Abstract | Keywords | Export | Availability | Bookmark

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Peptides are the smallest biological molecules of plant proteome often the size of 2 to 100 amino acids. Previously, many eukaryotic RNAs containing short open reading frames (sORFs) in eukaryotes were considered non-coding. However, several recent studies clearly indicate that sORFs can indeed translate into bioactive peptides. It was shown that several transcriptionally active regions (TARs) were induced upon by application of biotic (Botrytis cinerea) and abiotic stress (Paraquat) in Arabidopsis thaliana. These TARs could be translated into stress-induced peptides (SIPs). These SIPs can be either Botrytis cinerea induced peptides (BIPs) or Oxidative stress induced peptides (OSIPs) depending on the applied stress. Peptide-mediated interactions in which a short linear motif binds to a globular domain play a key role within the cell, mediating several cellular processes such as signal transduction and other regulatory pathways. These SIPs might play an important role in stress tolerance by interacting with a receptor protein. The main objective of this thesis is to carry out large scale predictions of peptide binding pockets on protein surfaces. In many cases, protein monomers have a specific function when interacting with a cofactor or ligand together. When a peptide encoded by a TAR binds at a known ligand or cofactor binding site, it may affect protein function. Moreover, for protein dimers or multimers, several peptides may bind at dimer interface of multi-chains and modulate its activity. In our current study, the peptide-protein docking method pepATTRACT-local was adopted, which combines a coarse-grained peptide docking protocol with atomistic refinement. The restraints used for dockings were obtained using peptide binding site prediction server PepSite2. 530 peptide-protein pairs were successfully docked and we screened out 104 top pairs using in-house scripts and manual observations. 30 peptides were found to bind to a known ligand/co-factor binding site and 17 peptides can bind to a pocket at the interface between two monomers of a multi-chain complex. To characterize peptide-protein interactions, top ten models corresponding to each of docked complex were extracted and analyzed using tools such as BINding ANAlyzer and HBPLUS. In an additional study, 104 screened peptide-protein pairs were docked without any restraints and the performance was tested against the previously docked complexes. It was found that 26 complexes were docked correctly in the same pockets by both pepATTRACT-local and pepATTRACT-blind docking. This dataset was classified as a high confidence set. Binding analysis revealed that most of the peptides interact with help of their sidechains on a specific pocket on the protein surface. Hydrophobic and hydrogen bonds stabilize much of the interactions between the receptor and peptides. Several peptides may target signaling proteins or interfere with biosynthetic enzymes thus, influencing their functions. We further investigated three types of proteins from the top docked complexes including herbicide target proteins, stress responsive enzymes like aldo-keto reductases and leucine-rich repeat receptor kinases. Molecular visualization software Chimera and PyMOL along with 2D ligand-protein interaction analysis software LIGPLOT were used for the analysis. In the current study, we have generated novel hypotheses on protein-peptide interactions which will help scientists to validate these interactions experimentally.