Nozomi Nishimura - US grants

DESCRIPTION (provided by applicant): Recently, cerebral microvessel disease has been identified as an important component of Alzheimer's disease. The mechanism of interaction between the diseases is still unclear in part because animal models of microvascular disease are lacking. The proposed work studies the interrelationship between microvascular damage and the accumulation of Ap, the dominant characteristic of Alzheimer's disease. This work builds on the clinical observation that the severity of dementia in Alzheimer's disease is often related to the presence of vascular disease. We use novel optical tools to induce microvascular lesions in transgenic mouse models of Alzheimer's disease and then image the progression of the resulting pathology. Our lesioning technique, femtosecond laser ablation, can disrupt individual microvessels as deep as 500 pm beneath the cortical surface. The study includes multiple types of microvascular lesions, including hemorrhages, ischemic occlusions and transient leakages, all of which potentially contribute to disease progression. Two-photon excited fluorescence microscopy is used to image amyloid plaque development and to measure blood flow and leakage in the microvasculature. This allows time-lapsed study of both the microvascular lesion and amyloid plaque. Post-mortem labeling with A(3 antibodies will be used to further elucidate the impact of the microvascular lesion on Ap accumulation. In Aim 1, we test whether microvascular clots and hemorrhages trigger rapid amyloid plaque formation at different locations in the vascular tree. In Aim 2, we ask if vascular lesions earlier in life can induce a predisposition to plaques later. In the final aim, we determine where plaques that are seeded by vascular lesions are relative to different cell types and determine whether inflammation or reactive oxygen species are factors through colocalization studies. In addition, we use histological and immunohistological assays to identify the affected cells and map the Ap accumulation. Our preliminary findings predict that the presence of a microvascular clot will accelerate the local deposition of Ap plaques. These data suggest that microvascular lesions could play an important role in Alzheimer's disease pathogenesis. Relevance ~ Alzheimer's disease is the most common cause of dementia in the elderly. Clinically, Alzheimer's disease is often entangled with vascular disease, suggesting that the two diseases are intimately interrelated. In many patients, sucessful treatment will have to address both aspects. This work investigates how the two conditions might worsen each other and will help identify strategies for preventing dementia.

Neurons in the brain interact with other cell types. One type, microglia, are star-shaped cells that survey the connections between neurons with their arm-like branches. It is important to understand these interactions because changes in neural connections underlie learning and memory formation. The Principal Investigator hypothesizes that microglia might regulate such changes in neural connectivity. This project will investigate whether microglia are directly involved in disconnecting neurons, and whether these actions are disrupted by injury. In response to injury, microglia lose their branches, and the Principal Investigator hypothesizes that this prevents them from properly regulating neural connections. The work will use a specially-designed microscope that images cells within the living brain. To make the cells visible, the microglia will be engineered to look green under the microscope, and, using a strategy developed in this project, each neuron will be labeled with a unique color. With this approach, it will be possible to visualize how microglia interact with neurons under normal conditions and after injury. Analyzing such images is complicated and requires neurobiologists who are trained to use high-level mathematics. This project will develop software and teaching tools to make the mathematics behind image analysis understandable to scientists who were not trained as engineers. The Principal Investigator will create an inquiry-based module about these ideas called "Beyond the Image," which will introduce high school students to image analysis and digital processing. The module will be made available to teachers everywhere through websites such as the Cornell Cooperative Extension School's 4-H STEM site.

Because neural branches are small and densely packed, the proposed imaging of microglial modulation of neural connections is challenging. A technical innovation will be to infect neurons using a combination of virus strains that each generate a different color of fluorescent protein. Each neuron will be labeled with its own unique color code, enabling its branches to be distinguished. An advantage of using viruses is that both the type and location of labeled cells can be varied to optimize color-coding. Rather than wait for neurons to disconnect by chance, a laser will be used to precisely cut a branch off of a neuron. Time-lapse imaging will capture the reactions of microglia and neurons to the now useless, severed branch. The prediction is that microglia will reach out with processes to engulf and remove defunct connections. However, with injury, the microglia might erroneously allow aberrant connections to linger or might inappropriately damage normal connections.

Measuring and understanding the activity of individual neurons is critical for understanding how neuronal circuits function and lead to behavior. Two-photon microscopy of calcium-sensitive indicators has produced insightful data on the role of individual neurons with populations. With the use of head-fixed or miniaturized versions, such optical techniques have lead to links between neural dynamics and behavior. However, these methods have not translated to voltage indicators, so that the understanding of how spikes across populations of cells affects circuits and behavior is lacking. Much of the difficulty is related to the need in two-photon microscopy to scan a small laser focus throughout a volume which results in limited time resolution. Signals from reporters are associated with a neuron because the fluorescence was generated when the laser focus was at the location of that neuron. An alternate strategy based on using multiple colors of indicators to color code neural output is proposed here. This strategy relies entirely on spectral information, so no location information or image formation is required. This enables high-speed data acquisition. This strategy takes advantage of the availability of multiple colors of genetically encoded voltage indicators and associates individual neurons with unique color combinations. A mix of adeno-associated virus vectors, each carrying DNA for a indicator of a particular color, is injected into the brain. Because the infection process is stochastic and neurons are infected by several particles, neurons are labeled by a random combination of colors. In this proposal, the number of colors, delivery methods and analysis of signals is optimized for the identification of individual neurons within a population. For the benchmarking this technology, these novel signals are compared to the performance of multiphoton microscopy and electrophysiology in assays of neural activity.

Interstitial fluid (ISF) flow has many functions including the maintenance of ionic balances, flushing of wastes and providing a route for migration of both cell signals and cells. Recently, the use of multiphoton microscopy, which enables in vivo studies with cellular resolution, has resulted in novel findings, especially in the brain, about dynamics and anatomy involved in ISF flow. Notably, ISF flow may be critical in dealing with protein accumulation in Alzheimer's disease and is regulated by sleep. Although much progress has been made, there remains controversy about some of the fundamentals regarding ISF flow. Much of this may be due to complications in the experimental methods. Studies to date require the injection of tracers which can be imaged by multiphoton microscopy or other imaging methods such as MRI. However, the process of injection of the tracers may itself affect the flow due to the delicate balance of pressures within the brain. In addition, injected tracers do not mimic the origin of proteins, wastes and cytokines made by the brain. Studies are limited to superficial sites accessible by current generation imaging technologies. This proposal generates optical and biological tools that address these short comings using in vivo multiphoton microscopy. First, in a new way to generate tracers in situ, cells within the tissue of interests will be transduced so that they secrete fluorescent proteins into the extracellular space. This will be used to resolve existing controversies about the route of ISF flow within the brain. Second, the newly discovered brain lymphatics are thought to link to the peri or paravacular spaces that serve as conduits for ISF. This work will use the new secrete tracers to answer whether and how these lymphatic channels link to the these spaces. This fluid flow may be altered in different conditions, so this will be studied in normal function mimicking sleep and waking, as well as with a stroke model. Third, three-photon microscopy now enables much deeper imaging than traditional two-photon microscopy. This enables imaging of anatomy previously not accessible. This work will study ISF flow in the hippocampus, a critical brain structure in memory and cognition, that seems to be particularly vulnerable to disruptions to ISF. This work will establish novel tools that can enable new experiments to address ISF in many systems.

Brain-machine interfaces (BMIs) have the potential to enhance quality of life by restoring functions lost to neurological disease. However, current minimally-invasive BMIs (e.g. cortical surface electrodes) do not provide enough information for complex control. BMIs that yield high information content (e.g. implanted electrode arrays) damage brain tissue, limiting their lifespan. This project will develop a next-generation BMI that uses light instead of electricity, is less invasive and longer-lasting than implanted electrode arrays, and is capable of high information-content recording. This proposal will enhance teaching materials and summer programs with strong emphasis on recruitment of UMR students (and those who did not have real lab research experience). This could positively influence more students and attract them to multidisciplinary sciences or STEM programs in general.

The project will virally express calcium-sensitive bioluminescent proteins in neurons, so light is emitted with neural activity. To distinguish different neurons, multiple colors of such sensors are stochastically expressed, so each neuron contains different amounts of each color of the bioluminescent sensor, yielding a unique spectral signature. To monitor neural activity, the wavelength and intensity of emitted light is detected as a function of time. These data are processed to yield information on the activity of a large ensemble of individual neurons. This approach could provide a path to a long-term, robust, and high-fidelity BMI.

Abstract Our objective is to investigate the clearance of interstitial fluid from the brain and improve our understanding of the accumulation of proteins in the walls of the vasculature in ageing brains. There is significant evidence suggesting that pathologies associated with the clearance of amyloid-? from the brain contributes to the occurrence of Alzheimer?s disease. Our hypothesis is that the interstitial fluid, driven by hydrodynamic forces exerted by the pulsation of the arterial basement membranes, transports amyloid-? along the perivascular spaces outside the cerebral arteries. We have developed a preferential transport theory where the driving forces stem from the superposition of forward- propagating waves and their associated reflections along the arterial basement membranes. Perivascular pathways for lymphatic drainage have been identified in both experimental animals and in humans. Interstitial fluid and solutes enter bulk flow pathways in capillary and artery walls to drain to the cervical lymph nodes. However, it has also been shown that there is transport of the cerebrospinal fluid in the opposite direction as evidenced by the injection of tracers into the cisterna magna. As part of this proposal, we seek to clarify these seemingly contradictory results. Central to this proposal is the use of multi-photon microscopy to measure the transport of fluid along the arterial perivascular spaces. The effects of waveform pulsatility and pathway geometry on the fluid transport will be elucidated by studying mouse models with and without cerebral amyloid angiopathy. Our theory predicts that the arterial wall deformations drive transport. Therefore, we propose to monitor the deformation of the arterial walls using the third harmonic generation obtained from multi-photon microscopy. This novel, label-free imaging methodology is ideal for capturing the arterial wall motion. We will measure the time-dependent radial position of the various compartments in multiple places along the vascular hierarchy by including some of the first few branches off of the vertically oriented arterioles and venules. The in vivo data will be used to refine and validate our preferential transport theory for drainage through the basement membranes. Once validated, this theory can be used to provide key insights on the physiological parameters that govern amyloid-? clearance and accumulation in the perivascular space.

As the primary controllers of epithelium regeneration, intestinal stem cells (ISCs) at the bottom of the crypt must maintain a balance between self-renewal and differentiation. However, it is still unclear how the stem cells maintain tissue homeostasis in response to daily variations in cell numbers or after injury. It is also well known that the ability to repair damage is reduced in aging, but it is not known what mechanism(s) underly this process. We have recently developed a chronic preparation for in vivo imaging with multiphoton microscopy which allows the monitoring the ISC niche in real-time in mice expressing green fluorescent protein in Lgr5+ ISCs. The goal is to directly track and identify how stem cells maintain their balance in the intestinal crypts. Next generation multiphoton microscopy with an in vivo imaging preparation with femtosecond laser ablation is used to ablate individual cells of a specific type to perturb the crypt. Time lapse imaging captures changes in cell number, position, motion and marker expression to identify how the various populations of stem cell respond. Upon ablation, the targeted cells lost their shape and moved out of the plane of the crypt base towards the intestinal lumen. Immediately adjacent cells appeared to move into the space left by the ablated cell, suggesting that the niche cells actively move around in response to the pattern disruption. This proposal tests the hypothesis that age and underlying pathology reduces the efficacy of these newly discovered dynamics, which can be rescued by age-delaying agents. The expectation is that these motions are involved in protecting the stem cells from damaging factors spilling from injured cells. The new optical tools have identified a potential new function of ISCs. In addition to generating daughter cells to replenish the epithelium, Lgr5+ ISCs appear to be mechanically active in eliminating damaged cells. This adds a new function to the repertoire of ISC actions. Collectively, the results suggest that there is an active process that involves cell migration in addition to cell division for maintaining homeostasis in the intestinal crypt and epithelium. !

Summary Blood flow in the brain of Alzheimer disease patients is substantially decreased as compared to age-matched healthy controls. Recently, it was found that blood flow in Alzheimer?s disease mouse models is reduced because neutrophils plug up capillaries resulting in a small, but impactful number of stalled capillaries. Removing these plugs by interfering with neutrophil adhesion improves blood flow in minutes and also improves performance on tasks involving short term or episodic memory within hours. This extremely rapid change in cognitive performance suggests that no alterations in structure or connectivity of neurons are needed to recover sizable amounts of cognitive function in Alzheimer disease. Instead, there must be a change in the neural activity that reflects the improvement in function. This phenomena enables the investigation of these changing activity patterns by recording from the same cells before and after improving the blood flow. Cutting-edge approaches, such as multiphoton microscopy of genetically encoded calcium indicators, will be used to record activity from numbers of neurons deep inside the brain while still resolving individual cells in an awake animal. Neurons with aberrant activity may underlie the observed cognitive deficits will be identified by measuring how the activity in each cell changes before and after the blood flow treatment that rescues cognition. This gives a direct measurement of how individual neurons and circuits change when an Alzheimer mouse?s cognitive symptoms are improved. Several types of aberrant neural activity are known to be associated with Alzheimer?s disease and also with impaired blood flow, suggesting that these types of neural activity patterns might change during blood flow rescue. Studies in Alzheimer disease models show some neurons are abnormally quiet or silent and other neurons fire spontaneously at excessively high rates. Imaging of neural activity is used to evaluated changes in such activity in the hours and days after blood flow rescue. Such aberrant firing patterns can contribute to a decrease in the accuracy of neural representations of stimuli. Neural fidelity is assayed by measuring tuning curves in neurons that respond to directional whisker stimulation. Imaging of somatosensory cortex tests whether blood flow rescue improves these tuning curves and sensitivity to weak stimuli. Alzheimer?s disease patients and experimental models exhibit seizure-like discharges, so EEG recording is used to evaluate epileptiform activity and any changes with blood flow rescue. Interestingly, the blood-flow-mediated cognitive changes are so fast that there is little time for slower changes such as rewiring of circuits. Understanding what are the changes that are caused by blood flow rescue that are correlated with the fast behavioral improvement is critical to understanding how the cognitive symptoms emerge in the disease. This work suggests that a metabolic component is critical to cognitive function, and treatment of the newly identified capillary stalling mechanism might provide a future target for improving patient cognition.

Summary Measuring the activity of many individual neurons at once while knowing their wiring diagrams would provide exciting information on how the components of a network interact. Knowledge of wiring diagrams has rapidly improved due to advances in the field of connectomics, and capabilities for simultaneous measurement of many individual neurons has increased exponentially with large-scale recording techniques. However, it is still difficult to combine such measurements. Registering high-resolution imaging for tracing neural projections with electrophysiological measurements, such as electrode arrays, is extremely difficult. With optical imaging, such tracing is possible, but neural activity measurements are often limited to particular geometries, most commonly a single plane in z. Although new imaging advances for volumetric imaging have eased this limitation somewhat, complicated instrumentation puts such technologies out of reach for most labs. This proposal addresses this challenge by using multicolor aequorin-fluorescent proteins (Aeq-FPs) as both fluorescent structural tracers and functional indicators for recording calcium activity. Aeq-FPs are bioluminescent indicators of calcium concentration that emit light from the entire cell including the dendritic and axonal arbors. In the proposed scheme, each neuron will express a unique combination of Aeq-FP colors so that it is color-coded to have its own spectral signature. The activity of individual neurons can be distinguished from the spectrum of the emitted bioluminescence without resolving the spatial position of the origin of the light. This enables simultaneous recording of the activity of many cells in arbitrary spatial arrangements including from different layers in the cortex. Connected networks are identified by limiting expression of the Aeq-FPs to neurons that are one synapse away from ?starter? cells using transsynaptic viral vectors (modified rabies for retrograde transport and adeno- associated viruses (AAVs) for anterograde transport). The unique color combinations expressed in each cell also facilitate structural tracing. With these combined technologies, the network of microcircuits defined by connectivity to a single ?starter? cell will be traced in three dimensions and correlated to measurements of activity in a single trial. In Aim 1, the starter cell is postsynaptic from the network, so this data will show how the presynaptic network involving multiple different types of cells from across cortical layers affects starter cell activity. In Aim 2, the starter cell is presynaptic to the labeled network and will express channelrhodopsin. Optically stimulating the starter cell will show how the network activity is affected by the modulation of the single cell. Such measurement capabilities will enable new types of experiments relating structure and activity and could be readily adopted by many labs.

Project Summary Cerebral blood flow (CBF) is reduced in Alzheimer?s disease (AD) patients and mouse models by ~20%, but there remains a limited understanding of the mechanisms causing this hypoperfusion or the potential therapeutic benefit of rescuing CBF deficits. Under the previous award, chronic in vivo two-photon excited fluorescence microscopy was used to study CBF in mouse models of AD. While no blood flow disruption in cortical arterioles or venules was observed, blood flow was found to be stalled in ~2% of cortical capillaries in mouse models of AD, as compared to ~0.4% in wild type controls. These capillary stalls appeared early in disease progression, were caused by arrested neutrophils, and had outsized impacts on CBF because they decreased flow speed in up- and down-stream vessels. Antibodies against the neutrophil surface protein Ly6G were serendipitously found to reduce the incidence of capillary stalls immediately, leading to a rescue of two-thirds of the CBF deficit, and, remarkably, to improved memory function within hours. Preliminary data further link this capillary stalling to cellular damage from reactive oxygen species (ROS). In this competitive renewal, the mechanisms underlying neutrophil arrest in capillaries in mouse models of AD and the consequences of improving CBF on AD-related pathology are explored. First, three different hypotheses about the mechanism of neutrophil arrest in capillary segments are tested: a focal constriction of the capillary by a pericyte that prevents neutrophil passage; binding of the neutrophil to increased inflammatory adhesion molecules on endothelial cells; or binding of the neutrophil to basement membrane and adhesion molecules exposed at widened gaps between endothelial cells. Second, the molecular and cellular origin of the ROS that leads to neutrophil arrest is determined using cell type-specific knockouts of ROS producing enzymes. Third, the impact of long-term CBF rescue on the deposition of amyloid- beta (A?), a driver of AD pathology, and on neuropathology will be quantified. Critical for this study are recently- developed knock-in mouse models of AD that may better capture the feedback of CBF reductions on expression of amyloid precursor protein (APP), which is cleaved to produce A?. Finally, cutting-edge three-photon excited fluorescence microscopy is used to enable imaging of the hippocampus to determine the role of capillary stalling in CBF deficits in one of the first regions of the brain that exhibits AD pathology. The hypothesis that brain hypoperfusion in AD is due to neutrophil arrest in capillaries is both novel and strongly supported by the findings under the previous award. The work proposed in this competitive renewal would uncover the mechanisms underlying that neutrophil arrest, which could suggest therapeutic targets to improve CBF that would be complementary to anti-amyloid and other treatment approaches for AD.