Abstract

Travelling theta oscillations and sharp wave-associated ripples (SWRs) provide temporal structures to neural activity in the CA1 hippocampus. The contribution of rhythm-generating GABAergic interneurons to network timing across the septotemporal CA1 axis remains unknown. We recorded the spike-timing of identified parvalbumin (PV)-expressing basket, axo-axonic, oriens-lacunosum moleculare (O-LM) interneurons, and pyramidal cells in the intermediate CA1 (iCA1) of anesthetized rats in relation to simultaneously detected network oscillations in iCA1 and dorsal CA1 (dCA1). Distinct interneuron types were coupled differentially to SWR, and the majority of iCA1 SWR events occurred simultaneously with dCA1 SWR events. In contrast, iCA1 theta oscillations were shifted in time relative to dCA1 theta oscillations. During theta cycles, the highest firing of iCA1 axo-axonic cells was followed by PV-expressing basket cells and subsequently by O-LM together with pyramidal cells, similar to the firing sequence of dCA1 cell types reported previously. However, we observed that this temporal organization of cell types is shifted in time between dCA1 and iCA1, together with the respective shift in theta oscillations. We show that GABAergic activity can be synchronized during SWR but is shifted in time from dCA1 to iCA1 during theta oscillations, highlighting the flexible inhibitory control of excitatory activity across a brain structure.

Introduction

Along the septotemporal axis of the hippocampus, critical differences in synaptic connectivity and molecular expression have been observed (Cenquizca and Swanson 2007; Dong et al. 2009). In addition, an increasing amount of evidence suggests a division of labor among different hippocampal subdivisions. Accordingly, the dorsal hippocampus encodes primarily spatial information, and the ventral hippocampus processes also nonspatial, for example, emotional information, while the intermediate hippocampus has been suggested to integrate spatial and nonspatial information and to translate learning into behavior (Moser and Moser 1998; Bannerman et al. 2003; Maurer et al. 2005; Bast 2007, 2011; Fanselow and Dong 2010). The integration and coordinated read-out of information require a temporal framework, which is provided by network oscillations occurring at behavior-dependent frequencies (Buzsaki and Draguhn 2004). Sharp wave-associated ripples (SWRs) occur during resting and sleep and allow the consolidation of memories via the compact replay of cell assemblies (Foster and Wilson 2006; Diba and Buzsaki 2007; Girardeau et al. 2009; Ego-Stengel and Wilson 2010). Theta oscillations reflect the online state of memory formation and retrieval and allow the coding of spatial information via place cells (O'Keefe 1976; O'Keefe and Nadel 1978; O'Keefe and Recce 1993; Skaggs et al. 1996). It has been reported that the firing of pyramidal cells and theta in the local field potential (LFP) are gradually shifting in time across the septotemporal axis of the CA1 hippocampus (Hartwich et al. 2009; Lubenov and Siapas 2009; Patel et al. 2012). This phenomenon has been referred to as the travelling theta wave (Lubenov and Siapas 2009).

The generation of network oscillations is mediated by a large diversity of GABAergic interneurons. Distinct types of GABAergic interneuron target different subcellular domains of pyramidal cells and contribute differentially to SWR and theta oscillations in the dCA1 hippocampus (Klausberger and Somogyi 2008). However, it remains unknown how GABAergic interneurons support the coordination of network oscillations across different parts of the septotemporal axis of the hippocampus. To address this question, we have recorded from identified parvalbumin-expressing (PV+) basket, axo-axonic, and oriens-lacunosum moleculare (O-LM) interneurons in the iCA1 area of anesthetized rats and determined their distinct axonal targets and molecular expression. The analysis of the in vivo firing patterns of these cells during travelling theta waves along dCA1 and iCA1 as well as during SWR oscillations forms the basis for our understanding of how these cell types contribute to the organization of neural activity and, therefore, to the information processing across a dynamic brain area such as the CA1 hippocampus.

Materials and Methods

All animal procedures were performed under licenses approved by the Austrian ministry of Science and in accordance with the relevant regulations of the Medical University of Vienna.

Juxtacellular Recording and Labeling

Experiments were performed on male Sprague Dawley rats (250–400 g) as described previously (Klausberger et al. 2003,  2005). Anesthesia was induced with isoflurane and maintained with urethane (1.25 g/kg of body weight; i.p.); additional doses of a mix of ketamine and xylazine (17 and 7 mg/mL, respectively; 0.02–0.1 mL; i.p.) were given as needed to control the level of anesthesia. During anesthesia, the body temperature of the rat was maintained at 37 °C with a heating pad, and the heart rate was continuously monitored (aimed to be between 3 and 6 Hz). Two glass pipettes, filled with 1.5% neurobiotin in 0.5 M NaCl solution and a silver wire electrode, were inserted at 2 separate surgery sites to record extracellular neural activity simultaneously from the right dCA1 and iCA1 (from Bregma, dCA1: 3.2 mm posterior, 2 mm lateral; iCA1: 5–6.6 mm posterior and 3–6 mm lateral). The dCA1 electrode was lowered at a 10° caudal-to-rostral angle, until it reached the stratum pyramidale, which is characterized by the occurrence of the largest upward SWR events. The electrode targeting the iCA1 was lowered until the hippocampus was reached (i.e., theta and SWR events were detected) and then slowly further advanced until the isolated action potential of a single neuron could be recorded, as confirmed by consistent spike shape and firing pattern. We aimed at recording the firing patterns of the neuron during SWR events and theta oscillations. After recording neuronal activity for 10–20 min, the neuron was labeled with the juxtacellular technique (Pinault 1996) by applying 200ms long positive current pulses for about 1–5 min, which entrained and increased action potential firing. After 1–2 h, the rat was perfused for a short time with a 0.9% saline solution followed by a 4% paraformaldehyde, 0.05% glutaraldehyde, and 15% saturated picrinic acid solution.

Brain Processing, Immunohistochemical Analysis, and Cell Identification

The brains were cut by a vibratome into 70-µm-thick slices. Several sections where we expected neuronal processes of the recorded cell were chosen to be labeled with Streptavidin conjugated to Alexa Fluor 488 (Invitrogen), aminomethylcoumarin acetate (AMCA) (Jackson Immuno Research Laboratories) or Dylight 488 (Jackson Immuno Research Laboratories). Using an epifluorescence microscope, we then detected the neuronal processes and confirmed whether our labeling attempt had been successful. Based on our observations of dendrite and axon distribution, we performed immunohistochemical labeling for different known neuronal markers of specific neuron types. The markers used in our experiments were investigated with the following primary antibodies: calbindin (raised in rabbit) used at 1:5000 dilution from Swant (Airaksinen et al. 1997), cholecytstokinin (CCK; raised in guinea pig) used at 1:500 and pro-CCK (raised in rabbit) used at 1:500 dilution from Dr M. Watanabe, Hokkaido University, Japan (labeling patterns as published with other antibodies), PV (raised in mouse) used at 1:5000 dilution from Swant (Constantinople et al. 2009), PV (raised in guinea pig) used at 1:2000 dilution from Synaptic Systems (labeling patterns as published with previous antibodies), somatostatin (raised in rat) used at 1:500 dilution from Chemicon (Kubota et al. 2011), neuropeptide Y (NPY, raised in rabbit) used at 1:2500 from Immunostar (labeling patterns as published with other antibodies), cannabinoid receptor type 1 (CBR1, raised in guinea pig) and CBR1 receptor (raised in rabbit) used at 1:1000 dilution from Dr M. Watanabe (Fukudome et al. 2004), metabotropic glutamate receptor 1α (mGluR1α, raised in guinea pig or rabbit) used at 1:1000 dilution from Dr M. Watanabe (Alvarez et al. 2000; Nakamura et al. 2004), ErbB4 (raised in mouse) used at 1:500 dilution from Thermo Scientific (Chen et al. 1996), muscarinic acetylcholine receptor M2 (raised in rat) at 1:250 dilution from Chemicon (Levey et al. 1995), mGluR7a (raised in rabbit) used at 1:500 dilution from Dr. Shigemoto (Shigemoto et al. 1996), and ankyrin G (raised in mouse) used at 1:2000 dilution from Antibodies, Inc. (labeling patterns as published with other antibodies). Secondary antibodies with the fluorescent labels Alexa 488, AMCA, Cy3, or Cy5 (Jackson Immuno Research Laboratories) were used to visualize the binding location of the primary antibody. In addition, HRP/DAB (horseradish peroxidase/3,3′-diaminobenzidine) reactions allowed a very detailed visualization and investigation of the neuronal structures under a light microscope. Three-dimensinoal reconstructions of the soma, dendrites, and axons were performed using the Neurolucida system (MBF Bioscience) mounted on an Olympus BX 51 microscope, and using an UPlanFL N ×100 oil immersion objective (NA 1.3). The proportion of axo-axonic boutons in direct contact with ankyrin G expressing axon initial segments of pyramidal cells was further analyzed with a confocal microscope (Leica TCS SP5).

Spike, Theta, and Sharp Wave-Associated Ripple Detection

For the offline analysis, spikes were detected with in-built functions of Spike2 (version 7.01; Cambridge Electronic Design). The LFP traces were bandpass filtered between 36 Hz for theta periods and 90–200 Hz for SWR events, and the distinct periods were identified with scripts developed in Spike2 by John Tukker (Lasztoczi et al. 2011). Each detected theta period, and SWR event was inspected by the experimenter and discarded, accepted, or corrected manually. Theta periods were accepted only with several subsequent theta cycles lasting at least 4 s. SWR events were discarded based on the interference of noise and spike artifacts or when they could not be clearly identified as such due to their small power. In order to remove the spike artifact from SWR events recorded in the iCA1 hippocampus, an artifact removal script from Cambrigde Electronic Design was applied to the LFP channel, and then each iCA1 SWR event was checked carefully whether the removal of the spike artifacts interfered with the detection of SWR. Owing to the different recording location ranging from the stratum oriens/alveus border to the stratum pyramidale, the size of the SWR often differed and the small but distinct SWR that could not be detected automatically were defined manually. To compare SWR events between the iCA1 and dCA1, the time difference of each iCA1 SWR event to closest dCA1 SWR event was calculated and binned. The peak of ripple (90–200 Hz) power was taken as the time point for each event.

Analysis of Firing Patterns During Theta Oscillations

The detected spikes and theta periods in Spike2 were further analyzed using MATLAB (7.9.0 R2009b; MathWorks) with scripts developed by the authors (Lasztoczi et al. 2011). Functions from the Circular Statistics toolbox by Philip Berens (Berens 2009) were also used. To determine the phase coupling of individual neurons to theta oscillations, the troughs of the detected theta periods were determined and assigned 360° while the phases in between were obtained by linear interpolation. Each spike was assigned a phase value, binned into 20 phase bins of 1 theta cycle, and significant couplings were statistically ascertained with a Rayleigh' test. Spike counts were converted to firing rates by dividing each bin value by the total duration of theta periods and multiplying it by the bin number. The mean phase angle of single cells and of different cell types was determined with circular statistics. The LFP theta shift between dCA1 and iCA1 was the calculated mean phase offset of iCA1 theta troughs to dCA1 theta troughs. For the phase coupling changes of cell TF26a during theta oscillations, the phase angle of each spike in the theta cycle was calculated and then corrected by adding (subtracting) 360° when the difference to the previous phase angle was greater (smaller) than 180°. Because cell TF26a fired 1 spike per theta cycle, we could visualize if spikes in consecutive theta cycles were advancing or receding. To compare the time interval between spikes and theta cycles, the time between spikes during theta oscillations and the time between theta peaks were calculated and compared.

Analysis of Firing Patterns During Sharp Wave-Associated Ripples

The detected spikes and SWR events in Spike2 were further analyzed in MATLAB (7.9.0 R2009b; MathWorks) with scripts developed by the authors (Lasztoczi et al. 2011). Spike numbers were detected in 8 bins during the defined SWR and in 12 bins before and after the SWR event. Firing rates during SWR were calculated by dividing each bin number by the total added time of all SWR events and multiplying it by the bin number.

Results

We recorded and labeled interneurons and pyramidal cells in the iCA1 hippocampus of anesthetized rats. There are no established clear borders of the iCA1 hippocampus and many of the changes occurring along the septotemporal axis are often described as gradual (Bast 2007). We aimed at recording and labeling single cells located at 5–6.6 mm posterior and 3–6 mm lateral from Bregma, corresponding to a more vertical orientation of the CA1 hippocampus (Fig. 1A), and regarded by most studies as intermediate hippocampus (Fanselow and Dong 2010). By analyzing the expression of known molecular markers and verifying the dendritic and axonal distribution of the labeled cells, we reported 3 types of identified GABAergic interneuron in the iCA1: PV+ (PV-immunopositive) basket, axo-axonic, and O-LM cells (Table 1). These cell types have previously been described in the rat dCA1 under similar conditions (Klausberger et al. 2003). On the basis of many important similarities, we conclude that the same cell types also exist in the iCA1, although we here describe several differences.

Table 1

Molecular expression of recorded and labeled interneurons

Cell type Cell name Immunohistochemical analysis
 
Cell positions
 
PV NPY SOM mGluR1α ErbB4 CB M2 CCK CB1R mGluR7a inputs AnkG targets Anteroposterior Mediolateral Dorsoventral 
PV basket TF15a + d    + d       −5.6 3.2 
TF29a + d −s −s  + d   −s    −5.6 4.2 2.9 
TF40b + d −s −s  + d    −s   −6.3 5.6 4.6 
O43b + d −s −s  + d    −s   −5.9 4.8 
O54a + d −s −s  + d       −6.3 5.2 3.8 
SL03a + s  −s         −5.8 5.8 4.5 
Axo-axonic TF31a + d    + d      −5.8 5.8 4.3 
TF35c + d    + d      −6.2 4.8 3.3 
O-LM TF12b −s,d −s + s + d −d −d −d   + d  −5.3 5.2 3.6 
TF16b + s  + s + d  −s    + s  −5.2 3.2 
TF26a + d −s + s + d −d  −d   + d  −5.2 5.1 3.4 
O22b + d nc nc + d        −6.1 4.8 3.1 
O33b    + d        −5.9 4.8 
SL15b + s −s + s + d −d       −6 
Cell type Cell name Immunohistochemical analysis
 
Cell positions
 
PV NPY SOM mGluR1α ErbB4 CB M2 CCK CB1R mGluR7a inputs AnkG targets Anteroposterior Mediolateral Dorsoventral 
PV basket TF15a + d    + d       −5.6 3.2 
TF29a + d −s −s  + d   −s    −5.6 4.2 2.9 
TF40b + d −s −s  + d    −s   −6.3 5.6 4.6 
O43b + d −s −s  + d    −s   −5.9 4.8 
O54a + d −s −s  + d       −6.3 5.2 3.8 
SL03a + s  −s         −5.8 5.8 4.5 
Axo-axonic TF31a + d    + d      −5.8 5.8 4.3 
TF35c + d    + d      −6.2 4.8 3.3 
O-LM TF12b −s,d −s + s + d −d −d −d   + d  −5.3 5.2 3.6 
TF16b + s  + s + d  −s    + s  −5.2 3.2 
TF26a + d −s + s + d −d  −d   + d  −5.2 5.1 3.4 
O22b + d nc nc + d        −6.1 4.8 3.1 
O33b    + d        −5.9 4.8 
SL15b + s −s + s + d −d       −6 

Note: Neurochemical expression profile and position of recorded and labeled interneurons in relation to Bregma. “+” indicates that the cell was tested immunopositive; “−” indicates that no specific immunoreactivity was detected for this cell although other immunopositive structures were observed within the same frame; nc, test was not conclusive; d, tested on a dendrite; s, tested on the soma; a, tested on axon.

Figure 1.

Theta oscillations, but not SWR events, are shifted in time between the dCA1 and iCA1 hippocampus. (A) Somata position of recorded and labeled cells based on Paxinos and Watson (2007) The Rat Brain. (B) Theta oscillations in iCA1 theta are shifted with respect to simultaneously detected theta oscillations in dCA1 (49° on average, range between 11 and 96°). This shift is correlated with the recording position in iCA1 along the septotemporal axis (r = 0.53, P = 0.04, Pearson correlation). (C) The shift of LFP theta in a recording for a single cell was in general equal to the difference of the mean phase coupling values to dCA1 and iCA1 theta (r = 0.96, P < 0.001, Pearson correlation). This indicates high coherence of theta oscillations between dCA1 and iCA1. (D) Cross-correlogram indicating that most iCA1 SWR events occur, on average, at the same time as dCA1 SWR events. (B) and (C) include data also from other recorded iCA1 cells, otherwise not presented in this study. D includes all detected SWR events from the recordings of the iCA1 cells shown in (A).

Figure 1.

Theta oscillations, but not SWR events, are shifted in time between the dCA1 and iCA1 hippocampus. (A) Somata position of recorded and labeled cells based on Paxinos and Watson (2007) The Rat Brain. (B) Theta oscillations in iCA1 theta are shifted with respect to simultaneously detected theta oscillations in dCA1 (49° on average, range between 11 and 96°). This shift is correlated with the recording position in iCA1 along the septotemporal axis (r = 0.53, P = 0.04, Pearson correlation). (C) The shift of LFP theta in a recording for a single cell was in general equal to the difference of the mean phase coupling values to dCA1 and iCA1 theta (r = 0.96, P < 0.001, Pearson correlation). This indicates high coherence of theta oscillations between dCA1 and iCA1. (D) Cross-correlogram indicating that most iCA1 SWR events occur, on average, at the same time as dCA1 SWR events. (B) and (C) include data also from other recorded iCA1 cells, otherwise not presented in this study. D includes all detected SWR events from the recordings of the iCA1 cells shown in (A).

Temporal Relations of Theta Oscillations and Sharp Wave-Associated Ripples Between the dCA1 and iCA1

To determine the contribution of the different iCA1 cell types to local network computations and compare their spike-timing to those of previously reported dCA1 cells (Klausberger et al. 2003), we recorded simultaneously the dCA1 LFP from the stratum pyramidale together with iCA1 neuronal firing and iCA1 LFP from stratum pyramidale or oriens. This approach allowed us also to analyze the coupling of the cell activity to both dCA1 and iCA1 LFP events, including SWR (90–200 Hz) events and theta oscillations (3–6 Hz). We found that, as previously shown (Lubenov and Siapas 2009; Patel et al. 2012), theta oscillations in the dCA1 and iCA1 were highly coherent but shifted in phase with an average shift of +49° ± 19° (dCA1 troughs occurred before iCA1 troughs). Thus, we tested the hypothesis that differences in phase shift increased with the distance from the dCA1 by recording at different locations of the iCA1 hippocampus along the septotemporal axis (Fig. 1A), and using distance between Bregma and the recording position as a measure. We found that with increasing distance the phase shift between the iCA1 and dCA1 theta oscillations also increased (P = 0.015, r = 0.59, Pearson's correlation; Fig. 1B). Moreover, we determined the phase of theta oscillations when different cells fired with the highest probability both in dCA1 and iCA1, and found that the same cell was coupled differently to the phases of dCA1 and iCA1 theta. We tested the hypothesis that the shift in LFP theta oscillations between dCA1 and iCA1 in each recording could account for the differences in phase coupling of the cell to both areas. Indeed, the differences in LFP theta shifts were equal, on average, to the difference in phase coupling of the cells to dCA1 and iCA1 theta (P < 0.001, r = 0.96, Pearson's correlation; Fig. 1C).

We analyzed the co-occurrence of SWR events in the dCA1 and iCA1 and found that 43% of SWR events detected in iCA1 (n = 269) occurred within a 40ms period surrounding a SWR event detected in dCA1 (n = 428) (Fig. 1D). Considering a longer time period of 400 ms around dCA1 SWR events, 77% of iCA1 events occurred close to dCA1 SWR events. This analysis might underestimate the co-occurrence of SWR events in the both areas because low-amplitude SWR events might fall below the detection threshold. This might also explain the different numbers of SWR events detected in the 2 areas. However, larger time shifts and differences in the size of SWR events were also observed in some cases (Fig. 1D and see SWR recordings in Figs 2C, 3C, and 4C). Overall, SWR events occur, on average, at the same time in the dCA1 and iCA1 (Fig. 1D).

Figure 2.

Visualization, molecular expression, and in vivo firing patterns of the identified iCA1 PV-expressing basket cell TF29a. (A) Reconstruction of axons (blue, from one 70-μm-thick section, boutons represented by blue spheres), soma, and dendrites (red, complete from 5 consecutive sections). St. or., stratum oriens; st. py., stratum pyramidale. (B) Schematic representation of the cell in a coronal brain slice [5.6 mm posterior from Bregma, after Paxinos and Watson (2007) The Rat Brain]. (C) Increased firing of the labeled cell during SWR events; note the simultaneous ripple occurrence in dCA1 and iCA1. (D) The cell fired strongest at the trough and descending phase of theta oscillations measured extracellularly in the pyramidal cell layer of dCA1 and iCA1, respectively. Note the constant phase shift between theta oscillations in dCA1 and iCA1. (E) Immunofluorescence micrographs of a neurobiotion-filled dendrite immunopositive for PV and ErbB4.

Figure 2.

Visualization, molecular expression, and in vivo firing patterns of the identified iCA1 PV-expressing basket cell TF29a. (A) Reconstruction of axons (blue, from one 70-μm-thick section, boutons represented by blue spheres), soma, and dendrites (red, complete from 5 consecutive sections). St. or., stratum oriens; st. py., stratum pyramidale. (B) Schematic representation of the cell in a coronal brain slice [5.6 mm posterior from Bregma, after Paxinos and Watson (2007) The Rat Brain]. (C) Increased firing of the labeled cell during SWR events; note the simultaneous ripple occurrence in dCA1 and iCA1. (D) The cell fired strongest at the trough and descending phase of theta oscillations measured extracellularly in the pyramidal cell layer of dCA1 and iCA1, respectively. Note the constant phase shift between theta oscillations in dCA1 and iCA1. (E) Immunofluorescence micrographs of a neurobiotion-filled dendrite immunopositive for PV and ErbB4.

Figure 3.

Visualization, molecular expression, and in vivo firing patterns of the identified iCA1 axo-axonic cell TF35c. (A) Reconstruction of axons (blue, from 1 section, boutons represented by blue spheres) and soma and dendrites (red, complete from 8 consecutive sections). (B) Schematic representation of the cell in a coronal brain slice [6.2 mm posterior from Bregma, after Paxinos and Watson (2007) The Rat Brain]. (C) Firing pattern during SWR events showing different ripple occurrences in dCA1 and iCA1 and inhibition of firing with the iCA1 SWR onset. (D) Firing pattern during theta oscillations. A shift between the dCA1 and iCA1 LFP theta and the strongly phase coupled high firing rate can be seen (cell fires around the trough of dCA1 theta and during the descending phase of iCA1 theta). (E) Immunofluorescence micrograph (overlay of 9 consecutive optical slices with 1 μm spacing) of neurobiotin-filled axonal boutons which are in direct contact with axon initial segments of pyramidal cells labeled with antibodies against ankyrin G. (F) Immunofluorescence micrographs showing a neurobiotin-filled dendrite immunopositive for PV (left) and the soma positive for ErB4 (right).

Figure 3.

Visualization, molecular expression, and in vivo firing patterns of the identified iCA1 axo-axonic cell TF35c. (A) Reconstruction of axons (blue, from 1 section, boutons represented by blue spheres) and soma and dendrites (red, complete from 8 consecutive sections). (B) Schematic representation of the cell in a coronal brain slice [6.2 mm posterior from Bregma, after Paxinos and Watson (2007) The Rat Brain]. (C) Firing pattern during SWR events showing different ripple occurrences in dCA1 and iCA1 and inhibition of firing with the iCA1 SWR onset. (D) Firing pattern during theta oscillations. A shift between the dCA1 and iCA1 LFP theta and the strongly phase coupled high firing rate can be seen (cell fires around the trough of dCA1 theta and during the descending phase of iCA1 theta). (E) Immunofluorescence micrograph (overlay of 9 consecutive optical slices with 1 μm spacing) of neurobiotin-filled axonal boutons which are in direct contact with axon initial segments of pyramidal cells labeled with antibodies against ankyrin G. (F) Immunofluorescence micrographs showing a neurobiotin-filled dendrite immunopositive for PV (left) and the soma positive for ErB4 (right).

Figure 4.

Visualization, molecular expression, and in vivo firing patterns of the identified iCA1 O-LM cell TF12b. (A) Reconstruction of axons (blue, from 1 section, boutons represented by blue spheres) and soma and dendrites (red, complete from 9 consecutive sections). (B) Schematic representation of the cell in a coronal brain slice [5.3 mm posterior from Bregma, after Paxinos and Watson (2007) The Rat Brain]. (C) Firing pattern during SWR events showing different ripple occurrences in dCA1 and iCA1 and activation of cell TF12b during iCA1 but not all dCA1 SWR events. (D) Firing pattern during theta oscillations. A shift between the dCA1 and iCA1 LFP theta and a consistent phase coupling to theta is shown (cell fired 1 or 2 spikes during early ascending phase of dCA1 theta and trough of iCA1 theta). (E) Immunofluorescence micrograph of neurobiotin-filled dendrites and soma (below) in the stratum oriens immunopositive for mGlurR1α and somatostatin, respectively. (F) Immunofluorescence micrograph overlay showing mGluR7a-positive inputs (red dots) to a neurobiotin-filled distal dendrite (green) in the stratum oriens.

Figure 4.

Visualization, molecular expression, and in vivo firing patterns of the identified iCA1 O-LM cell TF12b. (A) Reconstruction of axons (blue, from 1 section, boutons represented by blue spheres) and soma and dendrites (red, complete from 9 consecutive sections). (B) Schematic representation of the cell in a coronal brain slice [5.3 mm posterior from Bregma, after Paxinos and Watson (2007) The Rat Brain]. (C) Firing pattern during SWR events showing different ripple occurrences in dCA1 and iCA1 and activation of cell TF12b during iCA1 but not all dCA1 SWR events. (D) Firing pattern during theta oscillations. A shift between the dCA1 and iCA1 LFP theta and a consistent phase coupling to theta is shown (cell fired 1 or 2 spikes during early ascending phase of dCA1 theta and trough of iCA1 theta). (E) Immunofluorescence micrograph of neurobiotin-filled dendrites and soma (below) in the stratum oriens immunopositive for mGlurR1α and somatostatin, respectively. (F) Immunofluorescence micrograph overlay showing mGluR7a-positive inputs (red dots) to a neurobiotin-filled distal dendrite (green) in the stratum oriens.

Axonal Arborizations, Dendritic Branching, and Molecular Expressions of iCA1 PV+ Basket Cells

Six of the recorded interneurons were identified as PV+ basket cells. All cells expressed the calcium binding protein PV, tested on either dendrites or soma, and had most of their axonal boutons in the stratum pyramidale of iCA1 (Fig. 2A,E). The axons of all PV+ basket cells except those of cell TF29a were distributed symmetrically around the soma (see Fig. 2A for TF29a axon distributions). The somata of all PV+ baskets were located in the stratum pyramidale. The dendrites were oriented vertically, with one part reaching into the alveus and the other cone of dendrites reaching into stratum lacunosum-moleculare. The dendrites of some cells hardly entered stratum lacunosum-moleculare, while the dendrites of other cells extended into the middle of stratum lacunosum-moleculare. However, the dendrites of cell TF29a reached the hippocampal fissure and exhibited unusual extensive branching in stratum lacunosum-moleculare (Fig. 2A). Cell TF29a also differed from the other PV+ basket cells in its narrow dendritic tree in the stratum radiatum compared with a wider distribution of dendrites of the other iCA1 PV+ basket cells and as shown by dCA1 PV+ basket cells (Klausberger et al. 2003).

To further confirm PV+ basket cell identity and test if iCA1 PV+ basket cells can be distinguished from other cell types by their molecular expression profile, we tested for known markers of bistratified cells (somatostatin and neuropeptide Y) and cholecystokinin-expressing (CCK+) cells (CCK and CBR1) for which all the tested cells were negative (Table 1). In addition, PV+ basket cells in the iCA1 were tested immunopositive for ErbB4 (Table 1; Fig. 2E).

Firing Patterns of iCA1 PV+ Basket Cells During Sharp Wave-Associated Ripples and Theta Oscillations

The firing patterns of iCA1 PV+ basket cells were analyzed during SWR detected in iCA1 and dCA1. All PV+ basket cells were strongly active during iCA1 and dCA1 SWR (Fig. 2C) with average firing rates of 53 ± 28 Hz (mean ± SD) and 55 ± 24 Hz during SWR in iCA1 and dCA1, respectively (Fig. 6A,B). These cells significantly increased their firing rate during SWR and were active during each iCA1 SWR event (Fig. 6C).

Figure 6.

Distinct firing patterns of identified iCA1 interneurons during SWR. (A) Different cell types show distinct firing patterns during SWR events detected either in dCA1 (left) and iCA1 (right). (B) Firing rates of individual interneurons (colored crosses) during SWR events and mean rate for each cell type (bars). (C) Percentage of SWR events with firing of individual cells (colored crosses) and mean for each cell type (bars).

Figure 6.

Distinct firing patterns of identified iCA1 interneurons during SWR. (A) Different cell types show distinct firing patterns during SWR events detected either in dCA1 (left) and iCA1 (right). (B) Firing rates of individual interneurons (colored crosses) during SWR events and mean rate for each cell type (bars). (C) Percentage of SWR events with firing of individual cells (colored crosses) and mean for each cell type (bars).

The iCA1 PV+ basket cells were highly active during theta oscillations (Fig. 2D) with an average firing rate of 18 ± 11.8 Hz and their firing was significantly phase coupled to theta oscillations (for all cells: P < 0.001, Rayleigh test). All cells fired with the highest probability during the late descending phase of iCA1 theta oscillations (mean phase angle 328°; 0° and 360° mark theta toughs) and early ascending phase (mean phase angle 404°) of dCA1 theta oscillations (Figs 2C and 7A,B). Relative to the dCA1 theta cycle as a common reference, the spikes of these iCA1 basket cells occurred significantly later (two-sample permutation test, P = 0.013) than the spikes of dCA1 PV+ basket cells, reported in Klausberger et al. (2003).

Figure 7.

Distinct firing patterns of identified iCA1 interneurons during theta oscillations. (A) Different interneuron types show distinct phase coupling to theta oscillations detected in stratum pyramidale or oriens of dCA1 (left) or iCA1 (right). (B) Mean phase angles of individual interneurons during theta oscillations (same color code as in A). (C) Distinct types of dCA1 and iCA1 interneurons fire in the same order of cell types but shifted in time. This shift in spike-timing corresponds to the detected shift in theta oscillations in the dCA1 and iCA1 LFPs. Bold lines mark the average phase angle for each cell type, regular lines mark the mean phase angles for individual cells, and shadowed areas mark the range of mean phase firing for each neuron type. For pyramidal cells, the overall mean firing phase is indicated. Data for dCA1 interneurons and pyramidal cells were reported in Klausberger et al. (2003).

Figure 7.

Distinct firing patterns of identified iCA1 interneurons during theta oscillations. (A) Different interneuron types show distinct phase coupling to theta oscillations detected in stratum pyramidale or oriens of dCA1 (left) or iCA1 (right). (B) Mean phase angles of individual interneurons during theta oscillations (same color code as in A). (C) Distinct types of dCA1 and iCA1 interneurons fire in the same order of cell types but shifted in time. This shift in spike-timing corresponds to the detected shift in theta oscillations in the dCA1 and iCA1 LFPs. Bold lines mark the average phase angle for each cell type, regular lines mark the mean phase angles for individual cells, and shadowed areas mark the range of mean phase firing for each neuron type. For pyramidal cells, the overall mean firing phase is indicated. Data for dCA1 interneurons and pyramidal cells were reported in Klausberger et al. (2003).

Axonal Arborizations, Dendritic Branching, and Molecular Expressions of iCA1 Axo-Axonic Cells

Two cells were identified as iCA1 axo-axonic cells. Both cells were tested positive for PV and ErbB4 (Table 1; Fig. 3F). Both axo-axonic cells exhibited the typical axonal cartridges suggesting innervation of axon initial segments of pyramidal cells mostly in the deeper stratum pyramidale and neighboring stratum oriens (Fig. 3A), and the axons were symmetrically distributed around the soma. To verify the highly specific axonal targeting of these cells, we labeled the axon initial segments of pyramidal cells with antibodies against ankyrin G and showed they were enwound by the neurobiotin-filled axonal boutons of the labeled cells (Fig. 3E). We further quantified the targeting of the labeled cells with confocal immunoflourescence microscopy and found that 93% of boutons (434 of 465 randomly sampled axonal boutons) from cell TF31a and 97% of boutons (300 of 310) from cell TF35c were in direct contact with ankyrin G expressing axon initial segments of pyramidal cells. The soma of cell TF31a was located in the stratum pyramidale and the soma of cell TF35c was in the stratum oriens close to the alveus. The dendrites of both cells were oriented mostly vertically to the stratum pyramidale and reached deep into the stratum lacunosum-moleculare where they branched heavily (Fig. 3A). The dendrites in the stratum oriens reached deeper into the alveus compared with PV+ basket cells.

Firing Patterns of iCA1 Axo-Axonic Cells During Sharp Wave-Associated Ripples and Theta Oscillations

The 2 iCA1 axo-axonic cells fired strongly before the onset of both iCA1 and dCA1 SWR, but decreased their firing rate as the SWR events started (Figs 3C and 6A). During occasional SWR that occurred only in dCA1 but not in iCA1, cell TF35c was active (Fig. 3C). The iCA1 axo-axonic cells were highly active during theta oscillations (average firing rate: 20 ± 5 Hz), firing phase coupled (for both cells: P < 0.001, Rayleigh test) to the mid-descending phase of iCA1 theta and late descending phase/trough of dCA1 theta oscillations (Figs 3C and 7A,B; average of iCA1 mean phase angles: 273°, dCA1: 336°).

Axonal Arborizations, Dendritic Branching, and Molecular Expressions of iCA1 O-LM Cells

Six cells were recorded and identified as iCA1 O-LM cells. The iCA1 O-LM cells showed the largest variability in several respects but were defined as O-LM cells based on 1) their main axon, projecting into the stratum lacunosum-moleculare and heavily innervating this layer; 2) their spiny horizontal dendrites in the stratum oriens; and 3) the expression of mGluR1α, tested in the dendrites or soma (Table 1; Fig. 4). In addition, the soma of all examined O-LM cells expressed somatostatin and either their dendrites or their soma received mGluR7α expressing synaptic inputs (Table 1; Fig. 4F). Of 5 cells tested, 4 were weakly immunopositive for PV, but cell TF12b was tested immunonegative. Immunoreactions for ErbB4 were negative for 3 cells tested (Table 1).

The positions of the somata varied from the midstratum oriens to the oriens/alveus border. In cells with the soma located in the middle of the stratum oriens, such as TF12b shown in Figure 4A, the dendrites often turned first toward the alveus and then ran parallel to the stratum pyramidale in the stratum oriens. Cells with somata close to the stratum oriens/alveus border had dendrites running directly horizontally away from the soma. In all cells, the proximal dendrites were smooth and turned with distance increasingly varicose and spiny. The axonal distribution of the different cells varied considerably. The axons extended throughout an unexpected large range along the anteroposterior (average: 1143 ± 176 µm) and dorsoventral (average: 985 ± 103 µm) axis. In 3 O-LM cells (TF12b, TF16b, and TF26a), the somata were located in a similar position within the more anterior and dorsal part of the iCA1 (Fig. 1A), but their axonal projections showed some variability. Axons of cell TF12b strongly innervated the iCA1 and posterior dCA1/CA2 and CA3 stratum lacunosum-moleculare but, to a lesser extent, the stratum lacunosum-moleculare posterior to the soma. In contrast axons of cells, TF16b and TF26a did not extend as far into the dCA1, CA2, and CA3 with TF16b innervating more heavily the stratum lacunosum-moleculare in the iCA1 posterior to the soma position. We also observed that cell TF12b showed a weak but consistent innervation of axon collaterals with boutons across the hippocampal fissure into about half of the molecular layer of the dentate gyrus, only very rarely observed in another iCA1 O-LM cell TF26a.

Firing Patterns of iCA1 O-LM Cells During Sharp Wave-Associated Ripples and Theta Oscillations

The firing patterns of iCA1 O-LM cells also showed some variability. Some cells, such as TF12b (Fig. 4C), were active during some, but not all, SWR and showed a significant coupling to SWR events. In contrast to PV+ basket cells, activated O-LM cells fired only few spikes during SWR and were also less active during irregular hippocampal LFP outside of SWR (Fig. 4C). Other O-LM cells were not activated during SWR events (Fig. 6). We were not able to correlate this variable firing during SWR with the variability we found in our anatomical and immunohistochemical analyses.

The O-LM cells were coupled to theta oscillations (for all cells: P < 0.001, Rayleigh test), with all cells firing on average with the highest probability during trough/early ascending phase of iCA1 theta (average of mean angles 19°) and mid-ascending phase of dCA1 theta (average of mean angles 54°) but at a lower firing rate than PV basket or axo-axonic cells with 1 or 2 spikes per theta cycle (average firing rate: 3.4 ± 1.4 Hz, Figs 4D and 7A). The spike-timing of these iCA1 O-LM cells (relative to dCA1 theta as a common reference) occurred significantly later (P = 0.024; two-sample permutation test) compared with the mean firing phase of 3 O-LM cells in the dCA1 area, reported in Klausberger et al. (2003). Cell TF26a fired usually 1 spike per theta cycle, and we noticed that, occasionally, the spikes changed their theta phase angle from 1 cycle to the next. We investigated whether this change was directional and found that during some theta periods the spike advanced consistently from 1 theta cycle to the next (Fig. 5). A comparison of the interspike time interval to the time between the peaks of the theta oscillations during the whole recording also showed a significant difference (P < 0.01, Wilcoxon signed-rank test; Fig. 5D). However, on average, this cell was still coupled to the trough of iCA1 theta oscillations (Fig. 7A,B).

Figure 5.

Phase coupling changes during theta oscillation epochs of the O-LM cells TF26a. (A) TF26a fired in general 1 spike per theta cycle and the spike often advanced in phase as shown in this example period. (B) Plot illustrating the advancement of spike phase in A. The relative phase in comparison to the previous spike phase is plotted against the spike number over time. (C) Spike phase changes during theta oscillations throughout the whole recording. (D) The spike-to-spike interval from all theta periods is shorter than the theta peak-to-peak interval (P < 0.01, Wilcoxon signed-rank test).

Figure 5.

Phase coupling changes during theta oscillation epochs of the O-LM cells TF26a. (A) TF26a fired in general 1 spike per theta cycle and the spike often advanced in phase as shown in this example period. (B) Plot illustrating the advancement of spike phase in A. The relative phase in comparison to the previous spike phase is plotted against the spike number over time. (C) Spike phase changes during theta oscillations throughout the whole recording. (D) The spike-to-spike interval from all theta periods is shorter than the theta peak-to-peak interval (P < 0.01, Wilcoxon signed-rank test).

Firing of iCA1 Pyramidal Cells During Theta Oscillations

We also recorded, labeled, and identified pyramidal neurons in the iCA1 hippocampus. They differed from interneurons in their spike width >1.5 ms and low activity during theta oscillations (average firing rate 1.8 ± 1.5 Hz); in addition, pyramidal cells exhibited thick apical dendrites directed toward the stratum lacunosum-moleculare, where they formed a wide apical tuft of spiny dendrites. Consistent with recent reports describing different types of pyramidal cells (Klausberger 2009; Mizuseki et al. 2011; Graves et al. 2012), we also observed pyramidal cells with different characteristics in the iCA1, suggesting that they might belong to different types. In order to compare our findings to previous reports, we treated the pyramidal cells as one group and included only those 10 cells that were coupled significantly to theta oscillations (for all cells: P < 0.05, Rayleigh test). Those cells showed some variability in their phase coupling and their activity during theta but were coupled on average to the trough/early ascending phase of iCA1 theta and mid-descending phase of dCA1 theta (average of mean iCA1 phase angles: 42°, dCA1: 58°; Fig. 7), consistent with previous reports on iCA1 and dCA1 pyramidal cells (Hartwich et al. 2009; Patel et al. 2012).

Spike Timing Across the Dorsal and Intermediate CA1 Hippocampus

The simultaneous recordings and the coherence of theta oscillations in the iCA1 and dCA1 allowed us to combine our findings with data from dCA1 (Klausberger et al. 2003) and propose a model of the timing of interneurons during travelling theta oscillations across the dCA1 and iCA1 (Fig. 7C). This model represents a snapshot of 1 theta cycle and the firing patterns of 3 types of interneuron in relation to pyramidal cells in the dCA1 and iCA1 and is based on 3 observations made here: 1) Theta oscillations of the LFP in iCA1 are shifted in time reflecting the average lag in phase of +49°, which we measured in the iCA1 compared with the dCA1 theta phase. 2) The temporal sequence of mean firing of distinct cell types is the same in dCA1 and iCA1 during theta oscillations: first axo-axonic cells fire, followed by PV+ baskets, and then O-LM and pyramidal cells discharge. 3) The firing of distinct iCA1 neurons is shifted in time relative to the mean firing phase of the corresponding cell types in the dCA1. These findings indicate that, in a travelling theta wave, the entire local microcircuit consisting of pyramidal cells and distinct types of interneurons shift concertedly across different parts of the septotemporal axis.

Discussion

From the data obtained in this study, we conclude that distinct types of GABAergic interneurons contribute differentially to neuronal timing across the dCA1 and iCA1 hippocampus. We found that SWR events occur, on average, simultaneously in dCA1 and iCA1; in contrast, theta oscillations are coherent but shifted in phase. Distinct types of iCA1 interneuron contribute differentially to SWR events also in the iCA1. During theta oscillations, a defined firing sequence of distinct interneuron types is shifted in time between dCA1 and iCA1 governing the formation of a travelling theta wave across the septotemporal axis of the CA1 hippocampus. These data indicate that the cell type-specific firing of GABAergic interneurons provide the temporal frame for local oscillations as well as the coordination across the hippocampus.

Temporal Organization of Sharp Wave-Associated Ripple Activity Across the dCA1 and iCA1 Hippocampus

SWR events in the CA1 LFP reflect highly synchronous input from the CA3 area and spiking activity of local neurons (O'Keefe and Nadel 1978; Buzsaki et al. 1983, 1992; Schomburg et al. 2012), but it is not known how these events are organized along the septotemporal axis. In our recordings, we observed that the majority of iCA1 SWRs occurred around the same time as the SWR events in the dCA1. In the few cases in which we observed a difference between dCA1 and iCA1 SWR, the cells exhibited their typical SWR firing patterns during iCA1 but not the offset dCA1 events. It is possible that, during irregular LFP activity, the same general mechanism opens temporal windows for SWR events to occur across the dCA1 and iCA1, then the SWR can manifest locally recruiting cell assemblies and are organized by interneurons within an area. The firing patterns iCA1 PV+ basket and axo-axonic cells during SWR was similar to previous dCA1 recordings. PV+ basket cells are highly active and thereby contribute to the high-frequency oscillations, while axo-axonic cells are inhibited and thereby release CA1 pyramidal cells from inhibition (Klausberger et al. 2003). In the dCA1, O-LM cells are inhibited during SWR events in anesthetized rats (Klausberger et al. 2003), which differ from the variability in the response of the iCA1 O-LM cells described here. However, in the dCA1 of drug-free mice and in slices, O-LM cells have been described as occasionally activated during SWR events (Varga et al. 2012; Pangalos et al. 2013). It cannot be excluded that the variability we observed in the iCA1 could be due to a greater sensitivity of O-LM cells to the precise level of anesthesia explaining why some cells remained silent and others were activated during SWR events. Our observations, nevertheless, support that iCA1 O-LM cells do not have a fixed response during all SWR events such as PV+ basket or axo-axonic cells, but might be sometimes recruited to facilitate CA3 input (Leao et al. 2012).

Travelling Theta Wave and Theta Phase Coupling of Interneurons Across the dCA1 and iCA1 Hippocampus

Theta oscillations are thought to play an important role in the encoding of information by defining a temporal window during which space is represented by the highly organized timing and firing rate of pyramidal cells (O'Keefe and Nadel 1978; Soltesz and Deschenes 1993; Skaggs et al. 1996; Buzsaki 2002). This raises the question about the implications of traveling theta oscillations on information processing and representation in the CA1 hippocampus. The hypothesis has been put forward that the traveling theta wave increases the information content to downstream targets of the CA1 by representing a segment of space, rather than a single location, at each time point (Lubenov and Siapas 2009). Our analysis of the timing of neural activity allowed us to test how distinct types of GABAergic interneuron, which are considered as the major rhythm generators in the CA1 hippocampus, contribute to traveling theta waves. We found that during theta oscillations the firing of the interneuron types and pyramidal cells is shifted together with the shift in LFP theta oscillations across the dCA1 and iCA1 hippocampus. The temporal order of the firing of distinct interneuron types is preserved in the iCA1 suggesting that similar mechanisms of the local circuitry are operating in the dCA1 and iCA1. However, this interneuronal building block of a theta cycle is gradually shifting in phase across the septotemporal axis pointing to a higher level organization of local circuits in space and time in the CA1 hippocampus. This indicates that a downstream read-out of the hippocampal information can differentiate between information originating from different septotemporal parts of the CA1 based on the theta-related timing.

A mechanism that has been put forward and could generate traveling theta oscillations is based on coupled intrinsic CA1 theta oscillators (Lubenov and Siapas 2009). Although theta oscillations are thought to be mainly driven by inputs from the medial septum, diagonal band of Broca, and the entorhinal cortex (Bland and Oddie 2001; Buzsaki 2002), intrinsic theta oscillations can arise in in vitro preparations of whole hippocampi without any extrahippocampal inputs (Goutagny et al. 2009). In such preparations during the blockade of synaptic transmission along the septotemporal CA1 axis, multiple theta oscillators with different frequencies arise along the septotemporal axis. Therefore, intrinsic theta oscillations that are driven by feedback networks of pyramidal cells and GABAergic interneurons (Goutagny et al. 2009) might contribute to the theta wave. In our results, we also present the O-LM cell TF26a that consistently fired for long time periods at a higher frequency than the concurrent LFP theta oscillations, indicating the presence of intrinsic oscillators of different frequencies within the CA1 network.

Conclusion

We have investigated GABAergic networks in the iCA1, which allowed us to determine how dynamic network activities are organized by the different network components, namely the different GABAergic interneuron types and pyramidal cells. Previous studies have shown that different interneuron types temporally divide the activity during theta oscillations. Here, we have shown that the shift in phase of theta oscillations is linked to a corresponding shift of the spike timing of entire GABAergic networks composed of PV+ basket, axo-axonic, and O-LM cells. This further supports the relationship between the temporal division of network activity by interneurons and their control of the timing of pyramidal cells within theta oscillations. The iCA1 is also a functionally important structure as it receives spatial information but also mixed nonspatial sensory (e.g., olfactory) information from the intermediate band of the entorhinal cortex (Burwell 2000). Place fields gradually increase in their size along the septotemporal axis (Maurer et al. 2005; Kjelstrup et al. 2008) but decrease in size and increase in number in the iCA1 during the presentation of nonspatial information such as 3D objects placed into an environment (Burke et al. 2011). The temporal and axo-dendritic specificity of distinct types of iCA1 interneurons may serve these integrative functions of the iCA1 hippocampus by providing a dynamic and flexible GABAergic control for the local network operations.

Funding

This work was supported by grant 242689 of the European Research Council and grant SCIC03 of the Vienna Science and Technology Fund.

Notes

We thank Simone Latkolik for recording and labeling cells SL03 and SL15b. We also thank Romana Hauer, Erzsebet Borok, and Cornelia Schimke for their excellent technical assistance. Conflict of Interest: None declared.

References

Airaksinen
MS
Eilers
J
Garaschuk
O
Thoenen
H
Konnerth
A
Meyer
M
.
1997
.
Ataxia and altered dendritic calcium signaling in mice carrying a targeted null mutation of the calbindin D28k gene
.
Proc Natl Acad Sci USA
 .
94
:
1488
1493
.
Alvarez
FJ
Villalba
RM
Carr
PA
Grandes
P
Somohano
PM
.
2000
.
Differential distribution of metabotropic glutamate receptors 1a, 1b, and 5 in the rat spinal cord
.
J Comp Neurol
 .
422
:
464
487
.
Bannerman
DM
Grubb
M
Deacon
RM
Yee
BK
Feldon
J
Rawlins
JN
.
2003
.
Ventral hippocampal lesions affect anxiety but not spatial learning
.
Behav Brain Res
 .
139
:
197
213
.
Bast
T
.
2011
.
The hippocampal learning-behavior translation and the functional significance of hippocampal dysfunction in schizophrenia
.
Curr Opin Neurobiol
 .
21
:
492
501
.
Bast
T
.
2007
.
Toward an integrative perspective on hippocampal function: from the rapid encoding of experience to adaptive behavior
.
Rev Neurosci
 .
18
:
253
281
.
Berens
P
.
2009
.
CircStat: a MATLAB toolbox for circular statistics
.
J Stat Softw
 .
31
:
1
21
.
Bland
BH
Oddie
SD
.
2001
.
Theta band oscillation and synchrony in the hippocampal formation and associated structures: the case for its role in sensorimotor integration
.
Behav Brain Res
 .
127
:
119
136
.
Burke
SN
Maurer
AP
Nematollahi
S
Uprety
AR
Wallace
JL
Barnes
CA
.
2011
.
The influence of objects on place field expression and size in distal hippocampal CA1
.
Hippocampus
 .
21
:
783
801
.
Burwell
RD
.
2000
.
The parahippocampal region: corticocortical connectivity
.
Ann N Y Acad Sci
 .
911
:
25
42
.
Buzsaki
G
.
2002
.
Theta oscillations in the hippocampus
.
Neuron
 .
33
:
325
340
.
Buzsaki
G
Draguhn
A
.
2004
.
Neuronal oscillations in cortical networks
.
Science
 .
304
:
1926
1929
.
Buzsaki
G
Horvath
Z
Urioste
R
Hetke
J
Wise
K
.
1992
.
High-frequency network oscillation in the hippocampus
.
Science
 .
256
:
1025
1027
.
Buzsaki
G
Leung
LW
Vanderwolf
CH
.
1983
.
Cellular bases of hippocampal EEG in the behaving rat
.
Brain Res
 .
287
:
139
171
.
Cenquizca
LA
Swanson
LW
.
2007
.
Spatial organization of direct hippocampal field CA1 axonal projections to the rest of the cerebral cortex
.
Brain Res Rev
 .
56
:
1
26
.
Chen
X
Levkowitz
G
Tzahar
E
Karunagaran
D
Lavi
S
Ben-Baruch
N
Leitner
O
Ratzkin
BJ
Bacus
SS
Yarden
Y
.
1996
.
An immunological approach reveals biological differences between the two NDF/heregulin receptors, ErbB-3 and ErbB-4
.
J Biol Chem
 .
271
:
7620
7629
.
Constantinople
CM
Disney
AA
Maffie
J
Rudy
B
Hawken
MJ
.
2009
.
Quantitative analysis of neurons with Kv3 potassium channel subunits, Kv3.1b and Kv3.2, in macaque primary visual cortex
.
J Comp Neurol
 .
516
:
291
311
.
Diba
K
Buzsaki
G
.
2007
.
Forward and reverse hippocampal place-cell sequences during ripples
.
Nat Neurosci
 .
10
:
1241
1242
.
Dong
HW
Swanson
LW
Chen
L
Fanselow
MS
Toga
AW
.
2009
.
Genomic-anatomic evidence for distinct functional domains in hippocampal field CA1
.
Proc Natl Acad Sci USA
 .
106
:
11794
11799
.
Ego-Stengel
V
Wilson
MA
.
2010
.
Disruption of ripple-associated hippocampal activity during rest impairs spatial learning in the rat
.
Hippocampus
 .
20
:
1
10
.
Fanselow
M
Dong
H
.
2010
.
Are the dorsal and ventral hippocampus functionally distinct structures?
Neuron
 .
65
:
7
19
.
Foster
DJ
Wilson
MA
.
2006
.
Reverse replay of behavioural sequences in hippocampal place cells during the awake state
.
Nature
 .
440
:
680
683
.
Fukudome
Y
Ohno-Shosaku
T
Matsui
M
Omori
Y
Fukaya
M
Tsubokawa
H
Taketo
MM
Watanabe
M
Manabe
T
Kano
M
.
2004
.
Two distinct classes of muscarinic action on hippocampal inhibitory synapses: M2-mediated direct suppression and M1/M3-mediated indirect suppression through endocannabinoid signalling
.
Eur J Neurosci
 .
19
:
2682
2692
.
Girardeau
G
Benchenane
K
Wiener
SI
Buzsaki
G
Zugaro
MB
.
2009
.
Selective suppression of hippocampal ripples impairs spatial memory
.
Nat Neurosci
 .
12
:
1222
1223
.
Goutagny
R
Jackson
J
Williams
S
.
2009
.
Self-generated theta oscillations in the hippocampus
.
Nat Neurosci
 .
12
:
1491
1493
.
Graves
AR
Moore
SJ
Bloss
EB
Mensh
BD
Kath
WL
Spruston
N
.
2012
.
Hippocampal pyramidal neurons comprise two distinct cell types that are countermodulated by metabotropic receptors
.
Neuron
 .
76
:
776
789
.
Hartwich
K
Pollak
T
Klausberger
T
.
2009
.
Distinct firing patterns of identified basket and dendrite-targeting interneurons in the prefrontal cortex during hippocampal theta and local spindle oscillations
.
J Neurosci
 .
29
:
9563
9574
.
Kjelstrup
KB
Solstad
T
Brun
VH
Hafting
T
Leutgeb
S
Witter
MP
Moser
EI
Moser
MB
.
2008
.
Finite scale of spatial representation in the hippocampus
.
Science
 .
321
:
140
143
.
Klausberger
T
.
2009
.
GABAergic interneurons targeting dendrites of pyramidal cells in the CA1 area of the hippocampus
.
Eur J Neurosci
 .
30
:
947
957
.
Klausberger
T
Magill
PJ
Marton
LF
Roberts
JD
Cobden
PM
Buzsaki
G
Somogyi
P
.
2003
.
Brain-state- and cell-type-specific firing of hippocampal interneurons in vivo
.
Nature
 .
421
:
844
848
.
Klausberger
T
Marton
L
O'Neill
J
Huck
J
Dalezios
Y
Fuentealba
P
Suen
W
Papp
E
Kaneko
T
Watanabe
M
et al
2005
.
Complementary roles of cholecystokinin- and parvalbumin-expressing GABAergic neurons in hippocampal network oscillations
.
J Neurosci
 .
25
:
9782
9793
.
Klausberger
T
Somogyi
P
.
2008
.
Neuronal diversity and temporal dynamics: the unity of hippocampal circuit operations
.
Science
 .
321
:
53
57
.
Kubota
Y
Karube
F
Nomura
M
Gulledge
AT
Mochizuki
A
Schertel
A
Kawaguchi
Y
.
2011
.
Conserved properties of dendritic trees in four cortical interneuron subtypes
.
Sci Rep
 .
1
:
89
.
Lasztoczi
B
Tukker
JJ
Somogyi
P
Klausberger
T
.
2011
.
Terminal field and firing selectivity of cholecystokinin-expressing interneurons in the hippocampal CA3 area
.
J Neurosci
 .
31
:
18073
18093
.
Leao
RN
Mikulovic
S
Leao
KE
Munguba
H
Gezelius
H
Enjin
A
Patra
K
Eriksson
A
Loew
LM
Tort
AB
et al
2012
.
OLM interneurons differentially modulate CA3 and entorhinal inputs to hippocampal CA1 neurons
.
Nat Neurosci
 .
15
:
1524
1530
.
Levey
AI
Edmunds
SM
Hersch
SM
Wiley
RG
Heilman
CJ
.
1995
.
Light and electron microscopic study of m2 muscarinic acetylcholine receptor in the basal forebrain of the rat
.
J Comp Neurol
 .
351
:
339
356
.
Lubenov
EV
Siapas
AG
.
2009
.
Hippocampal theta oscillations are travelling waves
.
Nature
 .
459
:
534
539
.
Maurer
AP
Vanrhoads
SR
Sutherland
GR
Lipa
P
McNaughton
BL
.
2005
.
Self-motion and the origin of differential spatial scaling along the septo-temporal axis of the hippocampus
.
Hippocampus
 .
15
:
841
852
.
Mizuseki
K
Diba
K
Pastalkova
E
Buzsáki
G
.
2011
.
Hippocampal CA1 pyramidal cells form functionally distinct sublayers
.
Nat Neurosci
 .
14
:
1174
1181
.
Moser
MB
Moser
EI
.
1998
.
Functional differentiation in the hippocampus
.
Hippocampus
 .
8
:
608
619
.
Nakamura
M
Sato
K
Fukaya
M
Araishi
K
Aiba
A
Kano
M
Watanabe
M
.
2004
.
Signaling complex formation of phospholipase Cβ4 with metabotropic glutamate receptor type 1α and 1,4,5-trisphosphate receptor at the perisynapse and endoplasmic reticulum in the mouse brain
.
Eur J Neurosci
 .
20
:
2929
2944
.
O'Keefe
J
.
1976
.
Place units in the hippocampus of the freely moving rat
.
Exp Neurol
 .
51
:
78
109
.
O'Keefe
J
Nadel
L
.
1978
.
The hippocampus as a cognitive map
 .
Oxford
:
Clarendon
.
O'Keefe
J
Recce
ML
.
1993
.
Phase relationship between hippocampal place units and the EEG theta rhythm
.
Hippocampus
 .
3
:
317
330
.
Pangalos
M
Donoso
JR
Winterer
J
Zivkovic
AR
Kempter
R
Maier
N
Schmitz
D
.
2013
.
Recruitment of oriens-lacunosum-moleculare interneurons during hippocampal ripples
.
Proc Natl Acad Sci USA
 .
110
:
4398
4403
.
Patel
J
Fujisawa
S
Berenyi
A
Royer
S
Buzsaki
G
.
2012
.
Traveling theta waves along the entire septotemporal axis of the hippocampus
.
Neuron
 .
75
:
410
417
.
Paxinos
G
Watson
C
.
2007
.
The rat brain in stereotaxic coordinates
 .
6th edn
.
London
(
UK
):
Elsevier
.
Pinault
D
.
1996
.
A novel single-cell staining procedure performed in vivo under electrophysiological control: morpho-functional features of juxtacellularly labeled thalamic cells and other central neurons with biocytin or neurobiotin
.
J Neurosci Methods
 .
65
:
113
136
.
Schomburg
EW
Anastassiou
CA
Buzsaki
G
Koch
C
.
2012
.
The spiking component of oscillatory extracellular potentials in the rat hippocampus
.
J Neurosci
 .
32
:
11798
11811
.
Shigemoto
R
Kulik
A
Roberts
JDB
Ohishi
H
Nusser
Z
Kaneko
T
Somogyi
P
.
1996
.
Target-cell-specific concentration of a metabotropic glutamate receptor in the presynaptic active zone
.
Nature
 .
381
:
523
525
.
Skaggs
WE
McNaughton
BL
Wilson
MA
Barnes
CA
.
1996
.
Theta phase precession in hippocampal neuronal populations and the compression of temporal sequences
.
Hippocampus
 .
6
:
149
172
.
Soltesz
I
Deschenes
M
.
1993
.
Low- and high-frequency membrane potential oscillations during theta activity in CA1 and CA3 pyramidal neurons of the rat hippocampus under ketamine-xylazine anesthesia
.
J Neurophysiol
 .
70
:
97
116
.
Varga
C
Golshani
P
Soltesz
I
.
2012
.
Frequency-invariant temporal ordering of interneuronal discharges during hippocampal oscillations in awake mice
.
Proc Natl Acad Sci USA
 .
109
:
E2726
E2734
.

Author notes

Thomas Forro and Ornella Valenti contributed to this work to a comparable extent.