Biosynthetic cannabinoids

Cannabinoids are hybrid molecules: half of each comes from fatty acid metabolism and half from terpene metabolism. That structure explains the whole family and why a microbial route is attractive.

Building the precursor

The aromatic half begins with hexanoyl-CoA, a six-carbon acyl group. A polyketide synthase extends it with three malonyl-CoA units, and a cyclase folds the resulting chain into an aromatic ring, giving olivetolic acid — a resorcinol with a five-carbon side chain.

The terpene half is geranyl pyrophosphate, the standard ten-carbon isoprenoid building block. An aromatic prenyltransferase joins the two, producing cannabigerolic acid (CBGA).

CBGA is the branch point. Every major cannabinoid descends from it.

The branch, and what sets the ratio

Two enzymes compete for CBGA, and both are FAD-dependent oxidocyclases that close the terpene chain into a ring while releasing hydrogen peroxide.

THCA synthase closes it one way, giving tetrahydrocannabinolic acid. CBDA synthase closes it differently, giving cannabidiolic acid. They act on the same substrate, so the ratio of products is set by the relative activity of the two enzymes, which in the plant is a matter of which functional alleles it carries. This is why a plant’s chemotype is heritable and roughly binary, and why in a microbial system the ratio is a design choice — express one enzyme and you get one product, with no separation problem.

A third route leaves CBGA unconverted, which is how cannabigerol is obtained.

Why the plant makes acids and the product is not one

The enzymes produce carboxylic acids — THCA, CBDA, CBGA — not the neutral compounds. THCA is not psychoactive. Heating removes the carboxyl group as CO₂, converting THCA to THC and CBDA to CBD.

So decarboxylation is a deliberate processing step, and it is also why raw plant material behaves differently from a heated preparation. In a fermentation route the same step applies, and it is thermal chemistry rather than biology.

Changing the starter unit changes everything

The five-carbon side chain of the common cannabinoids comes from hexanoyl-CoA. Feed the polyketide synthase butyryl-CoA instead, and it builds divarinolic acid, with a three-carbon side chain — leading to the propyl series: CBGV, THCV, CBDV.

This is a clean demonstration of how these pathways work. The rest of the enzymes are unchanged; a different starter unit propagates through the whole sequence and produces a parallel family. In a plant, side-chain length depends on precursor availability and is hard to control. In an engineered organism it is set by feeding or by pathway design, which is why rare cannabinoids present in a plant at fractions of a percent are the strongest commercial case for the fermentation route.

Where it is hard

The prenyltransferase is the bottleneck. It is membrane-associated and expresses poorly in yeast, and its substrate GPP is an intermediate the host also wants for sterol synthesis, so it is drained away. Engineering therefore concentrates on GPP supply and on membrane-protein expression, not on the polyketide half.

Regulatory status is jurisdiction-specific and turns on the molecule and its concentration rather than its origin.