Overall Project Aims
The U19 is based on the observation that CSF movement is linked to large fluctuations in neural activity and vascular volume during sleep. The overarching hypothesis was that coordinated patterns of neural activity and neuromodulation drive arterial dilations and blood volume changes during sleep, and these vascular changes move cerebrospinal fluid (CSF) via periarterial pumping, removing solutes and fluid from the brain. Our rationale is that large-scale changes in neural activity are known to drive changes in blood volume (via neurovascular coupling), and that these blood volume changes within the confines of the cranial cavity must be compensated by reciprocal changes in CSF volume. Thus, we predicted that coordinated neural activity – due to global modulation of arousal and/or local neuronal activity –controls the dynamics of fluid and solute clearance from the brain. We proposed to investigate this hypothesis in both humans and mice, from the micro- to macroscale, considering neuronal and glial physiology along with smooth muscle cell activity, and integrate these measurements in realistic models of fluid flow to understand the movement of CSF across sleep and wakefulness, and the neural processes that control it.
We hypothesize that local neuronal activity can produce periarterial CSF pumping and solute clearance by driving large changes in arteriole dilation and thus CSF flow.
Project 2 has studied the effects of sensory stimulation (whisker) on periarterial pumping. We found that whisker stimulation accelerated periarterial CSF influx providing supporting the concept. To map on the micro- to brain-scale neurovascular coupling and its effect on periarterial CSF pumping, the analysis used both particle and front tracking. The data pointed toward the existence of a tight temporal and spatial dynamics of arterial dilations and constrictions with CSF movements but only during anesthesia. We found that awake brain is capable of suppressing CSF influx, likely by a valve mechanism involving the astrocytic perivascular end feet. Thus, the coupling of vasomotion with periarterial CSF pumping did not exist in in wakefulness 1. Ongoing studies are evaluation whether the 4th meningeal membrane, SLYM, contributes to state-dependent changes in brain fluid flow.
The key intermediary between neural activity and CSF flow is neurovascular coupling. In Project 3, we will therefore test the role of different populations of excitatory and inhibitory neurons in controlling arteriole dilations during sleep (Aim 1). We will also measure the temporal dynamics of neuromodulation relative to vascular changes during sleep, and mechanistically test their role in generating vascular dilations and constrictions during sleep (Aim 2). Lastly, we will determine the spatial dynamics of arterial dilations and constrictions on the micro- to brain-scale during sleep and test their role in generating CSF movements (Aim 3).
In parallel, Project 4 will establish this neural control of vascular and CSF dynamics at multiple spatial scales in the human brain. Our human imaging techniques allow perivascular, subarachnoid and ventricle CSF flow, as well as hemodynamic responses throughout the entire brain. We will use visual stimulation to noninvasively drive neural activity and modulate the spatial extent of neural activity. We can then test how driving neural activity in distinct patterns affects subsequent hemodynamics and CSF flow (Aim 1.1 and 1.2).
Project 1 will use fluid dynamics simulations to predict how arteriole dilation drives and affects CSF flow in PVSs (Aim 1), basing PVS shapes, inlet/outlet boundary conditions, and artery motion on experimental observations (Projects 2-3).
Both clearance and CSF flow rates increase during sleep. We will test the hypothesis that the coherent neural dynamics that appear during sleep induce CSF flow, by driving large oscillations in blood volume that displace CSF.
Project 3 will perform these cell-type specific manipulations of neurovascular coupling (Aim 1) and neuromodulation (Aim 2) during sleep, and test whether these effects are sleep-dependent.
Project 2 has used optogenetic stimulation of vascular smooth muscle cells. The analysis showed that vascular constriction-dilation can drive perivascular arterial pumping in the absence of neural activity 1.
In Project 4, we will test how the dynamics of spontaneous neural activity during sleep are linked to multi-scale CSF flow. Perivascular spaces and flow have never been measured in human sleep. We will test the prediction that perivascular spaces will expand and show pulsatile flow coupled to neural slow waves, and that this effect depends on sleep stage (Aim 2.1). We will then track the propagation of neural slow waves during sleep (Aim 2.2) and test whether their spatial patterns predict the velocity and direction of CSF flow.
Finally, Project 1 will model this effect, by using the vascular dynamics observed during sleep in Projects 2, 3, and 4, as an input to the brain-wide flow and clearance models (Aims 2-3). Project 1 will then compare predictions from the models to measured CSF flow and solute transport (Projects 2-4).
Acetylcholine (ACh) and noradrenaline, also called norepinephrine (NE) are critical arousal regulatory neuromodulators that shape neuronal activity during sleep. Furthermore, they also act directly on the vasculature, with opposing effects (dilation vs constriction), and could therefore induce the vascular oscillations that appear during sleep. Our projects are therefore designed to test how ACh and NE influence periarterial CSF pumping and clearance during sleep. While we focus on these two key arousal- and vascular-regulatory neuromodulators, we will also be ideally positioned from these studies to then initiate subsequent investigations of other neuromodulators, such as serotonin.
Project 3 will first measure spontaneous ACh and NE dynamics during sleep, along with arteriolar diameter (Aim 2.1). It will then test how manipulating NE or ACh levels via opto- or chemogenetics affects brain-wide hemodynamics and neurovascular coupling (Aims 2.2, 2.3). We predict that these two neuromodulators act in a push-pull fashion to induce large oscillations in vascular dilation during sleep.
Project 2 has in an extensive series of experiments mapped the interrelationship between NE signals are linked to the size and kinetics of the perivascular space and CSF flow 3,4. We find that NE change the physical and functional properties of the perivascular space during sleep that in turn enhance fluid and solute transport.
Project 4 will simultaneously image basal forebrain (BF) and locus coeruleus (LC), the nuclei that release ACh and NE, using BOLD fMRI during sleep and wakefulness. We will test how spontaneous dynamics in these nuclei are linked to both EEG slow waves (Aim 3.1), cortical BOLD responses, and multiscale CSF flow (Aim 3.2). We predict that activity in BF and LC will have distinct coupling patterns to EEG slow waves, global hemodynamics, and CSF flow during NREM sleep.
Project 1 will develop predictive, empirical relationships among ACh levels, NE levels, EEG signals, and the drivers of CSF motion: vasodilation and PVS size/properties. Those relationships, combined with microscale and brain-wide fluid dynamics simulations, will provide quantitative links from neuromodulator activity to CSF flow and clearance.