No direct hepatotoxic potential following a multiple-low doseparaquat exposure in rat as related to its bioaccumulation

Varaporn Podprasart a, Jutamaad Satayavivad a,b, Suda Riengrojpitak c,Prapin Wilairat d, Winai Wananukul e, Pranee Chavalittumrong f,

Songpol Chivapat f, Krongtong Yoovathaworn a,b,∗a Graduate Program in Toxicology, Faculty of Science, Mahidol University, Bangkok 10400, Thailand

b Department of Pharmacology, Faculty of Science, Mahidol University, Bangkok 10400, Thailandc Department of Pathobiology, Faculty of Science, Mahidol University, Bangkok 10400, Thailand

d Department of Chemistry, Faculty of Science, Mahidol University, Bangkok 10400, Thailande Division of Pharmacology and Toxicology, Department of Medicine, Faculty of Medicine, Ramathibodi Hospital,

Mahidol University, Bangkok 10400, Thailandf Department of Medical Science, Ministry of Public Health, Bangkok 11000, Thailand

Received 23 November 2006; received in revised form 8 March 2007; accepted 8 March 2007Available online 13 March 2007

Abstract

Paraquat (PQ) is a well-known toxic bipyridyl herbicide commonly used in agricultural countries. Pulmonary toxicity is the maincause of death but damage to other organs has also been reported. PQ is also classified as a “direct hepatotoxicant” following anacute high dose exposure. The evidence of multi-low dose toxicity of PQ was scarce. Therefore, the aim of this study was to examinethe effect of multiple low doses of PQ on the liver function and xenobiotic-metabolizing enzyme activities including CYP1A1,2E1, and 3A4, and to correlate the effects with its tissue accumulation. PQ, at the dose range 4.0–6.0 mg/kg day, was subcuta-neously administered to male Wistar rats for seven consecutive days. The prominent feature of toxic response was lung toxicity.Interestingly, PQ-treatment caused a dose- and time-dependent reduction of plasma transaminase activity. Hypobilirubinemia andhypoalbuminemia were also observed without significant alteration in the liver morphology. Of all the xenobiotic-metabolizingenzymes being studied, only the activity of CYP1A1-related 7-ethoxyresorufin-O-deethylase was reduced following the highestdose of PQ administration. Plasma and tissue concentrations and accumulation of PQ analyzed by HPLC were dose-dependentshowing much higher concentration (approximately 13 times) in the lung than that in the liver whereas it was undetectable in theplasma at the same time point.

It can be concluded that multi-low dose PQ might affect certain synthetic function of the liver or activity of some hepaticxenobiotic-metabolizing enzymes. Minimal PQ accumulation in the liver is one of the explanations for the lack of cytotoxic hepatic

injury in this study. Plasma PQ concentration may not be a good marker of exposure and toxicity after a prolonged exposure to PQ.© 2007 Elsevier Ireland Ltd. All rights reserved.

Keywords: Paraquat; Hepatotoxicity; Enzymes; Accumulation; CYP

∗ Corresponding author at: Department of Pharmacology, Faculty of Science, Mahidol University, Rama 6 Road, Bangkok 10400, Thailand.Tel.: +66 2 2015650; fax: +66 2 3547152.

E-mail addresses: sckyo@mahidol.ac.th, ky pharmaco@yahoo.com (K. Yoovathaworn).

0378-4274/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.toxlet.2007.03.006

1. Introduction

Paraquat (PQ; 1,1′-dimethyl-4,4′-bipyridinium ion,methyl viologen), a quick acting and non-selective con-tact herbicide, is one of the most extensively usedpesticides for the control of broad-leaved weeds andgrasses in crop land and aquatic area. PQ is one of themost commonly used agents that produce toxicity inhuman (Teixeira et al., 2004; Hong et al., 2000). The hall-mark target organ in systemic PQ poisoning is the lung,however, death is usually the result of multisystem failureincluding acute pulmonary edema, hepatorenal failure,and myocarditis (Conradi et al., 1983). Much atten-tion has been focused on acute PQ toxicity. However,occupational exposure especially by workers in the agri-cultural areas probably occurs at low doses of multipleexposures. The toxic effects of PQ during occupationalexposure have been mentioned (Canton and Young-Holt,2002).

Among toxic effects of high dose PQ in humans andanimals, hepatotoxicity has been one of them (Takegoshiet al., 1988; Mutsumori et al., 1985; Mullick et al.,1981) especially in a large number of cases with suici-dal attempt (Hong et al., 2000; Erickson et al., 1997).Sufficient dermal exposure to PQ also led to hepaticinjury (Bataller et al., 2000). PQ-induced hepatotoxic-ity in patients is characterized by jaundice and abnormalliver function tests illustrated as increases in serum lev-els of LDH, SGOT, SGPT, alkaline phosphatase, totalbilirubin, as well as a decrease in serum proteins, albuminand globulin (Hong et al., 2000; Erickson et al., 1997).The signs of liver injury, usually observed in severePQ-poisoned cases, were cholestasis and cholangiocel-lular damage involving the small and medium-sized bileducts in the portal areas (Conradi et al., 1983; Mullicket al., 1981). Portal infiltration with neutrophil, plasmacells and eosinophil was noted (Kodagoda et al., 1973).Recently, some researchers classified PQ as an agentinducing cholestasis (Meeks et al., 2000). The studyon the toxic effect of PQ after long-term exposure inhumans is lacking. A multi-low dose PQ exposure modelused herein might resemble occupational exposure. Thedoses of PQ chosen for this study were based on thatof the previous report from our group (Suriyo, 2001).They reported a decrease in the body weight gain as wellas an increase in the hepatic enzyme markers using thedosage regimen of 3.0 mg/kg BW day, sc, 5 days/weeks,for 3 months.

The present study was, therefore, conducted to inves-tigate the multi-dose effect of PQ on the liver functionand xenobiotic-metabolizing enzyme activities, and tocorrelate the effects with its tissue accumulation.

2. Materials and methods

2.1. Materials

Paraquat dichloride (methyl viologen) and enzymaticreagent kit for determination of plasma albumin were obtainedfrom Sigma (USA). All other chemicals, if not specified, wereat least analytical grade.

2.2. Animals

Male Wistar rats (120–140 g) were purchased from theNational Laboratory Animal Centre, Thailand. They werehoused individually with food and water available ad libitumin a room maintained under constant environmental conditionswith 12-h light–dark cycle. Rats were habituated to the vivar-ium for at least 1 week prior to commencement of experiments.All animals were cared for and treated in accordance withNIH and Mahidol University Animal Care and Use Committeeguidelines.

2.3. Effect of PQ on the plasma biochemistry

2.3.1. Animal treatmentDose–response study. Forty rats were randomly allocated

into four groups and treated with daily subcutaneous doses of0 (water control), 4.0, 5.0, and 6.0 mg of PQ/kg day for sevenconsecutive days. The body weight and clinical signs of toxicitywere observed.

Time-course study. Twenty-four rats were divided into twogroups: control and PQ (4.0 mg/kg day, sc, 10 days)-treatedgroups. They were treated for 3, 7, and 10 days.

2.3.2. Sample collection and biochemical analysesAt the end of each experimental period, blood sam-

ples were obtained for blood chemistry study. Aspartateaminotransferase (AST), alanine aminotransferase (ALT),alkaline phosphatase (AP), total bilirubin, BUN, and creati-nine were determined using an automatic hematology analyzer(Roche/Hitachi, Japan). Total protein and albumin concentra-tions were determined according to the methods of Lowry etal. (1951) and Doumas et al. (1971), respectively. Liver andlung specimens were collected for morphological evaluationusing standard method for H&E staining. The remainders ofthe plasma and tissue samples were kept frozen at −80 ◦C.

2.4. Effect of PQ on xenobiotic-metabolizing enzymeactivities

Hepatic post-mitochondrial supernatant and microsomalfractions were prepared for the determination of CYP activ-

ities (Lake, 1987). The O-deethylation of ethoxyresorufin,as CYP1A1 probe, was assayed by the fluorometric method(Prough et al., 1978). The p-nitrophenol hydroxylation,as CYP2E1 probe, was determined spectrophotometrically(Koop, 1986). Formaldehyde formation was measured for

tCc

2

tvpdw1SHCadTefs

2a

m2Ppds

2

f

creatinine levels, but not the total plasma protein level,

TT

P

BAAATATBC

Pnc

he determination of erythromycin N-demethylase activity,YP3A4 probe (Nash, 1953). Liver microsomal protein con-entration was also determined (Lowry et al., 1951).

.5. Bioaccumulation of PQ in the biological samples

The frozen plasma and homogenized tissue samples werehawed and prepared for the determination of PQ content. Ethyliologen, as an internal standard, was added into the sam-le prior to the process of extraction according to the methodescribed by Fuke et al. (2002). The obtained alkaline eluentsere introduced onto a C18 �BondapakTM column (10 �m,25 A, 3.9 mm × 300 mm) connecting with an Aligent 1100eries (Agilent Technologies, Germany) HPLC system. ThePLC conditions for separation of PQ were as described byorasaniti et al. (1990). The calibration curves of PQ in plasmand tissues were constructed by plotting PQ and internal stan-ard peak area ratio against the respective PQ concentrations.he standard curve data was subjected to least square lin-ar regression analysis and the resulting equation was utilizedor the calculation of PQ concentration in plasma and tissueamples.

.6. Study on the plasma and tissue PQ concentrationsfter single and multi-dose exposures

For a single-low dose PQ exposure (6.0 mg/kg, sc), the ani-als were sacrificed at pre-assigned time points: 0.5, 1, 3, and

4 h post-dose. For multi-dose PQ exposure, the same dose ofQ was administered for 7 days and sacrificed at the same timeoints as in a single dose study after the last dose. Additionalata at 24 h post-dose were obtained from the animals in thetudy of time-course of toxicity in 2.3.

.7. Statistical analysis

The SPSS for Windows 12.0 computer program was usedor statistical analysis. Statistical comparison of the results

able 1he effects of PQ given subcutaneously at various doses for 7 days on the bod

arameters PQ (mg/kg/day)

0 (10) 4.0 (10)

W gain (g) 38.55 ± 1.51 36.38 ±ST (U/l) 77.10 ± 4.47 71.00 ±LT (U/l) 26.30 ± 1.33 24.20 ±P(U/l) 156.0 ± 5.47 161.0 ±otal bilirubin (mg/dl) 0.09 ± 0.01 0.09 ±lbumin (mg/dl) 4.10 ± 0.11 3.85 ±otal protein (g/dl) 6.34 ± 0.19 6.41 ±UN (mg/dl) 19.00 ± 1.05 17.96 ±reatinine (mg/dl) 0.52 ± 0.01 0.52 ±Q was subcutaneously injected at the doses of 4.0, 5.0, and 6.0 mg/kg/day, oumber in the parenthesis indicates number of animals. The asterisks, * and *ompared with control.

from control and treatment groups were carried out by one-wayanalysis of variance (ANOVA) test and post hoc analy-sis of group differences was performed by LSD or TukeyHSD test. Each data point was based on duplicate deter-minations and expressed as mean ± S.E.M. In addition, thedifference between two groups was assessed by independentsamples t-test. p values less than 0.05 were considered to besignificant.

3. Results

3.1. General signs of toxicity of PQ

Piloerection and impairment of respiratory functionshown as sunken thorax during breathing were observedin PQ-treated animals at all doses after four injectiondays. Animal had reduced locomotion and responseto the stimuli with slight ataxia after 6 days of treat-ment. Over the 7-day treatment period, a rapid andsignificant weight loss was found in 30% PQ-treatedanimals.

3.2. Dose–response to PQ and time-course of PQeffect on the liver function

Interestingly, plasma ALT and AP tended to decreasewith an increase in the dose of PQ (Table 1). PQ pro-duced a dose-dependent decrease in the level of bothalbumin and total bilirubin, especially at the highestdose. A statistical significant decrease in BUN and serum

was observed only at the PQ dose of 6.0 mg/kg day. Lowdose PQ given up to 10 days did not affect BUN, crea-tinine, total protein, total bilirubin, and plasma albumin(Table 2).

y weight and blood chemistry of male Wistar rats

5.0 (9) 6.0 (9)

1.71 25.78 ± 7.01* 12.15 ± 3.68***3.96 71.22 ± 4.64 55.33 ± 3.28**0.89 25.56 ± 1.22 23.78 ± 2.136.63 157.6 ± 5.57 144.3 ± 8.160.01 0.07 ± 0.01 0.06 ± 0.01*0.11 3.72 ± 0.14* 3.33 ± 0.12***0.19 6.54 ± 0.18 6.72 ± 0.130.51 17.87 ± 0.22 16.51 ± 0.45*0.01 0.50 ± 0.01 0.43 ± 0.01***

nce daily, for seven consecutive days. Values are mean ± S.E.M. The**, represent p-value of less than 0.05 and 0.001, respectively, when

Table 2The time-course of effect of PQ on the body weight gain and plasma chemistry of male Wistar rats

Parameters Group PQ treatment period (days)

3 7 10

BW gain (g)Control 15.00 ± 1.26 42.39 ± 1.13 67.70 ± 7.80PQ4.0 14.85 ± 0.29 37.29 ± 1.57 24.40 ± 13.02*

AST (U/l)Control 80.50 ± 1.50 71.25 ± 4.11 76.25 ± 1.80PQ4.0 82.75 ± 2.75 76.75 ± 5.17 59.33 ± 3.28**

ALT (U/l)Control 27.00 ± 2.35 26.00 ± 0.71 25.50 ± 1.04PQ4.0 28.75 ± 2.87 24.25 ± 1.44 19.00 ± 2.31*

AP (U/l)Control 171.3 ± 8.86 157.8 ± 2.87 176.0 ± 6.52PQ4.0 165.5 ± 4.41 147.0 ± 6.34 153.7 ± 12.6

Albumin (mg/dl)Control 3.86 ± 0.15 4.02 ± 0.11 3.92 ± 0.15PQ4.0 3.97 ± 0.08 3.87 ± 0.27 4.05 ± 0.16

Total bilirubin (mg/dl)Control 0.12 ± 0.01 0.11 ± 0.02 0.12 ± 0.00PQ4.0 0.11 ± 0.01 0.12 ± 0.07 0.10 ± 0.02

Total protein (g/dl)Control 6.10 ± 0.26 6.23 ± 0.29 6.61 ± 0.18PQ4.0 6.47 ± 0.22 6.14 ± 0.44 6.43 ± 0.15

BUN (mg/dl)Control 20.20 ± 1.47 20.98 ± 0.70 19.65 ± 1.96PQ4.0 17.75 ± 0.96 19.18 ± 0.67 19.30 ± 1.44

Creatinine (mg/dl)Control 0.53 ± 0.02 0.52 ± 0.01 0.53 ± 0.01PQ4.0 0.49 ± 0.03 0.54 ± 0.02 0.51 ± 0.01

ly, for ulue of l

PQ was subcutaneously injected at the dose of 4.0 mg/kg/day, once daimean ± S.E.M. from 4 animals. The asterisks, * and **, represent p-va

3.3. Effect of PQ on gross and cell morphology ofthe lung and the liver

PQ treatment for 7 days resulted in extensive pul-monary hemorrhage, thickening of alveolar septum aswell as infiltration of inflammatory cells but no pul-monary fibrosis was observed (Fig. 1). The hepatocytes

of the control and PQ-treated groups were still intact,though slight atrophy of the hepatocytes around the cen-tral vein was observed (Fig. 2). No swelling, dilation,infiltration, or proliferation and degenerative changes

Table 3Effect of multi-low dose subcutaneous administration of PQ on the hepatic C

Parameter PQ (mg/kg/day)

0 4.0

EROD 0.11 ± 0.01 0.10 ± 0.0PNPH 5.07 ± 0.21 4.72 ± 0.2ERY 1.09 ± 0.06 1.29 ± 0.0Microsomal protein 19.13 ± 0.65 16.29 ± 0.8

PQ was subcutaneously injected, once daily, for seven consecutive days. Valrepresent p-value of less than 0.05 and 0.005, respectively, when compared wof resorufin formed/mg protein/min. Unit of p-nitrophenol hydroxylation (erythromycin-N-demethylase (ERY) = nmole of formaldehyde formed/mg pro

p to 10 days and sacrificed during treatment and on day 10. Values areess than 0.05 and 0.005, respectively, when compared with its control.

of bile ductule epithelia in the portal area were noted.Necrosis was not seen in the liver of PQ-treated rats.

3.4. Effect of PQ on xenobiotic-metabolizingenzyme activities

Significant decrease in the ethoxyresorufin-O-

deethylase activity, the CYP1A1 probe, was foundin both PQ 5.0 and 6.0 mg/kg day-treated groups(Table 3). There was no change in the activitiesof p-nitrophenol hydroxylase and erythromycin-N-

YP1A1, CYP2E1, and CYP3A4 activities

5.0 6.0

1 0.05 ± 0.01*** 0.03 ± 0.00***2 5.09 ± 0.46 5.87 ± 0.407 1.03 ± 0.04 1.11 ± 0.073* 17.42 ± 0.60 14.73 ± 0.73***

ues are mean ± S.E.M. from eight animals. The asterisks, * and ***,ith control. Unit of ethoxyresorufin-O-deethylation (EROD) = nmolePNPH) = nmole of 4-nitrocatechol formed/mg protein/min. Unit oftein/min Unit of microsomal protein content = mg/g wet liver weight.

Fig. 1. Light micrographs of the alveolar sac in the lung sections from the control (upper panel) and PQ 6.0 mg/kg/day-treated rats (middle andl ote exa 10–20×

dadc

3

c

ower panels). PQ was injected subcutaneously once daily for 7 days. Ns infiltration of inflammatory cells observed in PQ-treated rat (H&E,

emethylase, the CYP2E1 and 3A4 probes, respectively,s compared to those of the control group. PQ seemed toose-dependently reduce the hepatic microsomal proteinontent.

.5. Bioaccumulation of PQ in tissues

The means % recovery of IS spiked in the biologi-al specimens after solid-phase extraction were higher

tensive pulmonary hemorrhage, thickening of alveolar septum as well).

than 90%. Bioaccumulation of PQ in the plasma, liver,and lung are shown in Fig. 3. Except for the plasma,PQ accumulation in the lung and liver of rats treated for7 days was dose-related with the correlation coefficient(R2) in a range of 0.90–0.99. Very high PQ concentra-

tions (2.00–2.25 �g of PQ ion/g tissue) were detected inthe lung compared with that of the liver. At the highestdose, PQ concentration in the liver was only 8% of thatdetected in the lung.

Fig. 2. Light micrographs of the liver sections from the control (left panels: A, C and E) and PQ 6.0 mg/kg/day-treated rats (right panels: B, D andF). PQ was injected subcutaneously once daily for 7 days. Note slight atrophy of the hepatocytes around the central vein (CV) of PQ-treated groups(B) when comparing with the control (A). No centrilobular necrosis or change of bile ductule epithelia in the portal area was observed (D and F)

(H&E, 10–20×).

3.6. Comparison of the concentration–time profilesof PQ in the plasma, lung, and liver after a singleand multi-low dose

Interestingly, there was no difference in the plasmaconcentration–time profiles of PQ after single dose andmulti-dose exposures at any time point indicating a

nearly complete PQ elimination from the plasma within24 h after each dose (Fig. 4). In contrast, lung and liverPQ concentrations after multi-low dose PQ exposure

were much higher than those obtained from a single doseat all times. PQ concentrations in the lung and liver at 24 hpostdose were approximately two- to three-fold higherthan a single exposure (Fig. 5).

Fig. 3. Bioaccumulation of PQ in the plasma and tissues after multi-low dose administration (4.0, 5.0, and 6.0 mg/kg/day, sc) for 7 days.Bar graphs represent mean ± S.E.M., n = 8 each. The asterisks (***)represent p-value less than 0.005.

Fig. 4. Concentration-time profiles of PQ in the plasma, liver, and lungasoi

4

aahto

Fig. 5. Comparison of the concentration–time profiles of PQ in the

fter a single dose, 6.0 mg/kg (A), and multi-dose, 6.0 mg/kg/day (B),c, seven consecutive days, at 0, 0.5, 1, 3, and 24 h after the last dosef PQ. Each point represents the average from the number of animalsn each time point which are 2, 2, 2, 2, and 8, respectively.

. Discussion

The studies of PQ toxicity were conducted in variousnimal species such as rat, mouse, and guinea pig. The

ppropriate animal model for a long term study of theepatotoxic effect of PQ is not well defined. However,he study of Sato (1991) demonstrated that the severityf the liver damage, as shown by dose-related decrease

plasma (A), liver (B), and lung (C) after a single dose, 6.0 mg/kg andmulti-dose, 6.0 mg/kg/day, subcutaneously for 7 days. Each point rep-resents the average from the number of animals in each time pointwhich are 2, 2, 2, 2, and 8, respectively.

in the liver weight and increase in plasma AST activity.In addition, the results of chronic PQ toxicity study bySuriyo (2001) also gave support to the use of rats as ananimal model for PQ-induced hepatotoxic effect.

In agreement with a number of previous reports, lowdoses PQ resulted in dose-dependent reduction feedintake and the body weight gain (Dey et al., 1987;

Sharp et al., 1972). This was probably the effect of PQon the medulla oblongata and hypothalamus (Edmondsand Edwards, 1996). Other toxic manifestations suchas piloerection, hyperpnia, slight ataxia, or hypokinesia

indicated the disturbance of the nervous system function(Satayavivad et al., 1997).

As mentioned earlier, a single high dose of PQ inman resulted in many signs of PQ-induced hepatotoxic-ity (Hong et al., 2000; Bataller et al., 2000; Erickson etal., 1997). However, multi-low dose PQ did not seem tobe a hepatotoxic agent in this study. Other than quite nor-mal liver micrograph, PQ did not cause an increase in theactivity of the marker enzymes for liver damage. Simi-lar results were observed by other workers (Raja et al.,1992; Vuksa et al., 1983). In contrast, dose-dependentdecrease in either plasma albumin or total bilirubin lev-els was found. The etiology of body weight loss andhypoalbuminemia may include malnutrition, overhydra-tion (Emerson, 1989), and PQ-induced albumin damage(Jaiswal et al., 2002). Unaltered total protein contentwith a decrease in albumin level might be the result ofa compensatory increase of the levels of other plasmaproteins as an attempt to maintain the pivotal function ofprotein such as plasma osmotic pressure as seen in thePQ-poisoned patients (Hong et al., 2002). There is noevidence showing the direct effect of PQ on bilirubinmetabolism. However, the depletion of NADPH lev-els following PQ exposure may be one of the factorsinvolving in PQ-induced hypobilirubinemia through aninhibition of heme catabolism. Significant reduction inBUN and creatinine was observed only at the highestdose of PQ. Less diet intake following PQ treatment maybe a major cause (Wu, 2004).

The xenobiotic-metabolizing enzymes play a crucialrole in the detoxification processes and bioactivation ofseveral procarcinogens and a number of low molecu-lar weight pharmaceutical and xenobiotic compoundssuch as that by CYP1A and CYP2E (Robert-Thomsonet al., 1993; Lieber, 1997). CYP3A family accountsfor the metabolism of around 50% of the currentlyused pharmaceutical agents (Pakinson, 1996). This studyhas been conducted to investigate whether PQ couldalter these metabolizing enzymes. The existing evidenceshowed that PQ could interact with hepatic cytochromeP450 and depress the activities of aniline hydroxylase(Montgomery and Shamblin, 1984) as well as ben-zphetamine N-demethylase (Zychlinski et al., 1988). Thepresent results demonstrated that, at low dose, PQ did notaffect either CYP2E1 or CYP3A4 activity but it caused adose-dependent reduction in the CYP1A1 activity. In thisregard, PQ might act as an enzyme inhibitor of CYP1A1.Reduction of EROD-related CYP1A1 activity following

PQ exposure might be due to a destruction of microso-mal cytochrome P-450 or possibly via the disruption themicrosomal electron transport system (Bus and Gibson,1975). The effect on CYP could partially explain the

prolongation of hexobarbital sleeping time after multi-dose administration of PQ as observed in the preliminarystudy. In addition to the depletion of NADPH, PQ alsocaused a decrease in the microsomal protein (Aldrich etal., 1983). As CYP1A1 plays a role in the biotransfor-mation of drugs, environmental chemicals (e.g. aromaticand heterocyclic hydrocarbons) and natural compounds(e.g. aflatoxin), such metabolic inhibition might haveboth clinical and toxicological significances in termsof therapeutic outcome and risk of toxic exposure. Themechanism by which PQ reduces CYP1A1 activity isunclear and needs to be explored.

Investigation regarding the accumulation of PQ asrelated to the target organ toxicity after a multi-doseexposure, especially the liver, has not been reportedpreviously. Nearly undetectable PQ at 24 h post-doseindicated a rapid clearance from the circulation. Thisfinding demonstrated that plasma PQ concentration,which has been thought to be a good marker for PQexposure and toxicity after a single high dose in man,may not be as useful after low multi-dose exposure. Theaccumulation of PQ both in the lung and liver was dose-dependent but the degree of accumulation in the lung wasmuch higher than that detected in the liver. The differ-ence in distribution could be explained, at least in part, bythe differences in herbicide uptake which is an energy-dependent transport into the lung (Rose et al., 1976).High oxygen concentration in the lung enhanced PQ tox-icity via radical formation. PQ concentration greater than1–2 �g/g tissue appeared to be associated with lung dam-age (Litchfield et al., 1973) and multi-low dose exposurecould increase lung PQ concentration to that toxic range.Another cause of morbidity and mortality by PQ was animpairment of the gas exchange as shown by a decreasein saturated oxygen and PO2, and an increase in PCO2and HCO3

− in a dose-dependent manner (unpublishedresults).

Low hepatic PQ concentration, only 8% of that inthe lung, one of the explanations for the lack of hepato-toxic effect of PQ may either be due to the absence ofspecific transporters or the presence of elimination pro-cess in this organ. Rapid disappearance of PQ from theliver via biliary excretion (Danial and Gage, 1966) andbiotransformation to a less toxic compound, paraquat-monopyridone, by an unknown detoxifying system inrat hepatic cytosol portion distinct from the P450 family(Fuke et al., 1993) may also be possible reasons. Theresistance of the liver to PQ toxicity at low to moderate

doses might be due to less production of lipid peroxida-tion products comparing with the lung (Melchiorri et al.,1995). On the other hand, the oxidative stress, if therewas any, could be suppressed by the antioxidant enzyme

sPdPc

iastdemldooles

A

GEuMgofb

R

A

B

B

C

C

C

ystems that were plentiful in the liver. At very high dose,Q-induced production of reactive oxygen radicals and aepletion of reduced glutathione (Melchiorri et al., 1995;almeira et al., 1995) may be much more prominent andould result in hepatotoxic effect.

Our preliminary study on the multi-low dose PQ tox-city reported herein indicated that this herbicide mightffect certain synthetic function of the liver or activity ofome hepatic xenobiotic-metabolizing enzymes. Resis-ance to the toxic effect of PQ by the liver was probablyue to low bioaccumulation of PQ in this organ. How-ver, a longer period of PQ exposure even at low doseight cause a slight change in the liver function (unpub-

ished results). In addition, plasma PQ concentrationoes not indicate the degree of exposure and severityf PQ toxicity. To further clarify the toxic mechanismf multi-low dose PQ on the target organs including theung and the liver, the status of antioxidant system andxpression of certain gene encoding the target proteinshould be investigated.

cknowledgements

This research is supported by the grant from the Post-raduate Education, Training and Research Program innvironmental Science, Technology and Managementnder Higher Education Development Project of theinistry of University Affairs. The authors are very

rateful to Assoc. Prof. Dr. Chiaki Fuke, Departmentf Legal Medicine, University of the Ryukyus, Japan,or assistance in the technical part of PQ determinationy HPLC.

eferences

ldrich, T.K., Fisher, A.B., Forman, H.J., 1983. Paraquat inhibitsmixed-function oxidation by rat lung. J. Appl. Physiol. 54 (4),1089–1093.

ataller, R., Bragulat, E., Nogue, S., Gorbig, M.N., Bruguera, M.,Rodes, J., 2000. Prolonged cholestasis after acute paraquat poi-soning through skin absorption. Am. J. Gastroenterol. 95 (5),1340–1343.

us, J.S., Gibson, J.E., 1975. Postnatal toxicity of chronicallyadministered paraquat in mice and interactions with oxygen andbromobenzene. Toxicol. Appl. Pharmacol. 33, 461–470.

anton, A., Young-Holt, B., 2002. Pesticide-related symtoms amongfarm workers in rural Hondurus. Int. J. Occup. Environ. Health 8(1), 41–45.

onradi, S.E., Olanoff, L.S., Dawson Jr., W.T., 1983. Fatality due toparaquat intoxication: confirmation by postmortem tissue analysis.

Am. J. Clin. Pathol. 80 (5), 771–776.

orasaniti, M.T., Strongoli, M.C., Nistico, G., 1990. Determinationof paraquat in rat brain using ion-pair solid-phase extractionand reversed-phase high performance liquid chromatography withultraviolet detection. J. Chromatogr. 527, 189–195.

Danial, J.W., Gage, J.C., 1966. Absorption and excretion of diquat andparaquat in rats. Br. J. Ind. Med. 23, 133–136.

Dey, M.S., Krieger, R.I., Ritter, R.C., 1987. Paraquat-induced, dose-dependent conditioned taste aversions and weight loss mediated bythe area postrema. Toxicol. Appl. Pharmacol. 87, 212–221.

Doumas, B.T., Watson, W.A., Biggs, H.G., 1971. Albumin standardsand the measurement of serum albumin with bromocresol green.Clin. Chem. Acta 31, 87–96.

Edmonds, B.K., Edwards, G.L., 1996. The area postrema is involvedin paraquat-induced conditioned aversion behavior and neuroen-docrine activation of the hypothalamic-pituitary-adrenal axis.Brain Res. 712, 127–133.

Emerson, T., 1989. Unique features of albumin: a brief review. Crit.Care Med. 17, 7690–7694.

Erickson, T., Brown, K.M., Wigder, H., Gillespie, M., 1997. A caseof paraquat poisoning and subsequent fatality presenting to anemergency department. J. Emerg. Med. 15 (5), 649–652.

Fuke, C., Ameno, K., Ameno, S., Kiriu, T., Shinohara, T., Ijiri, I.,1993. In vitro studies of the metabolism of paraquat and diquatusing rat liver homogenates—isolation and identification of themetabolites of paraquat and diquat. Nippon Hoigaku Zasshi 47 (1),33–45.

Fuke, C., Arao, T., Morinaga, Y., Takaesu, H., Ameno, K., Miyazaki,T., 2002. Analysis of paraquat, diquat and two diquat metabolites inbiological materials by high performance liquid chromatography.Leg. Med. 4, 156–163.

Hong, S.Y., Hwang, K.Y., Lee, E.Y., Eun, S.W., Cho, S.R., Han, C.S., etal., 2002. Effect of vitamin C on plasma total antioxidant status inpatients with paraquat intoxication. Toxicol. Lett. 126 (1), 51–59.

Hong, S.Y., Yang, D.H., Hwang, K.Y., 2000. Associations betweenlaboratory parameters and outcome of paraquat poisoning. Toxicol.Lett. 118 (1–2), 53–59.

Jaiswal, R., Khan, M.A., Musarret, J., 2002. Photosensitizedparaquat-induced structural alteration and free radical mediatedfragmentation of serum albumin. J. Photochem. Photobiol. B: Biol.67, 163–170.

Kodagoda, N., Jayewardene, R.P., Attygalle, D., 1973. Poisoning withparaquat. Forensic Sci. 2 (1), 107–111.

Koop, D.R., 1986. Hydroxylation of p-nitrophenol by rabbit ethanol-inducible cytochrome P450 isozyme 3a. Mol. Pharmacol. 29,399–404.

Lake, B.G., 1987. Preparation and charecterization of microsomal frac-tions for studies of xenobiotic metabolism. In: Shell, K., Mullockl,B. (Eds.), Biochemical Toxicology: A Practical Approach. IRLPress, Oxford, pp. 183–215.

Lieber, C.S., 1997. Cytochrome P-450 2E1: Its physiological andpathological role. Physiol. Rev. 77, 517–544.

Litchfield, M.H., Daniel, J.W., Longshaw, S., 1973. The tissue distri-bution of the bipyridylium herbicides diquat and paraquat in ratsand mice. Toxicology 1 (2), 155–165.

Lowry, O.H., Rosenbrough, N.J., Fare, A.L., Randall, R.T., 1951. Pro-tein measurement with the folin phenol reagent. J. Biol. Chem.193, 265–275.

Meeks, R.G., Harrison, S.D., Bull, J.B., 2000. Hepatotoxicology. CRCPress Inc., Florida, USA.

Melchiorri, D., Reiter, R.J., Attia, A.M., Hara, M., Burgos, A., Nistico,G., 1995. Potent protective effect of melatonin on in vivo paraquat-

induced oxidative damage in rats. Life Sci. 56 (2), 83–89.

Montgomery, M.R., Shamblin, P.B., 1984. Ascorbic acid potentiatesthe substrate-specific inhibition of mixed-function oxidation andthe stimulation of NADPH oxidation caused by paraquat. J. Toxi-col. Environ. Health 13 (1), 69–81.

Mullick, F.G., Ishak, K.G., Mahabir, R., Stromeyer, F.W., 1981. Hep-atic injury associated with paraquat toxicity in humans. Liver 1 (3),209–221.

Mutsumori, H., Matsumoto, T., Ishkawa, H., 1985. Acute toxic effectof paraquat on ultrastructural of rat liver. Acta Pathol. Jpn. 34 (4),507–518.

Nash, T., 1953. The colorimetric estimation of formaldehyde by meansof the Hantzsch reaction. J. Biol. Chem. 55, 416–421.

Palmeira, C.M., Moreno, A.J., Madeirs, V.M.C., 1995. Thiolsmetabolism is altered by the herbicides paraquat, dinoseb and2,4-D: a study in isolated hepatocytes. Toxicol. Lett. 81, 115–123.

Pakinson, A., 1996. Biotransformation of xenobiotics. In: Klaassen,C.D. (Ed.), Casarette & Doull’s Toxicology: The Basic Science ofPoisons, 5th ed. McGrew-Hill, New York, pp. 113–186.

Prough, R.A., Burke, M.D., Mayer, R.T., 1978. In: Fleischer, S., Packer,L. (Eds.), Methods in Enzymology, vol. 52. Academic Press, NewYork, p. 372.

Raja, M., Al-Fatah, A., Ali, M., Afzal, M., Hassan, R.A., Menon,M., et al., 1992. Modification of liver and serum enzymes byparaquat treatment in rabbits. Drug Metabol. Drug Interact. 10 (4),279–291.

Robert-Thomson, S.J., McManus, M.E., Tukey, R.H., Gonzales, F.J.,Holder, G.M., 1993. The catalytic activity for four expressedhuman cytochrome P450s towards benzo[a]pyrene and the isomers

of its proximate carcinogen. Biochem. Biophys. Res. Commun.192, 1373–1379.

Rose, M.S., Lock, E.A., Smith, L.L., Wyatt, I., 1976. Paraquat accu-mulation: tissue and species specificity. Biochem. Pharmacol. 25(4), 419–423.

Satayavivad, J., Siripat, W., Thiantanawat, A., 1997. Neurologicaleffects of chronic exposure to low dose of paraquat in rats. Res.Commun. Pharmacol. Toxicol. 2, 269–282.

Sato, M., 1991. Dose-dependent increases in metallothionein synthe-sis in the lung and liver of paraquat-treated rats. Toxicol. Appl.Pharmacol. 107, 98–105.

Sharp, C.W., Ottolenghi, A., Posner, H.S., 1972. Correlation ofparaquat toxicity with tissue concentrations and weight loss of therat. Toxicol. Appl. Pharmacol. 22 (2), 241–251.

Suriyo, T., 2001. Modification of central cholinergic controlling motoractivity during subchronic exposure to paraquat in rats. M.Sc. The-sis in Toxicology. Faculty of Graduate Studies, Mahidol University,Bangkok.

Takegoshi, K., Nakanuma, Y., Ohta, M., Thoyama, T., Okuda, K.,Kono, N., 1988. Light and electron microscopic study of the liverin paraquat poisoning. Liver 8 (6), 330–336.

Teixeira, H., Proenca, P., Alvarenga, M., Oliveira, M., Marques, E.P.,Vieira, D.N., 2004. Pesticide intoxications in the centre of Portugal:three years analysis. Forensic Sci. Int. 143 (2–3), 199–204.

Vuksa, M., Neskovic, N., Vitorovic, S., Karan, V., 1983. Subacute tox-icity of paraquat in rats-biochemical effects. Ecotoxicol. Environ.Saf. 7 (5), 475–483.

Wu, J.S., 2004. Dr. Wu’s liver diseases for professionals (medicalstudent and residents) and consumers. Retrieved January 14, 2005,

from http://olddoc.tmu.edu.tw/chiaungo/profes/Transaminase.htm (posted January 15, 1997).

Zychlinski, L., Raska-Emery, P., Montgomery, M.R., 1988. Influenceof bipyridylium compounds on microsomal mixed-function oxida-tion activities. J. Biochem. Toxicol. 3, 173–189.