Periplasmic targeting of fluorobodies

Abstract

Molecular evolution is a key strategy in the development of novel diagnostics and therapeutics. Libraries of mutant variants are generated and those with desired properties selected after screening. However, large scale screening is expensive and time-consuming and has become a bottleneck. The fluorescent properties of the green fluorescent protein (GFP) are used in quantitative and qualitative screening. An attractive approach is to take advantage of the fact that a correlation exists between the solubility of an upstream fusion partner to GFP and observed fluorescence when the fusion protein is expressed in Escherichia coli (E. coli). The T-cell receptor (TCR) is a molecule with therapeutic and diagnostic potential. However, the single chain TCR (scTCR) shows toxic and aggregation-prone characteristics that may be alleviated by molecular evolution. When expressed in E. coli, scTCR and antibody (Ab) based molecules such as single chain fragment variable (scFv), need the oxidizing environment of the periplasm to form disulfide bonds and obtain a functional fold. The two scFvs, with specificity for two haptens phOx and NIP, are expressed in the periplasm at high and moderate levels, respectively. In this study, the scTCR as well as the NIP and phOx binders were genetically fused to GFP and three different mechanisms for translocation from the cytosol to the periplasm were compared. The results show that the two scFv fusions with pelB leader were transported to the periplasm, but cleaved in the linker region. The same result was obtained for anti-phOx-GFP with TorA leader. Neither over-expression of the chaperones GroEL-GroES nor FkpA, expression in host strains compensating for codon bias or lacking proteases, expression without induction of fusion proteins nor reducing the expression temperature to 16°C resulted in expression of full length fusion protein in the periplasm. The scTCR-GFP fusion was not transported to the periplasm with the use of neither pelB, DsbA nor TorA leader. Neither over-expression of the chaperones GroEL-GroES or FkpA, nor expression in a host strain compensating for codon bias resulted in periplasmic yield of fusion protein. In conclusion, the expression profiles of the scFvs were not preserved when expressed as fusions to GFP. The scTCR-GFP fusions were not transported to the periplasm, regardless of leader sequence. Further optimization of expression system or construct design is needed to obtain a screening platform linking functional fold to observed fluorescence.