Phage-displayed combinatorial peptide libraries in fusion to beta-lactamase as reporter for an accelerated clone screening: Potential uses of selected enzyme-linked affinity reagents in downstream applications

Abstract

Phage-display selection of combinatorial libraries is a powerful technique for identifying binding ligands against desired targets. Evaluation of target binding capacity of multiple clones recovered from phage display selection to a specific target is laborious, time-consuming, and a rate-limiting step. We constructed phage-display combinatorial peptide libraries in fusion with a beta-lactamase enzyme, which acts as a reporter. Linear dodecapeptide and cysteine-constrained decapeptide libraries were created at the amino-terminus of the Enterobacter cloacae P99 cephalosporinase molecule (P99 beta-lactamase). The overall and positional diversity of amino acids in both libraries was similar to other phage-display systems. The libraries were selected against the extracellular domain of ErbB2 receptor (ErbB2(ECD)). The target-selected clones were already conjugated to an enzyme reporter, therefore, did not require subcloning or any other post-panning modifications. We used beta-lactamase enzyme activity-based assays for sample normalizations and clone binding evaluation. Clones were identified that bound to purified ErbB2(ECD) and ErbB2-overexpressing cell-lines. The peptide sequences of the selected binding clones shared significant motifs with several rationally designed peptide mimetics and phage-display derived peptides that have been reported to bind ErbB2(ECD). beta-Lactamase fusion to peptides saved time and resources otherwise required by the phage-ELISA of a typical phage display screening protocol. The beta-lactamase enzyme assay protocols is a one-step process that does not require secondary proteins, several steps of lengthy incubations, or washings and can be finished in a few minutes instead of hours. The clone screening protocol can be adopted for a high throughput platform. Target-specific beta-lactamase-linked affinity reagents selected by this procedure can be produced in bulk, purified, and used, without any modification, for a variety of downstream applications, including targeted prodrug therapy.

Fig. (1). Schematic presentation of phagemid vector design and features

The linear dodecapeptide (X₁₂) and cysteine-constrained decapeptide (CX₁₀C) libraries were created at the N-terminal position of Enterobacter cloacae P99 cephalosporinase (BLA) molecule, between the signal peptide and enzyme protein. The random sequences of these libraries were linked to the N-terminal end of the β-lactamase molecule with a short linker (GGGS). Restriction sites SpeI (3326–3231 bp) and AgeI (3256–3261 bp) were used for cloning. The selected site was randomized using the nnk-scheme (n= a, t, c, or g; k=g or t). The N-terminal presence of pIII signal peptide helped in periplasmic translocation of BLA. Chloramphenicol acetyltransferase (CAT) and BLA provided antibiotic resistance to transformed host bacteria in presence of chloramphenicol and cefotaxime, respectively. This vector in suppressor host TG1 E. coli with the help of helper phage (KM13)-generated phage particles expressing BLA-linker-random peptide as an N-terminal fusion protein to phage coat pIII. A protease-cleavable linker between pIII protein of phage particle and BLA was used for trypsin elution of phage following panning. The transformation of non-suppressor strain Top10 E. coli, which recognizes the amber stop (tag) between the BLA and protease-cleavable linker, with the vector allowed the expression of free BLA protein with a random peptide library at the N-terminal end, and FLAG and 6xHis tags at the C-terminal end. The magnified insert region given at the lower left of the figure is inversed in relation to the vector because the insert oligonucleotide and the gene sequences are written 5′– 3′.

Fig. (2). Frequency of occurrence of different amino acids in the random regions of peptide libraries

Observed frequencies of amino acid occurrence in both the libraries were compared with their expected frequencies based on the availability of codons for a given amino acid. Observed frequencies were calculated in the following manner: the number of times a given amino acid appeared in the analyzed random inserts (n) ÷ the total count of amino acids (n × 10 or 12). Expected frequencies were calculated based on 32 codons available in the nnk-scheme of cloning. Amino acids are represented as single-letter codes.

Fig. (3). Positional amino acid diversity in peptide libraries

The diversity of different amino acids at each position (1 through 10 or 12) in the randomized regions of both linear dodecapeptide (X₁₂) and cysteine-constrained decapeptide (CX₁₀C) libraries as presented is a statistical measure of the proportion of the 20 possible amino acids that are observed at a given position. If a position is populated equally by all the 20 amino acids, the score is 1.0 (20/20). The data represent positional diversity in the occurrence of amino acids ± SD. Amino acids are represented as single-letter codes.

Fig. (4). Screening of clones for their binding to extracellular domain of ErbB2 (ErbB2-ECD)

Randomly selected 50 clones following the third round of panning were screened for their binding to purified ErbB2-ECD and bovine serum albumin (BSA) proteins. BSA was used as a negative control protein. The screening was based on assaying the protein-bound β-lactamase activities. The clones expressing wild-type β-lactamase (Wt-BLA; without fusion peptide) and the unselected library were used as negative clone controls. The bars represent β-lactamase activity as the change in relative fluorescence unit (RFU)/min ± SE. *p≤0.05, significance of difference in the clone binding to ErbB2-ECD and BSA were determined using Student’s T-test.

Fig. (5). Binding of selected ErbB2^ECDpositive clones to ErbB2 over-expressing cells

The clones that were found to bind purified ErbB2^ECD were studied for their binding to ErbB2 over-expressing MCF-7/ErbB2 and SK-BR-3 cell-lines. The MCF-7 cell-line that is known to exhibit minimal ErbB2 expression was used for negative control cells. The screening was based on assaying the cell-bound β-lactamase activities. The clones expressing wild-type β-lactamase (Wt-BLA; without fusion peptide) and unselected library were used as negative clone controls. The bars represent β-lactamase activity as change in relative fluorescence unit (RFU)/min ± SE. *p≤0.05, significance of difference in the binding of clones to MCF-7 in comparison to MCF-7/ErbB2 or SK-BR-3 cells were determined using Student’s T-test. ErbB2^ECD, extracellular domain of ErbB2.

Fig. (6). Fluorescence imaging of ErbB2 over-expressing cells for their binding to ErbB2^ECDpositive clones

The figure shows binding of representative clones 113 and 150 to MCF-7/ErbB2 and SK-BR-3 cells, which are known to over-express the ErbB2 receptor. High intensities of fluorescence are seen in both MCF-7/ErbB2 and SK-BR-3 cells in comparison to minimally expressive MCF-7 cells. Negative controls of wild-type β-lactamase (Wt-BLA) and unselected library clones did not show higher fluorescence in MCF-7/ErbB2 and SK-BR-3 cells. The precipitating green fluorescence is produced by the β-lactamase activity associated with clones.