Funded Grants


Identification of bistable network topologies associated with toxic aggregation thresholds found in conformational diseases

The toxic aggregation of proteins is closely linked to conformational diseases, which arise from the failure of a specific protein to adopt (or remain in) its native conformational state. Conformational diseases give rise to neurodegenerative conditions, such as Alzheimer's, Parkinson's, and Huntington's disease as well as prion encephalopathies, and non‐neurodegenerative conditions, such as diabetes, cataract, medullary carcinoma and pulmonary alveolar proteinosis.

The largest group of conformational diseases, amyloidosis, is associated with the conversion of native proteins to highly organized fibrillar aggregates known as amyloids. Protein aggregation into amyloids is initiated by an activated monomer, such as a misfolded protein. Monomers aggregate together to form oligomers. The oligomers form protofilaments and eventually organize into fibril structures. Protein aggregates can be toxic to cells and may ultimately lead to disease as a result of improper trafficking, premature protein degradation, or misfolding of native proteins.

The formation of protein aggregates in conformational diseases often displays a threshold phenomenon that is characterized by a sudden shift between nontoxic to toxic levels of protein aggregates. This phenomenon is observed, for example, in the pathogenesis of Huntington's disease where patients with the disease exhibit high aggregation levels and express a form of the huntingtin protein with 38-182 polyglutamine residues, while patients without the disease exhibit low aggregation levels and express a form of the protein with 8-37 polyglutamine residues. Thus, the disease appears when a critical length of 37 glutamine residues is exceeded. A similar threshold is correlated to the loss of motility in worms where toxic polyglutamine aggregation occurs in muscle cells. Another example can be found in the formation of yeast prion aggregates. Yeast aggregates exhibit a toxic aggregation threshold that is dependent on the concentration of the Hsp104 protein. In the biotechnology industry, bovine and human serum albumin exhibit an aggregation threshold that is dependent on a specific concentration of albumin. A key to controlling conformational diseases, therefore, is to understand the underlying mechanisms responsible for the threshold phenomenon associated with toxic protein aggregation.

Due to the importance of protein aggregation in both conformational diseases and the biotechnology industry, the underlying mechanisms of protein aggregation and their kinetic models have been studied for more than 50 years. These mechanisms often display differences in the interactions of key intermediates involved in the aggregation process. Existing models are typically used to discriminate protein aggregation mechanisms through in vitro assays. The existing mechanistic models cannot, however, explain the observed threshold phenomenon of protein aggregation.

In chemical reactions the appearance of a threshold phenomenon is characterized by the presence of two stable steady states (and a third unstable steady state), which coexist within a certain range of parameters. In dynamical and complex systems, this phenomenon is known as bistability. In biochemistry, metabolic and signaling pathways exhibiting bistability will switch between the two stable steady states in response to a chemical signal. For this reason, it is generally said that such pathways exhibit 'switch‐like behavior'. In conformational diseases, the aggregation levels can switch between non‐toxic (low concentration) and toxic (high concentration) protein aggregates in a thresholddependent manner. To date, there are only a few mathematical models which can describe the threshold of protein aggregation in conformational or prion diseases as a bistable system. These models are phenomenological, however, and do not explain the threshold for protein aggregation in a mechanistic manner.

Both theoretical and experimental evidence exists for bistable reaction networks in biology. Bistable networks have been extensively analyzed in a number of cellular processes, such as the cell cycle, cell differentiation, and cell death (or apoptosis). Biochemical switches have been observed in gene expression systems, such as the lac operon, sporlation inhibition operon, and oocyte maturation. Bistable switches are also implicated in stem cell fate decision and protein phosphorylation.

Members of the complex systems community have focused on mapping and dissecting networks with the objective of discovering design principles that can explain complex behaviors. Recurrent network motifs have been linked to specific functions, such as temporal expression of gene networks, cell decision, biological oscillators, and biochemical adaptation. Some of the motifs involved in transcriptional and developmental networks are negative and positive auto‐regulation, feed‐forward loops, multi‐output, single input module, and dense overlapping regulons. These findings suggest that the functional repertoire of a network is determined by its topology (i.e., reaction mechanism). Therefore, it is reasonable to hypothesize that there are a limited number of reaction network topologies that can exhibit bistable behavior. Our objective is to determine how many different ways an aggregation reaction network can be configured to produce threshold phenomena (bistability) in protein aggregation.

Despite numerous studies characterizing the behavior of bistable networks, it is not known which specific network topologies will exhibit bistable behavior. There are three conditions generally considered necessary for bistability: (i) positive feedback loops, (ii) ultrasensitive stimulus‐response curves, and (iii) a mechanism to prevent a concentration explosion of any chemical species in the network. There is evidence in the literature to suggest that these may not be absolute requirements, however. In a review, Pomerening (2008, Current Opinion in Biotechnology 19, 381) addressed the fact that there is discordant evidence concerning both the design and function of bistable networks. Ortega et al. (2006, FEBS Journal 273, 3915) and Ramakrishnan & Bhalla (2008, PLoS Computational Biology No. 4) demonstrated that bistable switches can appear in networks with no positive feedback loops. In some enzyme‐catalyzed reactions, enzyme saturation plays a critical role, while in others the balance between competing reactions is at the source of the bistable behavior.

To characterize bistability, one approach may be to search for bistable networks motifs in pathways, databases, or known aggregation reaction mechanisms. Neither KEGG nor BIOCARTA, two widely used databases of network interactions, report information on bistable behavior, however. In addition, no bistable behavior has been reported in existing kinetic mechanisms of protein aggregation. A computational exploration of random population networks with 8 molecules revealed that 2% of the mass action reaction networks exhibit bistable behavior. A different study began with a set of 12 biologically common reactions and performed a systematic exploration of all possible reaction configurations of (i) 1 to 6 reactions involving 2 or 3 molecules and (ii) 1 to 3 reactions involving 4 molecules. They reported that 10% of the reaction networks exhibited bistable behavior and discovered that autocatalysis is a good predictor of bistability. Interestingly, there is a correlation between prion diseases and autocatalysis. Unfortunately, autocatalysis is only a good predictor for networks with 3 molecules, which is not sufficient to describe the typical complexity of an aggregation reaction network.

Aims of the project

The primary objective of this project is to identify the chemical aggregation network topologies that give rise to bistability. The guiding hypothesis of our research is that as there are a limited number of network topologies capable of bistable behavior. To test this hypothesis, we will implement a set of complex systems based tools to discover the reaction topologies capable of exhibiting bistability.

Specific aims:

  • Aim [I]. Identify the key reaction motifs that lead to bistability in aggregation reactions.
  • Aim [II]. Characterize the parameter conditions that lead to bistability in aggregation reactions.
  • Aim [III]. Create a database of aggregation reactions, annotating those that exhibit bistable behavior, listing the key motifs and the necessary conditions for generating bistability.

Significance

In this project, we hypothesize the existence of a limited number of topologies capable of producing bistability in aggregation reactions. If this hypothesis is correct, it should be possible to discover the essential motifs that are the basis of bistability and which are otherwise hidden by the complexity of the aggregation phenomenon. The characterization of bistability (motifs and conditions) in protein aggregation will be very important in the search of therapeutic approaches aimed at modulating conformational diseases. In addition, the discovery of essential motifs of bistability will help scientists organize network and pathway databases according to classification by key motifs in both naturally and non-naturally occurring aggregation reactions.

We will validate the tools and methodology developed in this project by testing with a number of well-studied networks known to exhibit bistability. We intend to disseminate our methodology and the results of our research to the complex systems, computational biology, biochemistry, pharmacology, cellular biology and conformational disease communities.

Finally, the research activities of this project will provide interdisciplinary (biological, physicochemical, mathematical, and computational) training to undergraduate and graduate students as well as postdoctoral researchers. The research activities will help train young scientists to apply complex systems approaches to an unconventional biological problem (the characterization of aggregation thresholds in conformational diseases).