Inhibitory conductance controls place field dynamics in the hippocampus

Inhibitory conductance controls place field dynamics in the hippocampus Manuel Valero1,5, Andrea Navas-Olive2,5, Liset M. de la Prida2,*, György Buzsáki1,3,4,6,* 1Neuroscience Institute, Langone Medical Center, New York University, New York, NY 10016, USA 2Instituto Cajal, Consejo Superior de Investigaciones Científicas, Avenue Doctor Arce 37, Madrid 28002, Spain 3Center for Neural Science, New York University, New York, NY 10003, USA 4Department of Neurology, Langone Medical Center, New York, NY 10016, USA 5These authors contributed equally 6Lead contact SUMMARY Hippocampal place cells receive a disparate collection of excitatory and inhibitory currents that endow them with spatially selective discharges and rhythmic activity. Using a combination of in vivo intracellular and extracellular recordings with opto/chemogenetic manipulations and computational modeling, we investigate the influence of inhibitory and excitatory inputs on CA1 pyramidal cell responses. At the cell bodies, inhibition leads and is stronger than excitation across the entire theta cycle. Pyramidal neurons fire on the ascending phase of theta when released from inhibition. Computational models equipped with the observed conductances reproduce these dynamics. In these models, place field properties are favored when the increased excitation is coupled with a reduction of inhibition within the field. As predicted by our simulations, firing rate within place fields and phase locking to theta are impaired by DREADDs activation of interneurons. Our results indicate that decreased inhibitory conductance is critical for place field expression. In brief Valero et al. examine the influence of inhibition on place fields. They show that hippocampal neurons are dominated by inhibitory conductances during theta oscillations. A transient increase This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). *Correspondence: [email protected] (L.M.d.l.P.), [email protected] (G.B.). AUTHOR CONTRIBUTIONS Conceptualization, M.V., G.B., and L.M.P. Methodology: M.V., G.B., and L.M.P. Software: M.V. and A.N.O. Formal analysis: M.V. and A.N.O. Investigation: M.V. Data curation: M.V. Writing – original draft: M.V. and G.B. Writing – review & editing: all authors. Visualization: M.V. and A.N.O. Supervision: G.B., M.V., and L.M.P. Funding acquisition: G.B., L.M.P and M.V. DECLARATION OF INTERESTS The authors declare no competing interests. SUPPLEMENTAL INFORMATION Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2022.111232. HHS Public Access Author manuscript Cell Rep . Author manuscript; available in PMC 2022 October 25. Published in final edited form as: Cell Rep . 2022 August 23; 40(8): 111232. doi:10.1016/j.celrep.2022.111232. Author Manuscript Author Manuscript Author Manuscript Author Manuscript of excitation and drop of inhibition mediates place field emergence in simulations. Consistently, chemogenetic activation of interneurons deteriorates place cell properties in vivo . Graphical Abstract INTRODUCTION Understanding how local circuits change yet maintain their own dynamics in response to intermittently occurring inputs is necessary for revealing the transformation rules of local computation. The hippocampal CA1 region is a unique case because of its weak and sparse excitatory recurrence and because it receives unidirectional afferents from both layer 3 of the entorhinal cortex and the CA3 region (Andersen et al., 2009; Deuchars and Thomson, 1996). Based mainly on anatomical considerations, it has been postulated that the firing patterns of CA1 pyramidal neurons are largely “inherited” or induced by these upstream regions (Ahmed and Mehta, 2009; Grienberger et al., 2017; Hafting et al., 2005; Mizumori et al., 1989; Rolls et al., 2006; Savelli and Knierim, 2010; Skaggs et al., 1996; Solstad et al., 2006; Steffenach et al., 2005). One of the most studied dynamic patterns in the CA1 region is spatially tuned firing of pyramidal neurons (O’Keefe and Nadel, 1978). When an animal explores its environment, a fraction of pyramidal neurons becomes sequentially active along its movement path. Transient cell assemblies, presumably activated by external “spatial inputs,” fire Valero et al. Page 2 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript rhythmically and at a frequency faster than the summed membrane potentials of the majority of pyramidal neurons, as reflected by the local field potential (LFP) theta rhythm (O’keefe and Recce, 1993). Consequently, the interference between the faster oscillating active cell assembly and background theta rhythm is reflected by “phase precession” of place cell spikes relative to the ongoing LFP theta oscillation (O’keefe and Recce, 1993; Skaggs et al., 1996). This theta “phase coding” has been shown to be more spatially precise, especially at shorter time scales, than relating on positional firing rates (Lisman, 2005; O’keefe and Recce, 1993). An alternative formulation is that the CA1 circuit is endowed with the computational capacity to generate assembly sequences, which are detected secondarily as sequentially active place fields during maze performance (Dragoi and Tonegawa, 2010; Grosmark and Buzsáki, 2016; McKenzie et al., 2021; Valero et al., 2022; Zutshi et al., 2022). In support of this view, recent studies have shown that functional deafferentation of CA1 from upstream inputs has limited impact on the fraction of observable CA1 place fields (Fernández-Ruiz et al., 2021; Kanter et al., 2017; Latuske et al., 2018; Miao et al., 2015; Rueckemann et al., 2016; Zutshi et al., 2022). The implication of this latter view is that the main features of the CA1 dynamics largely emerge from its local circuit, while upstream inputs play a role in selecting the initial condition for sequences. For single place cells, these features include slow depolarization and increased amplitude of intracellular theta rhythm within the place field and phase precession of the emitted spikes relative to the extracellularly recorded theta cycle (Grienberger et al., 2017; Harvey et al., 2009). Phase precession appears to involve a consortium of circuit and single neuron mechanisms (Figure 1A; Burgess et al., 1994; Fernández-Ruiz et al., 2017; Harvey et al., 2009; Hasselmo et al., 2002; Jensen and Lisman, 2000; Kamondi et al., 1998; Magee, 2001; Skaggs et al., 1996). A key common variable in the different models is the relationship between excitation and inhibition within and outside the place field, which has been the subject of intense debate, due mainly to the paucity of observational data (Grienberger et al., 2017; Harvey et al., 2009; Lapray et al., 2012; Royer et al., 2012). Perisomatic inhibitory drive within the place field has been proposed to increase (Bhatia et al., 2019; Milstein et al., 2021), remain unchanged (Grienberger et al., 2017), or decrease (Geiller et al., 2022; Valero et al., 2022) inside the place field (Figure 1B). Yet, how competition between excitatory and inhibitory conductances contributes to the fundamental properties of place cells has remained largely unexplained (Figure 1C). To investigate the role of inhibition in theta rhythmicity and spatial modulation in the hippocampus, we combined intracellular and extracellular recordings with computational modeling and optogenetic/chemogenetic manipulations. We observed that outside the place field, the inhibitory conductance is several times stronger than the excitatory conductance, and that inhibition precedes excitation in each theta cycle. Our key observation is that decreased inhibitory conductance is an important contributor to spike discharge dynamics, in line with recent works showing a downmodulation of the inhibitory inputs alongside place fields (Geiller et al., 2022; Valero et al., 2022). Constrained by our observations, we analyzed how inhibitory and excitatory conductances interact with upstream spatial inputs in a single neuron model and, in turn, tested the model predictions by pharmacogenetic activation of interneurons. Our findings suggest that the spatial tuning of local inhibition is a fundamental mechanism of both rate and phase coding in the hippocampal CA1 region. Valero et al. Page 3 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript RESULTS Excitatory and inhibition conductances during theta oscillations To probe the synaptic mechanism of theta phase locking of spikes, we combined intracellular and multi-site silicon probe recordings (16–64 channels) in the dorsal CA1 region of anesthetized rats and awake head-fixed mice (Figure 2A; 11 cells from 10 rats and seven cells from four mice; Figures S1A–S1C). The intracellular membrane potential (Vm) oscillations were coherent with the LFP. As previously reported (Nuñez et al., 1990; Soltesz and Deschenes, 1993; Ylinen et al., 1995), the peak of the intracellular depolarization measured at the resting potential (0 nA holding current) lagged (76° ± 35°) behind the trough (negative polarity peak) of theta cycles (π) recorded in the CA1 pyramidal layer. Both the phase and amplitude of the intracellular theta depended on the holding potential of the neuron (Figure 2B). Depolarizing current injection increased the amplitude of the intracellular theta, whereas hyperpolarization decreased it, reaching a minimum at approximately −75 mV. With further hyperpolarization, the theta amplitude increased again, associated with the reversal of theta cycle phase (Figure 2B; Figures S1D–S1G; Soltesz and Deschenes, 1993; Ylinen et al., 1995). It has been proposed that these Vm-dependent changes arise from an interaction of phase-shifted rhythmic inhibitory synaptic inputs to the soma and excitatory inputs to the dendrites (Leung, 1984; Navas-Olive et al., 2020). To dissect these inhibitory and excitatory components, we first estimated the total synaptic conductance (G( t )) during theta cycles from the slope (I/V curve) of the linear regression of the estimated current flow (I = Vm divided by the measured input resistance, ΔVm( t )/Rm) and the holding potential (Vhold) (Valero et al., 2017). From this relationship, we derived the excitatory and inhibitory conductances (Gexc and Ginh) using their respective reversal potentials (−75 mV and 0 mV for inhibitory and excitatory reversal potentials, respectively; STAR Methods; Borg-Graham et al., 1998; Ylinen et al., 1995). On average, Ginh peak, measured in the soma, was larger and preceded the peak of Gexc (Figures 2C and S2A–S2C), corresponding to a total inhibition-to-excitation conductance ratio of 6.7 ± 1.4 (Figure S2D). The Gexc peak matched the reported firing phase of the CA3 pyramidal cells at the falling phase of theta (Mizuseki et al., 2009; Valero and de la Prida, 2018), while Ginh mirrored the preferred firing phase of perisomatic parvalbumin (PV) and cholecystokinin (CCK)-expressing basket cells and axo-axonic cells (Klausberger et al., 2005; Lapray et al., 2012; Valero and de la Prida, 2018; Viney et al., 2013). Because our estimation of Gexc and Ginh may fall short to integrate the contribution of poorly current-clamped distal dendrites (Borg-Graham et al., 1998) and to test the sufficiency of these experimentally obtained conductances, we reverse-engineered our results in a single neuron model. We built a single-compartment passive model, which included oscillatory inhibitory and excitatory inputs derived from the estimated conductances (Figure 2D). We then injected current (Ihold) to clamp the simulated neuron at different holding voltages as we did in our experiments (Figure 2E). The model replicated the same voltage-dependent changes of both theta amplitude and phase as observed in our intracellular recordings (Figures 2F and 2G). The theta phase reversal and the minimum Valero et al. Page 4 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript theta power were consistently present at Ihold near the simulated GABAA reversal −75 mV; Figures 2F and 2G). These experiments suggest that during theta oscillations synaptic inhibition dominates over excitation in the somata of pyramidal cells and that excitatory inputs lag behind somatic inhibition. This relationship is also supported by the reported phase preferences of the main excitatory and perisomatic inhibitory neuronal populations to the CA1 region (Klausberger et al., 2005; Mizuseki et al., 2009). Membrane voltage affects firing phase In our model, we investigated how competition between Gexc and Ginh controls spike timing in pyramidal neuron. Spike probability and timing depend on the combinations of somato-dendritic synaptic inputs, Vm trajectory, and the spiking history of the neuron. As previously reported, action potentials of the majority of CA1 pyramidal cells occurred at highest probability close the trough (π or 180°) of the theta cycle recorded in the pyramidal layer (Figures 3A and 3B; Buzsáki et al., 1983; Navas-Olive et al., 2020), although phase preference for the theta cycle peak (0°) has also been reported, depending on the brain state, input integration, and pyramidal cell subtype (Fernández-Ruiz et al., 2017; Mizuseki et al., 2011; Navas-Olive et al., 2020). Vm depolarization peaks correlated with the maximal spiking probability of the cells during theta oscillations (Figures 3A and 3B; Ylinen et al., 1995), where inhibitory conductance reached a minimum (Figures 3A and 3C). The holding potential (Vhold) also influenced the spiking phase preference. At more depolarized voltages, action potentials occurred at earlier theta phases, resulting in voltage-dependent phase precession (Figures 3A and 3B), in line with prior both in vitro (Magee, 2001) and in vivo (Kamondi et al., 1998) studies. This phase advance occurred because the spike threshold was reached at an earlier theta phase at more depolarized membrane potentials (see Vm line in Figure 3A). To reproduce the model-predicted correlation experimentally in a larger population of pyramidal neurons, we exploited a high-throughput optogenetic probing method to track for theta phase subthreshold dynamics in freely behaving CamKIIα-Cre:Ai32 mice (Valero et al., 2022). Mice (n = 4) were implanted with four-shank μLED probes with three μLEDs/shank (Valero et al., 2022; Wu et al., 2015) to induce 20-ms light pulses at random intervals (Valero et al., 2022). For these experiments, we included only neurons that showed significant spike responses to each of the three different light intensity levels (63 pyramidal cells). Optogenetic depolarization, similar to intracellular current injection, increased the firing rate of the neuron and advanced the phase of preferred spiking (Figures 3D and 3E). To evaluate the preferred phase of spiking at different Vhold in our single-compartment neuron model, firing probability was derived from the simulated membrane potential by a linear activation function (Figure S3 and STAR Methods). As expected, depolarization of the model neuron advanced the preferred phase of firing (Figure 3F), similar to the in vivo observations (compare with Figure 3A; Kamondi et al., 1998). In summary, our conductance-based model reliably mimicked the subthreshold and firing dynamics (Figures 2 and 3, respectively) observed in our intracellular and extracellular recordings from CA1 pyramidal cells during theta oscillations. Valero et al. Page 5 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Place field-tuned excitatory and inhibitory conductances Next, we sought to understand how variations in the excitatory (Davoudi and Foster, 2019; Zutshi et al., 2022) and inhibitory inputs (Geiller et al., 2022; Royer et al., 2012; Valero et al., 2022) of a pyramidal neuron could contribute to place-field-related activity. Spatial modulation (McClain et al., 2019; Tsodyks et al., 1996) was added to our model as a skewed Gaussian-like (Harvey et al., 2009) waveform of the theta-oscillating Ginh and Gexc (Figure 4A). We generated three classes (decreasing, constant, and increasing inhibition) of spatially tuned Gexc and Ginh curves by multiplying the basal theta dynamics (shown in Figures 2D and 3F, no holding current) with varying place modulated kernels (Figure S3D). Modulation of inhibition (ΔGinh) and excitation (ΔGexc) at the place field peak ranged from −1to +1-fold and from 0 to 5-fold, respectively. Whole-cell and extracellular recordings (Grienberger et al., 2017; Harvey et al., 2009; Lee et al., 2012; O’keefe and Recce, 1993) have identified three main signatures of place fields in CA1 pyramidal cells: (1) ~6 mV depolarization at the place field peak (ΔVmPF) associated with increased firing rate; (2) ~70% increase of Vm theta power from baseline (ΔPθPF), and (3) phase precession of the emitted spikes relative to the extracellularly recorded theta cycle (Table S1 and Figure 1C). To evaluate the relative contribution of ΔGexc and ΔGinh to these observed features, we quantified the three signatures (ΔVmPF, ΔPθPF, and phase precession slope) for all place field simulations (n = 101 ΔGexc steps, n = 41 ΔGinh steps, n = 4,141 simulations), and displayed the effects on the 2D parametric heatmap where the y and x axes show the values of ΔGexc and ΔGinh, so that each location of the heatmap represents a unique inhibitory-excitatory contribution to a particular feature of the place field (Figures 4B, 4C, and S3E). In the simulations, place field amplitude (ΔVPF) increased for both larger in-field upmodulation of the ΔGexc and for decreasing ΔGinh (Figures 4B–4D, left). Yet, both the intracellular theta power (Δθ power) and phase precession slope were more influenced by decreasing ΔGinh than increasing ΔGexc (Figures 4B and 4D, middle and right, Figure S3F). To ground these results, we compared the values obtained in our simulations with expected “target” values from our own and other’s previously published data (latticed areas in Figures 4B and 4C; Table S1 and Figures S3G and S3H). For all three signatures, relatively lower ΔGexc values matched the experimental values when combined with moderate disinhibition (i.e., decrease of ΔGinh; Geiller et al., 2022; Valero et al., 2022). Moreover, by overlapping the regions of the three signatures that fitted the experimental values, we found that a combination of excitation (ΔGexc ~ 2-fold) and reduced inhibition mimicked most closely the three place field signatures (Figure 4C, orange patches). This supports the hypothesis that the main features of CA1 place cell firing dynamics result from the competition of Gexc and Ginh. A similar combination of conductances was also a requirement to properly mimic the place field features in a biophysically more realistic neuron model (Navas-Olive et al., 2020), with multi-compartment CA1 pyramidal cell morphology, CA3 proximal dendritic excitatory and perisomatic PV, CCK, and axo-axonic GABAergic inputs (Figures 4E, 4F, and S4). Valero et al. Page 6 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Testing model predictions by pharmacogenetic perturbation of inhibitory neurons Our simulations predicted that appropriately coordinated excitation and disinhibition underlie place field properties (Royer et al., 2012). To test this hypothesis more directly, we experimentally manipulated inhibitory inputs on pyramidal neurons by combining pharmacogenetic and optogenetic methods. We virally expressed an “excitatory designer receptor exclusively activated by a designer drug” (Gq-DREADD) in all types of CA1 GABAergic interneurons by infusing Dlx5/6-Gq-DREADD adeno-associated virus (rAAVhDLX) in the dorsal CA1 region of CamKIIα-Cre:Ai32 freely moving mice (n = 3; Figure 5A; Rogers et al., 2021). Histological evaluation showed that virus infection successfully targeted CA1 interneurons in all sublayers (Figure 5B) (Dimidschstein et al., 2016). As expected, clozapine N-oxide (CNO) activation of interneurons led to a strong suppression of pyramidal neuron spiking (68.35% with respect to DMSO; Figures 5C and S5A). Paradoxically, the firing rate of many interneurons also decreased (66.07% with respect to DMSO; Figures 5C and S5A). Such “paradoxical effects” have been described in the hippocampus (Rogers et al., 2021) and different cortical areas (Mahrach et al., 2020; Tsodyks et al., 1997) and were explained by the mutual inhibition of interneurons and the reduced excitatory inputs from the suppressed pyramidal cells. In addition to suppressing spontaneously occurring spikes in pyramidal neurons, we also found that their excitability in response to optogenetic stimulation decreased several folds after CNO injection (but not after DMSO vehicle injection; Figures 5D and S5B). Importantly, this difference survived after normalizing for rate change (Figure S5B), indicating an increase in the local inhibitory transmission. To examine the role of inhibition in place field signatures of pyramidal neurons, we exploited the position-dependent firing rate changes of interneurons. While interneurons rarely show bona fide single place fields, they do have characteristic tuning curves along the animal’s travel on the track that remain consistent across trials (Figure 5E; Souza et al., 2018; Valero et al., 2022; Wilson and McNaughton, 1993). Mice ran on a linear track 100 min after the DMSO or CNO injection. The place-specific tuning patterns in interneurons were disrupted after Dlx-Gq DREADDs activation by CNO, as reflected by the decreased correlation of position-specific firing rates between control and CNO conditions compared with the effect of vehicle DMSO (Figures 5E and 5F). The altered spatially tuned features of interneurons after CNO were also demonstrated by the standard deviation of the firing maps (Figure 5G) and the mutual information between position and spiking activity (Figures 5H and S6). These changes could be dissociated from effects of the firing rates alone since even after downsampling of spikes in the DMSO condition the differences with CNO remained significant (see DMSO “resampled” in Figures 5G and 5H). Importantly, the altered spatial tuning of the inhibitory inputs (ΔGinh) caused by Dlx-Gq DREADDs allowed for testing our model predictions on the three place field signatures discussed above: place field amplitude (ΔVmPF), in-field theta power change (ΔPθPF), and phase precession slope (Figure 5I). To examine the model predictions on the pyramidal cell population, we quantified the aforementioned place field properties. Place fields on the track for all place cells were sorted Valero et al. Page 7 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript by the peak rate position before CNO (or DMSO) injection (Figures 6A and S5C). In line with the observed decrease of spatial modulation of interneurons, cofiring of pyramidal cells and interneurons was dampened around place fields during CNO (Figures S5D and S5E). As expected, both the standard deviation of the firing maps (Figure 6B), the in-field firing rates of pyramidal neurons (Figure 6C), and the difference between in-field and out-field firing (Figures S5F) decreased after CNO compared with DMSO injection, compatible with reduced place field amplitude (ΔVmPF), as predicted by the model. Mutual information between position and firing (Figures 6D and S6) and place field stability (Figure S5G) were also reduced, even beyond what was expected from the changes in the firing rate alone. We also observed a decrease in place field size, which could be explained by the overall decreased firing rate (Figure S5H). As an indirect measure of the intracellular in-field theta power change (ΔPθPF), we estimated the theta phase-locking of spikes (“mean vector length,” θ MVL) both in field and out field, and computed the in-field/out-field θ MVL log-ratio for all place fields. As expected from the larger intracellular Vm theta power in-field versus out-field (Grienberger et al., 2017; Harvey et al., 2009; Lee et al., 2012), theta phase locking of spikes (MVL) was stronger in field than out field in the intact animal resulting in ratios >1 (Figure 6E for DMSO). This in-field gain of MVL was reduced by CNO injection (Figure 6E), in line with our model prediction. Finally, we found that the phase precession slope became less negative during CNO session (Figures 6F and 6G), a change that was independent of the concomitant firing rate reduction (Figures 6G and S5I, resampled). Consequently, mutual information between spatial position and theta phase of spiking, a measure of phase coding efficacy, was also decreased by CNO (Figure 6H). In summary, in harmony with our model predictions and prior studies (Geiller et al., 2022; Royer et al., 2012; Valero et al., 2022), impairment of the spatial tuning of interneurons brought about by their pharmacogenetic perturbation resulted in reduced in-field firing, less in-field theta modulation of spiking (but not the overall θ MVL), and reduced slope of phase precession. Population level effects of interneuron perturbation Theta phase coordination of spiking activity has been implicated in organizing cell assemblies as passing through consecutive place fields (Dragoi and Buzsáki, 2006; Harris et al., 2003; Petersen and Buzsáki, 2020; Skaggs et al., 1996). The critical role of local GABAergic interneuron activity in shaping phase precession dynamics suggests assembly expression should be also affected. To test this prediction, we quantified the expression of CA1 neuronal assemblies (Lopes-dos-Santos et al., 2013) in 25-ms bins (Harris et al., 2003; Zutshi et al., 2022) before and during pharmacogenetic manipulation of interneurons (Figure 6I). As expected from the multiple impaired parameters of individual pyramidal neurons, CNO (but not DMSO) dramatically decreased the assembly expression (Figures 6I, 6J, S6C, and S6D), independent of the reduced firing rates of pyramidal cells (Figure 6J; resampled). Valero et al. Page 8 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript DISCUSSION Using intracellular recordings in vivo and fast-pulse optogenetic probing of hippocampal CA1 pyramidal cells, we found that excitation lags behind inhibition by approximately one-fifth of a theta cycle, and that Ginh dominates over Gexc across the entire cycle. In control conditions, the firing of intracellularly recorded cells occurred asymmetrically around the extracellular theta negativity with more spikes occurring on the ascending than on the descending phase, enabled by the relative level of Gexc over perisomatic Ginh. Tonic depolarization advanced the phase preference of spikes, producing a voltagephase precession that spanned part of the theta cycle. These results were reproduced by both a single-compartment passive model endowed with the experimentally observed conductances and by a multi-compartment Hodgkin-Huxley type model. By adding a spatially modulated depolarizing signal coupled with decreased inhibition to the model, we found that the canonical place field features (a slow depolarization ramp, increased amplitude theta rhythmicity, and spike phase precession) emerged from the temporally coordinated increase of excitation and disinhibition. Consistently, DREADDs perturbation of GABAergic interneuron spiking interfered with the key place field features and changed coordinated assembly expression. Overall, our results demonstrate the necessity of both fast and slow temporal coordination of Ginh and Gexc for maintaining local circuit dynamics in the CA1 region. Relationship between Ginh and Gexc affects theta oscillatory dynamics While previous intracellular experiments have already provided insights on the role of perisomatic inhibition in theta oscillations (Grienberger et al., 2017; Harvey et al., 2009; Kamondi et al., 1998; Nuñez et al., 1990; Soltesz and Deschenes, 1993; Ylinen et al., 1995), our study was designed to clarify both the importance and the precise temporal/phase relationship between Ginh and Gexc. Within the theta cycle, Ginh peak was several times larger compared with Gexc, and Ginh preceded the peak of Gexc (Figure 2C). Both Gexc and Ginh peaked on the descending phase of the LFP theta waves, corresponding to the largest amplitude of slow gamma (30–80 Hz) oscillations (Fernández-Ruiz et al., 2017). The maximal firing from both upstream CA3 pyramidal cells (Mizuseki et al., 2009) and feed-forwardly activated perisomatic inhibitory neurons (i.e., CCK and PV basket cells and bistratified interneurons) also occur at this phase (Klausberger et al., 2005; Lapray et al., 2012; Valero and de la Prida, 2018; Viney et al., 2013). These circuit features may explain the asymmetry of theta waves (Buzsáki et al., 1985), as well as the uneven distribution of firing along the descending and ascending phases of theta (Belluscio et al., 2012; NavasOlive et al., 2020; Skaggs et al., 1996). Targeted manipulation of the somatic Vm in our intracellular experiments allowed us to examine how changes of Gexc and Ginh affected membrane polarity and firing preference relative to the extracellular LFP theta. Varying the holding potential from subthreshold to suprathreshold Vm levels led to a phase reversal of intracellular theta waves at about −75 mV, which corresponds to the Cl− equilibrium potential of GABAa receptors (Borg-Graham et al., 1998; Ylinen et al., 1995). Another consequence of the depolarizing Vm was that spike threshold is reached at progressively earlier phases of the theta cycle (Kamondi et Valero et al. Page 9 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript A normalized version FRmem norm Vmem which ranged from 0 to 1 was also computed. In parallel, we added firing dynamics with accommodation by first finding spiking times, or phases ( θspike ), whenever a rising Vmem crossed a threshold ( Vthr = − 65 mV). We shifted a double exponential function defined as: FRaccomm θ,θspike =exp −θ−θspike φdecay −exp −θ−θspike φraise where θ are theta phases, φraise = 15 deg, and φdecay = 45 deg, to center its maximum over spikes. We normalized it by dividing by its maximum, and leaving a constant 10% activation, and 90% of dynamics: FRaccomm norm θ,θspike = 0.1 + 0.9 ⋅ FRaccomm θ,θspike max FRaccomm θ,θspike We then applied accommodation to firing rate distribution by combining both approaches: FR = FRmem norm Vmem(θ) ⋅ FRaccomm norm θ,θspike Biophysically realistic model To expand and further validate our simulations with the simple conceptual model, we also used a more biophysically realistic pyramidal cell model (Navas-Olive et al., 2020) that integrates theta-modulated glutamatergic and GABAergic inputs, distributed along its somadendritic axis according as described in published experiments. To this purpose, we modified our former model at Github (https://github.com/PridaLab/LCN-HippoModel/tree/ place-field). To simulate place fields, we modified the code used in the simple model by adding a place field asymmetric modulation to CA3, axo-axonic, PV-bc and CCK-bc synaptic conductances. Assuming a constant running speed of 30 cm/sec, and a linear track length of 60 cm, the place field started at 12 cm, the center was at 37.92 cm (at 72% of place field width, Harvey et al., 2009), and ended at 48 cm. As in the simple uni-compartmental model, theta modulation was multiplied by 1 outside the place field, and by the skewed place field modulation function (Figure S4A), whose ΔGinh ranged from −1 to 1, and ΔGexc from 0 to 3. However, simulations with ΔGinh > 0.3 were not considered because firing rate was too low to match the criteria to compute phase precession. These simulations were done using the synthetic cell with morphology n409 from the NeuroMorpho Turner archive, intrinsic individual 12 and synaptic individual 1, a generic synthetic cell that was not tuned to incorporate superficial (more CCK projections) or deep (more PV) microcircuits. Extracellular recordings and behavior Mice were initially anesthetized with 2% isoflurane and maintained under anesthesia with 0.75–1% isoflurane. 400 nL of pAAV-hDlx-GqDREADD-dTomato-Fishell-5 (Addgene plasmid #83897) was infused into the hippocampus (antero-posterior 2.0 mm, mediolateral 1.5 mm, dorsoventral 1.2 mm) at a rate of 25 nL/minute using a sharp glass pipette (15– 20 nm in diameter), which was left in place for 15 minutes to minimize the backflow Valero et al. Page 16 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript of virus. Following virus infusion, and in the same surgical procedure animals were implanted with 32-site, 4 shank μLED probes (Neurolight, Plexon) over the dorsal right hippocampus (antero-posterior 2.0 mm, mediolateral 1.5 mm, dorsoventral 0.6 mm), as described previously (Valero et al., 2021, 2022; Zutshi et al., 2022). Ground and reference wires were implanted in the skull above the cerebellum, and a grounded copper mesh hat was constructed to shield the probes. Animals were allowed to recover for at least two weeks. Probes were mounted on microdrives that were advanced to CA1 pyramidal layer in small increments over 5 to 8 days, while depth distribution of LFPs (SPW-R events and theta oscillations) and unit firing were used to identify CA1 pyramidal layer. After implantation, animals were handled daily and accommodated to the experimenter, recording room and cables for 1 week before the start of the experiments. Mice were trained to run laps in a PVC linear track (110 cm long, 6.35 cm wide) to retrieve water reward (5–10 μL) at each end. Water access was restricted and was only available as reward on a linear track, ad libitum for 30 minutes at the end of each experimental day and ad libitum for one full non-experimental day per week. The animal’s position was monitored with a Basler camera (acA1300–60 gmNIR, Graftek Imaging) sampling at 30 Hz to detect a head-mounted red LEDs. Position was synchronized with neural data with TTLs signaling shutter position. Water delivery and optogenetic stimuli during track were controlled by a custom-made, Arduino-based circuit (circuits and software are available in https://github.com/valegarman/HippoPlayground). Electrophysiological data were acquired using an Intan RHD2000 system (Intan Technologies LLC) digitized with 30 kHz rate. The wide-band signal was downsampled to 1.25 kHz and used as the LFP signal. For a typical recording session, mice were recorded continuously for ~500 min through 3 experimental blocks before and after chemogenetic manipulation: pre-track baseline (~100 min), Linear Maze task (~30 min), post-track baseline (~100 min), followed by a intraperitoneal injection of Clozapine N-oxide (CNO 5 mg/kg CNO in 1% DMSO) or Dimethylsulfoxide (1% DMSO in saline) injection and one more set of pre track-baseline, maze and post-track baseline recordings. CNO and DMSO sessions were interleaved, and we did not include in our analysis any recording sessions beyond the fourth day of CNO injection. μLEDs stimulation were conducted as described in (Valero et al., 2022). Briefly, μLEDs were controlled with current (2–4.5 μA generating 0.02–0.1 μW of total light power; Valero et al., 2022) provided by a 12-channel current generator (OSC1Lite, NeuroNex Michigan Hub) driven by an Arduino (https://github.com/valegarman/HippoPlayground), which delivered trapezoid (1 ms rise time) blue light (centered emission at 460 nm, emission surface area = 150 mm2) 20 ms pulses at random sites with a randomly variable (40–60 ms) offset. Stimulation protocol was delivered for ~40 min during the pre-track baseline and post-track baseline epochs. Unit clustering and neuron classification Spike sorting was performed semi-automatically as previously described (Valero et al., 2021, 2022). Briefly, we employed our own pipeline KilosortWrapper (https:// github.com/brendonw1/KilosortWrapper), a wrapper for KiloSort (https://github.com/cortexlab/KiloSort). This was followed by manual adjustment of the waveform clusters using Valero et al. Page 17 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript the software Phy (https://github.com/kwikteam/phy) and plugins for Phy designed in our laboratory (https://github.com/petersenpeter/phy-plugins). Following parameters were used for the Kilosortclustering: ops.Nfilt: 6 * numberChannels; ops.nt0: 64; ops.whitening: ‘full’; ops.nSkipCov: 1; ops.whiteningRange: 64; ops.criterionNoise-Channels: 0.00001; ops.Nrank: 3; ops.nfullpasses: 6; ops.maxFR: 20000; ops.fshigh: 300; ops.ntbuff: 64; ops.scaleproc: 200; ops.Th: [4 10 10]; ops.lam: [5 20 20]; ops.nannealpasses: 4; ops.momentum: 1./[20 800]; ops.shuffle_clusters: 1. Unit clustering generated two clearly separable group (Petersen et al., 2021; Valero et al., 2021, 2022) based on their spike autocorrelograms, waveform characteristics and firing rate. Putative pyramidal cells and interneurons were tentatively separated based in these two clusters. A more reliable cell identity was assigned after inspection of all features, assisted by monosynaptic excitatory and inhibitory interactions between simultaneously recorded, well-isolated units and light responses by using the suite CellExplorer (Petersen et al., 2021; https://cellexplorer.org/). Units were defined as optogenetically responsive cells (Valero et al., 2022) based on combination of three criteria: (i) an average firing response higher than 2 SD, (ii) significant modulation using a p value cutoff of 10−3, and (iii) against randomly shuffled pulse times (500 replicates) and testing for significant difference between the observed value and the random distribution. Place field and extracellular Spike-LFP coupling analysis Firing rate distribution within place fields (‘rate map’) was generated as in (Valero et al., 2022). In summary, spiking data was binarized into 2.2 cm wide bins, and spike counts were normalized by time occupancy (smoothing size: 2 bins) for epochs when the animal’s speed was >1 cm/s. Trials for forward and backward directions were evaluated separately. Place fields were defined from these rate maps as previously described (Fernández-Ruiz et al., 2017; Mizuseki et al., 2011; Valero et al., 2022). The field boundaries were defined when the rate decreased below 20% of the peak firing rate. Place fields cleared all criteria when they were between 8.75 cm and 75 cm wide, had a minimum peak firing rate of 3 Hz and a spatial coherence >0.7. Place field stability was defined by correlating rate/bin distributions (Spearman correlation) for the first and second maze exploration. Standard deviation was computed across all spatial bins for both run directions (100 bins in total). Spatial coherence per cell was estimated as the mean correlation between the firing rate in each spatial bin and the corresponding rates averaged over the ±4 nearest-neighbor bins (Muller and Kubie, 1989). We also calculated the spatial information content in bits per spike (Souza et al., 2018) as following: SPI=∑ k PK∣xilog2 PK∣xi PK where PK is the probability of observing a rate K and PK∣xi is the conditional probability of observing a rate K in position xi . The mutual information between the position and firing rate (MIrate) (Souza et al., 2018) was estimated as following: Valero et al. Page 18 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript MIRate =∑ i∑ j pi,jlog2 pi,j pi⋅pj where pi and pj are the probabilities of the ith position bin and the jth firing rate bin, respectively, and pi,j is the joint probability between position ith and firing rate jth . The theta-band phase of the LFP recorded at the highest theta power channel above CA1 stratum pyramidale was estimated as the Hilbert transform of the narrowband filtered LFP (5–11 Hz). Theta epochs were detected automatically using the ratio of the power in theta band (5–11 Hz) to the power of nearby bands (1–4 Hz, 12–14 Hz) of CA1 LFP (FernándezRuiz et al., 2017). Theta modulation indices for each neuron were estimated as described for intracellularly recorded cells. For phase precession analysis, the instantaneous theta phases (from the filtered CA1 pyramidal layer LFP) of spikes were plotted against the linearized positions in the track for each place field (Fernández-Ruiz et al., 2017). Circular-linear regression between position and theta phase was applied to calculate the phase-precession slope and correlation strength, where a correlation coefficient, similar to the Pearson’s correlation, was obtained. For further analyses, only fields displaying significant phaseposition correlation (p < 0.05) were considered. Mutual information between the position and firing phase was estimated as following: MI(Pℎase,Position) = H(Pℎase) + H(Position)−H(Pℎase,Position) where entropies H(position) and H(phase) are defined as: H(x) = − ∑ i pxilogp xi and joint entropy H(x,y) is: H(x,y) = − ∑ i∑ j pxi,yjlog2p xi,yj Mutual information function from MATLAB’s Information Theory Toolbox was used (Chen, 2022). Subsampling for rate maps were performed to replicate pyramidal spiking suppression after CNO with respect to DMSO injection (68.35%). All of the subsampled analyses were performed from scratch following the same code as with the original data but adding a step in which 30% of spikes were randomly discarded in a uniform way along all the session. The analysis was performed using the remaining 70% of spikes. The only analysis that was not performed from scratch was the slope computation: slopes were computed using the subset of the subsampled spikes that fell within place fields computed with the original data. Slopes that were not significant were discarded. Valero et al. Page 19 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Assembly analysis Cell assemblies were defined using an unsupervised statistical framework based on a hybrid PCA followed by ICA (Lopes-dos-Santos et al., 2013) as previously described (Valero et al., 2022; Zutshi et al., 2022). Spike trains of each neuron were binned in 25-ms intervals and zscored firing rates were calculated for each bin. Spike trains were convolved with a Gaussian kernel (standard deviation = 10 ms), and the matrix of firing correlation coefficients for all pairs of neurons was constructed. Next, using the fast-ICA algorithm, we determined the vector of weights (contribution of each neuron) for each assembly(component). Most of the detected assembly patterns consisted of a few neurons with high weights and a large group of neurons with weights around zero (Figure 6I). The fraction of assemblies increasing or decreasing in expression was defined by calculating a threshold of mean ±1 standard deviation during baseline sessions and measuring the fraction of assemblies whose expression changed greater than, or less than this threshold after DMSO/CNO injection or downsampling. Histological processing and microscopy Mice or rats were overdosed with pentobarbital injection (100 mg/kg body weight), perfused with saline and 4% paraformaldehyde before their brains were rapidly removed. After overnight post-fixation in 4% paraformaldehyde solution, the brain was washed 3 times in 0.1 M phosphate buffered saline. Coronal sections of 50–100 μm thickness were cut using a Leica Microsystems VT 1000S vibratome, wash 3–4 times in 0.1 M phosphate buffer (PB) and stored in 0.1 M PB with 0.05% sodium azide at 4°C. Sections were mounted on glass slides in VectaShield (Vector Laboratories) and observed with epifluorescence (Leitz DMRB microscope, Leica) or confocal imaging (Zeiss). For the identification of native EYFP + or streptavidin + neurons and layer markers we proceeded as previously described (35, 45). Sections containing Neurobiotin-labeled cells were localized by incubating them in 1:400 Alexa Fluor488–conjugated streptavidin (Jackson ImmunoResearch 016–540-084) with 0.5% Triton X-100 (vol/vol) in PBS (PBS-Tx) for 2 h at 22–25°C. QUANTIFICATION AND STATISTICAL ANALYSIS Statistical analysis All statistical analyses were performed with standard and custom-made MATLAB functions (https://github.com/valegarman/HippoCookBook). No specific analysis was used to estimate minimal population sample, but the number of animals, trials, and recorded cells were larger or similar to those employed in previous works (Valero et al., 2017, 2021, 2022). All data presented here were obtained from experimental replicates with at least three independent experimental repeats for each assay. All attempts of replication were successful. Data collection was not performed blinded to the subject conditions. Data analysis was performed blinded to the scorer or did not require manual scoring. All subjects underwent the same number of conditions (unless stated otherwise) in a randomly assigned fashion. Unless otherwise noted, for all tests, non-parametric two-tailed Wilcoxon’s paired signedrank test and Kruskal-Wallis one-way analysis of variance. When parametric two-ways ANOVA test were used, the data satisfied the criteria for normality (Kolmogorov–Smirnov test) and equality of variance (Bartlett’s test for equal variance). For multiple comparisons, Valero et al. Page 20 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Tukey’s honesty post hoc test was employed and the corrected *p < 0.05, **p < 0.01, ***p < 0.001 are indicated. p-values for Spearman’s correlations are computed using a Student’s t distribution for a transformation of the correlation. Results are displayed as boxplots representing median, 25th/75th percentiles and data range. Dispersion represents ±CI95 around mean. The exact number of replications for each experiment is detailed in the text and figures. Supplementary Material Refer to Web version on PubMed Central for supplementary material. ACKNOWLEDGMENTS We would like to thank T. Hainmueller, N. Nitzan, A. Fernandez-Ruiz, I. Zutshi, E. Cid, and the rest of the members of the Buzsáki and de la Prida laboratories for helpful comments on the project. This work was supported by the European Molecular Biology Organization (EMBO) postdoctoral fellowship (EMBO ALTF 1161– 2017) and Human Frontiers Science Program (HFSP) postdoctoral fellowship (LT0000717/2018) to M.V., NIH grants (R01MH122391, U19NS104590, and U19NS107616) to G.B., and a grant from the Spanish Ministry of Science and Innovation MCIN/AEI/10.13039/501100011033 “ERDF A way of making Europe” (RTI2018–098581B-I00) to L.M.P. A.N.O. is supported by a PhD fellowship (FPU17/03268) and by a short-term visit fellowship (EST19/00828) from the Spanish Ministry of Education. REFERENCES Ahmed OJ, and Mehta MR (2009). The hippocampal rate code: anatomy, physiology and theory. Trends Neurosci. 32, 329–338. 10.1016/J.TINS.2009.01.009. [PubMed: 19406485] Andersen P, Morris R, Amaral D, Bliss T, and O’ Keefe J (2009). The Hippocampus Book (Oxford University Press). Belluscio MA, Mizuseki K, Schmidt R, Kempter R, and Buzsáki G (2012). Cross-frequency phasephase coupling between θ and γ oscillations in the hippocampus. J. Neurosci. 32, 423–435. 10.1523/JNEUROSCI.4122-11.2012. [PubMed: 22238079] Bezaire MJ, Raikov I, Burk K, Vyas D, and Soltesz I (2016). Interneuronal mechanisms of hippocampal theta oscillations in a full-scale model of the rodent CA1 circuit. Elife 5, e18566. 10.7554/ELIFE.18566. [PubMed: 28009257] Bhatia A, Moza S, and Bhalla US (2019). Precise excitation-inhibition balance controls gain and timing in the hippocampus. Elife 8, e43415. 10.7554/ELIFE.43415. [PubMed: 31021319] Borg-Graham LJ, Monier C, and Frégnac Y (1998). Visual input evokes transient and strong shunting inhibition in visual cortical neurons. Nature 393, 369–373. 10.1038/30735. [PubMed: 9620800] Brun VH, Otnass MK, Molden S, Steffenach H-A, Witter MP, Moser M-B, and Moser EI (2002). Place cells and place recognition maintained by direct entorhinal-hippocampal circuitry. Science 296, 2243–2246. 10.1126/science.1071089. [PubMed: 12077421] Burgess N, Recce M, and O’Keefe J (1994). A model of hippocampal function. Neural Network. 7, 1065–1081. 10.1016/S0893-6080(05)80159-5. Buzsáki G, Leung LW, and Vanderwolf CH (1983). Cellular bases of hippocampal EEG in the behaving rat. Brain Res. 287, 139–171. [PubMed: 6357356] Buzsáki G, Rappelsberger P, and Kellényi L (1985). Depth profiles of hippocampal rhythmic slow activity (‘theta rhythm’) depend on behaviour. Electroencephalogr. Clin. Neurophysiol. 61, 77–88. 10.1016/0013-4694(85)91075-2. [PubMed: 2408867] Carrillo-Reid L, Lopez-Huerta VG, Garcia-Munoz M, Theiss S, and Arbuthnott GW (2015). Cell assembly signatures defined by short-term synaptic plasticity in cortical networks. Int. J. Neural Syst. 25, 1550026. 10.1142/S0129065715500264. [PubMed: 26173906] Chen M (2022). Information Theory Toolbox. Valero et al. Page 21 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Davoudi H, and Foster DJ (2019). Acute silencing of hippocampal CA3 reveals a dominant role in place field responses. Nat. Neurosci. 22, 337–342. 10.1038/s41593-018-0321-z. [PubMed: 30664772] Deuchars J, and Thomson AM (1996). CA1 pyramid-pyramid connections in rat hippocampus in vitro: dual intracellular recordings with biocytin filling. Neuroscience 74, 1009–1018. 10.1016/0306-4522(96)00251-5. [PubMed: 8895869] Dimidschstein J, Chen Q, Tremblay R, Rogers SL, Saldi GA, Guo L, Xu Q, Liu R, Lu C, Chu J, et al. (2016). A viral strategy for targeting and manipulating interneurons across vertebrate species. Nat. Neurosci. 19, 1743–1749. 10.1038/NN.4430. [PubMed: 27798629] Dragoi G, and Buzsáki G (2006). Temporal encoding of place sequences by hippocampal cell assemblies. Neuron 50, 145–157. 10.1016/j.neuron.2006.02.023. [PubMed: 16600862] Dragoi G, and Tonegawa S (2010). Preplay of future place cell sequences by hippocampal cellular assemblies. Nature 469, 397–401. 10.1038/nature09633. [PubMed: 21179088] Eichenbaum H (2014). Time cells in the hippocampus: a new dimension for mapping memories. Nat. Rev. Neurosci. 15, 732–744. 10.1038/nrn3827. [PubMed: 25269553] English DF, McKenzie S, Evans T, Kim K, Yoon E, and Buzsáki G (2017). Pyramidal cell-interneuron circuit architecture and dynamics in hippocampal networks. Neuron 96, 505–520.e7. 10.1016/ j.neuron.2017.09.033. [PubMed: 29024669] Fernández-Ruiz A, Oliva A, Nagy GA, Maurer AP, Berényi A, Buzsáki G, Buzsáki G, Grosmark A, Mao D, Mizuseki K, et al. (2017). Entorhinal-CA3 dual-input control of spike timing in the Hippocampus by thetagamma coupling. Neuron 93, 1213–1226.e5. 10.1016/j.neuron.2017.02.017. [PubMed: 28279355] Fernández-Ruiz A, Oliva A, Soula M, Rocha-Almeida F, Nagy GA, Martin-Vazquez G, and Buzsáki G (2021). Gamma rhythm communication between entorhinal cortex and dentate gyrus neuronal assemblies. Science 1979, eabf3119. 10.1126/SCIENCE.ABF3119. [PubMed: 33795429] Freund TF (2003). Interneuron Diversity series: rhythm and mood in perisomatic inhibition. Trends Neurosci. 26, 489–495. 10.1016/S0166-2236(03)00227-3. [PubMed: 12948660] Geiller T, Sadeh S, Rolotti S.v., Blockus H, Vancura B, Negrean A, Murray AJ, Rózsa B, Polleux F, Clopath C, and Losonczy A (2022). Local circuit amplification of spatial selectivity in the hippocampus. Nature 601, 105–109. 10.1038/S41586-021-04169-9. [PubMed: 34853473] Geisler C, Robbe D, Zugaro M, Sirota A, and Buzsáki G (2007). Hippocampal place cell assemblies are speed-controlled oscillators. Proc. Natl. Acad. Sci. USA 104, 8149–8154. 10.1073/ PNAS.0610121104. [PubMed: 17470808] Grienberger C, Milstein AD, Bittner KC, Romani S, and Magee JC (2017). Inhibitory suppression of heterogeneously tuned excitation enhances spatial coding in CA1 place cells. Nat. Neurosci. 20, 417–426. 10.1038/nn.4486. [PubMed: 28114296] Grosmark AD, and Buzsáki G (2016). Diversity in neural firing dynamics supports both rigid and learned hippocampal sequences. Science 351, 1440–1443. 10.1126/SCIENCE.AAD1935. [PubMed: 27013730] Hafting T, Fyhn M, Molden S, Moser M-B, and Moser EI (2005). Microstructure of a spatial map in the entorhinal cortex. Nature 436, 801–806. 10.1038/nature03721. [PubMed: 15965463] Harris KD, Csicsvari J, Hirase H, Dragoi G, and Buzsáki G (2003). Organization of cell assemblies in the hippocampus. Nature 424, 552–556. 10.1038/nature01834. [PubMed: 12891358] Harvey CD, Collman F, Dombeck DA, and Tank DW (2009). Intracellular dynamics of hippocampal place cells during virtual navigation. Nature 461, 941–946. 10.1038/nature08499. [PubMed: 19829374] Hasselmo ME, Bodelón C, and Wyble BP (2002). A proposed function for hippocampal theta rhythm: separate phases of encoding and retrieval enhance reversal of prior learning. Neural Comput. 14, 793–817. 10.1162/089976602317318965. [PubMed: 11936962] Itskov V, Curto C, Pastalkova E, and Buzsáki G (2011). Cell assembly sequences arising from spike threshold adaptation keep track of time in the Hippocampus. J. Neurosci. 31, 2828–2834. 10.1523/ JNEUROSCI.3773-10.2011. [PubMed: 21414904] Valero et al. Page 22 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Jensen O, and Lisman JE (2000). Position reconstruction from an ensemble of hippocampal place cells: contribution of theta phase coding. J. Neurophysiol. 83, 2602–2609. 10.1152/ JN.2000.83.5.2602. [PubMed: 10805660] Kamondi A, Acsády L, Wang XJ, and Buzsáki G (1998). Theta oscillations in somata and dendrites of hippocampal pyramidal cells in vivo: activity-dependent phase-precession of action potentials. Hippocampus 8, 244–261. 10.1002/(SICI)1098-1063. [PubMed: 9662139] Kanter BR, Lykken CM, Avesar D, Weible A, Dickinson J, Dunn B, Borgesius NZ, Roudi Y, and Kentros CG (2017). A novel mechanism for the grid-to-place cell transformation revealed by transgenic depolarization of medial entorhinal cortex layer II. Neuron 93, 1480–1492.e6. 10.1016/ J.NEURON.2017.03.001. [PubMed: 28334610] Klausberger T, Marton LF, O’Neill J, Huck JHJ, Dalezios Y, Fuentealba P, Suen WY, Papp E, Kaneko T, Watanabe M, et al. (2005). Complementary roles of cholecystokininand parvalbuminexpressing GABAergic neurons in hippocampal network oscillations. J. Neurosci. 25, 9782–9793. 10.1523/JNEUROSCI.3269-05.2005. [PubMed: 16237182] Kullmann DM, Moreau AW, Bakiri Y, and Nicholson E (2012). Plasticity of inhibition. Neuron 75, 951–962. 10.1016/J.NEURON.2012.07.030. [PubMed: 22998865] Lapray D, Lasztoczi B, Lagler M, Viney TJ, Katona L, Valenti O, Hartwich K, Borhegyi Z, Somogyi P, and Klausberger T (2012). Behavior-dependent specialization of identified hippocampal interneurons. Nat. Neurosci. 15, 1265–1271. 10.1038/nn.3176. [PubMed: 22864613] Latuske P, Kornienko O, Kohler L, and Allen K (2018). Hippocampal remapping and its entorhinal origin. Front. Behav. Neurosci. 11, 253. 10.3389/FNBEH.2017.00253/BIBTEX. [PubMed: 29354038] Lee D, Lin B-J, and Lee AK (2012). Hippocampal place fields emerge upon single-cell manipulation of excitability during behavior. Science 337, 849–853. 10.1126/science.1221489. [PubMed: 22904011] Lai-Wo SL (1984). Pharmacology of theta phase shift in the hippocampal CA1 region of freely moving rats. Electroencephalogr. Clin. Neurophysiol. 58, 457–466. 10.1016/0013-4694(84)90142-1. Lisman J (2005). The theta/gamma discrete phase code occuring during the hippocampal phase precession may be a more general brain coding scheme. Hippocampus 15, 913–922. 10.1002/ HIPO.20121. [PubMed: 16161035] Lopes-dos-Santos V, Ribeiro S, and Tort ABL (2013). Detecting cell assemblies in large neuronal populations. J. Neurosci. Methods 220, 149–166. 10.1016/J.JNEUMETH.2013.04.010. [PubMed: 23639919] Losonczy A, Zemelman B.v., Vaziri A, and Magee JC (2010). Network mechanisms of theta related neuronal activity in hippocampal CA1 pyramidal neurons. Nat. Neurosci. 13, 967–972. 10.1038/ NN.2597. [PubMed: 20639875] Luo X, Guet-Mccreight A, Villette V, Francavilla R, Marino B, Chamberland S, Skinner FK, and Topolnik L (2020). Synaptic mechanisms underlying the network state-dependent recruitment of VIP-expressing interneurons in the CA1 Hippocampus. Cereb. Cortex 30, 3667–3685. 10.1093/ CERCOR/BHZ334. [PubMed: 32080739] Magee JC (2001). Dendritic mechanisms of phase precession in hippocampal CA1 pyramidal neurons. J. Neurophysiol. 86, 528–532. 10.1152/JN.2001.86.1.528. [PubMed: 11431530] Magee JC, and Grienberger C (2020). Synaptic plasticity forms and functions. Annu. Rev. Neurosci. 43, 95–117. 10.1146/ANNUREV-NEURO-090919-022842. [PubMed: 32075520] Mahrach A, Chen G, Li N, van Vreeswijk C, and Hansel D (2020). Mechanisms underlying the response of mouse cortical networks to optogenetic manipulation. Elife 9, e49967. 10.7554/ ELIFE.49967. [PubMed: 31951197] Marshall L, Henze DA, Hirase H, Leinekugel X, Dragoi G, and Buzsáki G (2002). Hippocampal pyramidal cell–interneuron spike transmission is frequency dependent and responsible for place modulation of interneuron discharge. J. Neurosci. 22, RC197. 10.1523/JNEUROSCI.22-02J0001.2002. [PubMed: 11784809] Maurer AP, and McNaughton BL (2007). Network and intrinsic cellular mechanisms underlying theta phase precession of hippocampal neurons. Trends Neurosci. 30, 325–333. 10.1016/ J.TINS.2007.05.002. [PubMed: 17532482] Valero et al. Page 23 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Maurer AP, Cowen SL, Burke SN, Barnes CA, and McNaughton BL (2006). Phase precession in hippocampal interneurons showing strong functional coupling to individual pyramidal cells. J. Neurosci. 26, 13485–13492. 10.1523/JNEUROSCI.2882-06.2006. [PubMed: 17192431] McClain K, Tingley D, Heeger DJ, and Buzsáki G (2019). Position-theta-phase model of hippocampal place cell activity applied to quantification of running speed modulation of firing rate. Proc. Natl. Acad. Sci. USA 116, 27035–27042. 10.1073/PNAS.1912792116. McKenzie S, Huszár R, English DF, Kim K, Christensen F, Yoon E, and Buzsáki G (2021). Preexisting hippocampal network dynamics constrain optogenetically induced place fields. Neuron 109, 1040– 1054.e7. 10.1016/J.NEURON.2021.01.011. [PubMed: 33539763] Miao C, Cao Q, Ito HT, Yamahachi H, Witter MP, Moser MB, and Moser EI (2015). Hippocampal remapping after partial inactivation of the medial entorhinal cortex. Neuron 88, 590–603. 10.1016/ J.NEURON.2015.09.051. [PubMed: 26539894] Milstein AD, Li Y, Bittner KC, Grienberger C, Soltesz I, Magee JC, and Romani S (2021). Bidirectional synaptic plasticity rapidly modifies hippocampal representations. Elife 10, e73046. 10.7554/ELIFE.73046. [PubMed: 34882093] Mizumori SJ, McNaughton BL, Barnes CA, and Fox KB (1989). Preserved spatial coding in hippocampal CA1 pyramidal cells during reversible suppression of CA3c output: evidence for pattern completion in hippocampus. J. Neurosci. 9, 3915–3928. [PubMed: 2585060] Mizuseki K, Sirota A, Pastalkova E, and Buzsáki G (2009). Theta oscillations provide temporal windows for local circuit computation in the entorhinal-hippocampal loop. Neuron 64, 267–280. 10.1016/j.neuron.2009.08.037. [PubMed: 19874793] Mizuseki K, Diba K, Pastalkova E, and Buzsáki G (2011). Hippocampal CA1 pyramidal cells form functionally distinct sublayers. Nat. Neurosci. 14, 1174–1181. 10.1038/nn.2894. [PubMed: 21822270] Moser EI, Roudi Y, Witter MP, Kentros C, Bonhoeffer T, and Moser MB (2014). Grid cells and cortical representation. Nat. Rev. Neurosci. 15, 466–481. 10.1038/NRN3766. [PubMed: 24917300] Muller RU, and Kubie JL (1987). The effects of changes in the environment on the spatial firing of hippocampal complex-spike cells. J. Neurosci. 7, 1951–1968. 10.1523/ JNEUROSCI.07-07-01951.1987. [PubMed: 3612226] Muller RU, and Kubie JL (1989). The firing of hippocampal place cells predicts the future position of freely moving rats. J. Neurosci. 9, 4101–4110. 10.1523/JNEUROSCI.09-12-04101.1989. [PubMed: 2592993] Navas-Olive A, Valero M, Jurado-Parras T, de Salas-Quiroga A, Averkin RG, Gambino G, Cid E, and de la Prida LM (2020). Multimodal determinants of phase-locked dynamics across deepsuperficial hippocampal sublayers during theta oscillations. Nat. Commun. 11, 2217. 10.1038/ s41467-020-15840-6. [PubMed: 32371879] Nuñez A, García-Austt E, and Buño W (1990). Slow intrinsic spikes recorded in vivo in rat CA1–CA3 hippocampal pyramidal neurons. Exp. Neurol. 109, 294–299. 10.1016/S0014-4886(05)80020-2. [PubMed: 2209774] O’keefe J, and Recce ML (1993). Phase relationship between hippocampal place units and the EEG theta rhythm. Hippocampus 3, 317–330. [PubMed: 8353611] O’Keefe J, and Nadel L (1978). The hippocampus as a Cognitive Map (Clarendon Press). Pastalkova E, Itskov V, Amarasingham A, and Buzsáki G (2008). Internally generated cell assembly sequences in the rat Hippocampus. Science 321, 1322–1327. 10.1126/science.1159775. [PubMed: 18772431] Petersen PC, and Buzsáki G (2020). Cooling of medial septum reveals theta phase lag coordination of hippocampal cell assemblies. Neuron 107, 731–744.e3. 10.1016/J.NEURON.2020.05.023. [PubMed: 32526196] Petersen PC, Vestergaard M, Jensen KHR, and Berg RW (2014). Premotor spinal network with balanced excitation and inhibition during motor patterns has high resilience to structural division. J. Neurosci. 34, 2774–2784. 10.1523/JNEUROSCI.3349-13.2014. [PubMed: 24553920] Valero et al. Page 24 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Petersen PC, Siegle JH, Steinmetz NA, Mahallati S, and Buzsáki G (2021). CellExplorer: a framework for visualizing and characterizing single neurons. Neuron 109, 3594–3608.e2. 10.1016/ J.NEURON.2021.09.002. [PubMed: 34592168] Rogers S, Rozman PA, Valero M, Doyle WK, and Buzsáki G (2021). Mechanisms and plasticity of chemogenically induced interneuronal suppression of principal cells. Proc. Natl. Acad. Sci. USA 118, e2014157118. 10.1073/PNAS.2014157118. [PubMed: 33372130] Rolls ET, Stringer SM, and Elliot T (2006). Entorhinal cortex grid cells can map to hippocampal place cells by competitive learning. Network 17, 447–465. 10.1080/09548980601064846. [PubMed: 17162463] Royer S, Zemelman B. v, Losonczy A, Kim J, Chance F, Magee JC, and Buzsáki G (2012). Control of timing, rate and bursts of hippocampal place cells by dendritic and somatic inhibition. Nat. Neurosci. 15, 769–775. 10.1038/nn.3077. [PubMed: 22446878] Rueckemann JW, Dimauro AJ, Rangel LM, Han X, Boyden ES, and Eichenbaum H (2016). Transient optogenetic inactivation of the medial entorhinal cortex biases the active population of hippocampal neurons. Hippocampus 26, 246–260. 10.1002/HIPO.22519. [PubMed: 26299904] Savelli F, and Knierim JJ (2010). Hebbian analysis of the transformation of medial entorhinal grid-cell inputs to hippocampal place fields. J. Neurophysiol. 103, 3167–3183. 10.1152/JN.00932.2009. [PubMed: 20357069] Schlesiger MI, Boublil BL, Hales JB, Leutgeb JK, and Leutgeb S (2018). Hippocampal global remapping can occur without input from the medial entorhinal cortex. Cell Rep. 22, 3152–3159. 10.1016/J.CELREP.2018.02.082. [PubMed: 29562172] Skaggs WE, Mcnaughton BL, Wilson MA, and Barnes CA (1996). Theta phase precession in hippocampal neuronal populations and the compression of temporal sequences. Hippocampus 6, 6149–6172. 10.1002/(SICI)1098-1063(1996)6:2. Solstad T, Moser EI, and Einevoll GT (2006). From grid cells to place cells: a mathematical model. Hippocampus 16, 1026–1031. 10.1002/HIPO.20244. [PubMed: 17094145] Soltesz I, and Deschênes M (1993). Lowand 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. 10.1152/JN.1993.70.1.97. [PubMed: 8395591] Souza BC, Pavão R, Belchior H, and Tort ABL (2018). On information metrics for spatial coding. Neuroscience 375, 62–73. 10.1016/J.NEUROSCIENCE.2018.01.066. [PubMed: 29432886] Steffenach HA, Witter M, Moser MB, and Moser EI (2005). Spatial memory in the rat requires the dorsolateral band of the entorhinal cortex. Neuron 45, 301–313. 10.1016/J.NEURON.2004.12.044. [PubMed: 15664181] Tsodyks M. v, Skaggs WE, Sejnowski TJ, and McNaughton BL (1996). Population dynamics and theta rhythm phase precession of hippocampal place cell firing: a spiking neuron model. Hippocampus 6, 271–280. 10.1002/(SICI)1098-1063(1996)6:3. [PubMed: 8841826] Tsodyks M.v., Skaggs WE, Sejnowski TJ, and McNaughton BL (1997). Paradoxical effects of external modulation of inhibitory interneurons. J. Neurosci. 17, 4382–4388. 10.1523/ JNEUROSCI.17-11-04382.1997. [PubMed: 9151754] Valero M, and de la Prida LM (2018). The hippocampus in depth: a sublayer-specific perspective of entorhinal–hippocampal function. Curr. Opin. Neurobiol. 52, 107–114. 10.1016/ j.conb.2018.04.013. [PubMed: 29729527] Valero M, Cid E, Averkin RG, Aguilar J, Sanchez-Aguilera A, Viney TJ, Gomez-Dominguez D, Bellistri E, and de La Prida LM (2015). Determinants of different deep and superficial CA1 pyramidal cell dynamics during sharp-wave ripples. Nat. Neurosci. 18, 1281–1290. 10.1038/ nn.4074. [PubMed: 26214372] Valero M, Averkin RG, Fernandez-Lamo I, Aguilar J, Lopez-Pigozzi D, Brotons-Mas JR, Cid E, Tamas G, Menendez de la Prida L, Prida LM, et al. (2017). Mechanisms for selective single-cell reactivation during off-line sharp-wave ripples and their distortion by fast ripples. Neuron 94, 1234–1247.e7. 10.1016/j.neuron.2017.05.032. [PubMed: 28641116] Valero M, Viney TJ, Machold R, Mederos S, Zutshi I, Schuman B, Senzai Y, Rudy B, and Buzsáki G (2021). Sleep down state-active ID2/Nkx2.1 interneurons in the neocortex. Nat. Neurosci. 24, 401–411. 10.1038/s41593-021-00797-6. [PubMed: 33619404] Valero et al. Page 25 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript (F) Place field features; peak amplitude (ΔVmPF, left), intracellular in-field/out-field theta power difference (ΔPθPF, middle), and spike phase precession slope for the realistic multicompartment model. Valero et al. Page 32 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 5. Chemogenetic activation of CA1 interneurons perturbs place field features (A) Schematic of the combined chemogenetic/optogenetic manipulation experiments. CNO or DMSO (vehicle control) was intraperitoneally injected in CamKIIα-Cre:Ai32 mice (n = 3) expressing the excitatory DREADD hM3Dq in all CA1 interneurons after AAV-hDlx virus injection. (B) tdTomato-expressing DREADD interneurons (white) in CA1 layers and CamkII-ChR2EYFP in pyramidal cells (red). (C) Effect of CNO/DMSO injection (time 0) on the firing rate of pyramidal cells (left) and interneurons (right). Each row is the color-coded firing rate of a neuron. Orange lines show maze exploration epochs. (D) Average peristimulus histograms (PSTH) of pyramidal cell responses to optogenetic stimulation (20 ms) before and after CNO or DMSO injection. (E) Distribution of place-dependent firing rates on the track for all interneurons before and after DMSO (top, 110 fields from 55 interneurons) and CNO (bottom, 106 fields from 53 interneurons). Each field is a single row, sorted according to the location of their lowest Z scored rate. Middle and right panels, trial/track location raster plot for two representative interneurons (marked by triangles). Right insets, unit waveform and autocorrelogram. Valero et al. Page 33 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript (F) Place-dependent rate stability (Spearman’s ρ) after injection for DMSO and CNO groups (χ21,204 = 16.23, p < 10−4, KW test). (G) CNO decreases the standard deviation (SD) of the interneurons’ rate maps. Differences persist after subsampling to match average rate between groups (DMSO resampled) (χ22,323 = 17.86, p < 10−3, KW test). (H) Group differences of mutual information between location and firing rate (χ22,323 = 35.91, p < 10−7, KW test). (I) In the model (see Figure 4C), reduction of the spatial modulation of interneurons (increase toward zero in the ΔGinh axis) predicts a decrease of the three place field signatures of pyramidal neurons shown in Figure 2 (ΔVmPF, ΔPθPF, and phase precession slope). *p < 0.05, ***p < 0.001. Valero et al. Page 34 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Figure 6. Perturbation of interneuron activity impairs place field features (A) Rate maps of pyramidal cells before (PRE, left) and after (POST; middle) CNO (bottom) or DMSO (top) injection sorted by place field position in PRE (left panels). Right panels, place fields during CNO (bottom) or DMSO (top) sorted by POST. (B) Standard deviation (SD) of firing maps after CNO or DMSO injection. CNO decreases SD (χ22,2236 = 87.72, p < 10−20, KW test). (C) PRE-POST difference of in-field firing rate (ΔFRin). CNO decreases in-field pyramidal cell firing rates (χ21,346 = 3.49, p = 0.045, KW test). (D) Mutual information between space and rate after DMSO and CNO injection. DMSO sessions were subsampled to match average rate between groups (DMSO resampled). (E) Correlation between in-field and out-of-field theta mean vector length (MVL) after DMSO and CNO injections (in-field: χ21,346 = 7.07, p = 0.007; out-field: χ21,346 = 38.41, p < 10−10, KW test; diagonal difference: χ21,346 = 6.11, p = 0.013, KW test). The MVL ratio between in-field and out-of-field values is reduced by CNO (left panel; χ21,346 = 7.45, p = 0.006, KW test). (F) Spike theta phase precession of a representative CA1 place cell before (top) and after CNO administration (bottom). Each dot is a spike. The orange/red lines show phase-position correlation (“precession”) and spatial rate distribution. Valero et al. Page 35 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript (G) Phase precession slopes are less negative during CNO, compared with DMSO, and persist after downsampling spikes in DMSO condition (χ22,412= 17.01, p = 0.0002, KW test). (H) Same display as in (G) for mutual information between space and phase (χ22,2337 = 58.63, p < 10−12, KW test). (I) Top, relative weights of neurons for an example assembly. Z scored assembly activity before and after CNO injection for a representative session. Orange rectangles show epochs during track running. (J) Assembly expression is reduced after CNO injection but not DMSO vehicle (χ22,25 = 12.83, p = 0.001, KW test, n = 3 mice). *p < 0.05, **p < 0.01. Valero et al. Page 36 Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Author Manuscript Author Manuscript Author Manuscript Author Manuscript Valero et al. Page 37 KEY RESOURCES TABLE REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies Mouse anti-calbindin D-28k Swant Cat# 300; RRID: AB_10000347 Rhodamine Red goat anti-mouse IgG Thermo Fisher Cat# 115-295-003; RRID: AB_2338756 Alexa Fluor488-conjugated streptavidin Jackson Immunoresearch Cat# 016-540-084; RRID: AB_2337249 Bacterial and virus strains pAAV-hDlx-GqDREADD-dTomato-Fishell-4 addgene 83897-AAV1 Chemicals, peptides, and recombinant proteins Neurobiotin tracer Vector Labs Cat# SP-1120 DPX mountant VWR 360294H DAPI (4’,6-Diamidino-2-Phenylindole, Dihydrochloride) Thermo Fisher D1306 C&B Metabond Parkell Cat#S380 Clozapine N-oxide hydrochloride Sigma-Aldrich SML2304 Experimental models: Organisms/strains Mouse: B6.Cg-Tg(Camk2a-cre)T29-1Stl/J Jackson Laboratory RRID:IMSR_JAX:005359 Mouse: B6; 129S-Gt(ROSA)26Sortm32 (CAGCOP4*H134R/EYFP)Hze/J Jackson Laboratory RRID:IMSR_JAX:012569 Rat: Wistar Instituto Cajal Animal facility N/A Software and algorithms HippoCookBook toolbox (MATLAB toolbox for extracellular/intracellular recordings and behaviour) Manuel Valero https://github.com/cortex-lab/KiloSort (https://doi.org/10.5281/zenodo.6902376) LCN-HippoModel (python-based biophysically realistic model) Andrea Navas-Olive and de la Prida lab https://github.com/PridaLab/LCN-HippoModel (https://doi.org/10.5281/zenodo.6902418) KiloSort (template-based spike sorting MATLAB software) Pachitariu M & Cortex-lab https://github.com/cortex-lab/KiloSort KilosortWrapper Peter C. Petersen & Brendon Watson https://github.com/petersenpeter/ KilosortWrapper Phy (Python GUI for manual spike curation) Cyrille Rossant, Ken Harris et al. https://github.com/cortex-lab/phy Phy plugins Peter C. Petersen https://github.com/petersenpeter/phy1-plugins MATLAB MathWorks https://www.mathworks.com/ FMA Toolbox (MATLAB toolbox for Freely Moving Animal (FMA)) Michael Zugaro https://fmatoolbox.sourceforge.net/ CellExplorer (Cell classification pipeline and graphical interface) Petersen and Buzsaki, 2020 https://linkinghub.elsevier.com/retrieve/pii/ S0896627321006565 Other Silicon probes Neuronexus, Cambridge Neurotech A1x16, H3 4 shank mLED probes Plexon NeuroLight Optoelectrode Cell Rep . Author manuscript; available in PMC 2022 October 25. Author Manuscript Author Manuscript Author Manuscript Author Manuscript Valero et al. Page 38 REAGENT or RESOURCE SOURCE IDENTIFIER RHD2000 USB Interface Board Intan Technologies C3100 64 and 32 channel digital amplifiers Intan Technologies C3314, C3324 PulsePal v2 Sanworks N/A Axoclamp 900A Microelectrode Amplifier Molecular Devices N/A Data adquisition interface Power3A Cambridge Electronic Design Limited (CED) Power1401-3A 3D printed microdrives Mihály Vöröslakos, Peter Petersen and György Buzsáki https://github.com/buzsakilab/3d_print_designs Cell Rep . Author manuscript; available in PMC 2022 October 25.