Previously, we published an article providing an overview of the Rosetta suite of biomacromolecular modeling software and a series of step-by-step tutorials [Kaufmann, K. W., et al. (2010) Biochemistry 49, 2987–2998]. The overwhelming positive response to this publication we received motivates us to here share the next iteration of these tutorials that feature de novo folding, comparative modeling, loop construction, protein docking, small molecule docking, and protein design. This updated and expanded set of tutorials is needed, as since 2010 Rosetta has been fully redesigned into an object-oriented protein modeling program Rosetta3. Notable improvements include a substantially improved energy function, an XML-like language termed “RosettaScripts” for flexibly specifying modeling task, new analysis tools, the addition of the TopologyBroker to control conformational sampling, and support for multiple templates in comparative modeling. Rosetta’s ability to model systems with symmetric proteins, membrane proteins, noncanonical amino acids, and RNA has also been greatly expanded and improved.

Structural restrictions are present even in the most sequence diverse portions of antibodies, the complementary determining region (CDR) loops. Previous studies identified robust rules that define canonical structures for five of the six CDR loops, however the heavy chain CDR 3 (HCDR3) defies standard classification attempts. The HCDR3 loop can be subdivided into two domains referred to as the “torso” and the “head” domains and two major families of canonical torso structures have been identified; the more prevalent “bulged” and less frequent “non-bulged” torsos. In the present study, we found that Rosetta loop modeling of 28 benchmark bulged HCDR3 loops is improved with knowledge-based structural restraints developed from available antibody crystal structures in the PDB. These restraints restrict the sampling space Rosetta searches in the torso domain, limiting the φ and ψ angles of these residues to conformations that have been experimentally observed. The application of these restraints in Rosetta result in more native-like structure sampling and improved score-based differentiation of native-like HCDR3 models, significantly improving our ability to model antibody HCDR3 loops.

Since 2011, over 300 human cases of infection, especially in exposed children, with the influenza A H3N2 variant (H3N2v) virus that circulates in swine in the US have been reported. The structural and genetic basis for the lack of protection against H3N2v induced by vaccines containing seasonal H3N2 antigens is poorly understood. We isolated 17 human monoclonal antibodies (mAbs) that neutralized H3N2v virus from subjects experimentally immunized with an H3N2v candidate vaccine. Six mAbs exhibited very potent neutralizing activity (IC50 < 200 ng/ml) against the H3N2v virus but not against current human H3N2 circulating strains. Fine epitope mapping and structural characterization of antigen-antibody complexes revealed that H3N2v specificity was attributable to amino acid polymorphisms in the 150-loop and the 190-helix antigenic sites on the hemagglutinin protein. H3N2v-specific antibodies also neutralized human H3N2 influenza strains naturally circulating between 1995 and 2005. These results reveal a high level of antigenic relatedness between the swine H3N2v virus and previously circulating human strains, consistent with the fact that early human H3 seasonal strains entered the porcine population in the 1990s and reentered the human population, where they had not been circulating, as H3N2v about a decade later. The data also explain the increased susceptibility to H3N2v viruses in young children, who lack prior exposure to human seasonal strains from the 1990s.

Background Anti-Jo-1 autoantibodies which recognize histidyl-tRNA synthetase identify patients with the rare rheumatologic disease, anti-histidyl-tRNA synthetase syndrome (Jo-1 ARS), a phenotypically distinct subset of idiopathic inflammatory myopathies (IIM). Jo-1-binding B cells (JBCs) are implicated in disease pathogenesis, yet they have not been studied directly. We therefore aimed to characterize JBCs to better understand how they expand and function in Jo-1 ARS. Methods We enrolled 10 IIM patients diagnosed with Jo-1 ARS, 4 patients with non-Jo-1 IIM, and 8 age- and sex-matched healthy controls. We phenotypically characterized peripheral blood mononuclear cells (PBMCs) ex vivo using flow cytometry to define the B cell subsets in which JBCs reside. We further tested their ability to differentiate into antibody-secreting cells following stimulation in vitro. Results The majority of JBCs were IgM+ (not class-switched). Compared to non-JBCs in the same donors, JBCs contained a higher percentage of autoimmune-prone CD21lo cells and were increased in the CD21lo IgM+ IgD− CD27+ memory subset relative to healthy donor B cells. Whereas non-JBCs were present in the anergic BND B cell subset, JBCs were nearly absent from this compartment. JBCs were detected among plasmablasts in some donors, but a reduced frequency of JBCs differentiated into CD38hi24− plasmablasts compared to non-JBCs present in the same wells following in vitro stimulation. Conclusions JBCs are enriched for autoimmune-prone CD21lo B cells, some of which exhibit a memory phenotype in the peripheral repertoire of Jo-1 ARS patients. JBCs undergo limited class switch and show reduced capacity to differentiate into antibody-secreting cells. This suggests complex B cell biology exists beyond class-switched cells that differentiate to secrete anti-Jo-1 autoantibody (i.e., what is captured through serum autoantibody studies). New Jo-1 ARS therapies should thus ideally target non-class-switched JBCs in addition to those that have undergone IgG class-switching to most effectively block cross-talk with autoreactive T cells.

Optimization of the heavy chain/light chain interface could serve as an important tool for maximizing antibody/antigen binding affinity without altering antigen contact residues.