# Sweet proteins Brazzein, thaumatin and monellin: the wedge model of sweet-taste receptor activation, why sweetness onset is slow and lingering, the heat- and pH-stability differences between them, and why they cannot replace sugar's bulk. A protein a thousand times sweeter than sugar by weight — because it binds the receptor across a large surface instead of dropping into a pocket. Source: https://en.bioecon.ru/docs/agri-food/food-alt-protein/sweet-proteins-brazzein-thaumatin-monellin/ Updated: 2026-08-25 A handful of plant proteins taste intensely sweet to humans. Brazzein, thaumatin and monellin come from unrelated West African and Central American plants, share no sequence or fold with one another, and yet all activate the same receptor. They are interesting because they break the intuition that sweetness is a property of small sugar-like molecules. ## How a protein activates a sugar receptor Sweet taste is detected by a single receptor, a heterodimer of T1R2 and T1R3. Sugars and most small sweeteners bind in a pocket within the large extracellular domain of T1R2 — the "Venus flytrap" domain — and close it. A protein of ten to twenty kilodaltons cannot fit in that pocket. The accepted explanation is the wedge model: sweet proteins bind across an extended surface of the receptor, contacting the outside of the flytrap domain and the cysteine-rich region, and stabilise the active conformation from without rather than from within. Mutational work supports this — the residues that matter for sweetness are distributed over a broad, largely charged patch of the protein's surface, not clustered into a small binding epitope. Two everyday observations follow directly. Because a single large molecule occupies a large receptor surface, potency per unit mass is enormous — figures of hundreds to thousands of times sucrose by weight are typical, though the number depends heavily on the concentration at which it is measured, since these proteins do not follow sucrose's dose–response curve. And because a large protein associates with and dissociates from that surface slowly, the sweetness has a delayed onset and a long, lingering finish. That kinetic signature is a direct consequence of the binding mode, not a formulation defect. ## They are not interchangeable Stability differs sharply, and it tracks structure. Brazzein is small — 54 residues — and cross-linked by four disulfide bonds, which makes it unusually resistant to heat and to a wide pH range. Thaumatin is larger with eight disulfides and is stable to acid and moderate heat, though it unfolds under stronger thermal treatment. Native monellin is the fragile one: it consists of two separate chains held together non-covalently, so heat dissociates them and sweetness is lost irreversibly. Joining the two chains into a single polypeptide — single-chain monellin — recovers much of the thermal stability, which is a clean example of engineering a property from structure. ## What they cannot do Sugar performs several jobs at once. It sweetens, but it also supplies bulk and mass, it depresses water activity and so preserves, it browns via caramelisation and the Maillard reaction, it controls crystallisation in confectionery and freezing point in frozen desserts, and it feeds yeast in dough. A protein used at a few parts per million does none of these. Any reformulation replacing sugar with a sweet protein therefore also needs a bulking agent, and often a separate solution for browning and texture. This is why these ingredients appear in beverages and dairy before they appear in baked goods and confectionery. Being proteins, they are digested as protein, and their sweetness does not survive into the gut. Their allergenic potential is the open question that applies to any novel dietary protein: it is assessed by sequence comparison and stability testing rather than known in advance, and absence of reported reactions to a newly introduced ingredient is weak evidence.