The article explains how to surgically remove eyes from living zebrafish larvae as the first step toward investigating how retinal input influences optic tectum growth and development. In addition, the article provides information about larval anesthetization, fixation, and brain dissection, followed by immunohistochemistry and confocal imaging.
Zebrafish exhibit remarkable life-long growth and regenerative abilities. For example, specialized stem cell niches established during embryogenesis support continuous growth of the entire visual system, both in the eye and the brain. Coordinated growth between the retinae and the optic tectum ensures accurate retinotopic mapping as new neurons are added in the eyes and brain. To address whether retinal axons provide crucial information for regulating tectal stem and progenitor cell behaviors such as survival, proliferation, and/or differentiation, it is necessary to be able to compare innervated and denervated tectal lobes within the same animal and across animals.
Surgical removal of one eye from living larval zebrafish followed by observation of the optic tectum achieves this goal. The accompanying video demonstrates how to anesthetize larvae, electrolytically sharpen tungsten needles, and use them to remove one eye. It next shows how to dissect brains from fixed zebrafish larvae. Finally, the video provides an overview of the protocol for immunohistochemistry and a demonstration of how to mount stained embryos in low-melting-point agarose for microscopy.
The goal of this method is to investigate how retinal input influences the growth and development of the optic tectum, the visual processing center in the zebrafish brain. By removing one eye and then comparing the two sides of the optic tectum, tectal changes within the same specimen can be observed and normalized, enabling comparison across multiple specimens. Modern molecular approaches combined with this technique will yield insights into the mechanisms underlying visual system growth and development, as well as axonal degeneration and regeneration.
Sensory systems – visual, auditory, and somatosensory – gather information from external organs and relay that information to the central nervous system, generating "maps" of the external world across the midbrain1,2. Vision is the dominant sensory modality for nearly all vertebrates, including many fishes. The retina, the neural tissue in the eye, gathers information with a neuronal circuit consisting primarily of photoreceptors, bipolar cells, and retinal ganglion cells (RGCs), the projection neurons of the retina. RGCs have long axons that find their way across the inner surface of the retina to the optic nerve head, where they fasciculate and travel together through the brain, ultimately terminating in the visual processing center in the dorsal midbrain. This structure is called the optic tectum in fish and other non-mammalian vertebrates and is homologous to the superior colliculus in mammals3.
The optic tectum is a bilaterally symmetric multilayered structure in the dorsal midbrain. In zebrafish and most other fishes, each lobe of the optic tectum receives visual input solely from the contralateral eye, such that the left optic nerve terminates in the right tectal lobe and the right optic nerve terminates in the left tectal lobe4 (Figure 1). Like its mammalian counterpart, the superior colliculus, the optic tectum integrates visual information with other sensory inputs, including audition and somatosensation, controlling shifts in visual attention and eye movements such as saccades1,5,6. However, unlike the mammalian superior colliculus, the optic tectum continuously generates new neurons and glia from a specialized stem cell niche near the medial and caudal edges of the tectal lobes called the tectal proliferation zone7. Maintenance of proliferative progenitors in the optic tectum and other regions of the central nervous system contributes, in part, to the remarkable regenerative capacity documented in zebrafish8.
Previous work examining the brains of blind or one-eyed fishes revealed that optic tectum size is directly proportional to the amount of retinal innervation it receives9,10,11. In adult cave fish, whose eyes degenerate in early embryogenesis, the optic tectum is noticeably smaller than that of closely related, sighted surface fish9. Cave fish eye degeneration can be blocked by replacing the endogenous lens with a lens from a surface fish during embryogenesis. When these one-eyed cave fish are reared to adulthood, the innervated tectal lobe contains approximately 10% more cells than the non-innervated tectal lobe9. Similarly, in larval killifish that were incubated with chemical treatments to generate eyes of different sizes within the same individual, the side of the tectum with more innervation was larger and contained more neurons10. Evidence from optic nerve crush experiments in adult goldfish indicates that innervation promotes proliferation, with tectal cell proliferation decreasing when innervation was disrupted11.
Confirming and extending these classical studies, several recent reports provide data suggesting that proliferation in response to innervation is modulated, at least in part, by the BDNF-TrkB pathway12,13. Many open questions about optic tectum growth and development remain, including how a developing sensory system copes with injury and axon degeneration, which cellular and molecular signals enable retinal input to regulate optic tectum growth, when these mechanisms become active, and whether innervation-linked proliferation and differentiation enable the retina and its target tissue to coordinate growth rates and ensure accurate retinotopic mapping. In addition, there are much larger questions about activity-dependent development that can be addressed by interrogating the zebrafish visual system with surgical approaches such as the one described below.
To investigate the cellular and molecular mechanisms by which neural activity, specifically from visual input, alters cell survival and proliferation, the described approach directly compares innervated and denervated tectal lobes (Figure 1) within individual zebrafish larvae. This method allows for the documentation of RGC axon degeneration in the optic tectum and confirmation that the number of mitotic cells correlates with innervation.
Figure 1: Sketches of zebrafish larvae before and after unilateral eye removal. (A) Drawing of 5 dpf larvae as viewed under a dissecting microscope. Each larva is embedded in low-melting-point agarose and oriented laterally before a tungsten needle with a sharp, hooked tip is used to scoop out the eye facing up (left eye in this example). (B) Drawing of the dorsal view of a 9 dpf larva resulting from the surgery depicted in A. Only three highly schematized RGC axons from the right eye are shown defasciculating and connecting with neurons in the left tectal lobe. Abbreviations: dpf = days post fertilization; dps = days post surgery; RGC = retinal ganglion cells. Please click here to view a larger version of this figure.
The methods in this paper were conducted in accordance with guidelines and approval of the Institutional Animal Care and Use Committees of Reed College and University College London. See the Table of Materials for details about zebrafish strains used in this study.
1. Prepare materials and tools
2. Embryo collection and rearing
3. Prepare larvae for surgery
4. Eye-removal
5. Postoperative care until experimental endpoint
6. Fixing larvae
7. Dissecting larvae to reveal brains (adapted from 15)
8. Brain dehydration and storage
9. Immunohistochemistry
NOTE: Established protocols for many useful wholemount techniques in zebrafish can be found on ZFIN20. This manuscript provides examples comparing one-eyed and two-eyed larvae that were immunostained with antibodies that recognize either the red fluorescent protein (RFP), which is expressed in optic nerve axons in the Tg[atoh7:RFP] line (Figure 2), or phosphorylated histone H3 (PH3), which highlights mitotic cells (Figure 3). A standard immunohistochemistry protocol for wholemount embryos and larvae is summarized below.
10. Mounting and imaging
To confirm whether eye removal was complete and assess how the optic tectum changes, surgeries were performed in the Tg[atoh7:RFP] strain, which labels all RGCs with a membrane-targeted RFP and, thus, all axons that project from the retina and form the optic nerve24. Although using this strain is not absolutely necessary, it enables straightforward observation and visualization of the optic nerve termini in the optic tectum neuropil. Other approaches for labeling the optic nerve, such as axon tracing with lipophilic DiI or DiO dyes25,26 or multispectral labeling with brainbow27, would also be possible.
As shown in Figure 2, progressive degeneration of retinal axons in the optic tectum neuropil was observed following eye removal. By two days post surgery, RFP-labeled axons exhibited hallmarks of rapid Wallerian degeneration28 such as blebbing and fragmentation (Figure 2A,B). Recent work has shown that microglia engulf more myelin material when neurons are silenced29. Consistent with microglia engulfing bits of the degenerating RGC axons and migrating away from the source of degeneration, bright RFP-labeled puncta within and outside the tectal neuropil were noted (Figure 1B, arrows). By four days post surgery, fragmented axons and RFP-labeled axonal debris are considerably reduced in either tectal lobe, indicating relatively rapid clearance of the dying and degenerating axons (Figure 2C,D).
To perform wholemount immunohistochemistry and visualize the optic tectum of larval zebrafish, the brain was exposed by dissecting off the eye(s), jaw, ears, and skull cap of the skin and connective tissues. Ideally, the brain remains intact during this procedure (e.g., Figure 2C,D). However, parts of the forebrain, particularly the olfactory bulb, are likely to be damaged or completely removed when the skull cap of skin or jaw is removed (Figure 2A, arrowhead). Moreover, sometimes the lateral edge of the tectum can be sliced when piercing or pinching the skin to pull it off the brain (e.g., Figure 2A; tear on the right tectal lobe).
Immunohistochemistry was also used to assess the number of mitotic cells in innervated and non-innervated lobes of the optic tectum by staining whole brains with a PH3 antibody. These data show significantly fewer mitotic cells on the denervated side of one-eyed larval fish (p = 0.00033, Welch's t-test; Figure 3).
Figure 2: Eye removal at 5 dpf triggers degeneration of optic nerve axons. (A–D) Representative maximum-intensity projections showing dorsal views of wild-type brains from intact larvae fixed at 7 dpf (A) or 9 dpf (C), or one-eyed larvae from eye extirpation surgeries at 5 dpf, fixed 2 dps (B) or 4 dps (D). All nuclei are stained with DAPI (green), and optic nerve axons terminating in the neuropil of the optic tectum are labeled by the atoh7:RFP transgene (magenta). Asterisk indicates innervated tectal neuropil in larvae with only the right eye intact. Arrows indicate examples of fragments of degenerated RGC axons emanating from the denervated optic tectum. Arrowhead indicates missing parts of the forebrain. Scale bar = 50 µm. Abbreviations: dpf = days post fertilization; dps = days post surgery; RGC = retinal ganglion cells; DAPI = 4',6-diamidino-2-phenylindole. Please click here to view a larger version of this figure.
Figure 3: Number of mitotic cells in the optic tectum increases with innervation. (A–B) Representative maximum-intensity projections showing dorsal views of wild-type brains from intact (A) or one-eyed (B) 9 dpf larvae immunostained with antibodies for the mitotic marker PH3 (magenta) and nuclei counterstained with DAPI (green). Asterisk indicates innervated tectal neuropil from the intact right eye. (C) Box plot with all individual data points overlaid showing quantitation of the ratio of PH3+ cells in innervated (left) versus denervated (right) tectal lobe as a difference ratio (left-right/total) for one-eyed (n = 12) and two-eyed (n = 8) 9 dpf larvae. Means of the two conditions compared with Welch's t-test (***, p < 0.0005). Scale bar = 50 µm. Abbreviations: dpf = days post fertilization; dps = days post surgery; PH3 = phosphorylated histone H3; DAPI = 4',6-diamidino-2-phenylindole. Please click here to view a larger version of this figure.
Supplemental file: Outline and instructions for plot shown in Figure 3C. This file contains all data and code so that anyone can reproduce the box plot with all data points overlaid, as shown in Figure 3C. Please click here to download this File.
The techniques described in this paper illustrate one of many approaches for studying vertebrate visual system development in zebrafish. Other researchers have published methods to dissect the embryonic retina and perform gene expression analyses19 or visualize neuronal activity in the optic tectum30. This paper provides an approach for exploring how differential retinal input may influence cell behaviors in the optic tectum.
To ensure successful eye extirpation and larval survival after surgery, it is important that the larvae are healthy, adequately anesthetized, and immobilized in 1% agarose. It is also crucial that forceps and tungsten needles are very sharp and clean. Larvae survive best when they are not fed 24 h before or after the surgery, and rearing larvae in MMR supplemented with antibiotics speeds the healing process. Importantly, the higher ionic strength of MMR and especially the higher concentration of calcium promotes healing14. However, longer exposure to the higher salinity can decrease survival, similar to what has been documented in embryonic zebrafish31. A most critical step is fixation to ensure successful brain dissections, which are essential for data analysis of images of immunostained samples. Using freshly made 4% PFA supplemented with 4% sucrose (affectionately called sweet fix) tends to increase the ease of skin removal and brain tissue preservation. As with the surgeries, sharp and clean tools are a must for dissecting brains.
One of the biggest challenges of this protocol is embedding larvae for eye extirpations. It is important that each person find the way that works best for them so that they can confidently and quickly remove the larval eye. If the larva is accidentally stabbed during the surgery, humanely kill it by decapitation. Another challenge is determining the most effective antibody concentrations for wholemount immunostaining. When determining the appropriate antibody concentrations, use the published literature as a starting point and optimize the protocol as using published reagents on different tissues (especially on larger tissues or older animals) may require increased concentrations and/or incubation times.
The major limitation of this method is the time a single experiment takes from start to finish and the number of fish that can be studied at any given time. For example, all the data shown in Figure 3 took about one month from the initial pairing of fish to the final data collection and analysis. This relatively low-throughput approach could be complemented by heterologous genetic approaches that either selectively silence neural activity in the visual system32,33 or by mutants that lack RGC axons or activity34,35.
The significance of this method is that it brings a traditional embryological technique to bear on modern transgenic fish lines, allowing direct comparison of cellular and genetic mechanisms operating in each of the tectal lobes within the same individual. Moreover, this method is ripe for application to other experimental systems, including Astyanax surface fish36 and/or transgenic zebrafish larvae with fluorescently-labeled microglia37,38,39 and astrocytes40 to study how the developing fish nervous system responds to such a dramatic injury.
The authors have nothing to disclose.
Funding for this work was supported primarily by start-up funds from Reed College to KLC, Helen Stafford Research Fellowship funds to OLH, and a Reed College Science Research Fellowship to YK. This project began in Steve Wilson's lab as a collaboration with HR, who was supported by a Wellcome Trust Studentship (2009-2014). We thank Máté Varga, Steve Wilson, and other members of the Wilson lab for initial discussions about this project, and we especially thank Florencia Cavodeassi and Kate Edwards, who were the first to teach KLC how to mount embryos in agarose and perform zebrafish brain dissections. We also thank Greta Glover and Jay Ewing for help with assembling our tungsten needle-sharpening device.
Equipment and supplies: | |||
Breeding boxes | Aquaneering | ZHCT100 | |
Dow Corning high vacuum grease | Sigma or equivalent supplier | Z273554 | |
Erlenmeyer flasks (125 mL) | For making Marc's Modified Ringers (MMR) with antibiotics for post-surgery incubation | ||
Fine forceps – Dumont #5 | Fine Science Tools (FST) | 11252-20 | |
Glass Pasteur pipettes | DWK Lifescience | 63A53 & 63A53WT | For pipetting embryos and larvae |
Glass slides for microscopy | VWR or equivalent supplier | 48311-703 | Standard glass microscope slides can be ordered from many different laboratory suppliers. |
Glassware including graduated bottles and graduated cylinders | For making and storing solutions | ||
2-part epoxy resin | ACE Hardware or other equivalent supplier of Gorilla Glue or equivalent | 0.85 oz syringe | https://www.acehardware.com/departments/paint-and-supplies/tape-glues-and-adhesives/glues-and-epoxy/1590793 |
Microcentrifuge tube (1.7 mL) | VWR or equivalent supplier | 22234-046 | |
Nickel plated pin holder (17 cm length) | Fine Science Tools (FST) | 26018-17 | To hold tungsten wire while sharpening and performing surgeries/dissections. |
Nylon mesh tea strainer or equivalent | Ali Express or equivalent | For harvesting zebrafish eggs after spawning; https://www.aliexpress.com/item/1005002219569756.html | |
Paper clip | For Tungsten needle sharpening device. | ||
Petri dishes 100 mm | Fischer Scientific or equivalent supplier | 50-190-0267 | |
Petri dishes 35 mm | Fischer Scientific or equivalent supplier | 08-757-100A | |
Pipette pump | SP Bel-Art or equivalent | F37898-0000 | |
Potassium hydroxide (KOH) | Sigma | 909122 | For Tungsten needle sharpening device. Make a 10% w/v solution of KOH in the hood by adding pellets to deionized water. |
Power supply (variable voltage) | For Tungsten needle sharpening device. Any power supply with variable voltage will work (even one used for gel electrophoresis). | ||
Sylgard 184 Elastomer kit | Dow Corning | 3097358 | |
Tungsten wire (0.125 mm diameter) | World Precision Instruments (WPI) | TGW0515 | Sharpen to remove eye and dissect larvae. |
Variable temperature heat block | The Lab Depot or equivalent supplier | BSH1001 or BSH1002 | Set to 40-42 °C ahead of experiments. |
Wide-mouth glass jar with lid (e.g., clean jam or salsa jar) | For Tungsten needle sharpening device. | ||
Wires with alligator clip leads | For Tungsten needle sharpening device. | ||
Microscopes: | |||
Dissecting microscope | Any type will work but having adjustable transmitted light on a mirrored base is preferred. | ||
Laser scanning confocal microscope | High NA, 20-25x water dipping objective lens is recommended. Microscope control and image capture software (Elements) is used here but any confocal microscope will work. |
||
Reagents for surgeries and dissections: | |||
Calcium chloride dihydrate | Sigma | C7902 | For Marc's Modified Ringers (MMR) and embryo medium (E3). |
HEPES | Sigma | H7006 | For Marc's Modified Ringers (MMR). |
Low melting point agarose | Invitrogen | 16520-050 | Make 1% in embryo medium (E3) or Marc's Modified Ringers (MMR). |
Magnesium chloride hexahydrate | Sigma | 1374248 | For embryo medium (E3). |
Magnesium sulfate | Sigma | M7506 | For Marc's Modified Ringers (MMR). |
Paraformaldehyde | Electron Microscopy Sciences | 19210 | Dilute 8% (w/v) stock with 2x concentrated PBS (diluted from 10x PBS stock). |
Penicillin/Streptomycin | Sigma | P4333-20ML | Dilute 1:100 in Marc's Modified Ringers. |
Phosphate buffered saline (PBS) tablets | Diagnostic BioSystems | DMR E404-01 | Make 10x stock in deionized water, autoclave and store at room temperature. Dilute to 1x working concentration. |
Potassium chloride | Sigma | P3911 | For Marc's Modified Ringers (MMR) and embryo medium (E3). |
Sodium chloride | Sigma | S9888 | For Marc's Modified Ringers (MMR) and embryo medium (E3). |
Sodium hydroxide | Sigma | S5881 | Make 10 M and use to adjust pH of MMR to 7.4. |
Sucrose | Sigma | S9378 | |
Tricaine-S | Pentair | 100G #TRS1 | Recipe: https://zfin.atlassian.net/wiki/spaces/prot/pages/362220023/TRICAINE |
Reagents for immunohistochemistry: | |||
Alexafluor 568 tagged Secondary antibody to detect rabbit IgG | Invitrogen | A-11011 | Use at 1:500 dilution for wholemount immunohistochemistry. |
DAPI or ToPro3 | Invitrogen | 1306 or T3605 | Make up 1 mg/mL solutions in DMSO; 1:5,000 dilution for counterstaining. |
Dimethyl sulfoxide (DMSO) | Sigma | D8418 | A component of immunoblock buffer. |
Methanol (MeOH) | Sigma | 34860 | Mixing MeOH with aqueous solutions like PBST is exothermic. Make the MeOH/PBST solutions at least several hours ahead of time or cool them on ice before using. |
Normal goat serum | ThermoFisher Scientific | 50-062Z | A component of immunoblock buffer. Can be aliquoted in 1-10 mL volumes and stored at -20 °C. |
Primary antibody to detect phosphohistone H3 | Millipore | 06-570 | Use at 1:300 dilution for wholemount immunohistochemistry. |
Primary antibody to detect Red Fluorescent Protein (RFP; detects dsRed derivatives) | MBL International | PM005 | Use at 1:500 dilution for wholemount immunohistochemistry. |
Proteinase K (PK) | Sigma | P2308-10MG | Make up 10 mg/mL stock solutions in PBS and use at 10 µg/mL. |
Triton X-100 | Sigma | T8787 | Useful to make a 20% (v/v) stock solution in PBS. |
Software for data analysis | |||
ImageJ (Fiji) | freeware for image analysis; https://imagej.net/software/fiji/ | ||
Rstudio | freeware for statistical analysis and data visualization; https://www.rstudio.com/products/rstudio/download/ | ||
Adobe Photoshop or GIMP | Proprietary image processing software (Adobe Photoshop and Illustrator) are often used to compose figures). A freeware alternative is Gnu Image Manipulation Program (GIMP; https://www.gimp.org/) | ||
Zebrafish strains | available from the Zebrafish International Resource Centers in the US (https://zebrafish.org/home/guide.php) or in Europe (https://www.ezrc.kit.edu/). Specialized transgenic strains that have not yet been deposited in either resource center can be requested from individual labs after publication. |