One-Step Recovery of scFv Clones from High-Throughput Sequencing-Based Screening of Phage Display Libraries Challenged to Cells Expressing Native Claudin-1

Abstract

Expanding the availability of monoclonal antibodies interfering with hepatitis C virus infection of hepatocytes is an active field of investigation within medical biotechnologies, to prevent graft reinfection in patients subjected to liver transplantation and to overcome resistances elicited by novel antiviral drugs. In this paper, we describe a complete pipeline for screening of phage display libraries of human scFvs against native Claudin-1, a tight-junction protein involved in hepatitis C virus infection, expressed on the cell surface of human hepatocytes. To this aim, we implemented a high-throughput sequencing approach for library screening, followed by a simple and effective strategy to recover active binder clones from enriched sublibraries. The recovered clones were successfully converted to active immunoglobulins, thus demonstrating the effectiveness of the whole procedure. This novel approach can guarantee rapid and cheap isolation of antibodies for virtually any native antigen involved in human diseases, for therapeutic and/or diagnostic applications.

Figure 1

Library screening and analysis of sequences. (a) The panel shows the DNA fragments, gel-purified from sublibraries, after the indicated selection cycles. The corresponding plasmid preparations were digested with NcoI and XhoI restriction endonucleases, to release the DNA fragments encoding for the VH regions of the scFv fragments. The fragments were bar-coded and subjected to high-throughput sequencing, as described in the text. SM, size marker. (b) The chart reports the entropy values for the populations of fragments originating from the indicated selection cycles, after sequencing. (c) The reported values indicate the total number of clones, and the relative representation of the most abundant clone, within the corresponding selection cycles. (d) The chart indicates the relative distribution of clones, according to the number of counts observed, within the indicated ranges, for each of the 4 selection cycles. Cycles 3 and 4 show similar distributions.

Figure 2

Selection of scFv clones and strategy for recovery. (a) The chart reports the relative enrichments, within the indicated selection cycles, for 75 scFv clones. The threshold for inclusion was arbitrary set to a relative representation value of 1 × 10⁻³ (0.1%). Most clones were maximally enriched at cycle 3. Compared to cycle 4, cycle 3 also showed the highest number of different clones with a relative representation >1 × 10⁻³ (75 versus 63). Cycle 3 was accordingly selected for recovery of scFv clones. (b) The cartoon describes the strategy, implemented for recovery of scFv clones. The methylated template DNA from cycle 3 sublibrary was copied by Pfu DNA polymerase from overlapping primers (block arrows), corresponding to specific sequences within HCDR3 region of VH. The dashed lines represent the newly synthesized DNA, nonmethylated, since it was generated in vitro. After DpnI digestion, methylated and hemimethylated DNAs are removed, so that the nicked DNA originating from template copy is able to transform competent E. coli cells. The originating colonies thus represent the recovered, bona fide scFv clones. (c) The panel shows the products of the extension reactions, carried out on template from selection cycle 3, with overlapping primers for HCDR3 regions of clones 3_2, 3_5, and 3_67. The upper bands correspond to the supercoiled, methylated template; the lower bands represent the primer-extended, nicked products. SM, size marker; T, template DNA.

Figure 3

Evaluation of binding for scFv phages and IgG4. (a) The panel shows the binding of scFv phages, clones 3_2, 3_5, and 3_67, to HEK293 cells (gray bars) and to cells transduced with CLDN1 vector (HEK293-CLDN1, black bars). Clone 3_2 was discarded, because of nonspecific binding to HEK293-CLDN1 cells. (b) SDS-PAGE analysis of IgG4, converted from the scFv clones 3_5 and 3_67, as indicated. Samples in lanes 1 and 3 were run under nonreducing conditions, so that the whole IgG4 and the heavy-chain/light chain IgG4 dimers were accordingly visualized. Under reducing conditions, the IgG4 preparations showed the fully denatured light and heavy chains, as indicated. (c) The panels show the binding of the IgG4s, converted from scFv fragments 3_5 and 3_67, to HEK293 (gray lines) and to HEK293-CLDN1 (black lines) cells at increasing antibody concentrations.