CIVIL ENGINEERING 365 ALL ABOUT CIVIL ENGINEERING

[ad_1]

Materials for adaptive self-healing electronic epineurium (A-SEE)

Synthesis of SHP and Ag flake-SHP composite followed our previous reports32,33. The Ag flake-SHP composite with a weight ratio of 4:1 (Ag flake:SHP) was used in this study. Two different methods were tried to introduce AuNM to the Ag flake-SHP composite: (1) transfer and (2) direct deposition. AuNM of 60 nm thickness was first deposited on a SiO2 wafer using an e-beam evaporator (ULVAC Co.). Then, a Ag flake-SHP mixed chloroform solution with an appropriate viscosity33 was poured onto the AuNM-deposited wafer and left to dry at room temperature for 12 h. When the sample was dry, the AuNM-composite was easily detached from the wafer. This process resulted in the conformal transfer of the AuNM from the SiO2 wafer due to the relatively strong Au–Ag and Au–polymer interactions compared with the Au–SiO2 interaction. In the second method, the AuNM was directly deposited to the top surface of the composite supported on an octadecyltrimethoxysilane-treated silicon wafer. After comparing the mechanical and electrical performances of the transferred AuNM-composite and those of the directly deposited AuNM-composite, we used the transferred AuNM-composite for further experimentation. The fabrication process of the A-SEE based on self-bonding assembly using SHP and the AuNM-composite is explained in Supplementary Fig. 1.

For the interconnection between A-SEE and neural signal recording amplifier, a Au pad with a PI substrate was used. Each pad had a 1 mm width and a 4 mm length with a hole for self-bonding between the AuNM-composite and the SHP substrate (hole dimensions (w × l): 0.35 mm × 2 mm), and was composed of a PI, Ti, Cu, and Au layer of 80, 5, 30, and 10 nm thickness, respectively. After the self-bonding of the Au pad with the AuNM-composite layer and SHP substrate layer, a Teflon-coated wire was soldered onto the other end of the Au for connecting with the neural signal recording amplifier. Silicone casting compound (Kwik-cast, World precision instruments) was used for encapsulating the soldered wire.

Characterizations of A-SHE

SEM analysis

Field emission scanning electron microscopy (Inspect F50, FEI Co.) was used to observe the interfaces between the composite and the transferred or directly deposited AuNM. Each AuNM- composite film was cross-sectioned by using a razor blade and coated by platinum before imaging.

Stress relaxation characterization

For DMA analysis, the 0.3 mm thick SHP film, the 0.3 mm thick AuNM-composite film, the 0.9 mm thick encapsulated AuNM-composite (A-SEE) film, and the 0.3 mm thick PDMS (Sylgard 184, 20:1 PDMS:crosslinking agent weight ratio) film were each cut to have a length of 10 mm and a width of 5 mm. We chose PDMS with a 20:1 ratio as the control material in all experiments due to its value of Young’s modulus (131.8 ± 13.8 kPa), which is comparable with that of SHP (163.3 ± 42.1 kPa) (Supplementary Fig. 5). Then, each sample was stretched to a tensile strain of 30% with a stretching speed of 50% strain per minute and the tensile stress at the strain was measured for 60 min by DMA Q800 (TA Instruments). The stress relaxation properties of SHP, PDMS, composite, and the A-SEE were measured at two different temperatures (25 °C and 37 °C) to determine the temperature dependence.

Electrical characterization

For resistance-strain characterization, each AuNM-composite (3 mm length, 5 mm width, 0.3 mm thickness) self-bonded to the SHP substrate (3 mm length, 7 mm width, 0.2 mm thickness) (A-SEE) was stretched by using an automatic stretching machine with a stretching speed of 100% strain per minute and the resistance was measured using a four-point probe method (Keithley 2400, Tektronix). For the cyclic stretching test, a strain of 50% was repeatedly applied for 1000 cycles. For the nerve interfacing mimicking test, the AuNM-composite with a length of 10 mm, a width of 5 mm, and a thickness of 0.3 mm was repeatedly bent to reach a bending radius of 0.5 mm for 1000 cycles. For the self-healing test, the sample was completely cut using a razor blade and then self-healed by being heated at 60 °C for 1.5 h.

Electrochemical characterization

Electrochemical impedance and CV were measured using a potentiostat (CHI 760 C, CH Instruments) in the PBS solution. The exposed area of the samples was 7.5 mm2 and a platinum wire was used as the counter electrode in both measurements. Potentiostatic electrochemical impedance spectroscopy was conducted with a frequency ranging from 1 Hz to 1 MHz and an amplitude of 10 mV. For the CV measurement, an Ag/AgCl electrode (BASiAg/AgCl/3 M NaCl) was used as the reference electrode, and the curves were obtained from −0.65 V to 0.8 V at a scan rate of 100 mV/s. The charge delivery capacity was calculated using the following Eq. (1).

$${mathrm{CDC}} = frac{1}{{vA}}int_{Ec}^{Ea} left| i right|dEleft( {C/cm^2} right).$$

(1)

Here, v, A, Ea, Ec, i, and E are scan rate (V/s), geometrical surface area of the electrode (cm2), anodic/cathodic potential limit (V), measured current, and electrode potential (V vs Ag/AgCl), respectively.

Finite element analysis

A simplified three-dimensional model of PDMS- and SHP-interfaced nerve tissues was created for FEA using ANSYS Workbench (Release 16.1; ANSYS Inc. Canonsburg, PA). The simplified three-dimensional model contained 23,216 nodes and 4530 elements and was assumed to be bonded at contacting surfaces, homogeneous, isotropic, and linearly elastic. The mechanical properties of the nerve tissue and the materials were determined and are listed in Supplementary Table 1. The nerve tissue was assumed to be cylindrical with a 1 mm diameter and 10 mm height. The tested materials were also assumed to be hollow cylinders with 1 mm inner diameter, 1.6 mm outer diameter, and 5 mm height that were wrapped around the nerve tissue. The stress measured from the DMA results at a specified time point, shown in Supplementary Fig. 3 and summarized in Supplementary Table 1, was applied as pressure on the outer layer of the tested material to determine the resulting stress distribution on the nerve tissue. The two-dimensional cross-sectional distributions were obtained as a cross-section of the center of the nerve tissue. The dotted circles indicate the original diameter of the nerve tissue before it was interfaced with the materials.

In vitro biocompatibility test

RAW 264.7 and C2C12 cell lines were purchased from American Type Culture Collection and were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) with 10% fetal bovine serum at 37 °C. For testing the biocompatibility of the materials (SHP, composite, and AuNM-composite), the cells were seeded in 96-well at a concentration of 1 × 104 cells/mL. After 7 days of incubation of cells with the same size of each material, the cell viability was measured using an MTS assay solution (G3585, Promega, Wisconsin, USA). The absorbance of the reacted solution at a wavelength of 490 nm was recorded using a 96-well plate reader (Infinite 200 Pro, Tecan, Männedorf, Switzerland).

Sterilization and preparation for the animal study

The SHP, PDMS, and A-SEE samples were immersed in 70% ethanol solution and washed twice with PBS solution before implantation to prevent contamination and remove the PDMS precursor (see Supplementary Fig. 8)44,45.

Animal preparation and implantation of A-SEE

All animal experiments were performed and handled per the regulation of the Institutional Animal Care and Use Committee of the Korea Institute of Science and Technology (Approval No. 2018-067). The experimental procedure was performed according to the Guide for the Care and Use of Laboratory Animals46. For the A-SEE implantation, a Sprague-Dawley rat (male, 300 g) was anesthetized using an intraperitoneal injection of 50 mg/kg Zoletil and 10 mg/kg Xylazine mixture. After a deep level of anesthesia was achieved, the skin incision was extended to the dorsal aspect of the paw to expose the hind limb musculature. The lateralis and biceps femoris muscles were then identified and retracted to expose the sciatic nerve. Then the A-SEE was wrapped around the sciatic nerve after the surrounding tissue was removed. Self-locking process for wrapping the nerve takes within a minute in general. During the process, we just need to maintain contact between the folded SHP films (substrate of A-SEE) by using surgical tweezers for a while. After completing the self-locking, the self-bonded A-SEE is difficult to separate again unless it is amputated using scalpel. However, since the contacted SHP films can be detached within few seconds, careful adjustment of position or posture of the A-SEE is possible during the implantation.

In vivo nerve compression comparison of SHP with PDMS

We prepared PDMS and SHP cuff structures through a molding process. Both PDMS and SHP cuff structures have an inner diameter of 1 mm (similar to the diameter of the sciatic nerve) and a slit that allows them to be opened and applied to sciatic nerves and then closed autonomously. These were implanted onto the two sciatic nerves to confirm how their mechanical properties affect the nerves for a week. After implanting the PDMS cuff structure to the nerve, it was locked using silicone casting compound (Kwik-cast, World precision instruments). The SHP cuff structure was locked by the self-bonding assembly. After 1 week of implantation, each of the sciatic nerves were investigated using histological and immunofluorescence analyses (see Supplementary Fig. 9).

Neural signal data acquisition

For neural signal recording, a nerve cuff electrode system that was developed in our previous report47 was used (see Supplementary Fig. 18). This electrode system was composed of a preamplifier and an external module with a 39,601 gain and a −3 dB bandwidth from 425 to 5500 Hz. An A-SEE interfaced with a rat sciatic nerve was connected to a preamplifier and the final output of the external amplifier was digitized using an analog-to-digital converter board (PXI-6733, National Instrument, USA) with a sampling frequency of 25 kHz. Then, the acquired signals were processed using the LabVIEW software (National Instruments, USA).

For neural signal recording in a moving animal, the preamplifier was implanted in the subcutaneous layer of a rat’s back with a transcutaneous head connector. The neural signal and the two-dimensional motion data were recorded synchronously. The motion system was composed of a high speed digital camera (Marlin F033B, AVT, Germany) and grabber board with a 640 × 480 pixel resolution. The digital camera was positioned perpendicular to the walking trace and the images were recorded at 60 frames/s. The anatomical landmarks were defined to the lateral side of the hind limb at the knee, ankle, and fifth metatarsal joint41.

In vivo neural signal recording

The in vivo neural signal recording was performed for 7 weeks with the right leg of five animals. The animals were anesthetized with 50 mg/kg Zoletil and 10 mg/kg Xylazine mixture by the intraperitoneal injection. The glabrous skin of the right hind paw was mechanically stimulated by using a pig hair brush, and the stimuli were induced with two different pressure intensities. The stimuli were classified as strong and weak, which correspond to 12.5 g/cm2 (220 g–16 cm2) at 0.05 s/cm and 3.1 g/cm2 (50 g–16 cm2) at 0.05 s/cm, respectively. Each stimulus was measured by a force gauge (Mark-10, USA). Each mechanical stimulus lasted ~1 s and was repeated approximately every 5 s. For each trial, the neural signal was recorded for 60 s, and each trial included 10 stimuli. For quantification of the neural signals, the SNR was defined as the ratio of the root mean voltage of the neural signals during the stimulus-present period to that during the stimulus-absent period. For each amplifier, the average SNR was calculated from five trials and each amplifier was used in five rats.

Denoising process of the recorded neural signals

The recorded neural signals were denoised using a 1D wavelet denoising algorithm48. To secure robustness in wavelet transform, we applied a maximal overlap discrete wavelet transform method. In Supplementary Fig. 21, raw, and denoised data were compared with variance change behaviors of the signals as a function of time-window used in the wavelet transform.

Calculation of SNR ratio between the recorded neural signals under stimulation and at rest state

For time-dependent signal data(Sleft( t right)), which is a set of (N) discrete signal points ((Sleft( t right) = left{ {sleft( {t_i} right)} right},{mathrm{ where }}i = 1,2, cdots ,N)), the residual noise, (Nleft( t right)), can be expressed as

$$N(t) = S(t) – S_d(t),$$

where (S_d(t)) denotes the denoised time-dependent signal. Using (Nleft( t right)), the SNR can be expressed as

$$SNR = 10log _{10}left( {frac{{leftlangle {Sleft( t right)^2} rightrangle }}{{leftlangle {Nleft( t right)^2} rightrangle }}} right),,leftlangle {fleft( t right)^2} rightrangle equiv frac{1}{N}mathop {sum}limits_{i = 1}^N {sleft( {t_i} right)} ^2,$$

To compare SNR difference between two time-dependent signal data, (S_{stim}left( t right)) and (S_{rest}left( t right)), where the subscripts (stim) and (rest) denote periods under stimulation and of rest, respectively, we can calculate the relative SNR which can be expressed as

$$SNR_{rel} = SNR_{stim} – SNR_{rest}.$$

Using (SNR_{rel}), we can quantitatively compare the deterioration dynamics of the neural signal transduction affected by degradation of the flexible electrode. A detailed procedure is provided in Supplementary Table 2.

In vivo electrical stimulation

An electrical stimulator (AM 2200, AM-system) with a programmed pulse generator, which was developed in our previous report49, was used to induce the biphasic current pulse on the sciatic nerve. The biphasic pulse was set to a 0.25 ms pulse width and different repetition stimulation pulse periods (5, 10, 20, 50, and 100 Hz). The stimulation period and current were set to 0.5 s and 1.14 μC/cm2, respectively. The amplitude of the stimulation pulse of 80 μA was chosen based on the minimum thresholds for joint movements. The muscle force response from electrical stimulation was measured using a torque sensor (QWFK-8M, Honeywell) with a shoe shaft. The torque data were digitalized using an analog-to-digital converter board (PXI-6143, National instruments). The muscle forces were normalized to the values at frequency 100 Hz.

Demonstration of nerve-to-nerve interfacing

The nerve-to-nerve interface system was composed of neural signal amplifier, electrical stimulator and nerve-to-nerve controller, which were modified from our previous work49. For nerve-to-nerve interfacing demonstration, two A-SEEs were implanted to each ipsilateral and contralateral sciatic nerve, respectively, and each A-SEE interfaced with the ipsilateral and contralateral sciatic nerve performed the neural signal recording and stimulation, respectively. The neural signal recording and electrical stimulation were performed using the same experimental setup as in the in vivo neural signal recording and in vivo electrical stimulation. The recorded neural signals from the ipsilateral sciatic nerve were converted to the stimulation parameter via a nerve-to-nerve controller and the electrical stimulation pulses were applied to the contralateral sciatic nerve. To verify the demonstration of mimicking crossed extensor reflex, the ankle and knee joint position of the contralateral limb were measured during strong and weak mechanical stimulation on the ipsilateral site paw using the brush. The joint position measurement was performed using the same experimental setup as in the in vivo neural signal recording in a moving animal. The joint movements of the contralateral limb were triggered only when the amplitude of sensory neural signals recorded from the ipsilateral nerve was above the threshold that we programmed using a rectified bin integration method49. The threshold, the neural signal amplitude of 9 μV, was converted to the electrical stimulation amplitude of 50 μA. The overall system was processed by using LabVIEW software (National Instruments, USA).

Western blot

After scarification, nerve tissue samples were immersed in tissue protein extraction reagent (T-PERTM, #78510, Thermo Fisher Scientific) at a ratio of 1:20 (w/v) for blotting. The samples were sonicated for 5 min in ice and centrifuged at 13572 g and 4 °C for 15 min. We collected the supernatant and then determined total protein concentration using the PierceTM BCA Protein assay kit (#23225, Thermo Fisher Scientific, #23225) according to the manufacturer’s instructions. Next, 15 µg protein from each sample was loaded on an SDS-PAGE gel (Biorad) and subjected to electrophoresis in a 1X Tris-Glycine SDS buffer (T8053-101, GenDepot) at 130 V for 60 min. The separated proteins were transferred to a nitrocellulose membrane, 0.45 μm (#1620145, Biorad) in 1X Tris-Glycine Native buffer (HT2028, Biosesang) with 10% methanol at 4 °C. To detect the target protein, the membrane was incubated with specific primary antibodies; GAPDH (ab9485, Abcam, 1:2500), CD68 (ab53444, Abcam, 1:1000) and CTGF (ab6992, Abcam, 1:1000). Two types of horse-radish-peroxidase (HRP)-conjugated antibodies, anti-rabbit (ab6721, Abcam, 1:5000) and anti-rat (SC-2006, SantaCruz, 1:5000), were used as a secondary antibody. After a 1-h incubation with 2nd antibodies, the membrane was washed with PBS solution and then was treated with a detection agent (WSE-7120L, ATTO corporation). Finally, the desired bands were imaged by an ImageQuant LAS 4000 mini (GE Healthcare).

Histology and immunofluorescence analysis

The nerve tissues were fixed in a 4% paraformaldehyde solution (P2031, Biosesang) and embedded in paraffin blocks. The blocks were sliced into 5 μm thick sections with Microtome (RM2255, Leica Biosystem). The paraffin sections were deparaffinized with Xylene (#8587-4400, Daejung) and hydrated with decreasing concentration of alcohol. For tissue histology imaging, the sections were stained with hematoxylin and eosin and terminal deoxynucleotidyl TUNEL assay using standard protocols. The images were monitored using an optical microscope (LAS X Widefield, Leica). For immunostaining, hydrated paraffin sections were blocked with 0.3% H2O2 for 30 min and soaked in 0.01 M citrate (C0759, Sigma-Aldrich) buffer pH 6.0 at 95 °C for 20 min. The sections were blocked with 3% BSA in PBS solution (#10010023, Gibco) for 1 h and were incubated overnight at 4 °C with the primary antibodies, CD68 (ab53444, Abcam, 1:500) and CTGF (ab6992, Abcam, 1:500). The samples were incubated with the secondary antibodies, a Donkey Anti-Rat IgG Alexa Fluor® 647 (ab150155, Abcam, 1:5000) and Alexa FluorTM 488 goat anti-rabbit IgG (A-11008, Invitrogen, 1:5000), for 1 h at room temperature. The slides were mounted with Vectashield mounting medium with DAPI (H-1200, Vector Laboratories) and were observed using a confocal microscope (LSM 700, Carl Zeiss). For quantification of the obtained immunostaining and TUNEL staining images, the cross-section nerve tissue images were obtained with the same size (1069 × 1069 pixels for immunostaining images and 5184 × 3456 pixels for TUNEL stained images) at the same exposure time and were selected randomly (n = 3). Then, the fluorescence intensity of the images was measured using the ImageJ program, and the brightness intensities of the black and white threshold color-inverted TUNEL stained images were quantified. For the relative fluorescence intensity, a mean value of CD68 and CTGF divided by the value of DAPI was used.

Quantification of Au and Ag-ion leakage

For the in vitro Ag-ion leakage measurement, the composite and the AuNM-composite were incubated in cell culture medium (DMEM, Dulbecco’s Modified Eagle’s Medium) for 7 days, respectively. Then, we collected the culture media to measure the amount of Ag ions and determine the potential Ag-induced toxicity of the composites in physiological conditions. For in vivo quantification, the nerve tissues interfaced with an A-SEE for 6 weeks were digested in an acid mixture containing 30% HCl (Sigma-Aldrich) and 70% HNO3 (Sigma-Aldrich) at 75 °C. After the complete digestion of the material, the solution was diluted with Millipore water. The Au and Ag contents were measured using ICP-MS.

Statistical analysis

The statistical analysis was performed using Origin 2020 software.

Reporting summary

Further information on experimental design is available in the Nature Research Reporting Summary linked to this paper.

[ad_2]

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *