Interfacing with the Brain: How Nanotechnology Can Contribute

Abstract

Interfacing artificial devices with the human brain is the central goal of neurotechnology. Yet, our imaginations are often limited by currently available paradigms and technologies. Suggestions for brain-machine interfaces have changed over time, along with the available technology. Mechanical levers and cable winches were used to move parts of the brain during the mechanical age. Sophisticated electronic wiring and remote control have arisen during the electronic age, ultimately leading to plug-and-play computer interfaces. Nonetheless, our brains are so complex that these visions, until recently, largely remained unreachable dreams. The general problem, thus far, is that most of our technology is mechanically and/or electrically engineered, whereas the brain is a living, dynamic entity. As a result, these worlds are difficult to interface with one another. Nanotechnology, which encompasses engineered solid-state objects and integrated circuits, excels at small length scales of single to a few hundred nanometers and, thus, matches the sizes of biomolecules, biomolecular assemblies, and parts of cells. Consequently, we envision nanomaterials and nanotools as opportunities to interface with the brain in alternative ways. Here, we review the existing literature on the use of nanotechnology in brain-machine interfaces and look forward in discussing perspectives and limitations based on the authors' expertise across a range of complementary disciplines─from neuroscience, engineering, physics, and chemistry to biology and medicine, computer science and mathematics, and social science and jurisprudence. We focus on nanotechnology but also include information from related fields when useful and complementary.

Keywords: Nanoneuro interface; brain-on-a-chip; brain−machine interfaces; control of ion channels; deep brain stimulation; electrode arrays; extracellular recordings; nanostructured interface; neuro-implants; neuronal communication.

Figure 1

Nanoelectrodes made with different fabrication approaches and shapes. (a) Scanning electron microscopy (SEM) image of so-called gold-spine electrodes (adapted with permission from Hai et al. Copyright 2009 The Royal Society). The height of the structure is 1.56 μm. (b) SEM image of an iridium oxide nanotube electrode (adapted with permission from Lin et al. Copyright 2014 Springer Nature Limited). (c) SEM image of core–shell-type nanowires connected toward external contacts with encapsulated conductive lines (adapted with permission from Casanova et al. Copyright 2018 IOP Publishing Ltd). (d) SEM image of a silicon-based ultrasharp nanowire with an exposed Pt tip (adapted with permission from Liu et al. Copyright 2022 Wiley-VCH).

Figure 2

Example recordings from various groups with electro- and optoporation: Electrical recording from cardiac myocytes (a) before and (b) after electroporation. Image adapted with permission from Lin et al. Copyright 2014 Macmillan Publishers. (c) Electrical recording from cardiac myocytes after optoporation (upper graph) and electroporation (lower graph). Image adapted with permission from Dipalo et al. Copyright 2019 Wiley-VCH. (d) Extracellular recordings after electroporation. Excitatory postsynaptic potentials (EPSPs) and their triggering of an action potential (AP) are also visible. Image adapted with permission from Abbott et al. Copyright 2020 Springer Nature.

Figure 3

(a) Scanning electron microscope images of five nanostraws with 2 μm pitch on electrodes with a 6 μm diameter opening; the diameter of the nanostraws is on the order of hundreds of nanometers (adapted with permission from Shokoohimehr et al. Copyright 2022 Wiley-VCH). (b) Staining and resin embedding. (c) Inset of panel (b) [the red dotted region in (b) indicates the nanostraw on the right in (c)]. The fixed cells tightly engulf the nanostraws while the nucleus is being deformed at the tip of the nanostraws (unpublished images from the Offenhäusser group).

Figure 4

Simultaneous recording of the neuron’s activity using a patch-clamp electrode (black trace) and nanostraw–nanocavity–microelectrode array (NS-NC-MEA) (blue trace) focusing on a giant excitatory postsynaptic potential (EPSP) triggered by an action potential (AP) (left) and a spikelet (right). Bottom left: NS–NC–MEA detects distinct spikes that correspond to quenched and coinciding PSPs. Vertical scale bar corresponds to 40 mV (black) and 400 μV (blue). Time scale = 50 ms. Bottom right: details of postsynaptic potentiations (PSPs) in both patch-clamp and MEA traces. Amplitude scale is 20 mV (black) and 100 μV (blue). Time scale = 50 ms. Reproduced with permission from Shokoohimehr et al. Copyright 2022 Wiley-VCH.

Figure 5

(a) Left to right: scanning electron microscopy (SEM) images of a macroscopic electrode with attached, freestanding GaP nanowires (NWs) (shown in magnification). Photograph of the NW-based electrode attached to a micromanipulator to enable in vivo neuronal recordings from the rat’s cortex. (b) False-colored SEM image of an individual (micro)needle from an array that has been used to electrically contact the left whisker barrel area in the somatosensory cortex of a rat (schematic and photograph middle column). On the right, the recorded and amplified wideband signal of the cortex after stimulating the rat’s whisker is shown. Images are taken and adapted with permission from Suyatin et al. and Fujishiro et al. Copyright 2013 PLoS and 2014 Springer Nature, respectively.

Figure 6

(a) Schematic depictions of an injectable, flexible nanowire network. (b) Left: Scanning electron microscopy (SEM) images of a kinked field-effect transistor nanowire (1), which is contacted by metallic interconnects (2) and supported by a polymeric SU-8 mesh (3) to form a nanowire nanoelectronic scaffold (nanoES). Middle and right: Hybrid nanoES device (false-colored in brown) based on an alginate scaffold. Images are taken and adapted with permission from Liu et al. and Tian et al. Copyright 2015 and 2012 MacMillan Publishers, respectively.

Figure 7

Nanowire (NW)-mediated light excitation of neuronal cells in (a) the retina or (b) heart muscle cells. (a) From left to right: Schematics of the replacement of biological photoreceptors in the retina by Au-decorated TiO₂ NW arrays. Scanning electron microscopy image of the retina in contact with the NW array (scale bare 5 μm). Whole-cell patch-clamp recordings of the retinal ganglion cells (RCGs) upon UV light recording for wild-type, blind, and blind retinas in contact nanowires. (b) Upper row from left to right: Schematics of the NW implant at the porcine heart. Photograph of the flexible device consisting of an aluminum foil on which radial junction (RJ) nanowires have been grown. Lower panel: Heartbeat as a function of the light irradiation of an implant. Images are adapted with permission from Tang et al. and Liu et al. Copyright 2018 Springer Nature and 2020 Wiley-VCH, respectively.

Figure 8

Characterization of carbon nanotube (CNT) substrates and ultrastructural interaction between CNTs and cultured neurons. (a) Scanning electron microscopy (SEM) images of cultured hippocampal neurons on CNTs grown for 10 days in vitro (DIV). Note the healthy morphology of the neurons and the outgrowth of neurites attaching to the CNT surface. At higher magnifications, the intimate contacts between bundles of CNTs and neuronal membrane are observed. Adapted with permission from Mazzatenta et al. Copyright 2010 Society for Neuroscience. (b) Organotypic spinal cultures: impact of multiwalled CNT (MWCNT) interfaces on neurite outgrowth. (A) SEM image of a peripheral neuronal fiber of a control spinal explant grown on glass. Scale bar: 500 nm. (B) Scanning electron microscopy (SEM) image of a spinal explant peripheral neuronal fiber on a CNT substrate; note the tight and intimate contacts (red arrows) between the neurite membrane and the MWCNTs. Scale bar 500 nm. (C,D) Confocal microscopy image reconstructions of spinal slice cultures at 8 DIV under control and CNT growth conditions, respectively. Immunofluorescence of specific cytoskeletal components, F-actin, β-tubulin III, and glial fibrillary acidic protein (GFAP). Note the β-tubulin III positive neuronal processes radially exiting the growth area in both cultured explants. (E,F) High-resolution confocal magnifications of the framed areas highlighted in (C) and (D), respectively, visualize the bundles of fibers emerging from the growing belt located around the slices. (C–F): green, GFAP; red, F-actin; blue, β-tubulin III. In (C–D): scale bar 1 mm. In (E,F): scale bar 500 μm. Adapted with permission from Fabbro et al. Copyright 2012 American Chemical Society.

Figure 9

Three-dimensional carbon nanotube-based sponges (3D CNFs) guide the functional reconnection of ventral outputs in segregated spinal organotypic slices, cocultured in “Control” and in 3D CNFs after 14 days of growth. (a) Immunofluorescence is displayed for neuron-specific microtubules (b-tubulin III; red), neurofilament H (SMI-32; green), and nuclei (DAPI; blue). Scale bars 500 μm. (b) Sketch of the experimental setting for double-slice ventral recordings and dorsal stimulation. (c) Local field potential bursting induced by strychnine and bicuculline recorded simultaneously from left (L) and right (R) slices in Control and 3D CNF. (d) Bursting local field potentials (LFP) entrainment by dorsal electrical stimulation (dots) of left slices (arrow) in Control and 3D CNF slice pairs. Adapted with permission from Usmani et al. Copyright 2016 The Authors.

Figure 10

Tissue reaction to carbon nanotube-based sponges (CNFs)-based scaffolds implanted into the adult rat visual cortex as visualized by immunostaining of glial fibrillary acidic protein (GFAP) and Iba1. GFAP is a marker for reactive astrocytes, and Iba1 is a marker for microglial cells, the resident immune cells of the central nerves system. (a) GFAP-positive cells (green) are found surrounding the implant and within the material; the boxed areas indicate high-magnification images shown in (b); Inset, contralateral hemisphere used as a control. Scale bar 200 μm. (b) High magnification of GFAP reactivity at the implant edge demonstrating the minimal and irregular cellular localization around the scaffold. Scale bar 50 μm. (c) Iba1-positive cells (red) are dispersed consistently throughout the tissue and within the material; the boxed areas indicate the high-magnification images shown in (d); inset, contralateral hemisphere used as a control. Scale bar 200 μm. (d) High-magnification images of the ionized calcium-binding adapter molecule 1 (Iba1) reactivity demonstrate no obvious border at the implant edge to indicate scar formation. Scale bar 50 μm. Adapted with permission from Usmani et al. Copyright 2016 The Authors.

Figure 11

Carbon nanotube-based sponge (CNF) supports implantation in spinal cord injury animal models. (a) Confocal micrographs detail the lesion site at low (top) and high (bottom) magnification. Arrowheads indicate shredded remains and fibers in poly(ethylene glycol) (PEG) (left) and tortuous axons within the CNF (right). Scales top (left and right), 100 μm; bottom (left and right), 25 μm. (b) Fiber tracks in aged-matched naïves (Control) and spinal cord injury (SCI) (PEG and CNF) at 5 to 6 months after surgery, with fractional anisotropy (FA) values ranging from FA = 0 (in blue) to FA = 1 (in red). Right column: 3D representations of fiber tracts of five different examples of 5 to 6 months carbon nanotube (CNT)-implanted animals. Scale bars, 2 mm. (c) Fiber tracking analysis of diffusion tensor imaging (DTI) data constructed along the implant area of a CNF-treated rat (6 months post-SCI; only half spine presented to facilitate visualization) with the 2D MRI coronal plane through the implant. Colors represent fiber orientation following the conventional code for tensor directionality (blue, anterior–posterior; red, left–right; and green, dorsal–ventral directions). Scale, 1 mm. Adapted with permission from Usmani et al. Copyright 2020 The Author(s).

Figure 12

(a) Scanning electron microscope images of SH-SY5Y cells grown on poly(3,4-ethylenedioxythiophene)/carbon nanotube (PEDOT/CNT) scaffolds after 3 (top) and 7 (bottom) days of culture (DIV). The red arrows indicate cells. The scale bars for the images on the left and right correspond to 20 and 5 μm, respectively. (b) β-Tubulin class III and f-actin staining of SH-SY5Y cells grown on PEDOT and PEDOT/CNT scaffolds after 7 DIV. The scale bar represents 50 μm. (c) Amount of β-IIITub expressed in terms of “signal-to-noise ratio” of the incubated cells. Adapted with permission from Dominguez-Alfaro et al. Copyright 2020 American Chemical Society.

Figure 13

Graphene-based neuronal interfaces have been designed, fabricated, and quality-controlled to achieve reproducible functionality for brain signal recording, electrical neuronal stimulation, and biosensing. (a) Multifunctional graphene-based neuronal interface concept schematics. Modified with permission from Kostarelos et al. Copyright 2017 Wiley-VCH. (b) Examples of functional graphene-based neuronal surface probes fabricated using (i) chemical vapor deposition (CVD) graphene field-effect transistor technology and (ii) reduced graphene oxide membrane technology used on a murine cortex (bottom image). Reproduced in modified form with permission from Garcia-Cortadella et al. and Viana et al. Copyright 2021 Springer Nature and 2022 The Authors, respectively. (c) Example of a graphene-based intracortical probe using the graphene field-effect transistor technology. Reprinted with permission from Bonaccini Calia et al. Copyright 2022 Springer Nature.

Figure 14

Conductive hydrogel-based neuronal interfaces: from molecular structures to applications. (a) A nanocomposite hydrogel composed of polyacrylamide and plasmonic silver nanocubes. The constructs benefit from well-dispersed silver nanocubes inside the hydrogel network, contributing to the formation of conducting pathways. The nanocomposite hydrogel was surrounded by a silicon-based template and utilized as a neuronal interface for in vivo electrocorticography (ECoG) recordings on a mouse model, and the long-term neuronal signal acquisition was practiced. Reproduced with permission from Rinoldi et al. Copyright 2022 American Chemical Society. (b) A conductive semi-interpenetrating network (IPN) hydrogel based on polythiophene. The hydrogel was synthesized by blending polythiophene with a poly(N-isopropylacrylamide) [p(NIPAAm)] copolymer, along with a cross-linker and photoinitiator. Subsequently, UV light exposure in a controlled, cold environment facilitated the formation of a conductive semi-IPN hydrogel to offer enhanced electrical conductivity, thermoresponsiveness, and biocompatibility. Reproduced with permission from Tian et al. Copyright 2021 American Chemical Society. (c) A conductive IPN hydrogel based on poly(3,4-ethylenedioxythiophene)-MeOH:poly(styrenesulfonate)/polydopamine (PEDOT-MeOH:PSS/PDA). The design of this adhesive conducting interface involves the incorporation of a thin PDA layer to enable the formation of interpenetrating networks through electropolymerization. The fabrication procedure follows a simple two-step methodology. Initially, PDA is electropolymerized to create an adhesive conductive thin layer on the wire microelectrodes. Subsequently, EDOT-MeOH with PSS acting as the supported polyelectrolyte undergoes electropolymerization to generate the desired interpenetrating PEDOT-MeOH:PSS/PDA networks. Reproduced with permission from Tian et al. Copyright 2023 Elsevier Inc.

Figure 15

Different members from the rhodopsin family and their function. Image reproduced with permission from Zhang et al. Copyright 2011 Elsevier.

Figure 16

Three types of phototriggered rhodopsins: (left) light-triggered ion-channel thermally reverting to its initial state, e.g., channelrhodopsins; (middle) light-triggered ion pump thermally reverting, e.g., bacteriorhodopsin and halorhodopsin; (right) light-triggered sensory pigment requires retinal isomerase to revert to its initial state. For optogenetic use, channelrhodopsin is the best-suited molecule.

Figure 17

Demonstration of upconverting nanoparticles (UCNPs) as transducers for converting NIR light into green light for exciting light-gated ion channels. (a) Schematic of tetherless near-infrared (NIR) optogenetic control of brain activity using fully implantable upconversion microdevices. (b) Bright-field and fluorescent photographs of the implantable micro-optrodes containing UCNPs doped with Tm³⁺ (blue) or Er³⁺ (green). Scale bar, 500 mm. (c) Fluorescent images of the operating UCNP microdevices (Tm³⁺-doped, blue; Er³⁺-doped, green) excited by near-infrared (NIR) light. Scale bar, 2 mm. (d) Images of animals implanted with different types of micro-optrodes containing Tm³⁺-doped (top) or Er³⁺-doped (bottom) UCNPs. Scale bar, 1 cm. (e) Instrumentation design of a robotic laser projection system for automatic and consistent NIR irradiation of the heads of behaving animals. Figure and caption taken with permission from Wang et al. Copyright 2017 Elsevier.

Figure 18

Amplification model for endolysosomal Ca²⁺ signaling. (a) Laser irradiation in the near-infrared results in small and brief local Ca²⁺ signals that require amplification for global Ca²⁺ signaling, here as a regenerative Ca²⁺ wave caused by Ca²⁺-induced Ca²⁺ release (CICR) in MCF-7 cells. (b) Depletion of the endoplasmic reticulum results in pure lysosomal Ca²⁺ microdomains due to lack of amplification. (c) Future applications for endolysosomal nanoparticles: photothermally induced ultrasmall lesions (left), activation (or inhibition) of Ca²⁺ signals (middle), or luminal Ca²⁺ sensing (right).

Figure 19

Magnetomechanical actuation with an m-Torquer nanoparticle for remote control of electrical activity in genetically engineered neurons. (a) Mechanical torque force generation by an m-Torquer nanoparticle under a rotating magnetic field for the activation of a mechanosensitive ion channel, Piezo1, to transduce Ca²⁺ eliciting action potentials. (b) Long working ranges under constant magnetic field, |B|, suitable for large-scale in vivo studies. (c) Temporal control of Piezo1 activation in mouse cortical neurons via magnetic fields. (d) Mice motional behavior control by m-Torquers injected into the motor cortex M1 region of mouse brain. The motion trajectory shows the increased movements of a mouse treated with Piezo1 and m-Torquers with circular magnetic array (CMA) rotation. Figure adopted with permission from Lee et al. Copyright 2021 Springer Nature. (e) Neuron-type specific magnetic stimulations for neuronal circuitry control. Stimulation of glutamatergic and GABAergic neurons in lateral hypothalamus induces increase and decrease of food foraging behavior in mice, respectively. Figure adopted with permission from Choi et al. Copyright 2024 Springer Nature.

Figure 20

Quantum dot (QD)-based voltage sensing. (a) (i) Schematic of quasi type-II CdSe-seeded CdS QD nanorods decorated with amphipathic peptides. Rigid, lipophilic, α-helical regions are oriented parallel to the rod-shaped nanoparticle, whereas flexible, hydrophilic regions are oriented to and cap the nanorod ends. (ii) Schematic (top, center) depicting the potential membrane bilayer insertion orientations and cryo-scanning electron micrographs (bottom) of the NPs inserted into small unilamellar vesicles. The scale bars correspond to 30 nm. (iii) Spatially high-pass-filtered image of human embryonic kidney cells containing inserted rod-shaped NPs (left) and representative temporal, bandpass-filtered traces of changes in fluorescence normalized to the steady-state fluorescence (ΔF/F₀) for each region of interest from (right). Bolded traces are the means of the 19 overlaid traces. The scale bar indicates 10 μm. Image taken with permission from Park et al. Copyright 2018 The Authors. (b) (i) Schematics depicting the mechanism of action of QD voltage sensing using electron transfer. At resting potential, the QD is bright and becomes dimmer upon membrane depolarization. (ii) Schematic of CdSe/CdS/ZnS core/multishell QDs conjugated to peptide-fullerene (C₆₀) with the corresponding peptides tested. (iii) (left) Representative frame of mouse cortices injected with QD-JBD1-C₆₀ conjugates. Tungsten electrodes are depicted by white lines. (right) Time-resolved ΔF/F₀ traces at the various regions of interest from (left, A1–A4) showing the response of QD-JBD1-C₆₀ to electrical stimulation as an average of 50 trials. Figure taken with permission from Nag et al. Copyright 2017 American Chemical Society.

Figure 21

(a) Ca²⁺ channels are located in the plasma membrane (PM), the membrane of the endoplasmatic reticulum (ER), and the membrane of mitochondria (Mito). (b) For observing Ca²⁺via Ca²⁺-responsive fluorophores, optical detection is in general performed with confocal microscopy. In the case of imaging fluorophores close to the glass substrate, total internal reflection (TIRF) geometries can be used. (c) Sketch of a nanoparticle (NP) with attached Ca²⁺-responsive fluorophores (PEGCaRuby) and cell-penetrating peptides (CPPs) and (d) corresponding normalized fluorescence signal F/F₀ upon stimulation as recorded from the NPs localized in the two cells shown below. Image adopted from Zamaleeva et al. Copyright 2014 American Chemical Society.

Figure 22

Electronic small-molecule detection using aptamer-functionalized field-effect transistor (FET) sensors. (a) Responses of FET sensors functionalized with a dopamine aptamer (K_d = 150 nM, 1× PBS) or its scrambled sequence as a control compared to FET responses with a previously known dopamine aptamer (K_d = 1 μM, 0.1× PBS). (b) The dopamine aptamer–FET and scrambled sequence control responses to dopamine in 1× artificial cerebrospinal fluid (aCSF). (c) For serotonin aptamer–FETs, serotonin in 1× aCSF led to concentration-dependent responses, whereas scrambled serotonin control sequences showed negligible responses. (d) Dopamine aptamer–FET responses to 100 μM norepinephrine, serotonin, l-3,4-dihydroxyphenylalanine (-DOPA), and 3,4-dihydroxyphenylacetic acid (DOPAC) were negligible relative to dopamine (10 nM). (e) Serotonin aptamer–FET responses to 100 μM dopamine, norepinephrine, serotonin biological precursor l-5-hydroxytryptophan (l-5-HTP), or serotonin metabolite 5-hydroxyindoleacetic acid (5-HIAA) were negligible relative to serotonin (10 nM). (f) By altering ratios of amine-terminated/methyl-terminated silanes for surface tethering, serotonin aptamer–FET sensitivity ranges were shifted. (g) Serotonin aptamer–FETs after 1 to 4 h of incubation in serotonin-free brain tissue followed by addition of serotonin had reproducible responses with differentiable physiological concentrations. (h) Sphingosine-1-phosphate (S1P) aptamer–FETs showed concentration-dependent responses to S1P but not to a phospholipid with similar epitopes or a scrambled control sequence in 1× HEPES. (i) In tests of glucose sensing in 1× Ringer’s buffer, the responses of glucose aptamer–FETs were minimal for galactose, fructose, and a scrambled control sequence. (j) Concentration curves for glucose aptamer–FET responses in mouse whole blood diluted in Ringer’s. The red circle shows the response in undiluted whole blood. (k) Glucose aptamer–FETs were able to differentiate hyperglycemia in serotonin transporter–deficient (KO) mice versus wild-type (WT) mice by measuring glucose concentrations in diluted serum under basal and glucose-challenged conditions. Error bars are ±SEM with N = 6 [(a–c, h, i, and k)] or N = 3 samples per group [(d–g and j)]. In (d,e), ***P < 0.001 versus countertargets; in (g), ***P < 0.001, *P < 0.05 versus different serotonin concentrations (10 pM to 100 nM); in (k), **P < 0.01 KO versus WT. Figure used with permission from Nakatsuka et al. Copyright 2018 The Authors.

Figure 23

Scales of enzyme-based field-effect transistors (EnFETs) versus antibody-based FET sensors (ImmunoFETs) versus aptamer-FETs compared to the Debye length in vivo over which charge is screened because of the high ionic strength in the brain. EnFETs and ImmunoFETs are more commonly used in the laboratory, where extracted solutions can be diluted and/or desalted. The aptamer-FETs can be used both in vivo and in vitro.

Figure 24

(a) Schematic and (b) digital photographs of an in vivo experiment where a neuroprobe, Ag/AgCl reference electrode and stimulator were implanted into the brain of a head-fixed mouse. (c) Schematic of the stimulation and recording sites. The stimulating electrode was implanted into the serotonin cell body region, and the neuroprobe was implanted into a serotonin terminal region in the striatum. (d) Three consecutive overlapping output sweeps in vivo where gate-source voltage (V_GS) was swept while source-drain voltage V_DS was held at constant. (e) Calibrated responses and (f) areas under the curves for in vivo measurements of basal and postelectrical stimulation levels from the same mouse, respectively. Error bars in (e) and (f) are standard error of the mean. **P < 0.01 versus basal. Used with permission from Zhao et al. Copyright 2021 The Authors.

Figure 25

Nanosensor arrays used for chemical imaging: (a) fluorescent single-walled carbon nanotubes (SWCNTs) are made responsive to dopamine by attaching specific single-stranded DNA sequences (ssDNA) to them via noncovalent bonding. These nanotubes are then fixed onto a glass substrate used to culture dopamine-releasing PC12 cells on top. Upon stimulation, PC12 cells release dopamine and the fluorescence of SWCNTs changes. (b) Increase in fluorescence intensity of a single (GA)15-ssDNA/SWCNT (GA = guanine, adenine) induced by dopamine addition (10 μM). (c) The proposed sensing mechanism involves dopamine-pulling phosphate groups toward the SWCNT surface, resulting in the elimination of quenching sites and an enhancement in SWCNT fluorescence quantum yield (molecular dynamics, MD, simulations). Image adopted with permission from Kruss et al. Copyright 2017 National Academy of Sciences.

Figure 26

Adsorbed nanosensors detecting release of dopamine (AndromeDA) functions as a dopamine (DA) sensor. (a) A cultured DAergic neuron is coated with AndromeDA paint, which detects DA released upon neuronal stimulation. The interaction of DA with the paint leads to an elevation in nanosensor fluorescence, enabling the detection of the spatiotemporal pattern of DA release and diffusion. (b) Each of the nanosensors utilized in AndromeDA comprises a (5,6)-SWCNT-(GT)₁₀ complex (SWCNT = single-walled carbon nanotube, GT = guanine thymine). (c) Left: AndromeDA is composed of a dense layer of individual nanosensors, as visualized through atomic force microscopy (AFM). Right: A magnified view of a single nanosensor is displayed from a lower density nanosensor preparation. (d) Left: A magnified view of an endogenous green fluorescence protein (EGFP)-positive axon with a single varicosity is shown. Right: AndromeDA fluorescence is observed at different time points in the same field of view. Initially, the near-infrared (NIR) fluorescence is low, reflecting the absence of extracellular DA (labeled as Basal). Upon neuronal stimulation, a transient AndromeDA hotspot emerges adjacent to the varicosity (labeled as Hotspot). As DA diffuses, AndromeDA becomes activated over a broader area, leading to a more generalized increase in NIR fluorescence (labeled as Diffusion). Below: A side-view schematic illustrates a DAergic varicosity surrounded by AndromeDA on the glass coverslip (left), and a fluorescence trace (right) illustrates the NIR fluorescence change associated with the hotspot image above it. Taken with permission from Elizarova et al. Copyright 2022 National Academy of Sciences.

Figure 27

Near-infrared (NIR) catecholamine (nIRCat) sensor method is employed to detect striatal dopamine (DA) release induced by optogenetic stimulation. (a) A schematic illustrates channelrhodopsin-2 (ChR2) expression in cortical glutamatergic terminals forming synaptic contacts in the dorsal striatum. The abbreviations AMPA (α-amino-3-hydroxyl-5-methyl-4-isoxazolepropionate), NMDA (N-methyl-d-aspartate), and DAR (DA receptor) are used. (b) Stimulation of glutamatergic terminals did not result in any nIRCat fluorescence modulation. Confirmation of glutamate (GLU) release was achieved through excitatory postsynaptic current (EPSC) recordings on MSN (medium spiny neurons). (c) ChR2 expression is schematically illustrated in nigrostriatal dopaminergic terminals of the dorsal striatum. (d) Stimulation of dopaminergic terminals led to nIRCat fluorescence modulation. The specific stimulation protocol in (b) involved five pulses (5P) at 25 Hz with a power flux of 1 mW/mm², and each pulse lasted for 5 ms. Image reproduced with permission from Beyene et al. Copyright 2019 The Authors.

Figure 28

Acetylcholine (ACh) nanosensors’ structure and detection mechanism. The nanosensors are directed to ACh receptors by conjugation with fluorescent alpha-bungarotoxin (BTX). AChE, connected to the DNA scaffold, catalyzes the hydrolysis of ACh, leading to reduction in the local pH due to the production of acetic acid. Four pH-sensitive pHAb fluorophores are located near AChE, causing an increase in fluorescence emission when ACh is hydrolyzed. Alexa fluorophore 647 (AF647) attached to the BTX serves as an internal fluorescence standard, facilitating quantitative measurements. Taken with permission from ref (628). Copyright 2021 National Academy of Sciences.

Figure 29

(a) Surface-enhanced Raman spectroscopy (SERS) of acetylcholine at different concentrations and calibration curves for the SERS quantification, down to the attomolar regime, of diverse neurotransmitters. Reproduced with permission from Lee et al. Copyright 2021 The Author(s). (b) Schematic diagram illustrating the method to detect dopamine (DA) release from single live undifferentiated/differentiated neuronal stem cells (NSCs) using graphene oxide (GO)–hybrid SERS. (i) Schematic diagram depicting a strategy to detect DA released from single neuronal stem cells, which were differentiating to neurons for 20 days on a composite consisting of gold nanostructures coated with graphene oxide. (ii) Representative immunofluorescence images of the undifferentiated/differentiated NSCs from day 0 to 20 after induction of differentiation. Scale 50 μm. (iii) Representative SERS images corresponding to (ii) at 830 cm^–1 (malachite green). The dotted lines indicate the boundary of the cells. Scale bar 5 μm. Reproduced with permission from Choi et al. Copyright 2020 American Chemical Society. (c) The principle of the SERS probe for the simultaneous biosensing of carbonate concentration and pH in live brains and single neurons. Scanning electron microscope (SEM) images of the functionalized gold-coated tips introduced in the cortex of mice. Reproduced with permission from Wang et al. Copyright 2019 Wiley-VCH Verlag GmbH.

Figure 30

Organotypic slice culture of rat hippocampus. (a) Result of gene gun transfection with endogenous green fluorescence protein (EGFP). No scale bars available. Adapted with permission from Holbro et al. Copyright 2009 National Academy of Sciences. (b) Single-cell electroporation of CA1 pyramidal cells with a genetically encoded calcium sensor. No scale bars available. Adopted with permission from Wiegert et al. Copyright 2013 National Academy of Sciences.

Figure 31

Time-controlled spiking of two neuronal populations in organotypic culture. (a) Hippocampal slice culture with CA3 neurons expressing ChrimsonR (magenta) and CA1 neurons expressing CheRiff (green). Scale bar: 500 μm. (b) Photocurrent amplitudes measured at different wavelengths and intensities (1 ms light pulses). Typical responses to 405 and 625 nm light pulses (current clamp) are plotted below. Figure adopted from Anisimova et al. Copyright 2022 The Author(s).

Figure 32

Uptake of ultrasmall carboxyfluorescein (FAM)-labeled gold nanoparticles (Au-Click-FAM; green fluorescence) and dissolved FAM-alkyne by six-cell brain organoids over 30 min, 6 h, and 24 h. Scale bars 200 μm. Reproduced with permission from Sokolova et al. Copyright 2020 The Author(s).

Figure 33

Methodology and examples of data obtained for gated luminescence imaging of Si nanoparticles (GLISiN) in mouse brain tissue compared with steady-state imaging. (a) Schematics showing the instrumental setup. The intensified charge-coupled device (iCCD) camera and the light source were controlled by an external pulse generator. In the case of laser illumination, the laser fired under control of the laser’s internal pulse generator, and the camera was configured to slave to it via transistor–transistor logic (TTL) digital trigger. (b) Notional waveforms for illumination and camera gating used to acquire images. The light emitting diode (LED) was triggered “ON” by the pulse generator, maintained in the “ON” position for the duration of “Gate width,” and then image acquisition terminated (“CLOSE”) at the end of the “Gate width” period. For the laser experiments, the laser fired at the beginning of “Gate width” but was only “ON” for the duration of the natural pulse width of the laser (∼8 ns). For GLISiN imaging (“Gated”), the camera was preprogrammed to energize the intensifier screen (“OPEN”) at a time delayed by “Gate delay” relative to the end of the excitation pulse. For continuous wave imaging, the camera was again programmed to be “OPEN” for the “Gate width” period, but the “Gate width” period overlapped with the laser or LED excitation pulse to generate a pseudosteady-state measurement. (c) Digital color photograph (from an iPhone 5, Apple Inc.) and (d) grayscale image (from an Andor iCCD) of mouse brain obtained under ambient light. (e) Continuous wave and (f) GLISiN images of the same brain under UV LED excitation (λ_ex= 365 nm, λ_em= 460 nm long-pass filter; gate width, 400 μs, 40 accumulations; gate delay for continuous wave = 0 μs, gate delay for GLISiN = 5 μs). Phantom samples corresponding to 150 ng of porous Si NPs (PSiNP) and 2.5 ng of the molecular dye Alexa Fluor 647 (“AF647”) were added next to the brain for comparison, as indicated. Note that the signals from the AF647 sample (fluorescence) and the brain tissue (autofluorescence), readily visible at steady state (e), almost disappear in the GLISiN image (f), whereas the longer-lived luminescence from PSiNP is much stronger in the GLISiN image. (g) Normalized intensity decay of the photoluminescence/fluorescence signals from the samples in (c–f) as a function of time after excitation pulse (gate width, 10 μs; gate step increase, 10 μs; accumulation, 20 times). Note the nanosecond decay times of the organic dye and tissue autofluorescence are too short to be resolved at the measurement time scale. The orange box depicts the “Gate width” window used to obtain GLISiN images in (f). Reprinted (adapted) with permission from Joo et al. Copyright 2015 American Chemical Society.

Figure 34

Scanning electron microscopy (SEM) images of devices utilizing spatial restriction in 2D, 2.5D, and 3D during cell culturing to control neuronal growth. (a) 2D pathways are defined by rolled-up GaAs/InGaAs microtubes. (b) Cavities and grooves with steps (arrows) are prepared by photolithography and reactive ion etching, defining 2.5D pathways for neuronal guiding. (c) Direct laser writing (DLW)-prepared scaffold structure with towerlike cavities connected by free-standing tunnels. The scale bars represent 50 μm. Original images are modified with permission from (a) Bausch et al. Copyright 2013 AIP Publishing; (b) Fendler et al. Copyright 2019 Wiley VCH Verlag; (c) Fendler et al. Copyright 2020 Royal Society of Chemistry.

Figure 35

Schematic of the direct laser writing (DLW) process for fabricating a 3D scaffold structure for neuronal guidance. (a) A substrate, here a glass coverslip, with a droplet of liquid resin on top is loaded into the DLW setup. (b) Within the focal spot of a pulsed fs-laser, the resin is polymerized. The laser focus point can be moved in all dimensions through the polymeric resin leading to a 3D-defined object, (c) which is still covered with the liquid resin. (d) Developing and cleaning leads to a free-standing object on the carrier substrate. This image has been taken with permission from the Ph.D. thesis of C. Fendler, 2019.

Figure 36

(a) Confocal microscopy images of murine cerebellar granule neurons at 10 DIV in a direct laser writing (DLW)-printed scaffold overcoated with Al₂O₃ and internally functionalized with poly-d-lysine. (b) Recorded action potential (AP) of a murine cerebellar granule neuron at 10 DIV inside the scaffold. (c) An example of a patch-clamping experiment on a human-induced pluripotent stem cell-derived neuron grown inside the scaffold. The pipet (blue) is approaching the cell (green) from the left. (d) Trace of excitatory postsynaptic current (EPSC) events and magnified image of a single event. The scale bars represent 20 μm. These images have been taken and modified with permission from Fendler et al. (Copyright 2019 Wiley VCH Verlag) and Harberts et al. (Copyright 2020 American Chemical Society).

Figure 37

Sketch of the concept of a human–machine interface (HMI).

Figure 38

Electroencephalography (EEG) setup to control external devices showing a test person controlling a roboter arm via EEG communication (left) and the noninvasive high-density EEG-montage (right). Note the additional red–green–blue/corresponding depth (RGB-D) camera helping in interpreting the EEG signals. The left image is taken from Schröer et al. Copyright 2015 IEEE. The right image is Copyright 2024 Enker, Uniklinik Düsseldorf.

Figure 39

Modes of interception of neuronal activity. (a) Extracellular potential ΔΦ, (b) change in fluorescence (of genetically transfected neurons) ΔF/F₀, (c) magnetic fields ΔB, and (d) concentration changes Δc due to metabolic activity.

Figure 40

Neuronal activity can be stimulated by (a) electrical means (applied voltages Φ or currents I); (b) optical illumination F, leading to membrane depolarization, heating, or direct opening of light-gated ion channels; (c) inductive activation with oscillating magnetic fields B(t) or magnetothermal heating with NPs; (d) ultrasound activation leading to mechanical displacement Δx; and (e) chemical stimulation Δcvia perfused neurotransmitters.

Figure 41

Flexible, human-scale graphene-based microelectrocorticography (μ-ECoG) device for clinical investigations. (a) Digital photograph of the graphene-based device developed by INBRAIN (top) in comparison to the clinically used ECoG metal-based strip. On the right, high magnification of one stimulating electrode contact consisting of hundreds of 25 μm graphene membranes. (b) Schematic representation of the device position on the magnetic resonance imaging (MRI)-generated image of the motor and somatosensory cortex of ovine (sheep) brain. On the right is a digital photograph of the transparent and thin-film device placed epicortically on a sheep brain. (c) Illustration of the difference in conformity between reduced graphene oxide (rGO) cortical electrodes on thin film technology (20 μm thick polyimide) compared to a silicone-based, clinically used ECoG strip. Copyright 2024 INBRAIN Neuroelectronics.

Figure 42

Investigation of axonal, myelin, and brain structure with small-angle X-ray scattering (SAXS) tensor tomography. (a) Experimental setup with an X-ray microbeam that is scanned across a brain section. (b) False-color-coded orientation of the measured local 3D orientation of the nerve fibers with the color representing the in-plane orientation.

Figure 43

Signal flow for AI/ML-controlled neuronal prostheses.