Brain-inspired computing paradigms have led to substantial advances in the automation of visual and linguistic tasks by emulating the distributed information processing of biological systems1. The similarity between artificial neural networks (ANNs) and biological systems has inspired ANN implementation in biomedical interfaces including prosthetics2 and brain-machine interfaces3. While promising, these implementations rely on software to run ANN algorithms. Ultimately, it is desirable to build hardware ANNs4,5 that can both directly interface with living tissue and adapt based on biofeedback6,7. The first essential step towards biologically integrated neuromorphic systems is to achieve synaptic conditioning based on biochemical signalling activity. Here, we directly couple an organic neuromorphic device with dopaminergic cells to constitute a biohybrid synapse with neurotransmitter-mediated synaptic plasticity. By mimicking the dopamine recycling machinery of the synaptic cleft, we demonstrate both long-term conditioning and recovery of the synaptic weight, paving the way towards combining artificial neuromorphic systems with biological neural networks.
Subscribe to Journal
Get full journal access for 1 year
only $16.58 per issue
All prices are NET prices.
VAT will be added later in the checkout.
Rent or Buy article
Get time limited or full article access on ReadCube.
All prices are NET prices.
Furber, S. Large-scale neuromorphic computing systems. J. Neural Eng. 13, 051001 (2016).
Grahn, P. J. et al. Restoration of motor function following spinal cord injury via optimal control of intraspinal microstimulation: toward a next generation closed-loop neural prosthesis. Front. Neurosci. 8, 1–12 (2014).
Bonifazi, P. et al. In vitro large-scale experimental and theoretical studies for the realization of bi-directional brain-prosthese. Front. Neural Circuits 7, 1–19 (2013).
van Doremaele, E. R. W., Gkoupidenis, P. & van de Burgt, Y. Towards organic neuromorphic devices for adaptive sensing and novel computing paradigms in bioelectronics. J. Mater. Chem. C https://doi.org/10.1039/C9TC03247A (2019).
Fuller, E. J. et al. Parallel programming of an ionic floating-gate memory array for scalable neuromorphic computing. Science 364, 570–574 (2019).
Berco, D. & Shenp Ang, D. Recent progress in synaptic devices paving the way toward an artificial cogni‐retina for bionic and machine vision. Adv. Intell. Syst. 1, 1900012 (2019).
Hebb, D. O. The Organization of Behavior: a Neuropsychological Theory (Wiley, 1949).
Vassanelli, S. & Mahmud, M. Trends and challenges in neuroengineering: toward ‘intelligent’ neuroprostheses through brain-‘brain inspired systems’ communication. Front. Neurosci. 10, 438 (2016).
Chiolerio, A., Chiappalone, M., Ariano, P. & Bocchini, S. Coupling resistive switching devices with neurons: state of the art and perspectives. Front. Neurosci. 11, 70 (2017).
Rivnay, J. et al. Organic electrochemical transistors. Nat. Rev. Mater. 3, 17086 (2018).
van de Burgt, Y. et al. A non-volatile organic electrochemical device as a low-voltage artificial synapse for neuromorphic computing. Nat. Mater. 16, 414–418 (2017).
van de Burgt, Y., Melianas, A., Keene, S. T., Malliaras, G. & Salleo, A. Organic electronics for neuromorphic computing. Nat. Electron. 1, 386–397 (2018).
Gkoupidenis, P., Koutsouras, D. A. & Malliaras, G. G. Neuromorphic device architectures with global connectivity through electrolyte gating. Nat. Commun. 8, 15448 (2017).
Asplund, M. et al. Toxicity evaluation of PEDOT/biomolecular composites intended for neural communication electrodes. Biomed. Mater. 4, 045009 (2009).
Santoro, F., van de Burgt, Y., Keene, S. T., Cui, B. & Salleo, A. Enhanced cell-chip coupling by rapid femtosecond laser patterning of soft PEDOT:PSS biointerfaces. ACS Appl. Mater. Interfaces 9, 39116–39121 (2017).
Gualandi, I. et al. Selective detection of dopamine with an all PEDOT:PSS organic electrochemical transistor. Sci. Rep. 6, 2–5 (2016).
Obaid, A. M. et al. Massively parallel microwire arrays integrated with CMOS chips for neural recording. Sci. Adv. 6, 573295 (2020).
Khodagholy, D. et al. NeuroGrid: recording action potentials from the surface of the brain. Nat. Neurosci. 18, 310–315 (2015).
Hess, L. H. et al. Electrical coupling between cells and graphene transistors. Small 11, 1703–1710 (2015).
Juzekaeva, E. et al. Coupling cortical neurons through electronic memristive synapse. Adv. Mater. Technol. 4, 4–9 (2019).
Serb, A. et al. Memristive synapses connect brain and silicon spiking neurons. Sci. Rep. 10, 2590 (2020).
Gupta, I. et al. Sub 100 nW volatile nano-metal-oxide memristor as synaptic-like encoder of neuronal spikes. IEEE Trans. Biomed. Circuits Syst. 12, 351–359 (2018).
Buccelli, S. et al. A neuromorphic prosthesis to restore communication in neuronal networks. iScience 19, 402–414 (2019).
Burns, M. E. & Augustine, G. J. Synaptic structure and function: dynamic organization yields architectural precision. Cell 83, 187–194 (1995).
Calabresi, P., Picconi, B., Tozzi, A. & di Filippo, M. Dopamine-mediated regulation of corticostriatal synaptic plasticity. Trends Neurosci. 30, 211–219 (2007).
Gkoupidenis, P., Schaefer, N., Strakosas, X., Fairfield, J. A. & Malliaras, G. G. Synaptic plasticity functions in an organic electrochemical transistor. Cit. Appl. Phys. Lett. 107, 263302 (2015).
Li, Y. et al. Identification of two functionally distinct endosomal recycling pathways for dopamine D2 receptor. J. Neurosci. 32, 7178–7190 (2012).
Shahrokhian, S. & Bozorgzadeh, S. Electrochemical oxidation of dopamine in the presence of sulfhydryl compounds: application to the square-wave voltammetric detection of penicillamine and cysteine. Electrochim. Acta 51, 4271–4276 (2006).
Keene, S. T. et al. Optimized pulsed write schemes improve linearity and write speed for low-power organic neuromorphic devices. J. Phys. D Appl. Phys. 51, 224002 (2018).
Yakushenko, A., Kätelhön, E. & Wolfrum, B. Parallel on-chip analysis of single vesicle neurotransmitter release. Anal. Chem. 85, 5483–5490 (2013).
Li, X. et al. A nanostructure platform for live-cell manipulation of membrane curvature. Nat. Protoc. 14, 1772–1802 (2019).
Santoro, F. et al. Revealing the cell-material interface with nanometer resolution by focused ion beam/scanning electron microscopy. ACS Nano 11, 8320–8328 (2017).
Keene, S. T., Melianas, A., van de Burgt, Y. & Salleo, A. Mechanisms for enhanced state retention and stability in redox-gated organic neuromorphic devices. Adv. Electron. Mater. 5, 1800686 (2018).
Isaksson, J. et al. Electronic control of Ca2+ signalling in neuronal cells using an organic electronic ion pump. Nat. Mater. 6, 673–679 (2007).
A.S. and S.T.K. acknowledge financial support from the National Science Foundation and the Semiconductor Research Corporation, E2CDA Award no. 1739795. Additionally, S.T.K. thanks the Stanford Graduate Fellowship fund for support. A.M. gratefully acknowledges support from the Knut and Alice Wallenberg Foundation (KAW 2016.0494) for postdoctoral research at Stanford University. This work was in part performed at the Stanford Nano Shared Facilities (SNSF) and the nano@Stanford (SNF) labs, which are supported by the National Science Foundation as part of the National Nanotechnology Coordinated Infrastructure under award ECCS‐1542152. Y.v.d.B. gratefully acknowledges funding from the European Union’s Horizon 2020 Research and Innovation Programme, grant agreement no. 802615. F.S. thanks the staff of the Cleanroom Facility at the Center Lab of Istituto Italiano di Tecnologia for the use of the dual-beam machine, and the group of A. Offenhäusser at the Institute of Complex Systems (ICS−8) of Jülich Forschungszentrum for providing the PC-12 cell line.
The authors declare no competing interests.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended Data Fig. 1 Effect of dopamine concentration in cell culture media on neuromorphic device transfer characteristics.
a, Transfer curves and b, corresponding transconductance curves as a function of dopamine concentration in DMEM solution flowed through the microfluidic channel showing an increase in peak transconductance at roughly +0.2 V corresponding to oxidation of dopamine. Source data
CVs (scan rate = 10 mV·s-1) show the oxidation of dopamine with a peak of ca. +0.1 V vs a saturated silver/silver chloride (sat’d Ag/AgCl) electrode for low dopamine concentrations (20–200 mM). At higher concentrations (0.5–1 mM), dopamine oxidation peak shifts and a secondary oxidation reaction is observed. The lack of a reduction peak in the reverse scan shows that the oxidation reaction is irreversible. Source data
a, When dopamine (DA, pink circles) is oxidized to dopamine o-quinone (DQ, green circles), the oxidation products (2e-, 2 H+) can compensate the electronic and ionic charges in doped PEDOT:PSS, thereby de-doping the channel and gate (we note that other cations in solution such as Na+ or K+ may compensate PSS- instead of or in addition to H+). b, This reaction at the gate electrode (1) changes its potential, resulting in effective gating of the PEDOT:PSS postsynaptic channel, and results in a transfer of an electron and a proton to the postsynaptic channel (2) to maintain to a potential drop of Vpost. c, To test this hypothesis, a device structure as shown in b, is fabricated and the conductance of both the postsynaptic electrode and postsynaptic channel are measured before and after oxidation driven at the postsynaptic electrode. The conductance of both the postsynaptic channel and electrode decrease, confirming the transfer of protons (or cations) and electrons to both PEDOT:PSS electrodes. Source data
a, Schematic of the experimental setup for monitoring the potential of the PEDOT:PSS channel during device operation using a potentiostat operated in open circuit voltage mode. The channel is connected to a grounded working electrode (WE) and the potential is monitored in reference to a saturated silver/silver chloride (Ag/AgCl) reference electrode (RE) immersed in the electrolyte. b, The channel conductance (GCh), potential (EWE) and gate voltage pulses (VG) over time showing no change in GCh or EWE in the absence of dopamine (DA) in response to VG pulses. After DA is added to solution, GCh and EWE both decrease in response to VG pulses, showing the change in GCh (ΔGCh) is due to the potential change in the channel following oxidation (ΔEWE). At t = 275 s, we exchange the solution with fresh PBS containing no DA to show that both ΔEWE and ΔGCh are properties of the channel, not the solution. Source data
Extended Data Fig. 5 Dopamine immunohistochemistry and biocompatibility for PC-12 cells on PEDOT:PSS.
The figure reports a comparison between PC-12 cells seeded on PEDOT:PSS film a, and bare glass b. In both cases the green fluorescence confirmed the presence of dopamine released locally at the cell membrane (in red), showing that dopamine is produced in vesicles in the PC-12 cells and then released. c, The viability of PC-12 cells was tested by staining live cells with calcein-AM (green) and dead cells with propidium iodide (red) for PEDOT:PSS electrodes coated with d, collagen IV and e, poly-L-lysine films. Both substrates exhibited good cell viability, with the collagen-IV coated film showing higher biocompatibility. For this reason, the collagen-IV coating was employed in all of the following experiments. f, Live/dead staining performed on PC-12 cells inside the microfluidic channel after performing repeated pulsed electrical measurements, showing no difference between cells cultured on PEDOT:PSS e, prior to and f, following electrical measurements, indicating no alteration of cell viability from the electrical measurements. Source data
Comparison of pulsed measurements performed on the biohybrid synapse after 2 hours (light blue line) and after 24 hours (blue line) following cell plating. The efficiency of the device remains unaltered since synaptic plasticity behavior from dopamine oxidation is retained following extended exposure to the cell culture media. Source data
Steady-state measurements of the postsynaptic channel conductance (Vpost = +0.3 V, Vch = -0.2 V) under varied flow rates and dopamine concentrations. When the flow rate is increased from 200 μL min-1 to 300 μL min-1, the equilibrium Gpost is larger for the same dopamine concentration. Additionally, the recovery time for the higher flow rate is shorter due to the fact that the fresh solution carries oxygen which can oxidize the PEDOT:PSS postsynaptic channel at faster rates when the solution flow rate is increased. Source data
Extended Data Fig. 8 Calibration of the organic neuromorphic device with increasing dopamine concentration.
a, Pulsed measurements of the ex vivo neuromorphic device under microfluidic flow of DMEM solution at varied dopamine concentrations showing the postsynaptic conductance change following voltage pulses with increasing dopamine concentration. b, Calibration curve showing the conductance change during pulsing (left, dark blue) and peak transconductance during transfer measurements (right, light blue) as a function of increasing dopamine concentration. c, Postsynaptic channel conductance update as a function of gate voltage pulse width for the ex vivo neuromorphic device with varied dopamine concentrations showing a nearly linear time dependence for all concentrations. The minimum pulse width resulting in a change in postsynaptic current (LTP) also depends on the dopamine concentration; at 0.02 mM the pulse width must be >100 ms to elicit a response, whereas at 0.2 mM a pulse width of 10 ms is sufficient to cause LTP. Source data
As the neuromorphic device electrode area is decreased, the conductance change per pulse (Vpulse = 0.3 V, tpulse = 2 s) increases proportionally. Measurements are performed in the absence of microfluidic flow. Source data
Short-term and long-term modulation of the postsynaptic conductance under dopamine flow
Response of the neuromorphic device to dopamine released by PC-12 cells
Long-term potentiation of the artificial postsynaptic neuron
Effect of dopamine concentration on transfer characteristics
Cyclic voltammetry of dopamine on PEDOT:PSS
Conductance modulation due to dopamine oxidation
Potential change of PEDOT:PSS channel following dopamine oxidation
Viability of PC-12 cells on PEDOT:PSS
Long-term device stability
Steady-state measurements of channel conductance with varying flow rates and dopamine concentrations
Response calibration of the organic neuromorphic devices to varying dopamine concentration
Dopamine response of scaled neuromorphic devices
About this article
Cite this article
Keene, S.T., Lubrano, C., Kazemzadeh, S. et al. A biohybrid synapse with neurotransmitter-mediated plasticity. Nat. Mater. (2020). https://doi.org/10.1038/s41563-020-0703-y
MRS Communications (2020)