TECHNOLOGY EXPLAINER
Hydrogels as Cell Scaffolds
— how the design of a polymer network changes a cell's fate
A hydrogel is a polymer network that holds a large amount of water. In cell culture it serves as a scaffold that surrounds cells, supports them and passes signals to them. Stiffness, the way it relaxes under load, how it degrades, the size of its mesh: every one of these is a polymer design variable, and it has become clear that they can even change the direction in which stem cells differentiate. For materials engineers, this is the most accessible entry point into cell culture technology.
- What a hydrogel is, in three points
- Why cells need a scaffold
- A map of the materials: natural, semi-synthetic, synthetic
- How the network is crosslinked: physical, chemical, light
- A materials engineer's view (1): photo-crosslinking is a matter of matching absorption spectra
- Design variables and cell behaviour: modulus, stress relaxation, degradability, adhesion
- Our calculation: estimating mesh size from modulus
- A materials engineer's view (2): the mesh is nanoscale, the cell microscale, so breaking becomes part of the spec
- The problems with Matrigel, and synthetic alternatives
- Bioprinting and bioinks
- A materials engineer's view (3): a bioink is a rheology problem
- Open problems, and what this article could not confirm
- Glossary / References / Claim-to-source audit
Sourced = stated in published material or a paper (link given)
Our calculation = a value this article derived from assumptions it states
Not yet confirmed = a research-stage result or outlook with no confirmed practical track record
Structural readings and materials-design interpretations are marked separately as Commentary.
This article explains technology. It is not medical advice and does not recommend any product.
1. What a hydrogel is, in three points
A hydrogel is a network made by tying together (crosslinking) a polymer that would otherwise dissolve in water, at points along its chains, so that it keeps its shape while holding water. Think of jelly, or of konjac, a firm Japanese jelly made from a plant tuber (our commentary).
- What it is used for: in cell culture it acts as a scaffold that either encapsulates cells or carries them on its surface. It stands in for the extracellular matrix (ECM) that surrounds cells in the bodySourced
- What it is made of: there are natural materials such as collagen, gelatin, alginate, hyaluronic acid, fibrin and a basement membrane extract from a mouse tumour (Matrigel), and synthetic ones such as polyethylene glycol (PEG)Sourced
- What matters: stiffness (modulus), how the material relaxes under load (stress relaxation), how it degrades, the size of its mesh, and the landmarks cells grip (adhesion ligands). These have been reported to change how cells spread, how they multiply and which way they differentiateSourced
Stiffness, relaxation, degradability: all of them are polymer design variables. A 2016 paper in Nature Materials held the modulus constant and changed only the time taken for stress to relax by half, from about an hour to about a minute, and showed that this alone made a large difference to how stem cells differentiated into boneSourced. To a cell, the viscoelasticity of its material can be as strong a signal as nutrients or growth factors.
2. Why cells need a scaffold
Cells in the body do not float in liquid. Most are surrounded by a network of proteins and polysaccharides, the extracellular matrix, and live by gripping it. Chaudhuri and colleagues point out that natural extracellular matrices are viscoelastic and show stress relaxation, whereas the synthetic scaffold hydrogels used for 3D culture are usually elasticSourced.
2D culture, growing cells on a plastic dish, is convenient, but plastic is far stiffer than hydrogels or most living tissue (our commentary). The review by Antoine and colleagues notes that the moduli of living tissues span from about 10² Pa for brain to about 10⁸ Pa for tendonSourced. Hydrogels are among the few materials that can reproduce the soft end of that range while full of water (our commentary). The difference between 2D and 3D culture is also covered in our explainer on organoids and spheroids.
3. A map of the materials: natural, semi-synthetic, synthetic
| Material | Origin and preparation | How it forms a gel | Features and caveats in the literature |
|---|---|---|---|
| Collagen I | Mainly rat tail tendon; bovine dermis is used about as often | At neutral pH, fibres (12 to 120 nm in diameter) self-assemble and bundle into a network | Raising the concentration increases fibre density, shrinks the pores and raises the modulus. Raising pH or temperature speeds polymerisation, giving thinner fibres and smaller pores |
| Gelatin / GelMA | Hydrolysed collagen. GelMA is made by reacting gelatin with methacrylic anhydride in phosphate buffer at 50 °C | Gelatin alone forms a physical gel at low temperature. GelMA becomes a covalent network with a photoinitiator and light | Carries RGD sequences and MMP target sequences, so cells can attach to it and remodel it. RGD does not react with methacrylic anhydride, so adhesiveness is preserved |
| Alginate | Extracted from brown algae with alkali. A block copolymer of M (mannuronic acid) and G (guluronic acid) | Ionic crosslinking through an “egg-box” structure in which G blocks sandwich Ca²⁺ and other divalent ions | Mammals lack the enzymes to break it down, so it does not degrade on its own. It has no sites for cells to grip, so RGD is coupled to it. More than 200 kinds are manufactured |
| Hyaluronic acid | A polysaccharide of the body (a glycosaminoglycan). Molecular weight ranges from 100,000 Da in serum to 8 million Da in the vitreous humour | Reactive groups are added with methacrylic anhydride or similar, then crosslinked by radical polymerisation | Rapidly metabolised in the body by hyaluronidase, with a half-life in tissue of hours to days |
| Fibrin | Made from fibrinogen and thrombin, the proteins of blood clotting | Polymerises by an enzyme reaction, with controllable gelation time | Shows non-linear elasticity: soft at small strains, strong and stiff under large deformation. One of the earliest biomaterials, used for haemostasis and wound healing |
| Matrigel | Basement membrane matrix extracted from the Engelbreth-Holm-Swarm (EHS) mouse sarcoma | Liquid when cold, gels when warmed (our note on common practice) | About 60% laminin, about 30% type IV collagen, about 8% entactin and about 2 to 3% perlecan. Also contains growth factors such as TGF-β and FGF, and MMPs. Complex and undefined in composition, varying within and between lots |
| PEG | Synthetic polymer. Acrylate, norbornene, thiol or other groups are attached to the ends of multi-arm PEG | Crosslinked by photopolymerisation, thiol-Michael addition, or an enzyme (activated factor XIIIa) | Hydrophilic, bioinert and easy to modify chemically. Because the crosslinking reactions are largely non-toxic, cells can be encapsulated as the gel forms |
| Polyacrylamide | Acrylamide and bisacrylamide polymerised with ammonium persulfate and TEMED | Chemical crosslinking by radical polymerisation | Because the precursors and the polymerisation reaction are toxic, it cannot be used for 3D culture with encapsulated cells and is limited to 2D culture |
All Sourced (collagen: Antoine et al. [Ref. 12]; gelatin and GelMA: Yue et al. [Ref. 10]; alginate: Lee and Mooney [Ref. 6]; hyaluronic acid: Burdick and Prestwich [Ref. 13]; fibrin: Janmey et al. [Ref. 14]; Matrigel, PEG and polyacrylamide: Aisenbrey and Murphy [Ref. 1]). The exception is Matrigel being “liquid when cold and gelling when warmed”, a note on common practice for which this article did not confirm a primary source.
Matrigel, a natural material in a class of its own
Matrigel differs in character from the other natural materials. It is not a single polymer but an extract of the basement membrane produced by a tumour. According to the review by Aisenbrey and Murphy, it has been used for all kinds of cell culture for more than 40 years, yet because its composition is complex, undefined and variable, its use in cell biology, therapeutic cell manufacturing and drug discovery is limitedSourced. In 1986 Kleinman and colleagues reported the protein composition and biological activity of this basement membrane extract from the EHS tumourSourced. In 2010 Hughes and colleagues used proteomics to identify even the minor components of Matrigel, and noted that it is not a well-defined matrix and can be a source of variability in experimental resultsSourced. Section 9 looks at this problem and the synthetic alternatives in detail.
4. How the network is crosslinked: physical, chemical, light
How the “knots” of the network are tied changes the character of the gel a great deal.
Physical crosslinking: alginate and the “egg box”
According to the review by Lee and Mooney, alginate is a copolymer in which two sugars, M and G, are arranged in blocks, and only the G blocks are thought to form intermolecular crosslinks with Ca²⁺ and other divalent ions. The way G blocks on neighbouring chains join by sandwiching the ions is known as the “egg-box model”Sourced.
- Gelation speed is set by the solubility of the calcium salt: calcium chloride (CaCl2) dissolves readily, so gelation is fast and hard to control. Calcium sulfate and calcium carbonate dissolve poorly, so they slow gelation and widen the working window. Another method adds glucono-δ-lactone to calcium carbonate to release Ca²⁺ slowlySourced
- The raw material changes the properties: G-block content is 60% in the stems of one brown alga (Laminaria hyperborea) and 14.0 to 31.0% in other commercial grades. The longer the G blocks and the higher the molecular weight, the better the mechanical properties of the gelSourced
- Combining molecular weights separates viscosity from stiffness: combining high- and low-molecular-weight alginate raises the elastic modulus of the gel substantially while barely raising the viscosity of the solutionSourced
- Making it degrade: mammals lack an enzyme to cut alginate, so it does not break down on its own, but ionically crosslinked gels dissolve as their ions exchange with monovalent ions in the surroundings. Partial oxidation with sodium periodate makes it degrade in waterSourced
Chemical crosslinking: building PEG from parts
The basic approach with PEG, the leading synthetic material, is to attach reactive groups (acrylate, norbornene, thiol and others) to the ends of multi-arm PEG and use them as the junctions of the network. Besides photopolymerisation, crosslinking uses thiol-Michael addition or enzymatic crosslinking with activated factor XIIIa (a blood-clotting enzyme). Aisenbrey and Murphy note that because these polymerisation methods are largely non-toxic, cells can be encapsulated at the same time as the gel formsSourced. The same review gives many examples of PEG gels whose crosslinks are peptides cleavable by MMPs (protein-degrading enzymes secreted by cells) and PEG gels carrying adhesion peptides such as RGDSourced.
Photo-crosslinking: GelMA and photoinitiators
GelMA (gelatin methacryloyl) is gelatin with methacryloyl groups attached to its amino and hydroxyl groups; exposed to light in the presence of a photoinitiator, it forms a covalent network. According to the review by Yue and colleagues, the synthesis methods in use are all minor variations of the method reported by Van Den Bulcke and colleagues in 2000, and keeping the pH high during the reaction raises the degree of substitutionSourced. The properties of GelMA can be tuned through the degree of substitution, the GelMA concentration, the initiator concentration and the exposure time, and examples have been reported in which microstructural design took the modulus from under 1 kPa to 800 kPaSourced.
5. A materials engineer's view (1): photo-crosslinking is a matter of matching absorption spectra
Curing a gel with light while it contains cells comes with two constraints: using a wavelength that does little harm to cells, and having an initiator that generates radicals efficiently at that wavelength.
In 2009 Fairbanks and colleagues set out the problem with Irgacure 2959 (I2959), long the standard photoinitiator, as follows. I2959 has been the most widely used because it is moderately water-soluble, but its molar absorptivity at 365 nm is only 4 M⁻¹cm⁻¹, and it absorbs almost nothing above 370 nmSourced. They therefore used a lithium acylphosphinate initiator (LAP) and reported that at 365 nm the time to gelation of PEG diacrylate was about one tenth of that with I2959, and that even under 405 nm visible light the gel formed at low initiator concentration and low light intensitySourced.
The review by Yue and colleagues also compares water solubility: at least 5 mg/mL for I2959, and up to 8.5 wt% for LAPSourced. Taking 8.5 wt% as roughly 85 mg/mL, that is about 17 times the lower bound for I2959Our calculation (assumption: a rough estimate taking the solution density as 1 g/mL, compared against the “at least” lower bound given for I2959).
This is exactly the design task of matching the initiator's absorption to the wavelength of the light source that UV-curable coatings and stereolithography resins deal with every day. The only differences are that the solvent is water and there are living cells in it. Water solubility, visible-light absorption, radical generation efficiency and the cytotoxicity of the by-products: for companies with expertise in photoinitiator and sensitiser chemistry, this is a theme with a direct point of contact (our commentary).
6. Design variables and cell behaviour: modulus, stress relaxation, degradability, adhesion
When designing a hydrogel, there are five main variables that cells “sense”. Here they are, together with representative studies.
(1) Modulus: stiffness changes the direction of differentiation
In 2006 Engler and colleagues showed that mesenchymal stem cells specify their lineage with “extreme sensitivity to tissue-level elasticity”. On soft matrices that mimic brain they head towards neurons, on stiffer matrices that mimic muscle towards muscle, and on comparatively rigid matrices that mimic collagenous bone towards bone. In the first week the lineage can be redirected with soluble induction factors, but after several weeks the cells commit to the lineage specified by the matrix elasticitySourced. As prior work, the paper by Chaudhuri and colleagues cites a finding that mesenchymal stem cells in ionically crosslinked alginate differentiated mainly into fat at an initial modulus of 1 to 10 kPa, and mainly into bone at 11 to 30 kPaSourced.
(2) Stress relaxation: the same stiffness, but a different way of giving
In 2016 Chaudhuri and colleagues lowered the molecular weight of alginate from 280 kDa to 35 kDa and attached 5 kDa PEG spacers, changing the time for stress to relax by half (τ1/2) from about an hour to about a minute while holding the alginate concentration and the initial modulus constantSourced.
- Fibroblasts encapsulated in gels with an initial modulus of about 9 kPa were held back from spreading and proliferating in the slow-relaxing gel (τ1/2 about an hour), and proliferated more the faster the gel relaxedSourced
- In fast-relaxing gels with an initial modulus of 17 kPa, mesenchymal stem cells formed a bone-like, mineralised matrix rich in type I collagenSourced
- τ1/2 went from about an hour to about a minute, roughly a sixtiethOur calculation (60 minutes / 1 minute)
(3) Stiffness over time: each stage needs a different environment
In 2016 Gjorevski and colleagues cultured intestinal stem cells in modular synthetic hydrogels and reported that adhesion via fibronectin alone was enough for stem-cell survival and proliferation, and a stiff matrix strongly promoted proliferation through YAP, whereas differentiation and organoid formation needed a soft matrix and laminin-based adhesionSourced. They therefore built a “mechanically dynamic” matrix that is optimal for proliferation at first and later allows differentiation, and used it to culture organoids from mouse and human stem cells without any animal-derived matrixSourced.
(4) Degradability and (5) adhesion ligands
GelMA carries RGD sequences (landmarks that cells grip) and MMP target sequences inherited from gelatin, and has been shown to be degraded by type I and type II collagenasesSourced. Alginate, by contrast, has no sites for cells to grip and adsorbs little protein, so RGD peptides are coupled to it using carbodiimide chemistry. Lee and Mooney describe a minimum RGD concentration needed for cell attachment and proliferation, and an example in which cyclic RGD promoted bone differentiation better than linear RGDSourced.
7. Our calculation: estimating mesh size from modulus
The size of the network openings (mesh size) decides what can pass through a gel. The review by Li and Mooney gives typical mesh sizes reported for hydrogels as 5 to 100 nm, and for comparison gives the hydrodynamic diameter of lysozyme as 4.1 nm and of an antibody (IgG) as 10.7 nmSourced. It also presents, from the classical theory of rubber elasticity, a relation between the shear modulus G and the mesh size r: r = (6RT / (π × NA × G))^(1/3)Sourced.
Using the equation above, this article works through the range of stiffness commonly met in cell cultureOur calculation.
- Assumption: temperature 310 K (37 °C). RT / NA = Boltzmann constant × T = 4.28 × 10⁻²¹ J
- Assumption: the Young's modulus E commonly quoted in papers is converted with G = E / 3, assuming incompressibility (Poisson's ratio 0.5)
- E = 0.3 kPa (G = 100 Pa): r ≈ 43 nm
- E = 1 kPa (G ≈ 333 Pa): r ≈ 29 nm
- E = 3 kPa (G = 1 kPa): r ≈ 20 nm
- E = 10 kPa (G ≈ 3.3 kPa): r ≈ 13.5 nm
- E = 30 kPa (G = 10 kPa): r ≈ 9.4 nm
Assumptions and limits: the equation is a guide for a uniform network obeying ideal rubber elasticity. It does not apply as it stands to heterogeneous networks with bundled fibres, such as collagen or Matrigel, or to ionically crosslinked gels that relax stress. It is a calculation for seeing orders of magnitude, not a statement of the mesh size of any real material.
“100 times stiffer, about 4.6 times finer” comes from the cube root of 100 (about 4.64)Our calculation. At E = 30 kPa the mesh (about 9.4 nm) is smaller than IgG (10.7 nm), so by this calculation proteins the size of an antibody would have trouble moving through itOur calculation. Li and Mooney state that as the mesh size approaches the size of the drug (a ratio of about 1), steric hindrance becomes pronounced and diffusion slows greatlySourced. A stiff gel, in other words, is also a gel that lets larger molecules such as growth factors through less easily (our commentary).
8. A materials engineer's view (2): the mesh is nanoscale, the cell microscale, so breaking becomes part of the spec
In the Section 7 calculation, gels at the stiffnesses common in cell culture had meshes of roughly 10 to 40 nmOur calculation. Cells themselves, meanwhile, are measured in micrometres. The mesh is far finer than the cell, so an encapsulated cell can neither move through it nor spread out as things stand (our commentary).
How, then, do cells spread? The primary sources suggest two routes.
- Cut the network: as in PEG gels crosslinked with MMP-cleavable peptides, or GelMA with its MMP target sequences, design the junctions to come undone under enzymes the cells secreteSourced
- Push the network aside: in materials whose junctions detach and re-form, such as ionically crosslinked alginate, the faster the stress relaxation, the more the cells spread and proliferated. Chaudhuri and colleagues note that in fast-relaxing gels the matrix can be plastically deformed and mechanically remodelledSourced
For most industrial materials, not breaking and not creeping is the performance. In a cell scaffold, however, breaking to the right degree and flowing to the right degree become the specification. And even how fast it breaks affects how the cells behave.
Viscoelastic measurement (stress relaxation tests, frequency-dependent rheology), the design of degradable polymers, reversible crosslinks (ionic bonds, dynamic covalent bonds): all of these are already in the materials engineer's toolbox. What is missing is data that ties them to performance as the cell experiences it, and that can be read as the opening for entrants from the materials side (our commentary).
9. The problems with Matrigel, and synthetic alternatives
Matrigel, widely used for culturing organoids and pluripotent stem cells, has a serious weakness when viewed as a material. The review by Aisenbrey and Murphy gives the following values as examples of moduli measured by atomic force microscopySourced.
The same review notes that Matrigel also contains tumour-derived growth factors such as TGF-β and FGF, and enzymes such as MMPsSourced. It is, in other words, a material whose stiffness and biochemical signals both shift from lot to lot.
How far have synthetic alternatives come?
- Intestinal organoids: Gjorevski and colleagues reported building, with PEG-based synthetic gels, a fully defined system that can culture mouse and human intestinal organoids without any animal-derived matrixSourced
- Pluripotent stem cells: the review describes an example in which a poly(vinyl alcohol-co-itaconic acid) gel with a modulus of 25 kPa and a vitronectin-derived peptide maintained human ES and iPS cells under xeno-free conditions for more than 20 passages, on a par with MatrigelSourced. For 2D substrates, the recombinant laminin-511 E8 fragment (Nakagawa et al. 2014) and vitronectin (Chen et al. 2011) have also been reported as defined substrates that can replace MatrigelSourced
- Cost: the review states that the raw material cost of PEG hydrogels is about half that of Matrigel, while pointing out that the synthetic peptides that provide biochemical signals can become very expensive at large scaleSourced
Within the scope of this article, no primary source could be confirmed showing that synthetic alternatives have widely replaced Matrigel in large-scale manufacturing or clinical useNot yet confirmed.
10. Bioprinting and bioinks
Bioprinting is the technology of placing cells and materials at defined positions in three dimensions. In 2018 Groll and colleagues proposed definitions that distinguish a “bioink”, a formulation that contains cells and can be processed by an automated fabrication technology, from a “biomaterial ink”, which contains no cellsSourced.
The review by Gungor-Ozkerim and colleagues divides the main approaches into laser-assisted, inkjet (droplet) and extrusion printing, and lists as bioink materials alginate, gelatin, collagen, fibrin, gellan gum, hyaluronic acid, agarose, chitosan, silk, decellularised ECM, PEG and PluronicSourced. The evaluation criteria it gives are flow-initiation behaviour, the degree of shear thinning, yield stress, cell viability and recovery after printingSourced.
In 2015 Hinton and colleagues reported a method called FRESH (freeform reversible embedding of suspended hydrogels). The support bath is made of gelatin microparticles and acts as a “Bingham plastic”, behaving like a rigid body at low shear stress and like a viscous fluid at high shear stress. It flows only where the nozzle passes, and the deposited ink is held in place where it lands. When printing is finished, the bath is warmed to 37 °C to melt the gelatin and the print is removed. With this method they printed alginate, collagen and fibrin with moduli below 500 kPa at a resolution of about 200 µmSourced.
The review by Gungor-Ozkerim and colleagues also describes examples of “two-step crosslinking”, printing via physical crosslinking of gelatin at low temperature and then ionically crosslinking alginate with Ca²⁺ after printingSourced.
11. A materials engineer's view (3): a bioink is a rheology problem
Rearrange the requirements in Fig. 7 and a bioink is being asked for three things.
- It does not flow at rest (yield stress)
- It flows readily under force (shear thinning)
- It returns to its original state as soon as the force is removed (fast recovery, thixotropy)
These are almost the same flow properties asked of screen-printing pastes, dispensing adhesives and conductive pastes (our commentary). That the FRESH support bath is described as a “Bingham plastic”Sourced also shows that this is a field where the vocabulary of rheology carries straight over.
There is one difference: the ink contains cells that shear can damage. In 2016 Blaeser and colleagues reported using a fluid-dynamics model to control the shear stress at the nozzle of a microvalve-based bioprinter and systematically studying how cell viability and proliferative capacity changed with the level of shear stressSourced. So there is a trade-off: the more easily the ink flows, the easier it is to print, but the harder you push it, the more the cells suffer.
The thickeners, thixotropic agents and molecular-weight-distribution design used in paste formulation apply here. For example, the finding from Section 4 that combining high- and low-molecular-weight alginate raises the gel's modulus substantially while barely raising the solution's viscositySourced is precisely a formulation design that reconciles low viscosity during printing with high elasticity after setting (our commentary).
12. Open problems, and what this article could not confirm
(1) “Stiffness drives differentiation” results depend on conditions
Each response listed in Fig. 3 is a result for a specific cell, a specific material and a specific setting, 2D or 3D. Within the scope of this article, no general rule could be confirmed that the same modulus gives the same response in different material systems. The work by Chaudhuri and colleagues itself shows that the outcome changes with stress relaxation even at the same modulusSourced.
(2) Engler and colleagues' specific modulus values
The specific moduli (in kPa) corresponding to neural, muscle and bone lineages in the paper by Engler and colleagues are not given here because this article could not check the full text of the original paper. Only the qualitative statements in the abstract are used.
(3) How far Matrigel has actually been replaced
There are many reports of synthetic gels, but no primary source could be confirmed showing how far they have replaced Matrigel in manufacturing or clinical useNot yet confirmed.
(4) Clinical use of bioprinting
On the clinical use of bioprinted tissue, regulatory approvals and peer-reviewed clinical trial results were not checked within the scope of this article, so nothing is stated.
- A hydrogel is a water-holding polymer network that stands in for the extracellular matrix. There are natural, semi-synthetic and synthetic typesSourced
- Crosslinking is physical (ionic and so on), chemical or light-driven. Alginate sets through the “egg box”, in which G blocks sandwich Ca²⁺Sourced
- The photoinitiator LAP gels in about a tenth of the time of I2959 at 365 nm and also works under 405 nm visible lightSourced
- Even at the same modulus, the speed of stress relaxation alone (about an hour versus about a minute) changed bone differentiation of stem cellsSourced
- Gels at commonly used stiffnesses have meshes of about 10 to 40 nm. That is far finer than a cell, so designs that break or flow are neededOur calculation
- Matrigel is undefined in composition, and in one example its modulus differed by about 2x between lotsOur calculation
- Bioink requirements can be written in the same terms as paste formulation: yield stress, shear thinning and recovery (our commentary)
13. Glossary
- Hydrogel
- A crosslinked polymer network that keeps its shape while holding a large amount of water.
- Scaffold
- A supporting material to which cells attach or in which they are encapsulated.
- Extracellular matrix (ECM)
- The network of proteins and polysaccharides that surrounds cells in the body.
- Crosslinking
- Joining polymer chains to one another. The joins are called crosslinks.
- Egg-box model
- A model of how the G blocks of alginate sandwich Ca²⁺ to join chains together.
- Elastic modulus
- Resistance to deformation. Includes Young's modulus E and shear modulus G.
- Stress relaxation
- The fall in internal stress over time under a constant deformation.
- Mesh size
- The size of the openings in the network. Sets the size of molecules that can pass through.
- RGD
- The three amino acids arginine, glycine and aspartic acid. A landmark that cells grip.
- MMP
- Matrix metalloproteinase. A protein-degrading enzyme secreted by cells.
- GelMA
- Gelatin with methacryloyl groups added so that it can be crosslinked by light.
- Photoinitiator
- A compound that absorbs light to generate radicals and start polymerisation. Examples include I2959 and LAP.
- Matrigel
- Basement membrane matrix extracted from the mouse EHS sarcoma. Undefined in composition.
- Bioink
- A formulation that contains cells and can be processed by an automated fabrication technology.
- Shear thinning
- The property of losing viscosity and flowing more easily under shear.
- Bingham plastic
- A material that behaves like a solid up to a certain stress (the yield stress) and like a fluid above it.
14. References (primary sources)
- Aisenbrey EA, Murphy WL. “Synthetic alternatives to Matrigel”, Nat Rev Mater 5:539 (2020) (PMC) — pmc.ncbi.nlm.nih.gov
- Kleinman HK et al. “Basement membrane complexes with biological activity”, Biochemistry 25:312 (1986) — doi.org
- Hughes CS, Postovit LM, Lajoie GA. “Matrigel: a complex protein mixture required for optimal growth of cell culture”, Proteomics 10:1886 (2010) — doi.org
- Engler AJ, Sen S, Sweeney HL, Discher DE. “Matrix elasticity directs stem cell lineage specification”, Cell 126:677 (2006) — doi.org
- Chaudhuri O et al. “Hydrogels with tunable stress relaxation regulate stem cell fate and activity”, Nat Mater 15:326 (2016) (PMC) — pmc.ncbi.nlm.nih.gov
- Lee KY, Mooney DJ. “Alginate: properties and biomedical applications”, Prog Polym Sci 37:106 (2012) (PMC) — pmc.ncbi.nlm.nih.gov
- Li J, Mooney DJ. “Designing hydrogels for controlled drug delivery”, Nat Rev Mater 1:16071 (2016) (PMC) — pmc.ncbi.nlm.nih.gov
- Gjorevski N et al. “Designer matrices for intestinal stem cell and organoid culture”, Nature 539:560 (2016) — doi.org
- Van Den Bulcke AI et al. “Structural and rheological properties of methacrylamide modified gelatin hydrogels”, Biomacromolecules 1:31 (2000) — doi.org
- Yue K et al. “Synthesis, properties, and biomedical applications of gelatin methacryloyl (GelMA) hydrogels”, Biomaterials 73:254 (2015) (PMC) — pmc.ncbi.nlm.nih.gov
- Fairbanks BD et al. “Photoinitiated polymerization of PEG-diacrylate with lithium phenyl-2,4,6-trimethylbenzoylphosphinate: polymerization rate and cytocompatibility”, Biomaterials 30:6702 (2009) (PMC) — pmc.ncbi.nlm.nih.gov
- Antoine EE, Vlachos PP, Rylander MN. “Review of collagen I hydrogels for bioengineered tissue microenvironments”, Tissue Eng Part B 20:683 (2014) (PMC) — pmc.ncbi.nlm.nih.gov
- Burdick JA, Prestwich GD. “Hyaluronic acid hydrogels for biomedical applications”, Adv Mater 23:H41 (2011) (PMC) — pmc.ncbi.nlm.nih.gov
- Janmey PA, Winer JP, Weisel JW. “Fibrin gels and their clinical and bioengineering applications”, J R Soc Interface 6:1 (2009) (PMC) — pmc.ncbi.nlm.nih.gov
- Hinton TJ et al. “Three-dimensional printing of complex biological structures by freeform reversible embedding of suspended hydrogels”, Sci Adv 1:e1500758 (2015) (PMC) — pmc.ncbi.nlm.nih.gov
- Gungor-Ozkerim PS et al. “Bioinks for 3D bioprinting: an overview”, Biomater Sci 6:915 (2018) (PMC) — pmc.ncbi.nlm.nih.gov
- Groll J et al. “A definition of bioinks and their distinction from biomaterial inks”, Biofabrication 11:013001 (2018) — doi.org
- Blaeser A et al. “Controlling shear stress in 3D bioprinting is a key factor to balance printing resolution and stem cell integrity”, Adv Healthc Mater 5:326 (2016) — doi.org
- Nakagawa M et al. (Kyoto University CiRA and others) “A novel efficient feeder-free culture system for the derivation of human induced pluripotent stem cells”, Sci Rep 4:3594 (2014) — doi.org
- Chen G et al. “Chemically defined conditions for human iPSC derivation and culture”, Nat Methods 8:424 (2011) (PMC) — pmc.ncbi.nlm.nih.gov
15. Claim-to-source audit
| Claim in the text | Basis | Label |
|---|---|---|
| That Matrigel is basement membrane matrix from the EHS mouse sarcoma and has been used for more than 40 years; that its complex, undefined and variable composition limits its use; about 60% laminin, about 30% type IV collagen, about 8% entactin and about 2 to 3% perlecan; that it contains growth factors such as TGF-β and FGF, and MMPs; moduli of 400 to 420 Pa (two lots), 840 Pa (a third lot), 1 to 3 kPa locally and a median of about 650 Pa. That PEG is hydrophilic, bioinert and easy to modify, is crosslinked by photopolymerisation, thiol-Michael addition or factor XIIIa, and is largely non-toxic so cells can be encapsulated. That polyacrylamide has toxic precursors and is limited to 2D. Examples of MMP-degradable and RGD-coupled PEG. The PVA-co-itaconic acid gel (25 kPa, more than 20 passages). That PEG raw materials cost about half as much as Matrigel and synthetic peptides can become expensive. Examples of self-assembling peptide gels | Reference 1 https://pmc.ncbi.nlm.nih.gov/articles/PMC7500703/ | Sourced |
| That Kleinman et al. reported the composition and biological activity of the basement membrane extract from the EHS tumour (1986) | Reference 2 https://doi.org/10.1021/bi00350a005 | Sourced |
| That Matrigel is not a well-defined matrix and can be a source of variability in experimental results; and that proteomic analysis identified even its minor components | Reference 3 https://doi.org/10.1002/pmic.200900758 | Sourced |
| That mesenchymal stem cells specify lineage with sensitivity to tissue-level elasticity, heading towards neural on brain-like, muscle on muscle-like and bone on bone-like matrices; and that after several weeks they commit to the lineage specified by matrix elasticity | Reference 4 https://doi.org/10.1016/j.cell.2006.06.044 | Sourced |
| That natural ECM is viscoelastic and shows stress relaxation while synthetic scaffolds are usually elastic. That lowering the molecular weight from 280 to 35 kDa and adding 5 kDa PEG spacers changed τ1/2 from about an hour to about a minute. That at 9 kPa fibroblast spreading and proliferation were held back in slow-relaxing gels. That fast-relaxing 17 kPa gels produced a bone-like mineralised matrix. The prior findings of fat at 1 to 10 kPa and bone at 11 to 30 kPa. That stress relaxation in ionically crosslinked alginate arises from crosslink unbinding and flow. That the matrix can be plastically deformed and remodelled. That alginate has no cell-adhesion sites | Reference 5 https://pmc.ncbi.nlm.nih.gov/articles/PMC4767627/ | Sourced |
| That alginate is an M/G block copolymer from brown algae whose G blocks crosslink with divalent ions in an egg-box arrangement. That CaCl2 is fast and hard to control while CaSO4 and CaCO3 can slow gelation; CaCO3 plus glucono-δ-lactone. G-block content of 60% and of 14.0 to 31.0%. That longer G blocks and higher molecular weight improve mechanical properties. That combining high and low molecular weights raises the modulus while barely raising viscosity. Non-degradable in mammals, dissolution by ion exchange, degradability conferred by periodate oxidation. Carbodiimide coupling of RGD, minimum concentration, cyclic RGD. That more than 200 kinds are manufactured | Reference 6 https://pmc.ncbi.nlm.nih.gov/articles/PMC3223967/ | Sourced |
| Typical mesh sizes of 5 to 100 nm; lysozyme 4.1 nm and IgG 10.7 nm. The rubber-elasticity equation r = (6RT/(πNA G))^(1/3). That diffusion slows greatly through steric hindrance as mesh size approaches the size of the drug | Reference 7 https://pmc.ncbi.nlm.nih.gov/articles/PMC5898614/ | Sourced |
| That in a synthetic gel, fibronectin adhesion allowed intestinal stem cells to survive and proliferate, a stiff matrix promoted proliferation through YAP, and differentiation and organoid formation needed softness and laminin; and that a mechanically dynamic matrix enabled culture without animal-derived matrix | Reference 8 https://doi.org/10.1038/nature20168 | Sourced |
| That Van Den Bulcke et al. reported gels of methacrylamide-modified gelatin (2000) | Reference 9 https://doi.org/10.1021/bm990017d | Sourced |
| GelMA synthesis (methacrylic anhydride, phosphate buffer, 50 °C, a variation of the Van Den Bulcke method, higher substitution at high pH). That gelatin gels at low temperature. That it carries RGD and MMP target sequences and RGD does not react with methacrylic anhydride. Degradation by collagenase. The tuning parameters. The example from under 1 kPa to 800 kPa. The solubility of I2959 (at least 5 mg/mL) and LAP (up to 8.5 wt%) | Reference 10 https://pmc.ncbi.nlm.nih.gov/articles/PMC4610009/ | Sourced |
| That I2959 is the most widely used, with molar absorptivity of 4 M⁻¹cm⁻¹ at 365 nm and almost no absorption above 370 nm; and that with LAP the gelation time at 365 nm is about an order of magnitude shorter, with gelation also possible at 405 nm at low concentration and low light intensity | Reference 11 https://pmc.ncbi.nlm.nih.gov/articles/PMC2896013/ | Sourced |
| That the main source of collagen I is rat tail tendon (bovine dermis about as common); that fibres 12 to 120 nm in diameter self-assemble at neutral pH; the relationships between concentration, pH and temperature and the fibres, pores and modulus; and that tissue moduli range from about 10² Pa for brain to about 10⁸ Pa for tendon | Reference 12 https://pmc.ncbi.nlm.nih.gov/articles/PMC4241868/ | Sourced |
| The molecular weight of hyaluronic acid (100,000 to 8 million Da), its rapid metabolism by hyaluronidase (half-life of hours to days), and modification with methacrylic anhydride | Reference 13 https://pmc.ncbi.nlm.nih.gov/articles/PMC3730855/ | Sourced |
| That fibrin is made from fibrinogen and thrombin and was among the earliest materials used for haemostasis and wound healing; that gelation time can be controlled; and non-linear elasticity, soft at small strains and stiff under large deformation | Reference 14 https://pmc.ncbi.nlm.nih.gov/articles/PMC2575398/ | Sourced |
| The FRESH method: that the gelatin-microparticle support bath behaves as a Bingham plastic, that it is melted at 37 °C to release the print, and that alginate, collagen and fibrin with moduli below 500 kPa were printed at about 200 µm resolution | Reference 15 https://pmc.ncbi.nlm.nih.gov/articles/PMC4646826/ | Sourced |
| The three bioprinting approaches (laser-assisted, inkjet, extrusion); the list of bioink materials; the evaluation criteria (flow initiation, shear thinning, yield stress, cell viability, recovery); the two-step example of low-temperature physical crosslinking of gelatin plus Ca²⁺ crosslinking of alginate; and that Pluronic is listed as a bioink material | Reference 16 https://pmc.ncbi.nlm.nih.gov/articles/PMC6439477/ | Sourced |
| The definitions distinguishing bioinks from biomaterial inks | Reference 17 https://doi.org/10.1088/1758-5090/aaec52 | Sourced |
| That shear stress at the nozzle of a microvalve-based printer was controlled, and the relationship between shear stress and cell viability and proliferative capacity was studied | Reference 18 https://doi.org/10.1002/adhm.201500677 | Sourced |
| That the recombinant laminin-511 E8 fragment has been reported as a defined substrate that can replace Matrigel | Reference 19 https://doi.org/10.1038/srep03594 | Sourced |
| Culture of pluripotent stem cells on vitronectin-coated surfaces | Reference 20 https://pmc.ncbi.nlm.nih.gov/articles/PMC3084903/ | Sourced |
| Mesh sizes (about 43, 29, 20, 13.5 and 9.4 nm); about 4.6x finer for 100x stiffer; smaller than IgG at E = 30 kPa; τ1/2 falling to about a sixtieth; LAP solubility about 17 times the I2959 lower bound; Matrigel moduli differing by about 2x (840/410) and up to about 4.6x (3000/650) | Our calculation. The temperature of 310 K, G = E/3 (Poisson's ratio 0.5) and 8.5 wt% taken as about 85 mg/mL (density 1 g/mL) are assumptions set by this article. The mesh sizes are a guide assuming a uniform ideal network, not values for real materials | Our calculation |
| Whether synthetic alternatives have widely replaced Matrigel in manufacturing or clinical use | No primary source could be confirmed within the scope of this article; treated as research stage | Not yet confirmed |
| The practice of handling Matrigel as a liquid when cold and gelling it by warming | Included as a note on common practice. No primary source was checked for this article (our note) | Commentary |
| Engler et al.'s specific modulus values; a general rule that the same modulus gives the same response in different material systems; clinical use of bioprinted tissue | Not stated in this article because the full text of the original paper and regulatory documents could not be checked | Commentary |
| The three groups of materials; the three-way split of crosslinking; the grouping of design variables; the three stages of bioink requirements; the parallel between photoinitiators and the design of UV-curable coatings; the reading that breaking is part of the spec; the similarity to paste formulation; the comparison that cells, measured in micrometres, are far larger than the mesh | This article's own framing and commentary based on published content. Not views expressed by the authors | Commentary |
| That Figs. 1, 2, 3 and 7 are explanatory drawings; that Figs. 4 and 6 are drawings that include our calculation; and that the hero image and Fig. 5 are AI-generated images | A note by this article | Commentary |
Last updated 23 September 2026. Sources are limited to primary material (peer-reviewed original papers and peer-reviewed reviews). No market-size or market-share estimates are used. Because the article includes structural readings and materials-design interpretations, those are marked as Commentary and kept separate from sourced fact. Engler and colleagues' specific modulus values, how far synthetic alternatives have spread, and the clinical use of bioprinted tissue are not covered, because they could not be confirmed in published primary sources. All figures are for explanation. Figs. 1, 2, 3 and 7 are concept diagrams; Figs. 4 and 6 are drawings that include our calculation; the hero image and Fig. 5 are AI-generated images, and none of them shows a real material, micrograph or product. This article explains technology. It is not medical advice and does not recommend any product.