Safety assessment of glutaminase from Aspergillus niger

Abstract Glutaminase (glutamine aminohydrolase EC 3.5.1.2) is used in the production of food ingredients rich in l‐glutamic acid that are added to finished foods for the purpose of enhancing or improving the savory flavor profile of food. The glutaminase enzyme preparation evaluated in these studies, designated as Sumizyme GT hereafter, is obtained by fermentation of Aspergillus niger strain GT147. The safety of Sumizyme GT was evaluated in a series of standard toxicological studies, including a 90‐day oral toxicity study in rats, an in vitro bacterial reverse mutation assay, an in vitro mammalian chromosome aberration test, and an in vivo alkaline Comet assay. Sumizyme GT was not mutagenic or genotoxic, and administration of the enzyme by gavage at doses up to 2,570 mg total organic solids (TOS)/kg body weight (bw) per day for 90 days was without any systemic toxicity. The no‐observed‐adverse‐effect level was concluded to be 2,570 mg TOS/kg bw per day, the highest dose tested. Considering that A. niger has an established history of safe use in the food industry and its safety in the production of food ingredients and food enzymes is well documented, the results of these studies provide further support of the safety of glutaminase from A. niger when used in food production.

However, a strain of Aspergillus niger, a filamentous fungus, isolated from a food source has been found to have high levels of glutaminase activity, making it suitable for commercial use. A. niger has a well-established history of safe use in food production, with one of the earliest reported industrial uses of A. niger in the production of citric acid in 1919 (Max et al., 2010). Since then, A. niger has been used in the production of numerous food ingredients, including chitosan and chitin-glucan, and food enzymes, such as amylase, amyloglucosidase, asparaginase, catalase, glucose oxidase, glucanase, cellulases, lactase, lipase, inulinase, invertase, pectinase, pentosanase, phospholipase, acid proteases, and xylanases, many of which have been concluded to be generally recognized as safe by the United States Food and Drug Administration (U.S. FDA) for use in food processing (US FDA, 2011, 2012, 2018 or by other regulatory bodies (FSANZ, 2019;Health Canada, 2019).
Limited toxicological information on orally administered glutaminase was identified in the scientific literature. Ohshita et al. (2000) conducted a safety assessment of glutaminase derived from different strains of Cryptococcus albidus, a filamentous fungus, and reported a no-observed-adverse-effect level (NOAEL) of 7.5 and 7.9 g/kg body weight (bw)/day for male and female ddY-F mice, respectively, 8.8 and 9.9 g/kg bw per day for male and female F244/ DuCrj rats, respectively, and 1.1 and 1.2 g/kg bw per day for male and female Std:Wistar rats, respectively, the highest dose tested in all cases, based on the results of 13-week oral toxicity studies. The reported NOAEL for a 1-year feeding study in ddY-F mice was 13 g/ kg bw per day for males and 15.5 g/kg bw per day for females, the highest dose tested. Despite the difference in fungal species, these results suggest that glutaminase derived from A. niger may be without toxicological effects. The objective of this study was to evaluate the safety of a glutaminase enzyme preparation derived from A. niger, namely Sumizyme GT, for use in food processing. Accordingly, a battery of genotoxicity and mutagenicity studies was conducted to assess the genotoxic or mutagenic potential of Sumizyme GT, including an in vitro bacterial reverse mutation assay, an in vitro mammalian chromosome aberration test, and an in vivo alkaline Comet assay. The subchronic toxicity of Sumizyme GT was evaluated in a 90-day repeated-dose oral toxicity study in Sprague Dawley rats.
The details of these toxicological studies are outlined in detail below. per ml. The test article was stored in airtight containers at −30 to −5°C. The test article was demonstrated analytically to be stable as validated by consistent enzyme activity and ultraviolet absorbance (280 nm) throughout the study period in the 90-day study.

Sumizyme GT was prepared by standard culture methods with
A. niger strain GT147. This strain of A. niger was selected based on its ability to produce high levels of glutaminase activity, its viability, and suitability for industrial production, including its lack of mycotoxin production. A. niger strain GT147 has been deposited at the Biological Resource Center, National Institute of Technology and Evaluation (Deposit No. 00326).

| Bacterial reverse mutation test (Ames test)
The bacterial reverse mutation test was conducted in accordance with Organisation for Economic Co-operation and Development (OECD) Guideline No. 471 (OECD, 1997) using the preincubation method and a modified method (treat-and-wash method) that involves a washing step to remove free amino acids, such as histidine, released into the culture medium that have been associated with overall increase in bacterial growth and in additional spontaneous mutations to occur. The treat-and-wash method is considered to be a valid method to eliminate the potential for "false-positive" effects due to the presence of free amino acids in the test substance (EFSA, 2014;Thompson, Morley, Kirkland, & Proudlock, 2005).
Preliminary and concentration-finding studies were conducted using the preincubation method to establish suitable concentrations and to evaluate mutagenicity prior to the initiation of the main study.
A concentration-finding study was also conducted using the treatand-wash method to establish appropriate concentrations for the main study using the modified method.
In the main study using the preincubation method, Sumizyme GT concentrations were prepared using a dilution factor of 2 to obtain final test concentrations of 402, 803, 1,610, 3,210, and 6,430  Precipitation of the test article was observed macroscopically in all plates at the initiation of treatment and at the time of colony counting. Growth inhibition was examined using a stereoscopic microscope and revertant colonies counted using an automated colony analyzer (CA-11, System Sciences). Responses were considered positive if the mean number of revertant colonies was at least 2 times greater than the negative control, a concentration-dependent relationship was observed, or the effects were reproducible. Statistical analysis of the data was not conducted.

| In vitro mammalian chromosomal aberration test
The in vitro mammalian chromosome aberration test was conducted in accordance with OECD Test Guideline No. 473 (OECD, 2014).

Cultured Chinese hamster lung (CHL) fibroblast cells (National
Institute of Health Sciences, Japan) were used in either a 6-hr shortterm assay in the presence or absence of metabolic activation or a 24-hr continuous assay in the absence of metabolic activation. The S9 microsomal fractions were obtained as previously described for the bacterial reverse mutation test. Distilled water served as the negative control. Mitomycin C (MMC) (Lot No. 577AEE, Kyowa Hakko Kirin) served as the positive control at a concentration of 0.1 µg/ml in the short-term assay in the absence of metabolic activation and at a concentration of 0.05 µg/ml in the continuous treatment assay.
Cyclophosphamide (Lot No. 4,399,Shionogi) served as the positive control in the short-term assay in the presence of metabolic activation at a concentration of 12.5 µg/ml.
In the short-term assay, CHL cells were incubated with Sumizyme GT for 6 hr at final concentrations of 1,000 to 10,000 µg TOS/ml in the absence of metabolic activation and 250 to 8,000 µg TOS/ml in the presence of metabolic activation. In the continuous assay, CHL cells were incubated with Sumizyme GT for 24 hr at final concentrations of 500 to 10,000 µg TOS/ml in the absence of metabolic activation. The concentrations were chosen based on preliminary testing to determine mitotic index and cell growth inhibition (data not shown).
Microscopic examinations were carried out in the short-term test in the absence and presence of metabolic activation and the continuous treatment test with analyzable upper-limit Sumizyme GT concentrations of 4,000, 3,000 and 3,000 µg TOS/ml, respectively.
Precipitation of the test article and changes in the color of the culture medium were observed with the naked eye at the beginning of the study and at the end of the treatment.
The CHL cells were seeded in culture plates and incubated for 3 days, followed by treatment with the negative or positive control or Sumizyme GT. The cells were then incubated for 6 hr for the short-term assay and 24 hr for the continuous assay. Following the incubation period, the cells in the short-term assay were detached by an effect of the test article. The medium was removed from each plate, transferred into individual centrifuge tubes, and rinsed with Dulbecco's phosphate-buffered saline (PBS) (Lot No. RNBF4689, Sigma-Aldrich), and fresh medium was added to each plate. The suspension was then centrifuged (1,000 rpm, 5 min), the supernatant was removed, and the collected cells were washed with PBS.
Following removal of the supernatant, fresh medium was added to each tube and the cells were resuspended and returned to the original plate. The cells were incubated for an additional 18 hr (total incubation of 24 hr after the start of exposure).
Two hours prior to preparation of the slides for examination, 0.2 µg/ml of colcemid solution (Lot No. 1,776,482;Life Technologies) was added to each plate to inhibit mitosis in the metaphase. Following incubation, the culture medium was transferred to a centrifuge tube and the cells were detached from the plate with 2 ml of 0.25% trypsin-ethylenediaminetetraacetic acid (EDTA) (Lot No. 1,798,279;Life Technologies). The cell suspension was also added to the tube. Relative cell growth rate was calculated using a portion of the cell suspension, and the remaining cell suspension was centrifuged (1,000 rpm, 5 min), and the supernatant was removed. The cells were then exposed to a hypotonic treatment for 16 min in 5 ml of a 75 mmol/L KCl solution that was prewarmed to 37°C and centrifuged, and the supernatant was removed. Next, the cells were twice fixed with an ice-cold 3:1 mixture of methanol and acetic acid, followed by cell resuspension in the fresh fixative mixture. One chromosome slide per plate was prepared to confirm the cell density. After confirmation, each suspension was added to a slide and 3 slides per plate were prepared for chromosome analysis. The slides were stained with 1.2% Giemsa solution (Lot No. HX57878704;Merck). Slides with greater than 50% relative cell growth rates were not subjected to chromosome analysis as the slides contained the presence of many c-mitosis figures, and showed almost no analyzable mitosis. The analyzable upper-limit GT concentrations were therefore selected as the highest concentrations (−S9 assay: 4,000 µg TOS/ml; +S9: 3,000 µg TOS/ml; 24-hr assay: 3,000 µg TOS/ml); in addition, three concentrations at appropriate intervals (including the highest concentration) were selected for microscopic examination.
Microscopic evaluation of the selected concentrations involved examination of 150 metaphase cells per plate or 300 metaphase cells per concentration, for the following chromosomal or chromatid-type aberrations: chromosome and chromatid gaps, breaks, exchanges, and others. In addition, the number of polyploid cells (38 chromosomes or more) was counted by observing 300 metaphases for each concentration.

The incidence of aberrant cells was analyzed by a one-side
Fisher's exact test with a 2.5% level of significance. If significant differences were observed, a one-sided Cochran-Armitage test with a 2.5% level of significance was used to analyze concentration dependency. Responses were considered to be positive if the incidence of cells with chromosomal aberrations was statistically significantly increased in more than one of the test concentrations compared to the negative control, a significant concentration-dependent increase is observed, and the observed incidence is greater than the acceptable range calculated from the historical data of the negative control group. The final evaluation was made on the total incidence of aberrant cells minus the number of cells with only gaps.

| In vivo alkaline comet assay
The in vivo mammalian alkaline comet assay was conducted in ac- All animals were observed for clinical signs at 0.5, 21 (prior to second administration), 21.5, and 24 hr (prior to necropsy) after the first administration. Body weight was measured prior to necropsy.
The animals were euthanized by inhalation of CO 2 3 hr after the second administration of either test article, negative control, or positive control. The stomach (glandular stomach) and duodenum were resected and rinsed with homogenizing buffer prepared from Hanks' balance salt solution (Lot No. 1,843,038; Life Technologies) containing 8.93 g/L EDTA·2Na (Lot No. KP001, Dojindo Laboratories) and 10 vol% dimethyl sulfoxide (Lot No. DSG2253, Wako Pure Chemical Industries), pH 7.5, and macroscopically observed for abnormalities.
A portion of the glandular stomach and duodenum (approximately 1 cm from the pylorus) were collected for potential histopathological examination, and the remaining tissues were used for the comet assay.
For the comet assay, the glandular stomach was dissected and incubated with cooled homogenizing buffer for 15 to 30 min. The surface epithelium was gently scraped with a blade and discarded; the epithelial cells were released following scraping with homogenizing buffer. The duodenum was dissected and rinsed with homogenizing buffer. An adequate volume of homogenizing buffer was added to each of the tissue samples, and the samples were homogenized using a Dounce tissue grinder. The cells were then centrifuged (800 rpm, 5 min), the supernatant was removed, and the cells were resuspended in the remaining supernatant. Each suspension (10 µl) was put into a microbe tube. Slides were prepared in triplicates (2 slides for evaluation, 1 for spare) on a superfrosted glass slide precoated with 1.0% agarose gel. Ninety microliters of 0.5% low-melting agarose gel (Lot No. 0,000,587,240; Lonza Rockland) was added to the microbe tubes containing cells and mixed. Following mixing, 90 µl of the cell-agarose mixture was placed on the precoated slide and covered with a noncoated superfrosted glass slide. After the agarose solidified and the covered glass slide was removed, the slides were placed in lysing solution (pH 10) and left overnight under refrigerated and light-protected conditions. The slides were then rinsed with electrophoresis buffer and placed in a submarine-type electrophoresis chamber (BE-540, BIOCRAFT), and chilled electrophoresis buffer was gently added to the chamber until the slides were completely immersed and the slides were left for 20 min (unwinding). Electrophoresis was conducted at a constant voltage of 25 V (0.7 V/cm; initial current: 300 mA) for 20 min. Throughout the unwinding and electrophoresis steps, the electrophoresis chamber was cooled on ice in order to keep the electrophoresis buffer at a low temperature (at the start of unwinding: 0.1 to 0.6°C, at the start of electrophoresis: 1.2 to 2.4°C, at the end of electrophoresis: 3.5 to 5.1°C). Following electrophoresis, neutralization of the alkali in the gels was carried out by immersion of the slides into a neutralizing solution (pH 7.5) for 10 min. The neutralized slides were dehydrated for 10 min in ethanol (≥99.5%), and no damage was observed; thus, the spare slides were discarded without electrophoresis.
Prior to microscopic examination, all slides were coded at random and treated with 50 µl of SYBR ® Gold nucleic acid gel stain (Lot No. 1,832,446; Life Technologies) diluted 5,000-fold with Tris-EDTA buffer (pH 8.0; Lot No. 02234G; Nippon Gene). The images of DNA migration of cells were examined with a fluorescence microscope (objective: ×20) with IB excitation. The images were then imported to a computer through a CCD camera that is attached to the microscope. A Comet assay analyzer (Comet Assay IV system, Perceptive Instruments) was used for the analysis.
One-hundred fifty cells [75 cells per slide; that is, 750 cells per group (5 animals)] were analyzed using a fluorescence microscope (×200), and the number of hedgehogs was counted.
Histopathological examination of the stomach or duodenum was not conducted as no clear DNA damage was observed in the comet assay. The percentage of DNA in the tail relative to the total (% tail DNA:Tail % intensity) served as the indicator for DNA damage.
Each slide was analyzed to determine the media % tail DNA (slide value), and the mean of the slide values for each animal (animal value) was calculated. The % tail DNA (animal value) was logarithmically transformed prior to statistical analysis. A two-sided Dunnett's multiple comparison test was used to analyze the % tail DNA (mean of animal values) between the negative control and each test article-treated group at a 5% level of significance. If a significant difference was observed, a two-sided linear trend test was conducted to analyze dose dependence at a 5% level of significance. A one-sided Aspin-Welch's t test was used to analyze the % tail DNA (mean of animal values) for the comparison between the negative and positive control groups at a 2.5% level of significance. The test result was considered to be positive, if at least 1 test article-treated group demonstrates a statistically significant increase in % tail DNA compared to the control, a significant dose dependence is observed in the % tail DNA, and the % tail DNA (mean of animal values) of the test article-treated groups exceeds the historical data of the negative control group.
A second comet assay was conducted to confirm the results of Sumizyme GT in the duodenum of rats at doses of 1,260, 1,800, and 2,570 mg TOS/kg. The dose formulations were prepared as previously described. A second group with heat-inactivated Sumizyme GT was also tested at the highest dose. Following euthanization of the animals, the duodenum was resected as it is immediately exposed to the test article following oral administration and the test article is considered to be present at a relatively high concentration in this organ. The duodenum was dissected, rinsed, and prepared for microscopic examination as previously described. Statistical analysis was conducted using a two-sided Dunnett's multiple comparison test.
As no significant difference was observed, dose dependence (trend) was not analyzed. Following the analysis of the % tail DNA (mean of animal values) using the one-sided Aspin-Welch's t test, a two-sided F test was used to analyze the % tail DNA (mean of animal values) for homogeneity of variance of the negative control and inactivated Sumizyme GT groups and between the activated and inactivated Sumizyme GT high-dose groups at a level of significance of 5%. Since homogeneity of variance was determined to not be significant by the F test, a one-sided Student's t test was used to compare the same groups at a level of significance of 2.5%.

| 90-day repeated-dose toxicity study
The 90-day repeated-dose oral toxicity study was conducted in accordance with OECD Test Guideline No. 408 (OECD, 1998b).

| Preparation of dosing formulations
The low-dose and mid-dose formulations were prepared by fourfold serial dilution of the thawed Sumizyme GT stock solution (128.5 mg TOS/ml) with distilled water to obtain concentrations of 8.0 and 32.1 mg TOS/ml, equivalent to doses of 161 and 643 mg TOS/kg bw per day, respectively. The undiluted stock solution (128.5 mg TOS/ml) served as the high-dose formulation, equivalent to 2,570 mg TOS/kg bw per day. The stability of the dose formulations was validated by measurements of enzyme activity, and the results demonstrated the test doses were stable after storage under refrigeration temperatures for 7 days followed by 24-hr storage at room temperature at concentrations up to 128.5 mg TOS/ml.

| Animals and treatment
Forty-five male and 45 female 4-week-old Crl:CD(SD) [SPF] rats were obtained from Charles River Laboratories Japan, Inc. Upon receipt, animals were observed daily for general condition throughout the 7-day acclimation period and body weight was measured on the day of receipt (Day −7) and the final day of the quarantine and acclimation period (Day 1). No abnormalities in clinical signs and body weight were observed from Days −7 to 1. Before dosing (Days −2 and −1), the animals underwent ophthalmological examination. Unignorable congenital defects were observed in the eyes of 1 male and 2 females; these rats were excluded from the study prior to group assignment (surplus animals). On the first day of dosing (Day 1), animals were randomly assigned to groups (10/sex/group) based on body weights using a computer system package for safety studies (LATOX-F/V5, FFC). The body weight of the animals used in the study ranged from 129 to 155 g for males and 111 to 135 g for females.
The doses of Sumizyme GT were selected based on the results of a 14-day dose range-finding study in rats in which no treatment-related toxicological effects on general condition, body weight change, food consumption, hematological parameters, and necropsy findings were reported at doses up to 2,570 mg TOS/ kg bw per day (data not shown). As a result, the high-dose group was selected as the maximum dose and 161 and 643 mg TOS/kg bw per day were selected for the low-and mid-dose levels, respectively, in the 90-day study. The control animals received distilled water (vehicle solvent) at an equivalent dosing volume to the treatment group. The test article and the vehicle control were administered at a dosing volume of 20 ml/kg bw per day via gavage using a plastic syringe and Teflon gastric tube. The dosing solutions were prepared once or more within 7 days at concentrations of 8.0, 32.1, and 128.5mg TOS/ml for the low-dose (161 mg TOS/kg bw per day), mid-dose (643 mg TOS/kg bw per day), and high-dose (2,570 mg TOS/kg bw per day) groups, respectively.

| Clinical observations, body weights, food consumption, and ophthalmology
The general condition of the animals was observed and recorded twice per day (before and after dosing) and once before necropsy

| Clinical pathology and urinalysis
Hematological and blood chemical examinations were conducted in all surviving animals on the day of the scheduled necropsy (Day 91). Prior to necropsy, the animals were fasted overnight, and blood samples were collected the following day from the abdominal aorta under isoflurane anesthesia. Hematological examination was carried out using blood and plasma samples. Blood samples were collected in tubes containing an anticoagulant (EDTA dipotassium salt) and were analyzed with a Hematology System (ADVIA120; Bayer). The following parameters were analyzed: hematocrit, hemoglobin, red blood cell count, mean corpuscular volume, mean corpuscular hemoglobin, mean corpuscular hemoglobin concentration, reticulocyte ratio, reticulocyte count, platelet count, white blood cell count, differential leukocyte ratios, leukocyte count (neutrophil, lymphocyte, monocyte, eosinophil, basophil),and large unstained cell count.
Plasma samples were collected in tubes containing an anticoagulant (3.2% sodium citrate solution) and subjected to centrifugation at 1,700×g for 13 min at room temperature. The plasma samples were analyzed with a coagulation analyzer (STA Compact, Roche), and the following hematological parameters were evaluated: prothrombin time and activated partial thromboplastin time.
Serum samples were obtained from blood samples and collected into tubes containing a Gel and Clot activator (Venoject II, Terumo) for blood chemistry examination. The samples were subjected to centrifugation at 1,700×g for 7 min at room temperature and analyzed using an automatic analyzer (Hitachi 7,170, Hitachi).
The following parameters were evaluated: total protein, glucose, triglyceride, total cholesterol, blood urea nitrogen, creatinine, total bilirubin, aspartate aminotransferase, alanine aminotransferase, γ-glutamyl transpeptidase, calcium, and inorganic phosphorus. The samples were also analyzed using an electrolyte analyzer (EA06R, A&T) to examine concentrations of sodium, potassium, and chloride.
Urine samples were collected within 3 hr after urination (fresh urine) and 24 hr after urination (pooled urine) for urinalysis. The animals remained fed and were supplied with water during urine sample collections. Fresh urine was analyzed using Ames test strips (N-Multistix, SG-L, Siemens Healthcare Diagnostics) and an automatic strip reader (CLINITEK Advantus, Siemens Healthcare Diagnostics) for the following parameters: pH, occult blood, ketone bodies, glucose, protein, bilirubin, and urobilinogen. Following confirmation of volume and color of urine, the 24-hr urine samples were centrifuged at 400×g for 5 min. The supernatant was analyzed for electrolyte concentrations (sodium, potassium, and chloride) using an electrolyte analyzer (EA06R) and osmotic pressure using an osmotic pressure analyzer (AUTO&STAT ™ OM-6030, Arkray Factory). Urinary volume was also used to calculate the total excretion value of each electrolyte. Urinary sediments were stained by the Sternheimer method and examined microscopically.

| Bacterial reverse mutation test (Ames test)
The number of revertant colonies increased 2 times or more relative to the negative controls in strains TA100, TA1535, TA98, and TA1537 in the presence and absence of metabolic activation in the preliminary study and in the absence of metabolic activation in the concentration-finding study using the preincubation method (Table 1). The number of revertant colonies also increased 2 times or more in the concentration-finding study in strains In the main study conducted using the preincubation method, the number of revertant colonies increased 2 times or more in strain TA1537 at the highest concentration tested (12,850 µg TOS/plate) in the presence of metabolic activation relative to the negative control ( In the main and concentration-finding studies using the modified (treat-and-wash) method, the number of revertant colonies was comparable to that of the negative control in strains TA100, TA1535, TA98, and TA1537 in the presence and absence of metabolic activation (Table 3). No accelerated growth in background bacteria and bacterial growth inhibition was observed at any concentration in the assays (main and concentration-finding). As expected, the number of revertant colonies in the positive control   (Table 3).

| In vitro mammalian chromosomal aberration test
In the short-term assay in the absence of metabolic activation, the incidence of cells with structural chromosomal aberrations was 2.7%, 5.0%, and 4.7% at Sumizyme GT concentrations of 1,000, 3,000, and 4,000 µg TOS/ml, respectively, and a significant (p ≤ .025) concentration-dependent increase was observed (Table 4). The incidences at 3,000 and 4,000 µg TOS/ml were significantly higher compared to the negative control (0.7%). The incidence of polyploid cells treated with the test article was 0.3%, 0.7%, and 0.3% at Sumizyme GT concentrations of 1,000, 3,000 and 4,000 µg TOS/ml, respectively; however, statistical significance was not observed compared to the negative control group (0.3%). In the presence of metabolic activation, the incidence of cells with structural chromosome aberrations were 2.7%, 7.7%, and 11.7% at Sumizyme GT concentrations of 500, 1,000, and 3,000 µg TOS/ml, respectively, and a significant (p ≤ .025) concentrationdependent increase was observed (Table 4). The incidence even at 500 µg TOS/ml was significantly higher compared to the negative control (0.3%). The incidence of polyploid cells treated with the test substance was comparable to the negative control group (0.0%). The relative cell growth rates at the highest concentration in the presence (3,000 µg TOS/ml) and absence (4,000 µg TOS/ ml) of metabolic activation were 71.5% and 72.8%, respectively.
Microscopic examination was impossible at concentrations of 4,000 µg TOS/ml or more and 5,000 µg TOS/ml or more in the presence and absence of metabolic activation, respectively, due to an abundance of c-mitosis figures. The positive control group (MMC) showed a high incidence of cells with structural chromosomal aberrations in the absence and presence of metabolic activation (57.3% and 71.3%, respectively) and was significantly higher compared to the negative control group.
In the continuous assay, the incidence of cells with structural chromosomal aberrations were 7.3%, 13.0%, and 19.7% at Sumizyme GT concentrations of 500, 1,000, and 3,000 µg TOS/ml, respectively, and a significant (p ≤ .025)concentration-dependent increase was observed (Table 4). At Sumizyme GT concentrations of 500 µg TOS/ml or more, the incidence of aberrations was significantly higher than that of the negative control group (0.7%). The incidence of polyploid cells treated with the test substance was 0.7%, 0.3%, and 1.0% at Sumizyme GT concentrations of 500, 1,000, and 3,000 µg TOS/ ml, respectively; however, they were not statistically significantly different compared to the negative control group. The relative cell growth rate at the highest concentration (3,000 µg TOS/ml) was

| Clinical observations, body weights, food consumption, and ophthalmology
Throughout the administration period, no deaths or toxicological ef-

Frequency of hedgehogs (%) (mean ± SD)
Negative control (distilled water; 20 ml/kg) The mean % tail DNA in each group was calculated from that in each animal based on the median % tail DNA in each slide. b Inactivated enzyme.

TA B L E 6
Results of the confirmative in vivo mammalian alkaline comet assay conducted in male rats orally administered GT once daily for 2 days

| Clinical pathology and urinalysis
No treatment-related changes in hematology and ophthalmology were observed in any of the dose groups with the exception of a statistically significant decrease in reticulocyte ratio and count in the middose females compared to the control group (Table 7). A statistically significant increase in γ-globulin ratio and concentration in high-dose females was observed compared to the control group (Table 8). No other significant differences were observed in measured blood chemistry parameters. A significant increase was observed in protein-positive urine in mid-dose males and high-dose males and females (data not shown). Similarly, an increase in ketone body-positive urine in highdose animals was observed (data not shown).

| Pathology
No significant changes in relative and absolute organ weights were observed in any group (Table 9)

| D ISCUSS I ON
Sumizyme GT is a liquid concentrate enzyme preparation containing glutaminase from nongenetically modified A. niger. Glutaminase has significant commercial use in the production of food ingredients that improve the umami or savory flavor of foods. Traditionally, glutaminase is used in the production of soy sauce by converting l-glutamine that is formed from the hydrolysis of proteins from soybeans and wheat into l-glutamic acid, thus imparting a savory flavor (Ohshita et al., 2000). There exists a technological need for glutaminase by the food industry; without glutaminase, l-glutamine is chemically converted to pyroglutamic acid, which does not have any taste or flavor. The safety of glutaminase derived from nongenetically modified A. niger was evaluated in a series of genotoxicity and mutagenicity bioassays, including a bacterial reverse mutation test, mammalian chromosomal aberration test, and in vivo Comet tests, and a 90-day repeated-dose oral toxicity study in rats.  20.9 ± 5.9 14.5 ± 6.0 14.2 ± 2.9 12.9 ± 5.0 14.9 ± 6.5 Osmotic pressure ( While several slight, but statistically significant, positive findings were observed in the studies using the preincubation method (Tables 1 and 2), the enzyme preparation was demonstrated to be nonmutagenic in the bacterial reverse mutation assay at concentrations up to 12.85 mg TOS/plate when evaluated using the treat-and-wash procedure. The treat-and-wash method is considered to be a valid method to eliminate "false-positive" findings arising from free amino acids in the test substances, and has been successfully employed in the mutagenicity evaluation of other enzyme preparations and proteinaceous materials using the Ames assay (EFSA, 2014;Okado et al., 2015Okado et al., , 2016Okado et al., , 2019Thompson et al., 2005). The negative findings using the treat-and-wash method demonstrate that any positive or mutagenic effect observed in the studies using the preincubation method may be attributable to the presence of free amino acids, and likewise, the presence of these free amino acids in the test substance likely resulted in the increase in background lawn in the preincubation studies, a finding that was not observed in the treat-and-wash studies. Thus, the enzyme preparation was nonmutagenic under the conditions of the Ames test using the treat-and-wash method. In the chromosomal aberration assay, the statistically significant concentration-dependent increases in chromosomal aberrations observed in both the short-term test (with and without metabolic activation) and in the continuous 24-hr assay were unlikely to be attributed to a change in pH or osmolality of the test article, as the pH of the medium was comparable between the test concentrations and the negative control, and no precipitation of the test article was observed.
Considering that increasing incidence rates of chromosomal aberrations were generally observed at higher concentrations, it is possible that the clastogenic effect was due to disruption of cellular processes, as a result, cytotoxicity, at these high concentrations rather than an inherent toxicological effect of the enzyme (Galloway, 2000). However, significant increases in chromosome aberrations were noted even at concentrations at high relative growth rates (  (EFSA, 2009(EFSA, , 2011. The comet assay was selected as it is considered by EFSA to be a useful indicator to detect gene mutations and/or structural chromosomal aberrations and can be used with many target tissues (EFSA, 2011). Since Sumizyme GT, a proteinaceous/amino acid-containing material, is not likely to be systemically available due to proteolysis of the enzyme in the stomach and small intestine and would therefore not reach the bone marrow intact (hence a micronucleus test was not performed), the comet tests were performed with stomach and duodenal cells (i.e., target cells due to site of contact). These rapidly dividing cells were examined as they would address any potential site-of-contact genotoxicity concerns of the enzyme given the lack of systemic exposure to the enzyme (EFSA, 2011;OECD, 2016a). Adequate and direct exposure of the test article to target tissues/cells is an important consideration in the testing of chemicals for genotoxicity potential, and therefore, the comet tests using stomach and duodenal cells were considered acceptable to further evaluate the genotoxic potential of glutaminase. The significant increase in % tail DNA in the duodenum of rats observed in the initial alkaline comet assay at the highest tested dose (2,570 mg TOS/kg) was within the testing facility's historical control range (0.0 to 4.01%, n = 27) and was not reproducible in the second comet assay, and therefore, this finding was deemed to not be related to treatment with Sumizyme GT. Furthermore, no significant changes in hedgehog frequencies in stomach or duodenum cells were observed up to the highest dose that would be suggestive of a cytotoxic effect, and macroscopic examination of both stomach and duodenum did not reveal any toxic effect related to Sumizyme GT administration. Collectively, although the results of the in vitro mammalian chromosomal aberration test produced an unexplained positive effect, the results of the follow-up in vivo alkaline comet assays in stomach and duodenal cells, as well as the results of the bacterial reverse mutation test, support a lack of genotoxic and mutagenic potential of Sumizyme GT. This conclusion is supported by the stepwise, weight-of-evidence approach to genotoxicity testing as described by EFSA (2011).
In the 90-day repeated-dose oral toxicity study, the findings related to clinical observations were considered to be spontaneous and not related to Sumizyme GT administration as they were slight, were only observed in a single animal or dose, and were not dose-dependent.
A slight but significant decrease in mean daily food consumption in low-dose and an increase in mid-dose females only was observed in 1 sex. As these effects were reported in 1 sex only and not noted in the higher dose group, as such, they were not considered to be treatment-related.
Hematological findings were limited to significant decreases in reticulocyte ratio and count in mid-dose females and increases in γ-globulin ratio and concentration in high-dose females. As they were observed only in 1 test group and 1 sex and were not dose-dependent, these changes were considered to be unrelated to treatment with Sumizyme GT.
The findings from macroscopic and pathological examinations were considered to be spontaneous and incidental, as they were focally distributed, were observed only in a single animal, or were not dose-dependent. Likewise, no toxicological findings were reported in other subchronic toxicity studies conducted with glutaminase derived from Cryptococcus albidus (Ohshita et al., 2000) or Bacillus amyloliquefaciens (FSANZ, 2015b). Based on the absence of adverse effects of Sumizyme GT, the NOAEL is considered to be 29,560 U/kg bw per day, equivalent to 2,570 mg TOS/kg bw per day, the highest dose tested. The absence of any histological effects in the gastrointestinal tract and other organs, including the liver, within the 90-day study at the highest dose tested, as well as the comet assay, further supports the conclusion of a lack of genotoxicity potential.
The safety assessment of a food enzyme must consider the source organism in addition to the inherent hazard profile of the enzyme (Pariza & Johnson, 2001). The safety of A. niger for the production of food ingredients and enzymes is well established in the scientific literature and has been extensively reviewed by the U.S. FDA and the Joint FAO/WHO Expert Committee on Food Additives (JECFA) who have collectively raised no safety concerns (JECFA, 1990;US FDA, 2018). In 1989, JECFA established a single acceptable daily intake (ADI) of "not specified" for enzyme preparations of carbohydrases, amyloglucosidases, endo-1,3(4)-β-glucanase, hemicellulose, pectinases, and protease derived from A.
niger, citing that A. niger is a "common organism in food, that many strains have a long history of use as an enzyme source, and that numerous studies from various preparations from various strains have demonstrated no hazard to human health" (JECFA, 1990). The production strain used to produce Sumizyme GT was screened for the production of common secondary metabolites and mycotoxins, such as aflatoxin B1, B2, G1, and G2, sterigmatocystin, zearalenone, ochratoxin A, and T-2 toxin, and was demonstrated analytically to be free of these toxins (personal communication).
The results of the toxicological studies as described herein, together with the history of safe use of the production strain (A. niger) in food processing, support the safety of Sumizyme GT for use in food production for human consumption.

ACK N OWLED G EM ENTS
Intertek Health Sciences Inc. and AR Toxicology Inc. received financial support from Shin Nihon Chemical Co. to assist in the preparation of the manuscript.

CO N FLI C T O F I NTE R E S T
The authors of this publication have declared that they have no conflicts of interest.

E TH I C A L A PPROVA L
The 90-day study was conducted in compliance with the Act on