Tissue-specific transcript annotation and expression profiling with complementary next-generation sequencing technologies

Next-generation sequencing is excellently suited to evaluate the abundance of mRNAs to study gene expression. Here we compare two alternative technologies, cap analysis of gene expression (CAGE) and serial analysis of gene expression (SAGE), for the same RNA samples. Along with quantifying gene expression levels, CAGE can be used to identify tissue-specific transcription start sites, while SAGE monitors 3′-end usage. We used both methods to get more insight into the transcriptional control of myogenesis, studying differential gene expression in differentiated and proliferating C2C12 myoblast cells with statistical evaluation of reproducibility and differential gene expression. Both CAGE and SAGE provided highly reproducible data (Pearson's correlations >0.92 among biological triplicates). With both methods we found around 10 000 genes expressed at levels 2 transcripts per million (0.3 copies per cell), with an overlap of 86%. We identified 4304 and 3846 genes differentially expressed between proliferating and differentiated C2C12 cells by CAGE and SAGE, respectively, with an overlap of 2144. We identified 196 novel regulatory regions with preferential use in proliferating or differentiated cells. Next-generation sequencing of CAGE and SAGE libraries provides consistent expression levels and can enrich current genome annotations with tissue-specific promoters and alternative 3′-UTR usage.

A global view of gene activity and alternative splicing by deep sequencing of the human transcriptome

The functional complexity of the human transcriptome is not yet fully elucidated. We report a high-throughput sequence of the human transcriptome from a human embryonic kidney and a B cell line. We used shotgun sequencing of transcripts to generate randomly distributed reads. Of these, 50% mapped to unique genomic locations, of which 80% corresponded to known exons. We found that 66% of the polyadenylated transcriptome mapped to known genes and 34% to nonannotated genomic regions. On the basis of known transcripts, RNA-Seq can detect 25% more genes than can microarrays. A global survey of messenger RNA splicing events identified 94,241 splice junctions (4096 of which were previously unidentified) and showed that exon skipping is the most prevalent form of alternative splicing.

Stat3 is involved in control of MASP2 gene expression

Little is known about determinants regulating expression of Mannan-binding lectin associated serine protease-2 (MASP-2), the effector component of the lectin pathway of complement activation. Comparative bioinformatic analysis of the MASP2 promoter regions in human, mouse, and rat, revealed conservation of two putative Stat binding sites, termed StatA and StatB. Site directed mutagenesis specific for these sites was performed. Transcription activity was decreased 5-fold when StatB site was mutated in the wildtype reporter gene construct. Gel retardation and competition assays demonstrated that proteins contained in the nuclear extract prepared from HepG2 specifically bound double-stranded StatB oligonucleotides. Supershift analysis revealed Stat3 to be the major specific binding protein. We conclude that Stat3 binding is important for MASP2 promoter activity.

MMP-20 is predominately a tooth-specific enzyme with a deep catalytic pocket that hydrolyzes type V collagen

Matrix metalloproteinase-20 (MMP-20, enamelysin) has a highly restricted pattern of expression. In healthy tissues, MMP-20 is observed in the enamel organ and pulp organ of developing teeth and is present only as an activated enzyme. To identify other tissues that may express MMP-20, we performed a systematic mouse tissue expression screen. Among the non-tooth tissues assayed, MMP-20 transcripts were identified only in minute quantities within the large intestine. The murine Mmp20 promoter was cloned, sequenced, and assessed for potential tooth-specific regulatory elements. In silico analysis identified four promoter modules that were common to Mmp20 and at least two of three coregulated predominantly tooth-specific genes that encode ameloblastin, amelogenin, and enamelin. We asked if the highly restricted MMP-20 expression pattern was associated with a broad substrate specificity that might preclude its expression in other tissues. An iterative mixture-based random doedecamer peptide library screen with Edman sequencing of MMP-20 cleavage products revealed that, among MMPs previously screened, MMP-20 had unique substrate preferences. These preferences indicate that MMP-20 has a deep and wide catalytic pocket that can accommodate substrates with large aromatic residues in the P1' position. On the basis of matrices derived from the peptide library data, we identified and then confirmed that type V collagen is an MMP-20 substrate. Since type V collagen is not present in dental enamel but is an otherwise widely distributed collagen, and since only active MMP-20 has been observed in teeth, our data suggest that control of MMP-20 activity is primarily regulated by transcriptional means.

Comparative promoter analysis allows de novo identification of specialized cell junction-associated proteins

Shared transcription factor binding sites that are conserved in distance and orientation help control the expression of gene products that act together in the same biological context. New bioinformatics approaches allow the rapid characterization of shared promoter structures and can be used to find novel interacting molecules. Here, these principles are demonstrated by using molecules linked to the unique functional unit of the glomerular slit diaphragm. An evolutionarily conserved promoter model was generated by comparative genomics in the proximal promoter regions of the slit diaphragm-associated molecule nephrin. Phylogenetic promoter fingerprints of known elements of the slit diaphragm complex identified the nephrin model in the promoter region of zonula occludens-1 (ZO-1). Genome-wide scans using this promoter model effectively predicted a previously unrecognized slit diaphragm molecule, cadherin-5. Nephrin, ZO-1, and cadherin-5 mRNA showed stringent coexpression across a diverse set of human glomerular diseases. Comparative promoter analysis can identify regulatory pathways at work in tissue homeostasis and disease processes.

MatInspector and beyond: promoter analysis based on transcription factor binding sites

MOTIVATION

Promoter analysis is an essential step on the way to identify regulatory networks. A prerequisite for successful promoter analysis is the prediction of potential transcription factor binding sites (TFBS) with reasonable accuracy. The next steps in promoter analysis can be tackled only with reliable predictions, e.g. finding phylogenetically conserved patterns or identifying higher order combinations of sites in promoters of co-regulated genes.

RESULTS

We present a new version of the program MatInspector that identifies TFBS in nucleotide sequences using a large library of weight matrices. By introducing a matrix family concept, optimized thresholds, and comparative analysis, the enhanced program produces concise results avoiding redundant and false-positive matches. We describe a number of programs based on MatInspector allowing in-depth promoter analysis (DiAlignTF, FrameWorker) and targeted design of regulatory sequences (SequenceShaper).

Linking disease-associated genes to regulatory networks via promoter organization

Pathway- or disease-associated genes may participate in more than one transcriptional co-regulation network. Such gene groups can be readily obtained by literature analysis or by high-throughput techniques such as microarrays or protein-interaction mapping. We developed a strategy that defines regulatory networks by in silico promoter analysis, finding potentially co-regulated subgroups without a priori knowledge. Pairs of transcription factor binding sites conserved in orthologous genes (vertically) as well as in promoter sequences of co-regulated genes (horizontally) were used as seeds for the development of promoter models representing potential co-regulation. This approach was applied to a Maturity Onset Diabetes of the Young (MODY)-associated gene list, which yielded two models connecting functionally interacting genes within MODY-related insulin/glucose signaling pathways. Additional genes functionally connected to our initial gene list were identified by database searches with these promoter models. Thus, data-driven in silico promoter analysis allowed integrating molecular mechanisms with biological functions of the cell.

Regulatory modules shared within gene classes as well as across gene classes can be detected by the same in silico approach

Transcriptional regulation depends on the binding of transcription factors to their corresponding binding sites. The response to cellular signals is often mediated by the cooperative binding of transcription factors to well defined regulatory modules consisting of at least two transcription factor binding sites. Such regulatory modules can be responsible for the common regulation of genes within a gene class or confer a common function to promoters belonging to different gene classes. We developed in silico models representing a common framework of potential regulatory sites specific for one promoter class (actins). We also generated models for two different functional promoter modules both of which confer responsiveness to tumor necrosis factor (TNF) and interferon (IFN) to a variety of promoters. All models exhibited high selectivity, e.g. the mammalian muscle actin promoter model produced no false negatives in a database search.

In silico prediction of scaffold/matrix attachment regions in large genomic sequences

Scaffold/matrix attachment regions (S/MARs) are essential regulatory DNA elements of eukaryotic cells. They are major determinants of locus control of gene expression and can shield gene expression from position effects. Experimental detection of S/MARs requires substantial effort and is not suitable for large-scale screening of genomic sequences. In silico prediction of S/MARs can provide a crucial first selection step to reduce the number of candidates. We used experimentally defined S/MAR sequences as the training set and generated a library of new S/MAR-associated, AT-rich patterns described as weight matrices. A new tool called SMARTest was developed that identifies potential S/MARs by performing a density analysis based on the S/MAR matrix library (http://www.genomatix.de/cgi-bin/smartest_pd/smartest.pl). S/MAR predictions were evaluated by using six genomic sequences from animal and plant for which S/MARs and non-S/MARs were experimentally mapped. SMARTest reached a sensitivity of 38% and a specificity of 68%. In contrast to previous algorithms, the SMARTest approach does not depend on the sequence context and is suitable to analyze long genomic sequences up to the size of whole chromosomes. To demonstrate the feasibility of large-scale S/MAR prediction, we analyzed the recently published chromosome 22 sequence and found 1198 S/MAR candidates.

First pass annotation of promoters on human chromosome 22

The publication of the first almost complete sequence of a human chromosome (chromosome 22) is a major milestone in human genomics. Together with the sequence, an excellent annotation of genes was published which certainly will serve as an information resource for numerous future projects. We noted that the annotation did not cover regulatory regions; in particular, no promoter annotation has been provided. Here we present an analysis of the complete published chromosome 22 sequence for promoters. A recent breakthrough in specific in silico prediction of promoter regions enabled us to attempt large-scale prediction of promoter regions on chromosome 22. Scanning of sequence databases revealed only 20 experimentally verified promoters, of which 10 were correctly predicted by our approach. Nearly 40% of our 465 predicted promoter regions are supported by the currently available gene annotation. Promoter finding also provides a biologically meaningful method for "chromosomal scaffolding", by which long genomic sequences can be divided into segments starting with a gene. As one example, the combination of promoter region prediction with exon/intron structure predictions greatly enhances the specificity of de novo gene finding. The present study demonstrates that it is possible to identify promoters in silico on the chromosomal level with sufficient reliability for experimental planning and indicates that a wealth of information about regulatory regions can be extracted from current large-scale (megabase) sequencing projects. Results are available on-line at http://genomatix.gsf.de/chr22/.

Highly specific localization of promoter regions in large genomic sequences by PromoterInspector: a novel context analysis approach

We present a new algorithm called PromoterInspector to locate eukaryotic polymase II promoter regions in large genomic sequences with a high degree of specificity. PromoterInspector focuses on the genetic context of promoters, rather than their exact location. Application of PromoterInspector can serve as a crucial pre-processing step for other methods to locate exactly, or to analyze promoters. PromoterInspector does not depend on heuristics, because it is purely based on libraries of IUPAC words extracted from training sequences by an unsupervised learning approach. We compared PromoterInspector to in silico promoter prediction tools using the sequences from the review by J.W. Fickett. PromoterInspector compared favourably on Fickett's evaluation scheme. A true positive to false positive ratio of 2.3 was obtained, surpassing the best ratio of 0.6, reported for TSSG. The application of our method to several large genomic sequences of over 1.3 million base-pairs in total resulted in even more specific predictions. The coverage of annotated promoters was comparable to other in silico promoter prediction methods, while the true positive predictions increased by up to 100% of total matches. PromoterInspector scans 100 kb in less than one minute on a workstation, and thus is especially applicable for large genome analysis. The method is available at http://genomatix.gsf. de/cgi-bin/promoterinspector/promoterinspector.pl.

Functional promoter modules can be detected by formal models independent of overall nucleotide sequence similarity

MOTIVATION

Gene regulation often depends on functional modules which feature a detectable internal organization. Overall sequence similarity of these modules is often insufficient for detection by general search methods like FASTA or even Gapped BLAST. However, it is of interest to evaluate whether modules, often known from experimental analysis of single sequences, are present in other regulatory sequences.

RESULTS

We developed a new method (FastM) which combines a search algorithm for individual transcription factor binding sites (MatInspector) with a distance correlation function. FastM allows fast definition of a model of correlated binding sites derived from as little as a single promoter or enhancer. ModelInspector results are suitable for evaluation of the significance of the model. We used FastM to define a model for the experimentally verified NFkappaB/IRF1 regulatory module from the major histocompatibility complex (MHC) class I HLA-B gene promoter. Analysis of a test set of sequences as well as database searches with this model showed excellent correlation of the model with the biological function of the module. These results could not be obtained by searches using FASTA or Gapped BLAST, which are based on sequence similarity. We were also able to demonstrate association of a hypothetical GRE-GRE module with viral sequences based on analysis of several GenBank sections with this module.

AVAILABILITY

The WWW version of FastM is accessible at: http://www.gsf.de/cgi-bin/fastm. pl and http://genomatix.gsf.de/cgi-bin/fastm2/fastm.pl

Functional analysis of the fission yeast Prp4 protein kinase involved in pre-mRNA splicing and isolation of a putative mammalian homologue

The prp4 gene of Schizosaccharomyces pombe encodes a protein kinase. A physiological substrate is not yet known. A mutational analysis of prp4 revealed that the protein consists of a short N-terminal domain, containing several essential motifs, which is followed by the kinase catalytic domain comprising the C-terminus of the protein. Overexpression of N-terminal mutations disturbs mitosis and produces elongated cells, Using a PCR approach, we isolated a putative homologue of Prp4 from human and mouse cells. The mammalian kinase domain is 53% identical to the kinase domain of Prp4. The short N-terminal domains share <20% identical amino acids, but contain conserved motifs. A fusion protein consisting of the N-terminal region from S. pombe followed by the mammalian kinase domain complements a temperature-sensitive prp4 mutation of S. pombe. Prp4 and the recombinant yeast/mouse protein kinase phosphorylate the human SR splicing factor ASF/SF2 in vitro in its RS domain.

A sensitive method for the detection of murine C-type retroviruses

A RT-PCR assay was developed for group-specific detection of murine C-type retroviruses using a nested set of degenerated primers. To distinguish exogenous viruses from related, but silent endogenous viruses, a DNAse I pretreatment of supernatants is applied. This is followed by a heat inactivation/denaturation step. The PCR method is ultrasensitive. which enables the detection of 100 attogram of MoMuLV proviral DNA or up to 1-10 infectious mouse C-type retroviruses in 10 microl supernatant of infected cells. The high specificity of the method allows the differentiation between mouse C-type retroviruses and related retroviruses of the A, B, and D type and C-type retroviruses found in other species. It serves as a valuable tool for the screening of animal cell cultures for contaminations with mouse retroviruses, e.g. hybridomas or recombinant cell lines producing foreign proteins.