Neuroplasticity, the ability of the brain to form new neural pathways and synaptic connections, is gaining renewed interest for therapeutic purposes in stroke or spinal cord injury patients. Earlier work in this area in animal models has shown that behavior modulation was associated with the release of serotonin that subsequently leads to changes in gene expression and new neural growth [1]. A separate work has also demonstrated that using serotonin 1A agonists in rats with respiratory dysfunction due spinal cord injury results in better respiratory function [2]. Taken together, these studies indicate that neurotransmitters such as serotonin are closely linked to neuroplasticity, suggesting that changes in levels of these neurotransmitters can be correlated to synaptic changes.

Currently, there is growing interest in the use of electrical stimulation as a means for inducing neuroplasticity in stroke and spinal cord injury patients [3].  However, while the data suggests that plasticity could be induced by electrical stimulation, neither the mechanism nor the protocols of stimulation are understood. We submit that a possible path to understanding these mechanisms involves simultaneous real-time monitoring of neurotransmitters such as serotonin electrochemically and electrical signals during stimulation and finding a correlation between these electrical signals and neurotransmitters. In cases where detection of several key electroactive neurotransmitters like serotonin and dopamine is sought, fast scan cyclic voltammetry (FSCV) with its high temporal resolution is the preferred electrochemical method of obtaining real time data [4]. Now, to be able to simultaneously monitor these electrochemical signals along with electrical signals, new approaches in neural probe material and architecture selection are required, as almost all neural probes currently in use are capable of recording only electrical or chemical signals, but not both.  Thus, implementation of electrochemical monitoring techniques in animal models requires an integrated electrode array with a size suitable for an animal containing a counter electrode, working electrodes and a reference electrode. Furthermore, electrode materials capable of electrochemically detecting the presence of neurotransmitters without undergoing non-reversible reactions are sought. To meet this need, we investigate and report on the use of glassy carbon, a form of carbon that has been demonstrated to possess ideal characteristics, such as a wide electrochemical window with no irreversible reactions as a coupled electrical and electrochemical sensing platform. In this particular study, we introduce a 12-channel glassy carbon (GC) microelectrode array fabricated on a flexible substrate. The specific design introduced here consists of electrodes supported on a polyimide substrate with penetrating probes at their edges that can penetrate up to 1.25 mm into an animal's  cortex or spinal cord. Each probe contains a glassy carbon electrode that can be used to detect serotonin, stimulate a neuron, or read an electrophysiological signal. Our preliminary results show that serotonin can be detected using a polyimide-glassy carbon device down to at least 1 uM in vitro by stimulating through one portion of the device and reading through another (Figure 1). Furthermore, we have shown that our electrodes can read electrophysiological data in-vivo. Figure 2b shows the Fourier Transform of a few seconds of electrical brain signals from a song bird without stimulus. In this graph low frequency alpha waves are most prominent. Figure 2c, the stimulus graph, on the other hand, shows that the beta and gamma waves became more prominent under external, audio stimulation. We believe that the ability to couple this kind of electrical signal detection with electrochemical sensing through FSCV (fast-scan CV) in a single device for in-vivostudies give promise for the disambiguation behind the mechanisms involved in neuroplasticity. 

Figure 1: Background subtracted cyclic voltagramms for various concentrations of serotonin

Figure2: In vivo ECoG signals recorded with carbon electrodes. (a) Time lapse of the audio stimulus (b) Amplitude spectra of the animal during quiet time (c) Amplitude spectra of the signal during the stimulus phase. 

[1] Kandel, Eric R. Science 294, no. 5544 (2001): 1030-1038.

[2] Choi, Howard, Wei-Lee Liao, Kimberly M. Newton, Renna C. Onario, Allyson M. King, Federico C. Desilets, Eric J. Woodard et al. The Journal of neuroscience 25, no. 18 (2005): 4550-4559.

[3] Dimyan, Michael A., and Leonardo G. Cohen. Nature Reviews Neurology 7, no. 2 (2011): 76-85.

[4] Lama, Rinchen D., Karl Charlson, Arun Anantharam, and Parastoo Hashemi. Analytical chemistry 84, no. 19 (2012): 8096-8101.