Relevant Degree Programs
Graduate Student Supervision
Doctoral Student Supervision (Jan 2008 - April 2022)
Intrinsically disordered protein regions (IDRs) constitute a significant portion of our proteome but have traditionally received less attention than folded domains, making IDRs a focus of ongoing research. These protein regions that are not folded prior to binding have functional importance, contradicting the protein structure–function paradigm. One mechanism through which IDRs function is by forming interactions with protein partners through interaction-mediating elements, including molecular recognition features (MoRFs). Computational biologists have developed many protein-sequence-based methods for predicting IDRs and MoRFs and have applied them in proteome-wide studies, leading to the recognition of their significant roles in regulatory and signaling pathways, housekeeping proteins, and interaction network hubs. IDRs’ involvement in these processes made them attractive targets for research and therapy. However, the folded (globular) proteins interacting with IDRs have received less attention. We developed a structure-based protein interface predictor for binding sites of IDRs named IDRBind, which incorporated features specific to MoRF binding sites with ideas from existing globular protein interface predictors. IDRBind was developed using machine learning and was trained on MoRF–globular complex structures. It consists of two gradient boosted trees models that are combined using a conditional random fields (CRF) model. The structural data used for the development of IDRBind was also useful for characterizing and comparing IDR and globular interactions.In this thesis, I will cover the development and benchmarking of IDRBind and examine the properties of MoRF interactions with comparisons to those of globular proteins and peptides. IDRBind exhibits high performance on predicting both MoRF and peptide binding sites. Our analysis also revealed that MoRF binding sites are positioned between those of peptide and globular proteins on multiple measured properties, in agreement with the performance trends of IDRBind. The differentiating characteristics of IDR-mediated interactions were further investigated by comparing the localization patterns of mutations. Despite the flexibility of IDRs, the interaction surfaces of the IDR complex structures are just as enriched in disease-associated mutations as globular interactions. Their prominent roles in disease, especially in cancer, as well as attributes that favor drug targeting, make IDR interactions a fascinating topic for research.
The Mycobacterium tuberculosis ABC transporter Rv1747 is a putative exporter of cell wall biosynthesis intermediates. Rv1747 has a cytoplasmic regulatory module consisting of two pThr-interacting Forkhead-associated (FHA) domains connected by a conformationally disordered linker with two phospho-acceptor threonines (pThr). In chapter 2, I report the structures of FHA-1 and FHA-2 determined by X-ray crystallography and NMR spectroscopy, respectively. Relative to the canonical 11-strand beta-sandwich FHA domain fold of FHA-1, FHA-2 is circularly permuted and lacking one beta-strand. Nevertheless, the two share a conserved pThr-binding cleft. FHA-2 is less stable and more dynamic than FHA-1, yet binds model pThr peptides with moderately higher affinity (~ 50 uM versus 500 uM equilibrium dissociation constants). Based on NMR relaxation and chemical shift perturbation measurements, when joined within a polypeptide chain, either FHA domain can bind either linker pThr to form intra- and intermolecular complexes. Protein phase separation has been recently shown to be a fundamental mechanism underlying the clustering of some proteins at the eukaryotic cell membrane. In chapter 3, I show that upon multi-site phosphorylation of the linker by several Mtb serine/threonine protein kinases including PknF, the isolated regulatory module readily multimerizes and phase separates into dynamic liquid droplets with diagnostic properties similar to those exhibited by eukaryotic proteins. The process is reversed by the sole Mtb serine/threonine phosphatase PstP. In the absence of phosphorylation, the Rv1747 regulatory module can still undergo phase separation, albeit at higher protein concentrations and into droplets with more fluid properties. This points to a synergy between specific FHA-pThr binding and additional weak association of the ID linker and/or the FHA domains leading to the pre-requisite multivalent interactions for demixing. Droplet formation of the regulatory module was replicated in a pseudo-two-dimensional system on model supported lipid bilayers. Potential clusters of Rv1747 were also detected in Mtb using ultra-high resolution Direct Stochastic Reconstruction Microscopy (dSTORM). This is the first reported example of phase separation by both a bacterial protein and an ABC transporter, and suggests possible mechanisms for the regulation of Rv1747 within Mtb. I hypothesize that the tunable, phosphorylation-dependent multimerization described here regulates Rv1747 transporter activity.
Master's Student Supervision (2010 - 2021)
The full abstract for this thesis is available in the body of the thesis, and will be available when the embargo expires.
Amyloid fibril formation, believed to be a generic property of polypeptides, plays major roles in neurodegenerative pathologies such as Alzheimer’s, Parkinson’s and prion diseases, as well as in functional biological processes in many organisms including humans. Revealing specifics of their molecular architecture, conformational stability, mechanisms of formation and physical properties holds clues to devising effective methods to fight their associated pathologies. An increasing requirement has been the development of a detailed understanding of the nanomechanics of amyloid core structures due to their relevance in biomedicine and nanotechnology. Of special significance is the mechanism of fibril fracture and infectivity in disease as well as the mechanical stability for novel biomaterial design. Here, we use a series of steered molecular dynamics simulations on different amyloid fibrils to report a broad spectrum of mechanical properties ranging from a strong and stiff β-helical fibril to weak and soft amyloid such as those formed by the mammalian prion protein. We relate the strength and elastic modulus with hydrogen bond densities and van der Waals energies in the core of the fibrils and show that weakened side-chain interactions lead to fibrils with reduced tensile strengths as a result of modified molecular packing in the fibril core. This modulation might lead to a combination of exceptional mechanical attributes such as those of the human functional amyloids.
Given the wide and fundamental roles proteins play in cells, as well as their potential medical and industrial applications, a detailed understanding of the relationships between sequence, structure, dynamics, and function is of critical importance. Molecular models are required to solve this problem, as well as models of the associated conformational spaces. One of the most challenging aspects of modeling these vast ensembles is the computer power required to carry out the requisite simulations. Reduced solvent models, and particularly a class referred to as implicit solvent models, have been developed extensively; however, they make many assumptions and approximations that are likely to affect accuracy. Here, several implicit solvent models commonly used for protein modeling are evaluated by comparing the expected changes in free energies of solvation upon folding ΔGsolv derived from micro--ms simulations of fast folding proteins to those given by the implicit solvent models. In the majority of cases, there is a significant and substantial difference between the ΔGsolv values calculated from the two approaches, with the implicit solvent models excessively favouring the folded state over the unfolded state. This could only be remedied by selecting values for the model parameters -- the internal dielectric constant for the polar term and the surface tension coefficient for the apolar term -- that were system specific or physically unrealistic.