Coupling Cortical Neurons through Electronic Memristive Synapse - Aix-Marseille Université Access content directly
Journal Articles Advanced Materials Technologies Year : 2018

Coupling Cortical Neurons through Electronic Memristive Synapse

Azat Nasretdinov
  • Function : Author
Silvia Battistoni
  • Function : Author
Tatiana Berzina
  • Function : Author
  • PersonId : 917990
Salvatore Iannotta
  • Function : Author
Victor Erokhin
  • Function : Author
  • PersonId : 917989


of electrophysiological spike-sorting [4] and optogenetic [5] approaches enables an efficient readout and control over activity of single or groups of neurons leading to a development of prosthetic devices. For example, fMRI-guided electrophysiological recordings of single motor cortex neurons involved in specific motor tasks combined with muscle electrostimulation enabled to produce prosthesis with a remarkable alleviation of the neurological deficits in hemiplegic patients with a traumatic lesion of synaptic connections between the corticospinal neurons and motor neurons in spinal cord. [6] Restoration of synaptic connections as in the case of traumatic injury above as well as in other patholo-gies associated with a synaptic loss of function and various synaptopathies could be also solved through an introduction of electronic synapses to connect neurons directly, given that these artificial synapses recapitulate the main feature of natural synapses including their plasticity. Moreover, development of electronic synapses with unprecedented, due to biological restraints, features during evolution could result in creation of cyborgs with unprecedented capacities. We used patch-clamp recordings from nonconnected pairs of cortical layer 5 pyramidal neurons in rat brain slices (Figure 1a). Action potentials (APs) evoked by suprathreshold depolarizing current injection in either neuron failed to evoke any response in another cell in the pair (Figure 1c), indicating that these cells were not connected by natural synapses in either direction. These neurons were then connected through an electronic circuit with an OMD, playing the role of a synapse analog (Figure 1b). The structure of the OMD included a conducting polymer-polyaniline (PANI), with a solid electrolyte-lithium salt doped polyethylene oxide (PEO), and its memristive features were based on the high difference in PANI conductivity in the oxidized and reduced forms. [7,8] The ratio of PANI conductivity in oxidized and reduced forms is in the order of 10 8. [8] However, the need to have a medium for these reactions reduces the maximum reported ratio for the entire device till about 10 5. [9] After setting the OMD resistance initially at high values by negative voltage loading (Figure 1d, bottom), APs in a "presynaptic" Cell 1 (Figure 1d, plot 2) were induced by a suprathreshold depolarizing steps (Figure 1d, plot 1). However, these APs in Cell 1 evoked only a subthreshold depolarizing response in the "postsynaptic" Cell 2 (Figure 1d, plot 4, the first Functional coupling live neurons through artificial synapses is the primary requirement for their implementation as prosthetic devices or in building hybrid networks. Here, the first evidence of unidirectional, activity dependent, coupling of two live neurons in brain slices via organic memristive devices (OMD) is demonstrated. ODM is a polymeric electrochemical element, which has two terminals for the connection in electrical circuits and which displays hysteresis and rectifying features. OMD coupling is characterized by nonlinear relationships determined by the instantaneous values of OMD resistance that can be controlled by the neuronal activity, and the excitation threshold in the postsynaptic neuron. OMD coupling also has the spike-timing features similar to that of the natural excitatory synapses. Also, OMD-synapses support synchronized delta-oscillations in the two-neuron network. It is proposed that OMD-synapses may enable realization of prosthetic synapses and building hybrid neuronal networks endowed with a capacity of learning, memory, and computation. A synapse is a biological structure, which connects two neu-rons enabling specific and unidirectional information flow (excitation or inhibition) from one neuron to another. Synaptic connections are the key elements of the neuronal networks and their plasticity underlies learning and memory. Recent progress in building artificial neuronal networks is largely based on the elements mimicking features of natural synapses in silico or in electrico. [1-3] Hybrid networks, in which brain-computer systems read and control the activity of live cells also require interphase devices with cellular resolution. Use
Fichier principal
Vignette du fichier
Juzekaeva_et_al-2018-Advanced_Materials_Technologies.pdf (3.39 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01963794 , version 1 (21-12-2018)



Elvira Juzekaeva, Azat Nasretdinov, Silvia Battistoni, Tatiana Berzina, Salvatore Iannotta, et al.. Coupling Cortical Neurons through Electronic Memristive Synapse. Advanced Materials Technologies, 2018, pp.1800350. ⟨10.1002/admt.201800350⟩. ⟨hal-01963794⟩


38 View
237 Download



Gmail Facebook Twitter LinkedIn More