In bacteria, gene expression is subject to large fluctuations among single cells (noise), even when genetically equal cells are exposed to identical conditions. Stochastic processes underpin a variety of bacterial behaviors suggesting that bacteria can exploit noise to transition between alternative phenotypes. Conversely, other regulatory networks are known to be extraordinarily refractory to noise, enabling synchronized, coordinated population-wide behaviors. I will study how such robust gene expression is achieved in regulatory networks by investigating quorum sensing (QS), a process of bacterial cell-to-cell communication. To date, investigations of QS have focused on steady-state measurements of single transcriptional reporters. Importantly, QS is known to affect gene expression globally, demanding rapid, genome-wide profiling techniques to simultaneously track the time-dependent expression of all affected genes. I will define the complete set of genes controlled by QS, in what order each gene turns on or off, and which QS regulators determine each particular expression pattern. I will also identify which QS-controlled genes exhibit noise and which show robust regulation, and define the key regulatory components that confer these features. I will perform these analyses in both individual cells and in populations. These data will be used to develop theoretical models to predict mechanisms allowing QS-signal integration and network performance. Identifying the components that allow synchronized population-wide performance in bacteria will be crucial for understanding the evolution of multi-cellularity in higher organisms, where coordination among cells is vital.

Decision making is the process resulting in the selection of a course of action among several alternatives. Decisions that are mainly based on the sensory evidence (perceptual decisions) constitute a valid model for studying how and where in the brain this process is accomplished. The neuronal correlates of perceptual decisions have been studied mostly by recording the activity of single cells from different cortical areas while primates perform perceptual tasks. It has been found that the trial-to-trial variability of neuronal activity is correlated with behavioral choices and this finding has been interpreted as evidence of a causal effect of sensory neurons on choice. However, recent findings challenge this interpretation and suggest that the direction of causality could be the opposite. This has fundamental implications for neuronal coding: if correlated random fluctuations in the sensory periphery bias behavioral choices, then noise in neural activity significantly constraints information processing; if correlated fluctuations in sensory areas reflect cognitive operation unknown to the experimenter, then they need not generate any constraint on information processing. To clarify this issue, we propose to record the activity of large neuronal populations simultaneously in the auditory cortex and the medial prefrontal cortex of rats performing auditory discriminations. By recording the activity of large neuronal populations, we will be able to quantify the time-course of decision-related activity in single trials on each area. This will allow us to perform the critical comparison to determine causality: where does decision-related activity appear first?

The last two stages in eukaryotic translation, termination and ribosome recycling, are very distinct from the corresponding, well-characterized stages in bacteria. Thus, eukaryotic termination results from the complex functional interplay between two release factors, eRF1 and eRF3, in which GTP hydrolysis by eRF3 couples codon recognition and peptidyl-tRNA hydrolysis by eRF1, whereas eukaryotic ribosome recycling is mediated by ABCE1, a unique highly conserved member of the ATP cassette family of proteins that contains an iron-sulfur cluster domain, and which does not exist in bacteria. Moreover, in contrast to bacteria, eukaryotic termination and ribosome recycling are intimately linked because eRF1 plays a key role in both processes. Although these two stages of eukaryotic protein synthesis are of major importance in regulation of eukaryotic translation, and are linked to downstream events such as reinitiation and mRNA surveillance, their mechanisms have only recently begun to attract attention. However, future progress in elucidating the molecular mechanism of eukaryotic termination and ribosome recycling is strongly compromised by the lack of structural data. The primary aim of this proposal is therefore to obtain a comprehensive overview of the architecture of defined intermediate states in termination and ribosome recycling and to model the transition between them. To achieve this goal, we shall integrate the complementary expertise of four leading laboratories in cryo-EM, X-ray crystallography, computational biology and functional studies.
The specific aims of the proposed research include: (i) elucidation of the molecular basis for the omnipotent decoding capacity of eRF1 using X-ray studies of modeled eRF1-associated 80S ribosomal complexes containing different stop codons in the A-site, (ii) determination by cryo-EM of the structures of eukaryotic ribosomal complexes corresponding to different stages of termination, from initial ribosomal attachment of release factors through GTP hydrolysis by eRF3 to the final stage corresponding to post-termination complexes containing deacylated tRNA in the P-site, (iii) visualization by cryo-EM of ABCE1 in post-termination complexes and identification of its critical functional interactions with 80S ribosomes and eRF1, (iv) computational analysis and dynamic simulation of ribosomal transition between individual stages in termination and ribosome recycling based on X-ray and cryo-EM results, and (v) functional investigation of the mechanisms of mammalian reinitiation, regarding its importance as a regulatory post-termination event. In summary, the results that are anticipated to emerge from the proposed studies will for the first time provide an integrated structural and mechanistic outline of the last two stages in eukaryotic translation.

Amyotrophic Lateral Sclerosis (ALS) is a neurodegenerative condition that targets primarily motor neurons, resulting in progressive paralysis and death within a few years from onset. Despite decades of intensive research, no central pathway or mechanistic basis for this motor neuron degeneration has emerged. Consequently, there is still no successful treatment for ALS, which affects more than 1 in 100,000 individuals each year. Within the past 5 years a paradigm shift in our understanding of neurodegeneration has occurred with the implication of the RNA-binding proteins TDP-43 and FUS, in the pathogenesis of ALS and a broad spectrum of neurodegenerative diseases. TDP-43- or FUS-containing cytoplasmic inclusions are found in ALS, and several other neurodegenerative conditions, accompanied by disturbance of the subcellular localization of these proteins. Mutations in either TDP-43 or FUS cause ALS or, in rare cases, frontotemporal lobar degeneration (FTLD), the second most common type of dementia after Alzheimer’s disease. Emerging evidence highlights the role of prion-like, template-directed misfolding of pathogenic proteins associated with various neurodegenerative diseases, in the spread of protein aggregates, pointing to mechanistic parallels underlying neurodegeneration. I plan to capitalize on my expertise in both prion diseases and ALS in order to study the mechanisms of progressive neurodegeneration and spreading of TDP-43 and FUS protein misfolding. I intend to take advantage of the recently shown prion-like properties of these proteins to develop an innovative ex vivo model of ALS, using mouse organotypic slice cultures.

Understanding the neural basis of sensory percepts has been one of the main driving forces in neuroscience. One traditional approach has been to reduce complex sensory stimuli to its elemental components while monitoring activity from individual neurons. However, it has become clear that in order to fully understand how sensory percepts are created in higher-order brain regions such as the cortex, it is key to study behaviorally relevant sensory stimuli in awake behaving animals while simultaneously recording from large neural ensembles. Olfaction is one of the most dominant senses across many organisms. For rodents, the sense of smell is particularly crucial. The context in which odors are experienced is not only critical for decoding the behavioral significance of the odors but also powerfully influences their perceptual qualities. This proposal aims to understand how contextual information about odors is encoded in the primary olfactory (piriform) cortex. Recent advances in genetic engineering and mouse behavior provide unique and unprecedented opportunities to precisely examine this question. I will create a novel spatial context-dependent olfactory discrimination task in a virtual reality environment for head-restrained behaving mice. I will then perform in vivo extracellular multiunit and patch-clamp recordings with optogenetic manipulations during behavior. In combination with anatomy and brain slice recordings, I aim to determine neural mechanisms for encoding spatial contexts and odor identity in populations of piriform cortical neurons and thus reveal fundamental principles governing the emergence of holistic and integrated sensory percepts in the brain.

The dauer diapause is an alternative stage in Caenorhabditis elegans development that is entered in case of unfavorable environmental conditions. These “dauer larvae” can reside in this stage for many months without feeding or aging and develop into reproductive adults with normal lifespan in case of improved environmental conditions.
A complex signaling network regulates entry and exit of the dauer diapause. This network has been extensively studied for many years and key players like TGF-beta, insulin or steroid signaling as well as their interplay have been identified. As a next step, I propose to image for the first time larvae exiting dauer diapause in their entirety with high spatial and temporal resolution using Selective Plane Illumination Microscopy. I want to characterize morphological changes from tissue level down to the migration of individual cells. Using BAC-recombineering, I will fluorescently label key components of the signaling network, characterize their target cells and correlate their activity to the onset of morphological change. To identify the precise timings of transcriptional activation, I plan to directly detect synthesis of mRNA using the MS2-GFP system, which will be detected by high-resolution confocal microscopy. All measurements will be integrated into one common digital atlas that will be initially created using ion beam scanning electron microscopy. This system will allow a more comprehensive understanding of the complex signaling processes during dauer diapause and how they regulate morphological changes.

Listeria monocytogenes is a clinically important pathogen that causes enteric infections, which can sometimes spread to the brain or across the feto-placental barrier. Listeria can target post-translational modifications (PTMs) as an effective strategy to alter a wide variety of cellular functions. The best-characterized PTMs targeted by pathogens are phosphorylation and ubiquitination. However, Listeria was recently found to regulate SUMOylation of cellular proteins as well. Therefore, it is expected that other PTMs will be similarly altered during bacterial infection. ISG15 is an interferon-stimulated, ubiquitin-like protein (ubl), which has anti-viral activity. ISG15 is also upregulated in a number of human cancers. Currently, no one has thoroughly investigated whether bacterial infection influences ISGylation or whether ISGylation targets invading intracellular bacteria. We propose to capitalize on host-pathogen interactions to understand both the molecular mechanism of ISG15-mediated host defense and bacterial subversion of PTMs. We will focus on Listeria but will also broaden our findings to other important bacterial pathogens. Ultimately, since the molecular function of ISG15 remains enigmatic, characterizing how bacteria manipulate this polypeptide may shed light on its biological mechanism of action.

Levels of arousal govern our behavioural interactions with the world. In different states of arousal, different thresholds of external stimuli are required to elicit behavioural responses. Although this is an important aspect of behaviour, how arousal levels actually influence behavioural neural circuits is not known. I will use the small transparent zebrafish larva to dissect how arousal modulates behaviour. Arousal states in zebrafish are similar to those of mammals; sleeping zebrafish require stronger stimuli to respond, and arousal levels are modulated by the neuropeptide Hypocretin. The relative simplicity and experimental accessibility of the zebrafish nervous system make it an ideal mode to dissect the fundamental aspects of arousal.

I will use the well-characterized Mauthner cell dependent escape response as a model behaviour. I will test the hypothesis that the arousing neuropeptide Hypocretin modulates the excitability of the Mauthner cell through a population of hindbrain interneurons. This could provide the first insights into how arousal states tune behaviour at the level of defined neural circuits. In parallel, I will take a discovery-driven approach, and develop a behavioural-imaging platform to determine the electrophysiological correlates of arousal. This technique will allow for non-invasive recording of neural activity across the entire brain at single cell resolution, while simultaneously monitoring behavioural output and assessing the state of arousal. I will use this to determine how global neural activity changes as a function of arousal state, and to find neurons and networks governing arousal.

Biological systems are essentially analogical and respond to a dynamic environment therefore ongoing efforts are being made to identify the dynamic behavior of signaling pathways. Unfortunately, reality is much more complex and actually, signaling networks, rather than signaling pathways, operate to encode environmental information. A clear example is the innate immune system that is a complex network of signaling pathways able to integrate many different stimuli including cytokines or pathogen associated molecules. Up to date, multiple studies have focused in defining signaling dynamics in specific situations using population assays. In this research plan I propose to use the innate immune system as a model to analyze signaling network dynamics at the single cell level and in complex dynamic environments. In one hand I suggest to create a library of single cell reporters to monitor multiple steps in the signaling network. This library will contain upstream readouts, several key signaling molecules like NF-kB, JNK or p38 and downstream reporters. Clustering of those reporters in overlapping groups will create a picture of the signaling network dynamics in response to a given stimulus. On the other hand, I propose to use this library in a 4D live cell imaging platform coupled to a microfluidic device. Therefore, I will monitor the network dynamics in response to different types of stimulations like single steps, ramps, pulses or input combinations using different stimuli. Combining this data with mathematical modeling I pretend to generate a general picture of how signaling networks encode environmental information at the single cell level.

The actomyosin cortex, composed of a thin layer of crosslinked actin filaments and myosin molecular motors beneath the cell membrane, exemplifies a tight coupling between molecular mechanisms and physical properties. Involved in different morphogenetic processes, the dynamics and coherence of its assembly seems to play major determining roles. Here I propose to investigate how this cytoskeleton auto-assembles and turns over in relation to its different biochemical actors in the model organism C. elegans. To do so I will study the regrowth of the cell cortex after cortical laser ablation in the one stage embryo. By opening a hole in the cortex but not in the membrane, it gives rise to an area of ‘naked’ membrane where a novel cortex will assemble over the next 30 seconds. A precise spatio-temporal observation during this de-novo cortical growth could enable us to better understand the complexity of the molecular mechanism underlying actomyosin cortex assembly in general: molecular needs and cooperation as well as the tight coupling between molecular function and mechanical properties. Finally, we will investigate systematically how molecular mechanisms influence de-novo assembly, by perturbing cortical re-growth by RNAi and by overexpression of structural and regulatory components of the actomyosin cytoskeleton.

Biological diversity is created by the formation of new species, but debates persist about how speciation unfolds from beginning to end (i.e., the mechanisms of speciation) and what traits are involved during the process. One important way in which reproductive isolation may evolve is via reduced heterotypic mating due to divergence of mate preferences and traits (= sexual isolation), and recent studies have demonstrated highly variable levels of sexual isolation between and within species of Timema (small, wingless, and plant-feeding walking-stick insects). However, the actual traits and/or sensory modalities involved in establishing sexual isolation in Timema have yet to be characterized. I plan to conduct an interdisciplinary study that integrates behavioral, chemical, experimental, and genomic methods to investigate the role of chemical communication in reproductive isolation in this system. Specifically, I will use chemical ecology techniques to identify the extent of cuticular hydrocarbon compound (CHC) variation in Timema. I will then use manipulative mate choice experiments to investigate whether CHCs underlie sexual isolation, and finally, I will investigate the genetic basis of CHC variation using modern genomics tools. The work is novel in that it will be the first multi-species integration of CHC with behavioral and genomic methods to investigate the role of CHC variation in reproductive isolation, and it will further allow me to switch study systems and acquire new research skills that will help rounding out my scientific education. The results will shed insight into species formation, communication systems, and the genetic basis of CHCs.

Macrophage function ranges from pathogen clearance to resolution of inflammation. This diverse functional repertoire requires rerouting resources through specific metabolic pathways. Reactive oxygen species (ROS) are a particularly important component of the macrophage arsenal against pathogens such as Salmonella. During infection, ROS are generated by NADPH oxidase during a process resulting in increased oxygen consumption that imposes a substantial metabolic burden to macrophages. Recently, it was shown that ROS are also generated in mitochondria during infection with Salmonella. ROS produced through NADPH oxidase and mitochondria are both important for resistance of macrophages to bacteria, but how metabolism is reorganized to support these different effector mechanisms remains unclear. I will perform a comprehensive quantitative analysis of macrophage metabolism during Salmonella infection using pharmacologic, metabolomics, proteomics and in silico approaches to identify metabolic pathways involved in ROS generation during infection, and key metabolic enzymes regulating their production. I will test the contribution of candidate proteins to Salmonella growth control in vivo by gene deletion using a recently described primary macrophage system. Results will likely offer novel therapeutic strategies for modulating macrophage physiology in infectious disease, but potentially also for conditions characterized by deregulated macrophage metabolism or ROS generation such as atherosclerosis, obesity, cancer, and aging.

A striking feature of bilateral organisms is that left and right sides of the body are remarkably similar in size, despite the huge increase in body length during development. Arguably, the only reliable way of achieving this similarity is through an active mechanism that compares size on both sides. In vertebrates, longitudinal growth occurs primarily in the growth plate of vertebrae and long bones, and there is evidence that these bones are capable of considerable regulation to compensate for growth anomalies (a phenomenon known as catch-up growth). The prevailing model posits that as chondrocytes expand, proliferation decreases and they stop dividing after a fixed number of cell divisions, leading always to the same bone size. However, this model cannot explain situations in which the growth rate increases beyond normal levels. Thus, we favour a model in which there is an active left-right communication system capable of comparing bone size on both sides of the body and triggering compensation mechanisms. The potential mechanisms will be studied in detail by generating transgenic mouse lines in which the size of the long bones can be transiently modified in only one side of the body, what should lead to unilateral catch-up growth. A very likely candidate to mediate this left-right coordination is the nervous system. We will therefore explore (genetically and pharmacologically) which neural circuits (if any) mediate both normal growth coordination and experimental catch-up growth. If the hypothesis proves correct, we will also explore the size-sensing mechanism, i.e., identify the molecules in the bone that the nervous system responds to and serve as a readout for size.

Eukaryotic cells create internal organization using molecular motors that walk along cytoskeleton tracks. Predominantly, cytoskeleton motors are thought to function as transporters, moving specific cellular cargoes to specific cellular destinations. However, some motor species are hypothesized in addition to function as dynamic anchors for organelles and other subcellular structures. The tension thus arising in the local cellular structure is believed to contribute to structural integrity and to potentiate mechanical signal transduction. Although several indications for dynamic anchoring have been found in in vivo experiments, direct experimental verification has been missing due to the lack of a technique to measure molecular scale forces in live cells. I aim to develop a novel approach using an optogenetic molecular force sensor to directly measure the tension on molecular motors in living cells. As a model system, I will study the role of the motor Myosin VI in the organization of the Golgi complex. These measurements will provide the first direct verification of dynamic anchoring, and will establish a means to explore the roles of molecular motors in maintaining cellular structural integrity and in facilitating signal transduction.

The mechanism by which the earliest proteins evolved to the present diversity is still not fully understood. Gene duplication has been proposed as one the most important evolutionary mechanisms for generating new genes and thus understanding the existence and functional role of duplicated genes is crucial for the understanding of the emergence of novel functions.
We plan to study the role of gene duplication by looking through genes which have been conserved as a duplicate in one species but lost their redundancy in a related species and compare their function and regulation. As a model system, we will use multiple species of budding yeast, including S. cerevisiae, S. castelli and C. glabrata, whose shared common ancestor underwent a whole-genome duplication. However, since duplication a distinct subset of duplicate genes has been lost in each lineage. We would compare these species to a related species, S. kluyveri, that has not undergone whole-genome duplication.
We will focus on a subset of genes that are amenable to both measurement of level and function such as genes that encode glucose transporters and genes that encode ribosomal proteins. We plan to swap the promoters of paralogous genes and the genes themselves between species and measure the effect by a combination of single cell transcriptional read-outs, fitness, proteomics, and metabolomics. To support the experimental work, we will develop a theoretical evolutionary model to study the effect of plasticity of the ancestor gene on the ability of duplicated genes to acquire new functions.

Modern genomics have revealed hundreds of new candidate genes associated with squamous cell carcinomas (SQCCs), the most prevalent life-threatening cancers world-wide. Distinguishing candidates that influence tumor initiation and metastases from mere ‘bystander’ mutations is a fundamental prerequisite for developing new and improved cancer therapies. I propose to tackle this important problem using as my model skin SQCCs.

As in humans, mouse SQCCs arise in a multi-step process whereby loss of TGFb receptor signaling or elevated Ras-MAPK signaling dictates a strong predisposition to SQCC, but additional gene mutations are required for SQCC formation. I propose to model these two susceptibility pathways in mouse skin, to establish a pool of lentiviral shRNAs targeting candiate SQCC genes and to assess their functionality in SQCC progression. The pool will be targeted to the single-layered surface ectoderm of tumor-prone mouse embryos using a novel, non-invasive in utero lentiviral delivery system. Once integrated, lentivirus shRNA is faithfully expressed and maintained in mature skin. Cells that express shRNAs which compromise normal epidermal growth/development will die. Cells expressing shRNAs targeting essential tumor suppressors will have a growth advantage and form tumors and/or metastasis faster than controls. Conversely, if a targeted gene is selectively required for tumor growth, their shRNA-carrying cells will be depleted selectively in tumors. Relative shRNA-abundance in WT and tumors will be quantified by Solexa-sequencing, revealing novel tumor suppressors/promoters in vivo. Their biological importance during tumorigenesis will be investigated.

Adult stem cells are characterized by their ability to generate multiple differentiated cell types while maintaining their capacity for long-term self-renewal. In order to keep the number of stem cells balanced, a close equilibrium between both processes is crucial. Recent studies have demonstrated that in the mammalian intestinal epithelium this equilibrium is not regulated by asymmetric cell division, in which one daughter differentiates and the other daughter maintains a stem cell, but at the population level: niche factors assure homeostasis of symmetrically dividing stem cells. Subsequent studies showed that these factors are provided by the juxtaposed Paneth cells. Thus, the number of stem cells is regulated by Paneth cells, and communication between both cell types is necessary to control the niche size.
However, due to the lack of live imaging in the intestinal epithelium, only little is known about the dynamic interactions between Paneth cells and stem cells. For instance, the role of Paneth cells in the formation of new niches and niche expansion has not been characterized. Moreover, the signaling factors between Paneth cells and stem cells have not been identified. Thus, how Paneth cells regulate stem cell numbers remains elusive. I aim to answer these questions by live imaging of the intestinal stem cell niche in the primary organoid culture system. Moreover, I will perform an shRNA screen in organoids to identify genes involved in niche size regulation. These genes will be subject to in-depth and in vivo analysis.

While many of the layers of epigenetic regulation by modification of DNA and proteins are intensively explored, far less is known about the scope and role of RNA modifications, mostly due to technical challenges. N6-methyladenine (m6A) is the most prevalent modification occurring on mRNA. I have recently studied the first transcriptome-wide catalogue of m6A sites in human and mouse. I identified thousands of m6A sites, clustered at specific genomic regions which are highly conserved between human and mouse, strongly suggestive of an important regulatory role played by this modification. Here, I propose to use an integrated experimental and computational strategy to uncover the biological function of m6A in mammalian cells. I will establish improved protocols for m6A measurement, and apply them both to RNA from different cellular compartments and to RNA from dynamic physiological systems. I will analyze the data to determine when RNA methylation is introduced during the RNA lifecycle and to decipher how dynamic these methylations are. I will perturb m6A levels using mutagenesis in minigene constructs and by knocking down the methylating enzyme, to examine the effects of the methylation on mRNA lifecycle and/or translation. I will also perform assays to identify the proteins involved in ‘encoding’ and ‘decoding’ this modification, and use an shRNA library to test the effects of their knockdown. Complementary bioinformatic analyses will aim to associate knockdown of specific proteins with coherent changes in methylation patterns. Collectively, I aim to shed light on the function of this methylation, which has remained obscure for >40 years.

The first line of plant and mammalian innate immunity is constituted by pattern recognition receptors (PRRs) recognizing conserved microbial signatures. After ligand perception PRRs employ non-RD kinases to initiate intracellular signaling cascades. Non-RD kinases lack the highly conserved arginine (R) that precedes the catalytic aspartate (D) typical of most kinases. Despite their fundamental role in innate immune signaling the regulatory mechanisms governing non-RD kinases are still poorly understood. The proposed studies will address an important question: How is the structure of non-RD kinases advantageous for their biological function?

I will use the rice PRR XA21, a receptor kinase sensing a conserved microbial signature from a bacterium, as model non-RD kinase. Employing phospho-proteomic techniques I aim to identify XA21 phosphorylation patterns before and after ligand binding. I will complement this approach with structural biological studies of the XA21-intracellular domain. Such studies will facilitate the identification of interaction surfaces important for kinase regulation. Because an ATPase (XB24), kinase (XAK1) and phosphatase (XB15) are critically important for XA21-mediated responses, analysis of changes in the XA21 phosphorylation pattern in the presence of these proteins and its structural implication will allow me to further refine my model of XA21 activation. I will validate this model through structural and functional analysis of XA21. This research will lead to detailed knowledge of the molecular mechanisms regulating XA21, a model non-RD kinase, and be of broad significance to understanding how plants and animals initiate innate immune signaling.

Most of our social encounters rest on processing the faces of others to disclose their emotions, intentions and hidden attitudes. Detecting a face is thus a central visual task, and the entry step for more elaborate processing. Under natural conditions, however, sensory information is often too ambiguous or scarce to support face recognition, eg., when faces are camouflaged or occluded. How does the brain decide that a face is present when visual input is sparse? Recent computational models suggest that face detection cannot exclusively rely on bottom-up input, but needs to integrate prior knowledge with the incoming input to disambiguate the visual scene and thus find a face. Though appealing, there is to date no neural mechanism known implementing this model. Thus, two questions of fundamental importance for understanding brain function arise: where in the brain is prior knowledge stored? And how does it enable face perception? I will tackle these questions in the macaque brain, using a combination of functional magnetic resonance imaging, targeted pharmacology and electrophysiological recordings to map candidate regions involved in prior knowledge for face detection, to prove their causal relevance, and to study how they exert their influence on face processing. These investigations will make an important contribution to a mechanistic understanding of object recognition in general. Furthermore, this multi-node network investigation will provide novel insight into how information is integrated in modular, distributed brain systems.