Thirteen-Week Oral Toxicity Study of l -Glutamine in Rats

Abstract

l-Glutamine (Gln) is a semiessential amino acid used in enteral feeding in critically ill patients, and is contained in numerous dietary supplements available to the general public. This study evaluated toxicological effects of Gln in male and female Sprague-Dawley rats. Gln produced by Ajinomoto Co. (Tokyo, Japan) was incorporated into a standard diet at doses equal to 1.25%, 2.5%, and 5.0% (w/w), respectivelly. A control group of rats received only a standard diet. All diets were administered ad libitum for 13 consecutive weeks. To examine recoverability of any potential effects, the administration period was followed by a 5-week recovery period, during which only the standard diet was provided to all animals. Throughout the administration and recovery periods, no deaths were observed, and no changes in diet consumption, ophthalmologic findings, gross pathology, and histopathology were detected. Several changes in urine parameters (total protein, urine pH, and a positive incidence (±) of ketone bodies) were observed in the 2.5% and 5.0% groups at the end of the administration period. Minor increases were found in hematology parameters for the 5.0% group (platelet count, γ-globulin, lactate dehydrogenase [LDH]), but all changes were within physiological range. No effects of administration were observed in the 1.25% group. The no-observed-adverse-effect level (NOAEL) for Gln was estimated at 1.25% for both genders (males 0.83 ± 0.01 g/kg/day; females, 0.96 ± 0.06 g/kg/day).

l-Glutamine (Gln) is not usually described as an essential amino acid; however, mammals are unable to synthesize Gln sufficiently during catabolic stress (Lacey and Wilmore 1990). Consequently, enteral administration of Gln is a potential therapeutic agent in critically ill patients (Zhou et al. 2003). Gln is especially important for the rapidly multiplying cells of the intestinal tract, because the cells of both mucosal and lymphoid tissue use it as a fuel to maintain the integrity of the intestinal flora (Okabe et al. 1976; Duggan, Gannon, and Walker 2002). Similarly to severe catabolic stress (i.e., accidental injury or major surgery; Savarese et al. 2003), exhaustive exercise can also reduce plasma concentrations of Gln (Antonio and Street 1999), and the high incidence of illness in professional endurance athletes is supposedly linked also to this decrease (Castell 2003). In marathon runners, plasma levels of Gln, the branched-chain amino acids (l-leucine, l-isoleucine, and l-valine), and T lymphocytes profoundly decreased following a run (Castell et al. 1997). Runners receiving Gln supplements reported significantly lower incidences of cold than those in the placebo group (Castell, Poortmans, and Newsholme 1996). Dietary supplements containing Gln were also recommended for professional athletes in non-endurance sports, such as power lifting (Antonio et al. 2003).

In view of the already existing and potential therapeutic indications for Gln preparations, and their regular use without medical supervision (Garlick 2001), it was deemed prudent to evaluate the subchronic toxicity of Gln administered via dietary incorporation in rats. Furthermore, very limited data describing Gln toxicity in animals and humans (Ziegler et al. 1990; Hornsby-Lewis et al. 1994; Jian et al. 1999) were available. Although no adverse effects have been reported to date, a review of the available literature indicated that chronic toxicity had not been sufficiently evaluated (Garlick 2001).

MATERIALS AND METHODS

Animals and Feeding Protocols

The study was conducted in compliance with the Good Laboratory Practice Standards for Safety Studies on Drugs (Notification N. 313, March 31, 1982) and Guidelines for Toxicity Studies Required for Applications for Approval to Manufacture (Import) Drugs (Ordinance N.1, Article N.24, September 11, 1989). Seventy-five 4-week-old male and female Sprague-Dawley rats (Charles River Japan, Tokyo, Japan) were housed individually in conventional stainless-steel hanging cages (Lead Engineering, Tokyo, Japan), and provided with a standard diet (CRF-1 Oriental Yeast, Tokyo, Japan) and tap water ad libitum in an animal room with controlled temperature (22°C ± 2°C), humidity (55% ± 10%), and illumination (12 h illumination per day, from 7:00 am to 7:00 pm ). Following the 2-week acclimation period, healthy animals of both genders showing a normal increase in body weight were selected at 6 weeks of age and used throughout the study. The selected rats were divided by a computer block-placement method into four separate groups (12 rats per group) based on their weight, to ensure the homogeneity of the group mean body weight. The control group of 12 rats was continuously provided with a standard CRF-1 diet (Oriental Yeast) and water ad libitum, whereas the remaining three groups received a diet (ad libitum) supplemented with Gln (Lot no. 201AA06), which was provided by Ajinomoto (Tokyo, Japan). The purity of the supplemented Gln, which was stable in the conditions of this experiment, was 99.6%. Gln was supplied at three doses (1.25%, 2.5%, and 5.00% w/w). All diets were administered for 13 weeks. Following this period, six rats from both control and 5.0% concentration groups were randomly selected for the recovery study. During the 5-week recovery study, only a standard diet was given to the selected rats.

All rats were observed twice daily (morning and afternoon) during the 13-week testing period, and once daily (morning) during the 5-week recovery period. The rats were weighed twice each week in the morning (between 9:00 am and 12:00 noon). Twenty-four-hour diet intake was measured twice each week, and water intake was measured once each week. Estimated intake of the amino acid (mg/kg/day) for each week was calculated from the mean consumption for 1 day for that week.

Ophthalmologic Examination

Ophthalmologic examination was done prior to the start of administration. The ophthalmologic examination was repeated during week 13 of administration using six randomly selected rats from each group, and during week 5 of recovery for all rats on study.

Urinalysis

Urinalysis was conducted in all rats during weeks 5 and 13 of the administration period, and during week 5 of the recovery period. Animals were placed in metabolic cages and initially provided with water, and fasted. Urine samples were collected for 4 h. After the 4-h fasted period rats was complete and urine samples collected, food was returned to the rats’ cages. Urine was collected for 20 h; food and water were available ad libitum during the 20-h period. From the 4-hour samples, the following data were collected: pH, protein, ketone body, glucose, occult blood, bilirubin, urobilinogen (all parameters measured by Uriflet 7A; Kyoto Daiichi Kagaku, Kyoto, Japan), urine color, and sedimentation (microscopic examination). From the 24-h samples, the following parameters were measured: volume of urine (volumetry), specific gravity (refractometry), and electrolyte concentration.

Hematology and Blood Chemistry

Hematological evaluations were conducted using blood samples collected on the day following the final administration during week 13, and again at the end of the recovery period from fasted rats. Blood samples were collected from the abdominal aorta by laparotomy under ether anesthesia into blood collecting tubes (SB-41; Toa Medical Electronics, Tokyo, Japan) containing an anticoagulant (EDTA-2K). The following parameters were measured, red blood cell count (RBC; electronic counting method using Coulter 8 Item Automatic Blood Cell Analyzer T890; Japan Scientific Instrument, Tokyo, Japan), mean corpuscular volume (MCV; electronic counting method using automatic coagulometer ACL100), hemoglobin (Hb; cyanmethemoglobin method using automatic coagulometer ACL100) to reticulocyte ratio (Brecher method), platelet and white blood cell counts (electronic counting method using automatic coagulometer ACL100), differential leukocyte count (microscopic method using May-Giemsa staining), prothrombine and activated partial thromboplastin times (PT and APTT; clot method using automatic analyzer Monarch), and fibrinogen (thromboplastin method, using automatic analyzer Monarch). Hematocrit and mean corpuscular hemoglobin were calculated from the above-measured parameters. GOT, GPT, and lactate dehydrogenase (LDH) were obtained from blood samples collected from the abdominal aorta into tubes containing heparin. The plasma was obtained by centrifugation (3000 rpm, 10 min, 4.0°C). The sera parameters (total cholesterol, triglycerides, phospholipids, total bilirubin, blood glucose, urea nitrogen, creatine, uric acid, sodium, potassium, chloride, calcium, inorganic phosphorus, and total protein) were obtained from blood samples allowed to stand for approximately 1 h, and thereafter separated by centrifugation (3000 rpm, 10 min, 4.0°C).

Myelograms

At the autopsy at the end of the administration period and the recovery period, femoral bone marrow samples were collected from all animals and stained specimens by May-Giemsa staining were prepared and examined microscopically.

Pathology and Histopathology

Femoral bone marrow samples were collected at necropsy from all rats and May-Giemsa–stained specimen were prepared and examined microscopically. The rats were euthanized by exsanguinations via the abdominal aorta and examined grossly for any external abnormalities. Then, the organs and tissues in the cranial, thoracic, and abdominal cavities were examined grossly. The brain, pituitary, salivary, and thyroid glands, heart, lungs (including bronchia), liver, spleen, kidneys, adrenals, testes, prostate, ovaries, and uterus were excised and weighted. All the organs listed above, plus spinal cord, sciatic nerve, aorta, trachea, tongue, esophagus, stomach, duodenum, jejunum, ileum, cecum, colon, rectum, pancreas, thymus, mesenteric lymph nodes, cervical lymph nodes, epididymides, seminal vesicles, vagina, mammary glands, skin, eyes, optic nerve, Harderian glands, sternum (bone marrow), femur (bone marrow), femoral muscle, and gross lesions were excised and fixed in phosphate-buffered formalin. After paraffin embedding, the excised organs and tissues were prepared for microscopic examination by sectioning and staining with hematoxylin and eosin.

Statistical Analysis

Data were analyzed for homogeneity of variances using Bartlett’s test. Homogenous data were analyzed using the parametric one-way analysis of variance (ANOVA), and group differences were assessed using Scheffe’s test. Heterogeneous data converted to rank-sum were analyzed using the nonparametric test of Kruskal-Wallis and the significance of differences between the control group and each amino acid administered group was assessed using the method of distribution free multiple comparison (Gad and Weil 1982). Means ± standard deviations (SD) are shown.

RESULTS

No deaths and no clinical signs related to ingestion of Gln were observed. Excoriation of the dorsal neck and/or scapular skin was seen in one male from the 5.0% group, but this observation was considered incidental and unrelated to treatment because it was transient and no other animals on study were similarly affected.

During both the administration and recovery periods, body weights for male and females were comparable across the control and treated groups (Table 1). At 1.25% and 2.5%, food consumption was occasionally different than that observed for controls during the administration period. These differences were not considered related to treatment because, the effect was not dose dependent, and the overall diet consumption was comparable among the groups (data not shown). During the recovery period for the 5.0% group, a single transient decrease in food consumption was found on day 17, which was not considered related to treatment (data not shown).

All of the above observations were slight, transient, dose-independent phenomena with no apparent impact on the overall study values. The average intake of Gln over the course of the study is shown in Table 2. No significant administration-related changes in water intake were recorded for male rats. During the recovery period, the water intake for male rats from the 5.0% group was slightly higher than the control group, but the differences did not reach a significant level. No opthalmological abnormalities were observed.

Urinalysis

For females from the 2.5% and 5.0% groups, a tendency toward an increase in the number of positive incidence (+) for urine protein was observed (controls, 0 rats; 2.5% and 5.0% groups, 5 rats) was observed in week 5. No other group differences were found (data not shown).

For females in the 2.5% and 5.0% groups at the end of the administration period (week 13), a tendency toward a decrease in urine pH and increased incidences of positive level (+) of urine protein (controls, 1 rat; 5.0% group, 4 rats) were observed. These changes were not seen at the end of the recovery period. For females from the 5.0% group at the end of the recovery, significant decreases in the total potassium (controls, 2.98 ± 0.53 mEq/24 h; 5.0% group, 2.23 ± 0.51 mEq/24 h) and chloride (controls, 1.79 ± 0.24 mEq/24 h; 5.0% group, 1.29 ± 0.35 mEq/24 h) were noted.

Hematology and Blood Chemistry

At the end of the administration period, an increase in platelet count was observed for females from the 5.0% group (controls, 105.8 ± 9.6 10⁴/mm³; 5.0% group, 117.8 ± 10.7 10⁴/mm³). A significant increase in the LDH activity was observed in males in the 5.0% group (controls, 23.4 ± 6.1 U/L; 5.0% group, 33.6 ± 10.8 U/L), and a significant increase in γ-globulin fractions was seen in females in the 2.5% and 5.0% groups (controls, 6.1% ± 0.9%; 2.5% group, 7.6% ± 1.3%; 5.0% group, 7.7% ± 0.9%).

The changes seen at the end of the administration period were not observed at the end of the recovery period. However, a significant decrease in triglycerides was seen in males in the 5.0% group (controls, 133.5 ± 28.1 mg/dl; 5.0% group, 77.7 ± 21.1 mg/dl), and a significant increase in the A/G ratio (controls, 1.03 ± 0.04; 5.0% group, 1.11 ± 0.06), albumin (controls, 50.7% ± 1.1%; 5.0% group, 52.5% ± 1.3%), as well as a significant decrease in α₁-globulin fractions (controls, 20.3% ± 1.2%; 5.0% group, 16.5% ± 1.6%) were observed in females in the 5.0% concentration group. Finally, as ether was used for anesthesia, it is possible that ether might have interfered partly with some of the Gln effects on blood parameters.

Pathology and Histopathology

There were no significant treatment-related pathologies; minor changes were few and dose independent. Absolute weights of all organs at the end of the administration period are presented in Tables 3A and 3B. A significant increase in the relative weight of the lungs was found in the females in the 5.0% concentration group (controls, 0.37% ± 0.02%; 5.0% group, 0.40% ± 0.02%), but this increase was considered related to slightly lower body weights at the time of necropsy.

At the end of the recovery period, a significant decrease in the absolute weight of the brain in the males in the 5.0% concentration group (controls, 2.29 ± 0.06 g; 5.0% group, 2.15 ± 0.03 g) was observed, but this was considered to be a change caused by the slightly lower body weight at the time of autopsy, when compared to that measured in controls, as the relative weights of brains were comparable between the two groups (controls, 0.38 ± 0.04 g/%; 5.0% group, 0.39 ± 0.06 g/%).

Histopathological findings at the end of the administration and recovery periods were incidental or spontaneous, and for the sake of clarity are not listed here.

DISCUSSION

The current subchronic toxicity study of Gln in Sprague-Dawley rats was conducted to examine potential toxicity and potential subsequent reversibility of any effects. Gln was administered in the diet for 13 weeks, followed by a 5-week recovery period during which the animals were not exposed to Gln. Gln was mixed into the diet at concentrations of 0%, 1.25%, 2.5%, 5.0% (w/w) and given ad libitum. Throughout the administration and recovery periods, no deaths were observed and no changes in diet consumption, ophthalmologic findings, gross pathology, and histopathology were found. A slight tendency toward a decreased body weight was observed during the administration period in males in the 5.0% group; however, this minor effect was reversed during the 5 weeks of the recovery, and no changes in total body weight gain were found. The absence of gross adverse effects agrees with previous studies in humans (Ziegler et al. 1990; Jian et al. 1999), whereby toxicity at Gln doses as high as 50 to 60 g/day (Garlick 2001) was not reported. However, the chronic ingestion in healthy humans was not yet taken into account in human safety studies.

A tendency toward an increase in the number of positive incidences for urine protein was observed in females in the 2.5% and 5.0% groups; however, total protein in the blood did not increase, and no pathological changes in the tissue of the kidneys or the urinary tracts were observed. Gln triggered slight decline in a urine pH, suggesting a slight acidifying effect within the urinary tract. An increase in the number of positive incidence of ketone bodies was also observed in the 2.5% and 5.0% groups. Nevertheless, ketosis was not found. At the end of the fifth week of the administration period, a drop was measured in the level of total potassium and chloride excretion; yet as this change was not seen at the end of the administration period (week 13), it was considered to be unrelated to Gln administration.

A significant increase in the platelet count was found in females in the 5.0% group at the end of the administration period, but the individual values were within physiological ranges and were regarded as toxicologically irrelevant. A small enhancement of γ-globulin was seen in females in the 2.5% and 5.0% groups at the end of the administration period, but no variations in total protein, A/G ratio or lymphocyte ratio accompanied this finding, which disappeared following 5 weeks of recovery. An increase in LDH was observed in males in the 5.0% group at the end of the administration period, but the individual values were within the normal range recorded in the experimental facility.

Several alterations in hematology and blood chemistry were uncovered at the end of the recovery period (triglycerides, A/G ratio, albumin fractions, and α₁-globulin fractions), but not at the end of the administration period, therefore these alterations were considered accidental. Similarly, a minor increase in prostate weight at the end of the recovery was not accompanied by pathological or histopathological findings, and was not seen at the end of the administration period; hence it was judged as a Gln-independent change.

As summarized in the above sections, infrequent changes were witnessed in the urinalysis and blood chemistry in the 2.5% and 5.0% (w/w) groups. All of those changes were determined toxicologically insignificant. No effects of the administration were observed in the 1.25% (w/w) group. Therefore, the definitive toxic level for Gln was determined to be above 5.0% w/w), and the no-observed-adverse-effect level (NOAEL) was estimated at 1.25% (w/w) for both genders (males 0.83 ± 0.01 g/kg/day; females, 0.96 ± 0.06 g/kg/day).