A novel method to express Channelrhodopsin2-variants at active synapses

Increasing evidence points to the importance of dendritic spines in the formation and allocation of memories [1]. Alterations that affect spines number and physiology are associated to memory and cognitive disorders. Plasticity mechanisms are believed to be directly involved in the establishment of engrams, with potentiated synapses playing a major role in the building and retention of memories. Therefore, the intervention on those synapses is expected to affect memory processing and recall; however, a way to directly test this hypothesis is currently lagging behind due to technical and biological reasons. In fact, one should be able to directly stimulate (or inhibit) selectively only those synapses that were active during the establishment of the memory trace in order to evaluate their role on later memory recall. Whereas much progress has been made on neural circuits by means of optogenetic stimulation and/or inhibition of whole nodes (neurons), the link between memory allocation and the synapses onto which the triggering stimuli converge is still largely missing [2]. Here we describe a novel approach to express opsins at synapses in an activity-dependent manner by means of RNA trafficking sequences that regulate mRNA transport and translation [3]. We compare different constructs bearing (i) proteic localization elements that enrich Channelrhodopsin at synapses, (ii) RNA elements that regulate Channelrhodospin expression within dendrites and (iii) a combination of the two. We evaluate their performance in terms of synaptic enrichment and ability to tag synapses in an activity-dependent manner. We demonstrate that the combination of the two (construct (iii)) confers superior synapse tagging and labelling with little residual expression on dendrites in primary cortical and hippocampal neurons by means of high-resolution confocal microscopy (Figure 1). We confirm Channelrhodospin co-localization at post-synaptic densities with synaptic markers PSD95 and Homer1c. Moreover, tight regulation of Channelrhodospin expression allows tagging of active synapses: treatments that involve neuron activation or synapse potentiation determine an increase of its expression at synapses; conversely, reduction of neural activity exerts the opposite effect. Furthermore, we also compare multiple RNA regulatory sequences derived from different dendritic mRNAs involved in synaptic potentiation, looking at their localization and response to neuronal status. In conclusion, we present here an innovative method to drive Channelrhodopsin expression at neuronal synapses in an activity-dependent way, establishing a new paradigm for “synaptic optogenetics”. The method introduced here will find future use in the study of memory circuits and in the inquiry of the role of synaptic potentiation in learning and memory.


[1] T. Rogerson, D. J. Cai, A. Frank, Y. Sano, J. Shobe, M. F. Lopez-Aranda, A. J. Silva, Nat. Rev. Neurosci. 15 (2014) 157–169.

[2] X. Liu, S. Ramirez, P. T. Pang, C. B. Puryear, A. Govindarajan, K. Deisseroth, S. Tonegawa, Nature 484 (2012) 381–385.

[3] S. Kindler, H. Wang, D. Richter, H. Tiedge, Ann. Rev. Cell Dev. Biol. 21 (2005) 223.