POLYETHYLEMINE AND ETHYLENIMINE EXPLANATION Polyethylenimine is an immobilizing agent used in the production of enzyme preparations for food processing. The substance is also used in food packaging materials. Polyethylenime is produced by the acid- catalyzed homopolymerization of ethylenimine. The polymerized material is cross-linked with ethylene dichloride to give the 40,000 to 60,000 molecular-weight substance utilized in enzyme immobilization. POLYETHYLEMINE EXPLANATION This substance has not been previously evaluated by the Joint FAO/WHO Expert Committee on Food Additives. BIOLOGICAL DATA Biochemical aspects No information available. Toxicological studies Special studies on mutagenicity Two different forms of polyethylenimine were evaluated for mutagenic activity in the presence or absence of an Archlor-induced rat liver activation system. The indicator organisms used were Salmonella typhimurium strains TA1535, TA1537, TA1538, TA98, and TA100. A sample of polyethylenimine obtained prior to cross-linking with ethylene dichloride was mutagenic to strains TA1535 and TA100 both in the presence and absence of the activation system. The sample contained measurable amounts of ethylenimine which likely caused the mutagenic activity (Mortelmans & Shepherd, 1980). An ethylene dichloride cross-linked sample, the form utilized in enzyme immobilization (Corcat P-600), did not contain measurable amounts of ethylenimine and was not mutagenic in any of the tester strains in the presence or absence of the activation system. At concentrations of up to 5000 micrograms/plate, 2 other forms of polyethylenimine, Corcat P-12 and Corcat P-18, were not mutagenic under the same test conditions (Mortelmans & Miron, 1981). Polyethylenimine (P-1000) having a molecular weight of 70,000 was tested for mutagenicity with or without metabolic activation (source of the activating system not specified) using Salmonella typhimurium strains TA1535, TA1537, TA1538, TA98, and TA100 and E. coli strain wp2 uvrA. No mutagenic activity was found when concentrations of up to 5000 micrograms/plate were tested (Kajiwara et al., 1984). Acute toxicity LD50 Species Route (mg/kg b.w.) Reference Mouse oral 2,800 BASF, 1959 Mouse oral 8,000 Kobe University, 1974 Mouse i.p. 40 BASF, 1959 Rat oral (gavage Rushbrook & in corn oil) Jorgenson, 1981 Male 7550(a) Female 7500(a) Both sexes oral 2286(b) Male 2286(c) Female 1991(c) Rat gavage (aqueous 2000 Norris, 1973 solution) Rat oral 3,000 BASF, 1959 Rat i.p. 70 mg/kg BASF, 1959 Rabbit oral 2,000 (no BASF, 1959 injury) Rabbit i.v. 4 (lethal BASF, 1959 dose) Cat i.v. 10 (lethal BASF, 1959 dose) (a) Polyethylenimine type Corcat 600 (type used in enzyme immobilization) (b) Polyethylenimine type Corcat P-12 (c) Polyethylenimine type Corcat P-18 It was not possible to determine the lethal oral dose of polyethylenimine in dogs and cats because of vomiting. Cats vomited within 30 minutes of receiving 100-500 mg/kg, so only a small amount of the material remained in the stomach. Doses of 100 mg/kg caused vomiting in dogs. In cases where vomiting did not occur, no adverse effects were reported (BASF, 1959). In guinea-pigs, a single oral dose of 600 mg/kg was lethal to 6/10 animals and a dose of 800 mg/kg was lethal to 2/10 animals (BASF, 1959). Short-term studies Rats Groups of 40 male and 40 female white rats received 0, 0.25, 0.5, or 1.0 g/kg b.w. of polyethylenimine in the diet for 8.5 months. No compound-related effects were reported with respect to mortality, body weight, feed consumption, clinical signs, haematology, urinalysis, or absolute and relative organ weights. An elevation in blood alkaline phosphatase levels was observed in high-dose male rats as compared to the controls, but the values were within normal limits for the performing laboratory. No treatment-related changes were observed in gross and microscopic pathology studies carried out on 5 animals/sex/ dose (McCollister & Copeland, 1968a). Dogs Groups of 4 male and 4 female beagle dogs were fed 0, 0.25, 0.5, or 1.0 g/kg b.w. polyethylenimine in the diet for 9 months. No compound-related differences were noted with respect to haematology, urinalysis, clinical chemistry, or bromsulfophthalein dye retention. Group mean body-weight gains were reduced as compared to controls in high-dose males and females. Relative liver and kidney weights were increased in high-dose females and relative kidney weights were slightly increased in high- and mid-dose males and mid-dose females. The pathology report indicated that severe degenerative changes occurred in the kidney proximal convoluted tubules in the high-dose animals. The same lesion, but not as severe, was reported to occur in all mid-dose and 5 of 8 low-dose animals. No lesions of the proximal convoluted tubules were reported in control animals. All of the low- dose females and one of the low-dose males were reported to have the lesion to a "very slight" degree. Brown pigmentation of Kupffer's cells in the liver was considered to be compound-related. In the high- dose animals, it was present in 3 of 4 males at a moderate to very slight degree and in 2 of 4 females at a marked to moderate extent. In the mid-dose groups, 1 of 4 animals of each sex had the lesion present at a grade of "very slight". The lesion was not present in controls of either sex or low-dose males, but was present at a "very slight" grade in 1 of 4 low-dose females. There did not appear to be any other compound-related microscopic lesions (McCollister & Copeland, 1968b). Rabbits Rabbits were reported to tolerate several once-per-week oral doses of 1.0 g/kg polyethylenimine for several weeks without impairment of liver function. No injury to the liver or kidney was noted in rabbits receiving 1.0 g/kg daily, but the report stated that the rabbits tolerated a maximum of 6 doses. Daily doses of 0.5 g/kg were reportedly tolerated without any injury; however, details of the study were not provided (BASF, 1959). Comments Although not up to modern standards of toxicity testing, the 8.5-month rat study of free polyethylenimine did not reveal any adverse effects. In the 9-month dog study, compound-related changes in the kidney and liver were found. Since absorption and distribution studies are not available to show if appreciable uptake of this high- molecular-weight compound occurs, it is not clear what the mechanism of action is for the liver and kidney lesions. Polyethylenimine free of measurable levels of ethylenimine showed no mutagenic activity when tested with or without metabolic activation using Salmonella typhimurium strains. EVALUATION Estimate of acceptable daily intake for man Polyethylenimine is considered to be a suitable substance for use as an immobilizing agent in the production of immobilized enzymes (see ethylenimine). REFERENCES BASF (1959). Preliminary report on the toxicity of PEI. Unpublished report of Badische Analin & Soda Fabrik AG. Submitted to WHO by UOP, Inc. Kajiwara, Y., Oguru, S., & Takeyasu, K. (1984). Ames metabolic activation test to assess the potential mutagenic effect of polyethylenimine. Unpublished report of the Hita Research Laboratories, Chemicals Inspection & Testing Institute. Submitted to WHO by UOP, Inc. Kobe University (1974). Acute toxicity of P-1000. Unpublished report of Kobe University, Medical Faculty, Public Health Section. Submitted to WHO by UOP, Inc. McCollister, D.D. & Copeland, J.R. (1968a). Results of 8.5 month dietary feeding studies of polyethylenimine in rats. Unpublished report of the Dow Chemical Company. Submitted to WHO by Cordova Chemical Company. McCollister, D.D. & Copeland, J.R. (1968b). Results of 9 month dietary feeding studies of polyethylenimine in Beagle hounds. Unpublished report of the Dow Chemical Company. Submitted to WHO by Cordova Chemical Company. Mortelmans, K.E. & Miron, K.L. (1981). In vitro microbiological mutagenicity assays of Cordova Chemical Company's compounds Corcat P-12, Corcat P-18, and Corcat P-600. Unpublished report of SRI International. Submitted to WHO by Cordova Chemical Company. Mortelmans, K.E. & Shepherd, G.F. (1980). In vitro microbiological mutagenicity assays of Cordova Chemical Company's compound PEI prepolymer, Sample No. SWM D32-084-1. Unpublished report of SRI International. Submitted to WHO by Cordova Chemical Company. Norris, J.M. (1973). Acute toxicological properties of PEI-600. Unpublished report of the Dow Chemical Company. Submitted to WHO by UOP, Inc. Rushbrook, C.J. & Jorgenson, T.A. (1981). Acute toxicity studies of three Corcat compounds. Unpublished report of SRI International. Submitted to WHO by Cordova Chemical Company. ETHYLENIMINE EXPLANATION Ethylenimine has not been previously evaluated by the Joint FAO/WHO Expert Committee on Food Additives. Trace amounts of ethylenimine may potentially migrate into food from the presence of small amounts of unreacted monomer present as a contaminant in polyethylenimine. BIOLOGICAL DATA Biochemical aspects Distribution Five male rats (Dow-Wistar strain) were injected i.p. with 0.3 to 0.4 mg/kg b.w. of 14C-labelled ethylenimine and sacrificed after 24 or 96 hours. In both cases about 50% of the radioactivity was excreted in the urine and small amounts were present in faeces and exhaled air. A small amount of ethylenimine and a number of non-volatile metabolites were present in the urine and both ethylenimine and CO2 were present in expired air. About 2.5% of the radioactivity was present in the liver after 24 hours and about 1% after 96 hours. Smaller amounts were present in many other tissues. The authors concluded that the compound was generally distributed throughout the rat before reaction with tissue components occurred (Wright & Rowe, 1967). Excretion Monoethanolamine and ethylenimine were excreted in the urine of rats following introduction of ethylenimine into the stomach (presumably by gavage). The urinary excretion of these 2 compounds accounted for approximately 50% of the administered dose. The fraction of the administered dose that was excreted in the urine was dose- dependent. At a dose of 1/20 the LD50 (0.85 mg/kg according to the authors) about 60% of the administered dose was excreted as monoethanolamine and ethylenimine in the urine over a 6-day period; 50% of the dose was excreted in the first 24 hours (Sanotsky et al., 1977). Toxicological studies Special studies on renal toxicity Rats Groups of 6 or 7 female Sprague-Dawley rats were given single subcutaneous injections of 0.25, 0.5, 1.0, 1.25, 2.0, 4.0, 6.25, or 8.0 mg/kg of ethylenimine in water. Animals were sacrificed 4 days after treatment. Renal papillary necrosis was observed at doses of 1.25 mg/kg ethylenimine and greater, whereas none occurred at 1.0 mg/kg and below (Axelson, 1978). Rabbits Renal medullary necrosis occurred in male and female New Zealand strain rabbits injected i.v. with a single 0.005 ml/kg dose (about 5 mg/kg) ethylenimine as a 1% v/v solution in water (Davies 1969; Davies, 1970). Dogs Groups of 4 male beagle dogs were given either a single i.v. injection of 3 µl/kg (about 3 mg/kg) ethylenimine or an initial injection of 0.6 µl/kg ethylenimine followed by a second injection of 1 µl/kg 3 days later. In the animals receiving 2 injections of ethylenimine the tubules of the collecting ducts and loop of Henle were dilated and inflamation was found in the renal pelvis. A small number of dilated tubules were observed in the medulla, and dilation of the distal convoluted tubules and collecting ducts of the cortex were noted. Almost complete necrosis of the papilla was observed in 2 dogs that became moribund after receiving the single high dose (3 µl/kg) of ethylenimine. Functional and clinical chemistry studies showed that treatment was related to impairment of renal function, proteinuria, and elevated excretion of urinary enzymes (Ellis et al., 1973). Special studies on carcinogenicity Mice A carcinogenesis study of ethylenimine was carried out using 2 strains of mice, (C57BL/6 × C3H/Anf)F1 (strain x) and (C57BL/6 × C3H/AKR)F1 (strain y). Test animals were given the maximum tolerated dose (MTD), 4.64 mg/kg, daily by gavage in a 0.5% gelatin vehicle from day 7 through day 28 of age (the MTD for this study was the maximal dose giving no mortality when administered daily for 19 consecutive days). Thereafter, the animals received the same calculated daily dose of ethylenimine mixed in the feed. Groups of 18 animals/sex/strain were used. A number of other compounds were tested in this study. All the animals receiving a particular compound were placed in 1 of 4 rooms. Each room also contained an untreated control group of 18 animals/sex/strain. There was also 1 additional control group of 18 animals/sex/strain given a gelatin suspension during the time when the compounds were administered by stomach tube. Ethylenimine was actually employed as a positive control. There was at least 1 positive control group in each room. Ethylenimine-treated animals were on test for 77 or 78 weeks. The controls were on test for 78 to 89 weeks. Data on mortality, body weight, food and water consumption, haematology, clinical chemistry, and non-carcinogenic pathological effects were not reported. A list of the major organs examined microscopically was not provided. The report stated that all major organs and grossly-visible lesions were examined microscopically. However, the cranium was not dissected. Thyroid glands were sectioned in only 1 of the 5 control groups and not in the ethylenimine-treated animals. The tabulation of tumours listed only the following categories: hepatomas, pulmonary tumours, lymphomas, and total mice with tumours. The investigators lumped all 5 negative control groups together for purposes of comparison with positive controls and experimental compounds. In the ethylenimine-treated animals there was a significant increase in the relative risk for development of hepatomas, pulmonary tumours, and total tumours. In strain x males given ethylenimine, the incidence of hepatomas, pulmonary tumours, and lymphomas was 15/17, 15/17, and 0/17, respectively. Corresponding values for the control strain x males were 8/79, 5/79, and 5/79. For strain x females given ethylenimine, the incidence of the different tumours was 11/15, 15/15, and 0/15, respectively, while the corresponding values for controls were 3/92, 3/92, and 5/92. In strain y males given ethylenimine, the incidence of hepatomas, pulmonary tumours, and lymphomas was 9/16, 12/16, and 0/16, respectively. Corresponding values for strain y male controls were 5/90, 10/90, and 1/90. In strain y females given ethylenimine, the respective tumour incidences were 2/11, 10/11, and 2/11; the corresponding values for strain y female controls were 1/82, 3/82, and 4/82. The actual dose of ethylenimine the animals received from the diet in this study is unclear. The concentration of ethylenimine in the feed was not measured and some material would certainly have been lost by volatilization and reaction with dietary constituents. An experiment, aimed at mimicking conditions of mixing in the diet and storage of ethylenimine during the conduct of the study, indicated loss of a considerable amount of the test substance (Bionetics Research Labs, Inc., 1968; Green & Lowry, 1983; Innest et al., 1969). The same group of workers gave single s.c. injections of 4.64 mg/kg b.w. of ethylenimine to groups of 18 male and 18 female mice of strains (C57Bl/6 × C3H/Anf)F1 and (C57BL/6 × C3H/AKR)F1. The animals were then observed for 18 months. Tumours developed in 7 of the males of the (C57Bl/6 × C3H/Anf)F1 strain - 2 lymphomas, 2 hepatomas, and 5 pulmonary tumours. In the (C57BL/6 × C3H/AKR)F1 strain, 6 of 18 males each developed a lung tumour. For males of both strains the total number of tumours and the incidence of pulmonary adenomas was significantly greater than in the controls (P < .01). In the treated females, 1 animal of each strain developed a lung tumour. The controls consisted of 9 separate groups lumped together for purposes of comparison with a total of about 160 animals/sex/strain (Bionetics Research Labs., Inc., 1968). Fragmentary data are available from another study of ethylenimine administered s.c. at doses of 0.4, 1.3, or 4.0 mg/kg at weekly intervals for 48 weeks to 187 male and female C56BL × CBA mice. After 2 years, dosed animals had an increased incidence of sarcomas at the injection site, tumours of the harderian gland and lung, and malignant hepatomas (Linnik, 1980). Rats A series of ethylenimine derivatives was administered by s.c. injection to groups of 6 male and 6 female albino rats. Ethylenimine was injected twice weekly in an arachis oil vehicle. The dosing regimen occurred over a 67-day period, with a total dose of 20 mg/kg being administered. Sarcomas were found at the injection site in 5 of 6 males and 1 of the females. The tumours were discovered between 355 and 511 days after the beginning of dosing. The study was terminated at 546 days after the beginning of dosing. No tumours were found at locations remote from the point of injection. In controls injected with arachis oil for the same duration as the ethylenimine group, 1 of 10 males developed an injection-site sarcoma at 568 days while 2 tumours developed at remote sites. None of 9 control females developed sarcomas at the injection site; however, a fibroma was found in 1 animal at that site and tumours were found in 2 other animals at remote sites. In another experiment, groups of 6 males and 6 females were given ethylenimine dissolved in water twice weekly by s.c. injection. A total dose of 10 or 12 mg/kg was given to males and females, respectively, over a 59-day period (dosed 5 days/week). The animals were observed up to 540 days; 2 sarcomas were observed at the injection site in females and none in males, while a transitional cell carcinoma of the kidney was seen in 1 male. Concurrent control groups injected with a carbowax-300 vehicle had no injection-site sarcomas (Walpole et al., 1954). Special studies on mutagenicity Sex-linked recessive lethals and translocations were reported in a study in which day-old male Drosophila were treated with an i.p. injection of 0.4 µl of a 10-2 M solution of ethylenimine and subsequently mated. The authors stated that the compound was radiomimetic in producing chromosomal abnormalities - translocations were nearly as frequent as sex-linked lethals - but resembled some chemical mutagens, such as mustards, in producing delayed lethals and translocations which appeared in the F2 generation from F1 parents that appeared normal (Alexander & Glanges, 1965). Other studies have also found that ethylenimine induces recessive lethals and translocations in Drosophila (Lim & Snyder, 1968; Alexander, 1967). Ethylenimine was mutagenic in Salmonella typhimurium strains TA1535 and TA100 without metabolic activation (McCann et al., 1975). The compound was also reported to be mutagenic in spot tests on Neurospora crassa strain N-23 (reverted by base pair mutagens), but not by strain N-24 (reverted by frame shift mutagens) (Ong, 1978), Severe inhibition of replicon initiation and blocked-chain elongation occurred when 5 × 10-4 M ethylenimine was added to cultured HeLa 53 cells (Painter, 1978). Injection of 1 mg/kg ethylenimine (i.p.) into male mice resulted in about a 2/3 reduction in incorporation of 3H-thymidine into testicular DNA as compared to control animals (Seiler, 1977). Lymphocyte cultures from 10 workers exposed to workroom ambient-air concentrations of 0.5 ppm ethylenimine reportedly did not show an increase in chromosomal aberrations (Gaeth & Thiess, 1972). The effect of incubation with ethylenimine was studied in cultures of WI-38 cells and in leukocytes from an adult male volunteer. A concentration of 10-2 M ethylenimine was cytotoxic to the WI-38 cells. Concentrations of 10-3 and 10-4 M were associated with chromatid breaks, gaps, and exchanges. Radiolabel studies with leukocyte cultures incubated with 10-4 M ethylenimine also showed gaps, exchanges, and breaks, occurring primarily in the S-period of the cell cycle (Chang & Elequin, 1967). Acute toxicity LD50 Species Route (mg/kg b.w.) Reference Rat oral 17 Santoski et al., 1977 Rat oral 15 NIOSH, 1977 Rat i.p. 3.8 NIOSH, 1977 Short-term studies No information available. Long-term studies No information available. Comments Ethylenimine appeared to be carcinogenic when administered orally to 2 strains of mice. Both sexes were affected, with the liver and lung being the major target organs. When administered s.c. to rats and mice, the compound appared to be associated with sarcomas at the injection site as well as at other locations. Ethylenimine was found to be mutagenic in Neurospora, Salmonella, and Drosophila, and chromosomal aberrations occurred in cultured mammalian cells exposed to the compound. Use of the oral feeding study in mice for risk analysis is complicated because of the one-dose design of the experiment, questions about the actual dose of ethylenimine received by the animals, the appearance of tumours at 2 sites, and the very high tumour incidence. While clearly indicating that ethylenimine is carcinogenic, the mouse study is not up to modern standards of toxicity testing. Carcinogenicity studies by the oral route are available only in the mouse. Reproduction and teratology studies are not available. EVALUATION Level causing no toxicological effect Ethylenimine has been determined to be carcinogenic in mice. A "no-effect" level in experimental animals has not been established. Estimate of acceptable daily intake for man Acceptable on condition that human exposure to ethylenimine as a result of its migrations into food from immobilized enzyme preparations is reduced to the lowest level technically possible (see polyethylenimine). REFERENCES Alexander, M.L. (1968). Mosaic mutations induced in Drosophila by ethylenimine. Genetics, 56, 273-281. Alexander, M.L. & Glanges, E. (1965). Genetic damage induced by ethylenimine. Proc. Nat'l Acad. Sci., 53, 282-288. Axelson, R.A. (1978). Experimental renal papillary necrosis in the rat: the selective vulnerability of medullary structures to injury. Virchows Arch. A. Path. Anat. and Histol., 381, 79-84. Research Labs., Inc. (1968). Evaluation of carcinogenic, teratogenic, and mutagenic activities of selected pesticides and industrial chemicals. Volume I, Carcinogenic Study. Prepared for National Cancer Institute. Available from National Technical Information Service, U.S. Department of Commerce. Chang, T. & Elequin, F.T. (1967). Induction of chromosome aberrations in cultured human cells by ethylenimine and its relation to cell cycle. Mut. Res., 4, 83-89. Davies, D.J. (1969). The structural changes in the kidney and urinary tract caused by ethylenimine (vinylamine). J. Path., 97, 695-703. Davies, D.J. (1970). The early changes produced in the rabbit renal medulla by ethylenimine: electron-microscope and circulatory studies. J. Path., 101, 329-332. Ellis, B.G., Price, R.G., & Topham, J.C. (1973). The effect of papillary damage by ethylenimine on kidney function and some urinary enzymes in the dog. Chem-Biol. Interact., 7, 131-141. Gaeth, V.J. & Thiess, A.M. (1972). Chromosome studies on chemical workers, Zentralbl. Arbeitsmed. Arbeitsschutz, 22, 357-362. Green, D.R. & Lowry, J.R. (1983). Stability of ethylenimine in mouse chow and 0.5% gel solution. Unpublished report of Cordova Chemical Company. Submitted to the World Health Organization by Cordova Chemical Company. Innes, J.R.M., Ulland, B.M., Valerio, M.G., Petrucelli, L., Fishbein, L., Hart, E.R., Pallota, A.J., Bates, R.R., Falk, H.L., Gart, J.J., Klein, M., Mitchell, I., & Peters, J. (1969). Bioassay of pesticides and industrial chemicals for tumorigenicity in mice: A preliminary note. J. Nat. Cancer Inst., 42, 1101-1114. Lim, J.K. & Snyder, L.A. (1968). The mutagenic effects of two mono functional alkylating chemicals on mature spermatozoa of Drosophila. Mutation Res., 6, 129-137. Linnik, A.B. (1980). Study of the carcinogenic effect of ethylenimine on F1(C57BL × CBA) mice. Eksp. Onkol., 2, 67-68. McCann, J., Choi, E., Yamasaki, E., & Ames, B.N. (1975). Detection of carcinogens as mutagens in the Salmonella/microsome test: Assay of 300 chemicals. Proc. Nat'l Acad. Sci., 72, 5135-5139. NIOSH (1977). Registry of toxic effects of chemical substances. National Institute of Occupational Safety and Health, Washington, DC, USA. Ong, T. (1978). Use of the spot, plate and suspension test systems for the detection of the mutagenicity of environmental agents and chemical carcinogens in Neurospora crassa. Mutation Res., 53, 297-308. Painter, R.B. (1978). Inhibition of DNA replicon initiation by 4 nitroquinoline 1-oxide, adriamycin and ethylenimine. Cancer Res., 38, 4445-4449. Sanotosky, I.V., Muravieva, S.I., Zaeva, G.N., & Semiletkina, N.N. (1977). Urinary excretion of ethylenimine and its metabolite, monoethanolamine, under experimental conditions. Gig. Tr. Prof. Zabol., p. 10-14. Seiler, J.P. (1977). Inhibition of testicular DNA synthesis by chemical mutagens and carcinogens. Preliminary results in the validation of a novel short-term test. Mutation Res., 46, 305-310. Walpole, A.L., Roberts, D.C., Rose, F.L., Hendry, J.A., & Homer, R.F. (1954). Brit. J. Pharmacol., 9, 306-323. Wright, G.J. & Rowe, V.K. (1967). Ethylenimine: Studies of the distribution and metabolism in the rat using carbon-14. Toxicol. Applied Pharmacol., 11, 575-584.
See Also: Toxicological Abbreviations