Effect of Sulfite on Cognitive Function in Normal and Sulfite Oxidase Deficient Rats

Neurotoxicology and Teratology 27 (2005) 47–54 www.elsevier.com/locate/neutera

Effect of sulfite on cognitive function in normal and sulfite oxidase deficient rats

Vural Kqc¸qkataya,*, Feyza SavcVog˘lub,Gqlay HacVog˘lub, Piraye YargVc¸og˘luc, Aysel Ag˘arb

aPamukkale University, Faculty of Medicine, Department of Physiology, Kinikli, 20020, Denizli, Turkey bAkdeniz University, Faculty of Medicine, Department of Physiology, 07070, Antalya, Turkey cAkdeniz University, Faculty of Medicine, Department of Biophysics, 07070, Antalya, Turkey

Received 17 May 2004; received in revised form 14 October 2004; accepted 18 October 2004 Available online 13 November 2004

Abstract

Sulfites, which are commonly used as preservatives, are continuously formed in the body during metabolism of sulfur-containing amino acids. Sulfite is oxidized to sulfate ion by sulfite oxidase (SOX, EC. 1.8.3.1). The aim of this study was to investigate the possible toxic effects of sulfite on neurons by measuring active avoidance learning in normal and SOX-deficient rats. For this purpose, male albino rats used in this study were divided into eight groups such as control group (C), sulfite group (25 mg/kg) (S), vitamin E group (50 mg/kg) (E), sulfite (25 mg/kg)+vitamin E group (50 mg/kg) (SE), SOX-deficient group (D), deficient+vitamin E group (50 mg/kg) (DE), deficient+sulfite group (25 mg/kg) (DS) and deficient+sulfite (25 mg/kg)+vitamin E group (50 mg/kg) (DSE). Sulfite-induced impairment of active avoidance learning in SOX-deficient rats but not in normal rats. Sulfite had no effect on hippocampus TBARS levels in SOX normal groups. In SOX- deficient rats, TBARS levels were found to be significantly increased with sulfite exposure. Vitamin E reversed the observed detrimental effects of sulfite in the SOX-deficient rats on their hippocampal TBARS but not on their active avoidance learning. In conclusion, sulfite has neurotoxic effects in sulfite oxidase deficient rats, but this effect may not depend on oxidative stress. D 2004 Elsevier Inc. All rights reserved.

Keywords: Food additives; Sulfite; Cognition; Oxidative stress; Rat

1. Introduction ability to react with several molecules of biological importance, including DNA [14,28]. It has been suggested Sulfur dioxide (SO2) and sulfites are added to foods for a that sulfite radicals which are intermediate products of sulfite variety of important technical purposes, including the metabolism play an important role in damaging nucleic acids control of enzymatic and non-enzymatic browning and and cause mutation [14,26]. These radicals can also react antimicrobial actions [31]. Considerable quantities of sulfite with proteins and lipids [15,36]. Both an endogenously are also generated in the body by normal catabolic generated and an exogenously intake of sulfite must be processing of sulfur-containing amino acids and other detoxified because of its toxic properties. For this purpose, sulfur-containing compounds [8,31]. Sulfite is a toxic mammalian tissues contain sulfite oxidase (SOX, EC. molecule and can react with a variety of humoral and 1.8.3.1), which catalyzes the oxidative detoxification of cellular components and can cause toxicity. There are sulfite [4]. Moreover, central nervous system effects have substantial data from in vitro studies that sulfites have the recently been demonstrated following sulfite exposure in rodents. Increased latencies in both visual and somatosensory evoked potentials have been reported following sulfite * Corresponding author. Tel.: +90 258 213 40 30; fax: +90 258 213 inhalation [2,21]. Physiological importance of this detox- 28 74. ification is also seen by consequences of SOX deficiency, E-mail address: [email protected] (V. Kqc¸qkatay). which is a genetically inherited disease [24]. This syndrome

0892-0362/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.ntt.2004.10.002 48 V. Ku¨c¸u¨katay et al. / Neurotoxicology and Teratology 27 (2005) 47–54 consisting of mental retardation, ataxia, seizures and learning process will be impaired. It was shown that dislocated lenses is associated with the absence of inorganic learning deficits are observed in hippocampus damaged sulfate in the urine. There is no effective treatment of this rats regardless of whether the learning task is spatial or non- disease, and patients who were diagnosed as having SOX spatial [5,6]. Active avoidance learning was chosen in this deficiency died during the first 2 years of their life [24]. One study as a non-spatial task to determine the possible toxic explanation for these contradictory results could be the effects of exogenous sulfite on hippocampus in normal and efficiency of sulfite detoxification in mammalian animal SOX-deficient rats and linked with TBARS measures in models. There are significant differences among species in hippocampus. their SOX activity [32]. Most notable is the difference between rat and man. It was shown that rat liver has about a 20-fold greater SOX activity than human liver [17]. 2. Materials and methods In contrast to results from in vitro experiments, numerous feeding studies with sulfite in mammals species have found 2.1. Animals no evidence of chronic sulfite toxicity [11,29]. Studies have also shown an inverse correlation between SOX levels in Adult male albino rats weighing 180–200 g were used liver and sulfite toxicity in several species of laboratory throughout all experiments. They were provided from our animals [32]. Although rats have been used predominantly own breeding colony and housed four to five per cage at 22– in the past for evaluation of sulfite toxicity, this species may 25 8C with a 12-h light/dark cycle. Akdeniz University not be the most appropriate model for the prediction of Animal Care and Usage Committee approved all exper- sulfite toxicity in man. It has been suggested that SOX- imental protocols used in our work. deficient rats might be used as a model for the prediction of sulfite toxicity in human [9,16]. 2.2. Reagents Activity of SOX in mammalian tissues exhibits a large distribution, and its activity in tissues shows significant The reagents used for hepatic SOX activity assay, lipid differences even in the same species. For instance, liver, peroxidation (TBARS) and plasma S-sulphonate levels were kidney and heart tissues have high SOX activities, whereas obtained from Merck (Darmstadt, Germany) and Sigma (St. brain, spleen and testis have very low activities [3,35]. Louis, MO, USA). Among cells, therefore, neuronal cells can be considered to be especially vulnerable to sulfite because of their low 2.3. Experimental design SOX activity. As discussed above, the most striking findings in SOX deficiency are neurological symptoms. Eighty male albino rats were randomly divided into two Moreover, it has been found that sulfite has not only toxic so as to form SOX-competent (SOXC, n=40) and SOX- effects in the rat mesencephalic cell line but also deficient groups (SOXD, n=40). Each group was further potentiates neurotoxic effect of peroxynitrite radicals divided into four subgroups: control (C), vitamin E (E), [27]. There is little information about the mechanism of sulfite (S), sulfite+vitamin E (SE) for SOXC groups and sulfite toxicity in neurons, but its detrimental effect on deficient (D), deficient+vitamin E (DE), deficient+sulfite neurons may involve formation of sulfur and oxygen (DS) and deficient+sulfite+vitamin E (DSE) for SOXD centered free radicals [1,30]. groups. Each subgroup contained 10 rats (Table 1). We hypothesized that sulfite has a free radical mediated Rats in SOXC groups were fed ad libitum with standard toxic effect on neurons. If this is true, hippocampus which is rat chow and tap water. Rats in S and SE groups were an important region in CNS related with learning will be treated with sulfite in their drinking water (sodium affected by free radical mediated sulfite toxicity and metabisulfite, Na2O5S2, 25 mg/kg/day). Vitamin E (a-

Table 1 Daily diet regimen and sulfite oxidase status in experimental groups of normal and sulfite oxidase deficient rats Group Sulfite oxidase Diet Drinking water supplementation Vitamin E/vehicle status (olive oil) treatment Control (C) Normal Standard rat chow None Vehicle Vitamin E (E) Normal Standard rat chow None Vitamin E (50 mg/kg) Sulfite (S) Normal Standard rat chow Sulfite (25 mg/kg) Vehicle Sulfite+vitamin E (SE) Normal Standard rat chow Sulfite (25 mg/kg) Vitamin E (50 mg/kg) Deficient (D) Deficient Low-Mo diet W (200 ppm) Vehicle Deficient+vitamin E (DE) Deficient Low-Mo diet W (200 ppm) Vitamin E (50 mg/kg) Deficient+sulfite (DS) Deficient Low-Mo diet W (200 ppm)+sulfite (25 mg/kg) Vehicle Deficient+sulfite+vitamin E (DSE) Deficient Low-Mo diet W (200 ppm)+sulfite (25 mg/kg) Vitamin E (50 mg/kg)

Tungsten (W; 200 ppm) was added to the rats drinking water in the form of sodium tungstate (NaWO4). Vehicle and vitamin E (50 mg/kg) were given by gastric gavage. Sulfite (25 mg/kg) was also added to drinking water in the form of sodium metabisulfite (Na2O5S2). V. Ku¨c¸u¨katay et al. / Neurotoxicology and Teratology 27 (2005) 47–54 49 tocopherol) was administered as an antioxidant to determine fixed intertrial interval of 20 s. In addition to the hanging importance of oxidative stress on sulfite toxicity. Vitamin E wire test to investigate motor activity, the activity level was dissolved in olive oil at dose of 50 mg/kg was given to rats also assessed by measuring the number of crossings in E and SE groups by gastric gavage. Olive oil was between the chambers when no shock was present (intertrial administered as a vehicle to the groups not receiving the crossing). This parameter was continuously recorded by the vitamin E doses. This dose of vitamin E was selected recorder unit of automated shuttle box during all expe- because of its antioxidant effect. The same sulfite and rimental period (50 trials) and referred to crossing between vitamin E intake protocol was carried out in DE, DS and the chambers when no shock present. The results were DSE groups. These treatments were continued throughout expressed as the mean percent avoidance responses for each the experiment (6 weeks). Rats in SOXD groups were made daily shuttle box session. deficient in SOX by the administration of a low molybdenum (Mo) diet (AIN 76, Research Dyets Inc USA) with 2.5. Biochemical measurements concurrent addition of 200 ppm tungsten (W) to their drinking water in the form of sodium tungstate (NaWO4). 2.5.1. Tissue and blood collection These treatments were started 3 weeks before the beginning At the end of 6th week, heparinized blood was collected of sulfite and vitamin E dosing regimens. In order to from the abdominal aorta of rats under urethane anesthesia commence the study, at the end of the 3rd week, SOX and used for the determination of plasma S-sulphonate activity in the livers of SOXD rats was measured to confirm levels. The animals were then killed by exsanguination. that the rats were SOX deficient. Brain and liver tissues were removed immediately. The livers were used for the assay of SOX activity and the 2.4. Behavioral tests hippocampi were analysed for TBARS. The tissues were stored at À80 8C for later analysis. 2.4.1. Motor function testing Before starting active avoidance responses measure- 2.5.2. Tissue preparation for enzyme and Lipid peroxidation ment, we tested motor function of the rats by using the assays Hanging Wire Test. It is known that almost every behavior Hippocampi from the removed brains were carefully requires the animal to move. If motor functions are dissected to avoid tissue damage by using the anatomic impaired, the animal will not be able to perform complex description of Paxinos and Watson. Hippocampi and about tasks such as active avoidance, simply because it cannot 1 g of frozen livers were homogenized at high speed with cross between the compartments. In this study, the the homogenizer placed in the chipped ice in 50 mM Hanging wire test was used as a simple but important phosphate buffer. In order to remove precipitates, hippo- test to confirm that the possible alterations in active campal homogenates were centrifuged at 12000Âg for 10 avoidance task are not due to motor deficits. A standard min at +4 8C and liver homogenates were centrifuged at wire cage lid was used for this purpose. The rats were 2100Âg for 10 min at +4 8C. placed on the top of a wire cage lid and the lid was lightly shaken three times to cause the rats to grip the wires. The 2.5.3. Determination of hippocampus lipid peroxidation lid was then turned upside down and held at a height about levels 20 cm above the cage litter. The latency to fall off the wire Hippocampus is known to have a very important role in lid was measured. A 60-s cut off time was used for the the learning process. Detrimental effects of oxidative stress standard test session. on hippocampus may impair learning and memory. Hippo- campus TBARS levels, an indicator of lipid peroxidation, 2.4.2. Active avoidance responses and the activity levels were measured to determine if sulfite causes lipid Active avoidance behavior was studied in the middle of peroxidation and in this manner induces neurotoxicity the 6th week. The animals were trained by using an according to the method of Winterbourn et al [34]. Briefly, automated shuttle box (Ugo Basile 7502). The shuttle box after centrifugation of hippocampi homogenates, super- was divided into two chambers of equal size by a stainless- natants were mixed with 500 Al thiobarbituric acid (1% in steel partition with a gate providing access to the adjacent NaOH 50 mM) and 500 Al of HCl% 25. The samples were compartment. A light (60 W, 5 s) was switched on then heated in a boiling water bath for 10 min and, after alternately in the two compartments and used as a cooling, were extracted with 1.5 mL of butanol. The conditioned stimulus (CS). The CS preceded the onset of mixture was centrifuged at 12000Âg for 10 min at +4 8C the unconditioned stimulus (US) by 5 s. The US was an and the absorbance of the supernatant was determined at electric shock (1 mA for 5 s) applied to the grid floor. If the 532 nm by using a spectrophotometer (Schimadzu UV animal avoided the US by running into the dark compart- 1600). The medium used for tissue preparations contains a ment within 5 s after the onset of the CS, the microprocessor chelating agent (EDTA) and an antioxidant (butylated recorder unit of the shuttle box recorded an avoidance hydroxytoluene). The results were expressed as nmol/mg response. Each rat was given 50 trials daily for 5 days with a protein. 50 V. Ku¨c¸u¨katay et al. / Neurotoxicology and Teratology 27 (2005) 47–54

2.5.4. Plasma S-sulphonate analysis 3. Results Plasma S-sulphonate levels were measured as a biomarker of ingested sulfite according to the method of 3.1. General animal health Gunnison et al. [12]. Briefly, one milliliter of plasma was mixed with 0.2 ml of a solution containing 0.027 mmol All the animals were outwardly healthy. There were no NaOH and 125 mmol KCN. The mixture was incubated at signs of toxicity in any of the experimental groups. 35F1 8C under nitrogen for 1 h. Following incubation, the Treatment groups in both series (SOXC and SOXD) mixture was cooled in ice and transferred to a cellulose exhibited similar weight gains and survival. dialysis bag (Fisher Scientific, Pittsburgh, PA) and dialyzed at 4 8C against 5 ml of 10 mmol glycine–NaOH 3.2. Hepatic SOX activity buffers at pH 10.2 for 2.5 h. After dialyzing, 1.4 ml of dialysate was analyzed for sulfite as follows: 0.2 ml of Fig. 1 shows hepatic SOX activity at the end of the 0.15 N HCl, 0.2 ml of 180 mmol sodium tetrachloromer- 6th week. Hepatic SOX activity can be a representative curate (absorbing solution), 0.2 ml of distilled water, 0.2 for status of SOX in whole organism because of its high ml of Pararosaniline (PRA) reagent and 0.2 ml of level of SOX activity compared to other organs. formaldehyde reagent were added in the dialysate in that Induction of SOX deficiency by maintaining on the order and the solution was mixed. After mixing for 20 low-Mo diet with W supplementation (200 ppm) was min, absorbance of mixture was measured at 560 nm using very effective. Hepatic SOX activity in SOXD groups a plasma dialysate sample treated with hydrogen peroxide was significantly reduced (about 15-fold) compared to (H2O2) as a blank. Sulfite level in the dialysate was SOXC groups by maintaining high-W/low-Mo regimen estimated from a sulfite standard curve and total amount of [ F(7,18)=11,49, pb0.0001]. Sulfite and/or vitamin E sulfite released from plasma proteins was calculated. The treatment did not affect SOX activity in both SOXC results were expressed as nmol/ml. and SOXD groups.

2.5.5. Hepatic SOX assay 3.3. Plasma S-sulphonate levels Hepatic SOX actitivity was determined as an indicator of SOX status of the body according to the method of Fig. 2 summarizes the concentrations of plasma S- Cohen et al [4]. An aliquot of supernatant was mixed with sulphonate in all groups. Detection of elevated levels of 5% Triton X-100 and further diluted (1:10) with 50 mmol plasma S-sulphonate indicates that it has recently been phosphate buffer (pH 7.4). This diluted mixture was used exposed to sulfite. Plasma S-sulphonate level in both SOX for measuring SOX activity at room temperature by competent and SOX-deficient rats was found to be increased monitoring the reduction of cytochrome c at 550 nm. significantly by adding sulfite in their drinking water The mixture was added to the cuvette containing 10 mmol sodium sulfite, 0.2 mmol cytochrome c, Triton X-100 (% 5), 100 mmol Tris–HCl (pH 8.5) and KCN (10 mmol in Tris–HCl) in a final volume of 2.5 ml. The slow non- enzymatic rate of reduction of cytochrome c was first recorded and its rate was subtracted from the recorded total rate. One unit of SOX activity was defined as the amount of enzyme, which caused an absorbance change of 0.1/min under these conditions. The results were expressed as units/mg protein.

2.5.6. Protein determinations The protein concentrations were evaluated by the method of Lowry et al [22].

2.6. Statistical analysis Fig. 1. Hepatic sulfite oxide status in experimental groups of normal and sulfite oxidase deficient rats on the 6th week. Sulfite oxidase deficiency was formed by feeding the rats with low-Mo/high-W regimen. Vitamin E Results were expressed as meansFS.E. Statistical com- (50 mg/kg) was applied by gastric gavage. Sulfite (25 mg/kg) was added parisons for repeated measurements between groups were to the rats drinking water. Open bars represent sulfite oxidase competent made by repeated-measures ANOVA for multiple com- rats and dark bars represent sulfite oxidase deficient rats. Values are F b parisons, followed by TUKEY post hoc test. Differences expressed as means S.E.M. for 10 rats in each group. *p 0.0001 when compared with sulfite oxidase competent rats. SOXC=sulfite oxidase between mean values in study groups were evaluated by a competent, SOXD=sulfite oxidase deficient, C=control groups, E=groups one-way ANOVA followed by TUKEY post hoc test. p treated with vitamin E, S=groups treated with sulfite, SE=groups treated values b0.05 were accepted as statistically significant. with sulfite+vitamin E. V. Ku¨c¸u¨katay et al. / Neurotoxicology and Teratology 27 (2005) 47–54 51

Vitamin E. There were no statistical differences between the D and DE groups and the SOXC groups on active avoidance response patterns for each daily session. Given alone, sulfite significantly impaired avoidance per- formance starting from the 4th session in the DS group [ F(7,1)=5,53, pb0.05]. The impairing effect was also observed on day 3 when sulfite was combined with Vitamin E in the DSE group [ F(1,71)=133,8 pb0.005]. The DSE group was not statistically worse than the DS group on active avoidance for each daily session.

3.6. Lipid peroxidation levels in hippocampus of SOX Fig. 2. S-sulphonate concentrations in the plasma of normal and sulfite normal and deficient rats oxidase deficient rats. At the end of the 6th week, rats were killed by exanguination and plasma S-sulphonate concentration was determined. Fig. 4 summarizes the hippocampus lipid peroxidation Open bars represent sulfite oxidase competent rats and dark bars level in experimental groups of normal and SOX-deficient represent sulfite oxidase deficient rats. Values are expressed as means FS.E.M. for 10 rats in each group. *pb0.01 as compared with C and E rats. TBARS levels were measured in order to assess groups, #pb0.001 as compared with C and E groups, ##pb0.01 as whether sulfite treatment causes lipid peroxidation in compared with D and DE groups. SOXC=sulfite oxidase competent, hippocampus. Hippocampus TBARS levels were not SOXD=sulfite oxidase deficient, C=control groups, E=groups treated found to be affected by sulfite (S group) and vitamin E with vitamin E, S=groups treated with sulfite, SE=groups treated with (E group) alone or by their co-administration (SE group) sulfite+vitamin E. in SOXC rats. In SOXD rats, significant increment in hippocampus TBARS levels was observed as a result of [ F(3,19)=9,23, pb0.001]. Vitamin E had no effect on this sulfite administration in the DS group [ F(7,14)=5,16, parameter in both experimental series. Although it was not pb0.05]. This effect of sulfite was reversed by vitamin E statistically significant, plasma S-sulphonate level in D and DE groups yielded a trend of increment compared to C and E groups.

3.4. Results of motor function testing

All rats in this experiment could easily hang upside down from the wire cage lid for 60 s. All of them had a 60-s cut off time for hanging wire grip test. This result indicates that they did not have any motor coordination and balance abnormalities on this test. Data are not shown since the rats in all groups were able to achieve the test.

3.5. Effects of oral administration of sulfite on activity level and active avoidance responses in SOX normal and deficient rats Fig. 3. Mean percent active avoidance responses in rats with normal The number of crossings from one compartment of the sulfite oxidase activity and deficient sulfite oxidase activity. On the 6th week, the rats in each group were given 50 training trials per day for 5 shuttle box to the other showed that all experimental consecutive days. An avoidance response was recorded if the animal groups exhibited same crossing activity during 50 trials. avoided the unconditioned stimulus (US) by running into the dark In other words, the activity level was not affected in any compartment within 5 s after the onset of the conditioned stimulus (CS). of the experimental groups (data not shown). Fig. 3 The active avoidance responses of sulfite oxidase normal groups were reports the percent active avoidance responses for each not affected by treatment of sulfite and/or vitamin E. Sulfite significantly reduced mean percent active avoidance responses, starting from the 4th daily shuttle-box session in SOXC and SOXD groups. day in SOX deficient rats and this decrement was not improved by There was no difference among the SOXC groups in vitamin E treatment. Given alone, vitamin E had no effect on this active avoidance responses during the daily sessions. All parameter. Values are expressed as meansFS.E.M. for 10 rats in each the animals in these groups started learning to avoid the group. *pb0.05 as compared with D, DE and SOXC groups, #pb0.005 US by running into the other compartment on the 2nd day as compared with D, DE and SOXC groups. C=control group, E=group treated with vitamin E, S=group treated with sulfite, SE=group treated and reached about 40% success rate of avoidance with sulfite+vitamin E. D=deficient control group, DE=deficient group response on the 5th day. Their active avoidance perform- treated with vitamin E, DS=deficient group treated with sulfite, ances were not affected by treatment of sulfite and/or DSE=deficient group treated with sulfite+vitamin E. 52 V. Ku¨c¸u¨katay et al. / Neurotoxicology and Teratology 27 (2005) 47–54

Recognized as Safe" (GRAS) for use in foods and drugs. However, GRAS status of sulfiting agents is currently under reconsideration because of their potential harmful effects. Although the toxic effects of sulfur dioxide in the ambient air have been studied extensively, toxicity of ingested sulfite gained relatively less attention. In retrospect, Kochen [20] first described hypersensitivity to foodborne sulfites in 1973. Many studies have indicated toxicities of ingested sulfite since that time. Although most of these adverse effects are of an allergic nature [13], it was shown that there are also toxic effects of sulfite on many cellular components, including DNA [28]. Regardless of the source, sulfite Fig. 4. Hippocampus TBARS values in normal and sulfite oxidase deficient is oxidized to sulfate, a reaction catalyzed by SOX in order groups. Hippocampus TBARS levels were not found to be different from to protect the cell from its damaging effect. This enzyme is a each other in sulfite oxidase normal rats. Increased TBARS levels were molybdohemoprotein and is located in the mitochondrial only observed in sulfite oxidase deficient group (DS), which received intermembrane space [19]. In case of genetic deficiency of sulfite and this increment was reversed by Vitamin E treatment (DSE). Values are expressed as meansFS.E.M. for 10 rats in each group. *pb0.05 this enzyme in humans, sulfite is not converted to sulfate. as compared with the other groups. SOXC=sulfite oxidase competent, The importance of this conversion is seen by the con- SOXD=sulfite oxidase deficient, C=control groups, E=groups treated with sequences of SOX deficiency in humans. The main clinical vitamin E, S=groups treated with sulfite, SE=groups treated with manifestations of SOX deficiency are neurological abnor- sulfite+vitamin E. malities such as severe mental retardation, seizures, spastic quadriparesis, dislocated lenses, progressive destruction of (DSE group). Vitamin E alone had no effect on this brain tissue and early death [24]. These findings highlight parameter (DE group). the potential neurotoxicity of sulfite. In the present study, hepatic SOX activity was measured as an indicator of SOX status in the body. The results of this 4. Discussion study clearly demonstrate that feeding the rats with high-W/ low-Mo diet effectively reduced SOX activity. Rats are The results of this study show that there is a marked known to have a 10- to 20-fold greater SOX activity than detrimental effect of sulfite on active avoidance responses in humans [17]. Fig. 1 shows that the SOXD rats have a 15- SOX-deficient rats but not in normal rats. Although this fold decrease in SOX activity than SOXC. This makes SOX effect of sulfite seems to be associated with elevated activity in the SOXD rats in the range of human SOX TBARS levels in hippocampus of SOX-deficient rats, the activity. For this reason, these SOX-deficient rats were used antioxidant treatment (vitamin E) did not reverse the in this experiment to be a more appropriate model for decrement of active avoidance learning (DSE group). It studying sulfite toxicity in man due to manTs reduced SOX can be concluded that sulfite may act on central nervous activity. The results of our experiments have shown an system as a toxic agent and this detrimental effect of sulfite inverse correlation between the levels of SOX activity in the may not depend on oxidative stress in hippocampus of liver and plasma S-sulphonate level that is an indicator of SOX-deficient rats. ingested sulfite. There was a small but non-significant The acceptable daily intake (ADI) for sulfites was increment of plasma S-sulphonate level in deficient controls established as 0–0.7 mg/kg/body weight by The Joint (D group). This non-significant little increase in D group FAO/WHO Expert Committee on Food Additives [10]. was probably due to diminished oxidizing capacity for The ADI value was based on long-term studies in rats, endogen-produced sulfite. In accordance with previous including a three-generation work of reproductive toxicity, studies [11,12], we also demonstrated a statistically signifi- with a NOEL (no-observed-effect level) of 0.25% Na2S2O5 cant accumulation of S-sulphonate in plasma of normal and in the diet, equivalent to 70 mg/kg body weight per day of SOX-deficient rats receiving sulfite. SO2 equivalents [33]. By applying typical 100-fold safety Although there is little information about the mechanism factor, ADI value was determined for humans as 0–0.7 mg/ of sulfite toxicity on central nervous system, sulfur- and kg. However, the daily intake of sulfite may not be in oxygen-centered free radicals might play an important role agreement with this value in many cases. Studies have in the development of this phenomenon [1,30]. Previous shown that it is possible to consume 180–200 mg/body studies suggested that sulfite can be oxidized non-enzymati- weight from foods and beverages in a single day or meal cally or by various peroxidases [23,27]. Non-enzymatic [13,31]. For this reason, higher level of sulfite (25 mg/kg) autooxidation or peroxidase-catalyzed oxidation of sulfite was chosen in this study. Sulfite compounds (sulfur dioxide, occurs via one electron oxidation. In this case, a sulfur sodium sulfite, sodium and potassium bisulfite, and sodium trioxide radical which is an intermediate product of sulfite and potassium metabisulfite) have been listed as "Generally metabolism is produced from sulfite. It has been proposed V. Ku¨c¸u¨katay et al. / Neurotoxicology and Teratology 27 (2005) 47–54 53 that sulfur trioxide radical may be involved in some toxic In conclusion, we suggest that sulfite toxicity found in effects of sulfite such as destruction of amino acids, SOX-deficient rats can be attributed to sulfite alone and/or impairment of DNA synthesis and increment of lipid to cysteine-S-sulfate. Further studies are required to clarify peroxidation [14,15,29,36]. these aspects. Despite these reports, we have found that there was no significant effect of sulfite treatment on the level of hippocampus TBARS in SOX normal groups. This finding Acknowledgements does not agree with previous studies, which report that formation of sulfur- and oxygen-centered free radicals may This study was carried out as part of a PhD thesis by V. play important role in sulfite toxicity. In addition, we also Kucukatay presented to Akdeniz University Health Sciences have found no adverse effect of sulfite on the percent active Institute. avoidance responses in rats with normal SOX activity. This work was supported by a grant from Akdeniz These results obtained from rats with normal SOX activity University Research Fund 2002.0122.04. could be due to the efficiency whereby the SOX enzyme catalyses oxidation of sulfite to sulfate. Indeed, previous studies have shown that rat liver possessed approximately 10–20 times higher hepatic SOX activity than human liver References [17]. Furthermore, an inverse correlation has been found [1] Z. Abedinzadeh, Sulfur-centered reactive intermediates derived from between the levels of SOX activity in the liver and sulfite the oxidation of sulfur compounds of biological interest, Can. J. toxicity in some laboratory animal species [32]. Our results Physiol. Pharm. 79 (2001) 166–170. also demonstrated that high levels of sulfite can be tolerated [2] A. Agar, V. Kucukatay, P. Yargicoglu, B. Aktekin, D. Kipmen- by the rats with normal SOX activity without any sign of Korgun, S. Gumuslu, C. Apaydin, The effect of sulfur dioxide toxicity. On the other hand, the results of percent active inhalation on visual evoked potentials, antioxidant status, and lipid peroxidation in alloxan-induced diabetic rats, Arch. Environ. Contam. avoidance responses and hippocampus lipid peroxidation Toxicol. 39 (2000) 257–264. levels were found to be significantly different in SOX- [3] F. Cabre, C. Marin, M. Cascante, E.I. Canela, Occurrence and deficient rats exposed to sulfites compared to normal rats. comparison of sulfite oxidase activity in mammalian tissues, Biochem. Sulfite treatment significantly increased TBARS levels of Med. Metabol. Biol. 43 (1990) 159–162. hippocampus in SOX-deficient rats (DS group) and sig- [4] H.J. Cohen, I. Fridovich, Hepatic sulfite oxidase. Purification and properties, J. Biol. Chem. 246 (1971) 359–366. nificantly reduced percent active avoidance responses [5] D.M. Compton, Encoding of a nonmonotonic serial pattern: role of starting from the 4th session in DS group. Although the the dorsal hippocampus and amygdala, Physiol. Behav. 53 (1993) results of this study indicating significant elevation of lipid 657–665. peroxidation in hippocampus and reduction in percent active [6] D.M. Compton, W.F. McDaniel, K.L. Dietrich, Non-spatial learning avoidance responses in SOX-deficient groups (DS group) is following posterior parietal or hippocampal lesions, NeuroReport 16 (1994) 2189–2192. in agreement with free radical mediated toxicity of sulfite, [7] M. Conradt, W. Stoffel, Functional analysis of the high affinity, Na(+)- the reduced percent active avoidance responses was not dependent glutamate transporter GLAST-1 by site-directed muta- reversed by antioxidant therapy in DSE group. In other genesis, J. Biol. Chem. 270 (1995) 25207–25212. words, toxicity was still observed. [8] A.J. Cooper, Biochemistry of sulfur-containing amino acids, Annu. The results of sulfite toxicity in SOX-deficient rats derive Rev. Biochem. 52 (1983) 187–222. [9] L. Dulak, G. Chiang, A.F. Gunnison, A sulphite oxidase-deficient rat from a mechanism that does not involve oxidative stress. It model: reproductive toxicology of sulphite in the female, Food Chem. can be suggested that cysteine-S-sulfate, which is a brain Toxicol. 22 (1984) 599–607. damaging metabolite in SOX deficiency may be responsible [10] FAO/WHO, Expert Committee on Food Additives. Toxicological for observed detrimental effects of sulfite. Cysteine-S- Evaluation of Certain Food Additives with a Review of General sulfate is not normally present in human body [25]. This Principle and of Specification. 17th report, Food an Agriculture Organization, Rome, 1974. metabolite is produced by sulfite reacting with free cysteine [11] A.F. Gunnison, E.D. Palmes, Persistence of plasma S-sulfonates amino acids in SOX deficiency [18]. It has structural following exposure of rabbits to sulfite and sulfur dioxide, Toxicol. features in common with the excitotoxic amino acids such Appl. Pharmacol. 24 (1973) 266–278. as glutamic, aspartic, cysteic and homocysteic acids [25].It [12] A.F. Gunnison, D.W. Jacobsen, Sulfite hypersensitivity. A critical is suggested that these excitotoxic amino acids are not review, CRC Crit. Rev. Toxicol. 17 (1987) 185–214. T [13] A.F. Gunnison, A. Sellakumar, D. Currie, E.A. Snyder, Distribution, normally found in the neuron s external milieu except in metabolism and toxicity of inhaled sulfur dioxide and endogenously tiny quantities for transmitter purposes. Indeed, a high generated sulfite in the respiratory tract of normal and sulfite oxidase- affinity uptake system in brain has been shown for deficient rats, J. Toxicol. Environ. Health 21 (1987) 141–162. glutamate and aspartate but not for cysteine-S-sulfate [7]. [14] H. Hayatsu, R.C. Miller Jr., The cleavage of DNA by the oxygen- Although we did not measure the levels of cysteine-S- dependent reaction of bisulfite, Biochem. Biophys. Res. Commun. 46 (1972) 120–124. sulfate in our experiments, observed toxicity of sulfite in [15] B. Inouye, K. Morita, T. Ishida, M. Ogata, Cooperative effect of SOX-deficient rats can be attributed to its possible high sulfite and vanadium compounds on lipid peroxidation, Toxicol. Appl. level in these rats. Pharmacol. 53 (1980) 101–107. 54 V. Ku¨c¸u¨katay et al. / Neurotoxicology and Teratology 27 (2005) 47–54

[16] J.L. Johnson, K.V. Rajagopalan, Purification and properties of sulfite [27] M. Reist, K.A. Marshall, P. Jenner, B. Halliwell, Toxic effects oxidase from human liver, J. Clin. Invest. 58 (1976) 543–550. of sulphite in combination with peroxynitrite on neuronal cells, [17] J.L. Johnson, K.V. Rajagopalan, H.J. Cohen, Molecular basis of the J. Neurochem. 71 (1998) 2431–2438. biological function of molybdenum. Effect of tungsten on xanthine [28] E. Rencuzogullari, H.B. Ila, A. Kayraldiz, M. Topaktas, Chromosome oxidase and sulfite oxidase in the rat, J. Biol. Chem. 249 (1974) aberrations and sister chromatid exchanges in cultured human 859–866. lymphocytes treated with sodium metabisulfite, a food preservative, [18] B. Kagedal, M. Kallberg, B. Sorbo, A possible involvement of Mutat. Res. 490 (2001) 107–112. glutathione in the detoxication of sulfite, Biochem. Biophys. Res. [29] D. Ribera, D. Jonker, J.F. Narbonne, J. O’Brien, E. Antignac, Absence Commun. 136 (1986) 1036–1041. of adverse effects of sodium metabisulphite in manufactured [19] C. Kisker, H. Schindelin, A. Pacheco, W.A. Wehbi, R.M. Garrett, K.V. biscuits: results of subacute (28-days) and subchronic (85- Rajagopalan, J.H. Enemark, D.C. Rees, Molecular basis of sulfite days) feeding studies in rats, Food Addit. Contam. 18 (2001) oxidase deficiency from the structure of sulfite oxidase, Cell 91 (1997) 103–114. 973–983. [30] X. Shi, Generation of SO3.À and OH radicals in SO3(2À) reactions [20] J. Kochen, Sulfur dioxide, a respiratory tract irritant, even if ingested, with inorganic environmental pollutants and its implications to Pediatrics 52 (1973) 145–146. SO3(2À) toxicity, J. Inorg. Biochem. 56 (1994) 155–165. [21] V. Kucukatay, A. Agar, P. Yargicoglu, S. Gumuslu, B. Aktekin, [31] S.L. Taylor, N.A. Higley, R.K. Bush, Sulfites in foods: uses, analytical Changes in somatosensory evoked potentials, lipid peroxidation, and methods, residues, fate, exposure assessment, metabolism, toxicity, antioxidant enzymes in experimental diabetes: effect of sulfur dioxide, and hypersensitivity, Adv. Food Res. 30 (1986) 1–76. Arch. Environ. Health 58 (2003) 14–22. [32] I. Tejnorova, Sulfite oxidase activity in liver and kidney tissue in [22] O.H. Lowry, N.J. Rosenbrough, A.L. Far, R.J. Randel, Protein five laboratory animal species, Toxicol. Appl. Pharmacol. 44 (1978) measurement with folin–phenol reagent, J. Biol. Chem. 193 (1951) 251–256. 265–275. [33] H.P. Til, V.J. Feron, A.P. De Groot, The toxicity of sulphite: I. Long- [23] C. Mottley, T.B. Trice, R.P. Mason, Direct detection of the sulfur term feeding and multigeneration studies in rats, Food Cosmet. trioxide radical anion during the horseradish peroxidase–hydrogen Toxicol. 10 (1972) 291–310. peroxide oxidation of sulfite (aqueous sulfur dioxide), Mol. Pharma- [34] C.C. Winterbourn, J.M. Gutteridge, B. Halliwell, Doxorubicin-

col. 22 (1982) 732–737. dependent lipid peroxidation at low partial pressures of O2, J. Free [24] S.H. Mudd, F. Irreverre, L. Laster, Sulfite oxidase deficiency in Radic. Biol. Med. 1 (1985) 43–49. man: demonstration of the enzymatic defect, Science 156 (1967) [35] W.H. Woo, H. Yang, K.P. Wong, B. Halliwell, Sulphite oxidase gene 1599–1602. expression in human brain and in other human and rat tissues, [25] J.W. Olney, C.H. Misra, T. de Gubareff, Cysteine-S-sulfate: brain Biochem. Biophys. Res. Commun. 305 (2003) 619–623. damaging metabolite in sulfite oxidase deficiency, J. Neuropathol. [36] S.F. Yang, Destruction of tryptophan during the aerobic oxidation of Exp. Neurol. 34 (1975) 167–177. sulfite ions, Environ. Res. 6 (1973) 395–402. [26] T. Ozawa, A. Hanaki, Spin-trapping of sulfite radical anion, SO3À., by a water-soluble, nitroso-aromatic spin-trap, Biochem. Biophys. Res. Commun. 142 (1987) 410–416.