Molecular imaging would benefit from efficient discovery of novel reagents. Antibody size (150 kDa) impedes biodistribution and imaging at early time points. Conversely, affibody is a small (7 kDa), stable, three-helix bundle protein with great potential as an alternative scaffold. Yet, inefficient preclinical development – identifying lead clones and improving stability – hinder progress. These challenges can be overcome by improved combinatorial library design for directed evolution. Building on our recent work with hydrophilic fibronectin domains (Woldring, et al, 2015), we hypothesize that libraries benefit from a broad range of diversities with sitewise unique amino acid distributions. To test this in the context of affibody ligand evolution, an ensemble of starting points was generated whereby various levels of amino acid diversity were positioned throughout the paratope. Strong binders (≥ 5nM Kd) were selected against multiple targets and showed tolerable stability (Tm&gt;60°C). High-throughput sequencing identified &gt;63k unique proteins. Custom software, quantifying sitewise and epistatic trends, provided evidence for sitewise amino acid preferences. Shannon entropies within the evolved paratope sites range from 2.9 to 3.6 – aligning well with the hypothesis (see Figure 1). To confirm the observed sitewise trends, a second library was designed and selected against an expanded cohort of targets. Paratope sites were diversified based on contributions from parallel studies, namely: binder sequences; stability calculations (ΔΔGFolding) conducted in FoldX; natural homologs; and solvent accessibility. Upon high-throughput sequence analysis of these evolved clones, the sitewise diversity hypothesis will be further evaluated to quantify the accuracy of the second generation affibody design as well as the impact of each library design component. Engineered ligands for multiple cancer biomarkers will be discussed.

Rethinking affibody: Evolving small proteins using data-driven diversification

251st National Meeting (2016)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.