TRACER: a resource to study the regulatory architecture of the mouse genome

Abstract

Background: Mammalian genes are regulated through the action of multiple regulatory elements, often distributed across large regions. The mechanisms that control the integration of these diverse inputs into specific gene expression patterns are still poorly understood. New approaches enabling the dissection of these mechanisms in vivo are needed.

Results: Here, we describe TRACER (http://tracerdatabase.embl.de), a resource that centralizes information from a large on-going functional exploration of the mouse genome with different transposon-associated regulatory sensors. Hundreds of insertions have been mapped to specific genomic positions, and their corresponding regulatory potential has been documented by analysis of the expression of the reporter sensor gene in mouse embryos. The data can be easily accessed and provides information on the regulatory activities present in a large number of genomic regions, notably in gene-poor intervals that have been associated with human diseases.

Conclusions: TRACER data enables comparisons with the expression pattern of neighbouring genes, activity of surrounding regulatory elements or with other genomic features, revealing the underlying regulatory architecture of these loci. TRACER mouse lines can also be requested for in vivo transposition and chromosomal engineering, to analyse further regions of interest.

Figure 1

Genome organisation and TRACER. (A) Schematic representation of a genomic locus, with the different elements that control gene expression, and the specialized databases dedicated to their description (see Websites). The TRACER database displays information from a regulatory sensor that detects the regulatory influences active at its insertion site, outlining the characteristics and extent of genomic regulatory landscapes. (B) The GROMIT strategy. A transposon that contains the regulatory sensor can be remobilised in vivo from its start site by a transposase transgene (not shown) active only in the male germline [2]. The genomic position of the new insertions and the associated expression pattern defined by LacZ staining in mouse embryos is displayed in the TRACER database.

Figure 2

The different transposons used in TRACER. Within the left and right inverted/direct repeats of the SB transposon (black double arrows) different cargoes have been cloned that can be utilised for various purposes. Most lines contain an insertion of either the SB9 or SB8 transposon, comprising a LacZ reporter gene with a SV40 polyA sequence (blue rectangle), driven by a synthetic promoter composed of 50 bp of the human beta-globin promoter (black box with β), and a loxP site (red triangle). Newer transposons with additional features have been constructed, and mice with these transposons are being produced. Additional modules comprise sites for the PhiC31 integrase (attB), and the I-Sce1 meganuclease, which open possibilities to use the transposon as a docking site for incoming cassettes [46-48]. TetO binding sites can be used to recruit fusion proteins [49]. SBIL contains an insulator/enhancer-blocker element (Ins, orange block, from the chicken HSS4 element [50]) flanked by loxP and lox2272 (white triangle with red contour). Cre-mediated recombination can transform this transposon into different derivatives (SBL or SB2lox), depending on which lox sites are used.

Figure 3

Searching the TRACER database. (A) TRACER lines can be retrieved (“General search”) either based the genomic position of the insertion as defined by a genomic range or a region around a particular gene, or using the result of the expression assays, with the possibility to specify stage and expression domains of interest. (B) Alternatively (“Visual search”), the distribution of insertion sites along a chromosome is visualised, and those within the range defined by a dragable and expandable red rectangle will be returned.

Figure 4

Presentation of TRACER data. (A) Summary display of results for a TRACER search. For each line the TRACER name, genomic position, orientation of the loxP site, summary of expression (expressed/positive, not expressed/negative, not done; developmental stage(s) assayed), status (alive, cryopreserved, not maintained, newly created) and the transposon type are displayed. The final two columns allow users to select the data for download, or to register interest in the line. Clicking on the “open” icon gives access to the detailed information of an insertion of interest. Quick access to a thumbnail photo of a representative embryo and to the corresponding transposon is possible by a simple click on the corresponding zone (B) Detailed view of a TRACER mouse line. The top section reveals basic information about the insertion, including the type of insertion, the name of the parent insertion, the orientation of the loxP site and a visual representation of the transposon in the correct orientation. The first panel shows the genomic context of the insertion in a snapshot from the Ensembl genome browser (Ensembl Genomic view +/−0.5 Mb) along with links to view the insertion point in the Ensembl or UCSC genome browsers. The second panel shows photos and annotations of the expression patterns of the regulatory sensor. The images can be mouse-overed to reveal a high-resolution zoomed-in view, and the annotation of expression domains is displayed. The third panel shows information related to mapping and genotyping, including the sequence obtained from the mapping procedure, and the sequences of the primers used to genotype animals carrying this specific insertion. The interface in the panel on the left allows neighbouring insertions to be selected for detailed analysis of a region’s regulatory potential.

Figure 5

Evolution of TRACER dataset. (A) Genomic distribution of the insertions with expression annotations (red) and available as lines (blue) (generated with the Karyoview tool from Ensembl – status in July 2012). Insertions on the Y chromosome are not shown as their positions are ambiguous (B) The number of mapped insertions increases at a regular pace (light blue line), almost paralleled by expression reports (red cursor). A smaller proportion of these are kept alive or archived by cryopreservation (blue cursor). (C) The user wish list allows researchers to register interest for a genomic interval. It ensures that mice with new insertions in this region will be preserved for the scientist who indicated this request. The interval of interest can be defined by position or by gene name.

Figure 6

Examples of potential applications of TRACER. (A) Using TRACER to obtain information about the regulatory potential of a specific interval. Two loci (Ndst3 and Zfhx3) are given as an example, with photos of lacZ-stained embryos with insertions close or far from these genes. (B) Searching TRACER for insertions with specific expression patterns can identify genomic positions where an inserted transgene will display a predicted activity. Three insertions with different patterns of expression in the fronto-nasal process are shown as examples. Such specific patterns can subsequently be exploited to re-target other transgenes with expression in the predicted domains through different techniques. (C) Overlapping deletions (red bars) and duplications (blue bars) over 14q12 entered in the DECIPHER database (http://decipher.sanger.ac.uk/) are associated with similar clinical phenotypes. TRACER insertions in the orthologous interval show that it corresponds to a large regulatory landscape associated with forebrain (green) and facial (brown) expression domains in E11.5 (SB-202250e1 and SB-202250) and E12.5 (SB-177175emb7) mouse embryos. LacZ expression in the ear of embryo SB-177175emb7 (star) is due to the presence of an additional insertion, SBlacC [2]. Enhancer regions (ovals, color-coded depending on their pattern of activity as shown in the schematized embryo; data from VISTA enhancer browser [28]) present in the proximity may contribute to these activities. By local remobilisation and Cre-mediated recombineering, the mouse line SB-202250 may be used to produce mouse models of the human aneuploidies associated with this largely non-coding interval.