3D engineered neuronal networks coupled to 3D-MEAs: a new experimental model for in-vitro electrophysiology

Abstract

Event Abstract Back to Event 3D engineered neuronal networks coupled to 3D-MEAs: a new experimental model for in-vitro electrophysiology Nicolò Colistra1, MariaTeresa Tedesco1, Paolo Massobrio1 and Sergio Martinoia1, 2* 1 Università di Genova, Department of Informatics, Bioengineering, Robotics and System Engineering (DIBRIS), University of Genova, Via All’Opera Pia 13, 16145, Italy 2 Istituto di biofisica (IBF), Italy Background and aims In the last years, some studies have been devoted to the introduction of three-dimensional (3D) engineered in-vitro neuronal systems1,2. 3D neuronal networks coupled to Micro-Electrode Arrays (MEAs), represent a powerful in-vitro model capable of better emulating in-vivo neurophysiology than 2D cultures. However, the existing proposed 3D in-vitro neuronal models lack of appropriate microtransducers that are not able to record the electrophysiological activity of the 3D networks at different levels of the structure. Up to now, only planar devices have been used to record the electrophysiological activity only from the bottom layer (i.e. the one directly coupled to the planar MEAs). Thus, to further characterize and optimize such 3D neuronal network systems and to study how the 3D neuronal network activity can change in the different layers of the structure, new recording devices should be developed. In this work, we show preliminary results of 3D engineered hippocampal neuronal networks coupled to innovative three-dimensional MEAs (i.e. devices constituted by 3D gold ball-shaped microelectrodes positioned at different heights). We characterized first the electrical properties of such devices (i.e., impedance) and then we used the 3D-MEA to record the spontaneous electrophysiological activity of 3D neuronal networks. Material and Methods 3D cell culture. Hippocampal neurons were dissociated from rats E18 Sprague Dawley. MEAs supports and glass microbeads (40 µm in diameter) were sterilized and pre-coated with adhesion proteins (poly-lysine and laminin). The hippocampi were dissociated in a trypsin enzymatic solution and the cells were kept in a Neurobasal medium + B27 + Glutamax 1% and 1% Pen-Streptomycin. Microbeads were placed onto a porous membrane (Transwell®, Millipore) where they self-assembled forming a compact monolayer. We then cultured dissociated hippocampal neurons on such coated microbeads in order to obtain a cell density of about 1500 cells/mm2. The assembly of neurons with microbeads was then moved from the membrane and deposited onto the 2D neuronal network previously coupled to the 3D-MEA device to obtain a packed 3D assembly. 3D-MEAs. The three-dimensional Micro-Electrode Arrays (3D-MEAs) used in this work are composed of 60 gold ball-shaped microstructures positioned at different heights (diameter of a single ball = 50 µm). A PMMA (Poly(methyl methacrylate)) thin-film was used as passivation layer to isolate the microstructures sidewalls from the cells contact. The only conducting part of the device exposed to the direct contact of cells is the top side of such ball-shaped microelectrodes. Thus, recordings from in-vitro neuronal cell cultures were performed only from the tip of the ball-shaped microstructures. Figure 1 shows the 3D-MEA layout and the correspondences electrode-level of the 3D neuronal structure. Electrical characterization of 3D-MEAs. In order to estimate electrical properties of such 3D-MEAs, an electrochemical impedance spectroscopy (EIS) based-experimental set-up was employed to carry out impedance measurements of electrodes. Measures were performed in a PBS (Phosphate-burrefed saline) solution with an Ag/AgCl reference electrode. The process was repeated for ten selected microelectrodes and by sweeping the frequency from 10 Hz to 10 kHz to compute impedance at each frequency value. Data and Statistical analysis. Data analysis was performed off-line by using the custom software package SPYCODE3 developed in MATLAB. We characterized the spontaneous activity by means of first-order statistics like mean firing rate and mean bursting rate. Data were expressed as mean ± standard error of the mean. Statistical analysis was performed using MATLAB. Since data do not follow a normal distribution, we performed a non-parametric Mann-Whitney U-test. MEA set-up. Spontaneous electrophysiological activity of 3D hippocampal neuronal networks (19DIV) was recorded in all the performed experiments by means of 3D-MEAs, described before, and fabricated at the Bruno Kessler Foundation (FBK, Trento, Italy). Results We characterized the electrical properties of such 3D-MEAs to determine the impedance of electrodes. We observed that the impedance decreases as the frequency of the AC input signal increases. The impedance values of 3D-MEAs ranges from a few mega ohms at low frequency (10 Hz) to tens of kilo ohms at higher frequency (10 kHz). The obtained impedance values are in line with the literature4. Then, we analysed the network dynamics of 3D hippocampal neuronal networks coupled to 3D-MEAs. Figure 2 shows 5 min of spontaneous electrophysiological activity of a representative experiment at DIV 19. Each colour corresponds to a different layer (n=4) of the 3D structure. It can be noticed that the signature of the network dynamics presents a wide repertoire of activities characterized of both synchronous bursts and random spikes at the single channel-neuron level. We also evaluated the mean firing and bursting rate in the performed experiments (n=2) at DIV 19. We observed a higher level of firing rate in correspondence of the planar (2.3 ± 1 spikes/s) and second layer (2.2 ± 0.8 spikes/s) of the 3D structure than the first (1.2 ± 0.5 spikes/s) and third one (0.8 ± 0.2 spikes/s). Moreover, a higher frequency of bursts in correspondence of the second level (5 ± 2 bursts/min) than the other ones was observed. Conclusion The new developed experimental platform constituted by 3D engineered neuronal networks coupled to 3D-MEAs represents a complete 3D-3D in-vitro neuronal system. This model allows to record the electrophysiological activity of the 3D neuronal network from different locations in the 3D space and to study how the network dynamics changes in different layers of the 3D structure. Figure 1 Figure 2 Acknowledgements The authors thank Dr. Leandro Lorenzelli (Fondazione Bruno Kessler (FBK), Microsystems Technology (MST) Research Unit, Center for Materials and Microsystems, Trento) for providing 3D-MEAs chips.