# Secondary metabolites as drugs

Why plants make so little of their own medicines, how extraction fights variability for dose precision, and when synthesis wins over the plant — and when it does not.

The molecules worth extracting from a forest are its defence chemistry — potent by evolution, dilute and variable by the same evolution, and everything about the industry follows from that contradiction.

Source: https://en.bioecon.ru/docs/forestry-biomaterials/wood-construction/forest-biopharma-phytochemicals/
Updated: 2026-09-07



The pharmacologically interesting molecules of a forest are its defence chemistry: terpenoids, alkaloids and polyphenols made not for growth but to poison, deter or signal. Evolution tuned them to be potent, not to be abundant or consistent — exactly backwards for a medicine. Two quantities govern the whole industry: how dilute the compound is inside the organism, and how much that concentration wobbles from tree to tree and season to season.

## Why plants make so little

A secondary metabolite is paid for in carbon, through long enzymatic routes, by an organism that needs only enough to work ecologically. Paclitaxel, the yew's anti-herbivore chemistry, is assembled through nearly twenty enzymatic steps and sits at hundredths of a percent of dry bark weight; in the 1990s, before alternatives existed, a single patient's course meant the bark of several slow-growing yews — and stripping bark kills the tree, a text-book case of [defence capital mined faster than it regrows](../wild-harvest-bioeconomy/). Production is often inducible, peaking in stressed or attacked tissue, because a plant does not stockpile what it can synthesise on demand. Low and patchy is not a manufacturing defect; it is the ecological design, and extraction has to work inside it.

## Extraction and the variability problem

Extraction begins with polarity — choosing a solvent that dissolves the target and not the rest — but the rest is the problem: a crude extract carries thousands of compounds, among them a family of near-identical relatives of the target, and chromatography has to pick one apart from the crowd at industrial scale. The harder fight is against the plant itself. The same species at another site, age, season or genotype delivers another content, while a medicine needs dose precision to a few percent. Hence standardisation: marker compounds, chromatographic fingerprints, batch blending — chemistry imposed on biology to manufacture a consistency biology never had. Sometimes processing creates the actives outright: fresh garlic's allicin is violently unstable, and controlled ageing converts it into stable organosulfur compounds, so the product is defined as much by the process as by the plant.

## When synthesis wins — and when it does not

Given a structure and a market, industry climbs a ladder. Harvest gives way to semisynthesis from renewable precursors — for paclitaxel, deacetylbaccatin taken from yew needles, a renewable tissue richer than bark — then to plant cell fermentation, growing the producing tissue in reactors, then to full chemical or microbial synthesis. Each rung trades biology for consistency and scale. But the ladder is not automatic: semisynthetic artemisinin, made in engineered yeast from the 2010s, never displaced farm-grown wormwood on cost — when the biosynthetic pathway is too tangled for a cheap microbe, the plant, which grows itself for free, keeps winning. The honest pattern: synthesis takes purity-critical, structurally tractable molecules; extraction survives where the pathway is too long or the biomass too cheap; and increasingly the activity lies in the mixture rather than any single molecule, so the product is a reproducible fingerprint — a consistency problem again, not a chemistry one.