Hydrogel and ECM suppliers: the directory, and the case against the default

The short answer: the default matrix in this field is an undefined, batch-variable, mouse sarcoma-derived extract, and there are now defined human and synthetic alternatives with published prices — but they are not drop-in substitutes and no supplier pretends otherwise.

That is the whole category in one sentence. What follows is who sells what, in which species, at what price, and which of them will tell you the stiffness of the gel you are buying.

Why the default is a problem worth pricing

Matrigel and equivalent basement membrane extracts remain the first choice for most 3D culture. The 2025 Advanced Science review by Wolff and Hendrix sets out why that persists and why it is a liability, and it is worth quoting the structure of the argument because it maps directly onto sourcing decisions.

A basement membrane extract forms a stable scaffold from basement membrane proteins and growth factors that resembles native extracellular matrix. Its complex composition is both the asset and the drawback: it is undefined and varies batch to batch. Its murine origin raises ethical concerns and, more practically, interspecies variation hinders translation from experimental models to human clinical application. The review’s recommendation is not “stop using it” but something more useful — abandon the one-for-all approach and select or design matrices on a tissue- and model-specific basis, supported by a matrix selection checklist and quantitative criteria for assessing matrix-model compatibility.

Two operational facts follow, and both are sourcing decisions rather than science decisions:

1. A matrix lot is an experimental variable. Reserve a lot for the duration of a programme and treat a lot change as a protocol change. This is standard practice for antibodies and serum and is applied far less consistently to matrix.
2. Substitutes require validation. WiCell publishes the clearest supplier-neutral guidance we found: Cultrex can be used interchangeably with Matrigel in WiCell’s own protocols; Geltrex would need to be optimised for your application; and defined matrices such as vitronectin or laminin may work but may need a ROCK inhibitor for attachment. Three products marketed as equivalent, and only one is treated as interchangeable by an organisation with no stake in the answer.

Corning’s site returned HTTP 403 to every automated retrieval attempt on 2026-09-01, so there is no Matrigel price on this page. That is deliberate. An unchecked price must never be presented as a checked one.

The directory

Human-derived matrix, priced

Humabiologics is the most interesting entry in this directory because it addresses the review’s interspecies objection directly: the matrix is human. The catalogue is organised as collagen, gelatin, universal ECM, tissue-specific ECM and organ-specific ECM, and the tissue-specific range is unusually granular.

Live-verified through the public store API on 2026-09-01:

And the published price ranges across the wider catalogue:

The organ-specific coating range runs to eleven separate tissue ECMs at 1 mg/ml — muscle, cartilage, skin, bone, adipose, tendon, liver, pancreas, kidney, heart and lung — each sold as its own SKU.

Why this matters for engineered muscle specifically. A muscle-derived human ECM coat at $95 for 1 ml is one of the few products in this entire market that lets you match the matrix to the tissue you are trying to build without a custom project. Whether tissue-matched decellularised ECM outperforms collagen or fibrin for engineered muscle is not settled, and Humabiologics publishes no comparative force data — but the material is purchasable at a bench-scale price, which is a different situation from the one that existed a few years ago. See engineered skeletal muscle tissue for what the tissue side of that costs.

The bioink and collagen layer, priced

CELLINK exposes a public store API, which makes it the most transparent supplier in the category — and the pattern of what it does not price is as informative as what it does.

Quote-only on the same storefront, which is the interesting part: all four Lifeink collagen bioinks (Lifeink 200 and 220 pH-neutral, 240 and 260 acidic), the GelMA 95% DS stock solution, and PhotoCol methacrylated type I collagen. A vendor that publishes a price for sixteen matrix products and withholds it for six is telling you those six are sold into a different kind of conversation — typically higher-value, application-supported, or volume-negotiated.

Two arithmetic notes before comparing rows. First, concentration is the hidden variable: FibriCol at 10 mg/mL in 20 ml and PureCol at 3 mg/mL in 100 ml are 200 mg and 300 mg of collagen respectively, at $390 and $450 — so the apparently cheaper product per bottle is more expensive per milligram. Always normalise to mass of protein. Second, human collagen carries a modest premium over bovine: VitroCol human at 3 mg/mL, 20 ml, $375 against PureCol bovine at 3 mg/mL, 100 ml, $450 — roughly a four-fold premium per unit protein. Whether that is worth paying depends entirely on whether species provenance affects your readout.

The only supplier that treats stiffness as a specification

Cellendes is the outlier in this directory and the one worth understanding even if you do not buy it.

The 3-D Life system is two polymers that form a hydrogel by covalent crosslinking at ambient temperature, assembled by the user from named components: crosslinkers, thioreactive polymers, and cell adhesion peptides — RGD for generic integrin engagement, GFOGER-3 for collagen-mimetic binding. Dextran, PVA and hyaluronic acid backbones are offered, plus ToLuminate photogels for light-induced patterning. The system is explicitly free of animal-derived components.

What separates it commercially is that the mechanical property is a design parameter you set rather than a property you inherit. Cellendes publishes a hydrogel stiffness resource, a mesh size resource, a calculation tool and a selection guide. You can also incorporate matrix metalloprotease-cleavable crosslinks, so the gel is degradable by the cells on a schedule you chose.

For anyone whose experiment has mechanics in it, this is not a marginal difference. Stem cell fate, myotube alignment, organoid morphogenesis and contractile output are all stiffness-sensitive, and a basement membrane extract does not let you set stiffness independently of composition — raising the concentration changes both at once. A defined system decouples them.

The trade is real and should be stated: a defined gel gives you control and takes away the uncharacterised cocktail of growth factors that makes a BME work so easily in the first place. That is precisely the tissue-specific design problem the Wolff and Hendrix review argues the field should be solving deliberately instead of avoiding.

How to choose

A blunt mapping, offered as a starting point:

• Established organoid protocol you need to reproduce: use the matrix the protocol names, from the same vendor, and reserve the lot. Substitution is a validation project, not a swap.
• Human relevance is the claim you are making: Humabiologics for human-derived native ECM, or VitroCol for human collagen. Species provenance is otherwise a standing objection to your data.
• Stiffness or degradability is an independent variable in your experiment: Cellendes. Nothing else in this table lets you set it.
• You are printing: GelMA or a collagen bioink, and the choice is constrained by your printer’s crosslinking method more than by biology — see bioprinting suppliers.
• Tissue-matched matrix for a specific organ: Humabiologics’ organ-specific range is the only catalogue route we found; anything else is a decellularisation project.
• Cost per experiment is binding and the biology is forgiving: bovine collagen at 3 mg/mL is the cheapest defined-enough option per milligram in this directory.

What we would ask before ordering matrix

1. Species of origin, and whether that survives review for your intended claim.
2. Lot number, expiry, and whether a single lot can be reserved for the whole programme.
3. Concentration and total protein mass, so you can normalise price per milligram rather than per bottle.
4. Endotoxin specification.
5. For any mechanical endpoint: the stiffness achieved at your working concentration, and whether the supplier will state it.
6. Crosslinking chemistry, initiator, and light dose if photo-crosslinked — and whether that dose is compatible with your cells.
7. Whether the product is sterile as supplied or requires filtration, which for viscous gels is not always possible.
8. Whether the protocol performance was established with this matrix or a different one.
9. Storage and freeze-thaw tolerance, because matrix is frequently ruined by receiving procedure rather than by the supplier.

Related pages

For what the matrix goes into, see organoid suppliers and bioprinting suppliers. For the device-material equivalent of this analysis — where absorption rather than composition is the failure mode — see PDMS vs COP vs PMMA.

We maintain this directory because we broker across it. We are not a reseller for any supplier listed, and where the honest answer is that the matrix your protocol already names is the right one, we will say so.