5-Ethynyl-2'-Deoxyuridine/Phospho-Histone H3 Dual-Labeling Protocol for Cell Cycle Progression Analysis in Drosophila Neural Stem Cells

Cell cycle analysis with 5-ethynyl-2'-deoxyuridine (EdU) and phospho-histone H3 (pH3) labeling is a multi-step procedure that may require extensive optimization. Here, we present a detailed protocol that describes all steps for this procedure including image analysis and quantification to distinguish cells in different cell cycle phases.

In vivo cell cycle progression analysis is routinely performed in studies on genes regulating mitosis and DNA replication. 5-Ethynyl-2'-deoxyuridine (EdU) has been utilized to investigate replicative/S-phase progression, whereas antibodies against phospho-histone H3 have been utilized to mark mitotic nuclei and cells. A combination of both labels would enable the classification of G0/G1 (Gap phase), S (replicative), and M (mitotic) phases and serve as an important tool to evaluate the effects of mitotic gene knockdowns or null mutants on cell cycle progression. However, the reagents used to mark EdU-labelled cells are incompatible with several secondary antibody-fluorescent tags. This complicates immunostaining, where primary and tagged secondary antibodies are used to mark pH3-positive mitotic cells. This paper describes a step-by-step protocol for the dual-labeling of EdU and pH3 in Drosophila larval neural stem cells, a system utilized extensively to study mitotic factors. Additionally, a protocol is provided for image analysis and quantification to allocate labeled cells in 3 distinct categories, G0/G1, S, S>G2/M (progression from S to G2/M), and M phases.

The cell division cycle comprises a G1 phase (first gap phase), a replicative/S-phase, a G2 (second gap phase), and an M (mitotic) phase. Passing through these phases, the cell undergoes dramatic changes in cellular transcription, translation, and re-organization of cytoskeletal machinery^1,2. In response to developmental and environmental cues, cells may temporarily cease to divide and become quiescent (G0) or differentiate and permanently cease to divide³. Other scenarios, such as DNA damage, may cause premature differentiation or apoptosis^3,4. Response to such cues is mediated by cell cycle checkpoints, which act as a surveillance system to ensure the integrity of essential cellular processes before the cell commits to the next phase of the division cycle⁵. Therefore, studies on genes regulating DNA replication, checkpoints, and mitotic machinery need to analyze possible cell cycle progression defects that may occur in mutant cells or upon siRNA knockdown of these genes. Additionally, such analyses may be employed to test overall cell health as well as cellular responses to drug treatment.

5-bromo-2'-deoxyuridine (BrdU) is a thymidine analog that is incorporated into DNA during replication⁶. This method was used extensively to identify cells in S-phase. However, the cells are then subjected to harsh DNA denaturation procedures to allow detection of BrdU through the use of anti-BrdU antibodies⁶. This harsh treatment may damage cellular epitopes and prevent further characterization of the sample through immunostaining. EdU incorporation and subsequent detection by a copper-catalyzed, 'click reaction' with small, cell-permeable, fluorescently tagged azide dyes eliminates the need for harsh denaturation procedures⁷. This method, therefore, emerged as a more practical alternative to BrdU incorporation.

Further, pH3 has been described as a reliable marker for mitotic/M phase cells⁸. Histone H3 is a DNA-associated core histone protein that becomes phosphorylated in around the late G2 phase to early M phase and is de-phosphorylated toward the end of anaphase⁸. Several commercial antibodies can be used to detect pH3 using standard immunostaining protocols. Dual-staining of EdU and pH3 would therefore enable the detection of cells in S-phase as well as M-phase. Additionally, cells in the G1 and early G2 phase would not stain positively for either of the markers.

Drosophila neural stem cells or neuroblasts (NBs) offer a well-characterized stem cell model wherein cells divide asymmetrically to produce one identical self-renewing NB and a ganglion mother cell (GMC), which is fated for differentiation⁹. Additionally, several genetic tools and NB-specific antibodies make this system suitable for genetic manipulation and live-cell imaging. Consequently, several studies have utilized NBs to study genes regulating asymmetric divisions and cell fate determination⁹. Distinct populations of NBs exist in the central brain (CB) and the optic lobe (OL) of the larval brain⁹; CB NBs were used for the current study. These third instar larval CB NBs are large cells that are also suitable for studying factors regulating mitotic spindle assembly. A protocol to analyze cell cycle progression defects would be a vital tool in such studies.

Protocols published earlier employed commercial kits, such as Click-iT EdU Alexa Fluor Cell Proliferation Kit, which provide several reaction components and azide dyes tagged with a variety of Alexa Fluor dyes for EdU incorporation and detection¹⁰. However, the reagents supplied with such kits are not compatible with some fluorescent tags often used with secondary antibodies. This EdU detection kit (Click-iT EdU Alexa Fluor Cell Proliferation Kit supplied with Alexa Fluor 647-conjugated azide dye) was tested in Drosophila third instar larval NBs, and co-staining was attempted with antibodies against pH3 and Miranda, a marker for NBs. Further, Alexa Fluor 568- or Cy3-tagged secondary antibodies were used for detection of Miranda labeling on the plasma membrane of NBs¹¹. However, the expected signal intensity and staining pattern (unpublished results) were not observed with these secondary antibodies when immunostaining was performed after EdU detection.

For EdU incorporation, the protocol described by Daul and colleagues required feeding of the larvae with Kankel-White medium mixed with EdU and bromophenol blue (BPB)¹⁰. The larvae fed on the EdU and BPB-spiked food, which could be seen by its blue color upon ingestion in the larval gut. Although this method was used for EdU incorporation in Mms19 loss-of-function (Mms19^P) third instar larvae, the Mms19^P larvae apparently did not feed as hardly any blue color was detected in the larval gut (unpublished results). The Mms19^P larvae show drastic developmental deformities and eventually arrest in the third instar stage. This may somehow affect the feeding behavior of the third instar larvae and render the EdU-feeding protocol unsuitable for such cases.

After studying the available literature and working extensively on the standardization of essential steps, an alternative approach was proposed for EdU/pH3 dual-labeling in Drosophila NBs, which does not require feeding EdU to larvae. A previous study employed dual EdU/pH3 staining to analyze the cell cycle in NBs, but did not present a detailed protocol⁴. This presents an unnecessary hurdle for labs trying to implement this method. Furthermore, evaluating the compatibility of various reagents with the EdU kit and performing further optimization can be a time-consuming process. This paper presents a step-by-step protocol that covers EdU incorporation in dissected larval brains and immunostaining with anti-pH3 antibodies, followed by confocal microscopy and image analysis to allocate NBs to four distinct categories: G0/G1 phase, S phase, S>G2/M (progression from S to G2/M), and M phase. The steps that need optimization are outlined and tips provided for image analysis of large datasets. Additionally, the EdU/pH3 readout in wild-type NBs is analyzed and compared with Mms19^P NBs, which were recently reported to show a cell cycle delay¹¹.

1. Preparation of reagents and stocks for Click-it EdU assay

NOTE: Refer to the Table of Materials and Table 1 for details about the kit and reagents supplied with the kit.

1. Bring the vials to room temperature before preparing the solutions.
2. Prepare 10 mM EdU (component A) stock solution by adding 2 mL of dimethylsulfoxide (DMSO, component C). Mix well and store at -20 °C.
3. Prepare a working solution of Alexa Fluor 647-azide (component B) by adding 70μL of DMSO (component C). Mix well and store at -20 °C.
4. Prepare a 1x solution of Click-iT EdU reaction buffer (component D) by mixing 4 mL of this solution with 36 mL of deionized water. Store the remaining solution at 2-6 °C.
5. Make a 10x stock (200 mg/mL) of the Click-iT EdU buffer additive (component F) by adding 2 mL of deionized water. Mix well and store at -20 °C.
6. Store Hoechst 33342 (component G) at 2-6 °C. Store DMSO (component C) in a desiccator at -20 °C.
   ​NOTE: Gloves should be used to handle DMSO and Hoechst.

2. Dissection of third instar larval brains and EdU incorporation

NOTE: The protocol for brain dissections has been described previously¹². Before starting dissections, make sure sufficient amounts of the EdU and PFA solutions are prepared and thawed as described in 2.6 and 3.1.

1. Prepare 10x Phosphate-buffered saline (PBS) by adding 25.6 g Na₂HPO₄.7H₂0, 80 g NaCl, 2 g KCl, and 2 g KH₂PO4 to 1 L of deionized water. Adjust the pH to 7.4, and subsequently prepare a 1x solution in deionized water.
2. Add Schneiders dissection medium (SM) to two consecutive depressions of a glass (3 or 9) depression glass spot plate. Add 1x PBS to the next consecutive depression.
3. Using a pair of forceps, pick wandering third instar larvae and place on a tissue wetted with PBS to clean off fly food residue. Place the larva in the depression with PBS and then in the consecutive depression with SM.
4. Using a pair of fine forceps, remove 3/4^th of the lower body of the larva.
   1. Gently grab the larval mouth hooks with one pair of forceps, and hold the cuticle at the other end with the other pair of forceps.
   2. Turn the larval head inside out by pushing inward the mouth hooks and simultaneously peeling off the tissue at the other end.
5. Observe the larval brain with attached imaginal discs. Remove other tissues attached to the brain, and transfer the brain to the next consecutive depression with SM.
6. Thaw the 10 mM EdU stock solution. Prepare a 100 μM solution by diluting this 10 mM stock in SM.
7. Add 100 μL of this 100 μM EdU+SM solution in another depression on the Pyrex plate. Incubate ~5-10 dissected brains in 100 μL of 100 μM EdU+SM solution for 2 h at 25 °C.
   ​NOTE: As NBs divide once in ~2 h, this protocol aims to analyze one full cell cycle. Only use intact brains for further steps. Discard brains that are damaged during dissection.

3. Fixation and immunostaining

1. Prepare 4% paraformaldehyde (PFA) solution in a fume hood.
   1. Add 4 g of paraformaldehyde powder to 80 mL of 1x PBS in a beaker placed on a magnetic stirrer and heat to 60 °C. To dissolve PFA, add a few drops of 1 N NaOH. Check the pH, and adjust to 7.0 with a few drops of 1 M HCl if required.
   2. Adjust the volume to 100 mL with 1x PBS. Aliquot the PFA solution and store at 2-6 °C for up to 1 month. Add 0.3% non-ionic detergent (see the Table of Materials) to the PFA solution for efficient fixation of large tissues such as the larval brain.
2. After EdU incorporation, transfer the brains to 0.6 mL microcentrifuge tubes containing 4% PFA. Incubate at room temperature on a nutator for 15 min. Tap the tube on the laboratory bench so that the brains settle down at the bottom of the tube.
3. Remove the PFA and wash 3 times with PBS + 0.3% non-ionic detergent (PBST) at room temperature. Ensure that each wash lasts for at least 10 min on a nutator. After the last wash, remove PBST, add blocking solution (PBST + 5% bovine serum albumin), and incubate for 30 min at room temperature.
4. Dilute anti-Miranda antibody (1:250) and anti pH3 antibody (1:200) in PBST (both antibodies in the same tube). Remove the blocking buffer, and add the primary antibody solution. Incubate overnight at 2-6 °C on a nutator.
   NOTE: Miranda is a membrane protein present on CB NBs. It serves to identify these large round cells in the CB region¹³.
5. Remove the primary antibody solution and wash 3 times with PBST. Ensure that each wash lasts for 10 min on a nutator at room temperature.
6. Prepare secondary antibody solution by adding 1:500 diluted Alexa Fluor 488 anti-Rabbit and Alexa Fluor 568 anti-Rat in PBST. Remove the PBST, and add the secondary antibody solution to the dissected brains. Incubate overnight at 2-6 °C in the dark, on a nutator, and wash 3 times with PBST (step 3.5).

4. EdU detection, DNA staining, and mounting

1.  To prepare the EdU detection cocktail, mix the components (see Table 2) in a 1.5 mL tube.
2. After the last wash (step 3.6), remove PBST, and add the EdU detection cocktail to the dissected brains. Incubate at room temperature in the dark for 30 min on a nutator. Wash 2 times with PBST for 10 min each, in the dark.
3. For DNA staining, dilute Hoechst 33342 (component G) 1:2,000 in PBST to prepare a 5 μg/mL solution.
4. Remove PBST after the second wash (step 4.2), and add 500 μL of 5 μg/mL Hoechst solution to the dissected brains. Incubate in the dark for 10 min. Remove Hoechst, and wash once with PBST for 10 min.
5. Tap the tube onto the lab bench, and let the brains settle down. Remove the PBST from the tube, but leave behind only 50-100 μL.
6. Cut the end of a 200 μL micropipette tip, and carefully transfer the brains onto a clean glass slide. Remove excess PBST from the slide by blotting with filter paper strips. Be careful not to let the filter paper touch the brains.
7. Put a drop of water-soluble, non-fluorescing mounting medium onto the brains, and orient the brains such that the ventral nerve cord faces the slide, and the lobes face upwards. Arrange the brains in a single file so that it is easier to image the brains serially.
8. Gently place a coverslip on top of the brains. Place the slide at 2-6 °C overnight.
   ​NOTE: The mounting medium hardens upon overnight storage so that there is no need to seal the coverslip with nail polish.

5. Imaging

NOTE: See the Table of Materials for details on the laser scanning microscope and oil-immersion objective used in this protocol.

1. From the Acquisition software, select the 63x objective.
2. Put a drop of immersion oil on the coverslip just above the mounted brains to make it easier to locate the tissue through the eyepiece.
3. Using the DAPI/Hoechst 33342 channel, find the brain through the eyepiece, and then switch to the acquisition mode in the software.
4. Set up 4 channels to image Hoechst 33342 (DNA), Alexa Fluor 488 (pH3), Alexa Fluor 568 (Miranda), and Alexa Fluor 647 (EdU). Use the dye-assistant tool, which automatically sets up the excitation lasers and emission filters for the selected dyes.
5. Set the field of view such that it encompasses the entire brain lobe. Image the entire volume of the brain lobe by acquiring z-stacks spaced 0.8 μm apart. Store all images from an imaging session in a *.lif library format.

6. Image analysis

NOTE: The following steps describe the analysis of acquired images and how to sort cells into G0/G1 phase, S phase, S>G2/M (progression from S to G2/M), and M phase using the ImageJ software.

1. Download Fiji (Fiji is ImageJ) from the following URL: https://fiji.sc/. Open Fiji, then drag and drop the .lif files into Fiji.
   NOTE: Fiji is a version of ImageJ that comes pre-installed with several plugins¹⁴. Transferring .lif files into Fiji will open the Bio-formats plugin, which is needed to process the .lif files. Bio-formats plugin is also required to open image files generated from some other microscope brands, e.g., nd2 files generated from a Nikon microscope. This plugin comes pre-installed with Fiji.
2. Select Data browser from the stack viewing tab, and use virtual stack from the memory management tab in the bio-formats plugin.
   NOTE: Lif files with z-stacks from 8-10 brains are often 300-400 megabytes in size. On computers with low RAM, opening a few such files can rapidly deplete the available RAM on ImageJ and prevent any further image processing. Virtual stack is 'read-only', does not need high RAM for processing, and is an ideal option to load large datasets onto Fiji.
3. Observe the multichannel image displayed in ImageJ. Change the color of the channels from the menu bar using Image | color | channels tool. Observe the Miranda-labeled NBs as large round cells in the CB region.
4. Draw a region of interest (ROI) using the ellipse tool over each NB to avoid counting the NB twice.
   1. From the ImageJ menu bar, select analyze | tools | ROI manager. Mark all NBs in the current z-section, and press t after marking each cell.
5. Once all NBs in the current z-section are marked, change the channels to pH3 and EdU, and manually count the number of EdU-positive NBs, pH3-positive NBs, NBs staining positively for both EdU and pH3, and NBs not staining for both markers.
6. Search for NBs in subsequent z-sections, delete old ROIs, and add new ROIs to count NBs in different stacks.
7. Prepare a spreadsheet with the following columns: 1. EdU- pH3-: this section represents the double-negative cells that are in the G0/G1 phase of the cell cycle; 2. EdU+: the cells in this category have incorporated EdU and are undergoing DNA replication (S-phase); 3. EdU+ pH3+: these dual-positive cells have completed the S phase and have progressed to G2 or M phase; 4. pH3+: these NBs are undergoing mitosis.
8. Calculate percentages of NBs present in all of the above 4 categories for each lobe. Prepare a bar graph using the spreadsheet software showing the pooled data for percentages of NBs in each category.

The bi-lobed Drosophila third instar larval brain has been utilized as a model system to study fundamental cellular and developmental processes⁹. The focus of the current study was to present a protocol for analysis of cell cycle progression in EdU- and pH3-labeled NBs of the CB region (Figure 1). The CB NBs are sub-divided into type I and type II, and they display the characteristic asymmetric division pattern⁹. Each type I NB division generates an NB, which is capable of self-renewal, and another GMC that is fated for differentiation⁹. In contrast, type II NBs generate transit-amplifying, immature neural progenitor cells (INPs) that self-renew and generate 2 GMCs⁹. Although a distinct NB population exists in the OL region, this study focused on the CB NBs, which are large and easily identifiable with NB-specific markers.

For this study, Miranda was used to identify and analyze CB NBs (Figure 2). Although Miranda also stains other cell populations in the OL, CB NBs can be identified with Miranda due to their large size and CB location¹³. Other more specific markers can also be used to label CB NBs, such as Deadpan, which is a transcription factor regulating NB self-renewal. However, as both pH3 and EdU mark the chromatin, the membrane marker Miranda was used to make it easier to visualize and analyze NBs marked with pH3 and EdU. EdU incorporation, subsequent detection, and further characterization by immunostaining and image analysis is a complex procedure. It is important to standardize each step to determine optimal staining and imaging conditions. Although some published reports present EdU/pH3 labeling strategies, these reports do not elaborate on optimization and image analysis steps. The workflow outlined in Figure 1 summarizes the steps from EdU incorporation until image analysis.

We recently characterized the Mms19 gene, which is required by NBs for normal mitotic progression¹¹. Through live-cell imaging analyzes, NBs lacking functional Mms19 were shown to take twice as long as wild-type NBs to finish mitosis. This was clearly reflected in the EdU/pH3 analysis wherein a significantly higher proportion of Mms19^P NBs were found to be in the M-phase as compared to wild-type NBs (Figure 3A). Expression of the Mms19::eGFP fusion protein in the Mms19^P background had been shown to rescue phenotypic defects^11,15. This correlated well with the results from the cell cycle progression analysis wherein the proportion of cells in M-phase was rescued to wild-type levels upon Mms19::eGFP expression in the Mms19^P background (Figure 3A,B).

Figure 1: A simplified workflow for EdU/pH3 dual-labeling. Third instar larval brains were dissected and incubated with 100 μM EdU for 2 h. Subsequently, the brain tissue was fixed with PFA and immunostained with antibodies against Miranda and pH3. Images of the stained brains were obtained on a confocal microscope and further processed with ImageJ/Fiji to allocate cells to different cell cycle phases. Abbreviations: EdU = with 5-ethynyl-2'-deoxyuridine; pH3 = phospho-histone H3; PFA = paraformaldehyde; SM = Schneiders dissection medium; PBS = phosphate-buffered saline; PBST = PBS + 0.3% non-ionic detergent; BSA = bovine serum albumin. Please click here to view a larger version of this figure.

Figure 2: Allocation of Cells to G1/G0; S; S>G2/M and M phases. Cells progress through the G1, S, G2, and M phases to complete a full cell cycle. CB NBs from wandering third instar larval brains were marked with Miranda (red), EdU (gray), pH3 (green), and DNA was labeled with Hoechst 33342 (blue). Labeled NBs were analyzed in ImageJ and allocated to 4 distinct phases based on whether they stained positively for (A) neither EdU nor pH3 (G1/G0), (B) only EdU (S phase), (C) both EdU and pH3 (S>G2/M), and (D) only pH3 (M phase). Examples of individual NBs only from wild-type brains are shown here. Scale bars = 5 µm. Abbreviations: CB = central brain; NBs = neuroblasts; EdU = with 5-ethynyl-2'-deoxyuridine; pH3 = phospho-histone H3. Please click here to view a larger version of this figure.

Figure 3: Effect of Mms19^P on NB cell cycle phase distribution. (A) Upon analysis of wild-type (w;+;+) NBs, ~25% NBs were found to be in M-phase. In Mms19^P NBs, however, this proportion increased to almost 40%. Mms19::eGFP expression in the Mms19^P background is known to rescue Mms19^P phenotypes, and the proportion of M phase NBs in this background was comparable to wild type. (B) Percentages of NBs in 4 different categories corresponding to cell cycle phases G1/G0, S, S>G2/M, and M were compared across wild type; Mms19^P, and Mms19::eGFP, Mms19^P genotypes. The proportion of M-phase and G1/G0 phases differs considerably in Mms19^P NBs when compared to wild-type NBs. In contrast, the cell cycle distribution of NBs in Mms19::eGFP, Mms19^P brains are comparable to the one found in the wild type. Statistical significance was calculated using the Kruskal-Wallis test and multiple columns compared using Dunn's post-test; ***(P<0.001). This figure has been modified from ¹¹. Abbreviations: NBs = neuroblasts; eGFP = enhanced green fluorescent protein; WT = wild-type. Please click here to view a larger version of this figure.

Table 1: EdU kit components. Components provided with the Click-iT EdU Alexa Fluor Cell Proliferation Kit and the respective amounts/volumes. Abbreviation: EdU = with 5-ethynyl-2'-deoxyuridine.

Table 2: Preparation of EdU detection cocktail. EdU detection cocktail was prepared by mixing the indicated volumes of the kit components (prepared in section 2). Abbreviation: EdU = with 5-ethynyl-2'-deoxyuridine.

EdU incorporation and its subsequent 'click' reaction with cell-permeable azide presents practical advantages of this technique over the BrdU method used earlier⁷. However, this reaction is catalyzed by Cu(I) ions, and several dyes may be unstable in the presence of this copper catalyst, as is clearly advised by the Click-it EdU kit manufacturer. When immunostaining experiments had been performed after executing the EdU detection step, the expected signal intensity was not observed with the red channel dyes, Alexa Fluor 568 and Cy3. However, the protocol worked when immunostaining reactions were completed before the EdU detection step. With this protocol, four channels were used to visualize Hoechst, pH3, Miranda, and EdU.

Another optimized step in this protocol was the uptake of EdU by dissected brains as opposed to feeding EdU-spiked fly food to larvae. As the full Drosophila NB cell cycle lasts for approximately 2-2.5 h¹⁶, EdU uptake by brains for 2.5 h as in this protocol, enables the tracking of one full cell cycle of NBs. Cells that do not stain positively for either EdU or pH3 are resting cells. In the case of a third instar larva wherein the NBs are actively dividing, these resting cells would most likely be in the G1 phase of the cell cycle. However, from the late embryonic to the early second instar stages, the NBs temporarily become quiescent¹⁷. In these early developmental stages, double-negative staining would most likely represent a G0/quiescent phase.

NBs undergoing DNA replication should stain positively for EdU, whereas cells that have progressed from the S phase (EdU incorporation) to the late G2 or M phase should stain positively for both EdU and pH3. However, characterization of the G2 phase would be difficult with this protocol as early G2 cells do not stain positively for EdU or pH3. However, cells that have incorporated EdU and have progressed to early G2 would also stain positively for EdU and could complicate the analysis of the G2 phase. Several drugs that inhibit DNA replication cause cell cycle arrest at the S-phase^18,19,20. This protocol could be a convenient tool to analyze and identify the effect of such drug treatments as S-phase-blocked cells would not progress to the M-phase and the fraction of EdU, pH3 double-positive cells would be decreased or even absent in this case.

In a recently published study, we evaluated the effect of the Mms19 gene on mitotic progression in NBs¹¹. Through live-cell imaging, we demonstrated a dramatic delay of Mms19^P NBs in M-phase progression. While wild-type NBs complete the M-phase in ~10 min, Mms19^P NBs took around twice as much time to finish the M-phase¹¹. Accordingly, this EdU/pH3 dual-labeling protocol also reflected a mitotic delay as we observed almost 1.5x as many NBs staining positively for pH3 in Mms19^P brains as compared to wild-type brains. Data from this EdU protocol is thus comparable with the direct live-cell imaging results. As such direct, live-cell approaches are time-consuming and need extensive optimization, this protocol could be utilized to perform a rapid initial screening of such cell cycle mutants to understand defects at specific stages. Once a phase of interest is identified, this could then be followed up by direct cell visualization assays.

The duration of cell cycle phases can vary considerably between different cell types and developmental stages. For example, in developing Drosophila embryos, downregulation of mitotic kinases results in an extended duration of the S phase, while in cells fated for differentiation, such as ferret (Mustela putorius furo) brain neural progenitors, the S phase duration is markedly short^21,22. EdU pulse labeling time should thus be optimized depending on the length of S and G phases in the given cell type. Duration of the EdU pulse would also depend on the experimental objectives, e.g., a short pulse is sufficient to measure the fraction of cells currently in the S-phase, while a longer pulse enables the analysis of the progression of cells through the S phase.

Extensive optimization of EdU pulse may also permit estimation of S phase, as was elegantly demonstrated by Pereira and colleagues²³. These researchers demonstrated a novel, flow cytometry-based method in which HCT116 cells were pulse-labeled with EdU for incremental time periods. The authors demonstrated that the maximum fluorescent intensity of EdU is obtained when the pulsing time matches the length of the S phase²³. Moreover, analysis of the temporal progression of EdU-labeled cells also enabled the quantification of the G1 and G2/M phases. Although the current protocol for EdU/pH3 dual-labeling enables the analysis of S and M phase defects, it is not possible to measure the precise duration of phases with this protocol. The protocol described by Pereira and colleagues could be a suitable alternative for assays that require the determination of cell cycle phase lengths. Adapting this protocol with an additional pH3 labeling may also result in higher sensitivity in detecting M phase cells.

Apart from the EdU/pH3 approach, the fluorescent ubiquitin-based cell cycle indicator (FUCCI) method has also been utilized to study cell cycle progression in mammalian cells as well as in Drosophila tissues^24,25. This system utilizes two fluorescently tagged proteins, geminin and Cdt1, which contain motifs for specific ubiquitination and proteasomal degradation by APC/C and SCF^Skp2, respectively. As APC/C is only active from the end of mitosis through G1 and SCF^Skp2 is active in the S and G2 phases, the cell-cycle-stage-specific degradation of fluorescently tagged geminin and Cdt1 enables the determination of the cell cycle phase²⁴. A slightly modified 'Fly-FUCCI' method for Drosophila tissues relies, instead, on fluorescently tagged Cyclin B and E2F1, which are degraded by APC/C (during mitosis) and CRL4^Cdt2 (during S-phase onset), respectively²⁵. A previous study characterizing premature differentiation of Drosophila larval NBs in response to aneuploidy analyzed cell cycle defects by both Fly-FUCCI and EdU/pH3 methods⁴. Aneuploidy induces premature differentiation of NBs, and therefore, a high fraction of NBs exit the cell cycle.

This was accurately reflected in both Fly-FUCCI and EdU/pH3 data⁴. Data obtained from the EdU/pH3 method are, therefore, also comparable with other cell cycle-tracking methods such as the Fly-FUCCI. Fly-FUCCI is a powerful tool, and all the fly stocks with ubiquitously driven fluorescent markers or markers fused to tissue-specific promoters are available from stock centers. However, using these constructs to analyze cell cycle defects in null mutants or with siRNA knockdown would involve genetic recombination to create a fly stock that carries the FUCCI elements, tissue-specific drivers, a specific mutation, or an siRNA of interest. This process would delay the actual experiment for several weeks, whereas the EdU/pH3 method can readily be applied to fly lines with a genetic null background. Alternatively, for siRNA knockdowns, a one-step cross between a driver stock and a siRNA stock can be utilized for EdU/pH3 analysis.

One drawback of this approach is the imaging and image analysis pipeline, which involves manual quantification of a large number of cells from several brain samples. Recently, a flow cytometry approach has been presented as a high-throughput alternative to a conventional microscopy-based assay for cell cycle analysis²⁶. The rapid analysis of thousands of cells at the same time with this approach is advantageous, and the ability to detect multiple markers enables the quantification of specific cell types. Although readily usable with cultured cell lines, this protocol might be difficult to apply to large tissue samples such as the Drosophila larval brain. However, a few protocols have been published that describe brain tissue dissociation and analysis of isolated larval NBs by flow cytometry^27,28. Further innovative approaches in this direction may present novel opportunities for high-throughput cell cycle analysis in Drosophila tissues.

The authors have declared that no competing interests exist.

This work was supported by funding from the Swiss National Science Foundation (project grant 31003A_173188; www.snf.ch) and the University of Bern (www.unibe.ch) to BS. The funders had no role in study design, data collection, and analysis, decision to publish, or preparation of the manuscript.