Dynamics of Synaptical Transmission under in vivo Conditions: A new in vitro Approach

This project is part (C7) of the SFB 505 "Neuronale Differenzierung und Neurotransmission".

The main goal of this project is to characterize the influence of ongoing network activity in the neocortex (1) on the functional properties of synaptic transmission, (2) on dendritic integration and (3) on the firing behaviour of single neurons as well as (4) on the network dynamics as a whole.

To adress our questions we follow two major appraoches. On the one hand, by means of dynamic photostimulation, we create network activity in vitro in acute neocortical slice preparations [4,6] to "simulate" the in vivo situation. On the other hand, we record intracellular activity in vivo from neurons of the prefrontal cortex to statistically describe the membrane fluctuation resulting from ongoing synaptic bombardement and to study its influence on dendritic integration [3].

Further approaches include in vitro experiments in which we study the output statistics of single neurons in relation to a controlled input mediated through somatic current injection [6,8,9], the study of background statistics in large scale neural network models [Diesmann et al., Hehl et al.] and the investigation of activity in hybrid networks that combine in vitro experiments with in virtu simulations [10,11].

Network Activiation In Vitro through Dynamic Photostimulation

Until now the understanding of synaptic transmission, dendritic integration and intrinsic properties of neurons is largely based on multiple intra-and extracellular recordings from somatas and dendrites of single or pairs of neurons in vitro. The acute slice preparation offers several technical advantages over intracellular recordings in vivo such as visibility and accessibility of neurons and dendrites.
    However, because the excitatory afferences are cut off in this preparation, slices of the neocortex lack spontaneous background activity that is constantly present in vivo. Strong evidence from theoretical work [e.g. Diesmann et al., Kuhn et al.] as well as from in vivo recordings [e.g. Arieli, Azouz&Gray] suggests that background activity strongly influences the properties of single neurons as well as the properties of the whole network. It is therefore difficult to draw conclusions from in vitro slice experiments that are also valid in the in vivo situation.
    To at least partially overcome this drawback we designed an experimental setup for dynamic photostimulation which allows us to induce background activity in the slice and thus to perform experiments in a standard in vitro preparation under in vivo like conditions [4]. The stimulation technique uses the photoactivation of caged glutamate by means of a UV laser beam and is free of electrical artefacts.

Optical setup for photostimulation. Two-dimensional sketch of the optical arrangement. The UV-laser beam is focused into the slice in the recording chamber. The beam deflection system allows for an arbitrary positioning of the target spot in the 2 D plane of the slice. The shutter provides for a fast and precisely timed interruption of light irradiation, especially while the beam moves to a new position.

The laser-induced network activity as reflected in the fluctuations of membrane potential in a recorded neuron is assesed by comparison with intracellular recordings in vivo from rat neocortical neurons [7].

Synaptic Integration and Intracellular Recording In Vivo

We record the intracellular membrane potential in vivo from neurons in the prefrontal cortex of anesthesized rats. These neurons show periods of intensive activity characterized by a depolarized and highly variable membrane potential, interspersed by quiescent hyperpolarized states [3]. Using one or more extracellular stimulating electrode we can elicit population EPSPs in the intracellularaly observed neuron. This enables us to study the influence of synaptic bombardement on the integration of compound events by comparing amplitude, dynamics and summation of the population PSPs in the presence and absence of background activity. Furthermore, we can characterize the typical membrane fluctuations in vivo for a comparison to the fluctuations generated in vitro by means of dynamic photostimulation.

Dendritic Summation of Compound PSPs In Vitro

We designed a paradigm for an in vitro experiment that repeats the in vivo experiment described above. We combine 2 extracellular stimulating electrodes with an intracellular recording from a pyramidal neuron in the acute neocortical slice preparation, i.e. under a condition with no background activity. Electric stimulation of fibres evokes multiple synaptic inputs to the intracellular recorded neuron observed as compund EPSP. By varying the time interval between successive stimulation at the two stimulation sites we investigate the properties of integration and compare the results to those obtained in the in vivo situation. Similarly we use the setup for dynamic photostimulation to probe integration upon direct postsynaptic stimulation at multiple dendritic sites.

Spike Generation In Dependence on Input Statistics In Vitro

Using the patch clamp technique and by means of somatic injection of either current or conductance we can provide a cell with controlled synthetic input. This allows us to study the output spike statistics in relation to the input statistics of mimiced excitatory and inhibitory synaptic events. In particular we can study the influence of synchronous events on the spike probability and spike timing [9], and the degree of output irregularity in dependence on the temporal distribution and correlation of input events [7,8]. In hybrid networks we are able to embed the living cell into a network model of hundreds or thousands of neurons which, for the recorded neuron, creates a quasi-realistic background activity mediated through somatic current injection [10].

Martin Nawrot

This project is part of the SFB 505 "Neuronale Differenzierung und Neurotransmission"
and is funded by the German Research Foundation (DFG).

Contributors

Responsible

Colaborations

Prof. E. Stern