A transwell invasion assay measures how many cells can degrade an extracellular-matrix barrier and cross a porous membrane toward a chemoattractant, which makes it the standard in vitro model of the first step of metastasis. This workflow covers how to choose between a scratch assay, a chemotaxis assay, an ECM-coated transwell and a 3D spheroid format; an eight-step transwell protocol; cell migration analysis including the invasion index; troubleshooting; and the ELISA readouts that connect a migrated-cell count to its mechanism.
Migration vs. Invasion: What a Transwell Invasion Assay Actually Measures
Cell movement drives wound repair, immune surveillance, development and metastasis. In vitro cell migration and invasion assays measure that movement under controlled conditions, but the two terms are not interchangeable, and the assay design differs accordingly.
Migration is movement of cells across a two-dimensional surface or through an uncoated porous membrane. It answers the question: can these cells move, and how fast, in response to a stimulus? An in vitro cell migration assay typically uses an uncoated membrane or an open scratch in a monolayer.
Invasion adds a barrier. Cells must first degrade and remodel an extracellular matrix (ECM) layer, usually a basement membrane extract coated on the membrane, before they can pass through. Invasion therefore reports a composite phenotype: motility plus proteolytic ECM degradation, driven largely by matrix metalloproteinases (MMPs) and regulated by their tissue inhibitors (TIMPs) (Kessenbrock et al., Cell 2010).
A third variable is directionality. When cells move toward a soluble gradient (a growth factor, chemokine or serum in the lower chamber) the experiment is a chemotaxis assay. When cells move randomly with no gradient, the readout is chemokinesis. Most transwell experiments are, in practice, chemotaxis assays, because the chemoattractant gradient between chambers is what drives cells through the membrane.

Why it matters for interpretation: a compound that reduces “invasion” in a coated transwell may be blocking motility, ECM degradation, gradient sensing, or simply proliferation and viability. A well-designed workflow includes parallel migration (uncoated) and viability controls so each component can be separated. For a comparison of viability readouts to pair with the assay, see Cell Viability Assay Types: What Each Measures and When to Use It.
Choosing the Format: Scratch Assay, Chemotaxis Assay, Transwell Invasion or 3D Spheroid
Whether the plan is a commercial cell migration assay kit or an assembled in-house setup, the underlying formats fall into four families. Choose based on the biological question, throughput and available imaging equipment.
Wound healing (scratch) assay — 2D migration, label-free, lowest cost
Timeframe 8–48 h · Readout: % gap closure · Equipment: brightfield microscope. A confluent monolayer is scratched with a pipette tip (or grown around a removable insert for reproducible gap width), and closure of the gap is imaged over time. Simple and inexpensive, but confounded by proliferation over longer timeframes; mitomycin C pretreatment or serum reduction is commonly used to suppress division. Best for: collective migration of adherent cells; quick screening of treatments before committing to transwell formats.
Transwell migration / Boyden chamber chemotaxis assay — directed migration, uncoated membrane
Timeframe 2–24 h · Readout: cells on the underside of the membrane · Pore size 3–8 µm. Cells seeded in serum-free medium in the upper chamber migrate through a microporous membrane toward a chemoattractant in the lower chamber. The steepness of the gradient, pore size and incubation time are the main tuning parameters. This is the reference format for a quantitative chemotaxis assay and the “migration control” arm for invasion experiments. Best for: chemokine and growth-factor response studies; immune cell chemotaxis (3–5 µm pores); the migration baseline for calculating an invasion index.
Transwell invasion assay (ECM-coated) — migration plus ECM degradation, the metastasis model
Timeframe 24–72 h · Readout: cells that cross the ECM layer · Coating: basement membrane extract, collagen I or fibronectin. Identical geometry to the migration format, but the membrane is coated with a reconstituted basement membrane extract or a defined ECM protein. Only cells that secrete active proteases and remodel the matrix reach the underside. Run an uncoated migration insert in parallel with the same cells and chemoattractant: the ratio of invaded to migrated cells is the invasion index, which normalizes invasion to intrinsic motility. Best for: metastasis and EMT studies; MMP inhibitor characterization; comparing invasive potential across cell lines or knockdowns.
3D spheroid invasion assay — physiologic architecture, imaging-heavy
Timeframe 3–7 days · Readout: invasive outgrowth area · Equipment: inverted or confocal imaging. Tumor spheroids embedded in a 3D matrix extend invasive protrusions outward over days. More physiologically relevant than membrane formats (cell–cell junctions, hypoxic cores and matrix mechanics are all represented) but slower, harder to quantify and more sensitive to matrix lot variability. Best for: validating hits from transwell screens; studying collective versus single-cell invasion modes.
Format comparison at a glance
| Format | Measures | Gradient? | ECM barrier? | Typical duration | Throughput |
|---|---|---|---|---|---|
| Scratch / wound healing | 2D collective migration | No | No | 8–48 h | High |
| Transwell migration (chemotaxis) | Directed single-cell migration | Yes | No | 2–24 h | Medium–high |
| Transwell invasion | Migration + ECM degradation | Yes | Yes | 24–72 h | Medium |
| 3D spheroid invasion | Invasion in 3D architecture | Optional | Yes | 3–7 days | Low–medium |
Pore-size guidance for transwell formats: 8 µm for most epithelial and tumor lines; 5 µm for lymphocytes and monocytes; 3 µm for neutrophils. Larger pores with small cells allow passive drop-through and inflate background.

Transwell Invasion Assay Protocol: Eight Steps From Coated Insert to Stained Membrane
The workflow below describes a standard 24-well transwell invasion experiment with an uncoated migration control. Volumes and cell numbers are starting points; optimize per cell line.
- Prepare and rehydrate coated inserts. Thaw basement membrane extract on ice and coat inserts per the manufacturer's protocol (pre-coated inserts: rehydrate with warm serum-free medium for 1–2 h at 37 °C). Keep everything cold during handling; the extract gels above about 10 °C. Consistency: coating thickness is the single largest source of well-to-well variability. Use one BME lot per experimental series and record lot numbers.
- Serum-starve the cells. Culture cells to 70–80% confluence, then switch to serum-free or low-serum (0.1–0.5%) medium for 12–24 h before the assay. Starvation synchronizes the population and prevents residual serum from flattening the chemoattractant gradient. Check: confirm >90% viability after starvation (trypan blue or equivalent). Dead and dying cells fall through pores and score as false migrants.
- Harvest and seed the upper chamber. Detach cells gently, using an enzyme-free dissociation buffer where possible, because harsh trypsinization strips surface receptors needed for gradient sensing (see Trypsinization Protocol: Passaging Adherent Cells Step by Step). Resuspend in serum-free medium and seed 2.5 × 104 to 1 × 105 cells in 100–200 µL per insert.
- Add chemoattractant to the lower chamber. Fill the lower well with 500–750 µL of medium containing the chemoattractant. 10% FBS is the standard general-purpose attractant; defined growth factors or chemokines are used when the receptor pathway is the question. Avoid air bubbles under the membrane; they block migration locally. Controls: include a no-attractant well (serum-free below) for random-migration background, and a serum-free-above / serum-free-below pair to confirm the response is gradient-driven. This is the defining control of a chemotaxis assay.
- Incubate. 37 °C, 5% CO2. Uncoated migration inserts: 4–24 h. Coated invasion inserts: 24–72 h, since cells must degrade the ECM layer first. Pilot a time course for each new cell line; too short underestimates, too long saturates and re-adds proliferation as a confounder.
- Remove non-migrated cells. After incubation, swab the upper surface of each membrane firmly but gently with a cotton swab (twice, rotating) to remove cells that never crossed. Incomplete swabbing is the most common cause of inflated counts.
- Fix and stain. Fix membranes in methanol or 4% paraformaldehyde (5–15 min), then stain cells on the underside. Crystal violet is the classic endpoint stain; fluorescent nuclear or cytoplasmic dyes enable plate-reader quantification. Rinse thoroughly to reduce background.
- Quantify. Count stained cells in at least 5 random fields per membrane at 10–20× and average, image the whole membrane for software-based counting, or elute crystal violet in 10% acetic acid and read absorbance at 570–590 nm for a bulk readout. Run every condition in at least triplicate inserts.
Cell Migration Analysis: Invasion Index, Normalization and Counting Rules
Sound cell migration analysis is what separates a publishable result from a picture of purple dots. Three principles apply across formats (Pijuan et al., Front Cell Dev Biol 2019).
1. Report the right metric per format
- Scratch assay: percent wound closure over time, or gap area remaining at fixed timepoints, measured with image-analysis software from the same field imaged repeatedly.
- Transwell migration: mean migrated cells per field (with field count and magnification stated), or fold change versus the no-attractant control.
- Transwell invasion: the invasion index, i.e. invaded cells (coated insert) divided by migrated cells (matched uncoated insert). This normalizes for intrinsic motility differences between conditions and is the metric reviewers expect (Justus et al., J Vis Exp 2014).
- Spheroid invasion: invasive area (total outgrowth area minus initial spheroid core area) over time.

2. Normalize for proliferation and viability
Any treatment that slows division or kills cells will “reduce migration” artifactually over long incubations. Pair the assay with a parallel viability or proliferation measurement under identical treatment, and either normalize migrated counts to it or keep incubation times short enough (<24 h for most lines) that division is negligible.
3. Pre-register the counting rules
Decide before unblinding: number of fields, field selection method (random or fixed grid), what counts as a cell (nucleus visible on the underside), and how membrane-edge artifacts are excluded. Blind the counter to condition wherever manual counting is used. The same logic applies to any readout; see Planning a Signal Detection Experiment: Controls, Timing, and Readouts.
Invasion Assay Troubleshooting: Common Problems and Fixes
| Observation | Likely cause | Fix |
|---|---|---|
| Few or no cells on underside | Over-thick ECM coating; incubation too short; low chemoattractant potency; cells too sparse | Reduce coating concentration; extend time course; titrate FBS or attractant; increase seeding density stepwise |
| High background without gradient | Serum carryover in upper chamber; pore size too large for cell type; dead cells falling through | Wash cells twice in serum-free medium before seeding; match pore size to cell diameter; verify post-starvation viability |
| High well-to-well variability | Uneven coating; bubbles under membrane; inconsistent swabbing | Coat on a level surface with cold reagents; seat inserts at an angle to avoid bubbles; standardize swab technique and operator |
| Treatment “blocks invasion” but also looks toxic | Cytotoxicity or cytostasis confounding the readout | Add a matched viability assay; test at sub-toxic concentrations; report invasion index rather than raw counts |
| Scratch closes at different rates across replicates | Variable scratch width; proliferation contribution | Use insert-based gap creation for uniform width; pretreat with mitomycin C or reduce serum |
Pairing the Invasion Assay With ELISA Readouts: MMP/TIMP, EMT Markers and Migration Factors
A migrated-cell count says that cells invaded; molecular readouts say how. Conditioned media and lysates collected from the same experiment can be assayed with ELISA kits to connect the phenotype to its mechanism, which is often the difference between a descriptive figure and a mechanistic story.
ECM degradation: the MMP/TIMP axis
MMP-9 (gelatinase B) and MMP-2 degrade type IV collagen, the main structural component of basement membranes, and elevated MMP-9 secretion is the canonical molecular correlate of an invasive phenotype (Kessenbrock et al., Cell 2010). Quantifying MMP-9 in conditioned media from the upper chamber, for example with the Human MMP-9 ELISA Kit PicoKine® (EK0465), directly links invasion counts to protease output. The balance matters as much as the absolute level: tissue inhibitors of metalloproteinases oppose MMP activity, so pairing it with a TIMP measurement such as the Human TIMP-4 ELISA Kit EZ-Set™ (EZ0524) captures the net proteolytic balance rather than one side of it.
EMT status: the cadherin switch
Epithelial–mesenchymal transition (EMT) is the transcriptional program most often underlying gains in invasiveness. Its signature event is the cadherin switch: loss of E-cadherin, gain of N-cadherin (Loh et al., Cells 2019). Measuring both in lysates from the assayed cells, using the Human E-Cadherin/CDH1 ELISA Kit PicoKine® (EK0561) and the Human N-Cadherin/CDH2 ELISA Kit PicoKine® (EK0669), positions any invasion change on the EMT spectrum. Secreted fibronectin (Human Fibronectin ELISA Kit PicoKine®, EK0349) is a complementary mesenchymal marker measurable in the same conditioned media.
Migration-promoting factors
Several secreted and intracellular proteins act directly on the migration machinery and can serve as mechanistic or biomarker readouts alongside the assay: MIEN1 (Migration and Invasion Enhancer 1), which promotes filopodia formation and cancer cell motility; CEMIP (Cell Migration-Inducing Protein), a hyaluronan-degrading driver of migration; and MIF (Macrophage Migration Inhibitory Factor), a chemotaxis-regulating cytokine relevant when immune cell movement is the subject of the assay.

Related ELISA Kits at BioHippo: Ordering Information
All kits below are stocked on ebiohippo.com and were verified against the live catalog on 9 September 2026. Prices are list prices at that date; for multi-kit or bulk pricing, request a quote.
| Product | Vendor | Catalog # | Use in this workflow | Price |
|---|---|---|---|---|
| Human MMP-9 ELISA Kit PicoKine® | Boster Bio | EK0465 | Quantify gelatinase B in conditioned media as the ECM-degradation readout | $513.97 |
| Human TIMP-4 ELISA Kit EZ-Set™ (DIY Antibody Pairs) | Boster Bio | EZ0524 | Measure the inhibitory side of the MMP/TIMP balance | $515.00 |
| Human E-Cadherin/CDH1/Cadherin-1 ELISA Kit PicoKine® | Boster Bio | EK0561 | Epithelial marker of the EMT cadherin switch | $513.97 |
| Human N-Cadherin-2/CDH2/CD325 ELISA Kit PicoKine® | Boster Bio | EK0669 | Mesenchymal marker of the EMT cadherin switch | $513.97 |
| Human Fibronectin ELISA Kit PicoKine® | Boster Bio | EK0349 | Secreted mesenchymal/ECM marker in conditioned media | $513.97 |
| Human MIEN1 (Migration and Invasion Enhancer 1) ELISA Kit | ELK Biotechnology | ELK7821 | Migration/invasion enhancer readout in lysates | $595.40 (96T) |
| Human CEMIP ELISA Kit | Fine Test | EH9595 | Hyaluronan-linked migration driver | $520.00 (96T) |
| Human MIF ELISA Kit | Fine Test | EH0016 | Chemotaxis-regulating cytokine for immune migration studies | $520.00 (96T) |
Browse the full ELISA kit range, cancer research reagents and ECM and cell adhesion products to build the rest of the panel. Kits are sourced from authorized life science vendors with US-based support.
Transwell Invasion Assay FAQ
What is the difference between a migration assay and an invasion assay?
The barrier. A transwell migration assay uses an uncoated porous membrane and measures motility alone. A transwell invasion assay coats that membrane with basement membrane extract or another ECM protein, so cells must also degrade matrix to cross. Invasion is therefore motility plus proteolysis, and it takes longer (24–72 h versus 2–24 h).
What is the difference between a chemotaxis assay and a transwell migration assay?
Geometry is the same; the distinction is the gradient. A transwell migration experiment becomes a chemotaxis assay when a defined chemoattractant gradient exists between chambers and the design includes gradient controls (attractant-free and attractant-on-both-sides wells) proving that movement is directional rather than random.
How is invasion measured and calculated in a transwell assay?
Count the stained cells on the underside of the membrane (at least 5 random fields per insert, triplicate inserts per condition), or elute the crystal violet and read absorbance. Then calculate the invasion index: invaded cells on the coated insert divided by migrated cells on a matched uncoated insert run in parallel. Raw invasion counts mix motility and ECM degradation; the index isolates the ECM-degradation component, so a treatment that merely slows general motility is not misreported as an invasion blocker.
How long should cells be starved before a migration or invasion assay?
12–24 h in serum-free or low-serum (0.1–0.5%) medium for most adherent lines. Starvation synchronizes the population and stops residual serum in the upper chamber from flattening the gradient. Confirm viability is still above 90% afterwards; dead cells fall through the pores and inflate counts.
How long should an in vitro cell migration assay run?
Uncoated migration: 4–24 h for most tumor lines; as little as 2–4 h for fast-moving immune cells. Coated invasion: 24–72 h. Always pilot a time course on a new line. The goal is the linear phase of migration, before counts saturate and before proliferation contributes meaningfully.
How do you tell invasion from proliferation or toxicity effects?
Run a viability or proliferation assay on the same cells under the identical treatment and incubation time. If the treatment also reduces viability, either normalize migrated counts to the viable-cell number, shorten the incubation, or test at sub-toxic concentrations. Report the invasion index rather than raw counts.
How do you remove non-invading cells from the insert?
After incubation, swab the upper surface of the membrane with a cotton swab, twice, rotating the swab, before fixing. Cells that never crossed sit on top of the ECM layer and would otherwise be stained and counted. Incomplete swabbing is the most common cause of inflated counts.
Should I use a commercial cell migration assay kit or an assembled setup?
Kits bundle pre-coated inserts, standardized ECM lots and matched staining or fluorescence reagents, which mainly buys reproducibility; useful for multi-operator labs and long series. An assembled setup (inserts plus separately sourced ECM and stain) costs less per well and allows custom coatings. Either way, the workflow steps and controls above apply unchanged.
References
- Kessenbrock K, Plaks V, Werb Z. Matrix metalloproteinases: regulators of the tumor microenvironment. Cell. 2010;141(1):52–67. PMID 20371345
- Justus CR, Leffler N, Ruiz-Echevarria M, Yang LV. In vitro cell migration and invasion assays. J Vis Exp. 2014;(88):51046. PMID 24962652
- Loh CY, Chai JY, Tang TF, et al. The E-cadherin and N-cadherin switch in epithelial-to-mesenchymal transition: signaling, therapeutic implications, and challenges. Cells. 2019;8(10):1118. PMID 31547193
- Pijuan J, Barceló C, Moreno DF, et al. In vitro cell migration, invasion, and adhesion assays: from cell imaging to data analysis. Front Cell Dev Biol. 2019;7:107. PMID 31259172
For research use only. Not for diagnostic or therapeutic use.