1. Differential Fluorescent Cell Labelling

1. Gene tagging with CRISPR Cas9
   1. Use CRISPR Cas9 genome editing to tag rootletin (or other genes of interest) with the fluorescent proteins meGFP or mScarlet-I in human cancer cell lines.
      NOTE: Detailed protocols for genome editing are covered elsewhere^24,25,26.
2. Fluorescent dye labelling
   1. Grow Cal51 human cancer cells in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS), L-glutamine, and 100 µg/mL penicillin/streptomycin at 37 °C and 5% CO₂ in a humidified incubator.
      NOTE: It is important to ensure that cells are regularly checked for the absence of mycoplasma contamination, and for the identity of the cell line. Do not use cells that have been extensively passaged in culture (>15 passages). Cells must be healthy and growing exponentially. Avoid under-confluency or over-confluency.
   2. Seed each cell type (rootletin-meGFP, rootletin-mScarlet-I and parental untagged Cal51 cells) such that ~6 x 10⁶ cells are present the next day for each sample, at a confluency of 70−90%.
      NOTE: This is approximately one T75 tissue culture flask or one 10 cm dish per cell type. The parental untagged cells are required as a negative control later in the protocol. This protocol allows for the production of ~20,000 fused cells, but this number could be increased through scaling up the protocol with an increased number of batches.
   3. The next day, prewarm DMEM, trypsin and 1x phosphate-buffered saline (PBS) by placing them in a water bath at 37 °C.
   4. Wash rootletin-meGFP and rootletin-mScarlet-I cells 2x in PBS by pouring or aspirating off medium and replacing with 10 mL of PBS.
   5. Label Cal51 rootletin-meGFP cells by adding 500 nM violet cell dye (Table of Materials) in PBS for 1 min at room temperature (RT).
   6. Label Cal51 rootletin-mScarlet-I cells by adding 200 nM far red cell dye (Table of Materials) in PBS for 1 min at RT.
      NOTE: Keep samples shielded from light wherever possible after fluorescent staining, to avoid photobleaching of fluorescence.
   7. Stop the dye labelling reactions by adding 10 mL of DMEM for 5 min.
   8. Remove unlabeled parental Cal51 cells from the incubator (from step 1.2.2).
      NOTE: These cells will be used later as unlabeled negative controls (step 3.3.2).
   9. Wash all cells once by pouring off growth medium and replacing with 10 mL of PBS.
   10. Pour off PBS and incubate for 5 min at culture conditions with 1 mL of pre-warmed 1x trypsin.
   11. Transfer cells to a 15 mL plastic conical tube and centrifuge at 1000 x g for 5 min.
   12. Carefully remove and discard trypsin from the cell pellet with a pipette.
   13. Gently resuspend the violet and far red labelled cell pellets in 1 mL of PBS.
   14. Gently resuspend the parental (non-fluorescent) cell pellet in 1 mL of DMEM and return to the incubator.
       NOTE: This sample will not undergo cell-cell fusion but will be used later during flow cytometry.

2. Cell-cell Fusion

1. Mix 3 x 10⁶ violet labelled cells with 3 x 10⁶ far red labelled cells in a 15 mL tube by gently pipetting together 0.5 mL of each cell type using a 1 mL pipette.
2. Centrifuge the remaining unmixed cells from step 2.1 at 1,000 x g for 5 min, then pour off PBS, resuspend the pellet in 1 mL of DMEM and return to the incubator.
   NOTE: These remaining unmixed single-labelled samples will be used later as negative and compensation controls in section 3 of the protocol.
3. Centrifuge the mixed cells from step 2.1 at 1,000 x g for 5 min and carefully aspirate PBS using a 1 mL pipette.
4. Add 0.7 mL of 50% 1450 PEG solution in a dropwise fashion to the cell pellet (step 2.3) using a 1 mL pipette over a period of 30 s.
5. Leave for 3.5 min at RT.
   NOTE: Incubation time with PEG may be optimized depending on cell type, although note that longer incubation time increases cell toxicity.
6. Add 10 mL of serum-free DMEM dropwise for 30 s and leave in the incubator for 10 min.
7. Spin down at 1,000 x g for 5 min, discard supernatant, and resuspend gently in 1 mL of complete medium (DMEM containing FBS).
   NOTE: Proceed directly to section 3. There is toxicity associated with PEG exposure followed by flow cytometry and imaging, so keep cells in culture conditions as much as possible by proceeding rapidly between steps.

3. Fluorescence-activated Cell Sorting (FACS) to Enrich Fused Cells

1. Prepare all cells for flow cytometry by gently pipetting each sample separately through a 70 µm filter into a FACS tube.
   NOTE: Four samples are required: fused cells, single labelled far red cells, single labelled violet cells, unlabeled parental cells. Once removed from the sterile tissue culture hood, cells have the capacity to become infected, so minimize the exposure time of samples to a non-sterile environment from here onwards. Cells are sorted in complete DMEM medium.
2. Use a cytometer capable of single cell sorting.
   1. Set up the cell sorter for an aseptic sort. Perform the instrument quality control as per the manufacturer's recommendation.
   2. Draw a plot of forward scatter versus side scatter and a doublet discrimination plot.
3. Create gates to identify fused cells.
   1. Excite the fluorescent dyes at wavelengths of 405 nm and 635 nm. Detect at 450 nm and 660 nm (violet and far-red wavelengths, respectively). Plot far red versus violet wavelengths.
   2. Run unlabeled parental cells through the cytometer and record the fluorescence intensity.
      NOTE: Values above this baseline define positivity for dye staining.
   3. Run single labelled violet cells through the cytometer. Confirm that there is no spill-over of violet into the far red channel.
   4. Run single labelled far red cells through the cytometer and confirm that there is no spill-over of far red into the violet channel.
      NOTE: For the representative data shown, cells were sorted at 20 psi on a sorter equipped with a 100 µm nozzle.
   5. Briefly run the fusion sample to confirm that fused cells are visible with the gates created in step 3.3.
4. Align the sorting stream centrally into an 8-well imaging dish.
5. FACS sort fused cells, present as double positive violet and far red labelled cells directly into an 8-well imaging dish containing 100 µL of growth medium.
   NOTE: The maximum recommended number of cells is ~50,000 per 8-well dish. Ensure cell health during sorting. Cell confluency can influence cell health so keep cells between 20−90% confluency post-sorting by sorting an appropriate number of cells for the imaging dish. Each sorted droplet contains approximately 3 nL of sheath fluid. Ensure that sheath fluid does not significantly dilute the growth medium during sorting.
6. Directly after sorting, take cells back to the incubator and leave for >2 h or overnight.
   NOTE: Adherent cells will gradually re-adhere to the coverslip during this period, from sorting onwards, and hence their morphology will change over time if taken directly to the microscope.

4. Immunofluorescent Staining and Imaging of Cell-cell Fusions

NOTE: Fused cells can be imaged live or after fixation and further fluorescent staining (or both), depending on the experiments and measurements required.

1. Live cell imaging
   1. Replace growth medium with imaging medium without phenol red (Table of Materials) and proceed directly to imaging.
2. Fixation and staining
   1. Prepare fresh 4% paraformaldehyde (PFA) in PBS.
   2. Fix cells by incubating them in 100 µL of 4% PFA for 15 min at RT. Remove PFA after 15 min.
      CAUTION: PFA is toxic when inhaled so this step should be performed in a fume hood with appropriate personal protective equipment.
      NOTE: After fixation, the experiment can be paused and restarted later if required. Store the sample at 4 °C protected from light if required.
   3. Wash cells 3x in 200 µL of PBS at RT.
   4. Permeabilize cells for 10 min at RT in 200 µL of 0.1% nonionic surfactant (Table of Materials) and 0.1% nonionic detergent (Table of Materials) diluted in PBS.
   5. Block in 200 µL of 3% bovine serum albumin in PBS for 30 min at RT.
   6. Incubate with antibodies in 150 µL of PBS containing 3% bovine serum albumin and 0.1% nonionic surfactant and 0.05% nonionic detergent for 1 h at RT.
      NOTE: The primary antibodies used to generate the representative results are fluorescently conjugated anti-GFP nanobody (used at 1:400 dilution), and fluorescently conjugated anti-mScarlet-I nanobody (used at 1:500 dilution). These enhance the signal of the already fluorescent proteins.
   7. Wash 2x for 5 min in 300 µL of PBS and then leave in 200 µL of PBS.
      NOTE: Samples are imaged in PBS. Samples can be stored at 4 °C protected from light.
3. Image acquisition
   1. Acquire images with an appropriate fluorescence microscope capable of four-color imaging (e.g., confocal, widefield, structured illumination).
   2. Excite meGFP and mScarlet-I channels with 488 nm and 561 nm wavelength lasers, respectively. Set detectors to detect at ~505−550 nm and ~590−650 nm. Excite violet and far red dye channels with 405 nm and 633 nm lasers, respectively. Set detectors to detect at ~430−500 nm and ~660−750 nm.
   3. Empirically determine laser intensity such that significant photobleaching does not occur and that cells remain healthy (if live cell imaging).
   4. Empirically determine gain such that signal is obtained without saturating or artificially clipping pixel values.
   5. Collect Z-stack data, such that images are three-dimensional, covering a range of 20−60 µm with ~500 µm step size.
      NOTE: The images shown in representative data were acquired by either confocal (Figure 2B), Airyscan (Figure 2C) or structured illumination microscopy (Figure 2D). Each cell is a bona fide heterokaryon if it contains both violet and far red dyes overlapping in the cytoplasm.