Anticalins Lipocalins Reshaped via Combinatorial Biotechnology

Asp mentioned above and Lys Ser, which was introduced in order to remove a proteolytic cleavage site [39]. Thus, there were altogether four fixed amino acid replacements in addition to fhe randomized side chains. The mutagenized gene cassette was fhen inserted into an appropriate E. coli vector and a genetic library comprising 3.7x10 variants was prepared [40]. 

Fluorescein, a well known immunological hapten [43] with many applications in biochemistry and biophysics and a collection of commercially available derivatives, served as fhe prescribed ligand for BBP variants in the first selection study. The phagemid random library was used for panning on a plastic surface coated with a covalent conjugate of fluorescein with bovine serum albumin (BSA). After six cycles of adsorption, acid elution, and phagemid re-amplification, fhe specific enrichment of a mutant phagemid fraction was observed. 

The BBP variant FluA was subjected to detailed biochemical characterization. The engineered lipocalin could be produced at high yield in the periplasm of E. coli (9.1 mg per 2 L culture compared with 1.2 mg for wild-type BBP) and isolated to homogeneity in one step via the Strep-tag method [45]. According to the relative shift in electrophoretic mobility between oxidized and reduced state of the pro- 

Interestingly, when the titration was performed such that the hapten s own characteristic fluorescence was measured, almost complete quenching was observed. From this very accurate titration experiment, a K, value of 35.2 3.2nM was determined for the FluA fluorescein complex [40]. Using the same spectroscopic effect, the association kinetics between fluorescein and FluA could be measured by rapid mixing, yielding a Kon value of 5.28 0.05x10 M s (G. Beste and A. Skerra, unpublished). 

Clearly, this spectroscopic phenomenon was a serendipitous event because there was no corresponding selection applied during the generation of FluA. Also, the other mutants that were selected along with it did not show fluorescence quenching to the same high extent. Nevertheless, it is remarkable that such an efficient electron transfer process, which is even faster (by a factor 3-4) than the one measured between bacteriochlorophyll and bacteriopheophytin in the bacterial reac- 

---