Ultra-deep RNA sequencing (RNA-Seq) has become a powerful approach for genome-wide analysis of pre-mRNA alternative splicing. We previously developed multivariate analysis of transcript splicing (MATS), a statistical method for detecting differential alternative splicing between two RNA-Seq samples. Here we describe a new statistical model and computer program, replicate MATS (rMATS), designed for detection of differential alternative splicing from replicate RNASeq data. rMATS uses a hierarchical model to simultaneously account for sampling uncertainty in individual replicates and variability among replicates. In addition to the analysis of unpaired replicates, rMATS also includes a model specifically designed for paired replicates between sample groups. The hypothesis-testing framework of rMATS is flexible and can assess the statistical significance over any user-defined magnitude of splicing change. The performance of rMATS is evaluated by the analysis of simulated and real RNA-Seq data. rMATS outperformed two existing methods for replicate RNA-Seq data in all simulation settings, and RT-PCR yielded a high validation rate (94%) in an RNA-Seq dataset of prostate cancer cell lines. Our data also provide guiding principles for designing RNA-Seq studies of alternative splicing. We demonstrate that it is essential to incorporate biological replicates in the study design. Of note, pooling RNAs or merging RNA-Seq data from multiple replicates is not an effective approach to account for variability, and the result is particularly sensitive to outliers. The rMATS source code is freely available at rnaseq-mats.sourceforge. net/. As the popularity of RNA-Seq continues to grow, we expect rMATS will be useful for studies of alternative splicing in diverse RNA-Seq projects.RNA sequencing | alternative splicing | exon | isoform | transcriptome A lternative splicing generates tremendous transcriptomic and proteomic complexity in higher eukaryotes (1-4). Changes in alternative splicing underlie gene regulation in diverse biological and disease processes (5-7). However, it has been challenging to globally determine and compare gene splicing profiles among biological states. The RNA sequencing (RNA-Seq) technology has become a powerful tool for quantitative profiling of alternative splicing (3,4,8). Due to the high cost, earlier RNA-Seq studies of alternative splicing typically did not incorporate replicates in the study design (9-12). Nonetheless, it is important to note that biological variability remains a critical issue in high-throughput sequencing studies (13). Furthermore, as the cost of sequencing continues to decline, it has become feasible and increasingly common to carry out RNA-Seq on a large number of samples, with sufficient coverage to quantify alternative splicing in each individual sample. This creates an urgent need for new and robust analytic tools to detect alternative splicing changes from replicate RNA-Seq data.Although a variety of computational methods have been developed for RNA-Seq analysis of alternati...

A peripheral membrane protein that is interactive with lymphocytic choriomeningitis virus (LCMV) was purified from cells permissive to infection. Tryptic peptides from this protein were determined to be alpha-dystroglycan (alpha-DG). Several strains of LCMV and other arenaviruses, including Lassa fever virus (LFV), Oliveros, and Mobala, bound to purified alpha-DG protein. Soluble alpha-DG blocked both LCMV and LFV infection. Cells bearing a null mutation of the gene encoding DG were resistant to LCMV infection, and reconstitution of DG expression in null mutant cells restored susceptibility to LCMV infection. Thus, alpha-DG is a cellular receptor for both LCMV and LFV.

Fukuyama congenital muscular dystrophy (FCMD), muscle-eye-brain disease (MEB), and Walker-Warburg syndrome are congenital muscular dystrophies (CMDs) with associated developmental brain defects. Mutations reported in genes of FCMD and MEB patients suggest that the genes may be involved in protein glycosylation. Dystroglycan is a highly glycosylated component of the muscle dystrophin-glycoprotein complex that is also expressed in brain, where its function is unknown. Here we show that brain-selective deletion of dystroglycan in mice is sufficient to cause CMD-like brain malformations, including disarray of cerebral cortical layering, fusion of cerebral hemispheres and cerebellar folia, and aberrant migration of granule cells. Dystroglycan-null brain loses its high-affinity binding to the extracellular matrix protein laminin, and shows discontinuities in the pial surface basal lamina (glia limitans) that probably underlie the neuronal migration errors. Furthermore, mutant mice have severely blunted hippocampal long-term potentiation with electrophysiologic characterization indicating that dystroglycan might have a postsynaptic role in learning and memory. Our data strongly support the hypothesis that defects in dystroglycan are central to the pathogenesis of structural and functional brain abnormalities seen in CMD.

Dystroglycan is a central component of the dystrophin-glycoprotein complex (DGC), a protein assembly that plays a critical role in a variety of muscular dystrophies. In order to better understand the function of dystroglycan in development and disease, we have generated a null allele of dystroglycan (Dag1neo2) in mice. Heterozygous Dag1neo2 mice are viable and fertile. In contrast, homozygous Dag1neo2 embryos exhibit gross developmental abnormalities beginning around 6.5 days of gestation. Analysis of the mutant phenotype indicates that an early defect in the development of homozygous Dag1neo2 embryos is a disruption of Reichert's membrane, an extra-embryonic basement membrane. Consistent with the functional defects observed in Reichert's membrane, dystroglycan protein is localized in apposition to this structure in normal egg cylinder stage embryos. We also show that the localization of two critical structural elements of Reichert's membrane--laminin and collagen IV--are specifically disrupted in the homozygous Dag1neo2 embryos. Taken together, the data indicate that dystroglycan is required for the development of Reichert's membrane. Furthermore, these results suggest that disruption of basement membrane organization might be a common feature of muscular dystrophies linked to the DGC.

Basement membranes are composed of ordered arrays of characteristic extracellular matrix proteins, but little is known about the assembly of these structures in vivo. We have investigated the function of dystroglycan, a cell-surface laminin receptor expressed by cells contacting basement membranes in developing and adult tissues. We find that dystroglycan is required for the formation of a basement membrane in embryoid bodies. Our results further indicate that dystroglycanlaminin interactions are prerequisite for the deposition of other basement membrane proteins. Dystroglycan may exert its influence on basement membrane assembly by binding soluble laminin and organizing it on the cell surface. These data establish a role for dystroglycan in the assembly of basement membranes and suggest fundamental mechanisms underlying this process.

During metastasis, cancer cells enter the circulation in order to gain access to distant tissues, but how this fluid microenvironment influences cancer cell biology is poorly understood. A longstanding view is that circulating cancer cells derived from solid tissues may be susceptible to damage from hemodynamic shear forces, contributing to metastatic inefficiency. Here we report that compared to non-transformed epithelial cells, transformed cells are remarkably resistant to fluid shear stress (FSS) in a microfluidic protocol, exhibiting a biphasic decrease in viability when subjected to a series of millisecond pulses of high FSS. We show that magnitude of FSS resistance is influenced by several oncogenes, is an adaptive and transient response triggered by plasma membrane damage and requires extracellular calcium and actin cytoskeletal dynamics. This novel property of malignant cancer cells may facilitate hematogenous metastasis and indicates, contrary to expectations, that cancer cells are quite resistant to destruction by hemodynamic shear forces.

Dystroglycan is a central component of the dystrophin-glycoprotein complex implicated in the pathogenesis of several neuromuscular diseases. Although dystroglycan is expressed by Schwann cells, its normal peripheral nerve functions are unknown. Here we show that selective deletion of Schwann cell dystroglycan results in slowed nerve conduction and nodal changes including reduced sodium channel density and disorganized microvilli. Additional features of mutant mice include deficits in rotorod performance, aberrant pain responses, and abnormal myelin sheath folding. These data indicate that dystroglycan is crucial for both myelination and nodal architecture. Dystroglycan may be required for the normal maintenance of voltage-gated sodium channels at nodes of Ranvier, possibly by mediating trans interactions between Schwann cell microvilli and the nodal axolemma.

Metastatic colonization involves cancer cell lodgment or adherence in the microvasculature and subsequent migration of those cells across the endothelium into a secondary organ site. To study this process further, we analyzed transendothelial migration of human PC-3 prostate cancer cells in vitro. We isolated a subpopulation of cells, TEM4-18, that crossed an endothelial barrier more efficiently, but surprisingly, were less invasive than parental PC-3 cells in other contexts in vitro. Importantly, TEM4-18 cells were more aggressive than PC-3 cells in a murine metastatic colonization model. Microarray and FACS analysis of these cells showed that the expression of many genes previously associated with leukocyte trafficking and cancer cell extravasation were either unchanged or down-regulated. Instead, TEM4-18 cells exhibited characteristic molecular markers of an epithelial-to-mesenchymal transition (EMT), including frank loss of E-cadherin expression and up-regulation of the E-cadherin repressor ZEB1. Silencing ZEB1 in TEM4-18 cells resulted in increased E-cadherin and reduced transendothelial migration. TEM4-18 cells also express N-cadherin, which was found to be necessary, but not sufficient for increased transendothelial migration. Our results extend the role of EMT in metastasis to transendothelial migration and implicate ZEB1 and N-cadherin in this process in prostate cancer cells. INTRODUCTIONMetastatic prostate cancer is a lethal disease and the second most frequent cause of cancer-related mortality in men in the United States (Jemal et al., 2008). Unfortunately, for such a prevalent disease, much remains unknown about the cellular and molecular mechanisms underlying prostate cancer metastasis. Extravasation, a step within the metastatic cascade, is the process whereby cancer cells exit the circulation via migration through an endothelial monolayer into the parenchyma of a secondary organ site (Wood, 1958). For extravasation to occur, cancer cells, perhaps associated with a thrombus, must first contact the microvascular endothelium, becoming entrapped in small-diameter vessels or adhering specifically to the luminal surface of the endothelium (Warren and Vales, 1972;Kramer and Nicolson, 1979). Some evidence suggests that extravasation is an efficient process, whereas other studies indicate that in some cases it might not occur at all because cancer cells proliferate intraluminally before rupturing microvessels (Crissman et al., 1985;Lapis et al., 1988;Luzzi et al., 1998). Adding to this complexity is that the mechanism of extravasation may depend on the tumor type and vascular bed involved. Thus, as with many other aspects of metastasis, questions remain about the basic mechanisms and the overall role of extravasation in the metastatic cascade.The passage of cancer cells across the endothelium, or transendothelial migration, is thought to be conceptually similar to leukocyte diapedesis (for a recent review see Miles et al., 2008). The extent to which this is true remains controversial, but several classes ...