Anal. Chem. 2008, 80, 3205-3212
Development of a High-Affinity Anti-Domoic Acid Sheep scFv and its Use in Detection of the Toxin in Shellfish Iain Shaw,* Aoife O’Reilly, Margaret Charleton, and Marian Kane
National Diagnostics Centre, National University of Ireland Galway, Galway, Ireland
The potential of immunoassays as high-throughput screening tools for the detection of harmful substances in foods will only be realized when convenient methods are available for production of the high affinity antibodies needed for sensitive assay development. Recombinant antibodies offer advantages over traditional monoclonal antibodies in terms of ease of production, much greater antibody repertoire for selection, and versatility. We describe here the development of recombinant antibodies against the common shellfish toxin, domoic acid (DA), utilizing the sheep immunoglobulin system as an effective method for generating high affinity anti-hapten recombinant antibody fragments. A single-chain antibody fragment (scFv) library was generated from a sheep immunized with DA-bovine serum albumin conjugate, and anti-DA scFvs were isolated by phage-display. Three selected scFvs gave I50s of 2.6 to 58 ng/mL (8.3-186 nM) in competitive enzymelinked immunosorbent assay (ELISA). Assay optimization with one of these scFvs gave a very reproducible standard curve with a range of 0.3 to 5.6 ng/mL (1.0 to 17.9 nM), a mean limit of quantification (LOQ, defined as the I20) of 0.5 ng/mL (1.6 nM), and a mean I50 of 1.2 ng/mL (3.9 nM). When the assay was used for the analysis of crude methanolic extracts of scallop tissues, results obtained correlated well with standard HPLC assay results (R2, 0.90, n ) 40; R2, 0.81, n ) 34), although ELISA results were lower than HPLC results. Adjusting the cutoff point for DA concentration accordingly from the regulatory 20 mg/kg, the potential of the sheep scFv-based ELISA for use as a screening assay for DA in shellfish extracts was demonstrated. The accurate quantification of low levels of small complex molecules, such as drug residues, pesticides, mycotoxins, and the algal toxins, in environmental and food samples is critical for consumer protection. These compounds are typically measured using HPLC with detection by spectrophotometry, fluorimetry, or mass spectrometry.1 However, HPLC methods are often timeconsuming and expensive and therefore not always suited to highthroughput screening.2 In these situations, immunoassays are an * To whom correspondence should be addressed. Phone: +353 91 492090. Fax: +353 91 586570. E-mail:
[email protected]. (1) Lawrence, J. F.; Cleroux, C.; Truelove, J. F. J. Chromatogr. 1994, 662, 173. 10.1021/ac7024199 CCC: $40.75 Published on Web 04/03/2008
© 2008 American Chemical Society
attractive approach for rapid, sensitive, and highly specific screening of large sample numbers.3 Immunoassay techniques generally rely on murine monoclonal antibodies or hyperimmune polyclonal sera from rabbits, sheep, and other mammals. Polyclonal sera may be produced in relatively large quantities but can suffer from batch to batch variation which is problematic for immunoassay kit development. This is particularly so in the case of the more popular rapid immunoassay formats, such as lateral flow devices, or for immunoaffinity applications, where higher quantities of antibodies are needed. Monoclonal antibodies, by contrast, are more consistent and are more easily standardized. However, high affinity monoclonals are needed for development of sensitive immunoassays for small hapten molecules, and it is not always easy to generate a sufficiently high affinity immune response in mice or rats against such molecules, despite extensive efforts to create optimized immunization protocols and screening methods.3 The limitations of traditional techniques have led to the investigation and use of recombinant antibody technology to produce anti-hapten antibodies for assay development. The phagedisplay of recombinant antibody libraries is a robust ‘monoclonal’ antibody technology that is becoming increasingly attractive, as it allows the screening of antibody repertoires from any species in which the immunoglobulin DNA sequences have been characterized.4 Single-chain antibody variable region fragments (scFvs) are the most commonly used recombinant antibody format as they can be rapidly constructed, are typically well expressed in E. coli, and can exhibit high affinity and stability.5 Several scFvs have been isolated from phage-displayed libraries that are suitable for use in immunoassays for haptens. These libraries have been derived predominantly from immunized animals, including mice, rabbits, and chicken.6-8 A small number were generated from naı¨ve or synthetic libraries.9,10 (2) Hall, J. C.; Van Deynze, T. D.; Struger, J.; Chan, C. H. J. Environ. Sci. Health, Part B 1993, 28, 577. (3) Yau, K. Y. F.; Lee, H.; Hall, J. C. Biotechnol. Adv. 2003, 21, 599. (4) Maynard, J.; Georgiou, G. Ann. Rev. Biomed. Eng. 2000, 2, 339. (5) Barbas, C. F., 3rd.; Burton, D. R.; Scott, J. K.; Silverman, G. J. Phage display: a laboratory manual; Cold Spring Harbor Laboratory Press: NY, 2001. (6) Tout, N. L.; Yau, K. Y. F.; Trevors, J. T.; Lee, H.; Hall, J. C. J Agric. Food Chem. 2001, 49, 3628. (7) Li, Y.; Cockburn, W.; Kilpatrick, J. B.; Whitelam, G. C. Biochem. Biophys. Res. Commun. 2000, 268, 398. (8) Finlay, W. J.; Shaw, I.; Reilly, J. P.; Kane, M. Appl. Environ. Microbiol. 2006, 72, 3343.
Analytical Chemistry, Vol. 80, No. 9, May 1, 2008 3205
Figure 1. Chemical structure of domoic acid.
There has been a single report of the generation of a high affinity anti-hapten scFv from an immunized sheep.11 Sheep are well recognized to be particularly good at producing high affinity antiserum against haptens leading to the frequent use of sheep polyclonal antiserum in commercial hapten immunoassay kits. Satisfying the current trend toward rapid on-site or point-of-care immunoassay test systems would be simplified if a continuous and consistent supply of these antibodies were available. One field where there is a growing demand for rapid methods is in the detection of algal toxins in shellfish. Domoic acid (DA, Figure 1) is a potent neuroexcitatory toxin produced by the diatom Pseudonitszchia sp. that causes amnesic shellfish poisoning in humans through the ingestion of contaminated shellfish. In our efforts to produce a sensitive screening assay for detection of DA in shellfish extracts or seawater samples, we have therefore focused on the development of a suitable anti-DA sheep scFv. We describe here the generation of a high affinity anti-DA sheep scFv for the first time and its application for the determination of DA levels in shellfish extracts. This study demonstrates the usefulness of this approach for production of reliable, high affinity binding agents for hapten immunoassay kit development. MATERIALS AND METHODS Preparation of Conjugates. DA was conjugated to bovine serum albumin (BSA) and ovalbumin (OVA) as described by Kawatsu et al.12 Briefly, 1 mg of DA (Merck, Darmstadt, Germany) was diluted in 50 µL of DMSO and activated by the addition of 20 µL of EDC (1-ethyl-3-dimethylaminopropyl carbonimide, 25 mg/ mL in DMSO) and 30 µL of NHS (N-hydroxysuccinimide, 30 mg/ mL in DMSO). This mixture was stirred for 90 min at room temperature, after which 250 µL of OVA or BSA (both 20 mg/ mL in 0.085 M borate buffer) was added dropwise, and the mixture was stirred for a further 90 min. The conjugates were then purified over a PD-10 column (Amersham Biosciences, Uppsala, Sweden) and dialyzed twice against phosphate buffered saline (PBS). Immunizations and Serum Characterization. Three 4-yearold ewes were immunized subcutaneously with 1 mg of DA-BSA in Freund’s complete adjuvant. This was followed 6 months later with a booster dose of 500 µg of DA-BSA in incomplete Freunds adjuvant. Finally, 6 months after the first boost, 3 further boosts of 100 µg/mL of DA-BSA in incomplete Freunds adjuvant were (9) Moghaddam, A.; Lebersli, I.; Gebhardt, K.; Braunagel, M.; Marvik, O. J. J. Immunol. Methods 2001, 254, 169. (10) Strachan, G.; McElhiney, J.; Drever, M. R.; McIntosh, F.; Lawton, L. A.; Porter, A. J. R. FEMS Microbiol. Lett. 2002, 210, 257. (11) Charlton, K.; Harris, W. J.; Porter, A. J. Biosens. Bioelectron. 2001, 16, 639. (12) Kawatsu, K.; Hamano, Y.; Noguchi, T. Toxicon 1999, 37, 1579.
3206
Analytical Chemistry, Vol. 80, No. 9, May 1, 2008
administered at two-weekly intervals. Prior to the start of the immunization program and 7 days after each boost, serum was collected from the jugular vein of each animal to determine the polyclonal antibody response by enzyme-linked immunosorbent assay (ELISA) on microtiter plate. All animal work was carried out under license from the Irish Department of Health. Wells of a 96 well microtiter plate (Maxisorp, Nunc) were coated overnight at 4 °C with 100 µL/well DA-OVA at 250 ng/ mL, washed 4 times with 0.15 M NaCl/0.05% Tween 20 in a SLT PW96 plate washer, and blocked for 1 h at 37 °C with 200 µL/ well of 3% skimmed milk powder (Marvel) in PBS. Serum samples were serially diluted in 3% skimmed milk powder and added to the coated wells (100 µL/well), and plates were incubated at 37 °C for 1 h. Wells were washed as before, and bound antibody was detected by reacting with HRP-conjugated anti-sheep IgG antibody (DAKO) in PBS/0.1% BSA for 1 h at 37 °C. After washing, bound HRP was visualized with 100 µL/well of tetramethylbenzidine (TMB, DAKO) for 15 min followed by 100 µL/well of 1 M H2SO4. All analyses were performed in triplicate. The titer was taken as the dilution of serum resulting in absorbance 1.0 determined at 450 nm using a SLT spectra microplate reader. The specificity and sensitivity of the immune response for DA was determined in a competition ELISA. To coated wells prepared as described above were added 20 µL of serially diluted DA standards (100 µg/mL to 100 pg/mL in PBS/0.1% BSA; stock 1 mg/mL in PBS) followed immediately by 100 µL of diluted serum. Plates were incubated at 37 °C for 1 h and assay completed as described above. Percentage binding of serum (% B/Bo), determined by dividing the OD450 for each DA concentration by the OD450 with no DA added, was plotted against the DA concentration. Isolation of Antibody-Encoding RNA from Sheep Spleen. Seven days after the final boost, the sheep was killed by barbiturate overdose. The whole spleen was removed, and approximately one-third (300 g) was used for isolation of total RNA. Spleen tissue was passed through a fine mesh sieve (approximately 300 µm opening size) to produce a near singlecell suspension, and the cells were then pelleted at 3000g for 30 min at 4 °C. Each cell pellet was resuspended in 20 mL of TRI reagent (Sigma) and homogenized using an Ultra-Turrax (IKA) homogenizer for 1 min at room temperature. A further 20 mL of TRI reagent was added to the homogenate before pelleting the cellular debris at 2500g for 10 min at 4 °C. The supernatant was transferred to a clean tube, 12 mL of chloroform was added, and the mixture was vortexed and incubated at room temperature for 10 min, after which it was centrifuged at 10 000g for 15 min at 4 °C. Total RNA was isolated from the upper aqueous phase by the addition of 8 mL of isopropyl alcohol, vortexing, incubation at room temperature for 10 min, and recovery of precipitated RNA by centrifugation at 10 000g for 30 min at 4 °C. The pellet was washed with 70% ethanol and centrifuged as before. The pellet was airdried and resuspended in 500 µL of RNase- and DNase-free water (Sigma). The quantity and quality of the extracted RNA was determined spectrophotometrically by measuring absorbance at 260 and 280 nm.13 Amplification and Assembly of scFv Fragments. First strand cDNA was amplified directly from the total RNA using 20 (13) Warburg, O.; Christian, W. Biochem. Z. 1942, 310, 384.
Table 1. Sequences of Primers Used sheep heavy chain primers OVVH1BACK OVVH2BACK OVJH1FOR OVJH2FOR OVJH3FOR OVJH4FOR
5′-AGGTKCRRCTGCAGGRGTCGGG-3′ 5′-AGGTKCAGYTKCAGGAGTCGGG-3′ 5′-CTCAGAGCCAGAGCCAGAAGATTTACCTTCTGAGGAGACGGTGACCAGGAGTCC-3′ 5′-CTCAGAGCCAGAGCCAGAAGATTTACCTTCTGAGGAGRCGGWGAYYAGKAGTCC-3′ 5′-CTCAGAGCCAGAGCCAGAAGATTTACCTTCTGAGGAGAYRGTRASCAGGASTCC-3′ 5′-CTCAGAGCCAGAGCCAGAAGATTTACCTTCTGAAAGAACGCTGATCAGGAG-3′
OVVL1BACK OVVL2BACK OVVL3BACK OVVL4BACK OVVL5BACK OVJL1FORSFI OVJL2FORSFI
5′-GGCTCTGGCTCTGAGTCTAAAGTGGATGACCAGGCTGTGCTGACTCAGCCG-3′ 5′-GGCTCTGGCTCTGAGTCTAAAGTGGATGACCARGCTGTGCTGACYCARCYG-3′ 5′-GGCTCTGGCTCTGAGTCTAAAGTGGATGACCAGGCYSTGSTGACTCAGCCR-3′ 5′-GGCTCTGGCTCTGAGTCTAAAGTGGATGACMRGGTCRTGCKGACTCARCCG-3′ 5′-GGCTCTGGCTCTGAGTCTAAAGTGGATGACCAGKCTGYSCTGACTCAGCCK-3′ 5′-TTCTCGGGCCCCCGAGGCCCCCAGGACGGTCAGCCT-3′ 5′-TTCTCGGGCCCCCGAGGCCCCCAGGACGGTCAGYCK-3′
sheep lambda light chain primers
sheep kappa light chain primers OVVK1BACK OVVK2BACK OVJK1FORSFI
5′-GGCTCTGGCTCTGAGTCTAAAGTGGATGACGACATCCAGGTGACCCAGTCTCCA-3′ 5′-GGCTCTGGCTCTGAGTCTAAAGTGGATGACGACATCCAGCTCACCCAGTCTCCA-3′ 5′-TTCTCGGGCCCCCGAGGCCCCTTTGATTTCCACGTT-3′
ScfvFOR ScFvREV
5′-TTCTCGGGCCCCCGAGGC-3′ 5′-GCGGCCCAGCCGGCCATG-3′
overlap extension primers
sequencing primers PAK100F PAK100R
5′-CTATGACCATGATTACGAATTTC-3′ 5′-TCATCGGCATTTTCGGTCATAG-3′
µg of total RNA using the ImPromII reverse transcriptase reagent (Promega) using oligo dT18 as the synthesis primer. The regions encoding the antibody heavy and light chains (both lambda and kappa) were amplified using a modification of the procedure of Charlton et al.14 Changes were made to the previously described oligonucleotide primers and are detailed in Table 1. One significant change made to the primers was the replacement of the NotI endonuclease recognition site in the 3′ terminus of the light chain with an SfiI recognition site to allow insertion of the recombinant antibody scFv fragment into the pAK100 phage-display vector.15 Five microliters of cDNA was mixed with 25 pmol of each forward and reverse primer combination for both the heavy and the light chains, 5 µL of 10x PCR reaction buffer B (Promega), 1.6 mM MgCl2, 5 µL of 2 mM dNTPs (Promega), 1 unit Taq DNA polymerase (Promega), and DNase-free water up to a total volume of 50 µL. The reactions were carried out in an Eppendorf Mastercycler using the following conditions: an initial denaturation step of 94 °C for 2 min, followed by 30 cycles of 94 °C for 15 s, 56 °C for 15 s, and 72 °C for 1 min. A final extension of 10 min at 72 °C was employed. Amplification products were separated by electrophoresis on a 1% agarose gel in Tris-Acetate-EDTA (TAE) buffer. V region amplification products were isolated from the agarose gel using the Wizard SV gel purification kit (Promega) and quantified by determining the absorbance at 280 nm. To produce full-length scFv, equimolar amounts of heavy and light chain were linked via overlap extension PCR (OEPCR) in a (14) Charlton, K. A.; Moyle, S.; Porter, A. J. R.; Harris, W. J. J. Immunol. 2000, 164, 6221. (15) Krebber, A.; Bornhauser, S.; Burmester, J.; Honegger, A.; Willuda, J.; Bosshard, H. R.; Pluckthun, A. J. Immunol. Methods 1997, 201, 35.
reaction containing 5 µL of 10x PCR reaction buffer B (Promega), 5 µL of 2 mM dNTPs, 3 mM MgCl2, 25 pmol of each scfvFOR and scfvREV oligonucleotide primers (Table 1), 2.5 units Taq DNA polymerase (Promega), and water to 50 µL. The reactions were carried out in an Eppendorf Mastercycler using an initial denaturation step of 94 °C for 2 min, followed by 20 cycles of 94 °C for 15 s, 63 °C for 15 s, and 72 °C for 2 min with a final extension of 10 min at 72 °C. The OEPCR reaction was performed 10 times to produce sufficient scFv for subsequent cloning steps. Amplification products were separated by electrophoresis on a 1% agarose gel in TAE buffer. The products were purified from the agarose gel as previously described. Construction of Phage-Displayed scFv Library. In order to clone the fully assembled scFv into the pAK100 vector, 2 µg of scFv was digested with SfiI (Roche) in a 100 µL reaction containing 10 µL of reaction buffer M (Roche), 80 units of SfiI, and DNase free water to 100 µL. Twenty micrograms of pAK100 vector was digested at the same time using the same conditions as for the scFv. The digest was carried out at 50 °C for 4 h. Digested products were separated by electrophoresis at a maximum of 80 V on a 1% low-melting point agarose gel. The products were purified as described above. Digested scFv (700 ng) was ligated into digested pAK100 (1400 ng) in a 200 µL reaction containing 40 µL of 5x ligase buffer (Invitrogen) and 10 units of ligase (Invitrogen). The products were allowed to ligate for at least 18 h at room temperature. Ligation products were purified by the addition of 20 µL of 3 M sodium acetate, pH 5 and 440 µL of 100% ethanol followed by centrifugation for 30 min at 20 000g. The DNA pellet was washed with 200 µL of 70% ethanol, centrifuged as Analytical Chemistry, Vol. 80, No. 9, May 1, 2008
3207
Table 2. Panning Conditions
a
biopanning round
DAOVA concn (µg/mL)
1
binding time and conditions
washesa
10
120 min at room temp followed by overnight at 4 °C
5 × PBS + 0.1% Tween 5 × PBS
2
1
60 min at room temp
10 × PBS + 0.1% Tween 10 × PBS
3
0.1
30 min at room temp
20 × PBS + 0.1% Tween 20 × PBS
4
0.1
30 min at room temp
20 × PBS + 0.1% Tween 20 × PBS overnight in PBS at 4 °C
Washing of tubes was carried out by filling the tube with wash and then emptying it out immediately.
before, and allowed to air-dry before finally resuspending in 20 µL of DNase free water. The ligated scFv phagemid library was introduced into 300 µL of E. coli XL-1 Blue (Stratagene) by electroporation (2.5 kV, 25 µF, and 200 Ω.). Bacteria were allowed to recover and transcribe the chloramphenicol resistance gene for 1 h at 37 °C, with shaking at 250 rpm in 5 mL of SOC media. Ten milliliters of warm Super Broth (SB) media (Sigma) was added to the library, along with 6 µL of chloramphenicol solution (30 mg/mL) and 15 µL of 10 mg/mL tetracycline. Following 1 h incubation as before, more chloramphenicol (9 µL of the same solution) was added, and the library was incubated for a further 1 h. To perform the phage rescue of the scFv, approximately 1012 VCSM13 helper phage were added to the library, along with 185 µL of chloramphenicol solution and 185 µL of tetracycline (10 mg/mL), the volume was adjusted to 200 mL with warm SB media, and the mixture was incubated as before for 2 h. Finally, 280 µL of kanamycin solution (50 mg/mL) was added to the library, and the mixture was incubated at 37 °C with shaking at 250 rpm for 18 h. Phage were isolated from the culture media by first separating the cells by centrifugation followed by precipitation of the phage-containing culture media with PEG/NaCl.5 Precipitated phage could then be used for panning against immobilized DAOVA. Biopanning of Phage-Bound scFv. High affinity selection for DA-specific phage-bound scFv was carried out by four rounds of biopanning against immobilized DA-OVA. A Maxisorp immunotube (Nunc) was coated overnight at 4 °C with DA-OVA. The concentration of DA-OVA was dependent on the round of biopanning and is detailed in Table 2. The tube was blocked with 3% ovalbumin in PBS for 1 h at room temperature. The precipitated phage were resuspended in 1.5% OVA/1.5% BSA, added to the blocked panning tube, and allowed to bind; the phage binding time was decreased over the four rounds of selection as detailed in Table 2. After binding, the tube was washed with PBS0.1%Tween 20 and with PBS; the number of washes was increased over the four rounds of selection (Table 2). The bound phage were eluted by adding 1 mL of 100 mM glycine buffer pH 2.2 and incubating for 10 min at room temperature with rotation. The eluate was neutralized with 200 µL of 2 M Tris buffer, pH 8.8 and used to infect 2 mL of mid-log phase XL1-Blue E. coli, and the phage were rescued as previously described above. Characterization of the Polyclonal Phage Response. The level of DA-specific phage at each round of panning was deter3208
Analytical Chemistry, Vol. 80, No. 9, May 1, 2008
mined by ELISA. Plates were set up as previously described. Rescued phage from each round of panning were diluted 1 in 4 with 3% skimmed milk powder in PBS and added to triplicate wells at 100 µL/well and allowed to bind at 37 °C. Binding phage were detected by the addition of 100 µL/well of HRP-conjugated antiM13 monoclonal antibody (Amersham Pharmacia Biotech) diluted in PBS/0.1% BSA and incubated for 1 h at 37 °C. Antibody reaction was visualized with TMB as previously described. Isolation of Individual anti-DA scFv Clones. After the final round of panning, phage-infected bacteria were plated onto selective media plates and allowed to grow overnight at 37 °C. Ninety-six individual clones were picked from the selective plate into separate wells of a 96-well culture block (NUNC) and grown overnight in 400 µL of SB media supplemented with 30 µg/mL of chloramphenicol. One tenth of the overnight culture was subcultured into fresh SB supplemented with 30 µg/mL of chlormaphenicol and 2.5 × 108 pfu VCSM13 helper phage and grown for 2 h at 37 °C. Bacteria were pelleted and resuspended with SB supplemented with 30 µg/mL of chloramphenicol and 50 µg/mL of kanamycin and grown overnight at 37 °C. Cells were pelleted, and the phage containing supernatant was used in an ELISA as described above for detection of positive binding clones. Soluble Expression of scFv. To enable an efficient, high-level expression of the scFvs of interest it was necessary to transfer the scFv-encoding region of positive binding clones into the soluble expression vector, pAK400. Transfer of scFv fragment was carried out as follows. Two micrograms of pAK100 containing scFv was digested with 20 units of SfiI for 4 h. Digested fragments were separated on a 1% agarose/TAE gel until the scFv was well resolved. The band was excised and purified as previously described. The pAK400 vector was similarly digested with SfiI, and digested fragments were separated. The vector band was isolated and gel purified. Digested scFv and pAK400 were ligated together at a vector:insert ration of 1:1 using 4 units of ligase. Following overnight ligation, 5 µL of ligation product was used to transform TOP10 E. coli cells. Individual colonies were screened for the presence of insert by PCR prior to soluble expression. A single, insert-containing colony was picked and grown overnight in 10 mL of SB supplemented with 30 µg/mL of chloramphenicol. The 10 mL culture was then added to 1000 mL of SB supplemented with 30 µg/mL of chloramphenicol and 0.02 M MgCl2. The culture was incubated until 0.8 absorbance measured at 600 nm was achieved. Protein expression was induced by the addition of IPTG to a final concentration of 0.01
mM and incubation of the culture with shaking at 21 °C. Induced protein was harvested by centrifuging the culture at 3000g for 30 min at 4 °C and resuspending the cell pellet in equilibration buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole, 200 µM PMSF). Soluble antibody fragments were released from the bacteria by sonication, using 10 cycles of 10 s at full-power, with a 20 s recovery on ice between each cycle. Insoluble material was removed by centrifugation at 10 000g for 60 min at 4 °C. The soluble scFv-containing supernatant was filtered through a 0.2 µm filter before purification on a nickel-NTS column. The soluble scFv fraction was loaded onto an equilibrated 2 mL Ni-NTS column and washed with wash buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole). The scFv was eluted from the column with 5 mL of elution buffer (50 mM NaH2PO4, 300 mM NaCl, 250 mM imidazole) and concentrated 10-fold. Purified protein was visualized on a 10% SDS polyacrylamide gel.16 ScFv-Based Competitive ELISA. Checkerboard titrations of DA-OVA coating conjugate and the soluble scFv preparations were performed to determine the optimal dilutions of each for use in the competitive ELISA. The protocol was as previously described for titration of the polyclonal antisera except that wells were blocked with 3% ovalbumin in PBS, and this buffer was also used as diluent for the scFvs, and HRP-conjugated anti-His tag monoclonal antibody (Sigma Aldrich) diluted in PBS/0.1% BSA was used to detect bound scFv. For the final assay, a coating concentration of 4 µg/mL of DA-OVA was used, and DA standards were prepared over the range 0.2 to 25 ng/mL (0.64-80.3 nM), as previously described. To triplicate coated wells 20 µL of standard or sample was added followed immediately by 100 µL of the selected scFv diluted according to the titer. Well contents were mixed briefly and incubated for 1 h at 37 °C. Detection of bound scFv was as described above Analysis of Shellfish Extracts. Crude extracts of various scallop tissues (hepatopancreas, abductor muscle, gonads, and remaining tissue) were provided by the Marine Institute, Galway, Ireland from their routine DA screening program. Extracts had been prepared by homogenization of a weighed aliquot of tissue in 50% methanol using the Ultraturrax, followed by centrifugation. All extracts were assayed for DA by HPLC in the Marine Institute using an accredited procedure.17 A range of dilutions of each extract was prepared for analysis in the competitive ELISA. The DA concentrations of the extracts were determined from the dilution giving a reading nearest the midpoint of the standard curve and compared to HPLC values. RESULTS Following the final boost of the DA-immunized sheep, ELISA analysis of the serum IgG responses showed anti-DA titers ranging from 1 in 40 000 to 1 in 176 000. The I50s, defined as the standard concentration giving 50% inhibition of antibody binding, of these sera in competitive ELISA ranged from 2.8 to 27 ng/mL (8.686.4 nM), and limits of quantification (LOQ), defined as the standard concentrations corresponding to I20, ranged from 0.02 (16) Laemmli, U. K. Nature 1970, 227, 680. (17) Hess, P.; Kilcoyne, J.; Swords, D.; Mulcahy, N.; McCarron, M.; Keogh, M.; Gibbons, B.; Ronan, J. In Proceedings of the Fifth International Conference of Molluscan Shellfish Safety, Galway, Ireland; Henshilwood, K., Deegan, B., McMahon, T., Cusack, C., Keaveney, S., Silke, J., O’Cinneide, M., Lyons, D., Hess, P., Eds.; 2004; p 77.
Figure 2. Binding of polyclonal phage preparations to DA-OVA before and following successive rounds of panning. DA-OVA was coated at 250 ng/mL, 100 µL per well. Reaction of each phage preparation against immobilized OVA was negligible (data not shown). The error bars indicate the standard error of the mean for 3 determinations.
to 0.2 ng/mL (0.06 to 0.6 nM), showing the response was specific for DA and highly sensitive. Sheep 2481 gave the most sensitive standard curve (I50, 2.8 ng/mL (8.6 nM), LOQ, 0.2 ng/mL) with a titer of 1 in 122 000. A recombinant antibody library was generated from splenocytes of sheep 2481 and cloned into the phage-display vector pAK100, giving a library size of 2 × 106 transformants. PCR analysis of 16 randomly selected clones from this library showed 100% of the library carried a full length scFv insert (data not shown). Prior to biopanning of library, the nonenriched unpanned phage population was evaluated for binding to DA-OVA and a high response was obtained, reflecting the high titer of circulating antibodies at the time of slaughter. Biopanning of the phagedisplayed library against DA-OVA involved increasingly stringent conditions (reduction in antigen coating concentration and reduction in time and temperature of binding) over the four rounds of panning to select for high affinity binders (Table 2). Except for the output from the first pan, the anti-DA polyclonal phage response was very strong after the remaining pans (Figure 2). The I50 of the polyclonal phage response decreased from 1.1 µg/ mL (3.5 µM) for the unpanned library to 8 ng/mL (25.6 nM) after pan 4 (Figure 3), indicating a significant increase in affinity on panning. ScFv Selection and Characterization. Individual clones were picked randomly from the output phage of pan 4 and grown in 96 well plates. When tested for reactivity against DA, seven DAbinding clones were identified. Three clones (DA24cB7, DA24cC8, and DA24cG12) had binding greater than 20 times above background, and these scFv gene fragments were isolated and transferred to the pAK400 vector15 for soluble scFv antibody production. All three clones expressed functional soluble scFv, DA24cB7 giving the lowest I50 (2.6 ng/mL, 8.3 nM) on preliminary evaluation in competitive ELISA (Figure 4). The I50 given by DA24cC8 was 30.7 ng/mL (98.2 nM) and by DA24cG12 was 58.0 ng/mL (186.0 nM). ScFv DA24cB7 was therefore selected for assay development. Large scale production and purification by Ni affinity chromatography was found to yield routinely 5-10 mg of the scFv per liter culture. ScFv preparations were shown to be predominantly monomeric by SDS-PAGE under nonreducing conditions (Figure 5). Analytical Chemistry, Vol. 80, No. 9, May 1, 2008
3209
Figure 3. Evaluation of polyclonal phage preparations before and following successive rounds of panning in competitive ELISA. DAOVA was coated at 250 ng/mL, 100 µL per well. The error bars indicate the standard error of the mean for 3 determinations. Figure 5. SDS-PAGE under reducing and nonreducing conditions of scFv DA24cB7. Lane 1 molecular weight markers (kilodaltons). Lane 2, scFv preparation following purification by Ni affinity chromatography run under reducing conditions. Lane 3, scFv preparation following purification by Ni affinity chromatography run under nonreducing conditions.
Figure 4. Evaluation of the three soluble scFv preparations, DA24cB7, DA24cC8, and DA24cG12, in competitive ELISA. DA-OVA was coated at 250 ng/mL, 100 µL per well. The error bars indicate the standard error of the mean for 3 determinations.
Competitive ELISA with scFv DA24cB7. The coating concentration of DA-OVA conjugate, blocking conditions, scFv dilution, standard concentrations, and anti-His-HRP conjugate dilution were all examined to find the optimal conditions for use in the competitive ELISA based on scFv DA24cB7. The final assay conditions are given as previously described above. The resulting calibration curve shown in Figure 6 represents the mean of 19 independent standard curves set up over a period of 6 weeks with four different batches of scFv, each with triplicate determinations of standards. The assay was highly reproducible, with a mean I50 of 1.21 ng/mL (3.9 nM), a linear detection range, calculated as the I10 to the I80, of 0.30 ng/mL (1.0 nM) to 5.58 ng/mL (17.9 nM), and an LOQ, defined conservatively as the I20, of 0.5 ng/ mL. Detection of Domoic Acid in Shellfish Extracts. The critical regulatory concentration of DA in shellfish is 20 mg/kg tissue with subsidiary cutoffs at 4.6 and 250 mg/kg.18 These values are equivalent to 5.0, 1.2, and 62.5 µg/mL extract, respectively, using the normal DA extraction protocol with 50% methanol. Because of the high sensitivity of the scFv standard curve, all extracts had to be diluted at least 500-fold prior to assay for results to fall on (18) European Commission. Official J. Eur. Communities Commission 2002, L75 65.
3210 Analytical Chemistry, Vol. 80, No. 9, May 1, 2008
Figure 6. Composite standard curve (n ) 19) for the competitive ELISA for DA using scFv DA24cB7 and optimized assay conditions where plates were coated with DA-OVA at 250 ng/mL, 100 µL per well. The error bars indicate the standard error of the mean. Each value for individual standard curves was determined in triplicate.
the standard curve, thereby eliminating any matrix interference in the assay. Forty scallop tissue extracts were provided by the Marine Institute, Ireland with known concentrations of DA which ranged from 1.69-1281 mg/kg of shellfish tissue. Using the DA24cB7-based competitive ELISA, the concentration of DA in each sample was determined and compared to the results obtained by HPLC. The correlation of the results between the two assay methods for these 40 samples was high with a R2 value of 0.91 and a slope of 1.39 (Figure 7 (i)). When the six hepatopancreas samples with high DA content were omitted, the correlation was still good (R2 ) 0.80; slope ) 0.66), although the ELISA showed a definite tendency to underestimate the DA concentration compared to the HPLC method (Figure 7 (ii)). However, use of 15 mg/kg as a cutoff for the scFv-based competitive ELISA, instead of the regulatory cutoff of 20 mg/kg in the edible parts of molluscs, resulted in the identification of all samples which were positive by HPLC and a positive predictive value of 75% with this sample set. The absence of false negatives indicates that this assay
Figure 7. Regression analysis of the correlation between the results given by the HPLC assay and the DA24cB7-based competitive ELISA for DA in random shellfish extracts taken from the Irish shellfish testing program, i) including results for all 40 shellfish extracts analyzed and ii) including results for shellfish extracts containing DA (by HPLC) at 0-60 mg/kg. The dotted line represents the line of perfect agreement and the solid line the correlation plot.
would be suitable for use as a convenient screening assay for the detection of DA in scallop extracts. DISCUSSION The recombinant antibody library generated during this study yielded three clones with high affinity for DA, all of which would have been suitable for development of sensitive immunoassays for detection of DA in shellfish extracts. This was despite the fact that the size of the library was relatively small (2 × 106) compared to some reported phage display libraries against specific antigens.14,15,19 However, the animal used to generate the library had produced a highly specific, high titer immune response following immunization against DA, so that the circulating B-cell population had already been enriched for DA-binding activity prior to the generation of the phage display library. This was confirmed with the detection of anti-DA-specific phage in the unpanned phage population. Competitive immunoassays are the most common immunoassay configuration used for analysis of haptens, such as DA (MW, 311 Da). However, the assay sensitivity achievable in a competitive immunoassay is limited by the affinity of the antibody used, i.e., high affinity antibodies are needed to achieve low detection limits in competitive assays.20 The panning procedure used in this study, involving successively lower coating concentrations of antigen, (19) Zhao, A.; Qin, W.; Han, Y.; Wen, W.; Zhang, W.; Lian, Z.; Chen, G.; Zhang, Z.; Peng, J.; Wang, H.; and Guo, Y. Microbes Infect. 2007, 9, 1026. (20) Ekins, R. Nature 1980, 284, 14.
shorter incubation times, and more extensive washing steps, was designed to favor the selection of high affinity binders. The increase in affinity, as demonstrated by the resulting reduction in I50,21,22 of the output phage as pannings proceed shows the success of this approach (Figure 3). Alternative panning approaches have been used to target high affinity binders.23 Charlton et al.11 maintained a high coating concentration of their target antigen through seven rounds of panning, in the last five of which bound phage were specifically eluted by incubation with decreasing concentrations of antigen. It is possible that this approach might have yielded higher affinity scFvs in this study, but those generated using our simpler approach were already more than adequately sensitive for the intended application and comparable in competitive ELISA performance to the better anti-DA antibodies described, both in terms of the I50, 0.5 to 1.5 ng/mL versus 1.2 ng/mL for sheep scFv DA24cB7, and the LOQ (I20), 0.01 to 0.15 ng/mL versus 0.5 ng/mL for DA24cB7.12,24-25 Also, the I50 of DA24cB7 was lower than that given by the polyclonal antiserum from the animal used. The pAK100/pAK400 vector system, which was developed for the isolation of mouse scFvs,15 was devised as a complementary vector system to allow for the easy swapping of scFv inserts between a gpIII fusion version for phage display to a 6xHIS-tagged version for soluble expression. It was found in this study to be a very convenient system for use in isolation of good binding clones from a sheep scFv library and subsequently for production of soluble scFv in consistent amounts. The resulting scFv was also found to be stable for several months when stored at 4 °C. The sheep scFv produced by Charlton et al.11 was also very stable; this may be a feature of sheep scFvs unlike scFvs isolated from other species which are often not so robust and require further manipulation to improve stability.26 Further sheep scFvs will have to be evaluated to confirm their comparable stability to the majority of polyclonal and monoclonal antibodies. Ease of production and stability will be major factors in the decision to use scFvs for commercial immunoassay kit development. There is still debate on whether immunization is necessary to generate high affinity anti-hapten recombinant antibodies from phage libraries. Naı¨ve libraries have been used for selection of such antibodies, and some high affinity binders have been produced, e.g., an anti-microcystin scFv isolated from the Griffin semi-synthetic library, which gave an I50 of 13 nM.10 However, most antibodies described from naı¨ve libraries have much lower affinity and are not suitable for sensitive assay development.27-28 As shown in this study, preselected immune libraries provide a much greater chance of selecting suitable antibodies with the affinity required particularly for sensitive detection of haptens in competitive assays. However, both naı¨ve and immune phage (21) Muller, R. J. Immunol. Methods 1980, 34, 345. (22) O’Connor, T.; Gosling, J. P. J. Immunol. Methods 1997, 208, 181. (23) Sheedy, C.; MacKenzie, C. R.; Hall, J. C. Biotechnol. Adv. 2007, 25, 333. (24) Tsao, Z. J.; Liao, Y. C.; Liu, B. H.; Su, C. C.; Yu, F. Y. J. Agric. Food Chem. 2007, 55, 4921. (25) Kleivdal, H.; Kristiansen, S. I.; Nilsen, M. V.; Goksoyr, A.; Briggs, L.; Holland, P.; McNabb, P. J. AOAC Int. 2007, 90, 1011. (26) Dooley, H.; Grant, S. D.; Harris, W. J.; Porter, A. J. Biotechnol. Appl. Biochem. 1998, 28, 77. (27) Moghaddam, A.; Borgens, T.; Stacy, J.; Kausmally, L.; Simonsen, B.; Marvik, O. J.; Brekke, O. H.; Braunagel, M. J. Immunol. Methods 2003, 280, 139. (28) Lauer, B.; Ottleben, I.; Jacobsen, H. J.; Reinard, T. J. Agric. Food Chem. 2005, 53, 899.
Analytical Chemistry, Vol. 80, No. 9, May 1, 2008
3211
libraries suffer equally from the hazard of pulling out anti-hapten antibodies fragments that can bind to the conjugated hapten, but have very poor binding to the free hapten.29 This emphasizes the importance of checking the specificity of the binding response by competitive assay at each stage of the selection process. The final arbiter of any antibody/antibody fragment is a demonstration of its suitability for intended use. The DA24cB7based ELISA returned concentrations of DA in crude scallop extracts that correlated well with the accredited HPLC assay, used for routine analysis of samples for regulatory purposes, and allowed all samples above the regulatory threshold to be identified. The absence of false negatives indicates the suitability of this scFv for use in a shellfish DA-screening assay. Further investigations are, however, needed to explain the underestimation of DA compared to the HPLC assay and also more extensive validation of the assay must be carried out. While a number of immunoassays for DA have been described,12,24,30 to our knowledge, there is only one commercial immunoassay with high-throughput potential available for the determination of DA in shellfish samples.25 This ELISA kit is based on the sheep polyclonal antiserum described by Garthwaite et al.30 While consumption of a high affinity polyclonal antiserum by microtiter plate ELISAs
can be very low, assay formats for on-site monitoring applications require much more antibody and hence a continuous and consistent supply of reagents, which cannot be guaranteed if dependent on the use of polyclonal antibodies. This study has demonstrated a convenient method for immortalizing valuable sheep polyclonal antisera and should be of interest to the many manufacturers of commercial kits dependent on a limited supply of polyclonal serum from a small number of animals.
(29) Marks, J. D.; Hoogenboom, H. R.; Bonnert, T. P.; McCafferty, J.; Griffiths, A. D.; Winter, G. J. Mol. Biol. 1991, 222, 581. (30) Garthwaite, I.; Ross, K. M.; Miles, C. O.; Hansen, R. P.; Foster, D.; Wilkins, A. L.; Towers, N. R. Nat. Toxins 1998, 6, 93.
Received for review November 26, 2007. Accepted February 20, 2008.
3212
Analytical Chemistry, Vol. 80, No. 9, May 1, 2008
ACKNOWLEDGMENT We would like to thank Dr. Andreas Pluckthun, Department of Biochemistry, University of Zu¨rich, Winterthurerstrasse 190, CH-8057 Zu¨rich, Switzerland for the gift of the pAK100 and pAK400 vectors, Dr. Philipp Hess, Biotoxins Chemistry, Marine Institute, Rinville, Oranmore, County Galway, Ireland for supplying the shellfish extracts and HPLC results, and Dr. Gerard Wall, Department of Microbiology, National University of Ireland Galway, Ireland for useful discussions. Personnel at Teagasc Agricultural Research Station, Athenry, Galway, Ireland are also thanked for animal care. This work was funded by an Advanced Technologies Research Program grant from Enterprise Ireland.
AC7024199