Standard Protocols for the Construction of scFv Libraries

Abstract

Since the generation of the first human antibodies (Abs) by phage display (1,2), technology has evolved to allow the creation of large, nonimmunized fully human scFv repertoires that yield Abs with comparable affinities to those obtained using hybridoma technology (3,4). Using a variety of selection and screening strategies, the same single-pot library can be used to simultaneously derive many high-affinity Abs with different specificities in only a few weeks. Abs isolated from such large, fully human scFv repertoires have a multitude of applications, from immunological reagents for enzyme-linked immunosorbant assay, immunocytochemistry, Western blotting, or epitope mapping, to therapy. The first fully human therapeutic monoclonal Abs isolated from a phagedisplayed library for the treatment of rheumatoid arthritis and ocular scarring are currently progressing through late-stage clinical trials (5). Construction of such scFv libraries from naturally rearranged V genes in a phagemid vector ensures natural diversity in the length of the V_H CDR3, a higher number of functional scFvs, and soluble scFv expression without the need for subcloning (3). Furthermore, phage-display technology also provides a means by which a selected Ab can, if necessary, be affinity-matured for improved neutralization potency or binding kinetics. Maximum diversity is generated by amplifying V genes from peripheral blood lymphocytes (PBL) or lymphoid tissue isolated from several nonimmunized donors using polymerase chain reaction (PCR) primers that correspond to all known V_H, V_κ, and V_λ gene sequences. These are principally based on those published previously (6) with further information on the most recent V_H sequences obtained from the V-BASE directory (Tomlinson et al., MRC Centre for Protein Engineering). To ensure that all five Ab classes are likely to be represented and increase the overall size of the final library, random hexamers are employed in the primary first-strand cDNA synthesis from PBL mRNA. Component V_H and V_L gene segments are amplified in separate PCR reactions, and initially cloned into two different vectors, pCANTAB6 and pCANTAB3hⁱs₆ (see Figure 1). The latter is used for cloning the V_L repertoire because it has the appropriate polylinker cloning sites for the digested V_L fragments; the V_H repertoire is cloned into pCANTAB6. A short linker from an existing scFv is cloned (together with an irrelevant or “dummy” V_H) into the V_L repertoire, upstream of the V_L fragments. The V_H and linker-V_L repertoires are then amplified from their vectors, and the scFv construct is prepared using a simple two-fragment PCR assembly procedure. This construct is then cloned into pCANTAB6 to create the large naïve scFv library (3).