Claudia Angelini

Uncovering the Complexity of Transcriptomes with RNA-Seq

In recent years, the introduction of massively parallel sequencing platforms for Next Generation Sequencing (NGS) protocols, able to simultaneously sequence hundred thousand DNA fragments, dramatically changed the landscape of the genetics studies. RNA-Seq for transcriptome studies, Chip-Seq for DNA-proteins interaction, CNV-Seq for large genome nucleotide variations are only some of the intriguing new applications supported by these innovative platforms. Among them RNA-Seq is perhaps the most complex NGS application. Expression levels of specific genes, differential splicing, allele-specific expression of transcripts can be accurately determined by RNA-Seq experiments to address many biological-related issues. All these attributes are not readily achievable from previously widespread hybridization-based or tag sequence-based approaches. However, the unprecedented level of sensitivity and the large amount of available data produced by NGS platforms provide clear advantages as well as new challenges and issues. This technology brings the great power to make several new biological observations and discoveries, it also requires a considerable effort in the development of new bioinformatics tools to deal with these massive data files. The paper aims to give a survey of the RNA-Seq methodology, particularly focusing on the challenges that this application presents both from a biological and a bioinformatics point of view.

Time-course analysis of genome-wide gene expression data from hormone-responsive human breast cancer cells

BackgroundMicroarray experiments enable simultaneous measurement of the expression levels of virtually all transcripts present in cells, thereby providing a ‘molecular picture’ of the cell state. On the other hand, the genomic responses to a pharmacological or hormonal stimulus are dynamic molecular processes, where time influences gene activity and expression. The potential use of the statistical analysis of microarray data in time series has not been fully exploited so far, due to the fact that only few methods are available which take into proper account temporal relationships between samples.ResultsWe compared here four different methods to analyze data derived from a time course mRNA expression profiling experiment which consisted in the study of the effects of estrogen on hormone-responsive human breast cancer cells. Gene expression was monitored with the innovative Illumina BeadArray platform, which includes an average of 30-40 replicates for each probe sequence randomly distributed on the chip surface. We present and discuss the results obtained by applying to these datasets different statistical methods for serial gene expression analysis. The influence of the normalization algorithm applied on data and of different parameter or threshold choices for the selection of differentially expressed transcripts has also been evaluated. In most cases, the selection was found fairly robust with respect to changes in parameters and type of normalization. We then identified which genes showed an expression profile significantly affected by the hormonal treatment over time. The final list of differentially expressed genes underwent cluster analysis of functional type, to identify groups of genes with similar regulation dynamics.ConclusionsSeveral methods for processing time series gene expression data are presented, including evaluation of benefits and drawbacks of the different methods applied. The resulting protocol for data analysis was applied to characterization of the gene expression changes induced by estrogen in human breast cancer ZR-75.1 cells over an entire cell cycle.

Massive-Scale RNA-Seq Analysis of Non Ribosomal Transcriptome in Human Trisomy 21

Hybridization- and tag-based technologies have been successfully used in Down syndrome to identify genes involved in various aspects of the pathogenesis. However, these technologies suffer from several limits and drawbacks and, to date, information about rare, even though relevant, RNA species such as long and small non-coding RNAs, is completely missing. Indeed, none of published works has still described the whole transcriptional landscape of Down syndrome. Although the recent advances in high-throughput RNA sequencing have revealed the complexity of transcriptomes, most of them rely on polyA enrichment protocols, able to detect only a small fraction of total RNA content. On the opposite end, massive-scale RNA sequencing on rRNA-depleted samples allows the survey of the complete set of coding and non-coding RNA species, now emerging as novel contributors to pathogenic mechanisms. Hence, in this work we analysed for the first time the complete transcriptome of human trisomic endothelial progenitor cells to an unprecedented level of resolution and sensitivity by RNA-sequencing. Our analysis allowed us to detect differential expression of even low expressed genes crucial for the pathogenesis, to disclose novel regions of active transcription outside yet annotated loci, and to investigate a plethora of non-polyadenilated long as well as short non coding RNAs. Novel splice isoforms for a large subset of crucial genes, and novel extended untranslated regions for known genes—possibly novel miRNA targets or regulatory sites for gene transcription—were also identified in this study. Coupling the rRNA depletion of samples, followed by high-throughput RNA-sequencing, to the easy availability of these cells renders this approach very feasible for transcriptome studies, offering the possibility of investigating in-depth blood-related pathological features of Down syndrome, as well as other genetic disorders.

L-Proline Induces a Mesenchymal-like Invasive Program in Embryonic Stem Cells by Remodeling H3K9 and H3K36 Methylation

Summary Metabolites are emerging as key mediators of crosstalk between metabolic flux, cellular signaling, and epigenetic regulation of cell fate. We found that the nonessential amino acid L-proline (L-Pro) acts as a signaling molecule that promotes the conversion of embryonic stem cells into mesenchymal-like, spindle-shaped, highly motile, invasive pluripotent stem cells. This embryonic-stem-cell-to-mesenchymal-like transition (esMT) is accompanied by a genome-wide remodeling of the H3K9 and H3K36 methylation status. Consistently, L-Pro-induced esMT is fully reversible either after L-Pro withdrawal or by addition of ascorbic acid (vitamin C), which in turn reduces H3K9 and H3K36 methylation, promoting a mesenchymal-like-to-embryonic-stem-cell transition (MesT). These findings suggest that L-Pro, which is produced by proteolytic remodeling of the extracellular matrix, may act as a microenvironmental cue to control stem cell behavior.

Wavelet regression estimation in nonparametric mixed effect models

We show that a nonparametric estimator of a regression function, obtained as solution of a specific regularization problem is the best linear unbiased predictor in some nonparametric mixed effect model. Since this estimator is intractable from a numerical point of view, we propose a tight approximation of it easy and fast to implement. This second estimator achieves the usual optimal rate of convergence of the mean integrated squared error over a Sobolev class both for equispaced and nonequispaced design. Numerical experiments are presented both on simulated and ERP real data.

Understanding gene regulatory mechanisms by integrating ChIP-seq and RNA-seq data: statistical solutions to biological problems

The availability of omic data produced from international consortia, as well as from worldwide laboratories, is offering the possibility both to answer long-standing questions in biomedicine/molecular biology and to formulate novel hypotheses to test. However, the impact of such data is not fully exploited due to a limited availability of multi-omic data integration tools and methods. In this paper, we discuss the interplay between gene expression and epigenetic markers/transcription factors. We show how integrating ChIP-seq and RNA-seq data can help to elucidate gene regulatory mechanisms. In particular, we discuss the two following questions: (i) Can transcription factor occupancies or histone modification data predict gene expression? (ii) Can ChIP-seq and RNA-seq data be used to infer gene regulatory networks? We propose potential directions for statistical data integration. We discuss the importance of incorporating underestimated aspects (such as alternative splicing and long-range chromatin interactions). We also highlight the lack of data benchmarks and the need to develop tools for data integration from a statistical viewpoint, designed in the spirit of reproducible research.

ZFP57 recognizes multiple and closely spaced sequence motif variants to maintain repressive epigenetic marks in mouse embryonic stem cells

Imprinting Control Regions (ICRs) need to maintain their parental allele-specific DNA methylation during early embryogenesis despite genome-wide demethylation and subsequent de novo methylation. ZFP57 and KAP1 are both required for maintaining the repressive DNA methylation and H3-lysine-9-trimethylation (H3K9me3) at ICRs. In vitro, ZFP57 binds a specific hexanucleotide motif that is enriched at its genomic binding sites. We now demonstrate in mouse embryonic stem cells (ESCs) that SNPs disrupting closely-spaced hexanucleotide motifs are associated with lack of ZFP57 binding and H3K9me3 enrichment. Through a transgenic approach in mouse ESCs, we further demonstrate that an ICR fragment containing three ZFP57 motif sequences recapitulates the original methylated or unmethylated status when integrated into the genome at an ectopic position. Mutation of Zfp57 or the hexanucleotide motifs led to loss of ZFP57 binding and DNA methylation of the transgene. Finally, we identified a sequence variant of the hexanucleotide motif that interacts with ZFP57 both in vivo and in vitro. The presence of multiple and closely located copies of ZFP57 motif variants emerges as a distinct characteristic that is required for the faithful maintenance of repressive epigenetic marks at ICRs and other ZFP57 binding sites.

Rebalancing gene haploinsufficiency in vivo by targeting chromatin

Congenital heart disease (CHD) affects eight out of 1,000 live births and is a major social and health-care burden. A common genetic cause of CHD is the 22q11.2 deletion, which is the basis of the homonymous deletion syndrome (22q11.2DS), also known as DiGeorge syndrome. Most of its clinical spectrum is caused by haploinsufficiency of Tbx1, a gene encoding a T-box transcription factor. Here we show that Tbx1 positively regulates monomethylation of histone 3 lysine 4 (H3K4me1) through interaction with and recruitment of histone methyltransferases. Treatment of cells with tranylcypromine (TCP), an inhibitor of histone demethylases, rebalances the loss of H3K4me1 and rescues the expression of approximately one-third of the genes dysregulated by Tbx1 suppression. In Tbx1 mouse mutants, TCP treatment ameliorates substantially the cardiovascular phenotype. These data suggest that epigenetic drugs may represent a potential therapeutic strategy for rescue of gene haploinsufficiency phenotypes, including structural defects.

RNASeqGUI: a GUI for analysing RNA-Seq data

UNLABELLED We present RNASeqGUI R package, a graphical user interface (GUI) for the identification of differentially expressed genes across multiple biological conditions. This R package includes some well-known RNA-Seq tools, available at www.bioconductor.org. RNASeqGUI package is not just a collection of some known methods and functions, but it is designed to guide the user during the entire analysis process. RNASeqGUI package is mainly addressed to those users who have little experience with command-line software. Therefore, thanks to RNASeqGUI, they can conduct analogous analyses using this simple graphical interface. Moreover, RNASeqGUI is also helpful for those who are expert R-users because it speeds up the usage of the included RNASeq methods drastically. AVAILABILITY AND IMPLEMENTATION RNASeqGUI package needs the RGTK2 graphical library to run. This package is open source and is freely available under General Public License at http://bioinfo.na.iac.cnr.it/RNASeqGUI/Download. SUPPLEMENTARY INFORMATION Supplementary data are available at Bioinformatics online.

Computational approaches for isoform detection and estimation: good and bad news

BackgroundThe main goal of the whole transcriptome analysis is to correctly identify all expressed transcripts within a specific cell/tissue - at a particular stage and condition - to determine their structures and to measure their abundances. RNA-seq data promise to allow identification and quantification of transcriptome at unprecedented level of resolution, accuracy and low cost. Several computational methods have been proposed to achieve such purposes. However, it is still not clear which promises are already met and which challenges are still open and require further methodological developments.ResultsWe carried out a simulation study to assess the performance of 5 widely used tools, such as: CEM, Cufflinks, iReckon, RSEM, and SLIDE. All of them have been used with default parameters. In particular, we considered the effect of the following three different scenarios: the availability of complete annotation, incomplete annotation, and no annotation at all. Moreover, comparisons were carried out using the methods in three different modes of action. In the first mode, the methods were forced to only deal with those isoforms that are present in the annotation; in the second mode, they were allowed to detect novel isoforms using the annotation as guide; in the third mode, they were operating in fully data driven way (although with the support of the alignment on the reference genome). In the latter modality, precision and recall are quite poor. On the contrary, results are better with the support of the annotation, even though it is not complete. Finally, abundance estimation error often shows a very skewed distribution. The performance strongly depends on the true real abundance of the isoforms. Lowly (and sometimes also moderately) expressed isoforms are poorly detected and estimated. In particular, lowly expressed isoforms are identified mainly if they are provided in the original annotation as potential isoforms.ConclusionsBoth detection and quantification of all isoforms from RNA-seq data are still hard problems and they are affected by many factors. Overall, the performance significantly changes since it depends on the modes of action and on the type of available annotation. Results obtained using complete or partial annotation are able to detect most of the expressed isoforms, even though the number of false positives is often high. Fully data driven approaches require more attention, at least for complex eucaryotic genomes. Improvements are desirable especially for isoform quantification and for isoform detection with low abundance.