What Is the Best Matrix for Organoid Culture? A Decision Framework
Team Dynamic Matrices | 1 November 2025
"What is the best matrix for organoids?" is the question we get asked most, and we will not pretend it has one answer. It depends on what you want to learn from the organoid. What we can offer is the framework we use with customers, the evidence behind it, and an honest account of where each matrix class is and is not ready.
Start with the question, not the material
Before comparing products, write down what you are optimising for:
- A drug screen lives or dies on well-to-well and plate-to-plate consistency. Batch variability is the enemy, even if the matrix is biologically "realistic".
- A mechanotransduction study needs stiffness and stress relaxation as controlled variables, not emergent properties of an undefined gel.
- A developmental question may need the matrix to change over the culture period, because the tissue would.
- A translational program needs to rule out animal-derived components early. Regulators treat xenogeneic material as a liability regardless of dish performance, and NAM guidance from FDA and EMA now explicitly favours defined systems.
The main matrix classes, honestly compared
| Class | Examples | Composition | Mechanics | Batch consistency | Cell recovery | Best fit |
|---|---|---|---|---|---|---|
| Basement membrane extract | Matrigel, Cultrex, Geltrex | Undefined, animal | Fixed, ~0.1 to 1 kPa | Low | Cell Recovery Solution, cold, slow | Matching a published protocol |
| Natural single-protein gels | Collagen I, fibrin | Defined protein, animal or human | Concentration-coupled | Medium | Collagenase / plasmin | Stromal and vascular models |
| Decellularised ECM | Tissue-specific dECM | Undefined, animal or human | Variable | Low | Enzymatic | Tissue-specific cues |
| Synthetic elastic gels | PEG-based, e.g. with MMP-cleavable links | Defined | Tunable stiffness, mostly elastic | High | Protease or photo-cleavage | Reductionist mechanics |
| Synthetic viscoelastic gels | DNA-crosslinked (DyNAtrix), some alginate systems | Defined | Tunable stiffness and stress relaxation, separately | High | Nuclease or ion exchange, gentle | Reproducible, mechanically faithful culture |
The two reviews we point people to are Aisenbrey and Murphy (Nature Reviews Materials, 2020) on synthetic alternatives and Kozlowski et al. (Communications Biology, 2021) on moving organoid culture off Matrigel.
Three questions to ask before ordering anything
1. Do I need to control mechanics independently of composition?
Basement membrane extracts bundle stiffness, ligand density and degradability into one fixed, lot-variable package. Synthetic matrices let you set stiffness and viscoelasticity as separate dials. Gjorevski et al. (Nature, 2016) showed why this matters: intestinal organoids needed one stiffness to expand and another to differentiate.
2. Does my readout depend on optical clarity?
Confocal and light-sheet work needs a transparent matrix. Dense fibrous ECMs scatter light and complicate quantitative image analysis. Most synthetic gels are clear at working thickness.
3. Do I need live cells back afterwards?
For passaging, sorting, single-cell RNA-seq or re-embedding, recovery method matters. Cold depolymerisation of BME is slow and incomplete; mechanical disruption of dense protein gels damages cells. Gentle on-demand degradation, such as nuclease-triggered release in DNA-crosslinked gels, preserves viability and avoids the dissociation stress signature that contaminates single-cell data (van den Brink et al., Nature Methods, 2017).
Why we built around DNA crosslinking
Hydrogels crosslinked by short DNA duplexes decouple mechanics from biochemistry, which biological matrices structurally cannot do because their mechanics and composition come from the same source. Crosslink sequence and length set stress relaxation; polymer content sets stiffness; adhesion peptides set biology. Each is a separate dial. The matrix can also be softened or stiffened mid-culture by adding a DNA strand, without disturbing the embedded tissue. That opens designs, particularly for kidney, liver and brain organoids, that fixed ECMs cannot support (Peng et al., Nature Nanotechnology, 2023).
A practical shortlist
| If your priority is | Lean toward |
|---|---|
| Maximum batch consistency | Synthetic, chemically defined matrix |
| Mechanistic stiffness or relaxation studies | Independently tunable synthetic viscoelastic system |
| Fastest adoption of a published protocol | The animal-derived matrix the paper used, then migrate once the assay is stable |
| Clinical or NAM-aligned pipeline | Xeno-free synthetic matrix from day one |
| A niche lineage with no synthetic validation yet | Stay on BME, but run a small synthetic pilot in parallel |
There is no universal best matrix. There is a best matrix for your organoid type and your question, and for groups prioritising reproducibility and mechanical control that answer is increasingly a defined, tunable, synthetic system.
References
- Aisenbrey EA, Murphy WL. Synthetic alternatives to Matrigel. Nat Rev Mater 5, 539–551 (2020).
- Kozlowski MT, Crook CJ, Ku HT. Towards organoid culture without Matrigel. Commun Biol 4, 1387 (2021).
- Gjorevski N et al. Designer matrices for intestinal stem cell and organoid culture. Nature 539, 560–564 (2016).
- van den Brink SC et al. Single-cell sequencing reveals dissociation-induced gene expression in tissue subpopulations. Nat Methods 14, 935–936 (2017).
- Peng Y-H et al. Dynamic matrices with DNA-encoded viscoelasticity support the development of organoids and stem cells. Nat Nanotechnol (2023).
FAQs
Yes. It remains the literature baseline, and comparing to that baseline is a legitimate reason to use it. The trade-off is accepting lot variability and animal origin as a fixed cost, and reporting the lot number.
Not universally yet. Replacement is mature for intestinal, MSC and iPSC-derived, trophoblast and kidney-cyst systems, and progressing for liver and brain. It is less established for highly specialised niches with little validation data. We would rather tell you that than sell you a kit that will not work.
Keep medium and timing identical, swap only the matrix, and compare formation efficiency and one lineage marker at a fixed day. If you want help designing that comparison, talk to us.
There is no universal best matrix; start with the question. Basement membrane extracts support the widest range of lineages but vary by lot and are animal-derived; defined synthetic gels remove that variability and let you set stiffness and stress relaxation. DyNAtrix has been validated with human iPSCs (see DyNAtrix).