Novel approaches for producing purified active recombinant human proteases, and the inability of dipeptidyl peptidase-4 to bind SARS-CoV-2 spike protein

Background and Aim: Proteases catalyze irreversible post-translational modifications that often alter biological function of the substrate The protease dipeptidyl peptidase-4 (DPP-4) is a pharmacological target in type 2 diabetes and is abundant in liver DPP-4 and its sister protease, fibroblast activation protein (FAP), are potential pharmacological targets in steatosis, insulin resistance, cancers, and inflammatory and fibrotic diseases Recently, DPP-4 has been identified as a potential binding target of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike protein 1,2 A detailed understanding of protein structure, function, and molecular interactions requires a reliable protocol for the large-scale production of highly stable and pure protein We aimed to optimize production and purification strategies for soluble recombinant human DPP-4 and FAP from Spodoptera frugiperda 9 (Sf9) insect cells Methods: The Bac-to-Bac Baculovirus Expression System (ThermoFisher) was used Soluble DPP-4 (residues 29-766) was purified by a four-step procedure: differential ammonium sulfate precipitation, hydrophobic interaction chromatography, dye affinity chromatography in tandem with immobilized metal affinity chromatography, and ion exchange chromatography Soluble FAP (residues 27-760) was expressed similarly, except for adding a gp67 secretion signal, and then purified similarly The binding affinities of DPP-4 to the SARS-CoV-2 full-length spike protein and its receptor binding domain were measured using surface plasmon resonance Results: This optimized DPP-4 purification procedure yielded, on average, 1 36 mg of pure fully active soluble DPP-4 protein per liter of suspension insect cell culture with specific activity 36 4 U/mg DPP-4 activity greater than 20 U/mg indicates that the enzyme is completely pure No specific binding between DPP-4 and CoV-2 protein was detected For FAP, a novel baculovirus expression construct was designed and used to generate abundant active soluble recombinant human FAP expression in Sf9 insect cells for protein purification Unexpectedly, FAP and DPP-4 behaved differently in hydrophobic interaction chromatography Nevertheless, soluble FAP was partially purified using a similar purification protocol to DPP-4 Compared with our previous work,3 the DPP-4 now has greater yield and purity, and the FAP has greater yield Conclusion: A procedure for high-yield DPP-4 that is purified to homogeneity was achieved and used to show that, unlike in Middle East respiratory syndrome (MERS), SARS-CoV-2 does not bind DPP-4 A better understanding of FAP behavior in liquid chromatography was obtained and will be useful for developing an optimized purification strategy