Protocol to analyze Drosophila intestinal tumor cellular heterogeneity using immunofluorescence imaging and nuclear size quantification

Summary

Drosophila intestinal tumors show an extended cellular heterogeneity. We devise a protocol to assess tumor cell heterogeneity by employing nuclear size measurement and immunofluorescence-based cell lineage analysis. We describe steps for intestinal dissection, staining, and imaging, followed by detailed procedures for nuclear size analysis. This approach detects overall heterogeneity across the entire tumor cell population and deviations within specific cell populations. The procedure is also applicable for analyzing the heterogeneity of wild-type intestinal cells in various contexts.

For complete details on the use and execution of this protocol, please refer to Pranoto et al.¹

Graphical abstract

Highlights

• •A protocol to quantify tumor cell heterogeneity in Drosophila intestine
• •Detection of typical cell type proportions and the presence of atypical cells
• •Employs immunofluorescence imaging technique and simple nuclear area measurement
• •Can be used to assess intestinal cell heterogeneity in various perturbation contexts

Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

Drosophila intestinal tumors show an extended cellular heterogeneity. We devise a protocol to assess tumor cell heterogeneity by employing nuclear size measurement and immunofluorescence-based cell lineage analysis. We describe steps for intestinal dissection, staining, and imaging, followed by detailed procedures for nuclear size analysis. This approach detects overall heterogeneity across the entire tumor cell population and deviations within specific cell populations. The procedure is also applicable for analyzing the heterogeneity of wild-type intestinal cells in various contexts.

Before you begin

Adult Drosophila intestinal epithelium is composed of four cell types: intestinal stem cells (ISCs), enteroblasts (EBs), enterocytes (ECs), and enteroendocrine cells (EEs).^2,3,4,5,6,7 ISCs are diploid cells expressing Delta (Dl), capable of generating themselves and differentiating into other intestinal cell types.^2,3,4,5,6,7 ISCs can give rise to EBs, the diploid cells often characterized using the Su(H)GBE reporter, which further differentiate into the absorptive ECs.³ ECs are large polyploid cells characterized by their distinctive large nuclear size and Pdm1 expression. ISCs can also differentiate into the diploid secretory EEs, identified by the expression of Prospero (Pros).^8,9

Escargot (Esg) is a transcription factor expressed in ISC and EBs at the posterior midguts. The conditional GAL4 driver esg^ts (esg-GAL4, tub-GAL80^ts, UAS-GFP) has been used for the genetic manipulation in ISCs and EBs.^10,11,12 Given that GFP is also driven by esg-GAL4, the ISCs and EBs are marked with GFP. Expression of oncogenes, such as an active form of Yorkie (Yki^3S/A) and a mutant Ras (Ras^V12), with esg^ts induces intestinal tumors, which are marked by GFP (Figure 1A).^{13,14,15,16,17,18} Notably, tumor cells induced by the expression of Yki^3S/A are quite variable in size, and a significant portion of them is even larger than normal ECs, suggesting their heterogeneity (Figure 1B).

Figure 1.

Expression of yki^3S/A in ISCs and EBs induces intestinal hyperplasia

(A) Representative images of esg^ts>+ and esg^ts>yki^3S/A posterior midguts. yki^3S/A is expressed with esg^ts by shifting to 29°C for 8 days. The XY plane illustrates the top view of intestinal tissues. Scale bar, 50 μm. The XZ and YZ planes show cross-sectional views of the midguts. Scale bars, 50 μm (x/y-axis) and 25 μm (z-axis). The cells manipulated by esg^ts are marked with GFP (green), nuclei are stained with DAPI (blue), and visceral muscle is stained with Phalloidin (red).

(B) Representative magnified views illustrating the cell compositions in esg^ts>+ and esg^ts>yki^3S/A intestines. Arrowheads show the cell size/nuclear size variability of GFP⁺ cells in esg^ts>yki^3S/A compared to those in esg^ts>+. Scale bar, 10 μm.

Different cell populations in the Drosophila intestinal epithelium have been assessed in multiple ways. Immunofluorescence imaging for the cell type markers, such as Dl, Su(H)GBE-LacZ, Pros, and Pdm1, has been commonly used to detect the four cell types in the Drosophila intestine.^2,9,19,20,21 In particular, lineage analysis of mitotic clones has been used for studying how an alteration of ISC lineage impacts the composition of the four cell types in the intestine.^8,22,23 Given the ploidy difference between ECs and other intestinal cell types, some studies have used nuclear size and DAPI signal intensity as proxies for ploidy to distinguish polypoid and diploid cells, which represents ECs and the other cell types in wild-type intestines, respectively.^24,25,26,27 During the differentiation of EBs into ECs, cells grow and adopt an elongated and irregular morphology.²⁸ A cell-type marker distinguishing the EB population undergoing differentiation from the rest of EBs is not currently available. Thus, measuring their size and circularity provides a way to assess the differentiating EB population.²⁸

Here, we present a protocol to assess the cellular heterogeneity of intestinal tumors and wild-type intestines using a combination of nuclear size measurement and cell type-specific markers. Our protocol quantitatively assesses the overall cellular heterogeneity in the intestine and deviations in the heterogeneity within specific cell populations defined by the conventional intestinal cell type markers. Lastly, our protocol provides an efficient way to detect deviations in the intestinal tumor cell composition compared to wild-type intestines.

Inducing Drosophila yki^3S/A intestinal tumor formation

Timing: ∼6 days (see note)

• 1.Collect 10–15 female flies in a fresh vial.
• 2.Incubate collected flies in a 29°C incubator for 6 days to induce yki^3S/A intestinal tumors.

Note: Collect 0–3-day-old female flies for tumor induction (Step 1).

Note: Do not keep more than 15 flies in a vial to prevent overcrowding and ensure proper nutrient availability.

Note: Depending on the research interest, the duration of yki^3S/A induction could be adjusted to reach the desired tumor cell confluency.

CRITICAL: Fly food in the vial would get soggy upon the temperature shift to 29°C. To prevent flies from drowning in the food, transfer flies into fresh vials every other day and keep them incubated in 29°C incubator until ready for dissection.

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
anti-GFP antibody, Alexa Fluor 488 (rabbit polyclonal) (dilution 1:1,000)	Thermo Fisher Scientific	Cat# A21311; RRID: AB_221477
anti-Delta antibody (mouse monoclonal) (dilution 1:1,000)	Developmental Studies Hybridoma Bank	Cat# C594.9B; RRID: AB_528194
anti-Pdm1 antibody (rabbit polyclonal) (dilution 1:1,000)	A gift from Dr. Yu Cai at Temasek Life Sciences Laboratory	N/A
anti-β-galactosidase antibody (rabbit polyclonal) (dilution 1:1,000)	Cappel	Cat# 55976
anti-β-galactosidase antibody (mouse monoclonal) (dilution 1:1,000)	Developmental Studies Hybridoma Bank	Cat# 40-1a; RRID: AB_2314509
anti-Prospero antibody (mouse monoclonal) (dilution 1:1,000)	Developmental Studies Hybridoma Bank	Cat# MR1A; RRID: AB_528440
Goat anti-rabbit IgG, Alexa Flour 594 (dilution 1:1,000)	Thermo Fisher Scientific	Cat# A-11012; RRID: AB_2534079
Goat anti-mouse IgG, Alexa Flour 594 (dilution 1:1,000)	Thermo Fisher Scientific	Cat# A-11005; RRID: AB_2534073
Goat anti-rabbit IgG, Alexa Flour 647 (dilution 1:1,000)	Thermo Fisher Scientific	Cat# A-21244; RRID: AB_2535812
Goat anti-mouse IgG, Alexa Flour 647 (dilution 1:1,000)	Thermo Fisher Scientific	Cat# A-21235; RRID: AB_2535804
Chemicals, peptides, and recombinant proteins
10X PBS	Fisher Scientific	Cat# 70011044
Normal goat serum	Thermo Fisher Scientific	Cat# 10000C
Triton X-100	Sigma-Aldrich	Cat# T9284
16% paraformaldehyde	Electron Microscopy Sciences	Cat# RT15710
Alexa Fluor 594 phalloidin (dilution 1:1,000)	Thermo Fisher Scientific	Cat# A-12381
Alexa Fluor 647 phalloidin (dilution 1:1,000)	Thermo Fisher Scientific	Cat# A-22287
DAPI (dilution 1:2,000)	Sigma-Aldrich	Cat# D9542
VECTASHIELD	Vector Laboratories	Cat# H-1000
Experimental models: Organisms/strains
D. melanogaster. esgts: esg-GAL4, tub-GAL80ts, UAS-GFP	Perrimon Lab at Harvard Medical School	N/A
D. melanogaster. w1118	Bloomington Drosophila Stock Center	BDSC: 5905; FlyBase: FBal0018186
D. melanogaster. w∗; P{y[+t7.7] w[+mC] = UAS-yki.S111A.S168A.S250A.V5}attP2	Bloomington Drosophila Stock Center	BDSC: 28817; FlyBase: FBtp0051046
Software and algorithms
Fiji	ImageJ	http://fiji.sc/
Leica Application Suite X (LAS X) software	Leica Microsystems	RRID:SCR_013673
Microsoft Excel	Microsoft Office	N/A
Other
PYREX 9 depression spot plate, Corning	VWR	Cat# 89090-482
Dumont Tweezers Style 5, Electronic Inox 02	Electron Microscopy Science	Cat# 0302-5-PO

Materials and equipment

• •4% sucrose: 2 g of sucrose in ddH₂O up to 50 mL.
• •1X phosphate-buffered saline (PBS): 100 mL of 10X PBS in 900 mL ddH₂O.
• •10% Triton X-100: 5 mL of Triton X-100 in 45 mL ddH₂O.
• •PBS containing 0.2% Triton X-100 (PBST): 1 mL of 10% Triton X-100 in 49 mL of 1X PBS.
• •Blocking buffer: 5% (v/v) normal goat serum in PBST.
• •4% paraformaldehyde (PFA): 3.75 mL of 16% PFA in 11.25 mL of 1X PBS.

Note: Unused 4% PFA can be stored at 4°C for 1–2 months.

Note: Make fresh blocking buffer on dissection day.

CRITICAL: Paraformaldehyde is an irritant, toxic, and carcinogenic chemical. Prepare the 4% PFA solution in a fume hood with proper personal protective equipment (PPE).

Step-by-step method details

Intestinal lumen clearing and intestinal tissue dissection: Day 1

Timing: 3–4 h for intestinal lumen clearance + 10–20 min for dissection

Intestines are emptied and then dissected.

• 1.
  Remove auto-fluorescent food particles in the fly intestinal lumen.

  • a.
    Place a half-sheet of a Kimwipe to cover the bottom of an empty fly vial.
  • b.
    Thoroughly wet the Kimwipes with 4% sucrose using a disposable transfer pipette.
  • c.
    Remove excess sucrose solution from the vial.
  • d.
    Transfer flies to sucrose vials.
  • e.
    Incubate flies in sucrose vials for 3–4 h at 29°C.
• 2.Dissect fly intestines in 1X PBS in a 9-well round bottom glass dish. See Troubleshooting Problem 1.

Note: There are a few different ways to dissect a fly intestine, depending on certain preferences and personal adjustments. Some useful resources on the fundamentals of dissecting adult fly intestines are available online as described by Tauc et al. and Micchelli.^29,30

Note: Avoid leaving unfixed fly intestines in PBS for more than 20 minutes.

CRITICAL: Carefully remove the crop and the Malpighian tubules from the intestine without stretching and compromising the midgut. Handle the midguts by grabbing onto the hindgut region and the cardia to avoid damaging the posterior midgut region.

Intestinal tissue fixation and staining: Day 1

Timing: 2 h for fixation and blocking + 16–20 h for staining

Intestinal tissue is fixed using paraformaldehyde and stained with various primary antibodies and dyes. For standard nuclear size analysis, which assesses the degree of intestinal tumor cell heterogeneity, intestines are stained with DGP staining solution (DAPI, anti-GFP Alexa Fluor 488 conjugate, and Phalloidin). Whereas, for cell type characterization within a tumor, intestines are stained with primary antibodies against specific cell type biomarkers.

• 3.Incubate the intestines in 300 μL of 4% PFA solution for 20 min at 20°C–25ºC.

Note: Depending on the primary antibody, fixing duration could be adjusted for enhanced staining results and signal detection under a fluorescence microscope.

• 4.Wash intestines 3 times with 5-min intervals using 300 μL of PBS containing 0.2% triton X-100 (PBST).

Note: Use a P200 or P1000 micropipette to slowly remove the existing solution in the well and add the appropriate solution. Make sure not to withdraw the intestines out into the pipette tips.

• 5.Incubate the fixed intestines in 300 μL of blocking buffer for 1 h at 20°C–25ºC.
• 6.
  Incubate the samples in 300 μL of staining solution for 16–20 h at 4°C.

  • a.
    For standard nuclear size analysis, make a DGP staining solution by diluting DAPI, anti-GFP Alexa Fluor 488 conjugate, and Phalloidin in the blocking buffer.

    • i.
      DAPI is used at 1:2000 concentration.
    • ii.
      Anti-GFP Alexa Fluor 488 conjugate is used at 1:1000 concentration.
    • iii.
      Phalloidin Alexa Fluor 594 conjugate is used at 1:1000 concentration.
  • b.
    For cell-type-specific biomarker characterization, dilute the primary antibody in the blocking buffer.

Note: Most biomarker antibodies, including Anti-Delta for ISCs, Anti-β Galactosidase for EBs, and Anti-Prospero for EEs, are used at 1:1000 concentration. Concentration may be adjusted according to experimental needs.

Note: As an alternative to a 16–20 h primary antibody incubation, a one-hour incubation at 20°C–25ºC would yield sufficient signals.

Note: All staining steps are done in a 9-well round bottom glass dish. To keep the primary antibody solution from evaporating during the 16–20 h incubation, place the 9-well dish in an airtight container lined with a wet paper towel.

CRITICAL: To prevent intestinal tissues from drying out, quickly exchange the buffers/solutions during washing and antibody incubations.

Intestinal tissue staining and mounting: Day 2

Timing: 2 h for secondary staining + 1 h for total washing steps and mounting

• 7.Wash intestines 3 times with 5-min intervals using 300 μL of PBST.
• 8.If tissues are incubated in the DGP staining solution for standard nuclear size analysis (Step 6a), proceed to Step 10.
  If tissues are incubated in a primary antibody solution (Step 6b), incubate the samples in 300 μL of secondary antibody solution for 2 h at 20°C–25ºC.

  • a.
    DAPI is used at 1:2000 concentration.
  • b.
    Anti-GFP Alexa Fluor 488 conjugate is used at 1:1000 concentration.
  • c.
    Secondary antibody Alexa Fluor 594 conjugate is used at 1:1000 concentration.
  • d.
    Phalloidin Alexa Fluor 647 conjugate is used at 1:1000 concentration.

    Note: Ensure to use the correct secondary antibody according to the primary antibody host species.

• 9.Wash intestines 3 times with 5-min intervals using 300 μL of PBST.
• 10.
  Mount the intestines onto microscope slides. See Figure 2.

  • a.
    Add 15 μL of Vectashield mounting medium onto a microscope slide.
  • b.
    Using dissection forceps, place the intestines onto the Vectashield droplet on the slide and laterally align the intestines next to each other. See Troubleshooting Problem 1 and 2.
  • c.
    At a slight angle, gently place an 18 × 18 mm coverslip over the row of intestines. Allow the Vectashield to perfuse throughout the area under the coverslip. Avoid air bubbles.
  • d.
    Seal the perimeter of the coverslip with a thin coat of clear nail polish.

Note: If 15 μL of mounting medium is insufficient to cover the area under the coverslip, add 5 μL of Vectashield on the edge of the coverslip and let it permeate the remaining area under the coverslip.

Pause point: Slides can be stored for up to a week at 4°C with minimal light exposure before imaging. However, imaging within a few days of tissue mounting is highly recommended.

Figure 2.

Alignment of the intestines during mounting

(A) Example of mounted intestinal samples on a slide. Intestines are aligned in parallel on Vectashield medium, protected using a coverslip, and sealed with a thin coat of nail polish around the perimeter.

(B) A magnified view of intestines aligned along their anterior/posterior axis.

Intestinal tissue imaging: Day 3

Timing: 10–15 min per posterior midgut

• 11.Under a confocal/fluorescence microscope, find the posterior region of the midgut. The posterior midgut is located just above the pyloric ring, the beginning of the hindgut.

Note: Our study focuses on the posterior midgut region as cell differentiation programs and cell types in this region are better understood in wild-type intestines. In principle, this protocol can be adopted for studying different regions in the intestine.

• 12.Using the 40x objective lens, capture Z-stack images of the posterior midguts at a 1024 × 1024 resolution with a Z-step size of 0.8 μm.

Note: Set the scanning depth to capture the entire intestine, using the visceral muscle layer as a proxy for tissue boundary. The fluorescent signals at the deeper side of the intestinal leaflet will be weaker due to the thickness of the tissue.

Note: Fly intestinal tumor images are taken using a Leica SP8 LIGHTNING confocal microscope with a 40X oil immersion objective, using sequential laser scanning. While it is ideal to have consistent laser intensity and gain for the DAPI channel across all samples, it is okay to adjust the settings to best capture the nuclear signal. Make sure not to over-saturate the DAPI signal. Laser power for specific biomarkers, GFP, and Phalloidin signals across all intestinal tissues may be adjusted accordingly to achieve the desired signal intensity. Additionally, a Z-step size of 0.5 μm may be used for a better image quality for nuclear size measurement. See Troubleshooting Problem 3.

Note: The outlined method for dissecting adult fly intestine is intended as a starting point for potential modifications. Each lab may have different ways of preparing flies and dissecting the intestines for immunohistochemistry. Lab-specific tools and methods for dissecting the Drosophila intestine can replace what is outlined here.

Intestinal cell nuclear size analysis in ImageJ: Day 4

Timing: 5–10 min per posterior midgut

• 13.Import Z-stack images into Fiji (ImageJ).
• 14.
  Perform nuclear size analysis. See Figure 3 for a step-by-step visual guide.

  • a.
    Make a duplicate stack of a 50 μm × 50 μm region.

    • i.
      Using the rectangle tool, create a 50 μm × 50 μm box.
    • ii.
      Place the box near one end of the intestine.
    • iii.
      Create a duplicate Z-stack image of the selected region (Image – Duplicate). Make sure to check on the “Duplicate hyperstack box”.

      Note: We will take nuclear size measurements from 4 – 50 μm × 50 μm regions across the image to cover the length of the intestine. Since the control “wild-type” intestine is narrower in gut width compared to the intestinal tumor, make sure not to capture overlapping regions when making 50 μm × 50 μm duplicated regions.

  • b.
    Make a Z-stack projection (max intensity) of the most basal layer of the front epithelial layer. See Troubleshooting Problem 4.

    Note: While the control “wild-type” intestine is a monolayer, the intestinal tumor is a multilayer epithelium due to the over-proliferating tumor cells. Therefore, to make a Z-stack projection in the intestinal tumor, we typically use 2–3 stacks capturing the cells in the most basal layer (adjacent to the visceral muscle) of the front epithelium.

  • c.
    Using the Channels Tool (Image – Color – Channels Tool), remove signals from all channels except the DAPI channel, leaving the nuclear signal on the projected image. If necessary, adjust the brightness (Image – Adjust – Brightness/Contrast) of the DAPI signal to obtain an optimum signal. Avoid over-adjusting the brightness to prevent signal saturation.
  • d.
    Create an RGB image of the DAPI channel (Channels Tool – More>> – Create RGB Image).
  • e.
    Smoothen the outlines of nuclei using the Smooth tool under the Process tab.
  • f.
    Convert the RBG image into a binary image (Process – Binary – Make binary).
  • g.
    Separate the conjoint nuclei using the Watershed tool (Process – Binary – Watershed). See Troubleshooting Problem 4.

    Note: Due to tumor cell overcrowding, two adjacent cells might have their nucleus too close to each other, resulting in an optical artifact of conjoint nuclei. The Watershed function is a useful tool to correct this issue. See further in point 14h Note #3.

  • h.
    For standard nuclear size analysis, measure the nucleus area of all GFP⁺ cells in the 50 μm × 50 μm region (Analyze – Analyze particles) (See Note #2 below). If intestinal tumors are stained for specific biomarkers, measure the nucleus area of cells that are positive for the marker.

    Note: Set analyzer for area measurements (Analyze – Set measurements – click on Area).

    Note: Due to tissue overcrowding in the multilayered intestinal tumor, the Z-stack projection image created in step 14b may depict a partial view of cells residing underneath the captured layer. This creates an illusion of cells with smaller nuclei in the projected layer. To avoid incorporating measurements of these fractional nuclei, we compare the binary image to the duplicated Z-stack from step 14a and only take nuclear size measurements of those cells that are wholly captured at the projected plane.

    Note: While the Watershed tool works well in most cases, an incorrect or failure in component segmentation might occur. Hence, it is important to verify the resulting segmentation to the original Z-stack projection image created in step 14b and manually exclude those mis-segmented particles from measurements.

    Note: Due to the signal/background brightness adjustment, the particle analyzer tool might pick up random artifacts and include them in the area measurements. These area measurements are typically smaller than a representative nuclear size would be (in the 0–2 μm² range). To eliminate those false measurements, we compare the binary image to the original Z-stack projection image created in step 14b and manually filter out the artifacts. Alternatively, set the minimum size limit to 2 μm² instead of 0 μm² on the Analyze Particles panel. However, manual validation of the measured particles is still highly recommended.

  • i.
    Copy the measurement results to an Excel sheet for data processing.
  • j.
    Repeat step 14a-i for three additional 50 μm × 50 μm regions to cover the length of the intestinal epithelium.
• 15.Process nuclear area measurements into a histogram using the Data Analysis Plug-in on Microsoft Excel, starting at a nuclear size of 2 μm² up to 43 μm² with a bin size of 1 μm². Normalized cell frequency to the total number of cells measured in each sample. See Figure 4.

Note: Histogram’s start bin value, end bin value, and bin range are arbitrary and can be adjusted. See Troubleshooting Problem 5.

Note: We recommend measuring the nuclear area of at least 200 cells per sample group. See Troubleshooting Problem 5.

Figure 3.

Step-by-step guide of nuclear size analysis (Step 14) in Fiji

Figure 4.

Nuclear size analysis histograms displaying cell distribution in esg^ts>+ and esg^ts>yki^3S/A intestines

(A-A″) Nuclear size distribution of GFP⁺ (A), Su(H)GBE⁺ (A′), and Dl⁺ (A″) cells in control (esg^ts>+) midguts. Numbered red brackets in (A) indicate three different cell populations described in the text.

(B-B″) Nuclear size distribution of GFP⁺ (B), Su(H)GBE⁺ (B′), and Dl⁺ (B″) cells in esg^ts>yki^3S/A midguts. Transgenes are induced for 6 days. Figure reprinted and adapted from Pranoto et al.¹

Expected outcomes

This protocol describes a tool to analyze the cellular heterogeneity in intestinal tumors and its deviation from that in the normal intestine by measuring the nuclear size as a proxy of cell size. The nuclear size analysis on the GFP⁺ cells (esg⁺ cells) in wild-type intestines illustrates the degree of cellular heterogeneity for ISCs and EBs (Figure 4A). Meanwhile, nuclear size analysis of GFP⁺ cells in intestinal tumors represents the heterogeneity of tumor cells. Additionally, a combination of nuclear size analysis and immunohistochemistry for detecting the markers of the four different cell types in wild-type intestines can further extend the analysis.

A successful nuclear size analysis will present a histogram of nuclear size distribution for each cell population of interest (Figure 4). The nuclear size distribution of GFP⁺ cells (Esg⁺ cell) peaked at around 12 μm² and displayed a discernable population at around 7 μm² and a trailing nuclear population on the right side of the peak (Figure 4A, indicated with brackets with I, III, and II, respectively). Su(H)GBE⁺ cells (Figure 4A′) and Dl⁺ cells (Figure 4A″) display distinguishable histograms. The nuclei corresponding to the population III in GFP⁺ nuclei are largely missing in Su(H)GBE⁺ cells (Figure 4A′). In contrast, the nuclei corresponding to the population II in GFP⁺ nuclei are not present in the histogram for Dl⁺ cells (Figure 4A″). The GFP⁺ nuclei population I is reminiscent of the peak at 12 μm² in the Su(H)GBE⁺ nuclear size distribution (Figure 4A′, left bracket).

The trailing nuclei population in the Su(H)GBE⁺ nuclear size distribution (Figure 4A′, right bracket) resembles the population II in the GFP⁺ nuclei distribution. The Dl⁺ nuclei population at the major peak (Figure 4A″, left bracket) resembles the population I in the GFP⁺ nuclei distribution. The nuclei population resembling population III in the GFP+ nuclei distribution is also present in the Dl+ nuclei distribution (Figure 4A″, left bracket). Thus, we can infer how EBs and ISCs constitute the GFP⁺ cells by comparing their nuclear size distributions (Figure 4A-A″). The distribution of yki^3S/A nuclei size peaks at 15 μm² and displays a larger trailing nuclei population. Many of the nuclei are considerably larger than the GFP⁺ wild-type nuclei (Figure 4B). The nuclear size analysis enables us to assess the heterogeneity of the subpopulation of yki^3S/A cells defined by Dl or Su(H)GBE signals (Figure 4B-B″). A similar analysis for GFP⁺, Dl⁺, and Su(H)GBE⁺ yki^3S/A nuclei distributions can be performed to gain additional insights into how Su(H)GBE⁺ and Dl⁺ yki^3S/A tumor cells contribute to the overall tumor cell heterogeneity (Figure 4B-B″).

The figures in our original paper are adopted and modified to illustrate the expected outcome of nuclear size analysis. For a complete interpretation of nuclear size analysis on wild-type control and yki^3S/A intestines, refer to Pranoto et al.¹

Limitations

Our methodology employs advanced confocal microscopy imaging, capable of capturing up to four fluorescent channels per sample. Since DAPI is used for nuclear size analysis and GFP is used for marking wild-type esg⁺ cells or tumor cells, the red and the far-red channels can be used for detecting the signals for cell-type markers. This limitation hampers the ability to detect aberrant tumor cells that may express more than two cell-type markers concurrently.

For a more comprehensive assessment of intestinal tumor cell composition, we recommend adopting additional advanced techniques, such as flow cytometry or single-cell sequencing. These techniques offer a broader analytical scope, surpassing some constraints of the current protocol and enabling a molecular definition of tumor cell types.

Troubleshooting

Problem 1

Poor tissue quality (related to Step 2 and Step 10).

Potential solution

Since the intestinal tissue is an internal soft tissue, poor tissue handling can compromise the integrity of the intestine.

• •Handle dissection forceps and pipette tips with caution to prevent intestinal tissue piercing. Avoid including damaged intestines in the analysis, as it may lead to inaccurate nuclear size measurements.
• •Keep intestinal tissues hydrated at all steps by working quickly during buffer exchanges and preventing buffer evaporation.
• •Avoid stretching the tissue, especially during dissection and mounting, as intestines may become entangled. To untangle them, gently lift one intestine over another instead of pulling them through.

Problem 2

Obstructed view of the posterior midgut during confocal imaging and image analysis (related to Step 10).

Potential solution

The adult Drosophila intestinal tract is naturally folded to fit the abdominal cavity. Although the intestines are uncoiled during dissection, they tend to revert to their original state upon mounting, which may obscure the view of the posterior midgut region. Hence, it poses challenges in taking a proper Z-stack image of the intestine. To address this issue, we recommend carefully aligning the intestines parallel to each other. Before placing the coverslip, make sure to re-straighten the intestines.

Problem 3

Weak DAPI signals (related to Step 12).

Potential solution

This protocol relies on DAPI signals as a proxy for nuclear size. While minor adjustments in DAPI brightness using Fiji are generally effective, substantial corrections due to a weak DAPI signal can lead to imprecise delineation of nuclear boundaries, a cause of inaccuracy in nuclear area measurements. To ensure optimal DAPI staining across all intestinal samples, it is recommended to limit the number of intestines to 15 per staining. If more intestines are assessed, consider increasing the DAPI concentration during staining and use an adequate volume of Vectashield mounting medium for signal preservation. Additionally, setting optimal laser power during confocal imaging is crucial to capture appropriate DAPI signals.

Problem 4

Overlapping nuclei in the 50 μm × 50 μm duplicated region (related to Step 14b and 14g).

Potential solution

The over-proliferation of tumor cells within the intestinal tumor leads to the formation of a multilayer epithelial layer. In conducting nuclear size analysis, we create a duplicate image of a 50 μm × 50 μm region and generate a Z-stack projection to capture the nuclei of cells residing in proximity to the visceral muscle, aiming to include as many nuclei as possible in the projected image. Increasing the number of stacks will enhance the count of nuclei in the projection, however, it will also result in adjacent nuclei being projected as a single particle that often cannot be segmented using the watershed tool. To overcome this issue, for each 50 μm × 50 μm duplicated region, determine the fewest number of stacks to achieve an optimal Z-stack projection with the higher number of individually segmented nuclei particles.

Problem 5

Indistinguishable peaks of cell types in the nuclear size distribution histogram (related to Step 15).

Potential solution

For optimal results in generating a nuclear size distribution with prominent peak patterns for each cell type, we suggest analyzing a minimum of 200 cell nuclei. However, in the case of intestinal tumors, these distinctive peaks may be less recognizable due to the high total cell number and robust cell heterogeneity. To enhance the cell type patterns in the histogram, we recommend increasing the total cell number for nuclear size analysis. Furthermore, adjusting the histogram bin size to a smaller increment is recommended to facilitate a more precise observation of peak patterns in the nuclear size distribution.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Young V. Kwon (ykwon7@uw.edu).

Technical contact

Further information on the protocol and analysis may be directed to the technical contact, Inez K.A. Pranoto (inezp@uw.edu).

Materials availability

This study did not generate new unique reagents.

Data and code availability

This study did not generate/analyze datasets/code.