Electrically Controlled Vasodilator Delivery from PEDOT/Silica Nanoparticle Modulates Vessel Diameter in Mouse Brain

Abstract

Vascular damage and reduced tissue perfusion are expected to majorly contribute to the loss of neurons or neural signals around implanted electrodes. However, there are limited methods of controlling the vascular dynamics in tissues surrounding these implants. We utilize conducting polymer poly(ethylenedioxythiophene) and sulfonated silica nanoparticle composite (PEDOT/SNP) to load and release a vasodilator, sodium nitroprusside, to controllably dilate the vasculature around carbon fiber electrodes (CFEs) implanted in the mouse cortex. The vasodilator release is triggered via electrical stimulation and the amount of release increases with increasing electrical pulses. The vascular dynamics are monitored in real-time using two-photon microscopy, with changes in vessel diameters quantified before, during, and after the release of the vasodilator into the tissues. We observe significant increases in vessel diameters when the vasodilator is electrically triggered to release, and differential effects of the drug release on vessels of different sizes. In conclusion, the use of nanoparticle reservoirs in conducting polymer-based drug delivery platforms enables the controlled delivery of vasodilator into the implant environment, effectively altering the local vascular dynamics on demand. With further optimization, this technology could be a powerful tool to improve the neural electrode-tissue interface and study neurovascular coupling.

Graphical Abstract

Utilizing conducting polymer poly(ethylenedioxythiophene) and sulfonated silica nanoparticle composite (PEDOT/SNP) electrode coating, an electrically controlled vasodilator release system is developed for local modulation of vasculature. Sodium nitroprusside was loaded on microelectrode coating and the eclectically triggered release effectively increased vessel diameter in the mouse brain as revealed by 2 photon imaging.

1. Introduction

Implantable neural electrodes are indispensable tools for neuroscience research and have also shown tremendous potential as components of neuroprosthetic devices aimed at restoring or modulating neurological functions[1–3]. However, electrode implantation is inherently traumatic to the central nervous system (CNS) tissues. Neuronal health around an implant is the product of many factors, including the inflammatory response to the implantation, the mechanical damage from insertion, the electrode/tissue interface, and changes in blood flow following implantation. The electrode inevitably severs blood vessels, which in turn lowers perfusion of blood in the tissues[4], thus interfering with the delivery of oxygen and nutrients to CNS cells, finally leading to a reduction in blood oxygen dependent signaling[5]. Following implantation, remodeling of vascular networks occurs to mitigate the loss of tissue perfusion[6–8], but this process takes days to weeks and leaves the tissues around the electrode in an oxygen-deprived environment. Anoxic environments directly affect neurons, changing their resting membrane potential [9, 10] and leading to excitotoxicity [11–13]. Following neural electrode implantation, the lack of sufficient tissue perfusion is a principal component of neuronal silencing and maybe even acute cell death, but controlling the vascular dynamics following electrode insertion has not been attempted.

Much like the rest of the body, the CNS vasculature responds to compounds such as nitroglycerin[14–16] and sodium nitroprusside[17, 18], which cause rapid vessel dilation. However, direct investigation of the effects of local blood flow modulation on the CNS tissues is difficult. This is largely due to the necessity of spatially and temporally constrained drug delivery. Fortunately, localized and on-demand drug delivery methods have been developed in the field of neural engineering which may be employed to control vascular dynamics while recording neural activity. These controlled drug delivery systems involve numerous techniques, including microfluidics[19–24], degradable polymers[25–27], nanoparticles and liposomes [28–32], and conducting polymers[33–37]. Conducting polymers can be electropolymerized onto neural electrode sites to reduce site impedance and increase neural recording yield[38, 39]. Further, the electroactivity of the conducting polymer creates the basis for electrochemically controlled drug loading and release[33, 34, 40–44], which can be integrated onto the neural recording/stimulation devices to reduce inflammation or modulate neural activity[36, 43, 45–47]. Conducting polymer-based electronics can even be coupled with biomolecules,[48–50] increasing adhesion of cells to the electrode site and allowing for electrochemical detection of compounds such as glucose.

The first generation of conducting polymer-based electrochemical drug delivery systems directly load drug molecules into the film during electropolymerization and release the drug by electrochemically reducing the polymer and, therefore, dissociating the drug[34, 51, 52]. One limitation of such an approach is its incompatibility with many compounds. Compounds that are positively charged are poorly integrated into the film, while electroactive compounds such as antioxidants can be destroyed during the electropolymerization of the polymer. Another limitation is drug loading capacity. An alternative method of drug delivery is the organic electronic ion pump.[50, 53–55], allowing for the delivery of cationic compounds via electrical modulation of a conducting polymer doped with a polyanion. However, this method requires a fluid reservoir to draw from, which would require a redesign of many electrodes, and may still be restricted in its ability to load and release electroactive compounds. To address these limitations, we synthesized a mesoporous sulfonate-modified silica nanoparticle (SNP) that serves both as a dopant for the conducting polymer and a drug reservoir to increase loading capacity. Furthermore, the non-conductive nature of the silica helps shield the compounds from the electropolymerization current. This method has demonstrated success in allowing for the enhanced loading and release of negatively charged, positively charged, and electrochemically active compounds[35, 56].

In this work, we have adapted the SNP-doped conducting polymer polyethylenedioxythiophene (PEDOT/SNP) to locally deliver the vasodilator sodium nitroprusside (NaNP). Cyclic voltammetry (CV) was used to reduce and oxidize the polymer, driving the release of NaNP, which was characterized by mass spectrometry. We then implanted PEDOT/SNP/NaNP-coated carbon fiber electrodes (CFEs) into the cortical tissue and utilized two-photon microscopy to monitor the dilation of vessels following the drug release in ketamine-anesthetized mice.

2. Methods

2.1. Materials and Regents

All reagents were purchased from Sigma Aldrich unless otherwise specified. Double-deionized water was used for all experiments.

2.2. Nanoparticle Synthesis

Mesoporous sulfonate-modified silica nanoparticles were fabricated as previously described [35]. In brief, mesoporous thiol-modified silica nanoparticles were fabricated from a base-catalyzed condensation reaction between tetraethyl orthosilicate and mercaptopropyl trimethoxysilane with cetyltrimethylammonium bromide as the surfactant template. Thiolated nanoparticles were collected by centrifugation, the surfactant template was removed, and then the thiol groups on the particles were oxidized to form sulfonate-modified mesoporous silica nanoparticles (SNP) with 20% H₂O₂. Following oxidation and washing of the particles, the SNP were stored dry at 4°C until used. The resulting nanoparticles have been previously characterized by our group via dynamic light scattering and were found to have average diameters of 100–110 nm. Pore sizes of the nanoparticles have also previously been determined to be 0.55 cc g⁻¹ in volume and between 1.2 and 2.5 nm in diameter using Brunauer–Emmett–Teller measurements[35].

2.3. Drug loading and Electropolymerization

Drug loading was performed by dissolving 200mg NaNP into 1 mL water, followed by the addition of 10 mg SNP. The suspension was sonicated for 20 minutes to encourage the drug to enter the small pores, and the nanoparticles were collected by centrifuge. The supernatant was removed, and the particles were dried under an N₂ stream and stored dry.

Electropolymerization of the PEDOT/SNP was performed with an Autolab PGSTAT128N potentiostat (Metrohm Autolab B.V.), in a 0.01M aqueous solution of ethylenedioxythiophene (EDOT) containing 10mg/3ml of drug-loaded SNP. The solution was gently triturated prior to polymerization. For both carbon fiber microelectrodes (400μm long 7μm diameter) and glassy carbon macroelectrodes (3mm diameter), electropolymerization took place under a current density of 1.5mA cm⁻² for 1000 seconds with a three-electrode setup using an Ag/AgCl reference and a Pt counter electrode.

2.4. Drug Release and Detection

Mass spectrometry was used to measure the release of NaNP from the drug-loaded PEDOT/SNP films on 3 mm diameter glassy carbon electrodes. Release was performed in 1 mL of 0.009wt% tetrabutylammonium chloride solution with CV scanning from 0.6 V to −0.3 V vs Ag/AgCl at a rate of 0.1 V s⁻¹ for 200 scans. Quantification of released compounds was performed with a liquid chromatography-mass spectrometer (Shimadzu LCMS-2020, Shimadzu, Japan), using electrospray ionization in negative mode with a 10-minute gradient of 10%−90% 0.1% formic acid in acetonitrile and 0.1% formic acid in water. 107.9 m/z was used to detect NaNP, corresponding to the [M-2Na⁺]⁻² peak. Concentration was determined with a calibration curve.

2.5. Animal Experiments with 2 Photon Microscopy

All animal experiments were approved by the University of Pittsburgh Institutional Animal Care and Use Committee. CX3CR1-GFP mice were anesthetized with Ketamine/xylazine (65 mg kg⁻¹ and 7 mg kg⁻¹) and anesthesia was confirmed by the lack of a toe-pinch response. Breathing and heart rate were monitored throughout the surgery and the animal’s body temperature was controlled with a heating pad. Once anesthesia was achieved, the scalp was removed and a 2 mm x 2 mm section of the skull was removed. A small amount of dental acrylic was placed around the cranial window to form a well that would hold the saline required for the water immersion lens of the microscope. The mouse was then moved to the imaging platform and an 18 g steel needle was inserted into the back to act as the counter electrode. An Ag/AgCl reference was then placed in contact with the saline well on the brain surface. A PEDOT/SNP-coated CFE (7 μm diameter and 400 μm long) was then gradually inserted into the cortex at a 30° angle from the horizontal plane. This ensured that the electrode would not contact the lens during imaging. The mouse was then given a 0.03 ml IP injection of sulforhodamine 101 (SR101, 1 mg mL⁻¹ in saline) to label blood vessels. We have found that IP injection of SR101 label the blood vessel just as well as FITC-dextran, and the labeling can maintain for 1–2 hr and there is no sign of astrocytic labeling by SR101 (Supplemental Figure 4). When vascular labeling became dim, supplementary SR-101 injection was given. Two-photon microscopy imaging was performed with an OPO laser (Insight DS+; Spectraphysics, Menlo Park, Ca) tuned to 920 nm, photomultiplier tubes (Hamamatsu Photonics KK, Hamamatsu, Shizuoka, Japan), and a 16× 0.8 numerical aperture water immersion lens (Nikon Inc. Milville, NY). Images were collected with Prairie View software. Z-stacks, spaced by 2 μm, were taken of the region surrounding the electrode before, during, and after CV stimulation for drug release. Z-stacks were taken every 20 minutes. Drug release was performed with CV as described above, 0.6 V to −0.3 V vs Ag/AgCl at a rate of 0.1 V s⁻¹, resulting in 66 scans per Z-Stack. Vessel diameter was measured with the ‘diameter’ plugin on ImageJ. Ten animals were used for the imaging experiments, 5 for drug delivery from PEDOT/SNP/NaNP and 5 for the no-drug control from PEDOT/SNP.

2.6. Statistics

Statistics were performed in GraphPad Prism 8. Statistical comparisons for vessel diameters between NaNP-releasing electrodes and control electrodes were made with Student’s t-test. Changes in vessel diameter relative to the initial images were made with a one-sample Student’s t-test.

3. Results

To controllably release NaNP into the cortex, we have employed a conducting polymer-gated, electrochemically-controlled drug delivery platform. The NaNP was loaded into the sulfonate-functionalized mesoporous silica nanoparticles which were then used to dope the conducting polymer during its electropolymerization. The validation of the drug loading and release was performed with glassy carbon macroelectrodes (3 mm diameter). The electrodes were coated with PEDOT/SNP-NaNP and stimulated for up to 200 cycles or left undisturbed for 8 hours to serve as a diffusion control. The release was performed with CV with voltage sweeping between 0.6 V and −0.3 V for 200 cycles. The voltage range was chosen to avoid unwanted reduction of NaNP at −0.5 V vs Ag/AgCl as seen in the CV (Figure 1A). In vitro release was conducted in a dilute solution of tetrabutylammonium chloride instead of commonly used PBS to minimize exposure of the instrument to excess salt. After different numbers of CV cycles, solutions were collected to measure the NaNP concentration with mass spectroscopy.

Figure 1.

In Vitro release of NaNP. (A) Cyclic voltammogram of the 100 μM NaNP in water with a glassy carbon electrode, showing the reduction peak at −0.5 V and the oxidation peak at 0.15 V. (B) Mean ± SEM of the mass chromatograms at 107.90 m/z for solutions in which NaNP-coated electrodes were cycled 0 times (blue, n=4) or 200 times (red, n=6). While samples undergoing redox cycling exhibited a robust release of NaNP into solution, samples in which electrodes were soaked in solution but not cycled had a dramatically diminished response, indicating minimal NaNP release due to diffusion. (C) Release amount as a function of CV stimulus number (n=3). Mass spectrometer readings were calibrated and concentrations were calculated for each trial. ** p<0.01 compared to 0 stimuli.

NaNP has a molar mass of approximately 262 g M⁻¹, and mass spectroscopy measurements demonstrated that significantly higher NaNP concentrations were present in the release samples (Figure 1B), with a higher number of CV cycles releasing a higher amount of NaNP. Converting NaNP concentrations in the release solution to masses demonstrates that the PEDOT/SNP-NaNP coated electrode was capable of releasing 4.0 ±1.7 μg (mean ± SD) of NaNP, or 56±24 μg cm⁻² after 200 stimulations (Figure 1C).

Following the validation of the drug release, we performed in vivo experiments to examine the effects of controlled release of NaNP on the vasculature and microglia present in the cortex. We used a 7 μm diameter insulated CFE with a 400 μm long exposed conductive area, to perform the release experiments to minimize the insertion injury and maximize the electrochemical surface area. The electrode was coated under 1.5 C cm⁻² anodic charge density, then washed with water and saline prior to insertion into the animal. Following the polymerization of PEDOT/SNP, we observed a drop in the electrode impedance across all frequencies measured, characteristic of PEDOT (Figure 2A), indicating that the PEDOT/SNP coating was successfully grown on the CFE. Samples were taken for SEM imaging to examine the topography of the coating at different magnifications (Figure 2B-F). Overall, the PEDOT/SNP coating occurred uniformly over the surface of the CFE. At higher magnification, the shape of the nanoparticles is visible within the polymer matrix with qualitatively uniform size and distribution (Figure 2E,F). Carbon fibers with the PEDOT/SNP coating were inserted and removed from 0.6% agarose gel to validate that the coating does not delaminate (Supplemental Figure 1). Light microscopy images of the electrodes before and after implantation both show the characteristic appearance of a dark and rough surface on the fiber, distinct from the non-coated carbon fiber.

Figure 2.

Characterization of NaNP-loaded PEDOT/SNP-coated CFE. (A) Impedance spectroscopy of the bare CFE and the PEDOT/SNP-coated CFE, demonstrating the characteristic decrease in impedance of the CFE following PEDOT polymerization. (B-D) SEM images of the NaNP-loaded PEDOT/SNP-coated CFE. Images were taken at increasing magnification to visualize the coating uniformity and microstructure. Scale bars (B) 10 μm (C) 1 μm (D) 300 nm. (E, F) diameters of SNP within the PEDOT/SNP polymer, values are 74.96 (E) and 66.14 (F) Scale bars are 100μm.

CFE coated by PEDOT/SNP with PEDOT/SNP-NaNP and without NaNP loading (control group) were implanted into the mouse cortex and imaged under a two-photon microscope. Vessels were visualized following an injection of SR101. Z-stack images were taken prior to CV stimulation (Stack 0), during stimulation (Stacks 1 and 2), and following the end of stimulation (Stacks 3 and 4). Z stacks took approximately 20 minutes each. This allowed us to evaluate persistent diameter increases over a relatively large imaging volume at high resolution.

Changes in vasodilation were measured from the pre-stimulation (Z-Stack 0) values over four consecutive Z-Stacks, with stimulation applied during Stacks 1 and 2 (Figure 3A). Changes in vessel diameter were visualized by overlaying the cortical blood vessels before (yellow) and after (magenta) CV stimulation (Figure 3B). Small translations in the images in the x, y, and z directions were often necessary to directly compare vasculature. The representative control image (Figure 3B) shows that there is a nearly perfect overlap between the pre-and post-stimulation vessels, indicating that there were no dramatic changes in the vessel diameters caused by the CV stimulation to PEDOT/SNP without loaded NaNP. On the other hand, when stimulation was applied to PEDOT/SNP-NaNP, prominent magenta borders were observed on many larger vessels, which were deemed to be arteries based on their morphological shape, indicating that the vessel diameter increased following stimulation, as the result of NaNP release.

Figure 3.

Vessel dynamics after NaNP release. (A) Timeline of images taken for the experiment. Z-Stack 0 was used as a baseline, Stacks 1 and 2 were taken during stimulation, and 3 and 4 were taken immediately following the end of stimulation. (B) Overlays of cortical vessels before (yellow) and after (magenta) stimulation for control (top) and NaNP loaded (bottom) PEDOT/SNP-coated CFEs. Images were produced from Z-Stacks 0 and 2. Aproximate electrode locations are highlighted in blue (enlarged). Scale bars are 50μm (C,D) Changes in vessel diameters over time compared to the initial Z-Stack for control and NaNP-loaded PEDOT/SNP, respectively. (E) Changes in vessel diameter were measured between the pre-stimulation Z-Stack and Z-Stack 2. (F) Changes in vessel diameter were measured between the pre-stimulation Z-Stack and Z-Stack 4. *p < 0.05 with unpaired Student’s t-test, & p < 0.05 relative to 0% change with one-sample Student’s t-test. 20 vessels were analyzed for Control, 25 for NaNP.

Changes in the vessel diameters in comparison to pre-stimulation conditions were quantified over time (Figure 3C,D). Small fluctuations in the diameters of vessels in the control condition were observed, which were attributed to typical ongoing physiological changes in addition to noise from captured images. The experimental group experienced much larger and more positive changes than the control group, and a general upwards trend can be observed. The percent change in vessel diameter was measured between Stacks 0 and 2, and 0 and 4 (Figure 3E,F). There is a statistically significant difference in vessel diameter change between the NaNP and control, with the NaNP group showing an increase in the diameters as a result of the NaNP release. The mean diameter change is still higher than the pre-stimulation condition 20–40 minutes following the end of the release, but this change is no longer statistically different from the control animals. Stack 1 was scanned from the beginning of the stimulation onset, and no difference was found between control and NaNP group, which maybe a result of insufficient time for drug to reach the effective concentration and for vessel to respond (Supplementary Figure 2). Stack 3 was taken at the beginning of the simulation offset, when the NaNP is still present near the implant. A significant elevation in the vessel diameters of the NaNP treated animals was observed compared to the baseline vessel diameters (p<0.05, one-sample Student’s t-test in comparison to 0).

Several of the animals imaged were genetically modified to express green fluorescent protein in their microglia (CX3-CR1 transgenic mice). The microglia in the green fluorescence channels immediately extended processes towards the implant, (Supplementary Figure 3) which is a classical microglia polarization behavior in response to injury[57]. The cyclic voltammetric stimulation used to trigger NaNP release did not appear to elicit an immediate effect on microglia morphology, suggesting the safety of the drug-releasing stimulus. Due to limitations of imaging clarity and the acute implantation setup, we were not able to examine how the vessel modulation affects the microglia on longer timescales.

Next, we examined how vessels of different diameters responded to the local release of NaNP. Vessels were grouped based on their initial diameters, with large vessels having diameters > 60 pixels, medium vessels having diameters 60 ≤ D < 30 pixels, and the remaining vessels > 5 pixels in diameter being categorized as small vessels. It was apparent that small vessels have much greater variation in diameter changes than larger vessels, predominantly due to their lower signal-to-noise ratios. Significant differences in pre-stimulation vessel diameter and vessel diameter during the scan were observed following NaNP release for both large (p < 0.01, one-sample Student’s t-test) and medium vessels (p < 0.05, one-sample Student’s t-test), but not for the small vessels measured. No significant changes were observed for controls, regardless of vessel diameter.

4. Discussion

Vascular dynamics are thought to play a major role in the health of neural tissues in their native states and following injury. Changes in blood flow and blood oxygenation are the driving principles behind fMRI, and in addition to factors such as immune responses, inflammation, and mechanical damage, lack of blood flow can result in the rapid death of host cells. However, there are very few methods of controllably modulating cerebral vascular dynamics at microscales. There are fewer still that can enable simultaneous electrical recording and stimulation.

We have utilized mesoporous silica nanoparticles functionalized with negatively charged sulfonate groups as drug reservoirs for conducting polymers. The drug of choice was NaNP, known to rapidly affect local blood vessels by the release of nitric oxide. We then validated our drug delivery platform in vitro, measuring the concentrations of released NaNP with mass spectroscopy. Release was performed in a dilute solution of tetrabutylammonium chloride, which greatly increase our signal-to-noise ratio over standard saline. Traditionally, drug loading and release from conducting polymers is limited, however, the released masses of NaNP are impressive for conducting polymer-based platforms, on par with previously examined compounds such as fluorescein, rhodamine, and melatonin released from PEDOT/SNP[35].

We then validated that the PEDOT/SNP-NaNP coating was capable of controlling vascular dynamics in vivo. Drug release was performed after baseline measurements of the vasculature were taken, followed by additional measurements after the end of the release. Mice implanted with NaNP-loaded PEDOT/SNP exhibited prominent changes in vessel diameter, corresponding with the start of stimulation. Mice implanted with control electrodes did not have any significant vascular responses to the CV stimulation, indicating that any changes in vessel diameters were likely driven by the release of NaNP, not the electrical stimulation itself.

Measurements taken at different time points appear to indicate that the effects of NaNP release take place during the 20 minutes of stimulation and last for at least 40 minutes following the end of stimulation. Significant differences in the vessel diameters of NaNP-treated animals were apparent after stimulation and were significantly higher than the control animals, and after resting for 40 minutes, the vessel diameters were still significantly greater than the baseline diameters. These results are in line with the rapid onset of action of NaNP[17, 58], which is known to cause vasodilation within seconds of systemic administration through the release of nitric oxide. Interestingly, there was no longer a significant difference in the vessel diameters of NaNP-treated animals and control animals at 60 minutes post-stimulation. This may be due to natural fluctuations in the vessel diameters or diffusion of the NaNP from the site of release. It is important to note that other factors may influence vessel diameter over the course of this experiment, including blood pressure, tissue movement, and physiological responses to the implantation. While care was taken to account for these issues, it is impossible to completely account for all physiological responses.

Vessels were then grouped by size to isolate capillaries from arterioles and venules. We observe that the larger vessels have a far greater capacity for dilation, while smaller capillaries did not respond to the NaNP release. The mechanism of action of NaNP is through the release of nitric oxide, which in turn affects only smooth muscle cells through a signaling cascade originating with guanylate cyclase and the result of which is the relaxation of myosin light chains in the muscle cells[59]. Capillaries do not contain smooth muscle cells, and as such are not affected by the released nitric oxide. The effects of NaNP on vessels as a function of distance from the electrode was not quantified, because the number of the larger vessels was limited and heterogeneously distributed throughout the 820μm x820μm imaging field. Additionally, the largest diameter vessels responded most strongly than smaller vessels as demonstrated in figure 4, and this size effect would confound the distance effect.

Figure 4.

Changes in vessel diameter grouped by vessel size. (A) Percent diameter change of large vessels (diameters greater than 65 μm or 80 pixels, n=6 for control and n= 9 for NaNP). (B) Medium vessels (diameters between 25 and 50 μm or 30 and 60 pixels, n= 7 for control and n=8 for NaNP). (C) Small vessels (diameters less than 25 μm or 30 pixels, n=7 for control and n=8 for NaNP). *p < 0.05 one-sample Student’s t-test **p < 0.01 one-sample Student’s t-test.

We have previously observed that implanted microelectrode arrays for neural recording and stimulation cause vascular remodeling and restricted blood flow in the vicinity of the implant[4–8]. Lack of blood flow and oxygenation can cause neuronal death or silencing leading to interface failure. Local release of vessel dilator could be an effective treatment to mitigate these effects and promote neural tissue health and improve the quality of the neural interfacing. The work reported here was designed to demonstrate the effectiveness of the drug release in modulating vessel diameter in an acute preparation. Future experiments will examine how vessel modulation affects microglia reactivity and neuronal health on longer timescales.

Additionally, our experiments have also demonstrated a reliable method of modulating the vasculature dynamics of the brain, which may find applications in the field of neurovascular coupling and functional magnetic resonance imaging (fMRI). Changes in the energy, oxygen, and nutrient demands in the brain are managed by the central nervous system (CNS) vasculature through neurovascular coupling which is the principal component of fMRI. However, the exact relationships between cortical blood flow and neural activity are debated[60–65]. In addition, many animal studies are performed under anesthesia, which can often also impact blood flow in unique ways[66–69]. The lack of experimental methods of modulating local blood flow complicates the direct examination of neurovascular coupling. While the body of research utilizing fMRI to examine changes in neural activity is large, there has been less effort directed at studying how local changes in blood flow affect CNS cells. The PEDOT/SNP-NaNP microelectrode coating developed here can be used to precisely modulate the vessel diameter and blood flow at the microscale with high temporal control. With further optimization and integration of multielectrode arrays, the technology can provide a powerful tool for the study of neurovascular coupling.

5. Conclusions

We were able to achieve fine temporal control of vascular dynamics by utilizing a CFE coated with NaNP-loaded PEDOT/SNP. The polymer coating served as both a gate to the drug release and a drug carrier. Changes in vascular dynamics were evident within 20 minutes of stimulation. These findings present a novel tool that can be used to control tissue blood perfusion following the implantation of a neural electrode.