Anna Devor - US grants

DESCRIPTION (provided by applicant): Functional Magnetic Resonance Imaging (fMRI) has the potential of fulfilling a long held dream of the neuroscience community and of the public at large to view the human brain in action. However, the detailed relation between the neurovascular parameters mapped in fMRI, and the underlying neural activity, are poorly understood at present. In our previous work we demonstrated a nonlinear relationship between hemoglobin concentration and oxygenation, and neuronal spiking and synaptic electrical activity in rat somatosensory (Barrel) cortex. In this proposal we will extend our study to the cerebellar cortex in order to investigate whether the same general rules can be applied to neurovascular coupling in both structures. While the main focus of this proposal is on the cerebellum, we will in parallel continue and expand our previous work in Barrel cortex to facilitate the comparison between the two (2). We will perform simultaneous optical measurements of hemodynamic signals (blood flow, volume, oxygenation) and electrophysiological measurements of neuronal activity, and will use an empirically developed relationship to construct a "transfer function" of neurovascular coupling. Despite fundamental differences in neuronal and vascular organization between cerebral and cerebellar cortices we hypothesize a conservation of the main principles of coupling when taking into account the sum activity of Purkinje and granule cells, and differentiating the contribution of simple and complex spikes. The combination of optical imaging techniques and "gold standard" electrophysiology not only provides a tool for correlation of hemodynamic and neuronal signals, but also enables better spatio-temporal estimation of brain activity. Thus, in addition to addressing the question of neurovascular coupling, we will study key questions of cerebellar physiology centered around the influence of parallel fibers on spatial extend of cerebellar cortical activation. Investigations proposed here will provide essential information to bridge the gap between the growing body of fMRI data and generations of detailed electrophysiological studies in cerebrum and cerebellum. Ultimately, accurate interpretation of imaging data in terms of neuronal activity will play an important role in the early detection, diagnosis and treatment of human disease.

? DESCRIPTION (provided by applicant): The continuing overall goal of our research program is to understand the relationship between functional MRI (fMRI) signals and the underlying neural activity. During the previous period of support, we focused on elucidating specific neurovascular, neuroglial, and neurometabolic events that affect positive or negative hemodynamic response. We improved the methodology for quantitative imaging of calcium, validated a new technology for microscopic imaging of partial pressure of O2 (pO2), and took aboard optogenetics for manipulation of specific cell types that became available within the duration of the support period. With these methods in hand, we probed dilation/constriction of cerebral microvasculature following neuroglial calcium- dependent release of vascular mediators, intravascular and tissue pO2, and layer-resolved BOLD fMRI responses as a function of the underlying vascular dynamics. The key implication of our study is that cerebral blood flow (CBF) and cerebral metabolic rate of O2 (CMRO2) are being driven in parallel by neural activity, and potentially by different aspects of neural activity. For example, activation o the excitatory cells is associated with large metabolic costs of repolarization, glutamate recycling and calcium buffering but produces smaller dilation compared to activation of the inhibitory neurons. In parallel, similar conclusions have been reached by our co-investigator Dr. Buxton based on human studies that utilized the calibrated BOLD fMRI technique. Taken together with the insights from our microscopic imaging and manipulation, these results lead us to put forward the hypothesis that variability in the ratio of the fractional changes in CBF and CMRO2 reflects different proportions of inhibitory and excitatory evoked activity. If proven true, this would open a new direction in which quantitative fMRI may be able to provide information on the underlying neural activity. Thus, the goal of this proposal is to investigate the behavior o the CBF/CMRO2 ratio using microscopic imaging of the relevant underlying physiological parameters while directly controlling neural activity. Aim 1 is focused on exploring the CBF/CMRO2 ratio during selective optogenetic activation of excitation and inhibition in the mouse sensory cortex in vivo. In Aim 2, we will use optogenetics for modulating the baseline CBF and CMRO2 mimicking the neural response to a standard stimulus shaped by different brain states. The proposed project is an integral part our collaborative research program where animal and human efforts progress in a continual dialog. While the CBF/CMRO2 ratio hypothesis has been derived in part from human studies, the human data provide no direct information on the balance of excitatory and inhibitory activity. Our set of imaging and manipulation tools in the mouse, on the other hand, is perfectly suited to evaluate this hypothesis under well-controlled conditions. The proposed animal experiments will lay a mechanistic foundation for clinical applications of the calibrated BOLD methodology potentially enabling a paradigm shift in human fMRI studies: from simply asking where activation occurs to asking how much activation occurs.

DESCRIPTION (provided by applicant): Recent developments in two-photon laser scanning microscopy (TPLSM) in combination with improvement of fluorescent calcium indicators have opened an unprecedented possibility for in vivo imaging of neuronal activity. Today, TPLSM of calcium signals is a widely used tool for studying of neuronal circuits and neuron-glial-vascular interactions in health and disease. Current techniques of introduction of calcium indicators into the brain tissue result in non-selective labeling of neurons and glia. While glial cells can be distinguished from neurons by adding selective markers, different types of neurons (excitatory and multiple types of inhibitory cells) cannot be discriminated. Yet, neuronal behavior is cell-type specific and understanding of the physiology and pathophysiology of identified neuronal cell types is of prime importance for development of new strategies of treatment and prevention of human disease. Thus, in vivo identification and functional imaging of specific cell types is a central unresolved problem in biomedical imaging. In principle, this can be achieved through use of genetic methods. However, cell-type specific transgenic expression has been only partially successful. The goal of this proposal is to develop an alternative strategy for identification of neuronal cell types taking an advantage of differential behavior of calcium signals in specific neuronal cell types. TPLSM will be used for detailed assessment of unique calcium "signatures" of cell-type specific neurophysiological properties observable in vivo in wild type animals. The findings will be validated by measurements from identified neurons in transgenic mice and in vitro brain slice preparations. The main deliverable of this project, a tool for "on-line" identification of excitatory and multiple types of inhibitory neurons, will enable simultaneous calcium imaging from multiple identified cell types with no need in additional promoter- specific genetic labeling. This approach will significantly advance in vivo two-photon calcium imaging used in basic neuroscience research and for investigation of animal models of human disease. PUBLIC HEALTH RELEVANCE Physiological identification of the neuronal cell types that comprise the mammalian brain is a central unresolved problem in neuroscience. The proposed project will advance application of two-photon laser scanning microscopy to imaging of neuronal assemblies by providing a tool for identification of specific cell types. Simultaneous in vivo imaging of identified neuronal phenotypes is crucial to define the normal state of physiology in neuronal networks and the degree of deviation in brain disease.

DESCRIPTION (provided by applicant): Intravital imaging of cell-specific metabolic activity is of key importance for understanding of a wide range of clinical conditions. Among them is compromised blood perfusion following a stroke and a decrease in efficiency of single-cell respiratory processes that occurs in neurodegenerative diseases such as Alzheimer's and Parkinson's disease. However, no methods are available today for in vivo measurement of single cell metabolism with sufficient sensitivity to resolve fast metabolic events related to ongoing neuronal electrical activity and neuronal responses to stimulation. To meet this challenge, we will adapt an existing technology, 2- photon laser scanning microscopy, for high-resolution microscopic imaging of functional metabolism of single brain cells, taking advantage of intrinsic fluorescence of metabolic cofactor 2-nicotinamide adenine dinucleotide (NADH). This project combines a bioengineering effort with testing of biological hypotheses and brings together an interdisciplinary team of experts in neuroscience, physics, engineering and computational science. From the engineering perspective, we propose to (1) develop and validate 2-photon imaging of NADH with sufficient sensitivity to detect functional changes from single cortical neurons and astrocytes in response to sensory stimulation in living animals, and (2) optimize the optical design and experimental protocol for simultaneous in vivo 2-photon imaging of metabolic, neuronal and vascular activity with high spatial and temporal resolution. Using these technological developments, we will explore new concepts concerning the mechanisms of neuro-vascular-metabolic coupling. Specifically, we will (1) address the relative contribution of oxidative phosphorylation and glycolysis to the transient metabolic response in neurons and astrocytes, (2) test the relationship between astrocytic metabolic response and regulation of blood flow through astrocytic calcium-dependent mechanisms, and (3) establish functional NADH imaging as a biomarker for hypoxia and mitochondrial dysfunction. These goals will be achieved by focusing on temporal signal characteristics, by investigation of simultaneously acquired signals, and by using in vivo pharmacology. The main deliverable of the proposed project - a tool for simultaneous 2-photon imaging of metabolic, neuronal and vascular activity - has a potential to transform the investigation of rodent models of human brain disease by opening an unprecedented opportunity to study the homeostasis and functional interactions among neurons, glia, and capillaries of the living brain. In future, this tripartite imaging approach repeated at different stages of disease would allow establishing a defined set of in vivo imaging biomarkers characterizing the progression of neuro-vascular-metabolic pathology that could be used for objective screening of potential therapies. PUBLIC HEALTH RELEVANCE: We will adapt an existing technology, 2-photon laser scanning microscopy, to visualize functional metabolism of single brain cells in living animals taking advantage of the intrinsic fluorescence of metabolic cofactor 2- nicotinamide adenine dinucleotide (NADH). Furthermore, we will combine NADH imaging with 2-photon measurements of neuronal and vascular activity and will establish NADH as a microscopic imaging biomarker for hypoxia or a decrease in mitochondrial efficiency in a mouse model of human disease.

This is a development project to construct a scanning two- and three-photon microscope for deep imaging in the brain in support of activities related to neuronal circuit analysis and neurovascular coupling. The ability to image ever deeper in the brain with optical methods is a key enabling technology in our ability to decipher neuronal anatomy and circuit function as well as neurovascular function. Optical tools, together with labels of specific brain structures, are the only means to probe the geometry and state variables of single cells, e.g., voltage and second messengers, and the dynamics of brain vasculature in a noninvasive or partially invasive manner in vivo. The current method of choice for in vivo imaging makes use of two-photon microscopy with a 100-femtosecond pulsed laser sources to observe structure and dynamics throughout the upper ~ 500 micrometers of cortex of mice. Yet there is a clear need to image throughout the full depth of cortex, 1.0 to 1.2 micrometers in mice, to determine the complete flow of information in cortical processing. There is also a need to image deeper still into hippocampus and other subcortical structures without excavated overlying tissue, as well as to determine the loci of vascular control throughout gray and while matter. The initial proposed experiments, all of which depend on the proposed instrument, address topics in fundamental brain science as well biomedicine. Fundamental issues revolve around neuronal plasticity and memory formation and include: the formation of motor memories, where the learning of a behavioral task is believed to follow from the formation of patterns of correlated neuronal output in motor cortex; the transformation of sensory signals in cortex into memory traces, such as learned fear via the amygdala and induction of depression via the habenula; the role of specific gene products, known as inducible transcription factors, in synaptic plasticity; and understanding how the prodigious adult neurogenesis in the olfactory bulb is integrated into ongoing olfactory function. More applied issues concern the role of exposure to nicotine alone in changing the basis for memory formation, as well as issues in vasodynamics, including the locus for neuronal control of its own nutriment supply through the cortical vasculature and the impact of microinfarctions on cell death within the white matter, where myelinated fibers traffic information from sensory to motor areas that span the cortical mantle. Realization of this system will permit training of graduate students and postdoctoral fellows in state of the art in vivo optical imaging. UC San Diego, along with the greater La Jolla scientific community, supports a large and highly collaborative neuroscience community with graduate students and fellows who will pursue careers at institutes throughout the county, even the world. They will be inspired to think of new experiments based on the capabilities of imaging new vistas in the brain, as well as new associated technologies, particularly in the design of optical probes of yet unmeasured variables. Lastly, the high density of potential users within this community will facilitate unanticipated refinements of deep imaging and perhaps transform the proposed development project into a turn-key design for the benefit of the global neurosciences communities.


The PI proposes to build an instrument, whose design is motivated by three threads of work, that enables two- and three-photon imaging throughout the full depth of cortex and into deeper structures. First is the use of 100-fs pulsed laser light at wavelengths of 1.3 or 1.7 micrometers, where scattering is minimized but absorption by water is still weak; second is the use of an optical amplifier to increase the energy per pulse and drive fluorescence at greater depths, and third is the use of aberration corrective optics to counteract distortion of the incident beam with increasing depth into brain tissue.

DESCRIPTION (provided by applicant): The overarching goal of this proposal is to push the depth penetration of multiphoton microscopy targeting neuroscience applications in need of large-scale recording of cortical activity where high resolution requirement (in the order of singl microns) cannot be relaxed. To achieve deep high-resolution imaging while retaining sufficient signal-to-noise ratio of the measurements for imaging of activity (e.g., for detection of single spikes induced calcium transients), we will develop an unconventional non-degenerate 2-photon microscopy capitalizing on the recent practical demonstration of the advantage of using long wavelength light (~1700 nm) for deep penetration 3-photon microscopy but circumventing the low probability of 3-photon absorption (3PA). Our deliverables - complementary to engineering efforts elsewhere aimed at large-volume sampling - would have a transformative impact on our ability to reconstruct spatially distributed neuronal circuit activity providing unprecedented opportunities for tests of biological hypotheses that are currently unfeasible. The seed of the new technology is a well-known phenomenon where absorption of the second photon by the fluorophore molecule is enhanced through an intermediate state induced by absorption of the first photon. This warrants an increase in the excitation efficiency given the right combination of the wavelengths. For our goal of deep penetration, the IR beam will deliver high photon flux to the focal volume inside the cortical tissue. The second higher energy photon beam will have lower intensity. Thus, while the higher energy photon beam would experience higher scattering in the brain tissue, the flux requirement for this beam will be relaxed (compared to that in the conventional 2-photon microscopy) helping to achieve deep imaging. Importantly, by increasing the intensity of the IR beam while lowering the intensity of the shorter wavelength beam, we will decrease the unwanted out-of-focus excitation on the brain surface. This is because the shorter wavelength beam will not have enough photon density at the surface while the IR beam will lie outside the degenerate 2- photon absorption (2PA) range for visible emission fluorophores. Finally, we will implement an innovative Adaptive Optics strategy to correct the phase distortions that will be experienced by the beam delivering higher energy photons. Specifically, we use the IR beam, which can be focused well deep inside the tissue, as a reference point (guiding star) and adjusting the phase of the second beam to reach the maximum overlap. Overall, we expect to achieve ~1.6 mm penetration inside the cortical tissue while avoiding excessive laser power and retaining the excitation volume characteristic for 2PA of the IR beam alone in the degenerate excitation mode. Our endpoint deliverable will be a prototype device with the proof-of-concept demonstration of its performance for imaging of brain activity in vivo using synthetic calcium indicators and genetically encoded calcium-sensitive fluorescent proteins. This project will lay the groundwork for an academic-industry partnership proposal in 3 years to fully develop and deploy the new technology as a commercial product.

The computational properties of the human brain arise from an intricate interplay between billions of neurons connected in complex networks. However, our ability to study these networks in healthy human brain is limited by the necessity to use noninvasive technologies. This is in contrast to animal models where a rich, detailed view on the cellular level brain function has become available due to recent advances in microscopic optical imaging and genetics. Thus, a central challenge facing neuroscience today is leveraging these mechanistic insights from animal studies to accurately draw physiological inferences from human noninvasive signals. In the proposed project, we focus on the ?Calibrated? Blood Oxygenation Level Dependent (BOLD) fMRI technology asking the questions: ?Which aspects of the underlying neuronal activity can be reliably inferred from noninvasive cerebral blood flow (CBF) and Cerebral Metabolic Rate of O2 (CMRO2) observables?? and ?What further information can be obtained from combining Calibrated BOLD with Magnetoencephalography (MEG)?? Our central hypothesis is that specific neuronal cell types have identifiable ?signatures? in the way they drive changes in energy metabolism (CMRO2), blood flow (CBF) and contribute to macroscopic electrical signals (MEG current dipole dynamics). Because other factors may affect baseline flow and metabolism, our focus is on the evoked absolute CMRO2 and CBF changes associated with increased or decreased neuronal activity. We will perform parallel experiments in mice and humans to empirically connect the dots between the microscopic properties of brain's functional organization and their manifestation on the macroscopic level of noninvasive observables. Based on the experimental results, we will then develop a computational framework that will establish connections between scales and measurement modalities enabling robust estimation of the critical aspects of neuronal circuit activity from noninvasive measurements in humans. The proposed project will deliver a quantitative probe for neuronal activity of known cell types in human brain enabling a paradigm shift in human fMRI studies: from a simple mapping of fMRI signal change to the explicit estimation of the respective activity levels of specific neuronal cell types without confounding effects of the baseline state of flow and metabolism.  

Recent advances in pluripotent stem cell technology have enabled generation of neuronal cell lines and cerebral organoids from human embryonic stem cells (hESCs) as well as human induced pluripotent stem cells (hiPSCs) derived from peripheral tissues. These organoids are self-assembled, 3D cellular structures that resemble early developmental stages of the human brain opening unprecedented opportunities for investigation of human neuronal network-level dysfunction underlying developmental brain disease. However, the lack of the natural brain microenvironment in cultured organoids can influence the phenotype and maturation of the reprogrammed neurons. To mitigate this limitation, we recently transplanted human cerebral organoids into the mouse brain and demonstrated their differentiation and vascularization using 2- photon imaging through cranial ?windows? made of glass. In the proposed study, we will replace these windows with optically transparent graphene electrode microgrids, developed by members of our team, to enable multimodal longitudinal monitoring and interrogation of neuronal activity in the graft and the surrounding host neuronal circuits. The in vivo organoid transplantation and graphene electrode arrays are existing technologies providing Scientific Premise to the proposed project ? without these parts in place, we would be building a bridge too far. Our current goal is to combine these technologies creating a synergistic and transformative result. Critically, this combination will allow examination of cell-type specific spiking (detected with 2-photon imaging) referenced to large-scale network events arising either from the organoid or the host cortex (detected as Local Field Potentials, LFPs, by a graphene electrode grid). We know from our prior work that organoids achieve sufficient laminar organization and synaptic connectivity to generate LFPs, adding to the Scientific Premise. We will engineer implantable graphene devices that would adhere to the cortical surface and the organoid to enable stable, longitudinal recordings, imaging, and photostimulation in mice transplanted with human cortical organoids (Aim 1). Then, we will provide a proof-of-principle demonstration of the unique advantage of this multimodal technology for studying the evolution of organoid activity during its maturation in vivo. To this end, we will focus on participation of specific cell types in LFP events originating from either the organoid or the host (Aim 2). All experiments will be performed in awake mice without confounds of anesthesia. Since transplantation of human cerebral organoids in the mouse brain is still in infancy, this project will deliver a much needed tool for comprehensive functional assessment of this novel biological model system. Further along the road (outside the current scope), this model system will find its use for investigating aspects of human brain development and developmental disorders.

Abstract Ascending neuromodulation associated with cognitive functions, such as arousal, attention, learning, memory, decision making, evaluation of reward, are active in any conscious human subject participating in a Blood Oxygenation Level Dependent (BOLD) functional Magnetic Resonance Imaging (fMRI) study. And yet, our understanding of how these systems affect the BOLD signal remains rudimentary. In fact, our current knowledge of neurovascular and neurometaboic mechanisms that underlie the BOLD signal has been derived almost exclusively from studies in anesthetized animals where the state of neuromodulation was uncertain. Recently, we have developed optical reporters for dopamine (DA), norepinephrine (NE), and acetylcholine (ACh) applicable for high-resolution imaging of brain function in awake behaving mice. In the proposed project, we will combine these reporters with an integrated suite of the BRAIN Initiative tools, developed by us and others, to investigate the microscopic makeup of ?brain states? and their reflection in macroscopic BOLD fMRI signals. These tools (except fMRI) are only applicable to model organisms. Therefore, all experiments will be performed in awake behaving mice. Our Central Hypothesis is that ascending projections from one or more neuromodulatory systems contribute critically to generation of spontaneous (?resting-state?) hemodynamic fluctuations as well as task-induced hemodynamic responses. To test this hypothesis, we will investigate the relationship between neuronal, vascular and metabolic activity as a function of (i) intrinsic brain states (Aims 1-2), and (ii) exposure to cocaine ? a common drug of abuse that acts by affecting neuromodulation (Aim 3). Brain states will be operationally defined based on the readout of DA, NE, and ACh reporters referenced to electrophysiological/imaging measures of local cortical dynamics. These studies will be performed in the context of resting-state hemodynamic fluctuations as well as task-induced hemodynamic responses in the primary somatosensory and frontal cortices. The proposed project will (i) provide a stronger physiological foundation for resting-state and task-induced fMRI in healthy individuals; (ii) place the relationship between the state of neuromodulation and energy expenditure (cerebral metabolic rate of O2, CMRO2) on a quantitative footing; and (iii) examine the effects of cocaine on neuronal and hemodynamic brain activity. This study will also generate further hypotheses about the ways in which substance exposure may affect fMRI readouts.

Abstract The goal of the Administration Core is to ensure that the Projects are able to achieve their scientific and dissemination objectives while in parallel foster synergy and overseeing interactions between the Projects to ensure effective and coherent operation. Our central scientific and organizational leadership will rest with the Internal Advisory Committee (IAC). The Team Director (TD) Anna Devor will oversee IAC, provide overall scientific leadership and guidance, manage the sharing and dissemination efforts, and be assisted on a daily basis by administrative assistant (AA) Parya Farzam to achieve the goal of the Administration Core. For grants management including financials, progress reporting, and purchasing, the U19 team will work with their existing respective infrastructure at Boston University (BU), University of California San Diego (UCSD), Massachusetts General Hospital (MGH) and University of Illinois Chicago (UIC). This allows the Administration Core to focus on the administrative tasks specific to research activities of this U19. The Core, physically located at BU (the applicant institution), will implement the governance plan, put forward by the IAC and offer any course corrections to IAC if needed. In addition, we will establish a to-be-named External Advisory Board to be convened annually to assess progress and accomplishments, and to provide guidance and direction to the team (IAC). The TD and the AA will prepare the annual progress report, maintain a list of publications, presentations, other research products, and intellectual property that have resulted from this U19 funding. The Core will evaluate and maintain all materials transfer agreements and other legal agreements and activities related to sharing and dissemination that arise during the course of this program. The Core will maintain the IACUC and IRB approvals for the U19, including updating approvals and continuing review documents from the subcontract institution. Our sharing and dissemination plan is extensively described within the Data Science Core. The lead of the Data Science Core Dr. Freund will have primary responsibility over these operations and will be assisted by the Administration Core. Progress on fulfilling the goals of the sharing and dissemination plan will be reviewed at the monthly IAC meetings to identify any issues that may need attention and also to identify new opportunities and needs to broaden these activities.

PROJECT SUMMARY/ABSTRACT ? OVERALL We seek to understand the nature of the pial neurovascular circuit, whose dynamics is characterized by ultralow frequency oscillations near 0.1 Hz that parcellate into separate coherent regions across cortex. We will use this knowledge to form a mathematical relation between the hemodynamic patterns observed in optical and functional magnetic resonance imaging experiments and the underlying brain state. Our proposed studies propose to leverage our experimental expertise in in vivo optical microscopy in mouse and fMRI in mouse and human. These primary modalities for data acquisition are combined with behavioral training, electrophysiology, and data analysis. Our experimental effort is parallel by two theoretical efforts. One mixed analytical/computational effort is on coupled oscillator dynamics to formulate models, at varying levels of complexity, of the pial neurovascular circuit. A second solely computational effort concerns the modulation of the transport of oxygen, by regional oscillations of the pial neurovascular circuit. The pial neurovascular circuit is composed of a two-dimensional network of pial arterioles that undergo rhythmic oscillations in the ~ 0.1 Hz vasomotor band. Each element in this circuit - a segment of arteriole whose diameter is modulated by the constriction/dilation of smooth muscle, contains an intrinsic rhythm generator, much like intrinsic bursting neurons in central pattern generators. The pial arterioles integrate neuronal activity from neighboring arterioles, underlying neurons, subcortical neurons, and neuromodulatory centers to produce dynamic patterns of coherent oscillations in arteriolar diameter across the cortical mantle. These patterns contain regions that oscillate at slightly different frequencies, i.e., they parcellate into separate regions. The fascinating issue is that the parcellation only partially reflects input from the directly underlying neuronal input. We seek to understand, model, and exploit this parcellation. The PIs have collaborated on issues in neuroscience and neurovascular science for many years. This proposal is a result of their discoveries and converging interest in a structured collaborative effort. Project 1 will formulate an understanding of fundamental physiology of the pial neurovascular circuit. This includes determining if brain arterioles truly act as interacting non-linear oscillators, i.e., that they entrain and phase-lock rather than passively filter. Projects 1, 2, and 4 will explore experimentally and theoretically how four competitive interactions, viz, input from neighboring arterioles, (ii) input from underlying neurons, (iii) input from subcortical areas involved in homeostasis; and (iv) input from brain neuromodulatory centers, lead to the observed patterns of pial neurovascular activity. Projects 2 and 4 will explore and model the regulation of oxygen in subsurface vessels, while Project 3 will expand the resolution of MR imaging in humans to observe single vessels CBV changes and thus measure pial neurovascular dynamics with unparalleled resolution. A particular interest is to transform spatiotemporal patterns of vasomotion into predictions of internal brain state.

Abstract We propose to investigate the role of neuromodulation in the phenomenon of ?whole-cortex? activity of the pial neurovascular circuit. This circuit is composed of a network of pial arterioles that integrate neuronal activity with the intrinsic arteriolar vasomotion producing dynamic patterns of coherent oscillations in the arteriolar diameter effectively parcellating the cortical mantle. Prior research suggests that ascending neuromodulatory systems may work in parallel affecting the brain state and processing capacity of large-scale cortical networks. In the majority of these studies, however, the presence of neuromodulatory neurotransmitters in the cortex was not directly measured. Rather, their release was inferred from stimulation of the corresponding subcortical nuclei or indirect measures. To overcome this limitation, in the proposed project we will use direct, selective and sensitive optical probes for acetylcholine, norepinephrine, dopamine and serotonin and track the presence of these neurotransmitters in space and time across the cortical mantle in awake behaving mice. We will combine these probes with optical imaging of neuronal Ca2+, blood oxygenation, optically transparent electrode arrays, optogenetic manipulations and BOLD fMRI. Using these tools, including those pioneered by the members of our team, we will address the role of neuromodulation in generation of (i) large-scale spontaneous cortical neuronal activity observed with wide-field Ca2+ imaging, (ii) temporally coherent patterns of vasomotion in the pial neurovascular circuit, and (iii) the resultant spatiotemporal pattern of hemodynamic fluctuations. Further, we ask whether these spatiotemporal patterns of vasomotion and hemodynamics, which can be measured noninvasively, can be used to infer the underlying internal brain state and/or activity of specific neuromodulatory systems. We will collaborate with Project 1 to understand the rules of integration of the neuromodulatory drive with local neuronal activity and intrinsic oscillatory dynamics within the pial neurovascular circuit. We will also collaborate with Project 3 to ensure that our findings translate up the scale from mice to humans. A critical link to Project 3 will be simultaneous optical/fMRI studies in awake mice. Finally, we will work with Project 4 to devise a phenomenological mathematical model that captures the essence of a brain state from the standpoint of the vascular integrator producing large-scale patterns of coherent vascular/hemodynamic fluctuations. This Project will provide a novel, unprecedented view on the role of neuromodulation in orchestrating large- scale spontaneous neuronal and hemodynamic activity, explore the underlying mechanisms, and offer a strong physiological foundation for the interpretation of large-scale fMRI signals and better understanding of the mechanisms linking spontaneous neuronal activity to cognitive performance. In collaboration with other Projects, we will deliver a predictive, conceptual model of local and global control of the pial neurovascular circuit and inference of brain states and specific neuromodulatory circuits in humans.