First draft prepared by

Mrs I.M.E.J. Pronk
Centre for Substances and Integrated Risk Assessment, National Institute for Public Health and the Environment, Bilthoven, Netherlands


Dr C. Leclercq
National Research Institute for Food and Nutrition, Rome, Italy


Construction of production strain

Biochemical characterization

Biological data

Biochemical aspects

Toxicological studies

Acute toxicity

Short-term studies of toxicity

Long-term studies of toxicity and carcinogenicity


Reproductive toxicity

Observations in humans

Dietary intake





The enzyme preparation under evaluation contains the enzyme xylanase, which has not been evaluated previously by the Committee. Xylanase is produced by submerged fermentation of a strain of Fusarium venenatum that is non-pathogenic and non-toxigenic (under conditions consistent with good manufacturing practice), and which has been genetically modifed to carry a gene encoding a xylanase from Thermomyces lanuginosus, inserted by recombinant DNA techniques. The enzyme is subsequently partially purified and concentrated, resulting in a liquid enzyme concentrate (LEC). In the final preparation, this LEC is stabilized, standardized and formulated with sodium chloride, dextrin, sorbitol, and wheat solids.

The enzyme produced is an endo-xylanase, which hydrolyses xylosidic link-ages in the arabinoxylans into smaller oligosaccharides. The enzyme preparation is used in baking applications to increase the elasticity of the gluten network, improving handling and stability of the dough. The enzyme is denatured and inactivated during bread baking.

The activity of the petitioned xylanase enzyme preparation, called "Novozym 899", is expressed in fungal xylanase units (wheat), FXU(W), and is measured using a colorimetric method. Novozym 899 has a typical activity of 2500 FXU(W)/g, and has the following composition: total organic solids (TOS), approximately 4%; wheat solids, approximately 77%; ash (mainly sodium chloride), approximately 4%; dextrin, approximately 4%; sorbitol, approximately 1%; water, approximately 10%.

Novozym 899 is used as a processing aid in the baking industry to improve dough stability and crumb structure, resulting in a more uniform and softer crumb and increased volume of the bread. It can be used for all types of bread as an alternative for, or in combination with, emulsifiers. The enzyme preparation is added to the flour or to the liquid and is active during the preparation of the dough and the leavening of the unbaked bread. The recommended dosage is 216 g of the enzyme preparation per 100 kg flour, corresponding to 50400 FXU(W)/kg flour.

Toxicological studies have been performed with an LEC (batch PPQ 6125) omitting formulation, stabilization and standardization. The characteristics of test batch PPQ 6125 were as follows: enzyme activity, 8590 FXU(W)/g; TOS, 10.8% w/w; water, 88.7% w/w; ash, 0.5% w/w; density, 1.041 g/ml (Pedersen & Broadmeadow, 2000; Hvass, 2002).

1.1 Construction of production strain

The production strain LyMC4.B was developed from the F. venenatum host strain MLY3. The MLY3 strain is a spontaneous mutant of the CC1-3 strain that, in turn, is a spontaneous mutant of the F. venenatum strain used in the production of mycoprotein marketed for human consumption since 1985 under the trade name "Quorn." The MLY3 strain was transfected with the expression plasmid pjRoy36 containing the xylanase gene from T. lanuginosus and the bar gene from Streptomyces hygroscopicus. The bar gene confers resistance to the herbicide phosphinothricin and serves as a selectable marker. A single transformed colony producing xylanase was selected. The selected strain was designated as JRoy36-19.B.

The JRoy36-19.B strain was subsequently transfected with DNA containing the amdS gene from Aspergillus nidulans. The amdS gene was flanked by specific sequences of the F. venenatum tri5 gene and replaced the tri5 gene in the JRoy36-19.B strain. One of the transformants was designated as LyMC4.B and was used as a xylanase production strain.

1.2 Biochemical characterization

Deletion of the tri5 gene served to inactivate the biochemical pathway by which mycotoxins (trichothecenes) are synthesized. To confirm that the tri5 gene had been deleted, the transformed strain was evaluated for production of diacetoxyscirpenol (DAS), the major trichothecene produced by non-engineered strains of F. venenatum: no DAS was detected. F. venenatum is capable of producing other secondary metabolites, such as culmorins, enniatins and fusarins. Analyses performed on conventional F. venenatum under conditions known to be optimal for production of these secondary metabolites revealed them to be present only at very low concentrations. As it is very unlikely that the production strain LyMC4.B produces these secondary metabolites to the same extent under industrial fermentation conditions, the concentrations of these secondary metabolites, if present, are considered to be of no toxicological relevance.


2.1 Biochemical aspects

The xylanase was assessed for potential allergenicity by amino acid sequence comparison with known allergens listed in publicly available protein databases. No immunologically significant sequence homology was detected.

2.2 Toxicological studies

2.2.1 Acute toxicity

No information was available.

2.2.2 Short-term studies of toxicity


Groups of 10 male and 10 female CD rats (aged 3640 days) received water containing xylanase (batch PPQ 6125) at a dose of 0, 1, 3.3, or 10 ml/kg bw per day (equivalent to 0, 8942, 29 509, and 89 422 FXU(W)/kg bw per day and 0, 0.11, 0.37, and 1.1 g/kg bw per day of TOS) by oral gavage for 13 weeks. The study was performed according to OECD test guideline 408 (1998), and was certified for compliance with good laboratory practice (GLP) and quality assurance. Animals and cage-trays were inspected at least twice per day for reactions to treatment or ill health. All animals were observed individually before and after dosing daily during week 1 of treatment, twice per week during weeks 24, and once per week during weeks 513. In addition, all animals underwent a detailed physical examination, including palpation, each week. Body weight and food consumption were recorded weekly. Food conversion efficiency was calculated. Food and water were freely available. Functional observation battery tests were performed on all animals (hand and standard arena observations: before treatment and weekly during treatment; reflexes and motor activity: before treatment and during week 13). Ophthalmoscopy was carried out before treatment on all animals and in week 12 on animals in the control group and the group receiving the highest dose. During week 13, haematology and clinical chemistry were performed for all animals. Absolute weights of nine organs were determined and adjusted for body weight. All animals were examined macroscopically. Microscopy was carried out on about 35 organs and tissues of all animals in the control group and in the group receiving the highest dose, and on all macrospically abnormal tissues.

No effects on survival or behaviour were seen. Ophthalmoscopy was normal. Functional observation battery tests did not reveal any abnormalities. The slightly increased body-weight gain and food consumption of all treated females did not show any dose-response relationship. As male rats were also not affected, these changes were considered to be not biologically relevant. Haematology and clinical chemistry did not reveal any abnormalities, organ weights were normal and macroscopy and microscopy did not reveal any effects related to treatment. The Committee concluded that in this 13-week study in rats treated orally, the NOEL for xylanase (batch PPQ 6125) was the highest dose, 10 ml/kg bw per day (equivalent to 89 422 FXU(W)/kg bw per day and 1.1 g of TOS/kg bw per day) (Baguley, 1999; Pedersen & Broadmeadow, 2000).

2.2.3 Long-term studies of toxicity and carcinogenicity

No information was available.

2.2.4 Genotoxicity

The results of two studies of genotoxicity in vitro with xylanase (batch PPQ 6125) are summarized in Table 1. Both studies followed OECD test guidelines, 471 (1997) and 473 (1997), respectively, and were certified for compliance with GLP and quality assurance.

Table 1. Results of studies of the genotoxicity of xylanase (batch PPQ 6125)


Test object




In vitro


Reverse mutation

S. typhimurium TA98, TA100, TA1535, TA1537 and E. coli WP2uvrA

1565000 g/ml for Salmonella strains and
1565000 g/ plate for E. coli.
Solvent: sterile water


Pedersen (1999); Pedersen & Broadmeadow (2000)

Chromosomal aberration

Human lymphocytes

2450, 3500, and 5000 g/ml -S9; 1201, 2450, 3500, and 5000 g/ml +S9.
Solvent: sterile water


Burman (1999); Pedersen & Broadmeadow (2000)


In the presence and absence of metabolic activation from S9; no cytotoxicity was seen. Owing to the presence of free amino acids (e.g. histidine and tryptophan) in the xylanase preparation, the growth of Salmonella strains requiring histidine was significantly increased after direct-plate incorporation. Therefore, the Salmonella strains were exposed to the a-amylase preparation in a phosphate-buffered nutrient broth in liquid culture ("treat-and-plate assay") at six concentrations (highest dose, 5 mg/ml) for 3 h. After incubation, the test substance was removed by centrifugation before plating. Stimulation of growth of E. coli strains requiring tryptophan was only weak and insignificant


In the presence and absence of metabolic activation from S9. In the first experiment, cells were treated for 3 h in the absence and presence of S9 and were harvested 17 h later. Little or no mitotic inhibition (06%) was seen. In the second experiment, cells were exposed continuously for 20 h in the absence of S9 and then harvested (mitotic inhibition, 29% at 5000 g/ml), or treated for 3 h in the presence of S9 and harvested 17 h later (mitotic inhibition, 28% at 5000 g/ml)

S9, 9000 g supernatant of rat liver homogenate

2.2.5 Reproductive toxicity

No information was available.

2.3 Observations in humans

No information was available.


Xylanase enzyme preparations are used in the baking industry to increase the elasticity of the gluten network, in order to improve handling and stability of the dough (Association of Manufacturers and Formulators of Enzyme Products, 2003).

In Australia and New Zealand, xylanase enzyme preparations have been approved, as have preparations of other carbohydrate-modifying enzymes, for use in beer, spirits, glucose syrups, bread, sugar, enzyme-modified starches and fruit juices (Food Standards Australia New Zealand, 2003). A "generally recognized as safe" (GRAS) notice was received for this specific enzyme preparation in 2000 (Food & Drug Administration, 2003). A complete and comprehensive list of enzymes and their uses in food manufacturing in the European Union was not available, but an inventory of enzyme use in nine Member States was compiled for scientific cooperation (SCOOP) Task 7.4 (European Commission, 2000). It was reported that a number of xylanase enzyme preparations were in use, mainly for treatment of flour and for bakery goods, but not the specific enzyme preparation under review.

Recommended dosage and TOS content of the enzyme preparation were provided by the sponsor (Hvass, 2002).

A "worst-case" scenario was estimated on the basis of the following assumptions:

According to the budget method, the upper physiological intake of food is 50 g/kg bw per day (Hansen, 1979). If xylanase is used only in the baking industry, a "worst-case" scenario is that of ingestion of baking products at 25 g/kg bw per day, leading to an intake of TOS of 0.115 mg/kg bw per day (16 4% 0.025 100 / 140), i.e. 6.9 mg of TOS per day for a 60-kg person. When compared with the NOEL of 1.1 g of TOS/kg bw per day in the 13-week study of oral toxicity, the margin of safety is nearly 10 000.


Toxicological studies were conducted on the LEC. The materials added to the LEC for stabilization, formulation and standardization have either been evaluated previously by the Committee or are common food constituents and do not raise safety concerns.

In a 13-week study in rats, no significant treatment-related effects were seen when the LEC was administered at doses of up to and including 10 ml/kg bw per day by oral gavage. Therefore this highest dose tested (equivalent to 1.1 mg of TOS/kg bw per day) was the NOEL. The LEC was not active in an assay for mutagenicity in bacteria in vitro nor in an assay for chromosomal aberrations in mammalian cells in vitro.

A conservative estimate of daily intake resulting from the use of xylanase in bakery goods is 6.9 mg of TOS/day (equivalent to 0.12 mg/kg bw per day). Compared with the NOEL of 1.1 g of TOS/kg bw per day in the 13-week study of oral toxicity, the margin of safety is nearly 10 000.


The Committee allocated an ADI "not specified" to xylanase from this recombinant strain of F. venenatum, used in the applications specified and in accordance with good manufacturing practice.


Association of Manufacturers and Formulators of Enzyme Products (2003) Enzymes used in food (available at http://www.amfep.org/enzymes/enz3.html).

Baguley, J.K. (1999) Xylanase, PPQ 6125. Toxicity study by oral gavage administration to CD rats for 13 weeks. Unpublished report No. NVO166/992031 from Huntingdon Life Sciences Ltd., Suffolk, England. Submitted to WHO by Novozymes A/S, Bagsvaerd, Denmark.

Burman, M. (1999) Xylanase. Induction of chromosome aberrations in cultured human peripheral blood lymphocytes. Unpublished report No. 665/242-D5140 from Covance Labs. Ltd, Harrogate, England. Submitted to WHO by Novozymes A/S, Bagsvaerd, Denmark.

European Commission (2000) Report on task for scientific cooperation (SCOOP). Report of experts participating in Task 7.4. Study of the enzymes used in foodstuffs and collation of data on their safety (http://www.europa.eu.int/comm/food/fs/scoop/index_en.html).

Food & Drug Administration (2003) List of the substances that are the subject of each GRAS Notice (http://www.cfsan.fda.gov/~rdb/opa-gras.html).

Food Standards Australia New Zealand (2003) Approved genetically modified processing aids and food additives and their use. Canberra (http://www. foodstandards.gov.au/ whatsinfood/gmfoods/approvedgmprocessing1031.cfm).

Hansen, S.C. (1979) Conditions for use of food additives based on a budget for an accept-able daily intake. J. Food Protect., 42, 429434.

Hvass, P. (2002) Xylanase enzyme preparation produced by a strain of Fusarium venenatum containing the gene coding for xylanase from Thermomyces lanuginosus, inserted by recombinant DNA techniques. Unpublished report No. 2002-50717-01 from Novozymes A/S, Bagsvaerd, Denmark. Submitted to WHO by Novozymes A/S, Bagsvaerd, Denmark.

Pedersen, P.B. (1999) Xylanase (batch number: PPQ 6125): Test for mutagenic activity with strains of Salmonella typhimurium and Escherichia coli. Unpublished report No. 998012 from Novo Nordisk A/S, Bagsvaerd, Denmark. Submitted to WHO by Novozymes A/S, Bagsvaerd, Denmark.

Pedersen, P.B. & Broadmeadow, A. (2000) Toxicological studies on Thermomyces lanuginosus xylanase expressed by Fusarium venenatum, intended for use in food. Food Addit. Contam., 17, 739747.

    See Also:
       Toxicological Abbreviations