Theoden I. Netoff - US grants

Cells in the nervous system are often highly interconnected; however, their firing times are synchronized during certain events such as gamma oscillations and theta oscillations. During these oscillations, crosscorrelations can show synchronization of spike times. However, when the cells are not synchronized, it is our hypothesis that there may be more complex interactions, such as generalized synchrony, that cannot be detected using linear analysis (i.e. cross correlation). These interactions may give a much richer understanding of the interactions between cells. Currently, no technique exists to detect these complex interactions in neurons. We intend to record from two pyramidal cells in the hippocampus simultaneously with and without glutamate antagonists. The glutamate antagonists increases synchrony between cells. Analysis to detect nonlinear functional relations between the firing patterns of the cells will be applied to the time series. We will use a technique called mutual prediction. It has been used to show interactions between single cells and the population activity in the stimulated cat spinal cord that linear techniques could not detect. This work will also develop a new technique for detecting nonlinear interactions by comparing patterns of interspike intervals using unstable periodic orbits. If a functional relationship is found between the spontaneous firing pattern of two cells when they are spontaneously firing, these techniques will be used, in a collaboration with another laboratory, to analyze interactions between two cells in the olfactory bulb and how they are effected by different odorants.

Cells in the nervous system are often highly interconnected; however, their firing times are synchronized during certain events such as gamma oscillations and theta oscillations. During these oscillations, crosscorrelations can show synchronization of spike times. However, when the cells are not synchronized, it is our hypothesis that there may be more complex interactions, such as generalized synchrony, that cannot be detected using linear analysis (i.e. cross correlation). These interactions may give a much richer understanding of the interactions between cells. Currently, no technique exists to detect these complex interactions in neurons. We intend to record from two pyramidal cells in the hippocampus simultaneously with and without glutamate antagonists. The glutamate antagonists increases synchrony between cells. Analysis to detect nonlinear functional relations between the firing patterns of the cells will be applied to the time series. We will use a technique called mutual prediction. It has been used to show interactions between single cells and the population activity in the stimulated cat spinal cord that linear techniques could not detect. This work will also develop a new technique for detecting nonlinear interactions by comparing patterns of interspike intervals using unstable periodic orbits. If a functional relationship is found between the spontaneous firing pattern of two cells when they are spontaneously firing, these techniques will be used, in a collaboration with another laboratory, to analyze interactions between two cells in the olfactory bulb and how they are effected by different odorants.

DESCRIPTION (provided by applicant): The goals of this study are to measure how the dynamics of individual neurons and synapses contribute to synchronous oscillatory activity in populations of neurons, and to understand how the neuromodulator acetylcholine changes intracellular and network properties. Experiments will measure the dynamics of neurons in brain slices of the entorhinal cortex (EC) using whole-cell patch-clamp techniques, along with advanced real-time experimental control. The experiments will be supplemented by computational work. This study has four aims: (1) There are at least three distinct neuronal populations in the EC that have distinct electrophysiological properties. It will be tested if these neurons have different mechanisms of synchronization (e.g., excitation- or inhibition-based synchronization) using 'spike time response' (STR) methods from applied mathematics. In STR techniques, neurons are characterized in terms of how spikes in periodically firing neurons are advanced or delayed by artificial synaptic inputs. (2) The neuromodulator acetylcholine (ACh) is known to alter firing properties of neurons in the EC, and change population rhythms in brain slices. Therefore, it is hypothesized that ACh changes cellular intrinsic properties in a manner that supports enhanced synchronization. The effect of ACh agonists on neurons will be studied using STR measurements. (3) STR methods can be used to predict how small neuronal networks will synchronize. This hypothesis will be directly measured by using real-time control system to construct "hybrid" networks of coupled biological neurons and computer-modeled counterparts. (4) Network activity becomes complicated with increasing number and types of neurons and types of neurons. Models from STR based model networks can predict the behavior of large networks. Modeling can be used to understand networks with multiple cellular components and more complex patterns of synaptic coupling.

PI: Netoff, Theoden I.
Proposal Number: 0954797

The etiologies of pathological behaviors that emerge in networks are especially difficult to diagnose. The causes are usually subtle changes in the dynamics of the nodes that lead to changes in population behavior. These multi-scale problems are very general. Epilepsy is an example of disease where molecular level changes in neurons caused by genetic mutations lead to pathological neuronal activity generating seizures. While there are many hypotheses, very little is known about how and why these mutations cause seizures, which prevents us from developing better treatments. Understanding how synchrony in networks are affected by known epileptogenic mutations and antiepileptic drugs with known molecular effects will provide a model system in which multiple scales may be bridged. A synergistic approach using numerical simulations electrophysiology experiments and computational simulations will be used. Computational models of neurons will be used to predict how epileptogenic mutations and antiepileptic drugs change the phase response curve (PRC) of a neuron. The PRC is a measure of a neuron?s sensitivity to synaptic inputs. From the PRC it is possible to infer how changes caused by epileptogenic mutations and antiepileptic drugs would alter synchrony in a network of neurons. Predictions from the modeling will be tested using dynamic clamp experiments, where a computer running a real-time interface is interfaced to a neuron through a patch clamp amplifier and electrode. Dynamic clamp experiments will be used to measure the effects of epileptogenic mutations (introduced thorough electrical knock-in) and bath applied antiepileptic drugs on the phase response curve of the neuron. Hybrid networks will then be created using the dynamic clamp to simulate synaptic connections between two patch clamped neurons in which effects of epileptogenic mutations and antiepileptic drugs on synchrony will be measured directly. Physiological experiments will be used to provide parameters to run large scale simulations where synchrony will be measured. Preliminary data is presented from simulations and electrophsiological experiments that epileptogenic mutations in voltage gated sodium channels decrease synchrony and antiepileptic drugs increase synchrony. These findings are in contrast to the popular view of epilepsy that epilepsy is caused by hypersynchrony. By developing our understanding of how these mutations and drugs actually work, we may develop new and better approaches to treating this disease.

The goal of this proposed research is to test the hypotheses that changes in the dynamics of neurons caused by epileptogenic mutations increase network synchrony, and that the modulation of neurons by drugs that prevent seizures decrease network synchrony. By proving, or disproving these hypotheses, we will understand if developing new drugs or deep brain stimulation to prevent seizures should be optimized to decrease network synchrony. To test this hypothesis we propose the following specific aims: 1) use single cell modeling to identify effects of epileptogenic mutations and antiepileptic drugs on cell dynamics, 2) network modeling to assess the effect of epileptogenic mutations and antiepileptic drugs on network synchrony, and 3) characterize changes in cell dynamics caused by mutation of SCN1A channel using hybrid experiments with real neurons and virtual ion channels.

Intellectual Merit: The research proposed here will help elucidate how changes in neuronal dynamics and topology of network connectivity result in pathological neuronal activity such as seizures. How neuron dynamics are affected by epileptogenic mutations and antiepileptic drugs will be discovered to help develop better models of seizures. Effects of known epileptogenic ion channel mutations and antiepileptic drugs on network synchrony will be used to probe the role of neuronal population synchrony in epilepsy. With this knowledge we will develop more rational approaches to treating epilepsy.

Broader impact: Electrophysiolgy data acquired will be cataloged in a database available to any scientist interested in analyzing the data. To complete the electrophysiolgical experiment, we will generate many modules for the dynamicclamp which will be made available to the community using the RTXI dynamic clamp. Code developed to run network simulations using CUDA enabled machines for supercomputer performance on a desktop will be made available to the public. Outreach plan includes collaborations with the Bakken museum, the Epilepsy Foundation, the University of Minnesota?s ?Brain U?, its summer high school program ?Exploring Careers in Engineering and Physical Science?, and it?s North Star Alliance Program.

PI: Netoff, Theoden I. and Moehlis, Jeffrey M.
Proposal Number: 1264432 & 1264535

Intellectual Merit: Populations of neurons must dynamically synchronize and desynchronize for transmission of information within the brain. The disruption of this dynamic synchronization is thought to underlie the symptomatology of several neurological disorders. Deep Brain Stimulation (DBS) therapy is being used to treat many of these neurological disorders, such as Parkinsons disease (PD). It is generally believed that DBS leads are placed in regions of brain that are pathologically synchronous, and periodic DBS pulses then "over pace" these areas, blocking the pathological activity. The PIs have recently developed an alternative hypothesis for the mechanism of DBS which focuses on DBS's modulation of the firing times of neurons. Stimulation at certain frequencies can induce a chaotic response that desynchronizes a population; we term this chaotic desynchronization. The response of a neuron to a DBS pulse is characterized by its phase response curve (PRC), a measure of how the stimulus advances the phase depending on the phase the stimulus is applied at. The PRC can then be used to determine if two neurons in the population starting at nearly the same phase will entrain to the stimulus pulses, or will diverge and effectively become desynchronized. In this grant the PIS propose to use PRCs to determine the optimal stimuli to desynchronize population oscillations. Preliminary experiments show that small periodic stimulus pulses at certain frequencies can desynchronize populations; the frequency and amplitude that desynchronize can be predicted from the PRC of the neurons to the stimulus. Moreover, continuous stimulus waveforms can be designed that desynchronize populations with much less energy than the pulsatile stimuli. The aims of this grant are to further the theoretical work in designing these waveforms from measured PRCs, and then to test chaotic de-synchronization in physical and biological systems. Specific Aim 1 will use measured phase response curves and control theory to determine the optimal stimulus waveforms to maximize desynchronization of neuronal ensembles. Specific Aim 2 will be to apply this theory to desynchronize oscillations in a chemical oscillator model, the photosensitive Belousov-Zhabotinsky (pBZ) reaction, through pulsatile and continuous waveform photo stimulation. Specific Aim 3 will test the theory in neurons in vitro basal ganglia preparation. Neurons will be recorded and stimulated using a dynamic clamp experimental protocol. The PRCs from single neurons will be measured in response to DBS pulses, and we will test for chaotic behavior in their stimulus response patterns.

Broader Impacts: The motivation of this research is to 1) understand how behaviors relate to oscillatory synchronization in and between the basal ganglia and motor cortex, and 2) improve DBS treatment of PD, for which the selection of stimulus electrodes, frequency, and amplitudes are currently tuned manually by a clinician. The goal of this research is to determine the optimal stimulus properties based on simple physiological measures of the neurophysiological response to DBS. This approach will enable faster and more robust programming of neurostimulators and will decrease the amount of required injected current, which will reduce side effects and battery power consumption. This approach has high potential for closed loop control algorithms where DBS parameters are automatically tuned to maintain maximal efficacy. This approach may also be applied to seizure suppression and other neurological diseases. These studies leverage a recently funded IGERT training plan at UMN for neuromodulation. To maximize our clinical impact, we have discussed with Dwight Nelson (Neuromodulation department at Medtronic) what basic research will enable the next steps in developing new DBS stimulus parameters and the yet unmet clinical needs (letter of support included). The results from this research will be disseminated to the public through various education programs including ones focused on underrepresented undergraduate students, high school educators, high school students and junior-high school students. Finally, this award will train graduate students and undergraduates in interdisciplinary research activities, and enhance the education of other graduate students through results that will be incorporated into courses taught by the PI and co-PI.

The brain is an amazing organ which is responsible for a number of important functions including cognition, attention, emotion, perception, memory, and motor control. Many brain functions and disorders are believed to have a dynamical origin; for example, it has been hypothesized that some symptoms of Parkinson's disease are due to pathologically synchronized neural activity in the motor control region of the brain. Recent research suggests that an FDA-approved treatment for Parkinsonian tremors, called deep brain stimulation, is effective because it partially desynchronizes the neural activity via clustering, in which neurons in a subpopulation are synchronized with each other, but desynchronized with neurons in other subpopulations. This research will use engineering techniques, mathematical principles, computer simulations, and in vitro experiments to develop more energy-efficient electrical current stimuli which promote such clustering. Moreover, stimuli will be developed which enhance beneficial neural plasticity in which neurons change their connection strengths based on their activity patterns, work that may be important for treatment of diseases and for situations in which plasticity is desirable such as learning, memory, and recovery from strokes and spinal cord injury.

This research will use engineering techniques, mathematical principles, computer simulations, and in vitro experiments to develop efficient electrical stimuli for controlling neural populations in beneficial ways. This will include designing power-minimized stimuli which cause a neural population to split into balanced clusters, in which each cluster contains a nearly identical proportion of the overall population and neighboring clusters are roughly equally spaced in phase, a state of partial desynchronization which recent work suggests is responsible for the success of the standard protocol for deep brain stimulation treatment of Parkinson's disease. Moreover, Hebbian models for synaptic plasticity will be used in combination with optimal control theory to design stimuli which optimally promote plasticity to give beneficial long-term changes in synaptic connections, work which is expected to have important implications for Parkinson's disease and other disorders such as epilepsy and depression, and for situations in which plasticity is desirable such as learning, memory, and recovery from strokes and spinal cord injury. The plasticity studies will also include in vitro brain slice experiments in which neurons will be synchronized to an oscillating electric field and stimulation applied through an electrode to generate balanced clusters, whose effect on synaptic strengths will be measured.

This Research Experiences for Undergraduates (REU) Site from the University of Minnesota will support a diverse cohort of undergraduate students to pursue research in the interdisciplinary field of neural systems engineering. This program will highlight recent developments in neurotechnology for basic science research and healthcare. It is estimated that 100 million U.S. citizens will have a significant brain-related disorder in their lifetime. There is a need to train scientists and engineers who will research and understand the process for translating neural systems engineering research into human applications. This site will focus on recruiting a diverse group of students with an emphasis on recruiting women, underrepresented minorities, and undergraduates from colleges where opportunities to pursue STEM-based research are opportunities.

Over a three year period, this REU program will engage undergraduate students in a 10-week intensive, summer research experience aimed at further developing neurotechnology for brain imaging, decoding and modulation. Projects will focus on developing technology, running simulations or data analysis. Many of the projects will integrate multiple modalities of neurotechnology, such as multi-modal imaging, brain-computer interfaces and closed-loop neuromodulation. Students will have opportunities to investigate neuronal dynamics, plasticity, learning and attention as feedback metrics to optimize neurotechnologies. Research projects will provide opportunities for REU students to learn about large scale data analysis, computational modeling of the brain, and participate in experiments. The program will have group building projects to develop a cohesion between the students within the program and with their graduate student mentors to create lasting friendships and collaborations. The program will emphasize near-peer mentoring where the REU students will be directly mentored students in their second and third year of graduate school, and in turn will be offered the opportunity to meet and educate high school students about neural engineering. To broaden their perspective, students will participate in a bi-weekly neuroengineering seminar series hosted in conjunction with a NSF IGERT program and the Center for Neuroengineering. Multiple mechanisms will help guide students in their professional development and prepare them for graduate school. To develop communication skills, students will participate in outreach programs and present their research in a small forum at the end of the program.

PROJECT SUMMARY AND ABSTRACT The basal ganglia have a rich, functional topography composed of motor subcircuits and oscillatory networks that are thought to be critically important to the pathophysiology of Parkinson's disease (PD) and the successful application of deep brain stimulation therapy (DBS) in managing each cardinal motor sign of PD. There is a strong clinical need to better understand these processes and in turn harness them to deliver therapy that is tailored to an individual patient and a patient's own symptomatology. In this project, we seek to develop a novel spatiotemporally optimized DBS therapy and evaluate its efficacy in a non-human primate model of PD. The optimization approach leverages the unique capabilities of (1) high-field MR imaging (7T and 10.5T), (2) subject-specific computational models of DBS, (3) a high-density DBS lead with electrodes arranged along and around the lead shank, and (4) a real-time signal processing interface that can readily adapt stimulation parameters on the DBS array based on analysis of ongoing oscillatory activity at and downstream of the site of stimulation. High-density DBS arrays spanning the subthalamic nucleus (STN) and thalamic fasciculus (Array A) and the external and internal globus pallidus (GP) (Array B) will be implanted in each subject. Aim 1 will characterize the magnitude and time course of therapeutic effects on each parkinsonian motor sign when targeting electrical stimulation within and around the STN and GP. Aim 2 will investigate how targeted stimulation differentially affects oscillatory activity at and downstream of the site of stimulation and relates to improvement in each parkinsonian motor sign. Aim 3 will develop and apply a novel set of optimization algorithms, including chaotic desynchronization and real-time closed-loop phasic stimulation, to test the hypothesis that optimizing suppression of exaggerated phase amplitude coupling in the STN and GP will further increase the overall magnitude of DBS therapy. Together, this project will enhance our understanding of the pathophysiology of PD and provide critical data towards translating a patient-optimized DBS therapy that integrates high-density DBS leads with novel closed-loop stimulation.

Project Summary The overall objective of this effort is to convene ?ISPW8: Designing the next generation of closed loop seizure control?, that will take place August 20-23, 2017 at the University of Minnesota in Minneapolis. This conference is the next in a series of seizure prediction workshops that started in 2002, and has become the premier international venue for quantitative epilepsy research. The goal of this upcoming meeting is to focus on the development of all stages of a next-generation seizure control device. The first day will have two didactic sessions, one focusing on the clinical aspects of epilepsy, the other on Big Data analytic techniques. The body of the conference will be organized around themes motivated by the three stages of closed loop intervention: 1) input: sensing and biomarkers, 2) processing: system analysis, 3) output: intervention. The theme of the second day of will be on multimodal sensors and biomarkers of epilepsy, focused especially on developing new technology. The theme of the third day will be on utilizing advanced machine learning, statistics, and computational models, to understand and characterize the large datasets from high resolution technology. The final day will be on novel interventions and control theory: new strategies for implantable anti-seizure interventions and methods for optimizing them, as well as implementation into closed loop control devices. ?Provocative sessions? at the end of each day will focus on unpublished research and hypotheses followed by discussion and debate. Through this conference we aim to build bridges between the many disciplines that comprise this unique field: theoretical, computational, experimental, clinical, and industry, while also preparing the next wave of young researchers. The overall goal is to develop improved quantitative methods to predict, quantify, characterize, and control seizures. NIH funding is sought for travel support to encourage US participation.

PROJECT SUMMARY AND ABSTRACT Existing neurotechnologies continue to make significant clinical impact, but challenges remain for scientists and engineers in moving new devices from the bench to the bedside. This grant will support the development of a comprehensive educational training program on the best practices for successfully navigating the translational and commercialization pathway for neural medical devices. The educational program will include (1) a series of publicly available video lectures that will be assembled and curated into a certificate program by a diverse group of program faculty with significant experience in moving neurotechnologies from the bench to the bedside. These short-course video lectures will cover the following topics: preclinical model systems, safety and efficacy studies, good laboratory practices, device testing, quality system processes, regulatory agency interactions, steps in developing an investigational device exemption (IDE) application, reimbursement agency interactions, clinical trial design with an emphasis on quantitative outcome measures of target engagement, bioethical considerations that are specific to neural medical devices, techniques for securing strong intellectual property claims, funding opportunities available for technology development and clinical trials, and advice on moving neurotechnology into successful commercial ventures. In addition, the educational program will provide (2) an annual three-day workshop in which participants will work with program faculty to think through and develop submissions to an institutional review board (IRB), the FDA, and/or funding groups for their own devices or devices inspired by relevant case studies. The workshop will be open to academic researchers, clinician scientists, and small-business entrepreneurs who are interested in gaining expertise that will help them translate their own neural medical devices. Participants of the workshop will be encouraged to continue interacting with program faculty through regular online follow-ups. The hands-on workshop will be offered for two years during the grant funding period with the goal of extending the workshop in future years through collaborations with our ongoing relationships with neural interfaces conferences.

ABSTRACT The remarkable efficiency of human perception derives from the fact that we do not process each stimulus as a novel event. Instead, past experiences and scene context inform internal, working models of the world that allow us to generate predictions for our physical environment. A leading theory suggests that perceptual predictions are accomplished via flexible normalization: local inhibitory neuronal populations are regulated by long-range connections so that responses are suppressed when they do not provide helpful information about object boundaries. However, the precise neural mechanisms by which the healthy human brain accomplishes this flexible normalization are not known. In order to understand exactly how neural population responses are suppressed or enhanced in response to different scene contexts, we will perform 2-photon imaging in ferret primary visual cortex (V1) to quantify the responses of excitatory and inhibitory neural populations in superficial layers of cortex during several different visual stimulus paradigms. The ferret model is chosen because the imaging techniques necessary to quantify inhibitory neuronal responses are not yet well established in primate models, and while our current knowledge about neural morphology and connections has been derived from mouse models, mouse visual cortex lacks the ?columnar organization? (spatial grouping of neurons with similar response properties) that is a hallmark of primate visual cortex and is present in ferrets. Thus, the ferret model is well-positioned to bridge the gap between mouse models and primate models. First, in order to understand neuronal behaviors in the absence of contextual modulation, we will characterize interactions within a single hypercolumn to small, simple stimuli (sinusoidally modulated luminance gratings) at a range of orientations and contrasts. We hypothesize that parvalbumin-containing (PV+) inhibitory interneurons will demonstrate the sharpest orientation tuning, followed by somatostatin-containing (SOM+) and serotonin-positive (5HTR+) populations. Next. using a Cross Orientation Suppression paradigm, we will test the hypothesis that that SOM+ responses track the overall contrast energy in the stimulus, while PV+ populations reflect suppression of individual grating component representations. Additional experiments with naturalistic textures will test whether these behaviors generalize to stimuli with a broad range of contrasts, orientations, and spatial frequencies. Finally, we will use classical Orientation-Dependent Surround Suppression and Collinear Facilitation paradigms to study how the local inhibitory pool responds to scene context. We hypothesize that the responses of local 5HTR+ neurons will reflect the surrounding stimuli rather than the center stimuli. Together, these experiments will constrain an open-source computational model articulated at the level of the single neuron that will constrain hypotheses about how human perceptual behaviors are linked to specific neuronal populations; this model will be valuable for understanding how perceptual aberrations associated with psychosis might be mapped to the function of specific neuronal subpopulations.

Summary The last decade ushered in amazing advances in neuroimaging due to transformative developments in magnetic resonance and optical imaging techniques as well as in computational and modeling tools. A common thread in these advances is their multi-disciplinary nature, requiring collaborations among medical researchers, engineers, physicists, mathematicians, and data scientists, among many others. In order to continue the pace of technical advances in neuroimaging and to exploit their unique capabilities for brain research and medical applications, it is critical to train the next generation of neuroimaging specialists in a setting 1) with an abundance of state-of-the-art tools, 2) with a programmatic interest in developing novel neuroimaging technologies and biomedical applications, and 3) where trainees can carry out groundbreaking research under multi-disciplinary mentorship. The University of Minnesota (UMN) has an excellent tradition of training neuroimaging postdocs in its Medical School and its College of Science & Engineering. It is also the home of major, world-renowned efforts in neuroimaging technology development and novel biomedical applications of neuroimaging. The proposed Minnesota Neuroimaging Postdoctoral Training Grant aims to give 14 postdoctoral fellows ? at least 4 of whom will be from communities under-represented in STEM fields ? multi-disciplinary skills in neuroimaging technology development and advanced biomedical applications, guidance in career development, and social and networking support through intense two-year neuroimaging fellowships at UMN. Each fellow's primary research will be conducted on a multi-disciplinary project that combines their background with another field with the express goal of developing new neuroimaging technologies. Each fellow will be co-mentored by two faculty selected from the 40+ participating faculty in this grant: one that will directly supervise the research project, and one that represents a core area related to the research project. Fellows will take at least two courses to broaden their skillset and prepare for either an academic or industry research career. They will participate in an Annual Retreat and twice-monthly seminars that will cover research and career development topics such as responsible conduct of research, scientific rigor and reproducibility, grant writing, and other key subjects. They will also participate in UMN's numerous neuroscience conferences, symposia, and workshops, along with well-established UMN outreach programs to high schoolers and undergraduates from communities under-represented in STEM fields. The program will be managed by an Executive Board that represents the diversity of the participating faculty across the Medical School and the College of Science & Engineering. Management plans include a rigorous, ongoing evaluation process that incorporates an external Advisory Board and the University of Minnesota's internal research and assessment services.