Neurodegenerative diseases such as amyotrophic lateral sclerosis (ALS) affect motor neurons. The degeneration of these neurons is typically characterized by the abnormal cytoplasmic deposition of protein aggregates. These pathological aggregates are amyloid-like, and evidence suggests that they evolve from physiological stress granules (SGs). Thus, to understand how SGs mature into disease causing aggregates is highly important. In the proposed project, I aim to unravel these mechanisms by establishing a motor neuron-like cell line for the comprehensive analysis of SG maturation. To do so, I will take advantage of preliminary data in yeast and develop protocols for the maturation of physiological dynamic SGs into solid-like and aberrant amyloid-like states in motor neuron-like cells. This process will be characterized by fluorescent microscopy and cryo-tomography, and biochemical methods such as limited proteolysis coupled to mass-spectrometry (LiP) to visualize structural alterations. Physiological, solid-like and amyloid-like states will be purified and analysed by mass spectrometry to reveal alterations in the SG composition and in the post-translational modifications (PTMs) of SG components. For the observed changes, I will identify whether they are causative for SG maturation. Finally, applying an advanced mass-spectrometry method, the "sentinel protein assay”, will allow me to identify cellular pathways that are specifically activated during SG maturation. I believe that these results will not only provide important mechanistic insights into SG maturation but may also help to develop innovative therapeutic strategies to prevent this transition and the onset of ALS.

In this project, we will use model microbial communities based around the degradation of polymers abundant in marine ecosystems. Raman microscopy is a hyperspectral imaging method that can be used to profile the chemical environment of live samples, and will be used to record the progression of a model microbial community within a microfluidic device. Our device contains a single polymeric primary substrate as well as a number of ports that allow for introduction of soluble substrates. By integrating this microenvironment with Raman microscopy we intend to clearly document the biochemical progression of the microbial community during primary succession and colonization of the substrate. The use of this model environment will be leveraged into a general method identify and predict state transitions in real-time as the community moves through the phases of succession, with the goal of being able to elucidate the biochemical drivers of these transitions, and subsequently demonstrate the validity of these drivers by perturbing the natural progression of these state transitions.

The complexity of cellular machineries has been forged by the motors of evolution (mutation, natural selection, and genetic drift), with boundaries that are determined by the properties of biological molecules and the biophysical limits of the cell. Our goal in this project is to determine what are the relative contributions of biophysics and evolution in shaping this complexity. Our model of interest is cell signaling, which is rapidly evolving and involves highly interconnected networks of proteins that sense and process signals and make real-time decisions to satisfy cellular needs and promote survival and proliferation. Because typical cells contain from millions to billions of protein molecules, signals can reach out to unwanted proteins in an unspecific manner. How much can evolutionary forces optimize this process so that cells can reach biophysical optimality by minimizing unspecific interactions is currently unknown. This question can be answered neither from physics nor from biology alone, and thus requires a multidisciplinary team. Our international team with expertise in theoretical physics (Gasper Tkacik), evolutionary biology (Christian R Landry) and biochemistry (Judit Villen) is poised address this question. We will generate a mathematical model of signaling network evolution that includes key biophysical and evolutionary driving forces to examine how each contributes to create unspecific signaling interactions in organisms with finite population sizes. This model will guide experiments in which we will manipulate the amount and extent of unspecific interactions and measure its biochemical impacts and its effects on fitness using experiments and bioinformatics data integration to calibrate parameters of the model. In biophysics, it is often assumed that biological systems have reached an optimum attained by natural selection and are only limited by their physical characteristics. Evolutionary biology considers that natural selection works against genetic drift and mutations to bring biological networks near their optimum, which is unknown. Our work between these two fields will reveal how close to the biophysical limits can signaling networks get and how biophysical and evolutionary forces are integrated to shape accurate signal transduction.

Heat produced by living species has been known for centuries but viewed merely as a factor contributing to body temperature. Recent progress in ultra-local thermometry at a cellular scale has challenged this concept. Reports on heat release by active ion channels and pumps and other pioneering experiments and theoretical estimates have fuelled discussion about the significance and the magnitude of endogenous heat. Discoveries of the faster neurite growth aligned with a thermal gradient and heat-pulse induced muscle contractions without Ca2+ influx suggest that cell hot spots may channel signalling and thus contribute to cell and tissue differentiation. We will study the role of heat in three-dimensional cell shaping and differentiation and test the hypothesis that localized heat signals can affect morphology in cells other than neurons. We will aim to resolve the current controversy of intracellular thermometry, the drastic mismatch between theoretical estimates based on near-equilibrium thermodynamics and reported experimental values of temperature rise. We will investigate the signalling activity of artificial hot spots introduced into or close to cells and measure thermal activity of natural hot spots on cell membranes. We will develop a reliable intracellular nanothermometry system and an ultra-local heat-shock trigger-and-measure method to create and control the ultra-local hot spots with light. The method will use a synthetic complex of several nanoparticles, one as a heater and the others as temperature sensors. For the heater, we will use a gold nanorod whose heating can be controlled by the wavelength and polarization of light focussed on the rod. The temperature will also be measured optically, with nano-sized crystals of diamond doped with photoluminescent defects (nitrogen-vacancy centres being the primary candidate fro this role). The expertise of the applicants provides the required synergy between advanced nanotechnology, biophysics and cell biology to achieve these goals. The project starts simultaneously at four nodes and efforts will increasingly integrate as the investigation progresses.

Heteromeric amino acid transporters (HATs) mediate the antiport of amino acids across the cell membrane, and play pivotal roles in amino acid absorption, neurotransmission, and metabolite supply to cancer cells. HATs are unique among all solute carrier (SLC) transporters, in that they are composed of the two subunits, the ‘helper’ heavy chain and the ‘catalytic’ light chain, linked by a disulfide bridge. Malfunction of HATs are implicated in hereditary diseases like cystinuria and lysinuric protein intolerance, and thus HATs are important pharmacological targets for both hereditary and acquired diseases. However, the structures and mechanisms of HATs are poorly understood. The crystallization of HATs proved extremely challenging, likely due to extensive glycosylation and difficulty in preparing heteromeric complexes in general. I propose to use single-particle cryo-electron microscopy (cryo-EM) to solve the first three-dimensional structure of a HAT in its full dimeric assembly. Conventional cryo-EM suffers from a practical size barrier at ~150 kDa due to low image contrast, and a HAT (~125 kDa) represents one of the most challenging targets. I will employ cutting-edge techniques like Volta phase plate, antibodies and nanodiscs to overcome the issue. I aim to solve the structures in multiple conformations, (1) apo, (2) substrate-bound and (3) inhibitor-bound states, to gain fundamental insights into the heterodimer formation, substrate transport and pharmacological inhibition. With concomitant biochemical experiments, these studies will facilitate understanding and therapeutic application of this important SLC transporter.

Growing flat leaves is crucial for plant success, since it directly affects light interception and photosynthesis. This process is tightly regulated by endogenous signals revealing a strongly genetically defined developmental program, but is also modulated by the light environment. As shown using null mutants, both blue light receptors phototropins (phot) and red/far-red light receptor phytochromeB (phyB) affect leaf flattening, and it was recently found that plants lacking PKS3, a protein involved in phot signaling, suffer severe flattening defects. The main objective of this project is thus to determine how light signals interact with endogenous cues to control development of a flat leaf. First, I will perform an exhaustive temporal and spatial characterization of the phot and phyB-mediated light control of leaf development applying various light treatments and complementing phot1phot2, phyB and pks3 with tissue-specific promoters. Next, I will apply the latest microscopy techniques to determine the expression pattern of leaf development markers under various light conditions and in photoreceptor mutant backgrounds. Also, given the major role of auxin in leaf development, and the various ways in which phot and phyB control it, I will study auxin abundance, transport and sensitivity at the tissue and cellular level. Finally, I expect to identify new components of phot signaling evaluating the interactome of PKS3 through IP-MS. Thereby I expect to determine how a conserved developmental program and photosensory cues crosstalk to control leaf flattening.

Sexual behavior is fundamental for evolution and an important component of human well-being. Ejaculation is a critical mechanism in male sexual function, which is hypothesized to be controlled by a neural circuit known as the 'spinal ejaculation generator' (SEG), located in the lumbar spinal cord. The SEG has also been hypothesized to control the post-ejaculatory refractory period, a phase of sexual satiety, through ascending projections to the brain. However, the mechanisms by which the SEG controls ejaculation and the refractory remain elusive. We will tackle this problem first by using anatomical tracing to find the spinal neurons that provide input to the ejaculatory muscles and immunochemistry to characterize their molecular identity. We will use this information to provide specific genetic access to SEG neurons. By expressing fluorescent probes in these neurons, we will be able to obtain targeted electrophysiological recordings and use them to characterize the intrinsic and synaptic properties of the SEG circuitry in vivo. By expressing genetically encoded calcium indicators, which provide fluorescence probes of activity, we will monitor the activity of the SEG during sexual behaviour. Finally, by expressing and optically stimulating excitatory and inhibitory opsins in SEG neurons or their terminals, we will test their causal role in triggering both ejaculation and the post-ejaculatory refractory period. This project will provide the first mechanistic description of the key neural circuitry controlling male ejaculation and the refractory period, with potential implications for the treatment of sexual dysfunction.

Microbial communities affect us in many important ways, from health and disease to food production, industrial fouling, agriculture, and the environment. There is a great need, therefore, to both understand and control microbial communities. The challenge is that these communities often contain many dynamically-interacting species, making them complex systems that are difficult to predict. To understand microbial communities, we need breakthroughs in advanced theoretical methods. The current state-of-the-art treats microbes as static networks but microbial communities are dynamic systems where species interactions can rapidly change and shift a community to an undesirable state. To deal with this, I will develop a new dynamic network theory to both understand and control microbial communities. Firstly, I will analyse how the key properties of microbial communities are influenced by different sources of network dynamics. I will study both microbe-driven changes in network structures, e.g. ecological succession, and changes driven by external factors, including the abiotic environment and host factors for symbiotic communities. I will then apply control theory to these dynamic networks in order to understand how to stabilize dynamic networks and shift networks from an undesirable (e.g. disease) to a desirable state. As a final step, I will work with my host laboratory to design experiments and test the ability of my theory to both predict and engineer gut communities for specific outcomes. In this way, I aim to make major contributions to both microbiology and complex system theory.

Complex human diseases are the result of a combination of genetic, environmental and lifestyle factors, and the development of new treatments for such diseases has been challenging. Large-scale screens in pharmaceutical industry have essentially not changed over the last 50 years and many drug candidates fail in clinic, as animal testing is mostly performed on animals with a single genetic background, compromising translational utility. While multiplexing experiments in vertebrate animal models with different genetic backgrounds seems unrealistic, small model organisms, such as C. elegans, are emerging as valid alternatives to cells and inbred mouse models in drug development, significantly lowering time and costs of research. My host laboratory has developed an automated platform for roundworm culture and high-throughput phenomic data collection and analysis. I here propose to use this platform to map mitochondrial stress pathways, such as the UPRmt, in a large C. elegans genetic reference population and in depth characterize selected candidate hits from this genetic screen through the study of loss-of-function worm models in vivo. This will allow me at the same time to develop and validate protocols for high-content and high-throughput phenotypic analysis of C. elegans. Furthermore, my project will spur replacement, reduction, and refinement (3Rs) of vertebrate animal testing and will promote the roundworm (and eventually other small organisms) as new model organism for high-throughput biology as well as for drug discovery.

Mammalian pregnancy is a unique physiological context that undergoes substantial hormonal, metabolic, microbiota and immunological changes. Epidemiology studies have shown that pregnant women have altered susceptibility to various infections. These environmental exposures could have profound and long-term effects on the offspring with impaired growth and disorders. Despite ample evidence linking the maternal immune perturbation to offspring outcomes, the mechanistic pathways for this in utero influence remain unclear. Further, little is known about the impact of these microbial exposures on the offspring immune system. Moreover, how distinct microbiota could imprint the offspring immune system for the long term remains unknown. The objective of this grant is to set up murine maternal immune activation models via infectious diseases that commonly occur during human pregnancy, including influenza virus and Listeria monocytogenes infections, to understand the impact on immune ontogeny, microbiota composition, infectious and inflammatory diseases susceptibility of offspring in neonatal and adult stages. We will dissect the mechanistic links of maternal-fetal crosstalk through vertical transmission of maternal cytokines, microbiota and/or epigenetic regulation with an ultimate goal to rescue the potential detrimental consequence in offspring. Further integration of our findings from murine models to human samples will provide significant impact in the clinic. Insight into maternal-fetal immune crosstalk is a fundamental line of research with profound implications in understanding and prevention of complex immune disorders that are on the rise in high-income countries.

Cells continually encounter a variety of suboptimal conditions which restrict growth and proliferation. In response to such stressors, proper adaptive mechanisms are typically activated. Adaptive mechanisms can be categorized into two groups, specific and general. The specific responses, such as DNA repair or unfolded protein response, directly deal with the primary cause. By contrast, a general response is assumed to inhibit growth and render cells highly tolerant to the stress as a dormant state of an organism. While most studies have focused on the specific effectors, little is understood about how cells initiate and maintain the central stress response pathway. By analyzing high-throughput sequencing data available in our lab on various types of stress conditions, I have found several genes that are commonly regulated in mammalian cells. I hypothesize the common genes may modulate a general stress response, which protects cells from stressors. To understand the role of those candidates, I will 1) reveal molecular mechanisms underlying the common stress program and 2) evaluate their biological significance in vivo. With biochemical and computational approaches combined with loss of function studies, I will determine their targets to unveil regulatory networks established in the common stress program. Finally, I will perform in vivo experiments with various stresses to confirm whether the candidates function in the physiological context. By understanding the core stress response, this research will address an important but often overlooked aspect of stress response and provide a fundamental understanding of cell survival and maintenance.

Post embryonic growth and regeneration require new cells to integrate into the network of existing cells. In the case of the brain it is extremely important to maintain and restore circuit function, such that behaviors can be executed correctly. To study such processes a system that fulfills both post-embryonic growth as well as regeneration is needed. Furthermore, since neural circuit functions are highly dynamic the need for in vivo approaches is evident. In mammalian systems where neurogenesis occurs throughout life in restricted brain regions such approaches are technically challenging. Here, I propose to use the axolotl (Ambystoma mexicanum) brain to understand the dynamic remodeling of neural circuits through the addition of newborn neurons during post-embryonic growth and regeneration. I will establish a high resolution intra-vital in vivo imaging approach to address the dynamics of neurogenesis and remodeling of neural circuits in the telencephalon and the optic tectum. Using this setup in combination with genetically encoded calcium indicators I will generate activity maps of these regions in the presence or absence of defined stimuli and address their functional remodeling during growth and their restoration after injury. Furthermore, I will develop and adapt methods to visualize the functional connections of neurons in axolotl. Together, this work will provide fundamental insights into the functional maintenance and regeneration of neural circuits and the dynamics of how new neurons integrate in vivo which has not yet been achieved in any vertebrate system.

Basic biological processes such as cell migration, tissue morphogenesis, and hearing rely on mechanotransduction, i.e., the conversion of mechanical forces into electrical and biochemical signals that can be used by individual cells or multicellular organisms to survive. Adhesion complexes bearing mechanical force are often needed in mechanotransduction, with several protein families that include integrins and cadherins directly transmitting force signals to sites of transduction. Classical cadherins, key for tissue morphogenesis and integrity, provide calcium-dependent bonds involved in cell-cell contact formation, in signaling to the actomyosin cytoskeleton, and in mechanical stabilization of cell-cell adhesion. Non-classical members of the family are also involved in mechanotransduction, with cadherin-23 and protocadherin-15 forming an adhesive “tip link” essential for the vertebrate senses of hearing and balance. Multiple structural and biophysical studies have elucidated the unbinding, unbending, and unfolding strength of single molecules and adhesive complexes, including those formed by cadherins. However, to the best of our knowledge, nothing is known about how the mechanical properties of adhesion complexes have evolved. Here we aim to study the mechanical evolution of adhesion complexes using inner-ear cadherin tip links as a model system. Tip links initiate sensory perception by directly conveying tension to inner-ear mechanotransduction channels on a microsecond time scale. Fast and effective transmission of mechanical force must be one of the selective factors that dictate the molecular evolution of tip links. We will sequence, synthesize, and study tip-link proteins from diverse extant species and will carry out evolutionary analyses both to reconstruct the sequences of ancestral, “resurrected” tip links, and to identify potentially stronger tip links from species in which positive selection is reported. In parallel, we will use biophysical techniques to determine equilibrium binding affinities and structural properties of modern tip links from various species and ancestral tip links at different evolutionary stages. High-speed force spectroscopy and molecular dynamics simulations will be combined to probe tip link mechanics and strength across species and during evolution. Our interdisciplinary work will provide a unique evolutionary view of adhesion mechanics.

A driving goal of regenerative medicine is to use stem cells to replace worn-out cells, tissues and organs. Stem cells could in principle be instructed to generate all cell types required. However, two crucial questions must first be addressed: 1) How can we precisely target specific cell types with these signals to control differentiation? 2) What molecular signals instruct cells to differentiate, sequentially, into cell types of interest? These fundamental questions have been difficult to address in a generalizable way. However, two new advances open up powerful new ways to address them. First, a new design principle for signaling systems shows how combinations of seemingly redundant ligands and receptors could be used to target signals to specific cell types. Second, a new technique called MEMOIR enables cells to actively record their own signaling histories within their genomes, providing information on which signals are perceived within each cell lineage. Here, I propose to combine my own experience in stem cell research with Michael Elowitz’s expertise in synthetic biology to develop technology for mapping and manipulating the processes by which cells acquire their identities. More specifically, I will analyze promiscuous receptor-ligand interactions in the FGF or ErbB receptor families to better understand how the use of ligand combinations could enable more cell type specific activation of these pathways. In addition, I will implement MEMOIR in developing neurons to extract the signaling histories that accompany their differentiation. Knowledge of the way differentiating cells process signals may enable more precise control over cell identity.

The role of midbrain dopamine neurons in learning has been extensively studied over the last two decades, mainly focused on the commonly observed reward prediction error(RPE) activity, i.e. the discrepancy between expected and obtained reward. Still, the field is going through a critical turning point as increasing evidence shows heterogeneity of dopamine neurons’ responses, opening the possibility that dopamine could modulate learning in multiple ways besides through a RPE. This proposal aims to understand if distinct dopamine projections to the striatum, a structure involved in motivation and action selection, play differential roles in learning. For example, dopamine projections to the ventral striatum signaling a RPE could drive the encoding of value observed in this area, while those to the dorsal striatum could influence the encoding of performance vigor, and those projecting to the tail of the striatum could involve this area in attention. I will also test the hypothesis that corticostriatal plasticity follows common rules across the striatum, but striatal activity will vary across regions due to heterogeneous information conveyed by dopamine innervations. I will use molecular techniques to image simultaneously the activity of dopamine axons and of principal neurons in various striatal regions with two-photon microscopy while mice perform a sensorimotor go/no-go task to achieve rewards and avoid punishment by modulating their gait. This proposal will elucidate fundamental mechanisms of dopaminergic control of associative learning and motor activity, providing relevant results for the understanding of neurodegenerative diseases such as Parkinson and Huntington’s.

Trypanosomatids are parasitic protozoa causing a range of devastating diseases. Many mitochondrial genes of these pathogens are transcribed as precursor mRNAs that require insertion and/or deletion of uridylate residues to become functional. The RNA editing reactions are catalysed by the editosome, a large molecular machinery present in the trypanosomatids mitochondria. Editosomes are parasite-specific complexes representing ideal therapeutic targets, however their structural organisation and mode of action are not fully understood. The aim of the proposed research is to obtain a comprehensive structural and functional characterisation of the trypanosomatids RNA editing machinery. To achieve this, I will first produce genetically modified Trypanosoma brucei cell lines expressing tagged editosomal proteins that will allow for rapid and efficient purification of editosome variants under native conditions. I will then use cryo-electron microscopy and single particle analysis to determine high-resolution structural models describing the organisation and the interaction network of the purified editosomes. By determining the structure of the complexes actively engaged in the editing reactions, I also aim at elucidating the conformational states associated with the individual steps of the enzymatic cascade. The proposed studies will provide a broad characterisation of the molecular architecture of the editosomes that will be crucial both for the fundamental understanding of RNA editing and the development of therapeutic strategies against trypanosomatids pathogens.

Bacteria play a ubiquitous role in the ecological success of many plants, algae, and animals. Much of this success reflects the capacity for bacteria to provide resources to their hosts that are otherwise scarce or inaccessible in the immediate environment. Microalgae of the genus Symbiodinium are found globally in tropical and temperate coastal ecosystems, forming symbioses with many invertebrates but notably reef forming corals. Over 50 years of research has shown immense functional divergence amongst the genus Symbiodinium. This functional divergence has commonly been ascribed to fundamental differences in gene acquisition and expression of resource exchange and stress tolerance. However, I hypothesize that obligate associations with bacterial consortia regulate Symbiodinium fitness and functional diversity over space and time. As such, mutualistic bacteria may act as essential “resource surrogates” for Symbiodinium when they live transiently as free-living cells outside of their host corals. The type and health of the algae–bacterial community interactions are entirely unrecognized but in fact, determine the survival of Symbiodinium, and hence the future for reef systems worldwide as they are subjected to accelerating anthropogenic stress. To address this complete unknown, I will use a combination of state of the art techniques (e.g. NanoSIMS, chemotaxis flow imagers and bioreactors) to characterize key ecologically-relevant Symbiodinium-bacterial associations and identify key metabolic networks that promote Symbiodinium ecological resilience. This study will transform our understanding of what makes a healthy coral symbiont, and hence the fitness of reef ecosystems.

Perception is shaped both by sensory information and internal variables. Our brains uses regularities of the environment to make predictions, and combines them with sensory stimuli to form percepts. However, how and where is statistical knowledge of the environment stored and learned remains unknown. Visual perception arises from a set of hierarchically-organized cortical areas. Abundant feedback inputs project from higher to lower areas but their role in perception is unknown. In this proposal, I will test if knowledge of the world is stored in the connectional specificity of feedback projections. I hypothesize that the tuning-specific wiring pattern of feedback projections terminating in mouse primary visual cortex reflects spatiotemporal visual statistics of the environment. First, I will assess if sensory experience plays a role in visual feedback wiring. To do so I will visually-deprive mice or raise them in artificial environments with altered visual statistics. Using a novel combination of dual-color optical recordings, I will measure how connectional specificity of functionally-characterized feedback axons in primary visual cortex relates to the animal’s visual experience. I will also determine if spatiotemporal patterns of correlated neuronal activity are sufficient for instructing the observed wiring organization. These experiments will unveil if feedback inputs are a neural substrate for storing learned regularities of the world. By doing so, they will shed light on the computational role of this enigmatic connections while providing a mechanistic description of the internal factors shaping perception.

Specific regulation of genes is essential for cellular life. In mammals, transcription factors play a critical role in this process as they distinctly mark regions of the genome for expression and are the pillars of cell identity. Reprogramming of a terminally differentiated cell into a pluripotent cell can be accomplished by mere expression of a few transcription factors. These transcription factors regulate developmentally silenced regions of the genome and initiate reprogramming. Fundamentally, how a transcription factor is able to target specific DNA sequences within chromatin is still largely a mystery. Oct4 is one of the key factors in this ‘pluripotent cocktail’ and it can directly interact with nucleosomes. Most transcription factors cannot interact with nucleosomes because of the inaccessibility of DNA wrapped around histones. Pioneer transcription factors, such as Oct4, are able to access these restricted regions of DNA, yet we still do not know how. I propose to use biochemical and structural techniques to elucidate the molecular assembly of Oct4 and chromatin. The results of my proposed work aim to uncover the molecular determinants of a transcription factor for association with nucleosomes and the mechanism by which a transcription factor complex may initiate nucleosome rearrangement/eviction to recreate cell identity.

All of our movements require the coordination of dozens of muscles to ensure that we reach our desired goal. Indeed, the ability to perform complex motor corrections that depend on the context, termed flexible feedback control, is a fundamental aspect of human behavior. Three brain areas essential for the preparation and execution of reaching movements are the interconnected dorsal premotor, primary motor, and primary somatosensory cortices. These areas modulate their activity to reflect the context of upcoming actions, but the roles of each area and the underlying computational principles are poorly understood. To address this question, macaque monkeys will be trained to reach targets in virtual reality using a haptic feedback device in response to various mechanical loads and sensory expectations while neural activity is recorded from over 1000 sites simultaneously. To dissect the computational principals of this complex control problem, a modular neural network model of the circuit will be trained to complete goal-directed tasks by learning to control a 50-muscle biomechanical model. Finally, neural projections between these brain areas will be selectively manipulated using optogenetic stimulation to dissect the roles of each of these areas during feedback control. The results of these three experiments will provide a rich bi-directional transfer of information between theory and experiment and significantly improve our understanding of motor control.