Welcome to the Gerland group -
Physics of Complex Biosystem
Vision
In physics, interactions between particles follow laws. In biology, interactions between biomolecules serve a function. These very different points of view are beginning to merge as research over the past years has demonstrated how, in some exemplary cases, the laws of physics constrain the implementation of biological function.
We investigate several such cases. For instance, we study how the spatial arrangement and coordination of enzymes determines the efficiency of a multi-step reaction. These spatial arrangements can be natural (as in biomolecular complexes) or engineered with the modern methods of bio-nanotechnology. In both cases, fundamental functional tradeoffs emerge, which must be characterized to understand the optimization of such systems.
Methods from theoretical physics help to describe the functioning of these complex biomolecular systems on a quantitative level, while the biological function leads to new questions, with many parallels in the engineering disciplines. Seen from this perspective, a bacterium is a microscopic bioreactor programmed by evolution to rebuild itself from a variable set of resources and in fluctuating environments. How is this bioreactor programmed? Which strategies enable the control of a diverse set of physico-chemical processes in a way as to robustly produce a highly complex product? Quantitative analysis and modeling facilitates insight into the underlying design principles.
Recent Research Highlights
Gradient estimators for parameter inference in discrete stochastic kinetic models
Stochastic kinetic models are ubiquitous in physics, yet inferring their parameters from experimental data remains challenging. For deterministic models, parameter inference often relies on gradients, which can be obtained efficiently through automatic differentiation (AD). However, AD cannot be applied directly to the Gillespie stochastic simulation algorithm (SSA), since sampling from a discrete set of reactions introduces non-differentiable operations. In this work, we adopt three gradient estimators from machine learning for the Gillespie SSA: the Gumbel-Softmax Straight-Through (GS-ST) estimator, the Score Function estimator, and the Alternative Path estimator. We use the estimators to evaluate gradients of steady-state and time-dependent observables, and compare their performance in representative biophysical systems with relaxation dynamics (bimolecular association) and oscillatory dynamics (repressilator). We find that the GS-ST estimator generally yields well-behaved gradient estimates, but exhibits diverging variance in challenging parameter regimes, which can cause parameter inference to fail. In these cases, other estimators provide more robust, lower variance gradients. Our results demonstrate that gradient-based parameter inference can be effectively combined with the Gillespie SSA, with different estimators offering complementary advantages.
A quantitative framework for bacterial competition during starvation
Bacterial communities are inherently characterized by recurring cycles of feast and famine, which creates distinct forms of competition when cells compete for recycled necromass. In this work, we develop a quantitative framework for the competition of isogenic Escherichia coli populations whose starvation physiology is tuned by prior growth history. If starved individually, fast-grown populations have higher maintenance demands and die slightly faster, while slow-grown populations are better adapted to starvation, leading to a slower decline in cell density. However, when mixing both populations, these differences are exaggerated in a frequency-dependent manner: Populations with high maintenance demand die several-fold faster than in monoculture, whereas better adapted populations reduce their death rate below that of stationary-phase adapted monocultures. Our framework shows that survival in competition is governed by a self-amplifying necromass recycling feedback, in which the majority population sets the shared energy pool, constituted by released biomass from dead cells. This shows that phenotypic differences alone are sufficient to predict selection under starvation, and our framework provides a mechanistic and predictive basis as a starting point for modelling multi-species environments such as biofilms or trade-off scenarios between rapid growth and long-term survival.
Sequence and chemical specificity define the functional landscape of intrinsically disordered regions
Intrinsically disordered regions (IDRs) pervasively engage in essential molecular functions, yet they are often poorly conserved as assessed by sequence alignment. To explore the seeming paradox of how sequence variability is compatible with persistent function, we examined the functional determinants for a poorly conserved but essential IDR. We show that IDR function depends on two distinct but related properties: sequence and chemical specificity. Whereas sequence specificity operates via binding motifs and depends on the precise order and identity of residues, chemical specificity reflects the sequence-encoded chemistry of multivalent interactions across an IDR and depends on local and global chemical properties. Unexpectedly, a binding motif essential in the wild-type IDR can be removed when compensatory changes to the sequence chemistry are introduced, highlighting the orthogonality and interoperability of these properties, and expanding the sequence space compatible with function. Our results provide a general framework for the functional constraints on IDR evolution.
A quantitative model of enzyme-free copying of RNA with dimers
Copying genetic information without enzymes may have been a crucial step toward the origin of life, but known RNA copying reactions, such as template-directed primer extension, are slow and inefficient. We combined experiments and kinetic modeling to investigate the most effective enzyme-free RNA copying system reported so far, which uses strongly pairing RNA dimers as building blocks. Our analysis shows that the main bottleneck is the formation of new phosphodiester bonds on the template. The model reproduces the experimentally observed copying of up to 12 template bases through primer extension, dimer coupling, and fragment ligation, while revealing how material flows through competing reaction pathways. Our results provide a detailed mechanistic picture of protein-free RNA copying and identify the key limitations that must be overcome to develop more efficient self-replicating systems.
Force generation by enhanced diffusion in enzyme-loaded vesicles
The diffusion coefficient of some metabolic enzymes increases with the concentration of their cognate substrate, a phenomenon known as enhanced diffusion. In the presence of substrate gradients, enhanced diffusion induces enzymatic drift, resulting in a nonhomogeneous enzyme distribution. In this publication, we study the effects of enhanced diffusion on enzyme-loaded vesicles placed in external substrate gradients using a combination of computer simulations and analytical modeling. We observe that the spatially inhomogeneous enzyme profiles generated by enhanced diffusion result in a pressure gradient across the vesicle, which leads to macroscopically observable effects, namely deformation and self-propulsion of the vesicle. Our analytical model allows us to characterize the dependence of the velocity of propulsion on experimentally tunable parameters. The effects predicted by our work provide an avenue for further validation of enhanced diffusion, and might be leveraged for the design of novel synthetic cargo transporters, such as targeted drug delivery systems.
more research highlights ...