Transwell Boyden Chamber Invasion Assay Through Matrigel for MDA-MB-231 Breast Carcinoma Cells Driven by a CXCL12 Chemotactic Gradient

Conditions: (1) no chemoattractant (serum-free below), (2) CXCL12 gradient (100 ng/mL below), (3) CXCL12 + AMD3100 (cells pretreated, antagonist present), (4) 10% FBS positive chemoattractant. Each condition uses n = 3 inserts per biological replicate, with N = 3 independent biological replicates (separate cell preps). Five non-overlapping fields per membrane are counted, and field positions are predefined (center + 4 quadrants) to avoid edge bias. Membranes are coded for blinded counting. Matrigel coating thickness and cell input (5 x 10^4) are held constant.

BSL-2 recommended for human tumor cell lines; work in a class II biosafety cabinet. Paraformaldehyde is a sensitizer/carcinogen — use in a fume hood with gloves and goggles; collect PFA waste separately. Crystal violet and methanol are toxic/flammable; handle with care and segregate waste. Matrigel is animal-derived (handle as potentially infectious). Standard PPE: lab coat, nitrile gloves, eye protection. Decontaminate surfaces and aspirated media with 10% bleach.

Negative/random-migration control: serum-free lower chamber (no gradient) measures basal, non-directed invasion. Positive control: 10% FBS lower chamber drives maximal invasion and confirms membrane/Matrigel competence. Specificity control: AMD3100 with CXCL12 tests CXCR4 dependence; a vehicle (water) arm matched to AMD3100 controls for carrier. Migration-vs-invasion control: a parallel set of uncoated inserts (no Matrigel) distinguishes barrier degradation (invasion) from simple motility. Viability check by trypan blue on input cells ensures >90% viability before seeding.

No-gradient wells should show low basal invasion (e.g., ~20-60 cells/field). CXCL12 should increase invasion ~2-4 fold over no-gradient. AMD3100 should suppress CXCL12-driven invasion by >60%, returning counts toward baseline. FBS positive control should give the highest counts. A chemotactic index (CXCL12/no-gradient) > 2 indicates a genuine directed response. Inter-insert CV should be < 25%.

• MDA-MB-231 cells (ATCC HTB-26), passage 6-20
• Corning Transwell inserts, 24-well, 8.0 µm pore PET membrane
• Matrigel Growth Factor Reduced, phenol-red-free, diluted to 300 µg/mL in cold serum-free DMEM (~30 µL/insert)
• DMEM/F-12, serum-free (assay) and with 10% FBS (positive control)
• Recombinant human CXCL12/SDF-1alpha, 100 ng/mL working in serum-free medium
• AMD3100 octahydrochloride (plerixafor), 10 µM working
• 0.5% BSA in serum-free medium (carrier for chemokine)
• Crystal violet 0.5% in 20% methanol (fixation+stain) or calcein-AM for fluorescence readout
• Cotton swabs (to remove non-invaded cells)
• 4% paraformaldehyde
• Trypsin-EDTA 0.25%, Dulbecco's PBS

To measure the number of MDA-MB-231 cells that degrade and invade through a reconstituted basement-membrane (Matrigel) barrier on an 8 µm transwell membrane in response to a CXCL12 chemotactic gradient established in the lower chamber, and to determine whether the CXCR4 antagonist AMD3100 inhibits this directed invasion relative to chemokine-only and no-chemokine controls over 22 h.

1. Thaw Matrigel on ice; coat the upper surface of each insert with 30 µL of 300 µg/mL Matrigel and gel for 1 h at 37 C. Do not let it dry.
2. Harvest MDA-MB-231 (subconfluent), wash in serum-free DMEM/F-12, resuspend at 2.5 x 10^5/mL in serum-free medium with 0.5% BSA.
3. For the antagonist arm, pretreat cells with 10 µM AMD3100 for 30 min at 37 C and keep antagonist in the upper chamber.
4. Add 750 µL of the appropriate lower-chamber medium (serum-free; CXCL12 100 ng/mL; or 10% FBS) to each well.
5. Add 200 µL cell suspension (5 x 10^4 cells) to each insert. Avoid bubbles under the membrane.
6. Incubate 22 h at 37 C, 5% CO2.
7. Remove medium; gently swab the upper (non-invaded) side with a cotton swab to remove cells that did not invade.
8. Fix invaded cells in 4% PFA 10 min, stain with 0.5% crystal violet 15 min, rinse in water, air-dry.
9. Image 5 fields/insert at 20x and count invaded cells; optionally elute crystal violet in 10% acetic acid and read A595 for a bulk readout.

Independent variable: lower-chamber chemoattractant identity and presence of AMD3100. Dependent variable: number of invaded cells per field (or A595 of eluted stain). Controlled variables: Matrigel concentration/volume and gelling time, pore size (8 µm), cell input number, serum-free upper chamber, incubation time (22 h), swabbing technique, fields counted, passage range.

If MDA-MB-231 invasion toward CXCL12 is mediated by CXCR4 signaling, then a defined CXCL12 gradient (100 ng/mL in the lower well, serum-free upper well) will increase the number of cells invading through Matrigel relative to no-gradient controls, and pretreatment with AMD3100 (CXCR4 antagonist) will reduce invaded-cell counts toward the no-chemokine baseline without reducing total cell viability.

Stained cells are counted in Fiji using a threshold + Watershed + Analyze Particles workflow on five fields per membrane, averaged per insert; alternatively, eluted crystal-violet A595 gives a bulk metric. A chemotactic/invasion index is computed relative to the no-gradient control. File names are coded to blind counting. Per-insert means are averaged to one value per biological replicate before statistics.

No cells on the underside: gradient too weak, incubation too short, or membrane dried during coating — verify chemokine lot, extend to 24 h, keep inserts humidified. Cells on both sides / high background invasion: incomplete swabbing or leaky serum in upper chamber — swab twice and confirm serum-free upper medium. Matrigel layer detaches: coating too thick or insufficient gelling — reduce to 30 µL and gel a full hour at 37 C. Clumped cells block pores: ensure single-cell suspension (filter through 40 µm mesh). High variability between inserts: standardize swabbing and pipette cell input precisely; pre-warm all media.

Invaded-cell counts (per-biological-replicate means, N = 3) are compared by one-way ANOVA with Tukey's HSD across all four conditions (alpha = 0.05). The CXCL12 vs CXCL12+AMD3100 contrast is the primary comparison. If counts are non-normal, use Kruskal-Wallis with Dunn's correction. Power analysis (alpha 0.05, power 0.8) indicates N = 3 with 3 inserts each detects a 50% reduction in invasion given typical CV ~20%.