Transcriptional and translational regulation of gene expression noise

Riddhiman Dhar

Genetically identical cells living in the same environment often show differences in gene expression. This variation in expression is quantified as gene expression noise. Expression noise generates phenotypic heterogeneity within a cell population, and have important implications for antibiotic persistence, cellular decision-making, cancer proliferation and anti-cancer therapy resistance. Research over last two decades have investigated the molecular origins of expression noise, and have identified several molecular features that contribute to and regulate noise. However, the relative importance of these features in noise regulation had remained unknown.
Using an integrated statistical model, we have shown that the number of transcription factors regulating a gene is a key predictor of expression noise. In addition, cooperation and competition between transcription factors for binding to promoter can influence transcriptional bursts and thereby, can increase expression noise. Noise originating from transcriptional bursts is further amplified or suppressed by the process of translation. This eventually determines the noise at the level of proteins, which have a more direct impact on phenotypes. Interestingly, studies in bacteria, yeast and Arabidopsis have revealed a positive association between translational efficiency and protein expression noise, although the molecular basis of this association had remained unknown. Our work, using stochastic models of translation at the level of single mRNA molecule and empirical measurement of protein noise, shows that the demand for ribosomal machinery during translation drives the positive association between translational efficiency and protein noise. Taken together, our work reveals the role of transcription factors in noise regulation, and further shows how transcriptional bursts are translated into protein noise. Cooperative and competitive binding of transcription factors, and variation in demand for ribosomal machinery are universal that occur in all organisms. Therefore, these findings have important implications for investigation of noise and phenotypic heterogeneity across all biological systems.

Mechanisms for functional noise in gene regulation

James Locke

Gene expression in individual cells can be surprisingly noisy. In unicellular organisms this noise can be functional. For example, it may allow a few cells in a bacterial population to enter a stress prepared state that allows survival of a sudden environmental stress. The role of gene expression noise in multicellular organisms is less clear. I will first discuss our work using alternative sigma factor circuits in bacteria as model systems for understanding noisy output from genetic circuits. Using a combination of single-cell time-lapse microscopy, synthetic biology techniques and mathematical modelling, we have investigated the circuit properties required for functional noise in gene regulation. I will then look at the role of noise in a multicellular system. Our work is revealing an unexpected level of variability in gene expression between and within genetically identical plants. This noise in gene expression may act as a mechanism for generating functional phenotypic di
versity in plants, similar to how it does in bacteria.

 

 

Controlling evolutionary trajectories through community interactions

Akshit Goyal

The emergence of antibiotic tolerance (prolonged survival against exposure) in natural bacterial populations is a major concern. Since tolerance has been studied primarily in isogenic populations, we do not yet understand how ecological interactions in a diverse community impact its evolution. To address this, we studied the evolutionary dynamics of a synthetic bacterial community composed of two interacting strains. In this community, an antibiotic-resistant strain protected the other, susceptible strain by degrading the antibiotic ampicillin in the medium. Surprisingly, we found that in the presence of antibiotics, the susceptible strain evolved tolerance. Tolerance was typified by an increase in survival as well as an accompanying decrease in the growth rate, highlighting a trade-off between the two. A simple mathematical model explained that the observed decrease in the death rate, even when coupled with a decreased growth rate, is beneficial in a community with weak protective interactions. In the presence of strong interactions, the model predicted that the trade-off would instead be detrimental, and tolerance would not emerge, which we experimentally verified. By whole genome sequencing the evolved tolerant isolates, we identified two genetic hotspots which accumulated mutations in parallel lines, suggesting their association with tolerance. Our work highlights how ecological interactions can promote divergent evolutionary trajectories in bacterial communities.

 

The Evolution of Kinetic Proofreading

Kabir Husain

Since at least Schrodinger, life has been seen as a non-equilibrium process that has successfully evaded the second law of thermodynamics: maintaining order for four billion years. Yet, while we understand how extant biomolecular Maxwell Demons work, much less is known about how such they come into existence in the first place. Here, combining experimental and theoretical work on proofreading DNA polymerases, we suggest that surprisingly little might be needed. Counter to intuition, proofreading can potentially arise through selection for speed alone, without any selection for higher accuracy. Our results rely on a memory effect intrinsic to the biophysics of templated copying – namely, stalling due to misincorporations. We describe progress towards an experimental test of the hypothesis: developing a massively multiplexed Luria-Delbruck assay capable of measuring the fidelity and speed of thousands of polymerases in a single experiment. Our work sheds light on the origins of fidelity in biomolecular processes, and suggests a minimal scenario for the origins of non-equilibrium order in biology.

 

From molecules to tissues and back: shaping and scaling morphogen gradients

Zena Hadjivasiliou

How morphogen gradients are formed has been under debate since the term was first coined by Alan Turing. Can extracellular diffusion alone lead to the robust formation of morphogen gradients? Or are cell processes like transcytosis important to move molecules across tissues? In this work we develop a theoretical framework to establish how the combination of different processes at the molecular and cellular level can in principle lead to morphogen gradient formation and its scaling to tissue size. We use this framework together with a sequence of independent experimental assays to quantify the relative contribution of different processes to the formation of the Dpp gradient in the developing fly wing. For large discs we find that a combination of hindered diffusion and recycling of endocytosed molecules drive Dpp gradient formation. By studying wing discs of different sizes, we show that the relative contribution of endocytosed molecules to Dpp transport becomes progressively more pronounced during growth. Furthermore, the contribution of recycled molecules is diminished when the extracellular factor Pentagone is impaired. Our findings suggest that Pentagone modulates the contribution of recycled molecules to the diffusing pool. I will discuss potential mechanisms through which Pentagone encodes information about tissue size and their evolution. Finally, I will present a framework that takes into account cell and tissue architecture to ask how morphogenesis may impact the transport of morphogen molecules and patterning.

 

Generating neural diversity through spatial and temporal patterning of neural stem cells

Sonia Sen

The nervous system has an incredible array of cell types. These develop from a relatively limited group of neural stem cells (NSCs) that use both spatial and temporal information to achieve such diversity. In both vertebrates and invertebrates, these NSCs receive specific molecular signals during their development on the neuroectoderm. As a result, they gain distinct molecular characteristics and the capacity to produce specialized neural lineages. Additionally, NSCs are temporally patterned via genes that are expressed sequentially, allowing them to produce a series of neural identities. Spatial patterning creates diversity between lineages, while temporal patterning generates variations in cell identity within lineages. The integration of the two within NSCs, gives rise to the multitude of cell types seen in the nervous system.
In the recent past, much insight has been gained about the temporal aspect of NSC patterning, but the spatial aspect remains less explored. I will discuss how Drosophila NSCs obtain and preserve their distinct spatial characteristics and how spatial patterning could influence the integration of temporal information within NSCs to generate neural diversity.

 

A Theoretical Perspecive on Waddington's Genetic Assimilation Experiments

Archishman Raju

Genetic assimilation is the process by which a phenotype that is initially induced by an environmental stimulus becomes stably inherited in the absence of the stimulus after a few generations of selection. While the concept has attracted much debate  after being introduced by C.~H.~Waddington seventy years ago, there have been few experiments to quantitatively characterize the phenomenon. Here, we revisit and organize the results of Waddington's original experiments and follow-up studies that attempted to replicate his results. We then present a theoretical model to illustrate the process of genetic assimilation and highlight several aspects that we think require further quantitative studies, including the gradual increase of penetrance, the statistics of delay in assimilation, and the frequency of unviability during selection. Our model captures Waddington's picture of developmental paths in a canalized landscape using a stochastic dynamical system with alternative trajectories that can be controlled by either external signals or internal variables. It also reconciles two descriptions of the phenomenon -- Waddington's, expressed in terms of an individual organism's developmental paths, and that of Bateman in terms of the population distribution crossing a hypothetical threshold. Our results provide theoretical insight into the concepts of canalization, phenotypic plasticity, and genetic assimilation.

Garima Rani

Microbial communities were some of the first agglomerations of living beings on the planet but even today the biophysical principles underpinning their spatiotemporal evolution remain mysterious. Here, in an effort to understand the way cells are organised in microbial colonies, we study the intermixing dynamics of descendants of individual bacterial cells in colonies of surface associated bacterial species. For this, we implement a custom-built label-free algorithm to track the progeny of individual cells as they grow and divide in such colonies, revealing the formation of distinct enclaves within the colony of comprising descendants of individual cells, thus displaying a spatial affinity for close relatives, at the cost of an entropically favourable option of intermixing.

Dependence on biological activity is a distinctive feature of the intermixing dynamics at the enclave scale in the colony. Specifically, we demonstrate that in faster growing colonies, invasion of one enclave into its complementary territory is relatively broad while in slower growing colonies, narrow invasion of fingers of cells occurs. We show that the evolution of such enclaves, which display a high degree of self-similarity with the colony at large on several key phenotypic traits, is regulated by the dynamics of topological defects in the colonies. Coarse-grained lattice modelling of such colonies, combined with insights from the thermodynamics of phase separation suggests that the emergence of progeny enclaves results from the interplay of cell shoving in the colony and the tendency of daughter cells to remain together, presumably due to polar adhesion after division. The model also reveals the role of the level of stochasticity in division times of cells in the colony as a feature engineering the activity dependence of enclave invasion. Our study uncovers several so-far hidden features of emergent organisation within bacterial colonies at intermediate scales, ultimately pointing at the biological benefits of proximity to close kith and kin.

Precision of morphogen-driven tissue patterning during development is enhanced through contact-mediated cellular interactions

Shakti Menon

A characteristic feature of robust pattern formation in developing embryos is that regions of distinct cell fates are separated by sharp boundaries whose location is invariant for each species. This is achieved despite fluctuations in the extracellular concentration of "morphogens" that provide positional information to the cells, and the intrinsic stochasticity of the intracellular processes that interpret these signals. One mechanism by which the accuracy of cell-fate decisions can be improved is through the sharing of information between neighboring cells, for example via contact-mediated signaling. Such communication is widely observed between proximal cells in the developing embryo, most notably through the evolutionarily conserved Notch signaling pathway. While it has been suggested that such signaling can play a role in noise regulation, and consequently in enhancing the precision of fate-boundaries in developing tissues, an explicit mechanism for how this may occur has thus far not been established. Here, we propose a conceptual framework for how the impact of noise on boundary demarcation can be attenuated via contact-mediated cell-cell interactions. We demonstrate how such signaling can lead to a decrease in cell fate uncertainty near the boundary between two domains, which allows for robust developmental outcomes. In particular, we find that the nature of interaction between the downstream effector of the intercellular signaling mechanism and the patterning genes is important in determining the extent to which such coupling can improve precision at the boundary.

Tuned Junctional Contractility can drive Spatial and Orientational Order

Raj Ladher

The reception and transduction of auditory information by the cochlea of the inner ear requires exquisite levels of organisation of its constituent cells: hair cells (HCs) and supporting cells (SCs). Each HCis intrinsically polar as the mechanosensory hair bundle, consisting of kinocilium and stereocilia, is eccentrically positioned on the HC apex. This intrinsic polarity aligns across the tissue axis such that the epithelium displays tissue-wide orientational order, ensuring a concerted and sensitive response to sound. HCs themselves are always surrounded by SCs, a spatial organisation that is necessary to maintain ionic homeostasis during transduction. HC and SC differentiation occurs while the auditory epithelium undergoes extension through cell intercalation via junctional remodelling. Even in the midst of these rearrangements, both spatial ordering and tissue-wide planar polarity emerge.

We find that tissue-wide force patterning generates both positional and orientational order. Positional order is established by differences between the mechanical activity of cell-cell junctions. These differences are encoded by cell-specific genetic programmes restricting the di-phosphorylated form of non-muscle myosin-II regulatory light chain (RLC) to specific junctions. The alignment of planar polarity emerges through local modulation of RLC phosphorylation through both tissue-wide cues and HC-intrinsic activities. This tuning results in patterns of tissue-wide force that couple spatial to orientation order. Our findings show that tuned mechanical asymmetries are sufficient to generate order in epithelia that are made of more than one cell type.

Competition Across Scales In Biology

Siddharth Goyal

Many biological phenomena emerge from interaction and competition between its parts. I will share some examples across biological scales where data-driven theory can reveal new rules of biological competition. At the molecular scale competition between mitochondrial genomes within budding yeast depends on genome architecture; dynamics of adaptive immunity in microbes reveal different modalities of competition and coexistence of bacteria and its phages; in mammals' cell fate dynamics, competing trajectories may be precursors to plasticity in development and disease. Going beyond, I will present some preliminary ideas on understanding the rules of memory and inheritance at the organismal scale and if competition may play a role in it.

Information processing in gene regulation

Marianne Bauer

Cells express genes in order to respond to environmental changes, differentiate or decide their fates, and develop into a healthy organism. Gene expression is regulated by cues, such as changing transcription factor concentrations, whose concentrations are often low. This expression in response to a changing concentration can be viewed as a type of decision that can be analyzed in terms of an information-theoretic framework. In this talk, I will show, on the example of early fly embryo development, how such an information-theoretic inference approach can help us understand features of a complex transcriptional apparatus that may be difficult to model, due to the complexity of the contributing regulatory factors. I will compare the inferred optimal sensor to realistic, microscopic models for regions on the DNA that respond to transcription factors, and, finally, relate their architecture to features commonly found in efficient computing systems.

How does resource presentation change adaptive trajectories?

Neetika Ahlawat

Adaptive trajectories of populations are dictated by the make-up of the environment in which they evolve. For instance, the adaptive trajectory in a glucose-limiting environment is distinct from when adaptation takes place in presence of glycerol. However, how do adaptive trajectories of a population change when the same resource is presented to the population in different packaging?

To answer this question, we evolve E. coli in three similar but non-identical environments where glucose and galactose were presented in the form of lactose, or melibiose, or as a mixture of glucose and galactose. Six lines of E. coli were evolved in each of the three environments for 300 generations.  
Phenotypic characterization of the evolved lines shows that melibiose evolved lines behave qualitatively differently than the lactose- and glucose-galactose evolved lines. Specifically, melibiose-evolved lines have a higher fitness in not only melibiose, but also in non-native environments (lactose and glucose-galactose). Thus, from the context of phenotypic adaptive response, resource packaging and presentation dictates adaptive trajectories.

Genome sequencing of evolved lines shows that adaptation in glucose-galactose, in each of the six lines, happens via mutation in either RpoB or RpoC, thus exhibiting convergent adaptation. On the other hand, all melibiose-evolved lines exhibit mutations in distinct genes, leading to adaptation.  This genetic diversity provides insights into challenges associated with predictability of evolutionary processes.

More importantly, we report a novel phenomenon, that the nature of resource packaging might alter the evolutionary trajectories of evolving populations.

Using genomics to investigate the role of mitochondria in mediating neoplasticity of stem cells

Swati Agarwala

Mitochondria are multifaceted cellular organelles and have now been well-established to play a causative role in cancer through mitochondrial structural and functional dynamics. The lab combines bulk and single cell transcriptomics with other approaches to investigate contribution of mitochondria in cancer etiology. Our literature review on alteration of mitochondrial properties during stem cell differentiation led us to a hypothesis on mtDNA regulation during stem cell differentiation (Agarwal et.al., in revision). Previously, the lab demonstrated that particularly low dose exposure to the model oxidative carcinogen, TCDD, promotes enrichment of mitochondria-primed quiescent neoplastic stem cells. There, scRNAseq) analyses indicated that an increase in neoplastic stem cell population is linked to prevention of reduction in mtDNA gene expression levels (Spurlock et.al, eLife, 2021). We aim to address if and how carcinogen-driven enrichment of neoplastic stem cells is mediated by early impacts of the carcinogen on mitochondria. Indeed, we found that 1 hr exposure to particularly the stemness enriching low TCDD dose causes dramatic loss of mitochondrial potential that sustains mitochondrial energetics. To study the impact of such early mitochondrial depolarization on gene expression, we performed scRNA-seq of the cells after 2.5 hrs exposure to TCDD doses. From the scRNAseq data we confirmed dose dependent increase in canonical TCCD induced genes. Furthermore, we identified 4 marked cell clusters, where the abundance of the CellCycle-Hi cluster and the Cytokine cluster decreased, and the CellCycle-Lo and the Keratin cluster increased with exposure to both low and high doses of TCDD. Interestingly, the stem cell enriching low TCDD dose caused a particular  increase in gene expression from the heavy strand of mtDNA across all the clusters and also increased the expression of the nuclear encoded mitochondrial ribosomal genes only in CellCycle-Lo cluster. In contrast, such TCDD exposure reduced gene expression from the light strand of mtDNA and also of the majority of nuclear genes coding for mitochondrial proteins. We are currently performing Weighted Gene Correlation Network Analyses (WGCNA) analyses to reveal the overall alteration in gene expression network and how that can potentially explain the specific mtDNA gene regulation towards establishment of neoplastic stem cell gene expression profile by low TCDD exposure.

Nandita Chaturvedi

The question of adaptation to environmental change at long time scales remains an important area of theoretical investigation within evolution and ecology, but has been treated mainly in the context of single environments. However, organisms almost always deal with multiple environments and trade offs arising from them, in addition to the possibility of long term change. Here we combine the idea of repeated variation or heterogeneity, like seasonal shifts, with long term and directional dynamics. To address this complex situation, we extend the framework of fitness sets and study how the optimal phenotype in this situation can itself change with long term shifts. We consider selection from two distinct environments. We find that the behavior of a population under such a system is qualitatively different and more complex than that of a population responding to long term change in a single environment. The chance of survival or extinction depends crucially on the relative frequency of the two environments, the strength and asymmetry of their selection pressure in addition to population size, phenotypic diversity, fecundity and the rate of change of the environment. We study characteristics of the population under selection such as its phenotypic lag behind the optimal phenotype and the mean time to extinction.

Optimal reliability, robustness, and control of nucleus centering in fission, yeast is contingent on nonequilibrium force patterning

Ishutesh Jain

Cells, such as fission yeast, center their division apparatus to ensure symmetric cell division, a challenging task when the governing dynamics is stochastic. We show that the spindle pole body (SPB) positioning, which defines the division septum, is controlled by the patterning of nonequilibrium polymerization forces of microtubule (MT) bundles. We define two cellular objectives, reliability, the mean SPB position relative to the geometric center, and robustness, the variance of the SPB position, which are sensitive to genetic perturbations that change cell length, MT-bundle number/orientation, and MT dynamics. We show that the optimal control of reliability and robustness required to minimize septum positioning error is achieved by the wild-type (WT). A stochastic model for the MT-based nucleus centering, with parameters measured directly or estimated using Bayesian inference, recapitulates WT optimality. We use this to perform a sensitivity analysis of the parameters that control nuclear centering.

Establishing the fundamental limits of spatially resolved cellular-time decoding from noisy single-cell gene expression measurements

Shaon Chakrabarti

An oscillator driven by a periodic Zeitgeber gets entrained with a phase difference that varies depending on the underlying endogenous period of the oscillator. Intriguingly, this physical phenomenon is thought to give rise to variations in "body time" between individuals and likely even between tissues within an individual. An exciting open challenge in the field of chronobiology and chrono-medicine is to accurately measure body or tissue time, to then allow for personalized, circadian-time based development of therapeutics.

Here I will talk about our lab's recent efforts in establishing the fundamental limits of decoding time from gene expression measurements. Given the well-established phenomenon of super-Poissonian noise in single-cell gene expression, the lowest resolution at which cellular time can be accurately decoded from mammalian cells is currently unknown. To address this, I will first discuss the experimental approaches we have developed to simultaneously measure RNA levels of multiple genes in mouse embryonic fibroblasts, at single cell and single molecule resolution. To analyze such datasets we have developed algorithms based on stochastic processes, to first learn the patterns of noisy gene expression and then infer single cell time of test samples. Using these approaches, I will quantitatively demonstrate that decoding time from single cells is not possible. Remarkably however, averaging over as low as 20-30 cells allows for time-inference with just 1-hour error with only 3 genes, thereby providing a conceptual framework for spatially resolved circadian-time inference.

Physiological adaptation and homeostasis at the single-cell level

Yuichi Wakamoto

Exposing biological cells to environmental and genetic stresses can exert global changes in gene expression and metabolic profiles. These changes can vary even among genetically identical cells, and a fractionof cells eventually adapt and reach the states that permit cells long-term survival or even stable proliferation. To understand the mechanisms of cellular adaptation with global changes in gene expressionand metabolic profiles, we need to answer the following crucial questions:

(1) What factors determine which cells adapt and which do not? 2) Are adapted states to an identical stress solitary or plural? (3) Which parts of the molecular profile in cells are altered during  adaptation? To gain insight into these questions, we are studying bacterial and cancer cells' adaptation to chemical stress, such as exposure to antibiotics, at the single-cell level. In this workshop, we present experimental results on the adaptation of Escherichia coli, which expresses a chloramphenicol-resistant protein tagged with a fluorescent protein, to chloramphenicol exposures. We show the time-dependent significance of the expression noise of the resistant protein, the commonality and diversity of the adapted states, and the role of stoichiometry recovery of core components in cells. Additionally, we introduce a new Raman-based omics profiling method that can potentially be used for live single-cell analysis. We discuss how this technique can further our understanding of cellular adaptation.

Plasticity and stochasticity across spatial scales

Steffen Rulands

Biological systems operate far from thermal equilibrium. They import energy from the environment to build and maintain complex structures in the presence of noise. They also have the remarkable capacity to break up and rebuild these structures rapidly, such as during wound healing. In this talk, I will show how biological systems integrate processes on different spatial scales to regulate stochasticity and achieve plasticity. I will give specific examples ranging from social insects to  molecular signaling.,