Selection, Expression And Improvement Of Anti-Aflatoxin B₁ Single Chain Fragment Variable

Aflatoxins are a group of Aspergillus flavus metabolites that contaminate the food and feed supply due to their strong carcinogenicity. Among these aflatoxins, aflatoxin B1 (AFB1) is the most toxic. Dietary exposure to AFB1 is one of the major risk factors in the multi-factorial etiology of hepatocellular carcinoma (HCC), especially in populations exposed to hepatitis B virus (HBV). To effectively monitor the occurrence of AFB1 in food at low contamination levels, sensitive, reliable, and simple analytical methods are required. High performance liquid chromatograph (HPLC) and immunoassay are the most common methods for AFB1 detection in food and feed. High quality antibody is very important in the competitive enzyme-linked immunosorbent assay (ELISA). Antibody is also important in HPLC analysis, since immunoaffinity chromatograph (IAC) is often used for cleaning up the samples prior to HPLC analysis. Bypassing the immunization of animals, the recombinant antibody technology could produce low-cost antibodies with easy manipulation of affinity and specificity.In the study, we selected an anti-AFB1 single chain fragment variable (scFv-H4) with an improved screening method and the selected colon H4 was expressed in pET system. With homology molding, the 3-D structure of scFv-H4 was contructed. Referencing to the docking result between scFv-H4 and AFB1 and sequential diversity of Tomlinson I+J libaries, the affinity and the stability of scFv-H4 were improved.AFB1 was directly coated to Maxisorp plates and the Tomlinson I+J libraries were selected against these plats. Traditionally, the conjugation of haptens and carries might change antigenicity of haptens and the selection might be misled by immunodominant epitopes in carriers. These problems have been completely avoided in our methods. The competitive eluted phages were subseqentially treated by trypsin, which disable the phages without scFvs. Compared to the previously reports, our selection showed higher specificity for AFB1 (44 positives out of 46 random clones). In addition, polymerase chain chain-denaturing gradient gel electrophoresis (PCR-DGGE) was employed to describe the biopanning for the first time. The scFv-H4 with Ka of 9.8×10-11 L/mol to AFB1-BSA and Ka of 1.2×10-12 L/mol to AFB1 was sequenced. The scFv-H4 has cross reactivity of 12% with AFB2, 42% with AFG1, and 9% with AFG2.We employed four pET vectors (pET20b, pET22b, pET28a and pET32a) in the functional expression of scFv-H4 in Escherichia coli (E.coli). The comparison of functional scFv-H4 and its inclusion body indicated that pET20b and pET22b were able to express the functional antibody in the periplasm. The vector of pET28a seemed not suitable in the expression of scFv-H4 due to no detectable scFv-H4 in SDS-PAGE of total cell protein. There was only inclusion body detected in the expression with pET32. When expressed with pET22b, the amount of functional scFv-H4 was significantly affected by induction temperature, but not by inducer concentration. With the optimized conditions, E.coli BL21(DE3)/pET22b/H4 could express 35 mg/L cultural medium.The 3-D model of scFv-H4 was homologically built with WAM server. According to the evaluation of Ramachandran plot server, the orientation of the most amino acid residues in the model was allowed in the perspective of dihedral angles (ψ,φ), as well as fractional accessible surface area, fractional residue volume, stereo/packing quality index, the 3-D profile quality index. After the hard docking with Patchdock sever, the interaction between scFv-H4 and AFB1 was refined with Firedock sever in terms of side chains'orientation. The best solution was evaluatedd by Ligplot software and the hydrophobic interaction might be the dominant force between scFv-H4 and AFB1.Basing on the sequential diversity and the molecular docking, 18 sites of scFv-H4 (H50,H52,H52a,H53,H58,H95,H96,H97,H98,L50,L53,L91,L92,L93,L94 and L96) was saturated mutated in NNK mode. The mutants selected by affinity were sequenced. The results indicated that complementarity-determinning region H2 (CDR H2), CDR L2 were not evolved into the interaction between scFv-H4 and AFB1 while AL91 and HH95 might be necessary in the reorganizations. With CDR walking, an affinity-enhanced clone of A10 was obtained and its IC50 to AFB1 is 0.12 ng/mL. We stabilized scFv-H4 with interdomain disulphide bond and the effect of disulphide bond on affinity was studied. With homology modeling and molecular docking, we designed a scFv (dscFv-H4) containing interdomain disulphide bond between H44 and L100. The stability was increased from a GldHCl50 of 2.4 M for scFv-H4 to 4.2 M for dscFv-H4. The analysis with size exclusion chromatography revealed that dscFv-H4 existed primarily as monomer and no aggregates were detected. Competative ELISA indicated that scFv-H4 and dscFv-H4 had similar IC50 to AFB1.