Sweet Protein Brazzein Passes First Human Safety Trial as a Sugar Stand-In

In a 21-day randomized controlled trial, 60 healthy adults ate wafers sweetened with recombinant brazzein — a protein produced in yeast that is 500 to 2,000 times sweeter than sucrose — while a control group ate matched wafers containing 30 grams of ordinary dietary sugar. The brazzein group showed no adverse changes in clinical or biochemical parameters, no allergic reactions, and only modest, exploratory shifts in gut microbiota, according to the study published in Food Production, Processing and Nutrition.

Brazzein is not an artificial sweetener in the conventional sense. It is a compact protein of just 54 amino acids, with a molecular weight of 6,473 daltons, originally isolated from the fruit of Pentadiplandra brazzeana. Because the natural content of the protein in its plant source is extremely low and extraction would be impractical, the researchers produced their version biosynthetically in Pichia pastoris yeast. The recombinant sequence used in the trial was identical to the natural one, and the purified protein was indistinguishable from natural brazzein by chromatographic and electrophoretic mobility, exact molecular mass, and fragmentation spectrum.

The motivation for the trial rests on two converging concerns. Global sugar consumption has tripled over the past 50 years, contributing to obesity, diabetes, cardiovascular disease, and dental caries, and excessive sugar intake has been linked to gut dysbiosis in both animal models and humans. At the same time, widely used artificial sweeteners such as aspartame and erythritol have been associated in observational and experimental work with increased risks of carcinogenesis, anxiety, cardiovascular problems, thrombus formation, ischemic stroke, dementia, and adverse changes in gut microbiota. Sweet-tasting proteins offer a third path: brazzein, in particular, does not influence blood glucose or insulin levels and shows remarkable thermal stability across a broad pH range.

The authors describe the study as a first-in-human safety and tolerability study of recombinant brazzein as a sugar alternative, with microbiota and psychological measures. The 60 participants, 28 men and 32 women aged 18 to 44, were hospitalized for the entire 21-day period solely to ensure full dietary standardization, continuous safety monitoring, and precise sample collection; the hospitalization was not related to any medical condition. Participants were randomized one to one using sealed opaque envelopes into a control group receiving sucrose-sweetened wafers or a brazzein group receiving brazzein-sweetened wafers. The experimental group received 0.02 grams of brazzein orally once daily in a 100-gram wafer; by sweetness equivalence, this dose corresponds to approximately 30 grams of sucrose, which the authors describe as a typical sugar load for a single sweet snack portion. The control group received daily wafers containing 30 grams of dietary sugar (sucrose). Because of the organoleptic differences between sucrose and protein sweeteners, blinding of participants was not feasible, but laboratory personnel analyzing biochemical and microbiota data were blinded to group allocation.

All participants received the same standardized background inpatient diet of roughly 2,170 to 2,400 kilocalories per day, enriched with plant fiber and excluding refined carbohydrates, spicy seasonings, and smoked foods. Participants were not allowed to consume products brought from outside the clinic, and compliance with the diet was monitored daily by the attending dietitian. The inpatient environment was designed to be as nonrestrictive and comfortable as possible: participants had unrestricted access to personal electronic devices such as phones and laptops, could communicate freely with one another, received visitors during designated hours, could move freely within the hospital including an outdoor park area, and could take part in leisure activities such as group movie screenings, quizzes, and board games.

Across the three weeks, the safety picture was unremarkable in the best sense. Vital signs — respiratory rate, blood pressure, heart rate, and body temperature — showed no significant differences before, after, or throughout the study in either group, and all integral health indicators stayed within reference ranges. The control group logged one case of excessive sweating and a headache; the brazzein group recorded no adverse reactions at all. Complete blood counts and urinalysis showed no significant changes in either group, and biochemical blood tests likewise revealed no significant differences between groups. The single statistically significant finding was a small rise in alkaline phosphatase activity in the brazzein group at the end of the study — negligible in absolute terms and still within the reference range. Body mass index remained normal throughout, with only a slight downward trend in both groups.

Allergy monitoring also came back clean. Total IgE levels were measured at the start and end of the trial, and although median values drifted upward in both groups — from 24.70 to 45.90 IU/ml in the control group and from 8.70 to 14.20 IU/ml in the brazzein group — there were no statistically significant differences between groups, all values remained within reference limits, and local allergic manifestations such as hyperemia, urticaria, itching, or Quincke’s edema were never detected in either arm.

The microbiota results demand more cautious reading. Fecal samples from 24 participants per group (48 participants in total, 96 samples across two time points) were analyzed using full-length 16S rRNA gene sequencing on an Oxford Nanopore platform, with all samples processed in a single sequencing run to minimize potential batch effects. Taxonomic annotation identified 482 bacterial species, dominated by Firmicutes, Bacteroidetes, Tenericutes, Proteobacteria, and Actinobacteria. Crucially, there were no significant differences between groups in species counts or Shannon alpha diversity, and beta-diversity analysis showed composition was significantly associated with both group and time point, though inter-individual variability precluded visually distinct clustering. Within-group longitudinal analysis found that brazzein consumption was associated with an increased relative abundance of Dialister invisus, a species normally present in the human gut whose metabolic products, including the short-chain fatty acids acetate and propionate, have been implicated in regulating inflammatory processes. Brazzein was also associated with increases in Bifidobacterium adolescentis and Negativibacillus massiliensis, and decreases in Prevotella copri and Eubacterium siraeum, while sucrose produced its own distinct pattern of shifts. The authors stress that, given the absence of direct between-group statistical comparisons and the exploratory nature of the analysis, these findings do not permit causal inference or conclusions about health effects.

Psychological measures were similarly framed as exploratory, since the controlled inpatient setting could influence emotional state independently of sweetener type. Well-being, activity, and mood scores on the daily WAM questionnaire remained stable and within normal range in both groups. HADS anxiety and depression scores stayed below the clinical threshold throughout, though both groups showed significant decreases in anxiety by day 21 — a change the authors attribute plausibly to reduced stress and lifestyle changes accompanying hospitalization rather than to either sweetener. On the gastrointestinal 7 × 7 questionnaire, minor transient complaints such as bloating and flatulence appeared in a small number of participants in both groups during initial adaptation to the standardized diet, with no statistically significant differences between brazzein and sucrose. A survey embedded in the trial underscored the public health stakes: 88 percent of respondents consumed sugar daily, 85 percent saw no benefit in it, rating its harm at 7.64 out of 10, yet only 56 percent were ready to give it up, and just 17 percent had ever used sugar substitutes.

The authors are explicit about the study’s limits. The cohort consisted of younger, healthy individuals without metabolic comorbidities, limiting generalizability; recent antibiotic exposure was not a formal screening exclusion criterion, a limitation the team acknowledges, although no antibiotics were given during the study period. The 21-day intervention cannot capture long-term microbiota stability, chronic metabolic effects, or rare adverse events, and the investigational products necessarily differed in macronutrient composition because replacing sucrose with a high-intensity protein sweetener changes the wafer’s carbohydrate and calorie content. Within those boundaries, the conclusion is measured: recombinant brazzein was well tolerated, produced no adverse changes in biochemical or physiological parameters, and caused no major structural disruption of the gut microbiota in healthy adults. Whether a protein sweetener can deliver on that promise over months and years, and in people with obesity or diabetes who stand to benefit most, will require the longer follow-up trials the researchers say must come next.

Subject of Research: First-in-human safety trial of recombinant brazzein as a sugar alternative examining biochemical status, gut microbiota, and psychological well-being

Article Title: Effects of recombinant brazzein consumption as a sugar alternative on human gut microbiota, biochemical status, and psychological well-being

Article References: Sokolov, P., Veselovsky, V., Kotlobay, A., Fomenko, A., Strokach, A., Andreev, D., Zoruk, P., Devkota, M., Olekhnovich, E., Tsaregorodtsev, S., Babenko, V., Kanatbek, A., Polina, N., Andreev, N., Zaborovsky, A., Klimina, K., Dontsova, O., Sergiev, P., Lagarkova, M., & Maev, I. V. (2026). Effects of recombinant brazzein consumption as a sugar alternative on human gut microbiota, biochemical status, and psychological well-being. Food Production, Processing and Nutrition, 8(1), Article 63. https://doi.org/10.1186/s43014-026-00421-9

Image Credits: AI Generated

DOI: 10.1186/s43014-026-00421-9

Keywords: brazzein, sweet protein, sugar substitute, gut microbiota, randomized controlled trial, food safety, recombinant protein, Pichia pastoris, artificial sweeteners, clinical trial, microbiome, Dialister invisus