Microglial Calcium: An Exquisite Sensor for Neuronal Activity

With the development of in vivo imaging techniques, it was discovered early in this century that microglia are highly dynamic even in the resting state [1]. However, for over a decade, little has been learned about how the dynamic microglial motility is regulated, particularly in physiological and awake conditions [2]. Two recent studies have made a breakthrough in this regard, showing that the motility of microglial processes is regulated by a shift in the state of neuronal activity, and the neurotransmitter norepinephrine (NE) is one of the key neuronal signal mediators through acting on microglial β2 receptors [3]. Now, a new study has further revealed that Ca²⁺ microdomain signaling in microglial processes is an exquisite sensor for neuronal activity shifts, and this is associated with process extension/outgrowth (Fig. 1).

Fig. 1.

Elevation of calcium activity in microglial processes in response to biphasic neuronal activity.

Ca²⁺ is a common second-messenger in most cells and is broadly used to indicate neuronal and glial activity in neuroscience research [4]. In neurons, spiking activity in an all-or-none fashion actively spreads throughout the whole cell and propagates along neuronal circuits. As such, Ca²⁺ dynamics frequently emerges as a whole-cell event in neurons. In non-excitable glial cells such as astrocytes, however, Ca²⁺ signal events occur more locally in microdomains (soma, major branches, and minor branches), and both IP3-mediated Ca²⁺ release from the endoplasmic reticulum (ER) and Ca²⁺ influx through transmembrane channels have been found to contribute to astrocytic Ca²⁺ signaling [5]. Whether microglia exhibit the microdomain Ca²⁺ activity is largely unexplored. As generally known, the local event detection requires both high Ca²⁺ indicator sensitivity and spatial imaging resolution, which are major challenges for imaging in awake animals. Due to these limits, most previous studies on microglial Ca²⁺ signaling were from cultured cells/slices and focused on the soma. Limited available data with injury models and pharmacological operations showed that microglial Ca²⁺ transients are pre-dominantly mediated via purinergic receptors, which involve IP3-triggered ER Ca²⁺ release and subsequent Ca²⁺ release-activated channel activation. In addition, several in vivo studies of microglial Ca²⁺ signaling so far have only been performed in the anesthetized mouse, where spontaneous Ca²⁺ activity is nearly absent [6]. The question of whether and how Ca²⁺ signaling responds to neuronal activity in the awake brain was still open.

To answer this question, Wu and colleagues designed an exciting study using a new genetic mouse line in which microglia express the high signal-to-noise ratio Ca²⁺ indicator GCaMP6s by crossing Rosa26-CAG-LSL-GCaMP6s mice with CX3CR1-CreER mice [7]. The study first characterized the spontaneous microglial Ca²⁺ activity in the somatosensory cortex of head-restrained awake mice. Through a pre-installed chronic observation window, a low level of Ca²⁺ activity was detected in less than half of the microglial process areas and around one third of the somatic areas. This is consistent with previous reports from anesthetized mice [8].

To study how microglial Ca²⁺ signaling responds to neuronal activity, they used a variety of experimental approaches to manipulate neuronal firing [7]. First, they employed isoflurane anesthesia and kainate-generalized seizures to inhibit and increase the whole network neuronal activity, respectively. With the switch from awake to anesthesia, increased microglial process Ca²⁺ activity was first detected 6 min after isoflurane induction and reached peak levels after approximately 10 min–20 min. In addition, consistent with their previous report [9], increased microglial process growth occurred after anesthesia, and these areas of new process extension/outgrowth were strongly associated with both the increased Ca²⁺ transient intensity and probability. During the neuronal hyperactivity induced by kainate, they found that after the first seizure, the microglial process Ca²⁺ activity was greatly increased for both transient intensity and probability, which was also associated with process extension/outgrowth. Interestingly, strong Ca²⁺ activity was also found in microglial somata during status epilepticus. These findings demonstrate that approaches to rapidly decrease neuronal activity (isoflurane) and increase neuronal activity (kainate) similarly result in large-scale increases in the Ca²⁺ activity of microglial processes.

To determine whether a local controllable shift in network activity could induce similar microglial responses, they used the DREADD systems to precisely decrease or increase neuronal activity locally in the cortex [7]. Again, increased Ca²⁺ activity associated with process extension/outgrowth was observed after CNO injection to either decrease or increase neuronal activity. The extent of Ca²⁺ activity was dose-dependent and coupled with the degree of neuronal activity change. Unlike those in status epilepticus, these brief Ca²⁺ responses only occurred on microglial processes but not somata in response to local changes in neuronal activity. Therefore, this study uncovered a key reinforcing principle for microglia: these cells are bidirectionally attuned to neuronal activity shifts and use Ca²⁺ signaling only when changes in the microenvironment occur.

As a ubiquitous second-messenger, Ca²⁺ is known to trigger a variety of intracellular signals. Under disease conditions such as status epilepticus, microglia can be activated as shown by morphological changes and pro-inflammatory factor release. Microglial Ca²⁺ activity is thought to be involved in these activation events [10]. In this study, after the induction of a generalized seizure, they performed longitudinal observation over a 14-day period for changes in microglial Ca²⁺ signaling [7]. A greater proportion of microglial somata and processes displayed Ca²⁺ activity, with increased Ca²⁺ signaling lasting up to 10 days. The activity was also highly synchronized between the processes and soma, appearing as large, spreading Ca²⁺ waves. These results suggest that microglial Ca²⁺ signaling could be critically implicated in excitotoxicity and the injury environment in many disease contexts such as epilepsy, stroke, and neurodegeneration.

In summary, this work strengthened and extended the previous concept that microglia sense shifts in neuronal activity states in both the hypoactive and hyperactive directions. The study revealed that microdomains of the microglia (soma and processes) display clearly distinct Ca²⁺ signaling profiles in response to acute activity shifts, suggesting that microglial Ca²⁺ signaling is an important determinant of how microglia respond to brain state changes. However, the mechanisms regulating microglial Ca²⁺ signaling remains elusive and needs further investigation. For example, the mechanisms that trigger the microglial Ca²⁺ responses to neuronal hyperactive and hypoactive states could be different. ATP/ADP is a well-known molecule for transducing microglial Ca²⁺ signaling [8]. It is reasonable to suggest that kainate and Gq DREADD could increase the release of purinergic signaling molecules, but it would be illogical for Gi DREADD and isoflurane to increase purinergic signaling. Whether the previously-reported β2 disinhibition that triggers microglial process extension [9] also contributes to microglial Ca²⁺ signaling is yet unclear. In addition, the Ca²⁺ sources responding to neuronal activity shifts remain to be identified. Another important issue that is not addressed is the functional significance of microglial Ca²⁺ signaling. Although the study showed a strong correlation between Ca²⁺ elevation and process extension, the causal relationship is still largely unknown. Further study is needed to also manipulate Ca²⁺ levels in microglia to examine how Ca²⁺ signaling controls or regulates microglial activity, neuronal circuits, and even behaviors in normal and diseased brain.