Summary: Proteins are biochemical workhorses that are needed in many important biological functions. Lately one aspect of these macromolecules have gained significant attention, which is their ability to ¡¥stick to each other¡¦ to from ¡¥protein aggregates¡¦ or ¡¥amyloids¡¦. This behavior is more common when proteins fail to adopt a ¡¥correct¡¦ three dimensional shape, commonly known as a ¡¥misfolded¡¦ form. These aggregates can be both beneficial and toxic for cellular processes. Although seems simple, this process is extremely complicated and no precise molecular understanding has emerged. Also, the ability of the proteins to misfold and aggregate via multiple pathways leading to various forms of aggregates has never been explored. Such molecular-level details of the process are important to know since functional aspects of these aggregates are related to the molecular size, shapes, stability and the rates of the formation. Since many of these parameters of this stochastic process are extremely difficult to analyze via conventional biochemical means, molecular-level computational simulations can be valuable. A precise understanding of the protein aggregation phenomenon would broaden the fundamental knowledge of both pathological and functional aspects of biomolecular science besides throwing newer insights. In this proposal, we will use amyloid-?Ò (A?Ò) peptide as a model protein that is known to form misfolded aggregates to accomplish our goals. Our main objective is to establish a fundamental framework for stochastic molecular-level simulation of the ¡¥on-pathway¡¦ fibril formation process that will serve as a basis for analyzing more realistic models with competing pathways to precisely predict the dynamics and mechanisms of protein aggregation. We have initiated a collaborative effort between two PIs at University of Southern Mississippi, (USM) with expertise in computational and biophysical analysis, to achieve our truly inter-disciplinary objectives.
Intellectual Merit: Protein aggregation is a nucleation-dependent process, however, precise understanding of its kinetics is not yet known. Aggregation and fiber formation is often considered to be a stochastic process with large variations in macroscopic molecule behavior and hence, stochastic molecular-level simulations would be essential to understand their dynamics. Furthermore, it is not realistic to consider aggregation as an isolated event as there are many different factors that influence protein aggregation in a physiological environment. Broadly, these include molecules that may ¡¥interact¡¦ with the protein besides others such as ionic strength, temperature etc. Hence in this proposal, we are focused on developing a fundamental framework of modeling protein aggregation and amyloid formation phenomenon via molecular-level modeling and stochastic simulation methodologies. The biophysical experiments can show the cumulative effects of the aggregates whereas the simulation will be able to predict the concentration change dynamics with respect to time for every aggregate involved in the pathway. This will allow us to study the exact nature of each aggregate and their sensitivity to the over-all pathway dynamics.
Broader Impact: USM is an excellent place to conduct this research from a scientist training perspective; MS being among the states with the highest levels of poverty besides providing a truly diverse student population. The research will provide a broader impact to the scientific community in the form of a fundamental mechanistic knowledge about protein aggregation systems. Our educational outreach mechanisms will involve giving seminars at participating undergraduate institutions in MS, recruiting economically disadvantaged students (including women and minorities) to perform summer research in the PI and Co-PI laboratories at USM. It will also enhance our graduate program through the design of new inter-disciplinary courses (e.g. ¡§Systems Biology¡¨ and ¡§Computational biophysics¡¨).
The phenomenon of protein misfolding and aggregation play a significant role in neurodegenerative (including amyloid based) diseases. Although, mechanisms of amyloid formation are relatively well known via several experimental and theoretical methods, the ability of the proteins to misfold and aggregate via multiple pathways leading to various forms of aggregates has never been explored. In this project, we have developed a simulation software that can precisely predict the dynamics and mechanisms of such protein aggregation. Specifically, we have used amyloid-β (Aβ) peptide as a model protein that is known to form misfolded aggregates in patients suffering from Alzheimer's Disease. Our key project outcomes include a model identification method, parameter estimation and fine-tuning methods and model-based predictions for the entire aggregation pathway. The model identification method can also incorporate realistic parameters of the cerebro-spinal fluid in the human brain that affect the aggregation process. The parameter fine-tuning block suggests that there are in fact multiple solutions for the estimated parameters that will help in designing new methods to disrupt this pathway for drug discovery. Our framework identified a critical stage of this aggregation process, called the nucleation mass, which serves as the crucial gate-keeper of the entire aggregation dynamics. We also hypothesise that the most detailed models need to be implemented for this stage in order to create a diagnostic tool for Alzheimer's Disease patients. The project findings has been widely disseminated through several publications, presentations and the creation of a new International Workshop on Complex Networks Dynamics: Cross-disciplinary tools for modeling, analysis and design (CoNed). Graduate students involved in the project have been trained in both computational approaches and biophysical experimental techniques. These students are well-prepared for pharmaceutical/biotech/programming related jobs in the future. The project findings were brought into the classroom through the PI's courses on Parallel Programming at Virginia Commonwealth University and the Co-PI's Protein Biochemistry course at the University of Southern Mississippi. This project was also instrumental in supporting some undergraduate students in the PIs' laboratory. One of the undergraduate students involved in the project was selected to participate in a one-of-a-kind undergraduate research symposium called, 'Posters on the Hill' held at Capitol Hill, Washington DC, giving him an opportunity to present this work to the US senators. This nation-wide event organized by the Council for Undergraduate Research (CUR) invites only a select number of undergraduate researchers. This project is uniquely poised towards creating an automated diagnostic tool for neuro-degerative diseases that involve different protein aggregation pathways. The project also makes a very important translational impact: our developed framework can be easily extended to create a flexible, plug-and-play based diagnostic software for protein aggregation and misfolding in general and amyloid formation in particular that will aid drug discovery for a suit of neuro-degenerative diseases in the future.
