FishNet: an online database of zebrafish anatomy
© Bryson-Richardson et al; licensee BioMed Central Ltd. 2007
Received: 29 March 2007
Accepted: 17 August 2007
Published: 17 August 2007
Over the last two decades, zebrafish have been established as a genetically versatile model system for investigating many different aspects of vertebrate developmental biology. With the credentials of zebrafish as a developmental model now well recognized, the emerging new opportunity is the wider application of zebrafish biology to aspects of human disease modelling. This rapidly increasing use of zebrafish as a model for human disease has necessarily generated interest in the anatomy of later developmental phases such as the larval, juvenile, and adult stages, during which many of the key aspects of organ morphogenesis and maturation take place. Anatomical resources and references that encompass these stages are non-existent in zebrafish and there is therefore an urgent need to understand how different organ systems and anatomical structures develop throughout the life of the fish.
To overcome this deficit we have utilized the technique of optical projection tomography to produce three-dimensional (3D) models of larval fish. In order to view and display these models we have created FishNet http://www.fishnet.org.au, an interactive reference of zebrafish anatomy spanning the range of zebrafish development from 24 h until adulthood.
FishNet contains more than 36 000 images of larval zebrafish, with more than 1 500 of these being annotated. The 3D models can be manipulated on screen or virtually sectioned. This resource represents the first complete embryo to adult atlas for any species in 3D.
Zebrafish possess a number of attributes that have facilitated their uptake as a developmental model system. Zebrafish uniquely combine embryological manipulability, optical clarity of the early embryo and larvae (allowing simple visualization of cell biological events directly in vivo) and the ability to apply invertebrate-style forward genetics to questions of vertebrate development. More recently, research has extended into later aspects of zebrafish development and adulthood, examining aspects of organogenesis and tissue maintenance. Many of the same strengths that made zebrafish a superior model for the study of development also complement those of existing mammalian disease models. Thus a huge variety of human conditions are now being modeled in zebrafish, ranging from drug and alcohol addiction to cancer [1–8].
The usefulness of any model organism is limited by the available accurate anatomical information for that system. At present, there is a lack of a detailed anatomical reference for zebrafish. While the earliest stages of zebrafish development have been intensively studied over the last two decades, and an embryological staging series described , comparatively little is known about the development of larval, juvenile and adult anatomy. Whilst the vascular anatomy of the embryonic fish , the anatomy of the embryonic and adult brain [11, 12], and heart  have been elegantly described, there is a need for a detailed record of zebrafish development that spans the entire life history of the organism and presents its anatomy.
Traditionally, our understanding of anatomy has relied on the interpretation of two-dimensional (2D) images extrapolated into three-dimensions (3D). Data recorded in this manner are difficult to reinterpret, with the original anatomical representation conveying only one aspect of the data. Such data are also unable to be further interrogated in the context of the sample as a whole. If by contrast, data is captured for the entire sample in 3D it can be viewed, digitally sectioned, and interrogated in any way the researcher desires. A holistic understanding of the anatomy of an organism is critical to dissecting the development and function of different organs and tissues in space and time. We have therefore generated a database of zebrafish anatomy charting development at multiple time points in 3D.
While a number of different methods exist for the capture and generation of 3D images; such as the reconstruction of serial physical sections [14–17] from serial optical sections using confocal microscopy , or from magnetic resonance imaging (MRI); these methods all possess specific limitations that prevent their application to a 3D atlas of zebrafish development. MRI has not been adapted for use in the zebrafish due to the small size of its embryo and the current lack of resolution of this technique. Furthermore, the specialized equipment required for the application of MRI imaging has prevented its uptake for routine laboratory imaging. In the zebrafish, the 3D analysis undertaken so far has largely depended upon confocal imaging  or to a much lesser extent serial sectioning. However, neither of these techniques is currently suitable for the generation of a 3D atlas of zebrafish development. Conventional confocal microscopy is limited by the depth of tissue that can be examined, restricting its use to the earliest stages of embryonic development. While deeper penetration can be achieved using multi-photon imaging, this procedure is still insufficient to access all stages of zebrafish development, and all forms of confocal imaging rely exclusively on the use of fluorescent probes. In addition, 3D models created by the alternative method of serial sectioning are highly time consuming to generate and physical distortion of the sample can often occur during sectioning. Models from serial sections often have high quality images in the plane of section, however difficulties in aligning the serial sections can lead to poor quality images when viewed in other planes.
We have therefore decided to adapt the method of optical projection tomography (OPT; [19, 20]) to create a digital full life atlas of zebrafish development. OPT uses projection images of the sample to reconstruct a 3D model of the sample (demonstrated in Additional file 1), a process that overcomes many of the shortcomings of the other methods listed above. OPT can be used to create 3D models over a wide range of sample sizes, from 1–10 mm, covering the range of zebrafish development from the early embryo until adulthood. OPT can utilize both fluorescent or conventional brightfield labeling techniques, although prior to this publication its application has been limited to the generation of grayscale 3D models. As the OPT technique is non destructive there is no distortion of the sample similar to that induced by serial sectioning and consequently a smooth 3D model can be created, resulting in high quality images in all planes, not just that of the plane of section. Furthermore, the use of a single sample to generate all of the sections in any plane allows the user to examine the same structure at multiple angles, giving a clear representation of the 3D organization of the sample.
We have utilized OPT to create the first 3D lifespan atlas for any species, in the process generating 18 3D models spanning the entire period from embryo to adulthood. In order to make this data available to the research community we have created a specifically designed web interface  to allow the data to be viewed and manipulated. This resource contains more than 36 000 digital section images and the complete 3D models can be downloaded to allow interrogation in any plane the researcher desires. To further enhance the usefulness of this resource we have annotated more than 1 500 of the sections throughout the 3D models we have created with apparent anatomical structures. As such, FacishNet represents the first complete embryo to adult atlas in 3D for any species.
Results and Discussion
Generating 3D models of zebrafish development
We collected samples representing the entire development of the zebrafish from early embryo until adulthood. Samples were collected at 24, 48, 72, 96, and 120 h post-fertilization (hpf). After the 120 hpf time point embryos were collected at specific total lengths (jaw to tip of tail), as length was more representative of developmental stage than age in later development, at each 1 mm stage from 5–17 mm. Using the OPT method we have developed for zebrafish  we generated a series of 3D models charting the development of the zebrafish, Danio rerio, from embryo to adult.
Analysis and 3D display
FishNet: an interface for the database
The user can change the magnification of the viewed image to browse either the entire section or to closely examine a selected region. The image can also be dragged on screen to centre on a region of interest. Once a particular section of interest has been identified, the user can also open this in a separate new window while continuing to browse the database, facilitating comparison between stages and views.
To allow users of the website to view the 3D renderings of the models in an Internet browser we have generated rendered images of the models viewed at 162 different angles. These pre-computed renderings are then used to create a QuickTime virtual rendering (QTVR) for each model. These QTVRs can be viewed in a browser window and the sample moved in 3D using the mouse or cursor keys. As the renderings are pre-computed, once loaded, the QTVR can be manipulated instantly and the user can rotate the model freely. To ensure access to all users, we have also included smaller versions of the QTVR with 18 pre-computed views compared to 162 in the standard versions; these smaller versions will be more accessible to those with limited Internet connections.
An anatomical reference for the zebrafish
A major limitation of the current OPT system is that it is only able to capture grayscale images. Although fluorescent OPT models can be pseudo-colored following reconstruction, to represent the color of the fluorescent probe, transmission or brightfield OPT has so far been limited to producing grayscale models. This has prevented its use with many of the commonly used histological stains that rely on color differences to provide information about the intensity and levels of a signal and the use of multiple color stains together.
A holistic understanding of the anatomy of an organism is critical to dissect the development and function of different organs and tissues. Information about the spatio-temporal patterning of an anatomical structure can be invaluable in understanding its development and effect on surrounding structures. The most complete mechanism for generating this understanding is to represent development in four dimensions. Using OPT we have created an online database recording development in 3D at a complete range of stages from early embryo to adulthood.
Given the greater clarity with which anatomical information can be relayed in 3D it is clear that the future for data representation in whole organismal context is 3D. The relative ease of data capture by OPT, the establishment of rendering and display tools to display this captured data, suggests that OPT might become one of a standard set of tools for the analysis of zebrafish anatomy, gene, and protein expression. Recently, further improvements to OPT have been described that improve the resolution obtained through use of a frequency space filter to remove out of focus information prior to reconstruction . As these improvements and others are incorporated into OPT reconstructions, the quality and resolution of the images will continue to improve. The ability to create 3D models of color samples using cOPT greatly increases the number and combination of stains that can be analyzed by this method, and also allows the results to be presented in a more recognizable way to the user.
In order for the data to be of use it is vital that it is presented in an accessible and intuitive manner. Establishing user friendly Internet-based tools to allow the virtual sectioning of the 3D models and the manipulation of 3D models ensures the data is available to all without the need for specialist equipment or software. The establishment of this reference for zebrafish development and anatomy will facilitate research using this model system, in particular research into later stages of development and adulthood, which are of special interest for the modeling of human disease. The annotation of a large and representative fraction of the database makes FishNet a valuable resource not only for zebrafish researchers but also to those in other disciplines who wish to further examine results in fish or undertake comparative analyses. FishNet will also be of great use as an educational resource.
Our development of FishNet is the first step towards developing a fully integrated repository for 3D data for anatomy and development in the zebrafish. We plan to continue to build this resource adding models following organogenesis through the labeling of individual organs at multiple time points and hope that in the future FishNet will serve as a repository for 3D gene expression information. As more data are added to FishNet we will also continue to develop tools for the interrogation of this data set to make use of this 3D resource.
Zebrafish were obtained from wild-type or golden mutant lines. Samples were fixed in 4% paraformaldehyde overnight at 4°C, and then washed with PBS containing 0.1% Tween 20. Samples were used immediately or progressed through a methanol series for storage in 100% methanol at -20°C. Pigmented embryos were bleached in a 5% formamide, 10% H2O2, 0.5 × SSC solution, following dehydration and re-hydration in a methanol series. Samples were embedded in a 1.5% low melting point agarose gel and attached to a metal mount. The mounted samples were then cleared through a series of 25%. 50%, 75%, 2 × 100% methanol prior to immersion in a 2:1 benzyl alcohol:benyzl benzoate solution.
Optical projection tomography
The projection images were captured using a Retiga 1300Exi ccd camera (QImaging; 19535 56th Avenue, Suite 101, Surrey, BC, Canada. V3S 6K3) mounted on a Leica MZFLIII stereo-fluorescent microscope with HBO 100 fluorescent lamp source and a GFP1 filter for auto-fluorescent capture, and a LCD RGB filter (QImaging) for color samples. The samples were mounted in the 'Edinburgh OPT Scanner', and a series of 400 images captured over a 360° rotation. Images were captured at a 1360 × 1024 pixel resolution. A detailed description of the equipment required and method for preparation of samples and subsequent 3D reconstruction is given in Bryson-Richardson and Currie .
The 3D models were converted to .slc format files and rendered using the visualization toolkit either for direct display on screen, or to produce a series of rendered images of the rotating sample. These were then subsequently used to create QuickTime virtual renderings (QTVR rendering) using The VR Worx (VR Toolbox Inc; P.O. Box 111419, Pittsburgh, PA 15238 USA).
Color OPT reconstruction
Each color channel was reconstructed individually. The complete 3D models were then examined to see the range of grey values they contained. All three images were then converted to 8-bit using the same limits as the channel with the greatest range of values. Section images and volume renderings were created for each model individually and merged using ImageMagick (on-line resource ) convert to produce a color image.
We would like to thank the Edinburgh Mouse Atlas team, especially Richard Baldock and Bill Hill, for their support and assistance, and Paul Perry for advice on imaging. PDC was supported by the Welcome Trust and Human Frontiers Science Program, RJBR by the Welcome Trust, SB the NHMRC, TFS the National Institute of Health, TEH the Muscular Dystrophy Association,
- Amsterdam A, Sadler KC, Lai K, Farrington S, Bronson RT, Lees JA, Hopkins N: Many ribosomal protein genes are cancer genes in zebrafish. PLoS Biol. 2004, 2 (5): E139-10.1371/journal.pbio.0020139.PubMed CentralView ArticlePubMedGoogle Scholar
- Berghmans S, Murphey RD, Wienholds E, Neuberg D, Kutok JL, Fletcher CD, Morris JP, Liu TX, Schulte-Merker S, Kanki JP, Plasterk R, Zon LI, Look AT: tp53 mutant zebrafish develop malignant peripheral nerve sheath tumors. Proc Natl Acad Sci U S A. 2005, 102 (2): 407-412. 10.1073/pnas.0406252102.PubMed CentralView ArticlePubMedGoogle Scholar
- Carvan MJ, Loucks E, Weber DN, Williams FE: Ethanol effects on the developing zebrafish: neurobehavior and skeletal morphogenesis. Neurotoxicol Teratol. 2004, 26 (6): 757-768. 10.1016/j.ntt.2004.06.016.View ArticlePubMedGoogle Scholar
- Darland T, Dowling JE: Behavioral screening for cocaine sensitivity in mutagenized zebrafish. Proc Natl Acad Sci U S A. 2001, 98 (20): 11691-11696. 10.1073/pnas.191380698.PubMed CentralView ArticlePubMedGoogle Scholar
- Langenau DM, Traver D, Ferrando AA, Kutok JL, Aster JC, Kanki JP, Lin S, Prochownik E, Trede NS, Zon LI, Look AT: Myc-induced T cell leukemia in transgenic zebrafish. Science. 2003, 299 (5608): 887-890. 10.1126/science.1080280.View ArticlePubMedGoogle Scholar
- Lockwood B, Bjerke S, Kobayashi K, Guo S: Acute effects of alcohol on larval zebrafish: a genetic system for large-scale screening. Pharmacol Biochem Behav. 2004, 77 (3): 647-654. 10.1016/j.pbb.2004.01.003.View ArticlePubMedGoogle Scholar
- Patton EE, Widlund HR, Kutok JL, Kopani KR, Amatruda JF, Murphey RD, Berghmans S, Mayhall EA, Traver D, Fletcher CD, Aster JC, Granter SR, Look AT, Lee C, Fisher DE, Zon LI: BRAF mutations are sufficient to promote nevi formation and cooperate with p53 in the genesis of melanoma. Curr Biol. 2005, 15 (3): 249-254. 10.1016/j.cub.2005.01.031.View ArticlePubMedGoogle Scholar
- Shepard JL, Amatruda JF, Stern HM, Subramanian A, Finkelstein D, Ziai J, Finley KR, Pfaff KL, Hersey C, Zhou Y, Barut B, Freedman M, Lee C, Spitsbergen J, Neuberg D, Weber G, Golub TR, Glickman JN, Kutok JL, Aster JC, Zon LI: A zebrafish bmyb mutation causes genome instability and increased cancer susceptibility. Proc Natl Acad Sci U S A. 2005, 102 (37): 13194-13199. 10.1073/pnas.0506583102.PubMed CentralView ArticlePubMedGoogle Scholar
- Kimmel CB, Ballard WW, Kimmel SR, Ullmann B, Schilling TF: Stages of embryonic development of the zebrafish. Dev Dyn. 1995, 203 (3): 253-310.View ArticlePubMedGoogle Scholar
- Isogai S, Horiguchi M, Weinstein BM: The vascular anatomy of the developing zebrafish: an atlas of embryonic and early larval development. Dev Biol. 2001, 230 (2): 278-301. 10.1006/dbio.2000.9995.View ArticlePubMedGoogle Scholar
- Mueller T, Wullimann MF: Atlas of Early Zebrafish Brain Development : A Tool for Molecular Neurogenetics. 2005, Amsterdam , Elsevier B.V,, 1Google Scholar
- Wullimann MF, Rupp B, Reichert H: Neuroanatomy of the Zebrafish Brain: a Topological Atlas. 1996, Basel , Birkhauser Verlag, 1View ArticleGoogle Scholar
- Hu N, Sedmera D, Yost HJ, Clark EB: Structure and function of the developing zebrafish heart. Anat Rec. 2000, 260 (2): 148-157. 10.1002/1097-0185(20001001)260:2<148::AID-AR50>3.0.CO;2-X.View ArticlePubMedGoogle Scholar
- Brune RM, Bard JB, Dubreuil C, Guest E, Hill W, Kaufman M, Stark M, Davidson D, Baldock RA: A three-dimensional model of the mouse at embryonic day 9. Dev Biol. 1999, 216 (2): 457-468. 10.1006/dbio.1999.9500.View ArticlePubMedGoogle Scholar
- Kaufman MH, Brune RM, Davidson DR, Baldock RA: Computer-generated three-dimensional reconstructions of serially sectioned mouse embryos. J Anat. 1998, 193 ( Pt 3): 323-336. 10.1046/j.1469-7580.1998.19330323.x.View ArticleGoogle Scholar
- Streicher J, Donat MA, Strauss B, Sporle R, Schughart K, Muller GB: Computer-based three-dimensional visualization of developmental gene expression. Nat Genet. 2000, 25 (2): 147-152. 10.1038/75989.View ArticlePubMedGoogle Scholar
- Weninger WJ, Mohun T: Phenotyping transgenic embryos: a rapid 3-D screening method based on episcopic fluorescence image capturing. Nat Genet. 2002, 30 (1): 59-65. 10.1038/ng785.View ArticlePubMedGoogle Scholar
- Cooper MS, D'Amico LA, Henry CA: Confocal microscopic analysis of morphogenetic movements. Methods Cell Biol. 1999, 59: 179-204.View ArticlePubMedGoogle Scholar
- Bryson-Richardson RJ, Currie PD: Optical projection tomography for spatio-temporal analysis in the zebrafish. Methods Cell Biol. 2004, 76: 37-50.View ArticlePubMedGoogle Scholar
- Sharpe J, Ahlgren U, Perry P, Hill B, Ross A, Hecksher-Sorensen J, Baldock R, Davidson D: Optical projection tomography as a tool for 3D microscopy and gene expression studies. Science. 2002, 296 (5567): 541-545. 10.1126/science.1068206.View ArticlePubMedGoogle Scholar
- Walls JR, Sled JG, Sharpe J, Henkelman RM: Resolution improvement in emission optical projection tomography. Phys Med Biol. 2007, 52 (10): 2775-2790. 10.1088/0031-9155/52/10/010.View ArticlePubMedGoogle Scholar
- ImageMagick. [http://www.imagemagick.org]
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.