Professional Documents
Culture Documents
Dendritic Spikes and Activity-Dependent Synaptic Plasticity
Dendritic Spikes and Activity-Dependent Synaptic Plasticity
DOI 10.1007/s00441-006-0263-8
REVIEW
Received: 7 April 2006 / Accepted: 31 May 2006 / Published online: 1 July 2006
# Springer-Verlag 2006
Abstract Whereas the regenerative nature of action poten- generations of neuroscientists. Indeed, the identification of
tial conduction in axons has been known since the late the link between the structure and specific function of the
1940s, neuronal dendrites have been considered as passive various types of neurons remains an ongoing challenge.
cables transferring incoming synaptic activity to the soma. More than a century ago and without detailed physiological
The relatively recent discovery that neuronal dendrites knowledge, Ramón y Cajal postulated correctly that
contain active conductances has revolutionized our view of dendrites and axons serve as the anatomical correlates of
information processing in neurons. In many neuronal cell the input and the output devices of neurons, respectively
types, sodium action potentials initiated at the axon initial (Ramón y Cajal 1891). Thereafter, as a result of more
segment can back-propagate actively into the dendrite detailed investigations, dendrites were compared with a
thereby serving, for the dendrite, as an indicator of the telephone switchboard, a tree of passive electrical cables, or
output activity of the neuron. In addition, the dendrites a digitally computing unit (for a review, see Yuste and Tank
themselves can initiate action-potential-like regenerative 1996). One of the truly groundbreaking observations in the
responses, so-called dendritic spikes, that are mediated field of dendritic physiology was the discovery of a variety
either by the activation of sodium, calcium, and/or of active conductances in neuronal dendrites (for reviews,
N-methyl-D-aspartate receptor channels. Here, we review see Johnston et al. 1996; Häusser et al. 2000; Reyes 2001).
the recent experimental and theoretical evidence for a role This suggested that dendrites not only modulated postsyn-
of regenerative dendritic activity in information processing aptic potentials on their way to the soma (Magee and
within neurons and, especially, in activity-dependent Johnston 1995; Cash and Yuste 1999), but that they could
synaptic plasticity. also of themselves cause spike-like activity (Fig. 1). So far,
three major types of dendritic action potentials (also called
Keywords Dendritic spikes . Synaptic plasticity . LTD . LTP dendritic spikes) have been characterized; they are mediat-
ed by the regenerative activation of either sodium, calcium,
or N-methyl-D-aspartate (NMDA) receptor channels and
Introduction are therefore called sodium, calcium, or NMDA spikes,
respectively. Here, we review the possible roles of these
Since the pioneering work of Ramón y Cajal (1904), the different forms of regenerative activity in neuronal den-
diversity in the morphology of neuronal cells has fascinated drites for the integration of synaptic input, with special
emphasis on processes of activity-dependent synaptic
plasticity.
This work was supported by grants from the Deutsche
Forschungsgemeinschaft.
K. Holthoff (*) : Y. Kovalchuk : A. Konnerth Back-propagating sodium action potentials
Institute of Neuroscience, Technical University Munich,
Biedersteinerstrasse 29,
80802 Munich, Germany Whether one or multiple initiation zones exist for regener-
e-mail: knut.holthoff@lrz.tu-muenchen.de ative activity in neurons has long been debated. Early
370 Cell Tissue Res (2006) 326:369–377
strengthened, if the firing of neuron A repeatedly leads to dence detection underlying STDP. Because it is perme-
the firing of neuron B. In this context, the back-propagating able for calcium, and because its blockage by Mg2+ at
action potential in the postsynaptic cell (neuron B) could resting membrane potentials can be unblocked by post-
serve as a global feedback signal for the dendrite with synaptic depolarization, the NMDA-R is considered to be
regard to the output firing activity of its own axon. This the ideal candidate for detecting the coincidence between
would enable the neuron to correlate its input activity with pre- and postsynaptic activity and for translating it into a
the resulting output, a prerequisite of the Hebbian form of postsynaptic calcium transient. Indeed, compared with
learning. A recent and popular variation of a Hebbian form synaptic activation and postsynaptic action potential firing
of synaptic plasticity is the so-called spike-timing-depen- alone (Fig. 2a and b), the pairing of both events in the
dent plasticity (STDP; for reviews, see Tsodyks 2002; right order leads to a supralinear summation of the
Roberts and Bell 2002; Dan and Poo 2004). postsynaptic calcium transients (Köster and Sakmann
1998; Yuste et al. 1999; Nevian and Sakmann 2004;
Fig. 2c and d). If synaptic activation occurs first, a
Spike-timing dependent plasticity subsequent action potential arriving at the activated
synapse within a well-defined time windows can remove
In their pioneering work on cortical pyramidal neurons, the Mg2+-ion-blocking action of the NMDA-R channel
Markram, Sakmann, and colleagues discovered that the and, because glutamate is still bound, it will mediate a
pairing of a synaptic input (i.e., excitatory postsynaptic large calcium influx. Vice versa, if the action potential
potentials, EPSPs) with an action potential, fired during a arrives at the synapse before glutamate is released
narrow time window of several tens of milliseconds in the presynaptically, the NMDA-R channels will remain
same cell, produces a long lasting change in the efficacy of largely blocked by the Mg2+ ions, resulting in a moderate
synaptic transmission (Markram et al. 1997). The direction postsynaptic calcium increase.
of the change in synaptic transmission was shown to be The mechanisms of the induction and expression of
dependent on the sequential order of the two events, with a STDP have been studied extensively, but the precise
millisecond time precision. If the EPSP comes first and the mechanisms have not entirely been clarified. Thus,
postsynaptic action potential follows within a time window although initially introduced to be simply timing-depen-
of typically less than 20 ms, long term potentiation (LTP) of dent, the various pairing protocols for LTP induction have
synaptic transmission efficacy is induced. If the order is to be applied repeatedly and at sufficiently high frequen-
reversed, long term depression (LTD) is observed. The cies, otherwise they are not efficient in inducing LTP.
phenomenon of STDP has subsequently been confirmed in This suggests the notion that pairing alone is not
a wide variety of preparations in vitro and in vivo (Zhang et sufficient to induce LTP (Sjostrom et al. 2001; for
al. 1998; Bi and Poo 1998; for a review, see Dan and Poo reviews, see Sjostrom and Nelson 2002; Lisman and
2004). Spruston 2005). Whether this rate-dependence of STDP
Regardless of the mechanisms for induction, one major reflects the need for an additional depolarizing component
player in the expression of synaptic plasticity is the increase besides the back-propagating action potential is not clear.
in the postsynaptic calcium concentration during induction Nevertheless, in most studies dealing with STDP, the
(Lynch et al. 1983; Malenka 1991). Whether a strengthen- back-propagating action potentials are induced by somatic
ing or weakening of synaptic transmission is obtained current injections into the postsynaptic cell. These action
depends on the amplitude and/or duration of the postsyn- potentials ride on a depolarizing wave, which by itself
aptic calcium transient during induction (Sjostrom and might support the LTP induction process (Lisman and
Nelson 2002). The “threshold-hypothesis” proposes that Spruston 2005). Furthermore, whether action potential
long-lasting and moderate calcium elevations induce LTD, back-propagation is as effective in vivo as it is in slices
whereas short and high increases favour LTP (Lisman (see above) remains doubtful. However, this concern may
1989; Artola and Singer 1993; Hansel et al. 1997). So far, represent an advantage, because the limited efficacy of
however, quantitative estimates of the calcium signals action potential back-propagation leaves room for an
underlying the induction of LTD or LTP remain unavail- elaborated modulatory system determining the parts of
able, despite numerous attempts over the past few years the dendritic tree that will participate in Hebbian learning
(Connor et al. 1999; Conti and Lisman 2002; Malenka et al. (Magee and Johnston 1997; Migliore et al. 1999). In
1992; Neveu and Zucker 1996b; Yang et al. 1999; Wang et conclusion, STDP as a model of associative synaptic
al. 2000; Cho et al. 2001; Cormier et al. 2001; Ismailov et plasticity remains an attractive concept, but its physiolog-
al. 2004). ical significance for in vivo conditions is just beginning to
The postsynaptic NMDA receptor (NMDA-R) has emerge and needs further validation (Lisman and Spruston
been identified as the molecular device of the coinci- 2005).
372 Cell Tissue Res (2006) 326:369–377
Local dendritic spikes and information processing indeed initiate all-or-none action potentials in vitro (Houns-
gaard and Yamamoto 1979; Regehr et al. 1993; Golding and
The first evidence that dendrites themselves might generate Spruston 1998; Schiller et al. 2000; Wei et al. 2001; Golding
regenerative activity probably came from experiments in et al. 2002; Polsky et al. 2004; Holthoff et al. 2004) and in
hippocampal pyramidal neurons showing small all-or-none vivo (Pockberger 1991; Hirsch et al. 1995; Kamondi et al.
spikelets called fast pre-potentials (FPPs) preceding somatic 1998). Based on their electrophysiological properties, two
action potentials (Spencer and Kandel 1961). Although the main classes of dendritic spikes can be distinguished. The
FPPs were subsequently considered to be an artifact of conventional fast spikes are mediated by the regenerative
somatic action potentials fired from electrically coupled activation of voltage-sensitive sodium channels, can be
neighboring cells, there is now evidence that dendrites can initiated by dendritic current injection (Stuart et al. 1997a) or
Cell Tissue Res (2006) 326:369–377 373
synaptic stimulation (Turner et al. 1991; Golding and Spruston somatic sodium action potentials, because they are blocked
1998), and are sensitive to the sodium channel blocker by local application of the sodium channel antagonist TTX to
tetrodotoxin (TTX). On the other hand, the so-called complex the proximal part of the apical dendrite. The authors have
or long-lasting dendritic spikes are mainly caused by the convincingly shown that there is a strong correlation between
activation of voltage-dependent calcium and/or NMDA-R the incidence of dendritic spike firing and the successful
channels and are, therefore, referred to as “calcium spikes” induction of LTP. In layer V pyramidal neurons of mouse
(Schiller et al. 1997) and “NMDA spikes”, respectively visual cortex, local dendritic spikes can be induced by strong
(Schiller et al. 2000; Schiller and Schiller 2001; for a review, extracellular synaptic stimulation (Holthoff et al. 2004;
see Holthoff 2004). Polsky et al. 2004) or glutamate uncaging (Schiller et al.
The ability of dendrites to initiate locally regenerative 2000; Wei et al. 2001). Surprisingly, the synaptic activation of
activity, such as calcium spikes or NMDA spikes, considerably a single dendritic spike is sufficient to induce a LTD of excit-
enriches the repertoire of dendritic integration of synaptic atory synaptic transmission (Holthoff et al. 2004; Fig. 3a,b).
input. Whereas a linear or sublinear summation of postsynaptic This form of LTD, called single-shock LTD, does not need
potentials has been described for distributed synaptic input in somatic spiking and is largely saturated following single-
various preparations (Urban and Barrionuevo 1998; Cash and shock induction. The production of five consecutive dendritic
Yuste 1998, 1999; Magee 2000; Polsky et al. 2004), spatially spikes has no substantial boosting effect (Fig. 3c). Prelimi-
clustered synaptic input to the same neurons can induce non- nary results from the same preparation indicate that coinci-
linear responses in small dendritic sections (Polsky et al. dent pairing of a synaptically evoked dendritic spike with
2004; Holthoff et al. 2004; Losonczy and Magee 2006; back-propagating action potentials may evoke a supralinear
Fig. 1b). Depending on the spatial distribution and timing of increase in the amplitude of the intra-dendritic calcium
the synaptic input, the same neuron can therefore switch transient and subsequently lead to LTP (Holthoff et al.
between an essentially linear and highly non-linear mode of 2005). Thus, mounting evidence indicates that local dendritic
integration. Both modes of integration can be combined, spikes are able to induce bidirectional synaptic plasticity.
because the depolarization produced by the local non-linear What is the computational and physiological significance of
integrating at dendritic subunits sum up linearly at the soma local spike-evoked synaptic plasticity? Unlike STDP, which
(Polsky et al. 2004). Simulations suggest that the equipment requires the coincidence between pre- and postsynaptic
of dendrites with non-linearly acting subunits leads to an spiking, local spike-evoked plasticity involves the coinci-
increase in potential storage capacity by several orders of dence of synaptic activation of closely grouped inputs to the
magnitude (Poirazi and Mel 2001). In addition, the alliance of same part of a dendrite (Polsky et al. 2004; Losonczy and
non-linear and linear integration modes combines the Magee 2006). This represents a new learning rule. The new
advantages of the excellent signal-to-noise ratio of digital mechanism divides neurons into largely independent modules
processing with the speed and complexity of analog process- of integration and learning that do not depend on the usual
ing in a single cellular compartment, viz., the dendritic tree. feedback signal from the soma, viz., the back-propagating
sodium action potential. A major difference between STDP
and local spike-evoked plasticity is the induction speed. Thus,
Local dendritic action potentials and synaptic plasticity whereas STDP requires many repetitions of the conditioning
pre-post stimulation (Bear and Abraham 1996; Lisman and
A common feature of local dendritic calcium spikes or Spruston 2005), a single local spike is sufficient for LTD
NMDA spikes is that they are accompanied by a transient (Holthoff et al. 2004) or LTP (Holthoff et al. 2005; but see
increase in intracellular calcium concentration (Schiller et al. Golding et al. 2002) induction. The increased induction speed
1997, 2000; Golding et al. 2002; Wei et al. 2001; Holthoff et largely relies on the synergistic cross-activation of neighbor-
al. 2004). Assuming that the intracellular calcium concen- ing NMDA receptor channels located on the same dendrite
tration transient is necessary and sufficient to induce synaptic (Schiller et al. 2000; Holthoff et al. 2004). Therefore, local
plasticity (Neveu and Zucker 1996a), then synaptically spike-dependent plasticity may be the primary mechanism
evoked local dendritic spikes can be postulated to be able underlying the rapid acquisition of memories.
to induce synaptic plasticity (Goldberg et al. 2002). Indeed,
two recent reports show that, in hippocampal CA1 neurons
and cortical pyramidal neurons, locally restricted dendritic Conclusions and perspectives
spikes on their own can induce LTP and LTD, respectively.
Golding and colleagues (2002) have demonstrated, in an During the last few decades, the active properties of neuronal
elegant study, that synaptically evoked dendritic spikes dendrites have come into the realm of scientific interest.
induce LTP in hippocampal CA1 neurons. This form of Parallel to the capability of dendrites for analog computation,
synaptic plasticity does not need active back-propagation of the non-linear processing of synaptic inputs provides modular
374 Cell Tissue Res (2006) 326:369–377
50
4
5x
3
0
2
0 10 20 30 40
1 1x
Time (min)
0
before after
0 10 20 30 40 50 60 d
Time (min)
b 150
n=8
150
n=4
EPSP ampl (%)
50 50
1x 5x
0 0
0 10 20 30 40 50 60 0 10 20 30 40
Time (min) Time (min)
Fig. 3 A single local dendritic spike induces instantaneous long-term stimulus (bars 2 mV and 50 ms). b Summary LTDx1 experiments of
synaptic depression (LTD). a Induction of LTD by single dendritic eight cells. c LTD induced by five consecutive conditioning stimuli
spike in a pyramidal neuron from layer 5 of the mouse hippocampus delivered 15 s apart from each other. d Same experiments as in c, but
(LTDx1). Synaptic input to a basal dendrite located approximately cells were hyperpolarized to –80 mV during conditioning stimuli to
50 μm away from the cell body. Note the framed LTD induction signal prevent the initiation of local dendritic spikes. No significant change
(bars 50% ΔF/F, 250 ms for calcium signal and 5 mV, 50 ms for in strength of synaptic transmission was detected. Modified from
EPSP). Traces indicate mean EPSPs before and after the conditioning Holthoff et al. (2004)
integration, which increases the memory capacity of neuronal whether LTP and LTD are expressed at the postsynaptic or
tissue (Poirazi and Mel 2001) and enriches the processing presynaptic site are of great interest. The combination of high
complexity of dendrites by an implementation of additional resolution imaging and single synapse stimulation by two-
local learning rules (Golding et al. 2002; Holthoff et al. photon glutamate uncaging could refine this analysis down to
2004). Although the ability of neuronal dendrites to initiate the single synapse level. Early experimental evidence has
local dendritic spikes has been known for several years, indicated that local dendritic spikes can provide a means for
surprisingly few attempts have been undertaken to investi- the induction of synaptic plasticity in vitro (Golding et al.
gate their ability to induce synaptic plasticity. This may be 2002; Holthoff et al. 2004, 2005); however, the significance
because dendritic spikes are often barely detectable by of this mechanism for in vivo conditions remains obscure.
somatic recordings, making necessary the more difficult An important challenge for the future will be to
dendritic recordings or imaging techniques to pinpoint the investigate the physiological significance of local dendritic
initiation of dendritic spikes with respect to the induction of signaling for cortical information processing in the
synaptic plasticity (Golding et al. 2002; Holthoff et al. 2004). sensory system in vivo and to test whether it participates
Because a single dendritic spike, and therefore a single in synaptic plasticity during development and learning.
postsynaptic calcium transient, is able to induce long-term Some of these tests will probably involve the use of
synaptic plasticity, the detailed characterization of the multi-photon imaging (for a review, see Helmchen and
postsynaptic signals that lead to the potentiation or the de- Denk 2005) in combination with various neuron labeling
pression of a synapse will be possible. In particular, the techniques (Stosiek et al. 2003; Rathenberg et al. 2003;
expression mechanisms of plasticity and the question of Miesenböck 2004; Mank et al. 2006).
Cell Tissue Res (2006) 326:369–377 375
Magee JC (2000) Dendritic integration of excitatory synaptic input. Roberts PD, Bell CC (2002) Spike timing dependent synaptic
Nat Rev Neurosci 1:181–190 plasticity in biological systems. Biol Cybern 87:392–403
Magee JC, Johnston D (1995) Synaptic activation of voltage-gated Schiller J, Schiller Y (2001) NMDA receptor-mediated dendritic spikes
channels in the dendrites of hippocampal pyramidal neurons. and coincident signal amplification. Curr Opin Neurobiol 11:343–348
Science 268:301–304 Schiller J, Schiller Y, Stuart G, Sakmann B (1997) Calcium action
Magee JC, Johnston D (1997) A synaptically controlled, associative potentials restricted to distal apical dendrites of rat neocortical
signal for Hebbian plasticity in hippocampal neurons. Science pyramidal neurons. J Physiol (Lond) 505:605–616
275:209–213 Schiller J, Major G, Köster HJ, Schiller Y (2000) NMDA spikes in basal
Malenka RC (1991) The role of postsynaptic calcium in the induction dendrites of cortical pyramidal neurons. Nature 404:285–289
of long-term potentiation. Mol Neurobiol 5:289–295 Sjostrom PJ, Nelson SB (2002) Spike timing, calcium signals and
Malenka RC, Lancaster B, Zucker RS (1992) Temporal limits on the synaptic plasticity. Curr Opin Neurobiol 12:305–314
rise in postsynaptic calcium required for the induction of long- Sjostrom PJ, Turrigiano GG, Nelson SB (2001) Rate, timing, and
term potentiation. Neuron 9:121–128 cooperativity jointly determine cortical synaptic plasticity.
Mank M, Reiff DF, Heim N, Friedrich MW, Borst A, Griesbeck O Neuron 20:1149–1164
(2006) A FRET-based calcium biosensor with fast signal kinetics Spencer WA, Kandel ER (1961) Electrophysiology of hippocampal
and high fluorescence change. Biophys J 90:1790–1796 neurons. IV. Fast prepotentials. J Neurophysiol 24:272–285
Markram H, Lübke J, Frotscher M, Sakmann B (1997) Regulation of Spruston N, Schiller Y, Stuart G, Sakmann B (1995) Activity-
synaptic efficacy by coincidence of postsynaptic APs and EPSPs. dependent action potential invasion and calcium influx into
Science 275:213–215 hippocampal CA1 dendrites. Science 268:297–300
Miesenböck G (2004) Genetic methods for illuminating the function Stosiek C, Garaschuk O, Holthoff K, Konnerth A (2003) In vivo two-
of neural circuits. Curr Opin Neurobiol 14:395–402 photon calcium imaging of neuronal networks. Proc Natl Acad
Migliore M, Hoffman DA, Magee JC, Johnston D (1999) Role of an Sci USA 100:7319–7324
A-type K+ conductance in the back-propagation of action Stuart G, Häusser M (1994) Initiation and spread of sodium action
potentials in the dendrites of hippocampal pyramidal neurons. J potentials in cerebellar Purkinje cells. Neuron 13:703–712
Comput Neurosci 7:5–15 Stuart GJ, Häusser M (2001) Dendritic coincidence detection of
Miyakawa H, Kato H (1986) Active properties of dendritic EPSPs and action potentials. Nat Neurosci 4:63–71
membrane examined by current source density analysis in Stuart GJ, Sakmann B (1994) Active propagation of somatic action
hippocampal CA1 pyramidal neurons. Brain Res 399:303– potentials into neocortical pyramidal cell dendrites. Nature
309 367:69–72
Neveu D, Zucker RS (1996a) Long-lasting potentiation and depres- Stuart G, Schiller J, Sakmann B (1997a) Action potential initiation
sion without presynaptic activity. J Neurophysiol 75:2157–2160 and propagation in rat neocortical pyramidal neurons. J Physiol
Neveu D, Zucker RS (1996b) Postsynaptic levels of [Ca2+]i needed to (Lond) 505:617–632
trigger LTD and LTP. Neuron 16:619–629 Stuart G, Spruston N, Sakmann B, Häusser M (1997b) Action
Nevian T, Sakmann B (2004) Single spine Ca2+ signals evoked by potential initiation and backpropagation in neurons of the
coincident EPSPs and backpropagating action potentials in spiny mammalian CNS. Trends Neurosci 20:125–131
stellate cells of layer 4 in the juvenile rat somatosensory barrel Svoboda K, Denk W, Kleinfeld D, Tank DW (1997) In vivo dendritic
cortex. J Neurosci 24:1689–1699 calcium dynamics in neocortical pyramidal neurons. Nature
Palmer LM, Stuart GJ (2006) Site of action potential initiation in layer 385:161–165
5 pyramidal neurons. J Neurosci 26:1854–1863 Svoboda K, Helmchen F, Denk W, Tank DW (1999) Spread of
Pan E, Colbert CM (2001) Subthreshold inactivation of Na+ and dendritic excitation in layer 2/3 pyramidal neurons in rat barrel
K + channels supports activity-dependent enhancement of cortex in vivo. Nat Neurosci 2:65–73
back-propagating action potentials in hippocampal CA1. J Tsodyks M (2002) Spike-timing-dependent synaptic plasticity—the
Neurophysiol 85:1013–1016 long road towards understanding neuronal mechanisms of
Pockberger H (1991) Electrophysiological and morphological proper- learning and memory. Trends Neurosci 25:599–600
ties of rat motor cortex neurons in vivo. Brain Res 539:181–190 Turner RW, Meyers DE, Richardson TL, Barker JL (1991) The site for
Poirazi P, Mel BW (2001) Impact of active dendrites and structural initiation of action potential discharge over the somatodendritic
plasticity on the memory capacity of neural tissue. Neuron axis of rat hippocampal CA1 pyramidal neurons. J Neurosci
29:779–796 11:2270–2280
Polsky A, Mel BW, Schiller J (2004) Computational subunits in thin Urban NN, Barrionuevo G (1998) Active summation of excitatory
dendrites of pyramidal cells. Nat Neurosci 7:621–627 postsynaptic potentials in hippocampal CA3 pyramidal neurons.
Ramón y Cajal S (1891) Significación fisiológica de las expansiones Proc Natl Acad Sci USA 95:11450–11455
protoplásmicas y nerviosas de la sustancia gris. Rev Cienc Med Vetter P, Roth A, Häusser M (2001) Propagation of action potentials in
Barcel 22:23 dendrites depends on dendritic morphology. J Neurophysiol
Ramón y Cajal S (1904) La textura del sistema nerviosa del hombre y 85:926–937
los vertebrados, primera edicion. Moya, Madrid Wang SS, Denk W, Häusser M (2000) Coincidence detection in single
Rathenberg J, Nevian T, Witzemann V (2003) High-efficiency dendritic spines mediated by calcium release. Nat Neurosci
transfection of individual neurons using modified electrophysi- 3:1266–1273
ology techniques. J Neurosci Methods 126:91–98 Waters J, Helmchen F (2004) Boosting of action potential back-
Regehr W, Kehoe JS, Ascher P, Armstrong C (1993) Synaptically propagation by neocortical network activity in vivo. J Neurosci
triggered action potentials in dendrites. Neuron 11:145–151 24:11127–11136
Reyes A (2001) Influence of dendritic conductances on the Waters J, Larkum M, Sakmann B, Helmchen F (2003) Supralinear Ca2+
input-output properties of neurons. Annu Rev Neurosci influx into dendritic tufts of layer 2/3 neocortical pyramidal
24:653–675 neurons in vitro and in vivo. J Neurosci 23:8558–8567
Richardson TL, Turner RW, Miller JJ (1987) Action-potential Waters J, Schäfer A, Sakmann B (2005) Backpropagating action
discharge in hippocampal CA1 pyramidal neurons: current potentials in neurones: measurement, mechanisms and potential
source-density analysis. J Neurophysiol 58:981–996 functions. Prog Biophys Mol Biol 87:145–170
Cell Tissue Res (2006) 326:369–377 377
Wei DS, Mei YA, Bagal A, Kao JP, Thompson SM, Tang CM Yuste R, Majewska A, Cash SS, Denk W (1999) Mechanisms of
(2001) Compartmentalized and binary behavior of terminal calcium influx into hippocampal spines: heterogeneity among
dendrites in hippocampal pyramidal neurons. Science spines, coincidence detection by NMDA receptors, and optical
293:2272–2275 quantal analysis. J Neurosci 19:1976–1987
Yang SN, Tang YG, Zucker RS (1999) Selective induction of LTP Zecevic D (1996) Multiple spike-initiation zones in single neurons
and LTD by postsynaptic [Ca2+]i elevation. J Neurophysiol revealed by voltage-sensitive dyes. Nature 381:322–325
81:781–787 Zhang LI, Tao HW, Holt CE, Harris WA, Poo M (1998) A critical
Yuste R, Tank DW (1996) Dendritic integration in mammalian window for cooperation and competition among developing
neurons, a century after Cajal. Neuron 16:701–716 retinotectal synapses. Nature 395:37–44