Development of Local Circuit Connections to Hilar Mossy Cells in the Mouse Dentate Gyrus

Developmental changes in intrinsic physiologic and morphologic properties of mossy cells. A, Example responses to current injection (horizontal lines) in example mossy cells recorded from P7, P14, and P21 mice. B, C, Example morphology of biocytin-labeled mossy neurons (green) and surrounding DAPI-stained tissue (blue) at P7, P14, and P21. The labeled neurons in the P7 mouse had proximal dendrites with branches penetrated into the fascia dentate (arrows for P7 mouse in B) and were relatively smooth (image shown in C), while neurons recorded in P14 and P21 mice have obvious thorny excrescence (arrows in C) on proximal dendrites. Scale bar = 250 μM (B) and 25 μM (C). All labeled neurons had large, multipolar somata, and thick thorny proximal dendrites. A neuron (circle in B) in the P14 mouse is in CA3. D, The input resistance (left, measured in MΩ) decreased with age from P6–P7 to P13–P14, and P21–P28. The Cm (middle, measured in pF), is the capacitance of the cell membrane and increased with age from P6–P7 to P13–P14, and P21–P28. The spike rate (right, measured in Hz) is the number of evoked spikes in response to current injection of 100 pA during recording, and decreased with age from P6–P7 to P13–P14. * indicates the statistical significance (p < 0.05), and ** indicates p < 0.01. Also see Table 2.

LSPS mapping and data analysis. A, A horizontal hippocampal slice under a 4× objective with a patched neuron (red circle) and laser stimulation sites overlaid (cyan asterisks). B, Raw signal traces recorded from the patched neuron during laser stimulation. C, Examples of a direct response (top green trace) which has a large amplitude and a short response latency, and synaptic responses (bottom red) which have smaller amplitudes and longer latencies. D–F, Synaptic responses were detected and extracted using automatic software processing (Shi et al., 2010). Input amplitude (see Methods) (D), the number of evoked synaptic events, as defined by the number of EPSCs elicited per laser pulse (E), and the number of spontaneous events (F) were plotted in heat maps.

Spatial precision of LSPS mapping for neurons in the DG, hilus, and CA3c. A, B, Excitation profile of a recorded DG granule cell from a mouse hippocampal slice. The excitation profile shows the spatial distribution of uncaging sites that produce APs. The cell was held in current clamp mode. The cyan asterisks in A are the stimulation sites (75-µm spacing). The evoked APs were restricted to a small region (yellow square). Raw signal traces within the yellow square are shown in B. C, D, Excitation profile of a hilar mossy cell from a mouse hippocampal slice (100-µm spacing). E, F, Excitation profile of CA3 pyramid cells from mouse hippocampal slice (100-µm spacing). Excitation profiles show no spike-evoking sites distal from the perisomatic area of the recorded neuron, demonstrating that LSPS maps (e.g., Figures 5 and 6) represent input from monosynaptic connections.

Functional maps of excitatory inputs to mossy cells across developmental ages. A, Strength of excitatory input (input amplitudes) at P7, P14, and P24 for a representative mossy cell. B, Frequency of EPSC events (number of evoked synaptic events per second) for a representative mossy cell. C, D, Summed (averaged) amplitude of excitatory inputs to mossy cells. We recorded from 12, 9, and 10 cells from P6–P7, P13–P14, and P21–P28 mice, respectively. In the left two panels in C, the y-axis indicates input strength. In the right panel in D, the y-axis shows the ratio of input strengths. * indicates the statistical significance (p < 0.05), and ** indicates p < 0.01. Also see Table 2.