Phase segregation of medial septal GABAergic neurons during hippocampal theta activity

Abstract

Septo-hippocampal GABAergic neurons immunoreactive for parvalbumin are thought to play a crucial role in the generation of hippocampal theta oscillations associated with a specific stage of memory formation. Here we use in vivo juxtacellular recording and filling in the medial septum followed by immunocytochemical identification of the recorded cells containing parvalbumin to determine their firing pattern, phase relationship with hippocampal theta, morphology, and to thereby reveal their involvement in the generation of hippocampal theta activity. We have demonstrated that GABAergic medial septal neurons form two distinct populations exhibiting highly regular bursting activity that is tightly coupled to either the trough (178 degrees ) or the peak (330 degrees ) of hippocampal theta waves. Additionally, different types of bursting as well as nonbursting activity patterns were also observed. The morphological reconstruction of theta-bursting neurons revealed extensive axon arbors of these cells with numerous local collaterals establishing symmetrical synapses; thus, synchrony among the septal pacemaker units may be brought about by their recurrent collateral interactions. Long projecting axons could also be found running dorsally toward the hippocampus and ventrally in the direction of basal forebrain regions. We conclude that GABAergic neurons in the medial septum, which are known to selectively innervate hippocampal interneurons, are in a position to induce rhythmic disinhibition in the hippocampus and other theta-related subcortical areas at two different phases of hippocampal theta.

Figure 1.

Firing pattern of PV-IR neurons during theta (A) and high-frequency ripples (B). A, Theta-related activity. Left column, baseline recordings, right column, tail-pinch recordings. Below the raw traces (filtered EEG and unit, respectively, in all cases), the corresponding color-coded wavelet cross-spectra (1.5-6.5 Hz range) is shown. Warmer colors indicate higher correlation. For description of different types, see Results. Calibration in volts for the EEG-unit traces from left to right (shown as EEG/unit; see black vertical bar at the right of the topmost panel, right column, the unit recording of PV-IR neuron coded 250n1): 250n1, 1.5/2, 1.1/2.2; 258n2, 2.2/1.1, 2.2/1.1; 359n2, 1.7/2, 1/1.7. B, Ripple-associated firing of PV-IR neurons. Top, Filtered EEG; middle, unit; bottom, cumulative histogram showing firing probability. Most PV-IR neurons were inhibited during ripples, but, in one case, activation was observed. The activated PV-IR cell was classified as tail-pinch-responsive bursting neuron. The EEG and unit segments (n = 11 for both 250n1 and 318n1) corresponding to ripples were overlaid in a 500 msec window (480 is shown) centered at the peak of the ripples. Bin size, 20 msec. Calibration in volts for the traces (shown as EEG/unit; see black vertical bar at right): 250n1, 0.5/1.75; 318n1, 0.65/2.2).

Figure 2.

Identification of juxtacellularly labeled neurons. The cells from experiments (rows), numbered 250, 258, and 359 (constitutively bursting, tail-pinch-responsive bursting, and tonic firing neuron, respectively) are parvalbumin immunoreactive, whereas number 295 is PV negative (arrows), as shown in the fluorescent micrographs in the first column (PV) using different fluorochromes. The middle column indicates the biocytin (or Neurobiotin) labeling by fluorescent markers, and, in the right column, light micrographs of the same neurons are shown after nickel-intensified DAB reaction. Scale bars, 50 μm.

Figure 3.

PV-IR neurons fired different types of bursts. A, Normalized (by total event count) distributions of the analyzed burst parameters are shown. From left to right, Intraburst frequency (bin, 10 Hz), burst length (bin, 15 msec), and spike count per burst (bin, 1.5). B, The distribution of means of the above variables calculated for each cell is plotted. C, The different burst types can be seen. The top trace represents the instantaneous frequency function, and the bottom trace is the corresponding raw unit recording. Four burst types are demonstrated: (1) low-frequency, long burst; (2) low-frequency, short burst; (3) high-frequency, short burst; and (4) high-frequency, long burst. Transient increase in intraburst frequency can be observed within bursts, especially in case of high-frequency, long bursts. It accounts for the higher variability of intraburst frequency in the case of high-frequency bursts.

Figure 4.

Phase preference and phase-firing pattern relationship of PV-IR neurons. A, The phase relationship of the two PV-IR groups (□ and ▪) is shown on a polar plot. Strong phase preference is indicated by the distance from the center corresponding to the mean vector length (the plotted range: center, 0.3; periphery, 1.2). B, Weak correlation between burst length and phase preference of PV-IR neurons. Long-burst neurons tended to fire around the peak, whereas short-burst neurons fired around the trough of hippocampal theta. C, PV-IR neurons showed strong phase preference forming two well-separable groups. The two groups were significantly different from each other. In the top, two pyramidal layer theta cycles are shown. Below the theta cycles on the cumulative histograms, the phase distribution of PV-IR neurons (for details, see Materials and Methods) is demonstrated with mean phase values for all groups. PV-IR neurons have a bimodal phase distribution, exhibiting increased firing probability on the trough and slightly before the peak of pyramidal layer theta, and they avoid to fire on the falling edge of the cycle. For comparison, in previous papers (Klausberger et al., 2003), trough was 0°; in our calculation, trough was 180°.

Figure 5.

Reconstructions of four PV-IR, juxtacellularly filled neurons. The tail-pinch-responsive bursting (A, 258) and the three constitutively bursting cells (B, 250; C, 209; D, 173) show common morphological features, despite the differences in the extent of their labeling: dendritic arbor in an oval-shaped area (gray), a thick main axon emerging ventrally and taking a U-turn (wide gray arrows), leaving the septum dorsally. This main axon emits parallel-running, thick myelinated secondary fibers that leave the medial septum dorsally with the main axon or ventrally by entering or crossing the diagonal band of Broca. The thick myelinated axons emit thin unmyelinated collaterals bearing the terminals (3× magnified area in C, D). Scale bar: 200 μm; for the magnified insets, 66 μm.

Figure 6.

Electron microscopic features of the biocytin- or Neurobiotin-filled axons. The terminals of constitutively or theta-responsive bursting, PV-IR neurons formed symmetrical synapses (double arrows) frequently on proximal (thick) or middle-sized dendrites (A, B). Occasionally, somata (C) or thin, distal dendrites (D) were also targeted. Scale bars, 500 nm.

Figure 7.

A hypothetical scheme showing the interaction between medial septal PV-IR neurons and hippocampal perisomatic and dendritic inhibitory interneurons during the peak (left) and the trough (right) of pyramidal layer theta. The corresponding current source density maps (with permission from G. Buzsaki, 2002) and EEG recordings are also shown. The bottom EEG trace derives from the pyramidal layer, and the top trace is from the stratum lacunosum-moleculare. Left, The peak-preferring PV-IR cells of the medial septum are firing and inhibiting both the trough-preferring PV-IR neurons of the MS and dendritic inhibitory interneurons in the hippocampus, resulting in disinhibition of the distal dendritic region and the appearance of a marked current sink in the lacunosum-moleculare on the current source density map. Basket cells are active in this phase and induce a source in stratum pyramidale. Right, The trough-preferring neurons are active and suppress both the peak-preferring MS neurons and hippocampal perisomatic inhibitory cells. As a result, the basket cell-mediated active source disappears in the stratum pyramidale, and a new source (likely active) appears in stratum lacunosum-moleculare because of the activity of O-LM cells.