Conceived and designed the experiments: M. Tschopp OB SWW SCFN. Performed the experiments: M. Tschopp, M. Takamiya KLC GG. Analyzed the data: M. Tschopp, M. Takamiya KLC GG OB SWW UW SCFN. Contributed reagents/materials/analysis tools: M. Tschopp US SCFN. Wrote the paper: SCFN M. Tschopp.
Current address: Department of Ophthalmology, Inselspital, University of Bern, Bern, Switzerland
The authors have declared that no competing interests exist.
Funduscopy is one of the most commonly used diagnostic tools in the ophthalmic practice, allowing for a ready assessment of pathological changes in the retinal vasculature and the outer retina. This non-invasive technique has so far been rarely used in animal model for ophthalmic diseases, albeit its potential as a screening assay in genetic screens. The zebrafish (
A stereomicroscope with coaxial reflected light illumination was used to obtain fundus color images of the zebrafish. In order to find lens and retinal alterations, a pilot screen of 299 families of the F3 generation of ENU-treated adult zebrafish was carried out.
Images of the fundus of the eye and the anterior segment can be rapidly obtained and be used to identify alterations in genetically modified animals. A number of putative mutants with cataracts, defects in the cornea, eye pigmentation, ocular vessels and retina were identified. This easily implemented method can also be used to obtain fundus images from rodent retinas.
In summary, we present funduscopy as a valuable tool to analyse ocular abnormalities in adult zebrafish and other small animal models. A proof of principle screen identified a number of putative mutants, making funduscopy based screens in zebrafish feasible.
For centuries people wondered what lies behind the pupil. Unaided vision only reveals a black hole while the fundus of the eye itself remains invisible. Only in the 19th century did Helmholtz rationalize that by trying to look into someone else's eye one blocks the light needed to illuminate the eye cavity. Helmholtz overcame this problem by using a transparent mirror made of three thin parallel sheets of glass angled to reflect the maximum light in the eye
Although funduscopy as a non-invasive tool is well established in the clinics, its use to survey ocular alterations in animal model organisms is still not routine. This is particularly true for the zebrafish, a genetic model for vertebrate vision that is becoming increasingly popular
To the best of our knowledge, no method to routinely image the fundus of the zebrafish eye has been described in the literature. Instruments designed for humans cannot be used due to the small size of the zebrafish eyes (with a pupil size of around 1 mm
We report here that a stereomicroscope with “coaxial reflected light” or near-vertical illumination can be used for funduscopy in fish (and other small animals). The described setup is inexpensive and well suited to screen adult zebrafish for retinal alterations. In a pilot screen we demonstrate that anterior segment and fundus alterations can be found in living adult zebrafish carrying novel (induced) genetic mutations.
Adult wild type zebrafish (
Of the tested stereomicroscopes, the Olympus SZ-61 equipped with a “coaxial reflected light illuminator” (Olympus, SZ2-ILLC) was found to provide the best images and was therefore used in this study. The “coaxial reflected light illuminator” consists of a half reflecting mirror, a polarizer and a 1/4λ plate. The setup is shown in
(A) Photograph of the setup; (*) indicates the “coaxial reflected light illuminator” and (#) indicates the ball-and-socket stage. (B) Schematic drawing of the light path.
For funduscopy, fish were anaesthetized in a solution of 0.014% ethyl-m-aminobenzoate metanesulphonate (MESAB; Sigma) in housing water, taken out of the solution and placed on a wet paper towel. A ball-and-socket stage facilitated the rotation of the fish, which is necessary to see different parts of the retina. To reduce the refraction power of the fish eye, a cover slip or a small concave lens (R = 2.51 mm, diameter 3.5 mm¸ 77.004.41-520, FISBA Optik, St.Gallen, Switzerland,) was placed on the eye. 3% methylcellulose or lubricant eye gel (Viscotears, Novartis) was used to bridge the gap between the cornea and the cover slide, or small lens respectively. Color images of the retina were taken with the camera of the stereomicroscope (Olympus ColorView III or Leica DFC 300FX). As the camera takes square pictures and the magnification of the stereomicroscope is not large enough to fill the whole picture with the image of the fundus, interesting parts of the picture were clipped out using Adobe Photoshop. Magnification has been estimated by magnification of individual lenses and calibrated using known blood vessel thickness.
Mouse funduscopy was essentially the same with the exception that the animal was anesthetized by intraperitoneal injection of Hypnorm (30 µl/20 g body weight; Janssen Pharmaceutics, Beerse, Belgium) and Dormicum (60 µl/20 g body weight; Roche Pharmaceuticals, Basel, Switzerland) diluted with 210 µl Aqua ad inj. (Fresenius Kabi GmbH, Bad Homburg, Germany). To enlarge the pupil of the mouse, 1% atropine was applied on the eye. Again, a cover slide helped to improve image quality.
To find lens and retinal alterations, 299 families (8 fish per family, one eye per fish) of the F3 generation of ENU-treated fish were examined. These fish, aged one year and three months, were part of the European Community ZF-models project (see
The enucleated eyes of paraformaldehyde-fixed fish (4% paraformaldehyde in 0.2 M phosphate buffer, pH 7.4, for about one week at 4°C) were dehydrated in a graded series of ethanol-water mixtures, then incubated twice in 1∶1 ethanol and basic solution (Technovit 7100; Heraeus Kulzer, Hanau, Germany) for 1 hour. After overnight infiltration in basic solution, larvae were positioned in polymerization medium (Technovit 7100; Heraeus Kulzer) overnight at room temperature.
Microtome sections (3 µm) were prepared and mounted on slides (Menzel-Gläser, Braunschweig, Germany), air dried at 60°C, stained with modified Richardson solution (0.25% methylene blue, 0.25% borax, and 0.5% azure II in ddH2O), overlaid with rapid mounting medium (Entellan; Merck, Darmstadt, Germany), and cover-slipped.
After testing different illuminations, we found “coaxial reflected light” illumination gave the highest quality images of the zebrafish eye. Near-vertical illumination of operating microscopes also gave good results, although the magnification of the tested operating microscopes was rather small. Coaxial illumination (not reflected) is similar to near vertical illumination and should therefore also be suitable for funduscopy. However, the tested binoculars with coaxial illumination gave dark and dull pictures, likely due to the strong polarizers that are usually used with coaxial illumination to reduce reflection.
Both cover slips and small concave lenses can be used to reduce the refraction power of the fish eye by directly applying them on the cornea. The field of view obtained when using concave lenses is slightly larger compared to that obtained with cover slides. Although a large field of view is favourable when screening, it is more convenient to use cover slides since they are easier to handle and disposable.
The cornea and lens can be examined with standard illumination (other than coaxial or near vertical illumination). However, since light from standard illumination does not travel to the posterior part of the eye chamber, it does not allow viewing of the entire lens (in zebrafish, the lens extends almost to the retina). With coaxial or near-vertical illumination, the entire lens can be examined. Hence our funduscopy setup can be employed for biomicroscopy of the anterior segment. This may help to discriminate different forms of cataract and to find alterations in the posterior part of the lens.
In adult zebrafish (about 15 months old) of the ZF-models screening project, we found cataracts in 52 of 299 screened families (on average in 2.5 of 8 fish screened per family). Cataracts can be grouped according to the affected structure. Although the morphogenesis of the lens in zebrafish shows important differences to that in mammals, the overall morphology of the adult zebrafish lens is similar to that of other vertebrates
(A) Overview of anaesthetized fish with cover glass, (B) shows a normal, transparent lens, (C–F) examples of opaque lenses (cataracts): (C) small inclusion are present in an otherwise normal lens, (D) massive cataract, (E) including cracks in the lens, and (F) membranous cataract. Scale bar: 500 µm.
Our setup allows the examination of the retina, retinal blood vessels, and the optic disc in great detail. Due to the high magnification, it is not possible to have all layers of the retina simultaneously in focus. Focusing on the inner part of the retina, also images retinal blood vessels (
(A–D) wild type zebrafish fundus at the level of the mid-peripheral retina (A), optic disc (B and C) and photoreceptor level, showing the presumed photoreceptor mosaic (D). Hypopigmented fundus of the zebrafish
Hypopigmentation of the eye is readily apparent, as illustrated in the zebrafish
The assessment of altered retinal blood vessels is important, since blood vessel alterations are associated with many retinal and systemic diseases including diabetes mellitus and hypertension
Additionally, we found retinal alterations in three families with features analogous to mammalian retinal degenerations. One family (9 of 32 fish) showed tortuous arteries and a darkened, somehow bumpy or uneven retina (see
(A) Wild type zebrafish. (B) Right eye and (C) left eye of a fish with black spots in the retina; the funduscopy of this fish is shown in
In the diseased human retina, the most noticeable retinal changes besides blood vessel alterations and bone-spicule like bodies (e.g. in retinitis pigmentosa), are drusen (e.g. in age related macular degeneration). Drusen are conspicuous extracellular accumulations at the level of Bruch's membrane. In our screened fish, we did not find any evidence for drusen.
The described setup should work for all animals that can be placed below a stereomicroscope. As a cursory proof of principle, we tested the setup on one wild type mouse (
Funduscopy is one of the most important tools in ophthalmology, and will be of growing importance in the study of animal models of ocular diseases. It already proved its usefulness in mice research where it is routinely applied. For instance, this technique was applied to isolate strains with spontaneous
In this paper we describe a simple and cheap method to obtain fundus images of adult zebrafish. The zebrafish model system is increasingly popular to study ophthalmic diseases
We applied the principle of direct ophthalmoscopy, based on the use of the subject's lens as magnification device. Since the refractive power of aquatic lenses is higher than those of terrestrial animals, we obtain a much larger magnification than in the human eye. Therefore the magnification is easily sufficient to image blood vessels and even large enough to image the presumed photoreceptor mosaic of the retina (
To demonstrate the feasibility of zebrafish funduscopy for the isolation of adult fish strains with morphological eye alterations, we screened 299 lines of F3 adult zebrafish descended from chemically mutagenized founder fish. For each fish line, we screened at least 8 fishes and only phenotypes that were seen at least twice were considered to be potential mutants. Since this is a pilot proof-of-principle screen, we did not prove the heritability of these traits. However all described phenotpypes appear at least twice per family, and are therefore in the range of expected Mendelian ratio for recessive mutations. In few cases we found half of the assayed fish to be affected, suggesting dominant inheritance.
A total of 59 potential mutants were found, with the great majority (n = 52) displaying lens defects. This finding is in line with clinical data, considering the high incidence of cataracts in elderly humans (almost 50% in persons aged 75 years and more)
Four families displayed aberrations of the retinal vasculature in that the arteries appear abnormally twisted (tortuous). In humans, many disease classes may produce tortuosity, including high blood flow, angiogenesis and blood vessel congestion
Three families displayed retinal alterations. Some of these alterations resemble the bone-spicule-like degeneration seen in retinitis pigmentosa, both in funduscopy and in subsequent histological sections (
In summary, we have developed a simple and fast method to obtain fundus images of adult living zebrafish. This is the first report of funduscopy in zebrafish and its use in isolating adult ocular mutants. This procedure is simple and fast enough to be useful for forward genetics screens, as we have demonstrated in a pilot genetic screen, which identified several interesting fish lines with ocular alterations, affecting lens, blood vessels and outer retina. Since we have not demonstrated the heritable nature of these alterations, we can only assume them to be genetic mutations due to their approximate Mendelian ratio. The successful isolation of such fish in a screen based on funduscopy demonstrates that such screens are possible and are expected to lead to the isolation of interesting mutant lines, relevant as disease models of age related eye diseases.
We would like to thank Thomas Labhart for fruitful discussions and Peter Froesch and Maryam Rastegar for technical assistance.