Microglia, the principal immune cells of the brain, are thought to be the nervous system's roaming cleanup crew. When activated by injury or insult (including lesions, stroke, neurodegenerative disorders and tumors), microglia surround dead cells and clear cellular debris from the area. However, most of this work was done in vitro using brain slices, and as the slicing procedure inherently induces some injury, it remained unclear how microglia behave in vivo.

In a technical tour de force, two recent reports by Fritjof Helmchen and colleagues (published online in Science on 14 April) and Wen-Biao Gan (pp 752-758, this issue) describe the imaging of microglia in intact mouse cortex. Both groups took advantage of transgenic mice in which all the microglia were fluorescently labeled and used transcranial two-photon microcsopy to image the behavior of these cells through the thinned skull. Microglial processes were highly dynamic in the intact brain. Although the somata of microglial cells remained morphologically stable over hours, higher-order branches showed rapid extension and retraction over intervals of seconds to minutes. This high resting mobility may enable the microglia to act as vigilant sentries, constantly screening the surrounding parenchyma.

The microglia also responded rapidly to focal brain injury in both studies. Time-lapse imaging showed that after a small laser ablation, microglia near the site of injury responded within the first minute to extend their processes toward the damaged site. Gan and colleagues report that within 30 minutes after the laser-induced injury, the processes of nearby cells reached the damaged site and appeared to fuse together, forming a spherical containment around it and establishing a potential barrier between the healthy and injured tissue, as shown in the photo. Microglia responded similarly to mechanical injury.

What signals mediate this rapid microglial response? In culture, ATP signaling induces microglial migration. Gan and colleagues extend this work to the in vivo situation, and show that extracellular ATP and activation of P2Y receptors on microglia are necessary for the rapid microglial response toward the injury site. Simply inserting an electrode containing ATP allowed the authors to mimic—in time, range and kinetics—the rapid response of microglial processes observed following laser ablation. Furthermore, ATP-induced ATP release was essential for this response; when the authors applied apyrase (which degrades endogenous ATP in the extracellular space), and then released non-hydrolyzable ATP from a microelectrode, they observed no such rapid microglial response. Applying connexin channel inhibitors before laser ablation also inhibited the microglial response toward the laser ablation site. Interestingly, baseline motility of microglial processes in the intact brain seems to be modulated by the same ATP signaling mechanisms that mediate injury-induced responses, because apyrase and connexin channel inhibitors also significantly slowed microglial baseline dynamics.