{"id":15066,"date":"2026-11-04T10:26:54","date_gmt":"2026-11-04T09:26:54","guid":{"rendered":"https:\/\/bict.it\/?p=15066"},"modified":"2026-11-04T10:27:13","modified_gmt":"2026-11-04T09:27:13","slug":"when-a-transaminase-beats-a-reductive-amination","status":"publish","type":"post","link":"https:\/\/bict.it\/en\/when-a-transaminase-beats-a-reductive-amination\/","title":{"rendered":"When a transaminase beats a reductive amination"},"content":{"rendered":"
There is more than one way to a chiral amine, and the choice is not made on the reaction: it is made on the point where the process jams. In a reduction, the critical point is the carbonyl face the hydride arrives at. In a transamination, the critical point is the equilibrium. Moving the question from the enzyme to the equilibrium is already half the development work.<\/p>\n
A transaminase is worth it when the amine has to be enantiopure and the alternatives pay for that purity with a metal, with a pressure, or with half the yield.<\/p>\n
Chemical reductive amination<\/strong> reaches the amine in a single step, and it is robust; but on a prochiral ketone it returns the racemate. The enantiomer you need is recovered by a resolution, and half the material leaves through the wrong door, with its synthesis cost already paid.<\/p>\n Asymmetric hydrogenation<\/strong> solves the selectivity, and pays for it elsewhere: a transition-metal catalyst, usually rhodium or ruthenium, with its chiral ligand; pressure equipment; and a metal residue to keep under control in the final product, where the reference is ICH Q3D and scavenging is never free.<\/p>\n A transaminase<\/strong> does the same job in water, at ambient pressure, with a single enantiomer coming out and no metals. There is a price here too, and it sits somewhere other than where people usually look for it.<\/p>\n An ω-transaminase does not reduce: it transfers. It takes the amino group from a donor, parks it on pyridoxal phosphate as pyridoxamine phosphate, and passes it from there to the ketone. The mechanism is ping-pong, with the enzyme changing state between the two half reactions, and above all it is reversible in both.<\/p>\n So the technical question is not “is there an enzyme that accepts my ketone”, but where the equilibrium sits and what it costs to move it<\/strong>. On many ketones, and particularly on those with bulk near the carbonyl, the equilibrium sits on the reagent side: thermodynamic conversion stops at values that do not hold up economically, however good the enzyme is. A faster enzyme simply reaches the same wall sooner.<\/p>\n There are three tools for moving it, and on scale they are used together.<\/p>\n Choosing between these three routes is not a laboratory preference: it is the decision that sets the cost per kilo. The Transaminations<\/a> page sets out how we work through it on the real ketone and under process conditions, rather than on a model substrate.<\/p>\n This is the biggest difference from an enzymatic reduction, where the cofactor is regenerated and its total turnover number makes it almost irrelevant to cost. In a transamination the donor is stoichiometric<\/strong>: one mole per mole of product, as an absolute minimum. And the absolute minimum is never what you use, because the excess is what moves the equilibrium.<\/p>\n An order-of-magnitude calculation, derived from stoichiometry alone and stated as such: on an amine of 400 g\/mol, ten equivalents of isopropylamine (59.1 g\/mol) come to roughly 1.5 kg of donor per kilo of product<\/strong>. Of that, one equivalent reacts: around 1.33 kg of unconverted donor<\/strong> is left to recover or dispose of, together with about 145 g of acetone<\/strong> per kilo of product. With alanine the arithmetic changes in sign and in kind: more expensive per equivalent, but with the option of a coupled system that lets you work at low excess.<\/p>\n It is this number, not the specific activity of the enzyme, that decides whether a transamination process closes. Screening finds the selectivity; equilibrium engineering makes it payable.<\/p>\n This is the case worth looking at in full, because it is the hardest situation of all: not a resolution to replace, but an asymmetric hydrogenation already in manufacturing, and working<\/strong>.<\/p>\n The incumbent process, as described in the Presidential Green Chemistry Challenge documentation, ran an asymmetric catalytic hydrogenation of an unprotected enamine: a rhodium catalyst, hydrogen at 250 psi, dedicated and expensive equipment, and a stereoselectivity that was not sufficient, which forced a crystallisation step to bring the product to the required optical purity.<\/p>\n The biocatalytic route takes the precursor ketone directly to the chiral amine with an evolved R-selective transaminase. The declared figures:<\/p>\n Two qualifications that matter. First: the source speaks in terms of class<\/em> — “R-selective transaminase”, “evolved transaminase” — and names no specific enzyme, no source organism and no EC number. We do not attribute to a source what it does not say. Second, and this is the point of the case: yield was not the argument<\/strong>. The +10–13% on its own would not have justified touching a validated process. The argument was removing the rhodium, removing the pressure, removing a crystallisation, and freeing capacity on the existing plant.<\/p>\n Four questions, in the order worth asking them.<\/p>\n We start from the ketone and the enantiomer you need. A screen of (R)- and (S)-selective transaminases looks for selectivity on the real substrate; in parallel we set up the equilibrium-shifting system, because donor and removal method have to be chosen together with the enzyme, not after it. Then substrate loading goes up and we watch where the system gives way: co-product inhibition, enzyme stability over the reaction time, cosolvent effects when the ketone will not dissolve.<\/p>\n What you get is a sheet with the numbers a decision needs: conversion and enantiomeric excess, donor equivalents and the fate of the co-product, substrate loading and productivity per litre of reactor, stability and biocatalyst cost per kilo of product, and an impurity profile set against the incumbent route.<\/p>\n Three starting points, three different conclusions.<\/p>\n If the enantiopure amine comes today from a resolution of the racemate<\/strong>, the comparison almost always turns: the 50% thrown away is a structural cost, and the enzyme removes it by construction.<\/p>\n If it comes from a reductive amination with a chiral auxiliary<\/strong>, the question is how many steps leave with the auxiliary: often two, between introduction and removal.<\/p>\n If it comes from an asymmetric hydrogenation that is already validated<\/strong>, yield is not enough of an argument, and the sitagliptin case says it better than any reasoning: you change when the metal, the pressure and a purification step all come out together, and when the plant you already own produces more without being touched. If the longest chapter in your dossier is the one on metal residues, that case is about you.<\/p>\n The starting point is always the same: a ketone, the enantiomer you need, and the number that does not add up today. Let’s talk about your reaction<\/a>.<\/p>\n Because it transfers instead of reducing. A reduction with a regenerated cofactor is effectively irreversible: the regeneration system pulls the reaction to completion. A transamination is an equilibrium between two ketone\/amine pairs, and without outside intervention it stops where thermodynamics stops it. Full conversion is bought, with donor excess or with co-product removal.<\/p>\n Isopropylamine is cheap and gives acetone, which leaves on its own: it is the default on scale, provided the line can handle the stripping and the enzyme tolerates the excess. Alanine is used when the enzyme prefers it as a donor or when an isopropylamine excess is incompatible with the substrate; it needs a coupled system to remove the pyruvate, and therefore one more enzyme to manage.<\/p>\n You work with a miscible cosolvent or a biphasic system, and the variable to measure is not solubility but enzyme stability under those conditions for the whole reaction time. It is one of the parameters we put on the sheet, because an enzyme that loses activity halfway through changes the biocatalyst loading and therefore the cost per kilo.<\/p>\n There is more than one way to a chiral amine, and the choice is not made on the reaction: it is made on the point where [\u2026]<\/span><\/p>\n","protected":false},"author":7,"featured_media":15069,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[79],"tags":[87],"class_list":["post-15066","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-bioindustry-en","tag-insight"],"yoast_head":"\nThe constraint that decides is the equilibrium<\/h2>\n
\n
The donor is a cost line, not a service reagent<\/h2>\n
The case where the numbers are public: sitagliptin<\/h2>\n
\n
How to tell whether your amination is a candidate<\/h2>\n
\n
How we work<\/h2>\n
From here on<\/h2>\n
Frequently asked questions<\/h2>\n
Why does a transaminase not reach full conversion the way a ketoreductase does?<\/h3>\n
Isopropylamine or alanine?<\/h3>\n
What if the ketone will not dissolve in water?<\/h3>\n
Sources<\/h2>\n
\n