# Recombinant collagen and gelatin The Gly-X-Y repeat and the triple helix, why 4-hydroxyproline is required for thermal stability, what prolyl-4-hydroxylase co-expression solves, and why Bloom strength does not follow automatically from a pure protein. Collagen is one of the few proteins a cell cannot make without a second enzyme first modifying it, and that post-translational step is why expressing it in microbes is hard. Source: https://en.bioecon.ru/docs/forestry-biomaterials/fibers-textiles/recombinant-collagen-gelatin/ Updated: 2026-09-04 Collagen is the most abundant protein in the animal body and structurally the most repetitive. Its industrial derivative, gelatin, is simply collagen denatured by heat and acid or alkali. Both are supplied today almost entirely from bovine and porcine hide and bone, and both are unusually difficult to make recombinantly — for a reason that sits in the protein's chemistry rather than in fermentation engineering. ## The helix and the residue that stabilises it Collagen's sequence is a near-continuous **Gly-X-Y** repeat. Every third residue must be glycine because three chains wind into a right-handed triple helix in which the third position points into the crowded core, and only glycine's hydrogen side chain fits. X is frequently proline and Y frequently 4-hydroxyproline. That hydroxyproline is not decorative. Hydroxylation at the 4-position of proline in the Y position raises the melting temperature of the helix substantially; without it, a human collagen sequence denatures below body temperature and is useless. The stabilisation is stereoelectronic — the electron-withdrawing hydroxyl favours the ring pucker and backbone geometry the helix demands — with an additional contribution from ordered water bridging. ## Why the modification is the bottleneck Hydroxyproline is not encoded. It is introduced after translation by **prolyl-4-hydroxylase**, a 2-oxoglutarate-dependent dioxygenase that requires Fe(II), molecular oxygen and ascorbate as a reducing cofactor. Bacteria and yeasts do not have it. Expressing a collagen gene alone in *Escherichia coli* or *Pichia* therefore yields an unhydroxylated chain that will not form a stable helix at useful temperatures, and the standard solution is to co-express a functional prolyl-4-hydroxylase alongside the collagen and supply ascorbate — which turns a single-gene problem into a coordinated two-enzyme system whose stoichiometry and cofactor supply have to be balanced. Plant and insect systems have also been used. A separate route sidesteps the issue entirely: bacterial collagen-like proteins, such as Scl2 from *Streptococcus pyogenes*, form stable triple helices with no hydroxyproline at all, stabilised instead by charged-residue pairs — useful, but not human collagen. ## Purity is not performance Two limits are routinely understated. First, a recombinant collagen molecule is not tissue. Native fibril assembly, with its characteristic 67 nm banding, depends on the telopeptides and on subsequent enzymatic crosslinking by lysyl oxidase; a purified soluble helix does not reproduce it without additional processing. Second, gelatin's functional property is gel strength, reported as Bloom, and it comes from partial renaturation of triple-helical junction zones on cooling. That depends on chain-length distribution and sequence, so a defined recombinant gelatin gives excellent batch-to-batch reproducibility but not automatically a high Bloom value — the two are independent. What recombinant material does deliver is a defined, animal-free product with no TSE risk and no religious or dietary constraint, which is why the clearest adoption is in medical devices, cell culture and pharmaceutical capsules rather than in bulk food gelatin, where the price of extracted hide collagen is very hard to meet.