1098
Chem. Res. Toxicol. 1998, 11, 1098-1104
Mechanism for Inhibition of Thyroid Peroxidase by Leucomalachite Green Daniel R. Doerge,* Hebron C. Chang,† Rao L. Divi,†,‡ and Mona I. Churchwell Division of Chemistry, National Center for Toxicological Research, Jefferson, Arkansas 72079 Received December 22, 1997
The triphenylmethane dye, malachite green (MG), is used to treat and prevent fungal and parasitic infections in the aquaculture industry. It has been reported that the reduced metabolite of MG, leucomalachite green (LMG), accumulates in the tissues of fish treated with MG. MG is structurally related to other triphenylmethane dyes (e.g., gentian violet and pararosaniline) that are carcinogenic in the liver, thyroid, and other organs of experimental animals. The ability of LMG to inhibit thyroid peroxidase (TPO), the enzyme that catalyzes the iodination and coupling reactions required for thyroid hormone synthesis, was determined in this study. LMG inhibited TPO-catalyzed tyrosine iodination (half-maximal inhibition at ca. 10 µM). LMG also inhibited the TPO-catalyzed formation of thyroxine in low-iodine human goiter thyroglobulin (half-maximal inhibition at ca. 10 µM) using a model system that measures simultaneous iodination and coupling. Direct inhibition of the coupling reaction by LMG was shown using a coupling-only system containing chemically preiodinated thyroglobulin as the substrate. Incubation of LMG with TPO, iodide, and tyrosine in the presence of a H2O2generating system yielded oxidation products that were identified by using on-line LC/APCIMS as desmethyl LMG, 2desmethyl LMG, 3desmethyl LMG, MG, and MG N-oxide. Similar products from LMG were observed in incubations with TPO and H2O2 alone. These findings suggest that the anti-thyroid effects (increased serum thyroid-stimulating hormone and decreased serum thyroxine) observed in rats treated with LMG result from blockade of hormone synthesis through alternate substrate inhibition and that chronic exposure could cause thyroid follicular cell tumors through a hormonal mechanism. The observed TPO-catalyzed oxidative demethylation of LMG to a primary arylamine also suggests a genotoxic mechanism for tumor formation is possible.
Introduction 1
Malachite green (MG, see Figure 1) is a triphenylmethane dye that has been used traditionally in aquaculture to treat fish eggs and adult fish for fungal infestations and ectoparasites (for a review, see ref 1). Malachite green is not approved for use by the U.S. Food and Drug Administration (FDA); however, worldwide use of MG in aquaculture probably continues because of low cost, ready availability, and high efficacy (2). A further concern comes from reports that the product from reductive metabolism of MG, leucomalachite green (LMG), accumulates in the edible tissue of fish treated with MG * Corresponding author. Phone: (870) 543-7943. Fax: (870) 5437720. E-mail:
[email protected]. † Supported by a fellowship from the Oak Ridge Institute for Science and Education administered through an interagency agreement between the U.S. Department of Energy and the U.S. Food and Drug Administration. ‡ Current address: Laboratory of Chemical Carcinogenesis and Tumor Promotion, National Cancer Institute, 37 Convent Dr., Bethesda, MD 20892-4255. 1 Abbreviations: APCI-MS, atmospheric pressure chemical ionization mass spectrometry; DIT, 3,5-diiodotyrosine; GV, gentian violet N,N,N′,N′,N′′,N′′-hexamethylpararosaniline; HRP, horseradish peroxidase; LMG, leucomalachite green; MIT, 3-iodotyrosine; MG, malachite green N-[4-[[4-(dimethylamino)phenyl]phenyl]methylene-2,5cyclohexadien-1-ylidene]-N-methylmethanaminium chloride; PEG, poly(ethylene glycol); rT3, 3,3′,5′-triiodo-L-thyronine; T3, 3,3′,5-triiodoL-thyronine; T4, thyroxine [3-(4-hydroxy-3,5-diiodophenoxy)-3,5-diiodophenyl]-L-alanine; Tg, thyroglobulin; TPO, thyroid peroxidase; TSH, thyroid-stimulating hormone.
S0893-228x(97)00226-9
(3). Humans may also be exposed to MG through occupational exposure in fabric dyeing (1). Malachite green is related structurally to other triphenylmethane dyes that are carcinogenic. For example, gentian violet (GV, see Figure 1 for the structure of the principal component of the dye mixture) caused a statistically significant increase in the incidence of thyroid follicular cell adenocarcinoma in male and female rats in addition to a slight increase in hepatocellular adenomas (4). In a related study, these workers also observed an increased incidence of hepatocellular adenocarcinoma in female and male mice (5). It was concluded that occupational exposure to pararosaniline, a primary arylamine, was associated with increased risks of bladder cancer in dye industry workers (6). Benzyl violet 4B was found to be carcinogenic in several organs in rats (7). MG has been reported to enhance the development of preneoplastic lesions induced in livers from rats treated with N-nitrosodiethylamine (8). Subsequent work by this group suggests this tumor-promoting ability is related to radical formation and initiation of lipid peroxidation (9), leading to inhibition of DNA synthesis and cytotoxicity (10). In part on the basis of these findings, in 1993 the FDA nominated MG as a priority chemical for carcinogenicity testing by the National Toxicology Program. The transformation of triphenylmethanes by oxidative enzymes has been reported. Hepatic microsomes from
This article not subject to U.S. Copyright. Published 1998 by the American Chemical Society Published on Web 08/19/1998
Thyroid Peroxidase Inhibition by Leucomalachite Green
Chem. Res. Toxicol., Vol. 11, No. 9, 1998 1099
Experimental Procedures
Figure 1. Structures for cited triphenylmethane dyes.
several species catalyze the N-demethylation of GV to N,N,N′,N′,N′′-pentamethylpararosaniline, N,N,N′,N′′-tetramethylpararosaniline, and the primary arylamine N,N,N′,N′-tetramethylpararosaniline (11). Lignin peroxidase from the fungus Phanerochaete chrysosporium also catalyzed the N-demethylation of GV to secondary arylamines in the presence of H2O2 (12). MG was also a substrate for degradation by this peroxidase, although reaction products were not identified (12). The demethylation of GV catalyzed by horseradish peroxidase (HRP) produced formaldehyde, and it was postulated that a cation radical intermediate was involved because oxygen uptake, thiyl radical formation from glutathione, and an undefined ESR signal from GV were observed (13). It was also shown that reductive metabolism of GV to the corresponding radical species occurred upon treatment with hepatic microsomes supplemented with NADPH under an inert atmosphere (14) or to leuco-GV by treatment with anaerobic suspensions of human, rat, and chicken gut microflora (15). A recent study demonstrated a similar reductive metabolism of MG by human, rat, and monkey gut microflora (16). The thyroid carcinogenicity observed for GV may be relevant to MG, and there are two possible mechanisms: inhibition of TPO-catalyzed thyroxine synthesis and/or oxidative demethylation of GV in the thyroid or other organs to pararosaniline, a primary arylamine that is a suspected human carcinogen (6). The first possibility involves a nongenotoxic mechanism in which disruption of thyroid hormone homeostasis leads to a prolonged growth stimulus in the thyroid from the action of pituitary-derived TSH (17); the second possibility would require further metabolic activation by N-hydroxylation to form electron-deficient DNA-reactive intermediates (18). The potential for human exposure to MG in food and the environment prompted us to investigate the interactions of MG and metabolites with TPO, the enzyme that catalyzes the oxidative biosynthesis of iodothyronines in the thyroid.
Reagents and Chemicals. MG and LMG were synthesized by Chemsyn Science Laboratories (Lanexa, KS) and were determined before use to be 96 and 99% pure, respectively, using LC/APCI-MS and 1H NMR; 3,5-diiodotyrosine (DIT), horseradish peroxidase (HRP), glucose, glucose oxidase, hydrogen peroxide, 3-iodotyrosine (MIT), methimazole, poly(ethylene glycol) (PEG), potassium iodide, and tyrosine were purchased from Sigma Chemical Co. (St. Louis, MO). Porcine TPO used in this study was purified and quantified spectrophotometrically as described earlier (19). Low-iodine human goiter thyroglobulin, containing 0.038% I equivalent to an average of two atoms of I per molecule (20), was a generous gift from A. Taurog (University of Texas Southwestern Medical Center). Inhibition of Tyrosine Iodination by LMG. Tyrosine (150 µM in a 1 mL total volume), iodide (150 µM), and TPO (16 nM) were incubated with LMG (0-30 µM) at pH 6.5 in the presence of 2.5 mM glucose/glucose oxidase (5 milliunits/mL) for 5 min in 0.1 M potassium phosphate buffer at pH 7.0 and 37 ( 0.1 °C. Alternatively, incubations were carried out in the presence of H2O2 (200 µM) as indicated. Formation of iodotyrosines was determined in triplicate at various times (0-6 min) using HPLC with UV detection as previously described (19). The formation of H2O2 by glucose/glucose oxidase, measured using HRPcatalyzed scopoletin oxidation as previously described (21), was unaffected by the presence of LMG (30 µM). Inhibition of Iodination and Coupling in Goiter Thyroglobulin by LMG. Measurement of simultaneous TPOcatalyzed tyrosine iodination and thyroid hormone synthesis was measured in human goiter Tg using the procedure described previously (20). Briefly, low-iodine Tg (0.038% iodine, 0.76 µM) was incubated with TPO (24 nM), iodide (100 µM), 1.5 mM glucose, and 5 milliunits/mL glucose oxidase in phosphate buffer (pH 7.0) in a total volume of 1.0 mL. After incubation at 37 °C for various times (0-60 min), 0.5 mM methimazole was added to stop the reaction. The reaction mixture was digested with Pronase, and then MIT, DIT, and the thyroid hormones were extracted with ethyl acetate and quantified using LC with 230 nm absorbance detection (22). The inhibition of iodination/ coupling in Tg was determined in triplicate for concentrations of LMG ranging from 0 to 30 µM. Inhibition of TPO-Catalyzed Coupling in Preiodinated Goiter Thyroglobulin by LMG. Measurement of TPOcatalyzed formation of T4 and T3 was similarly carried out using preiodinated goiter Tg as the coupling substrate in the absence of iodide as described previously (20). Chemically preiodinated goiter Tg containing 7.5-9.7 residues of DIT and 15-16 residues of MIT per molecule (0.4-0.45 mg/mL) was incubated with TPO (26 nM), 1.5 mM glucose, and 5 milliunits/mL glucose oxidase in 67 mM phosphate buffer (pH 7.0) in a total volume of 1.0 mL. After incubation at 37 °C for various times (0-60 min), the reaction was stopped and analyzed for thyroid hormones as described above. The inhibition of coupling was determined using a single concentration of 30 µM LMG or MG over the course of 1 h with duplicate samples being analyzed for each time point. These results were also confirmed in a second experiment (not shown). The detection limit (s/n ) 3) for T4 was ca. 8 ng on-column or 0.003 newly formed residue/molecule of Tg, and the T4 content of the Tg before treatment with TPO was below this value. Analysis of Reaction Products from LMG. TPO-catalyzed oxidation of LMG and formation of related products were monitored by HPLC using 260 nm detection. The incubations were also performed in the absence of potassium iodide, tyrosine, or both. A Spherisorb S5 Nitrile column (80A, 5 µm particle size, 4.6 mm × 250 mm, Phase Separations Inc., Norwalk, CT) was used to analyze the LMG derivatives with a solvent system containing 40% acetonitrile in 50 mM ammonium acetate (pH 4.5) at a flow rate of 1.0 mL/min. MG formation was monitored spectrophotometrically using 620 nm absorbance (model 8452A, Hewlett-Packard, Palo Alto, CA). The incubation conditions were the same as those described above for tyrosine iodination.
1100 Chem. Res. Toxicol., Vol. 11, No. 9, 1998
Figure 2. Inhibition of TPO-catalyzed tyrosine iodination by LMG. TPO was incubated at 37 °C with tyrosine, potassium iodide, glucose, and LMG at the concentrations indicated in Experimental Procedures. The reaction was initiated by addition of glucose oxidase, and at the indicated times, an aliquot was removed for analysis of MIT using HPLC with 289 nm detection. Single determinations at each time point were made. All analyses were conducted in triplicate from 60 s incubations at room temperature. Incubations were carried out as described above using 30 µM LMG, and after a 60 s reaction time, a 100 µL aliquot was analyzed using on-line LC with APCI-MS detection. The LC separation was performed using a Prodigy ODS-3 column (5 µm, 4.6 mm × 250 mm, Phenomenex Co., Torrance, CA) at a flow rate of 1.0 mL/min with a mobile phase gradient consisting of 50% acetonitrile and 50% aqueous ammonium acetate (50 mM, pH 4.5) to 100% acetonitrile over a 10 min linear ramp period. MS experiments were performed using a Platform singlequadrupole mass spectrometer (Micromass, Altrincham, U.K.) equipped with an APCI interface. The total LC column effluent (1.0 mL/min) was delivered into the atmospheric pressure ion source (150 °C) through a heated nebulizer probe (500 °C) using nitrogen as the probe and bath gas (275 L/h). At a low sampling cone-skimmer voltage (15 V), mass spectra for LMG and related compounds consisted predominately of the respective protonated molecule; however, at higher voltages (60 V), diagnostic fragment ions were formed. Positive ions were acquired in the full scan mode (m/z 100-650, 2.1 s total cycle time) in series with a UV detector set at 260 nm. Two separate scan functions of 1.0 s each were used to acquire spectra sequentially at 15 and 60 V. Background-subtracted mass spectra were obtained by averaging spectra across the respective chromatographic peak and subtracting average background spectra immediately before and after this peak. The mass spectrometer was calibrated over the range m/z 85-1200 using a solution of poly(ethylene glycol)s [PEG 200 (25 µg/mL), 300 (50 µg/mL), 600 (75 µg/mL), and 1000 (250 µg/mL)] in 50% acetonitrile in aqueous ammonium acetate (5 mM). A switching valve (Rheodyne model 7030, Cotati, CA) was used to divert chromatographically unretained compounds to waste during the initial 3.5 min of each run. During this interval, a constant flow of the same mobile phase was delivered to the APCI probe by a Waters M6000 pump.
Results LMG inhibited the initial rate and maximal extent of TPO-catalyzed iodination of tyrosine as shown in Figure 2. LMG did not inactivate TPO either in the absence or in the presence of iodide (not shown). The kinetics of MIT formation showed no evidence for a lag phase characteristic of inhibition by alternate iodination substrates as
Doerge et al.
Figure 3. Inhibition of TPO-catalyzed iodination/coupling in thyroglobulin by LMG. TPO was incubated at 37 °C with lowiodine human thyroglobulin, potassium iodide, glucose, and LMG at the indicated concentrations. The reaction was initiated by addition of glucose oxidase, and at the indicated times, an aliquot was removed for analysis of thyroxine and triiodothyronine following protease digestion and solvent extraction using HPLC with 230 nm detection. The mean determinations with standard deviations (error bars) are shown.
previously described (19, 22, 23). Inhibition by LMG was observed when H2O2 was added in a single bolus or when it was generated enzymatically. It was further determined that LMG did not block glucose/glucose oxidasecatalyzed formation of H2O2. The concentration of LMG required for 50% inhibition of the control rate of MIT formation (IC50) was estimated to be 5 µM. MG also inhibited TPO-catalyzed MIT formation in this concentration range (data not shown). LMG inhibited TPO-catalyzed coupling of iodotyrosyl residues on Tg to form Tg-bound T4 and T3. The assay measures simultaneous iodination of tyrosyl residues and coupling (20) and was used because it approximates the conditions that exist in the intact thyroid gland (i.e., continuous uptake of iodide and rate-limiting formation of H2O2; see ref 24). LMG inhibited both the initial rate and the maximal amount of coupling products formed in a concentration-dependent manner (see Figure 3). There was no evidence for a lag phase in coupling from as early as 1 min in the presence of 15 or 30 µM LMG. The IC50 was estimated to be 17 µM using four LMG concentrations from 15 to 30 µM (not shown). Because in this model system iodothyronine formation is dependent on formation of DIT and MIT residues prior to coupling, coupling could be affected by inhibitors that affect only iodination. To further clarify the effect on coupling, inhibition by LMG and MG was determined in a couplingonly model system that uses chemically preiodinated Tg as the substrate. Figure 4 shows that MG and LMG similarly inhibited both the rate and extent of TPOcatalyzed coupling (ca. 50% of control) throughout the time course of 60 min. Under the conditions used in Figure 2, LMG was oxidized to a mixture of products including demethylated LMG species, MG, and a product with a mass spectrum consistent with MG N-oxide. The possible formation of a C-monooxygenation product, the pseudobase or carbinol (see Figure 1 and ref 1), was eliminated on the basis of the molecular weight (M + H+ ) m/z 347) determined using APCI-MS and the different LC retention time of
Thyroid Peroxidase Inhibition by Leucomalachite Green
Chem. Res. Toxicol., Vol. 11, No. 9, 1998 1101
Figure 4. Inhibition of TPO-catalyzed coupling in preiodinated thyroglobulin by LMG and MG. TPO was incubated with Tg that had been previously chemically iodinated using I3- as described in Experimental Procedures. The reaction was initiated by addition of glucose oxidase, and at the indicated times, an aliquot was removed for analysis of thyroxine (T4) and triiodothyronine (T3) using HPLC with 230 nm detection following protease digestion and solvent extraction. The total amount of coupling products is shown as T4 + T3, which at the 60 min control time point consisted of 0.60 T4 residue and 0.03 T3 residue per molecule of Tg. The data points shown are averages of closely agreeing duplicate determinations, and similar results were observed in an independent experiment.
an authentic synthetic standard. Consumption of LMG ranged from ca. 12-34% in the presence of an enzymatic H2O2-generating system to ca. 75% when 200 µM H2O2 was added directly (not shown). Product formation, estimated using the UV or APCI-MS response factors for LMG, ranged from 40 to 87% of the LMG consumed. This suggests either that additional unidentified products were formed or that the assumptions about response factors were inaccurate. The formation of desmethyl LMG, 2desmethyl LMG (symmetrical), 3desmethyl LMG, MG N-oxide, and MG was determined using APCI mass spectra (M + H+ or M+ ions at low sampling coneskimmer voltages and fragment ions at high voltages; see ref 25) and LC retention times (see Figures 5 and 6). The symmetrical structure of 2desmethyl LMG was inferred by comparing the mass spectrum and LC retention time with that from authentic synthetic symmetrical 2desmethyl LMG, formed by condensation of N-methylaniline with benzaldehyde (26). While the mass spectra for symmetrical and unsymmetrical 2desmethyl LMG are indistinguishable, oxidation with PbO2 to the respective 2desmethyl MG species gives spectra whose differences further confirm the proposed structures (not shown; see ref 25). Furthermore, the LC retention times are different (not shown). A signal was also observed for a product (s/n ∼ 3) with a mass spectrum and LC retention time consistent with desmethyl MG N-oxide (M + H+ ) m/z 331, tR ) 8.95 min). When H2O2 was added directly to the enzymatic reaction mixture, all these same products were formed in greater amounts. No signal corresponding to 4desmethyl LMG (M + H+ ) m/z 275) was observed under any conditions. The same LMG-derived products were observed at quantitatively different levels when iodide, tyrosine, or both cosubstrates were removed from the incubation mixture (not shown). Control incubations conducted in
Figure 5. LC/APCI-MS analysis of TPO-catalyzed LMG oxidation products. Incubation of LMG with TPO and cofactors was carried out as described in the legend of Figure 2, and after 60 s, aliquots were analyzed using on-line LC/APCI-MS. The mass chromatograms for the indicated respective protonated molecules (molecular cation for MG) were acquired at a low cone voltage (15 V) that produces minimal fragmentation.
the absence of H2O2 or TPO contained small amounts of desmethyl LMG present as an impurity from the LMG ( 2desmethyl > 3desmethyl). Evidence for N-oxidation was provided by the observation
Thyroid Peroxidase Inhibition by Leucomalachite Green Scheme 2. Proposed Mechanism for Inhibition of TPO-Catalyzed Iodination and Coupling by LMG
of compounds with LC retention times and APCI mass spectra consistent with MG N-oxide and a demethylated homologue; oxidation of LMG to MG was also observed (see Figure 4 for molecular ion chromatograms). Figure 5 shows the mass spectra obtained for the desmethyl LMG products at both low and high cone voltages. Using both acquisition conditions yields spectral information from the protonated molecules (low voltage) and diagnostic fragment ions (high voltage). The prominent fragment ion observed for each compound corresponds to the loss of methyl and phenyl moieties from the protonated molecule (M + H minus m/z 92)+. The proposed structure of MG N-oxide was assigned from the observed molecular weight, the shorter retention time as predicted from the more polar N-oxide relative to MG, and fragmentation to yield a mass spectrum similar to that produced by MG, including loss of m/z 16 to yield the MG cation molecule (m/z 329, not shown). A likely mechanism for formation of these products involves initial single-electron transfer from the LMG nitrogen to the electron-deficient heme moiety in TPO compound I (see Scheme 2). It is plausible that such a cation radical intermediate could yield the demethylated products via proton transfer to yield a carbon-centered radical prior to oxygen rebound and loss of formaldehyde, the N-oxygenated product via direct oxygen rebound, and a second one-electron transfer to an oxidized heme species (compound II) to give MG. While the reactivity of LMG with TPO-OI was suggested by the observation of iodinated LMG and iodinated desmethyl derivatives, the small amounts of these products formed suggest that the relative rate for this pathway is slow. The inhibition of TPO-catalyzed iodination and phenolic oxidations (e.g., coupling and guaiacol oxidation) is also consistent with the proposed mechanism. The ability of LMG to inhibit TPO-catalyzed iodination and coupling reactions demonstrates the potential for disruption of thyroid hormone homeostasis. The regulation of thyroid hormone levels involves feedback through pituitary-derived TSH by production of a direct growth stimulus to the thyroid. When thyroid hormone synthesis is blocked chronically, as would be predicted in a highdose lifetime exposure rodent bioassay, the prolonged growth stimulus from TSH could induce thyroid follicular tumors (17). The plausibility of this mechanism was recently confirmed by the results from 28-day dietary
Chem. Res. Toxicol., Vol. 11, No. 9, 1998 1103 Scheme 3. Hypothetical Reaction Sequence of LMG N-Demethylation and N-Hydroxylation Leading to Covalent Bonding with Biological Nucleophiles
feeding of 1160 ppm LMG to F344 male rats (S. Culp, personal communication). A statistically significant (p < 0.05) decrease in serum T4 (but not T3) and an increase in TSH were observed at 4 and 21 days of treatment. However, the relevance of thyroid follicular cell tumor formation, especially in male rats, to human thyroid cancer is unclear because of (1) very different exposure regimens (i.e., human exposure is typically low-dose, intermittent as opposed to high-dose, chronic administration in rodent bioassays), (2) the many documented differences in thyroid hormone biochemistry and physiology between rats and humans (e.g., different hormone binding proteins, serum half-life, and thyroid morphology), (3) differences in susceptibility to disruption of thyroid homeostasis by antithyroid chemicals, and (4) the different cellular origin of the most common tumor type (follicular vs papillary; see ref 30). However, the use of such a sensitive animal model is consistent with hazard identification in a regulatory risk assessment procedure. The demonstrated formation of a primary arylamine, 3desmethyl LMG, from TPO-mediated oxidation of LMG raises another possible mechanism for tumor formation in the thyroid and other organs. Many primary arylamines are suspected or actual carcinogens in humans and animals, including pararosaniline (see Figure 1). The mechanism of action typically involves metabolic activation to N-oxidized species that are capable of covalent bonding with DNA at critical sites (18). Scheme 3 shows a hypothetical reaction sequence of N-demethylation followed by N-hydroxylation that could lead to covalent bonding of a reactive LMG metabolite (e.g., the imine) to nucleophilic centers in macromolecules (e.g., DNA bases and protein side chains). It is also likely that oxidative metabolism in other tissues of an intact animal (e.g., liver and leukocytes) could provide additional pathways for demethylation of LMG to primary arylamine species. The validity of these proposed mechanisms will be tested in a 2-year chronic rodent carcinogenicity bioassay to be conducted at this institution. Residues of LMG in edible catfish muscle were observed at levels approaching 1 µg/g of tissue following treatment of catfish with MG under putative use conditions (3). Furthermore, we have observed LMG (up to
1104 Chem. Res. Toxicol., Vol. 11, No. 9, 1998
100 ppb) and demethylated/N-oxidized metabolites (ca. 60 ppb) in foreign farm-raised fish (D. R. Doerge et al., unpublished). Although a thorough assessment of LMG exposure to consumers through consumption of fish farmed in the United States has not been completed, these results do give an estimate for possible human exposure. It also will be important to assess human exposure to MG from occupational settings (i.e., fish farmers and dye industry workers). The combination of exposure data and hazard assessment results from the rodent carcinogenicity bioassay is necessary to accurately determine toxicological risks for MG. Mechanistic information regarding LMG inhibition of thyroid hormone synthesis through a direct action on TPO may be useful in clarifying a possible positive finding of MG/LMGinduced thyroid carcinogenicity in rodents, especially at high doses.
Acknowledgment. We recognize the generous contributions of Dr. Alvin Taurog (University of Texas Southwestern Medical Center), who provided human goiter thyroglobulin and many helpful discussions, and Dr. Sandra J. Culp (National Center for Toxicological Research) for providing preliminary results from rat feeding studies. This research was supported in part by Interagency Agreement 224-93-0001 between NCTR/FDA and the National Institute for Environmental Health Sciences/National Toxicology Program.
References (1) Culp, S. J., and Beland, F. A. (1996) Malachite green: a toxicological review. J. Am. Coll. Toxicol. 15, 219-238. (2) Schnick, R. A. (1988) The impetus to register new therapeutics for aquaculture. Prog. Fish-Cult. 50, 190-196. (3) Plakas, S. M., El Said, K. R., Stehly, G. R., Gingerich, W. H., and Allen, J. H. (1996) Uptake, tissue distribution, and metabolism of malachite green in the channel catfish (Ictalurus punctatus). Can. J. Fish. Aquat. Sci. 53, 1427-1433. (4) Littlefield, N. A., and Gaylor, D. W. (1989) Chronic toxicity/ carcinogenicity studies of gentian violet in Fischer 344 rats: Twogeneration exposure. Fundam. Chem. Toxicol. 27, 239-247. (5) Littlefield, N. A., Blackwell, B.-N., Hewitt, C. C., and Gaylor, D. W. (1985) Chronic toxicity and carcinogenicity studies of gentian violet in mice. Fundam. Appl. Toxicol. 5, 902-912. (6) Case, R. A. M., and Pearson, J. T. (1954) Tumours of the urinary bladder in workmen engaged in the manufacture and use of certain dyestuff intermediates in the British chemical industry. Br. J. Ind. Med. 11, 213-221. (7) Iheda, Y. S., Horiuchi, A., Imoto, Y., Aida, Y., and Kobayashi, K. (1974) Induction of mammary gland and skin tumors in female rats by the feeding of benzyl violet 4B. Toxicology 2, 275-284. (8) Fernandes, C., Lalitha, V. S., and Rao, K. V. (1991) Enhancing effect of malachite green on the development of hepatic preneoplastic lesions induced by N-nitrosodiethylamine in rats. Carcinogenesis 12, 839-845. (9) Pandanikar, A., Maru, G. B., and Rao, K. V. (1994) Dose-response effects of malachite green on free radical formation, lipid peroxidation and DNA damage in Syrian hamster embryo cells and their modulation by antioxidants. Carcinogenesis 15, 2445-2448. (10) Rao, K. V. (1995) Inhibition of DNA synthesis in primary rat hepatocyte cultures by malachite green: a new liver tumor
Doerge et al. promoter. Toxicol. Lett. 81, 107-113. (11) McDonald, J. J., Breeden, C. R., North, B. M., and Roth, R. W. (1984) Species and strain comparison of the metabolism of gentian violet by liver microsomes. J. Agric. Food Chem. 32, 596-600. (12) Bumpus, J. A., and Brock, B. J. (1988) Biodegradation of crystal violet by the white rot fungus Phanerochaete chrysosporium. Appl. Environ. Microbiol. 54, 1143-1150. (13) Gadelha, F. R., Hanna, P. M., Mason, R. P., and Docampo, R. (1992) Evidence for free radical formation during horseradish peroxidase-catalyzed N-demethylation of crystal violet. Chem.Biol. Interact. 85, 35-48. (14) Harrelson, W. G., and Mason, R. P. (1982) Microsomal reduction of gentian violet. Evidence for cytochrome P450-catalyzed free radical formation. Mol. Pharmacol. 22, 239-242. (15) McDonald, J. J., and Cerniglia, C. E. (1984) Biotransformation of gentian violet to leucogentian violet by human, rat, and chicken intestinal microflora. Drug Metab. Dispos. 12, 330-336. (16) Henderson, A. L., Schmitt, T. C., Heinze, T. M., and Cerniglia, C. E. (1997) Reduction of malachite green to leucomalachite green by intestinal bacteria. Appl. Environ. Microbiol. 63, 4099-4101. (17) Williams, E. D. (1990) TSH and thyroid cancer. Horm. Metab. Res. 23, 72-75. (18) Beland, F. A., and Kadlubar, F. F. (1990) Metabolic activation and DNA adducts of aromatic amines and nitroaromatic hydrocarbons. In Chemical Carcinogenesis and Mutagenesis (Cooper, C. S., and Grover, P. L., Eds.) pp 267-326, Springer-Verlag, Berlin. (19) Divi, R. L., and Doerge, D. R. (1996) Inhibition of thyroid peroxidase by dietary flavonoids. Chem. Res. Toxicol. 9, 16-23. (20) Taurog, A., Dorris, M. L., and Doerge, D. R. (1996) Mechanism of simultaneous iodination and coupling catalyzed by thyroid peroxidase. Arch. Biochem. Biophys. 330, 24-32. (21) Doerge, D. R., Divi, R. L., and Taurog, A. (1997) Mechanism for the anti-thyroid action of minocycline. Chem. Res. Toxicol. 10, 4958. (22) Divi, R. L., and Doerge, D. R. (1997) Identification, characterization, and mechanisms of anti-thyroid activity of isoflavones from soybean. Biochem. Pharmacol. 54, 1087-1096. (23) Doerge, D. R., Decker, C. J., and Takazawa, R. S. (1993) Chemical and enzymatic oxidation of benzimidazoline-2-thiones: A dichotomy in the mechanism of peroxidase inhibition. Biochemistry 32, 58-65. (24) Corvilain, B., Van Sande, J., and Dumont, J. E. (1991) Role of the cAMP and the phosphatidylinositol-Ca cascades in mediating the effects of thyrotropin and iodide on hormone synthesis and secretion in human thyroid slices. Endocrinology 128, 779-785. (25) Doerge, D. R., Churchwell, M. I., Bajic, S., and Preece, S. W. (1997) Analysis of malachite green using LC/MS and MS/MS. Proceedings of the 45th Conference on Mass Spectrometry and Allied Topics, p 1084, American Society for Mass Spectrometry, Santa Fe, NM. (26) Fierz-David, H. E. (1921) Triphenylmethane Dyes. In The Fundamental Process of Dye Chemistry, pp 145-148, Van Nostrand Co., New York. (27) Ohtaki, S., Nakagawa, H., Kimura, S., and Yamazaki, I. (1981) Analyses of catalytic intermediates of hog thyroid peroxidase during its iodination reaction. J. Biol. Chem. 256, 805-810. (28) Magnusson, R. P., Taurog, A., and Dorris, M. L. (1984) Mechanisms of thyroid peroxidase- and lactoperoxidase-catalyzed reactions involving iodide. J. Biol. Chem. 259, 13783-13790. (29) Taurog, A., Dorris, M. L., and Doerge, D. R. (1996) Minocycline and the thyroid: Anti-thyroid effects of the drug, and the role of thyroid peroxidase in minocycline-induced black pigmentation of the gland. Thyroid 6, 203-211. (30) McClain, R. M. (1995) Mechanistic considerations for the relevance of animal data on thyroid neoplasia to human risk assessment. Mutat. Res. 333, 131-142.
TX970226O
