Search

In the last decades, planar multi-electrode arrays (MEAs) have been widely used to record activity from in vitro neuronal cell cultures and tissue slices. Though successful, this technique bears some limitations, particularly relevant when applied to three-dimensional (3D) tissue, such as brain slices, spheroids or organoids. For example, planar MEAs signals are informative on just one side of a 3D-organized structure. This limits the interpretation of the results in terms of network functions in a complex structured and hyperconnected brain tissue. Moreover, the side in contact with the MEAs often shows lower oxygenation rates and related vitality issues. To overcome these problems, we empowered a CMOS high-density multi-electrode array (HD-MEA) with thousands of microneedles (μneedles) of 65-90 μm height, able to penetrate and record in-tissue signals, providing for the first time a 3D HD-MEA chip. We propose a CMOS-compatible fabrication process to produce arrays of μneedles of different widths mounted on large pedestals to create microchannels underneath the tissue. By using cerebellar and cortico-hippocampal slices as a model, we show that the μneedles efficiently penetrate the 3D tissue while the microchannels allow the flowing of maintenance solutions to increase tissue vitality in the recording sites. These improvements are reflected by the increase in electrodes sensing capabilities, the number of sampled neuronal units (compared to matched planar technology), and the efficiency of compound effects. Importantly, each electrode can also be used to stimulate the tissue with optimal efficiency due to the 3D structure. Furthermore, we demonstrate how the 3D HD-MEA can efficiently penetrate and get outstanding signals from in vitro 3D cellular models as brain spheroids. In conclusion, we describe a new recording device characterized by the highest spatio-temporal resolution reported for a 3D MEA and significant improvements in the quality of recordings, with a high signal-to-noise ratio and improved tissue vitality. The applications of this game-changing technique are countless, opening unprecedented possibilities in the neuroscience field and beyond.

The retina is a complex circuit of the central nervous system whose aim is to encode visual stimuli prior the higher order processing performed in the visual cortex. Due to the importance of its role, modeling the retina to advance in interpreting its spiking activity output is a well studied problem. In particular, it has been shown that latent variable models can be used to model the joint distribution of Retinal Ganglion Cells (RGCs). In this work, we validate the applicability of Restricted Boltzmann Machines to model the spiking activity responses of a large a population of RGCs recorded with high-resolution electrode arrays. In particular, we show that latent variables can encode modes in the RGC activity distribution that are closely related to the visual stimuli. In contrast to previous work, we further validate our findings by comparing results associated with recordings from retinas under normal and altered encoding conditions obtained by pharmacological manipulation. In these conditions, we observe that the model reflects well-known physiological behaviors of the retina. Finally, we show that we can also discover temporal patterns, associated with distinct dynamics of the stimuli.

Nanoparticles (NPs) are increasingly used in biomedical applications, but the factors that influence their interactions with living cells need to be elucidated. Here, wereveal the role of NP surface charge in determining theirneuronal interactions and electrical responses. We discoveredthat negatively charged NPs administered at low concentration(10 nM) interact with the neuronal membrane and at the synaptic cleft, whereas positively and neutrally charged NPs never localize on neurons. This effect is shape and material independent. The presence of negatively charged NPs on neuronal cell membranes influences the excitability of neurons by causing an increase in the amplitude and frequency of spontaneous postsynaptic currents at the single cell level and an increase of both the spiking activity and synchronous firing at neural network level. The negatively charged NPs exclusively bind to excitable neuronal cells, and never to nonexcitable glial cells. This specific interaction was also confirmed by manipulating the electrophysiological activity of neuronal cells. Indeed, the interaction of negatively charged NPs with neurons is either promoted or hindered by pharmacological suppression or enhancement of the neuronal activity with tetrodotoxin or bicuculline, respectively. We further support our main experimental conclusions by using numerical simulations. This study demonstrates that negatively charged NPs modulate the excitability of neurons, revealing the potential use of NPs for controlling neuron activity.

The control of neuron-substrate adhesion has been always a challenge for fabricating neuron-based cell chips and in particular for multielectrode array (MEA) devices, which warrants the investigation of the electrophysiological activity of neuronal networks. The recent introduction of high-density chips based on the complementary metal oxide semiconductor (CMOS) technology, integrating thousands of electrodes, improved the possibility to sense large networks and raised the challenge to develop newly adapted functionalization techniques to further increase neuron electrode localization to avoid the positioning of cells out of the recording area. Here, we present a simple and straightforward chemical functionalization method that leads to the precise and exclusive positioning of the neural cell bodies onto modified electrodes and inhibits, at the same time, cellular adhesion in the surrounding insulator areas. Different from other approaches, this technique does not require any adhesion molecule as well as complex patterning technique such as μ-contact printing. The functionalization was first optimized on gold (Au) and silicon nitride (Si3N4)-patterned surfaces. The procedure consisted of the introduction of a passivating layer of hydrophobic silane molecules (propyltriethoxysilane [PTES]) followed by a treatment of the Au surface using 11-amino-1-undecanethiol hydrochloride (AT). On model substrates, well-ordered neural networks and an optimal coupling between a single neuron and single micrometric functionalized Au surface were achieved. In addition, we presented the preliminary results of this functionalization method directly applied on a CMOS-MEA: the electrical spontaneous spiking and bursting activities of the network recorded for up to 4 weeks demonstrate an excellent and stable neural adhesion and functional behavior comparable with what expected using a standard adhesion factor, such as polylysine or laminin, thus demonstrating that this procedure can be considered a good starting point to develop alternatives to the traditional chip coatings to provide selective and specific neuron-substrate adhesion.

Electrophysiological signals in the brain are distributed over broad spatial and temporal scales. Monitoring these signals at multiple scales is fundamental in order to decipher how brain circuits operate and might dysfunction in disease. A possible strategy to enlarge the experimentally accessible spatial and temporal scales consists in combining the use of multiple probes with different resolutions and sensing areas. Here, we propose a neural recording system capable of simultaneous and synchronous acquisitions from a new generation of high-resolution CMOS probes (512 microelectrodes, 25 kHz/electrode whole-array sampling frequency) as well as from a custom-designed CMOS-based headstage. While CMOS probes can provide recordings from a large number of closely spaced electrodes on single-shaft devices, the CMOS-based headstage can be used to interface the wide range of available intra- or epi-cortical passive electrode array devices. The current platform was designed to simultaneously manage high-resolution recordings from up to four differently located CMOS probes and from a single 36-channels low-resolution passive electrode array device. The design, implementation, and performances for both ICs and for the FPGA-based interface are presented. Experiments on retina and neuronal culture preparations demonstrate the recording of neural spiking activity for both CMOS devices and the functionality of the system.

Substrate-integrated multielectrode arrays (MEAs) enable multisite, long-term, and label-free sensing and actuation of neuronal electrical signals in reduced cell culture models for network electrophysiology. Conventional, thin-film fabricated passive MEAs typically provide a few tens of electrode sites. New generations of active CMOS-based high-resolution arrays provide the capabilities of simultaneous recordings from thousands of neurons over fields of view of several square millimeters, yet allowing extracellular electrical imaging to be achieved down to the subcellular scale. In turn, such advancement in chip-based electrical readouts can significantly complement recently developed biotechnological and bimolecular techniques for neurobiology applications. Here, we describe (1) a simple method to fabricate passive MEAs and (2) protocols for preparing and growing primary rat hippocampal neuronal cultures and human iPS-derived neurons on MEAs. The aim is to provide reliable protocols for initiating the reader to this technology and for stimulating their further development and experimental use in neurobiology.

Neuronal responses to external stimuli vary from trial to trial partly because they depend on continuous spontaneous variations of the state of neural circuits, reflected in variations of ongoing activity prior to stimulus presentation. Understanding how post-stimulus responses relate to the pre-stimulus spontaneous activity is thus important to understand how state dependence affects information processing and neural coding, and how state variations can be discounted to better decode single-trial neural responses. Here we exploited high-resolution CMOS electrode arrays to record simultaneously from thousands of electrodes in in-vitro cultures stimulated at specific sites. We used information-theoretic analyses to study how ongoing activity affects the information that neuronal responses carry about the location of the stimuli. We found that responses exhibited state dependence on the time between the last spontaneous burst and the stimulus presentation and that the dependence could be described with a linear model. Importantly, we found that a small number of selected neurons carry most of the stimulus information and contribute to the state-dependent information gain. This suggests that a major value of large-scale recording is that it individuates the small subset of neurons that carry most information and that benefit the most from knowledge of its state dependence.

Substrate-integrated multielectrode arrays (MEAs) enable multisite, long-term, and label-free sensing and actuation of neuronal electrical signals in reduced cell culture models for network electrophysiology. Conventional, thin-film fabricated passive MEAs typically provide a few tens of electrode sites. New generations of active CMOS-based high-resolution arrays provide the capabilities of simultaneous recordings from thousands of neurons over fields of view of several square millimeters, yet allowing extracellular electrical imaging to be achieved down to the subcellular scale. In turn, such advancement in chip-based electrical readouts can significantly complement recently developed biotechnological and bimolecular techniques for neurobiology applications. Here, we describe (1) a simple method to fabricate passive MEAs and (2) protocols for preparing and growing primary rat hippocampal neuronal cultures and human iPS-derived neurons on MEAs. The aim is to provide reliable protocols for initiating the reader to this technology and for stimulating their further development and experimental use in neurobiology.

In this work we present a FPGA-based system for real-time processing of neural signals acquired by commercial high-density microelectrode array (HDMEA). The considered MEA features 4096 electrodes with 18kHz sampling frequency and 12-bit resolution, thus produces nearly 1 Gbps of data. Within the implementation, we considered low-latency as a main objective, to allow for closed-loop acquisition-stimulation experiments, that represent a novel promising frontier in neuro-physiology and in the development of brain-machine interfaces. The developed platform is implemented on a low-to-mid Zynq all-programmable SoC, and is able to perform all the required computation (from signal acquisition to response generation) with less than 2ms latency, enabling closed-loop applications in a wide range of experiments.

Understanding the interplay among the spectrum of electrophysiological signals in the brain, distributed over broad spatial and temporal scales, is fundamental to decipher how brain circuits operate and might dysfunction in disease. Here, we present a multiscale recording system for simultaneous and synchronous acquisitions from a new generation of im-plantable CMOS active probes (single-shaft, 512 microelectrodes, 25 kHz/electrode full-array sampling) as well as from implantable conventional electrode array probes interfaced with a custom-designed CMOS-based headstage. The platform can manage highresolution recordings from up to 4 differently located probes, and one 36-channels low-resolution passive electrode array. As presented here, and prior to complete in-vivo validation, the design and recording performance of high- and low-density electrode array ICs were experimentally tested ex-vivo on retinal whole-mounts.

The immature retina generates spontaneous waves of spiking activity that sweep across the ganglion cell layer during a limited period of development before the onset of visual experience. The spatiotemporal patterns encoded in the waves are believed to be instructive for the wiring of functional connections throughout the visual system. However, the ontogeny of retinal waves is still poorly documented as a result of the relatively low resolution of conventional recording techniques. Here, we characterize the spatiotemporal features of mouse retinal waves from birth until eye opening in unprecedented detail using a large-scale, dense, 4096-channel multielectrode array that allowed us to record from the entire neonatal retina at near cellular resolution. We found that early cholinergic waves propagate with random trajectories over large areas with low ganglion cell recruitment. They become slower, smaller and denser when GABAA signalling matures, as occurs beyond postnatal day (P) 7. Glutamatergic influences dominate from P10, coinciding with profound changes in activity dynamics. At this time, waves cease to be random and begin to show repetitive trajectories confined to a few localized hotspots. These hotspots gradually tile the retina with time, and disappear after eye opening. Our observations demonstrate that retinal waves undergo major spatiotemporal changes during ontogeny. Our results support the hypotheses that cholinergic waves guide the refinement of retinal targets and that glutamatergic waves may also support the wiring of retinal receptive fields.

Neurotoxicity and the accumulation of extracellular amyloid-beta1–42 (Aβ) peptides are associated with the development of Alzheimer’s disease (AD) and correlate with neuronal activity and network dysfunctions, ultimately leading to cellular death. However, research on neurodegenerative diseases is hampered by the paucity of reliable readouts and experimental models to study such functional decline from an early onset and to test rescue strategies within networks at cellular resolution. To overcome this important obstacle, we demonstrate a simple yet powerful in vitro AD model based on a rat hippocampal cell culture system that exploits large-scale neuronal recordings from 4096-electrodes on CMOS-chips for electrophysiological quantifications. This model allows us to monitor network activity changes at the cellular level and to uniquely uncover the early activity-dependent deterioration induced by Aβ-neurotoxicity. We also demonstrate the potential of this in vitro model to test a plausible hypothesis underlying the Aβ-neurotoxicity and to assay potential therapeutic approaches. Specifically, by quantifying N-methyl D-aspartate (NMDA) concentration-dependent effects in comparison with low-concentration allogenic-Aβ, we confirm the role of extrasynaptic-NMDA receptors activation that may contribute to Aβ-neurotoxicity. Finally, we assess the potential rescue of neural stem cells (NSCs) and of two pharmacotherapies, memantine and saffron, for reversing Aβ-neurotoxicity and rescuing network-wide firing.

A new method for automated spike sorting for recordings with high density, large scale multielectrode arrays is presented. Exploiting the dense sampling of single neurons by multiple electrodes, we obtain an efficient, low-dimensional representation of detected spikes consisting of estimated spatial spike locations and dominant spike shape features, which enables fast and reliable clustering into single units. Millions of events can be sorted in minutes, and the method is parallelized and scales better than quadratically with the number of detected spikes. We demonstrate this method using recordings with a 4,096 channel array, and present validation based on anatomical imaging, optogenetic stimulation and model-based quality control. A comparison with semi-automated, shape-based spike sorting exposes significant limitations of conventional methods. Our analysis shows that it is feasible to reliably isolate the activity of hundreds to thousands of neurons in a single recording, and that dense, multi-channel probes substantially aid reliable spike sorting.

The recent availability of human induced pluripotent stem cells (hiPSCs) holds great promise as a novel source of human-derived neurons for cell and tissue therapies as well as for in vitro drug screenings that might replace the use of animal models. However, there is still a considerable lack of knowledge on the functional properties of hiPSC-derived neuronal networks, thus limiting their application. Here, upon optimization of cell culture protocols, we demonstrate that both spontaneous and evoked electrical spiking activities of these networks can be characterized on-chip by taking advantage of the resolution provided by CMOS multielectrode arrays (CMOS-MEAs). These devices feature a large and closely-spaced array of 4096 simultaneously recording electrodes and multi-site on-chip electrical stimulation. Our results show that networks of human-derived neurons can respond to electrical stimulation with a physiological repertoire of spike waveforms after three months of cell culture, a period of time during which the network undergoes the expression of developing patterns of spontaneous spiking activity. To achieve this, we have investigated the impact on the network formation and on the emerging network-wide functional properties induced by different biochemical substrates, i.e. poly-dl-ornithine (PDLO), poly-l-ornithine (PLO), and polyethylenimine (PEI), that were used as adhesion promoters for the cell culture. Interestingly, we found that neuronal networks grown on PDLO coated substrates show significantly higher spontaneous firing activity, reliable responses to low-frequency electrical stimuli, and an appropriate level of PSD-95 that may denote a physiological neuronal maturation profile and synapse stabilization. However, our results also suggest that even three-month culture might not be sufficient for human-derived neuronal network maturation. Taken together, our results highlight the tight relationship existing between substrate coatings and emerging network properties, i.e. spontaneous activity, responsiveness, synapse formation and maturation. Additionally, our results provide a baseline on the functional properties expressed over three months of network development for a commercially available line of hiPSC-derived neurons. This is a first step toward the development of functional pre-clinical assays to test pharmaceutical compounds on human-derived neuronal networks with CMOS-MEAs.

We used electrophysiological signals recorded by CMOS Micro Electrode Arrays (MEAs) at high spatial resolution to estimate the functional-effective connectivity of sparse hippocampal neuronal networks in vitro by applying a cross-correlation (CC) based method and ad hoc developed spatio-temporal filtering. Low-density cultures were recorded by a recently introduced CMOS-MEA device providing simultaneous multi-site acquisition at high-spatial (21 μm inter-electrode separation) as well as high-temporal resolution (8 kHz per channel). The method is applied to estimate functional connections in different cultures and it is refined by applying spatio-temporal filters that allow pruning of those functional connections not compatible with signal propagation. This approach permits to discriminate between possible causal influence and spurious co-activation, and to obtain detailed maps down to cellular resolution. Further, a thorough analysis of the links strength and time delays (i.e., amplitude and peak position of the CC function) allows characterizing the inferred interconnected networks and supports a possible discrimination of fast mono-synaptic propagations, and slow poly-synaptic pathways. By focusing on specific regions of interest we could observe and analyze microcircuits involving connections among a few cells. Finally, the use of the high-density MEA with low density cultures analyzed with the proposed approach enables to compare the inferred effective links with the network structure obtained by staining procedures.

Recent developments in the neuroengineering field and the widespread use of the micro electrode arrays (MEAs) for electrophysiological investigations made available new approaches for studying the dynamics of dissociated neuronal networks as well as acute/organotypic slices maintained ex vivo. Importantly, the extraction of relevant parameters from these neural populations is likely to involve long-term measurements, lasting from a few hours to entire days. The processing of huge amounts of electrophysiological data, in terms of computational time and automation of the procedures, is actually one of the major bottlenecks for both in vivo and in vitro recordings. In this paper we present a collection of algorithms implemented within a new software package, named the Neural Signal Manager (NSM), aimed at analyzing a huge quantity of data recorded by means of MEAs in a fast and efficient way. The NSM offers different approaches for both spike and burst analysis, and integrates state-of-the-art statistical algorithms, such as the inter-spike interval histogram or the post stimulus time histogram, with some recent ones, such as the burst detection and its related statistics. In order to show the potentialities of the software, the application of the developed algorithms to a set of spontaneous activity recordings from dissociated cultures at different ages is presented in the Results section. Copyright © 2008 John Wiley & Sons, Ltd.

The microelectrode array (MEA) technology is continuously progressing towards higher integration of an increasing number of electrodes. The ensuing data streams that can be of several hundreds or thousands of Megabits/s require the implementation of new signal processing and data handling methodologies to substitute the currently used off-line analysis methods. Here, we present one approach based on the hardware implementation of a wavelet-based solution for real-time processing of extracellular neuronal signals acquired on high-density MEAs. We demonstrate that simple mathematical operations on the discrete wavelet transform (DWT) coefficients can be used for efficient neuronal spike detection and sorting. As the DWT is particularly well suited for implementation on dedicated hardware, we elaborated a wavelet processor on a field programmable gate array (FPGA) in order to compute the wavelet coefficients on 256 channels in real-time. By providing sufficient hardware resources, this solution can be easily scaled up for processing more electrode channels. Copyright © 2008 John Wiley & Sons, Ltd.

Visual Abstract, How a population of retinal ganglion cells (RGCs) encodes the visual scene remains an open question. Going beyond individual RGC coding strategies, results in salamander suggest that the relative latencies of a RGC pair encode spatial information. Thus, a population code based on this concerted spiking could be a powerful mechanism to transmit visual information rapidly and efficiently., How a population of retinal ganglion cells (RGCs) encodes the visual scene remains an open question. Going beyond individual RGC coding strategies, results in salamander suggest that the relative latencies of a RGC pair encode spatial information. Thus, a population code based on this concerted spiking could be a powerful mechanism to transmit visual information rapidly and efficiently. Here, we tested this hypothesis in mouse by recording simultaneous light-evoked responses from hundreds of RGCs, at pan-retinal level, using a new generation of large-scale, high-density multielectrode array consisting of 4096 electrodes. Interestingly, we did not find any RGCs exhibiting a clear latency tuning to the stimuli, suggesting that in mouse, individual RGC pairs may not provide sufficient information. We show that a significant amount of information is encoded synergistically in the concerted spiking of large RGC populations. Thus, the RGC population response described with relative activities, or ranks, provides more relevant information than classical independent spike count- or latency- based codes. In particular, we report for the first time that when considering the relative activities across the whole population, the wave of first stimulus-evoked spikes is an accurate indicator of stimulus content. We show that this coding strategy coexists with classical neural codes, and that it is more efficient and faster. Overall, these novel observations suggest that already at the level of the retina, concerted spiking provides a reliable and fast strategy to rapidly transmit new visual scenes.

High density multielectrode arrays (MEAs) based on CMOS technology (CMOS-MEAs) can simultaneously record extracellular spiking activity in neuronal cultures from 4096 closely spaced microelectrodes. This allows for a finer investigation of neuronal network activity compared to conventional MEAs with a few tens of electrodes. However, the sensing properties of these devices differ. To highlight this aspect, here we investigate and discuss the differences observed when quantifying spontaneous synchronized bursting events (SBEs) in datasets acquired with conventional MEAs and high-density MEAs from comparable hippocampal cultures. We found that datasets acquired with high-density MEAs exhibit collective dynamics similar to conventional arrays, but are characterized by a higher percentage of random spikes, i.e. spikes that are not part of a burst, most probably resulting from the larger recording capability. Additionally, the percentage of electrodes that record a burst is remarkably small on high-density MEAs compared to what can be observed on conventional MEAs and SBEs appear to be propagating in time across the electrode array, by involving shorter sequences of spikes per electrode. Overall, these results highlight a lower level of network synchronization involved in SBEs compared to what has been debated for several decades based on conventional MEA recordings from cell cultures.

Ion Beam Induced Deposition (IBID) is employed to fabricate three-dimensional nanoprotrusions on top of the recording pads of an active pixel sensor array (APS-MEA) featuring 4096 microelectrodes. Modified APS-MEAs are envisioned as enhanced tools to achieve real-time ''in-cell'' recordings from thousands of sensing elements, thus aiming to large-scale in-vitro registrations with unprecedented signal quality. A generalized electric model is proposed to address the revealed complexity of the neuron/electrode interface, and simulations have been conducted revealing the most advantageous cell/electrode coupling conditions. Preliminary results on the recording of spontaneous activity in cultured neuronal networks by means of nanostructured microelectrodes demonstrate the compatibility of IBID technology and APSMEA infrastructure. The interface between cultured mammalian neurons and modified microelectrodes is revealed by FIB/SEM analysis, fostering the employment of the proposed electrical model for interpretation of electrical recordings from nanostructured microelectrodes. ©2013 Elsevier B.V. All rights reserved.

Deciphering neural network function in health and disease requires recording from many active neurons simultaneously. Developing approaches to increase their numbers is a major neurotechnological challenge. Parallel to recent advances in optical Ca(2+) imaging, an emerging approach consists in adopting complementary-metal-oxide-semiconductor (CMOS) technology to realize MultiElectrode Array (MEA) devices. By implementing signal conditioning and multiplexing circuits, these devices allow nowadays to record from several thousands of single neurons at sub-millisecond temporal resolution. At the same time, these recordings generate very large data streams which become challenging to analyze. Here, at first we shortly review the major approaches developed for data management and analysis for conventional, low-resolution MEAs. We highlight how conventional computational tools cannot be easily up-scaled to very large electrode array recordings, and custom bioinformatics tools are an emerging need in this field. We then introduce a novel approach adapted for the acquisition, compression and analysis of extracellular signals acquired simultaneously from 4096 electrodes with CMOS MEAs. Finally, as a case study, we describe how this novel large scale recording platform was used to record and analyze extracellular spikes from the ganglion cell layer in the wholemount retina at pan-retinal scale following patterned light stimulation.

Neurons communicate in the brain via spikes. Understanding the balance between the fast-firing minority and slow-firing majority in a neuronal population is therefore a fundamental step to unravel the nature of communication and of operation within and across neuronal assemblies. Recent in vivo observations show that many functional and structural parameters of the brain follow a skewed nature and typically manifest lognormal distributions of patterns. Here, we show for the first time that high-density microelectrode array (HD-MEA) reveal such a lognormal-like distribution of the firing patterns also in in vitro grown hippocampal neuronal networks, and already after 10 minutes of recording. Additionally, we demonstrate that the electrode density plays a key role for obtaining such a distribution in cultures. Overall, our findings show that in vitro neural networks recorded with CMOS-MEAs might contribute in investigating the organization and function of neuronal networks by revealing, with similarities with results obtained in vivo, the relationships among different skewed distributions at multiple scales, i.e. from synapses, single cells and micro-circuits up to large-scale networks.

Neuronal signaling in brain circuits occurs at multiple scales ranging from molecules and cells to large neuronal assemblies. However, current sensing neurotechnologies are not designed for parallel access of signals at multiple scales. With the aim of combining nanoscale molecular sensing with electrical neural activity recordings within large neuronal assemblies, in this work three-dimensional (3D) plasmonic nanoantennas are integrated with multielectrode arrays (MEA). Nanoantennas are fabricated by fast ion beam milling on optical resist; gold is deposited on the nanoantennas in order to connect them electrically to the MEA microelectrodes and to obtain plasmonic behavior. The optical properties of these 3D nanostructures are studied through finite elements method (FEM) simulations that show a high electromagnetic field enhancement. This plasmonic enhancement is confirmed by surface enhancement Raman spectroscopy of a dye performed in liquid, which presents an enhancement of almost 100 times the incident field amplitude at resonant excitation. Finally, the reported MEA devices are tested on cultured rat hippocampal neurons. Neurons develop by extending branches on the nanostructured electrodes and extracellular action potentials are recorded over multiple days in vitro. Raman spectra of living neurons cultured on the nanoantennas are also acquired. These results highlight that these nanostructures could be potential candidates for combining electrophysiological measures of large networks with simultaneous spectroscopic investigations at the molecular level.