Optimization of primer sequences for mouse scFv repertoire display library construction.

In immunoglobulin repertoire library cloning, the homology between a particular primer sequence and that of its target template, and the diversity capacity of a primer pool are the two most important parameters which determine the cloning efficiency and the size of a resulting repertoire library. In this report we present an optimized set of primers, compatible with a commercially available vector, for mouse scFv repertoire phage display library construction. We used the INMUNO-ZAP Fab cloning system (1) in an initial effort to clone monoclonal hybridomas for functional characterizations of the various domains of human protooncogene ETS1 and ETS2 by intracellular targeting. In our hands, however, the Immuno-ZAP compatible primers were only effective in PCR-cloning of the heavy chain of U-244 (an ETS2-specific MAb) (2) and the light chains of T7 (an ETS2-specific MAb) (2) and E44 (an ETS 1-specific MAb) (3), but not their partner chains. This led us to try a scFv cloning system (4), which had been demonstrated effective in the construction of a human naive antibody repertoire display library (5). To develop a set of compatible primers for mouse antibody repertoire display library construction, we pooled and lined up the mouse antibody sequences collected in the Kabat-Wu data book (6). Similar sequences were grouped and a putative primer sequence was drawn from each group. All the putative primer sequences were then each compared against all the pooled sequences with the aid of a VAX computer and the best-fit primer sequences were selected. This process was repeated until all the sequences were covered. Flanking sequences compatible with the pHENl phagemid vector (4) and a linker (see notes for Table 1) were finally added to the VH and VL primers so that cloned VH or VL fragments can be used either in primary repertoire cloning or in affinity improvement by chain shuffling (7). Computer-aided sequence homology analysis showed that the ten resulting MHV.BACK VH repertoire-cloning 5' primers cover 55% of the 147 listed mouse heavy chain sequences at 100% homology, 84% at > 95%, 97% at > 90% and 99% at > 86%. In contrast, the eight 5' primers (H1-H8) from Immuno-ZAP system (1) only cover 58.5% of the listed mouse VH sequences at a homology level equal or greater than 82%. The nine resulting MKV.BACK VL kappa chain specific 5' primers cover 67% listed sequences at 100% homology, 84% at > 95%, 91% at > 90% and 99% at > 86%. We also designed a single MLV.BACK 5' primer for mouse lambda chain cloning, although lambda light chains constitute only less than 5% of the mouse repertoire antibodies (8). To maximize the possibility to reach the size of a naive mouse Ig repertoire, we further developed four MHV.FOR VH cloning 3' primers, four VL kappa chain 3' primers (three MKV.FOR primers derived from the VL kappa chain FR4 3' sequences and one MKC5.FOR primer from the CL 5' sequences) and one MLV.FOR VL lambda chain 3' primer. The MHV.FOR primers cover 94% of the listed VH 3' sequences at 100% match and 100% at 95% match. The three FR4 MKV.FOR primers cover 66% of the listed VL kappa chain 3' sequences with 100% homology, 97% at 95% match and the single MKC5.FOR primer