Journal of Visualized Experiments www.jove.com Video Article Live Imaging Of Drosophila melanogaster Embryonic Hemocyte Migrations

Many studies address cell migration using in vitro methods, whereas the physiologically relevant environment is that of the organism itself. Here we present a protocol for the mounting of Drosophila melanogaster embryos and subsequent live imaging of fluorescently labeled hemocytes, the embryonic macrophages of this organism. Using the Gal4-uas system1 we drive the expression of a variety of genetically encoded, fluorescently tagged markers in hemocytes to follow their developmental dispersal throughout the embryo. Following collection of embryos at the desired stage of development, the outer chorion is removed and the embryos are then mounted in halocarbon oil between a hydrophobic, gas-permeable membrane and a glass coverslip for live imaging. In addition to gross migratory parameters such as speed and directionality, higher resolution imaging coupled with the use of fluorescent reporters of F-actin and microtubules can provide more detailed information concerning the dynamics of these cytoskeletal components.

1. Dislodge embryos from the apple juice agar plate using a small amount of water and a soft-tipped paintbrush. Dislodged embryos can be seen easily with the naked eye. 2. Transfer embryos to a cell strainer (Fisher) or home-made basket 7 by pouring water from the apple juice agar plate into the basket held over a beaker to collect waste water. 3. Repeat step 2.2 until you are satisfied you have enough embryos transferred from the apple juice agar plate. 4. Wash embryos in cell strainer/basket using water. 5. Place cell strainer/basket in the Petri dish lid of the apple juice agar plate and add enough neat bleach to suspend embryos in the cell strainer/basket. 6. Follow dechorionation of the embryos on a dissection microscope under brightfield: dechorionation is complete when the dorsal appendages have dissolved, which should occur within two minutes. 7. Remove cell strainer/basket containing embryos from the bleach and wash off residual bleach using water. All traces of bleach should be removed before proceeding to step 2.8. One trick to assess whether all the bleach has been removed is to blot off residual water on blue-colored laboratory tissues -if there is residual bleach the blue color will be bleached white/pink. 8. Blot off remaining water using laboratory tissue/mediwipes applied to the underside of the cell strainer/basket. 9. Place a droplet of water in a Petri dish lid. With a fine paintbrush, collect all the dechorionated embryos from the embryo basket and resuspend them in the droplet. Next dry the embryos by aspirating water using a micropipette or carefully absorbing it with a laboratory tissue/mediwipes. 10. Once the embryos have been dried, add a drop of voltalef oil to cover all the embryos. Put a second small drop of oil adjacent to the droplet containing the embryos. N.B. we have been unable to find a UK-based supplier of voltalef oil; halocarbon oil 700 (Sigma) may be used instead.

Representative Result:
This protocol describes how to mount Drosophila embryos for live imaging of hemocytes on the ventral side of the embryo. If done correctly it will be easy to generate either stills or movies of hemocytes. The major determinant is the microscope used to image the hemocytes (in particular the objective lens), but the nature of the images acquired will also depend upon the stage of development, the temperature the embryos were raised at and the Gal4 and uas lines used.
Higher levels of fluorescent protein expression will enable hemocytes to be imaged with greater ease, therefore it is important to be able to see hemocytes when the embryos are at stage 2.11 of the protocol ( Figure 2 contains examples of clear hemocytes within embryos, taken with a camera fitted to a dissection microscope). Therefore increased numbers of Gal4 and uas constructs enable a greater signal-to-noise ratio. Furthermore this reduces the need for high laser intensities or increased exposure times when imaging, which in turn enables hemocyte behavior to be followed for longer periods of time.
Very high levels of GFP expression will show up fine details of hemocyte morphology, particularly the thin sheet-like lamellae that surround the circular cell body ( Figure 4A-B). Circular regions excluding GFP represent phagosomes ( Figure 4A-C). Finger-like filopodia can also be seen emerging from the lamellae ( Figure 4B). Two Gal4 drivers remain sufficient to see these processes ( Figure 4C), particularly if one or more is srp-Gal4 (see discussion), however slower scanning speeds or greater laser power on the confocal microscope may be required. As expression levels decrease it becomes more difficult to image the protrusions of hemocytes; nonetheless it is still possible to track the migration of hemocytes under these conditions as the cell body remains obvious even when protrusions are less clear ( Figure 4D).
At earlier stages of development (up to stage 13) hemocytes migrate in close contact to one another and it is often hard to distinguish individual cells. By the end of stage 13 hemocytes have formed a single line down the ventral midline ( Figure 5A), then, becoming more motile, migrate laterally to the edges of the ventral nerve cord ( Figure 5B). The actin cytoskeleton within the dynamic protrusions of hemocytes can be directly observed through expression of GMA ( Figure 5C) or cherry-moesin.
Mounting the embryos in this way allows gas exchange and prevents dehydration and embryos remain viable following imaging. If the embryo is damaged during mounting it is generally obvious as the embryo content will leak through its vitelline membrane. If an embryo does start to dehydrate then this can often been seen by deformations in the vitelline membrane. Occasionally an embryo will roll during the course of a timelapse movie, however this only tends to be problematic for longer timescale movies. Lastly, mounting several embryos at once gives the experimentalist the best chance of obtaining an embryo in the perfect orientation for their experiment. . Under a fluorescent dissection microscope select appropriately staged embryos of the desired genotype using a pair of watchmakers forceps (number 5) from the oil droplet. These forceps should be bent inwards ( Figure 1) in order to scoop up the embryos without puncturing their vitelline membrane. Transfer selected embryos to the second oil droplet. It is important that you are able to see fluorescent hemocytes on the dissecting microscope in order to be able to collect good images on the confocal microscope ( Figure 2). We typically mount stage 13/14 embryos to image lateral migration of hemocytes on the ventral midline or stage 15 embryos to image the motility of hemocytes following dispersal over the embryo. 12. Stick two coverslips (18x18mm, thickness 1) to the underside of a Petriperm/Lumox dish (Sarstedt), using 2 small drops of voltalef oil, leaving approximately 1cm between them ( Figure 3); these will be used to support a coverslip placed over the embryos, so as not to crush them. Petriperm dishes (50mm diameter) contain a hydrophobic, gas-permeable membrane. We find that the dishes become easier to use once they have been used several times (dishes can be wiped with 70% ethanol and reused). 13. Under brightfield on the dissection microscope, pick up selected embryos one by one with the bent forceps and line them ventral side up and parallel to the edge of the coverslips (Figure 3). It is possible to align up to 15 embryos in this way, depending on your dexterity and your patience. It is important to manipulate the embryos gently as both the embryos and Petriperm dish membrane are fragile and can be easily ruptured. 14. Once the embryos are aligned add a small drop of oil and let it spread to form a homogenous layer between the two coverslips. After the oil has spread (this may take a few minutes) check that the embryos are still ventral side up. If the embryos have rolled slightly, reposition them again with the forceps. 15. Finally, using tweezers (number 3) place a coverslip (18x18mm, thickness 1) over the embryos, resting upon the two previously adhered coverslips. Glue this coverslip to the coverslip supports using nail polish (Figure 3). 16. Take the Petriperm dish with mounted embryos to the confocal or wide-field microscope and mount the Petriperm dish on the stage using an appropriate adapter. Either an upright or inverted microscope may be used, with the objective lens focusing through the coverslip (as opposed to through the membrane).  Once the oil has spread to fill the gap between the two coverslips a third coverslip (18x18mm thickness 1) is gently placed on top of the oil-covered embryos using the two previously adhered coverslips as a bridge to prevent the embryos from being squashed. This coverslip is then glued to the two coverslip bridges using two small drops of nail polish. Once set, the embryos can be imaged on an upright or inverted microscope with the objective lens focusing down through the coverslip (as opposed to through the Petriperm membrane). Here hemocyte protrusions are less obvious due to lower expression of GFP but it is still possible to make movies and track hemocyte migration with this combination of Gal4 driver and uas construct. Images were taken on a Leica LSM510 confocal microscope; anterior is up in all images; the rings at the periphery of images are caused by vitelline membrane autofluorescence. move during the scan. The precise details of how images are subsequently collected will be determined by the experimental questions to be addressed.
We typically use uas-GFP or uas-GMA 3 to label hemocytes; uas-GMA is particularly useful since, in addition to marking hemocytes, it provides a read-out of actin filament dynamics. Other uas constructs may be used to probe hemocyte behavior: for example uas-cherry constructs 9 can be exploited as alternative fluorophores or uas-tau-GFP 10 could be used to visualize the microtubules within a hemocyte. Uas constructs with a nuclear label may be particularly valuable for automated tracking of hemocyte movements ( Figure 2E-F). Two-color imaging is also possible (e.g. one can label both the actin and microtubule cytoskeletons using uas-cherry-moesin and uas-tau-GFP, respectively, under the control of a hemocyte-specific Gal4 driver). As previously mentioned, a key determinant in live imaging of hemocytes is the Gal4 driver employed: as regards to current hemocyte-specific promoters srp-Gal4 2 > crq-Gal4 11 > pxn-Gal4 4 , with only srp-Gal4 sufficient to label hemocytes when heterozygous, whereas at least two copies of crq-Gal4 or pxn-Gal4 are needed to observe hemocytes live. Nonetheless optimal imaging with srp-Gal4 requires homozygosity (or the presence of an alternative Gal4 driver).
By following hemocytes live in this way it is possible to study their developmental dispersal from the head 8 and observe how they interact with the cues that pattern this process, deposit matrix and engulf apoptotic corpses. Hemocytes also represent an interesting system with which to probe the machinery of cell migration; by examining the migration of hemocytes in embryos that lack components of the actin or microtubule cytoskeletons it is possible to understand their in vivo function more clearly (e.g. Rho GTPases 4, 12 ). Exploitation of uas constructs to label these cytoskeletons provides yet more detailed information on their regulation.
We have modified this technique to probe hemocyte behavior in a variety of contexts such as their responses to laser-induced wounds4 and injection of fluorescently labeled bacteria 13 . Furthermore this method can easily be adapted to image other cell types by choosing different Gal4 drivers. For example we have previously imaged dorsal closure using epithelial Gal4 drivers 14,15 . The application of this technique to other cell types is limited by the strength of the Gal4 driver, the position of the cells to be imaged within the embryo and the type of microscope used for imaging; multiphoton confocal microscopes enable the user to image cell behavior deep within the embryo (e.g. germ cell transepithelial migration 16 ), whereas conventional confocal microscopes are unable to reach these depths (this is why we image along the ventral midline where hemocytes are very superficial, trapped between the developing ventral nerve cord and epidermis).
The strength of this protocol is that it provides an environment to maintain embryos in a healthy condition for live imaging. While a degree of manual dexterity is required to position the embryos, it is an easy technique to master and quickly gives reproducible results and can easily be modified using different Gal4 drivers and uas constructs to probe many different aspects of hemocyte biology. Additionally, the protocol is not restricted to hemocytes since the use of alternative Gal4 drivers enables the behavior of other tissue types to be analyzed.