Peptide reagents can serve as alternatives or replacements to antibodies in sensing or diagnostic applications. The passive adsorption of peptides onto polystyrene surfaces can limit the target binding capability, especially for short, positively charged, or hydrophobic sequences. In this report, we show that fusing a peptide with a previously characterized 12-amino acid polystyrene binding sequence (PS-tag) improves overall peptide solubility and enzyme-linked immunosorbent assay (ELISA) results using the peptide as a capture agent. Specific improvements for protective antigen (PA; Bacillus anthracis) protein binding peptides selected from bacterial surface display were compared with native or biotinylated peptides. The PS-tag was added to either peptide terminus, using a (Gly)₄ spacer, and comparable binding affinities were obtained. Fusion with the PS-tag did not have any negative impact on peptide secondary structure as measured by circular dichroism. The addition of the PS-tag provides a convenient method to utilize peptide reagents from peptide display libraries as capture agents in an ELISA format without the need for a biotin tag or concerns about passive adsorption of critical residues for target capture.

The number of peptide affinity reagents as antibody alternatives, therapeutics, and diagnostic agents has increased with the use of peptide display techniques, such as bacterial display (1-5), yeast display (6,7), phage display (4,5, 8-10), and synthetic libraries (11). Peptide reagents have shown great potential, particularly since they can be rapidly isolated, in as little as 2–4 days with bacterial display libraries (3), and can be mass produced synthetically in large quantities, making them an attractive alternative to antibodies. Recently, peptides have exhibited comparable results to antibodies and other detection assays (12,13), especially as an alternative diagnostic tool and low-cost fieldable test (14).

Currently, there are very few instances in the literature detailing enzyme-linked immunosorbent assays (ELISAs) that specifically use peptides as capture agents. Poor peptide performance in direct ELISAs is often attributed to peptide length, especially when shorter than 20 residues. Short polypeptide lengths limit the number of sidechains available for both adsorption and target recognition, restrict the number of available peptide conformations for target binding, and impact molar density during adsorption because of potential solubility issues, all of which could contribute to poorer target recognition in peptide ELISAs (15-17). There are a number of alternatives for peptide adsorption for direct ELISA. Direct covalent attachment of peptides to a modified plate surface (18,19) or using a biotinylated peptide with a streptavidin-or neutravidin-modified plate (20) can improve peptide adsorption and accessibility at the plate surface compared with direct peptide adsorption, but both methods require surface-modified plates, ultimately increasing assay cost. Indirect ELISA and competition ELISA (21-23) are also used in peptide analysis by ELISA, but require additional assays steps, such as the use of sandwich formats, enzyme-labeled secondary antibodies, or protein-peptide fusions, such as peptides covalently linked to BSA (24). The addition of a polystyrene surface binding sequence is expected to increase the adsorption of peptides to unmodified, polystyrene surfaces, for an improved, direct ELISA method using the peptides as capture agents.

Extending the N terminus of peptide sequences with a plate binding region has been shown to increase the surface coverage of peptides and improve the capture ability and function of peptides, as was previously shown with a (Lys)₇ N-terminal extension (25). Recently, Kumada, et al., identified a 12-amino acid polystyrene binding sequence (PS-tag) using the Flitrx random peptide display library that could be used for improved hydrophilic polystyrene plate binding of fusion proteins (26). This PS-tag (PS19–1; RAFIASRRIRRP) and PS19–6 (RIIIRRIR), derived from the original PS19–1 sequence, are able to maintain polystyrene binding capacity in the presence of detergents, such as Tween-20, and BSA used during adsorption steps (27). Both the PS19–1 and PS19–6 have high affinity for polystyrene, 169 and 86 nM respectively, and show similar maximum adsorption, 81 and 86 mg/m² respectively, with the PS19–1 sequence being more hydrophilic overall, -0.792, compared with the PS19–6, -0.500 (28). This PS-tag modification has been successfully demonstrated for use with antibodies (29), glutathione S-transferase (GST) (26,30), and Escherichia coli cysteine synthase complex (31). Since the measured affinity and adsorptions for the PS19–1 and PS19–6 are so similar, the PS19–1 PS-tag is more likely to improve both peptide solubility (more hydrophilic), while introducing a high-affinity polystyrene binding sequence for the first use of the PS-tag appended to a peptide sequence for an improved method for direct ELISA using the peptide as a capture agent.

Peptide reagents from display libraries are typically 15 amino acids or less, which can present solubility issues, thereby making it challenging for simple peptide-for-antibody substitutions in immunoassays. As the peptide chain length approaches 10 or more residues in length, the relative solubility decreases (32), especially for very hydrophobic sequences or sequences prone to adopt secondary structures in solution (33,34). Peptide sequences that can adopt and maintain a random coil structure at long lengths (32), such as short sequences of five residues or less or sequences that have a mix of hydrophobic and hydrophilic residues, would not have solubility issues in aqueous buffers. The overall net peptide charge determined by the number of acidic and basic residues (also accounting for the N and C termini charges), as well as the average hydropathy (35), or ratio of hydrophobic to hydrophilic residues, are good indicators of the aqueous stability and solubility of peptides. During affinity selection of these peptides, solubility is less problematic, since the sequence is coexpressed as a randomized region on the microbe surface, typically through a membrane protein in bacteria and yeast or coat protein in phage (4-6,36), which masks any potential solubility issues of the synthetic peptide reagent.