Live-Cell Time-Lapse Imaging of Scratch-Wound Collective Migration in MDCK Epithelial Monolayers with PIV-Based Velocity Field Analysis

Three independent biological replicates on separate days. Within each, 4 technical replicate wells per condition (control vehicle, 10 µM Y-27632) in a standardized scratch assay. Imaging spans 18 h with frames every 10 min at 37 C / 5% CO2 in an environmental stage chamber. Wound width is standardized using a culture-insert (500 µm gap) rather than a manual scratch to reduce edge-roughness variability. Two non-overlapping positions per well are imaged via a motorized stage. Wells are randomized across the plate to distribute positional/edge effects.

BSL-2 for MDCK culture; work in a biosafety cabinet. Mitomycin C is a potent mutagen/carcinogen — handle in a fume hood with double nitrile gloves and dedicated chemical waste; weigh and dissolve with minimal aerosolization. Y-27632 is a research chemical; avoid skin contact. HEPES-buffered medium reduces CO2 dependence but should still be handled sterilely. Phase-contrast LED illumination poses minimal hazard. Biological waste autoclaved or 10% bleach treated; mitomycin waste segregated per institutional hazardous-waste policy. Wear gloves, lab coat, eye protection.

Vehicle control: matched volume of water/DMSO in which Y-27632 is dissolved. Proliferation control: paired wells with and without mitomycin C to confirm that closure reflects migration, not division. Positive migration control: untreated confluent monolayer with a known closure rate (~25-35 µm/h for MDCK). Negative/no-gap control: an uninjured confluent region imaged in parallel to confirm baseline stability and absence of drift artifacts. A cell-free medium-only position controls for autofocus and illumination drift.

Control monolayers should close the 500 µm gap in roughly 14-18 h at ~25-35 µm/h with a smooth, coherent leading edge (order parameter ~0.6-0.8). Y-27632-treated wells should show faster closure (~35-45 µm/h) but a rougher, less coordinated front (order parameter ~0.3-0.5). PIV velocity fields should reveal organized front-directed flow in controls and more disordered, swirling flow under ROCK inhibition. Good data have stable focus throughout, <5% evaporation-driven medium-osmolarity change, and reproducible closure curves across technical replicates.

• MDCK (NBL-2) canine kidney epithelial cell line, mycoplasma-tested
• DMEM high-glucose, 10% FBS, 1% penicillin/streptomycin, 2 mM L-glutamine
• Culture-Insert 2 Well (defined 500 µm gap) or sterile p200 tip for manual scratch
• Y-27632 dihydrochloride (ROCK inhibitor), 10 µM working
• Mitomycin C, 10 µg/mL (optional proliferation block, 2 h pre-scratch)
• Phenol-red-free DMEM with 25 mM HEPES + 10% FBS for imaging medium
• 35 mm glass-bottom dishes or #1.5 glass-bottom 24-well plates
• PBS (Ca2+/Mg2+-free)
• Mineral oil or sealed lid to limit evaporation (if no humidified chamber)
• 0.05% trypsin-EDTA for passaging

To quantify collective cell migration dynamics during scratch-wound closure in a confluent MDCK monolayer using long-duration live-cell phase-contrast imaging. The assay yields both a population-level wound-closure rate (µm/h) and a spatially resolved velocity field via PIV, enabling assessment of how the ROCK inhibitor Y-27632 alters migration speed and front coordination.

1. Seed MDCK at 3 x 10^5 cells/well into culture-inserts placed in glass-bottom wells; grow to a confluent monolayer (~24-36 h).
2. Optional: treat with 10 µg/mL mitomycin C for 2 h to suppress proliferation, then wash 3x PBS.
3. Remove the culture-insert with forceps to create a clean 500 µm cell-free gap; gently wash once with PBS to remove debris.
4. Replace medium with phenol-red-free imaging medium containing 10 µM Y-27632 or vehicle.
5. Mount on a pre-equilibrated (37 C, 5% CO2, humidified) motorized stage; allow 15 min thermal equilibration to minimize focus drift.
6. Define 2 stage positions per well; enable hardware autofocus (definite-focus) to maintain focus over 18 h.
7. Acquire 10x phase-contrast images every 10 min for 18 h.
8. At endpoint, optionally fix and stain for F-actin/DAPI to confirm monolayer integrity.
9. Export time-lapse as a TIFF stack per position for analysis.

Independent variable: ROCK inhibition (10 µM Y-27632 vs vehicle). Dependent variables: wound-closure rate (µm/h), percent wound area closed over time, mean migration speed from PIV (µm/h), and the velocity order parameter (directional coherence, 0-1). Controlled variables: initial gap width (500 µm insert), monolayer confluency at scratch, imaging medium and HEPES buffering, temperature/CO2/humidity, frame interval, objective and illumination, and proliferation state (mitomycin C).

We hypothesize that ROCK inhibition with 10 µM Y-27632 increases the wound-closure rate by reducing rear-edge retraction tension while decreasing the directional coherence (order parameter) of the leading front, consistent with a more fluid but less coordinated migration mode compared to vehicle-treated monolayers.

Wound area per frame is segmented in Fiji using the MRI Wound Healing Tool or a variance-filter mask; closure rate is the slope of the linear closure phase. Velocity fields are computed with PIVlab (MATLAB) or the Fiji PIV plugin on consecutive frame pairs (interrogation window 64 -> 32 px, 50% overlap). Mean speed and the velocity order parameter (mean normalized velocity vector magnitude) are extracted per frame. Drift is corrected via the StackReg/rigid-body registration before PIV. All segmentation thresholds and PIV parameters are fixed across conditions and reported.

1. Focus drift over 18 h: enable hardware definite-focus, allow full thermal equilibration before imaging, and use a sealed/humidified chamber. 2. Ragged or variable initial wounds: use a defined-gap culture insert instead of manual scratching. 3. Closure dominated by proliferation rather than migration: add mitomycin C and confirm with a no-mitomycin comparison. 4. Medium evaporation increasing osmolarity and slowing cells: use a humidified chamber or mineral-oil overlay and monitor edge wells. 5. PIV noise from debris: pre-wash to remove floating cells and apply a debris-size filter and image registration before analysis.

Per-well closure rates and mean order parameters (n = 4 technical replicates nested within 3 biological replicates) are compared with a linear mixed-effects model or, if simplified, an unpaired two-tailed t-test on biological-replicate means (n = 3) at alpha = 0.05. Closure curves are compared by fitting and comparing slopes (ANCOVA). With 3 biological replicates and an expected ~30% rate difference (effect size d ~2), power exceeds 0.85. Technical replicates are averaged before the biological-level test to avoid pseudoreplication.