br Introduction L Glutamate is one

Introduction L-Glutamate is one of the most prevalent excitatory signaling molecules in the central nervous system (CNS) (Salazar et al., 2016) and plays a crucial role in a variety of Atropine synthesis functions such as memory and learning (Danbolt, 2001, Nedergaard et al., 2002). At elevated concentrations, glutamate exhibits excitotoxic properties that are implicated in a variety of neurological disorders including ischemic stroke (Camacho and Massieu, 2006), epilepsy (Babb et al., 1998), and neurodegenerative diseases (Mehta et al., 2013). Basal glutamate concentrations in vivo range from 0.9 to 3.7 µM (Day et al., 2006, Stephens et al., 2011, Vasylieva et al., 2013), however during pathological situations glutamate levels increase several fold (Wahl et al., 1994, Lee et al., 2009). Monitoring glutamate in real-time is key to understanding its normal and pathological functions. Traditionally, in vivo glutamate levels were monitored using microdialysis (Windels et al., 2000) but even the fastest measurements have a time resolution of only 15 s (Venton et al., 2006). Moreover, microdialysis probes cause significant damage to brain tissue due to the larger size, with a diameter from 200 to 500 µm and length of 1–4 mm (Hascup et al., 2009, Hascup and Hascup, 2014). Therefore, small sensors with a fast time response and large linear range are required to measure real-time glutamate release in vivo. Glutamate is not electroactive and cannot be directly detected by voltammetry but enzyme-based electrochemical biosensors can monitor glutamate continuously (Hu et al., 1994, Özel et al., 2014, Vasylieva et al., 2013). Many glutamate biosensors utilize glutamate oxidase (GlutOx) to enable real-time glutamate measurements in the brain. GlutOx metabolizes glutamate and releases hydrogen peroxide, which is electroactive and can be detected electrochemically (Özel et al., 2014). GlutOx sensors are dependent on oxygen, but operate well at normoxic conditions (Clay and Monbouquette, 2018). Glutamate dehydrogenase has also been employed in biosensor applications and is not oxygen dependent; however, it has longer response times, lower sensitivity, and requires a NAD+ cofactor (Hughes et al., 2016). Several different glutamate biosensors have been developed in recent years. For physiological applications, enzymes are typically coated on platinum wire electrodes (Batra et al., 2016, Hu et al., 1994, Özel et al., 2014). Commercially available GlutOx-based biosensor designs are available (Pinnacle Technology, Inc. USA), but they are larger than optimal for small model organisms (176 µm in diameter). Polymer-based, flexible glutamate sensors were developed for in vivo applications (Weltin et al., 2014) with a sensing core of only 100 µm, but an overall width of 500 µm. Microelectrode arrays for glutamate were designed with individual sensors that are less than 100 µm in size, and multiple sensors in a single assembly allow measurements of different analytes in parallel (Tolosa et al., 2013, Wassum et al., 2008). However, the overall width of the sensor at the site of implantation is on the order of a millimeter. Gerhardt's group developed multisite ceramic microelectrodes for in vivo glutamate detection in rats and mice (Burmeister et al., 2002, Hascup et al., 2008, Rutherford et al., 2007). These microelectrode arrays are triangular in shape with an active site on the order of 50–150 µm, but the actual taper starts from 1 mm. A smaller design would not only minimize the tissue damage during biosensor implantation and removal, but also would allow more precise targeting of brain regions. In this study, we developed a microelectrode that is only 50 µm in diameter for measuring glutamate in vivo and in brain slices. The biosensor design includes electrochemical deposition of PPD, immobilization of GlutOx entrapped in a chitosan matrix, and a layer of ascorbate oxidase. Analytical performance was tested in vitro and the electrodes exhibited excellent sensitivity and selectivity for glutamate. Electrodes were batch fabricated, which is important for making new, disposable sensors for in vivo use. The response time was fast (2 s) and the limit of detection (0.044 µM) was sufficient for in vivo measurements. The biosensor was tested for real-time measurement of glutamate release in vivo in the subthalamic nucleus as well as in brain slices.