Hydrogen bond rotations as a uniform structural tool for analyzing protein architecture

Abstract

Proteins fold into three-dimensional structures, which determine their diverse functions. The conformation of the backbone of each structure is locally at each C a effectively described by conformational angles resulting in Ramachandran plots. These, however, do not describe the conformations around hydrogen bonds, which can be non-local along the backbone and are of major importance for protein structure. Here, we introduce the spatial rotation between hydrogen bonded peptide planes as a new descriptor for protein structure locally around a hydrogen bond. Strikingly, this rotational descriptor sampled over high-quality structures from the protein data base (PDB) concentrates into 30 localized clusters, some of which correlate to the common secondary structures and others to more special motifs, yet generally providing a unifying systematic classification of local structure around protein hydrogen bonds. It further provides a uniform vocabulary for comparison of protein structure near hydrogen bonds even between bonds in different proteins without alignment.

Structural, functional and molecular dynamics analysis of cathepsin B gene SNPs associated with tropical calcific pancreatitis, a rare disease of tropics

“…The free energy (ΔΔ G ) calculations of mutated proteins by varied algorithms indicated that mutations are destabilizing. Further, the comparative analysis of H-bond distances between mutated and native 3D-structure of procathepsin B provided a unique information about the structural characteristics of motifs around main chain H-bonds which are altered in mutant protein, thereby affecting the function of protein (37). Additionally, MD simulation of mutated and native protein indicated that the mutations distinctly deviate the structural confirmation of procathepsin B , which can have deleterious effects in downstream signalling mechanism.…”

Moduli spaces and macromolecules

“…The utility of this will take some time yet to confirm, but the phenomenon of clustering and the attendant classification it entails are so robust that they are plainly visible to the naked eye already in the raw data in Figure 10. Moreover, the same basic clusters occur for other data sets extracted from the PDB using various cutoffs of primary sequence identity and PDB file quality; see the original paper [104] for details.…”

“…This is not our key result on protein; rather, using SO(3)-graph connections, we have discovered [104] new geometric constraints on proteins as we shall explain effectively passing from topology in [103] to geometry in [104]. The marvelous monograph [44] provides the broader view of protein structure and dynamics.…”

“…The resulting solution again lies in a region roughly 30 percent of the volume of SO(3) nicely roughly agreeing with that already discovered albeit without the fine detail of the 30 clusters. We refer the interested reader to the original paper [104].…”

“…These subsets are separately grouped into clusters according to methods discussed in [104] leading to 30 regions in SO (3). H-bonds that are short-range along the backbone, i.e., those with |Δ| ≤ 6, have of course been well studied in terms of the backbone conformational angles ϕ, ψ.…”