Cryopreservation of Organoids and 3D Cultures

Team Dynamic Matrices | 28 August 2025

Freezing a single-cell suspension is a solved problem: DMSO, one degree per minute, a good cell line, and you get acceptable viability. Organoids break that reliability in a specific way. The thing you are freezing is an architecture, often embedded in a hydrogel, where the spatial relationships between cells matter as much as whether each cell survives. We have lost more organoid stocks to a "standard" freeze than we would like to admit. Here is what we do differently now.

Freezing a structure is not freezing a suspension

Ice does not just threaten membranes. In a 3D structure, intracellular and extracellular ice growth shears cell-cell junctions and damages the matrix, giving you a sample that thaws with viable cells and a compromised organoid. Whether that matters depends on the downstream use: continued culture can often recover; a fixed-timepoint functional assay cannot.

Why slow-freeze underperforms for 3D

The default protocol assumes three things that stop being true for larger structures:

Cryoprotectant penetration. DMSO equilibrates through a suspension in minutes. Through a 300 µm organoid or a hydrogel dome it can take much longer, so the periphery is protected and the core is not. The result is a viability gradient across each structure that a bulk count will not show. Fick's law is unforgiving: doubling the radius quadruples the equilibration time.

Cooling rate. The rate that minimises ice for a small aggregate is not the rate for a larger organoid or a gel, because thermal conductivity and water content differ.

Matrix carryover. Freezing organoids inside basement membrane extract traps them in a gel that itself freezes unevenly. Most published protocols that work well (for example the intestinal and liver organoid protocols from the Clevers lab, and the widely used Driehuis, Kretzschmar and Clevers Nature Protocols paper from 2020) recover organoids from the matrix and freeze fragments, not whole domes.

Vitrification: worth evaluating, not a default

Vitrification cools fast enough that water forms a glass rather than crystals. It removes ice damage and is standard for embryos and oocytes. For organoids it is attractive for the same reason, but it needs high cryoprotectant concentrations (often 30 to 40 percent total), and those are toxic over the exposure time a larger structure needs to equilibrate. Cell Reports Methods and Biomaterials papers from 2021 onward report good results for small organoids and slices, but every one of them optimised concentration and exposure per tissue. It does not transfer cleanly from the literature. If you have the equipment, run it head to head against slow-freeze for your specific organoid at your specific size.

Match the strategy to structure size

The practical fix for the penetration gradient is to shrink the diffusion distance. Two ways:

1. Freeze earlier, at day 3 to 5 when structures are under 100 µm. This costs post-thaw culture time to reach the intended stage.
2. Fragment before freezing. Mechanically break mature organoids to 30 to 100 µm pieces. This sacrifices some pre-existing organisation but gives consistent viability across the stock.

Which trade-off is right depends on what the stock is for. Decide that before defaulting to freezing organoids at their most mature and least cryoprotectant-accessible state.

The protocol we use for biobanking

1. Dissolve the matrix gently. For DyNAtrix this is a 10 to 15 minute nuclease step at 37 °C; for BME it is 30 to 60 minutes cold depolymerisation with washes. Avoid enzymatic dissociation to single cells; you want fragments.
2. Fragment to 30 to 100 µm by pipetting through a P200 or a 21G needle. Count fragments, not cells.
3. Resuspend in cold freezing medium (we use 10 percent DMSO in a serum-free commercial medium; Recovery Cell Culture Freezing Medium or CryoStor CS10 both work). Keep on ice for no more than 10 minutes before freezing so the cryoprotectant equilibrates but toxicity stays low.
4. Freeze at 1 °C per minute in an isopropanol container or a controlled-rate freezer to minus 80 °C, then transfer to liquid nitrogen vapour within 24 hours.
5. Thaw fast: 37 °C water bath, under 2 minutes, until a small ice pellet remains.
6. Remove cryoprotectant stepwise: add pre-warmed medium dropwise to double the volume over 2 to 3 minutes, then dilute fully and spin gently (200 g, 3 minutes). A single rapid wash osmotically shocks larger fragments.
7. Re-embed immediately. Add ROCK inhibitor (Y-27632, 10 µM) for the first 48 hours for epithelial organoids; it consistently improves recovery.

Validate beyond a viability count

A live/dead stain on thaw day is a first checkpoint and an incomplete one. It does not tell you whether the recovered structure re-forms its architecture or function. Build in a recovery period matched to how long your organoid type takes to re-establish characteristic morphology, usually 5 to 10 days, then assess formation efficiency, morphology and a marker panel against a never-frozen control passaged in parallel. Slower than a same-day check, but for a workflow where the structure is what you are preserving, it is the only validation that counts.