Computational Prediction and Design for Creating Iteratively Larger Heterospecific Coiled Coil Sets.

A major biochemical goal is the ability to mimic nature in engineering highly specific protein-protein interactions (PPIs). We previously devised a computational interactome screen to identify eight peptides that form four heterospecific dimers despite 32 potential off-targets. To expand the speed and utility of our approach and the PPI toolkit, we have developed new software to derive much larger heterospecific sets (≥24 peptides) while directing against antiparallel off-targets. It works by predicting Tm values for every dimer on the basis of core, electrostatic, and helical propensity components. These guide interaction specificity, allowing heterospecific coiled coil (CC) sets to be incrementally assembled. Prediction accuracy is experimentally validated using circular dichroism and size exclusion chromatography. Thermal denaturation data from a 22-CC training set were used to improve software prediction accuracy and verified using a 136-CC test set consisting of eight predicted heterospecific dimers and 128 off-targets. The resulting software, qCIPA, individually now weighs core a-a' (II/NN/NI) and electrostatic g-e'+1 (EE/EK/KK) components. The expanded data set has resulted in emerging sequence context rules for otherwise energetically equivalent CCs; for example, introducing intrahelical electrostatic charge blocks generated increased stability for designed CCs while concomitantly decreasing the stability of off-target CCs. Coupled with increased prediction accuracy and speed, the approach can be applied to a wide range of downstream chemical and synthetic biology applications, in addition more generally to impose specificity on structurally unrelated PPIs.

Deriving Heterospecific Self-Assembling Protein-Protein Interactions Using a Computational Interactome Screen.

Interactions between naturally occurring proteins are highly specific, with protein-network imbalances associated with numerous diseases. For designed protein-protein interactions (PPIs), required specificity can be notoriously difficult to engineer. To accelerate this process, we have derived peptides that form heterospecific PPIs when combined. This is achieved using software that generates large virtual libraries of peptide sequences and searches within the resulting interactome for preferentially interacting peptides. To demonstrate feasibility, we have (i) generated 1536 peptide sequences based on the parallel dimeric coiled-coil motif and varied residues known to be important for stability and specificity, (ii) screened the 1,180,416 member interactome for predicted Tm values and (iii) used predicted Tm cutoff points to isolate eight peptides that form four heterospecific PPIs when combined. This required that all 32 hypothetical off-target interactions within the eight-peptide interactome be disfavoured and that the four desired interactions pair correctly. Lastly, we have verified the approach by characterising all 36 pairs within the interactome. In analysing the output, we hypothesised that several sequences are capable of adopting antiparallel orientations. We subsequently improved the software by removing sequences where doing so led to fully complementary electrostatic pairings. Our approach can be used to derive increasingly large and therefore complex sets of heterospecific PPIs with a wide range of potential downstream applications from disease modulation to the design of biomaterials and peptides in synthetic biology.