Environ. Sci. Technol. 2004, 38, 503-507
Structure Effect on the Interaction of Phenylurea Herbicides with Model Biomembrane as an Environmental Mobility Parameter V I T O L I B R A N D O , * ,†,‡ STEFANO FORTE,‡ AND MARIA G. SARPIETRO‡ INCA, Unit Catania 5, c/o Dipartimento di Scienze Chimiche, Universita` di Catania, Viale A. Doria 6, 95127 Catania, Italy, and Dipartimento di Scienze Chimiche, Universita` di Catania, V.le A. Doria 6, 95127 Catania, Italy
During recent years, intensive use of herbicides has raised increasing concern mainly due to their massive pollution of the environment. As these herbicides are directly or indirectly toxic to a wide range of organisms, their potential for contaminating soil, surface water, and groundwater makes these xenobiotics of special interest from a health and environmental point of view. Knowledge of the mechanisms by which they exert their toxic effects is becoming a need. Because of the herbicides’ lipophilicity, a possible site of interaction in the cell is represented by biomembranes. The interaction of four herbicides, difenoxuron, diuron, linuron, and metoxuron, with model membranes constituted of dimyristoylphosphatidylcholine multilamellar vesicles was investigated by the differential scanning calorimetry technique. The aim was to study the effects exerted by an increasing amount of the examined compounds on thermotropic behavior of the model phospholipid membranes and to correlate the obtained results with structural features of the herbicides due to their environmental mobility. Among the herbicides studied, linuron is the most effective in perturbing the ordinate structure of vesicles forming phospholipids, whereas metoxuron is the least effective and the others exert an intermediate effect. Linuron exerts its effect both on the transition temperature of the gel to the liquid crystalline phase and on the enthalpy change. Difenoxuron, diuron, and metoxuron cause a change in the transition temperature but have an insignificant effect on the enthalpy change. The calorimetric results, correlated with the structural features of the herbicides, are consistent with their partition coefficient, log Kow, suggesting that the more hydrophobic compound character causes a greater liposolubility and consequential cellular absorption with more effectiveness on the membrane order.
1. Introduction In the past few decades, the use of herbicides to increase yields and improve the quality of agricultural production * Corresponding author address: Dipartimento Scienze Chimiche, Viale A. Doria 6, 95127 Catania, Italy; phone: 0039-95-7385201; fax: 0039-95-580138; e-mail:
[email protected]. † INCA, Unit Catania 5. ‡ Dipartimento di Scienze Chimiche, Universita ` di Catania. 10.1021/es034459f CCC: $27.50 Published on Web 12/09/2003
2004 American Chemical Society
has been widely spread. These products, indispensable in a modern, profitable agricultural context, are nevertheless a major source of contamination of the natural environment, especially in intensive agricultural areas (1), but also they have relevance in the industrial emissions during their production. The biosphere has the potential for simultaneous or sequential exposure to these intentionally introduced environmental xenobiotics and is subjected to their toxic effects. Unfortunately, nontarget organisms, including humans, are affected by these compounds (2, 3). Therefore, exact knowledge of their toxic effects is a need in order to plan for environmental remediation, particularly for soils, where they accumulate. Over the past few decades, extensive work has been devoted to identify the precise biochemical mechanisms underlying insecticide toxicity. Defined acute biochemical interactions have been only assigned to organophosphorus and carbamate compounds (4, 5). However, because of the lipophilicity of most of these compounds, a possible target of their interaction with living organisms is represented by biomembranes where they may induce physical and chemical perturbations and, consequently, alterations of the native properties of biomembranes. Several studies demonstrate that insecticides induce perturbations of membrane fluidity and enzyme dynamics and that among these compounds, the most powerful toxicant is also the most effective in inducing membrane perturbations. On the other hand, the least toxic affects the membrane structure to a lesser extent. The insecticides possessing intermediate toxicity have shown to have also intermediate effects (6-8). Phenylurea derivatives are extensively used as herbicides. These substances are well-known to inhibit photosynthesis by entering the plants via the root (9). They are principally employed for selective control of germinating grass and broad-leaved weeds in many crops, but some of them are also used for total weed control of noncultivated areas such as roads, railways, and parks. In previous studies (10-12), we investigated, by the differential scanning calorimetry (DSC) technique, the interaction between some polycyclic aromatic hydrocarbon environmental pollutants and dimyristoylphosphatidylcholine (DMPC) multilamellar vesicles (MLVs). The DSC technique is a nonperturbative way to determine the variations occurring in thermodynamic parameters, like temperature and enthalpy changes, which are characteristic of a changing of phase or transitional process. The determination of such parameters is carried out by measuring the differential heat flow required to maintain the examined sample, placed inside an aluminum pan hermetically sealed, at the same temperature of a pan filled with an inert reference material. The sample and the reference pans are submitted to a heating or cooling program at a determinate rate in a defined temperature range. In this work, we report the effects of four phenylurea herbicides, metoxuron, difenoxuron, diuron, and linuron (whose chemical structures are shown in Chart 1), on the thermotropic behavior of DMPC MLVs, chosen as model membranes. Lipid membranes undergo a sharp phase transition from an ordered gel-like structure to a disordered fluidlike structure. This phase change happens at a characteristic temperature (Tm), and it can be detected by the DSC technique. The presence of foreign substances dissolved inside the lipid bilayer strongly affects the transition temperature and the enthalpy changes related to the endothermal peak (13-17). A substance dissolved in a lipid layer usually destabilizes the membrane ordered structure, decreasing the VOL. 38, NO. 2, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
503
CHART 1. Chemical Structures of Metoxuron, Difenoxuron, Diuron, and Linuron
enthalpy changes, by using indium, stearic acid, and cyclohexane by following the procedure of the Mettler TA STARe software. For each sample, at least four heating and cooling cycles were recorded to ensure the reproducibility of the thermotropic behavior. After calorimetric scans, all samples were extracted from the pan and aliquots were used to determine lipid concentrations by phosphorus analysis (21).
3. Results and Discussion
lipid transition temperature (18-20), and the destabilizing effect depends on the amount of chemicals dissolved in the lipid structure. There are several indications of the correlation between lipid bilayer structure and dynamics and membrane functionality; therefore the physicochemical changes induced by herbicides on the structure and organization of lipid membranes can be correlated with their effects. From this point of view, the main aim of this work is to correlate the effect of these herbicides on the thermotropic behavior of DMPC MLVs with the herbicides’ structural features in order to better understand the role of the substituents in interacting with the lipid bilayer, to make a correlation between the effect on model membrane thermotropic behavior and their lipophilicity and thereby to obtain information about the mechanism by which they exert their toxic action on cells.
2. Experimental Section 2.1. Materials. Synthetic L-R-dimyristoylphosphatidylcholine (purity 99.9%) was obtained from Genzyme (Switzerland). Lipids were chromatographically pure as assessed by twodimensional thin-layer chromatography (TLC). Lipid concentrations were determined by phosphorus analysis (19). Buffer solution consisted of 50 mM Tris(hydroxymethyl)aminomethane (Tris), adjusted to pH 7.4 with hydrochloric acid. Metoxuron (purity 99.6%), difenoxuron (purity 99.7%), diuron (purity 99.4%), and linuron (purity 99.5%) were purchased from Sigma (Germany). 2.2. Preparation of Liposomes. Stock solutions of DMPC and of the herbicides in chloroform/methanol (1:1) were mixed to obtain 8 mg of DMPC and different molar fractions of herbicides. The solvents were removed under nitrogen flow in order to allow the lipid film to form. The samples were then lyophilized for 1 h to eliminate the residual traces of solvents. Liposomes were formed by adding 192 µL of Tris (pH 7.4) to the film, heating in a water bath at 37 °C, and mixing three times for 1 min. To obtain homogeneous liposomes and to permit the herbicide to partition between the aqueous and lipid phases, the samples were further kept for 1 h in the water bath at 37 °C. 2.3. Differential Scanning Calorimetry. The thermal phase transition of the liposomes was detected by a Mettler TC-15 system equipped with a DSC-30 calorimetric cell and Mettler TA STARe V 6.10 SW software. Samples were prepared by loading 120 µL of the lipid dispersion (containing 5 mg of DMPC) into a 160 µL aluminum pan and sealing hermetically. The samples were submitted to calorimetric measurement performed in the temperature range of 5-37 °C at a scanning rate of 2 °C/min. A Tris loaded pan was used as a reference. The sensitivity was automatically chosen as the maximum allowed by the calorimetric system. The calorimetric system was calibrated, in transition temperature and 504
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 2, 2004
The influence of four phenylurea herbicides on the thermotropic behavior of DMPC multilamellar liposomes was studied by DSC. This technique reveals the changes in the transition temperature and in the associated enthalpy due to the presence of a foreign substance in the liposome bilayers. This provides information about the biophysical modifications occurring in the model membranes as a result of herbicide interaction. Figure 1 reports the calorimetric peak of pure DMPC liposomes (curve a) as well as that of DMPC liposomes containing increasing molar fractions of herbicide (curves b-g). When metoxuron is used (Figure 1A), the pretransition peak is still present at a 0.015 molar fraction of metoxuron but it disappears at higher molar fractions. A small shift of the main peak toward lower temperatures is visible, while the peak shape remains almost constant. The calorimetric curves of DMPC in the presence of difenoxuron (Figure 1B), diuron (Figure 1C), and linuron (Figure 1D) are similar to those of metoxuron because they are shifted to lower temperature. However, any molar fraction used causes the disappearance of the pretransitional calorimetric peak of DMPC liposomes. Thus, in the case of linuron the increasing molar fractions cause a larger shift of the transition temperature as well as a strong lowering and broadening of the main peak. Figure 2 shows the shift of the transition temperature, reported as ∆T/T°m (∆T ) T°m - Tm; where T°m is the transition temperature of the pure DMPC vesicles and Tm is that of DMPC in the presence of increasing molar fractions of the herbicide), versus the molar fraction of the herbicide present in the aqueous dispersion of vesicles. The most effective on the thermotropic behavior is linuron, followed by diuron, difenoxuron, and then metoxuron, which is the less effective. Thus, the effect of linuron increases with the increase of its molar fraction present in MLVs. Diuron decreases the Tm up to a molar fraction of 0.06; a further increase of the molar fraction has no effect on the Tm. The effect of metoxuron increases up to the molar fraction of 0.06; higher molar fractions have a more limited effect. The enthalpy changes of the calorimetric peaks shown in Figure 1A-C are reported in Figure 3 as a function of the herbicide molar fractions. Difenoxuron, diuron, and metoxuron show small effects on the enthalpy changes, while linuron is the only one strongly affecting the enthalpy variation. The interactions of the four studied herbicides with the model membranes are well correlated with their partition coefficients (log Kow). In fact, the bigger the herbicide effect on the DMPC thermotropic behavior, the higher the partition coefficient, being log Kow ) 1.64 (metoxuron) (22), 2.43 (difenoxuron) (23), 2.68 (diuron) (22), and 3.20 (linuron) (22). Moreover, a correlation with other physicochemical parameters has been found. For instance, the water solubility increases on going from linuron to metoxuron (passing through diuron and difenoxuron), demonstrating that the lower the water solubility, the higher the DSC effect. The correlation between our DSC data and the LD50 (lethal dose, 50% kill) parameter is interesting. The LD50 literature data of the studied compounds are rather spread, strongly
FIGURE 1. DSC heating curves of DMPC MLVs containing (A) metoxuron, (B) difenoxuron, (C) diuron, and (D) linuron in the following molar fractions: a ) 0.0, b ) 0.015, c ) 0.03, d ) 0.045, e ) 0.06, f ) 0.09, and g ) 0.12. depending on the adsorption route. However, at least qualitatively, the higher the DSC effect, the smaller the LD50. For instance, acute oral LD50 values in rats of 3200 mg/kg (metoxuron), >5000 mg/kg (diuron), and 1500 mg/kg (linuron) have been reported (24); regarding the route of adsorption by inhalation, LC50 (lethal concentration, 50% kill) values of >660 mg/m3/6 h and 48 mg/m3/4 h were found for difenoxuron (25) and linuron (26), respectively. If we consider the herbicide as an impurity dissolved into an ideal two-dimensional solution, the lowering of the lipid
bilayer transition temperature can be easily rationalized in terms of Van’t Hoff depressing of freezing temperature (27, 28). In that hypothesis, one should observe an almost linear decrease of the bilayer transition temperature with the concentration of the foreign molecule within the bilayer, while the associated enthalpy should remain constant. Concerning the temperature variation, this behavior has been indeed observed in Figure 3, which shows how the scaled transition temperature ∆T/T°m increases with the stoichiometric concentration of herbicide. Moreover, the Van’t Hoff VOL. 38, NO. 2, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
505
FIGURE 2. Transitional temperature variations, as ∆T/T°m (∆T ) T°m - Tm, where T°m and Tm are the transition peak temperature of the DMPC empty vesicles and that in the presence of herbicides, respectively), in heating mode as a function of diuron, difenoxuron, metoxuron, and linuron increasing molar fractions in the buffer lipid dispersion.
FIGURE 3. Enthalpy changes, as ∆∆H/∆H° (∆∆H ) ∆H° - ∆H, where ∆H° and ∆H are the enthalpy change for the DMPC empty vesicles and that in the presence of herbicides, respectively), in heating mode as a function of diuron, difenoxuron, metoxuron, and linuron increasing molar fractions in the buffer lipid dispersion. depressing of freezing temperature mechanism explains the good relationship between ∆T/T°m and herbicide solubility in the hydrophobic phase (proportional to the water/octanol partition coefficient): the higher the herbicide liposolubility, the greater the shift in the bilayer transition temperature. The real situation, however, is more complex. Looking at enthalpy data of Figure 3, we observe a dramatic decrease of the transition enthalpy only in the case of linuron, while the enthalpy decrease is much smaller for the remaining studied molecules. This behavior can be explained only by invoking a destabilizing effect of lipid-lipid interactions due to the presence of a foreign impurity. The strong interaction of the impurity with the lipid chain prevents the highly cooperative melting of the lipid tails, causing the gel to fluid phase transition of the bilayer to be less endothermic and less cooperative. As a result, the intensity of the calorimetric peak decreases and the peak shape broadens. The above behavior is typical of some substances such as cholesterol, a molecule able to lower the enthalpy of phospholipid bilayers even at small mole fractions (29), where a loss in the cooperativity between the lipid molecules, due to the 506
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 2, 2004
insertion between both the acyl chains as well as the polar group of lipids of the spacing acting herbicide molecule, is happening causing the above cited enlargement of the calorimetric peak. Therefore, we conclude that linuron (i) is more soluble inside the lipid membrane than closely related compounds (Figure 3) and (ii) exerts a greater perturbation of the lipid structure, while the other studied herbicides mainly behave as interstitial impurities. If one looks at the structures of the four herbicides, interesting considerations can be deduced demonstrating that the different effects exerted by these herbicides on the thermotropic behavior of the DMPC MLVs can be attributed to some particular structural features. Metoxuron, the compound that is less effective on the thermotropic behavior of DMPC vesicles, bears a chloro and a methoxy group in positions 3 and 4 of the aromatic ring, respectively, which make it different from diuron which instead bears two -Cl groups. The presence of the chlorine (the only difference between the two molecules) seems to make diuron more able than metoxuron to interact with model membranes. This is still true when linuron is considered; this molecule also possesses two -Cl groups on the aromatic ring like diuron. A comparison of the diuron and linuron structures shows that they are identical, apart from the presence of a N-methoxylic and a N-methylic group on the linuron molecule instead of two N-methylic groups of diuron. The fact that linuron exerts a stronger effect than diuron on the membrane thermotropic behavior can be related to the presence of the methoxylic group. However, the different effect of linuron on membrane thermodynamics is due not only to the methoxylic group but also to its location, as difenoxuron and metoxuron also contain methoxylic groups. The reported results can be considered also considering both the structural differences of the molecular structure (sterically larger molecules better fluidify the lipid bilayer than the small ones) and solubility in the membranes. Both parameters affect the shift of lipid phase transition temperature and the related enthalpy change (30-32). The interaction might be of relevance for polycyclic aromatic hydrocarbons’ carcinogenic and mutagenic properties even if such interaction is not caused by a simple passive transport through biological membranes, but they can be correlated with other mechanisms accompanying the transfer from a hydrophilic medium to a biological membrane. Although it is not clear which lipid membrane properties are critical for a given membrane function, bilayer structure and dynamics are crucial for membrane functionality (33, 34). Among the different physicochemical parameters (water/ octanol partition coefficient, water solubility, etc.), the correlation between biological response (e.g., LD50) and lipid bilayer transition temperature obtained by DSC seems to have the strongest conceptual link because of the closer similarity between the physicochemical model system and the biological reality. To get a relationship between the lipid organization and membrane functions, systematic studies have however to be carried out. They would provide new insights to understand the molecular mechanism of herbicide toxicity.
Acknowledgments This work was partially supported by MIUR (Progetti d’Ateneo) and L 488, Cluster Ambiente Terrestre, Progetto P6, “Soil remediation”.
Literature Cited (1) SRPV Bretagne, DRAF Bretagne; Ministere de L’Agriculture, 1991. (2) Ware, G. W. Pesticides: Theory and application; Zweig, Ed.; CRC Press: New York, 1983; pp 35-67.
(3) Metcalf, R. L. Pesticides in the environment; White-Stevens, Ed.; Marcel Dekker: New York, 1971; pp 515-539. (4) Eto, M. Organophosphorus pesticides: Organic and biological chemistry; CRC Press: Cleveland, OH, 1974; pp 123-133. (5) Kuhr, R. J.; Dorough, H. W. Carbamate insecticides: Chemistry, biochemistry and toxicology; CRC Press: Cleveland, OH; 1976; pp 41-70. (6) Lopez, V. I. C. F.; Antunes-Madeira, M. C.; Madeira, V. M. C. Toxicol. in Vitro 1997, 11, 337-345. (7) Videira, R. A.; Antunes-Madeira, M. C.; Lopes, V. I. C. F.; Madeira, V. M. C. Biochim. Biophys. Acta 2001, 1511, 360-368. (8) Suwalsky, M.; Benites, M.; Villena, F.; Norris, B.; Quevedo, L. Comp. Biochem. Physiol., C 1998, 120, 29-35. (9) Ducruet, J. M. Les herbicides: Mode d’action et principles d’utilisation; Les inhibiteurs du photosysteme II; Scalla, Ed.; INRA: Paris, 1991; pp 79-114. (10) Castelli, F.; Librando, V.; Sarpietro, M. G. Thermochim. Acta 2001, 373, 133-140. (11) Castelli, F.; Librando, V.; Sarpietro, M. G. Environ. Sci. Technol. 2002, 36, 2717-2723. (12) Librando, V.; Sarpietro, M. G.; Castelli, F. Environ. Toxicol. Pharm. 2003, 14, 25-32. (13) Mabrey-Gaud, S. Liposomes: From physical structure to therapeutic applications; Knight, Ed.; Elsevier/North-Holland Biomedical Press: Amsterdam, 1981; pp 105-138. (14) Silvius, J. R. Chem. Phys. Lipids 1991, 57, 241-252. (15) Bach, D. Biomembrane structure and function; Chapman, D., Ed.; MacMillan Press: London, 1994; pp 1-41. (16) Marsh, D. Handbook of nonmedical applications of liposomes; Barenholz, Y., Lasic, D. D., Eds.; CRC Press: Boca Raton, FL, 1996; Vol. II, pp 21-49. (17) Huang, C.; Li, S. Biochim. Biophys. Acta 1999, 1422, 273-307. (18) Jain, M. K.; Wu, N. M. J. Membr. Biol. 1977, 34, 151-201. (19) Castelli, F.; Valencia, G. Thermochim. Acta 1989, 154, 323-331. (20) Jain, M. K. Introduction to biological membranes; Jain, M. K., Ed.; John Wiley and Sons: New York, 1988; pp 122-165 and references therein.
(21) Rouser, G.; Fleicher, S.; Yamamoto, A. Lipids 1970, 5, 494-496. (22) Sabljic, A.; Gusten, H.; Verhaar, H.; Hermens, J. Chemosphere 1995, 31, 4489-4514. (23) Liu, J.; Qian, C. Chemosphere 1995, 31, 3951-3959. (24) Routt Reigard, J.; Roberts, J. R. Recognition and management of pesticide poisonings; U.S. Environmental Protection Agency: Washington, DC, 1999. (25) Pesticide manual; The British Crop Protection Council: Farnham, Surrey, U.K., 1991; p 278. (26) Izmerov, N. F. Toxicometric Parameters of Industrial Toxic Chemicals Under Single Exposure; Centre of International Projects, GKNT: Moscow, 1982; p 48. (27) Cevc, G.; Marsh, D. Phospholipid bilayers: Physical principles and models; Bittar, E. E., Ed.; John Wiley and Sons: New York, 1987; Vol. 5, p 441. (28) Jorgensen, K.; Ipsen, J. H.; Mouritsen, O. G.; Bennet, D.; Zuckermann, M. J. Biochim. Biophys. Acta 1991, 1062, 227238. (29) Estep, T. N.; Mountcastle, D. B.; Biltonen, R. L.; Thompson, T. E. Biochemistry 1978, 17, 1984-1989. (30) Castelli, F.; Caruso, S.; Giuffrida, N. Thermochim. Acta 1999, 327, 125-131. (31) Lohner, K. Chem. Phys. Lipids 1991, 57, 341-362. (32) Castelli, F.; Pitarresi, G.; Giammona, G. Biomaterials 2000, 21, 821-833. (33) Lee, A. G. Prog. Biophys. Mol. Biol. 1975, 29, 3-56. (34) Mouritsen, O. G.; Jorgensen, K. Pharm. Res. 1998, 15, 15071519.
Received for review May 9, 2003. Revised manuscript received October 20, 2003. Accepted October 20, 2003. ES034459F
VOL. 38, NO. 2, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
507
