-bleaching fluorochromes. This is expected to facilitate multiplex in situ transcript detection in routinely processed human clinical tissue.

Drug resistance remains a major obstacle to successful cancer treatment. A database of drug-associated gene expression profiles was screened for molecules whose profile overlapped with a gene expression signature of glucocorticoid (GC) sensitivity/resistance in acute lymphoblastic leukemia (ALL) cells. The screen indicated that the mTOR inhibitor rapamycin profile matched the signature of GC sensitivity. We tested the hypothesis that rapamycin would induce GC sensitivity in lymphoid malignancy cells and found that it sensitized to GC-induced apoptosis via modulation of antiapoptotic MCL1. These data indicate that MCL1 is an important regulator of GC-induced apoptosis and that the combination of rapamycin and glucocorticoids has potential utility in lymphoid malignancies. Furthermore, this approach represents a strategy for identification of promising combination therapies for cancer.