Comparison of hybridoma screening methods for the efficient detection of high-affinity hapten-specific monoclonal antibodies

Journal of Immunological Methods 329 (2008) 184 – 193 www.elsevier.com/locate/jim

Research paper

Comparison of hybridoma screening methods for the efficient detection of high-affinity hapten-specific monoclonal antibodies Christian Cervino a , Ekkehard Weber b , Dietmar Knopp a,⁎, Reinhard Niessner a a

b

Institute of Hydrochemistry, Chair for Analytical Chemistry, Technische Universität München, Marchioninistrasse 17, 81377 München, Germany Institute of Physiological Chemistry, Martin-Luther-Universität Halle-Wittenberg, Hollystrasse 1, 06114 Halle, Germany Received 16 July 2007; received in revised form 24 September 2007; accepted 19 October 2007 Available online 20 November 2007

Abstract This study compares diverse microplate-based hybridoma screening methods for the generation of hapten-(aflatoxin-) specific monoclonal antibodies (MAbs). Standard indirect enzyme-linked immunosorbent assay (ELISA) screenings (with immobilization of hapten–protein conjugate and use of enzyme-labeled anti-mouse IgG as tracer) were compared with direct ELISAs (with antibody immobilization and use of a hapten-enzyme conjugate as tracer). Although direct ELISA is rarely used for routine hybridoma screenings, it showed considerable advantages compared to the indirect assays. Standard indirect ELISA screening can lead to a considerable number of false positives (up to about 50% false positives of all 373 supernatants tested) if the antibody concentrations in the supernatants are too high. Direct ELISAs gave useful screening results for the different supernatant dilutions chosen. At most 3 false positives were detected out of 373 supernatants. However, the sensitivity of the direct ELISA screening is generally lower compared to indirect ELISA, and individual high-affinity MAbs might be classified as false negative. Therefore, a modified indirect ELISA screening was also developed. It includes pre-incubation of the supernatants in anti-mouse IgG-coated microplates which are then transferred into the (indirect) hapten conjugate-coated microplates. This screening method leads to excellent results with good overall selectivity and sensitivity. It can also be conveniently combined with the direct ELISA screening. Using these improved screening methods, aflatoxin-specific MAbs could be generated with IC50 values down to 3 ng/l (aflatoxin concentration). © 2007 Elsevier B.V. All rights reserved. Keywords: Monoclonal antibody; Hapten-specific; Hybridoma screening; ELISA; Aflatoxin; Mycotoxin

Abbreviations: AF, aflatoxin; BSA, bovine serum albumin; CMO, carboxymethoxyloxime; DCC, N,N′-dicyclohexylcarbodiimide; DMSO, dimethyl sulfoxide; ELISA, enzyme-linked immunosorbent assay; HAT, hypoxanthine/aminopterin/thymidine; HEPES, 4-(2-hydroxyethyl)-1piperazineethanesulfonic acid; HT, hypoxanthine/thymidine; IC50, concentration of analyte that causes a 50% decrease of the maximum response; LC, liquid chromatography; MAb, monoclonal antibody; MALDI, matrix-assisted laser desorption/ionization; MCD, mean coupling density; MS, mass spectrometry; NHS, N-hydroxysuccinimide; PBS, phosphate buffered saline; POD, horseradish peroxidase; RPMI, Roswell Park Memorial Institute; TG, bovine thyroglobulin; TMB, tetramethylbenzidine; TOF, time-of-flight. ⁎ Corresponding author. Tel.: +49 89 2180 782 52; fax: +49 89 2180 782 55. E-mail address: [email protected] (D. Knopp). 0022-1759/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jim.2007.10.010

C. Cervino et al. / Journal of Immunological Methods 329 (2008) 184–193

1. Introduction Monoclonal antibodies (MAbs) are useful biological tools for various analytical applications, e.g. in clinical chemistry, food analysis, and environmental monitoring. In addition, antibodies are increasingly used as human therapeutics. Immunization of animals, mainly mice, in combination with hybridoma technology is still the most common method for the generation of MAbs. Regardless of the intended application, the selection of high-affinity MAbs is often preferred. An efficient hybridoma screening procedure is a crucial step that usually has to be accomplished within about one day (Burrin and Newman, 1991). Thus, the ideal screening method should be fast, reliable, and easy to accomplish, especially if there is little or no equipment available in the laboratory for carrying out automated immunoassays. It should clearly detect high-affinity MAbs with a minimum of both false positives and false negatives. In addition, useful screening results must be obtained relatively independent of the MAb concentration in the supernatants, because optimization of the ELISA parameters (such as supernatant dilution and coating conjugate dilution) prior to the screening is usually much too time-consuming or even impossible, especially as the screening often only involves a single measurement per MAb. The microplate-based antigen-immobilized ELISA (indirect ELISA) is the de facto standard screening method for the detection of hapten-specific antibodies (Grol and Schulze, 1990), although other methods have been reported, e.g. BIAcore screening (Canziani et al., 2004), flow-based immunoassay (Sasaki et al., 2005), and time-resolved fluorescence assay (Daigo et al., 2006). It usually includes immobilization of a hapten– protein conjugate on the microplate surface, the addition of (diluted) hybridoma culture supernatant and use of a (enzyme-) labeled secondary (anti-mouse IgG) antibody. For the generation of anti-hapten MAbs, simultaneous non-competitive/competitive indirect ELISAs can be performed (Abad and Montoya, 1994; Mercader and Montoya, 1999; Moreno et al., 2001; Manclús et al., 2004). Yet an alternative immunoassay, although rarely used, is known for the screening of hapten-specific MAbs (henceforth referred to as direct ELISA) (Cho et al., 2005; Schetters, 1993; Hack et al., 1987). It includes immobilization of a capture (anti-mouse IgG-) antibody, addition of supernatant, and use of a haptenenzyme conjugate. Qualitative differences between these two screening methods have been suggested (Kane and Banks, 2000). We have had diverse experiences with the two immunoassays in our laboratories (Matschulat et al., 2005; Mangler et al., 1994;

185

Winklmair et al., 1997; Weller, 1992), which encouraged us to carry out a systematic comparative study. For this work, aflatoxin-specific MAbs were produced. Aflatoxins, a sub-group of mycotoxins, are low molecular weight secondary metabolic products of moulds (Aspergillus flavus and Aspergillus parasiticus) that can contaminate various food matrices. Due to their extreme carcinogenicity these toxins are of great concern. Strict maximum permissible limits exist in most countries worldwide (Food and Agriculture Organization, 2004). High-affinity aflatoxin-specific MAbs are therefore useful tools for analytical food chemistry (Eaton and Groopman, 1994). 2. Materials and methods 2.1. Safety note Aflatoxins are highly carcinogenic and should be handled with extreme care. Aflatoxin-contaminated material should be decontaminated with an aqueous solution of sodium hypochlorite (5%). 2.2. Materials, reagents, and equipment Flat-bottom, transparent 96-well polystyrene microplates were obtained from Greiner (Frickenhausen, Germany). Sephadex G-25 columns were purchased from GE Healthcare Bio-Sciences AB (Uppsala, Sweden). An IsoStrip mouse monoclonal antibody isotyping kit was purchased from Roche (Mannheim, Germany). Anti-mouse IgG (whole antiserum, produced in goat) was purchased from Sigma (St. Louis, MO). Monoclonal aflatoxin-specific reference antibodies were obtained from LCTech (Dorfen, Germany). Solid horseradish peroxidase (POD, EIA grade) was obtained from Roche. POD-labeled anti-mouse IgG antibody (H + L, produced in horse, affinity purified) was obtained from Axxora (Lörrach, Germany). Bovine serum albumin (BSA, fraction V, ∼ 99%) and casein were obtained from Sigma. Thyroglobulin (TG) from bovine thyroid glands was obtained from Fluka (Buchs, Switzerland). Aflatoxins B1, B2, G1, and G2 were obtained from Sigma. Aflatoxin standard stock solutions in acetonitrile (0.1 mg/ml) were prepared by a validated method (Nesheim et al., 1999). 3,3′,5,5′-tetramethylbenzidine (TMB, ≥ 99%), carboxymethoxylamine hemihydrochloride (∼ 98%), N,N′-dicyclohexylcarbodiimide (DCC, ≥ 99%), N-hydroxysuccinimide (NHS, ≥ 97%), dioxane (≥ 99.5%, H2O ≤ 0.01%), acetonitrile (HPLC grade), H2O2 (35%), Tween 20, and dimethyl sulfoxide (DMSO, ≥ 99.5%) were obtained from Sigma. All

186

C. Cervino et al. / Journal of Immunological Methods 329 (2008) 184–193

buffers were prepared from the corresponding anhydrous salts (purity ≥ 99%). Phosphate buffered saline (PBS, pH 7.6) consisted of a solution of 1.36 g potassium dihydrogenphosphate, 12.20 g dipotassium hydrogenphosphate, and 8.5 g sodium chloride in 1 l of water. The washing buffer (pH 7.6) was a solution of 55 mg potassium dihydrogenphosphate, 49 mg dipotassium hydrogenphosphate, 35 mg sodium chloride, and 202 μl of Tween 20 in 1 l of water. The coating buffer (pH 9.6) was a solution of 1.59 g disodium carbonate, 2.93 g sodium hydrogencarbonate, and 0.20 g sodium azide in 1 l of water. The substrate buffer (pH 3.8) was a solution of 46.04 g potassium dihydrogencitrate and 0.10 g potassium sorbate in 1 l of water. Substrate solution for ELISAs was always freshly prepared directly before use by adding 273 μl of TMB stock solution (375 mg TMB in 30 ml of DMSO) and 138 μl of 1% H2O2 to 15 ml of substrate buffer. Sulfuric acid (5%) was used as stop solution. The coupling buffer was a solution of 3.00 g sodium hydrogencarbonate in 50 ml of water brought to pH 7.5 by addition of 1.0 ml of semi-concentrated hydrochloric acid. For the ELISAs, a 96-channel plate washer (ELx405 Select) and a microplate reader (Synergy HT) from BioTek (Bad Friedrichshall, Germany) was used. 2.3. Preparation of aflatoxin–protein conjugates Aflatoxin B2 (AFB2) was coupled to BSA, TG, and POD by means of a carboxymethoxyloxime (CMO) spacer. AFB2-CMO was synthesized by reaction of AFB2 with carboxymethoxylamine and recrystallized (Hastings et al., 1989). Structure and purity were confirmed by LC-MS. AFB2–protein conjugates were synthesized by the NHS-ester method, avoiding unwanted coupling of activating reagent (DCC) with the proteins (Cervino et al., 2007). Briefly, 20 μl of a solution of 34 mg NHS in 1.0 ml of dry dioxane and 26 μl of a solution of 38 mg DCC in 1.0 ml of dry dioxane were added to a solution of 2.0 mg AFB2-CMO in 0.4 ml of dry dioxane and shaken overnight in a closed vial at room temperature (fine needles of the dicyclohexylurea by-product start crystallizing after about 12 h). Eighty microliters of this AFB2-CMONHS-ester solution were added to an ice-cold solution of (a) 2.0 mg BSA in 230 μl of coupling buffer and shaken for 6 h, (b) 1.6 mg TG in 160 μl of coupling buffer and shaken for 6 h, or (c) 1.0 mg of POD in 160 μl of coupling buffer and shaken for 1 h while warming up to room temperature (addition of the NHS-ester solution in dioxane raises the buffer's pH to 8.5–9.0). The protein conjugates (AFB2-CMO-BSA, AFB2-CMO-

TG, and AFB2-CMO-POD) were subsequently purified on a Sephadex G-25 column using PBS as mobile phase. Aflatoxin coupling was confirmed by measurement of the conjugates' UV spectra showing intensive absorption at the maximum for aflatoxin at around 360 nm. A mean coupling density (MCD) of about 7 mol AFB2 per mole of protein was determined by MALDI-TOF-MS for AFB2-CMO-BSA. Dilutions of the aflatoxin– protein conjugates are henceforth referred to stock solutions of 1 mg/ml. 2.4. Immunization and generation of hybridomas Four 10-week-old female BALB/c mice (Charles River, Sulzfeld, Germany) were immunized with 25 μg AFB2-CMO-BSA per mouse emulsified in 200 μl of sodium chloride solution (154 mM) and 300 μl of complete Freund's adjuvant (Sigma). The intraperitoneal immunizations were repeated after 6, 8 and 10 weeks (25 μg of AFB2-CMO-BSA with incomplete Freund's adjuvant (Sigma)). Mice were then rested for another 4 weeks. The affinities of the sera were determined analogously to the affinity measurements of the MAbs (Section 2.7). Four days and one day before the fusion, the mouse with the highest serum affinity to AFB2 (minimal IC50 value 10 ng/l) was boosted with another 25 μg of antigen in 200 μl of sodium chloride solution (154 mM). The fusion protocol was based on a standard method (Köhler and Milstein, 1975) with some modifications. Briefly, 108 spleen cells were mixed with 5 × 107 cells of the mouse non-secretory cell line P3X63Ag8.653 (CRL 1580, LGC Promochem, Wesel, Germany) and polyethylene glycol 1500 (Roche). Fused cells were grown in a RPMI-based HAT selection medium containing hypoxanthine (100 μM), aminopterin (400 μM), and thymidine (16 μM) (HAT supplement, 21060-017, Invitrogen, Karlsruhe, Germany) for 2–3 weeks followed by 1 week in HT medium. Hybridomas were then grown in RPMI 1640 medium supplemented with 10% fetal calf serum (Invitrogen), HEPES (4.77 g/l, acidic form, Roth, Karlsruhe, Germany), glucose (2.5 g/l, Merck, Darmstadt, Germany), 2-mercaptoethanol (1 mM, Serva, Heidelberg, Germany), L-glutamine (2 mM, Invitrogen), bovine insulin (5 mg/l, Roche), sodium pyruvate (50 mg/l, Invitrogen), succinic acid (65 mg/l, Invitrogen) and gentamicin (80 mg/l, Serva). Hybridoma supernatants were screened for aflatoxinspecific antibodies by ELISA (Sections 2.5 and 2.6) and cultures of interest (in this work 27 of the 373 hybridomas) were rendered monoclonal by 2–3 limiting dilutions.

C. Cervino et al. / Journal of Immunological Methods 329 (2008) 184–193

187

necessary). For some (e.g. four) wells, positive controls (aflatoxin-specific MAb, diluted 1/10,000 in PBS, 100 μl/well) and negative controls (PBS, 100 μl/well) were added instead. After incubation for 1.5 h, plates were washed three times and AFB2-CMO-POD (diluted 1/10,000 in 0.1% BSA in PBS (w/v), 100 μl/ well) was added and incubated for 1 h. After threefold washing, color development was carried out as described for the indirect ELISA screening procedure (Section 2.5).

2.5. Indirect ELISA screening Microplates were coated with AFB2-CMO-TG dissolved in coating buffer (dilution 1/10,000, 200 μl/well). After overnight incubation at 4 °C plates were washed three times with washing buffer and blocked with 1% casein in PBS (w/v, 250 μl/well) for 20 min. After threefold washing, the supernatants (diluted in PBS, 100 μl/well) were then added for the standard indirect ELISAs. For the modified indirect ELISA, supernatants were added after initial pre-incubation in anti-mouse IgG-coated microplates (for anti-mouse IgG coating see Section 2.6) for 1.5 h. For some (e.g. four) wells, positive controls (aflatoxin-specific MAb, diluted 1/10,000 in PBS, 100 μl/well) and negative controls (PBS, 100 μl/ well) were added instead of the supernatants. After incubation for 1.5 h, plates were washed three times and incubated with POD-labeled anti-mouse IgG (diluted 1/ 10,000 in 0.1% BSA in PBS (w/v), 100 μl/well) and incubated for 1 h. After threefold washing, color development was carried out by adding freshly prepared substrate solution (100 μl/well) and gentle shaking. Absorbance was measured at 650 nm after several time intervals (e.g. after 3, 10, and 25 min) while the enzymecatalyzed reaction was running. It is favorable to measure when the differences in color intensities between the different supernatants are clearly visible by eye. Finally, stop solution was added (100 μl/well) after 25 min and absorbance was measured at 450 nm.

2.7. IC50 measurements The affinity to AFB2 was estimated for selected supernatants by means of indirect competitive ELISAs analogously to the method described elsewhere (Winklmair et al., 1997; Weller, 1992). The dilutions of coating conjugate and supernatant were systematically optimized for each MAb examined to obtain a minimal IC50 value of the antibody response curve (Section 3.1). The configuration of a microplate for such a “two-dimensional titration” is shown in Table 1. Using this configuration, it is possible to optimize the most important ELISA parameters for one MAb on a single 96-well plate. The indirect competitive ELISA for a two-dimensional titration for one supernatant was carried out as follows. A 96-well plate was coated with AFB2-CMO-TG in coating buffer (dilution 1/ 10,000, 200 μl/well). After overnight incubation at 4 °C the plate was washed three times with washing buffer, blocked with 1% casein in PBS (w/v, 250 μl/well), and incubated for 20 min. After washing three times, AFB2 (diluted in water (for dilutions see Table 1), 100 μl/well) and supernatant (diluted in PBS (for dilutions see Table 1), 100 μl/well) were added and incubated for 1.5 h. For some (e.g. four) wells, positive controls (aflatoxin-specific MAb, diluted 1/10,000 in PBS,

2.6. Direct ELISA screening Microplates were coated with anti-mouse IgG (goat) dissolved in coating buffer (dilution 1/5000, 250 μl/well) and incubated overnight at 4 °C. After washing three times with washing buffer, supernatants (diluted in PBS, 100 μl/well) were added (blocking is not

Table 1 Ninety-six-well plate configuration for the estimation of a MAb's affinity showing the principle of a two-dimensional titration for minimization of the IC50 value Column a, c Dilution of coating conjugate Dilution of supernatant b

1 b

2

3

1/nx 1/my

5

6 2

1/n x 1/my

1/nx 1/m2y

7

8 2

1/n x 1/m2y

9 3

1/x 1/y

1/nx 1/y

Row c

A

B

C

D

E

F

G

H

AFB2 concentration (μg/l)

0

1.0 × 10− 4

1.0 × 10− 3

1.0 × 10− 2

0.10

1.0

10

100

a

1/x 1/my

4

1/n x 1/m2y

10 2

1/n x 1/m3y

1/n3x 1/m3y

Positive and negative controls can be added in columns 11 and 12 (not shown). The assays were mostly carried out at dilutions of x = 10,000 and y = 100, if a relatively strong color development was obtained for the respective supernatant in the preceding antibody screening. As dilution factors n and m 2, 3, or 4 were chosen. c The dilutions of coating conjugate (AFB2-CMO-TG) and supernatant are varied column-wise. For competition the AFB2 concentration is varied line by-line. As only single measurements are carried out, the results have to be validated afterwards (see Section 2.7). b

188

C. Cervino et al. / Journal of Immunological Methods 329 (2008) 184–193

200 μl/well) and negative controls (PBS, 200 μl/well) were added instead (without any AFB2). After washing three times, POD-labeled anti-mouse IgG (diluted 1/ 10,000 in 0.1% BSA in PBS (w/v), 100 μl/well) was added and incubated for 1 h. After further washing, color development was carried out by adding freshly prepared substrate solution (100 μl/well) and gentle shaking. Stop solution (100 μl/well) was added after sufficient color had developed (stop solution can be added column-wise after different time intervals). Absorbance was measured at 450 nm. For validation of the results, the competitive ELISA was reproduced with three measurements under the optimal conditions observed (i.e. at the dilutions of coating conjugate and supernatant that led to the minimal IC50 value but for which the color development was still sufficiently intense).

3. Results and discussion 3.1. IC50 measurements In order to evaluate and compare the screening methods, it was necessary to classify MAbs as interesting (i.e. of sufficiently high affinity) or not. The affinities of selected MAbs which gave a relatively strong signal in the preceding screenings (especially for the screenings shown in Fig. 2b, c, d, and e) were estimated by performing indirect competitive ELISA with optimization of the most important parameters, i.e. dilution of coating conjugate and supernatant (see Section 2.7 and Table 1). Fig. 1a shows antibody response curves for a high-affinity MAb with variation of the two parameters. It is obvious that the resulting IC50 value is a monotone function of both parameters.

Fig. 1. Example of a two-dimensional titration for minimization of the IC50 value of MAb 195 (a). Antibody response curves of indirect competitive ELISAs are shown for various ELISA parameters. (b) Antibody response curves for aflatoxins B1, B2, G1, and G2 for determination of the crossreactivities. A 1/90,000 dilution of coating conjugate and a 1/9000 dilution of supernatant were chosen. For corresponding data see Table 2.

C. Cervino et al. / Journal of Immunological Methods 329 (2008) 184–193 Table 2 Antibody types, midpoints, and cross-reactivities of the 5 high-affinity MAbs prepared with an aflatoxin B2 conjugate MAb no. 54 Antibody type Minimal IC50 (aflatoxin, ng/l) Cross-reactivity (%) AFB1 AFB2 AFG1 AFG2 a

62

195

287

357

– IgG2a, λ IgG2b, λ IgG1, κ IgG1, κ 53 ± 11 2.7 ± 0.2 30 ± 7 10 ± 5 3 ± 0.5 a

–a –a –a –a

100 98 72 30

37 100 4 6

32 100 5 7

100 42 88 19

Could not be determined due to unstable cell line.

The IC50 measurements were validated by repeating the competitive ELISA under optimized conditions with three replicates. In addition, the cross-reactivities to structurally similar toxins, i.e. AFB1, AFG1, and AFG2, were determined for the high-affinity MAbs (Fig. 1b). For 22 of the 27 MAbs examined, no AFB2 inhibition could be observed for the highest AFB2 concentration chosen (100 μg/l). Such MAbs were classified as “not interesting”. Five MAbs showed high affinity to AFB2. Table 2 summarizes their minimal IC50 values and cross-reactivities measured. 3.2. Comparison of ELISA screening methods The purpose of an optimal screening method, in general, is to find high-affinity hapten-(aflatoxin-) specific monoclonal IgG antibodies. In this study, microplate-based hybridoma screenings were carried out and compared under different conditions. The screening was accomplished using both a direct and an indirect non-competitive ELISA format. The following criteria were considered for evaluating the screening methods. High-affinity MAbs must be detected as positive in a reliable way, whereas MAbs of low or no affinity should be classified as negative, thus resulting in a good overall selectivity. The over-reporting (false positives) and under-reporting (false negatives) should be minimal. False negative means the negative detection of an interesting MAb and false positive means the positive detection of an uninteresting MAb. A MAb was classified as interesting if its affinity measurement resulted in a minimal IC50 value below 1 μg/l (AFB2 concentration). A screening method which results in a considerable number of false positives will lead to a lot of unnecessary work, as the affinities of uninteresting MAbs have to be measured and the corresponding clones are possibly recloned and cultured to no purpose.

189

One way to avoid a high number of false positives is to perform simultaneous non-competitive/competitive indirect ELISAs for all cell culture supernatants as described elsewhere (Abad and Montoya, 1994; Mercader and Montoya, 1999; Moreno et al., 2001; Manclús et al., 2004). The criterion for the selection of high-affinity MAbs is then the ratio of the two absorbances measured (non-competitive/competitive). However, this means a doubling of measurements. Thus, our strategy was to develop a non-competitive screening with optimal selectivity. Competitive ELISAs were then performed only for the positive MAbs. As the MAb concentration in a supernatant depends on the hybridoma cell line and its growth state (and thus is usually unknown), it is also important to have a screening method that leads to useful results relatively independent of the MAb concentrations. This is important because the ELISA parameters can hardly be optimized prior to the screening, since the screening itself often consists of only a single (one-well) measurement. Fig. 2 summarizes the screening results. The screening was carried out for a set of 373 supernatants. It is important to have both positive and negative controls on each microplate. Sufficient color development for the positive controls indicates that the immunoassay is working in general. If no reference MAb is available, the polyclonal mouse serum from the corresponding immunization can be used. Negative controls can generally be used to calculate a cut-off value in order to classify signals as positive or negative. The value of this cut-off is somewhat arbitrary. We suggest calculating a cut-off that is dependent on the precision of the screening, e.g. the median of all negative controls (e.g. 20 replicates) plus ten times the standard deviation. Alternatively, another cut-off was set at 15% of the maximum signal intensity without consideration of the precision of the screening (Fig. 2). In general, acceptable results can be achieved with both the indirect and the direct non-competitive ELISA screening methods (Fig. 2c, d, and e). However, the conventional indirect non-competitive ELISA screening (which is usually the standard method) can lead to a high number of false positives (we observed up to about 50% of false positives among all 373 supernatants tested), particularly if the MAb concentration is relatively high (Fig. 2a), which may lead to screening results that are all but useless. In fact, these screenings did not lead to acceptable results for any cut-off, as the signal intensities of the high-affinity MAbs were often below the false positives. However, the modified indirect screening with pre-incubation of the supernatants in

190

C. Cervino et al. / Journal of Immunological Methods 329 (2008) 184–193

anti-mouse IgG-coated microplates led to very good screening results (Fig. 2c). This modified indirect screening also proved to be superior for other hybridoma screenings in our hands. The improvement of the overall selectivity is likely to be caused by the

decrease of all MAb concentrations as a result of the additional pre-incubation step. However, this cannot explain changes of relative selectivities, e.g. clone 62 being more intense than clone 287 in Fig. 2b but less intense in Fig. 2c. These changes might be caused by

Fig. 2. Screening results for a set of 373 supernatants. Two standard indirect screenings (a, b), one modified indirect screening (c), and two direct screenings (d, e) were performed. Standard indirect ELISAs were carried out at a 1/10,000 dilution of coating conjugate (AFB2-CMO-TG) and supernatant dilutions of 1/10 (a) and 1/100 (b). For the modified indirect screening (c) the applied amount of antibody was reduced by initial preincubation of the 1/10 diluted supernatants in anti-mouse IgG coated microplates. Direct screenings were carried out at a 1/10,000 dilution of enzyme tracer (AFB2-CMO-POD) and supernatant dilutions of 1/10 (d) and 1/100 (e). Absorbance measurements were carried out after 2 min (a), 3 min (b), 10 min (c, d), and 15 min (e) (after having added substrate solution) at 650 nm. The horizontal lines mark possible (arbitrary) thresholds classifying signals as positive or negative. One threshold was set at the median of all negative controls from one screening (20 replicates) plus ten times the standard deviation (solid line), the other at 15% of the relative signal intensity (broken line). High-affinity MAbs are marked as solid triangles and are numbered; (uninteresting) MAbs (of low affinity) are marked as empty squares. The affinities of all other MAbs were not tested (small squares).

C. Cervino et al. / Journal of Immunological Methods 329 (2008) 184–193

191

Fig. 2 (continued ).

different affinities of the capture antibody to IgG1 and IgG2. For the direct screenings, at most 3 out of 373 samples were tested as false positive for two different supernatant dilutions (Fig. 2d and e). Thus, the overall selectivity was generally very good for the direct assay. The three screenings with good overall selectivities (Fig. 2c, d, and e) all clearly indicate three of the five interesting MAbs (62, 195, and 287). However, individual high-affinity MAbs such as MAb 357 might be overlooked (false negative) using the direct screening at high supernatant dilutions (Fig. 2e) compared to the modified indirect screening (Fig. 2c). In our experience, the sensitivity of the direct screening is generally somewhat lower compared to the indirect one. Nevertheless, the direct assay can be very useful because of the good overall selectivity obtained. MAb 54 was not detected positive by any of the three useful screenings, as its concentration in the supernatant was extremely low (the corresponding cell line was unstable and thus not interesting). The differences in relative color intensities (from supernatant to supernatant) between the indirect and direct immunoassays may be explained as follows. For

the indirect ELISA, AFB2-CMO-TG (a “polyhapten”) was immobilized. Due to the bivalency of IgG molecules, double bonding of the test MAb is possible (Plückthun and Pack, 1997, and references therein). This effect, also referred to as the “bonus effect of multivalency” (Roitt and Delves, 2006), is not helpful for the hapten-specific MAb screening, because the bonding strength of polyvalent antibody-antigen complexes is much higher compared to what the screening results should mirror, i.e. the affinity of one single IgG binding region to a small molecule. In other words, it might be likely that low-affinity MAbs bind to the polyhapten-coated surface. This may be one of the reasons for considerable over-reporting of the standard indirect ELISA screening (Fig. 2a and b). Such effects of bivalent binding may be important to a lesser extent for the direct ELISA, where a non-immobilized haptenenzyme tracer is used. There are at most only three accessible amino groups (where hapten is coupled) available on the same side of the POD molecule (Berglund et al., 2002) and, therefore, simultaneous binding of the IgG to two conjugated haptens might be sterically impossible.

192

C. Cervino et al. / Journal of Immunological Methods 329 (2008) 184–193

The lower sensitivity of the direct screening compared to the indirect one may be somewhat dependent on the capture antibody (anti-mouse IgG). Selectivity differences for individual high-affinity MAbs (here MAb 357) comparing the direct and indirect screening might be caused by excess of another, non-haptenspecific antibody in the same supernatant, thus leading to “IgG competition” in the (direct) anti-mouse IgGcoated microplates. Hence, some high-affinity MAbs might be overlooked using only the direct screening, even if the antibody concentrations in the supernatants and thus the overall sensitivity is sufficiently high. 4. Concluding remarks For the generation of hapten-specific MAbs, we recommend the direct ELISA screening (Fig. 2d) and the modified indirect ELISA screening method (Fig. 2c). We found these methods to be highly useful for the efficient detection of high-affinity MAbs. The number of false positives was very low (b1%) for the direct screening compared to the standard indirect screening, which can lead to a large number of false positives if the MAb concentrations are too high. Similar results comparing the two screening assays were also obtained for the highly selective detection of triazine herbicide-specific MAbs at our institute (Winklmair et al., 1997; Weller, 1992). However, a disadvantage of the direct screening is the lower sensitivity compared to the indirect method. Thus, high-affinity MAbs may be overlooked. In order to make sure that the number of false negatives is minimal, while maintaining a good overall selectivity, we suggest accomplishing the direct screening using low-diluted (or undiluted) hybridoma cell supernatants and/or the modified indirect screening. These two screening methods can also be conveniently combined by transferring the supernatants that have been pre-incubated in the (direct) antimouse IgG-coated microplates, into the (indirect) microplates that are coated with coating-antigen and blocked. Most of the research on various non-microplatebased hybridoma screening techniques concentrate on the standard indirect immunoassay format. However, it would be interesting or even preferable to develop similar MAb screening techniques using the direct immunoassay format or the modified indirect assay, especially for the generation of hapten-specific MAbs. Acknowledgements We thank Dr Michael G. Weller from the Federal Institute for Materials Research and Testing (BAM,

Berlin) for fruitful discussions. We would also like to thank Cédrik E.C. Schöne and Rita Medek for valuable assistance in the lab. References Abad, A., Montoya, A., 1994. Production of monoclonal antibodies for carbaryl from a hapten preserving the carbamate group. J. Agric. Food Chem. 42, 1818. Berglund, G.I., Carlsson, G.H., Smith, A.T., Szöke, H., Henriksen, A., Hajdu, J., 2002. The catalytic pathway of horseradish peroxidase at high resolution. Nature 417, 463. Burrin, J., Newman, D.J., 1991. Production and assessment of antibodies. In: Price, C.P., Newman, D.J. (Eds.), Principles and Practice of Immunoassay. Stockton Press, New York, p. 19. Canziani, G.A., Klakamp, S., Myszka, D.G., 2004. Kinetic screening of antibodies from crude hybridoma samples using Biacore. Anal. Biochem. 325, 301. Cervino, C., Knopp, D., Weller, M.G., Niessner, R., 2007. Novel aflatoxin derivatives and protein conjugates. Molecules 12, 641. Cho, Y.-J., Lee, D.-H., Kim, D.-O., Min, W.-K., Bong, K.-T., Lee, G.-G., Seo, J.-H., 2005. Production of a monoclonal antibody against ochratoxin A and its application to immunochromatographic assay. J. Agric. Food Chem. 53, 8447. Daigo, K., Sugita, S., Mochizuki, Y., Iwanari, H., Hiraishi, K., Miyano, K., Kodama, T., Hamakubo, T., 2006. A simple hybridoma screening method for high-affinity monoclonal antibodies using the signal ratio obtained from time-resolved fluorescence assay. Anal. Biochem. 351, 219. Eaton, D.L., Groopman, J.D. (Eds.), 1994. The Toxicology of Aflatoxins. Human Health, Veterinary, and Agricultural Significance. Academic Press, San Diego. Food and Agriculture Organization, 2004. Worldwide regulations for mycotoxins in food and feed 2003. FAO Food and Nutr. Paper 81. Grol, M., Schulze, M., 1990. Immunoassays für lösliche Antigene: Ein Überblick. In: Peters, J.H., Baumgarten, H. (Eds.), Monoklonale Antikörper. Springer, Berlin/Heidelberg, p. 321. Hack, R., Märtlbauer, E., Terplan, G., 1987. A monoclonal antibody to the trichotecene T-2 toxin: screening for the antibody by a direct enzyme immunoassay. J. Vet. Med. 34, 538. Hastings, K.L., Hagler, W.M., Harris, T.M., Voyksner, R.D., Dean, J.H., 1989. Production and characterization of aflatoxin B2 oximinoacetate. J. Agric. Food Chem. 36, 404. Kane, M.M., Banks, J.N., 2000. Raising antibodies. In: Gosling, J.P. (Ed.), Immunoassays - a Practical Approach. Oxford University Press, Oxford, p. 19. Köhler, G., Milstein, C., 1975. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256, 495. Manclús, J.J., Abad, A., Lebedev, M.Y., Mojarrad, F., Micková, B., Mercader, J.V., Primo, J., Miranda, M.A., Montoya, A., 2004. Development of a monoclonal immunoassay selective for chlorinated cyclodiene insecticides. J. Agric. Food Chem. 52, 2776. Mangler, J., Weller, M.G., Weil, L., Niessner, R., Hämmerle, H., Schlosshauer, B., 1994. New monoclonal antibodies to triazine herbicides. Fresenius J. Anal. Chem. 349, 346. Matschulat, D., Deng, A., Niessner, R., Knopp, D., 2005. Development of a highly sensitive monoclonal antibody based ELISA for detection of benzo[a]pyrene in potable water. Analyst 130, 1078. Mercader, J.V., Montoya, A., 1999. Development of monoclonal ELISAs for azinphos-methyl. 1. Hapten synthesis and antibody production. J. Agric. Food Chem. 47, 1276.

C. Cervino et al. / Journal of Immunological Methods 329 (2008) 184–193 Moreno, M.J., Abad, A., Montoya, A., 2001. Production of monoclonal antibodies to the N-methylcarbamate pesticide propoxur. J. Agric. Food Chem. 49, 72. Nesheim, S., Trucksess, M.W., Page, S.W., 1999. Molar absorptivities of aflatoxins B1, B2, G1, and G2 in acetonitrile, methanol, and tolueneacetonitrile (9+ 1) (modification of AOAC official method 971.22): collaborative study. J. Assoc. Off. Anal. Chem. Int. 82, 251. Plückthun, A., Pack, P., 1997. New protein engineering approaches to multivalent and bispecific antibody fragments. Immunotechnology 3, 83. Roitt, I.M., Delves, P.J., 2006. Roitt's Essential Immunology. Blackwell Publishing, Malden, MA. Sasaki, K., Glass, T.R., Ohmura, N., 2005. Validation of accuracy of enzyme-linked immunosorbent assay in hybridoma screening and proposal of an improved screening method. Anal. Chem. 77, 1933.

193

Schetters, H., 1993. Screening. In: Masseyeff, R.F., Albert, W.H., Staines, N.A. (Eds.), Methods of Immunological Analysis, Vol. 2. VCH, Weinheim, p. 246. Weller, M.G., 1992. Strukturelle und kinetische Untersuchungen zur Entwicklung und Optimierung von Hapten-Enzymimmunoassays (ELISAs) am Beispiel der Bestimmung von Triazinherbiziden. Dissertation, Technische Universität München. Winklmair, M., Weller, M.G., Mangler, J., Schlosshauer, B., Niessner, R., 1997. Development of a highly sensitive enzyme-immunoassay for the determination of triazine herbicides. Fresenius J. Anal. Chem. 358, 614.