Uncovering allosteric pathways in caspase-1 using Markov transient analysis and multiscale community detection

Abstract

Allosteric regulation at distant sites is central to many cellular processes. In particular, allosteric sites in proteins are a major target to increase the range and selectivity of new drugs, and there is a need for methods capable of identifying intra-molecular signalling pathways leading to allosteric effects. Here, we use an atomistic graph-theoretical approach that exploits Markov transients to extract such pathways and exemplify our results in an important allosteric protein, caspase-1. Firstly, we use Markov Stability community detection to perform a multiscale analysis of the structure of caspase-1 which reveals that the active conformation has a weaker, less compartmentalised large-scale structure as compared to the inactive conformation, resulting in greater intra-protein coherence and signal propagation. We also carry out a full computational point mutagenesis and identify that only a few residues are critical to such structural coherence. Secondly, we characterise explicitly the transients of random walks originating at the active site and predict the location of a known allosteric site in this protein quantifying the contribution of individual bonds to the communication pathway between the active and allosteric sites. Several of the bonds we find have been shown experimentally to be functionally critical, but we also predict a number of as yet unidentified bonds which may contribute to the pathway. Our approach offers a computationally inexpensive method for the identification of allosteric sites and communication pathways in proteins using a fully atomistic description.

Prediction of allosteric sites and mediating interactions through bond-to-bond propensities

“…Caspase-1 is a tetramer composed of two asymmetric dimers, each containing one active site. From the Protein Data Bank (PDB) atomic structure (2HBQ), we constructed an atomistic, energy-weighted graph representation of the protein based on interaction potentials, as described in ‘Construction of the atomistic graph'3839. To quantify how strongly each bond is coupled to the active site, we calculate the propensities Π b for all bonds in the protein (equation (8)), and we aggregate the bond propensities over each residue to obtain the residue score Π R (equation (9)).…”

“…An in-depth discussion of the construction of the graph can be found in refs 38, 39, and further details are given in Supplementary Method 4. Briefly, we use an atomistic graph representation of a protein, where each node corresponds to an atom and the edges represent both covalent and non-covalent interactions, weighted by bond energies derived from detailed atomic potentials.…”

“…We start by building an atomistic graph model of the protein: nodes are atoms, and weighted edges represent both covalent bonds as well as non-covalent bonds (hydrogen bonds, salt bridges, hydrophobic tethers and electrostatic interactions), with weights derived from interatomic potentials (see the section ‘Construction of the atomistic graph' and refs 38, 39). The resulting all-atom graph is analysed using the edge-to-edge transfer matrix M , a discrete Green's function in the edge space of the graph recently introduced in ref.…”

Current Knowledge on Procaspase-1 Variants with Reduced or Abrogated Enzymatic Activity in Autoinflammatory Disease

“…Hence, a stable interface is essential for optimal enzymatic activity of caspase-1. Only recently the construction of a fully atomistic biophysical graph representation of the caspase-1 tetramer led to the detection that R240 (which is mutated in the p.R240Q variant) belongs to the allosteric network points [43].…”

Network Analysis and Modeling in Systems Biology

“…These reactions are organised in metabolic pathways that are composed of metabolites such as carbon sources and intermediate precursors, and the enzymatic reactions that convert them into one another. Metabolism is naturally amenable to analysis with network science, which has also been successfully used to describe cellular systems such as protein-protein interactions (Thomas et al 2003), transcriptional regulation (Alon 2007), and protein structure (Amor et al 2014). In the case of metabolism, however, the insights gained from network science are somewhat more dispersed because of the lack of a widely-accepted method to construct metabolic networks.…”