Screening and Optimization of Peptides

Short bioactive peptides have emerged as a new trend in drug research, serving as vaccines, diagnostic reagents, therapeutics, and drug leads. There are multiple approaches to screen for bioactive peptides, including phage-display peptide libraries, randomized synthetic peptide libraries, antisense (complementary) peptide libraries, peptides derived from protein degradation (enzymatic hydrolysis or chemical cleavage), MHC–peptide complexes, and strategies based on protein structures/structure prediction.

Overview of Peptide Screening Methods

Phage-display peptide libraries are among the most mature screening technologies. Typically, ~3 rounds of biopanning yield preferred phage clones; sequencing directly reveals the displayed peptide sequences, which are then synthesized in a targeted manner to validate bioactivity. Limitations include library diversity being susceptible to various confounding factors and the often modest affinity of primary hits, necessitating subsequent optimization.

Randomized synthetic peptide libraries can be customized to experimental needs, usually offering short production cycles and large library capacities. For example, a 10-mer peptide library has a theoretical diversity of 20^10 ≈ 1 × 10^13; however, the practically accessible effective library size is constrained by synthesis and screening formats. Two common synthetic strategies are:

• Split-and-mix: tends to yield a more uniform representation of different sequences; each resin bead typically carries a single peptide species.
• Pre-mix (amino-acid premixing): can result in substantial differences in sequence abundance, introducing representation bias that may compromise screening outcomes. Given the enormous library sizes, structural identification (sequence determination) of active peptides discovered during screening can be considerably challenging.

Antisense (complementary) peptide libraries are rationally designed, target-focused libraries that require prior identification of the key amino-acid region(s) in the target protein. Bioinformatics tools such as Clustal X can be used to perform sequence alignments and pinpoint these critical regions. Once identified, antisense libraries can be designed according to the Mekler–Idlis (M–I) theory. They can be synthesized directly for screening, or first virtually down-selected using software (e.g., Discovery Studio) to obtain a smaller candidate set. Because the target is well defined and the library is relatively small (for a 10-mer, on the order of 2^10 ≈ 10^3), this approach facilitates hit discovery and sequence confirmation.

After identifying a target peptide, additional focused peptide libraries can be constructed to map functional regions and further enrich for more potent peptides.

Overview of Sequence Optimization Methods

For peptides already obtained, sequence optimization can be performed using library designs such as single-site scanning or multi-site randomized scanning to generate variants with enhanced activity.

• Single-site scanning employs a straightforward design: each selected position (or region) is systematically substituted with other amino acids. This helps reveal residues that critically influence activity or physicochemical properties.
• Multi-site randomized scanning is more complex: using the 20 natural amino acids in a “shotgun” fashion to simultaneously and randomly substitute selected residues, constructing as many combinatorial variants as possible at the chosen positions to identify sequences with improved activity. The library size and screening effort can be substantial—for example, simultaneous variation at three positions yields 20^3 = 8,000 theoretical combinations. Employing the proprietary PeptideGo™ combinatorial chemistry strategy can markedly reduce the number of syntheses and the screening burden; for a three-position scan, the practical synthesis/screening scale can be reduced to ~800 variants (depending on the amino-acid set chosen), which also enables simultaneous scanning across more positions.

 Aladdin：https://www.aladdinsci.com/