Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 145 (2015) 198–202
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa
Partitioning of organophosphorus pesticides into phosphatidylcholine small unilamellar vesicles studied by second-derivative spectrophotometry Shigehiko Takegami ⇑, Keisuke Kitamura, Mayuko Ohsugi, Aya Ito, Tatsuya Kitade Department of Analytical Chemistry, Kyoto Pharmaceutical University, 5 Nakauchicho, Misasagi, Yamashina-ku, Kyoto 607-8414, Japan
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
Derivative spectra of pesticides in
a r t i c l e
i n f o
Article history: Received 21 February 2014 Received in revised form 16 February 2015 Accepted 17 February 2015 Available online 3 March 2015 Keywords: Organophosphorus pesticide Partition coefficient Liposome Second-derivative spectrophotometry
5 Profenofos
4.5 Chlorpyrifos-methyl
log Poct
liposome suspensions showed isosbestic points. Partition coefficients (Kps) of 8 pesticides were determined as a lipophilic index. Kp values reflected the chemical structure difference of the pesticides. There was no linear relationship between the log Kp and log Poctanol values. Kp is a good index to predict the bioaccumulation of organophosphorus pesticides.
Isofenphos
4
Chlorfenvinphos Diazinon
Fenthion Pyraclofos
3.5 Fenitrothion
3 3
3.5
4
4.5
5
log Plip
a b s t r a c t In order to quantitatively examine the lipophilicity of the widely used organophosphorus pesticides (OPs) chlorfenvinphos (CFVP), chlorpyrifos-methyl (CPFM), diazinon (DZN), fenitrothion (FNT), fenthion (FT), isofenphos (IFP), profenofos (PFF) and pyraclofos (PCF), their partition coefficient (Kp) values between phosphatidylcholine (PC) small unilamellar vesicles (SUVs) and water (liposome–water system) were determined by second-derivative spectrophotometry. The second-derivative spectra of these OPs in the presence of PC SUV showed a bathochromic shift according to the increase in PC concentration and distinct derivative isosbestic points, demonstrating the complete elimination of the residual background signal effects that were observed in the absorption spectra. The Kp values were calculated from the second-derivative intensity change induced by addition of PC SUV and obtained with a good precision of R.S.D. below 10%. The Kp values were in the order of CPFM > FT > PFF > PCF > IFP > CFVP > FNT P DZN and did not show a linear correlation relationship with the reported partition coefficients obtained using an n-octanol–water system (R2 = 0.530). Also, the results quantitatively clarified the effect of chemicalgroup substitution in OPs on their lipophilicity. Since the partition coefficient for the liposome–water system is more effective for modeling the quantitative structure–activity relationship than that for the n-octanol–water system, the obtained results are toxicologically important for estimating the accumulation of these OPs in human cell membranes. Ó 2015 Elsevier B.V. All rights reserved.
Introduction ⇑ Corresponding author. Tel.: +81 75 5954659; fax: +81 75 5954760. E-mail address: [email protected] (S. Takegami). http://dx.doi.org/10.1016/j.saa.2015.02.061 1386-1425/Ó 2015 Elsevier B.V. All rights reserved.
The development and spread of synthetic pesticides have greatly contributed to an increase of food production worldwide.
199
S. Takegami et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 145 (2015) 198–202
One family of synthetic pesticides in particular, the organophosphorus pesticides (OPs), have been widely used for agricultural and domestic purposes in the control of insect pests. Given their widespread use, these agents have very high potential for human exposure and uptake, and they have been shown to affect most organs in the human body. In general, the molecular mechanism of the acute toxicity of OPs has been shown to involve the powerful inhibition of acetylcholinesterase. In addition, due to their lipophilic nature, OPs accumulate in the membranes of living cells and can thereby cause chronic toxicity. Therefore, it is crucial to investigate the interaction between OPs and biomembranes in order to obtain fundamental information related to their toxic effects. The n-octanol–water partition coefficient (log Poct) is widely used as a simple model system to evaluate the toxicity and bioconcentration factors (BCFs) of pesticides [1–3], although the n-octanol–water system has some thermodynamic differences from the actual uptake by biota due to the use of n-octanol as the model biological phase [4]. In fact, in the quantitative structure–activity relationship (QSAR) studies of drugs, it has been suggested that the partition coefficients obtained for the lipid bilayer vesicles (liposomes)-water system are more effective than log Poct [5–7]. Accordingly, previous studies have examined the partitioning of pesticides [8] and endocrine disrupting compounds [9] between lipid vesicles and water. However, in these studies, the partition coefficients of chemicals in the liposome–water system were determined by equilibrium dialysis and filtration techniques [10–12], which require difficult separation procedures that may disturb the equilibrium states of chemicals between the lipid bilayer membranes and water. In contrast, since derivative spectrophotometry can eliminate the effect of background signals caused by the light-scattering of vesicles [13], it has been successfully employed to determine the liposome–water partition coefficient without disturbing their equilibrium condition, e.g., second-derivative spectrophotometry has been applied to determine the partition coefficients (Kps) of psychotropic phenothiazines [14–18], benzodiazepines [19,20], anti-inflammatory steroid drugs [21], non-steroidal anti-inflammatory drugs (NSAIDs) [22], and several other drugs [23,24] between the phosphatidylcholine (PC) small unilamellar vesicles (SUVs) and water. In this study, we determined the Kps of eight OPs, chlorfenvinphos (CFVP), chlorpyrifos-methyl (CPFM), diazinon (DZN), fenitrothion (FNT), fenthion (FT), isofenphos (IFP), profenofos (PFF) and pyraclofos (PCF), between PC SUV and water by using second-derivative spectrophotometry.
where DDmax is the maximum DD value assuming all OPs are partitioned in PC SUV. The values of Kp and DDmax can be calculated from the experimental values of [L] and DD by applying a non-linear least-squares calculation to Eq. (2) [14]. The calculation was performed using a laboratory-made Visual Basic program on a personal computer. Chemicals and reagents CFVP, CPFM, DZN, FNT, FT, IFP, PFF and PCF (Fig. 1) were purchased from Supelco Inc. (Bellefonte, PA, USA). The buffer used was 50 mmol L1 NaCl-10 mmol L1 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (Hepes buffer, pH 7.4, 37 °C). L-a-PC (egg yolk) of 99% purity was supplied as a 2% (w/v) chloroform solution from Avanti Polar-Lipids Inc. (Alabaster, AL, USA) and used without further purification after the purity of PC was confirmed by thin-layer chromatography. Preparation of PC SUV Appropriate amounts of the PC stock solution were mixed and dried by using a rotary evaporator and then a vacuum pump. To the residue, 5 ml of the buffer was added to yield a PC concentration of ca. 40 mmol L1, and the mixture was vortexed to produce multilamellar vesicles. Then SUV was prepared by the sonication method as previously reported [14]. Measurement of the mean diameter and zeta potential of PC SUV The SUV size distribution was determined by a dynamic light scattering method using a submicron particle analyzer (Nicomp Model 380; Particle Sizing Systems, Santa Barbara, CA, USA) [15]. The zeta potential was measured by ZEECOM ZC-3000 (Microtec Co., Ltd., Chiba, Japan), which was based on the principle of electrophoresis. Phosphorus determination The exact PC concentration in SUV suspensions was calculated from phosphate analysis according to the phosphovanadomolybdate method [26]. O O Cl CH
Cl S Cl
O
Cl
P
N
O
CH3
O
CH3
Cl
S N
Chlorpyrifos-methyl (CPFM)
O
P
N
O
C2H5
O
C2H5
S O2N
(H3C) 2HC
O
P
O
CH3
O
CH3
H 3C
Fenitrothion (FNT)
Diazinon (DZN)
ð1Þ
where [OPm] and [OPw] represent the molar concentrations of OP in PC SUV and water, respectively, and [OPt] = [OPm] + [OPw], and [L] and [W] are molar concentrations of PC in SUV and water (55.3 mol L1 at 37 °C), respectively. If the background signal effect based on PC SUV is eliminated in the second-derivative spectra, the derivative intensity difference (DD) of OP before and after the addition of PC SUV at a specific wavelength is proportional to the concentration of OP in PC SUV. As described in a previous paper [14], the following Eq. (2) can be derived from Eq. (1):
K p DDmax ½L ½W þ K p ½L
C2H5
H3C
The molar partition coefficient (Kp) of an OP between PC SUV and water was defined as follows [14,25]:
DD ¼
C2H5
O
Chlorfenvinphos (CFVP)
Calculation of molar partition coefficients
ð½OPm =½OPt Þ=½L ð½OPw =½OPt Þ=½W
O
Cl
Methods and materials
Kp ¼
P
ð2Þ
S H3C
S
O
P
O
CH3
O
CH3
S O
H3C
P
NHCH(CH 3)2 O
C2H5
COOCH(CH 3)2
Isofenphos (IFP)
Fenthion (FT)
O Br
O
P
S
C3H7
O
C2H5
Cl
N N
O O
P
S
C3H7
O
C2H5
Cl
Profenofos (PFF)
Pyraclofos (PCF)
Fig. 1. Chemical structures of the eight organophosphorus pesticides studied.
200
S. Takegami et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 145 (2015) 198–202
Measurements of absorption and second-derivative spectra
Absorption and second-derivative spectra
Sample solutions with a suitable OP concentration in the range of 10–50 lmol L1 and various amounts of the SUV suspension were prepared in a manner similar to that of the previous papers [14,15]. The reference solutions were those prepared without OP. Each flask was shaken for a short time and incubated at 37 °C for 30 min. The absorption spectrum of the sample solution was measured against the reference solution by using a spectrophotometer (Hitachi U-3310) equipped with a temperature-regulated cell holder in a 1-cm light-pass length cuvette at 37 °C with a band pass of 2 nm, wavelength interval of 0.1 nm, and scan speed of 30 nm min1. The second-derivative spectra were calculated based on the Savitzky-Golay method [27] using a laboratory-made Visual Basic program [28]. The cubic polynomial convolution of 17 points and the wavelength interval (Dk) of 0.8 (CFVP), 1.0 (DZN and FT), and 1.5 (CPFM, FNT, IFP, PFF and PCF) nm were used in the calculation, respectively.
The absorption spectra of 10 lmol L1 FT and 20 lmol L1 PCF in the sample solutions containing various amounts of PC SUV at 37 °C are shown in Fig. 2. As seen in Fig. 2A and B, both spectra show spectral changes according to the increase of PC concentration. However, no isosbestic points could be observed because of the incomplete baseline compensation due to the intense light scattering of PC SUV. In general, when strong background signals are present, it is difficult to cancel their effects completely in order to obtain a flat and zero-level baseline. Thus, further spectral data for calculating the Kp values could not be obtained from these absorption spectra. The second-derivative spectra calculated from the absorption spectra in Fig. 2 are shown in Fig. 3A and B, respectively. In these spectra, derivative isosbestic points can be clearly seen at 251 and 263 nm for FT in Fig. 3A, and at 264 and 294 nm for PCF in Fig. 3B, thus confirming that the influences of the residual background signal of PC SUVs are entirely eliminated in the secondderivative spectra, and that FT and PCF exist in two states [29], i.e., in the bulk water and the PC bilayer of SUV. Similar results were obtained for all of the other OPs examined.
Results and discussion Characteristics of phosphatidylcholine small unilamellar vesicles
Calculated Kp values The PC SUVs prepared by sonication in 50 mmol L1 NaCl-10 mmol L1 Hepes buffer (pH 7.4) were characterized as follows; the diameters of more than 90% of the SUVs were in the range of 20–30 nm, and their zeta potential values and the final PC concentrations were 10 ± 2 mV and 38 ± 1 mmol L1, respectively.
0.5
Absorbance
A
(4) (5) (7)
(1) (2)
B
0.4
(1) Absorbance
0.2
0.16
The DD values for eight OPs were obtained from the derivative values at the wavelengths of 263 (CFVP), 300 (CPFM), 267 (DZN), 268 (FNT), 256 (FT), 287 (IFP), 251 (PFF) and 275 (PCF) nm, respectively. Using the DD values, the Kp and DDmax values were
0.12
(3) (6) 0.08
0.3
0.2
(7) 0.1
0.04
0
0
240
250
260
270
280
290
300
310
320
245
255
265
275
285
295
305
315
Wavelength (nm)
Wavelength (nm)
Fig. 2. Absorption spectra of 10 lmol L1 FT (A) and 20 lmol L1 PCF (B) in Hepes buffer (pH 7.4, 37 °C) varying concentrations of PC SUV. PC concentration (mmol L1): (A) (1) 0.000, (2) 0.057, (3) 0.114, (4) 0.285, (5) 0.456, (6) 0.855, (7) 1.710; (B) (1) 0.000, (2) 0.105, (3) 0.210, (4) 0.420, (5) 0.630, (6) 1.050, (7) 2.100 (in the direction of the arrow).
0.0012
0.0015
A
0.0009 0.0006
0.0005
0
(1)
-0.0003
2nd-Der.
2nd-Der.
0.0003
-0.0006 -0.0009
(1)
0 -0.0005 -0.001
-0.0012 -0.0015 -0.0018 -0.0021
B
0.001
-0.0015
(7) 240
250
260
270
280
290
Wavelength (nm)
300
310
320
-0.002
(7) 245
255
265
275
285
295
305
315
Wavelength (nm)
Fig. 3. Second-derivative spectra of FT (A) and PCF (B) calculated from the absorption spectra in Fig. 2A and B, respectively. The curve labels are equivalent to those in Fig. 2.
201
S. Takegami et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 145 (2015) 198–202 Table 1 The mean Kp values of eight OPs calculated from the Kp values at all OP concentrations in Fig. 4.
8 7
Kp (×10-5)
6 5 4 3 2 1 0 0
10
20
30
40
50
60
calculated. The Kp values of the eight OPs at various OP concentrations are plotted in Fig. 4; their relative standard deviations were below 10%, indicating the good precision of the second-derivative method. It is known that the Kp values of certain drugs show drug-concentration dependence due to the molecular association or formation of their micelles [11,30]. As seen in Fig. 4, similar Kp values were obtained for each OP within the range of OP concentrations employed, although the Kp values of IFP and PCF were slightly reduced with increasing OP concentration. Thus, the results demonstrate that neither of these OPs associated or formed micelles in the bulk water, which in turn confirmed the validity of applying partition theory to account for the interactions between these OPs and the PC bilayer. In Fig. 5, the fractions of eight OPs partitioned to the PC SUVs are shown as a plot of the DD/DDmax values versus the PC concentration. Solid lines represent the theoretical curves calculated from Eq. (2) using the obtained Kp and DDmax values. The experimental values for each OP show a good fit with the calculated curves, indicating the validity of the obtained Kp values. Similar results were obtained for the other OP concentrations employed. The means of the Kp values obtained at various OP concentrations are also shown in Table 1, and were in the order of CPFM > FT > PFF > PCF > IFP > CFVP > FNT P DZN. The results quantitatively reveal that the substitution of a chemical group in FNT affects its Kp value; i.e., FT, which is given by substitution of a thiomethyl group for the nitro group at the
Kp (105)a
CFVP CPFM DZN FNT FT IFP PCF PFF
1.38 ± 0.18 6.57 ± 0.32 0.98 ± 0.15 1.00 ± 0.15 5.17 ± 0.33 2.65 ± 0.33 3.78 ± 0.42 4.37 ± 0.38
a Each value is expressed as the mean ± S.D. (n = 6–12).
OP concentration (µ µmolL-1) Fig. 4. Dependence of the Kp values of eight organophosphorus pesticides on their concentrations. The plotted Kp values are the averages of independent triplicate determinations. Symbols: (d) CFVP, (N) CPFM, (j) PFF, () PCF, (s) DZN, (4) FNT, (h) FT, (}) IFP.
OPs
C4 position of FNT, has an approximately 5-fold greater Kp value than FNT itself. In addition, the OPs having halogen atoms (Cl and Br) in their molecular structures (closed symbols), e.g., CPFM, PFF and PCF, tend to show larger Kp values than the OPs without halogen atoms (open symbols) as shown in Figs. 4 and 5. Relationship between partition coefficients for the liposome–water system and the n-octanol–water system In order to examine the relationship between the partition coefficients for the liposome–water system and those for the n-octanol–water system, the Kp values of these OPs in Table 1 were converted to the Plip values based on the membrane and water volumes; i.e., Plip was calculated by dividing each Kp value by 42: Kp/ 42 = Plip [25]. In Fig. 6, the log Plip values of the eight OPs were plotted against the corresponding reported log Poct values [31], and the results show a rather scatted figure with an R2 value of 0.530 (the solid line shows the linear correlation equation of y = 0.915x + 0.519). Due to the difference in the partitioning mechanisms of the n-octanol–water system (from water to n-octanol) and the liposome–water system (from water to lipid membrane), the correlation between the log Plip and log Poct values was poor, as seen in Fig. 6. Meanwhile, the dotted line in Fig. 6 indicates equal partitioning in the liposome–water and n-octanol–water systems. Four OPs, CFVP, DZN, IFP and PFF, show high log Poct values compared with their log Plip values. In the liposome–water system, many drugs depend not only on their hydrophobicity but also on their specific interaction with membrane lipids. In fact, although the log Poct
5 1
log Poct
Fraction (-)
4.5
0.5
4
3.5 0
0
1
2
3
PC concentration (mmolL-1)
Fig. 5. Fraction (DD/DDmax) of CFVP (d), CPFM (N), PFF (j), PCF (), DZN (s), FNT (4), FT (h) and IFP (}) in PC SUV as a function of PC concentration. The solid lines show the theoretical curves calculated from Eq. (2) using the experimental values of Kp and DDmax. The symbols are the experimental values. The OP concentrations were as follows (lmol L1): 40 (CFVP), 10 (CPFM), 40 (PFF), 30 (PCF), 40 (DZN), 30 (FNT), 10 (FT), 40 (IFP).
3 3
3.5
4
4.5
5
log Plip Fig. 6. The relationship between log Kp and log P of eight organophosphorus pesticides. Symbols: (d) CFVP, (N) CPFM, (j) PFF, () PCF, (s) DZN, (4) FNT, (h) FT, (}) IFP.
202
S. Takegami et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 145 (2015) 198–202
values have been used as a standard hydrophobic index to predict their bioactivity of drugs, they generally produce poor correlations between bioaccumulation and biotransport in living systems. Therefore, the obtained log Plip values for the eight OPs may be a better index to predict the accumulation of OPs related to chronic toxicity in human cell membranes rather than the log Poct values determined using the n-octanol–water system. Conclusions In conclusion, the Kp values of eight organophosphorus pesticides between PC SUV and water were accurately and precisely determined by the second-derivative spectrophotometric method and were in the order of CPFM > FT > PFF > PCF > IFP > CFVP > FNT P DZN. Since the partition coefficient for the liposome–water system is more effective for modeling the quantitative structure– activity relationship than that for the n-octanol–water system, the obtained results are toxicologically important to estimate the accumulation of these organophosphorus pesticides in human cell membranes. References [1] J. de Bruijn, J. Hermens, Environ. Toxicol. Chem. 10 (1991) 791–804. [2] H. Saito, J. Koyasu, K. Yoshida, T. Shigeoka, S. Koike, Chemosphere 26 (1993) 1015–1028. [3] T. Tsuda, M. Kojima, H. Harada, A. Nakajima, S. Aoki, Biochem. Physiol. 116C (1997) 213–218. [4] A. Opperhuizen, P. Serne, J.M.D. van der Stern, Environ. Sci. Technol. 22 (1988) 286–292. [5] Y.W. Choi, J.A. Rogers, Pharm. Res. 7 (1990) 508–512. [6] H. Fujiwara, Y.Z. Da, K. Ito, T. Takagi, Y. Nishioka, Bull. Chem. Soc. Jpn. 64 (1991) 3707–3712.
[7] J.A. Rogers, Y.W. Choi, Pharm. Res. 10 (1993) 913–917. [8] M.C. Antunes-Madeira, V.M.C. Madeira, Biochim. Biophys. Acta 778 (1984) 49– 56. [9] H. Yamamoto, H.M. Liljestrand, Environ. Sci. Technol. 38 (2004) 1139–1147. [10] B. Bondy, J. Remien, Life Sci. 28 (1981) 441–449. [11] M. Luxnat, H.J. Galla, Biochim. Biophys. Acta 856 (1986) 274–282. [12] A. Zachowski, P. Durand, Biochim. Biophys. Acta 937 (1988) 411–416. [13] K. Kitamura, M. Takenaka, S. Yoshida, M. Ito, Y. Nakamura, K. Hozumi, Anal. Chim. Acta 242 (1991) 131–135. [14] K. Kitamura, N. Imayoshi, T. Goto, H. Shiro, T. Mano, Y. Nakai, Anal. Chim. Acta 304 (1995) 101–106. [15] S. Takegami, K. Kitamura, T. Kitade, K. Hasegawa, A. Nishihira, J. Colloid Interface Sci. 220 (1999) 81–87. [16] S. Takegami, K. Kitamura, T. Kitade, A. Kitagawa, K. Kawamura, Chem. Pharm. Bull. 51 (2003) 1056–1059. [17] A. Pola, K. Michalak, A. Burliga, N. Motohashi, M. Kawase, Eur. J. Pharm. Sci. 21 (2004) 421–427. [18] S. Takegami, K. Kitamura, T. Kitade, M. Takashima, M. Ito, E. Nakagawa, M. Sone, R. Sumitani, Y. Yasuda, Chem. Pharm. Bull. 53 (2005) 147–150. [19] A.A. Omran, K. Kitamura, S. Takegami, T. Kitade, A.Y. El-Sayed, M. AbdelMottaleb, J. Pharm. Biomed. Anal. 25 (2001) 319–324. [20] A.A. Omran, K. Kitamura, S. Takegami, T. Kitade, A.Y. El-Sayed, M.H. Mohamed, Abdel-Mottaleb, Chem. Pharm. Bull. 50 (2002) 312–315. [21] S. Takegami, K. Kitamura, T. Funakoshi, T. Kitade, Chem. Pharm. Bull. 56 (2008) 663–667. [22] H. Ferreira, M. Lucio, B. de Castro, P. Gameiro, J.L.F.C. Lima, S. Reis, Anal. Bioanal. Chem. 377 (2003) 293–298. [23] C. Rodrigues, P. Gameiro, S. Reis, J.L.F.C. Lima, B. de Castro, Biophys. Chem. 94 (2001) 97–106. [24] S. Takegami, A. Otani, Y. Otake, T. Kitade, Curr. Pharm. Anal. 8 (2012) 286–294. [25] R. Welti, L.J. Mullikin, T. Yoshimura, G.M. Helmkamp Jr., Biochemistry 23 (1984) 6086–6091. [26] A.J. Christopher, T.R.F.W. Fennell, Microchem. J. 12 (1967) 593–605. [27] A. Savitzky, M.J.E. Golay, Anal. Chem. 36 (1964) 1627–1639. [28] K. Kitamura, K. Hozumi, Anal. Chim. Acta 172 (1985) 111–118. [29] K.A. Connors, Binding Constants, John Wiley Sons, New York, 1987. pp. 142– 147. [30] W.N. Charman, C.S. Lai, B.C. Finnin, B.L. Reed, Pharm. Res. 8 (1991) 1144–1150. [31] C. Hansch, A. Leo, D. Hoekman, Exploring QSAR, American Chemical Society, Washington DC, 1995.
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa
Partitioning of organophosphorus pesticides into phosphatidylcholine small unilamellar vesicles studied by second-derivative spectrophotometry Shigehiko Takegami ⇑, Keisuke Kitamura, Mayuko Ohsugi, Aya Ito, Tatsuya Kitade Department of Analytical Chemistry, Kyoto Pharmaceutical University, 5 Nakauchicho, Misasagi, Yamashina-ku, Kyoto 607-8414, Japan
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
Derivative spectra of pesticides in
a r t i c l e
i n f o
Article history: Received 21 February 2014 Received in revised form 16 February 2015 Accepted 17 February 2015 Available online 3 March 2015 Keywords: Organophosphorus pesticide Partition coefficient Liposome Second-derivative spectrophotometry
5 Profenofos
4.5 Chlorpyrifos-methyl
log Poct
liposome suspensions showed isosbestic points. Partition coefficients (Kps) of 8 pesticides were determined as a lipophilic index. Kp values reflected the chemical structure difference of the pesticides. There was no linear relationship between the log Kp and log Poctanol values. Kp is a good index to predict the bioaccumulation of organophosphorus pesticides.
Isofenphos
4
Chlorfenvinphos Diazinon
Fenthion Pyraclofos
3.5 Fenitrothion
3 3
3.5
4
4.5
5
log Plip
a b s t r a c t In order to quantitatively examine the lipophilicity of the widely used organophosphorus pesticides (OPs) chlorfenvinphos (CFVP), chlorpyrifos-methyl (CPFM), diazinon (DZN), fenitrothion (FNT), fenthion (FT), isofenphos (IFP), profenofos (PFF) and pyraclofos (PCF), their partition coefficient (Kp) values between phosphatidylcholine (PC) small unilamellar vesicles (SUVs) and water (liposome–water system) were determined by second-derivative spectrophotometry. The second-derivative spectra of these OPs in the presence of PC SUV showed a bathochromic shift according to the increase in PC concentration and distinct derivative isosbestic points, demonstrating the complete elimination of the residual background signal effects that were observed in the absorption spectra. The Kp values were calculated from the second-derivative intensity change induced by addition of PC SUV and obtained with a good precision of R.S.D. below 10%. The Kp values were in the order of CPFM > FT > PFF > PCF > IFP > CFVP > FNT P DZN and did not show a linear correlation relationship with the reported partition coefficients obtained using an n-octanol–water system (R2 = 0.530). Also, the results quantitatively clarified the effect of chemicalgroup substitution in OPs on their lipophilicity. Since the partition coefficient for the liposome–water system is more effective for modeling the quantitative structure–activity relationship than that for the n-octanol–water system, the obtained results are toxicologically important for estimating the accumulation of these OPs in human cell membranes. Ó 2015 Elsevier B.V. All rights reserved.
Introduction ⇑ Corresponding author. Tel.: +81 75 5954659; fax: +81 75 5954760. E-mail address: [email protected] (S. Takegami). http://dx.doi.org/10.1016/j.saa.2015.02.061 1386-1425/Ó 2015 Elsevier B.V. All rights reserved.
The development and spread of synthetic pesticides have greatly contributed to an increase of food production worldwide.
199
S. Takegami et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 145 (2015) 198–202
One family of synthetic pesticides in particular, the organophosphorus pesticides (OPs), have been widely used for agricultural and domestic purposes in the control of insect pests. Given their widespread use, these agents have very high potential for human exposure and uptake, and they have been shown to affect most organs in the human body. In general, the molecular mechanism of the acute toxicity of OPs has been shown to involve the powerful inhibition of acetylcholinesterase. In addition, due to their lipophilic nature, OPs accumulate in the membranes of living cells and can thereby cause chronic toxicity. Therefore, it is crucial to investigate the interaction between OPs and biomembranes in order to obtain fundamental information related to their toxic effects. The n-octanol–water partition coefficient (log Poct) is widely used as a simple model system to evaluate the toxicity and bioconcentration factors (BCFs) of pesticides [1–3], although the n-octanol–water system has some thermodynamic differences from the actual uptake by biota due to the use of n-octanol as the model biological phase [4]. In fact, in the quantitative structure–activity relationship (QSAR) studies of drugs, it has been suggested that the partition coefficients obtained for the lipid bilayer vesicles (liposomes)-water system are more effective than log Poct [5–7]. Accordingly, previous studies have examined the partitioning of pesticides [8] and endocrine disrupting compounds [9] between lipid vesicles and water. However, in these studies, the partition coefficients of chemicals in the liposome–water system were determined by equilibrium dialysis and filtration techniques [10–12], which require difficult separation procedures that may disturb the equilibrium states of chemicals between the lipid bilayer membranes and water. In contrast, since derivative spectrophotometry can eliminate the effect of background signals caused by the light-scattering of vesicles [13], it has been successfully employed to determine the liposome–water partition coefficient without disturbing their equilibrium condition, e.g., second-derivative spectrophotometry has been applied to determine the partition coefficients (Kps) of psychotropic phenothiazines [14–18], benzodiazepines [19,20], anti-inflammatory steroid drugs [21], non-steroidal anti-inflammatory drugs (NSAIDs) [22], and several other drugs [23,24] between the phosphatidylcholine (PC) small unilamellar vesicles (SUVs) and water. In this study, we determined the Kps of eight OPs, chlorfenvinphos (CFVP), chlorpyrifos-methyl (CPFM), diazinon (DZN), fenitrothion (FNT), fenthion (FT), isofenphos (IFP), profenofos (PFF) and pyraclofos (PCF), between PC SUV and water by using second-derivative spectrophotometry.
where DDmax is the maximum DD value assuming all OPs are partitioned in PC SUV. The values of Kp and DDmax can be calculated from the experimental values of [L] and DD by applying a non-linear least-squares calculation to Eq. (2) [14]. The calculation was performed using a laboratory-made Visual Basic program on a personal computer. Chemicals and reagents CFVP, CPFM, DZN, FNT, FT, IFP, PFF and PCF (Fig. 1) were purchased from Supelco Inc. (Bellefonte, PA, USA). The buffer used was 50 mmol L1 NaCl-10 mmol L1 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (Hepes buffer, pH 7.4, 37 °C). L-a-PC (egg yolk) of 99% purity was supplied as a 2% (w/v) chloroform solution from Avanti Polar-Lipids Inc. (Alabaster, AL, USA) and used without further purification after the purity of PC was confirmed by thin-layer chromatography. Preparation of PC SUV Appropriate amounts of the PC stock solution were mixed and dried by using a rotary evaporator and then a vacuum pump. To the residue, 5 ml of the buffer was added to yield a PC concentration of ca. 40 mmol L1, and the mixture was vortexed to produce multilamellar vesicles. Then SUV was prepared by the sonication method as previously reported [14]. Measurement of the mean diameter and zeta potential of PC SUV The SUV size distribution was determined by a dynamic light scattering method using a submicron particle analyzer (Nicomp Model 380; Particle Sizing Systems, Santa Barbara, CA, USA) [15]. The zeta potential was measured by ZEECOM ZC-3000 (Microtec Co., Ltd., Chiba, Japan), which was based on the principle of electrophoresis. Phosphorus determination The exact PC concentration in SUV suspensions was calculated from phosphate analysis according to the phosphovanadomolybdate method [26]. O O Cl CH
Cl S Cl
O
Cl
P
N
O
CH3
O
CH3
Cl
S N
Chlorpyrifos-methyl (CPFM)
O
P
N
O
C2H5
O
C2H5
S O2N
(H3C) 2HC
O
P
O
CH3
O
CH3
H 3C
Fenitrothion (FNT)
Diazinon (DZN)
ð1Þ
where [OPm] and [OPw] represent the molar concentrations of OP in PC SUV and water, respectively, and [OPt] = [OPm] + [OPw], and [L] and [W] are molar concentrations of PC in SUV and water (55.3 mol L1 at 37 °C), respectively. If the background signal effect based on PC SUV is eliminated in the second-derivative spectra, the derivative intensity difference (DD) of OP before and after the addition of PC SUV at a specific wavelength is proportional to the concentration of OP in PC SUV. As described in a previous paper [14], the following Eq. (2) can be derived from Eq. (1):
K p DDmax ½L ½W þ K p ½L
C2H5
H3C
The molar partition coefficient (Kp) of an OP between PC SUV and water was defined as follows [14,25]:
DD ¼
C2H5
O
Chlorfenvinphos (CFVP)
Calculation of molar partition coefficients
ð½OPm =½OPt Þ=½L ð½OPw =½OPt Þ=½W
O
Cl
Methods and materials
Kp ¼
P
ð2Þ
S H3C
S
O
P
O
CH3
O
CH3
S O
H3C
P
NHCH(CH 3)2 O
C2H5
COOCH(CH 3)2
Isofenphos (IFP)
Fenthion (FT)
O Br
O
P
S
C3H7
O
C2H5
Cl
N N
O O
P
S
C3H7
O
C2H5
Cl
Profenofos (PFF)
Pyraclofos (PCF)
Fig. 1. Chemical structures of the eight organophosphorus pesticides studied.
200
S. Takegami et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 145 (2015) 198–202
Measurements of absorption and second-derivative spectra
Absorption and second-derivative spectra
Sample solutions with a suitable OP concentration in the range of 10–50 lmol L1 and various amounts of the SUV suspension were prepared in a manner similar to that of the previous papers [14,15]. The reference solutions were those prepared without OP. Each flask was shaken for a short time and incubated at 37 °C for 30 min. The absorption spectrum of the sample solution was measured against the reference solution by using a spectrophotometer (Hitachi U-3310) equipped with a temperature-regulated cell holder in a 1-cm light-pass length cuvette at 37 °C with a band pass of 2 nm, wavelength interval of 0.1 nm, and scan speed of 30 nm min1. The second-derivative spectra were calculated based on the Savitzky-Golay method [27] using a laboratory-made Visual Basic program [28]. The cubic polynomial convolution of 17 points and the wavelength interval (Dk) of 0.8 (CFVP), 1.0 (DZN and FT), and 1.5 (CPFM, FNT, IFP, PFF and PCF) nm were used in the calculation, respectively.
The absorption spectra of 10 lmol L1 FT and 20 lmol L1 PCF in the sample solutions containing various amounts of PC SUV at 37 °C are shown in Fig. 2. As seen in Fig. 2A and B, both spectra show spectral changes according to the increase of PC concentration. However, no isosbestic points could be observed because of the incomplete baseline compensation due to the intense light scattering of PC SUV. In general, when strong background signals are present, it is difficult to cancel their effects completely in order to obtain a flat and zero-level baseline. Thus, further spectral data for calculating the Kp values could not be obtained from these absorption spectra. The second-derivative spectra calculated from the absorption spectra in Fig. 2 are shown in Fig. 3A and B, respectively. In these spectra, derivative isosbestic points can be clearly seen at 251 and 263 nm for FT in Fig. 3A, and at 264 and 294 nm for PCF in Fig. 3B, thus confirming that the influences of the residual background signal of PC SUVs are entirely eliminated in the secondderivative spectra, and that FT and PCF exist in two states [29], i.e., in the bulk water and the PC bilayer of SUV. Similar results were obtained for all of the other OPs examined.
Results and discussion Characteristics of phosphatidylcholine small unilamellar vesicles
Calculated Kp values The PC SUVs prepared by sonication in 50 mmol L1 NaCl-10 mmol L1 Hepes buffer (pH 7.4) were characterized as follows; the diameters of more than 90% of the SUVs were in the range of 20–30 nm, and their zeta potential values and the final PC concentrations were 10 ± 2 mV and 38 ± 1 mmol L1, respectively.
0.5
Absorbance
A
(4) (5) (7)
(1) (2)
B
0.4
(1) Absorbance
0.2
0.16
The DD values for eight OPs were obtained from the derivative values at the wavelengths of 263 (CFVP), 300 (CPFM), 267 (DZN), 268 (FNT), 256 (FT), 287 (IFP), 251 (PFF) and 275 (PCF) nm, respectively. Using the DD values, the Kp and DDmax values were
0.12
(3) (6) 0.08
0.3
0.2
(7) 0.1
0.04
0
0
240
250
260
270
280
290
300
310
320
245
255
265
275
285
295
305
315
Wavelength (nm)
Wavelength (nm)
Fig. 2. Absorption spectra of 10 lmol L1 FT (A) and 20 lmol L1 PCF (B) in Hepes buffer (pH 7.4, 37 °C) varying concentrations of PC SUV. PC concentration (mmol L1): (A) (1) 0.000, (2) 0.057, (3) 0.114, (4) 0.285, (5) 0.456, (6) 0.855, (7) 1.710; (B) (1) 0.000, (2) 0.105, (3) 0.210, (4) 0.420, (5) 0.630, (6) 1.050, (7) 2.100 (in the direction of the arrow).
0.0012
0.0015
A
0.0009 0.0006
0.0005
0
(1)
-0.0003
2nd-Der.
2nd-Der.
0.0003
-0.0006 -0.0009
(1)
0 -0.0005 -0.001
-0.0012 -0.0015 -0.0018 -0.0021
B
0.001
-0.0015
(7) 240
250
260
270
280
290
Wavelength (nm)
300
310
320
-0.002
(7) 245
255
265
275
285
295
305
315
Wavelength (nm)
Fig. 3. Second-derivative spectra of FT (A) and PCF (B) calculated from the absorption spectra in Fig. 2A and B, respectively. The curve labels are equivalent to those in Fig. 2.
201
S. Takegami et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 145 (2015) 198–202 Table 1 The mean Kp values of eight OPs calculated from the Kp values at all OP concentrations in Fig. 4.
8 7
Kp (×10-5)
6 5 4 3 2 1 0 0
10
20
30
40
50
60
calculated. The Kp values of the eight OPs at various OP concentrations are plotted in Fig. 4; their relative standard deviations were below 10%, indicating the good precision of the second-derivative method. It is known that the Kp values of certain drugs show drug-concentration dependence due to the molecular association or formation of their micelles [11,30]. As seen in Fig. 4, similar Kp values were obtained for each OP within the range of OP concentrations employed, although the Kp values of IFP and PCF were slightly reduced with increasing OP concentration. Thus, the results demonstrate that neither of these OPs associated or formed micelles in the bulk water, which in turn confirmed the validity of applying partition theory to account for the interactions between these OPs and the PC bilayer. In Fig. 5, the fractions of eight OPs partitioned to the PC SUVs are shown as a plot of the DD/DDmax values versus the PC concentration. Solid lines represent the theoretical curves calculated from Eq. (2) using the obtained Kp and DDmax values. The experimental values for each OP show a good fit with the calculated curves, indicating the validity of the obtained Kp values. Similar results were obtained for the other OP concentrations employed. The means of the Kp values obtained at various OP concentrations are also shown in Table 1, and were in the order of CPFM > FT > PFF > PCF > IFP > CFVP > FNT P DZN. The results quantitatively reveal that the substitution of a chemical group in FNT affects its Kp value; i.e., FT, which is given by substitution of a thiomethyl group for the nitro group at the
Kp (105)a
CFVP CPFM DZN FNT FT IFP PCF PFF
1.38 ± 0.18 6.57 ± 0.32 0.98 ± 0.15 1.00 ± 0.15 5.17 ± 0.33 2.65 ± 0.33 3.78 ± 0.42 4.37 ± 0.38
a Each value is expressed as the mean ± S.D. (n = 6–12).
OP concentration (µ µmolL-1) Fig. 4. Dependence of the Kp values of eight organophosphorus pesticides on their concentrations. The plotted Kp values are the averages of independent triplicate determinations. Symbols: (d) CFVP, (N) CPFM, (j) PFF, () PCF, (s) DZN, (4) FNT, (h) FT, (}) IFP.
OPs
C4 position of FNT, has an approximately 5-fold greater Kp value than FNT itself. In addition, the OPs having halogen atoms (Cl and Br) in their molecular structures (closed symbols), e.g., CPFM, PFF and PCF, tend to show larger Kp values than the OPs without halogen atoms (open symbols) as shown in Figs. 4 and 5. Relationship between partition coefficients for the liposome–water system and the n-octanol–water system In order to examine the relationship between the partition coefficients for the liposome–water system and those for the n-octanol–water system, the Kp values of these OPs in Table 1 were converted to the Plip values based on the membrane and water volumes; i.e., Plip was calculated by dividing each Kp value by 42: Kp/ 42 = Plip [25]. In Fig. 6, the log Plip values of the eight OPs were plotted against the corresponding reported log Poct values [31], and the results show a rather scatted figure with an R2 value of 0.530 (the solid line shows the linear correlation equation of y = 0.915x + 0.519). Due to the difference in the partitioning mechanisms of the n-octanol–water system (from water to n-octanol) and the liposome–water system (from water to lipid membrane), the correlation between the log Plip and log Poct values was poor, as seen in Fig. 6. Meanwhile, the dotted line in Fig. 6 indicates equal partitioning in the liposome–water and n-octanol–water systems. Four OPs, CFVP, DZN, IFP and PFF, show high log Poct values compared with their log Plip values. In the liposome–water system, many drugs depend not only on their hydrophobicity but also on their specific interaction with membrane lipids. In fact, although the log Poct
5 1
log Poct
Fraction (-)
4.5
0.5
4
3.5 0
0
1
2
3
PC concentration (mmolL-1)
Fig. 5. Fraction (DD/DDmax) of CFVP (d), CPFM (N), PFF (j), PCF (), DZN (s), FNT (4), FT (h) and IFP (}) in PC SUV as a function of PC concentration. The solid lines show the theoretical curves calculated from Eq. (2) using the experimental values of Kp and DDmax. The symbols are the experimental values. The OP concentrations were as follows (lmol L1): 40 (CFVP), 10 (CPFM), 40 (PFF), 30 (PCF), 40 (DZN), 30 (FNT), 10 (FT), 40 (IFP).
3 3
3.5
4
4.5
5
log Plip Fig. 6. The relationship between log Kp and log P of eight organophosphorus pesticides. Symbols: (d) CFVP, (N) CPFM, (j) PFF, () PCF, (s) DZN, (4) FNT, (h) FT, (}) IFP.
202
S. Takegami et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 145 (2015) 198–202
values have been used as a standard hydrophobic index to predict their bioactivity of drugs, they generally produce poor correlations between bioaccumulation and biotransport in living systems. Therefore, the obtained log Plip values for the eight OPs may be a better index to predict the accumulation of OPs related to chronic toxicity in human cell membranes rather than the log Poct values determined using the n-octanol–water system. Conclusions In conclusion, the Kp values of eight organophosphorus pesticides between PC SUV and water were accurately and precisely determined by the second-derivative spectrophotometric method and were in the order of CPFM > FT > PFF > PCF > IFP > CFVP > FNT P DZN. Since the partition coefficient for the liposome–water system is more effective for modeling the quantitative structure– activity relationship than that for the n-octanol–water system, the obtained results are toxicologically important to estimate the accumulation of these organophosphorus pesticides in human cell membranes. References [1] J. de Bruijn, J. Hermens, Environ. Toxicol. Chem. 10 (1991) 791–804. [2] H. Saito, J. Koyasu, K. Yoshida, T. Shigeoka, S. Koike, Chemosphere 26 (1993) 1015–1028. [3] T. Tsuda, M. Kojima, H. Harada, A. Nakajima, S. Aoki, Biochem. Physiol. 116C (1997) 213–218. [4] A. Opperhuizen, P. Serne, J.M.D. van der Stern, Environ. Sci. Technol. 22 (1988) 286–292. [5] Y.W. Choi, J.A. Rogers, Pharm. Res. 7 (1990) 508–512. [6] H. Fujiwara, Y.Z. Da, K. Ito, T. Takagi, Y. Nishioka, Bull. Chem. Soc. Jpn. 64 (1991) 3707–3712.
[7] J.A. Rogers, Y.W. Choi, Pharm. Res. 10 (1993) 913–917. [8] M.C. Antunes-Madeira, V.M.C. Madeira, Biochim. Biophys. Acta 778 (1984) 49– 56. [9] H. Yamamoto, H.M. Liljestrand, Environ. Sci. Technol. 38 (2004) 1139–1147. [10] B. Bondy, J. Remien, Life Sci. 28 (1981) 441–449. [11] M. Luxnat, H.J. Galla, Biochim. Biophys. Acta 856 (1986) 274–282. [12] A. Zachowski, P. Durand, Biochim. Biophys. Acta 937 (1988) 411–416. [13] K. Kitamura, M. Takenaka, S. Yoshida, M. Ito, Y. Nakamura, K. Hozumi, Anal. Chim. Acta 242 (1991) 131–135. [14] K. Kitamura, N. Imayoshi, T. Goto, H. Shiro, T. Mano, Y. Nakai, Anal. Chim. Acta 304 (1995) 101–106. [15] S. Takegami, K. Kitamura, T. Kitade, K. Hasegawa, A. Nishihira, J. Colloid Interface Sci. 220 (1999) 81–87. [16] S. Takegami, K. Kitamura, T. Kitade, A. Kitagawa, K. Kawamura, Chem. Pharm. Bull. 51 (2003) 1056–1059. [17] A. Pola, K. Michalak, A. Burliga, N. Motohashi, M. Kawase, Eur. J. Pharm. Sci. 21 (2004) 421–427. [18] S. Takegami, K. Kitamura, T. Kitade, M. Takashima, M. Ito, E. Nakagawa, M. Sone, R. Sumitani, Y. Yasuda, Chem. Pharm. Bull. 53 (2005) 147–150. [19] A.A. Omran, K. Kitamura, S. Takegami, T. Kitade, A.Y. El-Sayed, M. AbdelMottaleb, J. Pharm. Biomed. Anal. 25 (2001) 319–324. [20] A.A. Omran, K. Kitamura, S. Takegami, T. Kitade, A.Y. El-Sayed, M.H. Mohamed, Abdel-Mottaleb, Chem. Pharm. Bull. 50 (2002) 312–315. [21] S. Takegami, K. Kitamura, T. Funakoshi, T. Kitade, Chem. Pharm. Bull. 56 (2008) 663–667. [22] H. Ferreira, M. Lucio, B. de Castro, P. Gameiro, J.L.F.C. Lima, S. Reis, Anal. Bioanal. Chem. 377 (2003) 293–298. [23] C. Rodrigues, P. Gameiro, S. Reis, J.L.F.C. Lima, B. de Castro, Biophys. Chem. 94 (2001) 97–106. [24] S. Takegami, A. Otani, Y. Otake, T. Kitade, Curr. Pharm. Anal. 8 (2012) 286–294. [25] R. Welti, L.J. Mullikin, T. Yoshimura, G.M. Helmkamp Jr., Biochemistry 23 (1984) 6086–6091. [26] A.J. Christopher, T.R.F.W. Fennell, Microchem. J. 12 (1967) 593–605. [27] A. Savitzky, M.J.E. Golay, Anal. Chem. 36 (1964) 1627–1639. [28] K. Kitamura, K. Hozumi, Anal. Chim. Acta 172 (1985) 111–118. [29] K.A. Connors, Binding Constants, John Wiley Sons, New York, 1987. pp. 142– 147. [30] W.N. Charman, C.S. Lai, B.C. Finnin, B.L. Reed, Pharm. Res. 8 (1991) 1144–1150. [31] C. Hansch, A. Leo, D. Hoekman, Exploring QSAR, American Chemical Society, Washington DC, 1995.
