The mammalian brain is among the most complex structures in nature. Still we are lacking an understanding even of fundamental principles of its operation. In particular, it is not clear how higher functions of the CNS, such as sensory feature discrimination or fine-tuned motor coordination, arise from the underlying cellular activities. Understanding how single cells and local neural networks compute will be a prerequisite of understanding information processing in the brain. Hence, key questions that we are addressing are:

• How is sensory information processed on the cellular level and the level of neural networks? How do individual cells integrate the synaptic input they receive? What is the functional organization of local neural networks?
• How are cellular and network properties modulated? How do they depend on attention, motivation, and behavioral state? What is the role of glial cells for information processing in the neocortex?
• How can memories be formed and maintained? What are the underlying cellular or network changes?
• How do neural circuits constantly adapts to the changing environment? How do they 'learn' and how do they reorganize following brain injuries? 

In our research group we employ modern neuroscientific techniques to record from cells and networks under in vivo conditions. By combining electrophysiology and fluorescence microscopy we are able to study electrical signaling, biochemical signaling (e.g. of intracellular calcium), as well as structural dynamics. In particular we apply 2-photon microscopy, which permits high-resolution fluorescence imaging several hundred micrometer below the tissue surface.

In vivo 2-photon imaging of neural activity

Two-photon laser-scanning microscopy has become the method of choice for imaging on the cellular level in the intact brain. The particular advantage of 2-photon microscopy is that it provides a large depth penetration due to its low sensitivity to scattering. For example, individual cells with their dendrites and dendritic spines can be resolved relatively deep (up to 600 µm) in the intact neocortex of anesthetized animals (Fig. 1, left). Various labeling techniques now permit fluorescence staining not only of individual cells (e.g. via intracellular filling through a patch pipette), but also of specific subtypes of cells (e.g. in transgenic animals expressing fluorescent proteins) or of large populations of cells (by unspecific loading with membrane-permeable calcium indicators; Fig. 1, middle). On all levels, from the synapse to the population , structural as well as functional dynamics can be studied. In particular, we use in vivo calcium imaging to reveal single-cell and population activity patterns (Fig. 1, right).

3D imaging of neural network activity

A major limitation of imaging cell dynamics so far has been that most techniques are restrcited to two-dimensional imaging. Recently, we have developed a novel method to enable imaging of large populations of cells with relatively high temporal resolution (Göbel et al., Nat Methods, 2006). The basic idea is to employ laser-scanning along a 3D trajectory to sample fluorescence signals from as many cell bodies as possible (Fig. 2). We are now able to record in vivo from several hundred cells with 10 Hz resolution. This novel technology opens a new field for the study of neural network dynamics on the scale of microcircuits, e.g. within single cortical columns.

Imaging of glial cells

Besides our focus on neuronal activity, we have become interested in the role of glial cells in neocortex. We have found a specific marker of astrocytes in vivo, which facilitates functional measurements from the identified astroglial network (Nimmerjahn et al., Nat Methods 2004). Moreover, we have found that microglial cells are highly dynamic structures that perpetually survey the surrounding tissue (Fig. 3; Nimmerjahn et al., Science 2005). This steady-state surveillance immediately switches to a focal, highly directed activation of microglial processes following local damage such as the disruption of a blood capillary. The ability to measure glial cell dynamics opens new avenues to investigate the interaction between neuronal and glial networks and the role of glial cells in various models of brain diseases.