Anal. Chem. 2009, 81, 7081–7086
Determination of Cocaine in Human Plasma by Selective Solid-Phase Extraction Using an Aptamer-Based Sorbent Benjamin Madru,† Florence Chapuis-Hugon,† Eric Peyrin,‡ and Vale´rie Pichon*,† Department of Environment and Analytical Chemistry, ESPCI ParisTech, UMR PECSA (ESPCI ParisTech-UPMC Univ Paris 06-CNRS), 10 rue Vauquelin, 75231 Paris cedex 05, France, and De´partement de Pharmacochimie Mole´culaire UMR 5063, Institut de Chimie Mole´culaire de Grenoble FR 2607, CNRS-Universite´ Grenoble I (Joseph Fourier), 38041 Grenoble cedex 9, France A complete characterization is presented of a highly selective solid-phase extraction (SPE) sorbent which exploits the properties of aptamers. An oligosorbent based on aptamers immobilized on a solid support was synthesized and tested for the selective extraction of cocaine from human plasma. Anticocaine aptamers were immobilized to CNBr-activated Sepharose, and an extraction procedure was developed in pure media. Specific retention of cocaine on the oligosorbent was demonstrated, and the capacity of the support was determined. This oligosorbent was then applied to the selective extraction of cocaine from plasma at a concentration of 0.4 mg L-1, i.e., corresponding to the plasma concentration reached after an intake of a single dose of cocaine. Extraction recovery close to 90% was obtained. Moreover, interfering compounds that perturbed cocaine quantification when using a standard SPE sorbent were not retained on the oligosorbent, thus allowing fast and reliable analyses of plasma samples with an estimated limit of detection of 0.1 µg mL-1. Despite advances in detector specificity and sensitivity, trace level analysis of a target analyte in biological or environmental samples often requires sample treatment. Solid-phase extraction (SPE) is commonly used to purify and concentrate target analytes from complex samples. However, hydrophobic SPE sorbents (e.g., C18, polymers) commonly used for the treatment of aqueous samples (environmental water, biological fluids) mainly develop hydrophobic interactions, so many interfering compounds with similar polarity to the analyte are coextracted, thus affecting the reliability of the analytical methods. Selective SPE methods have been recently developed to obtain an extract free from matrix interference in a single extraction step. These sorbents involve retention mechanisms based on molecular recognition of target molecule(s). The first approach uses immunosorbents (IS) made of antibodies immobilized on a solid support. The high affinity and selectivity of the antigen-antibody interactions allow the * Corresponding author. E-mail:
[email protected]. Fax: 33 (1) 40 79 47 76. † ESPCI ParisTech. ‡ CNRS-Universite´ Grenoble I (Joseph Fourier). 10.1021/ac9006667 CCC: $40.75 2009 American Chemical Society Published on Web 07/24/2009
selective extraction of the target analyte and its byproduct.1-3 However, the development of IS is expensive and time-consuming. Another approach is the synthesis of molecularly imprinted polymers (MIPs), which are sorbents possessing specific cavities complementary to the template in size, shape, and position of the functional groups.4,5 The synthesis of these sorbents is rapid and relatively easy. However, it necessitates a large amount of target analyte or of one of its analogues, and this production can be costly for scarce and expensive molecules. In this study, we evaluated a third selective SPE strategy based on molecular recognition using aptamers immobilized on a solid support. Aptamers are single-stranded oligonucleotides that bind target molecules with very high affinities, equivalent to those of antibodies.6 They can be generated against various targets such as divalent metal ions,7 small organic molecules,8 proteins,9 and cells.10 They are identified for each target molecule within randomly synthesized nucleic acid libraries containing up to 1015 different candidates by an iterative in vitro process of selection and amplification. This process is called “systematic evolution of ligands by exponential enrichment” (SELEX).11,12 The aptamers have many advantages over antibodies. While antibodies need one or two days to recover their native active conformation after denaturation, aptamers can be regenerated within minutes. Once the aptamer sequence is identified by SELEX, the aptamers are chemically synthesized resulting in little or no batch-to-batch variation. Chemical synthesis also allows the introduction of modifications to enhance aptamer stability, specificity, detection, (1) Hennion, M.-C.; Pichon, V. J. Chromatogr., A 2003, 1000, 29–52. (2) Delaunay-Bertoncini, N.; Pichon, V.; Hennion, M.-C. LCGC Eur. 2001, 162– 172. (3) Majors, R. E. LCGC Asia Pacific 2008, May 1. (4) Pichon, V. J. Chromatogr., A 2007, 1152, 41–53. (5) He, C.; Long, Y.; Pan, J.; Li, K.; Liu, F. J. Biochem. Biophys. Methods 2006, 70, 133–150. (6) Liss, M.; Petersen, B.; Wolf, H.; Prohaska, E. Anal. Chem. 2002, 74, 4488– 4495. (7) Hofmann, H. P.; Limmer, S.; Hornung, V.; Sprinzl, M. RNA 1997, 3, 1289– 1300. (8) Jenison, R. D.; Gill, S. C.; Polisky, B. Science 1994, 263, 1425–1429. (9) Bock, L. C.; Griffin, L. C.; Latham, J. A.; Vermaas, E. H.; Toole, J. J. Nature 1992, 355, 564–566. (10) Cerchia, L.; Duconge´, F.; Pestourie, C.; Boulay, J.; Aissouni, Y.; Gombert, K.; Tavitian, B.; de Franciscis, V.; Libri, D. PLoS Biol. 2005, 3, e123. (11) Sampson, T. World Pat. Inf. 2003, 25, 123–129. (12) Stoltenburg, R.; Reinemann, C.; Strehlitz, B. Biomol. Eng. 2007, 24, 381– 403.
Analytical Chemistry, Vol. 81, No. 16, August 15, 2009
7081
and immobilization. Moreover, aptamers are produced by an in vitro process which does not require animals. Various analytical aptamer-based formats have been used, including enzyme-linked oligonucleotide assay (ELONA, variant of ELISA with aptamers instead of antibodies),13 biosensors (“aptasensors”),14,15 affinity chromatography, electrochromatography, and affinity capillary electrophoresis16-18 or matrix-assisted laser desorption/ionization mass spectrometry.19 As far as we know, only one study demonstrates the feasibility of using aptamers for SPE,20 by applying it to the extraction of a mycotoxin from wheat grain extracts. Despite the high recoveries mentioned, the sorbent made was not fully characterized in terms of breakthrough volume, nonspecific retention, binding efficiency, capacity, or reusability. Moreover, the specificity of the fluorescence detection mode used does not allow complete evaluation of the selectivity of the sorbent. We describe the use of aptamers immobilized on a solid support as selective SPE sorbents, with full characterization of the resulting oligosorbent. The sequence of the anticocaine aptamer used has been published by Stojanovic et al.21 and has already been used for the design of several aptasensors.21-30 The aptamers were covalently immobilized on CNBr-activated Sepharose. An extraction procedure was then developed in pure media in order to determine the specific retention of cocaine on the resulting oligosorbent. The capacity of the support was also investigated, and cocaine was extracted from plasma to assess sorbent selectivity. The oligoextraction was then compared with a standard SPE procedure on C18 silica. EXPERIMENTAL SECTION Chemicals. Cocaine hydrochloride, benzoylecgonine solution, sodium phosphate (Na2HPO4), Trizma hydrochloride, sodium azide, sodium chloride, and CNBr-activated Sepharose (4B, 90 µm) were from Sigma-Aldrich (Saint-Quentin Fallavier, France). Magnesium chloride, potassium dihydrogen phosphate (KH2PO4), hydrochloric acid, and acetic acid were from VWR (13) Drolet, D. W.; Moon-McDermott, L.; Romig, T. S. Nat. Biotechnol. 1996, 14, 1021–1025. (14) Potyrailo, R. A.; Conrad, R. C.; Ellington, A. D.; Hieftje, G. M. Anal. Chem. 1998, 70, 3419–3425. (15) Song, S.; Wang, L.; Li, J.; Zhao, J.; Fan, C. Trends Anal. Chem. 2008, 27, 108–117. (16) Ravelet, C.; Grosset, C.; Peyrin, E. J. Chromatogr., A 2006, 1117, 1–10. ¨ zalp, V. C.; Sanchez, P. L.; Mir, M.; Katakis, I.; O’Sullivan, (17) Mairal, T.; O C. K. Anal. Bioanal. Chem. 2008, 390, 989–1007. (18) Tombelli, S.; Minunni, M.; Mascini, M. Biosens. Bioelectron. 2005, 20, 2424– 2434. (19) Dick, L. W.; McGown, L. B. Anal. Chem. 2004, 76, 3037–3041. (20) Cruz-Aguado, J. A.; Penner, G. J. Agric. Food Chem. 2008, 56, 10456–10461. (21) Stojanovic, M. N.; de Prada, P.; Landry, D. W. J. Am. Chem. Soc. 2001, 123, 4928–4931. (22) Stojanovic, M. N.; de Prada, P.; Landry, D. W. J. Am. Chem. Soc. 2000, 122, 11547–11548. (23) Stojanovic, M. N.; Landry, D. W. J. Am. Chem. Soc. 2002, 124, 9678–9679. (24) Li, Y.; Qi, H.; Peng, Y.; Yang, J.; Zhang, C. Electrochem. Commun. 2007, 9, 2571–2575. (25) Liu, J.; Lu, Y. Angew. Chem., Int. Ed. 2006, 45, 90–94. (26) Shlyahovsky, B.; Li, D.; Weizmann, Y.; Nowarski, R.; Kotler, M.; Willner, I. J. Am. Chem. Soc. 2007, 129, 3814–3815. (27) Liu, J.; Mazumdar, D.; Lu, Y. Angew. Chem., Int. Ed. 2006, 45, 7955–7959. (28) Baker, B. R.; Lai, R. Y.; Wood, M. S.; Doctor, E. H.; Heeger, A. J.; Plaxco, K. W. J. Am. Chem. Soc. 2006, 128, 3138–3139. (29) Li, T.; Li, B.; Dong, S. Chem.-Eur. J. 2007, 13, 6718–6723. (30) Swensen, J. S.; Xiao, Y.; Ferguson, B. S.; Lubin, A. A.; Lai, R. Y.; Heeger, A. J.; Plaxco, K. W.; Soh, H. T. J. Am. Chem. Soc. 2009, 131, 4262–4266.
7082
Analytical Chemistry, Vol. 81, No. 16, August 15, 2009
(Fontenay-sous-bois, France). Human plasma was provided by EFS (Strasbourg, France). HPLC-grade acetonitrile was from Mallinckrodt Baker (Deventer, The Netherlands). High-purity water was obtained using a Milli-Q purification system (Millipore, Saint-Quentin en Yvelines, France). The 5′-amino-modified DNA oligonucleotides (sequence 5′-GGGAGACAAGGAAAATCCTTCAATGAAGTGGGTCGACA-3′ for the aptamer and 5′AAGTGAACAGAAGGCGTCATAGAGCGAAGTACGATGTC3′ for the scrambled oligonucleotide) with a C6 spacer arm were synthesized and HPLC-purified by Eurogentec (Angers, France). The selection buffer contained 20 mM Trizma hydrochloride, 140 mM NaCl, 5 mM KCl, and 1 mM MgCl2, pH ) 7.4. Apparatus and Analytical Conditions. An Agilent 1200 series (Agilent Technology, Massy, France) RRLC (Rapid Resolution LC) system equipped with a binary pump, an autosampler, and a diode array detector controlled by Chemstation software was used. Cocaine was quantified using a Waters SymmetryShield RP18 column (150 × 2.1 mm i.d., 3.5 µm, Waters, Saint-Quentin-enYvelines, France) maintained at 35 °C with a column oven (Crococil, Interchim, Montluc¸on, France). An isocratic mode was chosen with a mix of 62% 5 mM phosphate buffer (pH ) 6) and 38% acetonitrile at a flow rate of 0.2 mL min-1. The detection of cocaine was carried out at 233 nm. Synthesis of the Oligosorbent. Aptamers were immobilized on CNBr-activated Sepharose using a procedure adapted from the immobilization of antibodies and enzymes previously described by our group.31 Prior to immobilization, the oligonucleotides were renatured by heating the aptamer solution (1 g L-1 in a 200 mM Na2HPO4 and 5 mM MgCl2, pH ) 8) at 75 °C for 5 min and leaving it to stand at room temperature for 30 min. Dry CNBractivated Sepharose (35 mg) was swollen and washed six times with 1 mL of 1 mM HCl. Aptamer solution (150 µL) was mixed with the gel overnight at room temperature. The oligosorbent was then packed between two frits in a 1 mL SPE cartridge and washed with 3 mL of 200 mM Na2HPO4 (pH ) 8). Remaining active groups of the support were blocked by a 0.1 M Trizma solution (pH ) 8) for 2 h at room temperature. The gel was then washed alternately three times with 2 mL of an aqueous saline buffer (0.1 M acetate + 0.5 M NaCl, pH 4) and 2 mL of a Trizma buffer (0.1 M + 0.5 M NaCl, pH 8) to remove noncovalently bound aptamers. To evaluate nonspecific interactions between cocaine and the Sepharose-based sorbent, a blank sorbent was prepared following the same procedure but without aptamers. To estimate the risk of nonspecific interactions between cocaine and single-stranded DNA, the same immobilization procedure was followed with a scrambled aptamer (with the same base composition as the active form but in a random order). Extraction Procedures. Before each extraction, the oligosorbent was conditioned with 5 mL of the selection buffer (20 mM Trizma hydrochloride, 140 mM NaCl, 5 mM KCl, and 1 mM MgCl2, pH ) 7.4, 4 °C). The optimized extraction procedure used consisted of the percolation of 200 µL of the selection buffer at 4 °C containing variable amounts of cocaine (8-4250 ng) followed by a washing step with 300 µL of the selection buffer at 4 °C. Cocaine was then eluted with 400 µL of a water/ (31) Cingo ¨z, A.; Hugon-Chapuis, F.; Pichon, V. J. Chromatogr., A 2008, 1209, 95–103.
Figure 1. (A) Cocaine structure. (B) Anticocaine aptamer (secondary structure) bound to cocaine (black ellipsoid) adapted from Stojanovic et al.21
acetonitrile solution (60/40, v/v) at ambient temperature. The elution fraction was diluted once with water, and the injection volume for HPLC was 50 µL. The standard SPE procedure on C18 silica was adapted from Maralikova et al.32 Briefly, an Isolute C18 cartridge (200 mg, Biotage, Courtaboeuf, France) was conditioned with 2 mL of MeOH, 1 mL of deionized water, and 1 mL of 4 mM ammonium carbonate buffer (AC buffer) (pH 9.4). Two milliliters of AC buffer was then added to 1 mL of plasma, and this mixture was applied to the SPE cartridge. A washing step of 1 mL of AC buffer was followed by the elution, using 1.5 mL of 2% (v/v) acetic acid in methanol. The extracts were evaporated to dryness under a nitrogen stream, and the residue was reconstituted with 8 mL of the HPLC mobile phase to match the same dilution factor as for the oligoextraction procedure. Treatment of Plasma Samples. One milliliter of human plasma spiked with 400 ng of cocaine was diluted with 1 mL of selection buffer for the oligoextractions and with 2 mL of AC buffer for the procedure on C18 silica. After filtration using a Millex-HV filter (0.45 µm, 13 mm, Millipore) and then a Millex-LG filter (0.2 µm, 13 mm, Millipore), the filtered supernatant was passed through the oligosorbent or the C18 silica SPE cartridge. RESULTS AND DISCUSSION Selective Oligo-Extraction of Cocaine from Pure Media. An anticocaine aptamer identified by Stojanovic et al.21 (Figure 1) was 5′-amino-modified and immobilized on a CNBr-activated Sepharose following a procedure adapted from the immobilization of antibodies and enzymes previously described by our group.31,33 The structure of cocaine and the secondary structure of the anticocaine aptamer are shown in Figure 1. The formation of the complex between an aptamer and its ligand is highly dependent on the 3D structure of the oligonucleotide, itself being dependent on the medium composition. A medium permitting high affinity is used during the selection of the aptamers in the SELEX process. Therefore, a 200 µL sample of selection buffer (20 mM Trizma hydrochloride, 140 mM NaCl, 5 mM KCl, 1 mM MgCl2; pH 7.4, 4 °C) was spiked with 25 ng of cocaine and percolated through the oligosorbent (35 mg packed in a 1 mL cartridge). In order to study the strength of (32) Maralikova, B.; Weinmann, W. J. Chromatogr., B 2004, 811, 21–30. (33) Maisonnette, C.; Simon, P.; Hennion, M.-C.; Pichon, V. J. Chromatogr., A 2006, 1120, 185–193.
the interactions between cocaine and immobilized aptamers, various washing fractions of the selection buffer at 4 °C were then percolated through the oligosorbent. To estimate the risk of nonspecific interactions between cocaine and the activated Sepharose, the same experiments were carried out in parallel on a blank sorbent prepared following the same procedure but without aptamers. The recovery yields of cocaine in the percolated and washing fractions are reported in Figure 2. The profiles obtained using the oligosorbent and the blank support were different. After the sample percolation and the first two washing fractions, cocaine was eluted from the blank support, whereas nearly 90% of the cocaine was still retained on the oligosorbent. This difference of retention demonstrates that immobilized aptamers can bind cocaine strongly in these conditions. These results indicate that a washing volume of 300 µL should break low nonspecific interactions generated by the sorbent without affecting the retention of cocaine by aptamers: 94% of cocaine was removed from the blank sorbent while more than 89% of the percolated cocaine was retained on the oligosorbent. This percolation and these washing conditions were then fixed for the selective extraction procedure. Elution of cocaine was improved by using a solution leading to alteration of the DNA tertiary structure to an inactive form. Denaturation of the oligonucleotide may occur using a hydro-organic solution, which could also disrupt the interactions between the target and its aptamer.21 The eluent was a water/acetonitrile mixture (60:40 v/v, 400 µL) at ambient temperature, which is compatible with the analytical conditions: a simple dilution with an equivalent volume of water is performed before the LC injection. To check for possible nonspecific interactions between an oligonucleotide and cocaine, a second control sorbent was synthesized. An aptamer with the same base composition as the active form but in a random order was also immobilized on Sepharose. Since this scrambled aptamer is expected to have no affinity for cocaine, nonspecific interactions between cocaine and a nonspecific DNA sequence could be estimated. Results for application of the final selective extraction procedure to the different sorbents are reported in Figure 3. The elution profile obtained with the oligosorbent was different from those obtained with the two other sorbents. Less than 10% of cocaine was recovered in the elution fraction of both control supports, indicating that few nonspecific interactions were developed during this extraction procedure. This comparison demonstrates a high specific retention of cocaine on the oligosorbent. Repeatability of this optimized extraction procedure is reported in Figure 4. Extraction recoveries of 86% were obtained for the oligosorbent, with RSD values lower than 10% (n ) 3), whereas 6% ± 1% was observed using the blank support. Five milliliters of selection buffer was sufficient to condition the cartridge between two consecutive uses, showing the simplicity of the regeneration procedure. Determination of the Oligosorbent Capacity. The oligosorbent was characterized in terms of capacity, which is directly linked to the number of active aptamers covalently fixed to the support. Various samples containing increasing amounts of cocaine (from 8 to 4250 ng) were percolated through the oligosorbent, and the same experiments were carried out in parallel on the blank sorbents. The cocaine amount found in the Analytical Chemistry, Vol. 81, No. 16, August 15, 2009
7083
Figure 2. Comparison of the retention of cocaine on the oligosorbent and on the blank support when selection buffer was passed through the sorbent. Percolation: 200 µL of selection buffer spiked with 150 µg/L cocaine. Washing with the selection buffer: 200 µL for W1 and 100 µL for W2-W9.
Figure 5. Capacity curves of the oligosorbent and the blank support. Mass of fixed cocaine corresponds to the amount found in the elution fraction after applying the extraction procedure described in Figure 3.
Figure 3. Extraction profile of cocaine from an aqueous medium on the oligosorbent and both control sorbents. All supports were conditioned by 5 mL of selection buffer at 4 °C. P: 200 µL of selection buffer at 4 °C spiked with 5 ng of cocaine; W: 300 µL of selection buffer at 4 °C; E: 400 µL of water/acetonitrile (60:40 v/v) at room temperature.
Figure 4. Repeatability of the extraction protocol on the oligosorbent and on the blank support. Same conditions as Figure 3 (n ) 3).
elution fractions of both sorbents for the various percolated samples are reported in Figure 5. A linear response was obtained for the blank sorbent with a slope of 5%, corresponding to the 7084
Analytical Chemistry, Vol. 81, No. 16, August 15, 2009
low recovery of cocaine by this sorbent. For the oligosorbent, at lower doses of cocaine, the resulting curve is linear with a slope of 86%, which corresponds to the extraction recovery previously obtained. At higher doses, the amount of bound cocaine reached a plateau with a slope around 5%, corresponding to a nonspecific retention of cocaine (saturation of the oligosorbent). The capacity corresponds to the amount of cocaine introduced at the upper limit of the steep part of the curve which is 733 ng, i.e., 69.2 nmoles per gram of oligosorbent. This point also represents the upper limit of the concentration range that can be used for quantitative extraction and is sufficient for extraction of trace amounts of cocaine from biological samples. Assuming one molecule of cocaine is retained by one aptamer, a theoretical binding efficiency can be calculated. The use of 150 µg of aptamers, i.e., 12.5 nmoles, permitted specific retention of a maximum of 733 ng of cocaine, i.e., 2.4 nmoles for 35 mg of support, which corresponds to a capacity of 69.2 nmol/g of sorbent. Therefore, the binding efficiency for the immobilization of active aptamers to the Sepharose was estimated to be 19.3%. This capacity is very close to those obtained with immunosorbents (from 4 to 93.6 nmol/g of sorbent, e.g., immunosorbents specific to pesticides).34 The (34) Pichon, V.; Bouzige, M.; Mie`ge, C.; Hennion, M.-C. Trends Anal. Chem. 1999, 18, 219–235.
Figure 6. HPLC chromatograms resulting from the analysis of elution fractions from an extraction by a conventional C18 silica column (A) and oligoextraction (B).
small size of aptamers compared with antibodies should allow higher binding density using optimized immobilization conditions. Furthermore, unlike randomly oriented immobilization of antibodies,1 the aptamers are linked by a defined part (3′ end or 5′ end) of their structure, which should lead to a very high ratio of active immobilized aptamers, permitting higher capacities. Optimization of the binding conditions would probably improve binding efficiencies and sorbent capacities. Selective Extraction of Cocaine from Plasma. To determine the ability of this oligosorbent to extract cocaine from complex samples, the extraction procedure developed above was applied to human plasma spiked with 0.4 µg mL-1 cocaine. This corresponds to the plasma cocaine concentration reached after the intake of a single dose of cocaine.35 Diluted and filtered plasma (200 µL) containing 40 ng of cocaine was percolated through the oligosorbent. An extraction recovery close to 90% as previously obtained in pure media indicated that oligosorbent efficiency was not influenced by a matrix effect. To check the integrity of the oligosorbent after the percolation of plasma containing DNases, which catalyze the cleavage of the DNA backbone, extraction in pure media was carried out to compare retention of the oligosorbent before and after application of the plasma sample. No loss in retention was observed, with an extraction recovery close to 90%, demonstrating that the sorbent was sufficiently stable for the treatment of biological samples. To check the selectivity of this oligoextraction procedure, it was compared with a standard solid-phase extraction procedure using a C18 silica SPE cartridge. The extraction protocol used was adapted from Maralikova et al.32 and resembles SPE protocols generally used for the analysis of plasma in cases of drug abuse. Chromatograms of both elution fractions are compared in Figure 6. The chromatogram corresponding to the oligoextraction (Figure 6B) indicated that fewer compounds were coextracted than by the C18 silica sorbent (Figure 6A). Moreover, when using C18 silica, the cocaine peak is coeluted with an interfering compound so that cocaine quantification is difficult. This compound, which has a similar polarity to cocaine, was not extracted by the oligosorbent. This improvement of the purification step brought about by aptamer selectivity could benefit several analytical methods, including MS detection, which often suffers from matrix effects. The chromatogram resulting from the oligoextraction also (35) Jenkins, A. J.; Oyler, J. M.; Cone, E. J. J. Anal. Tox. 1995, 19, 359–374.
revealed the presence of a coextracted compound at 4.2 min. This analyte might be a steroid, because cocaine aptamers are known to be cross-reactive to these hydrophobic ligands.36 However, the short separation used is sufficient to isolate cocaine from this coextracted compound. Further experiments in mass spectrometry would elucidate its identity. The reusability of the oligosorbent for plasma samples was also investigated. Three consecutive oligoextractions in plasma were carried out, revealing a decrease in extraction recovery (72%, 43%, and 41%). To check the stability of the oligosorbent, extraction in pure media was carried out but revealed no damage to the oligosorbent, with an extraction recovery of 87%. Decrease in extraction recovery cannot be attributed to a sorbent fouling by plasma proteins because no additional cleaning step was performed before this control extraction. It is probably due to progressive protein binding of cocaine in plasma as previously reported.37-39 Protein-bound cocaine is not recognized by the aptamers, and the sorbent extracts only the free form of cocaine. Improved plasma pretreatment with protein precipitation would circumvent the problem but was not carried out in this study, which concentrates on the improvement in selectivity conferred by the oligosorbent for the treatment of very complex matrices. The limit of detection obtained with this selective method is estimated to be 0.1 µg of cocaine/mL of plasma (by taking a signal-to-noise ratio of 10). The lifetime of the oligosorbent was also assessed, and no loss in retention was observed eight months after its synthesis, demonstrating the excellent stability of the sorbent obtained by covalent linkage of aptamer to Sepharose. CONCLUSIONS We have demonstrated that an anticocaine aptamer-based sorbent can be used for the selective extraction of cocaine from human plasma. Covalent binding of amino-aptamers to CNBractivated Sepharose was successfully carried out, leading to a (36) Stojanovic, M. N.; Green, E. G.; Semova, S.; Nikic, D. B.; Landry, D. W. J. Am. Chem. Soc. 2003, 125, 6085–6089. (37) Edwards, D. J.; Bowles, S. K. Pharm. Res. 1988, 5, 440–442. (38) Parker, R. B.; Williams, C. L.; Laizure, S. C.; Lima, J. J. J. Pharmacol. Exp. Ther. 1995, 275, 605–610. (39) Sukbuntherng, J.; Martin, D. K.; Pak, Y.; Mayersohn, M. J. Pharm. Sci. 1996, 85, 567–571.
Analytical Chemistry, Vol. 81, No. 16, August 15, 2009
7085
satisfactory capacity, an excellent sorbent time, and nuclease stability. An extraction protocol developed in pure media led to a high extraction recovery. This oligosorbent was then applied to the extraction of cocaine from human plasma spiked with cocaine. Extraction recovery was close to 90%, and the resulting chromatogram was free of all interfering compounds that perturbed cocaine determination when using SPE on C18 silica. These results confirm that the use of oligosorbents as selective SPE sorbents constitutes a very promising approach to selective sample treatment of complex matrices due to the advantages of aptamers. Other immobilization conditions and pressure-resistant supports should be evaluated to optimize
7086
Analytical Chemistry, Vol. 81, No. 16, August 15, 2009
capacity and to integrate the oligosorbent in online coupling with liquid chromatography. ACKNOWLEDGMENT This work was supported by the French National Research Agency (Micraptox Project, ANR SEST). We thank Jean-Jacques Toulme´ for his valuable contributions throughout this project.
Received for review March 31, 2009. Accepted July 14, 2009. AC9006667
