An improved phage-display panning method to produce an HM-1 killer toxin anti-idiotypic antibody

Abstract

Background: Phage-display panning is an integral part of biomedical research. Regular panning methods are sometimes complicated by inefficient detachment of the captured phages from the antigen-coated solid supports, which prompted us to modify. Here, we produce an efficient antigen-specific single chain fragment variable (scFv) antibody by using a target-related molecule that favored selection of recombinant antibodies.

Results: To produce more selective and specific anti-idiotypic scFv-antibodies from a cDNA library, constructed from HM-1 killer toxin (HM-1)-neutralizing monoclonal antibodies (nmAb-KT), the method was modified by using an elution buffer supplemented with HM-1 that shares structural and functional similarities with the active site of the scFv antibody. Competitive binding of HM-1 to nmAb-KT allowed easy and quick dissociation of scFv-displayed phages from immobilized nmAb-KT to select specific anti-idiotypic scFv antibodies of HM-1. After modified panning, 80% clones (40/50) showed several times higher binding affinity to nmAb-KT than regular panning. The major populations (48%) of these clones (scFv K1) were genotypically same and had strong cytocidal activity against Saccharomyces and Candida species. The scFv K1 (K(d) value = 4.62 x 10(-8) M) had strong reactivity toward nmAb-KT, like HM-1 (K(d) value = 6.74 x 10(-9) M) as judged by SPR analysis.

Conclusion: The scFv antibodies generated after modified subtractive panning appear to have superior binding properties and cytocidal activity than regular panning. A simple modification of the elution condition in the phage-display panning protocol makes a large difference in determining success. Our method offers an attractive platform to discover potential therapeutic candidates.

Figure 1

Schematic representation of the phage-display selection procedure: modified subtractive panning. Phage particles displaying the members of the library were produced. The specific binder (nmAb-KT) was allowed to bind with the target, and other variants were removed by washing. Molecular variants with specificity for the target were retrieved after multiple cycles of selection and were characterized in detail. Bound phages were eluted with HM-1 containing phosphate buffer (pH 7.0). HM-1 had high binding affinity to the immobilized nmAb-KT. This competition of binding favored quick dissociation of scFv-containing phages from the bound nmAb-KT and consequently increased the elution stringency of the infected phages.

Figure 2

Enrichment of phage-displayed scFvs by panning. Phage-displayed scFvs were selected against nmAb-KT and were reinfected into E. coli TG1. The number of phages eluted after each round of panning was counted based on the number of colonies (CFU/ml) formed after reinfection of the eluted phage particles in the host bacteria. The results were means of triplicates.

Figure 3

Identification of positive clones to immobilized antigen. The binding properties of randomly selected clones to nmAb-KT were measured by microtiter plate ELISA. The supernatant containing soluble scFv antibodies was added to each well of a microtiter plate that had been pre-coated with nmAb-KT and detection was done by HRP-conjugated anti-E-tag antibody. A. Regular panning: 12 of 50 (24%) tested clones of infected E. coli TG1 contained the anti-idiotypic scFv gene of HM-1 killer toxin, and so soluble scFvs showed positive ELISA signals after the fifth round of panning. B. Modified subtractive panning: 40 of 50 (80%) tested clones showed positive ELISA signals after the fifth round of modified subtractive panning. Negative controls consisted of microtiter plates coated with blocking buffer (N-1) or washing buffer (N-2). The results were means of duplicates.

Figure 4

DNA fingerprint analysis of specific scFv genes of positive clones. scFv inserts were amplified from individual colonies of all positive clones. Amplification was done with Ex-Taq DNA polymearase enzyme using pCANTAB5-R1 and pCANTAB5-S6 primers. The amplified products were then analyzed on agarose gels. A. scFv encoding inserts of DNA of 40 diversified positive clones B. Only five restriction patterns of scFv encoding inserts of DNA were identified against five different clones (scFv K1 - K5). Molecular weights of DNA ladder markers (kbp) were indicated by M on the left.

Figure 5

Alignment of amino acid sequences of scFv K (1-5) antibody fragments. Amino acid sequences were deduced from the nucleotide sequences. Positions of the respective complementarity determining regions for the variable domains of heavy chain (CDR H1-3) and light chain (CDR L1-3) were indicated by bold symbols. Framework (Fw) and CDRs as defined by Kabat et al. [44] were indicated. (Fw 1 - 4) were framework parts of antibody domains. To compare with scFv K1, amino acid differences in the other four clones were shown by gray symbols.

Figure 6

SDS-PAGE and Western blot analysis of scFv expression. A. Supernatant containing the soluble scFv antibodies for 40 diversified positive clones underwent SDS-PAGE and then Western blotting. B. Molecular weights of purified soluble scFv antibodies (K1 - K5) were detected by SDS-PAGE and were stained with silver stain. C. Western blotting of duplicate SDS-polyacrylamide gels was detected by probing with HRP-conjugated anti E-tag antibody. Molecular weights of marker proteins (kDa) were indicated by M on the left.

Figure 7

Functional characterization of recombinant scFvs by measuring in vitro antifungal activity. A. Biological activity of purified five clones scFv (K1 - K5). The effect of recombinant scFvs on the growth of S. cerevisiae A451 cells was measured by a cell suspension assay as described in Methods. Gray bars: Yeast Growth (%); black bars: Protein at 280 nm. B. Elution profile for scFv purification by using an anti-E-tag Sepharose column and the biological activity of the selected clone (scFv K1) against the growth of S. cerevisiae A451. Purified scFv K1 proteins per fraction (fr) were detected and visualized by SDS-PAGE and Western blotting with HRP-antiE-tag conjugate. A clear band of the expected size (30 kDa) for E-tag containing scFv was present in the purified elution fractions. C. Biological activity of scFv K1 against C. albicans ATCC 10231. For all cases fungal growth was measured at 600 nm. The results were means of duplicates.

Figure 8

Competitive binding analysis of the ability of HM-1 to inhibit the binding of scFv K1 antibody to nmAb-KT. ELISA plates coated with nmAb-KT was incubated with increasing concentrations of the selected scFv K1 antibody as described in Methods. The results were means of duplicates.