Design Of Human Smaller Antibody Against TNF-α Based On Antibody Consensus Frameworks And Antagonist Peptides

As an important cytokine, Tumor Neurous Factor-α(TNF-α) involves in many physiological function. However, aberrant production of TNF-αhas been discovered to mediate or be involved in many pathological processes including rheumatoid arthritis (RA), Crohn's disease (CD), etc. Neutralization of TNF-αhas become an effective therapeutic strategy for these diseases. Several TNF-αblockers have been approved by the US FDA and used successfully in clinic: infliximab (Mouse-human chimeric mAb, registered as Remicade), adalimumab (human mAb, registered as Humira) and etanercept (fusion protein of TNFR II with Fc fragment of human IgG, registered as Enbrel). However, the use of these large molecules was limited by expensive production cost and potential side effects. It is very attractive and useful to explore novel TNF-αantagonists. Human smaller antibodies are of particular interest for their therapeutic value.In previous study, we obtained a TNF-αneutralizing mAb Z12 and identified the epitope (i.e. from 141 to 146 aa in TNF) recognized by Z12. Moreover, we explored a series of TNF-αantagonist peptides (i.e. PT2, PT3, PT4 and PT7) based on the interaction between TNF-αand mAb Z12. To improve the activity of the above-mentioned antagonist peptides, a domain antibody was designed using V_h5 framework of antibody variable region as scaffold to display antagonist peptides (i.e. PT2, PT3 and PT4). Results showed that the bioactivity was improved evidently and the human antibody variable region was good at displaying peptides.We attempt to design conformational library of scFV(abbr. CLscFv) based on human antibody consensus frameworks and the antagonist peptides(PT2, PT3, PT4 and PT7)using computer modeling method. Then, the novel anti-TNF scFv fragment will be selected and the activity will be tested by biological experiments.In the present study, based on the existing domain antibody PTVH5 and the most appropriate framework V_K1 (antagonist peptide PT7 replaced CDR3 in V_K1), a novel anti-TNF scFv (named as TSA1) was designed. Theoretical analysis showed that TSA1 was stable and recognized the position 141-146 of TNF-α. Biological experiments showed that TSA1 could bind to TNF-α, competitively inhibit the binding of mAb Z12 to TNF-α, block the binding of TNF-αto TNFR and inhibit TNF-induced cytotoxicity and signaling. The novel scFv fragment displayed significantly improved activity over the domain antibody. However, the activity of TSA1 was weaker than S-Remicade(the scFv form of Remicade).Further, to enhance the potency of the scFv fragment, the design was optimized by the selection of the frameworks and the location of the peptides. In the study of TSA1, HCDR3 was found to play the key role in the binding with TNF-α. Consequently, HCDR3 was replaced by the best peptide (PT7). HCDR1, HCDR2 or LCDR3 was replaced by other peptides (PT2, PT3 and PT4) randomly. Based on this, CLscFv was constructed using computer modeling method according to the seven V_h and seven V_l frameworks. Theoretical analysis indicated that the interaction between TSA2 (based on frameworks V_H5 and V_λ1, where HCDR1, HCDR2, HCDR3 and LCDR3 were replaced by PT2, PT3, PT7 and PT4, respectively.) and TNF-αwas the strongest in all scFv fragments. TSA2 was better than TSA1 at interacting domain, binding free energy and intermolecular hydrogen bond. The biological experiments demonstrated the ability of TSA2 to bind with TNF-αand block the deleterious biological effects of TNF-α. The activity of TSA2 was significantly better than that of TSA1. Moreover, the activity of TSA2 was similar to that of S-Remicade.This study demonstrated that it is feasible to design CLscFv and select novel antagonists based on human antibody consensus frameworks and antagonist peptides. It also provided an alternative way to obtain human smaller antibody.