Alterations in Hippocampal Network Activity after In Vitro Traumatic Brain Injury

Abstract

Traumatic brain injury (TBI) alters function and behavior, which can be characterized by changes in electrophysiological function in vitro. A common cognitive deficit after mild-to-moderate TBI is disruption of persistent working memory, of which the in vitro correlate is long-lasting, neuronal network synchronization that can be induced pharmacologically by the gamma-aminobutyric acid A antagonist, bicuculline. We utilized a novel in vitro platform for TBI research, the stretchable microelectrode array (SMEA), to investigate the effects of TBI on bicuculline-induced, long-lasting network synchronization in the hippocampus. Mechanical stimulation significantly disrupted bicuculline-induced, long-lasting network synchronization 24 h after injury, despite the continued ability of the injured neurons to fire, as revealed by a significant increase in the normalized spontaneous event rate in the dentate gyrus (DG) and CA1. A second challenge with bicuculline 24 h after the first challenge significantly decreased the normalized spontaneous event rate in the DG. In addition, we illustrate the utility of the SMEA for TBI research by combining multiple experimental paradigms in one platform, which has the potential to enable novel investigations into the mechanisms responsible for functional consequences of TBI and speed the rate of drug discovery.

Keywords: electrophysiology; hippocampus; network synchronization; traumatic brain injury.

FIG. 1.

Images of an SMEA. (A) The SMEA featured 28 electrodes and two reference electrodes in a 49×49 mm package. (B) Image of a hippocampal slice culture on an SMEA before stretch injury. (C) Image of a hippocampal slice culture on an SMEA after stretch injury of approximately 0.2 strain and 2 s^–1 strain rate. (D) Image of the 28-electrode array in the center of the SMEA. The tips of the patterned conductors were exposed through 100×100 μm vias photopatterned in the encapsulation layer. The four small squares in the center are registration marks for aligning photolithographic masks. Individual electrode ID assignments are indicated in white. SMEA, stretchable microelectrode array. Color image is available online at www.liebertpub.com/neu

FIG. 2.

Neither network synchronization of spontaneous activity nor the normalized spontaneous event rate was significantly affected by injury. (A) Network synchronization, as measured by the normalized global synchronization index (GSI), was not significantly affected by injury either acutely or 24 h postinjury in DG, CA3, or CA1. (B) The normalized spontaneous event rate was not significantly altered by injury in DG, CA3, or CA1, either acutely postinjury or 24 h postinjury. All data were normalized to preinjury, pretreatment levels (mean±standard error of the mean). DG, dentate gyrus; CA, cornu amonis.

FIG. 3.

Representative traces of temporally aligned raw electrophysiology data from four electrodes in CA1 before bicuculline treatment and during bicuculline treatment from uninjured (A) and injured (B) slice cultures.

FIG. 4.

Changes in bicuculline-induced, long-lasting network synchronization of spontaneous activity in uninjured and injured slice cultures. Representative raster plots of spontaneous activity and heat maps of pair-wise synchronization c^τ(x|y) for every electrode pair are shown for uninjured and injured slice cultures at the indicated time points: preinjury (or sham exposure) and before bicuculline treatment (A and D), during bicuculline treatment (B and E), and 24 h after bicuculline treatment (C and F). Each line in the raster plots represent a distinct, identified neural event. Heat maps of pair-wise synchronization depict the event synchronization index for each electrode pair, ranging in value from 0 (completely uncorrelated, blue) to 1 (perfectly correlated, red). Color image is available online at www.liebertpub.com/neu

FIG. 5.

Changes in bicuculline-induced, long-lasting network synchronization of spontaneous activity in uninjured and injured slice cultures, quantified by the normalized GSI. Preinjury (or sham exposure) and bicuculline treatment, network activity was not synchronized in any region (DG, CA3, or CA1), with the normalized GSI below 0.01 (A–C). Acutely during bicuculline exposure, the normalized GSI increased significantly in all hippocampal regions in both uninjured and injured slice cultures, compared to their respective baseline recordings, indicating significantly higher network synchronization. Twenty-four hours after bicuculline exposure, the normalized GSI remained significantly higher in all hippocampal regions in uninjured slice cultures, compared to pretreatment baseline levels. In all regions of injured slice cultures, the normalized GSI was significantly diminished 24 h after bicuculline exposure, when compared to the normalized GSI during bicuculline treatment, and when compared to uninjured slice cultures 24 h after bicuculline treatment. Data are presented as mean±standard error of the mean. GSI, global synchronization index; DG, dentate gyrus; CA, cornu amonis.

FIG. 6.

Normalized spontaneous event rate before and after bicuculline treatment in uninjured and injured slice cultures. Twenty-hour hours after bicuculline exposure, the normalized spontaneous event rate was significantly increased in injured DG (A) and CA1 (C) compared to pretreatment, preinjury baseline levels and compared to uninjured DG and CA1 at the same time point. No significant changes in the normalized spontaneous event rate were observed in CA3 (B). All data were normalized to preinjury, pretreatment levels (mean±standard error of the mean). DG, dentate gyrus; CA, cornu amonis.

FIG. 7.

Changes in network synchronization of spontaneous activity and the normalized spontaneous event rate in injured slice cultures. (A) Second exposure to bicuculline 24 h after the initial bicuculline exposure significantly increased the normalized GSI compared to preinjury, pretreatment baseline levels and compared to 24 h postinjury and the initial bicuculline exposure in DG, CA3, and CA1. The normalized GSI was not significantly different between hippocampal regions after the second bicuculline exposure. (B) Second exposure to bicuculline 24 h after the initial bicuculline exposure produced different effects on the normalized spontaneous event rate depending on hippocampal region. Compared to 24 h, re-exposure to bicuculline significantly decreased the normalized spontaneous event rate in DG, while significantly increasing the normalized spontaneous event rate in CA3 and CA1. All data were normalized to preinjury, pretreatment levels (mean±standard error of the mean). GSI, global synchronization index; DG, dentate gyrus; CA, cornu amonis.