Here we describe a method for mounting zebrafish embryos for long-term imaging, two-photon imaging and tissue-damage techniques, and time-lapse confocal imaging.
Zebrafish have long been utilized to study the cellular and molecular mechanisms of development by time-lapse imaging of the living transparent embryo. Here we describe a method to mount zebrafish embryos for long-term imaging and demonstrate how to automate the capture of time-lapse images using a confocal microscope. We also describe a method to create controlled, precise damage to individual branches of peripheral sensory axons in zebrafish using the focused power of a femtosecond laser mounted on a two-photon microscope. The parameters for successful two-photon axotomy must be optimized for each microscope. We will demonstrate two-photon axotomy on both a custom built two-photon microscope and a Zeiss 510 confocal/two-photon to provide two examples.
Zebrafish trigeminal sensory neurons can be visualized in a transgenic line expressing GFP driven by a sensory neuron specific promoter 1. We have adapted this zebrafish trigeminal model to directly observe sensory axon regeneration in living zebrafish embryos. Embryos are anesthetized with tricaine and positioned within a drop of agarose as it solidifies. Immobilized embryos are sealed within an imaging chamber filled with phenylthiourea (PTU) Ringers. We have found that embryos can be continuously imaged in these chambers for 12-48 hours. A single confocal image is then captured to determine the desired site of axotomy. The region of interest is located on the two-photon microscope by imaging the sensory axons under low, non-damaging power. After zooming in on the desired site of axotomy, the power is increased and a single scan of that defined region is sufficient to sever the axon. Multiple location time-lapse imaging is then set up on a confocal microscope to directly observe axonal recovery from injury.
Part 1: Mounting zebrafish embryos for long-term imaging
Part 2: Two-photon axotomy using a custom built two-photon microscope with a Chameleon Ti-Sapphire laser
Part 3: Two-photon axotomy on Zeiss 510 confocal/two-photon microscope
Part 4: Confocal time-lapse imaging on Zeiss LSM 510
Part 5: Representative results
A successful experiment will result in an accurate representation of the cellular dynamics of axon recovery from injury. Your embryos will be healthy after imaging, with no visible degeneration and a strong heartbeat. The axotomy should result in precise damage, only severing the defined branch of the axon. There should be no injury to surrounding axons and minimal cell death. We believe we see the death of a single epidermal cell directly over the site of axotomy in ~50% of experiments. If you observe more damage, you should optimize your two-photon protocol as described in the discussion.
We have used the methods described to precisely axotomize peripheral sensory axons and to directly observe regeneration in the living zebrafish embryo. Long-term time-lapse confocal imaging in zebrafish can be used to observe many developmental processes in vivo. The two-photon axotomy procedure described can be modified for many different experimental goals. We have used the same general procedure to ablate entire trigeminal sensory neuron cell bodies, by zooming in on the cell body rather than on a branch of the peripheral axon. Any cell type identifiable with fluorescence can be precisely damaged or ablated with the focused power of the femtosecond laser. We were inspired to perfect these techniques for the zebrafish system by previous studies in several other systems where pulsed lasers were used to create localized damage or to ablate cells4,5,6. In control experiments we confirmed that axotomy is extremely precise: we have never damaged nearby axons, even when they are branches of the same cell, and only occasionally damaged epithelial cells in close juxtaposition to axons. This specificity can be explained by the fact that intensity from two-photon laser excitation drops off quadratically with distance from the focal point3,7. Moreover, since energy emitted from the excited fluorophore contributes to photodamage, surrounding unlabeled cells are likely to be spared.
The laser power required to damage an axon may vary depending on the set up of the laser, the depth of the tissue, and your experimental goal. If you wish to damage axons deeper in the embryo, more power will be required. It is best to attempt the axotomy at a low power, and then incrementally increase the power until you find the amount that will sever the axon. Once you have determined the appropriate amount of laser power for axotomy, you should be able to reproducibly use this laser power to cause local tissue damage. If you notice that over time more power is required to create an axotomy, the laser and microscope may require maintenance (for example, the laser may be out of alignment). Make sure your laser is properly aligned, clean your objective, and check the mirrors.
We thank Mark Terasaki for initial advice on using a two-photon microscope to create local tissue damage, Kathy Joubin for advice on mounting for time-lapse imaging, and the Sagasti and Portera-Cailliau labs for discussions. Initial experiments were performed by AS as a Grass Foundation Fellow at the Marine Biological Labs in Woods Hole, MA. Work in the Sagasti lab was supported by grants from the Whitehall Foundation, the Klingenstein Foundation, and a Burroughs Wellcome Fund Career Award in the Biological Sciences.