A. Korovaichuk, J. Makarova, V. A. Makarov, N. Benito and O.

Herreras
J Neurophysiol 104:484-497, 2010. First published May 12, 2010; doi:10.1152/jn.00297.2010

You might find this additional information useful...

Supplemental material for this article can be found at:

http://jn.physiology.org/cgi/content/full/jn.00297.2010/DC1

This article cites 55 articles, 24 of which you can access free at:
http://jn.physiology.org/cgi/content/full/104/1/484#BIBL

Updated information and services including high-resolution figures, can be found at:
http://jn.physiology.org/cgi/content/full/104/1/484

Additional material and information about Journal of Neurophysiology can be found at:
http://www.the-aps.org/publications/jn

This information is current as of September 3, 2010 .

Downloaded from jn.physiology.org on September 3, 2010

Journal of Neurophysiology publishes original articles on the function of the nervous system. It is published 12 times a year
(monthly) by the American Physiological Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2010 by the
American Physiological Society. ISSN: 0022-3077, ESSN: 1522-1598. Visit our website at http://www.the-aps.org/.
J Neurophysiol 104: 484 – 497, 2010.
First published May 12, 2010; doi:10.1152/jn.00297.2010.

Minor Contribution of Principal Excitatory Pathways to Hippocampal LFPs

in the Anesthetized Rat: A Combined Independent Component and Current
Source Density Study

A. Korovaichuk,1,* J. Makarova,1,* V. A. Makarov,2,* N. Benito,1 and O. Herreras1

1
Department of Systems Neuroscience, Cajal Institute–Consejo Superior de Investigaciones Científicas, Madrid; and 2Department
of Applied Mathematics, Faculty of Mathematics, Universidad Complutense of Madrid, Madrid, Spain
Submitted 26 March 2010; accepted in final form 6 May 2010

Korovaichuk A, Makarova J, Makarov VA, Benito N, Herreras Approaches are emerging to analyze LFPs, such as spectral
O. Minor contribution of principal excitatory pathways to hippocam- coherence, principal components, and laminar population anal-
pal LFPs in the anesthetized rat: A combined independent component ysis, and these have proven useful to identify important fea-
and current source density study. J Neurophysiol 104: 484 – 497,
2010. First published May 12, 2010; doi:10.1152/jn.00297.2010. tures in mixed deep electrical sources (for example, see
Einevoll et al. 2007; Kocsis et al. 1999). Recently we imple-

Downloaded from jn.physiology.org on September 3, 2010

Analysis of local field potentials (LFPs) helps understand the function
of the converging neuronal populations that produce the mixed syn- mented a blind source separation strategy (Makarov et al.
aptic activity in principal cells. Recently, using independent compo- 2010) that takes advantage of the spatial immobility of elec-
nent analysis (ICA), we resolved ongoing hippocampal activity into trical brain sources in the LFP’s time scale. This structural
several major contributions of stratified LFP-generators. Here, using feature arises from the fixed location of the synaptic afferents
pathway-specific LFP reconstruction, we isolated LFP-generators that
to target cells, producing transmembrane currents whose ex-
describe the activity of Schaffer-CA1 and Perforant-Dentate excita-
tory inputs in the anesthetized rat. First, we applied ICA and current tracellular part is circumscribed to the postsynaptic cell. The
source density analysis to LFPs evoked by electrical subthreshold approach uses independent component analysis (ICA) to sep-
stimulation of the pathways. The results showed that pathway specific arate the informative sources (Bell and Sejnowski 1995;
activity is selectively captured by individual components or LFP- Makeig et al. 2004; Stone 2004), and it allowed us to disen-
generators. Each generator matches the known distribution of axonal tangle intracerebral LFPs recorded in the hippocampal CA1
terminal fields in the hippocampus and recovers the specific time region into the contributions of a few major LFP-generators
course of their activation. Second, we use sparse weak electrical
(Makarov et al. 2010). In this study, we present a special
stimulation to prime ongoing LFPs with activity of a known origin.
Decomposition of ongoing LFPs yields a few significant LFP-gener- technique that permits the unequivocal association of the LFP-
ators with distinct spatiotemporal characteristics for the Schaffer and generators isolated with the presynaptic neuronal populations
Perforant inputs. Both pathways convey an irregular temporal pattern and its application to quantify the independent contribution of
in bouts of population activity of varying amplitude. Importantly, the Schaffer and Perforant inputs to the total hippocampal LFP.
contribution of Schaffer and Perforant inputs to the power of raw Thus we have obtained a tool for the parallel readout of the
LFPs in the hippocampus is minor (7 and 5%, respectively). The discrete activities of the working neuronal populations from
results support the hypothesis on a sparse population code used by standard LFPs.
excitatory populations in the entorhino-hippocampal system, and they ICA is a widely used approach to extract spatially distinct
validate the separation of LFP-generators as a powerful tool to explore independent sources of activity from mixed signals (for review,
the computational function of neuronal circuits in real time.
see Choi et al. 2005). Simultaneous multisite recordings are
required to estimate the different impact of several spatially
INTRODUCTION localized and distant generators into each sensor. This method
has been successfully applied to surface EEG recordings,
Multiple neuronal populations form a complex system of whereby deep strongly synchronized neural sources can be
local circuits and global networks whose activity constitutes discriminated (Jung et al. 2005), but their spatial localization is
the basis for information processing in the brain. The electrical only poorly assessed and it lacks cellular correlates. These
transmembrane currents produced by working neurons are problems are significantly diminished when using intracerebral
mixed together and generate local field potentials (LFPs), a local recordings, because the signals are generated in the
macroscopic variable that reflects the information processing at volume surrounding the array of recording sensors. The paral-
mesoscopic scales. Thus the decoding of LFPs may help us lel arrangement of principal cells and the pronounced stratifi-
understand the principles of information handling. To date, cation of inputs in the hippocampus also favor the application
resolving the electrical activity of converging neuronal popu- of ICA and interpretation of the results obtained.
lations into the original informative sources remains problem- Our recent application of ICA to LFPs in the rat CA1
atic (Mitzdorf 1985; Nunez and Srinivasan 2006). showed several dominant generators of LFPs, producing a
subcellular definition of their spatial localizations that was
* A. Korovaichuk, J. Makarova, and V. A. Makarov contributed equally to constant across animals (Makarov et al. 2010). A causal rela-
this work.
Address for reprint requests and other correspondence: O. Herreras, Cajal
tion between the firing of local units and the activity of specific
Inst.-CSIC, Av. Dr. Arce 37, Madrid 28002, Spain (E-mail: herreras@cajal. LFP-generators was also shown. This result suggests that each
csic.es). isolated LFP-generator may have a neuronal origin with dif-
484 0022-3077/10 Copyright © 2010 The American Physiological Society www.jn.org
 EXCITATORY GENERATORS OF HIPPOCAMPAL LFPS 485

ferent functionality. Indeed, one may expect that the activity of guidelines (86/609/EU) and Spanish regulations (BOE 67/8509-12
an afferent input to a restricted dendritic field in the principal 1988) regarding the use of laboratory animals.
cells may be associated with a single LFP-generator disentan-
gled by ICA from the complete LFPs. To test this hypothesis, Histology
here we have used selective electrical stimulation of known
hippocampal pathways such as the Schaffer input to CA1 and Silicon probes were labeled with DiI (Molecular Probes, Invitro-
the perforant path (PP) input to the dentate gyrus (DG). gen, Carisbad, CA) by soaking in a 1% solution of N-N-dimethyl-
formamide (Sigma-Aldrich, St. Louis, MO). At the end of the exper-
Selective stimulation elicits evoked LFPs whose well-known
iment, animals were perfused transcardially with saline/heparine and
spatial maps are used to crosscheck the ICA-derived LFP- paraformaldehyde (4%). Coronal cryostat sections (15 ␮m) were
generators in the first part of the study. obtained and labeled with bis-benzimide (Sigma), and they were
Hippocampal evoked potentials are regularly used to explore examined under a fluorescence microscope (560 – 610 nm) to identify
the functional properties of neuronal populations. Experimen- the probe position (see example in the Supplementary Fig. S1).1
tal and theoretical studies show they are reliable population
indices of the corresponding membrane events in single neu-
Independent component and current source density analyses
rons (Andersen et al. 1971; López-Aguado et al. 2002; Varona
of LFPs
et al. 2000). Evoked synaptic inputs produce inward and
outward currents covering the entire morphology of principal The mathematical procedure and detailed signal treatment of the
cells with pathway specific spatial distribution that can be ICA were described elsewhere (Makarov et al. 2010). Briefly, 16 or 32
precisely determined by current-source density (CSD) analysis LFP signals u共t兲 ⫽ 兵uk共t兲其k⫽1
K
recorded simultaneously can be repre-

Downloaded from jn.physiology.org on September 3, 2010

(Freeman and Nicholson 1975; Lorente de Nó 1947). As a first sented as the weighted sum of the activities of N neuronal sources or
step, we assessed the spatial and temporal performance of the LFP-generators2
ICA, as well as its capacity to discriminate different sources N
using evoked LFPs of known origin. u(t) ⫽ 兺 Vnsn(t) ⫺ ␴⌬Vn ⫽ In ∀ n 僆 [1, N]
n⫽1
(1)
Subsequently, we studied spontaneous LFPs primed by ran-
dom sparse subthreshold evoked activity to identify the disen- where V ⫽ [V1, V2 . . .,Vn]T is the mixing matrix composed of the
tangled components. The presence of evoked and well-known so-called voltage loadings or spatial weights of all LFP-generators;
spontaneous events specific for each pathway in the raw LFPs, 兵sn共t兲其n⫽1
N
are the time courses or activations of the LFP-generators; ␴
and in a single LFP-generator, guarantees its correct unequiv- is the conductivity of the extracellular space; and 兵In其n⫽1
N
are the CSD
ocal identification. We further quantified the specific contribu- loadings. Thus the raw LFP observed at the kth electrode tip is a linear
tion of LFP-generators to the global signal during irregular mixture of the electrical activity of several independent LFP-genera-
tors describing transmembrane currents in principle cells.
LFP patterns. Interestingly, the two major excitatory pathways,
Figure 1A shows an example of (evoked) LFPs recorded in the CA1
Schaffer and PP, contribute little to ongoing LFPs, producing by a 16-site probe along the pyramidal neuron axis. The raw LFPs can
scarce and irregular bouts of activity. be used to evaluate CSD maps by approximating the Laplace operator
by 1D finite differences (Fig. 1B)
METHODS 1 uk⫺1(t) ⫺ 2uk(t) ⫹ uk⫹1(t)
CSD ⫽ ⫺ ⌬u ⫽ ⫺ (2)
␴ ␴h2
Experimental methods
where h ⫽ 50 ␮m is the intersite distance. The CSD shows the
Female Sprague-Dawley rats (200 –220 g) were anesthetized with characteristic Schaffer-evoked pattern, where several neuronal events
urethane (1.2 g/kg, ip) and fastened to a stereotaxic device, maintain- can be appreciated overlapping in space and time, such as generation
ing their body temperature at 37°C with a heated blanket. The surgical of a propagating population spike.
and stereotaxic procedures performed were described elsewhere (Ca- The ICA of u(t) returns the generator’s activation 兵sn共t兲其n⫽1
N
(Fig.
nals et al. 2005; Makarova et al. 2008). Concentric bipolar stimulating 1C) and the mixing matrix V (Fig. 1D) for ⱕ16 LFP-generators.
electrodes were placed in at least two of the following locations: the Usually only a few of them (G1–G6 in the figure) have significant
alveus for antidromic activation (from bregma, midline, and cortical amplitude and different spatial distributions (voltage loadings). We
surface AP: ⫺5.5, L: 2.6, V: ⫺1.8 mm), the ipsilateral CA3 (septal noticed that, because of the ambiguity of the ICA (fortunately insig-
pole) for orthodromic activation of the CA1 region (AP: ⫺3.2, L: 2.6, nificant for our study), the voltage loadings and activations are given
V: ⫺3.3 mm), and the angular bundle (AP: ⫺8, L: 4 –5, V: 3.5– 4 mm) in arbitrary units (i.e., we can arbitrarily scale the loadings but we
for orthodromic activation of the granule cell population in the DG. must simultaneously apply inverse scaling to the activations). Once
These points were approached at a 30° angle in the sagittal plane. LFP-generators have been extracted from the raw LFPs, we can
Stimuli (0.07– 0.1 ms square pulses, 0.1– 0.8 mA) were applied to analyze them as if they were active alone. For example, we can
elicit the characteristic evoked potentials, which were used to guide construct virtual LFPs produced by a single generator, say kth, from
the placement of recording probes. Linear multisite probes (models: its voltage loading and activation
A1x16-5mm50-177 and A1x32-6mm50-413, Neuronexus Technolo-
gies, Ann Arbor, MI) were lowered into the hippocampus (AP:
1
4.5–5.5, L: 2–3 mm) and connected to a multiple high-impedance The online version of this article contains supplemental data.
2
headstage. The 16-site probes were used to record along the main axis Clarification of terminology. The term source is used with different
of CA1 pyramidal cells, whereas the 32-site probes spanned both CA1 implications in ICA and CSD analysis. In the former, source is used in a
comprehensive manner to denote active brain regions (i.e., large zones of
and DG/CA3 regions. A silver chloride wire in the neck skin served neural activity). Here a source in the ICA sense refers to each of the individual
as a reference for recording, and the signals were amplified and coherent components of the LFPs. Neither of these should be confounded with
acquired using low-noise MultiChannel System (Reutlingen, Ger- the formal notation of outward transmembrane currents or current sources used
many) recording hardware and software (50 kHz sampling rate). All in CSD analysis. For the sake of simplicity, we maintain the latter and adopt
the experiments were performed in accordance with European Union the term LFP-generator for the ICA-derived sources.

J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org

486 KOROVAICHUK ET AL.

A raw LFPs B raw CSD

1
2

electrode #
CSD analysis
8 8

15
16
10 ms 5 mV 10 ms 50 m

ICA
C D E
generator activations voltage loadings CSD loadings
G1 1

# of electrode
2nd
G8 8 derivative

G16 16
10 ms 8 a.u. 0 1
-1
(a.u.) -0.5 0
(a.u.) 0.5

Downloaded from jn.physiology.org on September 3, 2010

F virtual LFPs from G6
G
1
CSD for reconstructed LFPs
2
electrode #

CSD analysis
8 8

15
16
10 ms 5 mV 10 ms 50 µm

FIG. 1. General strategy of local field potential (LFP) decomposition and the identification of generators. The diagram shows the sequential steps from an
LFP spatial map recorded along the CA1 axis containing evoked responses to a twin pulse in the ipsilateral CA3 (Schaffer response). A similar procedure is
followed for spontaneous LFPs. The simultaneous recording of 16 sites along the main axis of the principal cell (pyramidal neurons) (A) is analyzed by
current-source density (CSD) (B) to obtain the map of inward and outward population currents (sinks are in blue and sources in yellow-red). Multiple membrane
events overlap spatiotemporally. The LFP profile is also analyzed by an independent component analysis (ICA) that provides a set of 16 generators each with
a temporal activation (C) and spatial distribution (D). Of the 16 generators, only a few (1– 6 in the example) have a significant variance, whereas the others are
rejected for analysis. The second spatial derivative of the spatial weight curves (or voltage loadings) provides the CSD loadings (E), a spatial index that reflects
the distribution of membrane currents along the conductor. Because the ICA does not ensure the correct polarity or absolute values of amplitude, we reconstructed
the LFP for each generator isolated (F). Such partial LFPs regain the correct polarity and true amplitude. A CSD can be applied to the reconstructed LFP, which
provides simplified spatiotemporal maps of inward/outward currents for unique spatially coherent membrane events (G). These maps can be compared with the
CSD of raw LFPs (curved dashed arrow) to find specific membrane events buried within the mixed CSD (closed dashed white lines). It can be shown that the
spatial profiles of partial CSDs (E) match the spatial profile of the CSD loadings (small dashed arrow) and hence their true polarity is assessed. Arrowheads mark
the time of stimulus (artifacts have been removed for analysis). a.u., arbitrary units.

uGk(t) ⫽ Vksk(t) (3) LFP-generators corresponding to the same population of presynaptic

neurons. Thus to define a measure of similarity between a pair (k, m)
Figure 1F shows LFPs contributed by G6 only. Note that the ICA of LFP-generators, we used the distance measure between their
ambiguity does not affect the virtual LFPs, i.e.: the reconstructed voltage loadings (Makarov et al. 2010)
LFPs are dimensional. The virtual LFPs can be used to evaluate the
CSD created by a single generator (Fig. 1G shows the CSD of G6), d(Vk, Vm) ⫽ 1 ⫺
ⱍ具V , V 典ⱍ
k m
(4)
and thus the raw and generator-based CSDs can be compared (Fig. 1, 储Vk 储 储 Vm 储
B and G). We also noticed that, using Eqs. 1–3, the CSD related to the
kth generator can be evaluated by CSDGk ⫽ Iksk共t兲.
具Vk, Vm 典 ⫽ 兰
⍀
V kV m ⫹ ␬ ⵜ V k ⵜ V m
For the ICA, we made use of the infomax algorithm proposed by ⫹ ␬2⌬Vk⌬Vmdx, 储 Vk储2 ⫽ 具Vk, Vk 典
Bell and Sejnowski (1995) and further modified by Amari et al. (1996)
where ␬ ⫽ 0.05 mm2 is the dimensional constant. The distance
and Lee et al. (1999). The algorithm is implemented in the EEGLAB
measure is bounded 0 ⱕ d ⱕ 1 and d ⫽ 1 for two orthogonal
Matlab toolbox (Delorme and Makeig 2004).
(completely different) loadings, whereas d ⫽ 0 for two equivalent
(identical) loadings.
Comparison of LFP-generators: Distance measure To partition a set of voltage loadings into disjointed clusters, we use
hierarchical clustering based on the introduced distance measure with
Each application of the ICA on LFPs, to either different recordings the average-linkage algorithm.
or for different time windows, provides several LFP-generators, and
thus we need a method to compare and classify them. Each generator Identification of stable LFP-generators
is characterized by its activation sk(t) and spatial loading Vk (Fig. 1, C
and D). Activations are time specific and therefore they cannot be An ICA may theoretically isolate as many LFP-generators as the
used for comparison. In contrast, the loadings depend on the spatial number of LFP signals (16 or 32 in these experiments), yet obviously,
distribution of an axon’s terminals and hence they must be similar for only a few of them will be statistically significant and stable. For

J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org

 EXCITATORY GENERATORS OF HIPPOCAMPAL LFPS 487

example, in Fig. 1C, the first 6 generators have high amplitude and the same spatial distribution as defined by the axon’s terminals. Thus
they are clearly related to the stimulus, whereas the other 10 may two types of LFPs with the same spatial characteristics must be
capture noise and therefore they may vary strongly from one record- captured by one and the same LFP-generator, and hence the presence
ing to another. In our former study (Makarov et al. 2010), we of evoked responses in the activation of the LFP-generator sk(t) assists
considered those LFP-generators contributing ⬎5% to the total LFP in its identification. In preliminary experiments, we found that sub-
variance (power) as statistically significant. The number of reliable threshold (⬍0.5 ⫻ threshold) stimuli delivered at an average rate of 1
components that can be extracted in practice depends on many factors Hz do not significantly change the variance of the total signal. We also
such as the particular region under study, the signal to noise ratio, and found that strong (near threshold and suprathreshold) evoked re-
the relative strength and spatial extension of active components. Some sponses produce a complex mixture of components when analyzed by
neuronal populations (e.g., hippocampal pyramidal cells) fire very ICA/CSD (see Fig. 1). Apparently such phenomena are not present in
rarely (Ranck 1973; Thompson and Best 1989) and hence, the the low-amplitude events of ongoing LFPs, and thus we chose low
postsynaptic activity of their targets may contribute little to the total stimulus intensities to prevent the generation of PS components (i.e.,
LFPs. In this study, we consider an isolated LFP-generator with the we achieve vanishing recruitment of intrinsic currents into field
variance contribution ⬍5% to be significant, as long as it is unequiv- excitatory postsynaptic potentials (fEPSPs) during synchronous acti-
ocally related to a known pathway and provides a clean CSD spatial vation, Herreras 1990). For Schaffer activation, we chose stimuli that
map (low noise and a flat baseline). To identify the most stable produced fEPSPs of ⱕ0.5 mV in the stratum radiatum, and for the PP
LFP-generators in spontaneous LFPs, we used the two-step algorithm input, the intensity was adjusted to raise a positive field in the Hilus
(Makarov et al. 2010). First, we identified “global” generators through ⱕ1 mV (below 0.5 mV in the DG molecular layer). To analyze the
an ICA of a sufficiently long recording (tens of seconds). Second, we LFPs, we used a two-epoch recording: 1) spontaneous and 2) spon-
divided the same recording into short-term contiguous epochs (each 1 taneous plus evoked activities. Thus we can examine whether evoked
s long), and we applied an ICA to each of these segments. We expect activity modifies the level of spontaneous activity on the same or

Downloaded from jn.physiology.org on September 3, 2010

that stable strong LFP-generators should be presented in all epochs, another LFP-generator. In accordance with the stability analysis of the
whereas noisy, unstable, temporal or weak generators will fluctuate, LFP-generators (Makarov et al. 2010), we found that mixing sparse
and hence, their spatial loading will differ from epoch to epoch. We subthreshold evoked activity into ongoing LFPs neither modifies their
identified those that are most stable over time by calculating the characteristics nor affects the spatial profiles of the LFP-generators, as
similarity measure (4) between the “global” and “local” generators. long as the probe array remained in place.

Power of LFP-generators Reconstruction of LFPs for specific pathways

Using Eq. 3, we reconstructed virtual LFPs corresponding to the Reconstructed LFPs recover the correct polarity and provide a good
generator Gk, and we defined the mean power of Gk (measured in quantitative estimate of the pathway-specific variance and power. The
mV2) as power was calculated in the electrode that showed maximum ampli-

冉 冊
T tude during irregular LFP patterns (i.e., excluding theta periods) over
1
PGk ⫽ max
m僆[1,K] T
兰u0
2
m(t)dt (5) a total of 3 min of 20 s randomly chosen segments within 1 h. A major
advantage of pathway-specific LFPs is that they provide simplified
CSD maps describing a unique spatial structure of inward and out-
where T is the time interval for averaging (about a few minutes in our ward currents. These can easily be matched to those obtained from
experiments). To evaluate the evolution of the power of the genera- customary hippocampal evoked fields, whose uncomplicated nature
tor’s activations over time, we use the convolution with the appropri- and well-defined distribution along the anatomy of principal cells
ately scaled square kernel H assist path identification (Herreras 1990; Herreras et al. 1987; Leung
PGk(t) ⫽ 兰 H(t ⫺ ␶)s (␶)d␶2
k (6)
et al. 1995; Lømo 1971).

RESULTS
Analysis of evoked LFPs
We first tested the spatial and temporal performance of the
Evoked LFPs were recorded for a series of stimuli of varying ICA on evoked LFPs obtained by activation of the Schaffer-
intensity (short, 0.1 ms, square pulses at a rate of 0.1–1 Hz), from a CA1 and PP-dentate inputs. Subsequently, we used the path-
low subthreshold to supramaximal [i.e., producing maximal popula-
way specificity of the ICA-derived LFP-generators to explore
tion spike (PS)]. The stimulus intensity was measured in units of the
threshold intensity to evoke a PS and stimulus artifacts were removed, the spontaneous population activity of these pathways in on-
except where indicated. Each series of evoked LFPs was analyzed as going LFPs.
one long signal containing either all stimuli or selected groups of
evoked responses. Only the period of interest (40 –50 ms) immediately Spatial and temporal precision of pathway-specific
after the stimulus was analyzed for each profile, and thus intermediate LFP-generators
periods were removed off-line and the time windows containing
responses were concatenated into a continuous unique signal. This SEPARATION OF ARTIFACTS AND NEURAL ACTIVITY. The capacity
procedure does not affect the applicability of ICA but rather improves of the ICA to suppress EEG artifacts is well established (see
its efficiency by increasing the relative variance of the signal of Castellanos and Makarov 2006; Vigario et al. 2000 and refer-
interest.
ences therein), and thus we assess here its applicability to
isolate stimulus artifacts (caused by electric shock) in deep
Priming of LFPs with evoked activity brain recordings. We antidromically stimulated the CA1 pyra-
To identify the presynaptic populations corresponding to LFP- midal population from the alveus (Fig. 2A, cartoon inset), and
generators, we stimulated known afferent pathways with random the electric shock produced a large negative spike-like poten-
sparse subthreshold pulses. The rationale was that the minimal electric tial (⬎40 mV, truncated at the black dot in Fig. 2Aa) that may
activation of a pathway emulates its spontaneous excitation. Therefore reverse polarity outside the hippocampal tissue. This artifact
the evoked and spontaneous activities of a given pathway must have was followed by evoked LFPs with the characteristic profile of
J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org
488 KOROVAICHUK ET AL.

A a raw LFP c volt. loadings

(Fig. 2Ae). Thus when evoked LFPs contain a large artifact, the
1
resolution of the mixed neuronal activity significantly deterio-
G2 rates. Hence, in subsequent analysis, we routinely cut off the
electrode #

stimulus artifacts before applying the ICA.

8
We found that the results shown in Fig. 2A were stable in all
G1 animals, and the clustering analysis of voltage loadings of all

5 mV
16 the LFP-generators in the seven animals provided a well-
0 1 2 3
(a.u.) defined partition into two groups (Fig. 2Ba; intracluster dis-
b CSD for raw LFP tance ⬍0.15, intercluster distance ⬎0.5). This is evident in Fig.
2 9 2Bb, which shows the mean ⫾ SD spatial loadings for the
artificial G1 and neuronal G2 generators averaged over load-
8 0 ings belonging to the corresponding cluster (n ⫽ 7).
SUBTHRESHOLD ACTIVATION OF AN AFFERENT PATHWAY IS CAP-
15 -9 TURED BY A SINGLE LFP-GENERATOR WITH PRECISE SPATIOTEMPO-
1 ms
RAL DEFINITION. Let us now examine the ICA performance on
LFPs evoked by activation of Schaffer-CA1 and PP-DG inputs
d activations e expl. variance (a detailed description of evoked LFPs and their CSD analysis
1.0
G1 can be found in the Supplementary Fig. S1). We first studied

Downloaded from jn.physiology.org on September 3, 2010

the excitatory Schaffer input to the CA1 that was activated by
0.5
G2 subthreshold (0.5 ⫻ threshold) stimuli to minimize the second-
40
1 ary intrinsic currents (Herreras 1990). Figure 3A shows raw
(a.u.) 0.0
G1 G2 G3-G16 profiles of evoked LFPs and the corresponding CSD map. It
can be appreciated a single active sink centered in the st.
B a clustering of generators b mean v. loadings radiatum, surrounded by two passive current sources in the
0.6 -0.25
G2 soma region and distal apical dendrites.
dissimilarity

Application of the ICA identified 16 LFP-generators, and

depth (mm)

0.4 0
only 1 of them, G1, had statistically significant amplitude (Fig.
0.2 0.25 3B). The activation of G1 matches the time course of the raw
G1
LFP at the site of maximum amplitude (electrode 9 in Fig.
0
2a 4a 3a 5a 6a 7a 1a1b 2b 5b 6b 7b 4b 3b 1c
0.5
-1 0 1 2 3
3Aa). G1 has a bell-shaped spatial distribution (Fig. 3Bc, blue
(a.u.) curve) that is maximal in the apical dendrites, and its polarity
FIG. 2. Segregation of artifacts and neural activity by the ICA. The figure reverses in the soma layer, in close correspondence to the
corresponds to the analysis of an LFP obtained during antidromic activation of spatial profile of LFPs calculated for the time instant marked
the CA1 field. The alvear electric shock produced a large voltage artifact (dot by the arrow in Fig. 3Aa (Fig. 3Bc, black line). The spatiotem-
in Aa) followed by an antidromic population spike. The artifact is nearly poral dynamics of the CSD obtained from the LFPs recon-
deleted in the corresponding CSD (Ab) while locally generated active sink and
passive sources are visible. The ICA showed only 2 significant generators (Bb). structed from G1 matched the CSD map calculated from the
G1 contains the stimulus artifact (Ba) that enters all electrodes with similar raw LFP profile (Fig. 3, Be vs. Ab). We noticed that the
weight (flat voltage loading in Ab). G2 contains all the neural activity, three-layered source-sink-source distribution caused by excita-
displaying a bell-shaped voltage loading curve spanning the activated soma- tory synaptic and return currents was not decomposed by the
todendritic domains (Ac). B: the cluster analysis of all voltage loading curves
obtained in 7 animals establishes 2 groups (blue and green) corresponding to
ICA into three different components, but rather, it was ex-
the artifacts and population spikes, respectively. The average spatial loading plained by the single G1 LFP-generator. The transmembrane
curves is shown in Bb (mean ⫾ SD). source/sink loop of current raised by a discrete patch of active
membranes is contained within the shape of the spatial weights
the antidromic PS (Varona et al. 2000). The corresponding of the ICA-isolated LFP-generator (loading vector V1 in Eq. 1).
CSD map (Fig. 2Ab) shows the somatodendritic sink (blue, Similar results were repeatedly obtained in five animals, and
electrodes 7–11) surrounded by passive sources (yellow/red). in a clustering analysis of the loading curves obtained in
The polarity of currents reverses in the ensuing slower tail, as different experiments, the interloading distance was ⬍0.1 for
expected for somatic recurrent inhibition (active somatic cur- all generators (Fig. 3Ca). In one experiment, we obtained two
rents flanked by passive basal and apical dendritic currents). significant generators: one of them (2a in Fig. 3Ca) had a
The ICA successfully isolated the artificial potential as a similar voltage loading as that described above (Fig. 3Bc),
single generator, G1, that was devoid of neuronal activity (Fig. whereas the other (2b in Fig. 3Ca) was significantly different.
2, Ac and Ad, blue curves). We noted that the spatial voltage Hence, in this experiment, it would seem that some spontane-
loading of the artificial generator is essentially flat (Fig. 2Ac, ous LFP event of a different origin overlapped with the evoked
blue line). Such a generator, which enters all electrodes at the activity, which was isolated into a separate LFP-generator. The
same strength or with a linear decay, has a null CSD (Eq. 2), mean ⫾ SD voltage loading for the Schaffer synaptic input
which means that its origin lies outside the recording zone and averaged over n ⫽ 5 experiments is shown in Fig. 3Cb.
that the generator is volume propagated, as expected. The In a representative experiment that examines PP-DG evoked
second strongest LFP-generator, G2, contains the main part of responses (Fig. 4), subthreshold (0.5 ⫻ threshold) activation of
the neuronal activity, including the PS and inhibitory currents. the PP produced the expected negative LFP in the molecular
The LFP variance explained by the remaining G3–G16 is ⬍5% layer of both blades and a strong positive potential across the
J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org
 EXCITATORY GENERATORS OF HIPPOCAMPAL LFPS 489

A a b
signals. Thus one ICA-isolated LFP-generator may contain
raw LFPs raw CSD synaptic activity from several separate populations, as long as
1
0.8 they are simultaneously activated by a common input.
Qualitatively, the same results were observed when we
electrode #

repeated this analysis in five animals. However, since the

8 0 separation between the GC layers varies because of the curved
arrangement of this population, the anatomical normalization
of the recording tracks across animals is not ideal. This pre-
16
-0.8
2 ms
cludes direct comparison of spatial loadings obtained in dif-
5 mV ferent experiments. Supplementary Fig. S2 shows the results
from the other four experiments.
B a activation of G1 b
1
explained variance
SEPARATION OF THE ACTIVITY EVOKED BY MULTIPLE PATHWAYS. Be-
0.5 cause of the volume propagation of extracellular currents, a
2 ms 0.3 (a.u.)
LFP may contain activity from several domains within the
0
G1 G1-G13 same neurons or even from different distant populations and at
c d a distance from the recording electrode. Thus we tested
volt. loading CSD loading e CSD for G1
1 whether the ICA can decompose LFPs with a contribution from
0.8
two pathways into the original Schaffer and PP inputs to the
CA1 and DG, respectively.
electrode #

Downloaded from jn.physiology.org on September 3, 2010

8 0 Conveniently, the pyramidal and granule cell populations
are spatially segregated, and they can be covered by different
G1
groups of electrodes in the 32-site linear probe. In the two
LFP -0.8 representative experiments shown in Fig. 5, a series of sub-
16
0 4 8 -1 0 1 2 2 ms threshold paired pulses of stimuli were delivered at a rate of 0.2
(a.u.) (a.u.)
Hz rate to the Schaffer and PP pathways. These pathways were
C a clustering of generators b mean voltage loading either stimulated with a 100 ms delay or simultaneously (left
-0.25
and right columns in Fig. 5, respectively). The stimulus inten-
0.6
sity was maintained constant for the Schaffer input, whereas
depth (mm)

0 for the PP input, it increased over time for noncoincident pairs

dissimilarity

0.4 of stimuli (Fig. 5A) and decreased for the coincident pairs (Fig.
5B), yet always remained below a 0.5 ⫻ threshold. The LFPs
0.2
0.25 recorded for the two stimulus configurations by the 32-site
linear probe extended from the CA1 st. oriens to the lower
GCL can be seen in Fig. 5A (see Supplementary Fig. S1 for
0 0.5
1a 5a 2a 3a 4a 2b -3 0 3 6 9 greater detail).
(a.u.)
NONCOINCIDENT STIMULI. An electric shock in the ipsilateral
FIG.
3. Precise spatial and temporal performance of ICA on LFPs produced CA3 produces the Schaffer fEPSP in CA1, as well as a local
during subthreshold activation of the Schaffer input to CA1. A: an electric
shock in the ipsilateral CA3 evoked a negative potential in the apical dendritic
multiphasic field in the CA3/DG. The CSD map shows the
tree (a). The schematic neuron indicates the approximate recording position. characteristic source-sink-source distribution of currents along
The CSD reveals inward currents (sinks) centered in stratum radiatum flanked the CA1 pyramidal cell axis and a local source-sink dipole
by passive sources in the soma layer and stratum lacunosum (b). B: the ICA (Fig. 5B, Sch). This latter dipole contains a mixture of the
applied to the LFP profile shows a unique significant component (b) with a antidromic sink in the area of the granule cell and of the local
temporal course matching the voltage change in central electrodes (a). C: the
voltage and CSD loadings (blue plots) coincide with the spatial profiles of the recurrent orthodromic fEPSPs in CA3 (this local response may
raw LFP and CSD (black plots; measured at the instant marked by the small contain PP activation in other experiments, depending on the
arrows in A). The CSD calculated for the reconstructed LFP for G1 (c) is position of the stimulating electrode). Activation of the PP
similar to the raw CSD (Ab). D: cluster analysis for voltage loadings obtained pathway produced an excitatory sink in the molecular layer
in 4 animals show a unique group (a). The average voltage loading (mean ⫾
SE) shown in b was built using the st. pyramidale as a spatial reference.
(100 ␮m above the CA3 sink), as well as the corresponding
somatic sources in both GCLs (Fig. 5B, PP). A weaker sink
Hilus (Fig. 4A). The CSD map showed strong sinks in both was also observed in the st. lacunosum of the CA1 region.
molecular layers (sk1, sk2) and a weaker sink (sk3) produced The ICA of the LFPs yielded three significant LFP-genera-
by CA1 pyramidal cells in the st. lacunosum, in close corre- tors, G1–G3, with little if any cross-contamination (Fig. 5C).
spondence to the known anatomical terminations of these Noticeably, the time course of G1 activation shows increasing
fibers. Passive sources (return currents) appeared in the GCL pulses in agreement with the increasing amplitude of the PP
(cartoon in Fig. 4A, right). The ICA of LFPs again rendered stimuli. In addition, the beginning of the G1 pulses coincides
only 1 significant LFP-generator, G1 (activation in Fig. 4Ba with the onset of the stimuli in this pathway. The other two
and loading in Fig. 4Bd), whereas the other 15 were negligible LFP-generators, G2 and G3, exhibit pulses of constant ampli-
(Fig. 4Bb). The CSD map of the virtual LFPs reconstructed for tude (as for the Schaffer stimulation), whose onset coincide
this unique generator (Eq. 3) matches the CSD of the raw with the beginning of the stimulation to the Schaffer input.
signal (Figs. 4, Bc vs. A). Likewise the voltage and CSD Thus the LFP-generator G1 exclusively represents the activity
loading curves (Fig. 4Bd) match the spatial profiles for the raw in the molecular layer of the DG (cf. Fig. 4), whereas the
J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org
490 KOROVAICHUK ET AL.

A B a b 1.0
explained variance
FIG. 4. Precise spatial and temporal performance of the
ICA on LFPs produced during subthreshold activation of the
e28 entorhinal input to the dentate gyrus (DG). A: an electric shock
0.5 in the ipsilateral perforant path (PP) evoked a negative poten-
1 mV 1 (a.u.)
G1 0.0
tial in both molecular layers of the DG. A sample record from
G1 G2-G16 electrode 13 of a 32-site probe is shown in a. The raw CSD (b)
raw CSD c CSD for G1 d CSD loadings volt. loadings shows active sinks in both molecular layers (sk1 and sk2) and
2 a weaker sink in the st. lacunosum of the CA1 (sk3). Passive
2
CA1
sources can be appreciated in the granule cell layers (GCL).
sk3 The schematic neurons show the approximate site of record-
electrode #

10
ings. B: the ICA applied to the LFP profile shows a unique
sk1 significant component (b) with a temporal course matching the
GCL 0 voltage change in the molecular layer (a). The CSD calculated
DG

20 for the reconstructed LFP for G1 (c) is similar to the raw CSD
GCL (Ab). The voltage and CSD loadings (d) coincide with the
sk2 spatial profiles of the raw LFP and CSD. Note that a unique
30 -2 ICA component captures coherent activity in physically sepa-
10 ms -0.8 0 0.8 0 2 4 rated cell generators as long as they are synaptically activated
(a.u.) (a.u) by a common input.

CA3-evoked LFPs were decomposed into two generators: G2 Schaffer and PP population activity in ongoing LFPs
responsible for the Schaffer-mediated CA1 excitation (cf. Fig.
We showed that ICA can separate Schaffer and PP evoked

Downloaded from jn.physiology.org on September 3, 2010

3) and G3 corresponding to the local CA3/DG activation. The activities mixed in LFPs, and we described their spatiotempo-
CSD maps of the virtual LFPs reconstructed from each of the ral characteristics. Neither of the spatial weight curves (load-
isolated generators (Fig. 5D) also confirm the pathway and ings) for these pathways matched those found in our former
population specific nature of the LFP-generators. Therefore study, which described the most powerful generators in spon-
ICA provides an accurate decomposition of multiple inputs that taneous LFPs (Makarov et al. 2010). This observation suggests
are not temporally coincident. that the LFP-generators describing the activity induced by the
excitatory pathway may contribute weakly to the spontaneous
COINCIDENT STIMULI. In principle, the tight temporal coinci-
LFPs. Thus we sought after the LFP-generators for the Schaf-
dence of multiple sources decreases the efficiency of ICA to fer and PP inputs in ongoing LFPs by mixing in subthreshold
separate them. Thus we checked its performance on two stimuli that facilitate the identification of the generators.
precisely synchronized different inputs. When the stimuli were
delivered simultaneously to the Schaffer and perforant path- IDENTIFICATION OF SCHAFFER AND PP POPULATION ACTIVITIES IN
ways (Fig. 5, right column), there seemed to be a complex ONGOING LFPS. A fragment of ongoing LFPs was analyzed
mixture of the sources and sinks in the raw CSD map. Notably, during which both pathways were activated in a random-like
the proximity of the CA3 and GC layers led to fusion of their manner at an average rate of 1 Hz each (Fig. 6A). The vertical
dashed red and blue lines mark the onset of the Schaffer and PP
respective sinks into a unique slightly drifting sink (Fig. 5B),
stimuli, respectively. Because of the weak intensity of the
which made the sinks indistinguishable and complicated its
electrical stimulation, the evoked responses are hardly discern-
direct interpretation. ible in the raw LFPs. By applying an ICA, we found five main
Even in this case, the ICA still rendered three significant LFP-generators (Fig. 6, B and C), two of which exclusively
LFP-generators. Each of the generators had voltage loadings contain the evoked responses to the Schaffer (G2) and PP (G4)
similar to those found in the noncoincident case (Fig. 5C, left stimuli (Fig. 6C). We noticed that the activations of the other
vs. right), although their separation was less effective. As such, LFP-generators (G1, G3, and G5) did not correlate with the
the strongest LFP-generator, G1, responsible for the PP input, onset of the stimulus, indicating that there was little if any
introduced some contamination into the other two. This cross- cross-contamination between these generators. The remarkable
contamination could be appreciated by comparing the loading specificity of the activities contained within G2 and G4 (Fig.
curves in the left and right columns. We noticed that in the 6B) was proof that both are truly independent and well-isolated
lower part of the loading curve (DG/hilus) for the Schaffer G2 LFP-generators responsible for the Schaffer and PP inputs,
(arrow pointing to red curve), there was a spatial weighting respectively. The shape of the voltage loadings (spatial weight
analogous (disregard the polarity) to that of the PP-specific G1 curves) for G2 and G4 are essentially identical to those found
generator (arrow pointing to blue curve). The CSD maps for in experiments where only evoked activity was assessed (cf.
the reconstructed LFPs (Fig. 5D) show that the Schaffer- Fig. 6C, red and blue curves to Figs. 3Bc and 4Bd, respectively,
specific G2 generator contains some current belonging to the and also to Fig. 5C, red and blue curves). The portion of the
CA3 and FD regions (left vs. right columns in Fig. 5D). variance explained by the two primed pathways (bars in red
Nevertheless, even in this extreme case of the perfect time and blue) is small (⬍5%; Fig. 6D). This reduced contribution
coincidence of different inputs, the specific activations of the of the Schaffer and perforant pathways to LFPs makes their
LFP-generators confirm an adequate separation. Indeed, only identification in ongoing spontaneous LFPs extremely difficult.
G1 exhibits decaying amplitude of activation in agreement As such, LFP-generators containing their activities fall below
with the decaying intensity of the stimulus applied to the PP the level of significance and hence they can be overlooked.
input, whereas G2 has stationary amplitude in agreement with RECONSTRUCTION OF PATHWAY-SPECIFIC LFPS AND THEIR SUBCEL-
constant stimulation of the Schaffer pathway. All these results LULAR SPECIFICATION. Disentangling the raw LFPs into LFP-
were reproduced in four additional animals. generators enables the reconstruction of virtual LFPs produced
J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org
 EXCITATORY GENERATORS OF HIPPOCAMPAL LFPS 491

by a specific pathway during ongoing activity and their sepa- generators and comparing their activations and voltage load-
rate analysis. An example is shown in Fig. 7 for an LFP ings to those shown in Fig. 6, B and C, respectively, we
fragment containing three evoked (subthreshold) pulses (ap- identified G2 and G4 as responsible for the Schaffer and
plied to Sch or PP pathways; the onset of each is marked by a perforant inputs, respectively. Because of the ambiguity of the
dashed line in Fig. 7A). The LFP was recorded from the ICA (see METHODS), the time course of the activations is given
pyramidal/oriens border in CA1 to the lower blade of the DG. in arbitrary units and is not directly comparable. However, we
The corresponding CSD map shows a confusion of sources and can reconstruct the virtual LFPs created by each single gener-
sinks produced by the ongoing spatiotemporal activity of ator. Figure 7D shows LFPs reconstructed for the Schaffer
different origins (Fig. 7B). generator (G2). The spontaneous population activity and two
The ICA of the LFP fragment rendered five LFP-generators of the evoked events can be tracked down in the raw signal.
(Fig. 7C), and their spatial loadings and time course of activa- The reconstructed evoked potential profile shows no difference
tion can now be used to identify pathway-specific evoked to its raw counterpart, as expected (see the ellipses in Fig. 7, A
activity. By searching for the evoked responses among all and D, amplified in the right insets). By visual inspection of the
raw and Schaffer reconstructed LFPs, it is patent that ongoing
A non-coincident coincident raw LFPs are most prominent in the hilus of the DG (lower
1
electrodes), whereas Schaffer mediated activity is more con-
spicuous in the CA1. Also some activity in the CA3/DG region
10
electrode #

that is probably caused by recurrent excitation of presynaptic

pyramidal cells in the CA3 region.

Downloaded from jn.physiology.org on September 3, 2010

20 As further confirmation of the specificity of the Schaffer
generator (G2) identified, we compared its activation (Fig.
5 mV

30 7B, G2) with the high-frequency activity (⬃100 –200 Hz)

. . . . . . . .* * . . . . . . . . . . . . . . . . . . .
0.2 s
* associated to sharp waves, an electrographic event known to
B
0.05 s
be caused by synchronous activation of the CA3 region (i.e.,
PP raw CSD Sch raw CSD PP + Sch raw CSD
2 x10
-4
Schaffer-mediated: Buzsáki et al. 1983). Indeed, the specific
8
bouts of activity contained within the Schaffer G2 generator
CA1

10 were coherent with ripple-like fast activity in the CA1

electrode #

sk3 sk6
sk2 0
perisomatic LFPs (black top trace in Fig. 7B; band-pass:
20 sk1 100 –500 Hz).
sk4 sk5
GCL To obtain the precise localization and magnitude of the
DG

30
H -8 underlying transmembrane currents, we evaluated the CSD of
GCL
5 ms 5 ms the reconstructed LFPs (Fig. 7E). As expected, the Schaffer
C activations activations population activity presented a main sink in the st. radiatum
G1 surrounded by passive sources (return currents) in the st.
G2 pyramidale and lacunosum, matching the locations of the
G3
. .. .. .. .* *
. .. .. .. .. . . . . . . * . . . . well-known CSD map for evoked Schaffer activity (Fig. 3).
2 (a.u.) 1 (a.u.)
The CSD maps of isolated pathway-specific LFPs contrasted
volt. loadings CSD loadings volt. loadings CSD loadings
1 with those obtained for the mixed signal (cf. Fig. 7, B and E).
It is worth mentioning that the temporal activation of isolated
electrode #

10 components corresponds to that of the population synaptic

currents (e.g., compare the evolution of the CSD map at
20

30
FIG. 5. Discrimination of multiple evoked synaptic inputs in mixed LFPs.
0 2 4 -0.4 0 0.4 -4 0 4 8 -1 0 1 2
(a.u.) (a.u.) (a.u.) (a.u.)
Subthreshold stimuli were applied to both the ipsilateral CA3 (Schaffer) and
the PP, either as paired pulses (100 ms delay, left column) or synchronously
D (right column). PP stimuli were modulated in intensity, whereas CA3 stimuli
2
G1 (PP) G1 (PP) were constant. A: raw LFPs recorded along a track spanning the CA1 and
10
CA3/DG up to the lower GCL. B: CSD maps for a sample response (raw CSD)
sk1
to either the PP, the CA3 (Sch), or both. Schematic neurons between panels
sk1
20 sk2 sk2 show the approximate locus of recording. PP produced a main sink in the st.
moleculare (sk1) and a weaker sink in the CA1 st. lacunosum (sk2). CA3
30 activation raised a sink in the CA1 st. radiatum (sk3) and a local sink (sk4).
2
G2 (Sch) G2 (Sch) Combined stimuli produced a complex sink in the DG (sk5) and the Schaffer
electrode #

10 sk3 sk3 sink in CA1 (sk6). C: the ICA yielded 3 significant generators in both cases. G1
and G2 correspond to the PP and Schaffer generators, respectively, whereas G3
20
corresponds to local activity in the CA3/DG. The top traces show the time
30
activation and the bottom traces show the corresponding spatial voltage and
-4
2 x10 CSD loadings. Note that activations vary in G1 with successive stimuli,
G3 (CA3) 4 G3 (CA3) whereas they remain constant in G2, precisely reflecting the intensity applied
10
in each case. The spatial loadings for G1 and G2 were identical as when each
0 pathway was stimulated alone. A small contaminant appeared during coinci-
20
sk4 sk4 dent stimulation (small arrows). D: virtual CSD maps for sample responses of
30 -4 each separate generator. Note the spatial coincidence of the separate virtual
5 ms 5 ms sinks with those in the raw LFPs.

J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org

492 KOROVAICHUK ET AL.

A raw LFPs 1s
C
5 mV voltage loadings
1

CA1
electrode #

10

20 GCL

CA3/H
DG
30 GCL
-0.2 0 0.2 0.4
B (a.u.)
D explained variance
G1 0.5

Schaffer
0.4
G2

Downloaded from jn.physiology.org on September 3, 2010

0.3
G3
0.2
Perforant
G4
0.1

G5 0
G1 - G5 G6 - G11
40 ms 2 (a.u)
FIG. 6. Identification of the Schaffer and PP population activities in ongoing LFPs. A: segment of raw LFPs recorded from the CA1 st. pyramidale to the lower
GCL in the DG. LFPs were primed with subthreshold stimuli applied in a random manner to the CA3 (Sch) and PP. The vertical dashed lines mark the timing
of stimulation (blue for PP and red for CA3). Evoked potentials are short and small, and hence, they do not introduce much variance in the raw signals (they
are hardly discernible over this time scale). B–D: the ICA yielded 5 principal generators, 2 of which specifically contain the activation corresponding to the
evoked potentials for the Schaffer (G2) and the PP (G4). B: an amplification of the responses after each stimulus. Note the pathway specific segregation of evoked
responses in the corresponding generators. The spatial loadings shown in C match those found for nonprimed LFPs (G1, G3, G5) or for separately stimulated
pathways (G2, G4). D: relative variance of the generators isolated.

electrodes 8 –9 in Fig. 7E to the corresponding activation in sponding LFP-generators (see METHODS and Eqs. 3 and 5). The
Fig. 7C, red trace). Thus LFP-generators isolated by the ICA Schaffer and PP inputs produce 0.045 and 0.031 mV2 on
enable the population-specific transmembrane current density average (7 and 5% of the total power), respectively (Fig. 8C),
to be examined, rejecting contamination of the other inputs in whereas the contributions of the other generators ranged be-
ongoing LFPs. tween 0.13 and 0.24 mV2 in four animals. This result corrob-
orates the relatively small contribution of the excitatory path-
MINOR CONTRIBUTION OF EXCITATORY PATHWAYS TO HIPPOCAM-
ways into the total LFP variance observed above (Fig. 6D), but
PAL LFPS. The Schaffer and PP generators (G2 and G4) iden-
now excluding the evoked activity. We also noticed that the
tified that describe the activation of the corresponding path- powers of the Schaffer and PP generators have smaller SE
ways enabled the spontaneous activity of these pathways to be relative to the powers of G1, G3, and G5, which suggests that
studied and quantified (i.e., the presynaptic CA3 and entorhinal the Schaffer and PP inputs described are more stationary in
afferent populations, respectively). In general, we found that time and between different animals than the other three
the spontaneous activity (during irregular LFPs) of both the LFP-generators.
Schaffer and PP generators represents an irregular pattern of
short bouts over a rather flat baseline. These can be appreciated
DISCUSSION
in a sample epoch in Fig. 8A along with their respective
wavelet spectra. They lasted a few hundred milliseconds, had Studying the mechanisms of neural information processing
variable amplitude and occurred at ⬍1 Hz rate. Smaller and requires certain knowledge of the circuitry and of the causal
shorter events can also be appreciated, which may be physio- and functional associations between activities of different neu-
logically relevant. The main frequency content of the activity ronal groups. Here, we showed that an ICA can effectively
concentrates in the 0.5– 4 Hz range for both generators. To separate standard LFPs in the hippocampus into pathway-
track the evolution of the generator’s power over time, we specific isolated activities, called LFP-generators, describing
calculated time envelopes of this activity (see METHODS and Eq. the inputs of multiple presynaptic populations. The main req-
6), where large events and baseline activity can be better uisite for the separation is the different spatial distribution of
distinguished (Fig. 8B). transmembrane currents along the somatodendritic axis of
The dimensional power (measured in mV2) that contributed principal cells, whereas the temporal pattern of the pathway’s
to the ongoing LFPs for each synaptic pathway can be esti- activation is less important (e.g., they can be rhythmic or
mated from the virtual LFPs reconstructed from the corre- irregular).
J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org
 EXCITATORY GENERATORS OF HIPPOCAMPAL LFPS 493

Using selective subthreshold electrical activation of major found were used to quantify the pathway-specific contribution
excitatory synaptic pathways, such as the Schaffer and PP to ongoing LFPs. Hence we have consistently shown in all
inputs, we described spatiotemporal features of LFP-generators studied animals that the contribution of the major excitatory
specific to each. Subsequently, the Schaffer and PP generators pathways to the power of ongoing LFP signals is minor (⬃7
and 5%, respectively). We also showed that both Schaffer and
A raw LFPs 0.2 s 20 ms perforant pathways have irregular temporal patterns, with
1
bouts of population activity of varying amplitude and with the
frequency concentrated in the 0.5– 4 Hz range.

CA1
electrode #

10 *
Methodological considerations
20 A critical consideration regarding the ICA is that the com-
GCL ponents found may have little functional relevance unless
CA3/H
accurate information on their anatomical basis is provided. All
30 the same, ICA can efficiently separate signals coming from
GCL 2 mV distant point sources (Choi et al. 2005). There are few studies
5 mV
on ICA of deep brain recordings (Makarov et al. 2010; Tans-
B 2
raw CSD
0.25 kanen et al. 2005). In deep recordings, the signals generated in
the volume surrounding the electrode array (i.e., the sources or
electrode #

LFP-generators) are essentially extended spatially, and their

Downloaded from jn.physiology.org on September 3, 2010

10
spatial scales may be larger than the interelectrode distance.
0
20 However, the same mathematical principles are applicable to
LFP recordings (Makarov et al. 2010). Disentangling such
LFP-generators is based 1) on the coincident activation of
30 -0.25 individual homologous fibers over the millisecond time scale
and 2) on their different spatial distributions. The first condi-
LFP [100-500 Hz]
0.1 mV

C tion provides the necessary synchronization of homologous

axons to raise significant macroscopic synaptic field potentials,
the main contributors to normal LFPs (Mitzdorf 1985; Nunez
activations volt. loadings and Srinavasan 2006). The second condition is ensured by the
G1 distinct distributions of afferent axons from different afferent
Schaffer populations. The accuracy of the spatial resolution is therefore
G2 optimized to the subcellular level of the active neurons. We
showed that the spatial CSD-loading curves of the Schaffer and
G3
PP LFP-generators match the configuration of the inward and
Perforant outward currents for known membrane events (Andersen et al.
G4
1971; Herreras 1990; Kandel et al. 1961). Admittedly, some
factors can modify the expected spatial pattern of field poten-
G5 -0.4 0 0.4 tials, such as the heterogeneous electrical properties of the
5(a.u.) (a.u.) conducting medium, spatial cancellation, and frequency scal-
D virtual LFPs reconstructed from G2 ing of currents in the volume conductor (Bédard et al. 2004;
1 López-Aguado et al. 2001; Makarova et al. 2008). In earlier
electrode #

10 FIG. 7. Characterization of pathway-specific activity in ongoing LFPs.

A: segment of raw LFPs recorded from the CA1 st. pyramidale to the lower GCL
in the DG. The CA3 and the PP were stimulated to prime LFPs with Schaffer and
20 PP evoked activity (instants marked by red and blue dashed lines, respectively).
The ellipse marks a Schaffer evoked response enlarged to the right. Note the
subthreshold field excitatory postsynaptic potential (fEPSP) in the st. radiatum
30 (electrodes 3–10, asterisk). B: CSD map estimated for the raw LFPs. Note the
5 mV 2 mV complex mixture of currents. The Schaffer sink in the evoked response is appre-
ciated in the amplification on the right (small arrow). C: ongoing activities of the
E CSD for Schaffer LFP-generator (G2) 5 main generators separated by the ICA. G2 and G4 correspond to the Schaffer and
2 0.25 PP generators, respectively. Note the characteristic spatial voltage loadings on the
right. The top (black) trace corresponds to the filtered (100 –500 Hz band-pass)
LFP recorded from the CA1 soma layer. Note that epochs of fast oscillations
electrode #

10 coincide with the bouts of activity in the Schaffer generator (G2), matching the
features of sharp wave events. D: reconstruction of the Schaffer-specific LFP. This
0 virtual LFP now contains only the virtual LFP corresponding to ongoing activation
20 of the CA3 to CA1 input. The amplification on the right shows the virtual fEPSP
for a sample Schaffer response. Note that all evoked and spontaneous activity
undergo changes of identical polarity for a given electrode. E: virtual CSD map for
the ongoing Schaffer activity. Schaffer evoked sinks and sources are now like the
30 -0.25 standard evoked responses (compare. with the raw CSD in B).

J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org

494 KOROVAICHUK ET AL.

A Schaffer (G2) Perforant (G4) C

5 (a.u.)

0.4
frequency (Hz)

32

8 0.3

Power (mV2)
2
0.2

Perforant
Schaffer
0.5
5s
B 0.1
2 (a.u.) 0.2 (a.u)
0.0
G1 G2 G3 G4 G5
FIG. 8. Minor contribution of the excitatory generators to the hippocampal LFPs. A: frequency content of the ongoing Schaffer (G2) and PP (G4) generators
during irregular activity in the raw LFP. The spectrogram shows dominant activity ⬍2–3 Hz. The white trace shows the temporal activation of the generators
(amplified above). The specific activity of spontaneous pathways was in both cases sparse and highly irregular. B: time envelopes of activity (1 s time window).
C: average power of reconstructed LFPs for specific generators in 6 animals (mean ⫾ SE) during irregular activity in the raw LFP. Note the small global
contribution of excitatory generators.

Downloaded from jn.physiology.org on September 3, 2010

experimental and modeling work, we reported some particular tions (Makarov et al. 2010), we found that mixing sparse sub-
cases where some of these may be relevant (López-Aguado et threshold evoked potentials in LFPs can be used to resolve the
al. 2001, 2002; Makarova et al. 2008; Varona et al. 2000). Such problem. The presence of evoked pulses in the time course of an
studies cannot easily be extrapolated on the distribution of LFP-generator is related to the stimulated neuronal population,
LFP-generators. Because this analysis was performed on lo- and it helps identify the correct polarity. The subsequent applica-
cally recorded LFPs, we speculate that the factor potentially tion of CSD analysis on virtual LFPs reconstructed from a single
most relevant is the mutual spatial cancellation between indi- generator enables the resultant partial CSD map to be directly
vidual synaptic inputs when these are scattered through ex- compared with the raw CSD and with CSD maps of well-known
tended dendritic domains (Makarova et al. 2010). The excita- evoked responses of specific pathways. It is worth noting that
tory Schaffer and PP inputs are strongly stratified; hence we CSD map of isolated generators show no alternations in terms of
would not expect a strong bias introduced by this factor either. the current’s sinks and sources, as expected for a specific synaptic
This approach is not dependent on the presence of rhythmic input. In contrast, raw CSD maps may contain such alternations
activity or stereotyped LFP events, and hence it can be applied to because of the spatiotemporal mixing of the activities of several
a wide variety of behaviors. Indeed, all the examples shown in this generators and DC filtering (Brankačk et al. 1993).
report were chosen among epochs of irregular activity, whose
behavioral correlates are very rich (Jarosiewicz and Skaggs 2004). When and what can an ICA discriminate in LFPs?
In neuroscience, ICA is frequently used to suppress artifacts in
EEG recordings (Castellanos and Makarov 2006). We showed To correctly interpret LFP-generators isolated by an ICA, it
that artifacts caused by electric shocks within brain tissue can also is relevant to ask whether each of them corresponds to postsyn-
be isolated and suppressed by the ICA. However, strong artifacts aptic activity from one or several neuron types. This problem
(ⱕ100 times higher than the neuronal activity) decrease the is simplified greatly by the use of linear probes in the hip-
efficiency of the ICA for the simultaneous separation of other pocampus where each subfield has only one type of principal
mixed components of neuronal origin. Hence we suggest that cell: namely pyramidal and granule cells. Each synaptic input
artifacts should be removed before performing an ICA. produces loops of inward/outward currents with a different
The weak contribution of the Schaffer and perforant pathways spatial distribution within the same core conductor. Typically
to ongoing LFPs makes their identification extremely difficult the maximum potential is attained at the center of the synaptic
(i.e., LFP-generators can be overlooked). Therefore we used contact made by the isolated pathway (Herreras 1990). How-
selective low-magnitude electrical stimulation of the known ex- ever, distant recording sites also contain activity, albeit of
citatory pathways to ensure the reliable identification of LFP- reduced magnitude. Therefore steadily correlated recording
generators describing the activity of these pathways. Thus the sites correspond to anatomical subdomains of one type of
identification was based on 1) a comparison of spatial voltage principal cell, whereas the subcellular precision can be
loadings of the LFP-generators found for evoked responses with achieved through a CSD analysis of the distribution of spatial
those found for ongoing LFPs and 2) the recognition of evoked weights in isolated generators. This procedure enables the
pulses in the time courses of LFP-generators obtained by ICA of separation of inputs into a single population according to
composite epochs (i.e., a recording consisting of 2 epochs: spon- known stratification along the dendritic arbor.
taneous LFPs and spontaneous plus a few evoked LFPs). However, when linear recording probes extend through several
The polarity of the CSD loading is essential to define whether subfields (e.g., CA1 and DG/CA3), the interpretation may not be
an LFP-generator describes excitatory or inhibitory input. How- that trivial because of volume propagation of currents and/or the
ever, the ICA cannot unequivocally determine this feature simultaneous activation of populations in both subfields. The
(Hyvärinen and Oja 2000). Although this problem is hard to solve electrical fields created by two physically separated cells extend
for spontaneous LFPs and requires special biophysical assump- beyond their respective physical limits, mixing in local record-
J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org
 EXCITATORY GENERATORS OF HIPPOCAMPAL LFPS 495

ings. This is not a handicap for this approach because the temporal recorded from membranes of cortical pyramidal cells (Rudolph et
activation for each generator has no spatial dimension (i.e., it al. 2005; see also Anderson et al. 2000; Borg-Graham et al. 1998).
contains the coherent activity of a group of electrodes regardless Besides the firing rate of several types of interneurons in the
of their location). Accordingly, the CSD examination shows hippocampus is much higher than that of pyramidal cells (Somo-
whether the activity can be circumscribed to one or several gyi and Klausberger 2005). Along with the functional coordina-
physically separated cell generators. In this study, we found that tion of some interneuron networks (Whittington et al. 1995), the
the two excitatory generators seem to have true activity in more hypothesis that inhibition prevails over excitation in hippocampal
than one region. Thus the activity of the entorhino-hippocampal LFPs becomes more plausible. Nonetheless, the precise percent-
pathway (PP) was contained in only one LFP-generator, but it was age of excitatory contribution to LFPs found here may vary in the
divided between both blades of the DG. The parsimonious expla- freely moving animal, because anesthesia modifies the overall
nation is that the presynaptic population activates two or more pattern of LFPs (Brankačk et al. 1993; Buzsáki et al. 1986).
physically separate target populations through bifurcating axons, In any case, the results presented here point to a very low output
in agreement with anatomical and electrophysiological findings rate of CA3 pyramidal cells during irregular LFPs, in agreement
(Hjorth-Simonsen and Jeune 1972; Lømo 1971). Hence, activa- with previous data (Ranck 1973; Thompson and Best 1989). In
tion is identical and simultaneous in both blades, and thus it can be terms of the PP generator, the overall behavior was similar
described by a single LFP-generator isolated by the ICA. With (occasional synchronous bouts of irregular activity), again indi-
regard to the Schaffer generator, although CA3 activation ren- cating a low output rate from entorhino-hippocampal projection
dered a single component in CA1 whose spatial loading curve cells during irregular patterns of LFP. The presynaptic origin of
matches the termination of Schaffer collaterals (Herreras 1990), this generator is suggested by the limited evoked responses after

Downloaded from jn.physiology.org on September 3, 2010

prominent activity was also obtained near the stimulating PP stimulation. The identity of the entorhinal projecting cells is
electrode in the CA3/DG. The origin of this local activity not clearly established in the literature, and difficulties may arise
varied somewhat in different experiments and was most from the fact that the entorhino-hippocampal projection is multi-
commonly found in the molecular layers of the DG for farious. Unitary studies have not clarified this issue, because most
evoked responses, probably because of backfiring of the PP interest has focused on the relationship of entorhinal units with
fibers (Wu et al. 1998). During spontaneous LFPs, the theta (Alonso and Llinás 1989; Chrobak and Buzsaki 1994;
CA3/DG activity associated to the Schaffer generator is less Frank et al. 2001). Thus stellate and pyramidal-like cells have
prominent, and it most likely reflects the synchronous ac- been deemed candidates for this projection. Indeed, interesting
tivity of CA3 pyramidal cells because of recurrent excitation functional data supporting the identity of the PP generator
in this area (Miles and Wong 1987). Thus the CA3 pyrami- shown here is that the bouts of activity are not related to
dal population is presynaptic for both CA1 and itself, and hippocampal sharp waves (Chrobak and Buzsaki 1994), which
therefore the ongoing activity in the Schaffer generator is consistent with the poor temporal correspondence between
should appear simultaneously in both regions, as happens to bouts of activity in the Schaffer and PP generators.
be the case. To conclude, we propose the combined use of ICA and CSD
as a high-resolution method for the subcellular identification of
Excitatory currents contribute little to hippocampal LFPs LFP components in regular structures and for the quantification
of their ongoing activity.
Some LFP events and rhythmic oscillations have been studied
intensively, and the neuronal circuits producing them have there- ACKNOWLEDGMENTS
fore been clarified. For example, sharp hippocampal waves are
We thank M. Sefton at BiomedRed for editorial support, S. Canals for
known to be mediated by the excitatory input from CA3 to CA1 critical reading and comments, and L. López-Mascaraque and N. Salvador for
populations (Buzsáki et al. 1983). Here, we also observed a help with histological processing.
correlation between sharp waves and the ICA isolated Schaffer
generator describing the synaptic activity produced by CA3 py-
GRANTS
ramidal cells in CA1. Some LFP events, such as the theta rhythm,
have a mainly inhibitory origin (Soltesz and Deschenes 1993), This work was supported by Spanish Ministry of Education and Science
Grants BFU2007-66621/BFI and FIS2007-65173 and Comunidad Autónoma
whereas others are still under study, such as gamma oscillations de Madrid Grant S-SEM-0255-2006.
(Csicsvari et al. 2003; Mann et al. 2005), fast ripples (Ylinen et al.
1995), and dentate spikes (Bragin et al. 1995). In light of the
proportion of excitatory and inhibitory neurons in the hippocam- DISCLOSURES
pus (Aika et al. 1994), the small contribution of the major No conflicts of interest, financial or otherwise, are declared by the author(s).
excitatory pathways to hippocampal LFPs might seem unexpected
(Schaffer and PP generators explain 7 and 5% of the power, REFERENCES
respectively). The other three LFP-generators show a significant
Aika Y, Ren JQ, Kosaka K, Kosaka T. Quantitative analysis of GABA-like-
phase-spike relationship, with ⬃60% of the putative interneurons immunoreactive and parvalbumin-containing neurons in the CA1 region of
in CA1 (Makarov et al. 2010), suggestive of their inhibitory nature the rat hippocampus using a stereological method, the dissector. Exp Brain
(i.e., the corresponding synaptic currents are produced by local Res 99: 267–276, 1994.
inhibitory circuits). Circumstantial data in the literature argue in Alonso A, Llinás RR. Subthreshold Na⫹-dependent theta-like rhythmicity in
favor of the dominance of inhibitory over excitatory currents in stellate cells of entorhinal cortex layer II. Nature 342: 175–177, 1989.
Amari S, Cichocki A, Yang H. A new learning algorithm for blind source
ongoing population activity (Hajos et al. 2000; Herreras et al. separation. Adv Neural Inform Process Syst 8: 757–763, 1996.
1988; Oren et al. 2010), although direct evidence is scarce. The Andersen P, Bliss TVP, Skrede KK. Lamellar organization of hippocampal
dominance of inhibitory currents was found in synaptic noise excitatory pathways. Exp Brain Res 13: 222–238, 1971.

J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org

496 KOROVAICHUK ET AL.

Anderson JS, Carandini M, Ferster D. Orientation tuning of input conduc- Kandel ER, Spencer WA, Brinley FJ Jr. Electrophysiology of hippocampal
tance, excitation and inhibition in cat primary visual cortex. J Neurophysiol neurons. I. Sequential invasion and synaptic organization. J Neurophysiol
84: 909 –926, 2000. 24: 225–242, 1961.
Bédard C, Kröger H, Destexhe A. Modeling extracellular field potentials and Kocsis B, Bragin A, Buzsáki G. Interdependence of multiple theta generators
the frequency-filtering properties of extracellular space. Biophys J 86: in the hippocampus: a partial coherence analysis. J Neurosci 19: 6200 –
1829 –1842, 2004. 6212, 1999.
Bell A, Sejnowski T. An information-maximization approach to blind sepa- Lee T, Girolomi M, Sejnowski T. Independent component analysis using an
ration and blind deconvolution. Neural Comput 7: 1129 –1159, 1995. extended infomax algorithm for mixed subgaussian and supergaussian
Borg-Graham LJ, Monier C, Frégnac Y. Visual input evokes transient and sources. Neural Comput 11: 417– 441, 1999.
strong shunting inhibition in visual cortical neurons. Nature 393: 369 –373, Leung LS, Roth L, Canning KJ. Entorhinal inputs to hippocampal CA1 and
1998. dentate gyrus in the rat: a current-source-density study. J Neurophysiol 73:
Bragin A, Jandó G, Nádasdy Z, van Landeghem M, Buzsáki G. Dentate 2392–2403, 1995.
EEG spikes and associated interneuronal population bursts in the hippocam- Lømo T. Patterns of activation in a monosynaptic cortical pathway: the
pal hilar region of the rat. J Neurophysiol 73: 1691–1705, 1995. perforant path input to the dentate area of the hippocampal formation. Exp
Brankačk J, Stewart M, Fox SE. Current source density analysis of the Brain Res 12: 18 – 45, 1971.
hippocampal theta rhythm: associated sustained potentials and candidate López-Aguado L, Ibarz JM, Herreras O. Activity-dependent changes of
synaptic generators. Brain Res 615: 310 –327, 1993. tissue resistivity in the CA1 region in vivo are layer-specific: modulation of
Buzsáki G, Czopf J, Kondákor I, Kellényi L. Laminar distribution of evoked potentials. Neuroscience 108: 249 –262, 2001.
hippocampal rhythmic slow activity (RSA) in the behaving rat: current- López-Aguado L, Ibarz JM, Varona P, Herreras O. Structural inhomoge-
source density analysis, effects of urethane and atropine. Brain Res 365: neities differentially modulate action currents and population spikes initiated
125–137, 1986. in the axon or dendrites. J Neurophysiol 88: 2809 –2820, 2002.
Buzsáki G, Leung LS, Vanderwolf CH. Cellular bases of hippocampal EEG Lorente de Nó R. Analysis of the distribution of action currents of nerves in

Downloaded from jn.physiology.org on September 3, 2010

in the behaving rat. Brain Res 287: 139 –171, 1983. volume conductors. In a study of nerve physiology. Stud Rockefeller Inst
Canals S, López-Aguado L, Herreras O. Synaptically-recruited apical currents Med Res Repr 132: 384 – 477, 1947.
are required to initiate axonal and apical spikes in hippocampal pyramidal Makarov VA, Makarova J, Herreras O. Disentanglement of local field
cells: modulation by inhibition. J Neurophysiol 93: 909 –918, 2005. potential sources by independent component analysis. J Comput Neurosci In
Castellanos NP, Makarov VA. Recovering EEG brain signals: artifact sup- press.
pression with wavelet enhanced independent component analysis. J Neuro- Makarova J, Gómez-Galán M, Herreras O. Layer specific changes in tissue
sci Methods 158: 300 –312, 2006. resistivity and spatial cancellation of transmembrane currents shape the
Choi S, Cichocki A, Park HM, Lee SY. Blind source separation and voltage signal during spreading depression in the CA1 vivo. Eur J Neurosci
independent component analysis: a review. Neural Inform Process Lett Rev 27: 444 – 456, 2008.
6: 1–57, 2005. Makarova J, Makarov VA, Herreras O. A model of sustained field poten-
Chrobak JJ, Buzsaki G. Selective activation of deep layer (V-VI) retrohip- tials based on gradients of polarization in single neurons. J Neurophysiol
pocampal cortical neurons during hippocampal sharp waves in the behaving 103: 2446 –2457, 2010.
rat. J Neurosci 14: 6160 – 6170, 1994. Makeig S, Debener S, Onton J, Delorme A. Mining event-related brain
Csicsvari J, Jamieson B, Wise KD, Buzsáki G. Mechanisms of gamma dynamics. Trends Cogn Sci 8: 204 –210, 2004.
oscillations in the hippocampus of the behaving rat. Neuron 37: 311–322,
Mann EO, Radcliffe CA, Paulsen O. Hippocampal gamma-frequency oscil-
2003.
lations: from interneurones to pyramidal cells, and back. J Physiol 562:
Delorme A, Makeig S. EEGLAB: an open source toolbox for analysis of
single trial EEG dynamics including independent component analysis. J 55– 63, 2005.
Neurosci Methods 134: 9 –21, 2004. Miles R, Wong RK. Inhibitory control of local excitatory circuits in the
Einevoll GT, Pettersen KH, Devor A, Ulbert I, Halgren E, Dale AM. guinea-pig hippocampus. J Physiol 388: 611– 629, 1987.
Laminar population analysis: estimating firing rates and evoked synaptic Mitzdorf U. Current source-density method and application in cat cerebral
activity from multielectrode recordings in rat barrel cortex. J Neurophysiol cortex: investigation of evoked potentials and EEG phenomena. Physiol Rev
97: 2174 –2190, 2007. 65: 37–100, 1985.
Frank LM, Brown EN, Wilson MA. A comparison of the firing properties of Nunez PL, Srinivasan R. Electric Fields of the Brain: The Neurophysics of
putative excitatory and inhibitory neurons from CA1 and the entorhinal EEG (2nd ed.). New York: Oxford, 2006.
cortex. J Neurophysiol 86: 2029 –2040, 2001. Oren I, Hajos N, Paulsen O. Identification of the current generator underlying
Freeman JA, Nicholson C. Experimental optimization of current source-density cholinergically induced gamma frequency field potential oscillations in the
technique for anuran cerebellum. J Neurophysiol 38: 369 –382, 1975. hippocampal CA3 region. J Physiol 588: 785–797, 2010.
Hajos N, Katona I, Naiem SS, MacKie K, Ledent C, Mody I, Freund TF. Ranck JB Jr. Studies on single neurons in dorsal hippocampal formation and
Cannabinoids inhibit hippocampal GABAergic transmission and network septum in unrestrained rats. Part 1. Behavioral correlates and firing reper-
oscillations. Eur J Neurosci 12: 3239 –3249, 2000. toires. Exp Neurol 40: 462–531, 1973.
Herreras O. Propagating dendritic action potential mediates synaptic transmis- Rudolph M, Pelletier JG, Paré D, Destexhe A. Characterization of synaptic
sion in CA1 pyramidal cells in situ. J Neurophysiol 64: 1429 –1441, 1990. conductances and integrative properties during electrically induced EEG-
Herreras O, Solís JM, Herranz AS, Martín del Río R, Lerma J. Sensory activated states in neocortical neurons in vivo. J Neurophysiol 94: 2805–
modulation of hippocampal transmission. II. Evidence for a cholinergic 2821, 2005.
locus of inhibition in the Schaffer-CA1 synapse. Brain Res 451: 303–313, Soltesz I, Deschenes M. Low- and high-frequency membrane potential oscil-
1988. lations during theta activity in CA1 and CA3 pyramidal neurons of the rat
Herreras O, Solís JM, Martín del Río R, Lerma J. Characteristics of CA1 hippocampus under ketamine-xylazine anesthesia. J Neurophysiol 70: 97–
activation through the hippocampal trisynaptic pathway in the unanaesthe- 116, 1993.
tized rat. Brain Res 413: 75– 86, 1987. Somogyi P, Klausberger T. Defined types of cortical interneurone structure
Hjorth-Simonsen A, Jeune B. Origin and termination of the hippocampal space and spike timing in the hippocampus. J Physiol 562: 9 –26, 2005.
perforant path in the rat studied by silver impregnation. J Comp Neurol 144: Stone JV. Independent Component Analysis: A Tutorial Introduction. Cam-
215–232, 1972. bridge, MA: MIT Press, 2004.
Hyvärinen A, Oja E. Independent component analysis: algorithms and appli- Tanskanen JM, Mikkonen JE, Penttonen M. Independent component anal-
cations Neural Netw 13: 411– 430, 2000. ysis of neural populations from multielectrode field potential measurements.
Jarosiewicz B, Skaggs WE. Level of arousal during the small irregular J Neurosci Methods 145: 213–232, 2005.
activity state in the rat hippocampal EEG. J Neurophysiol 91: 2649 –2657, Thompson LT, Best PJ. Place cells and silent cells in the hippocampus of
2004. freely-behaving rats. J Neurosci 9: 2382–2390, 1989.
Jung KY, Kim JM, Kim DW, Chung CS. Independent component analysis Varona P, Ibarz JM, López-Aguado L, Herreras O. Macroscopic and
of generalized spike-and-wave discharges: primary versus secondary bilat- subcellular factors shaping CA1 population spikes. J Neurophysiol 83:
eral synchrony. Clin Neurophysiol 116: 913–919, 2005. 2192–2208, 2000.

J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org

 EXCITATORY GENERATORS OF HIPPOCAMPAL LFPS 497

Vigario R, Sarela J, Jousmaki V, Hamalainen M, Oja E. Independent Wu K, Canning KJ, Leung LS. Functional interconnections between CA3
component approach to the analysis of EEG and MEG recordings. IEEE and the dentate gyrus revealed by current source density analysis. Hip-
Trans Biomed Eng 47: 589 –593, 2000. pocampus 8: 217–230, 1998.
Whittington MA, Traub RD, Jefferys JGR. Synchronized oscillations in Ylinen A, Bragin A, Nádasdy Z, Jandó G, Szabó I, Sik A, Buzsáki G. Sharp
interneuron networks driven by metabotropic glutamate receptor activation. wave-associated high-frequency oscillation (200 Hz) in the intact hippocam-
Nature 373: 612– 615, 1995. pus: network and intracellular mechanisms. J Neurosci 15: 30 – 46, 1995.

Downloaded from jn.physiology.org on September 3, 2010

J Neurophysiol • VOL 104 • JULY 2010 • www.jn.org