Article Figures & Data
Figures
- Figure 1.
mGlu2-activated outward current in GCs in rat acute slices. A, Average holding current and input resistance of GCs after application of 1 μm DCG IV. Data points indicate values averaged over 5 s and gray bars are the SEM (n = 23). B, Outward currents induced by the mGlu-II agonists DCG IV (1 μm), APDC (50 μm), and LY395756 (5 μm). Other columns show the block of DCG-IV-induced currents in GCs in the presence of the mGlu-II antagonists APICA (500 μm) and LY341495 (0.5 μm), the GIRK channel blocker tertiapin-Q (0.5 μm), and an intracellular solution containing the G-protein inhibitor, GDP-β-S (0.4 mm). C, DCG-IV-induced currents in anatomically identified hippocampal neurons. PC indicates pyramidal cells; PV+BC, parvalbumin-expressing basket cells in the DG; RSBC, regular spiking basket cells in the DG; NGFC, neurogliaform cells in the DG. D1, The mGlu-II specific antagonist APICA blocked the DCG-IV-induced current in GCs. After washout of the antagonist, agonist application resulted in an outward current. D2, Effect of tertiapin-Q on DCG-IV-induced currents during incubation with the toxin (left trace, n = 6 cells) and during subsequent application (right trace, n = 5 cells from separate experiments). D3, DCG IV effects in simultaneously recorded pairs of GCs with normal intracellular solution and with an added G-protein blocker, GDP-β-S. D4, DCG IV effects in simultaneously recorded pairs of GCs and neurogliaform cells. E, Reversal potential of the DCG-IV-induced current in GCs (average current between −100 and −55 mV, n = 8 cells). The gray line indicates the linear fit for the current values between −100 and −80 mV. Dashed line shows the calculated reversal potential for potassium. F, Inward rectification of the DCG-IV-induced current. Black and gray symbols mark the average whole-cell current (n = 3 cells) in control solution and in presence of DCG IV, respectively. The inset trace is the difference current.
- Figure 2.
mGlu2-activated outward current in GCs in organotypic slice cultures. A, Bath application of the mGlu-II-specific agonist DCG IV (2 μm) induced outward currents in GCs from organotypic slices from wild-type and mGlu3−/− mice. This current was not present in GCs from mGlu2−/− mice. B, DCG-IV-induced currents in GCs in organotypic slice cultures of C57BL/6 mice, white mice, rats, and mGlu2- and mGlu3-lacking mice. The lighter column shows the DCG-IV-induced currents in acute slices from C57BL/6 mice. C, The mGlu antagonist LY341495 prevented the outward current in C57BL/6 mice. An NMDA antagonist (APV, 40 μm), a sodium channel blocker (TTX, 1 μm), and a potassium channel blocker (α-DTX, 2 μm) did not affect the DCG-IV-induced current. The GIRK channel blocker tertiapin-Q (0.5 μm) reduced the DCG-IV-induced current. D, Voltage dependence of the DCG-IV-induced current in GCs in organotypic slice cultures (n = 3 cells). The black line indicates the linear fit for the current values between −85 and −65 mV. Dashed line shows the calculated reversal potential for potassium.
- Figure 3.
Dendritic origin of the mGlu2-activated outward current. A, Averaged holding currents from whole-cell-recorded GCs, large MFBs in the CA3 region, nucleated patches, and whole-cell-recorded GC dendrites from rat acute slices during DCG IV (1 μm) application. Insets show representative images of the recorded subcellular compartments. Scale bars, 50, 10, and 10 μm. B, The DCG-IV-induced outward current and relative change in input resistance in control GCs, in whole-cell recorded MFBs, in GCs with cut axons (<50 μm), in GC dendrites, and in nucleated patches pulled from GCs. C, Effect of DCG IV on isolated calcium currents recorded from MFBs. Representative traces are shown on the left and the voltage dependence of the absolute and normalized currents on the right. D, The simulated changes in local steady-state input resistance with various shunt locations (soma: black, 80 μm light gray, 120 μm dark gray, 200 μm gray) show the expected changes in local input resistance throughout the entire dendritic arborization of a mulicompartmental GC cell model (Schmidt-Hieber et al., 2007). Symbols show the measured changes in three different subcellular domains (open circles: individual experiments, black circles: averages).
- Figure 4.
Localization of mGlu2 receptors in the DG. A, Immunoreactivity for the mGlu2 receptor in the DG of a mGlu3-KO mouse. The immunolabeling is particularly intense in the molecular layer at ∼30 μm from the border of the GC layer and it decrements toward the outer layer. ML indicates molecular layer of the DG; LM, stratum lacunosum moleculare of CA1 region. B, Immunoreactivity for the mGlu3 receptor in the DG of an mGlu2-KO mouse. The staining appears homogeneous throughout the molecular layer. C, Immunoreactivity for mGlu2 and mGlu3 in the DG of a wild-type mouse. D, Left, Immunofluorescence labeling for the mGlu2 receptor in the hippocampal formation of an mGlu3-KO mouse. Middle, Surface plot traces of the immunofluorescence signal (relative optical density, R.O.D.) across the DG (sampled as shown by the blue bar in the left panel) obtained from different mGlu2-KO animals (n = 4). The distance was normalized between the borders of the hilus and the GC layer (0) and the DG molecular layer and CA1 (1). Right, Averaged R.O.D. distribution of mGlu2 immunoreactivity. E–F, Subcellular distribution of mGlu2 (i.e., in mGlu3-KO mice) in GC dendrites. Immunoreactivity for mGlu2, visible as an electron opaque peroxidase end product (arrows), can generally be detected at some distance from the cell body in the proximal portion of GC dendrites, whereas somata are mostly devoid of labeling. G, Immunoperoxidase end product for mGlu2 can be seen unevenly distributed at and near the plasma membrane of GC dendrites in the inner molecular layer (arrows). H, Immunometal particles marking mGlu2 receptors (arrows) in presynaptic terminals of putative entorhinal cortex axons were mostly observed outside the active zone. I, Similar to the immunoperoxidase end product, immunometal particles for mGlu2 were observed unevenly distributed at and near the plasma membrane of GC dendrites in the inner molecular layer (arrows). J, Labeling for mGlu2 was also sporadically detected in association with GC dendritic spines.
- Figure 5.
Shunting profiles of agonist-activated dendritic potassium conductances on dendritic events. A, Representative events evoked by MNI glutamate uncaging in control conditions (black) and in the presence of DCG IV (blue) from three different somatically recorded cells. B, Inhibition of glutamate-evoked dendritic events by DCG IV. Note that the amount of inhibition did not show a tendency of change for those events that originated >125 μm from soma. The right panel shows that the shunting effect did not depend on the control amplitudes of the evoked events. C, Average shunting profile of DCG IV on the somatically measured peak amplitude of glutamate-evoked events. The data are presented as a moving average in which both distance (x-axis) and the effect (y-axis) are averaged from 11 neighboring measurements. D, The DCG IV effect on integrated area (filled symbols) and half-width (open symbols) of the events. E, DCG IV effect on peak amplitude after normalization to the effect on the integrated area of the events within individual experiments. F, Average shunting profile of the GABAB agonist SKF97541 on the peak amplitude of glutamate-evoked events. G, SKF97541 effect on integrated area (filled symbols) and half-width (open symbols) of the events. H, SKF97541 effect on peak amplitude after normalization to the effect on the integrated area of the events within individual experiments.
- Figure 6.
Shunting profiles of dendritic conductances on simulated dendritic events. A, Simulated somatic voltage responses at three different dendritic locations (25, 125, and 195 μm) in a multicompartmental model of a GC in control conditions (black) and during activation of a shunting conductance at 80 μm from the soma (blue traces). B, Comparison of the peak amplitude and slope of the rising phase of events evoked by glutamate uncaging (red) and simulation (black). C, Average shunting profiles of 7 differently localized, spatially restricted shunting conductances on the somatic peak amplitude of simulated events (n = 3 cells, 4 dendrites each). The x-axis indicates the location of the simulated synapses. D, Effect of the differently localized shunts on integrated area (filled symbols) and half-width (open symbols) of the events. E, Shunting effects on peak amplitude after normalization to the effect on the integrated area. F, Average shunting profile of a distally distributed shunting conductance (simulating GABAB distribution) on the somatic peak amplitude of simulated events. G, Effect of the distributed shunt on integrated area (filled symbols) and half-width (open symbols) of the events. H, Shunting effect on peak amplitude after normalization to the effect on the integrated area.
- Figure 7.
Shunting of individual dendritic branches by spatially restricted dendritic conductances. A, Top, Average effect of shunting of the somatic peak amplitude of simulated AMPAR-mediated EPSPs originating from single shunted dendrites (positive x-axis values) or from neighboring dendrites (negative x-axis values). Middle, Same as above for the local peak amplitude of the EPSPs. Bottom, Effect on individual dendritic branches depending on the location of the shunt. Each horizontal line within a block represents the same dendritic branch with the shunt (positive x-axis values) and the same dendrite with the shunt located on one of the neighboring dendrites (negative x-axis values). The inhibitory efficacy on the peak amplitude was normalized to the shunting level of EPSPs at 195 μm on the shunted branch. The red color thus indicates a strong shunting effect, whereas the least inhibited regions are shown in black. B, Same as in A for compound EPSPs mediated by AMPA and NMDA conductances. C, Same as in A for compound EPSPs mediated by AMPA and NMDA conductances in spines with a 500 MΩ neck resistance. D, Selectivity of the inhibition of individual branches calculated as the ratio of shunting in the dendrites (90–195 μm) with the conductance compared with without the conductance. Each shunt location is color coded as in the previous panels. The various simulation conditions are indicated by the connected lines. E, Uniformity of the shunting within individual dendritic branches calculated as the proportion of the shunting of the EPSPs at 90–125 μm relative to the shunting of EPSPs at 195 μm. A higher percentage in this ratio thus indicates more uniform shunting along the dendrites.
