GENEPAINT
GENE
PAINT
G E N E P A I N T
G E N E P A I N T

Sorry

Genepaint currently does not support this display size. Please use a device with at least 768px width.

Advanced Search
  • Results 0
  • Selection 0
  • Viewer 0
    • Contact
    • Help
    • Admin
    • Login
    • Logout

GenePaint

Home of High Resolution Gene Expression Data

About Genepaint

GenePaint is a digital atlas of gene expression patterns in various tissues and species with strong focus on mouse embryos. Expression patterns are determined by non-radioactive in situ hybridization on serial tissue sections.

The database is gene-centric and entries can be searched either by gene name, accession number, sequence homology or site of expression. The website features a "virtual microscope" that enables zooming into images down to cellular resolution.

We acknowledge the contribution of the EURExpress consortium.

TUTORIALS

Search and Results

How to search and use the results page


Selection

Manage and share multiple genepaint sets


Viewer

How to use the Viewer


Advanced Search

Search genes by site of expression or sequence


E 14.5 ATLAS

Embryo Maps

To assist in the initial identification of sites of gene expression, maps of sagittal sections at embryonic day 14.5 are available in an interactive anatomy atlas. The E14.5 C57BL/6J embryo used in the atlas was prepared, sectioned and imaged identically to the embryos used for in situ hybridization (see Methods of Data Production). Section thickness is 25µm and inter-section distance is 150µm. Tissue was stained with Thioninacetate (Nissl-method). All sections were digitally scanned using a 5x objective. Structures annotated for gene expression are indicated in the maps with red pointers. Boundaries between brain regions are indicated with dashed yellow lines. Both, the in situ hybridization section after a performed search and the appropriate atlas section can be viewed side-by-side.

SHOWCASE

Col18a

This collagen is one of the multiplexins, extracellular matrix proteins that contain multiple triple-helix domains (collagenous domains) interrupted by non-collagenous domains.

Lef1

This gene encodes a transcription factor belonging to a family of proteins that share homology with the high mobility group protein-1.

Sox9

The transcription factor Sox9 acts during chondrocyte differentiation and is a pivotal factor in male sexual development.

Reln

This gene encodes a large secreted extracellular matrix protein thought to control cell-cell interactions critical for cell positioning and neuronal migration during brain development.

Cck

This gene encodes a member of the gastrin/cholecystokinin family of proteins.

Ccnd2

The protein encoded by this gene belongs to the highly conserved cyclin family, whose members are characterized by a dramatic periodicity in protein abundance through the cell cycle.

ABOUT

Methods Yaylaoglu M.B. et al. (2005); Eichele,G. and Diez-Roux, G. (2011)

Templates and Probes

Probes were designed and synthesized from freely available cDNA clone resources. For genes not represented in a public collection, templates were produced by PCR from cDNA using gene-specific primers. Templates and hence riboprobes were ∼1000 nt long. Transcript expression was detected with non-radioactive riboprobes tagged with digoxygenin by in vitro-transcription from DNA templates using appropriate primers and RNA polymerases.

Tissues

Mice were maintained on a 12-h light/dark cycle with light being turned on at 07:00. Animals were anesthetized with isoflurane (Abbott Laboratories) and either killed by decapitation (for brain preparations) or through cervical dislocation (for embryo collection).

Embryos

were briefly washed in cold phosphate-buffered saline, blotted dry with a filter paper and transferred into ice-cold OCT 4583 (Tissue-Tek, Sakura). Tissues were transferred to a freezing chamber (Figure 1) within 5 min after dissection. The custom-made freezing chamber consists of a square copper base and transparent Plexiglass side walls. The chamber was placed on a specially designed, movable stage fitted to a stereomicroscope. Through translation and rotation of this stage, the edges of the chamber were aligned parallel to a grid placed into one of the eyepieces of the stereomicroscope. The chamber was filled with OCT at room temperature and the tissue was submerged in OCT and oriented appropriately with the aid of a blunt dissecting needle. Orientation of the embryo was correct if the sagittal mid-plane of the embryo was parallel to a side wall of the chamber as judged by viewing the specimen frontally through the transparent side walls. In addition, the dorsal midline of the embryo had to be parallel to the gridlines of the eyepiece. In the case of brain the sagittal mid-plane was oriented parallel to the side walls. After orientation, usually within 1 min, the chamber containing the specimen is placed on the aluminum block immersed in ethanol contained in a cooler that was kept in a -80° C freezer. Alternatively, the freezing chamber was placed on a block of dry ice. Once the OCT was frozen up to the level of the specimen, the chamber was put onto the freezing shelf of a cryostat for a few hours. Thereafter, tissue axis and specimen number was written on the block, the chamber was disassembled and the blocks were placed into plastic bags and stored at –80°C for several months and even years.

Figure 1:

Right side shows an assembled, custom-made freezing chamber consisting of a copper base plate, four plexiglas side walls, a top plate and two springs to fasten the top plate to the vertical rods positioned at the four corners. Left side shows the parts. The 4 side walls are equipped with two glued in, protruding steel pegs and have a slanted lower edge. The latter is positioned into groves in the copper base (5× 5 cm) and the two steel pegs are inserted into holes drilled through the top plate. An arrow engraved in the top plate indicates a specimen axis. A cross is engraved in the copper base. The purpose of this cross is to align the chamber to a grid located in the optical path of a dissecting microscope. This alignment is performed prior to adding OCT and specimen. Before removing the chamber to add OCT and specimen, mark its position with flat corner iron brace fastened to the microscope base. Once filled with OCT and specimen, place the chamber back into its original position onto the microscope base and use the grid of the microscope for specimen orientation in the horizontal plane. Since the side walls of the chamber are translucent, the orientation in the two other planes can readily be achieved by visual inspection. A blunt needle should be used to manipulate the specimen. Once a satisfactory orientation has been achieved, place the chamber on top of a cold metal block.

Histology

At least one day prior to sectioning, OCT blocks were transferred to a –20°C freezer and then mounted on a cryostat chuck. Because of the orthogonal walls of the O.C.T. block, specimens can be accurately oriented. The chuck is fastened to the goniometer of the cryostat and oriented so that the edge of the knife cuts parallel to one of the cardinal planes of the specimen. A Leica cryostat (Model CM 3050S) is used and sections are 20µm or 25µm thick. Superfrost slides are placed into aluminum slide holders which permits precise positioning of sections onto the slides.

Following sectioning, slides are kept in slide racks in the cryostat until ready for fixation. Fixation, carried out in a Leica Autostainer XL, is in 4% paraformaldehyde for 20 min at room temperature. Subsequently, slides are washed for 5 min in PBS, acetylated and dried by passing them through an ethanol series ending with 100 % ethanol. After drying at 30°C, slides are stored at –80°C in boxes containing a drying agent and sealed with electrical tape.

Figure 2:

Customized slide holder made of aluminum alloy loaded with a slide with 4 frozen sections from a mouse embryo. A spring made of polyoxymethylene (right) keeps the slide in place by. The windows indicate the fields into which sections are placed which facilitates subsequent automated microscopy.

In situ Hybridization, Stringency Washes and Probe Detection

To increase sensitivity of in situ hybridization, tyramide-biotin amplification was used (also known as catalyzed reporter deposition; Speel, E.J. [1999]. Histochem Cell. Biol. 112, 89-113). Following hybridization of cellular mRNA with digoxigenin-11-dUTP-tagged RNA, peroxidase-coupled anti-digoxygenin antibody was added. The peroxidase moiety of the antibody activates a biotin-tyramide conjugate that covalent attaches biotin to proteins in the vicinity of the riboprobe. Biotin was subsequently detected by an alkaline phosphatase-based BCIP/NPT color reaction resulting in a precipitate covering the cell expressing the gene in question.

The slides carrying tissue sections were integrated into flow-through chamber (Figure 3), 48 of which are accommodated in a thermo rack located on a Tecan Genesis™ (or Freedom Evo™) liquid handling platform (Figure 4). Depending on the platform size, up to 4 thermo racks can be used. Turnaround time for a single run with 4 racks is ~24 h.

Figure 3:

Frontal view of an assembled hybridization chamber. Note, there is a slanted depression milled into the top portion of the 5mm thick glass plate, which, together with the slide and the spacer, forms a reservoir. Solutions pipetted into this reservoir, enter the narrow hybridization chamber by gravity, and flow partially through the chamber, but are also retained by capillary forces. If a new solution is added to the well, it displaces the previous one. In- and out-flows are uniform and flow rates are low, avoiding tissue shearing by liquid flow. All components of the hybridization chamber are made of heat- and solvent resistant materials and with the exception of the spacers can be reused.The five components of the flow-through hybridization chambers of Tecan’s GenePaint™ system are: the slide (shown with 4 stained embryo sections) metal holder, a spacer, the glass plate and two springs that hold this assembly together. The protruding ends of the spacer are trimmed with scissors.

Microscopy

Slides are coverslipped with a water-based medium and photographed in a bright field or DIC light microscope equipped with a motorized stage that moves the slide in front of the objective. The equipment used consists of a Leica DM-RXA2 microscope, a motorized Märzhäuser stage that accommodates up to eight slides, a Leica electronic focusing system, a Hitachi HV-C20A CCD color camera and a PC based custom-made controller that drives and coordinates stage and camera (Figure 5). Embryo sections are too large to be photographed as a whole. Therefore, the motorized stage moves the sections in a stepwise mosaic fashion in front of the objective and at each step an image is taken. Subsequently images are stitched together, cropped, saved as TIFF file and uploaded on the GenePaint database. Most images were captured with a 10 x objective in which case the numerical aperture of the objective lens was 0.4 and the pixel size was 1.6 µm/px. At an early phase of the GenePaint project, images were collected with a 5 x objective (numerical aperture 0.15) and 3.2 µm/px.

Figure 4:

A thermos rack (center) accommodates 48 hybridization chambers. Slides are in direct contact with the temperature-controlled surface of the rack. The temperature is maintained by a car radiator/water mixture circulating between the thermos racks and a heating/cooling thermostat (e.g. Lauda Integral T2200). 8 pipettes (gree, right) take up solutions from reservoirs situated on the platform and sequentially deliver solutions (kept at room temperature) into the top well of the hybridization chambers. Probes are added using the platform pipettes. Riboprobes and costly reagents can also be added using a manual single/repeater pipettes (e.g. Distriman, Gilson). The front row of the thermo rack shown is loaded with eight flow-through chambers. Solutions flow through the chambers and exit at the base, where they are collected in a tray (white) located underneath the rack and pumped into a waste container. Stringency wash solutions are preheated to 65°C in a hot water circulating thermos rack (right, equipped with hoses connecting to a circulating heating bath).

Figure 5:

Microscope used to scan slides after in situ hybridization. Note the aluminum frame that accommodates 8 slides. Because sections are positioned at stereotypic locations (Figure 2) regions of interest are well defined and hence scanning time is minimized. The frame can be horizontally adjusted with set screws so that across an entire embryo section the plane of focus is the same. This permits that the focal plane is determined only once per section in the center of the section and the same focal plane is used for all ~200 images that encompass a single section. This way a high throughput can be achieved.

Annotation of Expression

Expression patterns revealed by in situ hybridization were annotated by experts and volunteers. Emphasis was placed on those genes that exhibited regional expression. Expression strength and expression pattern for each anatomical structure were scored (see below) and entered into hierarchically organized collapsible and extendable tree of organs/tissues (Figure 6, top). For easier recognition an icon for strength and pattern was generated ant placed at the left margin (Figure 6, bottom). Annotation enables to query GenePaint.org for gene expression in particular anatomical regions.

Figure 6:

illustrates the different levels and patterns of expression the dorsal part of the spinal cord. At E14.5 this tissue appears quite homogeneous with regard to cell size and anatomical subdivisions. At a molecular level, however, one observes regionalization . Five different strengths of expression are defined:

  • weak expression (level 1), the colorimetric reaction generates a few small precipitates
  • medium expression (level 2) colored precipitate covers a sizable fraction of the tissue
  • strong expression (level 3), color precipitate completely covers the tissue
  • no color precipitate is detected (level N/D, not detected)
  • the anatomical structure has not been examined/annotated yet (level n/e, not examined)

The pattern of expression across the tissue is comes in three different types:

  • ubiquitous expression (u) pattern, all cells within the tissue express the gene
  • regional expression (r), signal is restricted to certain subregions of the tissue
  • scattered expression (s), signal is restricted to a subpopulation of cells, i.e. a “salt and pepper” pattern is observed.

Ubiquitous

Scattered

Regional

Weak

Medium

Strong

Figure 7:

Examples of expression patterns in the spinal cord. Although ISH data are not quantitative, the use of riboprobes of similar lengths and the automated ISH procedures allow an across the transcriptome comparison. An assessment of expression strength is facilitated if tissues that are “negative” are available as is the case for the examples shown. If, however, expression is characterized by a high background, a reliable assessment of the strength of expression is not possible. Therefore, such specimens were not annotated. Genepaint set ID to access to the full data set of the above examples are: Cmas (ES540), Scd2 (EH2751), Ctnna2 (EH3386), Serpinf1 (MH1163), Tgfbr1 (MH935), Fzd7 (MH680), Car2 (MH1443), Top2a (MH502), and Col4a (HD18). Of note, several of the expression patterns shown were determined by two different probes. In all cases, the main features are identical between the sets but there can be marked differences with regard to background. Thus Col4a (HD18) based on a specific primer-set derived probe has much less background than Col4a (EH4187), whose riboprobe was derived from a public resource cDNA. Abbreviations: expression strength: 1,2,3; expression pattern: u (ubiquitous), r, regional; s: scattered

Common Artifacts

Cartilage may occasionally display weak non-specific signal (Figure 8A). The myocardium of the embryonic heart can exhibit splotch-like round stains of uniform intensity that can readily be distinguished from the granular precipitate characteristic for true signal (Figure 8B). Single sections within a data set may exhibit streaks or smears across the whole specimen, usually resulting either from sectioning problems or from uneven exposure to reagents during in-situ hybridization (Figure 8C, D). These artifacts are easily identified because the stained regions do not agree with anatomical boundaries and are absent on adjacent sections.

Other GenePaint data

StageStrainType of SpecimenSections per SetPosition of Section Shown
E10.5NMRIwhole embryo1-5representative section
E14.5NMRIwhole embryo24100-175µm constant spacing
E15.5NMRIhead10selected standard planes (see atlas)
P7C57BL/6brain11selected standard planes (see atlas)
P56C57BL/6brain10selected standard planes (see atlas)

Figure 8:

Various artefacts. (A) Small spots of purple stain are located above somites of the tail and are distinct from the genuine Pax6 staining in the spinal cord. (B) Large spots in the myocardium are artefacts seen in a section hybridized with Tachykinine receptor 2 probe. (C) The narrow stripe is caused by a tissue scratch in this Shh in situ hybridization experiment. (D) A broad smear is seen in this Gm623 in situ hybridization experiment. Shh and Gm623 both show specific signal in the sections depicted in (C) and (D), respectively.

Literature

  1. Visel, A. et al. (2004). GenePaint.org: an atlas of gene expression patterns in the mouse embryo. Nucl. Acid Res. 32 D552-556.
  2. Yaylaoglu M.B. et al. (2005). Comprehensive expression atlas of fibroblast growth factors and their receptors generated by a novel robotic in situ hybridization platform. Dev Dyn.234:371-86.
  3. Diez-Roux, G. et al. (2011). A High-Resolution Anatomical Atlas of the Transcriptome in the Mouse EmbryoPLOS Biol 9:e1000582.
  4. Eichele, G. and Diez-Roux, G. (2011). High-throughput analysis of gene expression on tissue sections by in situ hybridizationMethods 53, 417-23.

RELATED DATABASES

METscout

This mouse metabolism database was designed to navigate complex metabolic maps with a keen eye on the gene expression landscapes of enzymes and transporters.

http://www.metscout.mpg.de/

EURExpress II

Funded by the EU this project was the first to achieve the transcriptome-wide acquisition of gene expression patterns by ISH in the mouse embryo.

http://www.eurexpress.org/ee/

EuReGene

The goal of this EU project was to discover genes responsible for renal development and disease. The original ISH-Atlas generated for EuReGene is now part of GUDMAP .

Allen Mouse Brain Atlas

A molecular map showing where all genes are expressed in all regions of the mouse brain.
http://brain-map.org .

Looking for the
old genepaint?

For developers:
API-Docs

Found any problems?
Known issues

Set of the day:
Twitter