The transcriptional landscape of the mammalian genome

Molecular predictors of response and outcome in ovarian cancer

A major problem in clinical management of patients with epithelial ovarian cancer (EOC) is the largely unpredictable response to first-line treatment and the occurrence of relapse after complete initial response, associated with broad cross-resistance to even structurally dissimilar drugs. During tumor development and progression, multiple genic alterations take place that might contribute specifically to the treatment response and eventually impact on disease outcome. One area of intense research is the identification of molecular markers to accurately assess the prognosis of EOC patients and to define innovative therapeutic strategies. A large survey of recent published data indicates the need to revisit traditional molecular markers with respect to their contribution to the assessment of overall survival in selected populations. Furthermore, recent technological developments that enable simultaneous measurement of many parameters ("omic" approaches) hold the promise of identifying new molecular prognostic and predictive markers.

More than 40,000 transcripts, including novel and noncoding transcripts, in mouse embryonic stem cells

To study the transcriptome of embryonic stem cells, we used a new gene expression profiling method that can measure the expression levels of unknown and rarely expressed transcripts precisely. We detected a total of 33,136 signal peaks representing transcripts in mouse embryonic stem cells, E14. Subsequent random cloning of the peaks suggests that mouse embryonic stem cells express at least 40,000 transcripts, of which about 2,000 are still unknown. In addition, we identified 1,022 noncoding transcripts, several of which change depending on differentiation in gene expression. Our database provides a high-resolution expression profile of E14 cells and is applicable to other mouse embryonic stem cell analyses. It includes most transcription regulation factor-encoding genes and a significant number of unknown and noncoding transcripts.

Diversity in transcriptional start site selection and alternative splicing affects the 5'-UTR of mouse striated muscle myosin transcripts

We have analyzed nearly 2,000 myosin heavy chain gene (Myh) clones representing over 30 different transcripts from seven of eight striated muscle Myh genes expressed in mouse. We also report the transcriptional start sites (TSS) for the mouse developmental Myh genes. The data reveal a previously unknown diversity of TSSs and 5'-end alternative splicing in these transcripts. The cardiac Myh6 gene had two major TSSs. Use of the major downstream site led to an alternatively spliced second exon. Each of the other Myh genes had one major TATA-directed TSS and one or more minor alternative TSSs, some associated with alternative splicing. The minor transcripts were associated with polysomes and their spatial-temporal expression largely mirrored that of the major transcripts in wild-type, Myh1 null, Myh4 null, injured, and uninjured muscle, except that one form of Myh7, detected in heart, was not detected in diaphragm, and the ratio of the two major Myh6 transcripts varied in some circumstances. These findings indicate that alternative TSS usage and alternative splicing in the 5'-UTR are a general feature of murine Myh gene expression and that Myh gene regulation is more complex than previously appreciated.

Gene expression microarrays: glimpses of the immunological genome

Successful microarray experimentation can generate enormous amounts of data, potentially very rich but also very unwieldy. Bold outlooks and new methods for data analysis and presentation should yield additional insight into the complexities of the immune system.

Identification and mapping of DNA binding proteins target sequences in long genomic regions by two-dimensional EMSA

Specific binding of nuclear proteins, in particular transcription factors, to target DNA sequences is a major mechanism of genome functioning and gene expression regulation in eukaryotes. Therefore, identification and mapping specific protein target sites (PTS) is necessary for understanding genomic regulation. Here we used a novel two-dimensional electrophoretic mobility shift assay (2D-EMSA) procedure for identification and mapping of 52 PTS within a 563-kb human genome region located between the FXYD5 and TZFP genes. The PTS occurred with approximately equal frequency within unique and repetitive genomic regions. PTS belonging to unique sequences tended to group together within gene introns and close to their 5′ and 3′ ends, whereas PTS located within repeats were evenly distributed between transcribed and intragenic regions.

Cellular transcriptomics -- the next phase of endocrine expression profiling

Transcriptome analysis, or global gene expression profiling, has become a commonly used and valuable tool in both basic and clinical endocrine research. Novel endocrine regulators have 'surfaced' and greater consideration is now given to understanding function at the level of gene networks. Recent developments have shown that the transcriptome is considerably larger and more divergently expressed than was previously thought. Endocrine cells express a great variety of coding and noncoding RNAs in a highly cell-specific manner. If further value is to be taken from this research area, then steps towards defined cellular transcriptomics must be taken. New sampling techniques that utilize novel genetic models are a key first step.

Non-coding RNAs in the nervous system

Increasing evidence suggests that the development and function of the nervous system is heavily dependent on RNA editing and the intricate spatiotemporal expression of a wide repertoire of non-coding RNAs, including micro RNAs, small nucleolar RNAs and longer non-coding RNAs. Non-coding RNAs may provide the key to understanding the multi-tiered links between neural development, nervous system function, and neurological diseases.

Evidence for variation in abundance of antisense transcripts between multicellular animals but no relationship between antisense transcriptionand organismic complexity

Given that humans have about the same number of genes as mice and not so many more than worm, what makes us more complex? Antisense transcripts are implicated in many aspects of gene regulation. Is there a functional connection between antisense transcription and organismic complexity, that is, is antisense regulation especially prevalent in humans? We used the same robust protocol to identify antisense transcripts in humans and five other metazoan genomes (mouse, rat, chicken, fruit fly, and nematode), and found that the estimated proportions of genes involved in antisense transcription are highly sensitive to the number of transcripts included in the analysis. By controlling for transcript abundance, we find that the probability that any given transcript is putatively involved in sense-antisense regulation is no higher in humans than in other vertebrates but appears unusually high in flies and especially low in nematodes. Similarly, there is no evidence that the proportion of sense-antisense transcripts is especially higher in humans than other vertebrates in a given subset of transcript sequences such as mRNAs, coding sequences, conserved, or nonconserved transcripts. Although antisense transcription might be enriched in mammalian brains compared with nonbrain tissues, it is no more enriched in human brain than in mouse brain. Overall, therefore, while we see striking variation between multicellular animals in the abundance of antisense transcripts, there is no evidence for a link between antisense transcription and organismic complexity. More particularly, we see no evidence that humans are in any way unusual among the vertebrates in this regard. Instead, our results suggest that antisense transcription might be prevalent in almost all metazoan genomes, nematodes being an unexplained exception.

Candidate genes and their regulatory elements: alcohol preference and tolerance

QTL analysis of behavioral traits and mouse brain gene expression studies were combined to identify candidate genes involved in the traits of alcohol preference and acute functional alcohol tolerance. The systematic application of normalization and statistical analysis of differential gene expression, behavioral and expression QTL location, and informatics methodologies resulted in identification of 8 candidate genes for the trait of alcohol preference and 22 candidate genes for acute functional tolerance. Pathway analysis, combined with clustering by ontology, indicated the importance of transcriptional regulation and DNA and protein binding elements in the acute functional tolerance trait, and protein kinases and intracellular signal transduction elements in the alcohol preference trait. A rudimentary search for transcription control elements that could indicate coregulation of the panels of candidate genes produced modest results, implicating SMAD-3 in the regulation of four of the eight candidate genes for alcohol preference. However, the realization of the many caveats related to transcription factor binding site analysis, and attempts to correlate between transcription factor binding and function, forestalled any definitive global analysis of transcriptional control of differentially expressed candidate genes.

Gene-breaking transposon mutagenesis reveals an essential role for histone H2afza in zebrafish larval development

We report a novel gene tagging, identification and mutagenicity ('gene-breaking') method for the zebrafish, Danio rerio. This modular approach consists of two distinct and separable molecular cassettes. The first is a gene-finding cassette. In this study, we employed a 3' gene-tagging approach that selectively 'traps' transcripts regardless of expression status, and we show that this cassette identifies both known and novel endogenous transcripts in transgenic zebrafish. The second is a transcriptional termination mutagenicity cassette assembled from a combination of a splice acceptor and polyadenylation signal to disrupt tagged transcripts upon integration into intronic sequence. We identified both novel and conserved loci as linked phenotypic mutations using this gene-breaking strategy, generating molecularly null mutations in both larval lethal and adult viable loci. We show that the Histone 2a family member z (H2afza) variant is essential for larval development through the generation of a lethal locus with a truncation of conserved carboxy-terminal residues in the protein. In principle this gene-breaking strategy is scalable for functional genomics screens and can be used in Sleeping Beauty transposon and other gene delivery systems in the zebrafish.

Silencing the transcriptome's dark matter: mechanisms for suppressing translation of intergenic transcripts

Large portions of the genomes of higher eukaryotes are transcribed into RNA molecules that are never destined for translation into proteins. Although some of these transcripts have clearly defined biological roles other than protein coding, most arise from genomic regions devoid of functional genes and many are antisense to regions containing annotated genes. A variety of mechanisms exist to prevent adventitious production of proteins from these transcripts, ranging from degradation within the nucleus to translational silencing in the cytosol.

Thousands of corresponding human and mouse genomic regions unalignable in primary sequence contain common RNA structure

Human and mouse genome sequences contain roughly 100,000 regions that are unalignable in primary sequence and neighbor corresponding alignable regions between both organisms. These pairs are generally assumed to be nonconserved, although the level of structural conservation between these has never been investigated. Owing to the limitations in computational methods, comparative genomics has been lacking the ability to compare such nonconserved sequence regions for conserved structural RNA elements. We have investigated the presence of structural RNA elements by conducting a local structural alignment, using FOLDALIGN, on a subset of these 100,000 corresponding regions and estimate that 1800 contain common RNA structures. Comparing our results with the recent mapping of transcribed fragments (transfrags) in human, we find that high-scoring candidates are twice as likely to be found in regions overlapped by transfrags than regions that are not overlapped by transfrags. To verify the coexpression between predicted candidates in human and mouse, we conducted expression studies by RT-PCR and Northern blotting on mouse candidates, which overlap with transfrags on human chromosome 20. RT-PCR results confirmed expression of 32 out of 36 candidates, whereas Northern blots confirmed four out of 12 candidates. Furthermore, many RT-PCR results indicate differential expression in different tissues. Hence, our findings suggest that there are corresponding regions between human and mouse, which contain expressed non-coding RNA sequences not alignable in primary sequence.

The types and prevalence of alternative splice forms

The finding that eukaryotic gene structures are extremely complex prompted the development of new experimental techniques for the accurate measurement of transcription start site usage and of the expression of alternative splice forms. On the computational side, analyses of large databases of splice variants revealed differences in the length, motif composition and selection pressure between constitutive and alternatively spliced exons. Such features are being incorporated into novel computational tools for gene structure prediction. The result of these investigations is a continuously improving catalogue of alternative splice forms. How the expression of these alternative splice forms is regulated remains one of the major open questions.

Evolutionary turnover of mammalian transcription start sites

Alignments of homologous genomic sequences are widely used to identify functional genetic elements and study their evolution. Most studies tacitly equate homology of functional elements with sequence homology. This assumption is violated by the phenomenon of turnover, in which functionally equivalent elements reside at locations that are nonorthologous at the sequence level. Turnover has been demonstrated previously for transcription-factor-binding sites. Here, we show that transcription start sites of equivalent genes do not always reside at equivalent locations in the human and mouse genomes. We also identify two types of partial turnover, illustrating evolutionary pathways that could lead to complete turnover. These findings suggest that the signals encoding transcription start sites are highly flexible and evolvable, and have cautionary implications for the use of sequence-level conservation to detect gene regulatory elements.

The transcriptional regulatory code of eukaryotic cells – insights from genome-wide analysis of chromatin organization and transcription factor binding

Control of eukaryotic gene expression involves combinatorial interactions between transcription factors and regulatory sequences in the genome. In addition, chromatin structure and modification states play key roles in determining the competence of transcription. The term 'transcriptional regulatory code' has been used to describe the interplay of these events in the complex control of transcription. With the maturation of methods for detecting in vivo protein-DNA interactions on a genome-wide scale, detailed maps of chromatin features and transcription factor localization over entire genomes of eukaryotic cells are enriching our understanding of the properties and nature of this transcriptional regulatory code. The rapidly growing number of maps has revealed the dynamic nature of nucleosome composition and chromatin remodeling at regulatory regions and highlighted some unexpected properties of transcriptional regulatory networks in eukaryotic cells.

Differential gene expression in the developing mouse ureter

In many instances, kidney dysgenesis results as a secondary consequence to defects in the development of the ureter. Through the use of mouse genetics a number of genes associated with such malformations have been identified, however, the cause of many other abnormalities remain unknown. In order to identify novel genes involved in ureter development we compared gene expression in embryonic day (E) 12.5, E15.5 and postnatal day (P) 75 ureters using the Compugen mouse long oligo microarrays. A total of 248 genes were dynamically upregulated and 208 downregulated between E12.5 and P75. At E12.5, when the mouse ureter is comprised of a simple cuboidal epithelium surrounded by ureteric mesenchyme, genes previously reported to be expressed in the ureteric mesenchyme, foxC1 and foxC2 were upregulated. By E15.5 the epithelial layer develops into urothelium, impermeable to urine, and smooth muscle develops for the peristaltic movement of urine towards the bladder. The development of these two cell types coincided with the upregulation of UPIIIa, RAB27b and PPARgamma reported to be expressed in the urothelium, and several muscle genes, Acta1, Tnnt2, Myocd, and Tpm2. In situ hybridization identified several novel genes with spatial expression within the smooth muscle, Acta1; ureteric mesenchyme and smooth muscle, Thbs2 and Col5a2; and urothelium, Kcnj8 and Adh1. This study marks the first known report defining global gene expression of the developing mouse ureter and will provide insight into the molecular mechanisms underlying kidney and lower urinary tract malformations.

Profiling Caenorhabditis elegans non-coding RNA expression with a combined microarray

Small non-coding RNAs (ncRNAs) are encoded by genes that function at the RNA level, and several hundred ncRNAs have been identified in various organisms. Here we describe an analysis of the small non-coding transcriptome of Caenorhabditis elegans, microRNAs excepted. As a substantial fraction of the ncRNAs is located in introns of protein-coding genes in C.elegans, we also analysed the relationship between ncRNA and host gene expression. To this end, we designed a combined microarray, which included probes against ncRNA as well as host gene mRNA transcripts. The microarray revealed pronounced differences in expression profiles, even among ncRNAs with housekeeping functions (e.g. snRNAs and snoRNAs), indicating distinct developmental regulation and stage-specific functions of a number of novel transcripts. Analysis of ncRNA–host mRNA relations showed that the expression of intronic ncRNA loci with conserved upstream motifs was not correlated to (and much higher than) expression levels of their host genes. Even promoter-less intronic ncRNA loci, though showing a clear correlation to host gene expression, appeared to have a surprising amount of ‘expressional freedom’, depending on host gene function. Taken together, our microarray analysis presents a more complete and detailed picture of a non-coding transcriptome than hitherto has been presented for any other multicellular organism.

RNA at the steering wheel

Noncoding RNAs transcribed from control elements in DNA provide an anti-silencing mechanism by targeting chromatin-modifying enzymes to these genes.

Expression of the genetic information encoded in our genomes is usually regulated by proteins interacting with the DNA. In some cases, however, noncoding RNAs transcribed from DNA control elements cooperate with histone-modifying enzymes to regulate gene expression, as has recently been shown for noncoding RNA originating from Polycomb- and Trithorax-group response elements.

Long homopurine*homopyrimidine sequences are characteristic of genes expressed in brain and the pseudoautosomal region

Homo(purine*pyrimidine) sequences (R*Y tracts) with mirror repeat symmetries form stable triplexes that block replication and transcription and promote genetic rearrangements. A systematic search was conducted to map the location of the longest R*Y tracts in the human genome in order to assess their potential function(s). The 814 R*Y tracts with > or =250 uninterrupted base pairs were preferentially clustered in the pseudoautosomal region of the sex chromosomes and located in the introns of 228 annotated genes whose protein products were associated with functions at the cell membrane. These genes were highly expressed in the brain and particularly in genes associated with susceptibility to mental disorders, such as schizophrenia. The set of 1957 genes harboring the 2886 R*Y tracts with > or =100 uninterrupted base pairs was additionally enriched in proteins associated with phosphorylation, signal transduction, development and morphogenesis. Comparisons of the > or =250 bp R*Y tracts in the mouse and chimpanzee genomes indicated that these sequences have mutated faster than the surrounding regions and are longer in humans than in chimpanzees. These results support a role for long R*Y tracts in promoting recombination and genome diversity during evolution through destabilization of chromosomal DNA, thereby inducing repair and mutation.

Chicken genomics resource: sequencing and annotation of 35,407 ESTs from single and multiple tissue cDNA libraries and CAP3 assembly of a chicken gene index

Its accessibility, unique evolutionary position, and recently assembled genome sequence have advanced the chicken to the forefront of comparative genomics and developmental biology research as a model organism. Several chicken expressed sequence tag (EST) projects have placed the chicken in 10th place for accrued ESTs among all organisms in GenBank. We have completed the single-pass 5′-end sequencing of 37,557 chicken cDNA clones from several single and multiple tissue cDNA libraries and have entered 35,407 EST sequences into GenBank. Our chicken EST sequences and those found in public databases (on July 1, 2004) provided a total of 517,727 public chicken ESTs and mRNAs. These sequences were used in the CAP3 assembly of a chicken gene index composed of 40,850 contigs and 79,192 unassembled singlets. The CAP3 contigs show a 96.7% match to the chicken genome sequence. The University of Delaware (UD) EST collection (43,928 clones) was assembled into 19,237 nonredundant sequences (13,495 contigs and 5,742 unassembled singlets). The UD collection contains 6,223 unique sequences that are not found in other public EST collections but show a 76% match to the chicken genome sequence. Our chicken contig and singlet sequences were annotated according to the highest BlastX and/or BlastN hits. The UD CAP3 contig assemblies and singlets are searchable by nucleotide sequence or key word ( http://cogburn.dbi.udel.edu ), and the cDNA clones are readily available for distribution from the chick EST website and clone repository ( http://www.chickest.udel.edu ). The present paper describes the construction and normalization of single and multiple tissue chicken cDNA libraries, high-throughput EST sequencing from these libraries, the CAP3 assembly of a chicken gene index from all public ESTs, and the identification of several nonredundant chicken gene sets for production of custom DNA microarrays.

Transcriptional network dynamics in macrophage activation

Transcriptional regulatory networks govern cell differentiation and the cellular response to external stimuli. However, mammalian model systems have not yet been accessible for network analysis. Here, we present a genome-wide network analysis of the transcriptional regulation underlying the mouse macrophage response to bacterial lipopolysaccharide (LPS). Key to uncovering the network structure is our combination of time-series cap analysis of gene expression with in silico prediction of transcription factor binding sites. By integrating microarray and qPCR time-series expression data with a promoter analysis, we find dynamic subnetworks that describe how signaling pathways change dynamically during the progress of the macrophage LPS response, thus defining regulatory modules characteristic of the inflammatory response. In particular, our integrative analysis enabled us to suggest novel roles for the transcription factors ATF-3 and NRF-2 during the inflammatory response. We believe that our system approach presented here is applicable to understanding cellular differentiation in higher eukaryotes.

The callipyge mutation enhances bidirectional long-range DLK1-GTL2 intergenic transcription in cis

The callipyge mutation (CLPG) is an A to G transition that affects a muscle-specific long-range control element located in the middle of the 90-kb DLK1-GTL2 intergenic (IG) region. It causes ectopic expression of a 327-kb cluster of imprinted genes in skeletal muscle, resulting in the callipyge muscular hypertrophy and its non-Mendelian inheritance pattern known as polar overdominance. We herein demonstrate that the CLPG mutation alters the muscular epigenotype of the DLK1-GTL2 IG region in cis, including hypomethylation, acquisition of novel DNase-I hypersentivite sites, and, most strikingly, strongly enhanced bidirectional, long-range IG transcription. The callipyge phenotype thus emerges as a unique model to study the functional significance of IG transcription, which recently has proven to be a widespread, yet elusive, feature of the mammalian genome.

RNA interference is not involved in natural antisense mediated regulation of gene expression in mammals

The study of two examples of endogenous genes with coding or non-coding natural antisense transcript partners provides evidence against the involvement of RNAi in the natural antisense-mediated regulation of mammalian gene expression.

Background

Antisense transcription, yielding both coding and non-coding RNA, is a widespread phenomenon in mammals. The mechanism by which natural antisense transcripts (NAT) may regulate gene expression are largely unknown. The aim of the present study was to explore the mechanism of reciprocal sense-antisense (S-AS) regulation by studying the effects of a coding and non-coding NAT on corresponding gene expression, and to investigate the possible involvement of endogenous RNA interference (RNAi) in S-AS interactions.

Results

We have examined the mechanism of S-AS RNA base pairing, using thymidylate synthase and hypoxia inducible factor-1α as primary examples of endogenous genes with coding and non-coding NAT partners, respectively. Here we provide direct evidence against S-AS RNA duplex formation in the cytoplasm of human cells and subsequent activation of RNAi.

Conclusion

Collectively, our data demonstrate that NAT regulation of gene expression occurs through a pathway independent of Dicer associated RNAi. Moreover, we introduce an experimental strategy with utility for the functional examination of other S-AS pair interactions.

A hidden Markov model approach for determining expression from genomic tiling micro arrays

Background

Genomic tiling micro arrays have great potential for identifying previously undiscovered coding as well as non-coding transcription. To-date, however, analyses of these data have been performed in an ad hoc fashion.

Results

We present a probabilistic procedure, ExpressHMM, that adaptively models tiling data prior to predicting expression on genomic sequence. A hidden Markov model (HMM) is used to model the distributions of tiling array probe scores in expressed and non-expressed regions. The HMM is trained on sets of probes mapped to regions of annotated expression and non-expression. Subsequently, prediction of transcribed fragments is made on tiled genomic sequence. The prediction is accompanied by an expression probability curve for visual inspection of the supporting evidence. We test ExpressHMM on data from the Cheng et al. (2005) tiling array experiments on ten Human chromosomes [1]. Results can be downloaded and viewed from our web site [2].

Conclusion

The value of adaptive modelling of fluorescence scores prior to categorisation into expressed and non-expressed probes is demonstrated. Our results indicate that our adaptive approach is superior to the previous analysis in terms of nucleotide sensitivity and transfrag specificity.

Highly conserved syntenic blocks at the vertebrate Hox loci and conserved regulatory elements within and outside Hox gene clusters

Hox genes in vertebrates are clustered, and the organization of the clusters has been highly conserved during evolution. The conservation of Hox clusters has been attributed to enhancers located within and outside the Hox clusters that are essential for the coordinated "temporal" and "spatial" expression patterns of Hox genes in developing embryos. To identify evolutionarily conserved regulatory elements within and outside the Hox clusters, we obtained contiguous sequences for the conserved syntenic blocks from the seven Hox loci in fugu and carried out a systematic search for conserved noncoding sequences (CNS) in the human, mouse, and fugu Hox loci. Our analysis has uncovered unusually large conserved syntenic blocks at the HoxA and HoxD loci. The conserved syntenic blocks at the human and mouse HoxA and HoxD loci span 5.4 Mb and 4 Mb and contain 21 and 19 genes, respectively. The corresponding regions in fugu are 16- and 12-fold smaller. A large number of CNS was identified within the Hox clusters and outside the Hox clusters spread over large regions. The CNS include previously characterized enhancers and overlap with the 5' global control regions of HoxA and HoxD clusters. Most of the CNS are likely to be control regions involved in the regulation of Hox and other genes in these loci. We propose that the regulatory elements spread across large regions on either side of Hox clusters are a major evolutionary constraint that has maintained the exceptionally long syntenic blocks at the HoxA and HoxD loci.

The challenges of gene expression microarrays for the study of human cancer

Large-scale genomic studies promise to advance our understanding of the biology of human cancers and to improve their diagnosis, prognostication, and treatment. The analysis and interpretation of genomics studies have faced challenges. The retrospective and observational design of many studies has rendered them susceptible to confounding and bias. Technological variations and advances have impacted on reproducibility. Statistical hurdles in relating a large number of variables to a small number of observations have added further constraints. This review considers the promise and challenge associated with the large-scale clinically oriented genomic analysis of human cancer and attempts to emphasize potential solutions.

Highly specific gene silencing by artificial microRNAs in Arabidopsis

Plant microRNAs (miRNAs) affect only a small number of targets with high sequence complementarity, while animal miRNAs usually have hundreds of targets with limited complementarity. We used artificial miRNAs (amiRNAs) to determine whether the narrow action spectrum of natural plant miRNAs reflects only intrinsic properties of the plant miRNA machinery or whether it is also due to past selection against natural miRNAs with broader specificity. amiRNAs were designed to target individual genes or groups of endogenous genes. Like natural miRNAs, they had varying numbers of target mismatches. Previously determined parameters of target selection for natural miRNAs could accurately predict direct targets of amiRNAs. The specificity of amiRNAs, as deduced from genome-wide expression profiling, was as high as that of natural plant miRNAs, supporting the notion that extensive base pairing with targets is required for plant miRNA function. amiRNAs make an effective tool for specific gene silencing in plants, especially when several related, but not identical, target genes need to be downregulated. We demonstrate that amiRNAs are also active when expressed under tissue-specific or inducible promoters, with limited nonautonomous effects. The design principles for amiRNAs have been generalized and integrated into a Web-based tool (http://wmd.weigelworld.org).

Chromosome 2 locus Nidd5 has a potent effect on adiposity in the TSOD mouse

We previously reported a quantitative trait locus for body weight, non-insulin-dependent diabetes 5 (Nidd5), on Chromosome 2 in the TSOD (Tsumura, Suzuki, Obese Diabetes) mouse, a model of polygenic obese type 2 diabetes. To find the gene responsible for a specific component of the pathogenesis, we used a marker-assisted selection protocol to produce congenic strains. These mice are designed to carry a control BALB/cA-derived genomic interval and a TSOD background to look for loss of phenotype. One of the strains with the widest congenic interval, D2Mit297-D2Mit304, showed reductions in both body weight and adiposity compared with TSOD mice. The phenotypic analyses of other congenic strains further narrowed the locus in a 9.4-Mb interval between D2Mit433 and D2Mit91, around which numerous loci for body weight and adiposity have been mapped previously. Although the locus showed a relatively modest effect on body weight, it had a major influence on fat mass that explains approximately 60% of the difference in the adipose index between parental TSOD and BALB/cA mice. Furthermore, the congenic strain with a minimal BALB/cA-derived region showed significantly smaller cell sizes of white and brown adipocytes compared with the control littermates. However, the locus did not primarily affect food consumption, general activity, or rectal temperature after cold exposure, although there are clear differences in these traits between the parental strains. The present work physically delineates the major locus for adiposity in the TSOD mouse.

The Genetic Basis of Type 2 Diabetes

Type 2 Diabetes results from a complex physiologic process that includes the pancreatic beta cells, peripheral glucose uptake in muscle, the secretion of multiple cytokines and hormone-like molecules from adipocytes, hepatic glucose production, and likely the central nervous system. Consistent with the complex web of physiologic defects, the emerging picture of the genetics will involve a large number of risk susceptibility genes, each individually with relatively small effect (odds ratios below 1.2 in most cases). The challenge for the future will include cataloging and confirming the genetic risk factors, and understanding how these risk factors interact with each other and with the known environmental and lifestyle risk factors that increase the propensity to type 2 diabetes.

Distinguishing protein-coding from non-coding RNAs through support vector machines

RIKEN's FANTOM project has revealed many previously unknown coding sequences, as well as an unexpected degree of variation in transcripts resulting from alternative promoter usage and splicing. Ever more transcripts that do not code for proteins have been identified by transcriptome studies, in general. Increasing evidence points to the important cellular roles of such non-coding RNAs (ncRNAs). The distinction of protein-coding RNA transcripts from ncRNA transcripts is therefore an important problem in understanding the transcriptome and carrying out its annotation. Very few in silico methods have specifically addressed this problem. Here, we introduce CONC (for “coding or non-coding”), a novel method based on support vector machines that classifies transcripts according to features they would have if they were coding for proteins. These features include peptide length, amino acid composition, predicted secondary structure content, predicted percentage of exposed residues, compositional entropy, number of homologs from database searches, and alignment entropy. Nucleotide frequencies are also incorporated into the method. Confirmed coding cDNAs for eukaryotic proteins from the Swiss-Prot database constituted the set of true positives, ncRNAs from RNAdb and NONCODE the true negatives. Ten-fold cross-validation suggested that CONC distinguished coding RNAs from ncRNAs at about 97% specificity and 98% sensitivity. Applied to 102,801 mouse cDNAs from the FANTOM3 dataset, our method reliably identified over 14,000 ncRNAs and estimated the total number of ncRNAs to be about 28,000.

There are two types of RNA: messenger RNAs (mRNAs), which are translated into proteins, and non-coding RNAs (ncRNAs), which function as RNA molecules. Besides textbook examples such as tRNAs and rRNAs, non-coding RNAs have been found to carry out very diverse functions, from mRNA splicing and RNA modification to translational regulation. It has been estimated that non-coding RNAs make up the vast majority of transcription output of higher eukaryotes. Discriminating mRNA from ncRNA has become an important biological and computational problem. The authors describe a computational method based on a machine learning algorithm known as a support vector machine (SVM) that classifies transcripts according to features they would have if they were coding for proteins. These features include peptide length, amino acid composition, secondary structure content, and protein alignment information. The method is applied to the dataset from the FANTOM3 large-scale mouse cDNA sequencing project; it identifies over 14,000 ncRNAs in mouse and estimates the total number of ncRNAs in the FANTOM3 data to be about 28,000.

Clusters of internally primed transcripts reveal novel long noncoding RNAs

Non-protein-coding RNAs (ncRNAs) are increasingly being recognized as having important regulatory roles. Although much recent attention has focused on tiny 22- to 25-nucleotide microRNAs, several functional ncRNAs are orders of magnitude larger in size. Examples of such macro ncRNAs include Xist and Air, which in mouse are 18 and 108 kilobases (Kb), respectively. We surveyed the 102,801 FANTOM3 mouse cDNA clones and found that Air and Xist were present not as single, full-length transcripts but as a cluster of multiple, shorter cDNAs, which were unspliced, had little coding potential, and were most likely primed from internal adenine-rich regions within longer parental transcripts. We therefore conducted a genome-wide search for regional clusters of such cDNAs to find novel macro ncRNA candidates. Sixty-six regions were identified, each of which mapped outside known protein-coding loci and which had a mean length of 92 Kb. We detected several known long ncRNAs within these regions, supporting the basic rationale of our approach. In silico analysis showed that many regions had evidence of imprinting and/or antisense transcription. These regions were significantly associated with microRNAs and transcripts from the central nervous system. We selected eight novel regions for experimental validation by northern blot and RT-PCR and found that the majority represent previously unrecognized noncoding transcripts that are at least 10 Kb in size and predominantly localized in the nucleus. Taken together, the data not only identify multiple new ncRNAs but also suggest the existence of many more macro ncRNAs like Xist and Air.

The human genome has been sequenced, and, intriguingly, less than 2% specifies the information for the basic protein building blocks of our bodies. So, what does the other 98% do? It now appears that the mammalian genome also specifies the instructions for many previously undiscovered “non protein-coding RNA” (ncRNA) genes. However, what these ncRNAs do is largely unknown. In recent years, strategies have been designed that have successfully identified hundreds of short ncRNAs—termed microRNAs—many of which have since been shown to act as genetic regulators. Also known to be functionally important are a handful of ncRNAs orders of magnitude larger in size than microRNAs. The availability of complete genome and comprehensive transcript sequences allows for the systematic discovery of more large ncRNAs. The authors developed a computational strategy to screen the mouse genome and identify large ncRNAs. They detected existing large ncRNAs, thus validating their approach, but, more importantly, discovered more than 60 other candidates, some of which were subsequently confirmed experimentally. This work opens the door to a virtually unexplored world of large ncRNAs and beckons future experimental work to define the cellular functions of these molecules.

A method for similarity search of genomic positional expression using CAGE

With the advancement of genome research, it is becoming clear that genes are not distributed on the genome in random order. Clusters of genes distributed at localized genome positions have been reported in several eukaryotes. Various correlations have been observed between the expressions of genes in adjacent or nearby positions along the chromosomes depending on tissue type and developmental stage. Moreover, in several cases, their transcripts, which control epigenetic transcription via processes such as transcriptional interference and genomic imprinting, occur in clusters. It is reasonable that genomic regions that have similar mechanisms show similar expression patterns and that the characteristics of expression in the same genomic regions differ depending on tissue type and developmental stage. In this study, we analyzed gene expression patterns using the cap analysis gene expression (CAGE) method for exploring systematic views of the mouse transcriptome. Counting the number of mapped CAGE tags for fixed-length regions allowed us to determine genomic expression levels. These expression levels were normalized, quantified, and converted into four types of descriptors, allowing the expression patterns along the genome to be represented by character strings. We analyzed them using dynamic programming in the same manner as for sequence analysis. We have developed a novel algorithm that provides a novel view of the genome from the perspective of genomic positional expression. In a similarity search of expression patterns across chromosomes and tissues, we found regions that had clusters of genes that showed expression patterns similar to each other depending on tissue type. Our results suggest the possibility that the regions that have sense–antisense transcription show similar expression patterns between forward and reverse strands.

Through the advancement of genome research, it is becoming clear that genes are not distributed on the genome in random order. Clusters of genes distributed at localized genome positions have been reported in several eukaryotes. Various correlations have been observed between the expressions of genes in adjacent or nearby positions along the chromosomes depending on tissue type and developmental stage. It is reasonable that genomic regions that have similar mechanisms show similar expression patterns. In this study, the authors analyzed gene expression patterns using the computational algorithm of similarity search for exploring systematic views of the mouse transcriptome. They found regions that had clusters of highly expressed genes in certain tissue types whose expression patterns showed strong similarity to each other. This work aims to provide additional insight into genome-wide mechanisms of transcription.

Genome-wide analysis of mammalian promoter architecture and evolution

Mammalian promoters can be separated into two classes, conserved TATA box-enriched promoters, which initiate at a well-defined site, and more plastic, broad and evolvable CpG-rich promoters. We have sequenced tags corresponding to several hundred thousand transcription start sites (TSSs) in the mouse and human genomes, allowing precise analysis of the sequence architecture and evolution of distinct promoter classes. Different tissues and families of genes differentially use distinct types of promoters. Our tagging methods allow quantitative analysis of promoter usage in different tissues and show that differentially regulated alternative TSSs are a common feature in protein-coding genes and commonly generate alternative N termini. Among the TSSs, we identified new start sites associated with the majority of exons and with 3' UTRs. These data permit genome-scale identification of tissue-specific promoters and analysis of the cis-acting elements associated with them.

Pseudo-messenger RNA: phantoms of the transcriptome

The mammalian transcriptome harbours shadowy entities that resist classification and analysis. In analogy with pseudogenes, we define pseudo–messenger RNA to be RNA molecules that resemble protein-coding mRNA, but cannot encode full-length proteins owing to disruptions of the reading frame. Using a rigorous computational pipeline, which rules out sequencing errors, we identify 10,679 pseudo–messenger RNAs (approximately half of which are transposon-associated) among the 102,801 FANTOM3 mouse cDNAs: just over 10% of the FANTOM3 transcriptome. These comprise not only transcribed pseudogenes, but also disrupted splice variants of otherwise protein-coding genes. Some may encode truncated proteins, only a minority of which appear subject to nonsense-mediated decay. The presence of an excess of transcripts whose only disruptions are opal stop codons suggests that there are more selenoproteins than currently estimated. We also describe compensatory frameshifts, where a segment of the gene has changed frame but remains translatable. In summary, we survey a large class of non-standard but potentially functional transcripts that are likely to encode genetic information and effect biological processes in novel ways. Many of these transcripts do not correspond cleanly to any identifiable object in the genome, implying fundamental limits to the goal of annotating all functional elements at the genome sequence level.

Our understanding of genetics has been dominated by the so-called central dogma: the theory that DNA is transcribed into RNA, which is translated via the genetic code to produce proteins. Thus, DNA is the inherited store of genetic information, proteins are the end products that carry out cellular functions, and RNA is a kind of passive intermediary, hence termed messenger RNA. However, evidence has been accumulating that RNA plays a much more dynamic role than this. This study provides an unprejudiced survey of “pathological” RNA molecules, which resemble protein-coding RNA except that they contain violations of the genetic code. These pseudo–messenger RNAs constitute a surprisingly large fraction of all transcripts, as much as 10%. These ghostly molecules have always been present in RNA surveys, but have stayed below the radar because they do not cleanly correspond to annotated elements in DNA, i.e., “genes”. Their prevalence demonstrates that RNA is a distinct continent that cannot be fully understood as a mirror of DNA or proteins.

A systematic analysis of disease-associated variants in the 3' regulatory regions of human protein-coding genes I: general principles and overview

The 3' regulatory regions (3' RRs) of human genes play an important role in regulating mRNA 3' end formation, stability/degradation, nuclear export, subcellular localization and translation and are consequently rich in regulatory elements. Although 3' RRs contain only approximately 0.2% of known disease-associated mutations, this is likely to represent a rather conservative estimate of their actual prevalence. In an attempt to catalogue 3' RR-mediated disease and also to gain a greater understanding of the functional role of regulatory elements within 3' RRs, we have performed a systematic analysis of disease-associated 3' RR variants; 121 3' RR variants in 94 human genes were collated. These included 17 mutations in the upstream core polyadenylation signal sequence (UCPAS), 81 in the upstream sequence (USS) between the translational termination codon and the UCPAS, 6 in the left arm of the 'spacer' sequence (LAS) between the UCPAS and the pre-mRNA cleavage site (CS), 3 in the right arm of the 'spacer' sequence (RAS) or downstream core polyadenylation signal sequence (DCPAS) and 7 in the downstream sequence (DSS) of the 3'-flanking region, with 7 further mutations being treated as isolated examples. All the UCPAS mutations and the rather unusual cases of the DMPK, SCA8, FCMD and GLA mutations exert a significant effect on the mRNA phenotype and are usually associated with monogenic disease. By contrast, most of the remaining variants are polymorphisms that exert a comparatively minor influence on mRNA expression, but which may nevertheless predispose to or otherwise modify complex clinical phenotypes. Considerable efforts have been made to validate/elucidate the mechanisms through which the 3' untranslated region (3' UTR) variants affect gene expression. It is hoped that the integrative approach employed here in the study of naturally occurring variants of actual or potential pathological significance will serve to complement ongoing efforts to identify all functional regulatory elements in the human genome.

SPA: a probabilistic algorithm for spliced alignment

Recent large-scale cDNA sequencing efforts show that elaborate patterns of splice variation are responsible for much of the proteome diversity in higher eukaryotes. To obtain an accurate account of the repertoire of splice variants, and to gain insight into the mechanisms of alternative splicing, it is essential that cDNAs are very accurately mapped to their respective genomes. Currently available algorithms for cDNA-to-genome alignment do not reach the necessary level of accuracy because they use ad hoc scoring models that cannot correctly trade off the likelihoods of various sequencing errors against the probabilities of different gene structures. Here we develop a Bayesian probabilistic approach to cDNA-to-genome alignment. Gene structures are assigned prior probabilities based on the lengths of their introns and exons, and based on the sequences at their splice boundaries. A likelihood model for sequencing errors takes into account the rates at which misincorporation, as well as insertions and deletions of different lengths, occurs during sequencing. The parameters of both the prior and likelihood model can be automatically estimated from a set of cDNAs, thus enabling our method to adapt itself to different organisms and experimental procedures. We implemented our method in a fast cDNA-to-genome alignment program, SPA, and applied it to the FANTOM3 dataset of over 100,000 full-length mouse cDNAs and a dataset of over 20,000 full-length human cDNAs. Comparison with the results of four other mapping programs shows that SPA produces alignments of significantly higher quality. In particular, the quality of the SPA alignments near splice boundaries and SPA's mapping of the 5′ and 3′ ends of the cDNAs are highly improved, allowing for more accurate identification of transcript starts and ends, and accurate identification of subtle splice variations. Finally, our splice boundary analysis on the human dataset suggests the existence of a novel non-canonical splice site that we also find in the mouse dataset. The SPA software package is available at http://www.biozentrum.unibas.ch/personal/nimwegen/cgi-bin/spa.cgi.

A prerequisite for the identification and analysis of splice variation in the transcriptomes of higher eukaryotes is the very accurate mapping of cDNAs to their genomes. However, current algorithms use ad hoc scoring schemes that cannot correctly trade off the likelihoods of different sequencing errors against the likelihoods of different gene structures.

In this paper the authors develop a Bayesian probabilistic approach to cDNA-to-genome mapping that combines explicit models for the prior probabilities of different gene structures with the likelihoods of different sequencing errors. The parameters of these probabilistic models can be estimated automatically from the input such that the mapping procedure is automatically adapted to the organism and sequencing technology of the data under study.

The authors implement their approach in a fast mapping algorithm called SPA and apply it to a dataset of human full-length cDNAs and the FANTOM3 dataset of mouse full-length cDNAs. Comparisons with four other mapping algorithms show that SPA produces mappings that are significantly more accurate, with the largest improvements in the mappings of the 5′ and 3′ ends of the cDNAs, and the mappings around splice boundaries. The authors also identify a novel set of putative splice sites in the human dataset.

A simple physical model predicts small exon length variations

One of the most common splice variations are small exon length variations caused by the use of alternative donor or acceptor splice sites that are in very close proximity on the pre-mRNA. Among these, three-nucleotide variations at so-called NAGNAG tandem acceptor sites have recently attracted considerable attention, and it has been suggested that these variations are regulated and serve to fine-tune protein forms by the addition or removal of a single amino acid. In this paper we first show that in-frame exon length variations are generally overrepresented and that this overrepresentation can be quantitatively explained by the effect of nonsense-mediated decay. Our analysis allows us to estimate that about 50% of frame-shifted coding transcripts are targeted by nonsense-mediated decay. Second, we show that a simple physical model that assumes that the splicing machinery stochastically binds to nearby splice sites in proportion to the affinities of the sites correctly predicts the relative abundances of different small length variations at both boundaries. Finally, using the same simple physical model, we show that for NAGNAG sites, the difference in affinities of the neighboring sites for the splicing machinery accurately predicts whether splicing will occur only at the first site, splicing will occur only at the second site, or three-nucleotide splice variants are likely to occur. Our analysis thus suggests that small exon length variations are the result of stochastic binding of the spliceosome at neighboring splice sites. Small exon length variations occur when there are nearby alternative splice sites that have similar affinity for the splicing machinery.

It has recently become clear that splice variation affects most mammalian genes. It is, however, less clear to what extent these splice variations are functional and regulated by the cell as opposed to simply a result of noise in the splicing process.

One of the most frequently observed forms of splice variation are small variations in exon length in which the boundary of an exon is shifted by small amounts between different transcripts. In this work the authors study the statistics of these splice variations in detail, and the results suggest that these variations are mostly the result of noise in the splicing process. In particular, they propose a simple physical model in which the last step of splicing involves the sequence-specific binding of the splicing machinery to the splice site. In this model, small length variations can occur when there are nearby splice sites with comparable affinity for the splicing machinery. The authors show that this model not only accurately predicts the relative abundances of different splice variations but also predicts which splice sites are likely to undergo small exon length variations.

The abundance of short proteins in the mammalian proteome

Short proteins play key roles in cell signalling and other processes, but their abundance in the mammalian proteome is unknown. Current catalogues of mammalian proteins exhibit an artefactual discontinuity at a length of 100 aa, so that protein abundance peaks just above this length and falls off sharply below it. To clarify the abundance of short proteins, we identify proteins in the FANTOM collection of mouse cDNAs by analysing synonymous and non-synonymous substitutions with the computer program CRITICA. This analysis confirms that there is no real discontinuity at length 100. Roughly 10% of mouse proteins are shorter than 100 aa, although the majority of these are variants of proteins longer than 100 aa. We identify many novel short proteins, including a “dark matter” subset containing ones that lack detectable homology to other known proteins. Translation assays confirm that some of these novel proteins can be translated and localised to the secretory pathway.

Living things work by the actions of proteins and other molecules that are encoded in their genomes. Genome projects aim to construct a “parts list” of life by cataloguing all of these molecules. However, some molecules are harder to identify than others. One difficult category is short proteins. This is because protein-coding nucleotide sequences have statistical features that, for large proteins, are very pronounced and readily distinguishable from non-coding sequence, but for small proteins are not. Thus, to avoid erroneous predictions, many genome projects have employed an artificial length threshold of 100 amino acids. Hence, short proteins are underrepresented in protein catalogues, although they are known to play important roles in immunity, cell signalling, and metabolism. This study clarifies the abundance of short proteins by exploiting two advantageous resources. The first is the large FANTOM collection of mouse transcript sequences: it is much easier to identify proteins in mature transcripts than in raw genome sequence. The second is a method to identify protein-coding sequences by examining how they differ between organisms. The redundancy of the genetic code implies that these differences will follow distinctive patterns, and this approach has the statistical power to break the 100-amino-acid barrier and reliably identify short proteins.

Heterotachy in mammalian promoter evolution

We have surveyed the evolutionary trends of mammalian promoters and upstream sequences, utilising large sets of experimentally supported transcription start sites (TSSs). With 30,969 well-defined TSSs from mouse and 26,341 from human, there are sufficient numbers to draw statistically meaningful conclusions and to consider differences between promoter types. Unlike previous smaller studies, we have considered the effects of insertions, deletions, and transposable elements as well as nucleotide substitutions. The rate of promoter evolution relative to that of control sequences has not been consistent between lineages nor within lineages over time. The most pronounced manifestation of this heterotachy is the increased rate of evolution in primate promoters. This increase is seen across different classes of mutation, including substitutions and micro-indel events. We investigated the relationship between promoter and coding sequence selective constraint and suggest that they are generally uncorrelated. This analysis also identified a small number of mouse promoters associated with the immune response that are under positive selection in rodents. We demonstrate significant differences in divergence between functional promoter categories and identify a category of promoters, not associated with conventional protein-coding genes, that has the highest rates of divergence across mammals. We find that evolutionary rates vary both on a fine scale within mammalian promoters and also between different functional classes of promoters. The discovery of heterotachy in promoter evolution, in particular the accelerated evolution of primate promoters, has important implications for our understanding of human evolution and for strategies to detect primate-specific regulatory elements.

Promoters are crucial to the regulation of gene expression. They are of considerable interest to molecular biologists from a functional perspective and to a much wider audience, as sequence changes within promoters are likely to be a substantial contributor to disease predisposition and the divergence of species. In mammals, promoters have been extensively studied in a case-by-case manner, but the more general mechanisms of promoter evolution are little understood. The authors have undertaken an extensive study of evolutionary trends across experimentally defined promoters. They have discovered that the relative rate of promoter evolution varies between lineages and is substantially accelerated in primates. The authors conclude that the predominant cause is variation in the mutation rate specifically within promoter regions. This finding has important implications for comparative genomics, in particular the identification of functional sites within promoters. The large datasets in this study have also allowed the pattern of evolution to be considered between different types of promoter, giving new insight into their distinct biology.

Non-coding RNA

The term non-coding RNA (ncRNA) is commonly employed for RNA that does not encode a protein, but this does not mean that such RNAs do not contain information nor have function. Although it has been generally assumed that most genetic information is transacted by proteins, recent evidence suggests that the majority of the genomes of mammals and other complex organisms is in fact transcribed into ncRNAs, many of which are alternatively spliced and/or processed into smaller products. These ncRNAs include microRNAs and snoRNAs (many if not most of which remain to be identified), as well as likely other classes of yet-to-be-discovered small regulatory RNAs, and tens of thousands of longer transcripts (including complex patterns of interlacing and overlapping sense and antisense transcripts), most of whose functions are unknown. These RNAs (including those derived from introns) appear to comprise a hidden layer of internal signals that control various levels of gene expression in physiology and development, including chromatin architecture/epigenetic memory, transcription, RNA splicing, editing, translation and turnover. RNA regulatory networks may determine most of our complex characteristics, play a significant role in disease and constitute an unexplored world of genetic variation both within and between species.

Genome-wide computational prediction of transcriptional regulatory modules reveals new insights into human gene expression

The identification of regulatory regions is one of the most important and challenging problems toward the functional annotation of the human genome. In higher eukaryotes, transcription-factor (TF) binding sites are often organized in clusters called cis-regulatory modules (CRM). While the prediction of individual TF-binding sites is a notoriously difficult problem, CRM prediction has proven to be somewhat more reliable. Starting from a set of predicted binding sites for more than 200 TF families documented in Transfac, we describe an algorithm relying on the principle that CRMs generally contain several phylogenetically conserved binding sites for a few different TFs. The method allows the prediction of more than 118,000 CRMs within the human genome. A subset of these is shown to be bound in vivo by TFs using ChIP-chip. Their analysis reveals, among other things, that CRM density varies widely across the genome, with CRM-rich regions often being located near genes encoding transcription factors involved in development. Predicted CRMs show a surprising enrichment near the 3' end of genes and in regions far from genes. We document the tendency for certain TFs to bind modules located in specific regions with respect to their target genes and identify TFs likely to be involved in tissue-specific regulation. The set of predicted CRMs, which is made available as a public database called PReMod (http://genomequebec.mcgill.ca/PReMod), will help analyze regulatory mechanisms in specific biological systems.

Complex Loci in human and mouse genomes

Mammalian genomes harbor a larger than expected number of complex loci, in which multiple genes are coupled by shared transcribed regions in antisense orientation and/or by bidirectional core promoters. To determine the incidence, functional significance, and evolutionary context of mammalian complex loci, we identified and characterized 5,248 cis-antisense pairs, 1,638 bidirectional promoters, and 1,153 chains of multiple cis-antisense and/or bidirectionally promoted pairs from 36,606 mouse transcriptional units (TUs), along with 6,141 cis-antisense pairs, 2,113 bidirectional promoters, and 1,480 chains from 42,887 human TUs. In both human and mouse, 25% of TUs resided in cis-antisense pairs, only 17% of which were conserved between the two organisms, indicating frequent species specificity of antisense gene arrangements. A sampling approach indicated that over 40% of all TUs might actually be in cis-antisense pairs, and that only a minority of these arrangements are likely to be conserved between human and mouse. Bidirectional promoters were characterized by variable transcriptional start sites and an identifiable midpoint at which overall sequence composition changed strand and the direction of transcriptional initiation switched. In microarray data covering a wide range of mouse tissues, genes in cis-antisense and bidirectionally promoted arrangement showed a higher probability of being coordinately expressed than random pairs of genes. In a case study on homeotic loci, we observed extensive transcription of nonconserved sequences on the noncoding strand, implying that the presence rather than the sequence of these transcripts is of functional importance. Complex loci are ubiquitous, host numerous nonconserved gene structures and lineage-specific exonification events, and may have a cis-regulatory impact on the member genes.

Differential use of signal peptides and membrane domains is a common occurrence in the protein output of transcriptional units

Membrane organization describes the orientation of a protein with respect to the membrane and can be determined by the presence, or absence, and organization within the protein sequence of two features: endoplasmic reticulum signal peptides and alpha-helical transmembrane domains. These features allow protein sequences to be classified into one of five membrane organization categories: soluble intracellular proteins, soluble secreted proteins, type I membrane proteins, type II membrane proteins, and multi-spanning membrane proteins. Generation of protein isoforms with variable membrane organizations can change a protein's subcellular localization or association with the membrane. Application of MemO, a membrane organization annotation pipeline, to the FANTOM3 Isoform Protein Sequence mouse protein set revealed that within the 8,032 transcriptional units (TUs) with multiple protein isoforms, 573 had variation in their use of signal peptides, 1,527 had variation in their use of transmembrane domains, and 615 generated protein isoforms from distinct membrane organization classes. The mechanisms underlying these transcript variations were analyzed. While TUs were identified encoding all pairwise combinations of membrane organization categories, the most common was conversion of membrane proteins to soluble proteins. Observed within our high-confidence set were 156 TUs predicted to generate both extracellular soluble and membrane proteins, and 217 TUs generating both intracellular soluble and membrane proteins. The differential use of endoplasmic reticulum signal peptides and transmembrane domains is a common occurrence within the variable protein output of TUs. The generation of protein isoforms that are targeted to multiple subcellular locations represents a major functional consequence of transcript variation within the mouse transcriptome.

Mice and men: their promoter properties

Using the two largest collections of Mus musculus and Homo sapiens transcription start sites (TSSs) determined based on CAGE tags, ditags, full-length cDNAs, and other transcript data, we describe the compositional landscape surrounding TSSs with the aim of gaining better insight into the properties of mammalian promoters. We classified TSSs into four types based on compositional properties of regions immediately surrounding them. These properties highlighted distinctive features in the extended core promoters that helped us delineate boundaries of the transcription initiation domain space for both species. The TSS types were analyzed for associations with initiating dinucleotides, CpG islands, TATA boxes, and an extensive collection of statistically significant cis-elements in mouse and human. We found that different TSS types show preferences for different sets of initiating dinucleotides and cis-elements. Through Gene Ontology and eVOC categories and tissue expression libraries we linked TSS characteristics to expression. Moreover, we show a link of TSS characteristics to very specific genomic organization in an example of immune-response-related genes (GO:0006955). Our results shed light on the global properties of the two transcriptomes not revealed before and therefore provide the framework for better understanding of the transcriptional mechanisms in the two species, as well as a framework for development of new and more efficient promoter- and gene-finding tools.

A high-resolution map of transcription in the yeast genome

There is abundant transcription from eukaryotic genomes unaccounted for by protein coding genes. A high-resolution genome-wide survey of transcription in a well annotated genome will help relate transcriptional complexity to function. By quantifying RNA expression on both strands of the complete genome of Saccharomyces cerevisiae using a high-density oligonucleotide tiling array, this study identifies the boundary, structure, and level of coding and noncoding transcripts. A total of 85% of the genome is expressed in rich media. Apart from expected transcripts, we found operon-like transcripts, transcripts from neighboring genes not separated by intergenic regions, and genes with complex transcriptional architecture where different parts of the same gene are expressed at different levels. We mapped the positions of 3' and 5' UTRs of coding genes and identified hundreds of RNA transcripts distinct from annotated genes. These nonannotated transcripts, on average, have lower sequence conservation and lower rates of deletion phenotype than protein coding genes. Many other transcripts overlap known genes in antisense orientation, and for these pairs global correlations were discovered: UTR lengths correlated with gene function, localization, and requirements for regulation; antisense transcripts overlapped 3' UTRs more than 5' UTRs; UTRs with overlapping antisense tended to be longer; and the presence of antisense associated with gene function. These findings may suggest a regulatory role of antisense transcription in S. cerevisiae. Moreover, the data show that even this well studied genome has transcriptional complexity far beyond current annotation.

A profile of alternative RNA splicing and transcript variation of CACNA1H, a human T-channel gene candidate for idiopathic generalized epilepsies

Highly alternatively spliced genes may provide complex targets for disease mutations. Structural changes created by missense mutations may differentially affect the activity of alternative gene products, whereas missense, silent and non-coding mutations may alter developmental regulation of splice variant expression. CACNA1H is a human gene encoding Ca(v)3.2 low-voltage-activated, T-type calcium channels associated with bursting behavior in neurons and has been linked to more than 30 mutations apparently predisposing to childhood absence epilepsy (CAE) and other idiopathic generalized epilepsies (IGEs). Biophysical properties, including the effects of missense mutations, have been evaluated previously for a single splice form of Ca(v)3.2 expressed in transformed cell lines. We here show that CACNA1H is alternatively spliced at 12-14 sites, capable of generating both functional and non-functional transcripts. Variable cytoplasmic and extracellular protein domains point to likely differences in gating behavior, sensitivity to neuromodulation and interactions with extracellular matrix. Biophysical profiles of selected physiological Ca(v)3.2 forms reveal variations in kinetics and steady-state gating parameters, most likely to affect membrane firing. These were comparable to or larger than changes reported for previously studied mutations. Missense CAE and IGE mutations were clustered near segments associated with anomalous splicing. Missense and silent mutations were found to destroy, create or change the regulatory specificity of predicted exonic splicing enhancer sequences that may control splicing regulation. We discuss a paradigm for CACNA1H expression of Ca(v)3.2 subunits, which may influence future basic and clinical studies.