Oligopeptides as Mimic of Acetylcholinesterase: From the Rational

Oct 24, 2008 - The cartridges prepared using His-Glu-Pro-Ser sequence was, as predicted, able to bind paraoxon and carbaryl with recovery values in th...
0 downloads 0 Views 134KB Size
Download PDF
Anal. Chem. 2008, 80, 9150–9156

Oligopeptides as Mimic of Acetylcholinesterase: From the Rational Design to the Application in Solid-Phase Extraction for Pesticides M. Mascini,*,† M. Sergi,† D. Monti,‡ M. Del Carlo,† and D. Compagnone† Department of Food Science, University of Teramo, 64023 Teramo, Italy, and Department of Chemical Technology Science, University of Rome “Tor Vergata”, 00133 Rome, Italy Three different peptides (His-Glu-Pro-Ser, His-Gly-SerAla and Glu-Pro-Ser-Ala) were selected and tested to be used as affinity binding receptors for organophosphate and carbamate pesticides. The peptides were rationally designed by mimicking acetylcholinesterase active site. The simulated binding energy of the three tetrapeptides versus one model of organophosphate (paraoxon) and one of carbamate (carbaryl) pesticide was calculated; a good correlation between shape designed and binding score was obtained. The binding properties of the peptidepesticide interaction were studied following the variation of UV-visible spectra in different solvents. The binding constants in water, which were nicely correlated with computational data, ranged from 506 ((115) to 36((2) M-1. All the peptides had a 5-fold decrease in binding by changing solvent, going from water to less polar ethanol. The binding affinity suggested the use of these ligands as a preanalytical tool in extraction cartridges. The tetrapeptides efficiency was tested linking the peptides to two different supports. The cartridges prepared using His-GluPro-Ser sequence was, as predicted, able to bind paraoxon and carbaryl with recovery values in the 72-88% range at pH 4.5. Intercolumn, interday RSD was in the 4-7% range. The columns were used for 80 cycles before losing retention ability. In the past decade, development of biomimetic ligands such as molecular imprinted polymers, aptamers, or peptides has enabled rapid advances as alternative candidates to antibodies in the development of affinity-based analytical procedures.1-7 Promising structures to obtain strong ligand-target interactions could be, in principle, made mimicking nature, which uses * To whom correspondence should be addressed. E-mail: mmascini@ unite.it. † University of Teramo. ‡ University of Rome “Tor Vergata”. (1) Navani, N. K.; Li, Y. Curr. Opin. Chem. Biol. 2006, 10, 272–281. (2) Tombelli, S.; Minunni, M.; Mascini, M. Biosens. Bioelectron. 2005, 20, 2424– 2434. (3) Tozzi, C.; Anfossi, L.; Giraudi, G. J. Chromatogr., B 2003, 797, 289–304. (4) Baines, I. C.; Colas, P. Drug Discovery Today 2006, 11, 334–341. (5) Gujraty, K.; Sadacharan, S.; Frost, M.; Poon, V.; Kane, R. S.; Mogridge, J. Mol. Pharm. 2005, 2, 367–372. (6) Wegner, G. J.; Wark, A. W.; Lee, H. J.; Codner, E.; Saeki, T.; Fang, S.; Corn, R. M. Anal. Chem. 2004, 76, 5677–5684. (7) Falciani, C.; Lozzi, L.; Pini, A.; Bracci, L. Chem. Biol. 2005, 12, 417–426.

9150

Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

amino acids as bricks to construct a large amount of variable and adaptable molecules for life.8 Peptides libraries can easily be synthesized by conventional methods both in solution and on solid phase.9,10 Moreover solid-phase peptide synthesis gives simultaneously the requisite sequence, reliability, and solid support itself for affinity chromatography application and should give more robustness compared to the use of complex molecules as antibodies or proteic receptors.29 However, the selection of a proper ligand to achieve some desired structure is largely a subject of trial and error. The huge number of candidates to be synthesized and tested using a combinatorial approach pushes for a radical change in the strategy to design biomimicking ligands. Maximizing diversity is now rarely considered a sufficient design criterion for a library. More and more strategies to build libraries are now intended to hit a single biological target or family of related targets.11 In these last years, many efforts have been carried out to make the selection method more efficient, and advances are in the area of computational chemistry.12-15 Progress in molecular modeling software has expanded the range of systems that can be effectively simulated. So, today, docking techniques contribute to design affinity ligands with a high probability of success.16,17 By using logical programming, the rational design of the peptide library can be really reduced to orders of magnitude. In this way, the ability to computationally design tailor-made peptides that bind specifically and with high affinity to any given target, and either activate or inhibit it, has led to more feasible, powerful and easier to obtain ligands.18 At the same time, increasingly numbers of proteins and ligand complexes have been resolved by means of X-ray crystallography, nuclear magnetic resonance, mass spectrometry (MS), and other spectroscopic techniques. Their structures are generally deposited (8) Kriplani, U.; Kay, B. K. Curr. Opin. Biotechnol. 2005, 16, 470–475. (9) Noppea, W.; Plieva, M. F.; Galaev, I. Y.; Vanhoorelbeke, K.; Mattiasson, B.; Deckmyna, H. J. Chromatogr., A 2006, 1101, 79–85. (10) Rose, S.; Stevens, A. Curr. Opin. Chem. Biol. 2003, 7, 331–339. (11) Turk, B. E.; Cantley, L. C. Curr. Opin. Chem. Biol. 2003, 7, 84–90. (12) Stahl, M.; Guba, W.; Kansy, M. Drug Discovery Today 2006, 11, 326–333. (13) Kortemme, T.; Baker, D. Curr. Opin. Chem. Biol. 2004, 8, 91–97. (14) Langer, T.; Wolber, G. Drug Discovery Today: Technol. 2004, 1, 203–207. (15) Laggner, C.; Schieferer, C.; Fiechtner, B.; Poles, G.; Hoffmann, R. D.; Glossmann, H.; Langer, T.; Moebius, F. F. J. Med. Chem. 2005, 48, 4754– 4764. (16) Labrou, N. E. J. Chromatogr., B 2003, 790, 67–78. (17) Erickson, J. A.; Jalaie, M.; Robertson, D. H.; Lewis, R. A.; Vieth, M. J. Med. Chem. 2004, 47, 45–55. (18) Sawyer, T. K. Chem. Biol. Drug Des. 2006, 67, 196–200. 10.1021/ac801030j CCC: $40.75  2008 American Chemical Society Published on Web 10/25/2008

into public repositories (e.g., Protein Data Bank), providing information on the key role of the structure-function properties.19 The identification of the binding site of acetylcholinerase (AChE), the target enzyme of carbamate (CM), and organophosphate pesticides (OP) has led to the knowledge of the inhibition mechanism, given in the work of Millard and co-workers,20 where structural and functional information on the amino acids involved in the pesticides binding is reported. Nerve agent detection is also an important current subject of interest that involves fields like food and environmental safety as well as public health benefits.21-23 In this area, there is a need for fast and inexpensive preanalytical devices for pesticides detection that can combine the high selectivity of biological receptors with greater stability in harsh environments. Typically, clinical, environmental, or food samples are characterized by high complexity and low levels of pesticide contamination.24 Therefore, analytical methods are mainly based on multiresidue analytical procedures consisting of an extraction and cleanup step followed by a chromatographic detection.25 Our research was oriented to develop preanalytical cartridges based on peptidic receptors able to mimic the active site of AChE. These receptors are envisaged to produce new, powerful ligands in separation methods, preparative isolation of the desired species, or preconcentration of trace compounds.26,27 Peptides for affinity chromatography were used in different analytical applications, most of them based on combinatorial approach.29-31 In this field, an interesting procedure was developed by Giraudi et al.32 that polymerized amino acid mixtures in aqueous medium in the presence of the template producing peptides with molecular recognition properties toward the target. Different authors optimized methods for pesticide isolation by means of cartridges mainly based on affinity columns using antibodies.33,34 A recent paper reported a preparation of a cartridge using sol-gel immunosorbents doped with polyclonal antibodies for selective extraction of malathion and triazines from aqueous samples.35 Starting from these considerations, we selected small peptides for trapping pesticides, designed by computational support, which (19) Berkessel, A. Curr. Opin. Chem. Biol. 2003, 7, 409–419. (20) Millard, C. B.; Kryger, G.; Ordentlich, A.; Greenblatt, H. M.; Harel, M.; Raves, M. L.; Segall, Y.; Barak, D.; Shafferman, A.; Silman, I.; Sussman, J. L. Biochemistry 1999, 38, 7032–7039. (21) Warren, N.; Allan, I. J.; Carter, J. E.; House, W. A.; Parker, A. Appl. Geochem. 2003, 18, 159–194. (22) Nasreddine, L.; Parent-Massin, D Toxicol. Lett. 2002, 127, 29–41. (23) Barr, D. B.; Needham, L. L. J. Chromatogr., B 2002, 778, 5–29. (24) Hajslova, J.; Zrostlıkova, J. J. Chromatogr., A 2003, 1000, 181–197. (25) Sanchez, F. G.; Diaz, A. N.; Herrera, R. G.; San Jose, L. P. Talanta 2007, 71, 1411–1416. (26) Zhang, X.; Martens, D.; Kra¨mer, P. M.; Kettrup, A. A.; Liang, X. J. Chromatogr., A 2006, 1133, 112–118. (27) Su, P.; Zhang, X. X.; Chang, W. B. J. Chromatogr., B 2005, 816, 7–14. (28) Tozzi, C.; Giraudi, G. Curr. Pharm. Des. 2006, 12, 191–203. (29) Clonis, Y. D. J. Chromatogr., A 2006, 1101, 1–24. (30) Ehrlich, G. K.; Bailon, P. J. Biochem. Biophys. Methods 2001, 49, 443–454. (31) Jacobsen, B.; Gardsvoll, H.; Funch, G. J.; Østergaard, S.; Barkholt, V; Ploug, M. Protein Expression Purif. 2007, 52, 286–296. (32) Giraudi, G.; Giovannoli, C.; Tozzi, C.; Baggiani, C.; Anfossi, L. Anal. Chim. Acta 2003, 481, 41–53. (33) Watanabe, E.; Yoshimura, Y.; Yuasa, Y.; Nakazawa, H. Anal. Chim. Acta 2001, 433, 199–206. (34) Rejeb, S. B.; Cle´roux, C.; Lawrence, J. F.; Geay, P.-Y.; Wu, S.; Stavinski, S. Anal. Chim. Acta 2001, 432, 193–200. (35) Vera-Avila, L. E.; Vazquez-Lira, J. C.; Garcia De Llasera, M.; Covarrubias, R. Environ. Sci. Technol. 2005, 39, 5421–5426.

may act as functional affinity model. Studying the threedimensional structure of the complex between CM/OP pesticides and the AChE, it was thought to reproduce a verisimilar shape of the AChE binding site with artificial oligopeptides. Three oligopeptides computationally designed were tested toward pesticides using genetic search algorithm.36,37 The simulated binding energy versus pesticides was evaluated in the experimental work. The binding properties of the peptide-pesticide were tested in solution and using solid-phase extraction. The resulting data were presented and discussed. EXPERIMENTAL PROCEDURE Reagents. All solvents were RS-Plus grade and purchased from Carlo Erba. The formic acid was purchased from Merck. The pesticides carbaryl and paraoxon as well as all other chemicals were supplied by Sigma. The three peptides were purchased from EspiKem Srl and were the following: (A) [N] His-Glu-Pro-Ser [C]; (B) [N] His-GlySer-Ala [C]; (C) [N] Glu-Pro-Ser-Ala [C], where N-terminus corresponds to the amine and C-terminus to the carboxyl group. Synthesis of Peptide PAL-PEG Resin. The PAL-PEG peptides were synthesized via a solid-phase method using Fmoc/ tBu chemistry. Peptide onto resin was provided as >85% purity product. Pal-Peg-PS resin (PAL, 5-(4-(9-fluorenylmethyloxycarbonyl)aminomethyl-3,5-dimethoxyphenoxy) valeric acid handle; PEG, poly(ethylene glycol); PS, polystyrene) was used as solid support attaching peptide to this via its carboxyl terminus. The substitution level was ∼10% obtaining 0.18 mmol of estimated peptide for 1 g of resin. Peptides for spectrophotometric analysis were after synthesis detached from the solid support, using trifluoroacetic acid, and further purified for a final purity of 95%. Synthesis of Peptide Amberlyte Resin. A proper amount of Amberlite IRC-50 was weighed and washed three times with N,N-dimethylformamide. A spacer arm was introduced on the surface of the resin by activating the carboxylic groups on the solid phase through the N-hydroxysuccinimide/N,N-dicyclohexylcarbodiimide method38 and then by reacting the activated beads with a mixture of ethanolamine and 4-aminobutyric acid (9:1 mol mol-1) in N,N-dimethylformamide overnight. The beads were then washed three times with N,N-dimethylformamide and activated again. The amino acid concentration used was 10 times that of the carboxylic groups present on the beads (1 mmol g-1 of dried resin). After 2 h of reaction, the beads were filtered, washed with N,N-dimethylformamide, and activated again. Then, the second amino acid was added to the beads and the procedure repeated again. Apparatus. All computational procedures were performed by using a Linux platform (Genuine Intel, model Intel Xeon). The structure of the AChE enzyme (PDB code 1VXO) along with the pesticide and peptide properties was studied using Openeye scientific software for academics. Peptides and pesticides were designed and minimized using the software package Sybyl 6.9.1 (Tripos Inc.). Pesticide-peptide binding score was calculated by Leapfrog algorithm, a module of the software package Sybyl 6.9.1. (36) Payne, A. W. R.; Glen, R. C. J. Mol. Graph. 1993, 11, 74–91. (37) Douguet, D.; Munier-Lehmann, H.; Labesse, G.; Pochet, S. J. Med. Chem. 2005, 48, 2457–2468. (38) Hosoda, H.; Sakai, Y.; Yoshida, H.; Nambara, H. Chem. Pharm. Bull. 1979, 27, 2147–2150.

Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

9151

UV-visible measurements for thermodynamic and kinetic data were obtained with a Varian Cary 1 spectrophotometer, using a 1-cm path quartz cuvette in a thermostated cell holder. Cartridges for solid-phase extraction were from Waters. Liquid chromatography was carried out using a HPLC/Autosampler, equipped with a 100-µL loop/vacuum degasser system, PerkinElmer Series 200. Analytes were separated with a reversed-phase C18 Alltima column (25 cm × 4.6 mm i.d.) packed with 5-µmdiameter particles equipped with an Alltima guard column. The identification and quantization of analytes were carried out using a tandem triple-quadrupole mass spectrometer PE-Sciex API-2000. Peak areas for selected ions were determined using the PE Sciex package Multiview 1.4. Computational Selection of the Receptors. After downloading the enzyme AChE methylphosphonylated from the Web site Protein Data Bank (PDB ID, 1VXO), the PDB file was imported and the binding site three-dimensional disposition analyzed via the open-eye module VIDA. Using Sybyl Software, the virtual tetrapeptides were drawn in zwitterionic mode, adding hydrogen, and charged with the Gasteiger-Hu¨ckel computational method. The minimization was executed using the Powell molecular mechanic method with Tripos force field. Sybyl software default parameters such as dielectric function and dielectric constant, which correspond to vacuum conditions, were used. The termination “gradient” was used, and convergence was reached when the difference in energy between one step and the other was less than 4.84 J (mol Å)-1, in a maximum of 2000 iterations. The same procedure was applied to design both the pesticides paraoxon and carbaryl that were used as templates for the subsequent computational step, where Leapfrog algorithm was used to screen peptide-pesticide interaction. This last program was applied in DREAM mode for 80 000 iterations. In this work, most of the categories were left as default with the exception of energy startup, including hydrogen-bonding energy, which was not active in default conditions. UV-Vis Spectroscopy. The spectrophotometric titrations were carried out in two bulk solvents, water and ethanol. The temperature was held constant at 298 (±0.2) K. The procedure for the titration was as follows: concentrated aliquots of 22.2 mM stock peptide solutions without solid support were added portionwise via a microsyringe to a fixed 60 µM paraoxon solution of 2.50 mL, placed in a 1-cm UV-vis quartz cuvette. The spectra were recorded from 230 to 330 nm (λmax ) 272 nm) against a reference solution containing the same amount of the added peptide. Further readings, carried out every 2 min for a total of 1 h, were monitored in order to check the stability of the adduct with time. The λmax were corrected for the dilution using as reference the calibration curve of the paraoxon in the Lambert-Beer linearity range. The absorbance variation was monitored at different concentration of added peptide. The stability constants (K) were calculated by using a standard equation for a 1:1 complexation: ∆A ) ∆ε × [Pa] × K × [Pe]/(1 + K × [Pe]), where ∆A ) A0 - Ai; ∆ε ) e0 - e∞; [Pe] was the concentration of the added peptide and [Pa] was the constant concentration of the pesticide paraoxon. The computer-aided nonlinear least-squares fitting analyses, to give K and ∆ε have been performed with the program Kaleidagraph with data of at least 8-10 measurements. The results were reproducible within 5%, unless otherwise 9152

Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

indicated. For better visual comparison between peptides, their corresponding plot were normalized. Solid-Phase Extraction and HPLC-MS/MS Detection. The PAL-PEG cartridges were packed with 20 or 50 mg of modified peptide resin dissolved in 5 mL of an ethanol/water solution (80:20, v/v) and kept at room temperature for 6-8 h. This suspension was slowly loaded into the cartridge with a Teflon frit on the bottom. During this procedure, the cartridge was continuously shaken in order to obtain an homogeneous packing. After loading, a second frit was used to cover the resin into the cartridge. Then the cartridge was conditioned and equilibrated by washing with ethanol. All elutions were carried out with flow rate gravity and the elute fractions collected and named progressively. The Amberlyte cartridges were assembled in a similar way using 1 g of derivatized resin in 75 mM bicarbonate buffer pH 7.8 and were then equilibrated in ethanol. Preconditioning of the column was found to be important to obtain acceptable reproducibility; each column was preconditioned with 10-20 cycles of 0.4 µM paraoxon-methylethanol at pH 4.5. Data were considered reliable when the recovery was over 70%. For the PAL-PEG columns, the pesticide sample was dissolved in 30 µM HCl solution, 4.5 pH (or adjusted with NaOH for other pH), at a final volume of 2 mL and loaded on the cartridge. This elute represented the second fraction. The third fraction corresponded of all unbound pesticides flushed out from the resin by using 0.5 mL of water. Fraction 4 was the elution of the pesticide bound by the functionalized peptide-resin recovered using 0.2 mL of ethanol. A further elution, fraction 5, was carried out using another 0.5 mL of ethanol in order to wash the cartridge and control the complete recovery of the pesticide. Fractions were collected and directly transferred into amber glass vials for HPLC-MS/MS analysis. Analysis on 50 mg of PAL-PEG and Amberlite was carried out using loading volume of 0.5 and 1 mL; fraction 4 volumes were 1 and 2 mL, respectively. The analysis of carbaryl and paraoxon was carried out using a procedure developed in an other work39 and here briefly reported. Chromatographic run was carried out using acetonitrile (55%) and water (45%). both added with 1 mM formic acid. The flow rate of the mobile phase in isocratic elution was 1 mL min-1, but only 150 µL min-1 was loaded into the mass spectrometer source. The analytes were detected using a TurboIonSpray source in positive ionization, with a capillary voltage of 5000 V. Quantitative analysis was led in selected reaction monitoring (SRM) by selecting one precursor/fragment ion transition for each analyte. The retention time for paraoxon and carbaryl was respectively 9.3 and 14.2, the mass transitions (SRM) were 276 f 220 and 202 f 145, the declustering potential was 42 and 56, and the collision energy was 28 and 13. RESULTS AND DISCUSSION Rational Design of Biomimetic Receptors. Starting from the biological structure of the methylphosphonylated AChE (PDB code 1VXO) analyzed in complex with the o-ethyl-s-[2-[bis(1methylethyl)amino]ethyl] methylphosphonothioate, we attempted to reproduce a verisimilar shape of template-complementary (39) Perret, D.; Gentili, A.; Marchese, S.; Sergi, M.; D’Ascenzo, G. J. AOAC Int. 2002, 85, 724–730.

Figure 1. Three-dimensional structures of the two pesticides and the three tetrapeptides used in this work. CR are positioned in correspondence of the amino acids label.

docking using the key amino acids in the active site of the enzyme. As reported in the paper of Millard et al.,20 the amino acids involved in the AChE-pesticide binding are Ser 200, His 440, and Glu 327 the so-called “catalytic triad”. The peptides were built taking into account the configuration of these three amino acids, but reducing the size of the ligand to the minimum, in order to control the possible shape of the receptor secondary structure during the computational simulation. In fact, keeping the size of the receptor to the minimum could minimize unexpected divergences between the synthesized peptide and the original shape predicted by molecular modeling. The three peptides were selected from a library proposed in a previous preliminary work,40 built studying the interaction of 24 tetrapeptides with 26 different organophosphate and carbamate pesticides. Computational data demonstrated that (a) the use of four amino acidic units represented a good compromise to have a certain degree of spatial similarity with the pesticide binding region of the active site of the enzyme and (b) the peptides carrying residues of the triad appeared to be important to have high calculated binding energies. In order to demonstrate the “proof of concept” that a computational approach can be used to reduce the experimental work for the screening of this receptors, we selected for the binding experiments the three tetrapeptides reported in Figure 1. The peptides were chosen according with the following simple criteria: the original catalytic triad of the AChE active site was maintained in peptide A with Pro used among the Glu and Ser residues. In peptides B and C, the triad was partially replaced keeping Ser and substituting Glu (peptide B) and His (peptide C) with Ala. Gly or Pro was used to obtain the proper distance among the residues in a linear or bended structure. The peptides were tested with two model pesticides, the organophosphate paraoxon-methyl and the carbamate carbaryl (Figure 1). (40) Mascini, M.; Del Carlo, M.; Compagnone, D. In Sensors and Microsystems: Proceedings of the 9th Italian Conference; World Scientific: Singapore, 2004; pp 44-49.

Table 1. Leapfrog Algorithm Results Using the Selected Tetrapeptides vs Carbaryl and Paraoxon-Methyl binding score (kJ mol-1) ID

peptide

carbaryl

paraoxon-methyl

A B C

His-Glu-Pro-Ser His-Gly-Ser-Ala Glu-Pro-Ser-Ala

94.2 85.8 DSa

93.4 72.7 20.7

a

DS, discarded by the algorithm.

The binding energy of the pesticide-peptide complex was studied using the Leapfrog algorithm run to screen each single ligand for its possible interaction with the two templates coming from both classes of pesticide (CM and OP). The empirical binding scores are reported in Table 1. The ligand giving the highest binding score was thought capable of forming the strongest complex with the template. It could be seen a good correlation between predicted peptide shape and binding score. In fact, the most similar structure represented by the peptide A, having all the three amino acids involved in the pesticide binding, showed the best interaction with both CM and OP templates. The geometry of the amino acids involved in the pesticides binding of the AChE (the catalytic triad Ser 200, His 440, Glu 327) and the designed peptides was then compared, using as parameters the angles and distances of the CR of each aminoacid (Table 2). The angles were calculated wherever the position in the peptide or AChE sequence using the CR of the central amino acid (in boldface in Table 2). In the peptides B and C, the amino acid Ala was chosen to substitute respectively Glu or His. Looking at the angles, it can be noticed that only the peptide A matched in part the geometry of the catalytic triad, showing a difference only in the angle formed from the triad His-Ser-Glu. The angles of peptide B were considerably different from the original spatial position in the enzyme AChE especially in the angle Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

9153

Table 2. Geometry of the Amino Acids Involved in the Pesticide Bindinga AChE Ser...Glu...His Glu...His...Ser His...Ser...Glu

Ser

Glu

AChE A B C AChE A B C

A B Angles CR-CR (deg)

55.7 98.1 26.3

50.6 85.6 43.8

89.6 34.7 55.6

C 42.4 77.9 59.7

Ser

Glu distances CR-CR (Å)

His

0 0 0 0 10.00 5.63

10.00 5.63

8.34 4.36 6.91

5.70

5.70 0 0 0 0

4.47 3.91

Figure 2. Comparison of the spectrum obtained using a water solution of 60 µM paraoxon (a) and the spectra recorded by adding to the paraoxon solution different concentrations of peptide A, obtaining a progressively increase of the absorbance following the increase of molecular ratio (MR) peptide/paraoxon, respectively, 5, 25, 50, 100, and 150. The spectrum pointed out with (b) is that obtained with MR 150.

a

The angles and distances were on the basis of the amino acids involved in the catalytic triad (Ser 200, His 440, Glu 327) selected from the structure of the AChE enzyme (PDB code 1VXO). The starting points of the rays were the CR of each external amino acid sharing as common end point the CR of the central amino acid (the vertex of the angle, in boldface type). In the peptides B and C, Ala was chosen to substitute respectively Glu or His.

formed by the amino acids Ala-His-Ser. Also, the angles in peptide C were quite different from the mimed structure, with a divergence, observed in the angle formed from the triad Ala-Ser-Glu, of 33.4°. Considering the distances between the CR of each amino acid involved in the binding, the AChE’s Ser distance from both His and Glu was twice that calculated using amino acids from peptides A, B, and C. Only the distance between Glu and His resembled the AChE’s structure. Despite the differences between the peptides’ shape and the structure of the enzyme AChE, the peptides A and B had relatively high binding scores as reported in Table 1. In fact, the potential pesticide receptors designed possess residues that are able to interact with template predominantly through electrostatic, van der Waals interactions, and hydrogen bonds. The binding can be explained as a synergic cooperation of the residue group in each amino acid that has a certain amount of freedom to move around the carbon backbone (much larger than in a protein), keeping the probability to interact via electrostatic interactions with the pesticide target atoms involved in the binding. UV-Vis Spectroscopy. In order to check the experimental ability of the tetrapeptides to bind the pesticides in solution and to have an idea of their features for possible solid-phase extraction, we explored the formation of pesticide-peptide complex changing the bulk solvent properties by following the variation with UV-visible spectroscopy. For this purpose, paraoxon was selected because of its nitro aromatic chromophore that in micromolar range obeys the Lambert-Beer law. Figure 2 reported the spectral variation of a 60 µM water solution of paraoxon, upon addition of increasing amount of peptide A. A progressive raise of absorbance was observed with no evident λmax shift. UV-visible spectra upon peptide addition were always hyperchromic and did not give any variation with time. In fact, by recording the spectra every minute for 30 min, no change either 9154

Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

Figure 3. Normalized changes of absorbance at 270 nm upon mixing 60 µM paraoxon with different concentrations of peptides A (solid line), B (dash-dotted line), and C (dotted line). (a) Using water as bulk solvent (b). Using ethanol as bulk solvent. For each point, at least 8-10 measurements were carried out. The results were reproducible within 5%, unless otherwise indicated.

in the λmax shift or in adsorption intensity was observed. This indicated for each of the peptides a fast complex formation, stable with time. In Figure 3, the Langmuir titration plots both in water and in ethanol for the interaction of paraoxon with the three selected peptides were reported. Data were normalized for sake of clarity because of the large differences in ∆ε variation.

Table 3. Spectrophotometric Parameters of the Complex Peptide-paraoxon Calculated Using a Langmuir-Type Model, Supposing 1:1 Complexationa peptide

K (M-1)

R

∆ε

peptide fraction

A, %

B, %

C, %

resin, %

1.3 ± 0.2 2.0 ± 0.2 4.0 ± 0.4

1 2 3 4 5

0.0 ± 0.0 10.0 ± 0.5 17.5 ± 0.5 72.5 ± 2.9 0.0 ± 0.0

0.0 ± 0.0 35.7 ± 1.4 17.2 ± 0.7 47.1 ± 2.4 0.0 ± 0.0

0.0 ± 0.0 24.2 ± 1.0 20.6 ± 0.7 55.2 ± 2.4 0.0 ± 0.0

0.0 ± 0.0 36.6 ± 1.8 16.5 ± 0.6 46.9 ± 2.3 0.0 ± 0.0

Water A B C

506 ± 115 116 ± 20 36 ± 2

0.99 0.99 0.99

a Experimental results of resin with or without tetrapeptides loading 200 µL of 4 µM paraoxon solution at pH 4.5.

Ethanol A B C a

109 ± 21 22 ± 2