Safety Assessment of the Food Enzyme D-Psicose 3-Epimerase from Genetically Modified Corynebacterium

Enzyme safety approvals rarely make headlines. For formulators working with rare sugars, this one matters. The EFSA panel evaluated the food enzyme D-psicose 3-epimerase, also called D-allulose 3-epimerase, produced by a genetically modified Corynebacterium strain. Their task was simple: decide if it is safe for use in making D-allulose from fructose.

D-allulose is a low-calorie monosaccharide. It delivers about 70% of the sweetness of sucrose, but with far fewer calories. That makes it attractive for beverage, bakery, and confectionery applications. The commercial route to D-allulose relies on enzymatic isomerization. Fructose goes in. D-allulose comes out. The enzyme that does the work is the subject of this EFSA safety assessment.

The production strain is genetically modified. That fact alone triggers extra scrutiny. EFSA evaluates the enzyme preparation, the production process, and the toxicological data. They do not just ask whether the enzyme works. They ask whether it harms anyone who consumes products made with it.

What the study did

EFSA’s panel reviewed a standard battery of safety studies. The central piece was a repeated-dose oral toxicity study in rodents. They fed test animals the enzyme preparation at increasing dose levels. The key metric for dosing is TOS, or Total Organic Solids. TOS represents the organic fraction of the enzyme preparation, excluding diluents and carrier materials. It is the standard way to express enzyme exposure.

The panel looked for adverse effects across multiple endpoints. They checked body weight, food intake, clinical signs, organ weights, hematology, and histopathology. No treatment-related effects appeared.

Key findings

The critical number is this: the no-observed-adverse-effect level, or NOAEL, was 1796 mg TOS per kg body weight per day. That was the highest dose tested. In plain terms, animals showed no harm even at the top dose.

The other major finding concerns carryover. During D-allulose production, the process removes TOS. The final D-allulose contains no recombinant DNA. That matters for buyers who need clean ingredient specifications.

Parameter	Finding
Production organism	Genetically modified Corynebacterium strain
Enzyme function	Converts fructose to D-allulose (D-psicose)
NOAEL	1796 mg TOS/kg body weight/day, highest dose tested
TOS in final D-allulose	Removed during processing
Recombinant DNA in D-allulose	Not present

What it means for manufacturers

This EFSA opinion gives you a regulatory anchor. It tells you the enzyme itself is safe for its intended use. That supports your own ingredient qualification work.

But read the scope carefully. The safety assessment covers the food enzyme preparation from this specific production strain. It is not a blanket approval for all D-allulose enzymes. Your supplier must match the strain, process, and purity profile to the assessed product.

The absence of TOS and recombinant DNA in D-allulose is a practical advantage. It simplifies your purity documentation. It also reduces the risk of unintended carryover into finished food products. You still need your own specifications for residual protein, endotoxins, and enzyme activity. The EFSA opinion does not replace those.

FAQ

Is D-allulose made with this enzyme a GMO?
No. The genetically modified organism is the production strain. EFSA states that TOS is removed and no recombinant DNA remains in the D-allulose. The final sweetener does not contain the GM organism.

What does 1796 mg TOS/kg bw/day mean?
It is the highest dose tested in the animal study. No adverse effects occurred at that level. Toxicologists call this the NOAEL. It gives regulators a safety margin when setting acceptable exposure limits.

Does this approval cover every D-allulose product?
No. It covers this specific enzyme preparation from this specific production strain. Always verify that your supplier’s enzyme matches the one in the EFSA opinion.

Closing

Read the full opinion at DOI 10.2903/j.efsa.2021.6870. Use it as a baseline. Then ask your supplier for batch-specific data on TOS, DNA, and enzyme activity. That is where real safety assurance lives.