Group A flavin-dependent monooxygenases catalyze the cleavage of the oxygen–oxygen bond of dioxygen, followed by the incorporation of one oxygen atom into the substrate molecule with the aid of NADPH and FAD. These flavoenzymes play an important role in many biological processes, and their most distinct structural feature is the choreographed motions of flavin, which typically adopts two distinct conformations (OUT and IN) to fulfill its function. Notably, these enzymes seem to have evolved a delicate control system to avoid the futile cycle of NADPH oxidation and FAD reduction in the absence of substrate, but the molecular basis of this system remains elusive. Using protein crystallography, size-exclusion chromatography coupled to multi-angle light scattering (SEC-MALS), and small-angle X-ray scattering (SEC-SAXS) and activity assay, we report here a structural and biochemical characterization of PieE, a member of the Group A flavin-dependent monooxygenases involved in the biosynthesis of the antibiotic piericidin A1. This analysis revealed that PieE forms a unique hexamer. Moreover, we found, to the best of our knowledge for the first time, that in addition to the classical OUT and IN conformations, FAD possesses a “sliding” conformation that exists in between the OUT and IN conformations. This observation sheds light on the underlying mechanism of how the signal of substrate binding is transmitted to the FAD-binding site to efficiently initiate NADPH binding and FAD reduction. Our findings bridge a gap currently missing in the orchestrated order of chemical events catalyzed by this important class of enzymes.

Solution stability is an important factor in the optimization of engineered biotherapeutic candidates such as monoclonal antibodies because of its possible effects on manufacturability, pharmacology, efficacy and safety. A detailed atomic understanding of the mechanisms governing self-association of natively folded protein monomers is required to devise predictive tools to guide screening and re-engineering along the drug development pipeline. We investigated pairs of affinity-matured full-size antibodies and observed drastically different propensities to aggregate from variants differing by a single amino-acid. Biophysical testing showed that antigen-binding fragments (Fabs) from the aggregating antibodies also reversibly associated with equilibrium dissociation constants in the low-micromolar range. Crystal structures (PDB accession codes 6MXR, 6MXS, 6MY4, 6MY5) and bottom-up hydrogen-exchange mass spectrometry revealed that Fab self-association occurs in a symmetric mode that involves the antigen complementarity-determining regions. Subtle local conformational changes incurred upon point mutation of monomeric variants foster formation of complementary polar interactions and hydrophobic contacts to generate a dimeric Fab interface. Testing of popular in silico tools generally indicated low reliabilities for predicting the aggregation propensities observed. A structure-aggregation data set is provided here in order to stimulate further improvements of in silico tools for prediction of native aggregation. Incorporation of intermolecular docking, conformational flexibility, and short-range packing interactions may all be necessary features of the ideal algorithm.

Tryptic hydrolysis of whey protein isolate was performed simultaneously during (in situ) and before (ex situ) fractionation by electrodialysis with ultrafiltration membrane (EDUF) to obtain bioactive peptides. Peptide migration to anionic (ARC –) and cationic (CRC +) peptide recovery compartments was strongly dependent on the digestion strategy used. Indeed, peptide migration to the ARC – was observed to be higher with in situ digestion while peptide migration to the CRC + was higher in an ex situ digestion: a final peptide concentration of 103.10 ± 2.76 μg/mL was found in the CRC + (ex situ) while it was 49.65 ± 6.13 μg/mL in the ARC – (in situ). HPLC-MS studies showed 23 major peaks that were generated by tryptic digestion of whey protein isolate. Seven of these peptides migrated to the ARC – while nine and eight peptides migrated to the CRC + for ex situ and in situ digestions, respectively. Among them, different antihypertensive, antimicrobial, and hypocholesterolemic peptides were recovered depending on the recovery compartment such as IDALNENK and VYVEELKP­TPEGDLEI­LLQK in ARC – and ALPMHIR, TKIPAVFK, VLVLDTDY­KK, and VAGTWY in CRC + regardless of the mode of enzymatic hydrolysis.

No abstract

The design of superior biologic therapeutics, including antibodies and engineered proteins, involves optimizing their specific ability to bind to disease-related molecular targets. Previously, we developed and applied the Assisted Design of Antibody and Protein Therapeutics (ADAPT) platform for virtual affinity maturation of antibodies (Vivcharuk et al. in PLoS One 12(7):e0181490, 10.1371/journal.pone.0181490, 2017). However, ADAPT is limited to point mutations of hot-spot residues in existing CDR loops. In this study, we explore the possibility of wholesale replacement of the entire H3 loop with no restriction to maintain the parental loop length. This complements other currently published studies that sample replacements for the CDR loops L1, L2, L3, H1 and H2. Given the immense sequence space theoretically available to H3, we focused on the virtual grafting of over 5000 human germline-derived H3 sequences from the IGMT/LIGM database increasing the diversity of the sequence space when compared to using crystalized H3 loop sequences. H3 loop conformations are generated and scored to identify optimized H3 sequences. Experimental testing of high-ranking H3 sequences grafted into the framework of the bH1 antibody against human VEGF-A led to the discovery of multiple hits, some of which had similar or better affinities relative to the parental antibody. In over 75% of the tested designs, the re-designed H3 loop contributed favorably to overall binding affinity. The hits also demonstrated good developability attributes such as high thermal stability and no aggregation. Crystal structures of select re-designed H3 variants were solved and indicated that although some deviations from predicted structures were seen in the more solvent accessible regions of the H3 loop, they did not significantly affect predicted affinity scores.