Session P42: Focus Session: Evolutionary Systems Biology II - From molecules to cells

Statistical Physics Approaches to RNA Editing

The central dogma of molecular Biology states that DNA is transcribed base by base into RNA which is in turn translated into proteins. However, some organisms edit their RNA before translation by inserting, deleting, or substituting individual or short stretches of bases. In many instances the mechanisms by which an organism recognizes the positions at which to edit or by which it performs the actual editing are unknown. One model system that stands out by its very high rate of on average one out of 25 bases being edited are the \textit{Myxomycetes}, a class of slime molds. In this talk we will show how the computational methods and concepts from statistical Physics can be used to analyze DNA and protein sequence data to predict editing sites in these slime molds and to guide experiments that identified previously unknown types of editing as well as the complete set of editing events in the slime mold \textit{Physarum polycephalum}.   

Biophysical Models of Evolution: Application to Transcription Factor Binding Sites in Yeast

There has been growing interest in understanding the physical driving forces of evolution at the molecular scale, in particular how biophysics determines the fitness landscapes that shape the evolution of DNA and proteins. To that end we study a model of molecular evolution that explicitly incorporates the underlying biophysics. Using population genetics, we derive a steady-state distribution of monomorphic populations evolving on an arbitrary fitness landscape. Compared to previous studies, we find this result is universal for a large class of population models and fully incorporates both stochastic effects and strong natural selection. This distribution can then be used to infer the underlying fitness landscape from genomic data. This model can be applied to a variety of systems, but we focus on transcription factor binding sites, which play a crucial role in gene regulatory networks. Since these sites must be bound for successful gene regulation, we consider a simple thermodynamic model of fitness as a function of the free energy for binding a transcription factor at the site. Using empirical energy matrices and genome-wide sets of binding sites from the yeast Saccharomyces cerevisiae, we use this model to infer the role of DNA-protein interaction physics in evolution.   

The Fitness of Genomic Order

Most bacteria have a single circular chromosome that can range in size from 160,000 to 12,200,000 base pairs. Considering the typical gene density, i.e. 1 gene per 1,000 base pairs, both the number of genes and the ways to arrange are huge. Intuitively, the arrangement of genes on the circle is not important if all of them can be replicated. However, there is typically one origin of replication, and when bacteria is attacked by genotoxic stress during replication, the whole replication process can not be finished. As a result, which gene is replicated first, which is second, ..., becomes very important. Experimentally, we found a broad increase of DNA copy number near the origin of replication (OriC) of bacteria E.coli ($\sim$3200 genes) under genotoxic stress. Since the genes near OriC are mostly efflux pump genes, we propose that there is fitness advantage for those rapid stress response genes got replicated first, because they can facilitate the replication of the rest of genome. Similar to bacterial evolution to present genomic order, in the somatic evolution of cancer, genomic shuffling was also frequently observed, especially under genotoxic chemotherapy. Such re-arrangement of genome can be viewed as a journey to optimal point in the rugged fitness landscape of genomic order.   

Plasticity of metabolic networks and the evolution of C4 photosynthesis

Over 50 groups of plants have independently developed a common mechanism (C4 photosynthesis) for increasing the efficiency of photosynthetic carbon dioxide assimilation. Understanding the high degree of evolvability of the C4 system could offer useful guidance for attempts to introduce it artificially to other plants. Previously, the nonlinear relationship between carbon dioxide levels and rates of carbon assimilation and photorespiration has prevented the application of genome-scale metabolic models to the problem of the evolution of the pathway. We apply a nonlinear optimization method to find feasible flux distributions in a plant metabolic model, allowing us to explore the plasticity of the metabolic network and characterize the fitness landscape of the transition from C3 to C4 photosynthesis.   

Uncovering principles of cellular decision-making

Cells can cope with unpredictable environmental conditions by differentiating into appropriate states. In this talk, I will present our recent attempts to understand the role of genetic circuits in regulating the underlying process of cellular decision-making. Specifically, we are interested in how interactions within and across genetic circuits enable cells to choose among alternative fates. To address this question my laboratory is employing systems and synthetic biology approaches. Our ultimate goal is to uncover possible evolutionary pressures that may have selected for specific gene circuit architectures, dynamics and noise properties.   

Role of Multisite Phosphorylation in Timing of a Yeast Cell Cycle Event

We study the biochemical network that triggers the S phase in yeast cell cycle. Key components of this network are three proteins: two kinases and an inhibitor. First kinase, Cln1/2-Cdk, acts as an input signal, phosphorylating the inhibitor, Sic1. The second kinase, Clb5/6-Cdk, is sequestered into an inactive complex by Sic1. Clb5/6-Cdk is the output signal of the circuit. Sic1 has nine phosphorylation sites, and phosphorylation of six or more of them causes it to degrade rapidly, leading to a sharp rise of free Clb5/6-Cdk. Our experiments indicate that multisite phosphorylation (MSP) is responsible for the timing robustness of this sharp transition. We study the role of MSP in timing using computer simulations. Preliminary results indicate that, MSP does not bring timing robustness when each kinase can phosphorylate each site with identical specificity. We employ in silico evolution to find the specificity configuration for the phosphorylation sites that leads to most robust timing under extrinsic noise.   

Precision of sensing, memory and fluctuating environments

Multiple cell types were recently shown to sense their chemical environment with astonishing accuracy, crucial for nutrient scavenging, mating, immune response, and development. It is currently unknown if this sensing near the single-molecule detection limit is due to highly precise single measurements, or due to learning over time. In this work, we analyze if cell memory can allow cells to sense beyond the current estimates of the fundamental physical limit. By merging Bayesian inference with information theoretic concepts, we derive analytical formulas which show that memory improves sensing in correlated fluctuating environments, but not in strongly uncorrelated environments. Despite many analogies with problem solving strategies in engineering, our theory shows fundamental differences in interpreting noisy stimuli in the microscopic and macroscopic world.   

Rapidly evolving microorganisms with high biofuel tolerance

Replacing non-renewable energy sources is one of the biggest and most exciting challenges of our generation. Algae and bacteria are poised to become major renewable biofuels if strains can be developed that provide a high,consistent and robust yield of oil. One major stumbling block towards this goal is the lack of tolerance to high concentrations of biofuels like isobutanol. Using traditional bioengineering techniques to remedy this face the hurdle of identifying the correct pathway or gene to modify. But the multiplicity of interactions inside a cell makes it very hard to determine what to modify a priori. Instead, we propose a technology that does not require prior knowledge of the genes or pathways to modify. In our approach that marries microfabrication and ecology, spatial heterogeneity is used as a knob to speed up evolution in the desired direction. Recently, we have successfully used this approach to demonstrate the rapid emergence of bacterial antibiotic resistance in as little as ten hours. Here, we describe our experimental results in developing new strains of micro-organisms with high oil tolerance. Besides biofuel production, our work is also relevant to oil spill clean-ups.   

A Cahn-Hilliard model of vascularized tumor growth in a complex evolving confinement using a diffuse domain approach

Understanding the spatiotemporal evolution of tumor growth is essential for developing effective strategies to treat cancers. Various studies have suggested that spatial heterogeneity during tumors growth is a key factor associated with subsequent tumor invasion and the effectiveness of chemotherapy. Spatial heterogeneity may arise due to morphological instability of the tumors and the complex tissue structure surrounding the tumors. In previous works, we have used a Cahn-Hilliard tumor growth model to study the morphological instability for tumors in non-resisting tissues. However, most tumors are surrounded by complex tissue structures and confined in the capsules of some organs or between certain basement membranes. The capsules and basement membranes may be distorted by interacting with the evolving tumors, affecting the morphological instability. Here we adopt a novel diffuse domain approach to adapt our previous Cahn-Hilliard model for tumor growth in such complex evolving environments. As an example, we apply the model to simulate the evolution of lymphoma in a lymph node, incorporating also the tumor-induced angiogenesis.   

Emergence of Information Transmission in a prebiotic RNA Reactor

A poorly understood step in the transition from a chemical to a biological world is the emergence of self-replicating molecular systems. We study how a precursor for such a replicator might arise in a hydrothermal RNA reactor, which accumulates longer sequences from unbiased monomer influx and random ligation. In the reactor, intra- and intermolecular base pairing locally protects from random cleavage. Analyzing stochastic simulations, we observe a strong bias towards long sequences with complex secondary structures, which would facilitate the emergence of ribozymes. Further, we find temporal sequence correlations that constitute a signature of information transmission, weaker but of the same form as in a true replicator.   

Towards molecular evolution with thermal traps

Live evolves by replication and selection of nucleotide polymers. Our experiments aim to drive molecular replication and selection with physical nonequilibrium boundary conditions. We discuss three approaches. Replication Trap. Molecules are exponentially accumulated by a combination of thermophoresis and convection, driven both by the same thermal gradient [1]. We have shown last year that with the help of a polymerase protein, concurrent replication and accumulation is possible [2]. Convection is melting and annealing the DNA in an oscillatory pattern, doubling the DNA in each cycle. Trapped polymerization. The chemical equilibrium of polymerization is expected to shift in the thermal trap. As the trap accumulates the monomers, polymerization yields longer polymers. However, since the trap is exponentially length selective, distributions beyond exponential tails are predicted. Replication by selective degradation. Replication typically is discussed as template directed polymerization. We showed that selective degradation and a thermal trap leads to replication-like behavior using only non templated polymerization [3]. The progression of information is given by the faster degradation of single stranded over double stranded RNA. [1] PNAS 104, 9346 (2007) [2] PRL 104, 188102 (2010) [3] PRL 107, 018101 (2011)