Impaired regulation of synaptic strength in hippocampal neurons from GluR1-deficient mice

Abstract

Neurons of the central nervous system (CNS) exhibit a variety of forms of synaptic plasticity, including associative long-term potentiation and depression (LTP/D), homeostatic activity-dependent scaling and distance-dependent scaling. Regulation of synaptic neurotransmitter receptors is currently thought to be a common mechanism amongst many of these forms of plasticity. In fact, glutamate receptor 1 (GluR1 or GluRA) subunits containing L-alpha-amino-3-hydroxy-5-methylisoxazole-4-propionate (AMPA) receptors have been shown to be required for several forms of hippocampal LTP and a particular hippocampal-dependent learning task. Because of this importance in associative plasticity, we sought to examine the role of these receptors in other forms of synaptic plasticity in the hippocampus. To do so, we recorded from the apical dendrites of hippocampal CA1 pyramidal neurons in mice lacking the GluR1 subunit (GluR1 -/-). Here we report data from outside-out patches that indicate GluR1-containing receptors are essential to the extrasynaptic population of AMPA receptors, as this pool was nearly empty in the GluR1 -/- mice. Additionally, these receptors appear to be a significant component of the synaptic glutamate receptor pool because the amplitude of spontaneous synaptic currents recorded at the site of input and synaptic AMPA receptor currents evoked by focal glutamate uncaging were both substantially reduced in these mice. Interestingly, the impact on synaptic weight was greatest at distant synapses such that the normal distance-dependent synaptic scaling used by these cells to counter dendritic attenuation was lacking in GluR1 -/- mice. Together the data suggest that the highly regulated movement of GluR1-containing AMPA receptors between extrasynaptic and synaptic receptor pools is critically involved in establishing two functionally diverse forms of synaptic plasticity: LTP and distance-dependent scaling.

Figure 3. Distance-dependent scaling of synaptic current amplitude is attenuated in GluR1 −/− mice

Schematic diagrams of (A) proximal and (B) distal dendritic voltage-clamp recording configurations. C-F, representative recordings of hypertonically-evoked synaptic activity from proximal and distal dendrites in wild-type and in GluR1 −/− mice. G, averages of 100-165 individual mEPSCs from each of the recordings shown above. p and d represent proximal and distal recordings, respectively. H, grouped data of mean mEPSC amplitude for all cells. Numbers of cells are shown above each bar. I, mEPSC rise-time constants in GluR1 −/− mice are slightly faster than in WT mice, and this increase is independent of synaptic location (WT: 175 ± 9 μs, n = 11; KO: 134 ± 7 μs, n = 14, * P < 0.001). J, no differences are detected in mEPSC decay-time constants (WT: 4.3 ± 0.2 ms, n = 11; KO: 4.7 ± 0.4 ms, n = 14).

Figure 1. The extrasynaptic pool of AMPA and NMDA receptors is severely altered in GluR1 −/− mice

Representative AMPA and NMDA receptor current traces, evoked by rapid glutamate applications to dendritic outside-out patches, are shown for proximal (A and C) and distal patches (B and D) from wild-type (A and B) and GluR1 −/− mice (C and D). Currents are the averages of 3-5 individual traces. Mean AMPA (E) and NMDA (F) receptor current amplitudes are shown for all groups. Representative non-stationary fluctuation analysis (NSFA) of distal patches from WT (G) and KO (H) mice. Note that patches from KO mice have a dramatic reduction in receptor number (N), and maximum open-probability (P_o,max) (I), whilst the single-channel conductance (γ) is the same in both groups (J). K, current-voltage relationships of AMPA receptor currents are similar between WT (squares) and KO (circles) mice. WT: E_rev = 5.0 ± 0.5 mV, n = 11; KO: E_rev = 2.0 ± 1.0 mV, n = 4.

Figure 2. Some kinetic properties of AMPA currents are altered in GluR1 −/− mice

A, rise times of outside-out patch currents are not altered (WT: 457 ± 34 μs, n = 16; KO: 568 ± 50 μs, n = 8; P > 0.05), whilst, as shown in B, the deactivation-time constants (WT: 3.2 ± 0.2 ms, n = 8; KO: 2.2 ± 0.2 ms, n = 12, * P < 0.001) and, as shown in D, the fast component of desensitization (WT: 8.1 ± 0.7 ms, n = 8; KO: 5.6 ± 0.4 ms, n = 12, * P < 0.05) are significantly faster in GluR1 −/− mice. G, the fast component of desensitization is also much more prominent KO mice (WT desensitization ratio 22 ± 2 %; KO desensitization ratio 82 ± 20 %, * P < 0.001). Representative AMPA receptor currents, evoked by 1 (C) and 100 ms (F) glutamate pulses (1 mM), are shown for WT and KO dendritic patches.

Figure 4. Synaptic current distributions from GluR1 −/− mice show altered properties

A, plot of CV for all four groups of mEPSCs. Numbers shown on plot are mean 1/CV², a measure that is indicative of quantal content. B, plot of σ²/x for the four groups of mEPSCs. This measure is indicative of quantal size and the plots indicate that the normal distance-dependent increase in this parameter is absent in GluR1 −/− mice. Increases in σ²/x from proximal locations to distal (distal/prox) are displayed on the plot. C-F, mEPSC amplitude distributions of proximal and distal synapses from wild-type (WT) and knockout mice (KO). Between 135 and 204 events were plotted using a bin size of 1 pA. Solid lines are fits of data by the sum of three Gaussians with individual mean peaks at WT-prox x1 = 6, x2 = 10, x3 = 20 pA; WT-distal x1 = 13, x2 = 25, x3 = 42 pA; KO-prox x1 = 4, x2 = 8, x3 = 15 pA; KO-distal x1 = 6, x2 = 12, x3 = 17 pA. G, plot of mean peak amplitudes from the multiple Gaussian fits for each group of distributions. Numbers on the plot are the ratio of WT distal to all other group peak amplitudes, suggesting again that distal synapses in GluR1 −/− mice lack an approximately 2.5-fold increase in quantal size. H, cumulative frequency distribution of all events from WT (blue) and GluR1 −/− (red) mice. Continuous lines are proximal and dashed lines are distal distributions. I, pair-pulse facilitation ratios (EPSC_2nd/EPSC_1st) are shown at four different interstimulus intervals, recorded and stimulated at proximal (filled circle) and distal (open circle) dendritic regions, from WT (blue) and KO (red) mice. Prox, proximal.

Figure 5. The synaptic pool of AMPA receptors is reduced in a location-dependent manner in GluR1 −/− mice

A, image stack spanning the entire apical dendritic arborization of another CA1 pyramidal neuron. Ovals represent distal and proximal recording sites in stratum radiatum (S. Rad). Dashed lines demonstrate the proximal stratum pyramidale (S. Pyr.) and the distal stratum lacunosum-moleculare (S. Lac. Mol.) borders of stratum radiatum. B, location-dependent AMPA currents from WT and KO mice, averages of 2-3 traces. Traces are fitted by the sum of two exponentials, with blue lines for WT and red for KO mice. Note that the distance-dependent increase in AMPA current is missing in the recordings from the KO mice. C, multi-photon image stack of a distal dendritic region of a CA1 pyramidal neuron filled with bis-fura2. The dendritic recording electrode is shown on the left and the coloured arrows indicate the isolated spines that gave the correspondingly coloured MNI-glu currents shown to the right. D, examples of AMPA receptor currents evoked by focal uncaging of MNI-glutamate (MNI-glu) onto isolated spines located on the main dendritic trunk and a nearby oblique dendrite (branch to the right), as indicated by coloured arrows shown in C. E, mean AMPA current amplitudes for wild-type (blue bars) and GluR1 −/− (red bars) mice. F, the spine head volumes were the same in both groups of mice in both locations. G and H, in all groups of MNI-glu currents have similar time constants of rise (WT: 1.95 ± 0.1 ms, n = 29; KO: 2.0 ± 0.1 ms, n = 23) and decay (WT: 5.85 ± 0.4 ms, n = 29; KO: 6.1 ± 0.5 ms, n = 23).

Figure 6. Comparison of alterations in the extrasynaptic and synaptic receptor pools

A, decrease in amplitude observed in GluR1 −/− mice for AMPA currents recorded in outside-out patches (patches), glutamate uncaging (MNI-glu) and mEPSCs. Note that mEPSC and MNI-glu currents are similarly reduced whilst patch currents are reduced to a much greater extent. p, proximal currents; d, distal currents. B, plot of the increase in distal current amplitude for mEPSCs and MNI-glu currents in both wild-type and GluR1 −/− mice. Notice the increases in mEPSCs amplitudes are mirrored by the increases in MNI-glu current.