Design of affinity tags for bare magnetic nanoparticles

In this project, peptide sequences that bind directly to unmodified magnetic nanoparticle (MNP) surfaces were designed. These peptides can be used as affinity tags for protein purification. Thereby, costly and degradable surface ligands commonly deployed in affinity systems for protein purification were avoided. This work included the identification of potential binders in peptide array experiments and the investigation of their adsorption and desorption behavior. Negatively charged peptides had the highest binding scores in buffers such as Tris buffered saline but showed almost no binding in phosphate and citrate based systems at the same pH. This strong buffer influence was explained by an adsorption of anionic buffer molecules onto the particles and thereby an alteration of the available surface evidenced by zeta potential measurements. The buffer effect was exploited to desorb MNP from negatively charged peptide spots. Positive peptides bound strongly to MNP but irreversibly. Glutamate based tags (E6, E4G4E4 and E8) and a (RH)4-tag were fused to green fluorescent protein (GFP) by molecular cloning. An untagged and a glycine tagged GFP variant were also produced and utilized as negative controls. The fusion proteins were chromatographically purified for further analysis. The difference in binding behavior indicated an exposed position of the tag. A combination of protein release by freeze-thaw cycles with anion exchange chromatography proved to be a very efficient purification protocol. The interaction behavior dependent on buffer conditions and tag composition anticipated from peptide array experiments could be confirmed for the tagged proteins in solution. As in the peptide array setup, E8-, E6-, and (RH)4-tagged protein variants had much higher static binding capacities on MNP than the negative controls G6-tagged or untagged GFP. In a competitive situation, the GFP-E6 load reached a maximum value of 0.17 +- 0.02 gxg-1 and was approximately 10 times higher compared to GFP-G6. GFP with negatively charged tags could be eluted with a buffer change to non-toxic and low-cost citrate buffered saline. The transfer to a bioseparation process for the purification of GFP-E6 from E. coli lysate was accomplished with a high-gradient magnetic separator. Without further optimizing, purities of 77% and yields of 82% could be obtained. The foundations of a new time- and cost-efficient purification process were therefore successfully laid.