We describe a method to separate organelles by density in living Drosophila embryos. Embryos are embedded in agar and centrifuged. This technique yields reproducible separation of major organelles along the anterior-posterior embryo axis. This method facilitates colocalization experiments and yields organelle fractions for biochemical analysis and transplantation experiments.
A major strategy for purifying and isolating different types of intracellular organelles is to separate them from each other based on differences in buoyant density. However, when cells are disrupted prior to centrifugation, proteins and organelles in this non-native environment often inappropriately stick to each other. Here we describe a method to separate organelles by density in intact, living Drosophila embryos. Early embryos before cellularization are harvested from population cages, and their outer egg shells are removed by treatment with 50% bleach. Embryos are then transferred to a small agar plate and inserted, posterior end first, into small vertical holes in the agar. The plates containing embedded embryos are centrifuged for 30 min at 3000g. The agar supports the embryos and keeps them in a defined orientation. Afterwards, the embryos are dug out of the agar with a blunt needle.
Centrifugation separates major organelles into distinct layers, a stratification easily visible by bright-field microscopy. A number of fluorescent markers are available to confirm successful stratification in living embryos. Proteins associated with certain organelles will be enriched in a particular layer, demonstrating colocalization. Individual layers can be recovered for biochemical analysis or transplantation into donor eggs. This technique is applicable for organelle separation in other large cells, including the eggs and oocytes of diverse species.
1) Preparing agar plates
Overview: This protocol employs agar plates for two distinct uses. First, the agar plates serve as egg collection surface; apple juice in the agar stimulates egg laying. Second, embryos embedded in the agar are held in a fixed orientation when the plates are centrifuged; a sufficiently high concentration of agar is necessary to prevent the plates from disintegrating during centrifugation. For simplicity’s sake, a single type of apple juice agar is employed for both uses.
Materials
Equipment
2) Harvesting embryos
Overview: During the first 3 hours of development (at room temperature), Drosophila embryos are single cells (not counting the germ-line precursors, the posteriorly located pole cells). When properly oriented embryos of these stages are centrifuged, the major organelles separate by density along the long axis of the egg. At the cellularization stage, this large single cell is converted into thousands of small cells; the centrifugation forces employed will not rupture these cells. For successful segregation of major organelles, it is therefore critical to centrifuge embryos that have not yet completed cellularization.
Materials
Equipment
3) Removing the egg shell from the embryos
Overview: Drosophila embryos are covered by two protective layers: an outer layer (chorion) made of protein and an inner layer (vitelline membrane) predominately made out of wax. They protect the embryo against mechanical insults and desiccation. The chorion has two extensions (the egg filaments or dorsal appendages) that are easily visible as distinct structures. In this step, we will remove the chorion by soaking the eggs in 50% bleach. Once the chorion is removed, the embryo becomes transparent in transmitted light, allowing selection of specific embryonic stages for centrifugation. As the chorion would also interfere with many post-centrifugation procedures (e.g., fixation in step 6), removal of the chorion now will allow quick processing later.
The 50% bleach treatment will dissolve the chorion within a few minutes, yet it does not harm the embryo. The vitelline membrane keeps the bleach out. We will continue the bleach treatment until the egg filaments are no longer visible. Afterwards, we will separate the embryos from the bleach by pouring the embryo/bleach slurry through a tiny sieve (a basket made of wire mesh). It is critical not to rush the bleach exposure step. If the bleach is washed off prematurely, later steps (e.g., removal of the vitelline membrane following fixation) might be compromised.
Materials:
Equipment:
4) Mounting embryos in agar
Overview: Drosophila eggs are much longer (~ 500 μm) than wide (~180 μm). To get maximal and consistent separation of organelles during centrifugation, we orient embryos such that their anterior ends point towards the center of rotation (Figure 1, 3). That way, the lightest intracellular structures (lipid droplets) accumulate at the anterior end in all embryos, while very dense structures (yolk components) accumulate near the posterior end. Embryos are inserted into small vertical holes in apple juice agar, with the anterior ends sticking up. The agar keeps the embryo in a fixed orientation during centrifugation.
Equipment:
Materials:
5) Centrifugation and recovery of embryos
Overview: The plates are transferred to swinging buckets of a Sorvall RT7 Plus centrifuge and centrifuged for 30 min at 4000 rpm, which corresponds to ~3000 g. After centrifugation, plates are covered with TSS. With a needle, embryos are dug out of the holes in the agar and transferred into appropriate vessels via a pipettor.
Equipment:
Materials:
6) Further processing
Depending on the application, embryos will subsequently be treated in various ways.
7) Representative results
If the centrifugation worked as expected, the major embryonic organelles will sort out by density 2. For example, the low-density lipid droplets will accumulate at the anterior end; the high-density yolk vesicles will accumulate at the posterior end (Figure 1, 2). Lipid droplets and yolk vesicles are storage organelles for lipids and for proteins.
Confirmation of density-dependent stratification is achieved via visual inspection of the living embryos by transillumination under a dissecting or compound microscope. The embryos should appear as in Figure 5A: The lipid droplets will form a solid brown layer at the very anterior tip (a “lipid cap”); yolk accumulation will result in a dark gray zone at the posterior end. These two dark layers are separated by a broad clear zone that represents other organelles and cytoplasm. If embryos are centrifuged after cellularization, layering will be minimal (Figure 5C).
If an epifluorescence microscope is available, these living, centrifuged embryos can be examined under UV excitation. Under these conditions, the yolk displays intense blue autofluorescence (Figure 5B). A compact layer of autofluorescent material at the posterior pole indicates successful centrifugation and serves as a marker for the posterior end.
Note that the appearance of these layers will change drastically if the embryos are fixed. For example, when embryos are fixed by standard heat- or formaldehyde fixation followed by heptane-methanol devitellinization 1, lipid droplets become translucent and the lipid layer appears whitish and fluffy by bright-light microscopy, rather than dark brown (Figure 5D). Yolk autofluorescence is partially or completely destroyed by fixation, becoming much less intense or even undetectable.
If in-vivo centrifugation is employed to highlight one particular organelle, it is helpful to label that organelle with a fluorescent marker for confirmation. For example, YFP fusion proteins targeted to mitochondria, the ER or the Golgi have been described 4. In centrifuged embryos, these markers accumulate at characteristic positions along the anterior posterior axis (Figure 1). Finally, certain histone fusion proteins localize both to lipid droplets and nuclei 2, and thus can be used to mark these cellular compartments (Figure 5 E, F) as well as to monitor the stage of the embryo (by the number of labeled nuclei). Fly strains expressing the markers mentioned in this paragraph are available from the Bloomington Drosophila stock center (http://flystocks.bio.indiana.edu/). For histone H2Av, both GFP and mRFP variants are available 5, 6
Figure 1. Separation of organelles by centrifugation (modified after 2). Centrifugation of oriented embryos results in distinct stratification (bright light). Major organelles accumulate at characteristic positions along the anterior (up) – posterior (down) axis: lipid droplets (detected in fixed embryos by Nile Red staining); endoplasmic reticulum, Golgi, and mitochondria (detected in living embryos via the YFP markers described in the main text); yolk (detected by autofluorescence). The fluorescence images were acquired by confocal microscopy.
Figure 2. Schematic depiction of embryos.
A. Egg with egg shell and the prominent egg filaments (dorsal appendages). As the egg shell is not translucent, the egg appears uniformly dark.
B. Egg with chorion removed. At the anterior end (left), the micropyle is visible. Depending on the development stage, the embryo inside the vitelline membrane will appear uniformly dark brown or show various structures. See Figure4 for examples.
C. Centrifuged embryo, oriented as in the protocol described here. Lipid droplets accumulate as a brown “cap” at the micropyle end; yolk forms a gray layer at or near the opposite end.
Figure 3. Schematic description of the procedure. Left: Embryos without chorion are placed on top of an agar plate and pushed with a needle into holes in the agar. This mounting results in a consistent orientation of all embryos, with anterior ends up. Right: After centrifugation, embryos are stratified. Agar is overlaid with TSS (blue), and embryos are recovered from the agar with the needle. Once suspended in TSS, embryos can be picked up with a pipettor.
Figure 4. Bright-light images of living Drosophila embryos, similar to how they will appear in step 4.1. Embryos are shown in standard orientation: anterior end to the left, posterior end to the right, dorsal side up, and ventral side down. A time-lapse movie of early embryogenesis is available as part of the video accompanying this protocol.
A. Cleavage stage embryos appear uniformly brown. Stages like this are ideal for in-vivo centrifugation.
B. Blastoderm embryos have a clear rim that appears similar on all sides. This clear periphery is in part due to accumulation of nuclei at the plasma membrane and in part due to inward transport of yolk vesicles and lipid droplets 7. This clear periphery expands until near the end of cellularization 1. Embryos can still be successfully stratified by centrifugation until the end of cellularization 8, though layering of yolk and nuclei is not as consistent as in cleavage stages.
C. After cellularization, embryos appear asymmetric; dorsal and ventral sides become distinct, and various folds develop. The embryo shown is in early germ-band extension. In these stages, centrifugation is no longer effective at stratifying the major organelles.
Figure 5. Centrifuged embryos. All embryos in this figure were centrifuged with their anterior end up and are shown in that orientation.
A. Bright-light image of a successfully centrifuged embryo. A brown layer at the very anterior end is due to lipid-droplet accumulation. A broad gray layer near the posterior end is due to yolk vesicles. There is frequently a clear zone below the yolk layer, of unknown origin. A yellowish zone is often apparent above the yolk layer; it probably represents mitochondria. Soluble proteins are found throughout the clear layer between lipid droplets and yolk 2.
B. The embryo in A viewed by epifluorescence microscopy under UV excitation. The blue autofluorescence characteristic of yolk helps identify the posterior end.
C. Embryo centrifuged during germ-band extension, i.e., after cellularization. Yolk may still accumulate at the posterior end, but otherwise layering is minimal.
D. Successfully stratified embryo, after heat fixation. The lipid-droplet layer appears white and fluffy, rather than dark brown as in unfixed embryos (A).
E. & F. Embryos expressing His2Av-GFP and imaged by confocal microscopy (modified after 2). E: centrifuged at cleavage stages; His2Av-GFP appears to only mark the lipid-droplet layer since few nuclei are formed by this time. Depending on the focal plane, nuclei may be visible just under the droplet layer (not shown here). F: centrifuged at blastoderm stages; His2Av-GFP marks both the lipid-droplet layer (top) and nuclei (in a broad zone below the droplet layer).
Critical steps
Make sure that the agar concentration is high enough. If it is too low, the agar will disintegrate during centrifugation, and the embryos will burst or be randomly oriented. If the agar concentration is just slightly too low or if the agar is wet (insufficiently dried or insufficient removal of excess TSS in step 4.3), the embryos will float out of their holes during centrifugation and become improperly oriented. If the agar is poured too thinly, it will also collapse or crack during centrifugation.
Make sure that embryos have not advanced beyond cellularization stages. Otherwise, separation of organelles will not work (Fig. 5C) since organelles will become sequestered within thousands of individual cells. In younger embryos at the one cell stage, organelles are free to migrate large distances and accumulate according to their characteristic density. To avoid analyzing embryos that are too old, make sure to (i) include a pre-collection to allow females to lay old eggs (step 2.4), (ii) collect embryos for 3 hours or less (step 2.5), (iii) visually select embryos before cellularization stages for embedding into agar (step 4.2), (iv) keep embedding time short to prevent development through the end of cellularization (step 4.6), and (v) discard embryos that did not layer correctly (step 5.6).
Do not skimp on the centrifugation time. Although some of the embryos will show nice layering even after only 10 minutes of centrifugation, the separation is inconsistent from embryo to embryo. 30 min spins give highly consistent results. Also, the longer the centrifugation time, the longer the layers are stable after the centrifugation has ended. This is important for those applications in which it takes time to process the embryos further before use (e.g., if they need to be processed for transplantation experiments or mounted for confocal analysis).
Possible modifications
Egg collection and chorion removal Many different protocols exist for steps 2 and 3 1, 3, 9, and will work as well as the ones described here as long as the embryos used are young, egg yield is high, and the fraction of mis-staged embryos is minimized.
Anterior-posterior alignment The protocol orients all embryos such that the posterior end is buried in the agar and the anterior end is up (Fig. 3). This orientation is good for consistency, so that it is possible to use the micropyle as a landmark which end pointed up during centrifugation (Fig. 1, 5). However, with practice, it is easy to pick out the orientation of centrifuged embryos by the distinct appearance of the lipid-droplet and yolk layers by bright-field microscopy. Then it becomes possible to mount embryos with either end up; this flexibility speeds up the mounting process.
Unoriented centrifugation The most-time-consuming and technically challenging aspect of the protocol is mounting the embryos in agar (step 4). It requires touching each embryo twice (once to insert it into the hole and once to recover it after centrifugation). As an alternative, one can transfer embryos at step 4.1 into microcentrifuge tubes filled with TSS. When these tubes are centrifuged in a microcentrifuge (13,000 rpm, 10-30 min), embryos will also typically layer well 10-13. This variation allows one to process hundreds of embryos at the same time very quickly. However, layering is not as consistent; a fraction of the embryos does not develop distinct layers despite much higher centrifugation forces (~16,000 g) than applied in the protocol above. More challenging is the fact that the orientation of the embryos is variable, and one can observe separation at various angles to the major embryo axis. This variability requires the experimenter to sort out after the centrifugation how a particular embryo was arranged in the centrifuge tube. With the use of the yolk autofluorescence as a marker for the most-dense part of the embryo, this is often possible, though it requires much practice and more judgment to do it reliably. If there are sufficient embryos in the sample, one may be able to sort out those instances in which embryos happened to be arranged like in Fig. 1.
Centrifugation of embryos in their egg shell It is possible to embed embryos in agar for step 4 without removing the chorion first 2. In this approach, embryos on the collection plate are covered with halocarbon oil instead of 50% bleach in step 3.2 (halocarbon turns the chorion translucent). Embryos of appropriate stages are selected and transferred to a new agar plate using tweezers; by grabbing only the egg filaments with the tweezers, damage to the embryo proper is avoided. Embryos are embedded as in step 4.4 through 4.8, with the anterior egg filaments sticking up out of the hole in the agar. The egg shell is removed by 50% bleach treatment after the embryos have been centrifuged and dug out of the agar as in step 5.4. Centrifugation in the presence of the egg shell may improve the rate of devitellinization after fixation (step 6.2), but selection of the correct stages for centrifugation (step 4.2) is more challenging.
Applications
Embryo centrifugation is particularly useful for colocalization experiments, i.e., to test if a particular protein localizes to a certain major organelle. For example, the Klar protein is present on early embryonic lipid droplets, yet this localization is challenging to demonstrate in intact embryos. In part, this is because both Klar puncta and lipid droplets are abundant in the embryo periphery 10, making spurious colocalization hard to rule out. However, centrifugation concentrates lipid droplets into a distinct layer, and thus, colocalization with Klar signal becomes obvious 10.
Similar analysis has unambiguously demonstrated the localization of other proteins to lipid droplets 8, 13-15, including surprising examples like certain histones 2 and in cases where a protein is present ubiquitously but enriched on lipid droplets 8. As embryos can be selected visually by developmental stage (step 4.2), it is even possible to detect developmentally regulated recruitment of proteins to lipid droplets 8, 15. Testing for colocalization with lipid droplets is particularly easy since lipid droplets accumulate in a visually distinct “cap”, so no further marker or stain is required to identify this layer. In fact, because of its ease and potentially striking visual results (Fig. 1, Fig. 5 D, E), our laboratory employs this technique now routinely as a first test whether a candidate protein is localized to lipid droplets. In principle, similar colocalization studies should work for the other organelles shown in Fig. 1, and likely for others (including intracellular parasites) whose distribution in centrifuged embryos has yet to be documented.
Since centrifugation concentrates organelles into tight bands, it facilitates isolation of a fraction enriched in these organelles. This approach has already been used to recover lipid droplets for transplantation into donor embryos 2. It may also be possible to biochemically analyze the enriched fraction (e.g., by cutting layers off embryos after fixation 15).
The method described here has applications beyond Drosophila embryos, as centrifugation can be used to stratify the eggs of many different species (including frogs, nematodes, tunicates, annelids, sea urchins, mollusks, and Cnidarians) 16. In our own laboratory, we have used the protocol above for embryos of the housefly Musca domestica 2 and have employed unoriented centrifugation in the analysis of embryos from the fungus gnat Sciara coprophila 17 and the snail Ilyanassa obsoleta 17. Finally, this method is not restricted to embryos: Centrifugation can also be useful to stratify other large cells, such as oocytes (e.g., in Drosophila 2, 18 and the milkweed bug Oncopeltus fasciatus 17). For different specimen, the exact centrifugation conditions need to be optimized: For example, 10 minutes at 3000 g is sufficient to efficiently stratify embryos of Musca domestica 2, and if centrifuged much longer, the embryos tend to break apart during fixation and immunostaining.
The authors have nothing to disclose.
We thank Susan Gerbi, Heidi Smith, and David Lambert for supplying material to test in-vivo centrifugation on Sciara coprophila and Ilyanassa obsoleta, respectively. We thank members of the Welte lab for comments on the technique and manuscript. Development and refinement of the in-vivo centrifugation assay was supported by NIGMS grant GM64687 to MAW. The images shown in Fig. 1 and Fig. 5E,F were previously published2 and are reprinted here with permission.
Material Name | Type | Company | Catalogue Number | Comment |
---|---|---|---|---|
60×15 mm Petri dishes | Becton Dickinson | #35-1007 | From Falcon | |
Wire mesh | Small Parts | CX-0150 | stainless steel type 304; mesh size #150 x150, wire diameter: 0.0026 inches | |
Tungsten needles | Fine Science Tools | 10130-10 | 0.25 mm | |
Moria Nickel Plated Pin/Needle Holder | Fine Science Tools | 26016-12 | ||
Fly Cages | Genesee Scientific | 59-100 or 59-101 | For 65 mm or 100 mm diameter Petri dishes, respectively | |
Halocarbon Oil 27 | Sigma | H8773 |