Abstracts

Rationale: Many gross anatomical and physiological features of post-traumatic epileptogenesis have been characterized (e.g. mossy fiber sprouting, a latent period of seizure freedom preceding chronic seizures). However, it has been technically infeasible to track epileptogenic changes in synaptic connectivity at the neuronal level. Anatomical synapses can be mapped at a high spatial resolution, but throughput is low and tissue must be fixed, making it difficult to track epileptogenic changes. Furthermore, although maps of anatomical synapses can be created with a high degree of certainty, they do not necessarily predict network function: Other factors (e.g. ion channel densities and posttranslational modifications, neurotransmitter release and binding) also influence neuronal and network activity. Alternatively, functional connectivity has been measured in living tissue by analyzing correlated activity in populations of neurons. In previous work, we have demonstrated that correlation-based functional connectivity changes dynamically with network state. For example, pre-ictal disinhibition unveils scale-free functional network connectivity in seizing brain slices. Thus, functional connectivity does not necessarily directly reflect anatomical connections – rather, it is a level of abstraction for quantifying network output. Methods: Here, we have developed a method for simultaneously eliciting and recording activity, with single-neuron resolution, across a field of view encompassing entire hippocampal slice cultures. We describe a method by which maps of synaptic connectivity are measured by repeatedly optically stimulating individual neurons while simultaneously imaging the evoked network response. To accomplish this, we designed a microscope with a field-of-view (FOV) large enough to capture an entire organotypic hippocampal slice culture. Individual neurons are stimulated by using a digital micromirror device, in the image plane of the excitation light pathway, to pattern blue light onto selected channelrhodopsin-expressing neurons. Implementing this over such a large area requires high N.A., low-magnification optics, a very bright LED, and high-sensitivity variants of channelrhodopsin (CoChr or CheRiff). Results: In preliminary experiments, we have used red-shifted genetically encoded calcium indicators to record activity. Individual neurons were repeatedly stimulated 50 times, while the network response was imaged. Stimulus-triggered mean images of evoked activity were used to identify the strongest synaptic connections. We identified two primary limitations of this method: off-target stimulation of nearby axons and poor sensitivity to sub-threshold evoked responses. However, advances in somatically targeted channelrhodopsins and voltage-sensitive fluorophores are rapidly increasing the feasibility of single neuron activation and imaging of sub-threshold and inhibitory post-synaptic responses. Conclusions: We have developed a system for all-optical mapping of synaptic connectivity by simultaneously manipulating and recording activity with single-neuron resolution. We demonstrate this technique in a brain slice model of post-traumatic epileptogenesis, but anticipate that it will be broadly useful for understanding how synaptic connectivity architecture relates to the network output of neural circuits in vitro and in vivo. Funding: This work was supported by NIH Grants 2R01NS034700, 5R01NS086364, 5R37NS077908.