Abstracts

Rationale: Localization-related epilepsies represent ~ 80% of the patients with seizures, > 1/3 of whom are refractory to medications. Seizure-free rates after epilepsy surgery range from ~40% to 70%, and less for those treated with implantable devices. A growing body of evidence suggests that understanding the dynamics of neural circuits involved in seizure generation and precisely locating the epileptogenic area to resect or stimulate requires monitoring the activity of microscale neural circuits at millisecond resolution. In this work we present the design, fabrication and initial in vivo validation of a minimally invasive neurotechnology based on microfluidic control and actuation of flexible microelectrodes (μFAEs) for dynamic mapping the activity of epileptic networks at high spatial and temporal resolution.  Methods: The µFAEs consist of thin microelectrodes with cellular scale size (20-50 µm width) enclosed in flexible, high-density microfluidic channels. The combination of ultrathin electrodes and micron-size channels allows to dynamically advance and retract the electrode while minimizing device footprint. Prototypes of two-channels μFAE were fabricated in polydimethylsiloxane (PDMS) by conventional replica-molding soft-lithography and equipped with on-chip valves for actuation and control of the enclosed microelectrodes. Electrochemical impedance spectra were acquired using a Gamry Reference 300 potentiostat. To evaluate whether our microfluidic-based approach could be used to implant the flexible microelectrodes and record physiological and epileptiform activity in vivo, we performed acute recordings experiments in naïve and epileptic rodents. Adult Sprague-Dawley rats were anesthetized and a 4x4 mm2 craniotomy was opened in the right hemisphere to expose the somatosensory cortex and remove the dura. A skull screw placed in the left parietal bone served as reference. Electrocorticography (ECoG) signal was initially recorded with the μFAE placed on the craniotomy and the microelectrode retracted inside the channel. Intracortical (IC) recordings were obtained by activating the flow control system to actuate and implant the microelectrode to a depth of ~1 mm. Following a similar experimental procedure, we recorded seizure-like discharges in mice induced by 4-aminopyridine (4AP), a selective blocker of voltage-activated K+ channels generating tonic-clonic ictal focal electrographic seizures when applied to neocortex.  Results: The average impedance at 1 kHz is 534 ± 114 kΩ (n=10, device fabrication yield 85%). In vivo LFP recordings showed slow waves (0.5-2 Hz) and alpha (8-15 Hz) oscillations, which are characteristic of the anesthetized state. IC recordings show MUA modulated by the phase of the slow wave (i.e., increasing on the upslope of each slow-wave cycle). Visual inspection of the microelectrode after disconnection from the fluidic and recording systems confirmed successful implantation into the cortical tissue. Acute mice recordings during 4AP-induced epileptiform activity show distinct interictal spikes and seizure-like discharges during actuation from the epidural surface to 1 mm inside the somatosensory cortex.  Conclusions: In conclusion, our results show that the microfluidic actuation technology allows precise insertion and control of flexible microelectrodes, while recording spontaneous and drug-induced epileptiform activity in vivo. Ongoing and future work will assess the long-term recording performance and biocompatibility of μFAE for dynamic mapping seizure discharges in chronic models of focal epilepsy. Funding: This work was funded by the Citizen United for Research in Epilepsy (CURE) Taking Flight Award to F.V.