Differential Astrocyte-Supplied NMDAR Co-Agonist for CA1 Versus Dentate Gyrus Long-Term Potentiation

Abstract

In the hippocampus, there is a region- and synapse-specific N-methyl-D-aspartate receptor (NMDAR) co-agonist preference for induction of long-term potentiation (LTP). Schaffer collateral (SC)-CA1 synapses, enriched in GluN2A-containing NMDARs, favor D-serine, while medial perforant path (MPP) to dentate gyrus (DG) synapses that are rich in GluN2B-containing NMDARs prefer glycine for LTP induction. This study investigated the role of astrocytes in providing these co-agonists. We confirmed in rat hippocampal slices that exogenous D-serine (10 μM) is sufficient to restore LTP at SC-CA1 synapses blocked under astrocyte calcium (Ca²⁺) -clamp conditions, consistent with previous findings. However, exogenous glycine (10 μM) also rescued the LTP. In contrast, at MPP-DG synapses, 100 μM exogenous glycine, but not 10 μM nor 100 μM D-serine, restored the LTP blocked by astrocyte Ca²⁺-clamping. Our findings support the view that, as for serine in CA1, astrocytes are the cellular source of the glycine required for LTP induction at MPP-DG synapses.

Keywords: D‐serine; LTP; calcium‐clamp; glycine; hippocampus.

FIGURE 1

(A) Representative confocal image of astrocytes labeled with SR101 using a 10× water immersion objective. (B) Representative images showing differential interference contrast (DIC) and corresponding fluorescence image of an SR 101 patched astrocyte in MML (glass pipette shown in white dotted line) using a 40× water immersion objective lens. (C) Schematic of 100 pA current steps used to generate V _m deflections and the corresponding traces of current–voltage relation from a single whole‐cell patch‐clamped astrocyte. (D) Representative evoked field response in MML recorded through an astrocytic membrane (AfEPSP) following test pulse stimulation of the MPP inputs. Red line indicates the slope measurement. The positive overshoot of the waveform after the initial field response was routinely observed in both CA1 and the DG and is likely due to the longer‐lasting inward potassium (K⁺) current (Henneberger and Rusakov 2012). (E, F) Representative average waveforms (4 sweeps) of evoked synaptic field potentials recorded through the astrocyte membrane at a 50‐ms inter‐pulse interval showed either paired‐pulse facilitation or paired‐pulse depression when two pulses were delivered at SC‐CA1 (E) and MPP‐DG (F) synapses, respectively. (G) Clamping astrocytic Ca²⁺ with EGTA blocked LTP at nearby SC synapses in SR of CA1 in a D‐serine (10 μM) as well as glycine (10 μM)‐dependent manner. (H) Clamping astrocytic Ca²⁺ blocks LTP at nearby MMP synapses in MML of DG in a specifically glycine‐dependent manner. LTP was measured in the presence and absence of clamping and by supplying either 10 μM, 100 μM D‐serine, or 100 μM glycine under Ca²⁺ clamped conditions. (G, H) Inset: Bar graphs summarizing the LTP measured via the whole‐cell patch‐clamped astrocyte in the presence and absence of Ca²⁺ clamping. Representative waveforms are an average of 10 synaptic responses prior to LTP induction (1) and at the conclusion of the experiment (2). All scale bars: 0.5 mV and 2 ms. The arrow indicates the timing of LTP induction. All data presented as mean ± SEM; ns, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.