Modality-specific neurovascular coupling via layer-segregated arteriole networks
INTRODUCTION
The brain continuously adjusts blood flow in response to elevations in neuronal activity through a process known as neurovascular coupling. These vascular responses form the basis of widely used brain imaging methods, such as functional magnetic resonance imaging (fMRI), which infer neuronal activity from local changes in blood oxygenation, flow, velocity, or volume. Although neurovascular signals are generally assumed to reflect neuronal activity in a consistent manner, different sensory experiences recruit distinct neural circuits across cortical layers. How these differences shape blood flow regulation within the brain remains poorly understood.
RATIONALE
We hypothesized that neurovascular coupling depends not only on the amount of neuronal activity but also on the type of input and how the activity is distributed across cortical layers. To test this idea, we compared cortical responses to multiple forms of neural input in mice, including gentle touch, optogenetic nociceptor activation, spontaneous activity, and motor-sensory feedback signals. Using widefield optical imaging and deep two-photon microscopy, we measured neuronal activity and vascular dynamics across cortical depth and spatial scales. We further combined these experiments with computational simulations to examine how arteriole topology shapes blood flow dynamics.
RESULTS
Touch and nociceptor stimulation generated markedly different blood flow responses across cortical depth. Although bulk neuronal activity in layers 2 to 5 was broadly similar between modalities, nociception-evoked blood flow responses were reduced by more than 50% in superficial cortical layers 2 and 3 (2/3). Consistent with this difference, touch produced larger decreases in deoxygenated hemoglobin (HbR)—indicative of greater blood oxygenation—and a more pronounced poststimulus undershoot (constriction) than did pain. By contrast, blood flow responses in deep layer 6 were similar for touch and pain. These effects arose because distinct forms of neuronal activity selectively recruited different classes of penetrating arterioles. Deep arterioles and their proximal branches dilated across all types of activity, whereas shallow arterioles dilated only in conditions when superficial cortical activity, particularly in layer 1, was sufficiently engaged—for example, layer 1 neuropil Ca2+ responses were ∼50% larger during touch than nociceptive stimulation. Computational simulations based solely on experimentally measured vessel diameter changes accurately reproduced the layer-specific perfusion patterns observed in vivo.
CONCLUSION
Our findings reveal that neurovascular coupling is modality-dependent and strongly shaped by vascular architecture. Different arteriole networks sample neuronal activity across distinct cortical layers and generate different spatial patterns of blood flow across cortical depth. As a result, similar levels of overall neuronal activity can produce markedly different local blood flow responses depending on how this activity is spatially distributed. Overall, our findings show that cortical blood flow patterns emerge from interactions between laminar neuronal circuitry and vascular network organization.
Web | DOI | PDF | Science | Paywall
Antoine Malescot; Milene R Malheiros-Lima; Laurianne Zana; Michael C Bennett; Éric Martineau; Franca Schmid; Ravi L Rungta
INTRODUCTION
The brain continuously adjusts blood flow in response to elevations in neuronal activity through a process known as neurovascular coupling. These vascular responses form the basis of widely used brain imaging methods, such as functional magnetic resonance imaging (fMRI), which infer neuronal activity from local changes in blood oxygenation, flow, velocity, or volume. Although neurovascular signals are generally assumed to reflect neuronal activity in a consistent manner, different sensory experiences recruit distinct neural circuits across cortical layers. How these differences shape blood flow regulation within the brain remains poorly understood.
RATIONALE
We hypothesized that neurovascular coupling depends not only on the amount of neuronal activity but also on the type of input and how the activity is distributed across cortical layers. To test this idea, we compared cortical responses to multiple forms of neural input in mice, including gentle touch, optogenetic nociceptor activation, spontaneous activity, and motor-sensory feedback signals. Using widefield optical imaging and deep two-photon microscopy, we measured neuronal activity and vascular dynamics across cortical depth and spatial scales. We further combined these experiments with computational simulations to examine how arteriole topology shapes blood flow dynamics.
RESULTS
Touch and nociceptor stimulation generated markedly different blood flow responses across cortical depth. Although bulk neuronal activity in layers 2 to 5 was broadly similar between modalities, nociception-evoked blood flow responses were reduced by more than 50% in superficial cortical layers 2 and 3 (2/3). Consistent with this difference, touch produced larger decreases in deoxygenated hemoglobin (HbR)—indicative of greater blood oxygenation—and a more pronounced poststimulus undershoot (constriction) than did pain. By contrast, blood flow responses in deep layer 6 were similar for touch and pain. These effects arose because distinct forms of neuronal activity selectively recruited different classes of penetrating arterioles. Deep arterioles and their proximal branches dilated across all types of activity, whereas shallow arterioles dilated only in conditions when superficial cortical activity, particularly in layer 1, was sufficiently engaged—for example, layer 1 neuropil Ca2+ responses were ∼50% larger during touch than nociceptive stimulation. Computational simulations based solely on experimentally measured vessel diameter changes accurately reproduced the layer-specific perfusion patterns observed in vivo.
CONCLUSION
Our findings reveal that neurovascular coupling is modality-dependent and strongly shaped by vascular architecture. Different arteriole networks sample neuronal activity across distinct cortical layers and generate different spatial patterns of blood flow across cortical depth. As a result, similar levels of overall neuronal activity can produce markedly different local blood flow responses depending on how this activity is spatially distributed. Overall, our findings show that cortical blood flow patterns emerge from interactions between laminar neuronal circuitry and vascular network organization.
Web | DOI | PDF | Science | Paywall