Combinatorially restricted computational design of protein-protein interfaces to produce IgG heterodimers

Redesigning protein-protein interfaces is an important tool for developing therapeutic strategies. Interfaces can be redesigned by in silico screening, which allows for efficient sampling of a large protein space before experimental validation. However, computational costs limit the number of combinations that can be reasonably sampled. Here, we present combinatorial tyrosine (Y)/serine (S) selection (combYSelect), a computational approach combining in silico determination of the change in binding free energy (ΔΔG) of an interface with a highly restricted library composed of just two amino acids, tyrosine and serine. We used combYSelect to design two immunoglobulin G (IgG) heterodimers-combYSelect1 (L368S/D399Y-K409S/T411Y) and combYSelect2 (D399Y/K447S-K409S/T411Y)-that exhibit near-optimal heterodimerization, without affecting IgG stability or function. We solved the crystal structures of these heterodimers and found that dynamic π-stacking interactions and polar contacts drive preferential heterodimeric interactions. Finally, we demonstrated the utility of our combYSelect heterodimers by engineering both a bispecific antibody and a cytokine trap for two unique therapeutic applications.

Nab3 nuclear granule accumulation is driven by respiratory capacity

Numerous biological processes involve proteins capable of transiently assembling into subcellular compartments necessary for cellular functions. One process is the RNA polymerase II transcription cycle which involves initiation, elongation, co-transcriptional modification of nascent RNA, and termination. The essential yeast transcription termination factor Nab3 is required for termination of small non-coding RNAs and accumulates into a compact nuclear granule upon glucose removal. Nab3 nuclear granule accumulation varies in penetrance across yeast strains and a higher Nab3 granule accumulation phenotype is associated with petite strains, suggesting a possible ATP-dependent mechanism for granule disassembly. Here, we demonstrate the uncoupling of mitochondrial oxidative phosphorylation by drug treatment or deletions of nuclear-encoded ATP synthase subunit genes were sufficient to increase Nab3 granule accumulation and led to an inability to proliferate during prolonged glucose deprivation, which requires respiration. Additionally, by enriching for respiration competent cells from a petite-prone strain, we generated a low granule-accumulating strain from a relatively high one, providing another link between respiratory competency and Nab3 granules. Consistent with the resulting idea that ATP is involved in granule accumulation, the addition of extracellular ATP to semi-permeabilized cells was sufficient to reduce Nab3 granule accumulation. Deleting the SKY1 gene, which encodes a kinase that phosphorylates nuclear SR repeat-containing proteins and is involved in efficient stress granule disassembly, also resulted in increased granule accumulation. This observation implicates Sky1 in Nab3 granule biogenesis. Taken together, these findings suggest there is normally an equilibrium between termination factor granule assembly and disassembly mediated by ATP-requiring nuclear machinery.

High-throughput identification of putative receptors for cancer-binding peptides using biopanning and microarray analysis

Phage-display peptide biopanning has been successfully used to identify cancer-targeting peptides in multiple models. For cancer-binding peptides, identification of the peptide receptor is necessary to demonstrate the mechanism of action and to further optimize specificity and target binding. The process of receptor identification can be slow and some peptides may turn out to bind ubiquitous proteins not suitable for further drug development. In this report, we describe a high-throughput method for screening a large number of peptides in parallel to identify peptide receptors, which we have termed "reverse biopanning." Peptides can then be selected for further development based on their receptor. To demonstrate this method, we screened a library of 39 peptides previously identified in our laboratory to bind specifically to cancers after irradiation. The reverse biopanning process identified 2 peptides, RKFLMTTRYSRV and KTAKKNVFFCSV, as candidate ligands for the protein tax interacting protein 1 (TIP-1), a protein previously identified in our laboratory to be expressed in tumors and upregulated after exposure to ionizing radiation. We used computational modeling as the initial method for rapid validation of peptide-TIP-1 binding. Pseudo-binding energies were calculated to be -360.645 kcal mol(-1), -487.239 kcal mol(-1), and -595.328 kcal mol(-1) for HVGGSSV, TTRYSRV, and NVFFCSV respectively, suggesting that the peptides would have at least similar, if not stronger, binding to TIP-1 compared to the known TIP-1 binding peptide HVGGSSV. We validated peptide binding in vitro using electrophoretic mobility shift assay, which showed strong binding of RKFLMTTRYSRV and the truncated form TTRYSRV. This method allows for the identification of many peptide receptors and subsequent selection of peptides for further drug development based on the peptide receptor.