Objective:

The primary objective of this study is to rigorously assess the utility and effectiveness of two distinct methodologies for antibody affinity maturation. Specifically, the study aims to compare the traditional approach of Error-Prone PCR (epPCR) with Ranomics' combinatorial NNK codon mutagenesis targeted at the Complementarity Determining Regions (CDRs).

​

Methodologies:

• The study utilized yeast surface display to create eight yeast single-chain variable fragment (scFv) libraries for affinity maturation of antibodies targeting immuno-oncology markers PD-1, PD-L1, CTLA-4, and OX40.

• This involved employing both error-prone PCR (epPCR) and degenerate NNK combinatorial codon mutagenesis. The combinatorial codon mutagenesis was facilitated by the VariantFind platform from Ranomics Inc. VariantFind libraries contained 3–5 × 10e5 possible scFv variants, and the epPCR libraries theoretically as many as 10e10 potential variants.

• The yeast libraries were screened for improved affinity, and long-read high throughput sequencing (Pacific Biosciences) before and after sorting was used to accurately quantify the frequency of each amino acid mutation across the full heavy and light chain V(D)J sequences.

• The iterative optimization process, guided by sequencing data, aimed at refining the libraries and identifying promising antibody variants.

 

This methodology, integrating advanced mutagenesis, yeast display, and sequencing technologies, contributed to the in-depth exploration of antibody affinity maturation.

​

Results and Comparative Analysis

1. VariantFind NNK-directed Mutagenesis:

• No restrictions on spatial separation between mutated codons.

• Absence of strong positional biases.

• Occasional low presence of specific amino acids in given positions.

• Substantial representation of amino acids observed across all templates.

​

   2. Comparative Analysis with epPCR

• Similar patterns across all four targets.

• Slightly higher tendency for epPCR to incorporate negatively charged and polar side chains.

• More substantial differences for individual amino acids.

​

   3. Amino Acid Class Distribution:

• VariantFind reflects the expected NNK codon distribution.

• Not detectably influenced by the original amino acid.

​

   4. epPCR amino acid base

• Biases observed for particular amino acids.

• Likelihood of a derived amino acid codon influenced by the parental sequence's codon.

​

   5. Key Distinction:

• VariantFind lacks parental sequence codon influence on derived amino acid codons.

• Flexibility and unbiased representation distinguish VariantFind from epPCR mutagenesis.

​

Significance and Implications

Gigagen's "Affinity Maturation of Antibodies by Combinatorial Codon Mutagenesis versus Error-Prone PCR" is a seminal work that not only highlights the nuanced intricacies of two prominent techniques but also contributes significantly to the evolving field of antibody engineering.

​

​