HCDR3 Antibody Library Generation via Natural Distribution Biasing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for constructing synthetic antibody libraries struggle to achieve the diversity and functionality of natural human antibody H-CDR3 regions, particularly in mimicking the natural amino acid distribution and length diversity found in human antibodies, leading to a high concentration of non-functional binders and cumbersome screening processes.
Innovation Solution
A collection of diverse human or humanized antibody H-CDR3 regions is generated by applying different diversity factors to varying ranges of amino acid lengths, mimicking the natural amino acid distribution by biasing the random distribution of amino acids in the H-CDR3 encoding DNA sequence, allowing for the construction of libraries that were previously deemed impossible to construct.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If randomization of H-CDR3 region with degenerate primers is used to generate sequence diversity, then enormous sequence diversity is achieved, but many structures are distorted and non-functional
Solution Approach 1:
The patent applies parameter changes by modifying the amino acid composition at different positions within the H-CDR3 region based on empirically determined diversity factors. Instead of uniform randomization, the method uses position-specific amino acid frequencies and length distributions derived from natural human antibodies to generate sequences that maintain structural integrity and functionality while achieving diverse antigen binding specificities.
Solution Approach 2:
The patent implements local quality by applying different diversity factors to different positions within the H-CDR3 region. Each position has its own amino acid frequency distribution based on natural antibody data, allowing certain positions to be more variable while others maintain stricter constraints. This ensures that critical structural positions retain appropriate amino acids while other positions contribute to diversity.
2Reliability
If very large synthetic antibody libraries are constructed to obtain good affinities, then functional antibodies can be found, but library screening becomes cumbersome
Solution Approach 1:
The patent reduces library size requirements by changing the parameters of sequence generation to match natural antibody characteristics. By using empirically determined amino acid frequencies and length distributions from human antibodies, the method generates sequences with higher probability of functionality, thereby reducing the total library size needed to find high-affinity binders and simplifying subsequent screening processes.
Solution Approach 2:
The patent applies partial action by focusing diversity generation on the H-CDR3 region specifically, rather than randomizing the entire antibody sequence. By concentrating variability where it matters most (the antigen-binding CDR3 region) while keeping framework regions intact, the method achieves effective diversity with smaller, more manageable libraries that are easier to screen.
3Reliability
If focused H-CDR3 libraries are constructed with limited V, D and J segments to reduce non-functional binders, then functional quality improves, but diversity at every amino acid residue is not optimized
Solution Approach 1:
The patent resolves this contradiction by using parameter changes based on comprehensive analysis of natural human antibody H-CDR3 regions. The method determines empirical amino acid frequency distributions and length variations from diverse natural sequences, then applies these parameters to generate synthetic libraries that achieve both high functional quality (by mimicking natural sequences) and maximum amino acid diversity (by capturing the full range of natural variation).
Solution Approach 2:
The patent applies copying by replicating the statistical properties and sequence characteristics of natural human antibody H-CDR3 regions in synthetic libraries. Instead of using limited germline segments, the method copies the empirical distributions of amino acid frequencies, sequence lengths, and positional variations observed in natural antibodies, thereby achieving both functionality and diversity.
Data Source
AI summary
The present invention is directed to the preparation and use of a collection of antibody heavy chain complementarity determining region 3 (HCDR3) members, where diversity of the collection is a function of the length of the HCDR3 members. The diversity of the collection of HCDR3 regions substantially represents the natural amino acid distribution of HCDR3 in the human repertoire. This natural amino acid distribution can be represented by biasing the complete random distribution of amino acids, accordingly, in the HCDR3 encoding DNA sequence by using trinucleotide mutagenesis (TRIM) technology. A collection of HCDR3 members of the invention each can be comprised within a variable region of an antibody (or fragment thereof) to form a library of synthetic antibodies or antibody fragments. The invention also provides nucleic acid molecules encoding such diverse collection and methods of making and using the same.