Selection of single-chain variable fragment antibodies against fenitrothion by ribosome display

Analytical Biochemistry 421 (2012) 130–137

Contents lists available at SciVerse ScienceDirect

Analytical Biochemistry journal homepage: www.elsevier.com/locate/yabio

Selection of single-chain variable fragment antibodies against fenitrothion by ribosome display Yihui Luo, Yuxian Xia ⇑ Genetic Engineering Research Center, College of Bioengineering, Chongqing University, Chongqing 400030, People’s Republic of China

a r t i c l e

i n f o

Article history: Received 15 July 2011 Received in revised form 1 October 2011 Accepted 25 October 2011 Available online 11 November 2011 Keywords: Fenitrothion Organophosphorus pesticide Single-chain variable fragment Ribosome display Immunoassay

a b s t r a c t A single-chain variable fragment (ScFv) complementary DNA (cDNA) library against fenitrothion was constructed, and ScFvs specific for fenitrothion were selected by ribosome display from the library. After three rounds of ribosome display, the ScFv genes were cloned into Escherichia coli for expression. The expressed ScFvs of 160 clones were analyzed by indirect enzyme-linked immunosorbent assay (ELISA). Of these, 40 clones produced antibodies with relatively high activity against fenitrothion, and 3 were selected for Biacore and ELISA analysis. These 3 antibodies—ScFv–AF50, ScFv–AF93, and ScFv–AF132— had IC50 values of 1.6, 3.4, and 2.2 ng/ml, respectively. Cross-reactivity with other organophosphorus (OP) pesticides was below 0.1% except for parathion-methyl (62.8%). The IC50 values and cross-reactivity were lower than achieved previously with polyclonal or monoclonal antibodies against fenitrothion. The equilibrium dissociation constant (KD) values determined by Biacore analysis were 4.56  1010 M for ScFv–AF50, 1.42  109 M for ScFv–AF93, and 2.66  1010 M for ScFv–AF132. These results demonstrate that the ribosome display has great potential in selection of ScFvs against pesticides. Recoveries of fenitrothion from fortified rice and cucumber were in the range 80.6 to 108%, indicating that the ELISAs with the isolated ScFvs can accurately determine fenitrothion in food samples after the simple and rapid extraction procedure. Ó 2011 Elsevier Inc. All rights reserved.

Pesticides have made major contributions to agriculture and disease control, but widespread use has created serious concerns regarding their effects on the environment and on human health. Organophosphorus (OP)1 pesticides are among the most widely used in horticulture, agriculture, and forestry. In high doses, OP pesticides can cause respiratory, myocardial, and neuromuscular impairments [1], necessitating high-sensitivity detectors for environmental and biological samples. Current methods, such as gas chromatography (GC) and high-performance liquid chromatography (HPLC), have been used successfully, with high sensitivity and reliability for analysis of many pesticides, including OP pesticides [2–4]. These methods are expensive and time-consuming, however, and can be performed only by skilled analysts at off-site laboratories. ⇑ Corresponding author. Fax: +86 23 65120490. E-mail address: [email protected] (Y. Xia). Abbreviations used: OP, organophosphorus; GC, gas chromatography; HPLC, highperformance liquid chromatography; ELISA, enzyme-linked immunosorbent assay; CDR, complementarity-determining region; ScFv, single-chain variable fragment; BSA, bovine serum albumin; OVA, ovalbumin; FE–BSA, fenitrothion–BSA; FE–OVA, fenitrothion–OVA; cDNA, complementary DNA; RT–PCR, reverse transcription–polymerase chain reaction; PBS, phosphate-buffered saline; mRNA, messenger RNA; RT, room temperature; His-tag, histidine tag; HRP, horseradish peroxidase; TMB, 3,30 ,5,50 tetramethylbenzidine; OD, optical density; CI–ELISA, competitive indirect ELISA; UV, ultraviolet; ARM, antibody–ribosome–mRNA. 1

0003-2697/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2011.10.044

Therefore, there is a growing demand for more economical, rapid, and portable methods for detection of OP pesticide residues [5–7]. Immunoassays such as enzyme-linked immunosorbent assay (ELISA) have the potential to become an alternative or complementary method because they have proven to be sensitive, fast, and costeffective tools for determining trace amounts of chemicals such as pesticides. Indeed, immunosensor technology is already an important tool for the detection of pharmaceuticals and pesticides, and the development and improvement of immunosensors remains a vital area of research [8–10]. Immunosensors and ELISA can yield quantitative results with similar or greater sensitivity, accuracy, and precision than other analytical methods. However, these two immunoassay techniques rely largely on the availability of high-quality antibodies against target analytes. Prior to the advent of recombinant DNA technology, antibodies with specificity and affinity could be obtained only from immunized animals (polyclonal antibodies) or from the culture supernatant of hybridoma cells (monoclonal antibodies). Development of polyclonal antibodies with high affinity and specificity is time-consuming and the quality of antibodies varies between antigen and host, whereas the isolation of monoclonal antibodies requires large-scale screening strategies and expensive production methods. These problems can be solved by recombinant DNA technology. Through recombinant DNA technology and

ScFv antibodies against fenitrothion / Y. Luo, Y. Xia / Anal. Biochem. 421 (2012) 130–137

antibody engineering, the specificity and affinity of an existing or novel antibody can be achieved at the molecular level by chain shuffling, site-directed mutagenesis, or complementarity-determining region (CDR) grafting [11]. Using a variety of expression systems, recombinant antibodies can be obtained cheaply and quickly. Single-chain variable fragments (ScFvs) are recombinant antibodies composed of only the variable regions of the heavy and light chains (bound by a short linker) that maintain the specific antigen binding properties of natural antibodies. The use of ScFvs is a new strategy for developing improved immunodetection tests for OP pesticides [12]. These ScFvs can be produced by conventional hybridoma or phage display technology. Although phage display has progressed considerably faster than hybridoma technology, deficiencies still exist. First, the library size is limited by the necessary transformation step. Second, selection in the context of the host environment cannot be avoided, possibly causing loss of potential candidates due to their growth disadvantage or even toxicity to Escherichia coli. Furthermore, difficulties in eluting phages carrying antibodies with very high affinity may be encountered [13]. Ribosome display technology was developed to overcome these deficiencies. Using a cell-free transcription, translation, and selection system, ribosome display has the benefits of very large size libraries while reducing candidate loss from cell toxicity and insolubility. Ribosome display has been used for selection and evaluation of ScFvs against bacteria [14], viruses [15], hormones such as progesterone [16], various drugs [17,18], and proteins [19–21], but not against OP pesticides. Here, we describe the isolation and characterization of ScFvs against fenitrothion, one of the most widely used OP pesticides. These recombinant antibody fragments were selected from an ScFv DNA library and are expressed for detection of fenitrothion in contaminated food samples. Materials and methods Synthesis of antigen The synthesis of fenitrothion antigens [22,23] is shown schematically in Fig. 1. Synthesis of hapten Fenitrothion [O,O-dimethyl-O-(3-methyl-4-nitrophenyl) phosphorothioae] (5 g) was dissolved in 40 ml of diethyl ether. The

131

solution was mixed with 30 ml of 0.1 M aqueous Na2CO3, and the inorganic phase was discarded. This procedure was repeated two more times, followed by the addition of 40 ml of hydrochloric acid/acetic acid (1:9) and 8.0 g of zinc powder to the organic phase. The reaction mixture was stirred and heated under reflux for 50 min. The solids were filtered out from the colorless solution and washed with chloroform. The combined organic phase was then mixed with 30 ml of distilled water and dried overnight with Na2SO4. The remaining solvent was evaporated under reduced pressure to yield a brown oil containing the hapten of fenitrothion [22]. Preparation of hapten–protein conjugates The hapten of fenitrothion was covalently attached to bovine serum albumin (BSA) or ovalbumin (OVA) by diazotization [23]. The antigen with BSA (FE–BSA) was used as the immunogen, whereas the hapten–OVA conjugate (FE–OVA) was used as the coating antigen for affinity selection and ELISA. The phosphorus content was determined by the molybdenum blue method [24], and the results were used to calculate the molar ratio of hapten to protein. Immunization Six female BALB/c mice (6 weeks old) were immunized by intraperitoneal injection with a 1:1 (v/v) mixture (200 ll) of 60 lg fenitrothion–BSA (FE–BSA) in phosphate-buffered saline (containing 8 g of NaCl, 0.2 g of KCl, 1.15 g of Na2HPO4, and 0.2 g of KH2PO4 per liter water, pH 7.4) and Freund’s complete adjuvant (Sigma, USA). Then, 2 and 4 weeks after the initial injection, booster injections were given with the same amount of immunogen and Freund’s incomplete adjuvant (Sigma). Next, 2 weeks after the final booster injection, mice were tail-bled and the antisera were tested by indirect ELISA using coating antigen (FE–OVA). The spleen of the mouse with the highest antibody titer was removed, and cells were isolated for RNA extraction [14]. Construction of VH/k-chain library The RNA was isolated from spleen cells using a Total RNA Purification Kit (Promega, USA) according to the manufacturer’s instructions. The complementary DNAs (cDNAs) encoding the mouse VH and kappa chains were amplified by reverse transcription–polymerase chain reaction (RT–PCR) using the primer pairs

Fig.1. Synthetic route for antigens of fenitrothion.

132

ScFv antibodies against fenitrothion / Y. Luo, Y. Xia / Anal. Biochem. 421 (2012) 130–137

T7/back (50 GCAGCTAATACGACTCAC TATAGGAACAGACCACCATG AGGTSMARCTGC AGSAGTCWGG30 ) and VH/for (50 CGATCCGCCAC CGCCAGAGCCACCTCCGCCTGAA CC GCCTCCACC GAGGA GACGGTG ACCGTGGTCCC30 ) for amplification of cDNA encoding the VH chain and Vk/back (50 GGTGGAGGCGGTTCAGGCGGAGGTGGCTCT GGCGG TGG CGGATCGGACATTGAGCTCACCCAGTCTCCA30 ) and Ck/for (50 GCTCTAGA ACACTCATTCCTGTTGGAGCT30 ) for amplification of cDNA encoding the k-chain. The T7/back primer included a ribosome binding site and a T7 promoter (underlined), and the Ck/for primer contained no stop codon. The Ck region of the k-chain served as a spacer that tethered the protein to the ribosome and maintained proper folding. The recombinant DNA encoding the single-chain antibody library in the VH/k format was constructed by randomly linking the VH fragments and k-chains by a 45-bp linker sequence (Gly4Ser)3 (underlined in VH/for and Vk/back above). A total of 3 ng of VH and 5 ng of Vk DNA were mixed in 25 ll of PCR mixture without primers. The mixture was subjected to 8 cycles at 94 °C for 1 min and 63 °C for 4 min to assemble the DNA fragments by overlapping PCR. After adding the outer primers, VH/back and Ck/for fragments were amplified by 30 cycles of 40 s at 94 °C, 1 min at 63 °C, and 1 min at 72 °C [16]. The capacity of the library was measured by a previously described method [25,26]. In vitro transcription and translation For library expression, 2.5 lg of VH/k PCR product and 0.02 mM methionine were added to a 50-ll mixture containing 40 ll of TNT T7 Quick for PCR mixture (Promega) and incubated at 30 °C for 90 min. Then, 60 ll of digestion buffer (400 mM Tris–HCl [pH 7.5], 60 mM MgCl2, and 100 mM NaCl) containing 5 U of DNase I was added to the mixture and incubated at 30 °C for 15 min. The reaction was stopped by cooling on ice and adding 100 ll of icecold PBSMO buffer (phosphate-buffered saline [PBS] with 5 mM MgCl2 and 5% [w/v] OVA). The messenger RNA (mRNA)–protein– ribosome complexes were then added immediately to antigencoated microtiter plates for affinity selection. Affinity selection and RT–PCR A polystyrene microtiter plate coated with 100 ll of coating antigen solution (40 lg/ml FE–OVA in PBS) was incubated at 4 °C overnight. The microtiter plate was washed with PBSMT (PBS with 5 mM MgCl2 and 0.05% [v/v] Tween 20) and blocked with 1% gelatin in PBS for 2 h at room temperature (RT). After washing three times in PBSMT, the plate was incubated on ice for at least 10 min. The prepared translation products in PBSMO were added to the wells coated with the FE–OVA and incubated on ice for 1 h. After three washes with ice-cold PBSTM and two washes with ice-cold PBSM (PBS with 5 mM MgCl2), the retained ribosomal complexes were dissociated with 200 ll of dissociating buffer (PBS with 20 mM ethylenediaminetetraacetic acid [EDTA]) for 15 min on ice. The mRNA in dissociating buffer was recovered by RT–PCR with primers T7/back and Ck/for. The obtained cDNAs were used for the next round of ribosome display following in vitro transcription and translation or were cloned into E. coli DH5a for expression [14].

DNA was digested with NcoI and BamHI and then ligated with vector PPOW3.0 using the T4 DNA ligase (Promega). The ligated products were transformed into E. coli DH5a, and soluble protein was expressed from each clone. Briefly, each clone was cultured under shaking (200 rpm) with 5 ml of 2 YT medium (containing 5 g of NaCl, 16 g of tryptone, 10 g of yeast extract, and 1000 ml of distilled water) with 100 lg/ml ampicillin at 30 °C until the absorbance at 600 nm reached between 0.8 and 1.0. The culture was then continued at 42 °C at 200 rpm for 4 h. The cultures were briefly centrifuged, and pelleted cells were dissolved in TBS buffer (1.5 M NaCl and 10 mM Tris–HCl, pH 7.4) with 100 lg/ml lysozyme and were lysed by sonication. ELISA Indirect ELISA was performed routinely to determine the binding activity of the isolated clones. All incubations except antigen coating were carried out at RT. Following each incubation, the plates were washed five times with PBST (10 mM PBS containing 0.05% Tween 20, pH 7.4). Microtiter plates were coated with 100 ll of coating antigen (FE–OVA) solution (50 lg/ml in PBS) overnight at 4 °C. The plates were blocked with 1% gelatin in PBS and incubated for 1 h. Cell extracts were diluted 1:1 with PBSO (PBS containing 2% [w/v] OVA), and 100 ll of this mixture was added to the antigen-coated wells and incubated for 1 h. Next, 100 ll of horseradish peroxidase (HRP)-conjugated anti-His-tag (Invitrogen, USA) was added and incubated for 1 h. After washing in PBST, 100 ll of 3,30 ,5,50 -tetramethylbenzidine (TMB) was added. The peroxidase reaction was stopped after an appropriate time (typically 10 min) by adding 50 ll of 2 M H2SO4. The optical density (OD) measured at 405 nm on a spectrophotometer was indicative of the level of antibody binding. Specificity and affinity analysis A competitive indirect ELISA (CI–ELISA) was performed to test the specificity and affinity of purified ScFv proteins. The soluble proteins were purified with the HisLink Spin Protein Purification System (Promega). The CI–ELISA procedure was similar to that described in the ‘‘ELISA’’ section above. Briefly, after coating with recombinant antibodies and blocking with 1% gelatin in PBS, 50-ll samples of serially diluted fenitrothion, 3-methyl-4-nitrophenol, and other OP pesticides (acephate, malathion, dimethoate, chlorpyrifos, parathion-methyl, and triazophos, obtained from the National Research Center for Certified Reference Materials, China) in PBS were added, followed by 50 ll of FE–BSA, and incubated for 1 h at RT. Finally, 100 ll of anti-BSA–HRP (Abnova, USA) was added and the absorbance was read at 405 nm. The affinity of ScFvs against fenitrothion was also assessed with a Biacore X system (GE Healthcare, Sweden). The purified coating antigen FE–OVA (40 lg/ml) was immobilized on a CM5 sensor chip (GE Healthcare). Sensograms for kinetic measurements were generated by injection of 40 ll of soluble ScFvs at different concentrations in Hepes buffer (pH 7.4) at a flow rate of 10 ll/min. The chip was regenerated by injection of repeated pulses of 10 mM Gly-HCl (pH 2.5). Analysis of food samples contaminated with fenitrothion

Cloning and expression After selection, the ScFv DNA was amplified using the forward primer VH/bs (50 GACATG CCATGGCCCATCATCACCATCATCACATGAG GTSMARCTGCAGSAGTC30 ) with the NcoI restriction site (underlined) and histidine tag (His-tag, italic) and the reverse primer VL/fn (50 AGCCG GAATTCTACTTACTTAGGATCCTGCAGCATCAGCCC GTTT30 ) with the BamHI restriction site (underlined). The amplified

Analysis by ELISA Solutions of fenitrothion in MeOH were prepared at 10, 100, and 1000 ng/ml and were used to contaminate samples of cucumber or rice. Here, 1 ml of this fenitrothion solution was added to 1 g of pesticide-free cucumber or rice that had been chopped or ground into fine pieces. After incubating the food–fenitrothion mixture for 24 h, the fenitrothion was extracted from the food sample in

ScFv antibodies against fenitrothion / Y. Luo, Y. Xia / Anal. Biochem. 421 (2012) 130–137

133

Fig.2. UV spectra of the hapten of fenitrothion (FE), BSA, and antigens FE–BSA and FE–OVA.

5 ml of MeOH for 10 min with four vigorous shakes. The container and residual food matter were rinsed with 5 ml of MeOH. Both MeOH extraction solutions were filtered to remove particulate material and then combined. The MeOH was evaporated to dryness under reduced pressure, and the residue was dissolved in 10 ml of 10% MeOH in 50 mM PBS. This solution was analyzed for fenitrothion by CI–ELISA using the selected and purified ScFv. Analysis by GC Fenitrothion recovered from contaminated rice and cucumber samples was also detected by GC. The residue remaining after MeOH extraction and solvent evaporation was dissolved in 10% MeOH–water and purified by solid phase extraction using water as the washing solvent and MeOH as the eluent. The final eluate was evaporated to dryness and then dissolved in acetonitrile for GC. An HP-5 column (5% phenyl methyl siloxane, 30 m  0.32 mm) was used from sample separation. An HP model 6890 series autoinjector was used to inject 1 ll of sample in splitless mode. The temperature of the injection port of the chromatograph was maintained at 250 °C. The oven was programmed to heat from 150 to 220 °C at 9 °C/min and then from 220 to 240 °C at 2 °C/ min. Helium was used as the carrier gas. The temperature of the electron capture detector (ECD) was maintained at 300 °C. Results and discussion

For example, the peak at around 350 nm was shown in FE–BSA and FE–OVA but not in hapten or BSA. The molar ratio of hapten to protein was 40:1 for FE–BSA and 22:1 for FE–OVA (primary data not shown). Immunization of mice Six female BALB/c mice were immunized with FE-BSA. The blood of each immunized mouse was collected after the second booster injection. The antibody titers were determined by ELISAs of diluted serum samples from each mouse. The spleen of the mouse with the highest titer was isolated, and RNA was extracted to construct the ScFv cDNA library. ScFv cDNA library The amplified k-chain (contained with variable k-chain and part of constant k-chain) and VH fragments (contained with variable heavy chain) were of the expected size (700 and 400 bp; Fig. 3A). The VH/k chain (1.1 kb; Fig. 3B) assembled from the VH and k-chain fragments was purified, cloned into the PMD18-T expression vector, and used to transform JM109 competent cells for amplification and sequence analysis. The VH/k-chain genes with T7/back, linker, and Ck/for belonged to antibody genes of mouse. The amount of the purified VH/k-chain fragment was 2.5 lg, and the capacity of the library was 2.1  1012.

Antigen High-affinity ScFvs The OP pesticides cannot be used as immunogens due to their low molecular weight and so must be conjugated to proteins through active groups (–OH, –NH2, and –COOH). Fenitrothion was conjugated to OVA and BSA through a diazo intermediate. The nitro group of fenitrothion was hydrogenized to an amino group, and the diazo bond between the two amino groups (one in the fenitrothion derivative and the other in the protein) produced the conjugate. The ultraviolet (UV)–visible spectra of the unreacted carrier protein and hapten conjugate were clearly distinct. The production of new chemical bonds was shown in UV–visible spectra (Fig. 2).

Individual ScFv species from the cDNA library were selected through three rounds of ribosomal display. The DNA fragments without 30 stop codons were used to produce the antibody–ribosome–mRNA (ARM) complexes using the rabbit reticulocyte lysate in vitro translation system (TNT T7 Quick for PCR, Promega). The ARM complexes were added to microtiter plates coated with FE– OVA and incubated. After washing, the retained ARM complexes were released by dissociation buffer and the mRNA was reverse transcribed by RT–PCR. Every round of ribosome display was performed under the same conditions (the spanning time and the

134

ScFv antibodies against fenitrothion / Y. Luo, Y. Xia / Anal. Biochem. 421 (2012) 130–137

Fig.3. Agarose gel electrophoresis of amplified k-chain, VH, and assembled VH/k fragments. (A) Lane M: DNA molecular weight marker; lane 1: k-chain DNA fragments; lane 2: VH chain DNA fragments. (B) Lane M: DNA marker; lanes 1–3: 1.1-kb assembled VH/k DNA fragments.

Fig.4. ELISA analysis of unselected clones (A) and clones selected after the third round of ribosome display (B). Each clone was expressed in E. coli to produce soluble ScFvs, and the binding activity of each ScFv was determined by ELISA (in triplicate). The error bars represent the standard deviations.

concentration of FE–OVA). For efficient RT–PCR, a high-performance reverse transcription enzyme (SuperScript II, Invitrogen) was used because only small amounts of mRNA are recovered by ribosome display. There are several available systems for in vitro transcription/ translation, including wheat germ, E. coli S30 extracts, and rabbit reticulocyte [27]. There are also a number of rabbit reticulocyte systems, including dithiothreitol (DTT)-deficient coupled transcription/translation for PCR, nuclease-treated, and non-nucleasetreated. TNT T7 Quick for PCR was chosen because it does not require a separate mRNA isolation step [28]. Clone analysis For soluble ScFv expression, the DNA outputs from the unselected antibody library and the antibody library after the third selection were cloned into express vector PPOW3.0. After transformation, 100 clones from the two libraries were picked randomly and their soluble proteins were expressed. The cell extracts from these clones were tested by indirect ELISA with the coating antigen FE–OVA. All tested clones from the unselected library showed little conjugation activity to FE–OVA, but approximately 25% of the iso-

Fig.5. Competitive ELISAs for fenitrothion detection using three ScFvs selected from a ribosome display library. CI–ELISAs were constructed as described in Materials and methods using the purified ScFv–AF50, ScFv–AF93, and ScFv–AF132. The values B0 and B are corrected absorbances in the absence and presence of fenitrothion, respectively. The error bars represent the standard deviations calculated from three replicate calibration curves obtained with the same set of standards.

lated clones from the selected library showed high binding activity (Fig. 4A and B). Furthermore, none of the clones exhibited significant binding activity to (unconjugated) OVA, indicating that the recombinant antibodies against fenitrothion were enriched by ribosome display. Approximately 160 clones from the cDNA library after the third selection were analyzed by ELISA. Of these, 40 clones reacted positively with fenitrothion.

ScFvs with high affinity and specificity After screening, three clones (ScFv–AF50, ScFv–AF93, and ScFv– AF132) were chosen for further study. The affinity and specificity of

135

ScFv antibodies against fenitrothion / Y. Luo, Y. Xia / Anal. Biochem. 421 (2012) 130–137 Table 1 Cross-reactivity of ScFv antibodies for compounds related to fenitrothion. Compound

a

Structure

Cross-reactivity (%)a AF50

AF93

AF132

Serum

Fenitrothion

100

100

100

100

Parathion-methyl

2.4

2.7

2.8

23

Acephate

0.082

0.093

0.090

0.77

Malathion

0.076

0.072

0.057

2.3

Dimethoate

0.033

0.024

0.017

3.1

Triazophos

<0.01

<0.01

<0.01

1.5

Chlorpyrifos

<0.01

<0.01

<0.01

3.8

Cross-reactivity (%) = (IC50 of fenitrothion/IC50 of other compounds)  100.

the three ScFvs were tested by CI–ELISA. Fenitrothion in PBS (104– 105 ng/ml) was detected by CI–ELISA using ScFv–AF50, ScFv–AF93, or ScFv–AF132 (Fig. 5). The IC50 values for fenitrothion with ScFv– AF50, ScFv–AF93, and ScFv–AF132 were 1.6, 3.4, and 2.2 ng/ml, respectively, with detection limits of 0.02, 0.07, and 0.01 ng/ml (10% inhibition [29]). The dynamic range of ScFv–AF50 was 0.1 to 78 ng/ml, and the dynamic ranges of ScFv–AF93 and ScFv–AF132 were very similar to ScFv–AF50’s dynamic range. The dynamic ranges were also similar to the previous ELISAs based on monoclonal antibodies for the detection of pesticides [29,30]. The cross-reactivity of the three ScFvs was determined by measuring IC50 values against other OP insecticides (Table 1). Parathion-methyl showed some cross-reactivity in the immunoassay, most likely because the structures of fenitrothion and parathionmethyl differ by only a single methyl group in the phenyl ring. For all other pesticides tested, cross-reactivities were below 0.1%. The negligible cross-reactivity indicated that the soluble ScFv fragment antibodies of the three positive clones had high specificity to hapten of fenitrothion. The low cross-reactivity of ScFv fragment antibodies with the serum from immunized mice (Table 1) indicated that the three ScFvs had low nonspecific protein binding, confirming that the comparative ELISAs selected highly specific ScFvs against fenitrothion. The cross-reactivities were lower than the previous ELISAs for the determination of fenitrothion with polyclonal [31,32] or monoclonal [33,34] antibodies. Real-time kinetic analysis by Biacore was used to determine the affinity of the three ScFvs to the antigen. The association rate of ScFv–AF50 was faster than those of ScFv–AF93 and ScFv–AF132, and the dissociation rates of ScFv–AF50 and ScFv–AF132 were slower than that of ScFv–AF93 (Fig. 6). The equilibrium dissociation constants (KD values) determined by Biacore analysis for ScFv–AF50, ScFv–AF93, and ScFv–AF132 were 4.56  1010, 1.42  109, and 2.66  1010 M, respectively (Table 2).

In this study, IC50 values and detection limits of the three positive clones were lower than those achieved previously with polyclonal antibodies against fenitrothion [31,32], monoclonal antibodies against fenitrothion [33,34], immunochromatographic assays against chlorpyrifos-methyl [6], a single-domain antibody against picloram [35], and a methamidophos-specific ScFv [12]. Moreover, the KD values were lower than the ScFv antibodies against methamidophos isolated by phage display [12]. Ribosome display may be superior to phage display for the isolation of high-affinity and specific antibodies due to the larger capacity of ribosome display libraries. Indeed, the capacity of the ScFv library was 2.1  1012, several orders of magnitude greater than the capacity of phage display libraries (107–109). Recovery studies Rice and cucumber were spiked with fenitrothion and analyzed by both CI–ELISA and GC. The ScFv–AF132 clone was used in the CI–ELISA because it possessed the highest affinity (Table 2). Recoveries of fenitrothion from food samples ranged from 80.6% to 108% for CI–ELISA and from 64% to 92% for GC (Table 3). Both methods demonstrated acceptable sensitivity, but the ELISA was superior, especially when the extraction solution contained fenitrothion concentrations near the IC50 of AF132. One possible reason might be that the extraction of ELISA was potentially much simpler than that of GC. Highly efficient recovery of fenitrothion from contaminated rice and cucumber samples indicated that there was no significant matrix effect, at least for food matter. In Europe, the acceptable limits for fenitrothion residues in food were 0.01 to 0.5 mg/kg (EC 149/2008), and the dynamic range in our study was 0.1 to 78 ng/ml. These results demonstrate that soluble ScFv fragments from these positive clones possess the sensitivity and selectivity

136

ScFv antibodies against fenitrothion / Y. Luo, Y. Xia / Anal. Biochem. 421 (2012) 130–137

for CI–ELISA-based detection of fenitrothion residues in environmental and agricultural samples. Sequence analysis

Fig.6. Sensogram demonstrating the association and disassociation of ScFv–AF50, ScFv–AF93, and ScFv–AF132.

Table 2 Kinetic constants determined by Biacore analysis for ScFvs. ScFv

ka (M1 s1)

kd (s1)

kd/ka (KD, M)

AF50 AF93 AF132

7.28  106 6.05  106 3.29  106

3.32  103 8.61  103 8.76  104

4.56  1010 1.42  109 2.66  1010

Note: ka, association rate constant; kd, dissociation rate constant; KD, equilibrium dissociation constant.

Table 3 Recovery of fenitrothion from contaminated food samples using CI–ELISA and GC.

a

Sample

Theoretical (ng/ml)

Recovery (%)a CI–ELISA

GC

Rice

10 100 1000

105 97.2 81.4

64 66 91

Cucumber

10 100 1000

108 95.3 80.6

78 67 92

Data from ELISAs are the means of triplicate measurements, whereas GC results are the means of duplicate measurements.

The amplified DNA fragments coding for the ScFv–AF50, ScFv– AF93, and ScFv–AF132 were characterized by sequence analysis. As shown in Fig. 7, all three sequences had different amino acid substitutions, and (as expected) most of these were in the CDRs. The VH and VL gene families of the ScFvs were designated based on the international ImMunoGeneTics (IMGT) database. The light chains of all three clones belonged to the same Vk 4 family. The heavy chains of ScFv–AF50 and ScFv–AF132 belonged to the VH4 gene family, whereas the ScFv–AF93 heavy chain belonged to the VH1 gene family. Sequencing alignment by the basic local alignment search tool (BLAST) demonstrated that the VH of the clones ScFv–AF50 and ScFv–AF93 had very similar sequences (97% homology) and that the VL of ScFv–AF50 and ScFv–AF132 shared 89% homology [12]. This is the first study to report the screening and isolation of specific ScFv antibodies to an organophosphate insecticide from a ribosome display library. These results demonstrate the power of ribosome display technology for antibody selection. Similar to other display technologies, the diversity and capacity of the cDNA library is paramount to the success of the selection experiments. The recombinant ScFv fragments acquired in this study may also prove to be invaluable for developing rapid and affordable immunoassays and immunosensors for quantification of fenitrothion in environmental and biological samples. The advantage of using recombinant antibodies such as ScFvs in immunosensors is that the recognition element can be produced inexpensively in a ‘‘ready to use’’ form. Moreover, high-affinity and specific ScFvs could enhance the sensitivity and accuracy of the immunosensors if they can be mass produced. In the next study, the stability of these three ScFvs will be improved by site-directed mutagenesis and by the addition of a stabilizer. The structure of these and other ScFvs should be analyzed to elucidate structure–binding relationships and as an aid to develop better antibodies against these pesticides. Using this same strategy, ScFvs against OP pesticides and other small molecule toxins can also be developed. Conclusions We have successfully obtained high-affinity recombinant antibodies against fenitrothion through ribosome display of an ScFv cDNA library derived from an immunized mouse. These antibodies

Fig.7. Alignment of the amino acid sequence of selected fenitrothion-specific ScFvs. The regions of CDR1 to CDR3 were deduced according to international ImMunoGeneTics (IMGT) database.

ScFv antibodies against fenitrothion / Y. Luo, Y. Xia / Anal. Biochem. 421 (2012) 130–137

appear to have binding properties superior to previously described polyclonal and monoclonal antibodies against fenitrothion and recombinant antibodies against other pesticides [12,34,35]. The IC50 values for fenitrothion with ScFv–AF50, ScFv–AF93, and ScFv–AF132 were 1.6, 3.4, and 2.2 ng/ml with detection limits of 0.02, 0.07, and 0.01 ng/ml, respectively. Since ribosome display technology was first reported in 1994 [36], several published reports have demonstrated antibody selection from a cDNA library [14–21,32,37]. However, this is the first example of the use of ribosome display for the selection of ScFvs against OP pesticides and for subsequent ELISA-based detection of fenitrothion in contaminated food samples. Overall, the ELISAs developed in this study can accurately determine fenitrothion in food samples after the simple and rapid extraction procedure. Acknowledgment This work was supported by the Ph.D. Programs Foundation of the Ministry of Education of China (20090191110031). References [1] V. Pardo-Yissar, E. Katz, J. Wasserman, I. Willner, Acetylcholine esteraselabeled CdS nanoparticles on electrodes: photoelectrochemical sensing of the enzyme inhibitors, J. Am. Chem. Soc. 125 (2003) 622–623. [2] Y.R. Tahboub, M.F. Zaater, Z.A. Al-Talla, Determination of the limits of identification and quantitation of selected organochlorine and organophosphorous pesticide residues in surface water by full-scan gas chromatography/mass spectrometry, J. Chromatogr. A 1098 (2005) 150–155. [3] C. Padrin-Sanz, R. Halko, Z. Sosa-Ferreraa, J.J. Santana-Rodriguez, Combination of microwave assisted micellar extraction and liquid chromatography for the determination of organophosphorous pesticides in soil samples, J. Chromatogr. A 1078 (2005) 13–21. [4] R.C. Martinez, E.R. Gonzalo, F.G. Garcia, J.H. Mendez, Automated highperformance liquid chromatographic method for the determination of organophosphorus pesticides in waters with dual electrochemical (reductive–oxidative) detection, J. Chromatogr. A 644 (1993) 49–58. [5] J.P.M. Wang, G. Collins, A. Mulchandani, Y. Liu, K. Olsen, Single-channel microchip for fast screening and detailed identification of nitroaromatic explosives or organophosphate nerve agents, J. Anal. Chem. 74 (2002) 1187– 1191. [6] X. Hua, G. Qian, J. Yang, B. Hu, J. Fan, N. Qin, G. Li, Y. Wang, F. Liu, Development of an immunochromatographic assay for the rapid detection of chlorpyrifosmethyl in water samples, Biosens. Bioelectron. 26 (2010) 189–194. [7] X. Jiang, D. Li, X. Xu, Y. Ying, Y. Li, Z. Ye, J. Wang, Immunosensors for detection of pesticide residues, Biosens. Bioelectron. 23 (2008) 1577–1587. [8] V.S. Morozova, A.I. Levashova, S.A. Eremin, Determination of pesticides by enzyme immunoassay, J. Anal. Chem. 60 (2005) 202–217. [9] W.J. Gui, Y.H. Liu, C.M. Wang, X. Liang, G.N. Zhu, Development of a direct competitive enzyme-linked immunosorbent assay for parathion residue in food samples, Anal. Biochem. 393 (2009) 88–94. [10] Y.H. Lin, G.D. Liu, Biosensor based on self-assembling acetylcholinesterase on carbon nanotubes for flow injection/amperometric detection of organophosphate pesticides and nerve agents, J. Anal. Chem. 78 (2006) 835–843. [11] K. Kramer, B. Hock, Recombinant antibodies for environmental analysis, Anal. Bioanal. Chem. 377 (2003) 417–426. [12] T. Li, Q. Zhang, Y. Liu, D. Chen, B. Hu, D.A. Blake, F. Liu, Production of recombinant ScFv antibodies against methamidophos from a phage-display library of a hyperimmunized mouse, J. Agric. Food Chem. 54 (2006) 9085–9091. [13] R. Schier, J.D. Marks, Efficient in vitro affinity maturation of phage antibodies using Biacore guided selections, Hum. Antibodies Hybridomas 7 (1996) 97–105. [14] Q. Yuan, Y.X. Xia, S.J. Nian, Y.P. Yin, Y.Q. Cao, Z.K. Wang, Selection of single chain fragments against the phytopathogen Xanthomonas axonopodis pv. citri by ribosome display, Enzyme Microb. Technol. 41 (2007) 383–389.

137

[15] X.L. Zhao, W.Q. Chen, Z.H. Yang, J.M. Li, S.J. Zhang, L.F. Tian, Selection and affinity maturation of human antibodies against rabies virus from a scFv gene library using ribosome display, J. Biotechnol. 144 (2009) 253–258. [16] M. He, M. Menges, M.A. Groves, E. Corps, H. Liu, M. Bruggemann, M.J. Taussig, Selection of a human anti-progesterone antibody fragment from a transgenic mouse library by ARM ribosome display, J. Immunol. Methods 231 (1999) 105–117. [17] X.H. Yan, Z.R. Xu, Production of human single-chain variable fragment (scFv) antibody specific for digoxin by ribosome display, Indian J. Biochem. Biophys. 42 (2005) 350–357. [18] Y. Qi, C. Wu, S. Zhang, Z. Wang, S. Huang, L. Dai, S. Wang, L. Xia, K. Wen, X. Cao, Y. Wu, J. Shen, Selection of anti-sulfadimidine specific ScFvs from a hybridoma cell by eukaryotic ribosome display, PLoS One 4 (2009) e6427. [19] A. Rothe, A. Nathanielsz, R.J. Hosse, F. Oberhauser, E. Strandmann, E.P. Engert, P.J. Hudson, B.E. Power, Selection of human anti-CD28 scFvs from a T-NHL related scFv library using ribosome display, J. Biotechnol. 130 (2007) 448–454. [20] F. Li, P. Su, C. Lin, H. Li, J. Cheng, J.D. Shi, Ribosome display and selection of human anti-placental growth factor scFv derived from ovarian cancer patients, Protein Pept. Lett. 17 (2010) 585–590. [21] M.S. Lee, M.H. Kwon, K.H. Kim, H.J. Shin, S. Park, H.I. Kim, Selection of scFvs specific for HBV DNA polymerase using ribosome display, J. Immunol. Methods 284 (2004) 147–157. [22] Y.H. Liu, M.J. Jin, W.J. Gui, J.L. Cheng, Y.R. Guo, G.N. Zhu, Hapten design and indirect competitive immunoassay for parathion determination: correlation with molecular modeling and principal component analysis, Anal. Chim. Acta 591 (2007) 173–182. [23] Y. Liang, X.J. Liu, Y. Liu, X.Y. Yu, M.T. Fan, Synthesis of three haptens for the class-specific immunoassay of O,O-dimethyl organophosphorus pesticides and effect of hapten heterology on immunoassay sensitivity, Anal. Chim. Acta 615 (2008) 174–183. [24] M.Y. He, M.J. Taussig, Ribosome display of antibodies: expression, specificity, and recovery in a eukaryotic system, J. Immunol. Methods 297 (2005) 73–82. [25] T. Lamla, V.A. Erdmann, Searching sequence space for high-affinity binding peptides using ribosome display, J. Mol. Biol. 329 (2003) 381–388. [26] M.Y. He, N. Cooley, A. Jackson, M.J. Taussig, Production of human single-chain antibodies by ribosome display, Methods Mol. Biol. 248 (2004) 177–189. [27] M.J. Tymms, P.J. Hertzog, Structure–function studies based on in vitro expression, Methods Mol. Biol. 37 (1995) 317–335. [28] P. Shepherd, C. Dean, Monoclonal Antibodies, Oxford University Press, New York, 2000. [29] W.Y. Lee, E.K. Lee, Y.J. Kim, W.C. Park, T. Chung, Y.T. Lee, Monoclonal antibodybased enzyme-linked immunosorbent assays for the detection of organophosphorus insecticide isofenphos, Anal. Chim. Acta 557 (2006) 169–178. [30] E. Watanabe, K. Baba, S. Miyake, Analytical evaluation of enzyme-linked immunosorbent assay for neonicotinoid dinotefuran for potential application to quick and simple screening method in rice samples, Talanta 84 (2011) 1107–1111. [31] E. Watanabe, Y. Kanzaki, H. Tokumoto, R. Hoshino, H. Kubo, H. Nakazawa, Enzyme-linked immunosorbent assay based on a polyclonal antibody for the detection of the insecticide fenitrothion: evaluation of antiserum and application to the analysis of water samples, J. Agric. Food Chem. 50 (2002) 53–58. [32] Y.A. Cho, J.A. Seok, H.S. Lee, Y.J. Kim, Y.C. Park, Y.T. Lee, Synthesis of haptens of organophosphorus pesticides and development of immunoassays for fenitrothion, Anal. Chim. Acta 522 (2004) 215–222. [33] Y.J. Kim, Y.A. Kim, Y.T. Lee, H.S. Lee, Enzyme-linked immunosorbent assays for the insecticide fenitrothion: influence of hapten conformation and sample matrix on assay performance, Anal. Chim. Acta 591 (2007) 183–190. [34] D.P. McAdam, A.S. Hill, H.L. Beasley, J.H. Skrritt, Mono- and polyclonal antibodies to the organophosphate fenitrothion: 1. Approaches to hapten– protein conjugation, J. Agric. Food Chem. 40 (1992) 1466–1470. [35] K.Y. Yau, M.A. Groves, S. Li, C. Sheedy, H. Lee, J. Tanha, C.R. MacKenzie, L. Jermutus, J.C. Hall, Selection of hapten-specific single-domain antibodies from a non-immunized llama ribosome display library, J. Immunol. Methods 281 (2003) 161–175. [36] L.C. Mattheakis, R.R. Bhatt, W.J. Dower, An in vitro polysome display system for identifying ligands from very large peptide libraries, Proc. Natl. Acad. Sci. USA 91 (1994) 9022–9026. [37] H. Yanagida, T. Matsuura, T. Yomo, Ribosome display for rapid protein evolution by consecutive rounds of mutation and selection, Methods Mol. Biol. 634 (2010) 257–267.