In Vitro Affinity Maturation Of An Anti-TNFα Antibody

In our early work , we cloned the variable genes of an anti-human TNF a monoclonal antibody from hybridoma cells secreting the antibody, and expressed its ScFv and Fab in E Coli. In order to improve the affinity of the anti-TNFa antibody , we tried to establish an effective in vitro affinity maturation procedure by investigating different mutation techniques with phage display technology. This procedure would be useful for improving affinity of other genetic engineered antibodies.Part one: Improving the affinity by random mutation. First, single chain Fv gene was mutated by two round error-prone PCR, and a mutated phage antibody library of 7X106 in size was constructed. The library was subjected to 4 round of selection with coated antigen TNFa, and seven different clones with slightly improved affinity were selected. In order to further improve the affinity of the mutants , two kinds of DNA recombination methods were used. Firstly seven selected clones were recombined by staggered extension process and a mutant library was constructed. After biopanning as above, two clones with further improved affinity were identified. Secondly on the basis of error-prone PCR generated mutation, a DNA shuffling procedure was performed to achieve a random exchange of the mutation spots and a library of 1. 12X107 was constructed. Through selection as before one clone with better affinity was identified.Part two: Improving the affinity by CDR3 mutagenesis. We first tried parsimonious mutagenesis of the CDR3. The oligonucleotides used for mutagenizing CDR3 of Fab were synthesized by two methods: one way was tomaintain 70% parental sequence bias for each codon within CDR3 by controlling the proportion of the 4 mixed bases during oligonucleotides synthesis; the other way was to maintain a 50% parental sequence bias for each codon by a reported codon-based mutagenesis procedure. As the H-CDR3 contains 10 residues, which is too long for one time mutation, it was separated into two parts , each contained 5 residues and was mutated respectively. Six phage antibody libraries containing L-CDR3 or H-CDR3 mutants were generated by by overlap extension PCR and were selected by biopanning. Mutants with improved affinity were sequenced and their affinity constants were determined by non-competitive ELISA. One specific mutant was selected from k mutated library with a 28-fold higher Ka. Ten specific mutants were selected from Fd mutated libraries with 2 ~ 65-fold higher Ka. To further improve affinities of the identified mutants, the mutated k and Fd genes and different mutated residues within H-CDR3 were recombined. The resulted mutants had further improved affinities than their original clones. Then we tried to mutate the hot spots within CDR3. According to the hot spots of L-CDR3 and H-CDR3, oligonucleotide primers containing ramdomized hot spots were synthesized. The mutated CDR3s were inserted into scFv gene by overlap extension PCR and a library of 8X104 was established. By biopanning selection one clone with significantly improved affinity was identified. In this work we tried 5 mutagenesis strategies to improve the affinity of antibody. Through Error-prone PCR we did not achieve obvious improvement of affinity of the antibody. Further recombination procedures of the mutated genes by DNA shuffling or staggered extension PCR resulted in clones with further improved affinities. Mutagenesis of CDR3 was successful in getting mutants with obviously improved affinities. Through mutating the hot spots within CDR3, we got one mutant with higher affinity, while by parsimonious mutagenesis of CDR3, more affinity improved mutantswere obtained. And by combine the mutation a mutant with 93 folds higher affinity was obtained. These results suggested that parsimonious mutagenesis of CDR3 seems to be the better strategy for in vitro affinity maturation. Thus our work, in addition to generating high affinity anti-TNFa antibodies, provided a useful starting procedure for in vitro antibody affinity maturation...