Stretchable transparent electrode array for simultaneous electrical and

dynamics of perturbed neural populations at the network level,. 5-9 while simultaneous and co-localized electrophysiology measurement and optical imag...
0 downloads 9 Views 1MB Size
Subscriber access provided by Kaohsiung Medical University

Stretchable transparent electrode array for simultaneous electrical and optical interrogation of neural circuits in vivo Jing Zhang, Xiaojun Liu, Wenjing Xu, Wenhan Luo, Ming Li, Fangbing Chu, Lu Xu, Anyuan Cao, Ji-Song Guan, Shiming Tang, and Xiaojie Duan Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.8b00087 • Publication Date (Web): 02 Apr 2018 Downloaded from http://pubs.acs.org on April 2, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Stretchable transparent electrode array for simultaneous electrical and optical interrogation of neural circuits in vivo Jing Zhang1,2,3,4, Xiaojun Liu1,4, Wenjing Xu5, Wenhan Luo2,3,6, Ming Li2,3,7, Fangbing Chu1,4, Lu Xu5, Anyuan Cao5, Jisong Guan6,8, Shiming Tang2,3,7, Xiaojie Duan1,3,4* 1

Department of Biomedical Engineering, College of Engineering, Peking University, 2

Beijing 100871, China.

Peking-Tsinghua Center for Life Sciences, Peking

University, Beijing 100871, China. 3Academy for Advanced Interdisciplinary Studies, Peking University, Beijing 100871, China. 4Center for Nanochemistry, Beijing Science and Engineering Center for Nanocarbons, Peking University, Beijing 100871, China. 5Department of Materials Science and Engineering, College of Engineering, Peking University, Beijing 100871, China. 6School of Life Sciences, Tsinghua University, Beijing 100084, China. Beijing 100871, China.

8

7

School of Life Sciences, Peking University,

School of Life Science and Technology, ShanghaiTech

University, Shanghai, 201210, China *Correspondence and requests for materials should be addressed to X. D. (email: [email protected]). Keywords: Stretchable electronics, transparent electrode, neural electrode array, brain activity mapping, optogenetics, calcium imaging, traumatic brain injury, nano-bio interface, carbon nanotube

1 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 25

Abstract Recent developments of transparent electrode arrays provide unique capability for simultaneous optical and electrical interrogation of neural circuits in brain. However, none of these electrode arrays possess the stretchability highly desired to interface with mechanically active neural systems, such as brain under injury, spinal cord and peripheral nervous system (PNS). Here we report a stretchable transparent electrode array from carbon nanotube (CNT) web-like thin films that retain excellent electrochemical performance and broad-band optical transparency under stretching, and are highly durable under cyclic stretching deformation. We showed that the CNT electrodes recorded well-defined neuronal response signals with negligible lightinduced artifacts from cortical surfaces under optogenetic stimulation. Simultaneous two-photon calcium imaging through the transparent CNT electrodes from cortical surfaces of GCaMP-expressing mice under epilepsy showed individual activated neurons in brain regions where the concurrent electrical recording was taken from, thus providing complementary cellular information in addition to the high temporal resolution electrical recording. Notably, the studies on rats show that the CNT electrodes remained operational during and after brain contusion that involves rapid deformation of both the electrode array and brain tissue. This enabled real-time, continuous electrophysiological monitoring of cortical activity under traumatic brain injury. These results highlight the potential application of the stretchable transparent CNT electrode array in combining electrical and optical modalities to study neural circuits especially under mechanically active conditions, which could potentially provide important new insights into the local circuit dynamics of spinal cord and PNS, as well as the mechanism underlying traumatic injuries of the nervous system.

2 ACS Paragon Plus Environment

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Combining optical and electrical modalities in neural interfacing is important to study the connectivity and function of neural circuits.1-5 Optogenetic stimulation with simultaneous multichannel electrophysiological readout enables mapping the dynamics of perturbed neural populations at the network level,5-9 while simultaneous and co-localized electrophysiology measurement and optical imaging of neural structure and function could leverage the spatial and temporal resolution advantages of both techniques, providing new insights into how neural circuits process information.10-12 Recent progress on direct optical stimulation of unmodified neurons provided an alternative to optogenetics and potentially for therapies involving neuronal photostimulation.13-15 Simultaneous electrophysiological recording with direct optical stimulation is important to validate and optimize these stimulation strategies and reveal the underlying mechanism. Conventional neural surface electrode arrays using opaque metal conductors are not suitable to be used in simultaneous electrical and optical neural interfacing, because they block the field of view and are prone to producing light-induced artifacts in the electrical recordings.16 Recent efforts of creating transparent electrode arrays based on graphene and indium tin oxide (ITO) films etc., enabled efficient light delivery through the electrodes and recording of electrophysiological signals concurrently with optical stimulation and imaging.17-19 The flexible nature of graphene based microelectrode array is especially attractive because it enables conformal and compliant interfacing with the soft, curvilinear surfaces of the brain tissue.17, 19, 20 However, these electrode arrays are not compliant

enough

to

tolerate

large

mechanical

deformations

that

occur

physiologically in many neural tissues. Developing electrode array sufficiently stretchable to adapt to various motions while maintaining reliable operation and stable contact with tissues is important to study the mechanically active neural systems. In humans, both the spinal cord and its meningeal protective membranes can experience as much as 10-20% tensile strain and displacement (relative to the spinal canal) during normal postural movements.21, 22 With each beat of the heart or due to movement of the head, the brain experiences up to 5% bulk strain,23,

24

and peripheral nerves may experience up to 15% strain.25

Except for these naturally happened deformations, brain tissue under traumatic brain injury (TBI) also experiences rapid deformation, including stretch, compression, and shear strain etc.26 Stretchable electrode array can move and deform with the tissue,

3 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 25

while maintaining their functionality and recording capabilities. This enables realtime, continuous electrophysiological recording from the same location on the nervous tissues when they are subjected to natural motion during daily activities or under injury, which is not achievable by conventional unstretchable electrode array. In this work, we have created a stretchable transparent electrode array from web-like carbon nanotube (CNT) thin film for use in simultaneous electrical and optical neural interfacing under mechanically active environments (Fig. 1A). The stretchable transparent CNT electrode retained high optical transparency over a broad wavelength spectrum and excellent electrochemical performance under stretching and can sustain multiple cycles of mechanical stretch. We verified the effectiveness of the CNT electrode array in electro-optic neural interfacing by performing simultaneous optical stimulation/imaging and electrical recording of cortical activities in rodent models. The ability of the CNT electrode array to operate under mechanically active environments was demonstrated by real-time, continuous monitoring of cortical activity prior to, during and after TBI from rats. Our results highlight the potential application of the stretchable transparent CNT electrode array for combining the optical and electrical modalities in studies of neural circuits of mechanically active neural systems, such as spinal cord, peripheral nervous system (PNS), and brain under injury etc. The stretchable transparent electrode array we developed integrates an elastic, transparent silicone substrate (polydimethylsiloxane, PDMS, 100 µm in thickness), CNT thin film interconnects and recording sites, and a top patterned SU-8 layer for interconnects insulation, as shown in Fig. 1A. The CNT thin film was prepared through floating catalyst chemical vapor deposition (CVD) and a solvent-induced condensation process (see Methods for details).27 This gives a web-like CNT film, consisting of strongly interconnected thin bundles (10-20 nm diameter) of singlewalled carbon nanotubes (SWNTs, 1-2 nm diameter) with high homogeneity and purity over large area (Fig. 1B).27 The void area in CNT film gives high optical transparency while the low tube-to-tube contact resistance, resulted from the strong interconnection between nanotubes, ensures high electrical conductivity. Plots of the light transmittance versus wavelength for bare PDMS and CNT film on PDMS substrate are shown in Fig. 1C. The CNT/PDMS complex maintained a transmittance over 85% across a wavelength range of 400 nm to 2.5 µm. This permits efficient light

4 ACS Paragon Plus Environment

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

delivery through the CNT electrodes. Importantly, the CNT/PDMS complex retained high optical transparency when being stretched to a strain of 20% (the biologically relevant strain happened in neural systems) (Fig. 1D), ensuring its application under mechanically active environments. By changing film thickness using different deposition parameters, the optical transmittance and electrical conductivity of the CNT film can be tuned. Thicker film possesses higher electrical conductivity but lower optical transparency.27 In this work, we used CNT film with transmittance at 550 nm in the range of 70% to 90%, which corresponds to film sheet resistance in the range of 150-500 Ω/sq. Except for the high optical transparency and electrical conductivity, the CNT thin film prepared this way showed superior mechanical strength and robustness, which can be handled in freestanding form and easily transferred to a variety of substrates.27 Furthermore, the entangled web-like structures could maintain their conductive pathways under mechanical deformation by reorientation and re-positioning of CNTs within the percolation networks, as schematically shown in Fig. 1E.28, 29 All of these make the as-prepared CNT web-like thin film an excellent candidate as conducting materials for stretchable transparent neural electrode array. The fabrication process of the CNT stretchable transparent electrode array involved patterning of the CNT film and SU-8 insulation layer on a sacrificial substrate, and subsequent coating of PDMS film on the CNT/SU-8 complex followed by removing the sacrificial substrate (Fig. S1). No opaque metal is needed for the final electrode array. A flexible cable connected to the CNT film on the electrode array was used to send the recorded signals to the outside preamplifier (Methods). As shown in Fig. 1F, the final electrode array exhibited high optical transparency in the whole electrode area, including the recording sites and entire connecting lines. We characterized the electrochemical performance of the CNT stretchable transparent electrodes and compared it to electrodes made of single-layer graphene. The graphene electrode array was designed same as the CNT electrode array, except for that an additional Au layer is applied on top of the graphene layer, covering the pads and partial portions of the connecting lines (Fig. S2). About 5 mm region from the recording sites was left only with graphene and SU-8 insulation layer to ensure optical transparency at and around the recording sites. This Au layer is necessary to ensure a

5 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 25

high yield of functional graphene electrodes in the array, and also is important for a reliable connection to the flexible cable used for reading the brain signals into the preamplifier, as will be discussed later. The CNT electrodes displayed lower impedance than the graphene electrodes with same recording site size of 100 × 100 µm2 (0.20 ±0.03 MΩ vs. 1.68±0.68 MΩ at 1 kHz, Fig. 2A) and larger cathodal charge storage capacity as calculated from the cyclic voltammogram (8.71 ± 3.3 mC/cm2 vs. 4.26 ± 1.6 mC/cm2, Fig. 2B), indicating improved electrochemical interfacial properties of the CNT thin film compared to graphene. To test the stretchability needed for use under mechanically active environments, we monitored the electrode impedance under increasing stretch load along the direction parallel to the electrode traces. The impedance of the CNT electrodes remained unchanged when the uniaxial tensile strain was below 5%, and then exhibited ~26% and ~55% increase upon a tensile strain of 20% and 50%. Differently, the impedance of graphene electrodes increased steeply under stretching, with over 10× increase upon a strain of ~3% (Fig. 2C). This high impedance under ~3% strain suggests electrochemical failure of the graphene electrodes as neural electrodes. Besides, the impedance spectra of the CNT electrodes showed great reversibility at stretching load of 50%, while for the graphene electrodes, in order for the impedance spectra to be reversible, the stretching loads can’t go beyond 0.9% (Fig. 2D). The change of cyclic voltammogram of the CNT electrodes was also modest under uniaxial strain of 20% and 50% (Fig. 2E). To demonstrate the robustness of the CNT electrodes against cyclic deformation experienced by many mechanically active neural tissues, we stretched the CNT electrodes to 20% strain over 10,000 cycles. The CNT electrodes withstood the cyclic deformation, displaying minimal variation in impedance over time (Fig. 2F). A minimum in impedance at ~2000 cycles of stretching was shown and also observed in three other tests, which we attribute to a period in which the nanotube bundles adopted their optimum morphology.29 These results demonstrate excellent interfacial property, mechanical robustness and stretchability of the CNT electrodes, fulfilling the challenging requirements for highresolution electrophysiological recordings under mechanically active environments. Due to its unique properties, graphene has recently been exploited as conducting materials in neural electrodes.30 With broad-spectrum transparency, graphene

6 ACS Paragon Plus Environment

Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

electrodes have successfully recorded ECoG response of the brain tissue to optogenetic stimulation.17,

20

Co-localized and simultaneous electrophysiological

recording and calcium imaging from hippocampal slices was also demonstrated using graphene electrodes.19 However, despite their superb optical and electrical properties, the graphene electrodes have limited strain sustainability. It was reported that a stretching induced tensile strain over 6% causes mechanical failure of graphene due to the fractures within the unique honeycomb structure.31 The lack of stretchability restricts the use of transparent graphene-based neural electrode arrays in mechanically active neural systems. Besides, the low mechanical/electrical reliability resulted from the limited strain sustainability makes the graphene vulnerable to damage during the fabrication and implantation processes. Previous studies used multi-layer graphene,17 in combination with the use of metal connection traces,17, 19, 20 to improve the success rate of the electrode fabrication and implantation. The use of multi-layer graphene decreases the cost- and time-effectiveness and the metal lead lines are undesirable because they make up opaque areas especially for high-density electrode array.17 Besides, the metal-graphene junction also provides another source for light-induced artifact under optogenetic stimulation.32, 33 The stretchability of graphene electrodes can be achieved by imparting some structural configurations to graphene films, such as wrinkled34 or crumpled35 structures. However, the wrinkled/crumpled graphene is not suitable for application of stretchable transparent neural electrodes because they are not mechanically stable and their optical transparency changes upon strain.34, 35 ITO was also used to fabricate ECoG electrodes for simultaneous electrical readout and optical stimulation.18 However, the brittle nature of ITO makes it unsustainable to deformation and precludes its use in stretchable transparent electrode arrays. The CNT electrodes maintained excellent electrochemical performances under a strain of up to 50% and during 10000 cycles of 20% stretching. The percolation network of CNTs plays a critical role in enduring mechanical deformation, because the strain can be accommodated by structural reconfiguration through rotating or sliding of CNTs against each other.36 The impedance of the CNT electrodes didn’t show substantial change when subjected to stretching, indicating that most of the CNTs and CNT connection junctions remained without fracture or breaking. The strong interlocking between neighboring CNT bundles is important for maintaining their connection junctions. The reversibility of the impedance change indicated re-

7 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 25

establishment of the CNT percolation network under strain release. It should be noted that due to the much smaller impedance of the CNT electrodes compared to the input impedance of most commercial preamplifiers, the ~26% increase of impedance under a tensile strain of 20% (the maximum biologically relevant strain for neural systems) should not cause a noticeable effect on amplitude or shape of recorded signals. This is important for continuous electrophysiological monitoring of neural activity under deformation. The CNT electrodes maintained a high optical transmission in the wavelength range of 400 nm to 2.5 µm under a strain of 20%, ensuring their capability of efficient light delivery through the electrodes under deformation. Besides, the CNT electrodes exhibited lower impedance than the graphene electrodes. This is advantageous for scaling down the electrode size for high spatial resolution electrophysiological mapping of neural activity. The transparent neural surface electrode arrays allow for delivery of light stimuli through the array directly to the brain region from which the electrical recordings are obtained, this enables unique research capabilities in combining optogenetics with electrophysiology.17, 20 With added stretchability, the transparent electrode array thus can find wide applications in deciphering neural circuits of mechanically active systems, such as the spinal cord of freely moving animals. We conducted concurrent optogenetic stimulation and neural recording from Thy1-ChR2-YFP mice using the CNT stretchable transparent electrode array. In a typical recording, a 4 × 4 CNT electrode array was placed over the retrosplenial cortex of a Thy1-ChR2-YFP mouse to record ECoG signals, under optogenetic stimulation of the exposed cortical surface with an optical fiber delivering blue light stimuli (488 nm) (Fig. 3A). The CNT electrode array closely and conformly adhered to the brain surface owing to its high compliance with soft tissues. The optical stimulation produced readily defined, large negative potential peaks in ECoG recording. These negative potentials persisted for tens of milliseconds after the light pulse turned off, and were highly reliable with similar amplitude being produced under each stimulus (Fig. 3B, C). No response was evoked in wild type (WT) mice under the same stimulation (Fig. 3C). This suggests that the negative potential originates from the ChR2 activation. The activation of ChR2 causes a net flux of cations into cells, which results in the depolarization of neurons, leaving behind a net negative charge.37 The normalized time-frequency spectral analysis shows a clear increase in power at the stimulation frequency during

8 ACS Paragon Plus Environment

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

the stimulation epochs. There were also concomitant increases of power at harmonics of the stimulation frequency (Fig. 3D). We suspect that these increases of power at harmonics were a result of the Fourier decomposition of the response waveform, rather than originating from a separate neuronal process or response.38 The evoked potentials were spatially uniform across different channels under photo-stimulation of the entire cortex window, and the amplitude of the signal depends on both the duration and intensity of the stimulus, with larger amplitude potentials produced under longer and stronger stimuli (Fig. 3E, F). Close-up views of the evoked responses from one representative channel under different stimulus duration and intensity is shown in Fig. S3. For simultaneous optogenetic modulation and electrophysiological recording, lightinduced artifact is a potential issue because it contaminates the neuronal response signals.16, 39 Because the photon energy of the stimulation light is normally below the work function of the electrode materials, the photoelectric effect was not a major concern. The main mechanism of the stimulus artifact is thought to be the photovoltaic effect (i.e. Becquerel effect), resulting from a differential in photoexcitation between the different conductive layers at a junction.32,

33

We

characterized the optical stimulus artifact produced from the CNT electrodes in saline. Au electrodes with same size of recording site were tested in parallel for comparison (Fig. S4). The artifact amplitude is dependent on the light stimulus power and duration (Fig. S5). For a 15 ms pulse with light intensity of 2.4 mW/mm2, the artifact produced in CNT electrode recording was undetectable (Fig. 3G), compared to the ~175 µV neuronal response signals observed in ChR2 mice under same illumination conditions (Fig. 3B, C). Differently, the Au electrodes showed a pronounced photovoltaic effect with artifact amplitude of ~60 µV under this illumination condition (Fig. 3G). Under increased light intensity, light-induced artifacts can be observed from CNT electrode recording, but even at 10 × intensity, the artifact from CNT electrode recording is still much smaller (-12 µV, Fig. S5). The small lightinduced artifact compared to traditional metal-based electrodes is advantageous for the application of the CNT electrodes in simultaneous optogenetic modulation and electrophysiological recording. The unique optical transmission behavior of SWNTs is critical for the small light-

9 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 25

induced artifact. The Becquerel effect requires photoexcitation of charge carriers into higher energy states.32, 33 SWNTs show Van Hove singularities in their band structure in contrast to continuous density of states (DOS) possessed by conventional threedimensional materials such as metal.40,

41

Dipole-allowed optical transitions only

occur between mirror image peaks in the DOS and crossover transitions are forbidden.42 This makes the photo-absorption and photo-excitation of SWNTs very selective of incident light wavelength and those with photon energy resonant with the electronic transition energy between mirror image peaks in the DOS can be absorbed.43 SWNTs electronic properties exhibit a strong structural dependence. An ensemble normally contains mixed SWNTs with different diameter and chirality [(n,m) values], and only a small fraction can be resonant with a single light wavelength and be photo-excited.41 This would cause a much smaller light-induced artifact under photo-stimulation than their metal counterpart. In recent years, metal nanomeshes have been explored for stretchable transparent electrode fabrication.44, 45 While it is interesting to test their potential application in electro-optic neural interfacing, we expect that a smaller light-induced artifact makes the CNT electrodes more unique and advantageous. The aforementioned results demonstrated the capability of the CNT electrode array in efficient readout of the optogenetic modulation, with the advantage of much smaller light-induced artifact. For future application, by applying complex focal photostimuli varying both in time and space, ECoG signals with high spatial resolution could be recorded with the CNT electrode array across extended cortical areas, allowing extraction of neuronal activation maps that are specific for each site of stimulation. This is not only important for studying the cortical dynamics, but also useful in determining the contribution of specific cell types to the ECoG signal which will further our understanding of the ECoG signal origin. Calcium imaging has the advantage of detecting neural activity at single-cell resolution from large or disperse neuron populations simultaneously, which can be complementary to electrophysiological recording in many applications.46 Besides, imaging calcium signal in the brain region from which the electrophysiological signals are recorded is also important for investigating the basis of the field potential, and evaluating the performance of calcium sensitive proteins etc. To perform

10 ACS Paragon Plus Environment

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

simultaneous transcranial two-photon calcium imaging and ECoG recording, we used transgenic mice that express improved genetically encoded green fluorescent proteinbased calcium sensor GCaMP3 in subsets of excitatory neurons in the brain under the control of the Thy1 promoter.47 In an acute study, a

5 × 5 mm cranial window was

opened over the right hemisphere of an anaesthetized transgenic mouse between Lambda and Bregma to expose the cortex. A 4 × 4 CNT electrode array was placed over the primary visual cortex above the dura mater, and covered by a glass coverslip around same size of the cranial window (Fig. 4A). The electrode array formed an intimate interface with the targeted neural tissues. The high broad-band optical transparency allowed the penetration of the excitation light (920 nm) as well as fluorescence emission light (495-540 nm) through the CNT electrodes, which enabled the calcium imaging of the brain region under the electrodes. Figure 4B shows the transcranial two-photon fluorescent images of the GCaMP3-expressing cortical neurons underneath a CNT recording site at two different depths. The existence of the CNT electrode on the cortex didn’t cause any obvious adverse effect on the twophoton calcium imaging. Individual GCaMP3-expressing neurons are clearly visible and resolvable in both layers, with GCaMP3 expression perimembrane and not detectable in the nucleus. Dynamic calcium imaging can be conducted while simultaneously recording the neural signals with the CNT electrodes. The K+ channel blocker 4-aminopyridine (4AP), which is used to induce seizure,48 was delivered into the cortex of a transgenic mouse at the edge of the glass coverslip, ~ 2 mm away from the electrode array (Fig. 4A). The calcium imaging didn’t cause any additional noise or artifact in the electrophysiological recordings with the CNT electrode. Upon application of 4-AP, the recording firstly showed normal basal brain activity which corresponds to the baseline fluorescence intensity of GCaMP3. Interictal spikes were then observed ~10 min after 4-AP application. Individual activated neurons with calcium transients can be identified in the region underneath the recording site coinciding with the interictallike event (Fig. 4C). The electrical signal recorded by the CNT electrode evolved into seizure-like ictal event, which are characterized by high-frequency, large-amplitude discharges, ~15 min after application of the 4-AP (Fig. 4D). The seizure discharges were generated by synchronously firing of neuronal clusters. We observed this epileptiform discharges from all functional channels in the CNT electrode array (Fig.

11 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 25

S6). As shown from the simultaneous two-photon calcium imaging, the calcium fluorescence within the recording site area (labeled as ROI) exhibited repeated oscillatory increases in intensity with respect to baseline and interictal period, resulting from the multiple dendritic and somatic calcium transients (Fig. 4E). Individual activated neurons with calcium transients can be easily defined in the area under the recording site during the ictal events (Fig. 4E, F). Calcium imaging with single-cell resolution can capture complex network contributions of individual neurons (Fig. 4E, F), while electrophysiological recording give high temporal resolution for detection of high-frequency population discharges (Fig. 4D). The simultaneous calcium imaging and electrical mapping of neuronal activities enabled by the transparent CNT electrode array thus provides an efficient way in leveraging the advantages of both techniques in neural circuits study without perturbing either sensing mode. To show the applicability of the CNT electrode array under mechanically active environments, we conducted continuous, real-time ECoG recording from rat cortical surfaces prior to, during and after TBI. Real-time electrophysiological monitoring of the neural activity under TBI is important for investigating the mechanism underlying the pathophysiology of cerebral contusion. For electrical recording under TBI, while it is possible to injure neural tissue after removing the electrode array, recording from the same locations and immediately after injury is very difficult. Stretchable electrode array lifts these limitations because the electrodes can move and deform with the brain tissues without degrading their recording capability, thus ensuring a stable functional interfacing with the brain tissues before, during, and after injury. In an acute study, a craniotomy was performed on an anaesthetized rat and a CNT electrode array was placed over the left cortical surface above the dura mater. The TBI was induced using a weight-drop model. Briefly, a 40 g rod with diameter of 4.5 mm was dropped from a 7 cm height onto the anaesthetized rat with impact point on the CNT electrode array (Fig. 5A). The resulting injury was relatively mild with no bleeding observed in the brain tissue. The impedance of all the functional channels in the CNT electrode array showed no obvious change, and no degradation of signal-tonoise ratio in the recorded signal was observed after the TBI (Fig. S7), despite the direct strike of the dropped rod on these channels. A representative real-time

12 ACS Paragon Plus Environment

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

recording of the ECoG activity before, during and after this mild TBI is shown in Figure 5B. A suppression of the low-frequency (1-6 Hz) ECoG activity was observed immediately following the injury, which persisted for ~15 s and then gradually recovered to pre-trauma level ~2 min after the injury (Fig. 5C, D). Similar suppression of the low-frequency ECoG activity under TBI was also observed from other channels in the array (Fig. S7). Rapid deformation of the brain tissue under TBI causes rupture of cell membranes and blood vessels, which is accompanied by a variety of effects, including an increase in intracranial pressure, failure of brain ion homeostasis, increased energy metabolism, changes in cerebral blood flow, and suppression of ECoG activity.49,

50

The ECoG suppression we observed here was

relatively minor which may be related with the level of the injury and the effects of anesthesia. For future application, we expect that the continuous, real-time electrophysiological monitoring of brain activity under TBI in awake, or even in freely moving animals using the stretchable CNT electrode array would provide important insights in understanding the pathophysiology of traumatic brain injury. The experiments here highlight the potential application of the CNT stretchable transparent electrode array in combining optical and electrical modalities to study neural circuits, especially that of mechanically active neural systems, including brain under injury, spinal cord, and PNS. TBI is associated with tissue damage, apoptotic and necrotic cell death, and blood-brain-barrier disruption etc. in the lesion. Optical imaging can assess morphological and molecular changes at the cellular level in brain tissue under injury. Combined with simultaneous, continuous electrical recording of neural signals from the injury site, important insights regarding the mechanisms, progression or the treatment of TBI could be provided. Optogenetics offers promise for dissecting the complex neural circuits of the spinal cord and PNS and has therapeutic potential for addressing many clinical needs. Although imposed with unique challenges, optogenetic control in the spinal cord and PNS has been successfully demonstrated.51 On the other hand, soft stretchable electrode array meets the demanding static and dynamic mechanical properties of spinal cord and PNS and has shown stable long-term neural interfacing with limited foreign-body reaction.52 With combined optical transparency and stretchability, the CNT electrode array could enable reliable observation of neural response to optogenetic modulation in spinal cord and PNS. This will increase our knowledge regarding the connectivity and

13 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 25

function of peripheral and spinal sensory neurons under both normal and pathological conditions. In conclusion, we have created an electrode array based on CNT web-like thin film which exhibited broad-spectrum optical transparency, excellent electrochemical interfacial properties, mechanical robustness, and stretching durability. With the capability of conducting electrical recording with negligible light-induced artifact under simultaneous optical stimulation/imaging, and real-time continuous recording under mechanically active environments, we expect that the electrode technology presented here could be a powerful tool in neuroscience to study neural mechanotransduction and traumatic injuries of the nervous system.

Supporting Information. Experimental details, supporting figures. This material is available free of charge via the Internet at http://pubs.acs.org. Corresponding Author *E-mail: [email protected]

Author contributions X.D. and J.Z. conceived the idea and designed the experiments. J.Z., W.X., L.X. and A.C. prepared the CNT film. J.Z. and F.C. fabricated the device. J.Z. did the characterization of the device. J.Z., X.L., W.L. and J.G. performed the optogenetics experiment. J.Z., X.L., M.L., and S.T. conducted the two-photon calcium imaging experiment. J.Z. and F.C. performed the TBI experiment. J.Z. and X.L. analyzed the

14 ACS Paragon Plus Environment

Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

data. X.D and J.Z. wrote the manuscript with inputs from all authors. All authors approved the final version of the manuscript. Notes The authors declare no competing financial interest. ACKNOWLEDGMENTS We thank the Microelectronics Research Center at Peking University for the microfabrication facility and support, the Animal Resources Center at Peking University for animal housing and care, and the Core Facilities at the School of Life Sciences, Peking University, for assistance with two-photon imaging. This work was supported by grants from the National Natural Science Foundation of China (No. 21422301, 91648207, 21373013), the National Basic Research Program of China (No. 2016YFA0200103, 2014CB932500), and China's 1000 Young Talent Award program.

15 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 25

References 1. Canales, A.; Jia, X.; Froriep, U. P.; Koppes, R. A.; Tringides, C. M.; Selvidge, J.; Lu, C.; Hou, C.; Wei, L.; Fink, Y. et al. Nat. Biotechnol. 2015, 33, (3), 277-284. 2. Scanziani, M.; Häusser, M. Nature 2009, 461, (7266), 930-939. 3. Wang, G.; Wyskiel, D. R.; Yang, W.; Wang, Y.; Milbern, L. C.; Lalanne, T.; Jiang, X.; Shen, Y.; Sun, Q. Q.; Zhu, J. J. Nat. Protoc. 2015, 10, (3), 397–412. 4. Zhang, J.; Laiwalla, F.; Kim, J. A.; Urabe, H.; Van, W. R.; Song, Y. K.; Connors, B. W.; Zhang, F.; Deisseroth, K.; Nurmikko, A. V. J. Neural Eng. 2009, 6, (5), 055007. 5. Anikeeva, P.; Andalman, A. S.; Witten, I.; Warden, M.; Goshen, I.; Grosenick, L.; Gunaydin, L. A.; Frank, L. M.; Deisseroth, K. Nat. Neurosci. 2012, 15, (1), 163170. 6. Packer, A. M.; Roska, B.; Häusser, M. Nat Neurosci 2013, 16, (7), 805-815. 7. Lee, J.; Ozden, I.; Song, Y. K.; Nurmikko, A. V. Nat. Methods 2015, 12, (12), 1157–1162. 8. Kim, T. I.; Mccall, J. G.; Jung, Y. H.; Huang, X.; Siuda, E. R.; Li, Y.; Song, J.; Song, Y. M.; Pao, H. A.; Kim, R. H. et al. Science 2013, 340, (6129), 211-216. 9. Park, S.; Guo, Y.; Jia, X.; Han, K. C.; Grena, B.; Kang, J.; Park, J.; Chi, L.; Canales, A.; Chen, R. et al. Nat. Neurosci. 2017, 20, (4), 612–619 10. Nikolenko, V.; Poskanzer, K. E.; Yuste, R. Nat. Methods 2007, 9, (6), 943– 950. 11. Gonçalves, J. T.; Anstey, J. E.; Golshani, P.; Portera-Cailliau, C. Nat. Neurosci. 2013, 16, (7), 903–909. 12. Carlson, G. C.; Coulter, D. A. Nat. Protoc. 2008, 3, (2), 249–255. 13. Ghezzi, D.; Antognazza, M. R.; Maccarone, R.; Bellani, S.; Lanzarini, E.; Martino, N.; Mete, M.; Pertile, G.; Bisti, S.; Lanzani, G. et al. Nat. Photonics 2013, 7, (5), 400-406. 14. Carvalho-De-Souza, J.; Treger, J.; Dang, B.; Kent, S. H.; Pepperberg, D.; Bezanilla, F. Neuron 2015, 86, (1), 207-217. 15. Jiang, Y.; Carvalhodesouza, J. L.; Wong, R. C. S.; Luo, Z.; Isheim, D.; Zuo, X.; Nicholls, A. W.; Jung, I. W.; Yue, J.; Liu, D. J. et al. Nat. Mater. 2016, 15, (9), 10231030. 16. Cardin, J. A.; Carlén, M.; Meletis, K.; Knoblich, U.; Zhang, F.; Deisseroth, K.; Tsai, L. H.; Moore, C. I. Nat. Protoc. 2010, 5, (2), 247–254. 17. Park, D. W.; Schendel, A. A.; Mikael, S.; Brodnick, S. K.; Richner, T. J.; Ness, J. P.; Hayat, M. R.; Atry, F.; Frye, S. T.; Pashaie, R. et al. Nat. Commun. 2014, 5, 5258. 18. Ledochowitsch, P.; Yazdan-Shahmorad, A.; Bouchard, K. E.; Diaz-Botia, C.; Hanson, T. L.; He, J. W.; Seybold, B. A.; Olivero, E.; Phillips, E. A. K.; Blanche, T. J. et al. J. Neurosci. Meth. 2015, 256, 220-231. 19. Kuzum, D.; Takano, H.; Shim, E.; Reed, J. C.; Juul, H.; Richardson, A. G.; Vries, J. D.; Bink, H.; Dichter, M. A.; Lucas, T. H. et al. Nat. Commun. 2014, 5, (5), 5259. 20. Park, D. W.; Brodnick, S. K.; Ness, J. P.; Atry, F.; Krugnerhigby, L.; Sandberg, A.; Mikael, S.; Richner, T. J.; Novello, J.; Kim, H. et al. Nat. Protoc. 2016, 11, (11), 2201-2222. 21. Lacour, S. P.; Courtine, G.; Guck, J. Nat. Rev. Mater. 2016, 1, 16063. 22. Harrison, D. E.; Cailliet, R.; Harrison, D. D.; Troyanovich, S. J.; Harrison, S. O. J. Manip. Physiol. Ther. 1999, 22, (4), 227-234.

16 ACS Paragon Plus Environment

Page 17 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

23. Sabet, A. A.; Christoforou, E.; Zatlin, B.; M.Genin, G.; V.Bayly, P. J. Biomech. 2008, 41, (2), 307-315. 24. Bayly, P. V.; Cohen, T. S.; Leister, E. P.; Ajo, D.; Leuthardt, E. C.; Genin, G. M. J. Neurotraum 2005, 22, (8), 845-856. 25. Topp, K. S.; Boyd, B. S. Phys. Ther. 2006, 86, (1), 92-109. 26. Cater, H. L.; Sundstrom, L. E.; Rd, M. B. J. Biomech. 2006, 39, (15), 28102818. 27. Li, Z.; Jia, Y.; Wei, J.; Wang, K.; Shu, Q.; Gui, X.; Zhu, H.; Cao, A.; Wu, D. J. Mater. Chem. 2010, 20, (34), 7236-7240. 28. Yang, L.; Zhao, Y.; Xu, W.; Shi, E.; Wei, W.; Li, X.; Cao, A.; Cao, Y.; Fang, Y. Nano Lett. 2016, 17, (1), 71–77. 29. Lipomi, D. J.; Vosgueritchian, M.; Tee, B. C.; Hellstrom, S. L.; Lee, J. A.; Fox, C. H.; Bao, Z. Nat. Nanotechnol. 2011, 6, (12), 788-792. 30. Zhao, S.; Liu, X.; Xu, Z.; Ren, H.; Deng, B.; Tang, M.; Lu, L.; Fu, X.; Peng, H.; Liu, Z. et al. Nano Lett. 2016, 16, (12), 7731-7738. 31. Kim, K. S.; Zhao, Y.; Jang, H.; Sang, Y. L.; Kim, J. M.; Kim, K. S.; Ahn, J. H.; Kim, P.; Choi, J. Y.; Hong, B. H. Nature 2009, 457, (7230), 706-710. 32. Grätzel, M. Nature 2001, 414, (6861), 338-344. 33. Honda, K. J. Photochem. Photobiol. A-Chem. 2004, 166, (1–3), 63-68. 34. Chen, T.; Xue, Y.; Roy, A. K.; Dai, L. ACS Nano 2014, 8, (1), 1039-1046. 35. Zang, J.; Ryu, S.; Pugno, N.; Wang, Q.; Tu, Q.; Buehler, M. J.; Zhao, X. Nat. Mater. 2013, 12, (4), 321-325. 36. Kim, K. H.; Vural, M.; Islam, M. F. Adv. Mater. 2011, 23, (25), 2865-2869. 37. Deisseroth, K. Nat. Methods 2011, 8, (1), 26-29. 38. Laxpati, N. G.; Mahmoudi, B.; Gutekunst, C. A.; Newman, J. P.; Zellertownson, R.; Gross, R. E. Frontiers in Neuroengineering 2014, 7, 40. 39. V, G.; KR, T.; F, Z.; M, M.; K, K.; MB, S.; K, D. J. Neurosci. 2007, 27, (52), 14231-14238. 40. Bachilo, S. M.; Strano, M. S.; Kittrell, C.; Hauge, R. H.; Smalley, R. E.; Weisman, R. B. Science 2002, 298, (5602), 2361-2366. 41. Dresselhaus, M. S.; Dresselhaus, G.; Avouris, P. Carbon Nanotubes: Synthesis, Structure, Properties, and Applications. Springer: Heidelberg, 2001. 42. Liu, K.; Deslippe, J.; Xiao, F.; Capaz, R. B.; Hong, X.; Aloni, S.; Zettl, A.; Wang, W.; Bai, X.; Louie, S. G. et al. Nat. Nanotechnol. 2012, 7, (5), 325-329. 43. Berciaud, S.; Cognet, L.; Poulin, P.; Weisman, R. B.; Lounis, B. Nano Lett. 2007, 7, (5), 1203-1207. 44. Wu, H.; Kong, D.; Ruan, Z.; Hsu, P. C.; Wang, S.; Yu, Z.; Carney, T. J.; Hu, L.; Fan, S.; Cui, Y. Nat. Nanotechnol. 2013, 8, (6), 421-425. 45. Kim, K.; Kim, J.; Hyun, B. G.; Ji, S.; Kim, S. Y.; Kim, S.; An, B. W.; Park, J. U. Nanoscale 2015, 7, (35), 14577-14594. 46. Cheng, A.; Gonçalves, J. T.; Golshani, P.; Arisaka, K.; Porteracailliau, C. Nat. Methods 2011, 8, (2), 139-142. 47. Chen, Q.; Cichon, J.; Wang, W.; Qiu, L.; Lee, S. J.; Campbell, N. R.; Destefino, N.; Goard, M. J.; Fu, Z.; Yasuda, R. et al. Neuron 2012, 76, (2), 297-308. 48. Carriero, G.; Uva, L.; Gnatkovsky, V.; Avoli, M.; De, C. M. J. Neurophysiol 2010, 103, (5), 2728-2736. 49. Mayevsky, A.; Zarchin, N.; Kamenir, Y.; Sonn, J.; Rogatsky, G. G. J. Neurotraum 2003, 20, (12), 1315-1325.

17 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 25

50. Lauritzen, M.; Dreier, J. P.; Fabricius, M.; Hartings, J. A.; Graf, R.; Strong, A. J. J. Cereb. Blood Flow Metab. 2011, 31, (1), 17-35. 51. Montgomery, K. L.; Iyer, S. M.; Christensen, A. J.; Deisseroth, K.; Delp, S. L. Sci. Transl. Med. 2016, 8, (337), 337rv5. 52. Minev, I. R.; Musienko, P.; Hirsch, A.; Barraud, Q.; Wenger, N.; Moraud, E. M.; Gandar, J.; Capogrosso, M.; Milekovic, T.; Asboth, L. et al. Science 2015, 347, (6218), 159-163.

18 ACS Paragon Plus Environment

Page 19 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figures

Figure 1. Stretchable transparent CNT electrode array. (A) Schematics of stretchable transparent CNT electrode array for simultaneous optical and electrical interrogation of neural circuits. (B) A scanning electron microscopy image of the CNT web-like thin film used in stretchable transparent electrode array. (C) Optical transmittance of a bare PDMS and CNT/PDMS complex, both with PDMS thickness of 100 µm. (D) Optical transmittance of a CNT/PDMS complex before stretching and under stretching to a strain of 20%. Thicker CNT film than that in C was used here. (E) Schematic illustration of the evolution of the CNT thin film under stretching. (F) A CNT electrode array positioned over a Peking University logo to show its optical transparency. The electrode array was bonded to a heat seal connector (HSC) for amplifier connection. Inset shows the layout of the CNT electrodes in the area marked by the red dashed box. The recording sites are highlighted by red color.

19 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 25

Figure 2. Characterization of the stretchable transparent CNT electrode array. (A) Electrochemical impedance magnitude of a CNT and graphene electrode recorded in saline solution (pH 7.4). (B) Cyclic voltammograms of a CNT and graphene electrode recorded in saline solution (pH 7.4). (C) Relative change of impedance at 1kHz of a CNT and graphene electrode under stretching strain. Insets show a close look of the data. Z0 denotes the impedance values before stretching. The strain is defined as (LsL0)/L0 × 100 (Ls: length after stretching, L0: length before stretching). (D) Impedance magnitude of a CNT (top) and graphene (bottom) electrode recorded in saline solution (pH 7.4) before stretching (0%), at elongation of 50% and 0.9% respectively, and after being released back to unstretched state [0% (REC)]. (E) Cyclic voltammograms of a CNT electrode before stretching and at 20% and 50% tensile strain. (F) Relative change of impedance at 1kHz of a CNT electrode at rest and after multiple stretching cycles to 20% strain. Z0 denotes the impedance value before stretching. The recording site size is 100 × 100 µm2 for all CNT and graphene electrodes.

20 ACS Paragon Plus Environment

Page 21 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 3. Simultaneous optogenetic stimulation and ECoG recording with stretchable, transparent CNT electrode array. (A) Schematics showing the experimental setup and position of the CNT electrode array on the cerebral cortex of a Thy1-ChR2-YFP mouse. The exposed cortical surface (marked by circle in the right schematic) was illuminated by blue light stimuli (488 nm). (B) Top, the train of applied rectangular photostimulus pulses (488 nm, 15 ms, 10 Hz, 2.4 mW/mm2, 30 mW applied onto cortical surface). Bottom, ECoG signal high-passed at 3 Hz recorded by a CNT electrode shows the light-evoked large negative potential peaks associated with every photostimulus. (C) Left, the close-up view of a light pulse (top) and an unfiltered light-evoked negative potential (bottom) shown in B. Right, ECoG signal recorded from a WT mouse under a light pulse with same stimulation condition as B. Three hundreds trials were averaged for each. WT mice did not show any response to the photo-stimulus. (D) Spectrogram of the ECoG signal shown in B. (E, F) Light-evoked potentials under photostimulus • are mapped according to channel location on the array under different stimulus intensity (E) and duration (F) (E, 15 ms, 10 Hz, 15 mW and 30 mW applied, which corresponds to 1.2 mW/mm2 and 2.4 mW/mm2; F, 10Hz, 15 mW applied, 1.2 mW/mm2). Stronger and longer stimuli evoked larger negative potentials. These potentials were spatially uniform with the entire window illuminated. Three hundreds trials were averaged for each trace. The electrode with 21 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 25

row number 3 and column number 3 was non-functional and the data is invalid. (G) The light-induced artifacts in CNT and Au electrode recording with recording site size of 100 × 100 µm2 tested in saline under a stimulus pulse (488 nm, 1 Hz, 30 mW, 2.4 mW/mm2, 15 ms). Grey rectangles indicate the start and duration of stimulation pulses. Inset shows the close-up view of the artifact from the CNT recording. X-scale bars represent 20 ms, y-scale bars represent 2 µV. The data shown in this figure were averaged from four hundreds stimulation trials. The light-induced artifact in CNT electrode recording is negligible compared to the large artifact from the Au electrode recording.

22 ACS Paragon Plus Environment

Page 23 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Figure 4. Simultaneous two-photon calcium imaging and ECoG recording with stretchable, transparent CNT electrode array. (A) Schematic showing the position of the CNT electrode array on the cerebral cortex of a Thy1-GCaMP3 transgenic mouse. The rectangle marks the cranial window and the blue dot indicates the position for 4AP delivery. (B) In vivo two-photon images of GCaMP3-expressing neurons underneath the recording site of a CNT electrode in the visual cortex of a Thy1GCaMP3 mouse. The depth below the pial surface is shown in each panel. (C) ECoG recording from a CNT electrode showing normal baseline brain activity (left trace) and interictal-like activity (right trace). The data was high-pass filtered at 3 Hz. Bottom images show the two-photon calcium images of the brain region underneath the recording site of the electrode used to record the top traces, taken at the time points marked by the red dashed lines. Individual activated neurons (indicated by the white dashed circles) with calcium transients can be identified coinciding with the interictal-like event. (D) ECoG signal recorded by the CNT electrode showing seizure-like ictal discharges. The data was high-pass filtered at 3 Hz. (E) ∆F/F0 traces for the entire brain region under the recording site (labeled as ROI) and several individual cells during the ictal-like events. (F) Two-photon calcium images of the 23 ACS Paragon Plus Environment

Nano Letters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 25

brain region underneath the recording site taken during the ictal-like events. The four images correspond to the time points indicated by the dashed lines in E. The individual activated neurons shown in E are indicated by the white dashed circles.

Figure 5. ECoG recording under TBI. (A) Schematic showing the CNT electrode array placement and weight drop location (indicated by the blue circle). The rectangle marks the position of the craniotomy. (B) A representative continuous ECoG trace recorded by a CNT electrode before, during and after the TBI high-pass filtered at 3 Hz. The TBI was induced at the time indicated by the arrow. (C) Power density spectrum of the recorded ECoG signal before TBI (red curve), after the TBI (blue curve, averaged from the data collected during the first 30 s after TBI), and 2 min after the TBI (black curve) where the ECoG activity was fully recovered to the preinjury level. (D) Time-frequency spectral analysis of the ECoG signal recorded by a CNT electrode before, during and after the TBI. The TBI was induced at the time indicated by the white dashed line. The white arrow indicates the period where the low frequency ECoG activity was suppressed due to TBI.

24 ACS Paragon Plus Environment

Page 25 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

TOC figure

25 ACS Paragon Plus Environment