Cancer Letters 88 (1995) 157-162 LETTERS ~__

Cytotoxic and mutagenic effects of ferric nitrilotriacetate on L5178Y mouse lymphoma cells

Shinya Toyokuni, Jose-Luis Sagripanti, Victoria M. Hitchins* HFZ-112, U.S. Food and Drug Administration, %OO Fishers Lane. Rockville, MD 2OgS7, US.4

Received 4 November 1994; accepted 8 November I994

Abstract

An iron chelate, ferric nitrilotriacetate (Fe-NTA), induces renal proximal tubular necrosis that leads to a high ina- dence of renal adenocarcinoma in rodents. Others have shown that Fe-NTA induces modified DNA base products both in vitro and in vivo. However, Fe-NTA is negative in the Ames Salmonella test with or without S9 activation. The goal of this project was to determine if Fe-NTA is cytotoxic and mutagenic using the L5178Y (TK ’ ..) mouse lymphoma assay. Our experiments showed a relationship between the concentration of Fe-NTA (0 to 1 mM) and the decrease in relative survival. An exposure-dependent increase in the number of mutations was observed with increasing concentrations of Fe-NTA. At 14% relative survival, there was about a 4-fold increase in mutations (triffuorothym~djne resistance) over unexposed, control cells. Ferric nitrate or nitrilotriacetic acid alone induced a relatively low t .S- or 1, l-fold increase in mutation, respectively. Our results establish that Fe-NTA is mutagenic in the L5 178Y mouse lym- phoma assay system.

Keywords: Iron; Iron chelate; Mutagens; Mouse lymphoma cells; Thymidine kinase _... --. -~I-- _...

1. Introductioo

Nitrilotriacetate (NTA) is a synthetic aminotricarboxylic acid that efficiently forms water-soluble chelate complexes at neutral pH with several metal cations, including ferric ions [2,22]. The primary industrial use of NTA is in the treatment of water to prevent mineral buildup in boilers [2,22]. In addition, NTA has been widely used as a substitute for polyphosphates in

* Corresponding author, Tel.: +l 301 4437170; Fax: +I 301 5946775.

household and hospital detergents in the United States, Canada and Europe 12,221. Because chronic exposure to high levels of NTA produces toxicity and neoplasia in the urinary tract [l, 121, probably through chelation of metal ions, the U.S. Food and Drug Administration preseotIy limits the use of this substance in food processing plants to 5 ppm [6].

Particularly important is the need to establish the potential risk of NTA when complexed to iron This metal is normally present in Mood and tissue. Even higher levels are present in several patbdogic conditions, including both genetic and secondary

0304-3835/95/$09.50 0 1995 Elsevier Science Ireland Ltd. All rights reserved SSDI 0304-3835(94)03641-U

158 S. Toyokuni et al. ! Cancer Letters 88 (1995) 157-162

hemochromatosis [lo]. Furthermore, appreciable amounts of iron can be found leaching or cor- roding from metallic medical implants [ 161. In this respect, there have been clinical reports on the in- duction of malignant neoplasms near orthopedic implants manufactured from stainless steel [7]. Supporting this observation, several animal stud- ies have shown that osteosarcoma was induced after implanting stainless steel [27].

Studies to address the risk of NTA and Fe-NTA have been inconclusive. Ad libitum per OS adminis- tration of 10.6 mM Fe-NTA to male Wistar rats produced no renal tumors [24]. By another admin- istration route, repeated intraperitoneal injections of Fe-NTA (5 to 15 mg Fe/kg of Fe-NTA for rats, 1.8-2.7 mg Fe/kg of Fe-NTA for mice) induced acute and subacute tubular necrosis [ 13,18,30] and a subsequent high incidence (60-92%) of renal ad- enocarcinoma in male rodents [11,18,26]. No tumors resulted in rats from intraperitoneal ad- ministration of NTA, alone or chelated to alumin- um [11,18].

Toyokuni and Sagripanti have shown in vitro that Fe-NTA in the presence of reductant, induced DNA single- and double-strand breaks [31,32]. Toyokuni et al. [29] and Umemura et al. [33] have observed formation of 8-hydroxydeoxyguanosine in rat kidney DNA after intraperitoneal adminis- tration of Fe-NTA. These molecular changes in the DNA induced by Fe-NTA may be related to the renal tubular carcinogenesis seen in vivo [11,18,26].

NTA has been found to be non-mutagenic in various microbial assays, including Escherichiu coli, Neurospora crassa, Saccharomyces cerevisiae, and S. pombe [28], and did not induce mutations in Drosophila or rodents [17]. However, NTA did cause chromatic aberrations at sublethal concen- trations (2.5-6 mM) in rat kangaroo cells and human leukocytes in culture [17]. Furthermore, complexes of NTA with heavy metals, such as chromium or mercury, have induced gene muta- tions in the Ames Salmonella typhimurium bacter- ial mutagenesis assay system and increased sister chromatid exchanges in Chinese hamster ovary cells [8,19,20,34].

Fe-NTA was found to be non-mutagenic with the Ames Salmonella assay [ 181. Opposite results

were obtained on tests carried out at the HGPRT locus of V79 Chinese hamster cells in which Fe- NTA appeared mutagenic [25].

The objective of this work was to assess the cytotoxicity and mutagenicity of NTA, Fe(II1) and Fe-NTA using the mouse lymphoma forward mu- tation (MLM) assay at the thymidine kinase (TK) locus [9], which has been selected by the U.S. Na- tional Toxicology Program to evaluate the muta- genic activity of chemicals [23].

2. Materials and methods

2.1. Cell culture L5178Y mouse lymphoma cells, subclone

3.7.2C, heterozygous for thymidine kinase (TK&) [9] were grown in Fischer’s medium for leukemic cells of mice (Gibco Laboratories, Grand Island, NY) containing 10% donor horse serum (HyClone Lab Inc., Logan, UT), 0.1% Pluronic F68 (a gift from BASF Wyandotte Corp., Parsippany, NJ), 0.1 g/l sodium pyruvate (Eastman Chemicals, Rochester, NY) and 1.125 g/l sodium bicarbonate (Fischer Scientific, Fair Lawn, NJ)( 15); FlOP me- dium. Cells were maintained at 37°C in a humidified (90%) atmosphere with 5% COZ and 95% air.

To determine plating efficiency, cytotoxicity and mutagenicity, cells were plated in agar medium made of FlOP and 0.24O/o BBL granulated agar (Becton Dickinson Microbiology Systems, Cockeysville, MD) and 50 IU penicillin-50 pg streptomycin/l (Flow Laboratories, McLean, VA). After 14 days of incubation at 37°C 5% CO*-95% air, clones were counted with an Artek colony counter (Model 982B; Artek Systems Corp., Farmingdale, NY). The cloning efficiency of control L5178Y cells was 23-52%.

2.2. Preparation of Fe-NTA solution The Fe-NTA solution was prepared immediate-

ly before use. Ferric nitrate enneahydrate (Mallin- crodt, Paris, KY) and nitrilotriacetic acid disodium salt (Sigma Chemical Co., St. Louis, MO) were each separately dissolved in doubly distilled deionized water to form 300 mM and 600 mM solutions, respectively. They were then mixed together at a volume ratio of 1:2 (molar ratio, 1:4)

S. Toyokuni et al. /Cancer Letters 88 /i995/ 157-162 L59

with magnetic stirring at room temperature to make a 100 mM Fe-NTA solution 141. The pH was adjusted with sodium bicarbonate to 7.4. The Fe- NTA solution was then filter-sterilized and was added to the medium to obtain the final concentra- tions indicated. After exposing cells to varying concentrations of Fe-NTA for 24 h, cells were cen- trifuged and resuspended in fresh FlOP medium. The same procedure was used for cells exposed to ferric chloride or NTA solution alone.

2.3. Survival assay After cells were exposed to Fe-NTA, an aliquot

was removed, counted and plated in triplicate in 100 x 15 mm plastic petri dishes (Falcon, Cockeysville, MD) containing soft agar medium to determine plating efficiency and cytotoxicity.

2.4. Mutagenesis assay Control and exposed cells were grown in FlOP

medium for 2 days (in nonselecting medium) to allow expression of resistance to trifluorothymi- dine (TFT) (Sigma). Cells (5 x 105) were plated in triplicate in 100 x 15 mm plastic petri dishes (Falcon) containing soft agar medium and 3.0 &ml TFT, and clones were counted after 14 days of growth at 37”C, 5% CO*-95% air [21]. Cells (IO*) were plated in triplicate in soft agar medium to determine plating efficiency. Mutagenicity was expressed as the number of TFT-resistant clones per 10’ survivors. The data for the survival and mutation curves were the means from 7 separate experiments. The background mutation frequency of unexposed cells ranged from 1.5 to 8.2 TFT- resistant clones per lo5 survivors.

3. Results

L5178Y mouse lymphoma cells were exposed for 24 h to water (control), NTA or ferric ions alone and the cytotoxic and mutagenic effects were independently evaluated. The Table shows that NTA alone exhibits a cytotoxic LD% at a concen- tration smaller than but close to 4000 PM. Ferric chloride alone was not cytotoxic up to the highest concentration of 1000 PM. Exposure of cells to NTA alone results in a mutant frequency not significantly different from that observed in un-

Table 1 Effects of NTA or FeCl, on survival and TFT’ mutants

Surviving fraction TFT’ .MutantdlO’ ---- ____.I__

NTA OlrM

2000 pM MJO PM

FeCI,

1.00 f 0.17 5.x f iJ.29 0.83 f 0.13 6 2 f 0.90 0.41 f 0.01 5.I f i.23

500 pM 1.06 f 0.14 8.6 ciz 1.75 1000 pM 0.95 f 0.09 8.9 * i.29

.--- _.-. ..__--I_

treated or controls. The mutation frequency was slightly elevated (approximately 1.5-fold increase) by ferric chloride concentrations between 500 and 1000 PM.

The survival curve for Fe-NTA treatment of L5178Y cells can be seen in Fig. la. The curve shows that cell death is linearly related to the Fe- NTA concentration in the range O-1000 PM (r = 0.991, n = 7). The cell mortality at the max- imal concentration studied (1000 PM) was I4%, still within the useful range of the mutagenicity assay [9]. Fig. 1 b shows the number of TFT- resistant mutants produced in L5178Y mouse lym- phoma cells exposed to various Fe-NTA concen- trations for 24 h in 7 independent experiments. The increase in number of mutants with increasing Fe-NTA concentrations follows the relationship: mutants/lo5 survivors = 0.011 x Fe-NTA (in FM) + 3.90 (correlation coefficient, r = 0.904; number of data used in the correlation, n = 7 x 6 = 42). The mutation d&a were analyz- ed by two different statistical methods. Data were analyzed using unpaired r-test for statistical signif- icance. There is a statistical difference b&ween 0 and 1000 PM Fe-NTA (P < 0.05) but not 0 PM and 1000 PM Fe (bidirectional). Mutation data were also analyzed with a General Linear Model ANOVA, using NCSS Number Cruncher. The effect of dose of Fe-NTA from 0 to 1000 CLM on mutation frequency is statistically significant (P = 0.0017) but not with Fe alone or with NTA alone. The highest Fe-NTA concentration studied (1000 PM) resulted in 14% relative cell survival and induced a 4.1-fold increase in the mtiativn frequency over that observed in unexposed control cells.

160 S. Toyokuni et al. /Cancer Letters 88 (1995) 157-162

6 .s 0.5. 0 ,m

IL

P '5 '5

5 m

0.1 -

0 100 250 500 750 1000

Fe-NTAt@l,

!! 20-

9 ‘5 5

UJ 6 15-

b

jj lo-

; e

i 5-

0; 0 100 250 5Oa 750 1000

Fe-NTA(@l)

Fig. la. Effect of Fe-NTA on survival of L5178Y mouse lym- phoma cells, subclone 3.7.2C. The values given are means and SE. from 7 experiments with triplicate points. Fig. lb. Effect of Fe-NTA on mutagenicity to TFT resistance in L5178Y mouse lymphoma cells. The values given are means and SE. from 7 experiments with triplicate points.

4. Discussion

A test substance is considered mutagenic in the L5178Y TK +‘- mouse lymphoma cell system when the resulting mutation frequency is twice or greater than the background rate [9]. Our data show that Fe-NTA, a renal carcinogen in rodents, is a mutagen in the L5178Y TK +‘- mouse lym- phoma cell system. This finding disagrees with re-

sults obtained by the Ames test [18], but is consistent with recent tests done on the V79 Chinese hamster cell assay 1251.

Our results indicate that NTA and ferric ions by themselves are not mutagenic, but when they are combined, Fe-NTA can induce mutations at the thymidine kinase locus. Our data are also consis- tent with the reports that Fe-NTA plus reducing agents induced DNA strand breaks [31,32] and the formation of modified DNA base products by Fe- NTA [3,29] can be detected with the LS178Y mouse lymphoma assay. In addition, our mutagen- ic data correlate with the observations that Fe- NTA induces adenocarcinomas in rats [ 11,18,26]. These findings suggest that the L5178Y system is more sensitive for evaluating weak mutagens than the Ames bacterial test.

It is important to determine the risk of NTA in humans. NTA is found in drinking water [2]. NTA may even be associated with medical conditions in which the level of iron is either elevated metabolically [5,10,14] or leached or corroded from implanted metallic medical devices [7,16,27]. Data from the L5178Y mouse lymphoma assay and the in vivo results with rodents should definitely raise concerns about the presence of NTA in drinking water and food.

This work was in part done while Shinya Toyokuni held a U.S. Food and Drug Adminis- tration-National Research Council Research Associateship. Present address of ST is Depart- ment of Pathology, Faculty of Medicine, Kyoto University, Sakyo-Ku, Kyoto 606, Japan. We thank John M. Dawson for help with statistical analysis of the mutation data.

References

[I] Anderson, R.L., Alden, C.L. and Merski, J.A. (1982) The effect of nitrilotriacettic on cation disposition and urinary tract toxicity. Fd. Chem. Topic, 20, 105-122.

[2] Anderson, R.L., Bishop, W.E. and Campbell, R.L. (1985) A review of the environmental and mammalian toxicology of nitrilotriacetic acid. Crit. Rev. Toxicol., IS, l-102.

[3] Aruoma, O.I., HaBiwell, B., Gajewski, E. and Diz- daroghr, M. (1989). Damage to bases in DNA induced by hydrogen peroxide and ferric iron chelates. J. Biol. Chem., 264, 20509-20512.

S. Toyokuni et al. / Cancer Letters 88 (199s) 157-162 161

[4] Awai, M., Narasaki, M., Yamanoi, Y. and Seno, S. (1979) Induction of diabetes in animals by parenteral ad- ministration of ferric nitrilotriacetate: a model of experi- mental hemochromatosis. Am. J. Pathol., 95, 663-674.

[5] Batey, R.G., Fong, P.L.C., Shamir, S., and Sherlock, S. (1980) A nontransferrin-bound serum iron in idiopathic hemochromatosis. Dig. Dis. Sci., 25, 340-346.

[6] Boiler water additives. (1984). In: Code of Federal Regulations, Title 2 1, Section 173.310, Food and Drugs, pp. 115-I 16. The Office of Federal Register, Washington, DC.

[7] Brien, W.W., Salvati, E.A., Healey, J.H., Bansal, M., Ghelman, B. and Betts, F. (1990). Osteogenic sarcoma arising in the area of a total hip replacement. J. Bone Joint Surg., 72(A), 1097-1099.

[8] Celotti, L, Furlan, D., Seccati, L. and Levis, A.G. (1987). Interactions of nitrilotriacetic acid (NTA) with Cr(VI) compounds in the induction of gene mutations in cul- tured mammalian cells. Mutat. Res., 156, 219-228.

[9] Clive. D. (1973) Recent developments with the L5178Y TK heterozygous mutagen assay system. Environ. Hlth. Perspect., 6, 119-125.

[IO] Cotran, R.S., Kumar, V. and Robbins, S.L. (1989) Rob- bins Pathologic Basis of Disease, 4th ed. pp. 950-953. W.B. Saunders Company, Philadelphia, PA.

[1 I] Ebina, Y., Okada, S., Hamazaki, S., Ogino, F., Li, J-L. and Midorikawa, 0. (1986) Nephrotoxicity and renal cell carcinoma after use of iron- and aluminum- nitriotriacetate complexes in rats. J. Natl. Cancer Inst., 76, 107-113.

[12] Coyer, R.H., Falk, H.L., Hogan, M., Feldman, D.D. and Richler, W. (1981) Renal tumors in rats given triso- dium nitrilotriacetic acid in drinking water for 2 years. J. Natl. Cancer Inst., 66, 869-880.

(13) Hamazaki, S., Okada, S., Ebina, Y., and Midorikawa, 0. ( 1985) Acute renal failure on glucosuria induced by ferric nitrilotriacetate in rats. Toxicol. Appt. Pharmacol., 77, 267-274.

[14] Hershko, C, and Peto, T.E.A. (1987) Annotation: Non- transferrin plasma iron. Br. J. Haematol., 66, 149-151.

[ 151 Jacobson, E.D., Krell, K., Dempsey, M.J., Lugo, M.H., Ellingson, 0. and Hench, C.W.11 (1978) Toxicity and mutagenicity of radiation from fluorescent lamps and a sunlamp in mouse lymphoma cells. Mutat. Res., 51, 61-75.

(161 Kawahara, H. (1983) Cellular responses to implant mate- rials: biological, physical and chemical factors. Int. Dent. J., 4, 350-375.

[ 171 Kramers, P.G.N. (1976) Mutagenicity studies with nitri- lotriacetic acid (NTA) and citrex 5-5 in Drosophila. Mutat. Res., 40, 277-280.

[18] Li, J-L., Okada, S., Hamazaki, S., Ebina, Y. and Midorikawa, 0. (1987) Subacute nephrotoxicity and in- duction of renal cell carcinoma in mice treated with ferric nitrilotriacetate. Cancer Res., 47, 1867-1869.

[19] Loprieno, N. Boncristiani, G., Venier, P., Mantaldi, A., Majone, F., Bianchi, V., Paglialunga, S. and Levis, A.G.

(1985). Increased mutagenicity of chromium compounds by nitrilotriacetic acid. Environ. Mutagen., 7, 185-200.

[20] Montaidi, A., Zentihn, L., Venier, P., G&t, I,, Bianchi, V., Pagliahmga, S. and Levis, A.G. (1985). Interaction of nitrilotriacetic acid with heavy me&Is in the induction of sister chromatid exchanges in cultured mammalian cells. Environ. Mutagen,, 7, 381-390.

[21] Moore, M.M. and Clive, D. (1982) The quaatitation of TK-‘- and HGPRT- mutants of L5178Y/TK”- mouse lymphoma cells at varying times post-treatment. En- viron. Mutagenesis, 4, 499-519.

[22] Mottola, H.A. (1974) Nitrilotriaoetic acid as a chelating agent: applications, toxicology, and b&environmental impact. Toxicol. Environ. Chem. Rev., 71, 99- 161.

[23] Myhr, B., McGregor, D., Bowers, L., Riach, C., Brown, A.G., Edwards, I., McBride, D., Martin, R. and Caspary, W.J. (1990). L5178Y mouse lymphoma cell mu- tation assay results with 41 compounds. Environ. Molec. Mutagen., 16 (Suppl. 18) 138-167.

[24] Nakano, K., Okada, S., Toyokuni, S. and Midorikawa, 0. (1989) Renal changes indused by chronic oral admin- istration of potassium bromate or ferric nitrilotriacetate in Wistar rats. Jpn. Arch. Intern. Med., 36, 41-47.

1251 Nakatsuka, S., Tanaka, H. and Namba, M. (1990) Muta- genic effects of ferric nitrilotriacetate (Fe-NTAl on V79 Chinese hamster cells and its inhibitory effects on cell-cell communication. Carcinogenesis, 11, 257-260.

[26] Okada, S. and Midorikawa, 0. (1982) Induction of rat renai adenocarcinoma by Fe-nitrilotriacetate (Fe-NTA). Jpn. Arch. Intern. Med., 29, 485-491.

[27] Pedley, R.B., Mea&m, G. and Williams, D.F. (1981). Tumor induction by implant materials. pp. 175-202. In: Fundamental Aspects of Biocompatibihty, Vol. II Editor: D.F. Williams. CRC Series in Biofompatibility (Series ed. Williams, D.F.) CRC Press. Inc.. Boca Raton. FL.

[28] Stine, G.J. and Adams, P. (1974) The effect of nitrilotri- acetic acid on development of Neurospom crus.su. Microb. Genet. Bull, 36, 8-10.

[29] Toyokuni, S., Mori, T. and Dizdaroglu, M. ( 1994) DNA base modifications in renal chromatin of Wistar PdtS

treated with a renal carcinogen, ferric nitrilotriacetate. Int. J. Cancer, 57, 123-128.

[30] Toyokuni, S., Okada, S., Hamazaki, S., Minamiyama, Y., Yamada, Y., Liang, P., Fukunaga, Y. and Midori- kawa, 0. (1990) Combined histochemical and biochemi- cal analysis of sex hormone dependence of ferric nitrilotriacetate-induced renal lipid peroxidation in ddY mice. Cancer Res., 50, 5574-5580.

1311 Toyokuni, S. and Sagripanti, J-L. (1992) Iron-mediated DNA damage: sensitive detection of DNA strand breakage catalyzed by iron. J. Inorg. Biochem.. 47, 241-248.

[32] Toyokuni, S. and Sagripanti, J-L. (1993) DNA single- and double-strand breaks produced by ferric nitrilotri- acetate in relation to renal tubular carcinogenesis. Car- cinogenesis. 14, 223-227.

162 S. Toyokuni et a/. /Cancer Letters 88 (1995) 157-162

[33] Umemura, T., Sai, K., Takagi, A., Hasegawa, R. and Kurokawa, Y. (1990) Formation of I-hydroxydeoxy- guanosine (8-OH-dG) in rat kidney DNA after in- traperitoneal administration of ferric nitrilotriacetate (Fe-NTA). Carcinogenesis, 11, 345-347.

[34] Venier, P., Montaldi, A., Gava, C., Zentilin, L., Tecchio, G, Bianchi, V., Paglialunga, S., and Levis, A.G. (1985). Effects of nitrilotriacetic acid on the induction of gene mutations and sister-chromatic exchanges by insoluble chromium compounds. Mutat. Res., 156, 219-228.