A Sweet Protein from a Fungus: HT-AC from the Honey Truffle and the Human Sweet-Taste Receptor

When we hear the phrase “sweet protein,” it may sound counterintuitive.

Most familiar sweet substances, including sucrose and glucose, are small molecules. Yet nature also contains proteins that can produce intense sweetness at very low concentrations. Well-known examples include thaumatin, monellin and brazzein, most of which were originally discovered in tropical plants.

In recent years, however, a very different member has joined this group:

a sweet protein from the fungus Mattirolomyces terfezioides, an underground fruiting ascomycete known as the Hungarian honey truffle or Hungarian sweet truffle.

The protein was identified at the molecular level in 2024. By 2026, studies had reported activation of the human sweet-taste receptor, safety data, an NMR structure and high-resolution X-ray crystal structures.

In September 2026, those crystal structures provided the clearest structural view yet of this unusual fungal sweet protein.

The sweet truffle Mattirolomyces terfezioides

Mattirolomyces terfezioides is an underground ascomycete in the order Pezizales and family Pezizaceae. It is often called the Hungarian honey truffle or Hungarian sweet truffle.

Taxonomically, it does not belong to the true truffle genus Tuber. It belongs to Mattirolomyces, but forms truffle-like fruiting bodies below ground.

The fungus has long been known for its remarkably sweet taste.

In 2024, McFarland and colleagues reported the molecular basis of that sweetness in the Journal of Agricultural and Food Chemistry.

By tracking the sweet activity in fruiting-body extracts, the researchers identified a novel 121-amino-acid sweet protein. It was initially described as Honey Truffle Sweetener (HTS), while later studies refer to the active protein as Honey Truffle Active Component (HT-AC).

Its molecular mass is approximately 13.4 kDa.

About as sweet as 4% sucrose at 20 µg/mL

HT-AC is a high-intensity sweetener.

A 2026 review of sweet-receptor biology summarized sensory data showing that 20 µg/mL Honey Truffle Sweetener produced sweetness comparable to a 4% (w/v) sucrose solution.

On a simple mass-concentration basis, that corresponds to roughly 2,000-fold the sweetness of sucrose. A 2026 safety study similarly described HT-AC as approximately 500–2,500 times sweeter than sucrose by weight, depending on the food or beverage matrix.

That places it in the same general range as other high-intensity sweet proteins such as thaumatin and brazzein.

Crucially, the story does not end with a sweet extract from an unusual fungus.

In the 2024 study, the gene encoding HTS was expressed heterologously in the yeast Komagataella phaffii (formerly Pichia pastoris), and the recombinant protein retained its sweet taste.

This strongly supports the conclusion that the protein itself is responsible for the sweetness.

In 2026, direct activation of the human sweet receptor was demonstrated

How can a protein taste sweet?

The principal human sweet-taste receptor is the class C GPCR heterodimer T1R2/T1R3.

In 2025, cryo-EM structures of the human T1R2/T1R3 receptor were reported, revealing how ligands such as sucralose and aspartame are recognized at the molecular level. Structures including PDB 9NOR became available.

Using this structural framework, Vo and colleagues investigated HT-AC and other sweet proteins in 2026.

The key experimental result was that HT-AC activated the human T1R2/T1R3 receptor in a cell-based assay.

Structural docking calculations further suggested that HT-AC does not simply occupy the same small-molecule pocket used by sugars or artificial sweeteners. Instead, the authors proposed an allosteric binding model in which the protein interacts with a broader receptor surface.

An important distinction is necessary here.

A direct experimental structure of the HT-AC–T1R2/T1R3 complex has not yet been determined by cryo-EM or X-ray crystallography.

Receptor activation is supported experimentally, but the precise protein–receptor binding pose remains a structure-based computational model.

Different folds, same sensation of sweetness

Another striking point from the 2026 work is the comparison between HT-AC and established sweet proteins.

HT-AC, thaumatin, brazzein and monellin all produce strong sweetness.

Yet they do not share strong overall sequence similarity or a common global protein fold.

One feature that appears repeatedly is the presence of positively charged regions on the protein surface.

This suggests that there may be no single structural fold that defines a “sweet protein.”

Instead, proteins with very different overall architectures may activate the same T1R2/T1R3 receptor through combinations of surface charge, local geometry and residue positioning.

In other words, sweetness may depend less on the overall protein scaffold than on the physicochemical properties of the receptor-contacting surface.

That does not mean that simply making a protein more positively charged will make it sweet. Small amino-acid substitutions and post-translational modifications can substantially alter sweet-protein activity, and the 2026 study also compared HT-AC isoforms and modified forms.

September 2026: high-resolution crystal structures

The solution NMR structure of HT-AC was released in April 2026 as PDB 9P6W.

The entry describes a monomeric protein of 121 residues with a molecular mass of approximately 13.39 kDa.

Then, on September 15, 2026, Pitkanen and colleagues published X-ray crystal structures of HT-AC in Acta Crystallographica Section F.

Four structures were deposited as 11OS, 11OT, 11OW and 11OX and released by the Protein Data Bank on September 23, 2026. The structures include complexes or crystallization conditions involving HEPES, 1,4-butanediol and glucose.

All were determined at resolutions in the 1 Å range and are broadly consistent with the NMR solution structure.

These structures now allow side-chain positions on candidate receptor-interacting surfaces to be discussed at near-atomic resolution.

One structure, 11OX, contains bound glucose. This should not be interpreted as evidence that glucose binding generates the sweetness of HT-AC. The major advance is the high-resolution structure of HT-AC itself; a direct experimental structure of HT-AC bound to the sweet receptor remains unavailable.

It can be made by yeast rather than by cultivating the truffle

The practical implications are also notable.

Natural M. terfezioides is not an easy organism to cultivate at industrial scale. HT-AC, however, can be produced through precision fermentation using K. phaffii.

A 2026 safety study evaluated this fermentation product as Honey Truffle Sweet Protein (HTSP).

Bacterial mutagenicity testing and a mammalian-cell micronucleus assay did not indicate genotoxicity. In a 13-week repeated-dose oral study in Wistar Han rats, the authors identified 5,400 mg HTSP/kg body weight/day, corresponding to approximately 1,500 mg HT-AC/kg/day, as the NOAEL.

This does not mean that humans can simply consume 1,500 mg/kg safely.

It is an animal-study NOAEL and must be distinguished from a human dietary exposure limit or a regulatory safety determination.

Another point worth noting is that many of the key HT-AC studies include researchers affiliated with the developer, MycoTechnology. The 2026 Biochemistry paper explicitly reports company funding. Independent replication will therefore remain important as the field develops.

Commercialization has already begun

This research is no longer purely academic.

MycoTechnology has introduced HTSP in the United States under the brand name Zukora and announced commercial-scale production in 2026.

Regulatory wording, however, requires care.

The company describes the ingredient as self-affirmed GRAS in the United States. Separately, FDA GRAS Notice GRN 1330, listed as “mycodulcein preparation produced by Komagataella phaffii,” remained Pending as of September 28, 2026.

It would therefore be inaccurate to describe the ingredient as already having received an FDA “no questions” response or FDA safety approval on the basis of that notice.

The search for sweet proteins may be far from over

Historically, the best-known high-intensity sweet proteins—thaumatin, brazzein and monellin—came from tropical plants.

HT-AC is different. It comes from a fungus.

It also differs substantially from established plant sweet proteins in both amino-acid sequence and overall structure, yet it activates the same human T1R2/T1R3 receptor.

That suggests that the sequence space capable of producing sweet proteins may be considerably broader than previously assumed.

Fungal genomes and transcriptomes may contain additional proteins with sweet-taste activity that have never been recognized because no one has tasted or functionally tested them.

Now that the structure of HT-AC is available, searches for new sweet proteins need not rely only on sequence homology. They can combine protein structure, surface electrostatics, local residue geometry and docking to T1R2/T1R3.

The field may be shifting from “find organisms that taste sweet” toward “find protein structures capable of activating the sweet receptor.”

The honey truffle could therefore represent not just another natural sweetener, but a starting point for a new way of discovering sweet proteins.

---

Key papers

McFarland CT. et al. (2024) Discovery, Expression, and In Silico Safety Evaluation of Honey Truffle Sweetener, a Sweet Protein Derived from Mattirolomyces terfezioides and Produced by Heterologous Expression in Komagataella phaffii. Journal of Agricultural and Food Chemistry 72: 19470–19479. DOI: 10.1021/acs.jafc.4c04368

Vo P. et al. (2026) Sweet Protein Allosteric Binding and Activation of the Human T1R2/R3 Sweet Receptor: A Simulation Model Validated by in Vitro Receptor Activation Assay. Biochemistry 65: 399–416. DOI: 10.1021/acs.biochem.5c00622

Potter SM. et al. (2026) Safety evaluation of honey truffle sweet protein produced from Komagataella phaffii. Food and Chemical Toxicology 211: 115987. DOI: 10.1016/j.fct.2026.115987

Pitkanen TT. et al. (2026) Crystal structures of the sweet-tasting protein honey truffle active component from Mattirolomyces terfezioides. Acta Crystallographica Section F. DOI: 10.1107/S2053230X26008745