A protocol for cell cycle analysis of live Drosophila tissues using the Attune Acoustic Focusing Cytometer is described. This protocol simultaneously provides information about relative cell size, cell number, DNA content and cell type via lineage tracing or tissue specific expression of fluorescent proteins in vivo.
Flow cytometry has been widely used to obtain information about DNA content in a population of cells, to infer relative percentages in different cell cycle phases. This technique has been successfully extended to the mitotic tissues of the model organism Drosophila melanogaster for genetic studies of cell cycle regulation in vivo. When coupled with cell-type specific fluorescent protein expression and genetic manipulations, one can obtain detailed information about effects on cell number, cell size and cell cycle phasing in vivo. However this live-cell method has relied on the use of the cell permeable Hoechst 33342 DNA-intercalating dye, limiting users to flow cytometers equipped with a UV laser. We have modified this protocol to use a newer live-cell DNA dye, Vybrant DyeCycle Violet, compatible with the more common violet 405nm laser. The protocol presented here allows for efficient cell cycle analysis coupled with cell type, relative cell size and cell number information, in a variety of Drosophila tissues. This protocol extends the useful cell cycle analysis technique for live Drosophila tissues to a small benchtop analyzer, the Attune Acoustic Focusing Cytometer, which can be run and maintained on a single-lab scale.
Flow cytometry can be used for measurements of cell viability, relative cell size, DNA content and fluorescent protein expression in live cell populations. Due to the replication of nuclear DNA during S-phase, information about DNA content in a population of cells can be used to infer relative percentages in different cell cycle phases 1-3. This method has become a cornerstone of cell cycle analysis in model systems from yeast to mammals.
The fruit fly Drosophila melanogaster has become an excellent model system for genetic in vivo analyses of cell cycle regulation. The extensive genetic tools available in flies allow for elegant tissue specific and temporally regulated manipulations of cell cycle regulators along with in vivo fluorescent protein-based lineage tracing 4-6. Flow cytometry has been used to study DNA content in a number of Drosophila cell types, including endoreplicating cells and cultured mitotic cells 7,8. An important advance for in vivo cell cycle studies was made by de la Cruz and Edgar, with the development of a protocol for flow cytometric analysis of live diploid Drosophila imaginal discs 9,10, a protocol which has been used and adapted by many labs. This technique, when coupled with genetic in vivo lineage tracing via inducible fluorescent protein expression and tissue specific labeling, allows one to obtain information about gene manipulation effects on overall cell doubling time, cell size and to determine precise timing of cell cycle phases in vivo 9,11. However this method has thus far relied on the use of the cell permeable Hoechst 33342 DNA-intercalating dye to stain and quantify DNA in live cells, which has limited users to flow cytometers with a UV laser capable of exciting the Hoechst dye. These are generally found only in sorters (i.e. BD FACS Vantage, BD FACSAria) or expensive multicolor benchtop systems (i.e. BD LSR), usually requiring support by institutional flow core facilities.
We have modified the Hoechst-based protocol to use a new live-cell DNA dye from Invitrogen, Vybrant DyeCycle Violet. This dye is compatible with a violet 405 nm laser, more common in smaller benchtop analyzers and available in the small self-contained benchtop analyzer, the Attune Acoustic Focusing Cytometer. Here we present a detailed protocol for cell cycle analysis that can be coupled with cell type, cell size, cell number and lineage analysis in a variety of Drosophila tissues during various stages of development using DyeCycle Violet and the Attune. This protocol expands the number of cytometers suitable for such analysis with Drosophila tissues and provides examples of how this type of live cell cycle analysis can be modified for additional tissue types and developmental stages.
1. Fly Husbandry
The rate of cell division and timing of dissection determines clone size in each tissue. Under normal conditions, we find that clones induced in the larval wing with a 20 min heat shock (using the hs-flp transgene on the second chromosome) and dissected 48 hr later, contain approximately 20-30 well separated clones with sizes ranging from 10 cells per clone to two cells per clone, with an average around four cells per clone. In contrast, a 7 min heat-shock with the same hs-flp transgene of pupae in a Petri dish at 0 hr APF dissected 36 hr later, yields well separated clones (approximately 15-25) ranging from one to four cells with an average of 2.2 cells per clone. Such data can be used to determine the average cell doubling time for the tissue under study.
2. Dissection
3. Tissue Dissociation and DNA Staining
4. Flow Cytometry
5. Data Analysis
Figure 2 shows representative results for a larval wing sample, expressing GFP in the posterior half of the tissue, using the provided GFP template. Similar results are obtained with the same tissue type and expression pattern for RFP using the provided RFP template (Figure 3A). The provided templates and voltages (Table 2) are suitable for analysis of larval eyes (Figure 3B), brains and wings, as well as pupal eyes, brains (Figure 3D) and wings. Histograms of relative cell size provided by plotting FSC for populations in Gates 5 and 6 can also be generated (Figure 3C). Gates may need to be adjusted slightly due to differences in cell size for different tissues or differences in fluorescence intensity for different GFP or RFP transgenes. This can be done while running a small test volume of sample (50-100 μl), before hitting record.
Allowing the y-axis (counts) for the cell cycle or cell size histograms to auto-scale (y-axis scale set to automatic) in the Attune software, results in histograms showing the global maximum for each population. Figure 3B-3D shows representative results with GFP positive and GFP negative histograms with y-axis set to the global maximum for each population. Setting a specific y-axis maximum (y-axis scale set to manual with a user defined value), allows for comparison of absolute numbers between populations, which can be useful for comparing proliferation rates, but makes it difficult to compare cell cycle phasing in populations when numbers of GFP positive cells vs. GFP negative cells are vastly different.
Figure 3B shows cell cycle profiles for GFP positive and negative cells in larval eyes, expressing GFP in the posterior using the GMR-Gal4, UAS-GFP transgenes. The overlay reveals the differences in cell cycle phasing of the posterior GFP expressing cells. Figure 3C shows the relative change in cell size between GFP positive and negative cells in B, by plotting cell size as measured by FSC. Figure 3D shows a cell cycle profile from pupal brains at 46h APF. GFP positive cells are lineage tracing clones expressing the G1-S regulators Cyclin D and E2F that disrupt quiescence and lead to S-phase entry, indicated by the red arrow and confirmed by S-phase labeling with incorporation of EdU (Figure 3D inset). This is in contrast to the non-expressing GFP negative cells, which at this stage are normally arrested in G1 (black trace, Figure 3D).
Most cell cycle profiles are used to obtain information about percentages of a population in different cell cycle phases. This can be approximated in the Attune software by creating user defined regions on the histograms delineating G1, S and G2 phases (Figure 3E). The Attune software automatically generates a statistics table for each run, giving the absolute counts of cells within the user defined regions and percentages (Figure 3F). This method works well when comparing relative changes in phase distribution between a control and experimental population. Alternatively, modeling software can be used to estimate percentages in each phase as well as apoptotic cells. We used ModFitLT (Verity Software House) in Figure 3G, which can directly open the .fcs data files generated by the Attune software.
Figure 1. Dissection and staging of larval and pupal tissues. A. Dissection of larvae (of genotype w1118 shown) at 110 hr of development showing the anterior third before (top) and after (bottom) inversion. White arrowhead indicates a wing attached to the larval cuticle and a red arrowhead indicates the position of the brain B. Wings (white arrowhead; note wing at left is attached to a leg disc, wing at right is torn at the dorsal notum), eyes (yellow) and brain (red) removed from the larvae. C. Dissection of pupa showing positions of cuts with forceps placed at the airspace anterior to the head (red dotted lines) D. Pupa removed from cuticle before washing to remove fat and gut. E. Cleaned pupa showing process of lifting wings for removal F. Cleaned pupa showing wings (dotted) and partially removed eye-brain complex (red and yellow arrowheads). G. Dissected larval wings and eyes showing pipette for transfer. H. Dissected post-mitotic pupal eye-brain complex. I. Correct staging of animals at the larval/prepupa transition; animal 1 is still moving and is too young; animal 2 is at the correct stage to collect, 0hAPF (after pupa formation) it is white and immobile; animal 3 is too old to collect, the darkening of cuticle is evident by 1hAPF; animal 4 is too old to collect already tanned at 2hAPF. Click here to view larger figure.
Figure 2. Representative cell cycle analysis of tissues expressing GFP using Vybrant DyeCycle Violet and the Attune. A representative workspace, containing 4 dot plots and 2 histograms is shown for a larval wing sample expressing GFP and Cyclin D in the posterior wing, driven by an engrailed-Gal4 transgene. A. Dot plot of cell size with Gate 1 to exclude debris and clumps. B. Dot plot to discriminate singlets with Gate 2 to exclude unstained cells, sub-G1 DNA content (apoptosis), and clumps. C. Dot plot of GFP vs. DNA content. GFP negative (Gate 3) and positive (Gate 4) cells can be distinguished and 2N and 4N DNA content is evident based upon position along x-axis. D. Dot plot of cell size as measured by GFP vs. forward scatter (FSC) to generate Gates 5 and 6. Note that cells within Gate1 are colored red for visual assistance in determining Gates 5 and 6 for this scatter plot. E. Histogram of DNA content vs. counts for GFP or RFP negative (population set to Gate 3). F. Histogram of DNA content vs. counts for GFP or RFP positive (population set to Gate 4). Histograms can also be included for cell size if desired. For cell size histograms, populations should be set to Gates 5 and 6. Click here to view larger figure.
Figure 3. Representative histograms of cell cycle and cell size analysis for various developmental stages and tissues. (A) Representative histogram data is shown for comparing cell cycle in specific cell types using RFP expression with DyeCycle Violet (inset shows scatter plot of RFP fluorescence vs. DNA content). This example contains larval wings expressing RFP in the posterior, driven by an engrailed-Gal4 transgene. (B) Overlays of GFP positive vs. GFP negative DNA content histograms in larval eyes with GFP expression driven by the GMR-gal4 promoter17, showing most GFP positive cells in G1. (C) Smaller relative cell size of GMR driven GFP+ cells compared to GFP- controls, as measured by forward scatter using Gates 5 and 6. (D) In the pupal brain, GFP positive cells indicate heat-shock induced lineage tracing clones via flp-out method5 induced during the 2nd larval instar, over-expressing the G1-S cell cycle regulators Cyclin D and E2F, causing aberrant S-phase entry (red arrowhead) in the normally postmitotic G1 arrested tissue. The abnormal S-phase entry observed by flow cytometry was confirmed by incorporation of EdU, which labels cells in S-phase (inset). (E) Example of regions used to estimate relative percentages in G1, S and G2 phases. Boundaries were estimated based upon S-phase incorporation of EdU in Drosophila cells in a separate experiment. (F) Example of statistics table generated by Attune software showing relative percentages in the regions defined. (G) Example of cell cycle modeling of representative Attune data using ModFitLT. Click here to view larger figure.
The protocol described here allows for analysis of cell cycle, relative cell size and relative cell number in live Drosophila tissues at various developmental stages. When this analysis is coupled with cell-type specific fluorescent protein expression or lineage tracing, detailed information can be obtained about cellular responses to discreet cell cycle or growth perturbations. As proof of principle, we disrupted quiescence in the pupal fly brain by expressing G1-S cell cycle regulators in GFP labeled cell clones, leading to aberrant DNA replication at a developmental stage when brain cells are normally over 90% G1 arrested (Figure 3D).
However there are some important limitations to this live cell cycle analysis method. First, one cannot distinguish between cells in the various stages of mitosis using the protocol described here. Thus, the analysis described here should be supplemented with immunofluorescence staining for a mitotic marker, such as Ser10-phosphorylated Histone H318 to quantify mitotic index in the tissues of interest. In addition the method described here cannot distinguish between cells in G1 vs. a G0 arrest, as both states have the same DNA content. Since DyeCycle Violet is taken up by live cells, identification of cells in later apoptotic stages is also limited. Lastly, it is important to note that measurements of cell size based upon forward scatter are relative cell size measurements rather than absolute quantification of size. For absolute quantification of cell size we recommend a volumetric approach such as that used in a Coulter Counter (Beckman Coulter).
When cell cycle phase and size data is coupled with lineage tracing to measure cell doubling time, detailed information about the cell cycle and growth can be calculated, such as the length of the G1, S and G2 phases and cell growth rate9,11. This coupled with discreet genetic manipulations of cell cycle regulators in vivo can provide detailed information about cell cycle and growth regulation for quantitative cell cycle, cell growth and tissue morphogenesis modeling.
The authors have nothing to disclose.
We thank Aida de la Cruz for developing and teaching the original protocol on which this version is based10. Work in the Buttitta Lab is supported by NIH grant GM086517.
Name of the reagent | Company | Catalogue number | Comments (optional) | |||||||||||||||
12×75 mm Polystyrene Round-Bottom 5 ml Test Tube | BD Falcon | 352058 | 5 ml tubes | |||||||||||||||
Attune Acoustic Focusing Cytometer | Life Technologies/ Applied Biosystems | 4445315 | Blue / Violet configuration | |||||||||||||||
Attune Cytometer Software (version 1.2.5) | Life Technologies/ Applied Biosystems | Free | PC only | |||||||||||||||
Attune Performance Tracking Beads (5 x 106 beads/ ml) | Life Technologies/ Applied Biosystems | 4449754 | For daily performance test | |||||||||||||||
Dumont #5 Inox forceps | Fine Science Tools | 11251-20 | ||||||||||||||||
Embryo dishes 30 mm x 12mm | Electron Microscopy Sciences | 70543-30 | Glass dissection dishes | |||||||||||||||
Eppendorf Thermomixer | Eppendorf | 022670051 | ||||||||||||||||
Trypsin-EDTA Solution (10x) | Sigma | T4174 | ||||||||||||||||
Vannas-Tübingen Spring Scissors | Fine Science Tools | 15003-08 | Straight 5mm Cutting Edge | |||||||||||||||
Vybrant DyeCycle Violet Stain | Life Technologies/ Invitrogen | V35003 | ||||||||||||||||
Table 1. Required reagents and instruments. | ||||||||||||||||||
Live DNA Stain Solution (10 ml): 1 ml 10X Ca2+ Mg2+ free PBS (pH7.2) 10X Ca2+ Mg2+ free PBS (pH7.2): 1.37M NaCl, 27 mM KCl, 100mM Na2HPO4 (dibasic), 20mM KH2PO4 (monobasic) adjusted to pH 7.2 |
||||||||||||||||||
Table 2. Threshold and voltage setting for the analysis in Figure 2. |