Drosophila and mammalian hematopoietic systems share many common features, making Drosophila an attractive genetic model to study hematopoiesis. Here we demonstrate dissection and mounting of the major larval hematopoietic organ for immunohistochemistry. We also describe methods to assay various larval hematopoietic compartments including circulating hemocytes and sessile crystal cells.
Many parallels exist between the Drosophila and mammalian hematopoietic systems, even though Drosophila lack the lymphoid lineage that characterize mammalian adaptive immunity. Drosophila and mammalian hematopoiesis occur in spatially and temporally distinct phases to produce several blood cell lineages. Both systems maintain reservoirs of blood cell progenitors with which to expand or replace mature lineages. The hematopoietic system allows Drosophila and mammals to respond to and to adapt to immune challenges. Importantly, the transcriptional regulators and signaling pathways that control the generation, maintenance, and function of the hematopoietic system are conserved from flies to mammals. These similarities allow Drosophila to be used to genetically model hematopoietic development and disease.
Here we detail assays to examine the hematopoietic system of Drosophila larvae. In particular, we outline methods to measure blood cell numbers and concentration, visualize a specific mature lineage in vivo, and perform immunohistochemistry on blood cells in circulation and in the hematopoietic organ. These assays can reveal changes in gene expression and cellular processes including signaling, survival, proliferation, and differentiation and can be used to investigate a variety of questions concerning hematopoiesis. Combined with the genetic tools available in Drosophila, these assays can be used to evaluate the hematopoietic system upon defined genetic alterations. While not specifically outlined here, these assays can also be used to examine the effect of environmental alterations, such as infection or diet, on the hematopoietic system.
The complex mechanisms regulating the transcription factors and signaling pathways that coordinate the development of the hematopoietic system and that malfunction in hematological diseases remain poorly understood. These transcription factors and signaling pathways, as well as their regulation, are highly conserved between Drosophila and mammalian hematopoiesis1-5. Thus the Drosophila hematopoietic system represents an excellent genetic model to define the molecular mechanisms controlling hematopoiesis and underlying hematological diseases.
Similar to mammals, Drosophila generate blood cells, called hemocytes, in spatially and temporally distinct phases of hematopoiesis. Traditionally, Drosophila hematopoiesis was thought to be restricted to phases in the embryonic mesoderm and in the larval lymph gland. Recent studies provide evidence that hematopoiesis also occurs in larval sessile clusters and in the adult abdomen6-8. All hematopoietic phases produce two types of mature hemocytes: plasmatocytes and crystal cells. Plasmatocytes are macrophage-like cells involved in phagocytosis, innate immunity, and wound healing. Crystal cells contain pro-phenoloxidases required for melanization, a reaction used in insect immune responses and wound healing. Larval hematopoiesis can generate a third mature hemocyte type, called a lamellocyte, in response to certain immune challenges such as parasitoid wasp infection9,10. Lamellocytes are large, adherent cells that function, in conjunction with plasmatocytes and crystal cells, to encapsulate and neutralize wasp eggs laid in Drosophila larvae. In the absence of parasitization, lamellocytes are not found in wild-type larvae. Melanotic masses resemble melanized, encapsulated wasp eggs; many mutant Drosophila strains develop melanotic masses in the absence of parasitization. The presence of lamellocytes and/or melanotic masses can be indicative of hematopoietic abnormalities. In fact, the melanotic mass phenotype has been used to identify genes and pathways involved in hematopoiesis11-14.
The larval hematopoietic system is the most extensively studied to date. It is comprised of hemocytes circulating in the hemolymph, sessile hemocyte clusters patterned under the cuticle, and hemocytes residing in the lymph gland. The lymph gland is a series of bilateral lobes attached to the dorsal vessel. Each primary lobe of the lymph gland is divided into three main zones. The outermost zone is known as the cortical zone and contains maturing hemocytes. The innermost zone is called the medullary zone and is comprised of quiescent hemocyte precursors. The third zone, the posterior signaling center, is a small group of cells at the base of the lymph gland that act as a stem cell-like niche. Early work established critical functions for Notch15-18, Hedgehog19,20, JAK-STAT18, and Wingless21 activity to regulate larval lymph gland development. More recent studies have demonstrated that BMP22, FGF-Ras23, and Hippo24,25 signaling also function within the larval lymph gland.
Four larval hematopoietic assays outlined here describe 1) measuring circulating hemocyte concentration, defined as number of cells per unit volume, 2) isolating and fixing circulating hemocytes for immunohistochemistry, 3) visualizing crystal cells in vivo, and 4) dissecting, fixing, and mounting lymph glands for immunohistochemistry. These assays can be used as hematopoietic readouts to assess the functions and regulations of signaling pathways in the larval hematopoietic system. While these methods have been used previously in the field, visual documentation of these assays has begun only recently8,26-30. Several publications cited here are helpful resources describing similar methods and hematopoietic markers26,31-33. Additionally, Trol and Viking are useful markers of the lymph gland basement membrane.
1. Circulating Hemocyte Concentration
2. Circulating Hemocyte Immunohistochemistry
3. In Vivo Crystal Cell Melanization
4. Larval Lymph Gland Immunohistochemistry
NOTE: The lymph gland is located approximately one-third length from the anterior end of a larva slightly below the brain on the dorsal side. (See arrow in Figure 3B.) The lymph gland flanks the dorsal vessel and is most easily dissected attached to the mouth hooks or to the brain. Wild-type, third instar lymph glands are very small structures; the primary lobes are approximately 100 – 200 µm in length. (See Figure 4A.)
5. Imaging
Solution | Composition | Storage | Yorumlar |
1x PBS | 200 mg/L potassium chloride | room temperature | |
200 mg/L potassium phosphate monobasic | |||
8,000 mg/L sodium chloride | |||
1,150 mg/L sodium phosphate dibasic | |||
dH2O | |||
Fixative | 3.7% or 7.5% formaldehyde in 1x PBS | room temperature in the dark | Formaldehyde is toxic. |
Permeabilization solution/antibody diluent | 0.4% Triton | 4 °C | The standard formula uses 0.4% Triton but the authors use 0.1% Tween 20 with success. Use to dilute primary and secondary antibodies according to providers' recommended concentrations. |
5% bovine serum albumin, normal goat serum, or normal donkey serum | |||
1x PBS | |||
70% ethanol | 70% 200 proof ethanol in dH2O | room temperature | |
Mounting buffer | 0.5% N-propyl gallate | 4 °C in the dark | N-propyl gallate is harmful. DAPI is a mutagen. |
80% glycerol | |||
Optional: 1 μg/ml DAPI (4’,6-diamidino-2-phenylindole) | |||
1x PBS | |||
0.1% PBST | 0.1% Tween 20 in 1x PBS | room temperature | |
0.01% PBST | 1:10 dilution of 0.1% PBST | room temperature |
Table 1. Solutions used in this Protocol.
Circulating Hemocyte Concentration
Hemocyte numbers increase throughout larval development35. To illustrate that this method detects differences in hemocyte numbers and concentration, regardless of the biological cause, we measured hemocyte concentrations of delayed and non-delayed larvae. Loss of prothoracicotropic hormone (ptth) by genetic ablation of ptth-producing neurons (ptth>grim) produces a delay in larval development36. For each genotype, hemocyte concentrations were measured as described in Protocol 1 for at least 8 individual larva raised at 25 °C. At 120 hr after a 2 hr egg collection, the average hemocyte concentration per delayed larva (ptth>grim) is less than the average hemocyte concentration per control larva (ptth). Only after 9 days does the average hemocyte concentration per delayed larva approach that of controls (Figure 1B). Another example in which this method is used to detect differences in hemocyte concentrations has been published37.
Images taken from an automated cell counter show a clean, desirable hemolymph sample and a sample containing debris, which might lead to inaccurate measurement (Figure 1C). Minimum and maximum cell sizes were set to 2 µm and 22 µm, respectively. Circularity was set to 75-80% roundness. These parameters are intended as guidelines and should be empirically optimized.
Circulating Hemocyte Immunohistochemistry
Circulating hemocytes expressing green fluorescent protein (GFP) were collected, fixed, and incubated with a mixture of plasmatocyte-specific antibodies (P1a and P1b, István Andó)31 as described in Protocol 2 (Figure 2). The image was taken on a standard fluorescence microscope and is shown unaltered and after constrained iterative deconvolution. In this instance, deconvolution did not drastically improve the image quality.
In Vivo Crystal Cell Melanization
Larvae were placed at the bottom of PCR tubes prior to heat-induced crystal cell melanization as described in Protocol 3 (Figure 3A). Red arrows indicate tubes in which larvae are too far from the bottom and might be heated unevenly. Uneven distribution of heat across individual larva can increase variability in the melanization of crystal cells within the larva.
A wild-type larva imaged on a standard stereomicroscope shows the typical pattern of melanized crystal cells in sessile clusters after heat exposure (Figure 3B). Melanized crystal cells in the lymph gland are sometimes seen.
Larval Lymph Gland Immunohistochemistry
Third instar larval lymph glands were dissected, fixed, and mounted as described in Protocol 4. A differential interference contrast (DIC) image taken on a standard fluorescence microscope shows the primary and secondary lymph gland lobes flanking the dorsal vessel (Figure 4A).
A lymph gland in which the medullary zone and the posterior signaling center were genetically marked with enhanced blue fluorescent protein (EBFP2) and GFP, respectively, was stained with an antibody against the Notch intracellular domain (C17.9C6, Developmental Studies Hybridoma Bank) as described in Protocol 4. Z-stack images were taken on a standard fluorescence microscope. A single maximum intensity projection image appears unaltered and after constrained iterative deconvolution (Figure 4B). In this instance, deconvolution drastically improved the quality and detail of the image.
Figure 1. Circulating Hemocyte Concentration. A) A portion of standard Drosophila medium was removed (right) to promote egg laying during a 2 hr egg collection period. B) The average hemocyte concentration per developmentally delayed larva (ptth>grim) was lower than the average hemocyte concentration per control larvae (ptth) 120 hr after egg laying (AEL). The average hemocyte concentration per delayed larva approached control level 9 days AEL. Error bars represent ± s.e.m. C) Hemolymph samples without (left) and with (right) debris. Scale bars represent 0.1 mm. Please click here to view a larger version of this figure.
Figure 2. Circulating Hemocyte Immunohistochemistry. Hemocytes genetically expressing GFP were collected and fixed to a coverslip. A plasmatocyte-specific antibody was used to stain plasmatocytes (red). DAPI staining is shown in blue. Unaltered image taken with a fluorescence microscope (left). The same image after deconvolution (right). Scale bars represent 50 µm. Please click here to view a larger version of this figure.
Figure 3. In Vivo Crystal Cell Melanization. A) Larvae were placed individually in PCR tubes prior to heating. Red arrows indicate larvae that are not at the bottom of the tubes. B) A typical crystal cell melanization pattern after heating, which sometimes reveals the lymph gland (arrow; dorsal side shown, anterior is left). Scale bar represents 1 mm. Please click here to view a larger version of this figure.
Figure 4. Larval Lymph Gland Immunohistochemistry. A) A DIC image of a third instar larval lymph gland showing the dorsal vessel (dv), primary lobes (1°), and secondary lobes (2°). Scale bar represents 100 µm. B) A representative third instar larval lymph gland from a larva genetically expressing EBFP2 in the medullary zone and GFP in the posterior signaling center. The lymph gland was stained with an antibody against the Notch intracellular domain (red). Unaltered image taken with a fluorescence microscope (top). The same image after deconvolution (bottom). Scale bars represent 20 µm. Please click here to view a larger version of this figure.
Upon genetic or environmental alteration, the four methods described here can be used individually or in conjunction to analyze distinct processes during hematopoiesis such as signaling, survival, proliferation, and differentiation. Drosophila hematopoiesis is a dynamic process; the number of hemocytes per animal increases35 and the structure and gene expression of the lymph gland changes32 during development. Prior to performing these assays, therefore, it is critical to restrict egg collections by allowing females to lay eggs for a fixed amount of time and to confirm the desired larval developmental stage. For shorter egg collections (less than 6 hr) or for instances in which females are unhealthy or scarce, egg laying can be encouraged by creating crevices in the food. This can be accomplished by removing a portion of food with an ethanol-cleaned spatula. If excess liquid accumulates in the food, use tissue wipes to soak up the liquid. (See Figure 1A.)
Alone, the methods described here have limitations. For instance, measuring circulating hemocyte concentration captures changes in hemocyte survival or proliferation but provides no information about hemocyte lineage distribution or any lineage disruption that might be consequent to genetic or environmental alteration. Conversely, immunohistochemistry of circulating hemocytes reveals changes in specific hemocyte lineages in the hemolymph but only in relative, not absolute, numbers. Crystal cell melanization in vivo is difficult to quantitate as crystal cell distribution is variable and sessile hemocytes are dynamic8,10. Rather, multiple individuals should score crystal cell melanization blindly. Finally, observations made in one hemocyte compartment might not necessarily hold true in other compartments.
When used together, these assays can be applied to distinguish genetic alterations that regulate proliferation or survival from those that regulate gene expression or differentiation. For instance, a genetic alteration can increase proliferation of hemocytes such that hemocyte concentration increases but the relative proportions of hemocyte lineages remains the same. Alternately, a genetic alteration can promote changes in hemocyte differentiation into specific lineages with no effect on overall hemocyte concentration. The crystal cell population can be interrogated quickly and easily using the in vivo melanization method, facilitating genetic screens and genetic interaction studies. This method can be used in corroboration with genetic studies that utilize prophenoloxidase 1 (PPO1, also called black cells, Bc) mutant alleles or immunohistochemical assays that utilize crystal cell-specific antibodies. The larval lymph gland can be used to address similar questions regarding hematopoietic cellular processes. Numerous signaling pathways are involved in establishing and maintaining the lymph gland and its major zones. Each zone has distinct gene expression and function in larval hematopoiesis. Additionally, utilizing the lymph gland can reveal autonomous and non-autonomous functions within the lymph gland zones or lymph gland hemocytes.
Relevant methods for studying Drosophila hematopoiesis have been compiled in comprehensive text-based resources only recently26,27. Though indispensable to the field, these resources are limited in scope. Methods for measuring hemocyte concentration, for example, were not included. Written methods, especially those describing dissection techniques, can be challenging to master quickly. Additional contributions have been made, providing visual resources to assist with methods, but were still limited in number and in scope28,29. The methods described here, while also not comprehensive, add to the resources available to aid in the study of Drosophila hematopoiesis.
We offer modifications and alternatives to existing protocols. For example, there are several advantages to measuring hemocyte concentration with an automated cell counter rather than a manual hemocytometer, including increased speed, ease, and objectivity. In our experience, using a water bath to heat larvae for crystal cell visualization resulted in widely variable results, especially if the larvae are heated in the vials in which they are raised. Heating larvae in individual PCR tubes yielded less variable and more reproducible results. The lymph gland dissection method described here, though previously outlined in a written protocol26, provides an alternative to existing visual references28. Finally, if access to a confocal microscope is limited, we suggest that deconvolution of standard fluorescence images might be a suitable substitute for confocal images. Deconvolution algorithms either remove or reassign out-of-focus light captured by conventional fluorescence microscopy, improving resolution and contrast similar in principle to the elimination of out-of-focus light by confocal microscopy, and can be applied to 2-dimensional and 3-dimensional (Z-stack) images. For some applications, as demonstrated here, deconvolution can dramatically improve images taken with conventional fluorescence microscopes.
The authors have nothing to disclose.
We thank Matthew O’Connell, Maryam Jahanshahi, and Andreas Jenny for assistance. We thank István Andó for plasmatocyte-specific antibodies, Utpal Banerjee for dome-meso-EBFP2 flies, Julian Martinez-Agosto for antp>GFP flies, and Michael O’Connor for ptth and ptth>grim flies. These methods were developed with support by the Kimmel Foundation, the Leukemia & Lymphoma Society, NIH/NCI R01CA140451, NSF 1257939, DOD/NFRP W81XWH-14-1-0059, and NIH/NCI T32CA078207.
PBS tablets | MP Biomedicals | 2810305 | |
dissecting dish | Corning | 7220-85 | |
microcentrifuge tube | Denville | C2170 | |
silicone dissecting pad, made from Sylgard 184 kit | Krayden (distributed through Fisher) | NC9644388 (Fisher catalog number) | Made in petri dish by mixing components of Sylgard elastomer kit according to manufacturer instructions. |
stereomicroscope | Morrell Instruments (Nikon distributor) | mna42000, mma36300 | Nikon models SMZ1000 and SMZ645 |
tissue wipe | VWR | 82003-820 | |
forceps | Electron Microscopy Sciences | 72700-DZ | |
p200 pipette | Eppendorf | 3120000054 | |
Countess Automated Cell Counter | Invitrogen | C10227 | |
Countess cell counting chamber slides | Invitrogen | C10283 | |
hemocytometer | Hausser Scientific | 3200 | |
trypan blue stain | Life Technologies | T10282 | |
formaldehyde | Fisher | BP531-500 | |
Triton | Fisher | BP151-500 | |
Tween 20 | Fisher | BP337-500 | |
bovine serum albumin | Rocky Mountain Biologicals | BSA-BSH-01K | |
normal goat serum | Sigma | G9023-10ML | |
normal donkey serum | Sigma | D9663-10ML | |
200 proof ethanol | VWR | V1001 | |
N-propyl gallate | MP Biomedicals | 102747 | |
glycerol | VWR | EM-4750 | |
DAPI (4’,6-diamidino-2-phenylindole) | Fisher | 62248 | |
6-well plate | Corning | 351146 | |
12-well plate | Corning | 351143 | |
microscope cover glass, 22 mm square | Fisher | 12-544-10 | |
microscope cover glass, 18mm circular | Fisher | 12-545-100 | |
glass microscope slides | Fisher | 22-034-980 | |
thermal cycler | Eppendorf | E950010037 | Mastercycler EP Gradient S |
PCR tubes | USA Scientific | 1402-2700 | |
24-well plate | Corning | 351147 | |
disposable transfer pipet | Fisher | 13-711-9AM | |
fluorescence microscope | Zeiss | Axio Imager.Z1 |