Michaels Projects
The Michaels lab welcomes applications of motivated students for the following master projects
Project 1: How do liquid condensates affect biochemical reactions?
Type: theoretical/computational
Overview: Living cells rely on a variety of compartments to organise and regulate biochemical reactions in space and time. Traditionally we think of these compartments as having a membrane. Recently, however, it has emerged that many of these compartments do not possess a membrane. These structures are called membrane-less organelles or liquid condensates. These condensates form by liquid-liquid phase separation of proteins in the cytoplasm. The biological function of liquid condensates remains a subject of intense research. Proposed functions include environmental sensing, stress response, signalling control, concentration buffering and, in general, compartmentalisation of biochemical reactions in the cell. It has been proposed that liquid condensates could accelerate the kinetics of chemical reactions by locally concentrating reactants. However, the increased concentration also has potentially inhibitory effects due to higher viscosity and crowding and reduced reactant mobility. Therefore, it remains unclear whether and to what extent liquid condensates can accelerate biochemical reactions.
Scientific aims: In this project we will use mathematical modelling to explore the effectiveness of liquid condensates at accelerating chemical reactions. This work will shed light on the physical limits on how much liquid condensates can affect biochemical reactions.
Project 2: Optimal control strategies to inhibit protein aggregation in Alzheimer’s disease
Type: theoretical/computational
Overview: A range of medical conditions, such as Alzheimer’s disease, Parkinson’s disease, and Type-II diabetes, have been linked to the aggregation of specific proteins into amyloid fibrils. Thus, it is imperative to develop effective strategies to combat protein aggregation. However, the development of effective therapeutics to prevent protein-aggregation-related diseases has been very challenging, in part due to the complex nature of the aggregation process itself, which involves several microscopic steps. A promising and recent strategy proposes the use of drugs to inhibit specific molecular events during the aggregation process. However, viable treatment protocols require that we balance the efficacy of the drug with its toxicity, since not only the aggregates are toxic but the inhibitors themselves are also toxic in large doses. This raises the natural question: what is the optimal strategy (dose of inhibitor and time of administration) that emerges from a balance between the degree of inhibition and the toxicity of the inhibitor while accounting for the mechanism of inhibition and the effect of metabolic clearance of the drug? This information is key for the design of rational therapies to control pathological protein aggregation and for seeding clinical trials, in particular for extrapolating the most desirable protocol and dose from a model system of disease to clinically relevant situations.
Scientific aims: In this project we will address the above question by combining mathematical models of protein aggregation kinetics and drug metabolism with control theory and reinforcement learning ideas. This combination will allow us to devise optimal treatment protocols that emerge directly from an understanding of the molecular mechanisms of aggregation, its inhibition by inhibitors and the metabolic clearance of the drug. Theoretical predictions will be tested directly using experimental data on the inhibition of amyloid-beta aggregation (a key protein involved in Alzheimer’s disease) in a C. Elegans model system of Alzheimer’s disease. This approach could guide the rational design of effective therapeutic strategies to target protein aggregation diseases.
Project 3: Optimal control of aging and cancer in multicellular organisms
Type: theoretical/computational
Overview: Effective multicellularity requires both cooperation and competition between constituent cells (1,2). Cooperation between cells involves sacrificing individual-cell fitness in favour of communal activities to build and maintain the multicellular organism (2), while competition eliminates unfit senescent cells, whose accumulation is liked to ageing and a number of ageing-related pathologies (1). However, when competition between cells is too fierce, multicellular organisms become susceptible to cancer and cancer-like phenomena, which correspond to a breakdown of cell cooperation in form of ‘cheating’ (2,3). These cheater cells have a selective advantage because they do not cooperate and so multiply faster than other cells. Cheating has devastating effects on the multicellular organism and manifests itself in the form of cancer (2,3). Thus, multicellularity experiences a constant conflict between cheating and cooperation, making aging inevitable (either due to senescence or cancer) and resulting in complex aging dynamics (3). This complexity makes it challenging to understand the action of therapeutic interventions in aging, as one strategy may slow down some mechanisms but accelerate others. Quantitative models can be very helpful in this context, providing a way to understand system-level effects of specific strategies.
References: (1) Aktipis et al, Phyl Trans Roy Soc B 370, 20140219 (2015); (2) Campisi, Cell 120, 513 (2005); (3) Nelson et al, PNAS 114, 12982 (2017).
Scientific aims: The goal of this project is to develop a mathematical model for describing the dynamics of aging in multicellular organisms because of cooperation and competition and to utilise this model to rationalise and optimise a range of possible therapeutic interventions in aging. Specifically, we will develop a simple mathematical framework for multicellular aging via a generalized replicator dynamical system, where cell genotypes are described in terms of two parameters: vigor (a measure of resources available to cells) and cooperation (fraction of resources allocated to community activities). Healthy cells have high vigor and high cooperation. Somatic mutations can cause progressive loss of vigor, or cooperation. Cells with low vigor are senescent, while cells with low cooperation correspond to cancer. A master equation describes the dynamics of this system because of cell proliferation (competition) and somatic mutations, which are irreversible transitions that lower vigor or cooperation. This project thus aims to use this model as a basis to understand the effect of a range of therapeutic strategies with the goal to delay the loss of organismal vitality. Strategies will include for instance drugs that clear either senescent or cancerous cells or biological manipulations are able to reverse senescence or cancer, including inactivation of specific tumour suppressor genes. Once the impact of the different strategies on vitality is understood, we will use optimal control theory to determine optimal treatment schedules (timing and dose of drug administration) that maximise organismal vitality at minimum cost of intervention. An alternative strategy is to ask whether the system can learn the optimal protocols via an iterative procedure that mimics evolution. To this end we will determine optimal treatment protocols using reinforcement learning (Q-learning algorithm), a process by which a system is able to optimize its actions by interacting with its environment. The resulting optimal protocols will be compared with the theoretical predictions from optimal control.
Interested? Apply!
Prerequisites: The successful candidate should possess high level of self-motivation and enthusiasm and should be open minded, flexible, and keen to learn new concepts. Experience in and aptitude for computational programming and/or physical modelling are an advantage for this project. No prior knowledge of biophysics is necessary.
Selection criteria: Interested students should email the supervisor Prof. Dr. Thomas Michaels (). The email should briefly outline the student’s interest and motivation in the project (why do you want to work on this specific project?) and include a short CV and a transcript of grades.