Enhanced astrocytic d-serine underlies synaptic damage after traumatic brain injury
Enmanuel J. Perez, … , Joseph T. Coyle, Daniel J. Liebl
Enmanuel J. Perez, … , Joseph T. Coyle, Daniel J. Liebl
Published July 17, 2017
Citation Information: J Clin Invest. 2017;127(8):3114-3125. https://doi.org/10.1172/JCI92300.
View: Text | PDF
Research Article Neuroscience

Enhanced astrocytic d-serine underlies synaptic damage after traumatic brain injury

Abstract

After traumatic brain injury (TBI), glial cells have both beneficial and deleterious roles in injury progression and recovery. However, few studies have examined the influence of reactive astrocytes in the tripartite synapse following TBI. Here, we have demonstrated that hippocampal synaptic damage caused by controlled cortical impact (CCI) injury in mice results in a switch from neuronal to astrocytic d-serine release. Under nonpathological conditions, d-serine functions as a neurotransmitter and coagonist for NMDA receptors and is involved in mediating synaptic plasticity. The phasic release of neuronal d-serine is important in maintaining synaptic function, and deficiencies lead to reductions in synaptic function and plasticity. Following CCI injury, hippocampal neurons downregulated d-serine levels, while astrocytes enhanced production and release of d-serine. We further determined that this switch in the cellular source of d-serine, together with the release of basal levels of glutamate, contributes to synaptic damage and dysfunction. Astrocyte-specific elimination of the astrocytic d-serine–synthesizing enzyme serine racemase after CCI injury improved synaptic plasticity, brain oscillations, and learning behavior. We conclude that the enhanced tonic release of d-serine from astrocytes after TBI underlies much of the synaptic damage associated with brain injury.

Authors

Enmanuel J. Perez, Stephen A. Tapanes, Zachary B. Loris, Darrick T. Balu, Thomas J. Sick, Joseph T. Coyle, Daniel J. Liebl

×

Figure 4

Synaptic dysfunction and learning deficits are not observed in aSRKO CCI-injured mice at 7 dpi.

Options: View larger image (or click on image) Download as PowerPoint
Synaptic dysfunction and learning deficits are not observed in aSRKO CCI...
(A) I/O curve showed reduced fEPSP amplitude at high intensities for cSR CCI-injured mice as compared with sham-treated and CCI-injured aSRKO mice. Individual traces shown. (B) PPF showed a reduced ratio for cSR CCI-injured mice as compared with sham-treated and CCI-injured aSRKO mice. (C) LTP was reduced in CCI-injured cSR control mice, but not sham-treated cSR, sham-treated aSRKO, or CCI-injured aSRKO mice. Individual traces shown. (D) Histograph of LTP early phase showed greatest reductions in CCI-injured cSR mice and reductions to a lesser extent in aSRKO mice. (E) Histograph of LTP late phase showed greatest reductions in CCI-injured cSR mice and reductions to a lesser extent in aSRKO mice. Early and late phase slopes of CCI-injured aSRKO mice were significantly greater than CCI-injured cSR mice at 7 dpi. (F) θ rhythm frequency was increased in CCI-injured cSR mice and was increased to a lesser extent in sham-treated and CCI-injured aSRKO mice at frequencies between 4 and 5 Hz as compared to sham-treated cSR mice. θ rhythm frequency of CCI-injured aSRKO mice were significantly less than CCI-injured cSR mice. θ rhythm traces were shown for amplitude versus time. Fear-conditioning tests showed significantly decreased freezing behavior in CCI-injured cSR mice for contextual (G) and cued (H) learning that was not observed in sham-treated or CCI-injured aSRKO mice. Data represent mean ± SEM. (A–F) n = 7–10/group. (G and H) n = 16–18/group. (A–C, F–H) Two-way RM ANOVA. (D and E) One-way ANOVA. *P < 0.05, **P < 0.01; *** P < 0.001, compared with cSR control mice. #P < 0.05; ##P < 0.01; ###P < 0.001, compared with cSR at 7 dpi.