We present a technique for labeling single neurons in the central nervous system (CNS) of Drosophila embryos, which allows the analysis of neuronal morphology by either transmitted light or confocal microscopy.
In this article we describe how to individually label neurons in the embryonic CNS of Drosophila melanogaster by juxtacellular injection of the lipophilic fluorescent membrane marker DiI. This method allows the visualization of neuronal cell morphology in great detail. It is possible to label any cell in the CNS: cell bodies of target neurons are visualized under DIC optics or by expression of a fluorescent genetic marker such as GFP. After labeling, the DiI can be transformed into a permanent brown stain by photoconversion to allow visualization of cell morphology with transmitted light and DIC optics. Alternatively, the DiI-labeled cells can be observed directly with confocal microscopy, enabling genetically introduced fluorescent reporter proteins to be colocalised. The technique can be used in any animal, irrespective of genotype, making it possible to analyze mutant phenotypes at single cell resolution.
Knowledge of neuronal morphology at the level of individual cells is a key prerequisite for understanding neuronal connectivity and CNS function. Thus, from the earliest days of neuroscience, researchers have sought to develop single cell labeling techniques (see 7 for a historical treatment of this issue). Classical methods such as Golgi staining provide excellent resolution of neuronal morphology but are not suitable if one seeks to label a particular type of neuron in a directed way, as staining occurs in a random fashion. The development of methods for staining single cells by intracellular or juxtacellular injection of dyes from a microelectrode addressed the requirement of specific labeling.
The application of single neuron dye injection to Drosophila presented a major challenge, because of the small size of both the organism and its neurons. Nonetheless, single neuron staining of embryonic Drosophila neurons was achieved in the mid-1980s in the laboratory of Corey Goodman 15. While the method has been eminently useful and has provided some key insights into mechanisms for neuronal development in Drosophila over the last 20-30 years (e.g. 2, 10), many workers have shied away from it, largely because of its technical demands.
The availability in more recent years of genetic techniques for neuronal labeling in Drosophila has also contributed to the unpopularity of single neuron dye injection. GAL4-directed expression of membrane-targeted GFP constructs can provide excellent resolution of neural morphology 1, 20. However, this method has certain limitations: the often-unavoidable expression of GFP in multiple cells can obscure the structure of individual neurons and a GAL4 driver line may not be available to drive expression in a particular neuron of interest. The MARCM (Mosaic Analysis with a Repressible Cell Marker) method 8 can provide labeling of essentially any neuron at the individual cell level, but cannot be successfully used in the embryo and early larva because of the slow turnover of the GAL80 protein.
Given these limitations of genetic labeling, we believe that single neuron dye injection in the Drosophila embryo remains a valuable technique and deserves broader application. To promote this goal, we provide here a detailed description of the method. An illustration of its power is provided by our recent account of the morphology of the complete set of interneurons in abdominal neuromeres of the late Drosophila embryo 14. This study, which revealed both the morphological variability of individual neuronal cell types and the principles of neuromere organization in the CNS of the embryo, would not have been possible with any other currently available labeling method.
We have used two slightly different variants of the single neuron labeling method in our laboratories. The differences relate to the embryo collection, dechorionisation, devitellinisation and embryo filleting steps. Figure 1 gives an overview of the common and diverging steps of the variants.
1. Preparation of Micro-needles for Embryo Dissection and Micropipettes for Dye Injection
2. Collection, Mounting and Dissection of Embryos
Devitellinisation and filleting are done manually using either of the two methods described in 2.3A and 2.3B.
3. Filling of Injection Micropipettes
Half fill a 0.5 ml Eppendorf microfuge tube with a 0.1% solution of the carbocynanine dye DiI (Molecular Probes, Eugene, OR) in 100% ethanol (be sure to use ‘dry’ absolute EtOH to avoid DiI precipitation). Make a narrow hole in the lid of the tube and insert the blunt end of the micropipette through the hole in the lid. Allow the DiI to ascend up the filament for at least 5 min. (Always cover the hole in the lid when not filling a micropipette to avoid evaporation of ethanol).
Equipment required for neuron dye injection (Figure 3).
Microscope. A fixed stage microscope must be used for neuron dye injection to avoid vertical movement of the embryo during focusing. Any such movement would displace the pipette after it has come into contact with the embryo. We have found both the Zeiss Axioskop FS and the Olympus AX50/BX50 fixed stage models to be well suited for this purpose. The microscope should be equipped with transmitted light DIC optics for visualization of neurons before staining. The microscope should also be an upright rather than an inverted model. With an upright microscope, the tip of the micropipette is on the same side of the embryo as the viewing objective, whereas with an inverted microscope the two are on opposite sides of the embryo. The latter arrangement compromises the quality of DIC optics. The microscope should be set up for fluorescence microscopy, with a suitable filter set for DiI observation. Since DiI has a comparably broad spectrum with excitation and emission maxima at 549 and 565 nm many filter sets in this range will work, e.g. those for Alexa 568, Cy3, Rhodamine, Texas Red or TRITC. Although it is convenient to fit an electronically controlled shutter in the path of the fluorescence light source, to allow operation of the shutter without hand contact with the microscope, we also effectively labeled on a setup which was only equipped with a manual shutter. Finally, a high magnification, high numerical aperture water immersion objective is required for observation during injection. This objective should be designed for use without a coverslip. We have successfully used the Zeiss Achroplan 100x/1.0W, Olympus LUMPlan FL 100x/1.0W, and Olympus LUMPlan FL/IR 60x/0.9 W objectives.
A micromanipulator is required to position and move the micropipette during neuron dye injection. We have used both the Leica micromanipulator and a stage-mounted Narashige 3-axis hydraulic micromanipulator.
An intracellular DC amplifier, with the facility for current injection, is required for iontophoretic injection of DiI from the micropipette.
4. Procedure for Dye Injection
5. Photoconversion for Examination Using DIC Optics
The principle of photoconversion is the (photo-)oxidation of DAB by the fluorescent light emitted by the stained cell during illumination with the appropriate wavelength. This means bright cells are photoconverted in shorter time compared to weakly stained cells. If multiple cells labeled in one specimen differ too much regarding their brightness this can lead to difficulties in photoconverting all of them in same quality (see Figure 4C): in this case one may have to decide on a compromise between neglecting the weaker cells (using short illumination) or accepting that the brighter cells begin to swell (using longer illumination). If all cells are too weakly labeled (thus needing very long illumination), background caused by endogenous peroxidases can become a problem. Since DAB is toxic it should be handled with care. Waste can be ‘deactivated’ with 7% chlorine bleaching.
6. Examination by Confocal Microscopy
Figure 4 illustrates typical results of the technique, we describe here. Figure 4A shows an example of a DiI filled single interneuron that was cleanly photoconverted. It nicely demonstrates the amount of detail these preparations offer. When viewed under DIC optics the spatial context of the labeled cell within the non-labeled surrounding tissue becomes visible, e.g. the position of the cell body within the cortex and of the fiber projection within the neuropile.
Figure 4B is a case where the dye drop was a little too big resulting in several neighboring cells becoming labeled simultaneously. This often makes it difficult to relate individual projections to distinct cell bodies. It also shows that when directly viewed under the fluorescent microscope (no photoconversion) background resolution is much lower.
The specimen in Figure 4C illustrates a case where multiple cells being were individually labeled in neighboring segments. Upon photoconversion the preparation was subsequently stained with an antibody against Fas2 5following standard protocols 13 to provide landmarks (fascicles) in the neuropile. Although in this preparation most cell bodies are well separated, and allowing reliable mapping of their projections, it also demonstrates what may occur when the photoconversion period is prolonged (in order to also convert the weakly labeled cells): the more intensely labeled cells tend to get overstained and start to swell (compare with A).
Figure 4D shows that observation by confocal microscopy reveals the morphology of labeled cells in detail. DiI labeling was performed in a strain that carried a GFP reporter construct. This may provide important information about the spatial context (and identity) of the dye filled cell within specific populations of neuronal or glial cell types (as defined by gene expression).
As a paradigm to evaluate morphological variability of an identified neuron we have chosen one of the Apterous positive neurons 14. This neuron lies in a dorsal and medial position close to the neuropile (dAP, Figure 5A, Lundgren et al., 1995) separate from the other Apterous positive neurons and is therefore easily identifiable in apGal4-UASGFP animals. Figure 5B shows a dAP cell that was filled with DiI and photoconverted. The cell body lies at the level of the anterior commissure and shows an axon that first grows towards the midline and then turns anterior in a medial connective region. It has growth cone like swellings at the tip and the turning point. 19 labels of this cell, drawn as maximum projections from image stacks, are summarized in Figure 5C. Cell morphology becomes visible to great detail. Only few cells show growth cone like structures; six of the 19 cells also send a branch posterior that is always shorter than the anterior branch and may well be a transient feature. In Figure 5D all 19 dAP cells labeled are stacked with each cell having only an opacity of 12%. This results in areas being darker the more cells share them. While the cell body varies more than one cell diameter around the center position, the mediolateral position of the axon within the neuropile varies much less – when the neuropile width is divided into nine parts it resides always in one of the medial two positions.
Figure 1. Schematic overview of the steps of the neuron labeling method and its two variants.
Figure 2. Diagram illustrating the various steps in preparing Drosophila embryos for neuron dye injection, according to variant B. Click here to view larger figure.
Figure 3. Schematic representation of the setup for dye injection. The upper right insert is an enlarged view of the arrangement of the specimen, the injection micropipette and the bath electrode. A: Micromanipulator, B: Fixed stage microscope. The micromanipulator is set at a shallow angle – approximately 10°. C: Micropipette holder with micropipette, D: Specimen on slide, E: Bath electrode fixed with plasticine, F: Connection to DC amplifier.
Figure 4. Examples of DiI-labeled interneurons in the ventral nerve cord of stage 17 embryos. Dorsal views, anterior is to left. Midline: dotted lines.
Figure 5. Targeted DiI labeling (marked by GFP expression) reveals the range of morphological variability of an identified neuron 14. Click here to view larger figure.
One major advantage of Drosophila as a model system is that it allows analysis of development and function on the level of single cells. This is especially helpful regarding the nervous system, where the diversity of cell types is exceptionally high and the function and morphology of neighboring cells can be totally different.
The method we present here allows the labeling of individual neurons with a dye that can either be transformed into a permanent stain or examined directly by fluorescence microscopy. It reveals the morphology of neural processes in great detail (Figure 4). One of the major advantages of the method is that it enables the morphology of virtually any type of neuron in the CNS to be examined (for the embryonic ventral nerve cord, see Rickert et al., 2011; for the embryonic brain, see Kunz et al., 2012). In addition, it does not require complex combinations of genetic elements to be present in the embryo, as several of the genetic labeling techniques do. Dye injections may be performed into cells of animals that carry reporter constructs, thereby allowing individual cell morphologies to be visualized in the context of particular gene expression patterns (Figures 4D and 5) in either wild type or mutant backgrounds. We have also used the technique to monitor the dynamics of axon outgrowth in living embryonic neurons 11, 12.
The method can be readily adapted to label individual neurons in embryos of other organisms, provided the embryo is transparent enough to be examined under DIC illumination. We have used it to visualize neural morphology in a wide variety of arthropod embryos, including grasshoppers 18, centipedes 17, crustaceans 19 and silverfish 16.
The major disadvantage of the technique lies in its technical difficulty. We trust that the current detailed account of the method will assist interested researchers in mastering it.
The authors have nothing to disclose.
This work was supported by a grant from the DFG to G.M.T.
Name of Reagent/Material | Company | Catalogue Number | Comments |
REAGENTS | |||
DAB | Sigma-Aldrich | D-5905 | 2-3 mg/ml in 100 mM TRIS-Hcl, pH 7.4 |
DiI | Invitrogen/Molecular Probes | D-282 | 1 mg/ml in ethanol |
Formaldehyde | Merck Millipore | 7,4 % in PBS | |
Glycerol | Roth | 3783 | 70% in PBS |
Heptane Glue | Beiersdorf AG | Cello 31-39-30 * | dilute ca. 1 to 1 with n-Heptane |
PBS | 1x | ||
TRIS | Roth | 4855 | |
Vectashield | Vector Laboratories | H1000 | |
EQUIPMENT | |||
Confocal Microscope | Leica | TCS SP2 | to view and document labelings in fluorescence |
DC – amplifier | Dragan Corporation | Cornerstone ION-100 | to perform the labelings |
Coverslips | Menzel | BB018018A1 | 18 x 18 mm |
Coverslips | Menzel | BB024060A1 | 24 x 60 mm |
Dissecting Microscope | Leica | MZ8 | to prepare and disect embryos |
Flat Capillaries | Hilgenberg | outer diameter 1 mm; glas thickness 0.1 mm | |
Injection Capillaries | Science Products | GB 100 TF 8P | |
Micromanipulator R | Leica | to perform the labelings | |
Model P-97 | Sutter Instruments | to pull capillaries | |
Object Slides | Marienfeld Superior | 1000000 | |
Scientific Microscope | Zeiss | Axioskop 2 mot | to view labelings after photoconversion |
Scientific Microscope | Olympus | BX50 | to perform the labelings |
Sony MC3255 Video Camera | Sony/AVT Horn | Sony MC3255 | to record labelings after photoconversion |
Table 1.