Nonrandom local circuits in the dentate gyrus

Abstract

The dentate hilus has been extensively studied in relation to its potential role in memory and in temporal lobe epilepsy. Little is known, however, about the synapses formed between the two major cell types in this region, glutamatergic mossy cells and hilar interneurons, or the organization of local circuits involving these cells. Using triple and quadruple simultaneous intracellular recordings in rat hippocampal slices, we find that mossy cells evoke EPSPs with high failure rates onto hilar neurons. Mossy cells show profound synapse specificity; 87.5% of their intralamellar connections are onto hilar interneurons. Hilar interneurons also show synapse specificity and preferentially inhibit mossy cells; 81% of inhibitory hilar synapses are onto mossy cells. Hilar IPSPs have low failure rates, are blocked by the GABA(A) receptor antagonist gabazine, and exhibit short-term depression when tested at 17 Hz. Surprisingly, more than half (57%) of the mossy cell synapses we found onto interneurons were part of reciprocal excitatory/inhibitory local circuit motifs. Neither the high degree of target cell specificity, nor the significant enrichment of structured polysynaptic local circuit motifs, could be explained by nonrandom sampling or somatic proximity. Intralamellar hilar synapses appear to function primarily by integrating synchronous inputs and presynaptic burst discharges, allowing hilar cells to respond over a large dynamic range of input strengths. The reciprocal mossy cell/interneuron local circuit motifs we find enriched in the hilus may generate sparse neural representations involved in hippocampal memory operations.

Figure 1.

Synaptic connections formed by dentate hilar neurons. A, Schematic representation of the dentate gyrus and CA3 subfield of the hippocampus. The dentate hilus is located between the granule cell layer (GCL) and CA3 and contains primarily two classes of neurons: MCs and HIs. B₁, Paired recording between two hilar neurons. Both spontaneous action potentials in the bottom recording are correlated with EPSPs in the top trace (EPSP onset latencies of 1.8 and 1.6 ms; putative evoked EPSPs enclosed within black circles). B₂, Paired recording between two hilar neurons with one apparent evoked EPSP in the top recording (2.8 ms onset latency) in a different experiment. Both postsynaptic recordings show frequent spontaneous EPSPs that are characteristic of hilar neurons; 400 pA current steps. C, Average postsynaptic response from the paired recording shown in B₂ [average of 39 consecutive spike-aligned recordings; paired recording between a mossy cell (black square) and an unclassified hilar neuron (gray diamond)]. D, Average postsynaptic response from a hilar paired recording between an inhibitory interneuron (black circle) and an unclassified hilar cell (average of 31 consecutive spike-aligned episodes); 250 pA current steps.

Figure 2.

Monosynaptic connections between mossy cells and hilar interneurons. A₁, Monosynaptic EPSPs evoked by single APs in the MC. Eight consecutive HI postsynaptic responses (shown in red) including four failures. A_2, Eight consecutive monosynaptic IPSPs (blue traces) recorded in the mossy cell and evoked by single APs in the interneuron. Same mossy cell and interneuron pair in A₁ and A₂. B, Distribution of unitary EPSP (red) and IPSP (blue) amplitudes in the reciprocal pair shown in A. The failure rate was 62.6% for the EPSP and 9.5% for the IPSP. Inset shows average postsynaptic response before and after bath application of 10 μm gabazine in the same paired recording. C, Distribution of EPSP and IPSP onset latencies from the reciprocal pair shown in A. D, Summary of unitary EPSP (red) and IPSP (blue) amplitudes from 9 excitatory and 22 inhibitory connections. E, Summary of onset latencies from all monosynaptic hilar connections identified. Hilar IPSP onset latencies are significantly shorter than EPSPs (***p < 0.001). Open red circles in D and E represent values from the one mossy cell-to-mossy cell connection identified; filled red circles represent values from postsynaptic responses in hilar interneurons.

Figure 3.

Mossy cells and hilar interneurons can be distinguished by their intrinsic properties. A, Responses to depolarizing steps in a MC that evoked an EPSP in a paired recording (left trace, red) and in a HI (right trace, blue) that evoked an IPSP. Mossy cells typically fire a burst at the beginning of the response and then generate single APs with small AHPs at irregular intervals, whereas hilar interneurons tend to fire intermittent clusters of APs with prominent AHPs. B, Mossy cells have a significantly smaller mean AP time (*p < 0.05), smaller spike clustering fraction (***p < 0.001), and smaller spike AHP slope (***p < 0.001) than hilar interneurons. C, Mossy cells (red spheres; n = 8) and interneurons (blue spheres; n = 14) can be distinguished by using all three metrics. All three axes reflect the same metrics and units as the plots in B. Separation plane shown in green.

Figure 4.

Mossy cells selectively innervate hilar interneurons. A, Analysis of 270 hilar neurons based on distance from separation plane shown in Fig. 3C reveals a bimodal distribution. Hilar neurons with positive distance metrics are classified as “presumptive mossy cells” and those with negative metrics as “presumptive interneurons.” Filled red and blue circles represent distance metrics for known mossy cells and hilar interneurons, respectively. Distance metrics for postsynaptic targets of known mossy cells and interneurons indicated by open circles. Gray arrows indicate distance metrics of the known mossy cell and hilar interneuron presented in Fig. 2A. B, Summary plot of the frequency of the four possible mossy cell/interneuron connection types. We found 1 of 206 possible mossy cell-to-mossy cell connections, 7 of 114 possible mossy cell-to-interneuron connections, 17 of 114 possible interneuron-to-mossy cell connections, and 4 of 68 possible interneuron-to-interneuron connections.

Figure 5.

Structured polysynaptic local circuit motifs in the dentate hilus. A, Simultaneous quadruple hilar cell recording with five monosynaptic connections, including two reciprocal mossy cell/interneuron connections (gray rectangles). Recordings also illustrate excitatory synaptic convergence (MC1 and MC2 both excite HI1; EPSP onset latencies = 2.1 and 1.5 ms, respectively) and inhibitory divergence (HI1 inhibits MC1, MC2, and HI2; latencies, 1.7, 1.8, and 2.0 ms). All known phenotypes except HI2 (distance metric, −4.23). Average postsynaptic response shown in bold trace superimposed on example unitary responses (gray traces). Vertical dashed lines indicate presynaptic AP timing (middle of rising phase). B, Simultaneous triple hilar cell recording demonstrating inhibitory convergence (HI1 and HI2 both inhibit MC; IPSP onset latencies, 2.4 and 1.8 ms) and a reciprocal mossy cell/interneuron connection (EPSP onset latency, 1.8 ms). All known phenotypes. C, Diagram representation of hilar circuitry expected on the basis of monosynaptic connection preferences alone (MC/HI reciprocal motif frequency expected = 0.9%; 1 in 114 pairs; top) and observed enrichment of reciprocal motifs in the dentate hilus (observed motif frequency = 3.5%; p < 0.05; binomial cumulative distribution; bottom).

Figure 6.

Morphology of hilar neurons. A, Morphology of intracellularly filled mossy cell. Montage of maximal projections from Z-stacks acquired through two-photon imaging of a live brain slice. Gray arrows indicate axonal processes. Enlargement of distal dendrite segment shown in right inset. Enlarged area indicated by white dashed rectangle in main image. Gray circle indicates proximal dendritic segments with thorny excrescences. Bottom inset shows example response of neuron to 3 s depolarizing step (350 pA). Distance metric, 1.04. B, Morphology of filled hilar interneuron. Gray arrows indicate axonal processes. Bottom inset shows example response of interneuron to a 200 pA depolarizing step. Distance metric, −1.31. Calibration: 20 mV, 500 ms (both insets).

Figure 7.

Functional effects of inhibitory hilar synapses. A, Spontaneous discharges in a depolarized hilar interneuron transiently inhibit a postsynaptic mossy cell. Interneuron discharges are correlated with pauses in the mossy cell spontaneous firing. Mean firing frequencies for the initial two interneuron discharges are shown above the recording. Mossy cell action potentials truncated. B₁, Large-amplitude, summating IPSPs evoked by high-frequency hilar interneuron discharges. Interneuron activated by trains of α-function current waveforms. B₂, Unitary IPSPs depress with paired-pulse stimuli (60 ms interval). B₃, IPSPs evoked by high-frequency interneuron bursts summate. Same paired recording in B₁–B₃. Vertical dashed lines indicate timing of action potentials in the presynaptic interneuron driven by an α function current stimulus. The mean presynaptic firing frequency in response to this stimulus was 75 Hz.

Figure 8.

Functional effects of excitatory hilar synapses. A₁, Presumptive mossy cell-to-mossy cell excitatory connection with a mean onset latency of 2.5 ms. Average postsynaptic response shown in bold trace superimposed on example unitary responses (gray traces). A₂, The postsynaptic hilar neuron (distance metric, 1.57; MC2) in this paired recording responded to a depolarizing step with an initial burst followed by irregular spiking and small AHPs, characteristic of mossy cells. B, Averaged postsynaptic response to paired-pulse activation of a mossy cell. Hilar EPSPs showed modest paired-pulse facilitation. C, Summary of short-term plasticity in hilar synapses. Inhibitory responses segregated by phenotype of postsynaptic target; excitatory connections pooled. *p < 0.05. D₁, Spike transmission through mossy cell synapses. The presynaptic mossy cell (bottom traces) fired repetitively at ∼50 Hz when activated by an α-function current waveform, evoking a series of summating EPSPs and a single AP (truncated) in the postsynaptic hilar interneuron (top trace). D₂, Summary plots of the probability of firing in the presynaptic and postsynaptic cells, versus time, when the presynaptic neuron was driven by trains of α-function current waveforms as in D₁. Repetitive mossy cell firing at 50 Hz reproducibly evoked firing in postsynaptic neurons. E₁, Average spike-aligned responses to the first 4 APs in the α-function-triggered presynaptic bursts. Example presynaptic mossy cell burst shown in inset. The EPSP evoked by the second presynaptic AP was larger than the response to the first AP; subsequent EPSPs were smaller than the initial EPSP response. E₂, Summary plots of EPSP amplitude and postsynaptic spike probability versus presynaptic spike number. Same stimulus protocol as D₁ and E₁.