The ability to visualize important developmental processes in an in vivo vertebrate has contributed to making the zebrafish a key model for studying normal and disease conditions (reviewed in ^1,2). In particular, the neural retina is an accessible part of the central nervous system. The retina lends itself to easily perform studies of neurogenesis due to its highly organized, yet relatively simple structure, and its highly conserved neuron types across vertebrate species ³. Dynamics of cellular behaviors such as proliferation, cell cycle exit, asymmetric cell division, fate specification, differentiation, and neural circuitry formation can be followed throughout the entire process of retinogenesis, which is completed in the central retina of the zebrafish by 3 d postfertilization (dpf) ^4,5,6,7.

Furthermore, the functional requirements of different genes in each of the above mentioned stages can be concomitantly assessed in the zebrafish retina, providing an advantage over other vertebrate models in which phenotypes resulting from application of gene knockout techniques can only be assessed upon post-mortem examination of fixed tissues. In particular, the use of transgenic lines in which we can visualize and monitor the expression of fluorescent proteins as reporter transgenes in the retina, allows us to obtain temporal resolution of gene expression that underlies the genesis of a particular neuronal cell type. Due to the rapid development of zebrafish, these events can be visualized during the entire developmental period, thereby providing deeper insights into the temporal importance of gene expression in relation to neuronal cell identity acquisition and cell behavior.

Finally, these approaches can be combined efficiently in the zebrafish with the generation of chimera via transplantations, resulting in insights into two key aspects of gene function. Firstly, examining cells transplanted from a donor embryo, in which a particular gene was knocked down while they develop in an unlabeled wild type environment, allows us to obtain relevant information about gene function in a cell-autonomous manner. This leads to important insights about the function of retinal fate determinant factors within the progenitor cells they are normally expressed in. This is exemplified by the examination of the developmental fate outcome of progenitors that can no longer generate functional protein from these genes ^4,6,8,9. Utilizing this approach, we have shown that many fate determinant factors (e.g., Vsx1, Atoh7, Ptf1a, Barhl2) act cell-autonomously to drive specific retinal neuronal fates; the lack of gene expression primarily leads to a fate switch, such that the cells with gene knockdown remain viable by adopting an alternate cell fate ^4,6,7,10. Secondly, such chimeric experiments can be used to assess how wild type progenitors behave when they develop within different genetic environments. For example, by comparing the development of WT cells that usually express a gene of interest (and reporter transgene) in a WT versus manipulated host environment (e.g., gene knockout / knockdown), the resulting effects on gene expression and cell fate can be assessed. The lack of certain neuron types in the host environment, for instance, has been shown to influence wild type progenitor behavior in a cell non-autonomous manner, to bias them towards differentiating into the underrepresented or missing neuron types ^4,7,11,12. Given that retinal neurons are born in a conserved histogenic order by the sequentially timed expression of specific neuronal fate determinant genes (fate gene expression) (reviewed in ¹³), we used these methods to demonstrate how the timing of fate gene expression in wild type progenitors is affected when such progenitors develop in retinal host environments with induced aberrant cellular compositions. Here, we outline these approaches as evidence for how the combination of relatively standard and widely used techniques enables the examination of the timing of fate gene expression in developing retinal progenitors ^8,9.

This protocol describes an experimental approach combining time-lapse imaging with the ease of performing transplantation in the ex vivo developing zebrafish embryo to follow individual mosaically labeled cells throughout the entire period of developmental retinogenesis. By performing functional gene manipulations either in the host embryo, donor embryo, both or neither, one can assess the cell autonomy of gene function. This approach can be adapted widely to similar research questions in any other system for which the individual components outlined here are suitable.

All procedures were carried out according to the provisions of the Australian National Health and Medical Research Council code of practice for the care and use of animals and were approved by the institutional ethics committees.

1. Preparation of Zebrafish

1. Adult fish pairing
   1. Set up adult fish as pairs (or two pairs) in appropriately sized breeding tanks the evening prior to use.
   2. To keep the female separated from the male, use a divider to enable the fish to see, but not touch each other.
   3. In the morning, after switching on the light, remove the dividers and allow the fish to mate undisturbed.
   4. After successful mating put adult fish back into their original tanks.
      NOTE: Host and donor strains may be the same, e.g., any wild type (WT) strains. For this study the donor embryos derived from the transgenic lines Tg(atoh7:gapRFP), Tg(atoh7:gapGFP) or Tg(vsx1:GFP) in which the expression of fate determinant factors driving early or late neural fate can be visualized, respectively ^14,15,16. In order to compare the effects of differing genetic environments, host embryos can be specific mutants or morphants as described below.
2. Embryo collection
   1. Use a tea strainer to collect the eggs from the breeding box. Check for eggs every 15 min to determine the time of birth and ensure single-cell stage embryos are collected for subsequent steps.
   2. Rinse embryos with E3 medium (5 mM NaCl, 0.2 mM KCl, 0.4 mM CaCl₂, 0.9 mM MgCl₂ and 2% methylene blue in distilled water) and place them in Petri dishes containing ~ 40 mL E3 ¹⁷ at a maximum density of 50 eggs per 90 mm diameter Petri dish.
   3. Label the Petri dishes with the strain name, date and time of birth. Use a Pasteur transfer pipette to remove any debris whilst viewing the embryos under the dissecting microscope.
   4. Keep the embryos at 28.5 °C until ~3 h postfertilization (hpf) for transplantation, or continue with microinjections.
      NOTE: For transplantations, aim for donor and host embryos at similar ages. Therefore, utilize different rearing temperatures between 25 - 32 °C according to Kimmel et al. ¹⁸ to slow down or speed up one clutch of embryos to match the other.
   5. Keep ~20 of the transgenic donor embryos uninjected at 32 °C from the time of collection until time-lapse imaging (24 - 28 h later).
      NOTE: These will develop faster than the experimental cohort and will express fluorescent markers earlier. They can thus be used to set up the parameters (laser power, gain) for the imaging. In this particular example, imaging of the experimental cohort begins prior to transgene expression (to assess the exact timing of this event).

2. Preparation for Microinjection

1. Microinjection needle preparation
   1. Use a needle puller to pull needles from each borosilicate glass capillary tube (1.0 mm O.D. / 0.78 mm I.D / 100 mm long glass capillary) in the center to obtain two needles of equal length. Aim for needles that taper with long shafts.
      NOTE: Example settings for the needle puller used here are supplied in the Materials List.
   2. Store the pulled needles in a Petri dish and secure by pressing them into a strip of moldable putty or adhesive tape.
   3. Prior to the injection bring the tip of the needle in focus under a dissecting microscope and use forceps to break open the needle at the tip. Aim for a sharp tapered tip with a small opening (Figure 1C).
2. Morpholino preparation for host embryos
   1. Order morpholino oligonucleotides specifically designed for the gene of interest as well as a standard control morpholino or an appropriate 5 base-pair mismatch to control for procedural and handling effects.
      NOTE: Here, a translation blocking Ptf1a morpholino with the sequence 5'-TTGCCCAGTAACAACAATCGCCTAC-3' that inhibits the development of the intermediate born horizontal and amacrine cells, and a standard control zebrafish morpholino targeting human beta-globin intron mutation (5'-CCTCTTACCTCAGTTACAATTTATA-3') ^7,19 were used.
   2. Prepare a 1 mM morpholino stock solution (~8 - 8.5 ng/nL) and store at room temperature.
   3. On the day of injection, heat an aliquot (10 µL) of morpholino at 65 °C for 10 min, spin briefly and place on ice to cool down from 65 °C (after which it can be kept at room temperature during use). This will reverse any potentially aggregation within the morpholino solution, which can decrease effectiveness of the knockdown.
      1. For Ptf1a morpholino and equivalent standard control morpholino, prepare a working solution of 6 ng/nL by diluting the morpholino stock solution with distilled water. Keep at RT.
         NOTE: Use 2 nL of the morpholino per embryo as the Ptf1a morpholino requires a relatively high working concentration of 12 ng per embryo as determined previously ⁷. For most morpholinos, 2 - 5 ng per embryo is effective, so dilute the stock to 2 - 5 ng/nL and inject 1 nL into each embryo.
3. Preparing fluorescent histone fusion reporter constructs for labeling donor embryos
   NOTE: In order to distinguish / visualize donor cells within the host embryos, there are different approaches to label the donor embryo, including microinjection of H2A-GFP (if using transgenic donors with red fluorescent reporters) or H2B-RFP (if using transgenic donors with green fluorescent reporters) mRNA into the yolk of the single-cell stage donor embryo.
   1. Linearize at least 1 µg plasmid containing the reporter DNA.
      NOTE: Here pCS2+ plasmid (generated and kindly provided by Prof. David Turner from the University of Michigan and Prof. Ralph Rupp from the Biomedical Center Munich) containing H2A-GFP or H2B-RFP (generated by Dr. Christopher Wilkinson, Royal Holloway, University of London) was linearized with NotI restriction enzyme digestion.
   2. Purify DNA using commercially available kits as per manufacturer's instructions.
   3. Transcribe mRNA using commercially available kits with appropriate polymerase as per manufacturer's instructions. For the pCS2+ plasmid, use a SP6 polymerase kit.
   4. Purify mRNA using commercial kits or phenol-chloroform extraction followed by isopropanol precipitation as per manufacturer's instructions. Dilute the mRNA in ultrapure water (20 - 30 µL), measure the concentration with a spectrophotometer and store at -80 °C.
   5. On the day of injection, prepare a working solution of mRNA in sterile water with a final concentration of 100 ng/µL and keep on ice. Return unused mRNA stock back to -80 °C.

3. Microinjection of Morpholinos and/or mRNA into Single-cell Stage Embryos

NOTE: Microinjections are used to mark all donor cells and to prepare the different host environments (control and gene knockdown) and involve injections of mRNA or morpholino antisense oligonucleotides ^20,21.

1. Loading and calibrating the injection needle
   1. Backload the injection needle with 3 μL of the morpholino and/or histone-fluorescent fusion mRNA solution using microloader pipette tips.
   2. Shake the solution towards the injection needle tip until there are no bubbles remaining. Use injection needles with an internal glass insert such that the solution will be primarily drawn to the tip by capillary action.
   3. Attach the injection needle to a holder mounted in a micromanipulator (3D manual micromanipulator) and ensure a tight seal within the tube housing.
   4. Turn on the power and gas supply to the pressure regulated microinjector.
   5. Immerse the needle tip in a drop of mineral oil in a dish (if using a graticule eyepiece) or hover the needle tip directly over a micrometer slide to calibrate the drop volume. Measure the injection volume by pressing the foot pedal and adjust the gas pressure and/or duration time until the injection volume is equivalent to 1 nL (125 μm diameter).
      NOTE: The advantage of the micrometer slide is that measurements are independent from the magnification used at the microscope, however by using the eyepiece graticule instead, the injection drop and scale can be in focus simultaneously.
2. Embryo microinjection
   1. Place a microscope slide in a Petri dish and use a transfer pipette to deposit single-cell stage embryos along the slide edge. Align into a single column using a microloader pipette tip with half of the thin tip cut off. Remove as much liquid as possible with a fine transfer pipette to prevent movements during injections (Figure 1A).
   2. Lower the needle towards each embryo, and penetrate the chorion and yolk in one smooth stroke (Figure 1B).
   3. Once the needle tip is centered in the yolk, microinject by pressing the foot pedal. Inject 1 nL of H2A-GFP or H2B-RFP mRNA into the yolk of a single-cell stage donor embryo (transgenic). Inject 2 nL of standard MO or Ptf1a MO into the yolk of a single-cell stage wild type host embryo by pressing the foot pedal twice. Successful injections can be visualized by the small spatial deformation in the yolk.
   4. Remove the injection needle, move the dish containing the embryos to bring the next embryo into position and continue injecting until the entire column of embryos are injected (50 - 60 embryos).
   5. After injecting all embryos, lift the dish up at a 30 - 45° angle and use a gentle stream of E3 medium (squeeze bottle) to transfer the injected embryos into a new clean Petri dish.
   6. Place the embryo dish in an incubator (either at 28.5 °C or other temperature to ensure that the donor and host embryos are at equivalent ages).
   7. Repeat steps 3.2.1 - 3.2.6 as necessary to obtain ~100 donor embryos or ~200 - 250 host embryos.
   8. At 1 - 2 hpf, check the embryos under a dissecting microscope and remove any unfertilized eggs that have not advanced to the 2- or 4-cell stages with a Pasteur transfer pipette.

4. Preparation for Transplantations

NOTE: Transplantations are used to generate chimeric embryos allowing for the effects of different genetic host environments to be assessed within equivalent WT donor progenitors ^9,22,23.

1. Donor embryo selection
   1. At 3 hpf, check the donor embryos at 2 - 5X magnification for labeling signal under a fluorescence microscope. Select well-developed (symmetric cell mass with evenly sized cells) embryos with a strong fluorescent signal using the GFP plus filter (460-500 nm excitation, 510 nm long pass emission) or Texas Red filter (540 - 580 nm excitation, 610 nm long pass emission) (Figure 1D). The transgenic reporters are not yet expressed.
2. Agarose injection plate and Petri dish preparation
   1. Pour molten 2% agarose diluted in E3 medium into a 90 mm diameter Petri dish and let it cool down until the dish is safe to touch. Hot agarose can otherwise deform the mold.
   2. Float a plastic mold with wedge-shaped protrusions (6 x 25/mold) on the agarose (Figure 2D, E). Avoid air bubbles by placing one side of the mold onto the agarose and slowly lower down the other side. Make sure the mold is centered in the dish or towards the side of the deepest point of the wedge-shaped protrusions to allow enough clearance space for the transplantation needle.
   3. Allow the agarose to completely solidify at RT.
   4. Place the agarose dish at 4 °C for 30 min and remove the molds by gently lifting from one corner (e.g., with forceps). Prepare agarose injection plates the day prior to the experiment and store at 4 °C with a small volume of E3 medium and paraffin film seal to prevent them from drying out.
   5. Prior to the transplantation, fill the agarose injection plate with E3 medium and place into a 28.5 °C incubator for at least 30 min.
   6. Since dechorionated embryos stick to plastic, either use glass dishes or coat the inside of Petri dishes (90 mm diameter) with a thin layer of 2% agarose in E3. Prepare one dish for dechorionated host embryos, one dish for dechorionated donor embryos and one dish for every 40 transplanted embryos. Store as described for the injection plate (4.2.4).
3. Transfer pipette preparation
   1. Cut the tip of a long form finely pulled glass tip Pasteur pipette (2 mL volume, 230 mm length, 100 mm long tip) to a length that allows for comfortable maneuvering (optional) by scratching with a diamond knife whilst rotating until the tip falls off (Figure 1F).
   2. Ensure that the cut is straight. Fire polish the glass transfer pipette by rotating the cut surface in a Bunsen burner to generate smooth ends so as to not damage the dechorionated embryos (Figure 1F).
      NOTE: Keeping the pipette tip in the flame for too long will result in the opening being sealed up or becoming too small for the embryos.
4. Dechorionating blastula stage donor and host embryos
   1. Dechorionate embryos either manually with two fine forceps by tearing open each chorion without touching the embryo, or enzymatically using protease to allow simultaneous rapid dechorionation of a large number of embryos ⁹.
      1. Prepare a 100x stock solution of 50 mg/mL protease mixture, make 100 μL aliquots and store them at -20 °C.
      2. Add 100 μL of the stock protease to 10 mL E3 medium (final concentration of 0.5 mg/mL protease).
      3. Place well developed host and donor embryos in two separate small Erlenmeyer glass beakers (10 mL) and remove as much liquid as possible.
      4. Add 5 mL of the 0.5 mg/mL protease solution to each beaker and gently swirl the embryos whilst looking at them under a dissecting microscope. Watch for any signs of deformation of the chorion.
      5. Keep swirling. As soon as the first embryo is out of the chorion (Figure 1D, asterisks), perform three quick rinses with E3 medium to remove the protease solution by pouring out most of the liquid and refilling the flask by gently pouring in E3 along the side. Do not allow dechorionated embryos to come in contact with the air in any of the subsequent steps until the end of transplantation (section 5), as they will burst. Leave adequate solution in the flask between rinses.
      6. With the fire polished glass transfer pipette, transfer the embryos from the Erlenmeyer beaker into glass Petri dishes or agarose (2% in E3 medium) coated plastic dishes. Gently pipette during transfer to remove any remaining weakened chorion.
5. Transplantation needle preparation
   1. Use a needle puller to pull two needles from each borosilicate thin wall glass tube (1.0 mm O.D. / 0.78 mm I.D / 100 mm length) with the same settings as the microinjection pipettes.
      NOTE: Example settings for the needle puller used here are supplied in the Materials List. Transplantations needles do not have a glass insert, as this damages the cells sucked into the needle.
   2. Pull needles in advance, store in a Petri dish and secure by pressing into a strip of moldable putty or adhesive tape.
   3. Prior to transplantation bring the tip of the needle in focus under a dissecting microscope and use forceps to break open the needle at the tip. Use fine forceps to adjust the shape of the needle to aim for a flute-shape (Figure 2G), or use alternate methods described to generate the tip shape ²³.

5. Chimera Generation via Blastula Transplantation

1. Transplantation setup
   NOTE: Here, a self-assembled transplantation apparatus was used (Figure 2A).
   1. Mount the transplantation needle in the same needle holder used for microinjection above (mounted on micromanipulator, step 3.1.3) (Figure 2A).
   2. Disconnect the injecting tubing that links the end of the micropipette holder to the pressure regulated microinjector used for microinjections. Attach the injection tubing that is linked to a 1 mL syringe (Figure 2A, B), which represents the transplantation rig.
      1. Ensure that all connections are tightly sealed using paraffin film or moldable putty. Prevent embryos from coming in contact with air by ensuring that there is always some E3 medium in the transplantation needle itself. Operate the syringe manually to suck cells / liquid into and out of the transplantation needle.
2. Embryo alignment
   1. Transfer the donor embryos with the glass pipette into the first column of the agarose mold. Transfer the host embryos into columns 2 to 6 of the agarose mold (Figure 2F). Position the embryos with the pipette so that the blastomere side faces up.
3. Transplantation
   1. Gently rest the transplantation needle onto a blastomere cell of a donor embryo and slowly suck up about 20 - 50 cells into the transplantation needle (Figure 2H). Avoid sucking up the yolk.
   2. Lift the transplantation needle from the donor embryo using the micromanipulator. Move the agarose mold dish to the left and insert the transplantation needle tip through the side of the cell mass into the animal pole of the first host embryo. Deposit 5 - 10 cells near the surface according to fate mapping showing that this area gives rise to the forebrain / eye field ²⁴ (Figure 2I). Keep the needle tip immersed in the E3 medium throughout the whole procedure.
4. Care of transplanted embryos
   1. Remove the donor embryos and leave the host embryos in the agarose mold for 1 - 2 h at 28.5 °C. Fill glass or plastic (coated with 2% agarose) Petri dishes with E3 medium containing 0.003% N-phenylthiourea (PTU) to prevent pigment formation as this will interfere with imaging. Transfer the host embryos into these dishes with the fire polished glass pipette and incubate at 28.5 °C O/N.
      CAUTION: Wear gloves when handling PTU.

6. Live Imaging Setup

NOTE: The small size and optical transparency of zebrafish combined with rapid development have allowed it to become a key vertebrate model for in vivo imaging of different cells and organs. Imaging can be performed on a variety of microscopes, which will differ in setup and parameters. The following describes a suitable confocal imaging setup for imaging of retinal development ^5,8.

1. Chimeric embryo selection
   1. At 24 - 26 hpf, screen the transplanted embryos under a fluorescent dissecting scope to select healthy well developed embryos that show the transplanted donor cells (H2A-GFP or H2B-RFP labeled) in the developing eye primordium.
   2. Set aside up to 40 embryos for mounting by pipetting into a new dish and add 0.4 mg/mL tricaine methanesulfate (MS222) to anesthetize the embryos.
      CAUTION: Wear gloves when handling MS222.
2. Mounting embryos
   NOTE: Use the appropriate mounting dish, depending on whether imaging will be performed on an inverted ⁸ or upright microscope (described here).
   1. Prepare a few 1.5 mL plastic tubes containing 1 mL of 1% low melt agarose in 0.4 mg/mL MS222 in E3 medium and keep molten in a 40 °C water bath.
   2. Suck up 5 embryos into a plastic transfer pipette and let the embryos accumulate in the tip by gravity. Touch the transfer pipette onto the surface of one of the prepared plastic tubes containing 1% low-melt agarose and 0.4 mg/mL MS222 allowing the embryos to sink into the tube whilst limiting the amount of E3 transfer.
   3. Mount embryos in a 90 mm diameter Petri dish for imaging in upright microscope and glass bottom Petri dish (any size) for imaging in inverted microscope. Use a permanent marker to visually divide either dish into two halves to image donors in the WT hosts (on one side) or morphant hosts (other side) in the same experiment.
   4. Transfer the five embryos from the agarose plastic tube to either Petri dish with 0.5 - 1 mL agarose. Remove excess agarose with a transfer pipette so that only a small drop (~200 - 300 μL) remains. If imaging with upright microscopy, avoid placing embryos within 1 cm of the Petri dish circumference. The water immersion dipping lens used for imaging may not be able to be centered in that area due to the Petri dish wall (Figure 3A).
   5. Use a microloading pipette (with the tip broken off) to align the embryos laterally until the agarose solidifies. For both types of dishes, the embryos are correctly angled laterally if the two eyes at either side of the head are completely overlapping (aligned in X-Y dimension) (Figure 3B). For inverted microscopy use, the eye near the coverslip must be pushed as close to the coverslip as possible, to enable imaging through the z-dimension within the constraints of the possible focus levels.
   6. Continue mounting five embryos at a time in individual agarose drops. Use more 1% low melt agarose to join the agarose drops to each other and the side of the dish to prevent any dislodging and lateral movement during imaging (Figure 3A).
   7. Once fully solidified, cover the dish with E3 medium (containing 0.003% PTU, 0.4 mg/mL MS222) prewarmed to 28.5 °C.
   8. Using the same approach described in steps 6.2.2 - 6.2.8, mount ~20 transgenic embryos raised at 32 °C (step 1.2.5) in a separate dish.
3. Imaging parameters
   NOTE: Imaging to analyze the onset of fluorescent transgenes does not require high spatial intracellular resolution. It can be performed with a W Plan-Apochromat 20X / 1.0 DIC M27 70-mm objective at 1.5 zoom.
   1. Set up the laser power and exposure time based on the transgene intensity for the ~20 transgenic embryos with accelerated development (i.e. raised at 32 °C; step 6.2.8).
      NOTE: For studies of onset of transgene expression, the experimental embryos show no fluorescence at the time the time-lapse is set up and thus, these accelerated embryos are crucial for determining the parameters of the imaging.
   2. For the experimental dish, visualize each embryo and save the position and focus levels of the embryos.
      1. Choose embryos that have the correct mounting angle (i.e. with the lateral side of the eye as horizontal as possible) and a strong heartbeat (an indicator of health). Choose embryos with a few transplanted donor cells (identified by H2A-GFP or H2B-RFP labeling) primarily in the anterior ventral quadrant of the developing retina (where the wave of gene expression begins) (Figure 3C).
         NOTE: In zebrafish embryos, the average heartrate is 120 - 180 beats/minute ²⁵. In the anesthetized embryos, a healthy heartbeat may be in the range of 60 - 120 beats/minute. Slower heartbeat may be indicative of reduced viability and those embryos should be excluded from the study.
   3. Set z-stacks of optical sections at 2 μm intervals through the retina of up to 15 embryos.
      NOTE: The total number of embryos that can be imaged in any given experiment will depend on individual parameters (exposure time and z-stack size). The time taken to image 1 time-point for all of the chosen embryos must be shorter than the interval between time points. By sacrificing some depth resolution, more embryos can be imaged within the time intervals. The limits of the z-stack are chosen as lateral and medial extremes of the mounted embryos eye and 30 μm is added to the lateral side in the upright microscopy setup, as the eyes of the developing embryos in this setup shift in the z direction as the embryos grow overnight. This results in stacks 100 - 170 μm thickness (50 - 85 images/eye).
   4. Take images every 24 min at 32 °C (equivalent to 0.5 hpf intervals) using the laser/channel appropriate to the transgene.
      NOTE: The embryos remain on the stage throughout the entire time-lapse within a heated chamber encompassing the entire microscope. While the experiment can be conducted at 28 °C, warmer temperatures are used to speed up development and ensure that the relevant time-points in the experiment are imaged within the 16 - 24 h time-lapse period.
   5. Image channels representing the donor label (H2A-GFP or H2B-RFP - optional) and transgene (Atoh7:RFP or Vsx1:GFP) using sequential scanning. For analysis, utilize the donor label only to choose appropriate embryos as described (take one z-stack at the beginning of the imaging) and then perform time-lapse only on the channel to capture the transgene expression.
      NOTE: Alternatively, use the donor label only to select appropriate embryos (i.e. those with confirmed transplanted cells in anterior ventral quadrant) and image only the transgene channel, roughly halving imaging time and almost doubling the number of embryos that can be imaged and analyzed per experiment (Figure 4).

This work presents an experimental protocol to assess changes in gene expression timing when wild type retinal progenitors develop within a morphant host embryo. The experimental host embryos are Ptf1a morphants, which lack the intermediate born horizontal and amacrine interneurons of the retina ^7,26. These have been compared to control host embryos, which were injected with a standard control morpholino.

Since neurogenesis of different types of neurons across the central nervous system (including the retina) occurs in a highly conserved histogenic order ^{27,28,29,30,31} the progression of the subsequent fate may depend on cell non-autonomous feedbacks from accumulating earlier born neuron types. This hypothesis was tested by assessing the timing of the generation of later born neural types in transplanted progenitor cells developing in the absence of intermediate born neuron types. As a control, the timing of early born cells was also quantified, which should be unaffected independently of whether the intermediate cells are generated or not. To this end, the Tg(atoh7:RFP) or Tg(atoh7:gapGFP) lines were used to visualize the onset of the first born ganglion cell neurogenesis while the Tg(vsx1:GFP) line was used to analyze the onset of the late born bipolar population. Other than additionally expressing the reporter transgenes, the donor cells from these embryos are genetically WT.

Microinjection of 1 nL mRNA (to label donor cells) into the yolk of single-cell stage embryos results in fluorescent reporter protein expression by 3 hpf, at which stage the brightest donors are selected (Figure 1). Prepare transplantation setup when embryos are around 2 hpf (Figure 2). At 24 hpf, embryos with transplanted donor cells in the relevant retinal locations are chosen and mounted in 1% low melt agarose in groups of five (Figure 3). After saving the embryo position, setting the z-stack and the time parameters, time-lapse imaging is performed for 16 - 24 h. Timing of first born ganglion cells is indicated by the onset of the Atoh7:RFP or Atoh7:gapGFP transgene and timing of the last born bipolar generation is indicated by the strong upregulation of the Vsx1:GFP transgene, which is also expressed at distinctly lower brightness levels in the developing progenitors. The timing of first transgene expression, such as Atoh7:gapGFP is identified in individual cells at each timepoint of the imaging (white arrowheads, Figure 4). Atoh7:gapGFP and ganglion cell differentiation occurs at equivalent times regardless of whether the host embryo is WT or a Ptf1a morphant lacking intermediate (i.e. after ganglion cell birth) horizontal and amacrine cells.

The data presented demonstrate that this experimental approach provides a powerful tool to assess the role of particular retinal environments on gene expression timing, with relevant implications, not only for understanding normal developmental processes but also for future applications of cell reprogramming strategies in normal and diseased conditions.

Figure 1: Microinjection into the Yolk of Single-cell Stage Embryos used to Label Donor Cells and to Generate Different Host Environments. A) Embryos are aligned against the edge of a glass slide with most of the surrounding liquid removed. B) The needle is inserted into the center of the yolk and either H2A-GFP or H2B-RFP mRNA (for donor) or morpholino (standard MO or Ptf1a MO for host) are injected. C) The injection needle is generated using a needle puller that generates two symmetrical needles from each capillary. Each pulled needle subsequently needs to have the tip broken with forceps to open the needle whilst providing a sharp angled tip (not blunt) that can be easily inserted into embryos. D) Donor embryos injected with H2B-RFP mRNA at the single-cell stage start expressing the fluorescent reporter proteins by 2.5 hpf, at which stage the brightest and most evenly labeled embryos can be selected. E) Around 2.5 - 3 hpf, donor and host embryos are enzymatically dechorionated using proteases. Swirling of the Erlenmeyer flask provides the mechanical force necessary to break open weakened chorions and brings dechorionated embryos into the center. As soon as the first one (asterisk) or a few dechorionated embryos are observed, rapid but gentle rinsing ensures that the protease solution is removed. F) For transfer of dechorionated embryos, glass transfer pipettes are utilized. These can be initially shortened (optional) to any length that is comfortably maneuvered by the user, by scratching the glass at the appropriate position with a diamond knife until it snaps off. This results in a sharp cut tip, which needs to be fire-polished to smoothen any rough edges that may damage cells being sucked into the pipette during transfer. Scale bars A, B, D, E = 1 mm. Please click here to view a larger version of this figure.

Figure 2: Transplantation Setup. A) The transplantation consists of a transplantation needle mounted in a micropipette holder attached to a micromanipulator. The end of the micropipette holder is attached to tubing that leads to a 1 mL syringe, which is manually operated to suck up or deposit cells. B) The transplantation tubing shown indicates how potential leaks at joints are simply prevented using moldable putty and paraffin film. C) A higher magnification view of the connection between the injection tubing and any suitable tubing to connect to the syringe. While any tubing can be used, the tube utilized here has a diameter that tightly fits a 10 μL pipette tip. Use paraffin film to ensure a tight seal. D) The mold has 150 (6 x 25) wedges to generate the triangular divots in the agarose injection plate. E) Approximate dimensions of each wedge shape are shown. The company measurements are proprietary. F) Prepared agarose injection plates with wedge-shaped indentations are used with donor embryos added to the first column and columns 2 - 6 filled with host embryos. Transplantations are performed across the row, whereby cells collected from one donor in each row can be transplanted into 5 host embryos. Within each mold, cells from up to 25 donors can be transplanted into up to 125 host embryos. Note the mold is placed asymmetrically with the deeper end of the wedge closer to the Petri dish edge. This allows the transplantation needle from the right hand side without hitting the Petri dish edge. G) The transplantation needle is pulled with the same parameters as the injection needle to result in a long tapered tip shown here. The needle must be broken to generate a larger internal diameter that can take up cells without deforming them. The higher power inset shows that the tip of the transplantation needle should be angled and ideally have a "flute" shape as shown in this example. H) Higher power view shows orientation of host embryos with cells at the top. The transplantation needle is inserted into the cell mass of the blastula stage host embryo laterally (white asterisks) and pushed into the future eye as shown here. Donor cells can be seen in the transplantation needle (arrow) and thus the rough number of cells transplanted can be controlled. I) Two examples (left and right) showing that immediately after transplantation, embryos with successfully transplanted cells in the correct location can immediately be sorted out using the donor label (e.g., H2B-RFP). A fluorescent channel was added to the brightfield using "linear dodge (add)" option under the layer menu in the raster graphics editor. Scale bar (for H, I) = 500 μm. Please click here to view a larger version of this figure.

Figure 3: Embryo Mounting for Time-lapse Imaging. A) Petri dish (90 mm) showing multiple low melt agarose drops containing five embryos each connected by agarose to form a firm network to prevent individual drops from detaching. B) Higher power view of 24 hpf embryos aligned laterally and embedded within solidified agarose. For the ideal angle, the two eyes should be positioned on top of each other (in the z-dimension). Depending on the gene of interest, mounting should be delayed until just prior to relevant timing (e.g., 26 hpf for ganglion cell differentiation or 35 hpf for the start of the bipolar cell differentiation). C) Confocal image of a single optical z-section showing overlay of brightfield and red channels (using "linear dodge (add)" option under the layer menu in the raster graphics editor) of the developing eye with a few labeled donor cells (red) within an otherwise unlabeled host environment. Scale bar B = 1 mm, scale bar C = 50 μm. Please click here to view a larger version of this figure.

Figure 4: Timing of Neurogenesis can be Visualized in Transgenic Donor Cells Transplanted into Different Host Environments. Micrographs from time-lapse image of the developing WT donor cells from Tg(atoh7:gapGFP) transgenic embryos transplanted into unlabeled WT host. Transgene expression onset is indicated for a few cells by white arrowheads. Scale bar = 10 μm. Please click here to view a larger version of this figure.

Understanding the extent to which neighboring cells influence timing of the expression of crucial cell fate determining factors is essential when aiming to efficiently instruct embryonic or induced multipotent stem cells to differentiate into a specific post-mitotic cell type or even patterned tissue. Furthermore, examining these molecular events in the developing cells of the living animal additionally provides relevant dynamic (temporal and spatial) information on the particular cellular contexts associated with these events in vivo. Such studies cannot be easily achieved in all vertebrate models.

In the presented protocol, the amenable zebrafish embryo is exploited as a vertebrate model to address these questions in vivo, in the relatively simple, yet highly organized three-dimensional patterned tissue of the vertebrate retina. This is exemplified by testing the hypothesis that depleting a particular retinal cell type influences gene expression timing and developmental fate of retinal progenitor cells. We can additionally utilize the donor cell marker to track specific cells back in time to assess, for instance, how long before the expression of a transgene a postmitotic cell underwent its last division, or assess the general behavior of this cell (e.g., migration) at various stages prior to the transgene expression in different environments. For these studies, a combination of relatively easy to perform techniques are presented, which include morpholino-mediated gene knock down, transplantation and live imaging. For the transplantation method, the protocol introduces a novel, relatively easy and inexpensive transplantation rig and demonstrates the ease of this operating setup.

This protocol is not restricted to the retina. It can also be applied to study a number of different dynamic cellular processes, that require disentangling of cell-autonomous developmental programs from cell non-autonomous environmental influences. However, while little troubleshooting is required, a number of critical steps need to be taken into account prior to starting an experiment. One important prerequisite for the successful application of the technique is that the organ or tissue of choice must be easily targeted in the early embryo using the fate map. If the efficiency of tissue targeting has not been previously tested, this needs to be first optimized and standardized. The number of microinjected and transplanted embryos can be increased and screened prior to live imaging to obtain sufficient embryos with donor cells correctly integrated and positioned within the relevant tissue. Additionally, while this protocol can be easily applied to early embryonic zebrafish stages, application to later stages such as the growing larvae become more difficult for two reasons. Firstly, the efficiency of gene knockdown obtainable with morpholinos decreases over time (depending on each morpholino, it is generally most efficient before 3 dpf). To override this problem, the use of genetic mutants as donor embryos for transplantation is preferable. Secondly, as the zebrafish grows, deeper parts of the larvae become less accessible to imaging with a confocal microscope, especially at higher magnifications (high power objectives) where the working distance of the microscope objective is more limited. Thus, each tissue of interest at relevant age needs to be first assessed for its amenability to imaging.

The time-lapse imaging can be performed either on an inverted or upright confocal microscope.

There are a number of advantages of the upright microscope, if available. At the required temperature, evaporation of the immersion substance used (e.g., water or immersion oil with water refractive index) on an inverted microscope can lead to complete drying out. This requires constant monitoring and eventually refilling, with the risk of placing the sample out of focus. This can be partially avoided by using automation to lower the objective a defined distance, adding the immersion substance and raising the objective again, though in our experience, refocusing is often still required, which should happen before the beginning of the next time point of imaging. Secondly, the largest coverslip bottom dishes we sourced are still much smaller (50 mm in diameter) than the standard 90 mm diameter Petri dishes utilized for upright microscopy. This means that the number of embryos that can be mounted and the embryo medium volume added are more limited when using inverted microscopes. Nevertheless, the time-lapse parameters usually constitute the main limiting factor for the number of embryos that can be imaged and therefore the size of the dish is still suitable. When using very high magnification objectives, inverted microscopy may allow a greater depth in the z-dimension as the eye and objective are only separated by a thin coverslip. When combining very high magnification with upright microscopy, the agarose drop covering the embryo must be kept as thin as possible. Overall, the data obtained using upright or inverted microscopy is of equivalent quality.

For generation of high quality data, there are additional critical steps that need to be considered. Firstly, selection of brightly labeled donors is crucial for obtaining transplanted embryos with an appropriate number of transplanted cells in the eye at the start of the time-lapse movies (a few, but not too many). The protease enzymatic dechorionation is a critical step and incubation in protease should be kept to a minimum followed by quick, but gentle rinses to not affect the subsequent health of the embryos. Age matching of donor and host embryos allows for accurate tissue targeting according to the fate map and for efficient integration of the cells. A well-shaped transplantation needle is important, ensuring that the tip is small and sharp enough to insert into host embryos without causing much damage, but large enough in internal diameter to not cause mechanical damage to the cells being transplanted. During transplantation, if the equipment is not fully sealed (especially at connection points), it is very difficult to manually control the movement of E3 medium and cells into and out of the transplantation needle. This is most obvious when there is flow either into or out of the transplantation needle without manually changing the syringe volume. In this case all joints must be re-checked and resealed. This should be tested in advance of the experiment.

Lastly, during the elapsed time between transplantation and time-lapse setup, it is crucial that the transplanted embryos are allowed to develop at low density whilst meticulously cleaning out any unhealthy embryos. This is particularly important, when the timing of development or gene expression is being investigated. When assessing the timing of any process (e.g., gene expression), controls must be used to rule out non-specific side effects of the morpholino technology. In our work, we use the timing of transgene expression in an unrelated tissue (e.g., the muscle) unaffected by the specific gene targeted by the morpholino.