Probing synaptic connectivity and function using high-density microelectrode arrays and whole-cell patch-clamp recordings

Abstract

Event Abstract Back to Event Probing synaptic connectivity and function using high-density microelectrode arrays and whole-cell patch-clamp recordings Julian Bartram1*, Manuel Schröter1, Silvia Ronchi1, Vishalini Emmenegger1, Jan Müller1 and Andreas R. Hierlemann1 1 ETH Zürich, Department of Biosystems Science and Engineering, Switzerland Motivation Synaptic efficacy and synapse number of monosynaptic connections between neurons are often regulated by the spiking activity of the respective pre- and postsynaptic cell. Progress towards a better understanding of the rules and mechanisms that underlie such modifications has been limited due to the difficulties associated with simultaneously studying plasticity at multiple synaptic inputs. Here, we provide a solution to this problem by combining cutting-edge high-density microelectrode array (HD-MEA) technology with the patch-clamp technique. While the latter allows for accurate measurement of postsynaptic currents or potentials, evoked by individual synaptic activation, the HD-MEA technology provides large-scale information about unit activity and allows for selective stimulation of neurons, including multiple presynaptic cells. The proposed approach has been applied to comprehensively examine forms of homeostatic plasticity – a collection of crucial processes acting at different temporal scales in order to stabilize neuronal firing rates. We report on a characterization of classic synaptic scaling operating in mature cortical networks and propose a novel model for the study of homeostatic plasticity during natural network states. Material and Methods In this study, we investigated homeostatic plasticity in cultures of primary cortical neurons. Many variants of this form of plasticity were originally identified in reduced preparations, but have since been confirmed in vivo [1,2], which supports the choice of such a cell culture model system. Isolation, plating on MEA chips, and treatment of primary cortical neurons were previously described [3]. A complementary-metal-oxide-semiconductor- (CMOS) based HD-MEA system with 26’400 microelectrodes, an electrode pitch of 17.5 μm and 1024 reconfigurable readout channels [4] was employed for all extracellular recording and stimulation experiments. Patch-clamp recordings in current-clamp or voltage-clamp mode were conducted in conjunction with HD-MEA stimulation and recordings or, in some cases, independently. Homeostatic plasticity was initially induced by pharmacological alteration of global network spiking starting at 14 - 21 days in vitro. Effects on network spiking, induced in this way, were investigated by performing array-wide activity scans. Some of these results were further complemented with miniature excitatory postsynaptic current (mEPSC) recordings, which provide more direct information about synaptic efficacies. Selective stimulation of cells with the HD-MEA during whole-cell patch-clamp recordings enabled the assessment of synaptic strength at multiple inputs. Finally, recording of the spiking activity of all identified presynaptic cells and the postsynaptic neuron during the induction of homeostatic plasticity will permit the examination of correlations between spike patterns and changes in synaptic strength. Results Array-wide network activity was assessed before, during and after the induction of homeostatic plasticity with 12 – 48 h of tetrodotoxin (TTX; 1 μM) treatment. The relatively limited cell migration in mature cultures facilitated the tracking of units over the multiday recording period. Clear increases in network activity could be observed following the washing out of TTX relative to controls. mEPSC recordings from before and after TTX treatment suggest that this increase was due to homeostatic adjustments of excitatory synaptic efficacy. Interestingly, while the network remained in an elevated activity state for several hours following TTX withdrawal, spiking subsequently decreased within a 24 h period. This likely reflected a homeostatic down-regulation of synaptic strength due to the high network spiking activity. Finally, using the electrical stimulation capabilities of the HD-MEA system, we have successfully measured the strength of multiple synaptic inputs. Discussion We demonstrated in this study how the combination of HD-MEA technology and the patch-clamp technique provides unique abilities to study synaptic homeostasis, and synaptic plasticity in general, by providing crucial information about spontaneous global network spiking, exquisite control over spiking of individual cells, and the ability to record subthreshold postsynaptic responses. A homeostatic down-regulation of network activity and (excitatory) synaptic efficacy, following the elevation of network spiking activity by TTX treatment, is a very attractive form of homeostatic plasticity, as no artificial stimulus or agent (e. g. TTX) is present during the induction phase. Moreover, with global spiking activity being consistent with synaptic efficacies, the network is in a more natural state. We are now attempting to link the observed changes in synaptic efficacy with the respective pre- and postsynaptic spike patterns, recorded during the induction phase, promising new insights into the rules and mechanisms underlying homeostatic plasticity. Acknowledgements This work was supported by the ERC Advanced Grant 694829 ‘neuroXscales’.