# Terpenes and terpenoids as platform molecules

The choice between the mevalonate and MEP pathways, the carbon price of each, the three ceilings on terpene biosynthesis, and artemisinin as the honest case of technical success and commercial failure.

Here biology competes with difficult organic synthesis rather than with cracking — and wins on chemistry without necessarily winning on market.

Source: https://en.bioecon.ru/docs/biochem-industrial/platform-chemicals/terpenes-terpenoids-platform-molecules/
Updated: 2026-09-06



Terpenes are built from a single five-carbon block, isopentenyl diphosphate and its isomer. Prenyltransferases join these into C10, C15 and C20 chains, and terpene synthases cyclise those chains into skeletons that would take a multi-step stereocontrolled synthesis to reach in a flask. Cytochrome P450s then oxidise specific positions. This is why terpenes are the rare class where fermentation competes not against cheap petrochemistry but against expensive fine synthesis and against agricultural extraction.

## Two routes to the same block

The mevalonate pathway runs in the eukaryotic cytosol: three acetyl-CoA condense to mevalonate, which is phosphorylated twice and decarboxylated to isopentenyl diphosphate. The MEP pathway runs in most bacteria and in plant plastids: pyruvate and glyceraldehyde-3-phosphate condense with the loss of a carbon dioxide, and the chain arrives at the same block.

The difference in carbon price is large. In MEP, six feedstock carbons give five product carbons — one lost. Mevalonate starts from acetyl-CoA, which is itself made from pyruvate with a carbon lost before the pathway even begins: nine carbons of glucose become five carbons of block. The mevalonate route is structurally about half again as wasteful. Industrial processes are nevertheless usually built on it, because it works in yeast, whereas MEP requires reduced ferredoxin, is tightly regulated at its first enzymes, and transplants badly into a convenient host. That is a choice of tractability over stoichiometry, and it is worth naming as such.

## Three ceilings

The first is pathway flux. Isopentenyl diphosphate and its derivatives are regulatory metabolites, and a cell is not built to push half its carbon through them; getting there needs feedback-insensitive variants, amplification of the limiting enzymes, and balancing the demand for NADPH and acetyl-CoA.

The second is toxicity. Terpenes are lipophilic, partition into the membrane and compromise its barrier function; for monoterpenes such as limonene, inhibition appears at fractions of a gram per litre.

The third is volatility — light terpenes leave with the off-gas.

The standard answer to the second and third at once is a second phase: an immiscible organic layer is added to the fermenter, the product partitions into it, the aqueous concentration stays low, and the product arrives partly separated already. This is the one case in the cluster where recovery from dilute aqueous broth is not the problem, because the product removes itself from the water.

## Artemisinin: the honest case

The semi-synthetic route to artemisinin is a first-rank technical achievement: yeast carrying an amplified mevalonate pathway and amorphadiene synthase, P450 oxidation to artemisinic acid at titres on the order of tens of grams per litre, then a photochemical conversion involving singlet oxygen. The process was taken to industrial production. Commercially it did not hold: the price of plant-derived artemisinin extracted from sweet wormwood fell below what the launch case assumed, and production was stopped. The lesson is the same one the rest of the cluster teaches, put more sharply — working molecular biology and a working process do not secure a market position if the alternative's price moves faster than yours.