Altered resting-state functional connectivity after cortical spreading depression in mice

Abstract

Cortical spreading depression (CSD) underlies some neurological disorders. Previous imaging work suggests that CSD is associated with functional and structural alterations in the cerebral cortex. However, the changes in cortical functional network following CSD are poorly understood. The present study examines the changes in resting-state function connectivity (RSFC) of the mouse sensorimotor cortex after the onset of CSD by using optical intrinsic signal imaging. Our results show that RSFC between ipsilateral sensorimotor cortex (the cortex where CSD spreads) and contralateral sensorimotor cortex (the cortex where CSD does not spread) was significantly reduced after CSD. Moreover, a marked connectivity increase was found after CSD not only within contralateral somatosensory cortex and contralateral motor cortex themselves, but also between contralateral somatosensory cortex and contralateral motor cortex. Amplitude of low-frequency fluctuation (ALFF) analysis revealed an increase in ALFF in the ipsilateral cortex but a decrease in the contralateral cortex after CSD, indicating different effects of CSD on the neural activity in the ipsilateral and contralateral sensorimotor cortexes. These results suggest that CSD would alter the RSFC in the sensorimotor cortexes, and functional connectivity analysis may help to understand the effect of CSD on the cerebral functional network.

Introduction

Cortical spreading depression (CSD) is a self-propagating wave of profound neuronal and glial depolarization, associated with a depression of synaptic activity (Leao, 1944, Martins-Ferreira et al., 2000, Somjen, 2001). Clinical and experimental investigations indicate that CSD plays a crucial role in some neurological disorders, such as migraine with aura, epilepsy, stroke and traumatic brain injury (Gorji, 2001, Lauritzen et al., 2011). In mice, severe cerebral blood flow (CBF) reduction and cerebral metabolic rate of oxygen decline are found during and after CSD (Yuzawa et al., 2011). Cerebral microstructures are also affected by CSD, and in vivo two-photon microscopic imaging reveals that CSD could induce the dendritic beading and spine loss (Risher et al., 2010, Takano et al., 2007). As a basic functional unit of integration of neural circuits, distortion of dendritic spine would cause the functional network to be changed (Hasbani et al., 2001). CSD could spread across the unilateral hemisphere, and the integrity of cerebral regions supporting normal brain function would be changed, which might alter cerebral functional circuits. However, CSD-related alterations in functional circuits, especially interactions between ipsilateral cortex (the cortex where CSD is elicited and spreads) and contralateral cortex (the cortex where CSD does not spread) are rarely examined.

The traditional approach to studying functional connectivity is to probe task-elicited neurobiological responses. Recently, there has been increasing interest in neuroimaging of resting-state functional connectivity (RSFC) to investigate the ongoing neuronal processes. The coherence in spontaneous, low-frequency fluctuations (i.e., < 0.1 Hz), which arises from neurovascular mechanisms, is believed to reflect intrinsic functional connectivity (Biswal et al., 1995). Resting-state functional magnetic resonance imaging (fMRI), which examines temporal correlations in the blood-oxygen-level-dependent signal, has been widely used to explore the functional connectivity, and distinct networks supporting vision, auditory, motor and cognitive function in the human brain have been identified (Beckmann et al., 2005, Biswal et al., 1995, Hampson et al., 2002). As the disruptions in functional connectivity are considered to be related to pathological states (Pizoli et al., 2011), many animal models are developed to study resting-state networks (Hutchison et al., 2010, van Meer et al., 2010b), which afford experimental manipulations not possible in humans. However, it is difficult to replicate human fMRI findings of RSFC in small animals, for example, mice, due to the small size of the brain. High spatial resolution is necessitated for small-animal-specific fMRI scanners, and the requirement is also expensive. Optical intrinsic signal imaging (OIS), which measures neural activity through the neurovascular response by converting changes in reflected light intensity from the surface of the cortex to changes in local hemoglobin concentrations including oxy-hemoglobin (HbO), deoxy-hemoglobin (HbR) and total hemoglobin (HbT), has been shown to have high signal-to-noise ratio and spatial resolution (Dunn et al., 2005, Grinvald et al., 1986, Sun et al., 2011, Ts'o et al., 1990). And, functional connectivity with OIS (fcOIS) has been used to examine the RSFC in mice (Bero et al., 2012, White et al., 2011). The approach of fcOIS has a similar hemodynamic contrast as fMRI but with higher spatial resolution and much lower cost.

Connections between homologous regions in the left and right hemispheres have been shown to be among the strongest relationships evaluated by functional connectivity (Stark et al., 2008). CSD is known to spread in the ipsilateral sensorimotor cortex but not in the contralateral regions. Here, we used fcOIS to examine the changes of RSFC between ipsilateral cortex and contralateral cortex after CSD in mice. Furthermore, to explore the effects of CSD on the neural activity in bilateral sensorimotor cortexes, amplitude of low-frequency fluctuation analysis (ALFF) was performed (Zou et al., 2008).