A Comparative Study of Binding Sites for Diisopropyl Phosphorofluoridate in Membrane and Cytosol Preparations from Spinal Cord and Brain of Hens

NeuroToxicology 22 (2001) 191±202

A Comparative Study of Binding Sites for Diisopropyl Phosphoro¯uoridate in Membrane and Cytosol Preparations from Spinal Cord and Brain of Hens Ryo Kamata1,3,*, Shin-ya Saito2, Tadahiko Suzuki1, Tadashi Takewaki3, Hisayoshi Kofujita4, Michikazu Ota4, Haruo Kobayashi1 1

Department of Veterinary Pharmacology, Faculty of Agriculture, Iwate University, 3-18-8 Ueda, Morioka 020-8550, Japan 2 Department of Pharmaceutical Molecular Biology, Faculty of Pharmaceutical Sciences, Tohoku University, Sendai 980-8578, Japan 3 United Graduate School of Veterinary Sciences, Gifu University 1-1 Yanagido, Gifu 501-1193, Japan 4 Department of Wood Science and Technology, Faculty of Agriculture, Iwate University, 3-18-8 Ueda, Morioka 0208550, Japan Received 23 February 2000; accepted 19 October 2000

Abstract Biochemical events in the initiation of organophosphorus induced delayed neurotoxicity (OPIDN) are not well understood. To ®nd new putative target(s) for OPIDN, we investigated the biochemical and pharmacological characteristics of [3 H]diisopropyl phosphoro¯uoridate (DFP) binding to membrane and cytosol preparations from the brain and spinal cord of hens in vitro. [3 H]DFP binding to both preparations was determined by the speci®c binding obtained by subtracting non-speci®c binding from total binding. The speci®c binding sites of [3 H]DFP were found not only on membrane but also in cytosol. Kd values were higher and Bmax values were lower in cytosol than in membrane. Moreover, the Kd values in both membrane and cytosol preparations from spinal cord were lower than those of brain. The Bmax values in membrane and cytosol were similar between brain and spinal cord. The speci®c binding to both preparations was markedly displaced by unlabeled DFP. The speci®c binding of DFP to the membrane was highly or partly displaced by organophosphorus compounds (OPs) or a carbamate, respectively. However, both the OPs and the carbamate had considerably weaker blocking effects on the speci®c binding of DFP to cytosol. None of the compounds known to interact with neuropathy target esterase (NTE) had a strong blocking effect on the speci®c binding of DFP to either membrane or cytosol. These results show that the speci®c binding of DFP to the membrane may be binding with cholinesterase (ChE). However, cytosol, especially in spinal cord, may have DFP binding sites other than ChE and NTE. # 2001 Elsevier Science Inc. All rights reserved.

Keywords: Organophosphorus induced delayed neurotoxicity (OPIDN); Diisopropyl phosphoro¯uoridate (DFP); Cytosolic protein; Hen

INTRODUCTION A number of organophosphorus esters produce delayed neuropathy (OPIDN) in humans and other * Corresponding author. Tel.: ‡81-19-621-6215; fax: ‡81-19-621-6515. E-mail address: [email protected] (R. Kamata).

sensitive avian and mammalian species (Abou-Donia, 1981). OPIDN is still observed following accidental or suicidal acute poisoning with some organophosphorus compounds (OPs), even though the organophosphate insecticides currently used have signi®cantly reduced toxicity (Abou-Donia and Lapadula, 1990). Although numerous studies have attempted to ®nd the pathogenesis of OPIDN, the mechanism is still almost unknown.

0161-813X/01/$ ± see front matter # 2001 Elsevier Science Inc. All rights reserved. PII: S 0 1 6 1 - 8 1 3 X ( 0 0 ) 0 0 0 1 3 - 9

192

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

OPIDN has been long thought to be initiated by phosphorylation of neuropathy target esterase (NTE) with the subsequent modi®cation (``aging'') of this phosphorylated enzyme (Johnson, 1970, 1982). However, attempts to associate the inhibition of NTE activity with speci®c endogenous substrates and changes in neuronal biochemistry and function have not been successful in de®ning the pathogenesis of OPIDN (Abou-Donia, 1981; Johnson, 1990; Kamata et al., 1999). For example, Abou-Donia and Lapadula (1990) have pointed out that NTE activity represents approximately 6% of the total hydrolysis of phenyl valerate, an arti®cial substrate, in hen brains. This enzyme is also present in most tissues assayed. The only evidence that NTE might be involved in the pathogenesis of OPIDN is a correlated change in NTE activity prior to the disease. Furthermore, there is no hypothesis as to how the inhibition and aging of NTE leads to neuronal damage characteristic of OPIDN. Diisopropyl phosphoro¯uoridate (DFP) is one of the typical compounds capable of producing OPIDN and inhibiting the activity of NTE. Many investigations of the DFP labeling sites (proteins) of hen brain have been reported. Most of these have investigated the active sites of NTE or phenyl valerate esterases on membrane proteins (Williams and Johnson, 1981; Pope and Padilla, 1989; Meredith and Johnson, 1989). However, there are probably many unexplored DFP binding sites, other than NTE, in neural tissue, which may include some sites with important physiological functions. Carrington and Abou-Donia (1985) have reported that the number of DFP binding sites does not necessarily correspond to phenyl valerate-activity in membrane fractions separated by sodium dodecyl sulfate/polyacrylamide-gel electrophoresis from hen brain. Therefore, it was important to explore new DFP binding site(s) which may participate in the production of OPIDN, and to clarify the pharmacological and biochemical properties of those sites. One of the objectives of the present study was to attempt to differentiate DFP binding on cholinesterases (ChE) and NTE from other potentially important sites. Recently, it has been reported that there are DFP binding sites on membranes from the spinal cord of chickens by means using radioactive DFP (Konno et al., 1994, 1999). However, it is likely to be dif®cult to differentiate those sites from the expected binding to acetylcholinesterase (AChE) or NTE. Therefore, in the present study, saturation binding assays of [3 H]DFP in membrane and cytosol preparations from the brain and spinal cord of hens were used

to ®nd new target site(s) for the initiation of OPIDN in neural tissue. Additionally, to differentiate potentially important new sites from simple binding on ChE or NTE, the target site(s) were further examined by analyzing displacement of speci®c [3 H]DFP binding by compounds that are known to interact with the activities of ChE or NTE in vitro. MATERIALS AND METHODS Chemicals DFP, diethyl p-nitrophenyl phosphate (paraoxon), phenylmethylsulfonyl ¯uoride (PMSF), O,Odiethyl-O-3,5,6-trichloro-2-pyridyl phosphorothioate (chlorpyrifos), N,N0 -diisopropyl phosphorodiamidic ¯uoridate (mipafox), and eserine (physostigmine) salicylate were purchased from Wako Pure Chemical Industries (Osaka, Japan), Aldrich Chemical Co., Inc. (Milwaukee, WI), Nacalai Tesque Co., Ltd. (Kyoto, Japan), Kanto Chemical Co., Inc. (Tokyo, Japan), Oriza Laboratories, Inc. (Newburyport, MA), and Sigma Chemical Co. (St. Louis, MO), respectively. Phenyl valerate was synthesized and puri®ed by the method of Johnson (1977). [3 H]DFP (111 GBq/mmol) was purchased from Dupont/New England Nuclear (Boston, MA). Animals and Tissues Adult white Leghorn laying hens (Gallus gallus domesticus, body weight 1.6±2.2 kg) were used. Tissue preparations were made using a modi®ed method of Konno et al. (1994). The medulla and cerebellum, regions with neuropathological changes in OPIDN, were removed from the brain to compare with the spinal cord, a susceptible tissue. The remaining brain tissue and spinal cords were cleaned free of blood vessels and meninges, and homogenized on ice in 10 vol (w/v) of ice-cold HEPES buffer (50 mM, pH 7.4) using a Te¯on-glass homogenizer. These homogenates were centrifuged at 1000g for 10 min at 48C, and the supernatants were centrifuged again at 50,000g for 10 min at 48C. The resulting supernatants were used as cytosol preparation. Furthermore, the pellets from the last 50,000g centrifugation were resuspended in HEPES buffer and recentrifuged at 50,000g for an additional 10 min at 48C, and the supernatants were discarded. HEPES buffer was then added to these washed pellets, which were then resuspended using Polytron homogenizer, so

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

the concentration of these preparations were about 50 mg tissue/ml. The resulting suspensions were used as the membrane preparation. Assay of [3 H]Diisopropyl Phosphorofluoridate Binding To assay the speci®c binding of [3 H]DFP to the membrane and cytosol preparations, the membrane and cytosol preparations were further diluted to 5 or 10 vol, respectively, with 10 mM sodium phosphate buffer (pH 7.4). Diluted preparations of 100 ml each and 100 ml of 10 mM sodium phosphate buffer or 100 ml of 25 mM unlabeled DFP (®nal concentration of 10 mM) were incubated with 50 ml of [3 H]DFP at 378C for 60 min. For experiments examining concentration dependent (saturation) binding of [3 H]DFP or in displacement experiments, the solution of [3 H]DFP was adjusted to 2±128 nM (®nal concentrations of 0.4±25.6 nM) or 50 nM (10 nM), respectively. In the displacement experiments, 100 ml of a solution of each appropriate displacer were used instead of 100 ml of

193

10 mM sodium phosphate buffer. The reaction was terminated by ®ltration through a GF/B glass ®lter (24 mm in diameter, Whatman International Ltd., Maidston, UK). The ®lter was washed twice with 5 ml of ice-cold Tris±HCl buffer (10 mM Tris, 140 mM NaCl, pH 7.4). The ®lters were soaked in 0.3% polyethylenimine for 1±24 h before use and were placed on the ®lter apparatus without washing according to the method of Bruns et al. (1983). After ®lters had been soaked overnight in 5 ml of scintillation cocktail (ACS-II; Amersham, Bucks, UK), the radioactivity attributable to tritium was counted (LSC-5100, Aloka Co. Ltd., Tokyo). Speci®c binding of [3 H]DFP was de®ned as difference between the total binding and non-speci®c binding measured in the absence and presence of unlabeled DFP, respectively. Protein Assay Concentration of protein was measured by the method of Lowry et al. (1951) with bovine serum albumin as the standard.

Fig. 1. Concentration dependent (saturation) curve for the binding of [3 H]DFP to preparations used in the present study: (A) brain membrane; (B) spinal cord membrane; (C) brain cytosol; (D) spinal cord cytosol. The curves are representative of results from three separate experiments, and each point represents the mean value of triplicate determinations. Specific binding of [3 H]DFP is calculated as the difference between total and non-specific binding. Triplicate experiments gave similar results.

194

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

Data Analysis Saturation binding data were converted to Scatchard plots, from which af®nity constants Kd and Bmax values were determined as 1/slope and the x-axis intercept, respectively. IC50 and slope factor (pseudo Hill coef®cient) values from displacement experiments in vitro were determined as slope and the x-axis intercept from pseudo Hill plots, respectively. pKi values were calculated from the IC50 values using the equation pKi ˆ logfIC50 =…1 ‡ …‰FŠ=Kd ††g, where [F] is the concentration of free ligand (Cheng and Prusoff, 1973). Comparisons of the slope factor and pKi values were performed by using a one-way analysis of variance (ANOVA). Post hoc analysis was performed with Dunnett's test.

RESULTS Saturation Binding of [3 H]DFP Concentration-dependence of [3 H]DFP binding to membrane and cytosol preparations from brain and spinal cord was examined. Representative saturation curves of [3 H]DFP binding to each preparation are shown in Fig. 1A±D. In all preparations, although total and non-speci®c binding increased in a concentrationdependent manner, speci®c binding reached a plateau at concentrations of 5±15 nM [3 H]DFP. The Kd and Bmax values of [3 H]DFP binding in each preparation were determined using Scatchard analysis and listed in Table 1. The Kd values were lower in membrane than in cytosol, and in spinal cord than in brain. The Bmax

Fig. 2. Displacement of [3 H]DFP binding to membrane preparations from brain (A) and spinal cord (B) by ChE inhibitors. Data are plotted as a percentage of specific binding of control. Control values of [3 H]DFP bound to each preparation in brain and spinal cord were 1:03  0:10 and 1:91  0:18 pmol/mg protein, respectively (mean  S:E:M:, n ˆ 5). The data are representative of results from three to four separate experiments, and each point represents the mean value of duplicate determinations.

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

195

Table 1 Specific binding of [3 H]DFP to membrane and cytosol preparations from brain and spinal cord of hensa

Effects of Various Compounds on [3 H]DFP Binding In Vitro

Tissue

Kd (nM)

Bmax (pmol/mg protein)

Brain Membrane Cytosol

3.36 7.41

1.49 0.41

Spinal cord Membrane Cytosol

1.40 4.03

1.71 0.57

In displacement experiments in vitro, the ability of various compounds to displace speci®c [3 H]DFP binding to membrane or cytosol from brain and spinal cord was investigated. The concentration of [3 H]DFP used in each experiment was determined by the results of saturation binding to each preparation. Displacement curves of [3 H]DFP binding to each preparation are shown in Figs. 2±5, and all ®tted data are summarized in Table 2. Furthermore, the percentages of maximal displacement are expressed in Table 3, since speci®c [3 H]DFP binding was not always completely displaced by various compounds. The most potent compound in each preparation was unlabelled DFP. Its slope factor values were near to 1 and pKi values were about 8±9 (Table 2).

a Kd and Bmax values were determined as intercept in Scatchard plots, respectively.

1/slope and the x-axis

values were higher in membrane than in cytosol, and were similar in brain and spinal cord. The highest af®nity and maximum binding were observed in the membranes from spinal cord (Table 1).

Fig. 3. Displacement of [3 H]DFP binding to cytosol preparations from brain (A) and spinal cord (B) by ChE inhibitors. Data are plotted as a percentage of specific binding of control. Control values of [3 H]DFP bound to each preparation in brain and spinal cord were 0:33  0:02 and 0:51  0:04 pmol/mg protein, respectively (mean  S:E:M:, n ˆ 5). The data are representative of results from three to four separate experiments, and each point represents the mean value of duplicate determinations.

196

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

Fig. 4. Displacement of [3 H]DFP binding to membrane preparations from brain (A) and spinal cord (B) by interactive compounds with NTE. Data are plotted as a percentage of specific binding of control. Control values of [3 H]DFP bound to each preparation in brain and spinal cord were 1:03  0:10 and 1:91  0:18 pmol/mg protein, respectively (mean  S:E:M:, n ˆ 5). The data are representative of results from three to four separate experiments, and each point represents the mean value of duplicate determinations.

Displacement by Cholinesterase Inhibitors Displacement of [3 H]DFP binding by cholinesterase inhibitors in membrane and cytosol preparations was examined and the response curves from brain and spinal cord are shown in Figs. 2 and 3, respectively. In membrane preparations from brain and spinal cord (Fig. 2A and B, respectively), DFP was most effective in displacing itself. The second most potent compound was paraoxon, an organophosphate (OP) which does not induce OPIDN. The ability of paraoxon to displace [3 H]DFP was only slightly lower than that of DFP itself in the membranes of both brain and spinal cord. The slope factor and pKi values of paraoxon in membranes were, in fact, not signi®cantly different from those of DFP (Table 2). The percentages of

maximal displacement by paraoxon approached 100% and were not signi®cantly different from those of DFP in membranes from both tissues (Table 3). Mipafox, a compound that produces OPIDN, was also very potent and displaced almost 100% of speci®c DFP binding in both brain and spinal cord membranes (Table 3). In the case of mipafox, however, the slope factor in brain membranes (but not spinal cord membranes) and pKi values in both brain and spinal cord membranes were signi®cantly lower than those of DFP (Table 2). In contrast to mipafox, chlorpyrifos, which is also proposed as a delayed neurotoxic compound, exhibited weak blocking of DFP binding in both brain and spinal cord membranes. Although chlorpyrifos had signi®cantly lower pKi values than DFP (Table 2), the percentages of maximal displacement were still very

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

197

Fig. 5. Displacement of [3 H]DFP binding to cytosol preparations from brain (A) and spinal cord (B) by interactive compounds with NTE. Data are plotted as a percentage of specific binding of control. Control values of [3 H]DFP bound to each preparation in brain and spinal cord were 0:33  0:02 and 0:51  0:04 pmol/mg protein, respectively (mean  S:E:M:, n ˆ 5). The data are representative of results from three to four separate experiments, and each point represents the mean value of duplicate determinations.

high and not signi®cantly different from DFP (Table 3). Eserine, a short acting carbamate anticholinesterase, showed a weak displacement of speci®c DFP binding in both brain and spinal cord membranes, and had signi®cantly lower slope factor and pKi values than DFP in both tissues (Table 2). Even at the highest concentration (1 mM), eserine never exhibited complete displacement of speci®c DFP binding and had signi®cantly lower maximal displacement from membranes of either tissue (Table 3). In cytosol preparations from brain and spinal cord (Fig. 3A and B, respectively), as compared to membrane, the inhibitory effects of paraoxon on the binding of [3 H]DFP were lower than those of DFP. The pKi values in brain cytosol, and the slope factor and pKi values in spinal cord cytosol for paraoxon were sig-

ni®cantly different from those of DFP (Table 2), even though the maximal displacement was high and not signi®cantly different from DFP (Table 3). Mipafox also had slope factor values in spinal cord cytosol and pKi values in cytosol from both tissues that were signi®cantly lower than corresponding values obtained with DFP (Table 2). There was, however, no signi®cant difference in percentages of maximal displacement between mipafox and DFP in cytosol from either brain or spinal cord (Table 3). As compared to membrane, chlorpyrifos exhibited a weak ability to displace DFP from cytosol from both brain and spinal cord. Chlorpyrifos had signi®cantly lower pKi values in both brain and spinal cord (Table 2) and showed signi®cantly less percentage of maximal displacement in brain (but not spinal cord) as compared to DFP (Table 3). Eserine

198

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

Table 2 Displacement of specific [3 H]DFP binding to membrane and cytosol preparations from brain and spinal cord by interactive compounds with ChE and NTEa Tissue

Preparation

Compound

Slope factor

Brain

Membrane

DFP Paraoxon Mipafox Chlorpyrifos Eserine Phenyl valerate PMSF DFP Paraoxon Mipafox Chlorpyrifos Eserine Phenyl valerate PMSF

0.83 0.75 0.46 0.57 0.34 1.18 0.44 0.90 0.88 0.47 0.49 0.13 1.10 0.54

             

0.07# 0.14 0.03* 0.09 0.05** 0.05## 0.06 0.07 0.07 0.05 0.03 0.01* 0.27# 0.03

DFP Paraoxon Mipafox Chlorpyrifos Eserine Phenyl valerate PMSF DFP Paraoxon Mipafox Chlorpyrifos Eserine Phenyl valerate PMSF

1.00 0.65 0.72 0.66 0.30 1.22 0.30 0.76 0.44 0.44 0.52 0.12 0.88 0.31

             

0.01 0.11 0.04 0.06 0.03** 0.20# 0.03** 0.03# 0.09* 0.04* 0.08 0.07**,## 0.01## 0.04**

Cytosol

Spinal cord

Membrane

Cytosol

pKi 8.25 7.46 6.72 4.96 5.90 3.73 4.15 8.09 5.22 5.04 4.28 3.63 4.14 3.85

             

0.01## 0.28 0.29** 0.13**,## 0.30** 0.14**,## 0.05**,## 0.17## 0.25** 0.18** 0.10** 0.40**,## 0.11** 0.01**,#

8.91 8.19 7.87 5.69 5.70 4.22 5.22 7.96 6.39 5.91 4.73 <3 4.14 4.45

          

0.33# 0.11 0.19* 0.20**,## 0.27**,## 0.15**,## 0.01**,## 0.17## 0.61* 0.34** 0.16**

 0.12**,#  0.15**

a Values are expressed as means  S.E.M. for three to four experiments. Asterisks and sharps indicate significant differences from DFP (*: P < 0:05, **: P < 0:01) and mipafox (#: P < 0:05, ##: P < 0:01), respectively. Slope factor and pKi values were determined from pseudo Hill plots and the IC50 values as described in Materials and Methods.

exhibited very weak displacement of DFP from cytosol even at high concentrations. The slope factors, pKi values and percentages of maximal displacement obtained with eserine were all signi®cantly lower than corresponding values obtained with DFP (Tables 2 and 3). Displacement by Neuropathy Target Esterase Inhibitors Displacement of [3 H]DFP binding by various compounds known to interact with the activity of NTE in membrane and cytosol preparations was examined and the response curves from brain and spinal cord are shown in Figs. 4 and 5, respectively. Mipafox is a delayed neurotoxicant as well as an NTE inhibitor, so the displacement curves for mipafox were also included in these ®gures. In membrane preparations from brain and spinal cord (Fig. 4A and B, respectively), phenyl valerate,

a substrate of NTE, exhibited very weak ability to displace DFP. The pKi values and percentages of maximal displacement of speci®c DFP binding were very low and signi®cantly different from those of DFP in membranes from both tissues (Tables 2 and 3, respectively), even though the slope factor values were very high in membranes from both tissues (Table 2). PMSF, an antagonist of NTE, exhibited very weak blocking of speci®c DFP binding even at high concentrations. The pKi values and percentages of maximal displacement in membranes from both tissues and the slope factor values in membranes from spinal cord were signi®cantly different from those of DFP (Tables 2 and 3). Displacement of speci®c DFP binding by phenyl valerate in cytosol preparations from brain and spinal cord is shown in Fig. 5A and B, respectively. The inhibitory effects of phenyl valerate on DFP binding in cytosol were similar to those in membrane preparations

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

199

Table 3 Percentages of maximal displacement of [3 H]DFPa Tissue

Preparation

Compound

Maximal displacement (%)

Brain

Membrane

DFP Paraoxon Mipafox Chlorpyrifos Eserine Phenyl valerate PMSF DFP Paraoxon Mipafox Chlorpyrifos Eserine Phenyl valerate PMSF

100.00 95.24 94.77 83.94 73.21 56.16 65.40 100.00 97.18 88.97 73.25 48.37 86.43 63.47

DFP Paraoxon Mipafox Chlorpyrifos Eserine Phenyl valerate PMSF DFP Paraoxon Mipafox Chlorpyrifos Eserine Phenyl valerate PMSF

100.00 99.88 100.00 89.17 64.52 68.41 77.85 100.00 92.24 90.71 81.07 39.86 76.69 71.23

Cytosol

Spinal cord

Membrane

Cytosol

     

2.38 1.84 4.44 4.70**,# 7.86**,## 2.49**,##

     

2.82 1.61 2.36* 9.96**,## 8.03 4.27**

 0.12    

2.76 3.44**,## 7.37**,## 4.59**,##

     

6.85 3.23 4.07 6.70**,# 8.59* 0.65*

a Values are expressed as means  S.E.M. for three to four experiments. Asterisks and sharps indicate significant differences from DFP (*: P < 0:05, **: P < 0:01) and mipafox (#: P < 0:05, ##: P < 0:01), respectively. Control values of [3 H]DFP bound to membrane and cytosol in brain were 1.03  0.10 and 0.33  0.02 pmol/mg protein, respectively, and in spinal cord 1.91  0.18 and 0.51  0.04 pmol/mg protein, respectively.

as shown by all values (Tables 2 and 3), even though the percentages of maximal displacement in brain cytosol were high and not signi®cantly different from those of DFP (Table 3). The ability of PMSF to block speci®c DFP binding in cytosol was also similar to the PMSF effects in membrane preparations. All values obtained with PMSF, except for the slope factor in brain, were signi®cantly lower than those of DFP in cytosol from both tissues (Tables 2 and 3). Although mipafox is a relatively potent inhibitor of NTE activity, the ability of mipafox to block speci®c DFP binding was signi®cantly less than that of DFP (pKi values). Moreover, although mipafox exhibited weaker blocking of speci®c DFP binding in cytosol than in membrane, as shown by the slope factor and pKi values, the blocking effects of mipafox in spinal cord were relatively higher than that in brain (Table 2).

DISCUSSION Many investigations of DFP binding sites have been reported (Williams and Johnson, 1981; Pope and Padilla, 1989; Meredith and Johnson, 1989). Most of them have used the membrane preparations from neural tissue, probably because of the presence of membrane bound esterases such as AChE. In fact, it is true that the activities of ChE and NTE in cytosol are generally far lower than those in membrane (Sogorb et al., 1994). However, it is important to note that NTE is present in most tissues (Husain, 1994) and the DFP binding sites, other than NTE, may also be present in neural tissue. Therefore, to ®nd the speci®c [3 H]DFP binding site(s) other than the membrane-bound structures and to clarify the biochemical and pharmacological characteristics of non-membrane speci®c DFP binding site(s), we compared cytosol preparations from

200

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

brain and spinal cord of hens with the corresponding membrane preparations. In the present study, we found that [3 H]DFP binds not only to the membrane preparation but also to the cytosol preparation of brain and spinal cord. Carrington and Abou-Donia (1985) have reported that there are several DFP binding sites in both membrane and soluble fractions separated by sodium dodecyl sulfate/polyacrylamide-gel electrophoresis from hen brain. The existence of the DFP binding sites in cytosol may support the existence of new target site(s) of OPIDN. However, the Kd and Bmax values in cytosol were higher and lower, respectively, than those in membrane in present study. These results suggest that both the af®nity of [3 H]DFP binding and the number of binding sites in the cytosol preparation were lower than those of membrane, even though the cytosol preparation contains a large number of proteins. The Bmax values for speci®c DFP binding were similar in brain and spinal cord. However, it is important to note that the Kd values in both membrane and cytosol preparations from spinal cord were lower than those in brain. The higher af®nity of DFP binding in spinal cord than in brain is consistent with the distribution of neuropathologic lesions in OPIDN, which are seen in spinal cord and peripheral nerves but not in higher brain (Carrington et al., 1988). The organophosphorus compounds such as DFP, paraoxon and mipafox exhibited strong displacement of speci®c [3 H]DFP binding to the membrane preparations of both brain and spinal cord in vitro. The pKi values and the percentages of maximal displacement were very high in both brain and spinal cord membranes. These compounds are all organophosphates and the membrane preparation from neural tissue, including glial or satellite cells, is rich in ChEs such as AChE (Lefkkowitz et al., 1996). Membraneassociated AChE from hen and rat brain was previously reported to show similar kinetic constants (Vmax and Km) describing the hydrolysis of acetylthiocholine as a substrate (Kemp and Wallace, 1990). The reported IC50s of paraoxon for AChE activity in both hen and rat brain (2.8 and 2:6  10 8 M, respectively) would be consistent with our result of displacement by paraoxon (pKi, 7.46). Furthermore, in rat cerebral cortex, IC50s of DFP and eserine for AChE activity have also been reported to be comparatively low (5:2  10 7 and 1:25  10 9 M, respectively) (Ward et al., 1993; Hirai et al., 1997). Therefore, these organophosphate compounds would be expected to strongly inhibit speci®c DFP binding to the membrane preparation. The carbamates such as eserine would effectively compete with DFP for the hydroxyl residue of serine at the catalytic

center of ChE, because of a more potent ability in the formation of compound±enzyme complex than organophosphates. Therefore, the partial inhibitory effects of eserine on the [3 H]DFP binding to membrane shows that a portion of the DFP binding sites in the membrane preparation is due to ChE. In addition, paraoxon and eserine have both been reported to maximally displace [3 H]oxotremorine-methiodide (oxo-M) binding from membrane of rat brain stem and/or cortex by 63±83% (Van Den Beukel et al., 1997). Furthermore, anticholinesterase compounds, including paraoxon and DFP, maximally displaced [3 H]cis-methyldioxolane (CD) binding from membrane of rat frontal cortex and hippocampus by 50±80% (Ward et al., 1993). Because oxo-M and CD do not or hardly interfere with AChE activity from rat brain membrane preparations, paraoxon and eserine may also bind to muscarinic acetylcholine receptor which may also contribute to a small portion of speci®c [3 H]DFP binding sites in the membrane preparations. In contrast to the effects observed in membranes, the blocking effects of paraoxon and mipafox on speci®c DFP binding in cytosol preparations, especially in brain, were considerably lower than those of unlabelled DFP. As eserine also hardly had any inhibitory effect, the speci®c DFP binding site in cytosol is expected to be different from ChEs. DFP may play a different role from a ChE inhibitor in the cytosol preparations. The weak blocking effects of mipafox, a compound that produces OPIDN, on speci®c DFP binding in cytosol are not clearly consistent with cytosol DFP binding sites in the initiation of OPIDN. However, the blocking effects of mipafox on DFP binding in membrane and cytosol preparations of spinal cord were about 10-fold superior (pKi values) to those in brain. This suggests that a delayed neuropathic compound such as mipafox may interact with OPIDN-sensitive speci®c DFP binding sites in spinal cord rather than OPIDN-insensitive speci®c DFP binding sites in brain. Such a proposal is in agreement with the neuropathology of OPIDN. Therefore, the target site(s) for DFP induced delayed toxicity may be present in cytosol rather than on membrane, especially in spinal cord. The target site(s) in cytosol would be different from ChE. Although chlorpyrifos is an OP, the blocking of speci®c DFP binding on membrane and cytosol preparations of both tissues (slope factor and pKi values) was lower than those of other OPs used in this study, but the percentages of maximal displacement were higher than those of eserine. The acute toxicity of chlorpyrifos results from its reactive metabolite, termed `oxon-form' (Sultatos et al., 1984). Therefore,

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

the blocking effect of chlorpyrifos would require the conversion to the `oxon-form'. By comparison, the other OPs, which already exist in the `oxon-form', would be expected to show stronger blocking of DFP binding on ChE than chlorpyrifos. Chlorpyrifos has been reported to cause OPIDN in man (Lotti and Moretto, 1986) and in hen (Capodicasa et al., 1991). Capodicasa and coworkers (1991) concluded that the minimal neuropathic dose of chlorpyrifos was 60± 90 mg/kg p.o., corresponding to four to six times the estimated LD50. In the present study, chlorpyrifos may have displaced the DFP binding at only higher concentrations due to the characteristics as an organophosphate or as a delayed neuropathic compound. NTE has been reported to represent a small proportion of total phenyl valerate hydrolyzing activity in hen brains (Abou-Donia and Lapadula, 1990; Lotti, 1992). Phenyl valerate is also an arti®cial substrate of carboxylesterase, including NTE. In the present study, phenyl valerate and PMSF, an antagonist of NTE, weakly displaced the DFP binding to both membrane and cytosol preparations. This would be in agreement with the observation that NTE activities exist not only in membrane but also in cytosol preparations of hen brain and sciatic nerve (Vilanova et al., 1990). However, the inhibitory effects of phenyl valerate on the binding of DFP were observed at only higher concentrations in the present experiment. Conversely, the inhibitory effects of DFP on carboxylesterase and NTE activities have been reported at lower concentrations (Maxwell, 1992; Milatovic et al., 1997). The pKi value of phenyl valerate was about 4 in the present study, while the IC50 of DFP for NTE activity in both brain and sciatic nerve was 5  10 7 M (Milatovic et al., 1997). These indicate that carboxylesterase activity, including NTE, has high sensitivity to DFP but phenyl valerate has only low af®nity to speci®c DFP binding sites. Therefore, carboxylesterase and NTE may not be speci®c binding sites but a part of non-speci®c binding sites. In addition, the low inhibitory effects of PMSF on the binding of DFP would be in agreement with this hypothesis. The IC50s of PMSF on NTE activity in both brain and sciatic nerve were 1  10 4 M (Milatovic et al., 1997), which would be consistent with our results of displacement by PMSF (pKi, about 4±5). The low pKi values and partial displacement by PMSF suggest that NTE is different from the main part of the speci®c DFP binding sites. Therefore, both NTE and carboxylesterase may not play an important role in the initiation of DFP induced delayed toxicity. In summary, we demonstrated the existence of the speci®c binding sites of [3 H]DFP not only in

201

membrane but also in cytosol preparations from brain and spinal cord of hen. Furthermore, we attempted to clarify a correlation of the speci®c binding sites of DFP with ChE and/or carboxylesterase, including NTE. We found that the binding sites on membrane corresponded with the active sites of ChE but the binding sites in cytosol were different from the active sites of both ChE, carboxylesterase and NTE. Therefore, the target site(s) for the initiation of DFP induced delayed toxicity may be present in cytosol rather than on membrane, especially in spinal cord. The speci®c binding sites of DFP in cytosol, which were found in the present study, may play a signi®cant role in the initiation of OPIDN without the inhibition of ChE and NTE. ACKNOWLEDGEMENTS The authors wish to thank Dr. Donald E. Moss for reviewing the manuscript and valuable advise, and Ms. Kaori Yotsuya for useful advice on statistical analysis. REFERENCES Abou-Donia MB. Organophosphorus ester-induced delayed neurotoxicity. Annu Rev Pharmacol Toxicol 1981;21:511±48. Abou-Donia MB, Lapadula DM. Mechanisms of organophosphorus ester-induced delayed neurotoxicity: type I and type II. Annu Rev Pharmacol Toxicol 1990;30:405±40. Bruns RF, Lawson-Wendling K, Pugsley TA. A rapid filtration assay for soluble receptors using polyethylenimine-treated filters. Anal Biochem 1983;132:74±81. Capodicasa E, Scapellato ML, Moretto A, Caroldi S, Lotti M. Chlorpyrifos-induced delayed polyneuropathy. Arch Toxicol 1991;65:150±5. Carrington CD, Abou-Donia MB. Characterization of [3 H]diiso3 H]diisopropyl phosphorofluoridate-binding proteins in hen brain. Rates of phosphorylation and sensitivity to neurotoxic and non-neurotoxic organophosphorus compounds. Biochem J 1985;228:537±44. Carrington CD, Brown HR, Abou-Donia MB. Histopathological assessment of triphenyl phosphite neurotoxicity in the hen. NeuroToxicology 1988;9:223±34. Cheng YC, Prusoff WH. Relationship between the inhibition constant (Ki) and the concentration of inhibitor which causes 50 per cent inhibition (IC50) of an enzymatic reaction. Biochem Pharmacol 1973;22:3039±46. Hirai K, Kato K, Nakayama T, Hayako H, Ishihara Y, Goto G, Miyamoto M. Neurochemical effects of 3-[1-(phenylmethyl)-4piperidinyl]-1-(2,3,4,5-tetrahydro-1H-1-benzazepin-8-yl)-1propanone fumarate (TAK-147), a novel acetylcholinesterase inhibitor, in rats. J Pharmacol Exp Therap 1997;280:1261±9. Husain K. Neurotoxcic esterase. Asia Pacific J Pharmacol 1994;9:119±28. Johnson MK. Organophosphorus and other inhibitors of brain

202

R. Kamata et al. / NeuroToxicology 22 (2001) 191±202

``neurotoxic esterase'' and the development of delayed neurotoxicity in hens. Biochem J 1970;120:523±31. Johnson MK. Improved assay of neurotoxic esterase for screening organophosphates for delayed neurotoxicity potential. Arch Toxicol 1977;37:113±5. Johnson MK. The target for initiation of delayed neurotoxicity by organophosphorus esters: biochemical studies and toxicological applications. Rev Biochem Toxicol 1982;21:141±212. Johnson MK. Organophosphates and delayed neuropathy Ð is NTE alive and well? Toxicol Appl Pharmacol 1990;102:385± 99. Kamata R, Suzuki T, Saito S, Kofujita H, Ota M, Kobayashi H. Lack of correlation of organophosphorus-induced delayed neuropathy with neuropathy target esterase in hens and Japanese quails. J Health Sci 1999;45:209±16. Kemp JR, Wallace KB. Molecular determinations of the speciesselective inhibition of brain acetylcolinesterase. Toxicol Appl Pharmacol 1990;104:246±58. Konno N, Horiguchi H, Fukushima M. Delayed neurotoxicity of diisopropylfluorophosphate (DFP): autoradiographic localization of high-affinity [3 H]DFP binding sites in the chicken spinal cord. Environ Health Prev Med 1999;4:92±6. Konno N, Suzuki N, Horiguchi H, Fukushima M. Characterization of high-affinity binding sites for diisopropylfluorophosphate (DFP) from chicken spinal cord membranes. Biochem Pharmacol 1994;48:2073±9. Lefkkowitz RJ, Hoffmen BB, Taylor P. Neurotransmission: the autonomic and somatic motor nervous systems. In: Hardman JG, Limbrid LE, Gilman AG, editors. Goodman and Gilman's the pharmacological basis of therapeutics. 9th ed. New York: McGraw-Hill, 1996. p. 105±39. Lotti M. The pathogenesis of organophosphate polyneuropathy. Toxicology 1992;21:455±87. Lotti M, Moretto A. Inhibition of lymphocyte neuropathy target esterase predicts the development of organophosphate polyneuropathy in man. Hum Toxicol 1986;5:114. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem

1951;193:265±75. Maxwell DM. The specificity of carboxylesterase protection against the toxicity of organophosphorus compounds. Toxicol Appl Pharmacol 1992;114:306±12. Meredith C, Johnson MK. Species distribution of paraoxonresistant brain polypeptides radiolabelled with diisopropyl phosphorofluoridate ([3 H]DiPF): electrophoretic assay for the aged polypeptide of [3 H]DiPF-labelled neuropathy target esterase. J Neurochem 1989;52:1248±52. Milatovic D, Moretto A, Osman KA, Lotti M. Phenyl valerate esterases other than neuropathy target esterase and the promotion of organophosphate polyneuropathy. Chem Res Toxicol 1997;10:1045±8. Pope CN, Padilla SS. Chromatographic characterization of neurotoxic esterase. Biochem Pharmacol 1989;38:181±8. Sogorb MA, Viniegra S, Reig JA, Vilanova E. Partial characterization of neuropathy target esterase and related phenyl valerate esterases from bovine adrenal medulla. J Biochem Toxicol 1994;9:145±52. Sultatos LG, Shao M, Murphy SD. The role of hepatic biotransformation in mediating the acute toxicity of the phosphorothionate insecticide chlorpyrifos. Toxicol Appl Pharmacol 1984;73:60±8. Van Den Beukel I, Dijcks FA, Vanderheyden P, Vauquelin G, Oortgiesen M. Different muscarinic receptor binding of acetylcholinesterase inhibitors in rat brain, human brain and Chinese hamster ovary cells expressing human receptors. J Pharmacol Exp Therap 1997;281:1113±9. Vilanova E, Barril J, Carrera V, Pellin MC. Soluble and particulate forms of the organophosphorus neuropathy target esterase in hen sciatic nerve. J Neurochem 1990;55:1258±65. Ward TR, Ferris DJ, Tilson HA, Mundy WR. Correlation of the anticholinesterase activity of a series of organophosphates with their ability to compete with agonist binding to muscarinic receptors. Toxicol Appl Pharmacol 1993;122:300±7. Williams DG, Johnson MK. Gel-electrophoretic identification of hen brain neurotoxic esterase, labelled with tritiated diisopropyl phosphorofluoridate. Biochem J 1981;199:323±33.