Reproductive and developmental toxicity screening study of an acetone extract of rosemary.

In 2017, JECFA requested reproductive and developmental toxicity studies to finalize an acceptable daily intake for solvent rosemary extracts. Thus, an OECD 421 reproductive/developmental toxicity study was conducted using an acetone rosemary extract that complied with JECFA and EFSA food additive specifications. Rosemary extract was provided to rats at dietary concentrations of 0 (control), 2100, 3600, or 5000 mg/kg, for 14 days before mating, during mating, and thereafter (throughout gestation and up to Lactation Day 13 for females) until necropsy. General toxicity (clinical signs, body weight, food consumption) and reproductive/developmental outcomes (fertility and mating performance, estrous cycles, anogenital distance, thyroid hormones, reproductive organ weights, thyroid histopathology) were assessed. There were no signs of general toxicity and no effects on reproduction; thus, the highest concentration tested (equivalent to mean daily intakes of 316 or 401 mg/kg bw/day [149 or 189 mg/kg bw/day carnosol and carnosic acid] for males and females, respectively) was established as the no-observed-adverse-effect level for general and reproductive toxicity. Dose-related reductions in T4 were observed for Day 13 pups (not seen on Day 4) but were not accompanied by thyroid weight changes or histopathological findings; further investigations are required to determine the biological relevance of these T4 reductions.

Safety evaluation of arabinase (arabinan endo-1,5-α-L-arabinanase) from Aspergillus tubingensis.

Arabinase is an enzyme recognized for its ability to degrade arabinan, a plant cell wall constituent. It has been applied in the food industry most commonly for juice processing. One commercial source of arabinase is Aspergillus tubingensis (A. tubingensis), a black Aspergillus species. Given the intended use in food for human consumption, and noting its potential presence at trace levels in finished products, a series of safety studies including in vitro Ames and chromosome aberration assays, in vivo mammalian erythrocyte micronucleus and alkaline comet assays, and a 90-day rat oral toxicity study were conducted. No test article-related mutagenic activity was observed in the Ames assay. Although positive activity was observed in the chromosome aberration assay, this was not replicated in the in vivo genotoxicity assays including in preabsorptive cells. In the subchronic toxicity study, no test article-related adverse effects were observed following oral administration of arabinase at doses of 15.3, 153, or 1,530 mg total organic solids (TOS)/kg body weight/day to Sprague Dawley rats. The no-observed-adverse-effect level was considered to be the highest dose tested (1,530 mg TOS/kg body weight/day). The results of the genotoxicity studies and the subchronic toxicity study support the safe use of arabinase from A. tubingensis in food production.

Safety evaluation of highly-branched cyclic dextrin and a 1,4-alpha-glucan branching enzyme from Bacillus stearothermophilus.

Highly-branched cyclic dextrin (HBCD), a dextrin food ingredient presently only used in Japan, was investigated for digestibility and potential toxicity. HBCD was readily hydrolyzed in vitro to maltose and maltotriose by human salivary and porcine pancreatic alpha-amylases. Incubation of HBCD with a rat intestinal homogenate, containing digestive enzymes, resulted in the formation of maltose, maltotriose, and maltotetraose, and with longer incubation times, resulted in the formation of glucose. In an acute toxicity study, Wistar rats orally administered a single-dose of 2000mg/kg body weight of HBCD did not display mortality or any signs or symptoms of toxicity or abnormalities upon necropsy. Transient loose stools were observed, but were resolved within 24h of HBCD administration, and therefore, were not considered as compound-specific adverse effects. In the Ames assay, HBCD was non-mutagenic with or without metabolic activation. Toxicity testing of the branching enzyme (BE) involved in the synthesis of HBCD showed that the BE also was not acutely toxic when orally administered to rats and was non-mutagenic in the mouse lymphoma assay. The results of this study demonstrate that HBCD is digested to normal and safe products of carbohydrate digestion, and therefore, support the safety of HBCD for human consumption.

Safety evaluation of a peptide product derived from sardine protein hydrolysates (valtyron).

The peptide product, Valtyron, is obtained via enzymatic hydrolysis of sardine muscle. Although the safety and efficacy of the sardine peptide product have been evaluated in human studies, sardine peptides have not been identified as the subject of toxicological testing. In this study, the sardine peptide product did not exhibit any mutagenic activity in Salmonella typhimurium or Escherichia coli WP2uvrA. Likewise, the sardine peptide product was not associated with clastogenic properties in mouse bone marrow cells in a micronucleus assay. An oral rat LD(50) value of greater than 10,000 mg per kilogram of body weight was determined for peptide alpha-1000, and in rats administered peptide alpha-1000 by gavage at levels up to 5000 mg per kilogram of body weight per day for 28 days, no compound-related differences were observed in standard toxicological parameters. The results of these studies support the safety of the sardine peptide product for use in food for human consumption as a dietary source of peptides available from sardines.

An evaluation of the maximized survey-derived daily intake (MSDI) as a practical method to estimate intake of flavouring substances.

Realistic estimates of intake are essential for risk assessments of flavouring agents, since substantial over or underestimations introduce inaccuracies into such evaluations. The objectives of this study were to examine the relationship between intakes estimated using methods based on the reported volume of production [e.g., maximized survey-derived daily intake (MSDI)] versus use-level data [e.g., possible average daily intake (PADI) and modified theoretical added maximum daily intake (mTAMDI)]. The impact of volatility, self-limiting organoleptic properties and whether 10% of the population are eaters, an assumption in the MSDI calculation, on intake estimates were investigated. Analyses on 221 flavouring substances showed that intake estimates derived from MSDI correlated with values determined from detailed 14-day menu-census data, PADI, and mTAMDI. Comparisons of menu-census intake data adjusted to account for factors such as volatile losses showed that MSDI estimates are realistic and sufficiently conservative, whereas mTAMDI results in substantial overestimates of intake. Very few flavours have less than 10% eaters, and in the worst case, this assumption underestimates percent eaters by a factor of about 4. This investigation supports the use of MSDI as a conservative yet practical method to estimate intake of flavouring substances.