Accurate quantitation for in vitro refolding of single domain antibody fragments expressed as inclusion bodies by referring the concomitant expression of a soluble form in the periplasms of Escherichia coli.

Single domain antibody fragments from two species, a camel VHH (PM1) and a shark VNAR (A6), were derived from inclusion bodies of E. coli and refolded in vitro following three refolding recipes for comparing refolding efficiencies: three-step cold dialysis refolding (TCDR), one-step hot dialysis refolding (OHDR), and one-step cold dialysis refolding (OCDR), as these fragments were expressed as 'a soluble form' either in cytoplasm or periplasm, but the amount were much less than those expressed as 'an insoluble form (inclusion body)' in cytoplasm and periplasm. In order to verify the refolding efficiencies from inclusion bodies correctly, proteins purified from periplasmic soluble fractions were used as reference samples. These samples showed far-UV spectra of a typical ß-sheet-dominant structure in circular dichroism (CD) spectroscopy and so did the refolded samples as well. As the maximal magnitude of ellipticity in millidegrees (θmax) observed at a given wave length was proportional to the concentrations of the respective reference samples, we could draw linear regression lines for the magnitudes vs. sample concentrations. By using these lines, we measured the concentrations for the refolded PM1 and A6 samples purified from solubilized cytoplasmic insoluble fractions. The refolding efficiency of PM1 was almost 50% following TCDR and 40% and 30% following OHDR and OCDR, respectively, whereas the value of A6 was around 30% following TCDR, and out of bound for quantitation following the other two recipes. The ELISA curves, which were derived from the refolded samples, coincided better with those obtained from the reference samples after converting the values from the protein-concentrations at recovery to the ones of refolded proteins using recovery ratios, indicating that such a correction gives better results for the accurate measure of the ELISA curves than those without correction. Our method require constructing a dual expression system, expressed both in periplasm as a soluble form and cytoplasm as an insoluble form; application of the different refolding recipes due to sequence-by-sequence-difference could be precisely monitored using CD spectra with the concomitant soluble samples as a reference.