Effects of rumen-protected glutamate supplementation during the periparturient period on digestibility, inflammation, metabolic responses, and performance in dairy cows.

The objective of this study was to evaluate the effects of feeding rumen-protected glutamate during the periparturient period (d -21 ± 3 to d 21 ± 3 relative to calving) on apparent total-tract digestibility (ATTD), inflammation, metabolic responses, and production performance of dairy cows. Fifty-two multiparous Holstein cows were blocked by parity, body condition score, and expected calving date, and randomly assigned to one of the experimental diets with rumen-protected monosodium glutamate (RP-Glu; intestinally available Glu = 8.8%) or without RP-Glu (control) at d -21 ± 3 relative to expected calving date. The RP-Glu was fed at 4% and 3% of dietary dry matter, before and after calving, respectively. Prepartum diets contained 17.1% and 16.5% crude protein, and 13.1% and 13.3% starch, and postpartum diets contained 18.8% and 18.3% crude protein, and 22.5% and 22.7% starch on a dry matter basis, respectively for RP-Glu and control treatments. A subset of 19 cows was used to measure ATTD. Cows fed the RP-Glu had greater ATTD of dry matter (70.6 vs. 69.1%), crude protein (75.1 vs. 72.6%), and ether extract (66.0 vs 61.2%) on d 5 ± 1 after calving. Cows fed the RP-Glu also had greater dry matter intake (15.7 vs. 13.7 kg/d) on d 1 after calving. Cows fed the RP-Glu had greater plasma concentrations of Glu (4.60 vs. 3.89 µmol/dL) and insulin-like growth factor-1 (44.2 vs. 30.1 mg/mL), lower serum concentrations of free fatty acids (670 vs. 981 µEq/L) and total bilirubin (0.22 vs. 0.34 mg/dL), and lower plasma 3-methylhistidine concentration (1.28 vs. 1.50 µmol/dL) on d 4 after calving. However, these treatment effects observed between d 1 and d 5 ± 1 immediately after calving did not continue until d 21 after calving. Concentrations of serum amyloid A, serum haptoglobin, and plasma lipopolysaccharide binding protein were not affected by the treatment. In addition, no differences were observed for serum ß-hydroxybutyrate concentration and milk yield during the postpartum period between the 2 groups, and cows fed the RP-Glu had a decreased lactose yield. These findings suggest that feeding RP-Glu during the periparturient period can increase digestive capacity and feed intake, and decrease mobilization of body fat and protein immediately after calving without increasing milk production.

Advantame--an overview of the toxicity data.

Advantame is an N-substituted (aspartic acid portion) derivative of aspartame that is similar in structure to neotame, another N-substituted aspartame. An extensive series of studies, were conducted on advantame to define the pharmacokinetics and metabolism in various species, subchronic and chronic toxicity in the rat and dog, carcinogenicity in the rat and mouse, genotoxicity, reproductive, and developmental toxicity, and human tolerability studies. The results of these studies, presented in overview in the present publication, and in greater detail in the accompanying publications, show that advantame is well tolerated by both animals and humans and does not possess systemic toxicity. The metabolic data demonstrate that the animal species used in the toxicity testing are relevant to the evaluation of human safety. The no-observed-adverse-effect levels (NOAELs) identified in the animal studies in which advantame was administered in the diet were generally the highest doses tested. Under the anticipated conditions of use, the predicted intakes of advantame are about 20,000- to 70,000-fold lower than the identified animal study NOAEL values. The results of the animal toxicology and human trial data support the safety of use of advantame in food.

In vitro and in vivo assessment of the mutagenic activity of N-[N-[3-(3-hydroxy-4-methoxyphenyl) propyl]-α-aspartyl]-L-phenylalanine 1-methyl ester, monohydrate (advantame).

Advantame (N-[N-[3-(3-hydroxy-4-methoxyphenyl) propyl]-α-aspartyl]-L-phenylalanine 1-methyl ester, monohydrate), an N-substituted analogue of aspartame, has been developed as a high-intensity sweetener. It is approximately 100 and 20,000 times sweeter than aspartame and sucrose, respectively. In this study the safety of advantame has been evaluated using a series of in vitro and in vivo genotoxicity assays including, bacterial mutation, mammalian cell mutation, and mouse micronucleus tests. Advantame did not induce reverse mutations in Salmonella typhimurium and Escherichia coli at concentrations of up to 5000 µg/plate. In the mammalian cell mutation assay, advantame did not induce mutation at the Hprt locus of L5178Y mouse lymphoma cells in two independent experiments, either in the absence or presence of S9. In vivo, there was no effect on the incidence of micronucleated immature or mature erythrocytes in bone marrow after oral administration of the test substance at any dose level (up to 2000 mg/kg body weight) or sampling time (24 and 48 h). The results of these studies demonstrate that advantame is without genotoxic potential.

Chronic toxicity and carcinogenicity of N-[N-[3-(3-hydroxy-4-methoxyphenyl) propyl]-α-aspartyl]-L-phenylalanine 1-methyl ester, monohydrate (advantame) in the rat.

Groups of 55 male and 55 female Han Wistar rats were administered advantame (98.9-99.8% purity) in the diet at concentrations of 0, 2000, 10,000, or 50,000 ppm for 104 weeks, following parental exposure to the same levels from prior to mating and throughout gestation. Additional groups of 20 rats/sex and 10 rats/sex were dosed for a period of 52 weeks and constituted the toxicity and reversibility phases of the study. Achieved doses of advantame over the carcinogenicity study were 0, 97, 488, and 2621 mg/kg body weight/day in males and 0, 125, 630, and 3454 mg/kg body weight/day in females, respectively. A high incidence of a pale and swollen anus and changes in fecal composition were observed in the high-dose groups. There was no effect of treatment on mortality. Body weight gain in the high-dose males (50,000 ppm) was slightly reduced compared to controls after 52 and 104 weeks of treatment; the decrease was not considered to be of toxicological significance, but due to the non-nutritive nature of the high dietary concentration of advantame. During the toxicity phase, food conversion efficiency was slightly decreased in both sexes, at the 50,000 ppm dose level. Given the non-nutritive content of the diet, this finding was not considered biologically significant. There were no relationships between treatment and the results of hematological or urinalysis investigations. Clinical chemistry evaluations showed consistently lower plasma urea concentrations in both sexes treated at 50,000 ppm, which was reversed during the 6-week recovery phase following 52 weeks of treatment, indicating a lack of permanent effects. Terminal investigations at both the 52 and 104-week revealed a number of intergroup differences in absolute and/or relative organ weights; however, the differences did not show dose-response relationships, were minor in nature, and/or occurred only in one sex, and were not associated with any pathological findings, and they were considered not to be treatment-related. Evaluation of the histopathology of the carcinogenicity phase animals revealed an increased incidence of pancreatic islet cell carcinomas in males (incidence rates of 0/55, 1/55, 2/55, and 3/55 in the 0, 2000, 10,000, or 50,000 ppm groups, respectively) and of mammary gland adenomas in the high-dose females (incidence rates of 0 in the control through 10,000 ppm dose groups and 4/41 in the 50,000 ppm dose group). The incidence rates of these tumors did not attain statistical significance and/or remained within background historical control values, and they were considered to be unrelated to advantame treatment. The no-observed-adverse-effect level was considered to be 50,000 ppm in the diet, the highest concentration tested, equivalent to 2621 and 3454 mg/kg body weight/day in males and females, respectively. Advantame was concluded to be without carcinogenic activity.

Chronic oral toxicity of N-[N-[3-(3-hydroxy-4-methoxyphenyl) propyl]-α-aspartyl]-L-phenylalanine 1-methyl ester, monohydrate (advantame) in the dog.

Advantame (N-[N-[3-(3-hydroxy-4-methoxyphenyl) propyl]-α-aspartyl]-L-phenylalanine 1-methyl ester, monohydrate), an N-substituted analog of aspartame, has been developed as a high-intensity sweetener. Groups of 4 dogs of each sex were treated at 0, 2000, 10,000, or 50,000 ppm of advantame in the diet for 52 weeks. Additional groups of 2 dogs/sex at the control, and mid- and high-dose groups were treated for 52 weeks followed by a 6-week recovery period. There was no effect of treatment on mortality, body weight, organ weights, food consumption, or the results of ophthalmological, electrocardiographic, haematological, clinical chemistry or urinalysis examinations. No histopathological changes were associated with advantame treatment. The NOAEL was considered to be 50,000 ppm, the highest concentration tested, which was equivalent to 2057 and 2139 mg/kg body weight/day in males, and females, respectively. The results of the study support the safety of advantame for use as a high-intensity sweetener.

Evaluation of the teratogenic potential of N-[N-[3-(3-hydroxy-4-methoxyphenyl) propyl]-α-aspartyl]-L-phenylalanine 1-methyl ester, monohydrate (advantame) in the rat and rabbit.

To assess its teratogenic potential, advantame (N-[N-[3-(3-hydroxy-4-methoxyphenyl) propyl]-α-aspartyl]-L-phenylalanine 1-methyl ester, monohydrate) was administered to mated rats (22/group) in the diet at 0, 5000, 15,000, and 50,000 ppm (providing approximately 465, 1418, and 4828 mg/kg body weight/day), and to mated rabbits (24/group) via oral gavage at 0, 500, 1000, and 2000 mg/kg body weight/day throughout gestation. Shortly before delivery (rats: day 20; rabbits: day 29), animals were killed and subjected to a detailed necropsy. Fetuses were examined for external, visceral, and skeletal alterations. Atypical coloration of the feces and cage liners seen with test diets in both rats and rabbits was attributed to excretion of test material/metabolites in the feces and urine. Advantame had no adverse effect on rat offspring survival or development. The no-observed-adverse-effect level (NOAEL) for both maternal and developmental toxicity in rats was 50,000 ppm, the highest dietary concentration tested. Due to adverse effects associated with reduced food intake and fecal output, approximately 20% of mated rabbits receiving 200 0mg/kg body weight/day and 1 animal at 1000 mg/kg body weight/day had to be terminated before scheduled necropsy. A NOAEL of 500 mg/kg body weight/day was established for maternal toxicity in rabbits. No teratogenic effects were observed in any animals, and based on a slightly increased incidence of fetal deaths at 2000 mg/kg body weight/day, a finding that was considered to be indirectly related to advantame treatment, 1000 mg/kg body weight/day was considered the NOAEL for developmental toxicity.

A two-generation reproductive toxicity study of the high-intensity sweetener advantame in CD rats.

Rats received diets containing 0, 2000, 10,000, or 50,000 ppm advantame (N-[N-[3-(3-hydroxy-4-methoxyphenyl) propyl]-α-aspartyl]-L-phenylalanine 1-methyl ester, monohydrate) for 2 generations. F(0) animals (30/sex/group) were treated from 10 weeks before pairing. Males continued until week 16; females through gestation and lactation. Once weaned, F(1) animals (25/sex/group) continued receiving the same diet until F(2) pups were weaned. Mean advantame intakes from each of the diets were 164, 833, and 4410 mg/kg bw/day among F(0) males, and 204, 1036, and 5431 mg/kg bw/day among F(1) males. F(0) and F(1) females had comparable intakes up to lactation, when intakes increased (up to 8447 mg/kg bw/day from 50,000 ppm diet). No treatment-related effects on mortality, body weights, reproduction, litter observations, or postnatal offspring development were noted. Atypical coloration of the feces and cage liners seen with test diets was attributed to excretion of test material/metabolites in the feces and urine. Slightly higher food consumption was seen in F(0) and F(1) animals, especially males, receiving 50,000 ppm. However, these differences were considered to be a secondary response to the high levels of non-nutritive material in the diet. The no-observed-adverse-effect level for reproductive and developmental toxicity was considered to be 50,000 ppm, the highest dietary concentration tested.

Purification and properties of extracellular chitinases from the parasitic fungus Isaria japonica.

Two chitinases (P-1 and P-2) induced with colloidal chitin were purified from the culture supernatant of Isaria japonica by chromatography on DEAE Bio-Gel, chromatofocusing and gel filtration with Superdex 75 pg. The enzymes were electrophoretically homogeneous and estimated to have a molecular mass of 43,273 (+/-5) for P-1 and 31,134 (+/-6) for P-2 by MALDI-MS. The optimum pH and temperature was 3.5-4.0 and 50 degrees C for P-1 and 4.0-4.5 and 40 degrees C for P-2. P-1 acted against chitosan 7B (degree of deacetylation, 65-74%) = glycol chitin > colloidal chitin = chitosan 10B (degree of deacetylation, above 99%) and P-2 against chitosan 7B > glycol chitin = chitosan 10B > colloidal chitin in order of activity. The products of hydrolysis of chitin and chitosan hexamer were analyzed by MALDI-MS. The products from the chitin hexamer obtained with P-1 were almost all dimers with only a small amount of trimer whereas those obtained with P-2 were mainly trimers with some dimer and tetramer. No hydrolysis of chitosan hexamer was observed. High homology in the amino-terminal sequence for chitinase P-1 was exhibited by chitinases from Trichoderma harzianum, Candida albicans and Saccharomyces cerevisiae in the range of 48-39%. The highest homology for Chitinase P-2 was shown by an endochitinase from Metarhizium anisopliae of 66%, while 44% homology was exhibited by chitinases of Leguminosae plants.

Influence of abomasal infusion of high levels of lysine or methionine, or both, on ruminal fermentation, eating behavior, and performance of lactating dairy cows.

Four multiparous late-lactation Holstein cows were fed a basal ration designed to be co-limiting in intestinally absorbable supplies of methionine and lysine. Cows were supplemented with no amino acids, lysine by abomasal infusion to 140% of the calculated intestinally absorbable requirement, methionine by abomasal infusion to 140% of requirement, or both amino acids in a 4 x 4 Latin square design with 28-d periods. Unsupplemented cows consumed 23.8 kg/d of dry matter and produced 36.9 kg/d of milk containing 3.70% fat, 3.22% protein, and 4.82% lactose. Cows ate less dry matter and produced less milk and milk lactose, and tended (P = .06 or .08) to produce less milk protein when abomasally infused with methionine alone or together with lysine. Infusion of lysine alone resulted in production values numerically between those of unsupplemented cows and those cows supplemented with methionine alone or together with lysine. Evaluation of the results with two metabolic models of dairy cows indicated that performance of unsupplemented cows may have been limited by delivery of metabolizable or digestible protein, or intestinally absorbable lysine, isoleucine, or histidine, depending on the metabolic model used to evaluate animal performance. Regardless, results are consistent with those using nonruminant species, which have shown that imbalanced profiles of intestinally absorbable amino acids are associated with reduced dry matter intake and animal performance. Results also show that negative effects on performance of lactating dairy cows can occur if methionine is supplied at levels substantially in excess of calculated intestinally absorbable requirements, either alone or together with lysine.

Influence of postruminal supplementation of methionine and lysine, isoleucine, or all three amino acids on intake and chewing behavior, ruminal fermentation, and milk and milk component production.

Four multiparous Holstein cows were fed a basal diet balanced with the Cornell Net Protein and Carbohydrate System (CNCPS). Diets were formulated to be co-limiting in intestinally absorbable supplies of methionine, lysine, and isoleucine. Cows were supplemented with no amino acids (control); lysine and methionine in a ruminally protected form; isoleucine by abomasal infusion; or lysine, methionine, and isoleucine in a 4x4 Latin square arrangement of treatments with 28-d periods. Performance of cows on all treatments was lower than expected due to low intake of DM that could have been caused by the high fiber level of the basal diet. This high fiber level was likely responsible for the high daily chewing times for cows fed all diets, which was consistent with the high ruminal pH values. Intake of DM and its components were not influenced by any treatment. Milk protein percentage tended to be higher when cows were fed diets supplemented with ruminally protected lysine and methionine; however, production of milk, milk fat, and milk lactose were not affected by any treatment. Cows tended to have a higher milk lactose proportion and tended to produce more milk and milk lactose when they were abomasally infused with isoleucine alone. However, when cows were supplemented with all three amino acids, milk production and composition did not differ from that of cows fed the unsupplemented diet. Use of the CNCPS to evaluate the performance of the cows fed the unsupplemented diet suggested that these cows may have been colimited by intestinally absorbable supplies of lysine, isoleucine, and methionine in addition to metabolizable protein. Evaluation of the unsupplemented diet with an alternate model, Shield, suggested that cows fed the unsupplemented diet may have been colimited by intestinally absorbable supplies of lysine, isoleucine, and arginine. Results suggest that enhanced delivery of intestinally absorbable isoleucine may stimulate milk lactose synthesis.

Ruminally protected lysine or lysine and methionine for lactating dairy cows fed a ration designed to meet requirements for microbial and postruminal protein.

The purpose of this study was to separate the effects of ruminally protected Lys from effects of ruminally protected Met on the performance of lactating dairy cows fed a ration calculated to be first-limiting in intestinally delivered Lys and second-limiting in intestinally delivered Met. Thirty multiparous Holstein cows were examined in a 20-wk study that started on wk 5 postpartum. Rations contained timothy silage, corn silage, barley, corn, corn gluten meal, and soybean meal. Treatments were 1) no supplemental amino acids, 2) 21 g/d of intestinally available Lys, and 3) 22 g/d of intestinally available Lys and 6 g/d of intestinally available Met. Post-experimental calculations suggested that, in contrast to the objective, the unsupplemented ration was colimiting in intestinally available His (0.96 of requirement), followed by Lys (1.00), digestible ruminally undegraded protein (1.01), Ile (1.03), Arg (1.04), Val (1.10), and Met (1.14). In this context, the virtually identical performance of cows fed the unsupplemented ration and cows fed the ration supplemented with ruminally protected Lys demonstrated that dairy cows did not respond to enhanced intestinal supplies of Lys when Lys was not calculated to be the first-limiting nutrient. In contrast, for cows fed rations supplemented with both ruminally protected Lys and ruminally protected Met, the production of both milk protein (40 g/d) and fat (40 g/d) was numerically increased to an extent that was consistent with earlier reported studies, although calculations did not indicate that performance was limited by intestinal supplies of Lys or Met. This result, which may be disputed because of a lack of statistical significance, suggests that Met, apparently unlike Lys, may enhance the production of milk components beyond an enhancement expected because of its role as a limiting amino acid.

The effect of ruminal bypass lysine and methionine on milk yield and composition of lactating cows.

Fifty-six multiparous Holstein cows were assigned at 3 wk prepartum to rations based on grass silage with 1) corn distillers grains to provide 86 and 90% of estimated required metabolizable Lys and Met, respectively; 2) a blend of blood meal, fish meal, and meat and bone meal as amino acid (AA) sources to provide 112 and 103% of required metabolizable Lys and Met, respectively; 3) ruminally protected Lys and Met added as a top-dressing to ration 1 to provide 27 g/d of Lys and 8 g/d of Met as available AA at the duodenum postpartum; and 4) ruminally protected AA for 8 wk postpartum as a top-dressing to ration 1 to provide 40 g/d of Lys and 13 g/d of Met as available AA at the duodenum. Cows fed rations 3 and 4 were offered 13.5 g/d of duodenally available Lys and 4 g/d of duodenally available Met for 3 wk prepartum. The total length of the study was 43 wk. Cows fed ration 4 consumed 3 to 4 kg more dry matter than did cows fed the other three rations, and milk yield and the percentage of milk protein and fat were significantly increased during the first 8 wk of lactation. In early lactation, cows fed ration 3 had a greater milk fat percentage but similar dry matter intake, protein percentage, and yield of 4% fat-corrected milk compared with cows fed ration 2. The concentrations of blood serum glutamic oxaloacetic transaminase, serum glutamic pyruvic transaminase, triglyceride, and nonesterified fatty acids were lower for cows fed ration 4 during the first 8 wk of lactation than they were for cows fed the other three rations. The mammary arteriovenous difference of whole blood AA indicated that Met along with His and Arg may be the most limiting AA for milk yield.

H-aggregation of Methyl Orange at the Interface between the Water Phase and Oil Phase in a Water-in-Oil Microemulsion

Methyl orange dissolved in a water-in-oil microemulsion was examined by UV-VIS absorption spectroscopy and differential scanning calorimetry (DSC). Methyl orange showed two temperature-dependent peaks at 416 and 354 nm in the microemulsion. The former was attributed to a methyl orange monomer, and the latter to aggregates of the dye in the parallel orientation (H-aggregates) at the interface between the water phase and the oil phase in the microemulsion. The position of methyl orange in the microemulsion was verified by DSC analysis.

Ruminally protected lysine and methionine for lactating dairy cows fed a diet designed to meet requirements for microbial and postruminal protein.

Dairy cows, 20 at each of two sites, were used to determine responses to ruminally protected Lys and Met in a full lactation study. Cows were fed corn silage twice daily for ad libitum intake and a concentrate four times daily in proportion to milk production. At Truro, cows were fed 2.7 kg/d of alfalfa and timothy hay DM at 0600 and at 1500 h. At Fredericton, cows were fed 2.7 kg of timothy silage DM at 0600 h and 2.7 kg of alfalfa hay DM at 1500 h. Diets were designed to meet, but not to exceed, recommendations for ruminally degradable CP and intestinally digestible protein. Ten cows at each site were fed ruminally protected L-Lys.HCl (19 g/d) and DL-Met (6.5 g/d). Cows fed AA at each site produced more milk, lactose, protein, and fat; milk protein and fat percentages were also higher. No time x treatment interactions occurred for any production parameter. In spite of similar production responses between sites, cows fed AA consumed more DM at Truro, but those at Fredericton did not. Thus, gross efficiency of utilization of dietary N for milk N was increased with AA at Fredericton but not at Truro. However, considering the increased intake of CP by cows fed AA at Truro, an event that would have been expected to depress efficiency of utilization of dietary N, the lack of difference at Truro between treatments can be interpreted as an improvement, relative to expectations, because of AA feeding. High producing dairy cows fed a diet that was adequate in CP responded to ruminally protected Lys and Met primarily with increased production of milk protein and fat throughout the full lactation.

Monoclonal antibodies define antigenic regions on the major internal protein p24 of bovine leukemia virus (BLV).

The major core protein p24 of bovine leukemia virus (BLV) was characterized by ten monoclonal antibodies. Competitive binding assays were performed in order to analyze the topography of the antigenic determinants by enzyme-linked immunosorbent assay. At least three independent antigenic regions were demonstrated on the BLV p24 molecule.

Polymyositis associated with dissecting aneurysm of arteries and intracerebral hemorrhage.

Dissecting aneurysm and stroke are a rare complication of polymyositis. The present report describes an autopsy case of a 53-year-old woman, who suffered from polymyositis accompanied by dissecting aneurysms of right external iliac and right renal arteries, and furthermore, intracerebral hemorrhage. The latter was caused by necrotizing angiitis. The patient had no abnormal anatomical changes such as coarctation of aorta, aortic stenosis, bicuspid aortic valve or Marfan's syndrome. Atherosclerosis in the patient was mild, and cystic medial necrosis of aorta and arteries was not found. Since the patient was attacked by dissections of arteries following long-term steroid therapy, a possibility can be raised that arterial dissection was attributable to steroid treatment besides necrotizing angiitis complicating in polymyositis.

Sulfate esters of hydroxymethyl-methyl-benz[a]anthracenes as active metabolites of 7,12-dimethylbenz[a]anthracene.

7-Hydroxymethyl-12-methylbenz[a]anthracene (7-HMBA) and 12-hydroxymethyl-7-methylbenz[a]anthracene (12-HMBA), carcinogenic major metabolites of 7,12-dimethylbenz[a]anthracene (DMBA) in untreated rat liver, showed high mutagenic activities toward Salmonella typhimurium TA 98 after preincubation with a sulfotransferase-PAPS system consisting of ATP, sodium sulfate, and a post-mitochondrial fraction (S-9) or a soluble supernatant fraction (S-105) from untreated rat liver. The 7- and 12-HMBAs themselves induced His+ mutation in TA 98 only slightly after preincubation with S-9 in the presence of an NADPH-generating system. Mutagenicity of DMBA toward TA 98 after preincubation with S-9 in the presence of the NADPH-generating system was remarkably enhanced by the addition of ATP and sodium sulfate. The active metabolites, 7-HMBA sulfate and 12-HMBA sulfate, were isolated from these preincubation systems and identified by comparison with the corresponding synthetic specimens. The sulfuric acid ester conjugates were potent mutagens toward TA 98 in the absence of rat liver subcellular fractions. The conjugates bound covalently at significant rates to calf-thymus DNA as well as to S-105 proteins at 37 degrees and pH 7.4 through the 7- or 12-methylene carbon with concomitant loss of their sulfate group. In the presence of S-105, glutathione inhibited the mutagenicity of the metabolically formed or exogenously added 7- and 12-HMBA sulfates. The non-mutagenic glutathione conjugates were isolated from the incubation mixtures and identified as S-(12-methylbenz[a]anthracen-7-yl)methylglutathione from 7-HMBA or its sulfate and S-(7-methylbenz[a]anthracen-12-yl)methylglutathione from 12-HMBA or its sulfate.