In the abpc48 antibody presented earlier, Karl Proba and Arne Wörn also removed the VL disulfide bridge to obtain a completely disulfide-free scFV. By gene-shuffling (Stemmer mutagenesis) and phage panning they selected mutated scFV fragments which, by additional mutation, had regained sufficient stability to be functional, and indeed were as stable as the original scFV fragment.

Interestingly, most of the mutations isolated and shown to be beneficial to stability and/or folding efficiency do not affect the hydrophobic packing of the core, but affect more or less buried charged residues, turn forming residues and exposed hydrophobic residues